SCIENCE JOURNAL TRANSPORTATION

SCIENCE JOURNALOF

TRANSPORTATION

Especial Issue No. 01International cooperation Journals

MADI – SWJTU – UTC

Moscow - Chengdu - Hanoi01 - 2009

Dear researchers, colleagues and readers, Transportation is the means by which all people are connected,all human activities occur. Nowadays, in strong globalizing process,community activities have not been limited by countr ies’ borders;thus transportation becomes non - confrontiers. We, transportation makers, in this moment, have had a commonforum to together discuss, contribute, share and dedicate. First Especial Issue of international co -operating transportationscience journals of State Technical University (MADI) - Russia,Southwest Jiaotong University (SWJTU) - China and University ofTransportation and Communication (UTC) - Vietnam is published inspring – season of blooming and developments. We wish you and our transportation career were achieved,prosperous and fruitful. Science is non - limitation, Transportation is non - boder, Friendship is non - confrontiers,Aim toward the future, we will do our best to make transportation:

More intelligent and effective, Faster and safer, Cleaner and Greener,With that objective, by this forum, we together connect, endeavour,research, create, contribute, share and devote.

Moscow – Chengto - HanoiBoard of Editors - in - Chief

ISSN 2410-9088

SCIENCE JOURNAL

OF

TRANSPORTATION

Especial Issue No. 01

International cooperation Journals

State Technical University-MADI Southwest Jiaotong University (SWJTU)

University of Transport and Communication (UTC)

Moscow - Chengdu - Hanoi 06-2009

INTERNATIONAL COOPERATION ISSUE OF TRANSPORTATION – Especial Issue – No.01 1

Pages 3 Prof. VAYCHESLAV PRIKHODKO

Prof. ALEXANDRVASILYEV

Prof. MIKHAIL NEMCHINOV

Prof. VLADIMIR NOSOV

Prof. PAVEL POSPELOV

Prof. VALENTIN SILYANOV

Prof. VIKTOR USHAKOV

(State technical University -MADI)

Modern tendencies in development of highwaysin Russia

Pages 16 Assoc. Prof. Dr. TRAN TUAN HIEP

University of Communication andTransport, Vietnam

Researching on position of calculation slidecenter in computing the stability of road bed byslide circular arc method

Pages 24YANJUN QIU, A.M. ASCE

School of Civil Engineering,

Southwest Jiaotong University

Chengdu, 610031, China

Theoretical development and engineeringpractice of pavements in China

Pages 39 Dr. CESAR QUEIROZ

World Bank


State Technical University-MADI

Dr. ALEKSEY AKULOV

Ministry of Natural Resources of Russia

Launching public-private partnerships for highwaysin transition economies

Pages 50 Prof. DR. DO DUC TUAN

Meng. LE LANG VAN

University of Transport andCommunications

Diesel engine diagnosis by vibration oil analysis

Pages 54Prof. VYACHESLAV M.

PRIKHODKO

Prof. VICTORIA D.GERAMI

Prof. ALEXANDER V. KOLIK

State Technical University -MADI,Moscow, Russia

Features of creation of logistic centers inconditions of siberia and far east of RussiaPages 63 Prof. MIKHAIL V. NEMCHINOV

Dr. ALEXEI S. MEN’SHOV State Technical University - Madi Dr. DMITRY M. NEMCHINOV The Association of Road Design

Institutions of Russia Dr. VERONIKA OSINOVSK AYA

Bryansk State TechnologicalAcademy, Bryansk, Russia

The environmental problems connected withhighway construction and maintenancePages 77 NGUYEN THANH SANG, Doctoral

studentPHAM DUY HUU, Professor

Institute of Science and Technology for Transport construction


An experimental research on sand concrete inMekong deltaPages 86 Prof. M.V. NEMCHINOV

Ph.D. student VU TUAN ANH

(State Technical University – MADI,Moscow, Russia)

Design of diversion ditch es for highway roadbedsPages 93 GHAZWAN AL-HAJI, ASP

KENNETH

Department of Science andTechnology (ITN), Linköping

University, 601 74 Norrköping,Sweden

The evolution of international road safety benchmarkingmodels: towards a road safety development in dex (rsdi)

TABLE OF CONTENT

INTERNATIONAL COOPERATION ISSUE OF TRANSPORTATION – Especial Issue – No.012

Pages 109 WEIHUA MA, SHIHUI LUO

Traction Power State Key Laboratory,Southwest Jiaotong University,

Chengdu 610031, China

RONG-RONG SONG

College of computer science andtechnology, Southwest University

for nationalities,Chengdu

Sichuan 610041. China

Influence of track irregularity on ongitudinalvibration of wheelset and correlationperformancePages 125

TRAN VIET HUNG

Msc., Dept. of Civil Eng., Universityof Transport and Communication

Caugiay, Dongda, Hanoi, Vietnam

NGUYEN VIET TRUN G

Dr. of Eng., Professor, Dept. of CivilEng., University of Transport and

Communication,


Seismic resistance of multi-spans pc bridgeunder earthquake occur in Vietnam

Pages 135 Dr. EVANGELOS BEKIARIS

Research Director of CERTH/HIT

Forum for European Road TrafficSafety Institutes (FERSI) President

Sustainable traffic safety policies and researchpriorities for safe and secure european roads

Pages 146 Lecturer JAMSHID SODIKOV

Tashkent Automobile and RoadInstitute,

20, street Movounnahr,Tashkent,100020

Advisor Economist

ZIYODULLO PARPIEVUNDP Uzbekistan Country Office

Preliminary road cost studies in developingcountries

Pages 165 XUAN BINH LUONG; VIKHONESAYNHAVONG

THANH THUY HOANG

Department of Civil Engineering

University of Transport andCommunications, Vietnam

MEIKETSU ENOKI


Tottori University, Japan

Development of generalized limit equilibriummethod for the failure of retaining wallsunder seismic loadingsPages 172

Dr. BUI NGOC TOAN


A method of determining social and economic

benefits of transportation construction projects

Pages 175DR.-ING. KHUAT VIET HUNG

Institute of Transport Planning and

Management

University of Transport and

Communications

Transport sector in vietnam: current issues andfuture agendaPages 187

MSC. NGUYEN THI TUYET TRINH

Ph.D Candidate, University of

Transport

and Communications, Vietnam

DR. TAKEHIKO HIMENO

Kawaguchi Metal Industries, J apan

Comparison between janpanese specification

and aashto 1998 specification in sesmic design


I. CURRENT STATE AND TENDENCIESIN THE DEVELOPMENT OF ROADS INRUSSIA

In the last 15 years the rate ofmotorization in the country increased. Rapidmotorization in Russia will continue in theforeseeable future. According to thepredictions of the Ministry of Transport of theRussian Federation, in year 2010 the car park

will amount to approximately 36 -39 millionautomobiles (table 1). Traffic volume in the

road network has increased by 5%, andgrowth of the traffic intensity on the mainroads has reached 26.2%.

The motorization growth ratio predeterminesthe necessity to speed up the road networkmodernization development.

Table 1. The number of vehicles in Russia (at theend of the year)

1990 2001 2010

Lorries (including pick-ups andvans), thous. pcs.

2744 3329 4927…5321

Passenger cars, thous. pcs. 8964 21152 31299…33805

MODERN TENDENCIES IN DEVELOPMENTOF HIGHWAYS IN RUSSIA

Prof. VAYCHESLAV PRIKHODKOProf. ALEXANDR VASILYEVProf. MIKHAIL NEMCHINOVProf. VLADIMIR NOSOVProf. PAVEL POSPELOVProf. VALENTIN SILYANOVProf. VIKTOR USHAKOV(State technical University-MADI)

Abstracts: In the report the data on the level of motorization and the road network length in

the Russian Federation, the rates of its growth from 1995 till 2010 are presented. The necessity to

enhance the pavement is validated and the new approach to road pavement design and

construction are discussed. The current system for diagnosing and monitoring the road conditions

used in Russia is defined.

Key words: road network, state and growth prospects, pavement, new tendencies in design,

diagnostic system, road condition estimation


including owned by the citizens,thous. pcs.

8677 19984 29571…31938

Number of automobiles per 1000of citizens, pcs.

80 169 250…270*

Average increase in vehicle parkper annum, %

7,7 (from 1990 till2001)

5…6* (from 2002 till 2010)

Increase in the total number ofvehicles, thous. pcs.

12733 (from 1990till 2001)

11745…14645* (from 2002till 2010 )

Note. * - Forecast. (The proportion of the passenger cars in year 1990 ≈ 76,56%; in year 2001 ≈ 86,4%).

Over the period of 1995-2000 the length

of the road network increased from 519 to 584

thous. km including increase in the federal

road network from 41 thous. km to 46.3 thous.

km. Within these years 33.9 thous. km. of

roads were built and reconstructed, including

18 thous. km. of newly built roads, 183.2

thous. km. of reconstructed roads and 290 km

of bridges; the public road network was

replenished with 47 thous. km. of roads

supervised by farm producers.

In the Programme “Moderniza tion of the

Transport System in Russia (for the period of

2002-2010)” it is envisaged that by year 2010the public roads network will increase by 1.1

times. The length of the public roads network

in 2010 will total 670 thous. km including 50

thous. km of federal and 620 thous. km of

regional roads. The length of the roads with

capital type of pavement will amount to 428

thous. km, and those with the transition type –to 212 thous. km. The length of the roads with

four and more traffic lanes will increase fr om

4.3 to 8 thous. km, i.e. nearly twofold, that

will drastically reduce the possibility of traffic

jams.

Traffic capacity of the most congested

sections of the primary interregional and

international routes will increase 1.5 -2 times,

and on an average, by 10-12% within the

network. Level of congestion, characterizing

correspondence between the road network

technical level and its traffic volume, will on

average amount to 0.4-0.6 within the network

what is optimal according to comfort, safety

and efficiency of transportation.

Under these circumstances the road

network development and road service quality

management is the task of primary

importance, its solution exercises direct

influence over the pace of the socioeconomic

development of the country. Moderni zation of

the existent public roads is becoming of great

importance, i.e. bringing their application

properties and service state into line with the

requirements of the country’s car park andfactual traffic volume. The application

properties include speed, fluidity, safety and

comfort of driving, road capacity and the level

of congestion, the capacity to carry


automobiles and road-trains with the

permissible axial load, total weight and

clearance limit as well as environmental

safety.

The efficiency of road works can be

significantly enhanced if in the process of

planning one proves and optimizes the

management decisions based on the

estimation of the factual road situation. To

adequately utilize facilities and physical

resources meant for reconstruction, r epairs

and maintenance of roads, the Ministry of

Transport of the Russian Federation and State

Technical University-MADI have developed

and implemented the system of road situation

maintenance based on the results of the

systematical monitoring, diagnostic s and

evaluation of the real road service state. This

system is based on the complete, objective and

reliable information on parameters, characteristics

and conditions of roads and road buildings

operation, availability of faults and the terms of

their emerging, traffic flow characteristics that

can be obtained in the course of diagnosis,

inspection, acquisition, analysis and

organization of the data bank on road service

quality. The management system created and

the data bank make it possible to objectivel y

estimate and predict the road and road

buildings state in the process of further

operation. The complex index of the road

service conditions takes into account

influence of the following key elements,

parameters and characteristics: width of the

main reinforced surface of carriageway and

width of the bridge; road shoulder width and

condition; traffic volume and composition;

longitudinal grade and sight distance; curvature

in plan and superelevation slope; longitudinal

surface roughness; skid resistance coefficient;

pavement condition and durability; roughness in

transverse direction (rut depth); traffic safety;

engineering equipment and instruments; level of

maintenance management.

Over a period of years the federal road

repairs have been planned and carri ed out on

based on the results of the diagnostics. For the

first time in the road sector of the Russian

Federation an evidence based, objective and

reliable system of road service monitoring and

diagnostics have been developed and widely

used, it is necessary for effective management

of the application properties by developing

projects and performing works on road

reconstruction, total overhaul and maintenance.

The road application properties are

provided by real level of their maintenance,

geometrical parameters, technical

characteristics, engineering equipment and

instruments. Diagnostic makes it possible to

detect origin of faults, justify the road repair

and maintenance type and scope. The

diagnostics system generalizes (synthesizes)

of the main regulations requirements to the

road application properties and contributes to

the increase in the quality of service.

The diagnostic system makes it possible

to effectively control the operational condition

of the certain road sections, certain routes or

certain roads and the road network altogether .

II. NEW TENDENCIES IN DESIGN ANDCONSTRUCTION OF ROADPAVEMENTS

The following factors predeterminemodern tendencies in development ofpavement calculation and constructionmethods in Russia:


Growing demands of the currentroad traffic which consists of a significantportion of lorries and a high number (m orethan 65%) of passenger cars.

The sudden increase in the numberof cars (according to the Ministry of Transportof the Russian Federation over the periodfrom 1991 to 2025 – 4-5.5 times, including anincrease in the number of lorries – 4.8 times)created necessity in radical modernization ofthe road network and, in the first instance,modernization of the federal roads.

Analysis of the real life time ofpavement in the Russian Federation and othercountries shows that when put in the sameoperational conditions rigid pavement has thelife time of 1.6-2 times longer than that offlexible pavement.

Modern tendencies in improvementof road engineering, development of n ewroad-building materials such as modifiedhigh-quality cement concrete, fiber andpolymer concrete, high-strength compositesand so on; evolution of structural conceptsand techniques contribute to expansion ofrigid and composite pavement (bituminousconcrete pavement with cement concrete subbase).

The current road network has mostlyflexible pavements including bituminousconcrete pavement with the sub base made ofmaterials that are unsuitable for bendingtension (this is mostly broken stone); suchpaving materials don’t provide for thestandard life time of the pavement as in suchstructures only a relatively thin bituminousconcrete pavement works in bending; in otherwords, in the current operational conditionsflexible pavement of the capital type should

have a sub base of binding agent -treated andsteady working in bending materials that willlast till the pavement total overhaul.

In 2001 new designing standards offlexible pavements were introduced in theRussian Federation. The peculiarity of thenew method of flexible pavement strengthcalculation lies in the fact that the structurecalculation is based on the three failurecriteria (aggregate pavement elastic modulus,shear resistance of the constructional layersand subfoundation, bending tensi on resistanceof indistinguishable constructional layers) thatare determined considering replicationinfluence (aggregate computational number ofapplications) on the pavement life time. Theold method presumed stress calculating basedon the influence of the perspective (as at thelife time end) daily average number of stressapplications (reduced to the equivalent thrustload).

Synthesis of the 60-year-old practice ofrigid pavements application in the formerUSSR and the Russian Federation makes itpossible to consider that in the overwhelmingmajority of cases the imperfections causingtheir early aging are connected not with thedesigning and constructional miscalculations,but with the failure to comply with theregulations on constructional and opera tionalmethods of the rigid pavement, and in singlecases with failure to comply with theregulations on the materials used.

At the same time it is worth mentioningthat within the period stated we continued toimprove pavement designing, calculation andconstruction methods including moreadvanced rigid pavement construction.

In a number of countries the traditional


indistinguishable cement concrete pavement withmetal pegs in the contraction joints isgradually replaced by solid-drawn continuousor fiber reinforced pavement what makes itpossible to lengthen the life time 1.5 -2 timesas compared to the unreinforced pavement. Inthe 1970-s some control sections withcontinuous reinforced pavement were laid inMoscow (Profsoyuznaya Street, AltufyevskoeShosse) that are in effective operation evennowadays.

Continuous reinforced pavement wasalso laid at some airfields of the formerUSSR.

Continuous reinforced pavement andsubgrade calculation is much different fromthat of indistinguishable unreinforcedpavement and subgrade. The most significantdifferences lie in the validation theoreticalreinforcement percentage, identification of theopening width of the reinforced breaches,reinforcement durability calculation,engagement stop resistance calculation and soon.

Based on the results of the experimentalconstruction, building regulations fordesigning and construction of continuousreinforced pavement and subgrade weredeveloped.

At the moment the following aspects ofpavement design are of relevance in theRussian Federation:

Contraction crack resistance analysisfor the bituminous concrete pavement .

Analysis of cohesionless or slightlycohesive subgrade for rigid constructionallayers.

Analysis of unreinforced indistinguishable

joint-free subgrade made of low-modulus cementconcrete for rigid and flexible constructional layers.

Analysis of fiber-reinforcedindistinguishable joint-free pavement andsubgrade.

Analysis of bituminous concretepavement with indistinguishable cementconcrete subgrade including joi nt-freepavement made of low-modulus cementconcrete and so on.

One of the most important aspects ofpavement durability implementation andenhancement in the Russian Federation is theanalysis of cement concrete durability limitsunder the joint action of repeated dynamictraffic loading and alternate periodic freezingand throwing. Due to the fact that at themoment the theoretical bases for cementconcrete cold resistance analysis aredeveloped insufficiently, the problem statedshould be studied and solved on theexperimental basis. The experimentations willmake it possible to validate the rational andeffective strength reserves for the rigidpavement what provides for the necessarylevel of durability (i.e. the life time). Anotheracute problem lies in the enhancement of themethods for the objective technical andeconomic comparison of rigid and flexiblepavement that takes into consideration notonly expenses on construction and repairs, butalso on transportation under differentoperational conditions of the pavementscompared.

At the moment the construction expensesfor the top class roads with the rigid andflexible pavement differ insignificantly –within 5%. Nevertheless, the current methodsfor technical and economic comparison


usually underestimate the influence of thedifference in the life times of the rigid andflexible pavement on the transportation cost.

Comparative analysis of the domestic andforeign designing and construction decisionsconcerning pavement make it possible toconsider that the technical level of suchdecisions in our country is rather close tothose in the highly developed countriesabroad. The engineering decisions andmodern road machinery differ insignificantlyfrom the foreign analogs. The practice of rigidpavement construction showed that:

1. Rigid pavement has the highest liferatio and according to this criterion has nosubstantial alternative under the currentconditions in the Russian Federation.

2. In the process modernization of theprimary roads it is advisably to use on a largescale the following types of pavement:

Monolithic cement concretepavement.

Monolithic cement concretesubgrade for rigid and flexible pavement,providing stable bending tension resistance.

Continuous reinforced pavement andsubgrade.

Fiber reinforced pavement andsubgrade.

Rolled concrete subgrade.

Modified concrete pavement andsubgrade, including thin continuousreinforced layers to strengthen flexiblepavement.

Polymer concrete reinforcementproviding low traffic noise level.

3. At the moment we advise to applyrigid pavement in the following fields:

Primary multilane roads.

Federal and regional roads, class II-III.

Access ways and by-passes.

Toll roads.

Industrial roads.

Municipal roads.

4. To expand the scope of application ofthe new rigid pavement it is necessary toorganize and systematically performoperational monitoring, to carry out laboratoryand in situ field experiments concerningcement concrete durability limits under thejoint action of repeated dynamic trafficloading and alternate periodic freezing andthrowing.

III. THE CURRENT SYSTEM FORDIAGNOSING AND ESTIMATING THEROAD CONDITIONS

General provisions diagnostics and

estimation of the road conditions are two

interconnected steps in the process of road

conditions management.

Diagnostics is inspection, accumulation and

analysis of the information on the parameters ,

characteristics and conditions of road

operations, on imperfections and the reasons

of their emergence and other data necessary to

estimate and forecast the road conditions in

the course of the further operation.

Road conditions estimation is the

determination of the degree of compliance of

the real road characteristics and parameters to

the regulations satisfying traffic demands.


The scope and description of th e

diagnosing works depend on the method of

the road conditions evaluation.

Stages in the development of the

methods for road conditions diagnosing and

estimation

Starting with the first half of the last

century the road conditions diagnosing and

estimation was performed according to the

availability, nature and number of faults

(deformations, corrosions, departing from the

regulations), that characterized the road from

the position of durability, efficiency, life time

and so on,

In various evaluation method s the

number of such factors fluctuated between 10

and 40. Usually not the entire road conditions

but the conditions of the pavement was

estimated. The necessary information was

accumulated applying visual method.

The second generation of the methods for

road conditions diagnosing and evaluation is

represented by composite and complex

methods, the core of which lies in the fact that

the road is considered engineering

transportation construction meant for safety

traffic with specified speed and loads.

Road conditions are estimated not only

according to technical parameters and

characteristics but also according to transport

quality figures (TQF) that a road assures:

speed, safety, admission rate, axel weight

limit etc.

Herewith each parameter, characteristic and

value is evaluated separately. As a result for

every road section there are from 20 to 80

absolute or relative numbers with different

units of measurement illustrating compliance to

or deviations from regulations, here they help

to solve the problem of function and most

important maintenance actions.

Diagnostic data are collected both

visually and using measuring equipment and

laboratories.

Third generation of methods for

generalized or complex evaluation of road

conditions is based on the concept of pu rpose

of function of a road as a means for customer

service, consumers of road services that in

some way pay for these services.

From consumers’ point of view the mostimportant are the transport and operational

characteristics assured by a road: continuit y,

speed, convenience and traffic safet, traffic

capacity and a level of congestion, axel

weight limit and other figures that are relevant

to consumer characteristics of roads.

One of the first methods of this

generation is HDM-III (Highway Design and

Maintenance) and its further modification

HDM-IV where there is an extended

evaluation of parameters of plan, longitudinal

profile and pavement conditions according to

their influence on an average speed of

vehicular traffic.

Since 1990 in Russia is widely used a

method developed by Prof. A.P.Vasiliev, a

method of complex evaluation of road quality

and condition according to its consumer

characteristics based on this principle.

In some cases methods of the first and

the second generation are used.

The basic indices of an evaluation of


quality and condition of roads according to

consumer characteristics

An integral index that most fully reflects

main consumer characteristics is a traffic

speed evaluated through a coefficient of

provision of design speed rating:

rV

Va.maxCsr (1)

Csr – coefficient of provision of speed

rating; Va.max – actual maximal possible and

safe speed of a single passenger car assured

by road at its actual parameters and condition,

km/h; Vr – design speed, km/h.

For calculation convenience as a base of

speed rating is assumed speed that equals to

120 km/h. Then:

120a.maxV

srC (2)

As an additional index is taken a number

that shows permissible carrying capacity and

axel weight limits that were reduced to the

coefficient of provision of speed rating

reasoning from proportionality of speed and

carrying capacity influence on vehicle

productivity.

Assured by a road level of serviceability,

comfort and traffic safety are estimate d with

the help of the coefficient of engineering

structure and provision of the necessary

facilities (Cnf) and with the coefficient of the

maintenance degree (Cm).

For the generalized complex estimation of

road quality and its maintenance degree they

determine index of road quality and condition

(I) which includes a complex parameter of

transport quality and operation condition (CI),

a parameter of engineering structure and

provision of the necessary facilities (C nf) and a

the coefficient of the maintenanc e degree

(Cm):

I = CI . Cnf . Cm (3)

Procedure of estimation of transport

quality and operation condition of road

In the process of development of this

procedure the most complicated methodical

problem has been solved. The problem was to

find a mode of reduction of effect of various

parameters and figures on the consumer

characteristics to one quantity indicator

describing these characteristics.

Transport quality and operation condition

of each characteristic road section i s estimated

by a total coefficient of provision of speed

ratingtotalpciC which is assumed as a complex

index of road transport quality and operation

condition on a given road section

totalpciCiCI .

For deriving total TQF index they

calculate partial indices taking into account

the effect on the main index of following

parameters and characteristics: widths of

carriageway - Ccl1; width and condition of

shoulders - Ccl2; traffic volume and traffic

composition - Ccl3; longitudinal grades and

visibility of a road area - Ccl4; radiuses of

horizontal curves and superelevation - Ccl5;

roughness of pavement - Ccl6; skid resistance

coefficient the pavement - Ccl7; condition and

strengths of road base - Ccl8; depths of ruts -

Ccl9; traffic safety - Ccl10.


Maximum traffic speed is determined by

a computed method or by measuring the speed

on the road. Most often in use is a computed

method when partial indices of rated speed

provision are determined on the basis of data

about parameters and characteristics of the

road gained by direct measurements and road

conditions surveys.

During primary survey they collect data

on all parameters and characteristics of a road,

and during the following ones they collect

data only on variable parameters and

characteristics. All the information enters in a

line graph which later on is only adjusted.

The problem of an estimation of the

effect of each parameter on traffic speed is in

determining the physical sense and

mechanism of such effect, choosing a

calculation scheme and giving some

mathematical description that allows defining

a top speed of a design car. By dividing

maximal possible speed by a base design

speed they get a coefficient of rated speed

provision.

As an initial one is taken a scheme of

movement of a single or a first car in a group

of cars that goes along the lane with a

maximal possible speed which is determined

by the effect of parameters, characteristics and

a condition of road on interacting of the car

with the road, on driver’s psychophysiological condition and his perception of

the circumstances on the road, side effect of

the cars going on adjacent lane and possible

restrictions from cars going ahead on the lane

(longitudinal effect).

During measurements on the road they

accept as a maximal possible speed a speed of

85% of provision for a single car ahead of the

group of cars or traffic speed of 95% of

provision.

Explicit numbers of indices of rated

speed provision are first gained by calculation

on the base of known or again determined

interrelations, schemes and formulas where

the parameter to be estimated is an argument,

and function is a speed of a car, and then

indices are mustered experimentally by

measurement of actual speed of cars on the

roads.

It is determined that all the estimated

parameters of the road according to their

character of the effect on the scheme of car

movement can be divided into 4 groups:

1. Parameters that effect cars or driver

and through them they influence on traffic

scheme and first of all on speed not

interrelating with other parameters. Among

such parameters are:

Roughness and strength of

pavement.

Traffic flow and its composition on

the main lane (longitudinal restrictions).

Axel load (total permissible vehicle

weight).

2. Parameters that in correlation with one

or more others effect driver’s perception ofroad and decision about driving mode. Among

such combinations are:

Widths of carriageway and strength

edge strip.

Width and condition of shoulder .

Widths of carriageway and traffic


volume on adjacent lane.

Visibility and skid resistance

coefficient on a horizontal section etc.


or more others effect on interaction of a car

with the road and on speed. Among such

parameters are:

Longitudinal grade, rolling

resistance and skid resistance coefficient on

upgrade.

Radiuses of horizontal curve,

superelevation and transversal skid resistance

coefficient on a horizontal curve and others.

4. Parameters and their combination that

simultaneously affect on interaction of a car

with the road, driver’s perception of road andthrough these factors on the car moving.

Among such parameters are:

Longitudinal grade, skid resistance

coefficient and visibility of road surface on

the downgrade movement.

It is determined that a driver chooses

speed mode estimating the entire situation on

the road in total. But in difficult conditions

one single parameter or their combination

have here the most influence. That is why an

estimation of effect of each parameter in

different combinations was carried o ut and the

most unfavorable of them were determined.

On each characteristic road section they

get 10 partial indices of rated speed and on the

base of these indices was defined transport

quality and operation condition CI:

CI = f(Ccl1, Ccl2, …, Ccl10).

To solve the problem calculations in

three models were analyzed and compared:

The first model is a multiplicative one

where summarized values of CI are

determined by multiplying all partial indices.

Theoretical premise of this model is a

conjecture that parameters of road render

cumulative effect on traffic speed according to

distributive law.

The second model is where the

summarized value of index of design speed

provision is obtained as the least value from

10 partial indices of design speed i.e. one of

the parameters of the road or a combination of

parameters unified in one partial index on this

section effects most of all on traffic speed or

safety.

The third model is an expansion of

function of summarized index of design speed

provision in a Taylor series limiting the series

to member that are not higher than third order.


conjecture that road parameters render

complex aggregate effect on traffic speed that

can be estimated through enumeration of their

different combinations with limits regarding

dual and triple interaction. And at the same

time on every section 3 coefficients that have

the least value are chosen from 10 partial

coefficients.

The functional test of models has been

carried out by comparison of the results of

calculation from all three models for the same

road sections to the results of measurements

of actual traffic speed on these sections. In

total there were chosen about 50 road sections

of different categories in various regions of

Russia.


The results of the test showed that the

first model gives significantly understated

values of summarized index of rated speed

provision especially in range of low values.

Difference of calculated values from actual

attains 40-50% and more.

The third model gives results that are

most closely congruent with the results of

actual measurements. The mathematical

expectation of discrepancies is 0,004, with a

variance of a discrepancy is 0,003, and a mean

square deviation of discrepancies is 0,057.

The second model gives resul ts that are

close enough to the results of actual

measurements. The mathematical expectation of

discrepancies is 0,045, with a variance of

discrepancies is 0,004, a mean square

deviation of discrepancies is 0,06. However

the second model in comparison with the third

has essential virtue which consists in its

elemental simplicity and accessibility. At the

same time results of determining the

summarized index of rated speed provision

coincide well with actual results. Therefore

the second model has been accep ted as a

working one.

Thus, value of the summarized

coefficient of design speed provision on each

road section is accepted equal to one of 10

partial coefficients that is the least in value.

A graph is drawn to visualize reduced

design and cutting of road, critical parameters

and characteristics, partial and summarized

values of coefficients of design speed

provision on each road section, and al so

superimpose lines of normative and maximum-

permissible least parameter of CI.

The analysis of this graph easily allows

to determine road sections that match and

mismatch the demands for consumer

characteristics of roads, to specify

quantitatively extent of misfit, to determine its

reasons and simultaneously to assign on each

section a complex of provisions on

elimination of all or some part of the

determined deficiencies and on adjustment of

road to complete or partial correspondence to

normative demands.

On the basis of stated above the

procedure of estimation of transport quality

and operation condition of road ne twork

maintained by a road agency, road network of

separate territory or a road network of the

country as a whole is developed.

Determining the index of engineeringequipment and provision of the necessaryfacilities. An index of provision of the

necessary facilities and engineering

equipment (Ieq) is determined on the base of

the value of index of imperfection of

compliance of engineering equipment and

provision of the necessary facilities on the

road (Ic).

Imperfection of compliance here is

absence, insufficient amount or misfit to

normative demands to parameters,

constructions and to a feature placement of

engineering system and provision of the

necessary facilities.

The coefficient of imperfection of

compliance of engineering system and

provision of the necessary facilities is

determined by results of diagnostic study of

roads under formulas:


)sIi(I8

1eq.I (4)

721i iIiIiII (5)

Ii – partial index of imperfection of

compliance of places for rest which functional

influence is spread for significant road

sections; Ii1 … Ii7 – partial indices of

imperfection of compliance of elements of

engineering equipment which functional

influence is spread on short road sections

(crossings, junctions, entries and at -grade

intersections, bus stops, barriers, sidewalks

and foot - paths in settlements, road

carriageway marking, road lighting, traffic

signs).

The value of index of provision of the


equipment (Ieq) for each kilometer of road is

assumed depending on the value of Ic

according to Fig. 1 and place in a line graph

of motor road quality and condition

estimation.

Fig 1. Graph for determining the value of I eq: I, II,

III, IV, V – categories of roads

Determination the index ofmaintenance (Io).This index is determined on

the ground of the results of the road

conditions estimation performed by a

specially appointed commission in

compliance with the active regulations on the

road conditions estimation.

To estimate the road on the ground of the

visual examination the commission uses rates

5, 4, 3, 0 depending on the defects of the main

elements of the road revealed: the roadbed

and drainage system, pavement, engineering

structures, environment and engineering

machinery, landscaping and planting.

First, each element of the road on each

section is estimated, and then the aggregate

index of maintenance on each particular

section or kilometer is rated:

554321 РРРРРР

(6)

where Р1, Р2, Р3, Р4, Р5 – index of

maintenance of each road element.

The index of maintenance is then

rendered into the index of operational

maintenance (Fig. 2).

Fig 2. Diagram for determination of (Io)

Overall estimation of highway qualityand condition

Overall index of quality and condition


for each road (road section) is determined

from the formula 3.3.

The degree of compliance of factual

road application properties with the

regulations is estimated by the relative index

of road quality and condition:

rCI

rIrQ (7)

where Ir - overall index of road quality and

condition; CIr – regulatory requirements to

aggregate factor.

The road fully meets the regulatory

requirements if Qr ≥ 1.

Increase in the overall index of road

quality and condition is determined from the

following formula:

100%rI

brIe

rIrΔI

(8)

where erI , b

rI - overall indexes of road quality

and condition as at the beginning and the end

of the period under report.

On the ground of the road and network

quality and condition analysis main

tendencies in enhancement of traffic

operational characteristics, the order and

sequence of maintenance, repair and

rehabilitation works are planned.

The methods of repair and

reconstruction planning as well as the

methods of strategy validation under the

limited financial and material resources were

developed based on the results of the road

condition diagnostics and estimation.

There is software making it possible to

perform the condition estimation and

technical and economic analysis using PC.

IV. CONCLUSION

The method for diagnosing and

estimation of the road condition according to

the degree of compliance with the demands of

the road and service users was developed, the

end objective of the above method is to

enhance the road application properties.


estimation of highway quality and condition

according to the application properties was

approved as official regulation in Russia,

Belarus and Kazakhstan.

Alone on the territory of Russia

approximately 30-40 thous. km of roads are

yearly estimated applying this method. A

system of diagnostic centers a full set of

instruments, equipment, mobile laboratories

and computing, making it possible to

implement all the aspects of the method, was

set up. The practical work in the field of

diagnosing and estimation of federal highway

network condition is performed under

supervision of Road Research Institute of

Russia where the federal data bank on roads

was established. The annual plans for repair

and maintenance works are developed based

on the results of road conditions estimation.

References

[1]. Nadezhko A.A. etc. Road Science. ReferenceEncyclopedia. Vol. IV/A.P.Vasi lyev,V.D.Kazarnovskiy, etc. Science editorA.A.Nadezhko. –M.: Publ.House

“INFORMAVTODOR”, 2006. -393 p

16

CT 2


I. APPROACH

Investigating the slope stability of the road is a problem which has been researched sincelong time. There had existed many methods but the most popular in over the world isW.Fellenius’ proposal based on “slide circle arc” suppose [1],[2],[ 3],[4],...

A specially improtant issue when applying this method is determining position of the mostdangerous slide arc that is its center, hereby called “caculation slide center” and correspondentlyis “caculation slide arc”. However, it is so complicat e that up to now there have been many roadscientists proposing different methods to specify this :

According to W. Fellenius: “calculation slide center” bases on a line and he proposed themanner to specify it as fig. 01.

Some experts as Gonstein, D.W. Taylor, N.N. Maslov, N.A. Txytovitr, G.Pilot…established the monograps, tables or shew the lines on which calculation slide center existing.

In the past, that calculation based on experience, is complex and large quantity; it is limitedby computation facilities, thus the quantity of investigation had been not so enough. In otherhand there has not been instruments which has ability to specify whether minimum factor ofsafety (or calculation slide centre”) is correct or not?

Stability of road slope is influenced by many factors: road bed elevation, talus grade;height, width of banquette lateral; gravity unit, friction angle, cohesive force of soil; ponding

RESEARCHING ON POSITION OF CALCULATION SLIDE CENTERIN COMPUTING THE STABILITY OF ROAD BED

BY SLIDE CIRCULAR ARC METHOD

Assoc. Prof. Dr. TRAN TUAN HIEPUniversity of Communication and Transport, Vietnam

Abstract: Investigating the slope stability of the road is a problem which has been

researched since long time. There had existed many methods but the most popular in over the

world is W.Fellenius’ proposal based on “slide circle arc” suppose. A specially improtantissue when applying this method is determining position of the most dangerous slide arc tha t is

its center, hereby called “caculation slide center” and correspondently is “caculation slidearc”. However, it is so complicate that up to now there have been many road scientistsproposing different methods for that. This presents the result of resea rching to discover the

position of calculation slide center in computing the stability of road bed by slide circular arc

method.

Key Word: Road bed stability; Embankment stability; road slope stability


CT 2

situation…

It is problem that: situations of different embankments, essentially, how different stabil ityof road bed are?

Specially, key issue of this problem: how to properly determine the position of “calculationslide arc”?

Methodology:

Firstly, we establish utility software for automatically determining factor of safety of roadbed. This program is flexible instrument for computing and then being used to investigate thestability of various cases such as factor of safety, the position of calculation slide arccentre…according to the various parameters of embankment.

II. SOME MANNERS TO SPECIFY POSITION OF CALCULATION SLIDE CENTER

2.1. Proposal of W. Fellenius

According to W. Fellenius, position of calculation slide center is determined as below (refer tofig. 01):

Firstly, determining line EF; E is specified as figure with a depth H and a length 4.5H fr omtoe of talus; F is specified by the angles 1 and 2

Correspondingly with fill slope (referring to table 01)

Table 01. Values of 1 and 2

Talus grade 1 (Deg.) 2 (Deg.)

1:5

1:3

1:2

1:1,5

1:1

25

25

25

26

28

37

35

35

35

37

Fig 01. Determining Calculation slide center by W. Fellenius

18

CT 2


On line EF lengthening, to suppose the points as the centers of slide arcs: O1, O2, O3 ...Oi,.... On…(distance between them may be 0.25H -0.3H). Corresponding to center Oi, establishingthe slide arcs then specifying their factor of safety (K) and determining minimum value of them(Kmin). With n of supposed centers we will get n of Kmin . Using EF as coordination axle,based on the centers Oi to establish the graph of the values of Kmin then determine minimumvalue Min [Kmin(i)] and we will have center Omin1 on EF, correspondingly.

Basing on Omin1 to draw line CD being perpendicular with EF. Basing on CD to repeatgradually trial process as above, we will determine the point Omin2, center of the mostdangerous slide arc.

This method has not instructed the location of firstly supposed centers O1, O2… on EF,more over computation quantity is so large.

2.2. Some other proposals

To determine calculation slide arc (the arc which has minimum facto r of safety), there areothers manners being presented as: graphic, tabulate,… by D.W. Taylor, Gonstein, G. Pilot [1],[2], ... among them attention to two conclusion should be paid:

1. As proposal of G.Pilot, D.Taylor: calculation slide centers will ba se in vertical line MN whichis across middle point of fill slope and perpendicular with bottom line of road bed (Fig.02 ), [3], [4].

2. As proposals of other scientists, scope of calculation slide center is bisector PQ of angleNPK; PK is perpendicular with talus at middle point P (refer to Fig. 3)

Fig 02. Scope of calculation slide center by Pilot, Taylor

Fig 03. Scope of calculation slide center by experiment


CT 2

In spite of gradually trial determi nation of calculation slide arc, the manners above

mentioned helped to limit the scope of finding calculation center out and pointed out that scope

are the lines. However, they just only considered particular cases and the results were not

coincided. In fact, by our calculation many cases, it is proved that calculation centers have not

been in scopes pointed out by those methods.

III. ORIENTATION FOR INVESTIGATING POSITIONS OF CALCULATIONSLIDE CENTERS

Using Utility Program, after pilot trial calculating mo re than 200 problems of different road

beds which are various by dimension, fill soil mechanic, natural soil properties; we discovered that

there are many cases their calculation center position were relatively suitable with G. Pilot’sProposal, but so much other one, they are different. In other hand, should be noted that G. Pilot only

investigates with fill up soils being grain (C=0); and natural soil are merely cohesive soil ( =0); the

influence of flooding permeable water compressor also is neglected. T hus the result may not be

correct with general case (fill up soils and natural soils have ,C; and ponding water compressor…).

By experimental computation on PC, we also invent that the scope of Calculation slide

center can be determine definitely by bas ing on two orient lines (refer to Fig.4):

- First line (line I) is perpendicular with base line of embankment and across through

“equivalent talus”

- Second line (line II) is perpendicular with base line of embankment and pass through

middle point of banquette lateral

Position of lines I and II are specified by xI and xII

“Equivalent talus” can be understood as below:

In general form, embankment has banquette (with general talus form: MNEFH) and can be

converted in to form of homogeneous talus with out banquette (line IPQK). P and Q in turn are middle

point of sub taluses NE and FH; so we have two couples of triangle which are equal each other: S1=S2

and S3=S4. It is proved by Varinogn Law that two these forms are equivalent in slope stability.

Fig 04. Orient schema for investigating the position of calculation slide center

XI

XII

I II

20

CT 2


IV. EXPERIMENTAL COMPUTING TO DETERMINE THE POSITION OFCALCULATION SLIDE CENTER

After orientation of investigation scope, experimental computation and ste p 2 of research

are conducted. In this research, all parameters such as height of embankment; height and width

of banquette; talus grade; traffic load; gravity unit, internal friction angle, cohesive force unit of

fill soil and natural soil,… are in turn v ariable; then by using utility software, corresponding to

those cases, the investigation were implemented to specify the rule of the scope which will

cover all calculation slide centers.

Because the rule of position of calculation slide centers have been d etermined yet, so in this

stage we have investigate on very large grid of slide center to cover all points may be

calculation center.

The cases of experimental computing and investigating are:

1. Position of Calculation Slide Center (PCSC) is in cases o f various slope grades;

2. PCSC is in cases of various height of road bed (H);

3. PCSC is in cases of various height of banquette;

4. PCSC is in cases of various width of banquette;

5. PCSC is in cases of various cohesive unit of fill soil of road bed;

6. PCSC is in cases of various internal friction angle of fill soil;

7. PCSC is in cases of various cohesive unit of natural soil;

8. PCSC is in cases of various internal friction angle of natural soil;

Deriving from experiment computing results we discovered a rule:

It exist an Area of Calculation Slide Center (ACSC), in which, all positions belonging to

are ACSC have minimum factor of safety.

The coordination of ACSC is as below:

(xI - 1) O.x (xII + 1)

H O.y 2.5H

V. RELIABLE EVALUATION OF ESTIMATION

To evaluate the correct and reliable level of estimation, many computing problems are


CT 2

implemented and planned as below:

B = 10 20 (m)

H = 2 8 (m)

L = 0 20 (m)

h/H = 0 0.5

Tg = 0 1

hd = hh = 0

hd - hh = 0 0.4

H1/H = 0.5 2.0

Cdry = 2.0 5.0 ( T/m2 )

dry = 14 22 ( deg. )

Cn1 and Cn2 = 0.5 6.0 ( T/m2 )

n1 and n2 = 8 20 ( deg.)

Of which:

B - road bed width

H - road bed height

L - banquette width

h - banquette height

Tg- slope grade

hd and hh- highest and lowest water level

Cdry - cohesive force unit of dry fill soil

dry - internal friction angle of dry fill soil

Cn1 and Cn2 – cohesive force unit of natural soil lay 1 and lay 2

n1 and n2 – internal friction angle of natural soil lay 1 and lay 2

Criteria of natural soils of the layers are presented in table 02

22

CT 2


Table 02. Criteria of natural soilsStae of soil Criteria Clay Loamy Sandy

C 6.0 40 2.0Semi-hard0< B 0.25 20 23 28

C 4.0 2.5 1.5Plast-firm0.25< B < 0.5 18 21 26

C 2.0 1.5 1.0Plast-soft0.25< B <

0.75 14 17 24

C 1.0 1.0 0.5Plast- liquit0.75< B < 1.0 8 13 20

That planning can cover all problems existing in fact. The data are random selected.

Dimension of test samples is: n = 250; of which: 6 cases are out of estimated area withdeflection is 1m; thus error is less than 1%.

Error ratio is: f = 6/250 = 0.024.

Deducing from that: to gain the correctness of estimation: = 0.02 then reliability of

estimation () is as below:

2,07f)f(1

nεt

Referring to table of Laplast integral:

zz(t) = 0.4808

= 2 (t) = 0.962

Thus, the estimation of rule of ACSC above mentioned has error about 0.7% 4.7%; andreliability is: 96.2%.

VI. RECOMMENT AND CONCLUSION

Based on experiment research, investigating many cases we have t he conclusions:

1. Position of calculation slide center is various when geometric parameters of road bed,mechanic and physical criteria of soil (fill soil and natural soil) are changed. However almost ofthem is located in specific scope. Coordinati on of this area is specified as below:

Abscises are limited by two lines which are perpendicular with base of road bed and passthrough middle point of banquette and “equivalent talus”;

Ordinate is limited by height of road bed (H) and 2.5 H

To have high reliable this area need to be wider, thus it should be:

(xI - 1) O.x (xII + 1)

H O.y 2.5H


CT 2

Hn

4,5Hn

Hn1/1,5

1 601 2 3 4 5 6 7 8 9 10

7

8

9

10

11

12

13

14

15

16

1 87

2 12

2 53

2 87

2 98

3 07

3 20

3 33

3 92

1 47

1 54

1 66

1 84

2 12

2 39

2 68

2 72

2 89

2 99

1 43

1 45

1 49

1 56

1 66

1 84

2 11

2 29

2 56

2 66

1 42

1 41

1 44

1 48

1 53

1 62

1 74

1 86

2 83

2 28

1 44

1 44

1 44

1 47

1 50

1 56

1 64

1 73

1 82

1 92

1 48

1 47

1 46

1 47

1 50

1 54

1 61

1 67

1 76

1 84

1 52

1 50

1 49

1 51

1 52

1 55

1 59

1 66

1 71

1 80

35°

26°

Fig 05. Result of Computing case No.110

2. To review the position of calculation slide center proposed by W. Felenius and other autho rs:

Using utility software, by experiment computing many cases, drawing the grids withminimum factor of safety corresponding to their intersect -points; synchronously establishing thecalculation slide center by W. Fellenius for comparing.

It is shown in Fig. 05 as an example. Those experiment computations presented that:

W. Fellenius’ proposal has not given minimum factor of safety as request and has adeflection of position of calculation slide center . More over, Fellenius’ proposal is only appli edfor particular case of talus is line (with out banquette).

3. By the experiment computations we discovered that:

It is existed an area of calculation slide center in which any point has calculation factor ofsafety is similar and equal minimum factor of safety.

We called that area is Area of Calculation Slide Center (ACSC) and determined it asabove mentioned.

The ACSC is able to apply correctly for general case of road bed with banquette. In caseof talus is merely line, the proposal of Taylor and Pilot is similar.

References[1]. G.Pilot, M.Moreau, Remblais sur sols mous equipe de banquettes laterales, CPC, Paris, 1973 .[2]. D.W. Taylor, Fundamentals of soil mechanics, Newjork - London, 1954.[3]. Tran Tuan Hiep, Research to automize optimization design of road bed, Ph.Dr thesis . University ofCommunication and Transportation, Hanoi, 1993.[4]. Piere Lareal, Nguyen Thanh Long, Nguyen Quang Chieu, Vu Duc Luc, Le Ba Luong . Remblaisroutiers sur sols compressible dans les cond itions du VietNam, Insa de Lyon, 1989

Hn

4,5Hn

Hn


During last two decades, China haswitnessed rapid developments in pavementengineering with upgrading of transportationmarket and expansion of transportationinfrastructures including highways, city roadsand airports. Total length of highways inChina reached 1,930,500 km with 132,674 kmof National Trunk Road connecting majorcities at the end of 2005[1].Construction ofthe first freeway, Hu-Jia Freeway in MainlandChina started in 1984 was completed in 1988linking Shanghai municipal center to Jiadin g,a suburban district 20 km away. The “7918Program” [2], China’s National FreewaySystem as shown in Figure 1, was approvedby the State Council on Dec. 17, 2004 and hasa total mileage of 85,000 km. The “7918Program” was made to meet the nationalmodernization goal in the middle of 21st

century, and China will have a freeway of0.83 km/100 km2 by then, a standard roughlyequal to the current freeway system in US[]given in Figure 2. According to China ’s 2005highway statistics[1] issued by Ministry ofCommunications in May, 2006, total mileageof China’s freeway has reached 41,005 km,which is ranked the second longest freewaysystem in the world just after US with afreeway system of 91,285 km (56,699 miles)in 2005’s Highway Statistics[3] issued by USfederal Department of Transportation. Besidesthe huge task of completing domestic highwaynetwork, the Asia Highway Route[4] asillustrated in Fig. 3 was approved inNovember, 2003 connecting 32 countries inAsia with a mileage of over 140,000 km,among which China has 26,000 km in Fig. 4,almost one-fifth of the whole program.

THEORETICAL DEVELOPMENT AND ENGINEERINGPRACTICE OF PAVEMENTS IN CHINA

YANJUN QIU, A.M. ASCESchool of Civil Engineering,Southwest Jiaotong UniversityChengdu, 610031, China

Abstract: This paper presents a comprehensive review of historical theory development

and current construction practice of pavement engineering in C hina. Mechanical models,

design guides, construction techniques, evaluation methods and maintenance standards are

elaborated for PCC pavements and AC pavements. Differences in design methodology among

pavements of rural highways, urban roads and airport fields are discussed based on service

requirements. Lessons and experiences based on past 20 years ’ construction practice and

pavement performance are summarized. Current research areas in pavement engineering

associated with unconventional geological and/or landscaping in China ’s highway

construction and national strategic plan for pavement engineering are also covered in this

paper.

Key words: Pavement Engineering, Asphalt Concrete Pavement, Portland Cement

Concrete.


At the same time, pavements for urbanroads expanded to 247,000 km in China ’s 661cities by the end of 2005 with an average roaddensity of 10.93 m2 per capita[5]. Current cityresidents accumulated to 358,940,000 whilecity land covers 412,700 km2 in 2005[5].Routine maintenance, rehabilitation, resurfacing,reconstruction, and new construction of roadpavements will continue to grow with the fastdevelopment of urbanization andindustrialization in next decades. Urbanizationratio will be increased from current 41.7% to60% during the national program of “Socialist

New Village Program”[5], which obviouslyrequires more urban roads to be built.

In airfield sector, there are 142 certifiedairports with operating air routes connectingmajor cities in the world and/or majordomestic cities in mainland China by the endof 2005[6]. Among 133 airports, nearly 60 areclose to design capacity or oversaturated withrapid expansion of air traffic. Air passengersreached 241,935,000 and cargo transportationexceeded 5,526,00 tonnage with a platoon of863 commercial jetliners. Annul increase inair transportation is expected to be 14% in the

Figure 1. China’s National Freeway System Figure 2. Eisenhower Interstate System

Figure 3. Asian Highway (AH) Route Map Figure 4. China’s AH routes

Transportation


next 5 years in China, which leads to anincrease of commercial jetliners to 1580 anddozens of runways to be expanded,rehabilitated, and/or newly built[6]. A Boeingprojection in 2002[7] expected that 2320commercial jetliners will be required in 2022to support the second largest air transportationmarket in the world, right after US. Severalmetropolitan airports such as ShanghaiPudong International Airport, Beijing CapitalInternational Airport, Guangzhou BaiyunInternational Airport, and Chengdu ShuangliuInternational Airports are potential candidatesto develop hub-and-spoke system (HSS) inorder to become a key node in world air routenetwork.

I. DEVELOPMENT OF DESIGN THEORY

Portland cement concrete (PCC)

pavements and asphalt concrete (AC)

pavements are the two major pavement

categories in China’s pavement engineering.

With the decrease of pavement’s share in total

highway costs and understanding of life -cycle

cost analysis, AC pavements have gained

more and more applications in city roads and

airports as well as rural highways. Data from

highway statistics in China show that there are

532,697 km of high-type pavements with AC

pavement of 226,075km and PCC pavements

of 306,622 km, which account for 42.44% and

57.56%, respectively. All mid-type pavements

are asphalt roads with a total length of

461,901 km. PCC pavements used to be the

only type of paving of airfield including

runways, taxiways, and aprons in China ’sairports including civil transport airports and

military airports. However, the successful

application of AC pavements in runway in

Beijing Capital International Airport in 2001

proved an effective alternative in design of

pavement types in runway construction. AC

overlay became a competitive alternative in

airport upgrading since then due to its better

serviceability. With the development of

structural analysis of pavements, design

philosophy evolved from empirical methods

in early days into mechanistic -based

methodologies. Table 1 lists pavement design

guides of various sectors current used in

China, structural analysis theory of pavements

varying from CBR (California Bearing Ratio)

method to FEM (Finite Element Method)

method. It should be noted that highways in

China refers to roads in rural areas while

urban roads is used to address in cities.

Ministry of Communications of China’s State

Council and corresponding DOT (department

of transportation) authorities in provincial and

county governments are responsible for

“rural” highways including issuing design

specifications for highway pavements. At the

same time, Ministry of Construction and

corresponding municipal authorities are

responsible for “city” roads including

publishing design guides for urban roads. In

the early days, design guide of highway

pavements was also used for urban roads,

forest road, airport and other sectors of

industry. Between middle of 1980s and early

of 1990s, separate design guides for urban

road, forest road, industrial road and airport

were developed and most of them borrowed

ideas of highway guide, AASHTO guide [8],

FAA guide and guides of other countries.

Basically, AC pavements were analyzed based

on Burmister’s layer theory[9-12] while PCC

pavements were designed based on

Westergaard’s solution or elastic solid

foundation[13-21].


II. PCC PAVEMENTS

Five types of PCC pavements, namelyJPCP (jointed plain concrete pavement),JRCP(jointed reinforced concrete pavement),CRCP(continuous reinforced concretepavement), RCCP(rolling compacted concrete

pavement), and PRCP (prestressed reinforcedconcrete pavement) have been constructed inChina. JPCP is the most widely-used PCCpavement for highways, urban roads, andrunways. Design theory was based on elastictheory of Kirchhoff thick plate with a modulusof rigidity D, which is governed by Equation(1)[12,14,21]. Base layer and underlying

Table 1. List of pavement design guides of various sectors in China

Sector CategoryPavementCategory

Current Design Guides Code Updated Year

AC Specifications for design of highwayasphalt pavement

JTG D50-2006

2006

Highway (rural)

PCCSpecifications of cement concretepavement design for highway

JTG D40-2002

2002

AC Specifications of asphalt concretepavement design of civil airports MH 5011-1999

1999

Commercial airport

PCCSpecifications for cement concretepavement design for civil transportairports

MHJ 5004-95

1995

Military airport PCC Specifications of design of cementconcrete pavement for militaryairport

GJB 1278-91

1991

Urban road AC, PCC

Specifications of design of industrialroads

CJJ 37-90 1990

Forest road AC, PCC

Design specifications of roadsurface for forest road

LYJ 131-92 1992

Industrial road AC, PCC


GBJ 22-87 1987

Note: (1) New design guide for AC highway pavement JTG D50-2006 implemented starting Jan. 01, 2007.

(2) Industrial roads refer to roads in factories and mines


subgrades can be assumed as elast ic half space(Hogg’s solution), Winkler foundation (denseliquid foundation, Westergaard ’s solution),Pasternak foundation, or multi -layer elasticfoundation, leading to different equations offoundation reaction, p(r) under axisymmetricloading of q(r) in Equation (1).

rprqrwD 4 (1)

Where

dr

d

rdr

d

dr

d

rdr

d 11 22224

Design guide for highway pavements wasfirst published in 1958 with 1954 version ofthen Soviet Union’s pavement design guide asblueprint[14]. The guide was revised in 1966(JT1004-66) and served as the only existingguide for pavements including AC and PCCpavements for highways, urban roads andairports. A separate PCC pavement designguide, Specifications of cement concretepavement design for highway JTJ012 -84, wasissued in 1984. Structural analysis was basedon theory of elastic think plate over elasticsolid foundations. Critical loading positionwas situated at the middle of transverse joints.Thickness calculation is solely based onloading stress without considering curlingstress. In 1994, the guide was revised asSpecifications of cement concrete pavementdesign for highway JTJ012-94 and curlingstress was included into thickness designconsideration. Critical loading position wasselected in the middle of longitudinal joints.In 2002, design guide was revised again usingreliability concept instead of factor of safety.Thickness is determined by limiting the

combination of loading fatigue stress pr and

curling fatigue stress tr to concrete strength

fr with a reliability coefficient of r inEquation (2).

r (pr + tr) ≤ fr (2)

Pavement maintenance has to meetstandards in “Technical specifications ofcement concrete pavement maintenance forhighway JTJ 073.1-2001”. RQI (ride qualityindex) in terms of IRI (international roughnessindex), TD (textural depth), PCI (pavementcondition index) and DBL (ratio of breakingplates), and representative deflection are usedto evaluate roughness, skid resistance, distressand strength of pavement structures,respectively. For freeway system, a separatespecification, “Expressway maintenancequality evaluation standards”, should be usedfor quality requirements in maintenance work.

Distress is evaluated based on type,amount and severity and different weighfactors are allocated to specific type ofdistress with certain amount and severity incalculating PCI. Major distresses found inPCC pavement include fatigue cracking,pumping, faulting, corner breaking, jointspalling, scaling, load-transfer deterioration,depression, durability cracking, pop-outs andreactive aggregate distress.

III. AC PAVEMENTS

Asphalt pavements used in China includeasphalt concrete (AC), asphalt macadam(AM), asphalt penetration, and asphalt surfacetreatment. Other types such as SMA (stonemastic asphalt), OGFC (open graded frictioncourse), cold mix, color mix, slurry seal,micro-surfacing, and recycled asphalt


pavement. Burmister’s layer foundationtheory[15-17] was used these days for asphaltpavement design for highways and urbanroads. Airport engineers usually use CBRmethod to design AC runway pavements inaccordance with MH 5011-1999.

Design guide for asphalt pavementswere included 1958 version of “Specificationof pavement design ” and revised version“Specification of highway pavementdesign(JT1004-66)”. In those two versions ofdesign guides, an elastic half space wasassumed and M.H.Yakunin’s solution in 1941based on Boussinesq’s theory was adopted tosolve structural responses in Equation (3).Vertical stress at a depth of z from half space

surface z under load intensity of p appliedover a single circular area of diameter z wasapproximated using a =1.5 in Equation (3).Practical application of design guide in 1950 ’sand 1960’s found that Yakunin’s solution wasnot sufficient to conduct structural analysis offlexible pavements.

2

D

zα1

pzσ

(3)

Field investigation, road tests, theoretical

analysis and laboratory research had been

conducted since 1968 and a new design guide

was issued for practical use, “Design

specification of flexible pavements for

highways (interim guide)”, in 1978. The

structural theory of 1978’s guide was based on

Burmister’s two-layer system. Surface

deflection in the middle of tandem tires was

used to check thickness. A nomograph was

provided in the guide. With the development

of computers and construction practice of

pavement in highways and urban roads, a

1986 version, “Design specifications of

flexible pavements for highways (014 -86)”,

was issued. Burmister’s three-layer system

was used to solve structural re sponses.

Allowable surface deflection, together with

bending stress at the bottom of asphalt layer

was used to calculate pavement thickness.

Different interlayer status could be assumed

based on construction process and specified

methods in analyzing pavements. In order to

accommodate tire application characteristics

in urban roads, a shear index at the pavement

surface was introduced with Mohr -Coulomb

strength criterion for AC surface layer. In

1997, another version of pavement design

guide, “Specifications for design of highway

asphalt pavement”, was published using

design software APDS (asphalt pavement

design software) based on Burmister ’s multi-

layer system. Design deflection instead of

allowable deflection was used to evaluate

pavement structural integrity. In this guide

which played a key role in the rapid expansion

of China’s freeway system, asphalt pavement

with semi-rigid base was emphasized to

reduce AC surface layer by increasing base

strength. In 2004, new AC design guide based

on engineering feedback, research inputs,


lessons and experience of pavement industry

abroad was drafted to seek reviews from

industry and institutes. This draft,

“Specifications for Design of Highway

Asphalt Pavement (JTG D50-2004)”, attracted

controversial discussions from various sides

based on different research backgrounds and

versifying engineering practices. Many new

ideas based on field investigations, research

findings and pavement performance in last 10

years were incorporated. Four base types,

semi-rigid base, flexible base, rigid base and

composite base are recommended. In

accordance with new construction guide,

“technical specifications for construction of

highway asphalt pavements (JTG F40 -2004)”,

gradation was adjusted compared to JTJ014 -

97. Finally, JTG D50-2006, was approved at

the end of 2006 by Ministry of Transportation

after long discussions and/or debates.

Maintenance work is governed by

“technical specifications for maintenance of

highway asphalt pavement JTJ 073.2-2001”

and PQI (pavement quality index) is used to

designate maintenance levels. PCI, RQI,

SSI(structural strength index), and SFC (side

friction coefficient) are calculated to evaluate

pavement condition, ride quality, structural

integrity, and traffic safety, respectively.

Automatic distress survey equipments, such

FWD (falling weight deflectometer) for

deflection and RTRRMS (response type road

roughness measuring system) roughometer or

profilometer for roughness, are recommended

for fast investigation of pavement condition.

Fatigue cracking, rutting, reflection cracking

and water damage are the primary four

distress types. Other distresses include low

temperature cracking, structural cracking,

bleeding, slippage cracking, and pumping.

IV. PAVEMENT RESEARCH IN CHINA

Major research topics in ear ly daysconcentrated in analytical solution ofpavement structures, based on theory ofelasticity of Boussinesq half space, Winklersystem, Burmister multilayer system, elasticsolid system. Influence charts, nomograph anddesign tables were published for designreference[6]. With the fast development offreeway system in China, asphalt pavementsgained more and more attention in researchfields as well as in industry. Mix design with aSHRP (Strategic Highway Research Program,of US) Superpave® understanding, pavementevaluation with PCI and PSI (present serviceabilityindex), transportation management system basedon Web-GIS (geographic information system),automatic pavement data acquisition, preventivemaintenance, and pavement recycling, all thesetopics became interests of research whichresulted in updating of specifications ofpavements with respect to policy, design,construction, supervision, acceptance, andmaintenance fields.

One special research interest in China isnon-conventional subgrade deformation andpavement responses in the context that Chinahas many mountainous areas, soft ground


zones, and existing highways with lowstandards. Literature reported research effortin embankments in sloped ground, cut -fillsubgrades, road widening over soft ground,crack propagation, and structured subgrades,see Figures 5-22 (note: All these illustrationswere cited from thesis-dissertation works ofMaster and PhD students of the author ’s).Figure 5 and Figure 6 show the difference indeformation characteristics of embankmentsover flat and sloped ground, which is acommon alignment result of roads inmountainous areas. It can be seen thatdeformation concentration at the downside toeof embankment constitutes a great concern inslope stability.

Figure 5. Deformation characteristics

of flat ground


of sloped ground

Soft and weak ground is one of the

traditional geotechnical challenges. Residual

settlement after construction completion poses

significant influence on structural

performance of pavements over embankment

in soft ground. Step by step consolidation

prediction can be predicted as shown in Figure

7 and Figure 8 and can be a practical reference

in embankment construction monitoring.

Figure 7 and 8 are simulation results of

embankment with a top width of 34.5m in

Chengdu 4th Ring Road Freeway (Chengdu

Bypass Freeway).

Cut-fill transitional subgrade sections arecommon practice in highway design inmountainous areas due to geometriclimitations and geological conditions. Trafficloading in pavements over cut -fill subgradesin Figure 9 will produce additional structuralresponses as shown in Figure 10 which in turnwill result in shortened pavement service life.

Major problems in road widening projects

as illustrated in Figure 11 are differential

deformation between existing roads and newly-

added section, especially in soft and weak

ground. Additional structural responses produced

in pavements as a result of differential

deformation behavior of subgrades will influence

pavement performance. In figure 12, it can be

seen that different widening scenarios will cause

different deformation curve across transverse

section.


II. NỘI DUNG

Nhằm mục đích phân tích tài chính vàphân tích kinh tế - xã hội, có thể phân chia cácloại dự án theo tính chất của công tr ình mà nóxây dựng là (i) các dự án xây dựng công tr ìnhcông cộng và (ii) các dự án xây dựng côngtrình dân dụng và công nghiệp. Các loại dự ánnày có các đặc điểm khác nhau cần tính đếntrong quá trình phân tích.

-20

-15

-10

-5

0

5

10

0 20 40 60

Embankment bottom widthm

Gro

und

settl

emen

tcm

layer 1 layer 2 layer 3 layer 4layer 5 layer 6 layer 7 layer 8

- 7

- 6

- 5

- 4

- 3

- 2

- 1

00 5 10 15 20 25 30 35

Top embankment wi dt h（m）

Emba

nkme

nt s

urfa

ce s

ettl

emen

t（cm

）

Load st ep 1 l oad st ep 2l oad st ep 3 l oad st ep 4

Figure 7. Consolidation process during

embankment fill

Figure 8. Settlement development

under traffic

0 1 2 3 4 5 6

-1400

-1200

-1000

-800

-600

-400

-200

0

200

400

600Ho

riza

onta

l st

ress

top

of

subg

rade /

kPa

X /m

1# 2# 3# 4# 5# 6#

Figure 9. Traffic loading on cut-fill sectionFigure 10. Structural responses

of cut-fill section

h

1:m

Y

X

-4 -3 -2 -1 0 1 2 3 4 5 6

-2.8

-2.6

-2.4

-2.2

-2.0

-1.8

-1.6

-1.4

Surfa

ce D

efor

mat

ion

(cm

)

Transver di st ance f rom cent er l i ne (m)

one-si de wi deni ngt wo-si de asymmet r i cal wi deni ngt wo-si de symmet r i cal wi deni ng

Centerline of roadway

Figure 11. Pavement over road-widening section Figure 12. Structural responses

of widened roads

Cut – fill boundary

1#

2#

3#

4#

5#

6#


Semi-rigid base structure has beenmainstream type of asphalt pavements since1980s. Main idea of semi-rigid base is toprovide a strong base to bear traffic loadingwhile surface AC layer is in a compressivestress state, which is very different fromtraditional flexible AC pavement structureswhere tensile stress produced at the bottom ofsurface AC layer controls fatigue cracking[5].Major problems with semi-rigid base ACpavement structures are reflection cracking.Figure 13 shows that AC layer in flexiblestructure will produce tensile stress at lowdepth while only compressive stress existed insurface AC layer in semi-rigid base. Figure14 suggests that tensile stress existed inflexible AC bottom will move down to semi-rigid base and surface AC thickness can thusbe reduced to produce economical benefits.

0

0.05

0.1

0.15

-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4

Horizontal StressSX/MPa

Surf

ace

Lay

er T

hick

ness

/m

Flexible

Semi-Rigid

Figure 13. Maximum horizontal stress of AC layer

0

0.1

0.2

0.3

0.4

0.5

0.6

-0.60 -0.50 -0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20Horizontal Stress

Dep

th/m

Flexible

Semi-Rigid

Figure 14. Horizontal stress at deflection point

Figure 15 and Figure 16 shows differencein structural responses of AC pavements

between contact layer interface andcontinuous layer interface using Burmister ’slayer model. China’s AC pavement designguide use only continuous model since 1986.Surface deflection between these two modelsexhibit significant state as shown in F igure 15while vertical stress are relatively similar.

Figure 15. AC pavement surface deflection

Figure 16. AC pavement vertical stress

Influence of contact area on pavementresponses is given in Figure 17 and Figure 18.Circular loading has been used in China ’s ACpavement design guides since 1958, which isalso a common assumption in other countriesas a result of availability of analytical elasticsolution in earlier days. Rectangular loadingcan be used with numerical methods tocompare the difference of pavement responsesamong various assumption of contact areabetween tires and pavement surface.

Crack propagation simulation usingcomputer simulation has been o ne ofpavement research fronts in last decade withthe development of damage mechanics and

X – Distancem

X – Distancem

Y–

Smax

.Pri

ncip

al P

aY

– D

efle

ctio

n m


simulation program. Figure 19 and Figure 20illustrate HMA crack propagation processusing APCPPS2D program and computertechnique used in the program.

Figure 21 and Figure 22 illustrateapplication of image recognition technique inpavement cracking research. Using imageenhancing method, raw pavement crackingimage can be processed to produce highquality data for computer recognition.

Figure 17. Shearing stress of circular loading

Figure 18. Shearing stress of rectangular loading

Figure 19. Crack propagation of testing

HMA beam

Figure 20. Element technique of simulating crack tip

Figure 21. Image Enhancing technique

Figure 22. Image threshold in MATLAB

Figure 23 shows the scanning process of

asphalt samples using computer tomography

(CT) method. This new methodology proves

to be an effective tool in analyzing the

cracking propagation process of asphalt

concrete under load application. F igure 24 is

the scanned cross section of one Marshall

specimen.

X – Distance m

X – Distance mY–

Max

She

arin

g St

ress

Pa

Y–

Max

She

arin

g St

ress

Pa


Figure 23. CT process of pressured AC specimen

Figure 24 . CT scanned AC specimen

V. LESSONS AND EXPERIENCE

Major experience in pavementengineering learnt from last 20 years ’ rapidexpansion of China’s transportationinfrastructures can be summarized as follows.

1. Overloaded truck axles have been topone factor to cause premature failure ofpavements in highways. Orchestrated effortsfrom various government authorities andrelated industries must be taken to ensure thetrucks rolling in pavements not to exceedspecified axle load standards. WIM (Weigh -in-motion) and strict truck and transportationpolicy can help reduce heavy axles.

2. Drainage is the key factor to producequality subgrade and sound pavementstructures. Drainage facilities should besufficiently designed since later adding afterpavement completion and/or during operatingproves to be an inefficient alternative.

3. Structural analysis and material designshould be integrated to reach a balancedpavement quality. Thickness sufficiency canprevent pavement from premature failure instructural distresses such as fatigue cracking,while mix design plays a central role to avoidunexpected distresses such as AC shoveling,high-temperature rutting, water damage,potholes and other types of material -relateddistresses that cannot adequately addressed bystructural design, especially for ACpavements.

4. Construction techniques in ACpavements such as application of tack coatand prime coat have to be fully implementedto form a layer structural which is close to theabstracted mechanistic model in design guide,usually based on Burmister ’s multi-layerelastic theory, otherwise, constructiondeficiency compared to structural modelmight lead to premature failure.

5. Thickness of PCC plates governsexpected pavement life. A small increase inPCC thickness can lead to significant increasein pavement life. China’s new design guidefor PCC pavements, JTG D40-2003, reflectssuch understanding which is also based fieldinvestigation of PCC pavement performance.

6. Joint sealing in PCC is almost thelatest field work in construction of PCCpavement which often leads to negligence inconstruction supervision. However, jointsealing is one of governing factor to preventPCC from premature failure such as jointspalling, corner breaking, pumping andswelling. Care should be taken to ensure 100percent joint sealing to produce soundpavement structure.

7. Deformation characteristics andengineering behavior of subgrades have to beadequately understood in design process,especially for multi-lane freeway roads inmountainous areas where cut -fill transition

Scanned Data

Strain (%)

Stre

ss (

MPa

)


and sloped ground are frequently found.Structural subgrade, such as pile -supportedembankment, column-supported embankment,and other subgrade structures can limit soildeformation with the help of embeddedstructures.

8. Preventive maintenance, together withroutine maintenance, proves an effectivepractice to keep pavement quality over time.Pavement distress of light severity may easilyescalate to high severity during short period oftime without timely engineering measures.Preventive maintenance such as crack sealing,slurry treatment, and seal coat are more costeffective than corrective maintenance.

9. Stage construction in highways used tobe an engineering alternative in soft groundarea to produce adequate consolidation rateand avoid excessive settlement aftercompletion. However, experience in past twodecades proved that stage construction is not apractical choice based on life -cycle costeffectiveness consideration.

10. Road widening projects, especially inweak ground and/or sloped ground, needspecial considerations in differentialdeformation between existing subgrade andnewly-added section. Additional structuralresponses resulted from differentialdeformation will lead to unexpected distressor even premature failure. Geosynthetics, ifproperly designed and laid between new andold subgrades, proved to be effective materialto tackle subgrade problems.

11. Vegetation is the primarylandscaping type which promotes highwayquality, environmental friendliness, visualrelaxation, sight guidance, and drive safety.One concern that should be addressed is thatvegetation in medians and side slopes maycause drainage problems and weak subgrade.Therefore, effective drainage system forsubgrades and other pavement layers must by

fully designed, especially in vegetatedmedians.

12. Heavy trucks traveled at low speedsclimbing long grades pose a great challenge tosurface layer of pavements, especially for ACpavements in high temperature environment.In the summer of 2006 when temperatureloomed historical high for several weeks, deeprutting, shoveling, and slippage cracking tookplace in several routes with AC pavements.Material design can address part of this problem.However, shear resistance of AC surface alonglong grade section is recommended for check instructural design, which is part of the code forpavement design for urban roads.

13. Structural evaluation of geometricdesign is one of the most important aspects indesign stage and post-design stage. Analysisof highway location and geometric designbased on statics, kinetics and dynamics willlead to a consistent design with respect togeometrical outputs, structural responses, andservice life predictions.

14. It is not necessary to designpavement shoulders different from lanethickness. A reduced thickness of pavementstructures in shoulder does not producesignificant economical benefits compared tototal costs of highway investment. Shoulders,usually used as emergency lanes andmaintenance work space, can serve as trafficlanes when necessary. Furthermore, a singlethickness along traffic lanes and shoulderswill facilitate easy construction, especially forAC pavements.

15. Transitional sections of subgrades,such as bridge approaches, culvertapproaches, tunnel approaches, short subgradesection between road structures, high -fill-deep-cut and cut-fill transitions, are commondesign scenarios in mountainous areas anddeserve special consideration in construction


and supervision to form a pavement support ofconsistency along the route. Strict compactionstandards, reliable subgrade drainage systemand quality granular fills are effectivemeasures to construct subgrades in transitionalsections.

16. Temperature is the key factor inHMA mix production and paving. Climateprediction has to be tracked daily to assureadequate temperature at paving, compaction,as well as proof rolling. Sudden change inclimate, such as temperature drop, should beavoided in the whole process. Mix has to bediscarded if temperature cannot meet specifiedstandards at every stage of construction frommixing to completion.

17. Adequate compaction is required forHMA surface layer for the mix to form asurface layer with sufficient strength to limitstructural distresses such as fatigue crackingand rutting. However, over compaction is notrare due to inadequate understanding of ACpavement as well as HMA mix design.According to Superpave® mix designphilosophy, over compaction is not allowedand a 2% of air voids at the end of pavementservice life has to be expected. Overcompacted AC pavements do not havesufficient voids to deform under the repetitiveapplication of traffic loads and will exhibitrutting, shoveling and other types ofpremature distress.

18. Reflection cracking is the primarydistress in AC pavements with semi -rigidbase. It is crucial for all highway-relatedagencies responsible for design, construction,supervision, acceptance, maintenance, and/ormanagement to understand that adequate baseis sufficient to support pavement under trafficloading. “Over-strong” base, especially semi-

rigid pavement will crack easily and l ead toreflection cracking in surface layer. Flexiblebase, which constitute “flexible pavement”together with AC surface layer, isrecommended instead of semi-rigid base.Literature reported that micro-cracks in thebased layer produced by heavy rollercompaction before surface AC layerplacement prove to be an effectiveengineering practice to avoid reflectioncracking caused by cracks in succeedinglayers.

VI. CONCLUDING REMARKS

Pavement engineering in China is still abooming industry. Unpaved roads wit h amileage of 935,945 comprised of 48.48% outof 1,930,500-km highway networks at the endof 2005. National Freeway System of China isstill under halfway construction which isexpected to be completed in 2040 and at thesame time another 1,000,000 km newhighway will be added to China ’s highwaysystem. There are hundreds of runwaypavements and other airport pavements willbe newly built and rehabilitated during next20 years to meet the rapid expanding of airtransportation and hundreds of big cities wil lcontinue road upgrading work in the contextof China’s national policy of urbanization.Due to different environmental requirements,local industry and economy status, marketfluctuation of raw materials including asphaltcement and Portland cement, emerging of newmaintenance technologies, development ofdesign guides and better understanding ofroad engineers, PCC pavements and ACpavement will continue to be the two majorcategories in the future. AC pavements willfind more and more applications in ur banroads and airport runways. Modified asphaltincluding color asphalt will be tailored to meetdifferent transportation needs besides


engineering requirements. Internationalcooperation and exchange will play an evenmore important role in pavement indus trywhich will promote the continuousdevelopment of theoretical understanding andengineering practice of pavement with betterquality for roads and airports.

Acknowledgement Figures 5 to 24 areresearch illustrations of master and doctoralstudents of the author’s. It has been a greatpleasure to work with such smart youngtalents. The author pays special thanks tothese graduate students for providingmaterials in the completion of this manuscript.

Reference

[1]. Statistics of China’s Highway and WaterwayDevelopment, (in Chinese)http://www.moc.gov.cn/zhuzhan/tongjixinxi/, 2007-03-18

[2]. National High Speed Highway System PlanningProgram (in Chinese),http://www.moc.gov.cn/zhuzhan/jiaotongguihua/guojiaguihua/quanguojiaotong_HYGH/, 2007-03-18

[3]. Highway Statistics,http://www.fhwa.dot.gov/ohim/ohimstat.htm, 2007-03-18

[4]. Aisan Highway Program,http://www.unescap.org/TTDW/index.asp?MenuName=AsianHighway, 2007-03-18

[5]. Statistics of Municipal Construction(in Chinese),http://www.cein.gov.cn/news/show.asp?rec_no=14462,2007-03-18

[6]. Statistics of Civil Aviation Airports,http://www.caac.gov.cn/I1/, 2007-03-18

[7]. Boeing Projection on China’s Aviation Market,http://news.xinhuanet.com/fortune/2004-10/27/content_2144873.htm, 2007-03-18

[8]. AASHTO(1993), Guide for design of pavementstructures, American Association of State Highway andTransportation Officials.

[9]. Burmister. D.M.(1943), The theory of stresses anddisplacements in layered systems and applications to thedesign of airport runways, Proceedings of HighwayResearch Board, pp. 126-148.

[10]. Burmister. D.M.(1945), The general theory ofstresses and displacements in layered soil systems,Journal of Applied Physics, Vol. 6, No. 2, pp. 89-96, No.3, pp. 126-127; No. 5, pp. 296-302.

[11]. Burmister. D.M.(1945), Stress and displacementcharacteristics of a two-layer rigid base soil system: influencediagrams and practical applications, Proceedings of HighwayResearch Board, pp. 773-814.

[12]. Huang, Y.H. (2004), Pavement analysis anddesign, 2nd, Pearson Prentice Hall.

[13]. Ioannides, A M; Thompson, M R; Barenberg, E J,“WESTERGAARD SOLUTIONS RECONSIDERED” ,Transportation Research Record N1043,pp. 13-23.

[14]. Lin, X. (1988), Design methodology of flexiblepavement structures (in Chinese). Renming JiaotongPublishing House.

[15]. Westergaard. H.M.(1926), Analysis of Stresses inConcrete Pavements Due to Variations of Temperature,Proceedings, Highway Research Record, Vol. 6, pp201-215.

[16]. Westergaard. H.M.(1926), Stresses in ConcretePavements Computed by Theoretical Analysis PublicRoads, V. 7, No. 2, pp.25-35

[17]. Westergaard. H.M.(1927), Theory of ConcretePavement Design, Proceedings, Highway ResearchRecord, Part I, pp175-181.

[18]. Westergaard. H.M.(1933), Stresses in ConcreteRunways of Airports, Public Roads, V. 14, No. 10,pp.185-188

[19]. Westergaard. H.M.(1943), Stress concentrations in PlatesLoaded over Small Areas, ASCE Transactions, Vols. 108,pp.831-856

[20]. Westergaard. H.M.(1948), New Formulas forStresses in Concrete Pavements of Airfields, ASCETransactions, Vol.113, pp.425-444

[21]. Yoder & Witczak, Principles of PavementDesign, 2nd Edition by John Wiley & Sons, Inc.,

1975

http://www.moc.gov.cn/zhuzhan/tongjixinxi/

http://www.moc.gov.cn/zhuzhan/jiaotongguihua/guojiag

http://www.fhwa.dot.gov/ohim/ohimstat.htm

http://www.unescap.org/TTDW/index.asp

http://www.cein.gov.cn/news/show.asp

http://www.caac.gov.cn/I1/

http://news.xinhuanet.com/fortune/2004-


I. INTRODUCTION

Many governments do not have all the financial resources required to expand, maintain,and operate their country’s highway networks and other transport infrastructure. The overallresources needed are enormous. In the United States, for example, it is estimated that $55billion will be required annually over the next 20 years simply to maintain the highway andbridges in their current condition.

In many countries, the private sector has been involved in financing infrastructure throughconcessions under a public-private partnership (PPP) program. Broadly defined, a concession isa legal arrangement in which a firm obtains from the government the right to provide aparticular service (Kerf, 1998). PPP schemes, however, are somewhat under utilized in transitioneconomies, and there seems to be an enormous potential for more private sector involvement inthe financing and operation of highway assets in these countries.

With many countries increasingly interested in attracting private capit al to infrastructureprojects, institutions such as the World Bank can contribute through greater use of their

LAUNCHING PUBLIC-PRIVATE PARTNERSHIPS FORHIGHWAYS IN TRANSITION ECONOMIES

Dr. CESAR QUEIROZWorld BankProf. VALENTIN SILYANOVState Technical University-MADIDr. ALEKSEY AKULOV


Abstracts: In many countries, the private sector has been involved in fi nancing infrastructure

through concessions under a public-private partnership (PPP) program. This paper reviews

potential applications of partial risk guarantees, the required legal framework (e.g., concession

law) for attracting private capital for PPP sc hemes, possible steps for a country to launch a

program of private participation in highways, the concept of Greenfield and road maintenance

concession programs, and the treatment of unsolicited proposals. It also summarizes potential

applications of the World Bank Toolkit for PPP in Highways as an instrument to help decision -

makers and practitioners to define the best PPP approach for a specific country .

Key words: a public-private partnership (PPP) program; concession law; private capital; “build -own-operate-transfer” (BOOT) contract; “build-own-operate” (BOO) contract; “build-operate-transfer”(BOT) contract


guarantee power. Partial risk guarantees are particularly relevant in the context of seeking moreprivate involvement in the financing of road i nfrastructure.

This paper reviews potential applications of partial risk guarantees, the required legalframework (e.g., concession law) for attracting private capital for PPP schemes, possible stepsfor a country to launch a program of private participa tion in highways, the concept of Greenfieldand road maintenance concession programs, and the treatment of unsolicited proposals. It alsosummarizes potential applications of the World Bank Toolkit for PPP in Highways as aninstrument to help decision-makers and practitioners to define the best PPP approach for aspecific country.

II. CONCESSION LAWS

An appropriate concession law is fundamental for a country to establish an enablingenvironment for PPPs and it also serves as a possible marketing tool for i nvestors. It shouldapply to construction, expansion, rehabilitation and maintenance of assets providing a publicservice, aiming at improving the efficiency and modernization of public services. In general, aconcession law should include provisions for:

Definition of concepts and terms.

Transparent competitive bidding.

Allowing for bid evaluation on a net present value (NPV) basis .

Assurance of national treatment to foreign investors .

Assurance of compensation in the event of expropriation .

Assurance of access to international arbitration for foreign investors .

A general deference to the terms of specific contracts, which creates scope for flexibleapproaches between sectors and projects .

Public disclosure of concession agreements

A concession law can be kept relatively simple and general, while specific regulation withdetailed guidelines about the ways in which the procurement process will be conducted, criteria,contract award, selection committees, etc. should be documented in operational guidelines (ordecrees). A separation between law and regulation provides more flexibility for amendmentsduring the implementation of a PPP program.

It is usually beneficial to have a draft concession law reviewed by a law firm with a strong

international project finance practice and with a strong local knowledge base.

Public disclosure of concession agreements should be supported. In recent years a growing

number of countries have taken the step of publishing the concession agreements. This has

several benefits: (a) it provides a further check on corruption, which in addition to its direct


benefits can enhance the legitimacy of private sector involvement in often sensitive sectors; and

(b) when the concession agreement relates to the provision of services to the publ ic, it provides

consumers with a clearer sense of their rights and obligations, and can facilitate public

monitoring of concessionaire performance.

It is usual practice for concessions law to contemplate the concept of “negativeconcessions” in which the bidding criterion is minimum public contribution rather thanmaximum payment to the public authority. The law would make explicit the right of the publicauthorities to enter into multi -year contracts to pay the concessionaire the required stream ofpayments.

A concession law needs to link with other laws, such as:

Laws regulating the provision of public services

It is common to find aspects of utility services governed by sector -specific laws, some ofwhich establish specialist regulatory bodies. The relationship between those laws and bodiesand concession agreements needs to be spelt out, for example regarding tariffs and servicestandards.

Procurement laws

In order to provide a clear legal framework, the regime for bidding for concessions needs tobe clear vis-a-vis other procurement laws.

Laws governing foreign investment

The provisions of a concession law need to be clear relating to other laws that mightinclude restrictions of some kind on foreign participation. It is important that regardingconcessions there is no separate treatments for local and foreign investors.

Many countries distinguish between concessions for public works, concessions for thedelivery of public services, and concessions for the exploitation of natural resources. Aconcession law would need to reflect such distinctions. While concessions for public worksrequire investments, under many concessions for the delivery of public services the mainobligation of the concessionaire is to provide the service, rather than make specific investments.

There are situations in which one of the bidding criteria is based on minimum public

contribution (or “subsidies”) to construction or reconstruction costs. A Concession Agreement

may include: (i) re-build1 and/or build; (ii) operate; and (iii) maintain. It may cover the whole

spectrum of PPP, i.e., management and maintenance contracts, operation and maintenance

contracts, and Build-Operate-Transfer (BOT) concessions.

1 Re-build may include, for example, rehabilitation, modernization, refurbishing.


Other useful provisions in concession laws include (relevant definit ions can be found inrecent concession agreements):

The concept of “cannon” or “entry ticket fee,” which is a current practice worldwide.It is usually through the “cannon” that the concession’s high transaction costs are reimbursed bythe concessionaire.

The concept of periodic independent assessment of the concession assets to be carriedout by an expert acceptable to both parties and paid, preferably, by the concessionaire.

Amendments of concession agreements. International experience illustrates twogeneral approaches: (a) provide no special rights for the grantor to unilaterally amend orterminate, and so leave this to be determined by the parties by agreement; or (b) provide thegovernment with such special rights, but with carefully defined safeg uards for theconcessionaire.

Required land to provide the public services. Responsibilities of the grantingauthority may include providing adequate site condition, right of access, expropriation andacquisition of land, contingent environmental liabil ities, etc.

The concept of contract renegotiation, as it is better to be prepared to manage theprocess when renegotiation may be necessary. Concession laws should establish clearmechanisms for renegotiation and amendments (as a way to minimize contract distress andcancellation). The renegotiation of projects is not an unusual occurrence (Harris et al. 2003).

Provision for international arbitration.

Award of contracts through a two-step approach in which the qualitative requirements(e.g., experience, financial capability, management plan) and some of the quantitativerequirements (e.g., investment plans) are judged on a pass/fail basis. All bidders that pass thisstage are by definition qualified and step two judges the financial offers.

Exceptions to competitive bidding. For example, most countries permit sole sourcingin the case of very small contracts (where the costs of a tender would be disproportionate to thebenefits) and in emergency situations (where there is no reasonable time to conduct a tender ---which may be a particular concern when it relates to the delivery of public services.

III. UNSOLICITED PROPOSALS

Unsolicited proposals, which seem attractive to some governments in their wish toaccelerate road or motorway construction in the country, tend to be so controversial (usuallyinvolving allegations of corruption), that in fact they may take longer to negotiate than an open,competitive tender procedure. In theory, unsolicited proposals could generate beneficial ideas;in practice, there have been a number of unfavorable experiences, mostly as a result of exclusivenegotiations behind closed doors (in a recent case, a contract signed between a government anda private company included a clause that prohibits any leakage of the signed contract).


Several countries have adopted specific legislation to deal with such proposals, while somegovernments have simply forbidden unsolicited proposals to reduce public sector corruption andopportunistic behavior by private sector companies. The general experience with unsolicitedproposals is often negative, reflecting the fact that projects of this type have usually representedpoor value for money, were frequently incompatible with the actual development needs of thecountries, and their ability to pay. They also often elicit allegations of corruption. Corruptionhas been shown to be associated with the lack of adequate transport infrastructure in a country,as well as low economic development (Queiroz and Visser 2001). It is essential to eli minate orminimize the perception of corruption in PPP programs so that such programs can bestcontribute to a country’s economic development.

Some governments have adopted procedures to transform unsolicited proposals for privateinfrastructure projects into competitively tendered projects. Such countries include Chile, theRepublic of Korea, the Philippines, and South Africa (Hodges 2003).

IV. STEPS TO LAUNCH A PPP PROGRAM IN HIGHWAYS

A first step in launching a PPP program in highways in a country is to define the priorityprojects where the government envisages to elicit private investors financing of the total orpartial cost of the project. In the case of Russia, for example, several high priority projects forpotential PPP in highways have been desc ribed, such as Moscow-St. Petersburg motorway,outer Moscow ring road, Moscow -Minsk highway, access to Domodedovo airport, St.Petersburg bypass, bridge on Volga river at Volgograd.

Other steps to launch a PPP program would include (some of these steps ca n be done inparallel):

a. Enact relevant legislation (e.g., concession and toll road laws) .

b. Carry out feasibility study of priority projects. Employ reputable consultants, using wellprepared terms of reference (TOR). Identify/quantify social and ec onomic benefits; carry outfinancial assessment to help check the potential for attracting private capital (e.g., relativelyhigh overall financial rate of return and return on equity) .

c. Carry out environmental and social assessment, including mitigation plan and land

acquisition plan for the right of way

d. Prepare bid documents to select the concessionaire (or concessionaires) .

e. Assess the willingness of users to pay; review tolling / payment options (e.g., actual tolls,

shadow tolls, vignette system).

f. Draft concession and other agreements; define performance standard for the new

investment and the service standards during the operation period .

g. Carry out prequalification of potential concessionaires .


h. Invite prequalified firms/consortiums to su bmit bids.

i. Sign concession agreement with the best evaluated bidder .

j. Reach financial closure.

k. Monitor the performance of the concessionaire over the life of the concession .

International financial institutions (IFI) such as the World Bank can coo perate and assist inall of these steps. Forms of assistance may include:

a. Technical assistance to all required processing stages, including establishing a goodregulatory capability.

b. In case the project requires government subsidies (e.g., governmen t contribution to partof the construction cost), the IFI could consider financing a part of the subsidies .

c. The IFI could consider providing a partial risk guarantee (PRG) to support the selectedconcessionaire so it can borrow at lower interest rate an d longer maturity.

V. WORLD BANK PARTIAL RISK GUARANTEES

The World Bank through its guarantee instruments can help accelerate growth in transitionand developing countries by mobilizing private financing for infrastructure development andother projects of national importance.

By covering government performance risks that the market is not able to absorb ormitigate, the World Bank’s guarantee mobilizes new sources of financing at reduced financingcosts and extended maturities, thereby enabling commercial /private lenders to invest in projectsin transition and developing countries. Guarantees can mitigate a variety of critical sovereignrisks and effectively attract long-term commercial financing in sectors such as power, water,transport, telecom, oil and gas, and mining. Guarantees can also enhance private sector interestin public private partnerships. It can also help sovereign governments access the financialmarket.

The World Bank’s presence in transactions is seen by investors as a stabilizing factorbecause of the World Bank’s long term relationship with the countries and policy support itprovides to the governments. The World Bank Guarantees help catalyze private financing,

which leads to greater job and income opportunities for people, and thus co ntribute to the

achievement of the Millennium Development Goals’ overall challenge of reducing poverty. Awebsite dedicated to World Bank guarantees is available at:

http://web.worldbank.org/WBSITE/EXTERNAL/PROJECTS/EXTFININSTRUMENTS/E

XTGUARANTEES/0,,contentMDK:20267847~hlPK:545970~menuPK:64143502~pagePK:6 41

43534~piPK:64143448~theSitePK:411474,00.html



A World Bank operational policy regarding its guarantee program, OP 14.25, states that aguarantee objective is to mobilize private sector financing for development purposes. The Bankmay guarantee private loans with or without an associated Bank loan; the Bank does notguarantee equity investments. The Bank provides guarantees only to the extent necessary. Theoperational policy is available at:

http://wbln0018.worldbank.org/institutional/manuals/opmanual.nsf/toc1/A505EC4B4C9EB1658525672C007D0976?OpenDocument

Although guarantees may be structured in different ways, there are two basic kinds. Partialcredit guarantees cover debt service defaults on a specified portion of a loan, normally for apublic sector project. Partial risk guarantees cover debt service defaults on a loan, normally for aprivate sector project, when such defa ults are caused by a government's failure to meet itsobligations under project contracts to which it is a party. The nature and scope of governmentcontractual undertakings that the Bank backs vary depending on specific project, sector, andcountry circumstances. The Bank requires that the underlying contracts for partial riskguarantees contain appropriate dispute resolution procedures; if there is a dispute about thegovernment's obligations, the Bank's guarantee is triggered only after the government's liabilityhas been determined in accordance with such procedures. Both kinds of guarantees may coverscheduled interest as well as principal payments on a loan.

Both governments and the private sector benefit from a guarantee. Governments benefitbecause it:

Catalyzes private financing in infrastructure .

Provides access to capital markets .

Facilitates privatizations and public private partnerships .

Reduces government risk exposure by passing commercial risk to the private sector .

Improves impact of private sector participation on tariffs .

Encourages cofinancing.

The private sector benefits because it:

Reduces risk of private transactions in emerging countries .

Mitigates risk that the private sector does not control .

Opens new markets.

Lowers the cost of financing.

Improves project sustainability.

http://wbln0018.worldbank.org/institutional/manuals/opmanual.nsf/


The World Bank guarantee instruments have proved to be a powerful instrument in

catalyzing private financing to frontier markets. Twenty two guarantees with about US$ 1.4

billion exposure to the Bank have achieved a remarkable leverage by catalyzing more than US$

12 billion of private resources for projects worth US$ 24 billion (see examples in Figure 1).

Each dollar of guarantee financing has catalyzed close to 5 dollars of private finance.

Partial risk guarantees are particularly relevant in the context of seeking more private

involvement in the financing of road infrastructure. Such guarantees cover specific government

obligations spelled out in a support agreement with the project entity. Example of such

agreements include concession agreement, implementation agreement, build -own-operate-

transfer (BOOT) contract, build -own-operate (BOO) contract and, the most common form,

build-operate-transfer (BOT). Partial risk guarantees are appropriate for enhancing a pr oject’slimited recourse project financing, the most common method of financing concessions for

transport infrastructure. Figure 2 provides an illustration of how a partial risk guarantee can

apply to a highway concession contract (Queiroz 1999).

Figure 1. Examples of guarantees’ leverage in catalyzing private resources

0 500 1000 1500 2000

China Yangzhou

Philippines Leyte

PakistanHub

China Zheijiang

Jordan Telecom

China Ertan

Pakistan Uch

Lebanon Power

Morocco Jorf Lasfar

Thailand EGAT

Cote d'Ivoire Power

ArgentinaPBG

Colombia PBG

Bangladesh Haripur

Vietnam Phu My 2 -2

US$ Million

Guaranteed Amount

Private Capital Mobilized


Government

Obligations:

• Toll rate

• Permits/consents• Forex

• Change in law• Political events

• Termination

Private

Lenders

Concessionaire Loan Agreement

WBGuarantee

CounterGuarantee

Concession

Project company obligations: Construct and operate

highway; maintain and rehabilitate to keep up quality.

World

Bank

Figure 2. Structure of a Highway Concession Contract and World Bank Guarantee

VI. GREENFIELD AND ROAD MAINTENANCE CONCESSION PROGRAMS

Greenfield PPP projects include investment in new construction, usually on a newalignment, by the concessionaire, while in road maintenance/rehabilitation/operation (RM/R/O)concessions the concessionaire agrees to assume responsibility for an existing road or pa rt of aroad network. Several concession options are available and each country should select the mostappropriate for its prevailing conditions. Through the most typical forms of concession, acountry can transfer to the private sector the responsibilit y to: (i) build, operate and transfer(BOT) back to the public sector (at the end of the concession period) a road facility (e.g., amotorway, bridge, tunnel), or (ii) maintain, rehabilitate, operate (RM/R/O concessions) anexisting road or road links. Each concession can include individual links or a set of links in agiven area of the country (i.e., area -wide concessions).

Steps in the process of launching a road concession program include drafting of all relevantdocuments, a competitive selection of concessionaires, evaluation of proposals, as well as awardof the concession contract. When the main purpose of the concessions is to obtain extrabudgetary funding for roads, or release limited public funds for use on other roads (e.g.,secondary and rural roads), shadow-tolls (whereby payment to concessionaires are made out ofthe budget, based on traffic volumes and classification) would not be a feasible option.

All concessions require an institutional means, such as a unit, to monitor the private s ectorperformance, including compliance with the performance standards defined in the concessionagreement. The concession contract and public - private sector arrangements for a newinvestment concession and for a maintenance concession may be different , but a country canpursue both types of concessions at the same time.


VII. WORLD BANK TOOLKIT FOR PPP IN HIGHWAYS

The main objective of the World Bank Toolkit for PPP in Highways is to provide policy

makers from economies in transition with some guidance in the design and implementation of a

Public Private Partnership (PPP) in the highway sector. The Toolkit is structured under five

headings (or modules) and is navigated through a series of tree diagrams under each of these

headings. It also includes a library and interactive financial models. It is a multimedia product

available on a CD ROM and also available from the World Bank's web site at:

www.worldbank.org/transport

or

http://rru.worldbank.org/Documents/Toolkits/Highways/start.HTM

Using basic assumptions about a specific motorway project, the financial simulation tool of

the Toolkit is helpful to answer key questions on t he financial feasibility of the project. For

example, questions such as the ones below can be answered with minimum effort using the

Toolkit:

- What is the internal financial rate of return (IRR) of the project?

- In the absence of Government subsidies, c eteris paribus, what would be the return on

equity (ROE)?

- While subsidies may be paid by the Government during the construction period, it

recovers some of this payment through taxes during the operation period. What would be the

Government contribution to the proposed project that would lead to a financial balance for the

government throughout the concession period?

- In the absence of Government subsidies, ceteris paribus, what would be the required

initial toll rate to yield a return on equity (ROE) of 16%?

- Assuming that an initial average toll rate of US$0.06 per vehicle -km is the highest

acceptable by road users, an investment cost of US$3 million per km (typical for a four -lane

road on flat terrain), and an initial traffic volume (AADT) of 15,00 0 vehicles per day, what is

the minimum concession life that would generate a return on equity (ROE) likely to attract a

private sector concessionaire (say an ROE of 15% or higher)?

A recent update of the financial simulation tool is particularly appropri ate to answer the

above questions. The Excel file with the updated Tool is available on the World Bank website

at:

http://wbln0018.worldbank.org/ECA/Transport.nsf/ECADocByLink/01C97A272081983D

85256FD20061ECB8?Opendocument



http://wbln0018.worldbank.org/ECA/Transport.nsf/


VIII. CONCLUSIONS

This paper discussed potential applications of partial risk guarantees to assist countries intransition to seek more private involvement in the fi nancing of road infrastructure. It alsopresented a review of the required legal framework (e.g., concession law) for attracting privatecapital for PPP schemes, possible steps for a country to launch a program of privateparticipation in highways, the concept of greenfield and road maintenance concession programs,and the treatment of unsolicited proposals. The paper also summarized potential applications ofthe World Bank Toolkit for PPP in Highways as an instrument to help decision -makers andpractitioners to define the best PPP approach for a specific country

IX. DISCLAIMER

This paper reflects only the authors’ views, and should be used and cited accordingly.The findings, interpretations, and conclusions are the authors’ own. They should not beattributed to the World Bank, its Board of Directors, its management, or any of its membercountries.

References

[1]. Harris, C., Hodges, J., Schur M., and Shukla , P. 2003 “Infrastructure Projects: A Review of CanceledPrivate Projects” Public Policy for the P rivate Sector, Note No 252. Washington, D.C.: The World Bank

http://rru.worldbank.org/Documents/PublicPolicyJournal/252Harri -010303.pdf

[2]. Hodges, J. 2003 “Unsolicited Proposals - Competitive Solutions for Private Infrastructure” PublicPolicy for the Private Sector, Note No 258. Washington, D.C.: The World Bank

http://rru.worldbank.org/Documents/PublicPolicyJournal/258Hodge -031103.pdf

[3]. Hodges, J. 2003 “Unsolicited Proposals - The Issues for Private Infrastructure Projects” Public Policyfor the Private Sector, Note No 257. Washington, D.C.: The World Bank


[4]. Kerf, Michel, et.al. 1998. “Concessions for Infrastructure: A Guide to Their Design and Award.”World Bank Technical Paper No. 399. World Bank, Washington, D.C.

[5]. Queiroz, C. and Visser , A. 2001 "Corruption, Transport Infrastructure Stock and Economic

Development." Infrastructure and Poverty Briefing for the World Bank Infrastructure Forum, CD -ROM.

World Markets Research Centre Ltd. Washington, D.C.: The World Bank.

[6]. Queiroz, Cesar. 1999. “Highway Concessions and World Bank Guarantees.” International RoadFederation Regional Conference on European Transport and Roads. Lahti, Finland, 1 4-16 June 1999.

[7]. World Bank Toolkit for Public Private Partnerships (PPP) in Highways.


http://rru.worldbank.org/Documents/PublicPolicyJournal/252Harri-010303.pdf

http://rru.worldbank.org/Documents/PublicPolicyJournal/258Hodge-031103.pdf



CT 2

I. INTRODUCTION

Vibration and oil analysis can reveal a

great deal of information about machine’shealth. Therefore, vibration and oil analysis

are two key components for machine

condition monitoring. Oil analysis has been

used for at least fifty years in determining the

wear condition of machinery. Rails road

companies in the late 1940s and early 1950s

found that the metals in a sample of used oil

revealed the condition of the wearing parts in

their locomotive engines. Vibration and oil

analysis today are used to monitor the

condition of everything from aircraft jet

engines and helicopter gearboxes to

construction equipment industry, commercial

transportation, and industrial plants. Oil

analysis is taking place alongside vibration

monitoring as an indispensable and valuable

predictive maintenance tool in industry. This

paper presents a study of integrating vibration

and oil analysis for diesel engine condition

monitoring based on the result of oil and

vibration analysis at Hanoi Locomotive

Entreprise, Vietnam.

II. CONTAINS

2.1. Oil analysis applicationThe wearing parts of a machine such as

the gears, hydraulic pistons, bearings, andwear rings generate fine metal particles duringnormal operation. At the onset of a severewear mode the particle size increases and theappearance of the particles change.Knowledge of the particles and how theyrelate to the mode of wear permits a trainedanalyst to determine the wear status in amachine by measuring the fine and coarsemetal particles and then examining theparticles under a microscope. The testing forwear metals for condition monitoring andpredictive maintenance is testedpredominantly in spectrometric analysis or inwear debris analysis.

The advantages of oil debris monitoringcompared with other monitoring methodsinclude:

- The evidence in the oil is to be foundnowhere else.

- The cost benefit ratio is better thanother technique.

- The oil carries evidence of faults fromvarious parts.

2.2. Chemical identification of debris

Quantitative measurement i s oftenrequired for many machine condition

DIESEL ENGINE DIAGNOSISBY VIBRATION OIL ANALYSIS

Prof. DR. DO DUC TUANMEng. LE LANG VANUniversity of Transport and Communications

Abstract: This paper presents a research of oil and vibration analysis in machine

condition monitoring and evaluating the use of oil and particle analysis in practice. The paper

also mentions to capable of this method in diesel engine diagnostic based on the result of oil

and vibration analysis at Hanoi Locomotive Entreprise, Vietnam.


CT 2

monitoring applications. Quantifying debrisgives a feel for the likely wear that isoccurring in machines. The measured mass ofdebris is determined to know any change inthe trapped quantity, such as weight per ml,intensity per ml, or shape of the sizedistributions. Chemical identificationinstruments used in this research is Alloy Pro9388 Metal Analyser (Figure1).

Figure1. ALLOY PRO 9388 Metal Analyser

Figure 2. Chemical identificationof piston of D12E engine

2.3. Wear particle image analysisOil samples and vibration data which

were collected regularly at the HanoiLocomotive Entreprise over a period of 14months were carefully examined andcompared. A particle analyzer was used todetermine oil sample to assess the generaltrend of diesel engine conditions. Normalwear process from oil samples of locomotiveNo 660, 657 and 658: number of particles andparticle dimensions were small (from 0.2 to 5

micromet).

Figure 3. Particles in oil sample(Locomotive No 660)

Figure 4. Particles in oil sampleis increasing in numbers

(Oil sample on Locomotive No 658)Various types of wear process give

various types of particle shapes: spheres,fibers, slabs, curls, spirals and slivers, rolls,strands and fibers.

Figure 5. Sphere shape particle

(In oil samples of locomotive No 660)

Sphere shape: the presence of spheres in

2.298 μm

7.423 μm

13.64 μm

2.347 μm

2.381 μm

2.298 μm


CT 2

oils is quite frequent. Sometime spheres arefound in the new lubricating oil from acontainer.

Figure 6. Pebble shape particle

(In oil sample of locomotive No 658)

Figure 7. Chunks and slabs shape particle


The smooth sphere is often found in therunning-in period. The rough seems to involvemore severe wear. The other shapes include:distorted smooth ovoid/pebble shape, c hunksand slabs, curls, spirals and slivers, rolls,strands and fibers. The research process drawsthe following conclusion:

Rubbing wear and normal wear: are regularwear particles which are formed from betweenlubricated sliding surfaces. They would take theshape of ‘platelets’ up to 10 µm.

Cutting wear: These particles are formedby the metal parts digging into each other [2].

Rolling fatigue: Spherical particles

appear quite often. It would have to beassumed that the spheres come from fatiguecracks in bearings. Chunks of metal canappear from fatigue with the size up to 100µm. Another form of particle is the platelet,perhaps up to 50 µm.

Severe sliding wear: These are also largeparticles depending on the magnitude of thesliding action load and speed. Usually in theform of platelets with surface st ress acted onthe wearing surfaces. The higher the stresslevel, the larger the ratio of large particles tosmall particles [4].

2.4. Particle Dimension

Dimension of the particle changes fromseveral micromets to 100 or 300 micromets.Particle shape varies depends on wearprocces. A large dimension of particle informssevere wear process. The research indicates:

Normal wear procces produces particleswith sphere shape and their dimensions arebetween 5 and 10 micromet.

Cutting wear is caused when an abras iveparticle has imbedded itself in soft surface ofcopper alloy wear.

Fatigue wear occurs when cracks developin the component surface that leads togeneration of particles. Particle dimension isup to 100 micromet.

Sliding wear evolves during equipmentstress. The dimension of particle is more than10 micromet.

The particle dimensions from oil samplesof fault engines on locomotives were between100 micromets and 120 micromets.

2.5. Number of wear particles

Determining the number of wear particlesis one of demands for diagnosis procces. Thenumber of wear particles per millilitre gives

7.423 μm

3.347 μm


CT 2

an explicit, easy-to-understand format forcharacterising wear conditions of lubricatedtribosystem monitored. The numbers of wearparticles per millilitre counted from oil samplecollected from fault engines on locomotiveswere over 100 particles per millilitre.

Vibration Analysis

The equipment used for measurement andanalysis includes: vibration meters, analyzer andDasyLab software. The experiments werecarried on Hanoi Locomotive Entreprise overperiod of 14 months.

Figure8. Vibration spectum measured

on engine of locomotive No 643



The vibrations are taken in the threeCartesian directions. In vibrationnomenclature, these are the vertical,horizontal and axial directions. This isnecessary due to the construction of machines– their defects can show up in any of three

directions and hence each should bemeasured. The data collector can collect andstore the data for comparision and trending.The database program stores vibration dataand makes comparisons between currentmeasurements, past measurements andpredefined alarm limits. Alarm limits forlocomotives D12E are 20mm/s and 7mm/s 2.

III. CONCLUSION

The diagnosis technique of combustionengines by oil and vibration analysis has beenresearched in several countries and has somesuccess. This research integrates vibration andoil analysis for diesel engines of K6S230DRused on D12E locomotives. When wearparticle dimensions exceeds 120 micromet,when concentration of metals (Cu, Fe, Cr) hasbeen increasing, shapes of oil debris areunnormal, vibration velocity and accelerationexceeds 20mm/s and 7 mm/s 2, engine faultsmay occur on piston, valves or piston rings.Oil analysis confirms the results of vibrationanalysis. Both wear debris and vibrationanalysis techniques were used to assess thediesel diagnose problems during this research.Oil debris analysis confirm the conclusionsabout the faults of diesel engine elementswhen had vibration alarms.

Reference[1]. Calder, N., Marine Diesel engines:Maintenance, troubleshooting, and repair,International Marine, 1992.[2]. Bowen, E.R. & Westcott, V.C. , Wear particleatlas, Naval Air Engineering Center, 1976.[3]. Doebelin, E. O., Measurement systems,McGraw- Hill Companies, 1990.[4]. Hunt, Trevor M., Handbook of wear debrisanalysis and particle detection, Elsevier AppliedScience, 1993.[5]. Rao, J.S., Vibratory condition monitoring ofmachines, Published by Addison -Wesley,America, 2000


Within many years in the economicplans and forecasts the regions of the Asianpart of Russia - Siberia and Far East - areconsidered as the most perspective.

The main precondition is their significantnatural-raw potential. In the Asian part of thecountry, where on 43 % of territory 6 percentsof the population live only, more than 60percents of ores of non-ferrous metals, almost100 percents of diamonds, third wood and twothird of fish resources are concentrated.

Just here there are strategic stock s ofcarbohydrates, due to which Russiastrengthens the global political and economicpositions.

One more feature of the Asian part ofRussia - direct boarder with China and Japan.This factor, in opinion of a number of theexperts, creates opportunities o f directregional integration in actively growing

economic systems of Asia-Pacific region.There is the possibility of construction ofcircuits of deliveries in a direction Asia -Europe on the basis of transit potential ofRussia, first of all – “Transsib Railway”.

The transport development of Siberia andFar East objectively requires a scientificsubstantiation, development and realizationunique paradigm. It should answer not onlypriorities of socio economic development ofthe Asian part of the country, but also strategyof development of Russian Federation as awhole.

Major element of this paradigm willbecome creation in Siberia and in Far East ofthe country the system of the logistic centers.

The concept of logistic centers grows outof searches of alternative ways ofdevelopment of transport system, which beactively ordered in the advanced countriessince the seventieth years.

FEATURES OF CREATION OF LOGISTIC CENTERS INCONDITIONS OF SIBERIA AND FAR EAST OF RUSSIA

Prof. VYACHESLAV M. PRIKHODKOProf. VICTORIA D.GERAMIProf. ALEXANDER V. KOLIKState Technical University-MADI, Moscow, Russia

Abstracts: In the paper the transportation problems of Siberia and Fa r East of Russia are

discussed. It was suggested to create a number of transport logistic centers in the region on the

basis of hub system (“Hub-Spoke”) to increase efficiency of transportation of goods.

Key Words: Transport of Russia, Siberia and Far East, logistic centers, hub system (“hub -

spoke”), logistic operation, articulated lorry, transit, Public Private Partnership (PPP)


The realization of the concept of logistic

centers in conditions of Siberia and Far East

will have a number of features.

First of all, logistic centers of the Asian

part of Russia should play a role intermodal

and unimodal transport hubs.

The organization of transportations in

system of hubs (differently, in system “hub

and spoke”, is effectively applied in sea

container business and in aircraft. This system

is used and in practice of automobile freight

traffic of a number of countries, for example,

American continent. Recently it all use in

system ground intermodal transport more

active.

In similar system a through servi ce

between items of a transport network are

replaced with a combination of transportations

between the allocated items - so-called hubs -

and traffic between points of origination

(destination) and hubs serving the appropriate

zone.

It is not enough for creation of system

“hub and spoke” only intermodal terminals.

The uniform system of organization of

transport process is necessary, at which the

disorder competition of the transport operators

is inadmissible. It should be replaced by

cooperation in conditions of deep functional

specialization and division of the market

between segments of main and regional

transportations.

The governing advantage, which is

provided by the system “hub -spoke” consists

in reduction of general number of transport

connections and due to it - concentration of

freight flows (Fig. 1). It allows to achieve

economy of scale and reduction of total costs.

For transport development of the Asian part of

Russia with its rather weak freight flows this

factor is represented critically impor tant.

The second feature of development of

creation of logistic centers in regions of

Siberia and Far East lies in accommodation

and specialization of these objects by an

essential image will determine development

of an infrastructure of different modes of

transport and distribution of cargo bases

between them.

By virtue of the marked above features ofa transport network of the Asian part ofRussia (fragmenting and weak developmentof a modal infrastructure) logistic centers willbecome natural units of joining existing andrecreate transport infrastructures. Thedecisions on accommodation and thesequences of input in build separate logisticcenters, in turn, will define(determine)priorities of selection and realization of theprojects of the transport communications.Differently, the network in considered regionshould develop in many respects not by aprinciple “from cities to city”, and by aprinciple “from logistic centers to logisticcenter”.


It is obvious, that in view of

characteristic for Siberia and Far East of

distances a key main kind in system of logistic

centers should be a railway transportation. In

view of it logistic centers should be created,

first of all, in those points, where is possible

and is necessary it effective joining with lines

of other modes of transport (Fig. 2).

Internal water transport, by keeping the

role of an alternative kind of communications

in the great Siberian rivers basins, will receive

reliable connection with ground transport

system through logistic centers placed in the

largest ports.

Cargo hubs of air transport, which

creation is actively discussed last years, first

of all, with reference to region of Siberia,

should by a natural image “blend with” in

system of logistic centers.

As to road transport, its task, first of all,

should become effective transport service of

zones of gravitation of logistic centers with

granting to clientele of maximal volume of

logistic services adding intermodal

transportation on a site supply- conveyance.

a)

Hub

Hub

b)

Hub

Hub


Fig 2. Transport service of the system of logistic centers

Regular rail transportation

Transportations by the internal water ways

Air Transportations

Supply – Deliver by road transport

Main road transportations

For logistic centers of Siberia and Far

East a zone of gravitation will be much more

extensive, than for similar objects in the

European part of Russia. ”First” and “last

mile”, as sometimes call supply - conveyance

of transportation abroad, can be stretched on

tens kilometers. Therefore separate important

task will become development and essential

increase of quality of a road network in zones

of gravitation of logistic centers - and it, in

turn, can become the essential factor of

regional development and tool of the decision

of a task of connection of the isolated today

occupied points with a basic transport network

of the country. In this case speech will go

about stable connection of the occupied points

not with a transport network in general, and

with national logistic system.

At the same time, in conditions of Siberiaand Far East the role of road transport can notbe limited to regional service. The mainautomobile transportations will becomenecessary on those directions, where logisticcenters for whatever reasons will not h aveamong themselves of regular railwaycommunication. In these conditions willbecome economically justified the realizationin Russia of the concept of trailers especiallyof large carrying capacity, which finds inworld practice rather wide application.

Now in European Union the idea ofincrease of allowable length and maximalcomplete weight of the trailers for theinternational transportations is activelydiscussed within the limits of the EU.


The most serious reason for the benefit of

this offer is the expected reduction of number

of vehicles on a road network, that critically

important for the overloaded European

highways. The efficiency and safety of

heavier trailers is proved by successful

practice of a number of the European

countries (Sweden, Finland and others), where

the national requirements already now allow

their application.

The idea of “road train” for a long time

and effectively is used also in a number of

states of USA, in Australia, Mexico, Brazil

and in other countries. The experience of

Australia, in particular, shows, that the

operation of similar vehicles is real at the

lowest axial loading of 6 tons.

In Russian Federation currently in use

road restrictions are established enough

arbitrary. Macroeconomic approach to

parameterization of the complex “road costs -

cost of a vehicle - cost of transportation” and

task solution of joint optimization of strength

properties of highways and such parameters of

lorries, as fully loaded mass and axial

loadings, undoubtedly, would allow to

achieve significant economic benefit in scales

of economy as a whole. However, not waiting

for statement and decision of this problem at a

national level, the introduction of the special

system of the road specifications for the

certain regions of the country, in particular -

for Siberia and Far East is represented quite

pertinent and useful. Primary factor of

efficiency should become significant

reduction of the cost price of automobile

transportations on those directions (including

between separate logistic centers), where the

vehicle is no alternative one. The

transportations by “road trains”, thus, will

become effective addition to system of main

rail transportation.

One more feature of logistic centers of

Siberia and Far East will be, obviously to

consist and that their creation will require a

new level of PPP in development of an

infrastructure.

The PPP idea is very popular both in

West, and in Russia. In our country the first

legal preconditions for its development,

including, for transport are created.

More often PPP assumes a role of the

private partner, first of all, as investor, which

considers the PPP project as an opportunity of

effective investments of free financial assets

at the certain guarantees and support on the

part of the state. But in the Asian part of

Russia it is required, probably, some other

approach to a choice of the partners in the

PPP projects.

First, the speech goes about the largest

enterprises of region, which will be interested

in creation of industrially logistic centers, first


of all, for maintenance of requirements of own

manufactures and chain of deliveries.

The prototype of such model can serve

logistic centers created in Germany in

partnership of authorities of city

Ludwigshafen and the chemical concern

BASF for transport service of new industrial

complex of the company. One figure -

processing of 300 thousand containers per one

year testifies only to capacity this largest in

the sort of object. But its efficiency is caused,

in many respects, that it is simultaneously

object of for general usage. Freight flows of

BASF are integrated on it with flows of other

users located in the given region.

Such model completely answers city -

forming function of the large enterprises,

typical for Siberia and Far East, and will

allow to effectively realize this function in

conditions of market economy.

The second group of the target partners

are large companies and groups of the

companies, which today independently

develop the marketing and transport networks

in the Asian part of Russia. The association of

their potential in frameworks of PPP will

create additional network effect both for all

participants of such partnership, and for

territories.

Characterizing the special role logistic

centers in transport system of the Asian part

of Russia, it is impossible to bypass a theme

Trans-Siberian transit.

The mention of huge unused potential of

Trans-Siberian transport bridge for a long

time has become a general place. The

development of transit within many years is

imperishable priority of the high level and is

considered as a major point of growth of

economy of Siberia and Far East.

Between that, from middle of the

ninetieth years, when this theme has become

actively discussed by the experts in the

transport markets there were changes, which

essentially have changed a ratio of the tariffs

on competing Asian - European routes not for

the benefit of Russia.

The sea container operators continued toincrease individual tonnage of linear ships,achieving a scale effect and stabilization - andin a number of cases, and reduction - tariffsfor sea container transportations.Simultaneously, during reforming the Russianrailways the internal cross subsidizing offreight traffic was liquidated which wasdistributed, in particular, to container transit.It objectively has resulted in increase of thetransit tariff on transportations of containersthrough Russia.

Thus, such advantages Trans-Siberianroute as, for example, shorter transit time,were appreciably shown on is not present byaction of the price factor. There are all basesto assume, that within the framework of thetraditional scheme of transit transportation atthe usual structure of costs and tariffs to


Import of goods from countries-exportersLogistic dealership

Export of goods to the countries of Europewith additional added costRegional distribution of goods withadditional added cost

Fig 3. Variant of Trans-Siberian transit “with additional added cost”

achieve essential growth of Trans -Siberiantransit it will be not possible. But the creati onin the Asian part of Russia logistic centersallows to realize a little bit other form of theAsian - European transit communication.

The cargoes addressed in the countries ofEurope, can be exposed at these centersadditional logistic processing. The speechgoes about such operations, as palleting,regrouping, packing, marks and others, whichnow are carried out by the countries inWestern Europe and consequently manage tothe importers extremely dearly. The inclusionof these operations in the Russian scheme ofdeliveries of cargoes from Asian -PacificRegion will add simple transit transportation

by much cheaper services in escalating theadded value of a final product (Fig. 3).

Other important function of logistic

centers of Siberia and Far East connected with

Asian freight flows, should become unloading

of Moscow and St.-Petersburg as “obligatory”

points of transshipment of the Asian import in

Russia. Thus the integration of import and

transit flows, and also logistic operations will

allow achieving additional economic benefit.

The enterprises and inhabitants of region

should not in addition pay transportation of

inward cargoes in the Moscow region and

Import of goods from countries - exporters

Logistic dealership

Regional distribution of goods withadditional added cost

Export of goods to the countries of Europe withadditional added cost


then again on East of Russia.

THE CONCLUSION

The economic development of Siberia

and Far East of Russia should carry advanced

character. For modern transport and logistic

technologies in this process the special place

should be assigned.

The transport development of the Asian

part of Russia has a number of features, which

do not allow to apply in this region model of

simple escalating of extent and increase of

density of transport networks. Paradigm of

“dotty” transport development of Siberia and

Far East should be based on creation of

system of regional logistic centers.

Logistic centers should be created as

compact technological objects, on which the

independent operators will carry out a

complex of the functions directed on

coordination and integration of logistic flows

and on increase their added cost. The

important factor of efficiency should become

optimum accommodation regional logistic

centers on a transport network and

organization of the stable communication

between them.



will have a number of features, in particular:

- Logistic centers of the Asian part of

Russia should play a role of transport hubs. It

will allow achieving a high degree of

concentration of freight flows and will raise

efficiency of transport process.

- Accommodation and specialization of

these objects in many respects will be defined

by priorities of selection and realization of the

projects of development of an infrastructure of

separate modes of transport and distribution of

cargo base between them. Thus the transport

network of the Asian part of the country will

develop by a principle “from logistic center to

logistic center”.

- The development and increase of

quality of a road network in zones of

gravitation logistic centers can become and

tool of the decision of a task of connection of

the isolated occupied points with a basic

transport network of the country and with

national logistic by system as a whole.

- The transportations between logistic

centers should be carried out, first of all, by

railway transportation. However on those

directions, where the regular railway

communication for whatever reasons will be

absent, the main automobile transportations

will be claimed. Thus the realization of the


concept of truck trains especially of large

carrying capacity is expedient .

- The creation logistic centers on the

basis of PPP in the Asian part of Russia is

expedient with attraction of the largest

enterprises which are carrying out city -

forming functions. The appropriate objects

(industrial logistic centers) will have the

“mixed” character, providing requirement of

large industrial complexes and working

simultaneously as logistic centers of general

usage.

- The creation in the Asian part of Russia

logistic centers will allow to realize the

modified variant Asian - European of transit

communication, at which the cargoes

addressed in the countries of Europe, will be

exposed additional logistic processing, and the

transit transportation will be complemented

by rather cheap services in escalating the

added value of a final product.

Reference

[1]. Materials of session of State Council of

Russian Federation and official documents.

October, 2003 - Moscow, State Council of Russian

Federation, 2003.- 514 p. (in Russian).

[2]. Prikhodko V.M., Kolik A.V., Gerami

V.D.Scientific The Profblems of Motor

Transportation Ensuring of National Logistic

System.- M.: Publ.House

“Technopoligraphcenter”, 2006. - 91 p. (in

Russian).

[3]. Prokofyeva T., Platonov S . Formation of

Transport-Logistic Infrastructure of Russia. –“Container Business”, № 1, 2005. - pp. 10-17. (in

Russian).

[4]. Belyaev V.M. Terminal Systems of

Transportations of Cargoes by Road Transport. -

M.: Transport Publ.House, 1987. - 282 p. (in

Russian).

[5]. Oreshin V.P. Planning of an Industrial

Infrastructure. The complex approach. - M.:

Economy Publ.House, 1986, -142 p. (in Russian).

[6]. Best Practice Handbook for Logistics Centres

in the Baltic Sea Region, 2003,

www.neloc.net/reports/Best _ Practice _

Handbook.pdf < http: //


Handbook.pdf >

[7]. Dirk Berendt, Importance of Harbors in

Logistic Chain. Tasks and opportunities, materials

of a conference “Transport and Logistic inInternational Trade”, Euroasian Transport Union(ЕАТС), 2004, www.eatu.org < http: //www.eatu.org >.

[8]. J.C.Johnson, D.F.Wood, D.L.Wardlow, Paul

R.Murphy. Contemporary Logistics. - Upper

Saddle River, NJ.: Prentice Hall, 1999. - 590 p.

[9]. Technological change and Multimodal Freight

Competition. By J.L.Courtney. Proceedings of the

Transportation Research Forum. - Boston, MA,

1984. - p.p. 116-121

[10]. White Paper - European Transport Policy for

2010: Time to decide. - Luxembourg, EC Official

Publications Department, 2001. - 150 p.p

www.neloc.net/reports/Best


www.eatu.org


CT 2 I. ENVIRONMENT PROBLEMS ON ROADS

The increase in the traffic volume (up to 40 -70 and more thous. cars per day at the largecities exits and 10-20 thous. on the majority of the federal roads), construction of new andreconstruction of the existing roads ha ve aggravated the problem of environmental protection.When considering the ecological problems in Russia considerable attention is traditionally anddeservedly paid to the automobiles. The result of it is the significant progress in the sphere ofengine-building accompanied by the sharp reduction in emission of harmful substances. In theyears coming Europe is planning to introduce standard EURO -5, Russia - standard EURO-2 andlater on - standard EURO-3. In the USA the President George Bush announced the d evelopmentof the new ecologically friendly car engine. But apart from the cars there are other factors thatcontribute to the environmental pollution. The most significant of them are highways affectingthe level of pollution coming from traffic. Moreover , the road itself has negative influence withthe roadside territory. Such influence is exerted by:

Road engineering elements: roadbed, bridge crossings and flyovers, water intakestructures and culverts.

Separate road engineering structures: pavement, roa dbed, shoulders.

Road infrastructure units: rest area, gasoline stations, food stations, public transportstations.

THE ENVIRONMENTAL PROBLEMS CONNECTED WITH

HIGHWAY CONSTRUCTION AND MAINTENANCE

Prof. MIKHAIL V. NEMCHINOVDr. ALEXEI S. MEN’SHOVState Technical University - MadiDr. DMITRY M. NEMCHINOVThe Association of Road Design Institutions of RussiaDr. VERONIKA OSINOVSKAYABryansk State Technological Academy, Bryansk, Russia

Abstracts: In the paper the environmental problems connected with road construction and

maintenance, trends and ways of their solving are formulated and presented. The problems with

deformation of the road embankment slopes are considered.

Key words: ecological safety of roads; paramet ers of ecological safety, road embankment,

slope of road embankment


CT 2

The above listed sources of highway influence over the environment affect all the natureelements: air, soil, water, biosphere. Air pollutio n over the highway is influenced by theroadbed, pavement surface material and texture, even interchanges design and peculiarities oftraffic. The roadbed in the form of a high embankment affects the thermal, humidity and windconditions of the roadside territory.

Roadway paving influences the quality and composition of the automobile exit gases, thequantity of wear debris of the automobile parts, including automobile tires, air pollution by thewear debris of the roadway covering, dust and garbage and it is an important factor of theformation of the level of traffic noise. Constructional features of the road crossing, means andmethods of organization and traffic control of automobiles also influence the quantity of the exitgases released by the automobi le engines.

The impact of automobile road on soil and water is no less diversified. The landscape ofthe area changes as a result of exemption of territory for the engineering constructions of theroads, careers, earth-deposits, construction sites, industr ial approaches. As a consequence ofdevelopment of the road network there occurs the fragmentation of the territory, change ofterrain and flora. The construction of grade level, bridges and crossovers is followed bydeformation of sub-base, development and strengthening of erosion processes. The regime ofrunoff of surface and ground offers is often broken, which is followed by drainage oroverdamping of territories, up to formation of marshes. It often leads to the erosion of the bed ofthe water streams, formation of ravines. The soil gets contaminated not only by the componentsof the exit gases of the automobiles but grade level erosion products, wear of roadway covering,by the materials used during the winter maintenance of roads (antiglaze reagents). Watercontamination of rivers and lakes occurs as the result of pollutant emission and impact oferosion products, wear of roadway coverings and automobile tires, dust, garbage, oil -productsand human wastes (in the locations of infrastructure facilities) .

Such kind of impact also has its consequences in biosphere: flora, fauna, including humans.The habitat of plants is limited as a result of the change of the regime of soil watering, drainageor underflooding of the territories, change of soil fertility as well as the presence ofcontaminative chemical agents. In the locations of rest areas occurs trampling and vegetationdamage, repacking of soil. The habitat of animals limits, natural migration ways change,acoustic environment becomes more complicated.

According to the law «On environment protection» in Russia the following items aresubject to protection from contamination, depletion, degeneration, damage, extermination andother negative impact of economic and other activity:

Land, the Earth's interior, soils.

Surface and underwater.

Forests and plants, animals and other organ isms and their genetic heritage.

Ambient air.


CT 2

Automobile roads belong to the objects of environmental threat. Depending on the level ofenvironmental threat they are divided in th ree classes. The first class are large objects, whichconsiderably influence the environment: federal and regional motorways and speedways of the Iand II technical categories with not less than four lanes and constructive works on them,separate bridges and crossovers with the length of more than 500 m. According to theInternational standards and Federal documents the construction of road objects of the first classbelongs to ecologically destructive activity categories. The second class represents the ob jects,which considerably influence the environment. They include the roads of II and III categorieswith predicted (perspective) rate of traffic more than 2000 veh. per day and the constructions ontheir surface, separate areas of other roads in populatio n centres and particularly protected areas.The third class is represented by the objects which have insignificantly influence, local actionon the environment: automobile roads with predicted traffic rate less than 2000 veh. per daytransport constructions on their surface, repair works.

Under environmental threat (safety, environmentally safe state) of the automobile roadthere is understood the ability of the road to provide the minimum of hazardous, formed byengineering constructions and constructions of the automobile road, impacts and pollution ofnature of the areas attached to the roads, their influence on the work of the road transport. Thelevel of environmental threat (safety) of the automobile road depends on its technical conditionand the technical state of the road buildings, the level of contamination of the naturalenvironment of the wayside, as well as influence of the technical condition of the road on thepollutant emission of the road transport.

With the purpose of quantitative estimatio n of the level of environmental security(environmentally safe state) of the road there are proposed special rates, which are divided intotwo groups – ecological and ecologically significant. The ecological ones include the rates,which characterize the level of air, water, soil pollution, bioenvironmental effect (human, flora,fauna) and reflect the cooperative effect on the nature of road transport as well as engineeringconstructions of automobile road. Ecologically significant rates include those, whic hcharacterize the technical condition of elements (constructions) of roads or maintenance works,which reflect the influence and environmental effect of the road and the effect of the latter onthe ecological rates of road transport. The level of ecologic al safety of the road is evaluated bycomparing factual and regulatory values of ecological and ecologically significant rates, statedin quantitative or qualitative form.

The state of road will be considered environmentally safe if:

There is no violation and pollution of the roadside territory, formed and caused byengineering constructions and road constructions, or they are as low as practicable with theexisting technologies and modern requirements.

There are created conditions, which provide the minima l possible (with the existingtechnologies and modern requirements) impact on nature from the side of road transport,which is at the road.


CT 2

Quantitative values and qualitative assessments of the environmentally safe state of road,its engineering facilities and constructions are represented in the branch regulatory document«Rates and norms of ecological safety of the road», prepared by the Road Agency of Ministry ofTransport of the Russian Federation and set in force since January 1, 2003.

With the purpose of environmental safety improvement of the roads in Russia there havebeen worked out the rules and norms of environmental design of road elements and roadsconstructions. The examples of such rules are the rules and standards of design and constructionof rest areas at the roads. The rules efficient up to the present moment both in Russia and abroadare made solely on the basis of requirements of road traffic safety ensuring. However the growthof traffic volume lead to the fact that rest areas are overl oaded by road transport, theenvironment of roadside territory cannot resist the excess human load. As a result rest areas donot carry the assigned functions – provide rest neither for the drivers nor for pedestrians and, bythis means, do not contribute to safety improving of road traffic.

The research of the recent years showed the significant impact of roadway coverings on thefuel consumption of automobile engines, and, in such a way, on the volume of exit gases.Beside the environment-oriented values the right choice of the material and texture options ofpavement also has an energy-conservative value. The research carried out, the results of whichare presented on the Fig. 1, revealed quite a complicated character of interrelation between thematerial and surface texture and fuel consumption in the whole actual speed ranges ofautomobile traffic. It is estimated that, on the road sections with the average speed of traffic of80 and more km/hour the minimum fuel consumption are observed on the cement -concretepavement in comparison with the asphalt -concrete. In case of traffic motion of less than 80 kmper hour there is observed quite an opposite picture.

Figure 1. Impact of the material and texture options of the roadway covering

on the motor car fuel consumption

Evenbituminousconcrete

Surfacetreatment

Cementconcrete

Fuel

con

sum

ptio

n, g

/km

Speed, km/h


CT 2

II. DEFORMATION OF THE ROAD EMBANKMENT SLOPES

The practice of construction and reconstruction of roads showed that the basic types ofdeformation of the earth embankments (roadbed), made from granular materials (sands, sandand gravel ground etc.) are surface erosion and local shear deformations in the form oflandslides, earthflows, caused by the impact of water on the ground. Such kind of deformationone may see in the regions with quite a cold climate, in the regions with snow falls, snowstormsand cold winter. This is the Northern and Central parts of the European territory of the RussianFederation, the whole territory of Siberia and Far East of the Russian Federation, Alaska (USA),high mountain areas of China (Tibet). It is confirm ed by the observations of numerous authors(Fig. 2) [1,2,3,4].

а) б)

в)


CT 2

г)

Figure 2. Local deformations of subgrade embankment:

а) road in Alaska [4]; б) Qinghai - Tibet railroads [3]; в) Russia - road «Yamal» [1]; г) Russia - road «Don» 104 km [1]

Deformations of grade level, caused by water erosion, are developed in the period, whenthe surface of formation is still not hardened and caused by considerable overspeeding of thewater flowing from the ground surface (usually dur ing the rains) of the standard (not eroding)speeds for the ground. The ways of prevention of such kind of deformations are well -known – itis timely embedment of the traffic way, waysides and slopes of grade level with the materials,which are highly resistant to the washaway.

The situation is more difficult with the deformations of the second type – shifts on theslopes. Local deformations of this type can be observed on the embankment slope of all types ofground. What is particularly interesting is the fact of appearance of shear deformations on thefill slopes from the cohesionless soil. Besides such deformations develop on the slopes of evenhigh fills (up to 8 and more m), with asphalt and cement -concrete pavements at the carriagewayand shoulders, with good grassy turfs at the slopes. Particularly often these deformations occurin the first 1-3 years of grade level service.

The possible schemes of local deformation developments are presented at the Fig. 3. In allcases there are slip lines of the defi nite coat massif of soil in the coating surface of the slope:

а) б)


CT 2

в)

Рис. 3. Development scheme of shear deformations on the fill slopes [6]:

а – due to identity element; б - inplane slip with uplift; в – destruction of the whole slope in the belt of

weathering on the circular cylindrical surface ; 1 - face of slope; 2 – capacity of the active zone ha; 3 –shift surface; 4 – assumed blocks; 5 - retaining prism in uplift zone

The condition of the slope stability is the balance or excess of restraining forces over theshear forces. Stability coefficient is:

tgαZitgψ

tgαZiγ

nСntgZiγзапR

ZiγnC

ntgZitg ψ

where γ - soil density; Zi - running coordinate of the active zone capacity of the slope

perpendicularly its surface; Zitgψ - coefficient of soil shift of the active zone h at depth Zi;

ntg , Cn - correspondingly calculated values of angle of repose and soil cohesion at depth Zi;

- rate of slope.

The analysis of the complex of restraining force showed that, the main role in the loss oflocal soil stability on the slopes is played by water, which causes decrease of angle of reposeand cohesion between the particulates and dynamically effects the soil grains.

Structural cohesion Сп in graded materials takes place only in case of high density and soilcompactness and predominantly in case of low homogeneity on grain -size classification and ispredetermined, mainly, by interlocking grain arrangement [9].

Table 1. Dependence of cohesion and angle of internal friction of soilfrom its porosity [9]

Cohesion С (МPа) and angle of internal friction (grade) withthe porosity factor Type of refuse stone

0,45 0,55 0,65 0,75

0,02 0,01 - -Gravel and coarse sand

43 40 38 -

0,03 0,02 0,01 -Sands of average coarseness

40 38 35 -

Fine sand 0,06 0,04 0,02 -

H


CT 2

38 36 32 28

0,08 0,06 0.04 0,02Dust sand

36 34 30 26

Note: Upper line - cohesion, lower - angle of repose.

The water gets into the soil on the slopes as a result of percolation in case of storm eventand snow melting.

In winter the soil of grade level freezes (after the temperature fall below -5°С). Isothermalcurve of zero temperature falls lower and lower from the surface of grade level. Temperaturedistribution in depth gives evidence of the character of the soil straight -freezing: maximal underthe roadway paving and lesser on the slopes of fi lls (Fig. 4 /8/).

Isotherm of February

Isotherm of March

Isotherm of April

Figure 4. Isotherms (°С) of the coat of grade level during the winter -spring months

(Moscow and Moscow region)


CT 2

In spring there is started a constant soil temperature rise i n the upper part of the grade level.Heat current changes its direction, moreover before the start of melting. Soil frost retreat startsfrom two sides: from above, from the surface of grade level, and from below, from the side ofthawed ground (in the mess or the ground of the grade level). The speed of frost retreat fromabove is more or less identical on all the areas and averages (for Moscow region) to 4 cm/day.Frost retreat from within averages to 0,6 -0,7 cm/day. On the whole the thickness of layer,melted from within, amounts - in relation to the whole thickness of frost -bound layer - to 7 up to34%.

After the start of snow melting the water from the upper coating of the snow cover, subjectto the forces of gravitation, passes through the snow to the soil slope. Under the influence ofmelt-water there is started gradual coat frost retreat. The part of melt -water gets into the pores ofthe unfrozen soil, the remaining part flows through the slope – through the face of slope, underthe snow cover. As the snow melts and the soil thaws the major part melt -water gets into the soilpores and the smallest part of it flows down the slope surface. At last there comes a moment,when the depth of the melted soil -work at the slope surface reaches the value, wherein the allamount of melt-water which enters the soil goes to the soil pores. The flow down the surface ofthe slope stops. Melt-water through the soil pores under the gravity forces reaches the surface ofthe soil still not melted. In case of quite a large openness there appears the water flow in thesoil. Gradually takes place the formation of seepage, which flows in the soil above the border ofthe section «thawed ground - frozen ground» (Fig. 5.)

Figure 5. Formation of the seepage on the slopes of the grade level during the snow melting

As a result there happens a considerable soil overwetting, followed by the decrease offorces, which secure the soil grains from the shift. In the zone of water filtration on soil grainsoperates the hydrodynamic head hB, which appears as the result of penetration of elementary

Snow cover

Water seepageDry ground

Frozen groundThawed ground


CT 2

rate of water flow q, and the following formation of seepage with the rate iqQ . Elementary

rate of flow qi is formed by water, which penetrates into the soil during the sno w melting on theslopes, and, in case of storm event, rainwater. The water flows through the surface of aquifuge –the surface of still frozen soil -work (in spring) or the surface of a more solid soil -work whichlies lower (in summer and in autumn).

The melting surface is not plain. That is why in some places, because of the outflowobstacle, the local additional body of water may occur, which increases weighing water impactand therefore decreases restraining forces.

The water of rains which fall during th e snow melting period accelerates and increases theprocess of snow melting, therefore leveling up the water flow in the coat. The rainwater itselfalso penetrates into the soil pores (because of the infiltration) increasing more the filtration flowand soil dampness. Because of the accelerated snow melting and soil frost retreat in the zone ofthe shelf of grade level there is possible a situation, when the water from the overdamping zoneunder the roadway paving through the unfrozen coat under the wayside and the upper part of fillslope comes into the filtration flow, which flows in the surficial belt of the slope.

At some time the soil overdamping reaches the level when, the shearing force exceedrestraining forces. So there happens a shift - local deformation in the form of slope gutter.

As regards sand the possibility of shift deformation is worsened by its tendency toattenuation in aqueous state. Attenuation often happens [5] under the influence of filtration flowon the sand structure, in particular in case of dynamic character of filtration forces. Recentlysettled refuse stone of earthworks is very sensitive to the dynamic forces. Dynamic effectsusually cause small shift of sand-grains, which cause sand fluidization.

In case of sand fluidization on the slopes, instead of vertical displacement of sand -grains inthe process of sand settlement, there occurs considerable relative flat and vertical displacementof values as a consequence of running ground dispersion. In case of sufficient surface slope t heburdens rush in the form flows to the lower areas, forming the covers, filling the cavities andhollows.

Deformation ratio depends on the rate of dynamic effects. Earthquakes can cause passingof sand into dilute state on the large area. The effects of explosions and vibration are causedonly by local fractures of area structure, quite close to the whence of dynamic effects. Veryoften the fluidization event happens in comparatively small scales, for example, in the event ofpeople walking or vehicle passing over the surface of loose water - saturated sands [5].

Fluidization is native to all quite loose granular soils of any grain size. However due to alarger permeability to water the retention time of coarse -grained soils in dilute state is less, thanthat of compact-grained and that is why the fluidization practically never occurs there.

The danger of fluidization for the resistibility and structural competence is defined not bythe fact of fluidization, but by the character of its flow. The dwelling tim e of sand in dilutecondition and toughness of burdens influences the possible construction displacement.


CT 2

The rightfulness of theoretical considerations concerning the reasons for formation of localdeformations on the slopes of grade level of roads was confirmed by the results of full -scalemeasurements of water content and soil (sand) density on the fill slopes of roads «Don» (km103-104), built from fine sand of the borrow pit «Martemianovo» of Tula region (filtrationcoefficient 1-3 m/day, gradation factor -1,67). Embankment height- from 1,5 to 8 m. Theresearch was carried out in the years 2001 -2005. Water content and soil density on the slopeswere defined at depth 0, 20, 40 and 60 cm. during different times of the year. The depth wascounted from the lower surface of top soil. The research was carried out in field and laboratoryconditions with the use of certified appliances. The character of coat moisture gradient of theslope part during the spring months is shown on Fig. 6.

Рис 6. Coat moisture gradient of the slope part of grade level in spring 2004 year. km 104 а/r «Don». 1 -5

- point on slope contour: 1 – on the edge, 5 - embankment foot, 2-3 - passing points

Table 2. The character of placement of thawed and frozen layers on the slope of the fill in the first

half of the day March, 2004, km. 104 а/r «Don», depth of fill 8 m, slope ratio 1:1,75. Slope orientation -

south, air temperature at night -10°С, day + 5°С

№ layer Soil state in the layer Layer height, m

1 Frozen ground 0,03-0,05

2 Hydromorphic soil 0,05-0,10

3 Fluidity soil 0,05-0,10

4 Frozen soil 0,15-0,20

5 Non frozen soil -

Embankment edge Center Embankment toe

Dep

th o

f so

il sa

mpl

ing,

cm


CT 2

The fact of water flow in the soil (filtration flow) was photographed (Fig. 7)

Fig 7. Water filtration on the border of frozen and unfrozen soil (km 104 а/ r «Don», 14.03.2004)

Density measurements, carried out simultaneously with the humidity estimation showedthat the soil on the slopes is in quite friable state (table 3).

Table 3. Soil density and humidity of the

fill slope on 104 km road «Don» (average rates). Slope orientation - south. 14.03.2004

№ measurement point Depth of measurement point, cm Coat density g/cm3 Coat humidity %

1

0-20-40-60

1,831,801,761,87

14,513,817,315,7

2

0-20-40-60

1,851,851,831,87

10,012,515,115,7

3

0-20-40-60

1,831,801,761,87

9,112,516,017,5

4

0-20-40-60

1,831,801,761,87

10,912,215,615,0

5

0-20-40-60

1,761,831,831,85

13,014,016,217,5

Note: Optimum density for this sand is 1,89 g/cm3, optimum humidity - 10,9%.


CT 2

Actual values of compacting factors (0,93 -0,95-0,97) during the first 2,5 years of slopework (0,93-0,95-0,97) turned out to be lower than regulatory value (min 0,99), which certifiessoil high porosity on the fill slopes.

The research of dynamic effects of automobile transport on the soil of the slope fill partswas carried out at 104 km а/r «Don» (depth of sand fill - 6-8 m) и9и!4km МКАR (depth of sandfill 2 m). On the а/r «Don» convulsion in coat generated by passage of single -unit truck of themass 22 t with the speed of 50, 60 and 80 km/h, at Moscow Ring Motorway there was moving areal traffic flow with the intensity (in one direction) 6480 veh/h (km 9) and 7200 veh./h (km14). The carriageway of the road «Don» has 4 lanes (two lanes in each direction), at MoscowRing Motorway - 4 lanes in each direction. Shoulder s at 1,0 m from the upper edge ofembankment are hardened by plant formation. In both cases the the road pavement made ofasphaltic concrete, the roadbase - of low cement content concrete, base – of sand.

Vibrational impact of automobiles on the soil of gr ade level was studied in dry weather, inJuly under the temperature of + 23°С and in November under the temperature of +4°С. Therewere registered mean square and peak heights (X, Y, Z) of vibration acceleration. Axle X isdirected perpendicularly to the road axle. The measurement time amounted to 5 to 10 minutesand included the automobile drive to the measurement point and automobile removal.

There were made measurements (the values of vidroaccelerations), processing of the resultsreceived according to the finite element method enabled to define, that the vibration impact onthe slopes for the conditions discussed in case of problem solving on normal stress amounted,average, from 0,1 up to 0,044 kg/cm2, tangentially - from 0,04 to 0,001 kg/cm2. Value pea ksfall within the upper and lower slope part, which suggests the increased load in these zones. Themovement of soil parts amounts to 0,6 up to 0,2 mm and on the whole uniformly decreases inproportion to the standing off pumping source (from the cover of the road) (Fig. 8).

III. CONCLUSION

The results of the research carried out enables to make a conclusion that local soil

deformation on the embankment slopes are determined by the combination of the range of

factors: low soil density in the slope surficia l belt, high soil moistening in spring period, the

presence of filtration stream of melted (and rain – in case of rain fall) water in the slope part of

the roadbed. Vibrations generated in the soil of the roadbed by the cars passing by contribute to

the disturbance of equilibrium of restraining and shearing forces.

Only one from the abovementioned factors is subject to control by the roads constructors -

soil density of slope parts of embankment. However at the present time the embankment

construction method implies that the slope soil is not compacted The technology of soil

compacting of slope of embankments still is not worked out. The recommendations concerning

the following overcutting of the unconsolidated slope soil, which one may find in references,

cannot be considered as rational due to many reasons. As a consequence, the tools for the works


CT 2

execution on slope soil stabilization are not available (one cannot view a small road roller,

which rolls down the slope as a major compaction tool).

Рис 8. Curve of the soil particles flow (mm) as a result of operation of single load. Summary constituent,

disturbing frequency 10 Hz

References

[1]. Evdokimov V.I. The nature of Yamal and roads: complicated interaction. The Journal «Russian Roadsof XXI century», № , 2005.

[2]. Kupachkin B., Radkevich А . Introduction to the new soil mechanics. The Journal «Russian roads ofthe XXI century», № 1, 2006

[3]. Kondratiev V.G. Qinghai-Tibetan railroads: new experience of the construction of grade level ondeep-frozen soil. The Journal «Transport construction», № 4, 2005.

[4]. G.Grondin, A. Guimond , G. Dore Impact of permafrost thaw on airfield and road infrastructures inNunavik - Quebec. «ROADS» (PIARC), № 332, 2006.

[5]. Ivanov P.L. Dilution of sandy ground. М., SEI, 1962

[6]. Nemchinov М.V., Men’shov А.S . Influence of vibration of the road transport on the local slopestability of road bed. The Journal «Science and Engineering for Roads», № 4, 2005.

[7]. Men’shov А.S. Provision of local slope stability of hi gh fills of roads made of granular soil. Ph.D.Thesis. Мoscow, МАDI(SТU), 2006.

[8]. Zolotar’ I.А., Puzakov N.А., Sidenko V.М . Aquatic- heating parameters of the grade level androadway paving. М., Publ. House “Transport”, 1971

[9]. SNiP 2.02.02-83. Base of buildings and constructions. М., 1985

[10]. Beliaev D.S., Yushkov B.S., Kychkin V.I., Rukavishnikova N.Е . Development and approbation of thevaluation method of technical condition of the grade level of the coats of transport works. The Journ al«Russian Roads of the XXI century», №3, 2005

Sand

1:1.75


CT 2I. INTRODUCTION

Many countries have, for many recently years, been establishing policies aimed to utilise as

much as possible the use of local materials in building construction. Sand concrete is a family of

cement concretes which can be used to overcome limitations about environmental or economic

problems in the use of coarse aggregate. Especially, some certain regions it has the depletion of

coarse aggregate deposits.

Back in 1869, concrete without coarse aggregate was used for buildings, the 52m high

lighthouse of Port-Said (Egypt), which is still in used, was built with sand only [1]. In Russia in

vast areas sand is the only building material to be used in concrete. Here, a lot of buildings have

been constructed with sand concrete since fifties. The und erground station in St.Petersburg was

built with precast arches of sand concrete. In Germany first investigations to increase the sand

content (more than 60 %) in a concrete mix design were made in 1971[2]. In France, since 1998

the national “SABLOCCRETE” project done with the cooperation of Russian, Algeria, Maroc

is a big project sand concrete [1].

In the Mekong Delta in Vietnam, there is an abundance of sand but lack of coarse

aggregates to produce traditional concrete. Therefore, the use of sand concre te to substitute for

AN EXPERIMENTAL RESEARCH ON SAND CONCRETEIN MEKONG DELTA

NGUYEN THANH SANG , Doctoral studentPHAM DUY HUU, ProfessorInstitute of Science and Technology for Transport constructionUniversity of Transport and Communications

Abstract: This paper presents an experimental research on mechanical properties of sand

concrete for possible use in the Meko ng Delta. The research has been car ried out at the

University of Transport and Communications, Hanoi, Vietnam. By applying experimental

method, several typical scenarios with different proportions of water, cement, sand, fillers,

and additives have been examined to determine the following m echanical properties of the

sand concrete: compressive strength, flexural strength, splitting strength, elastic modulus. The

obtained results showed that the sand concrete can be usable for different construction

projects in the Mekong Delta.


CT 2

traditional concrete is considered since transport of coarse aggregates from other regions to the

Mekong Delta is very costly.

Sand concrete is fine concrete consisting of a mixture of sand, cement, filler, and water.Besides these basic components, sand concrete typically includes one or more admixtures.When fine gravel is incorporated with sand and their ratio G/S remains below below 0.7 (withG= Gravel, S=Sand) the mix then can also be socalled as sand concrete. The sand concrete i sdistinguished from a traditional concrete by using high proportion of sand; with a smallproportion or without using fine gravel and the incorporation of filler. It is also distinguishedfrom the mortar by its composition (mortar generally contains high cement content) andespecially by its destination, as sand concrete are primarily intended for more traditional uses.

In order to apply this material into building purpose in Vietnam, several mechanicalproperties have been carried out. The result of this study will be characterized hereafter.

II. EXPERIMENTATION

2.1. Characterization of the used materials

2.1.1. Sand

Fine sand (FS) extracted from Vinh Long province near of Ho Chi Minh city and featuringby a maximum particle diameter of proximately 2.3 6mm. The proportion of grains (smaller than0.075mm) is 2%; organic impurities are lighter than standard colour.

Coarse sand (CS), from Tri An lake (region around Ho chi Minh city), presentscontinuous particle size distribution ranging from 0.075 to 4.7 5mm. However the fractionsmaller than 0.30mm remains very small; the proportion of grains smaller than 0.075mm is1.3%; organic impurities are lighter than standard colour.

Fine sand and coarse sand were mixed with a ratio FS/CS equals to 1.7 by mass. Theparticle size distributions of the various sands used are shown in Fig 1. Table 1 lists the set ofphysical characteristics for three types of sand. The modulus of fineness of mixed sand equals to2.75, and packing density is 60.5%.

Table 1. Physical properties of the various used sands

SandBulk density

(kg/m3)Specific density

(kg/m3)Finenessmodulus

Packing density(%)

Sandequivalent

FS

CS

FCS

1399

1510

1534

2500

2560

2536

1.74

3.30

2.75

56

59

60.5

84

87

86


CT 2

9.54.75

2.36

1.18

0.15 0.60.30.0

750

10

20

30

40

50

60

70

80

90

100

Grain diameter (mm)

Pass

sing

%

FS CS 'Ideal' curve CFS

Figure 1. Granular curve of the different sands

2.1.2 Cement

The cement used is Nghi Son PCB40 cement (similar to CEM I); it chemical analysis andcomposition are given in Table 2. The physical characteristics are following: specific density3100 kg/m3 and Blaine specific surface area 3690 (cm 2/g).

Table 2. Chemical analysis of the cement used

SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O free CaO

21.29 5.72 3.30 63.18 1.1 1.9 0.12 0.30 0.193

2.1.3 Fillers and admixture

The fillers used have been obtained by sifting (passing of 80 m sieve), from Hoa An

quarry (region around Ho Chi Minh City), and are mainly composed of limestone (98 mass % of

CaCO3). Its characteristics are following: specific mass 2740 kg/m3 and Blaine specific surface

area 3210 cm2/g.

The admixture used is a super plasticizer (a typical Sika product), with a dry matter content

of 1% of cement mass.

Drinking water is suitable for use in this concrete.

2.2. Preparation of sand concrete samples and testing

15 proportions of sand concrete mixes wich contains a mix of these two sands were used.

They have three various factors includes: water/cement ratio, filler content, and age of sand

Pass

sing

%

Grain diam eter (mm)

FS CS ‘Ideal’ curve


CT 2

concrete. The factors effected on mechanical propert ies of sand concrete were investigated.

The specimens produced have been cured in air at 27±2 oC for 24 hours. Then, they were

removed from the moulds and immesed in water until the day of testing.

Compressive strength, splitting strength, Young’s modulus o f elasticity were determined on

cylindrical specimens with 150mm diameter and 300mm height. Flexural strength was

determined (using three-points method) for each mix on three 100x100x400 mm prismatic

sample.

2.3. Experimental plan

In this experimental study, compressive strength (fc), flexural strength (f r), splitting strength

(fsp), and elastic modulus (E c) vary from water/cement ratio (w/c), filler content (f), and age of

sand concrete (t). The experimental design theory [11] is used to establish an opt imum

experimental procedure [8, 11] and to elaborate empirical models considering both experimental

parameters (Input: w/c, f, t) and results (Out put: f c, fr, fsp, Ec). True values (w/c, f, and t) and

normalize ones (x1, x2, x3) in [-1, 0, 1] interval is given in Table 3, 4. This transformation

enables to analyze in the same manner, both qualitative and quantifiable data [11]. In the

following, all capital letters (C: Cement, W: Water, S: Sand, F: Filler, A: additives) refer to

component weight per cubic meter of mix. When using experimental design, priority semi -

empirical models are built from mathematical expansion of the outputs as following:

y = a0+a1x1+a2x2+a3x3+a12x1x2+a12x1x2+a23x2x3+a11x12+a22x2

2+a33x3

These coefficients of the model are identified through regression analysis and all possible

parameters are ordered such as to keep only the most influent. In this case, 15 mixes are

required to establish the mathematical formulation.

Table 3. Field of parameters

Truevalues

Field of parameters Middle values Interval Normalize values

w/c 0.38 to 0.52 0.45 0.07 x1

f 100 to 150 125 25 x2

t 0.845 to 1.447 1.15 0.30 x3

The experiments were designed based on the orthogonal array technique and ac cording toStandard NF P 18 - 500, proportion of sand concre te are given in table 4.


CT 2

Table 4. Parameters their values

No Normalize values True values

x1 x2 x3 w/c f t

1 1 1 1 0.52 150 28

2 1 1 -1 0.52 150 7

3 1 -1 1 0.52 100 28

4 -1 1 1 0.38 150 28

5 1 -1 -1 0.52 100 7

6 -1 1 -1 0.38 150 7

7 -1 -1 1 0.38 100 28

8 -1 -1 -1 0.38 100 7

9 1.732 0 0 0.57 125 14

10 0 1.732 0 0.45 168 14

11 0 0 1.732 0.45 125 56

12 -1.732 0 0 0.33 125 14

13 0 -1.732 0 0.45 82 14

14 0 0 -1.732 0.45 125 4

15 0 0 0 0.45 125 14

Table 5. Sand concrete mix proportion for e xperiment

Compositions ( per 1m3)

Cement Water Sand Filler Additive TotalNo Ratio

(w/c)

Fillercontent

(f)

Age(t)

C (kg/) W (l) S (kg) F (kg) A (l) (kg)

1 0.52 150 28 413 215 1,507 150 4.55 2,290

2 0.52 150 7 404 210 1,507 150 4.44 2,275

3 0.52 100 28 413 215 1,532 100 4.13 2,264

4 0.38 150 28 513 195 1,444 150 5.64 2,308

5 0.52 100 7 404 210 1,532 100 4.04 2,250

6 0.38 150 7 513 195 1,444 150 5.64 2,308


CT 2

7 0.38 100 28 513 195 1,510 100 5.13 2,323

8 0.38 100 7 513 195 1,510 100 5.13 2,323

9 0.57 125 14 350 200 1,515 125 3.50 2,193

10 0.45 168 14 467 210 1,469 168 5.13 2,319

11 0.45 125 56 456 205 1,503 125 4.56 2,293

12 0.33 125 14 563 185 1,476 125 6.47 2,356

13 0.45 82 14 433 195 1,528 82 4.33 2,242

14 0.45 125 4 444 200 1,503 125 4.44 2,277

15 0.45 125 14 456 205 1,503 125 4.56 2,293

Each sand concrete proportion made in the laboratory has chosen from the experimentalplans. Concrete-mixer has revolving-paddle and 180 dm3 in volume and the rate of rotation is 37revolutions per minute. Mixing time is from 4 to 6 minutes. Making and curing concrete testspecimens, experimenting according to ASTM are given in Table 6. The results of theexperiment are given in Table 7.

Table 6. Measurement according to

No Testing Standard

1 Making and curing ASTM C192/C 192M-02

2 Slump of fresh concrete ASTM C143/ C143M-00

3 Compressive strength ASTM C39/C39-01

4 Flexural strength ASTM C78-02

5 Splitting strength ASTM C496-96

6 Young’s static modulus ASTM C469-02

Table 7. The results of experiment

S fc fr fsp EcNo x1 x2 x3

(cm) (MPa) (MPa) (MPa) (MPa)fc/fr fc/fsp fr/fsp

1 0.52 150 28 8.5 35.17 4.29 3.39 28,750 8.20 10.38 1.27

2 0.52 150 7 7.9 29.90 3.97 2.20 23,321 7.53 13.58 1.80

3 0.52 100 28 8.6 37.91 4.33 3.58 28,376 8.76 10.58 1.21

4 0.38 150 28 4.5 47.83 5.01 4.62 35,303 9.55 10.35 1.08

5 0.52 100 7 7.3 24.38 3.61 2.32 20,861 6.75 10.52 1.56


CT 2

6 0.38 150 7 8.8 37.21 4.46 3.20 29,202 8.34 11.65 1.40

7 0.38 100 28 5.5 55.42 5.32 4.63 34,255 10.42 11.96 1.15

8 0.38 100 7 12 37.54 4.37 3.59 28,647 8.59 10.46 1.22

9 0.57 125 14 8.5 28.60 3.55 2.87 22,184 8.06 9.95 1.24

10 0.45 168 14 8 42.28 4.07 3.04 32,270 10.39 13.93 1.34

11 0.45 125 56 9.4 44.02 4.36 4.04 36,423 10.10 10.89 1.08

12 0.33 125 14 3.5 48.33 5.39 3.83 35,803 8.97 12.62 1.41

13 0.45 82 14 4.5 37.24 3.65 2.87 32,557 10.20 12.97 1.27

14 0.45 125 4 5.5 29.26 3.81 2.52 23,840 7.68 11.61 1.51

15 0.45 125 14 3.5 39.57 4.12 3.02 32,239 9.60 13.09 1.36

2.4. Results and discussion

As the results in table 7, determination of regression model of mechanical properties is

completed by the aid of a computer with Maple software. This way is a quite simple and very

quickly to gives precise values of coefficients. After the regression models were determined.

Equation are verified as following [8]:

+ To verify the suitable results of experiment in accordance with the Cochran law.

+ To estimate coefficients to fit with by statistical method (at a 95% confidence level) in

accordance with the Student law.

+ To verify the optimal process quantities (factors) through the confirmation of

experiments the Fisher law.

Solving and fitting regression models are following:

Fitted model of compressive strength:

+ With normalized values: y1 = 39.57 – 6.06 x1 + 5.21 x3

+ With true values:fc = 58.69 – 86.53 (w/c) + 17.29 lg(t)

Fitted model of flexural strength:



CT 2

+ With true values:fr = 5.99 – 6.27 (w/c) + 0.83 lg(t)

Fitted model of splitting strength:


+ With true values:fsp = 3.81 – 6.33 (w/c) + 1.80 lg(t)

Relation between compressive strength and elastic modulus was established:

Ec= 1808.fc0.767

From the obtained results in Table 7 and regression models several discussio ns can be

withdrawn:

In valid selective factors, types of sand concrete obtained compressive strength from

25MPa to 55MPa; flexural strength from 3.55 MPa to 5.93 MPa; splitting strength from 2.20

MPa to 4.6 MPa, and elastic modulus from 20861 MPa to 36423 MPa respectively. This results

point out the rate of an increase in compressive strength as fast as the rate of an increase in

flexural strength and splitting strength. The investigation of types of fracture shows : 69% cone

type (a), 22% cone and split type (b), 4% shear type (d), and 4% columnar type [ASTM C39].

Almost specimens of sand concrete have the type of fracture similary to the type of fracture of

traditional concrete. This manner availably says that “theories of mechanical properties of sand

concrete similar to those of traditional concrete”.

Regression models point out that the sensibility of w/c and t to the strength of sand

concrete is more than the sensibility of f .

The ratios of fc/fr, fc/fsp of sand concrete and traditional concrete which have compressive

strength from 15MPa to 55MPa are from 6.7 to 10.5, and from 9.9 to 13.9 in sand concrete;

from 9 to 12, and from 10 to 15 in traditional concrete, respectively [10]. Therefore, the rate of

an increase in flexural strength and splitting st rength of sand concrete is more fast than the

those of traditional concrete.

III. CONCLUTION

* The obtained results from this research, would hopefully be initial contributions to the

use of sand concrete in the Mekong Delta region in Vietnam (especially, aiming at enhancing

the reuse of local materials in some regions of Vietnam).


CT 2

* An experimental work has been performed to study how water/cement ratio (w/c), filler

(f) content, age of sand concrete (t) can modify some mechanical properties of sand concr ete.

* In this study, fitted models of experimental theory were simple equation to using to

design the proportion of sand concrete.

* Theories of mechanical properties of traditional concrete can be used for sand concrete.

* Mixing time is a half as long as compared to traditional concrete, and vibrating time is

shorter. As results, enery comsumption can be reduced.

Reference

[1]. Presses de l’Ecole Nationale des Ponts et Chaussées, Paris, ISBN: 2 -85978-221-4 (1994). Béton desable, caractéristiques et pratiques d’utilisation, Synthése du Projet National de Recherche etDéveloppement SABLOCRETE. Vol. 237 (in French).

[2]. Pilly, F. ; Eschke, K. Sandreicher Beton, Beton und Stahlbeton. Heft 12/71. p.298-302.

[3]. AFNOR Standard NF P 18-500 (1995). Bétons de sables. 12 p.

[4]. Nguyen Thanh Sang (2005) . Research into design of component and strength of powdered -sandconcrete; Transport and Communications Science Journal; No 12; pp 106 -112.

[5]. Nguyen, S.T, Pham, H.D (2007) , Study of the effect of limestone powder on plasticity and strength of

sand concrete in Vietnam, The Transport Journal, No 7, pp 30-32.

(http://www.cauduong.net/forum_posts.asp?TID=1940 ; http://www.moc.gov.vn

http://english.vista.gov.vn/english/st_documents_abstract/ )

[6]. NCS. Nguyen Thanh Sang, GS.TS. Ph ạm Duy Huu (2008) . The rerults of an experimental stud y on

Mechanical Properties of Sand Concrete In the Mekong Delta, The Transport Journal, No 05, pp 33-35.

[7]. J.J Chauvin, G. Grimaldi ; (1998); Les bétons de sable; Bull. Liaison Lab. Ponts Chaussées 157 (9 –15 (in French).

[8]. To cam Tu, Tran Van Dien, Nguyen Đinh Hien, Pham Chi Thanh; (1999) ; Experimental Design and

Statictical Analysis; Science Technology Publishing House.

[9]. Sinan H n sl olu; Osman Ünsal Bayrak (2004). Optimization of early flexural strength of pavement concrete with

silica fume and fly ash by the Taguchi method. Civil Engineering and Environmental Systems, Volume 21, Issue 2 June,

pages 79 – 90. http://www.informaworld.com/smpp/content~content=a713947294~db=all .

[10]. M.S. Shetty (2003), Concrete Technology (Theory and practice). RAM Narga, New Delhi -110 055.

[11]. G.Taguchi (1987). System of experimental design. Unipub/Kraus international Publication, 2 24p

http://www.cauduong.net/forum_posts.asp

http://www.moc.gov.vn

http://english.vista.gov.vn/english/st_documents_abstract/


CT 2

During construction of highway roadbeds at zero elevations, and of small embankments,the ditches are being constructed in excavations and they are the roadside diversion ditches. Thepurpose of these structures is to drain the roadside area and the roadb ed soil.

However the role the ditches play in order to facilitate the roadbed construction operation isinsufficiently studied. This statement can be confirmed by the fact that mostly often designingof roadside ditches is quite a formal process - shape and sizes of the roadside ditch crosssections are specified in compliance with standard plans, trapezoidal shape with 0.3m bottomwidth and 0.4m depth, with slope inclination of 1:1.5. Hydraulic calculations for ditches areusually not done. Currently valid standards and recommendations / 1 / suggest that hydrauliccalculation be done for the ditch lined with concrete slabs (fig. 1), and as a rule there are noditch liners of such kind on flatland highways and on those ones in semi -rough terrains.

Fig 1. Structural model of a roadside ditch

1 - concrete slab; 2 - sand-gravel bedding for slabs; 3 - longitudinal seams filled with mastic

The fact which confirms the above statement is that roadside ditches are shutoff when rampsare being installed, especially in inhabited localities. In these cases a culvert (usually its diameterdoes not exceed 0.5m) is placed in the ditch and the area around is filled with soil or concrete.The ditch cross-sectional area decreases by several times. As the result ther e appears a block for waterflow in the ditch which causes water accumulation on the upstream side. These ramps are also the spotswhere different litter accumulates too and it blocks the water flow in the ditch even more.

Another confirmation of insuff icient study of role the ditch plays in facilitating the

DESIGN OF DIVERSION DITCHES FOR

HIGHWAY ROADBEDS

Prof. M.V. NEMCHINOVPh.D. student VU TUAN ANH(State Technical University – MADI,Moscow, Russia)

WATER LEVEL


CT 2

highway roadbed operation and consequently of not understanding their significance is themaintenance of roadside diversion ditches in the highways operation period. In warm season ofa year ditches become overgrown with tall grass and bushes and are not being cleared out ofthem (the grass is not mowed in the ditches) (fig. 2). The ditches are the most choked upsections of a road's right-of-way.

Technogenic litter, scrapped automobile parts (old ti res, rags, etc.) are being accumulated there.

And as a result the ditches turn from being lotic systems to being water detention basins. Water isremoved from them only by infiltration into soil and into roadbed too and it is also removed byevaporation. In temperate climate conditions of middle Russia still water in a ditch in summer season of ayear may be observed up to 1 - 2 weeks (in rainy years even longer), in spring during snow melt periodand in autumn during rain period it can be observed up to se veral months. In the monsoonal climaticregions (the Far East of Russia, Vietnam, the south of China, the Indochinese countries) in the rainyperiod of the year water stagnation in the ditch system (fig. 3) may be observed during some months.

Fig 2. The ditches overgrown with grass and bushes. Roads of Vietnam Republic, Russia, Poland


CT 2

Fig 3. Slack-water in a ditch

Stagnation of water for a long time in a roadside ditch results in an increased humidity ofroadbed soil. By infiltration water penetrates into roadbed soil and under it ( fig. 4) overwettingit and thus decreasing roadbed stability and subbase load -carrying capacity under it.

Fig 4. Relative humidity of roadbed soil under the impact of ditch slack -water

at State Highway №1A, Lang Son, Vietnam.

Note: numbers on a cross profile show relative humidity of the earth roadbed

In figure 4 it is seen that in case of a long period of slack -water in a ditch, the water

gradually infiltrates and penetrates into the roadbed body over wetting its soil. Under the road

pavement relative humidity of soil has reached 60-65% gradually increasing with the depth and

reaching 80% at the depth of 30cm (from the pavement bottom), and over 80% at the depth of

70cm, i.e. already in the soil, at the embankment base it is over 80%. Relative humidity optimal

for ensuring maximum soil density is 5 0-60% on the road №1А

Figure 4 illustrates the case of overwetting the roadbed soil which had been constructed in

zero elevation or in excavation. When stagnating for a long time in a ditch the water gradually

reaches the bottom of a road pavement. By d efinite combinations of water level and duration of

its stagnation in a ditch, soil water permeability, relative humidity of soil may reach 100%.

Little by little in roadbed soil under the pavement a free water level is being formed whichis dependent on the ditch water level. Under certain conditions there might appear a situation


CT 2

when under the road pavement a certain water head is being formed. It is in upright position andis equal to difference in the heights of ditch water level and road pavement b ottom elevation.

When a roadbed is in the form of an embankment of small height, the scheme of soilwetting shown in Figure 4 is accompanied by soil wetting which is performed by an upwardcapillary rise of water. In this situation sufficiency of the em bankment height depends on soilwater permeability and its degree of compaction during the construction process. However apossibility of overwetting and decreasing of the embankment load -carrying capacity ispreserved.

In the above considered operating conditions of roadbed and ditches (in case of waterstagnation) the decisive factor for estimating the degree of soil wetting in the embankmentbody and under it becomes the factor of time during which water remains in a ditch. This factoris determined by quickness (velocity) of water evaporation.

In operating conditions, i.e. in a situation when water flows along the ditch, the water too isbeing infiltrated into soil. Water in ditches is observed during rains and some time after. This iscaused by the water runoff coming from water catchment areas adjacent to the road a nd also theone coming from carriageway surface. The depth of water stream depends on hydrometricparameters of rain and water catchment area. In most cases the rains intensity is much le ss thanthat of a designed one and the water stream depth in a ditch is not big (in observations made bythe author the depth varied from 2 - 4 up to 10 - 12 cm). However under these conditions wateris being infiltrated into soil and overwettens it. Figur e 5 represents soil humidity data, obtainedduring rainy period on the highways of Vietnam Republic. It is seen in the diagrams that anactive water infiltration into soil of the embankment body and its base is going on.

It appeared that when ditches are in operation, i.e. when they ensure rainwater runoff alongthem, soil humidity is much less inspite infiltration process, than in the case when waterstagnates in the ditches.

Fig 5. Relative humidity of roadbed soil


CT 2

Water flow duration in a dit ch beside the hydrometric parameters of rain and water

catchment area depends on conditions of water flow. If there exists any obstacle the water flow

velocity decreases and the flow depth increases.

Thus when designing the roadside diversion ditches it is necessary to take into account

their "operating" condition, i.e. the presence of any obstacles for water flow in ditches - things

that litter ditches both of man-made and of natural origin (grass and bushes vegetation). This

can be done by improving the hydraulic calculation of a ditch.

Hydraulic calculations for the ditch are being done on the basis of the classical hydraulic

equation: w.vQ

where "Q" is capacity (flow rate), "w" is flow area,

"v" is the flow velocity.

In this formula flow velocity appears to be the key factor.

According to the Chezy equation iRCv

where "i" is the longitudinal flow inclination, "C" is the coefficient, "R" is the hydraulic

radius the value of which is determined by cross-sectional parameters of water flow in a ditch:

w/χR

where χ is the wetted perimeter.

The hydraulic radius R is the overall index of geometrical sizes and shape of the ditch cross

section (in the limits of cross section of th e streamflow going along this ditch).

An important indicator reflecting conditions of water streamflow is the coefficient C in the

Chezy formula. The value of this coefficient in the present time is determined by a series of

empirical formulas. There exist up to 136 of such formulas (for different conditions of water

streamflow) but the main ones are:

For turbulent conditions of water movement (which is typical of the situation with water

flow in a ditch) the coefficient C may be determined by formulas of different types. For open

channels with absolutely rough walls and for open natural beds the most used formulas are:

- The Basen formula: )R/1n87/(1C

where n1 is the Basen's hydraulic roughness coefficient. The Basen formula is used when

calculating water diversion ditches.

- The N.N. Pavlovsky formula: 2nyRC

where n2 is the N.N. Pavlovsky's coefficient of hydraulic roughness.


CT 2

R0,12n0,750,132n2,5y

This formula is to be the main one at calculating beds with absolutely ro ugh sides.

- The Manning formula: 3n61

RC

where n3 is the Manning's coefficient of hydraulic roughness.

- The I.I. Agroskin formula: lgR)(k2g4C

where "k" is the coefficient which characterizes absolute roughness both qualita tively and

quantitatively, i.e. according to its type and sizes.

Values of the hydraulic roughness coefficient differ considerably based on data obtained by

the authors mentioned above. For instance, Basen's n 1 depending on channel sides and bottom

roughness is n1 = 0.50…4.00; N.N. Pavlovsky's and Manning's n 2 = n3 = 0.012…0.150; I.I.Agroskin's k = 3.15…1.90.

Research data of Basen, N.N. Pavlovsky, Manning, I.I. Agroskin and of other authors / 3 /

show that:

1. Bed and sides roughness (i.e. the degree of their overgrowing with grass and bushes,

littering with stones and etc.) has a great impact on water flow velocity (up to 7 -8 times and

more).

2. When designing the ditches and doing hydraulic calculations it is necessary to carefully

estimate the future operating condition of ditches and in connection with that to choose value of

the design coefficient of bed sides roughness.

Obstructions found in ditches and blocking water flow in there may be divided to two

groups (according to character and degree of im pact on water flow in a ditch). One group is the

grass growing on the bottom and on the slopes of a ditch. It might be considered to be the

roughness of bottom surface and bed sides and in this case its impact is to be estimated

according to the hydraulic roughness value of bed. The other group are things of technogenic

origin (automobile parts, tires, and other kind of litter) and bushes growing on the bottom and

on the walls of ditches. In this case, depending on location, quantity, character of things th at

choke up the bed and their impact on water flow may be considered when doing calculations as

for the case when streamflow comes over a threshold or also when it comes through one or

several holes in the wall and etc.

The most frequent (practically countrywide) thing which chokes up a ditch is the grass (it

has different height in various periods of warm season of a year). Technogenic litter is sooner or

later being cleared out of ditches by the road maintenance services. Usually this is performed


CT 2

once a year before road spring acceptance (except bushes which are cleared out very seldom;

but bushes grow not so often).

Taking into consideration all the facts mentioned above a conclusion should be made that

the improvement of ditch design method must in clude:

- An obligatory determination of runoff volume on the designed section of a roadside

diversion ditch (for runoff coming from terrains of natural water catchments the well -known

methods are to be used / 3 /, for runoff coming from carriageway surface s of highways the M.V.

Nemchinov formulas are to be used / 4 /);

- Hydraulic calculations for a ditch considering its maintenance peculiarities: overgrowing

with grass and bushes vegetation (in a varying degree) or in conditions of periodic mowing

when grass is not taller than 2 - 5cm and doesn't influence considerably the flow velocity of

water; when doing calculations the Basen formula is to be used as the one which reflects to the

fullest extent the conditions of streamflow in the roadside diversion di tches (it follows from the

fact that the Basen formula was in the first place obtained for conditions nearest to the roadside

diversion ditch conditions).

- Determination of cross-sectional parameters of a ditch with consideration of character and

degree of roadbed soil wetting;

- Specifying the operating mode of a roadside ditch - whether it should be a detention

structure (perhaps on road sections located in the rural inhabited localities where there exist

numerous access ramps to the adjacent agricultur al lands or absolutely no ditch maintenance

works are conducted) or it should be a structure for transferring the collected runoff to a

discharge point.

Reference

[1]. G.A.Fedotov, P.I. Pospelov and others . Encyclopedia for a Highway Engineer. Volume V. HighwayDesign M: Informavtodor, 2007 (in Russian) .

[2]. Research Data of Highways Roadbed Hydrothermal Conditions in Vietnam Academy ofTransportation and Technology. Hanoi, 2003

[3]. Guide to Hydraulic Calcu lations of Small Artificial Structures. Under the General Editorship of G.Y.Volchenkov. M., Transport, 1974. (in Russia)

[4]. M.V. Nemchinov Adhesive Qualities of Road Pavements and Automobile's Security of Service. M.,

Transport, 1985.(in Russian)


I. INTRODUCTION

The road safety situation is a complex issue and there are high number of accidents factorsand indicators involved. The characteristics between factors and road safety (e.g. exposureversus risk) have been studied at microscopic and macr oscopic levels in several articles andmodels (i.e. OECD, 1997). Several model have attempted to represent the complexity of roadsafety problem. For instance the Haddon Matrix (1972) focused on three factors: driver, vehicle,environment (road design) at three different time phases of the crash: pre -crash, crash, and post-crash. Rumar (1999) described the road safety problem as a function of three dimensions:exposure, accident risk and consequences.

Many countries recognise the importance of internationa l benchmarking to measure andcompare their own achievements and progress in road safety with other countries. This willallow countries to learn/improve based on existing practices and lessons in other countries.

Benchmarking can identify the strengths an d weaknesses in road safety performance fromcountry to another. This can increase the awareness of the problem among public and policymakers. This will also help policy makers to take appropriate actions to solve their countryproblems.

THE EVOLUTION OF INTERNATIONAL ROAD S AFETY

BENCHMARKING MODELS: TOWARDS A ROAD

SAFETY DEVELOPMENT INDEX (RSDI)

GHAZWAN AL-HAJI, ASP KENNETHDepartment of Science and Technology (ITN),Linköping University, 601 74 Norrköping, Sweden

Abstract: Since the publication of Smeed’s model in 1949, the research on road safetybenchmarking has progressed nationally and internationally. There are mainly three types of

benchmarking: product of safety, practices, and strategic benchmarking. They differ depe nding on

the type of indicators that model is trying to compare. It is not relevant to review all literature,

however, our focus in this paper is on the background to some of the main benchmarking models

that have been used in the past and very recent. In our brief review, we describe the evolution

process of benchmarking into four generations of development. The latest generation realised the

necessity of combining up all the three types of benchmarking into one index. The Road Safety

Development Index (RSDI), as an example, provides a broad picture compared to the traditional

(earlier) models in road safety.

Keywords: Road safety, RSDI, international comparisons, ranking, composite index, macro -

performance indicators


A number of benchmarking models are already being developed and they range fromrelatively simple models to highly complex depending on the number of indicators involved,details of data and complexity of methods used in calculations and analysis.

The model in general depends on what is designed for and how is it designed? In roadsafety benchmarking between countries, three types of models, which generally used :

1. Product Benchmarking is used to compare accident rates.

2. Practices Benchmarking is used to compare activities related to human-vehicle-roadperformance (e.g. seat belts use, crash helmets use, motorways level, etc.)

3. Strategic Benchmarking is used to compare National Road Safety Programme (NRSP),management and organisational framework .

The major obstacle in constructing any benchmarking model is the lack of data fromdifferent countries, especially in developing countries. To have meaningful benchmarking, it isnecessary to have reliable, valid, and available data.

There are several reasons for reducing the n umber of input variables in most earlybenchmarking models. These include the simplification needs in the model, reducing the errors,and also reducing the cost and time of the data collection and analysis.

In our perspective, we can describe the evolution of road safety benchmarking models intofour stages of generations, which is simplified in the following description and illustration of“generations” in figure 1:

The first generation is characterized with models that compare countries’ road safetyperformance in terms of risk and exposure indicators such as accident rates and

motorisation (Product Benchmarking). These models are cross -sectional models, which

observed at the same year.

The second generation takes the time into account. Theses models bench mark the road

safety product over time series. These models are useful to monitor the trends in road

safety in countries and indicate the direction of progress ahead.

The third generation is realized the need for increased integration between product

(accident rates) and other indicators in the same model (Product and Practices Benchmarking).

The fourth generation focuses on all three types of benchmarking: Product, Practices

and Strategic Benchmarking. One approach is RSDI (Al -Haji, 2005), which integrates

much of macro-performance indicators in road safety into a single value.

Most of the early models are still in use and being applied in different studies. However,

today computers are developing rapidly, which simplifies the work and analysis of a large

amount of road safety data that was not available before. This development has made the work

in the third and fourth generations become easier and closer to reality.


PracticesBenchmarking

ProductBenchmarking

StrategicBenchmarking



ProductBenchmarking

RSDI

The first, second and third generation The fourth generation Time

Figure 1. The Road Safety Benchmarking Evolution towards RSDI

In particular, it was reasonable to start with simplified benchmarking models. Picking upideas (i.e. performance indicators) from the first three generations was very useful in reachingthe fourth generation (i.e. RSDI).

The aim of this paper is to make a selected review of the main benchmarking models inroad safety that has been used in the past and very recent.

II. THE FIRST GENERATION: LINKING MOTORISATION, TRAFFIC RISK ANDPERSONAL RISK

An early study in 1949, R.J.Smeed compared twenty countries, mos tly European for theyear 1938, where he developed a regression model (log -linear model) and he found an inverse(or negative) relationship between the traffic risk (fatality per motor vehicle) and the level ofmotorisation (number of vehicles per inhabita nt). This regression represented the best estimatesof the mean values of traffic risk for each given value of motorisation (what is called leastsquare). This shows that with annually increasing traffic volume, fatalities per vehicle decrease(see Figure 2). Smeed concluded that fatalities (F) in any country in a given year are related tothe number of registered vehicles (V) and population (P) of that country by the followingequation:

F/V = α (V/P)- β (1)

Where F = number of fatalities in road accidents in the country

V = number of vehicles in the country

P = population


Fatalities Rate(per vehicle orkilometre driven) Developing

Countries

Less DevelopedCountries

Highly DevelopedCountries

α = 0.003, β = 2/3

This formula became popular and has been used in many studies. It is often called asSmeed's formula or equation despite some authors preferring to call it a law.

Motorisation

Figure 2. Motorisation and fatalities rate internationally (based on Smeed’s formula)

This nonlinear relationship can be translated to a linear one by taking the logarithms o f thetwo sides:

Log Y = log α + β log X (2)

Where Y is F/V and X is V/P

The number of fatalities can be derived from Smeed’s formula as: F = c.V α.Pβ, where c, α,β are parameters and they are estimated from data by using the least square method . For theSmeed data (year 1938) the formula was:

F = 0.0003 P2/3 V1/3 (3)

Personal Risk (fatalities per population) is obtained by multiplying both sides of Smeed’sequation (1) by V/P as follows:

F/P = a (V/P)1-b

or F/P = 0.0003(V/P)1/ (4)

Since 1949, many studies have discussed Smeed’s equation (1) or they made a reference tothis equation. Some authors followed the equation of estimating the regress ion parameters (α, β)of the data by calculating the country road safety performance in comparison to other countries;see Jacobs and Hutchinson (1973), Jacobs (1982), Mekky (1985). They found that Smeed’sformula can give a close estimation of the actual data and it can be applied to different samplesizes of countries and years with the use of different values of α and β. Jacob and Fouracre


(1)

Learning andsociety force

Engineering andeconomy force

Individuals’learning force

Low Medium High Motorisation

(1977) applied this formula to the same sample of countries used by Smeed for the years 1968 -1971 and they found that the formula remains stable. Jacobs and Hutchinson (1973) examinedthe data for 32 developing and developed countries from the year 1968. Mekky (1985) foundthat the equation significantly captures the relationship between motorisation and traffic risk; h eused cross sectional data for the Rich Developing Countries (RCDs). Al Haji (2001) compared26 countries around the world with different levels of development. The results from this studysupport Smeed’s view of the relation between motorisation and fata lity rates. The correlationwas high, 96% of the variations are explained for the low motorised countries and 93% for thehighly motorised countries.

Some authors have tried to develop Smeed’s formula and its accuracy further by includingseveral socio-economic variables in the model. Fieldwick (1987) has included speed limits inthe same model. The number of registered vehicles has been replaced by the total vehiclekilometre driven in many late studies (e.g. Silvak, 1983). This measure (vehicle kilometredriven) was not available at the time of Smeed’s study.

Nevertheless, some other studies have tried to explain why the curve of development(fatality rates) declines downwards as been noted in many countries and shown in Smeed’sformula. The studies have analysed the factors and measures that influence the development ofthe curve of road safety. A review of these studies is reported by (Elvik & Vaa, 2004) and(Hakim, 1991). Besides, Minter (1987) and Oppe (1991b) showed that Smeed’s law is a resultof a national learning process over time. The development in society at the national level is theresult from the developments at the local level. In other words, the individuals (road users) canlearn by experience in traffic where they improve their driving ski lls and knowledge, while thewhole society can learn by better national policy and action plans. The Figure shown here illustratesthese factors on the development curve of road safety.

Figure 3. The influencing factors on the development curve of road safety


At the same time, many studies have criticised Smeed’s model because it only concentrateson the motorisation level of country and ignores the impact of other variables, see (Broughton, 1988),

(Andreassen, 1985), (Adam, 1987), where according to Smeed’s model, population and vehicles are the

only country values, that influence the number of fatalities. This means that road safety

measures have no meaning because road fatalities can simply be predicted from population and

vehicle numbers in any country and any year. Andreassen (1985) criticised the model’s accuracybecause there would always be a decline in traffic risk for any increase in the number of

vehicles, but generally in non-linear way. Andreassen proposed relating fatalities to (V) B4 where

B4 is a parameter highly related to each particular country, even to countries with a similar

degree of motorisation. Furthermore, Smeed’s study analysed data for one year, it was a cross -

sectional analysis with no time series analysis (Adam, 1987). Smeed’s formula expected thedowntrend in fatalities rate but not the number of absolute fatalities, which has occurred in

almost most western countries in the seventies (Broughton, 1988). In other words, the trend

failed to fit and predict the same as the real figures in HDCs. Broughton has concluded that:

“Smeed’s formula has no generally validity”

In later years Smeed (in Oppe, 1991a) has commented on some of these remarks that:“…We must be guided by the data and not by our preconceived ideas...The numb er of fatalities

in any country is the number that the country is prepared to tolerate…”

Also, Haight (in Andreassen, 1985) has referred to Smeed’s equation that: “… When the

formula disagrees with the observations we tend to assume that the particular area underinvestigation is safer or less safe than it ought to be… ”

Regardless of whether one agrees or disagrees with Smeed’s model, the fact remains thatthe model gave a simplified and fairly good representation between traffic risk and motorisationof different parts of the world during the earlier stages of road safety development.

At the same time, there are many other curves developed and presented in different studies

in a simple way and with a small number of indicators (motorisation, personal risk an d traffic

risk), which can describe the development of road safety in different countries. For instance,

Koornstra & Oppe (1992) have suggested the model shown in (figure 4) to describe the long -

term development of the number of fatalities over time in hig hly developed countries (HDCs).

There is an increasing S-shaped curve with regard to the development of motorisation (referring

to the number of vehicle kilometres per year). There is a decreasing curve for the development

of the fatality rates per year (t raffic risk). Together, by multiplication the values of motorisations

and fatality rates, they result in the increase and decline of the number of fatalities that have

been noticed in HDCs in recent decades.


Motorisation

Time

Number offatalaties

Time

Traffic Risk

Time

Curve C:total fatalities

Curve B:Fatalaties per

population

Curve A:Fatalaties per unit travel

Figure 4. Road safety development in HDCs (Koornstra & Oppe, 1992)

Haight (1983) illustrated the development of road safety in developing countries as shownbelow. The total number of fatalities increases, the fatalities per unit of travel decreases, and thefatality per population remains al most stable or with some decline over time.

Time

Figure 5. Road safety development over time for developing countries (Haight, 1983)

The long-run trends which are shown in (Figure 4) and (Figure 5) based mainly on repeated

cross-section surveys from different countries for different years. The objective is to show

whether the change (development) of data varies over time.

The Timo model (1998) shows curves of number of fatalities and total national mileage by

time in many eastern and western Europe countries according to the development levels of

mobility (Figure 6). At the beginning of the growth of motorisation, total fatalities are very high,

but decline continuously at a declining rate when mobility increases. When the mobility reaches

the saturation level, the decrease in the number of fatalities has slightly stopped or fluctuated.


total mileage

total fatalities

Level of mobilityI II III IV

Less DevelopedCountriesFatalities per

inhabitantsDevelopingCountries

Figure 6. Total fatalities based on the development of mobility (Timo, 1998)

The correlation between traffic risk (fatalities per number of ve hicle kilometres) andpersonal risk (fatalities per number of population) is shown in Figure 7. With a growing numberof vehicles per population, countries move from the right to the left across the curve (Fred,2001). An early level of motorisation, first leads to a growing number of traffic -related deaths,but not necessarily with the same high growth in the number of population -related deaths.However, later at a medium level of motorisation, traffic and personal risks increase and bothvalues are high. At the third higher stage of motorisation, when a country is completelymotorised, traffic and personal risks decrease. The change between the three stages is due tobetter engineering of vehicles and roads and greater understanding of the system by the ro adusers.

Figure 7. Traffic risk and personal risk in different countries where countries move

from the right to the left across the curve (Adapted from work by Fred, 2001)

As we discussed earlier, the personal risk is a function of traffic risk and motorisation.

Navin (1994) has converted this function into the following equation (see also Figure 8):


Fatalaties perpopulation

Vehicles perpopulation

0MM

efTT

(5)

Where M0 is the value of motorisation at maximum personal risk,

Tf is the point where the exponential curve meets the T-axis,

T is the traffic risk, fatalities per number of vehicles, and

M is the motorisation, vehicles per population

Figure 8. Three-dimension model of motorisation and fatality rates (Navin, 1994)

The models previously mentioned are in some way based on regression models or multipleregression models or quadratic regression models. They employ more than variables to checkthe goodness of fit to data from different countries and to find the appropriate relatedequation(s).

III. THE SECOND GENERATION: LINKING TRAFFIC RISK, MOTORISATION ANDPERSONAL RISK WITH TIME

In this generation, many benchmarking models have been developed to describe andpredict safety development between countries on the basis of time serie s models and theories.They relate the variables to a function of time to determine the long run change in safety levelover time either in a monthly form or annually. These models attempt to find the smoothedcurves to the time series data.

Koornstra (1992) has shown that motorisation is considered to be dependent on time, andthe relationship between deaths and population should include time. To measure the correlationbetween the output and input variables, one should take into account the trends in the model. Hefound the following formula for approximating the number of fatalities for a country in aparticular year:


y

1c)ktV

maxV(w

ktVxtzVtF

(6)

Where Ft is the number of fatalities for a country in a year t,

Vt is the number of vehicle kilometres travelled in the year t,

Vmax is the maximum number of vehicle kilometres,

k is the time lag in years, and

x, w, z, y, and c are constants

Oppe (1989) assumes that fatality rates follow a negative exponential learning function inrelation to the number of vehicle kilometres and time. This method has been found to be mosteffective when the components describing the time series behave slowly over time as follows:

ln (Ft/Vt) = ln (Rt) = αt + β

Or, equivalently:

Rt = eαt+ β (7)

Where the ln function is the natural logarithm,

Ft is the number of fatalities for some country in a year t,

Vt is the number of vehicle kilometres travelled in that year,

Rt is Ft/Vt and

α, β are constants

This means that the logarithm of the fatality rate decreases (sign of improvement) if α isnegative proportional with time. This model is called the negative exponential learning model,where α is supposed to be less than zero. Both α and β are the parameters to fit.

Oppe (1991a) assumes that the amount of vehicle kilometres per year is related to time andit is assumed that traffic volume will develop over time by a logistic function of a saturationmodel. This assumption indicates that th e growth rate of traffic volume is a percentage of theratio between the traffic already existing and the remaining percentage of Vm as follows:

βαt)tVmV

tVln(

Or, equivalently:

β)(ααe1

mVtV (8)

Where Vt is the number of vehicle kilometres travelled in that year, and

Vm is the maximum number of vehicle kilometres


This formula shows that countries with a large α should have a fast growth in traffic. Thetraffic volume will increase quickly first and at the end it will reach its saturation level, which

differs from country to country.

Oppe has applied the two formul as (7 and 8) to data from six highly motorised countries

over the time period 1950-1985. He found that both models describe the data fairly well. He

concluded that the development in road safety is a result of the development (learning) of the

traffic system in the country, which is more or less similar to Smeed’s conclusions. However,Oppe’s theory in estimating the remaining growth of traffic is questionable, particularly whenwe know that many European countries are currently discussing the possibility t o stop or reduce

the increase rate of motorisation. It is uncertain whether the number of fatalities can be

predicted simply from the fitted curves or from the number of vehicle -kilometres. The question

is therefore whether this decreasing equation (7) ass umes that the fatality rate reduces to zero in

the end or not, and in this case what is the predicted year for one particular country according to

its current level of mobility? Besides, what will happen to the expected number of fatalities if

the country’s trend becomes fully motorised to 100%.

Adams (1987) has stated a similar relation between fatalities (F) and vehicle kilometres

(V), which was presented: Log (F/V) = a + b*y where y = year – 1985. Broughton (1988) has

tested this logarithmic model on data from Britain between 1950 and 1985 and the results fitted

well. In the same study Broughton applied the same model to data from four western countries:

U.S.A (1943-85), West Germany (1965-85), Norway (1947-85) and New Zealand (1948-83).

He found that this model describes the data pretty well.

(Broughton 1991) and (Oppe, 2001a) they developed another technique, the ‘singular valuedecomposition method’, in comparing road safety trends between different countries. Thistechnique investigates the similarit ies and dissimilarities between different groups of countries

regarding fatality trend. They compared various time series of data of countries jointly to

investigate the correlation between these series. This technique is useful in classifying the

countries that are similar (accidents patterns) to each other.

The more detailed time series data have led to advanced and sophisticated ways of fitting a

curve to data, especially with the current use of computer packages. For example, auto -

regressive integrated moving average (ARIMA) techniques are used to fit and forecast the time

series that are changing fairly quickly. ARIMA models should be stationary; otherwise we need

to transform the data to make them stationary. The first part of the model is the auto -regressive

(AR). This means that the Y factor is a relation of past values of Y. The second part is the

moving average (MA). This means that the Y factor is a function of past values of the errors;

see Frits et al. (2001). For instance Scott (1986) has appli ed this method to model the accidents

in England (seasonal and annual data). Oppe (2001b) has applied this method to a model that

predicts the accident data from Poland (1980 -2010).


IV. THE THIRD GENERATION: THE NEED FOR INCREASED INTEGRATION WITHMANY VARIABLES INVOLVED

Most early development efforts for international road safety development have focused on

one or a few indicators by means of risk indicators (accident rates), which are few and isolated.

The third generation has realised the need for increa sed integration between product (accident

rates) and other indicators in the same benchmarking model.

Page (2001) has compared safety situations and trends in the OECD countries from 1980 to

1994. He developed a statistical model using pooling cross -sectional time series. The model

gives a rough estimate of the safety performance of a country regarding some variables such as:

population levels, vehicle fleet per capita, percentage of young people, and alcohol

consumption. Based on this model, countries that are showing the best levels are Sweden, the

Netherlands and Norway.

Bester (2001) has developed a model by means of stepwise regression analysis. The criteria

will indicate the variables that should be added or removed in the model. The study used

collected data from different international sources and the variables used are: national

infrastructure and socio-economic factors (e.g. GDP per capita).

(Elvik & Vaa, 2004) has used techniques for evaluating the effectiveness of various road

safety measures (output) in different countries by using what is called the “before and afterstudy” evaluation technique. Similar techniques might be used to show the effectiveness of roadsafety measures that countries have taken.

(Asp & Rumar, 2001) developed the ‘Road Sa fety Profile (RSP)’. It includes all possiblequantitative and qualitative variables that may have been important in describing, explaining

and comparing road safety situations in different countries. This technique illustrates the

development in a country over time in a quick and easy illustration. RSP uses both types of

quantitative and qualitative indicators. The quantitative data obtained from international

sources. The qualitative indicators are derived from a survey of questionnaires to experts in eac h

country. Respondents’ countries were asked to answer questions regarding key road safetyissues. The answers are used to measure the RSP level for each country.

The countries were divided into three different groups of motorisation (low, medium and

high). The RSP technique includes more than 20 direct and indirect road safety indicators. Each

indicator is normalised on a scale from +2 to –2. Then the results are illustrated as a profile (see

Figure 9). This made the comparisons between countries simpler a nd easier. The Road Safety

Profile was seen as a successful tool for identifying the problems in the country where actions

are needed.


Level of Motorisation: Low Country 1 Country 2

Personal RiskTraffic Risk

Road safety statistics

Road safety trend

Road safety R&D

Road safety organisation

Road safety program

Road safety legislation

Traffic police

Driver education

Alcohol in traffic

Speed

Seat belts

Road standard

Paved roads

Road expenditure per total

Etc.......

-2

Country 3

Direct safety measures

Indirect safety measures

0 2 -2 0 2 -2 0 -2

1

2

2

0

2

-2

2

-1

1

1-2

1

2

1

0

2

0

-2

-2

0

2

0

1

1

0

0-1

-2

0

1

N.A

-1

2

-2

0

2

1

-1

1

-2

0

-1

0

0

0

-1-2

-2

N.A

-2

2

0

Figure 9. Illustration of Road Safety Profiles (Asp & Rumar, 2001)

The Globesafe database (Asp, 2004) is pre sently being constructed by means of IT andInternet. It facilitates the illustration of Road Safety Profile across countries.

V. THE FOURTH GENERATION: LINKING PRODUCT, PRACTICES AND STRATEGICBENCHMARKING- RSDI AS AN EXAMPLE

The latest generation realised the necessity of having a systematic way to add up all the

potential indicators of human, vehicle, road, environment, and regulation combined with

weights into one index. This will give a broad picture of benchmarking and not focus on one or

few particular aspects.

RSDI, as an example, combines all three types of benchmarking together: Product,

Practices and Strategic Benchmarking. This will be useful to tell success from failure in a

country. RSDI is capable to compare the road safety level and progres s across a large number of

countries and regions worldwide

Each benchmarking type is a sum of indicators and dimensions. There will be as balanced

and important indicators within each dimension as available of data as possible.

The following figure shows the main components involved in RSDI where each component

comprises a number of indicators:


Personal risk"deaths perpopulation"

Socioeconomicperformance

Organisationalperformance

Enforcementperformance

ProductBenchmarking



RSDI

Safer road users"behaviour"

Safer roadsperformance

Safer vehiclesperformance

Figure 10. RSDI conceptual framework (overall road safety performance)

The major steps used in the process of constructing RSDI are the following (Al -Haji,2005):

Finding the key indicators and dimension,

Normalising (standardising) the indicators,

Weighting the indicators,

Combining the chosen indicators into (RSDI) by using different techniques,

Applying RSDI for a sample of countries and performing an a nalysis of the results, andfinally

Testing the uncertainty and comparing the methods used with the obtained results.

The composite index of RSDI takes the form:

n

1i iw

iXn

1i iwRSDI

Where: Xi: normalised indicators for country i

wi: the weights of the Xi

n number of dimensions

The weights ranged from 0 to 1 and the sum of weights is one.

The RSDI ranged from 0 till 100. The higher values indicate a higher level of safety in thecountry. The lower values indicate the worst performance in country in term s of road safetylevel and vice versa. The target value of RSDI is 100 and it shows how far the country has to bedeveloped to provide safer roads.


VI. CONCLUSIONSInternational benchmarking models in road safety are in the interest of most countries and

international bodies since they will show the scale of the problem. This paper has reviewed inbrief the development of the benchmarking models and how they have been used.

Road safety is a complex issue and there are a high number of performance indicator s thatcan be used for benchmarking. However few of these indicators were generally included in theearlier modelling process (generation one and two).

There are mainly three types of benchmarking: product of safety, practices, and strategicbenchmarking. They differ depending on the type of indicators, which the models are trying tocompare. The literature review has shown the importance of having a large number of factorsinvolved in the benchmarking model. The road safety level in a country is a result of the wholedevelopment in society (e.g. health, education, enforcement, engineering, etc.).

The RSDI seems to provide a broader picture compared to the traditional early models inroad safety.

Today computers are developing rapidly, which simplifies th e work and analysis of cross-sectional data and time series data, which was not available before (e.g. to Smeed in 1949). Thisdevelopment has made the work in the third and fourth generations become easier and closer to reality.

AcknowledgementThe authors would like to express their appreciation to the projects work and teams of

Traffic Safety and Environment, TechTrnas project (Developing E -learning Applications andCourses in Road Safety to Russian Universities) at the Department of Science and Technol ogy,Linköping University in Sweden and State Technical University -MADI, Moscow, Russia.

References[1]. Adams, J., (1987). Smeed's law: some further thoughts. Traffic Engineering and Control 10 (7), pp. 70 -73.

[2]. Al-Haji, G., (2005). Towards a Road Sa fety Development Index (RSDI) - Development of anInternational Index to Measure Road Safety Performance. Licentiate Thesis No. 1174, ISBN 91 -85299-70-7, Linköping University, Sweden, pp. 15 -28.

[3]. Al-Haji, G., (2001). Traffic Safety in Developing Countri es- New Approaches in TechnologyTransfer by Using Distance Education Technique. Master’s thesis, LITH -ITN-EX-2001:156-SE,Linköping University, Sweden, pp. 7 -23.

[4]. Andreassen, D.C., (1985). Linking deaths with vehicles and population. Traffic Engineeri ng andControl, Vol. 26, pp. 547–549.

[5]. Asp, K., (2004). Globsafe database- A database for road safety analysis. Some parts of this database isavailable on-line: http://www.globesafe.org/ (last visited 2005 -05-16).

[6]. Asp, K., Rumar K., (2001). The Road Safety Profile. The RetsNet News “Regional Traffic Safety Network. No. 1”.[7]. Bester, C.J., (2001). Explaining national road fatalities. Accident Analysis and Prevention. Vol. 33, pp. 663 –672.

[8]. Broughton, J., (1991). Forecasting road accident cas ualties in Great Britain. Accident Analysis andPrevention 23 (5), pp. 353–362.

[9]. Broughton, J., (1988). Predictive Models of Road Accident Fatalities. Traffic Engineering andControl, May 1988, ISSN:0041-0683, pp. 296-300.

[10]. Elvik, R., Vaa, T., (2004). The Handbook of Road Safety Measures. Elsevier Amsterdam, ISBN: 0 -08-044091-6, pp.66 & pp. 676-803.

[11]. Evans, L., (1991). Traffic Safety and the Driver, New York: Van Nostrand Reinhold, P25 -43, pp. 60-95.

http://www.globesafe.org/


[12]. Fieldwick, R. and Brown R.J . (1987). The effect of speed limits on road casualties. TrafficEngineering and Control, Vol. 28, pp 635 -640.

[13]. Haddon, W. (1972). A logical framework for categorizing highway safety phenomena and activity.The Journal of Trauma, 12, pp. 193 -207

[14]. Haight, F., (1983). Traffic Safety in Developing Countries, Journal of Safety Research, Vol. 14, No.1, pp. 1-12.

[15]. Hakim, S., Shefer, D., Hakkert, A.S., Hocherman, I . (1991). A Critical Review of Macro Models forRoad Accidents. Accident Analysis & Prevention, Vol. 23, No. 5, pp. 379-400.

[16]. Jacobs, G.D., (1982). The potential for road accident reduction in developing countries. TransportReviews 2, pp. 213–224.

[17]. Jacobs, G.D., Fouracre, P.R ., (1977). Further research on road accident rate in developingcountries. TRRL report LR 270. Transport and Road Research Laboratory, Crowthorne, Berkshire.

[18]. Jacobs, G.D., Hutchinson, P., (1973). A study of accident rates in developing countries. TRRLreport LR 546. Transport and Road Research Laboratory, Crowtho rne, Berkshire.

[19]. Koornstra, M.J., (1992). The evolution of road safety and mobility. IATSS (InternationalAssociation of Traffic and Safety Sciences), Research, Vol.16, No.2, pp. 129 -148.

[20]. Koornstra, M.J., Oppe, S . (1992). Predictions of road saf ety in industrialised countries and EasternEurope; an analysis based on models for time series of fatality rates and motorised vehicle kilometres oramounts of passenger cars. Proceedings of International Conference ‘Road Safety in Europe’, VTI,Linköping, Sweden, pp.49-70.

[21]. Mekky, A., (1985). Effect of road increase in road motorisation levels on road fatality rates in somerich developing countries. Accident Analysis and Prevention 17 2, pp. 101 –109.

[22]. Minter, A. L., (1987). Road casualties- Improvements by Learning Processes. Journal of the TrafficEngineering and Control, Feb. 1987, pp. 74 -79.

[23]. Navin, F., Bergan, A., Qi , J., (1994) A Fundamental Relationship for Roadway Safety: A Model for GlobalComparisons. Transportation Research Board, Transportation Research Record 1441, Washington D.C., pp. 53 -60.

[24]. OECD ‘Organisation For Economic Cooperation and Development’, (1997). Road SafetyPrinciples and Models: Review Of Descriptive, Predictive, Risk and Accident Consequence Models.OCDE/GD(97)153, Paris.

[25]. Oppe, S., (2001b). Traffic Safety Development in Poland. SWOV, Leidschendam, Netherlands.

[26]. Oppe S., (2001a). International comparisons of road safety using Singular Value Decomposition.Leidschendam, SWOV Institute for Road Saf ety Research, The Netherlands, pp. 8 -18.

[27]. Oppe, S., (1991a). The Development of Traffic and Traffic Safety in Six Developed Countries.Accident Analysis and Prevention, Vol.23, No.5, pp. 401 -412.

[28]. Oppe S., (1991b). Development of traffic and traf fic safety: global trends and incidental fluctuations.Accident Analysis and Prevention; Vol.23: pp. 413 –22.

[29]. Oppe S., (1989). Macroscopic models for traffic and traffic safety. Accident Analysis andPrevention, Vol.21, pp. 225-232.

[30]. Page,Y., (2001). A statistical model to compare road mortality in OECD countries. AccidentAnalysis & Prevention, 33, pp. 371–385.

[31]. Rumar K. (1999). Transport safety visions, targets and strategies: beyond 2000. First EuropeanTransport Safety lecture, Brussels, European Transport Safety Council.

[32]. Smeed, R.J., (1949). Some statistical aspects of road safety research. Journal of Royal StatisticalSociety Series A 112, pp. 1–34.

[33]. Timo, E., (1998). Risks, exposures and accident data, VTI Conference 9A, part 6, nr 9A:6, s 85-96,

Linköping, Sweden


I. INTRODUCTION

Much attention has been focused onlateral performance of railway vehicles for along time in order to achieve higher speed andbetter comfort. But the longitudinal dynamicperformance was neglected except for thestudy on traction and braking performance.Since it has nothing to do with lateralperformance, forward velocity was normallyconsidered uniform. The results obtained inthis paper show that it may be an importantcause of wheel tread spalling and rail

corrugation with certain wavelength.

There are many puzzled enigma exist ingin rail vehicles, such as the rail vehicle mayhave some tremble in the course of speedup orat a not high speed; the ride index of railvehicle in some low speed might be worsethan that in the higher speed; the wheel treadspalls and the track wave wears with somewavelength; the vibration of rail vehicle wasquite bigger on some tracks. Thesephenomena indicate that the performance ofrail vehicle has some relation with the trackirregularity and the longitudinal dynamic

INFLUENCE OF TRACK IRREGULARITY ON ONGITUDINALVIBRATION OF WHEELSET AND CORRELATION

PERFORMANCE

WEIHUA MA, SHIHUI LUOTraction Power State Key Laboratory,Southwest Jiaotong University, Chengdu 610031, ChinaRONG-RONG SONGCollege of computer science andtechnology, Southwest Universityfor nationalities,ChengduSichuan 610041. China

Abstract: A longitudinal vibration of wheelset with respect to bogie frame often exist s with a

high acceleration magnitude and relative high frequency. At first a simplified model with a single

wheelset moving at a constant speed on a tangential track with irregularity is used to investigate

the longitudinal vibration dynamics. Results of the longitudinal vibration study indicate that the

longitudinal vibration frequency of the wheelset is most sensitive to the primary longitudinal

stiffness and the mass of the wheelset. As to the locomotive model, the longitudinal vibration was

concerned with cross-level irregularity and vertical profile irregularity. A method to estimate the

resonance speed is presented. Finally, the paper shows a possible solution to extend wheel -rail

service life by eliminating longitudinal vibration of the wheelset. The solution is simply arranging

the primary vertical damper with a forward angle, so that its damping component can be applied to

longitudinal direction.

Keywords: Track irregularity; Acceleration; Longitudinal vibration; Dynamics


behavior of the wheelset. In this paper, weinvestigate the relation between longitudinalvibration and the track irregularity from thelongitudinal dynamic point of view.

Lots of researches have been carried outin the field of wheel-rail contact fatiguemechanism in last ten years [1, 2, 3, 4].Wheel-rail wear is a necessary expense ofrailway transportation because of adhesion incontact patch between track and vehiclesystem. Total wear between wheel and rail isdue to stress generated by static, dynamic andthermal loadings [6]. The best way to reducewheel-rail wear is to reduce the unnecessarystress, which is caused by dynamic interactionbetween track and vehicle due to trackirregularity and vehicle vibration. But it isdifficult to determine which one is the mainfactor. According to the following analysis,wheelset longitudinal dynamic behavior maybe a possible source of the severe wheel-raildynamic interaction.

Since track irregularity is one of the mainreasons of rail vehicles vibration [13], a largenumber of researches have been made on theresponse of rail vehicles to the trackirregularity [10, 11]. They all aimed at thevertical or lateral vibration of rail vehicles, notreferred to longitudinal vibration of thewheelset. In this paper, we analyse theinfluence of tack irregularity on longitudinalvibrations.

II. THE TRACK IRREGULARITY

There always exists track irregularity inthe realistic track from the track setup and theinteraction wear between the railway vehiclesand the track. Researches show that the trackirregularity is a random process [11]. And it’sdifficult to decide what kind of the track

spectrum should be adopted in the vehicledynamic evaluation in our country.

Usually the spectrum density used todescribe track irregularity [5] can beexpressed as:

6ω3a4ω2a

2ω1a0a

6ω3b4ω2b

2ω1b0bS(ω(

(1)

In which, )3,2,1,0i(b,a ii is the

constant coefficient of the angular frequencyof space .

The American Federal Rail wayManagement Bureau (FRA) gets the trackspectrum according to a great deal ofmeasured data. It is divided into 6 Grades,among which Grade 1 is the worst and Grade6 is the best. Germany has formulated thespectrum of the high-speed trunk with lowinterference and high-speed trunk with highinterference. Our country has not formulatedthe standard track spectrum yet, but thedepartment concerned has carried out a fewresearch works, and offered some expressionsfor track spectrum according to the measureddata. But the acquired data in our country’sresearch is so few that it is unable to representthe statistical characteristic of trackirregularity in our country [11], so we analysethe problem by using the Germany trackirregularity.

According to the accumulativeexperience, aiming to the speedup track in ourcountry, we consider the vertical profileirregularity of the medium track iscorresponding to the vertical profileirregularity of Germany high-speed trunk withlow interference; the alignment irregularity ofthe medium track is worse than the alignmentirregularity of Germany high-speed trunk with


high interference; the cross-level irregularityof the medium track is correspond ing to thecross-level irregularity of Germany high-speed trunk with low interference. To make asimple and proper calculation , we use themedium track irregularity in this paper.

Time-domain excitation is used both inthe locomotive model and simpl ified model. Itis calculated by a polynomial express ion asfollowing:

2ω2aω1a0a

ω1b0bF(ω(

(2)

The coefficients of excitation used in themodel and the parallels with Germany high-speed trunk with high interference are listed intable 1

Table 1. Coefficients in polynomial expressiona0 a1 a2 b0 b1

Germanyalignment

0.016987 0.8452 1 0.001144 0

Germanyvertical

0.016987 0.8452 1 0.001519 0

Germanycross-level

0.00744 0.38718 1.283 0 0.00203

Modelused

alignment0.034 1.6904 1 0.0057 0

Modelused

vertical0.016987 0.8452 1 0.00093 0

Modelused

cross-level

0.00744 0.0774 1.283 0 0.0015

III. CALCULATION MODEL

3.1. Simplified model

The longitudinal vibration of wheelsetwas found in the locomotive model. It isnecessary but much time-consuming to makemore analysis for all kinds of railway vehiclesto study the phenomenon. In order to provethe result obtained by the locomotive modeland find the mechanism of longitudinalvibration, a simplified model with just onewheelset and minimum freedoms is set up

(see Fig.1).

Giv

en m

ass

v

Forwardvelocity

Cpx

Kpx,kpy

Wheelset

Rail

Irregularity

y x

z

Kpx-primary longitudinal stiffness; Kpy-the primary lateral stiffness; Cpx-primarylongitudinal damper;

Fig.1. The Simplified model with one wheelset

In the simplified model, mass ofwheelset: 2500 kg;

Roll mass moment of inertia of wheelset:500 kg/m2;

Pitch mass moment of inertia ofwheelset: 100 kg/m2;

Yaw mass moment of inertia of wheelset:500 kg/m2;

Mass of given mass: 19500 kg;

Degree of freedom of wheelset: 6;

Degree of freedom of given mass: 1;

Kpx: 1.2×107 N/m;

Kpy: 6.0×106 N/m;

Cpx: 1000 N.s/m.

Rail: 60kg/m rail;

Track gauge: 1435mm;

Rail cant: 1/40;

The lateral clearance between wheel andrail: 14 mm;

Tread profile: JM3 profile;

Wheelset

Irregularity

Rail


In the model, the friction coefficient μbetween wheel and rail is supposed to beconstant, and Kalker’s simplification theory(FASTSIM) is adopted to calculate the wheel -rail creepage-creep force. The equivalentconicity of the wheel tread within the scope of6 mm is about 0.12, and is about 0.2 withinthe scope of 6-8 mm, and will be greater than0.25 beyond the scope of 8 mm.

3.2. Locomotive Model

The analysed locomotive is designed tomeet the operation requirements of 200km/h.In order to reduce the axle load to 22 t, atrailer wheelset is considered besides twodriven wheelsets in a bogie (Fig.2). Eachdriven wheelset has a set of driving unit thatconsists of one traction motor with a gearboxconnected it rigidly. The driving unit can beconsidered mounted on bogie frame rigidly ormounted on it elastically in lateral direction inmodel.

CSYCSX

Trac

tion

Unit

KSZ

Trac

tion

Unit

TR

BOGIE

DW TW DW

TR

KSZ

CSY

CSX

DW—driven wheelset; TW—trailer wheelset;CSX —secondary suspension longitudinal damper;CSY—secondary suspension lateral damper;KSZ—secondary suspension vertical stiffness;TR—traction rod.

Fig.2. Model of A-1-A bogie

The model consists of car body, 2 bogies,

4 traction motors and 6 wheelsets. The

longitudinal damper and the lateral damper

are arranged in secondary suspension between

each bogie and the car body. Four coil springs

are arranged on each side of bogie as

secondary spring suspension to keep good

lateral running performance. The longitudinal

force will be transferred from bogie frame to

car body through a four-linking-rod

mechanism. It causes a rigid connecti on

between car body and front and rear bogies in

longitudinal direction. The primary

suspension is considered as a compact force

element at each axle box with st iffness in

three directions and vertical damping. T o

improve the speed of the calculation, t he axle

box itself will not be considered as a body

element, and it doesn’t influence the precision

of the result.

VI. ANALYSIS OF SIMPLIFIED MODEL

4.1. Longitudinal dynamic performance ofwheelset

In order to study longitudinal dynamic

performance in principal, an analysis is

conducted for the simplified model shown in

Fig.1. Longitudinal resonance vibration is

revealed on the track with irregularity at the

speed of 20km/h, and the longitudinal

vibration acceleration is quite large at the

frequency of 10.6Hz (see Fig.3). The

longitudinal creepage and adhesion coefficient

are shown in Fig.4. In Figs.3 and 4, we can

see that the longitudinal vibration resonance is

obvious and the adhesion coefficient always

approaches the presumed saturation value

0.25. So we can say that longitudinal

resonance vibration of the wheelset really

exits in the simplified model.

DWTWDW

BOGIE

Tra

ctio

n U

nit

Tra

ctio

n U

nit


0 5 10 15 20-150

-100

-50

0

50

100

150

Long

itudi

nal A

cc [

m/s2 ]

Time [ s ]

(a)In the time domain

0 5 10 15 20 2505

101520253035

Long

itudi

nal A

cc [

m/s2 ]

Frequency [ Hz ]

(b) In the frequency domain

Fig 3. Longitudinal acceleration in the simplified

model at

0 5 10 15 20-0.75

-0.50

-0.25

0.00

0.25

0.50

0.75

Long

itudi

nal C

reep

age

Time [ s ]

(a) Longitudinal creepage

0 5 10 15 200.00

0.05

0.10

0.15

0.20

0.25

0.30

Adhe

sion

Coef

ficien

t

Time [ s ]

(b) adhesion coefficient

Fig 4. Longitudinal creepage and adhesioncoefficient in the simplified model at 20km/h

4.2. Root loci analysis

10.62

yaw displacement

lateral displacement

longitudinalvibration

- 1. 25 - 1. 00 - 0. 75 - 0. 50 - 0. 25 0. 00 0. 25

40. 0

35. 0

30. 0

25. 0

20. 0

15. 0

10. 0

5. 0

0. 0

Root loci

Natural damping

Freq

uenc

y [H

z]

Fig 5. Root loci analysis of the simplified model

To find the inherent frequency of thesimplified model, the linear analysis of thesimplified model was carried out, and theresults are shown in Fig.5. The speed isranged from 1 km/h to 401 km/h and theincrement is 5 km/h. We can clearly see thatthe longitudinal vibration frequency nearlyunchanged with the speed, but the lateraldisplacement frequency and yaw displacementfrequency of the wheelset changed with thespeed. From the speed of 1km/h to 401km/h,the frequency of the wheelset longitudinalvibration is nearly 10.62Hz.

In the simplified model, the inherentfrequency is related to the system, and notvaried with the speed. But the frequency willchange according to the mass of the wheelsetand the primary longitudinal stiffness. Theinherent frequency could be estimated by thefollowing expression:

m

2

2

1f L

xk

(3)

Where m is the mass of wheelset and kxis the longitudinal stiffness of each axle box.Let kx = 1.2×107N/m, m = 5000kg, then


Hz03.11

5000

102.12

2

1f

7

L

If this vibration cannot be dampedeffectively, the rail corrugation with the

wavelength is developed on rail, and the

wavelength is:

m5.003.11

6.3/20

f

v

L

.

As the longitudinal resonance vibrationonly occurs at a given speed, so the

wavelength doesn’t increase with theincrease of the speed. The possible cause ofrail corrugation has been discussed in manypapers [8]. Here a new possible solution isproposed on the formation of rail corrugationwith certain wavelength from the longitudinaldynamic point of view.

The longitudinal vibration resonance ofwheelset doesn’t occur at other speeds. Forexample, at the speed of 10km/h or 30km/h,the longitudinal vibration resonance ofwheelset doesn’t happen (see Fig.6). So wecan draw the conclusion that the longitudinalvibration resonance is corresponding to thespeed; higher or lower than 20km/h thelongitudinal vibration resonance does nothappen in the simplified model, although thelongitudinal vibration is still very high.

0 5 10 15 20-15

-10

-5

0

5

10

15

Long

itudi

nal A

cc [

m/s

2 ]

T im e [ s ]

10 km /h

(a)In the time domain at 10 km/h

0 5 10 15 20 250.0

0.4

0.8

1.2

1.6

2.0 10 km/h

Long

itudin

al Ac

c [m

/s2 ]

Frequency [ Hz ]

(b)In the frequency domain at 10 km/h

0 5 10 15 20-40

-20

0

20

40

Long

itudin

al Ac

c [m

/s2 ]

Time [ s ]

30 km/h

(c)In the time domain at 30km/h

0 5 10 15 20 250.0

0.5

1.0

1.5

2.0

2.5

3.0 30 km/h

Long

itudin

al Ac

c [m

/s2 ]

Frequency [ Hz ]

(d)In the frequency domain at 30km/h

Fig 6. Longitudinal vibrations of wheelset under

different speeds

4.3. The speed of longitudinal vibrationresonance

Several frequencies can be recognized in

Figs. 3 and 6. A fixed frequency of about

10.6Hz is corresponding to longitudinalvibration the wheelset with respect to bogie

frame. A velocity depending on frequency is

the rolling angular velocity of the wheelset .


With the forward speed increasing from

10km/h to 20km/h and 30km/h, this valuevaried from 5.56rad/s to 11.1rad/s and

16.66rad/s respectively. Suppose nominal

radius of the wheelset is 500mm, then the

value of 2.0rad/s equals 1cycle/s of wheelrolling. So above three angular velocities are

corresponding to 2.78Hz, 5.55 Hz and 8.33

Hz of wheel rolling frequency. Dividing themby the frequency of 10.6Hz, we can get the

results of 3.81, 1.91 and 1.29. Thus a ratio is summarized as:

w

L

f

f , (4)

where fL is longitudinal vibration

frequency of wheelset according to bogieframe; fw is rolling frequency of wheelset

depending on forward speed and nominal

radius, which is expressed as:

s/circleR2

6.3/vf

ww (5)

As the vibration resonance only occurs

when the two vibration frequencies are closeor when one is an integer multiple of the

other, so a resonance vibration of wheelset in

longitudinal direction tends to be i nducedwhen the vibration frequency is an integer

multiple of wheel rolling speed, which means is an integer.

From Eqs.3-5, we can get :

NL

www

fR2.7R2f6.3v (6)

According to Eq.6, the resonance forward

speed for the simplified model is:

km/h12.192

62.105.02.7

fR2.7v L

w N

Suppose that the nominal radius of wheel

is 500 mm. The resonance vibration occurs in

20km/h considering influence of other factors

such as adhesion.

To further learn the influence of mass of

the wheelset and the primary longitudinal

stiffness on the longitudinal resonance

vibration of the wheelset, we carried out the

calculation of the speed at which resonance

vibration occurs with different mass and

primary longitudinal stiffness in the simplified

model. The results with changing mass and

constant primary stiffness are shown in table 2

and the results with changing primary

longitudinal stiffness and constant mass are

shown in table 3. In table 2-3, f1 means the

inherent frequency of the wheelset attained

through the root loci analysis; f2 means the

inherent frequency calculated through Eq.3;

v1 means the speed of longitudinal resonance

vibration calculated by incorporating f1 into

Eq.6; v2 means resonance speed calculated by

incorporating f2 into Eq.6; V means the

resonance speed gained by the simulation

calculation of the simplified model.

Table 2. The resonance speed gained by differentmethods with different wheelset mass

Mass

(kg)f1(Hz) f2(Hz)

v1

(km/h)

v2

(km/h)

V

(km/h)

1000 20.86 24.66 37.54 44.38 45

2000 15.93 17.43 28.67 31.38 32

3000 13.38 14.24 24.08 25.62 26

4000 11.76 12.33 21.17 22.19 23

5000 10.62 11.03 19.11 19.85 20

6000 9.75 10.07 17.55 18.12 19

7000 9.07 9.32 16.32 16.77 17

8000 8.52 8.72 15.32 15.69 16


Table 3. The resonance speed gained by differentmethods with different primary

longitudinal stiffness

Kpx

(N/m)f1(Hz) f2(Hz)

v1

(km/h)

v2

(km/h)

V

(km/h)

1.0e7 9.69 10.07 17.44 18.12 18

1.1e7 10.16 10.56 18.30 19.00 19

1.2e7 10.62 11.03 19.11 19.85 20

1.3e7 11.04 11.48 19.89 20.66 21

1.4e7 11.47 11.91 20.64 21.44 22

1.5e7 11.87 12.33 21.36 22.20 23

1.6e7 12.26 12.73 22.07 22.92 23

1.7e7 12.64 13.12 22.74 23.62 24

Through table 2 and table 3, we can seethat the softer the stiffness or the lighter thewheelset, the larger the difference between theapproximated value (v1,v2) and the practicalvalue V.

As the results of v2 and v are very close,so we can use Eqs.3 and 6 to quickly calculatethe resonance speed, and the result is quiteclose to that of the multi-body dynamiccalculation. If we replace the track irregularityused in the model with the track irregularity ofGermany high-speed trunk with highinterference or Germany high-speed trunkwith low interference, we can see that thelongitudinal resonance vibration at the sp eedof 20km/h, and the difference betweenresonance vibrations on different trunk is onlythe amplitude of the vibration.

For the simplified model that just has onewheelset, the longitudinal resonance vibrationwill happen on the smooth, level and tangen ttrack without irregularity as well. But theamplitude is smaller compared with theamplitude with track irregularity. It also

indicates that if the wheelset has occurredlongitudinal vibration resonance, thelongitudinal vibrations will not converge at aconstant speed.

V. ANALYSIS OF LOCOMOTIVE MODEL


The root loci analysis of the locomotivemodel is shown in Fig.7. The movement ofwheelset in front and rear bogies is out ofphase at the frequency of 19.2Hz, and theirinteraction makes this vibration not shifted tothe car body; whereas the movement ofwheelsets in front and rear bogies is in phaseat the frequency of 20.6Hz, and theirinteractions on car body is overlapping, whichleads to a strong longitudinal vibration of carbody. Moreover, as the gravity center of bogieframe doesn’t coincide with geometricalcenter, the longitudinal forces from axle boxesto bogie frames will cause nodding movementon both bogie frames, and a vertical excitationdue to the longitudinal force i s thus producedon both ends of car body which leads to astrongly nodding vibration of car body. Thefrequencies of the discussed model will notchange under an uniform running speed.

200km/h~400km/h, 41 points

longi. vib. of wheelsets w.r.tbogie frame in phase

longi. vib. of wheelsets w.r.tbogie frame against phase

-1.00 -0.75 -0.50 -0.25 0.00

0.00

10.0

20.0

30.0

40.0

50.0Root locii

Natural damping

Freq

uenc

y [H

z]

Fig 7. Root loci analysis of the locomotive model

Natural damping

Freq

uenc

y [H

z]


Substitute kx=2.0×107N/m; m=2500kg

into Eq.3

Hzf L 13.20

2500

1022

2

1 7

Suppose the nominal radius of wheelset

is 525mm. Then resonance speed for the

locomotive model is:

v=3.6× (2Rwfw) = 3.6 × [2Rw × (fL/N)] =

3.6 × [2× 0.525 × (20.6/1)]=77.87km/h

Further research showed that in the

locomotive model the nominal mass used to

calculate the inherent frequency should be

slightly smaller than the wheelset mass; and

the nominal stiffness used to calculate the

inherent frequency should be slightly bigger

than the primary longitudinal stiffness. In the

locomotive model, the wheelset longitudinal

inherent frequency is calculated just by the

mass of wheelset and the primary suspension

longitudinal stiffness, without consider ation

on the influence of the second suspension. So

the actual frequency would be bigger than that

calculated by Eq.3, and the actual speed

which is 100km/h would bigger than that

calculated by Eq.6.

5.2. Wheel-Rail Longitudinal Interaction

Vibration corresponding to above

eigenvalues is actually a small local rol ling

vibration of wheelset with respect to bogie

frame, either in phase or out of phase (see

Fig.8). Because of the track irregularit y, a

dynamic component is added to the nominal

forward speed for wheelset.

lo ca l ro lling v ib ra tion o f w h ee lse t

Fig 8. Local rolling vibration of wheelset

within a bogie

That means the reference velocity, whichis often considered constant for creepagecalculation is a variable. According to thedefinition of longitudinal creepage [6], it isexpressed as:

v/)cv(x (7)

where v is reference velocity and c is thecircumferential velocity of wheel at contactpoint. There are different definitions ofreference velocity [7], which lead to differentcalculation results of longitudinal creepage.

0 5 10 15 2027.2

27.4

27.6

27.8

28.0

28.2

28.4

Refe

renc

e Ve

locity

[m

/s]

Time [ s ]

(a) Reference velocity

0 5 10 15 2027.027.227.427.627.828.028.228.4

Circ

umfe

rent

ial V

eloc

ity [

m/s

]

Time [ s ]

(b) Circumferential velocity

Fig 9. Reference velocity and circumferential

velocity at contact point at 100km/h


For the analysed locomotive model , the

reference velocity and circumferential

velocity at contact point are given in Fig.9.

The Longitudinal creepage and adhesion

coefficent of wheelset are given in Fig.10.

0 5 10 15 20-5.0

-2.5

0.0

2.5

5.0

7.5

Long

itudin

al Cr

eepa

ge [

10-3]

Time [ s ]


0 5 10 15 200.00

0.05

0.10

0.15

0.20

0.25

0.30

Adhe

sion C

oeffic

ient

Time [ s ]

(b) Adhesion coefficient

Fig 10. Longitudinal creepage and adhesion

coefficient of wheelset

Longitudinal creepage of leading wheel

at contact patch for example is thus obtained

according to physical measurements in Figs.7

and 10, and at several points maximum

adhesion coefficient reaches saturation value

0.25, see figure 10(b).

5.3. Effect of wheelset longitudinal

vibration on carbody vertical vibration

0 5 10 15 20-7.5

-5.0

-2.5

0.0

2.5

5.0

7.5

Long

itudin

al Ac

c [m

/s2 ]

Time [ s ]

(a) In the time domain

0 5 10 15 20 250.0

0.5

1.0

1.5

2.0

2.5

3.0

Long

itudin

al Ac

c [m

/s2 ]

Frequency [ Hz ]


Figure 11. Time history and frequency response

of longitudinal vibration acceleration of car body

The time history and the frequency

response of longitudinal vibration acceleration

of the car body were shown in Fig.11, in

which the longitudinal vibration acceleration

of the locomotive is 7.5m/s2, and the main

frequency is 20.6Hz.


0 5 10 15 20-75

-50

-25

0

25

50

75

Long

itudi

nal A

cc [

m/s

2 ]

Time [ s ]


0 5 10 15 20 250

5

10

15

20

Long

itudin

al Ac

c [m

/s2 ]

Frequency [ Hz ]


Fig 12. Time history and frequency response of

longitudinal vibration acceleration of wheelset

The time history and the frequencyresponse of longitudinal vibration accelerationof the first wheelset were shown in Fig.12, inwhich the amplitude of the longitudinalacceleration reached 70m/s2. It has two mainfrequencies that are very close and evenintense-coupling. The vibration whichfrequency is 20.6Hz is transferred to car body.

The maximum frequency of thelongitudinal vibration of car body equals thatof the wheelset. The vibration will not betransferred to the car body when thevibrations of two wheelsets in the same bogieare out of phase, as the vibrations of the twowheelsets within a bogie will be counteractedby the interaction. When the vibrations of thetwo wheelsets in the same bogie are in phase,it will cause high frequency vibration, which

would be transferred to car body, and producethe longitudinal oscillation and nod oscillationof the car body. From Figs.11and12, we cansee that the effect of wheelset longitudinalvibration on locomotive longitudinal vibrationis very large.

The time history and frequency respons eof vertical acceleration at the end of car bodyand the middle of car body were shown inFigs.13 and 14. The concerned resultindicated that the vertical vibration of thelocomotive was very intense, and clearlylongitudinal oscillation occured. Through theroot loci analysis, we can see that longitudinaland nodding vibration occurred actually.

0 5 10 15 20-4

-2

0

2

4Ve

rtica

l Acc

[m

/s2 ]

Time [ s ]

End of carbody


0 5 10 15 20 250.00

0.25

0.50

0.75

1.00

1.25

Vertic

al Ac

c [m

/s2 ]

Frequency [ Hz ]

End of carbody


Fig 13. Time history and frequency response of vertical

vibration acceleration at the end of car body


0 5 10 15 20-0.4

-0.2

0.0

0.2

0.4

Verti

cal A

cc [

m/s

2 ]

Time [ s ]

Middle of carbody


0 5 10 15 20 250.00

0.02

0.04

0.06

Vertic

al Ac

c [m

/s2 ]

Frequency [ Hz ]

Middle of carbody


Figs 14. Time history and frequency response of

vertical vibration acceleration at the middle of car body

The spectral analysis of the longitudinalvibration could find that a forced vibrationwith the frequency of 20.6Hz was applied atthe both ends of the car body. The spectralanalysis proved that the vibration frequenciesleading to up and down and nodding of the carbody are the values of 1.67 Hz and 1.37 Hzrespectively.

Although the longitudinal vibration andvertical vibration of car body are extremelybig, the amplitude of lateral acceleration isquite small (see Fig.15). It indicates that thecalculation is stable and the longitudinalresonance vibration of the wheelset is notcaused by the instability of the calculation.

0 5 10 15 20-1.0

-0.5

0.0

0.5

1.0

1.5

Later

al Ac

c [m

/s2 ]

Time [ s ]

End of carbody


0 5 10 15 20 250.00

0.03

0.06

0.09

0.12

0.15

Later

al Ac

c [m

/s2 ]

Frequency [ Hz ]

End of carbody


Fig 15. Time history and frequency response of Lateral

acceleration at the end of car body

5.4. The influence of wheelset longitudinalvibration on wheel treads spalling

Research results show that the strongwheel-rail dynamic force is the main reason

that causes the wheel treads contact fatigue

and spalling. Since the wheel-rail contact is athree-dimensional behavior on the contact

point, the lateral wheel-rail dynamic force can

be divided into three separate dimensionalforces that are lateral force, vertical force and

longitudinal force. On the condition of

reasonable control on the lateral force andvertical force, reducing the longitudinal force

can effectively reduce the wheel -rail dynamic

force. Many factors have influences on


wheelset longitudinal vibration. A strong

vibration, especially a resonance, may causeserious wheel-rail contact wear and shorten

service life. If the vibration is suppressed, one

of sources of wear on wheel-rail contact patch

will be eliminated. It is hopeful to improve thewheelset operation condition through reducing

the longitudinal force, and solve the wheel

tread spalling to some degree. The relevanttests are presently underway.

5.5. Influence of the track irregularity onwheelset longitudinal vibration in thelocomotive model

Through carrying out the linear vehicle

system analysis adopting the Germany high -speed trunks with high interference and low

interference and the medium irregularity as

the track irregularity separately, we get thesystem response depending on the speed [14].

The speed corresponding to the vibration

frequency that approximates to the inherentfrequency is the very speed at which

resonance vibration occurs.

The resonance frequencies with the three

irregularities have slightly differentamplitudes at the speed of 100km/h. So we

can say that the calculated resonance speed is

100km/h. The difference of the locomotivemodel and the simplified model is that the

longitudinal resonance vibration of the

locomotive wheelset will not occur on thetrack without irregularity.

In this paper, vertical profile irregularity,

alignment irregularity and cross -levelirregularity are considered, and we find the

phenomenon of longitudinal resonance

vibration of the wheelset clearly. Now disposeone of the three irregularities respectively to

find the source inducing longitudinal

vibrations.

First dispose vertical profile irregularityand reserve alignment irregularity and cross -

level irregularity, the result is shown in

Fig.16. It is shown that the longitudinal

acceleration of the wheelset and verticalacceleration of the car body is quite small,


wheelset does not happen. It indicates that ifthere is not vertical profile irregularity, the


wheelset will not take place.

0 5 10 15 20-0.4-0.20.00.20.40.60.81.0 Wheelset

Long

itudin

al Ac

c [m

/s2 ]

Time [ s ]

(a) At the wheelset

0 5 10 15 20-0.03-0.02-0.010.000.010.020.030.04

Verti

cal A

cc [

m/s

2 ]

Time [ s ]

End of carbody

(b)At the end of car body

Fig 16. Longitudinal acceleration without vertical

profile irregularity

Second dispose cross-level irregularityand reserve vertical profile irregularity andalignment irregularity, the result is shown inFig.17. In this case, longitudinal acceleration


of the wheelset and vertical acceleration of thecar body is much small, it doesn’t causelongitudinal resonance vibration too . It provesthat without the cross-level irregularity, thelongitudinal resonance vibration of thewheelset will not take place, and the influenceof cross-level irregularity is much bigger thanthat of the vertical profile irregularity.

0 5 10 15 20-8-6-4-20246 Wheelset

Long

itudi

nal A

cc [

m/s

2 ]

Time [ s ]

(a) At the wheelset

0 5 10 15 20-1.0

-0.5

0.0

0.5

1.0

1.5

Vertic

al Ac

c [m

/s2 ]

Time [ s ]

End of carbody

(b) At the end of the car body

Fig 17 . Longitudinal acceleration

without cross - level irregularity

At last, dispose alignment irregularityand reserve vertical profile irregularity an dcross-level irregularity, the result is shown inFig.18. Obviously, with the vertical profileirregularity and cross-level irregularity, stronglongitudinal resonance vibration take s place

on the wheelset.

0 5 10 15 20-75

-50

-25

0

25

50

75

Long

itudi

nal A

cc [

m/s

2 ]

Time [ s ]

Wheelset

(a) At the wheelset

0 5 10 15 20-4

-2

0

2

4

Vertic

al Ac

c [m

/s2 ]

Time [ s ]

End of carbody


Fig 18 . Longitudinal acceleration without

the track alignment irregularity

The three above results show that the

longitudinal resonance vibration is the product

of interaction between vertical profileirregularity and the cross-level irregularity,

and the alignment irregularity has little

relation with the resonance. Single irregularitywill not cause longitudinal resonance

vibration of the wheelset at all.

If we replace vertical profile irregularityand cross-level irregularity with the

corresponding Germany irregularity, the


longitudinal resonance vibration will take

place at the same speed. But on a smoothtrack under the same speed, the longitudinal

vibration resonance will not taken place,

which also validate the above conclusion. As

the vertical profile irregularity and the cross-level irregularity always exis ts, the

longitudinal vibration of wheelset always

exists, and can even be developed into strongresonance within some ranges of the speed. If

the vehicle runs under the condition of

resonance vibration, wear of wheelset and railwill be very serious.

VI. A POSSIBLE SOLUTION FORLONGITUDINAL VIBRATION

The control method is reducing the

source causing the longitudinal resonance

vibration. There are two kinds of methods:

one is to improve the track quality, the

condition of the track and reduce the track

irregularity, but this method will cost a lot; the

other is to improve the locomotive

adaptability to the track irregularity, which

can be easily achieved through changing the

suspension parameter.

As the damper can reduce the vibrations,

we can apply longitudinal damp er to the

primary suspension of the locomotive.

Usually the locomotive do not have the

primary longitudinal damper and a small slant

angle is assigned to the damper in the vertical

direction.

. If the vertical damper has certain

inclinations to longitudinal direction, it will

produce a certain damping component in

longitudinal direction. Generally the slant

angle is 25°. Results for different damper

arrangements are shown in Fig.19.

0 5 10 15 20-75

-50

-25

0

25

50

75

Long

itudi

nal A

cc [

m/s2 ]

Time [ s ]

Vertical arrangement

(a) With a vertical angle

0 5 10 15 20-0.50

-0.25

0.00

0.25

Long

itudin

al Ac

c [m

/s2 ]

Time [ s ]

25 degree ahead incline

(b) With a slant angle of 25°

Fig 19. Longitudinal accelerations with different

damper arrangement

Result shows that the longitudinal

damper is very effective for reduc ing

longitudinal vibration. With the vertical

arrangement, an extremely high resonance

occurred, and its amplitude approached

70m/s2. With the slant angle of 25°, 28kNs/m

damping is produced in longitudinal direction

that eliminated resonance totally, and random

vibration with much smaller amplitude

become obvious. Comparing Fig.20 (right)

with Fig.20 (left), we can conclude that a

much longer service life of the wheel could

then be expected.


VII. CONCLUSION

Longitudinal vibration of the wheelsetwith respect to bogie frame always existswhile a vehicle operating on track. Theamplitude of the vibration is influenced bymany factors, such as vehicle structure,suspension and mass parameters, frictioncoefficient in wheel-rail contact patch andtrack irregularity, etc. In the worst case, astrong longitudinal resonance of the wheelsetmay be developed and cause severe wheel -railcontact fatigue, which lead to faster wheelspalling or rail corrugation with certainwavelength. In most cases, this vibration willnot be transferred to car body but still existwithin bogie with large amplitude and highfrequency, which makes contribution onreducing wheelset service life. Anapproximate approach is presented to estimatethe resonance speed according to the study onthe simplified model.

The longitudinal resonance vibration ofthe wheelset will worsen the vertical dynamicperformance and bring longitudinal shake ofthe locomotive. It relates to the cross-levelirregularity and the vertical profileirregularity, and will not take place on thesmooth track. Arranging the primary verticaldamper with a forward angle is an effectivesolution to eliminate longitudinal vibration.

References

[1]. Connon, D F, Pradier, H. Rail rolling contactfatigue research by the European Rail ResearchInstitute[J], Wear, 1996,191(1): 1-13.

[2]. Sun, J, Sawly, K J, et al. Progress in thereduction of wheel spalling[A]. in: Proceeding ofthe 12th International Congr ess on Wheelset[C],Qingdao, 1998: 18-29.

[3]. Sato, Y, Matsumoto, A. Review on rail

corrugation studies[A], in: Proceeding of the 5thInternational Conference on Contact Mechanics andWear of Wheel/Rail System[C], Tokyo, 2000: 74-80.

[4]. Jin X S, Shen Z Y. Rolling contact fatigue ofwheel/rail and its advanced research progress [J].Journal of the China Railway Society, 2001, 23(2):92-108.(in Chinese)

[5]. SIMPACK documentation, 2003, SIMPACK-Track Module, VIII-TE8.

[6]. Garg, V K, Dukkipati, R V. Dynamics of RailwayVehicle System [M]. Academic Press, 1984.

[7]. Knothe, K. A Contribution to thestandardization of definitions for contactphenomena in wheel-rail-systems [J], in: ZEV railGlas.Ann., 2003, 127: 204-211.

[8]. Grassie, S L, Kalousek, J. Rolling contactfatigue of rails: Characteristic, causes andtreatments [A]. in: Proceedings of the 6thInternational Heavy Railway Conference,Southafrica, 1997: 381-404.

[9]. Wickens, A H. Fundamental of Rail VehicleDynamics: Guidance and Stability [M] .Loughborough University, UK, 2003.

[10]. Zhai W M, Wang K Y , et al. Applications ofthe Theory of Vehicle-track Coupling Dynamics tothe Design of Modern Locomotives and RollingStocks [J]. Journal of the China Railway Society,2003, 26(4):24-30. (in Chinese)

[11]. Zhai W M. Vehicle-track Coupling Dynamics[M]. China Railway Publishing Company, BeiJing,2001.

[12]. Lin J H, Chen J Z, et al. Theory Analysis andTest Research of Chinese Main Track IrregularitiesPSD [J]. Chinese Journal of MechanicalEngineering, 2004, 40(1): 174-178. (in Chinese)

[13]. Zhang M. Dynamic of Response of aLocomotive to Random Lateral Rail Irregularities[J].Journal of Southwest Jiaotong University,1994, 29(1): 39-44. (in Chinese)

[14]. Qi F L, Luo L, et al. Application of ModalAnalysis Technology in the Study of Track SystemDynamics [J]. China Railway Sciences, 1999,

20(1): 1-8. (in Chinese)


I. INTRODUCTION

In present, many transport system constructions are now making rapidly progress accordingto social and economic development in Vietnam. Large and many bridges have beenconstructed, are underway and in planning. Vietnam does not have hist ory of large scaleearthquakes, however some small earthquake have been recorded by seismometers installed inVietnam. According to analysis of earthquake records obtained by the seismometers, Vietnam islocated moderated seismic activity area. Seismic des ign Specification in Vietnam (a part of22TCN 272-05) is established based on AASHTO LRFD 1998. Many bridges are design by thisSpecification.

Recently, some bridges have been constructed by Japanese economic support (ODA). Alsosome bridges are constructed in very soft ground or potentially liquefaction sand area.Evaluation by Japanese Specification is required for these conditions. It is necessary in order toconfirmation of seismic performance of the bridge and also to evaluate seismic performance bydynamic response analysis based on Japanese Specification. The dynamic response analysisconsidering material non-linearity can evaluate seismic performance on not only resistance ofmembers but also deformation, unseat, etc…In this paper, unseat of the girders is mainlyevaluated to secure of the existing bridge.

SEISMIC RESISTANCE OF MULTI-SPANS PC BRIDGE UNDER

EARTHQUAKE OCCUR IN VIETNAM

TRAN VIET HUNGMsc., Dept. of Civil Eng., University ofTransport and CommunicationCaugiay, Dongda, Hanoi, VietnamNGUYEN VIET TRUNGDr. of Eng., Professor, Dept. of Civil Eng.,University of Transport and Communication,Caugiay, Dongda, Hanoi, Vietnam

Abstracts: The response of bridges when subjected to seismic excitation can be evaluated

by dynamic response analysis methods. A preliminary seismic response analysis of a multi -

spans highway bridge in Vietnam was performed using d ynamic analysis procedures to

identify the potential for nonlinear response of bridge structure. In this study, a typical

concrete girder bridge (multiple frames in the longitudinal direction) in Vietnam is used to

seismic resistance evaluations. These eva luations are performed by performing nonlinear time

history analyses on FEA method. It is found that the typical bridge and structure

characteristic influence on response of the bridge during earthquake .

Keywords: Seismic analysis, non-linear analysis, bridge in Vietnam, earthquake

engineering, dynamic response analysis.


II. DYNAMIC RESPONSE ANALYSIS

The dynamic response of a structure depends on its mechanical characteristics and thenature of the induced excitation. Mechanical properties which are efficient to mi tigate thestructure’s response when subjected to certain inputs might have an undesirable effect duringother inputs. In a dynamic analysis, the number of displacements required to define thedisplaced positions of all the masses relative to their origina l positions is called the number ofdegrees of freedom (DOF). The equation of motion of an MDOF system is similar to the SDOFsystem, but the stiffness k, mass m, and damping c are matrices. The equation of motion to anMDOF system under ground motion can be written as

[M]{ u } + [C]{ u }+[K]{ u }= -[M]{B} gu (1)

The stiffness matrix [K] can be obtained from standard static displacement -based analysismodels and may have off-diagonal terms. The mass matrix [M] due to the negligible effect ofmass coupling can best be expressed in the form of tributary lumped masses to thecorresponding displacement degrees of freedom, resulting in a diagonal or uncoupled massmatrix. The damping matrix [C] accounts for all the energy-dissipating mechanisms in thestructure and may have off diagonal terms. The vector {B} is a displacement transformationvector that has values 0 and 1 to define degrees of freedom to which the earthquake loads are

applied. Value gu is ground acceleration.

For the purpose of analysis, energy absorbed by inelastic deformation in a structuralcomponent may be assumed to be concentrated in plastic hinges and yield lines. The location ofthese sections may be established by successive approximation to obtain a lower bound solutionfor the energy absorbed. For these sections, moment -rotation hysteresis curves may bedetermined by using verified analytic material models. In addition to a lin ear analysis, it iscommon practice to perform a capacity analysis associated with the desired inelastic response inwhich ductile flexural response occurs at selected plastic hinge regions within the structure. Theplastic hinge regions are detailed to en sure plastic behaviour while inhibiting nonductile failuremodes. The hysteresis properties were the Takeda hysteresis properties. The Takeda degradingstiffness (see Fig. 1) and bilinear elasto -plastic hysteresis (see Fig. 2) were considered to derivethe inelastic spectra.

Fig 1. Takeda model Fig 2. Elasto-plastic hysteresis


III. ANALYTICAL MODELING

The bridge is multi-span continue bridges, representative of typical bridges in Vietnam,evaluated under this study. For superstructure was a continuou s hollow slab beam bridge with 8spans, a total length of 250m. Pier columns and abutments have bored cast -in-place pilefootings. The bridge arrangements are 3 rigid frame piers (P2, P3, and P4) and 4 bent piers isinstalled by rubber bearings support (P0 , P1, P5 and P6). The pier columns are all circular withspiral or circular lateral reinforcements. Elevation for bridge is shown in Fig. 3. The soilconditions are almost medium sand, fine sand and gravelly sand.

Fig 3. Profile of bridge

G ird e r ( lin e a r b e a m e le m e n t)

B e a rin g s p rin gS to p p e r

L o n g .= 0 .1 2 m

S p a c in g

A c tin g fo rc e

D is p .

K = 1 0 k N /m

H o z i.= 0 .0 5 m

8

K = 1 0 k N /m-1

C o lu n m p ie rN o n lin e a r b e a m e le m e n t

N o n lin e a r ro ta tin gP la s tic h in g e s p rin g

B e a m e le m e n tF o o tin g

s p r in g e le m e n t

F o u n d a tio n g ro u n d s p rin g

P

D is p .

k

k 1

Fig 4. Modeling of a bridge pier

In the pier column, a linear rotating spring that modeled a plastic hinge, the column bodywas a nonlinear beam element. The rubber bearing used bilinear model is a nonlinear spring inhorizontal direction. The dynamic analysis was perf ormed using the Newmark β method andintegration time interval was 0.01 second. The nonlinear behavior of the columns is presentedby the Takeda model with the potential plastic hinge zone located at bottom of the column asshown in Fig. 4. The stress vs. strain relation of reinforcing bars is idealized by an elastic perfectplastic model.


Lp = plastic hinge length (Lp = 0.2h – 0.1D; 0.1D ≤Lp ≤ 0.5D) with h = height of thecolumn pier (2)

Table 1. Plastic hinge length of pier, Lp (m)P0, P6 P1, P5 P2, P4(*) P3(*)

h (m) 3.9 4.8 5.7 5.9Lp m) 0.5 0.7 0.4 0.4 0.5 0.5

(*) The piers P2, P3, P4 have 2 locations of the plastic hinge length at bottom and top of pier

The inelastic dynamic analysis is performed by incorporating the non -linear rotationalspring (Kx, Ky, Kθ). The dynamic characteristic value of the surface ground is 0.82s , i.e. type IIIground according to the seismic design specified in Japanese Specifications for Highway Bridge(Part V Seismic design). In the analysis, the damping model used was Rayleigh damping. Thedamping of a structure is related to the amount of energy dissipated during its motion. It couldbe assumed that a portion of the energy is lost due to the deformations, and thus damping couldbe idealized as proportional to the stiffness of t he structure. Another mechanism of energydissipation could be attributed to the mass of the structure, and thus damping idealized asproportional to the mass of the structure. In Rayleigh damping, it is assumed that the damping isproportional to the mass and stiffness of the structure.

[C]= a0 [M] + a1 [K] (3)

in which [C] = damping matrix of the physical system; [M] = mass matrix of thephysical system; [K] = stiffness matrix of the system; a0 and a1 are pre -defined constants. Thegeneralized damping of the nth mode is then given by:

Cn = a0 Mn + a1Kn (4)

Cn = a0Mn + a1 ωn2Mn (5)

nn

nn M2

C

(6)

n1

n

0n 2

a1

2

a

(7)

Fig 5. Rayleigh damping variation with natural frequency


Fig. 5 shows the Rayleigh damping variation with natural fre quency. The coefficients a0and a1 can be determined from specified damping ratios at two independent dominant modes(say, ith and jth modes). Expressing Eq. (8) for these two modes will lead to the followingequations:

i1

i

0i 2

a1

2

a

(8)

j1

j

0j 2

a1

2

a

(9)

When the damping ratio at both the i th and jth modes is the same and equals ξ, it can beshown that:

ji

ji0

2a

;

ji1

2a

(10)

It is important to note that the damping ratio at a mode between the ith and jth modes isless than ξ. And in practical problems, the specified damping ratios should be ch osen to ensurereasonable values in all the mode shapes that lie between the ith and jth modes shapes. In theanalysis, the damping model used was Rayleigh damping. The Rayleigh damping coefficientwas set based on the vibration mode of the structure. We s tudy in case of natural frequencies are2.337Hz and 7.305Hz for damping ratio are 0.0322 and 0.0568, respectively. The values area0 = 0.45829 and a1 = 0.00225.

Table 2. Damping ratio of the members

Member Damping ratio

Bridge column – pier

(nonlinear member)2%

Bridge column – pier, footing

(linear member)5%

Girder (linear member) 3%

Bearing spring 4%

Foundation spring 10%

This analysis used two ground motion records in Japan are acceleration records in theDorokyou Shihousho (Level 1 in Japa n code) and Kushirogawa-1994 (Level 2 in Japan code)with the peak ground acceleration of ground motion records are 1.41m/s 2 (following to lowearthquake occurs in Vietnam) and 4.38m/s 2, respectively. The main seismic attack on moststructures is the set of horizontal inertial forces acting on the structural masses, these forcesbeing generated as a result of horizontal ground accelerations. For most structures, verticalseismic loads are relatively unimportant in comparison with horizontal seismic loads. T herefore,in this study the structure is excited in the horizontal (longitudinal) direction. The records have avariety of peak ground acceleration as shown in Fig. 7.


0

5000

10000

15000

20000

25000

00.0

010.0

020.0

030.0

040.0

050.0

060.0

07

φ (1/ m)

Mom

ent

(kN.

m)

P1, P5P0, P6P2, P4P3

a) Moment vs. curvature relation

0

5000

10000

15000

20000

25000

00.0

005 0.001

0.0015 0.0

020.0

025 0.003

0.0035 0.0

040.0

045

θ (rad)

Mom

ent

(kN.

m)

P1, P5P0, P6P2, P4P3

b) Moment vs. rotation relation

at the plastic hinge

Fig 6. M - and M – θ relationships of column piers

Max acc. is 1.41m/ s2

- 1.5

- 1.0

- 0.5

0.0

0.5

1.0

1.5

2.0

0 10 20 30 40 50

Time (s)

Acce

lera

tion

(m/s

2 )

a) Horizontal acc. in the Dorokyou Shihousho

(Maximum acc. is 1.41m/s2)


- 6.0

- 4.0

- 2.0

0.0

2.0

4.0

6.0

0 10 20 30 40 50 60

Time (s)

Acc.

(m/s

2 )

b) Horizontal acc. in the Kushirogawa, 1994


Fig 7. The ground motion records used in this analy sis

IV. RESPONSE OF BRIDGE STRUCTURE IN A LOW -MODERATE SEISMIC ZONE

The main objective is to determine relative superstructure -substructure displacementsobtained from an dynamic analysis for typical highway bridges located in a low to moderateseismic zone. Fig. 8 shows the response of the bridge for peak magnitude of the input wave is1.41m/s2 while install stopper and non-stopper. The accelerations of girder are 1.62m/s 2 and2.47m/s2 while maximum displacements of girder are 1.4cm and 2.2cm for install s topper andnon-stopper, respectively. The small displacement of girder can result in frame rigid structure atpier P2, P3, P4. With the results shows under a poor ground excitation, the bridge structure cannot damage of girder and bearing support. In this case, stopper and no-stopper weren’t influentto safety of structure when poor earthquake occurs. Fig. 9 shows the hysteretic response at theplastic hinge of the pier P1 (rubber bearing) and pier P2 (rigid pier and girder). The maximumrotations of P1, P2 are 1.072×10-4 rad, 4.149×10-5 rad and 3.134×10 -4 rad, 2.062×10-4 rad forinstall stopper and non-stopper, respectively. The difference can be caused by piercharacteristics and typical structure. However, the possibility of the undesired behavior, such asunseating failure of the superstructure, is still low since the absolute value of the relativedisplacement is very small.


- 3

- 2

- 1

0

1

2

3

0 10 20 30 40 50 60

Time (s)

Acce

larat

ion

(m/s

2 )

Stopper Non- stopper

a) Acceleration of the girder at the end girder

(A1 side)

- 0.025

- 0.020

- 0.015

- 0.010

- 0.005

0.000

0.005

0.010

0.015

0.020

0.025

0 10 20 30 40 50 60

Time (s)

Disp

lacem

ent (

m)


b) Displacement of the end girder

(A1side)

0

0.005

0.01

0.015

0.02

0.025

P0 P1 P2 P3 P4 P5 P6

Horiz

onta

l Dis

plac

emen

t (m

) Non stopper Stopper

c) Displacement of top’s pier

Fig 8. Response of the bridge with wave 1.41m/s 2

- 8000

- 6000

- 4000

- 2000

0

2000

4000

6000

8000

- 0.0002 - 0.0001 0 0.0001 0.0002Rotat ion, θ (rad)

Mom

ent

(kN

.m)

Pier P1

- 8000

- 6000

- 4000

- 2000

0

2000

4000

6000

8000

- 0.0001 - 0.00005 0 0.00005 0.0001Rotat ion, θ (rad)

Mom

ent

(kN

.m)

Pier P2

Fig. 9a) Case of stopper

- 15000

- 10000

- 5000

0

5000

10000

15000

- 0.0004 - 0.0002 0 0.0002 0.0004Rotat ion, θ (rad)

Mom

ent

(kN

.m)

Pier P1

- 15000

- 10000

- 5000

0

5000

10000

15000

- 0.0004 - 0.0002 0 0.0002 0.0004Rotat ion, θ (rad)

Mom

ent

(kN

.m)

Pier P2

b) Case of non-stopper

Fig 9. Hysteretic response at the plastic hinge of pier under input wave 1.41m/s2


V. RESPONSE OF THE BRIDGE WITH STRONG GROUND EXCITATIO N

Following to 22TCN 272-05, longitudinal stoppers shall be designed for a force calculated

as the acceleration coefficient times the permanent load of the lighter of the two adjoining spans

or parts of the structure. If the stopper is at a point where rel ative displacement of the sections of

superstructure is designed to occur during seismic motions, sufficient slack shall be allowed in

the stopper so that the restrainer does not start to act until the design displacement is exceeded.

Fig. 10 and Fig. 11 shows the response of the bridge for peak ground acceleration is 4.38m/s 2

when installed stopper and non-stopper. The accelerations of girder are 38.6m/s 2 and 4.52m/s2

while maximum displacements of girder are 10.13cm and 23.91cm for install stopper and non -

stopper, respectively. The acceleration of girder in case of install stopper larger non -stopper is

caused by the sudden exchange of velocity at the time pounding. Longitudinal displacement of

girder can be restrained by stopper in the top of pier.

- 0.05

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0 10 20 30 40 50 60 70

Time (s)Horiz

onta

l disp

lacem

ent (

m)

Non stopper Stopper

Fig 10. Displacement of the end girder (abutment A1) with input wave 4.38m/s 2

-0.06

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0 10 20 30 40 50 60 70

Time (s)

Disp

lacem

ent (

m)

Pier P0 Pier P1 Pier P2 Pier P3

P0P3P2P1

a) Case of stopper

-0.05

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0 10 20 30 40 50 60 70

Time (s)

Disp

lacem

ent (

m)


P0

P1P2

P3


Fig 11. Displacement of top’s pier with input wave 4.38m/s2


- 40000

-30000

-20000

-10000

0

10000

20000

30000

-0.04 -0.03 -0.02 -0.01 0 0.01

Rotat ion, θ (rad)

Mom

ent

(kN.

m)

Resistance Response

Mc,θ c

Mu,θ u

Mc=6520.4kN.m θ c=8.54x10- 5rad My=20756.4kN.m θ y=9.29x10- 4rad Mu=22224.0kN.m θ u=4.18x10- 3rad Mc,θ c

My,θ y

Mu,θ u

My,θ y

PierP1

- 50000

- 40000

- 30000

- 20000

- 10000

0

10000

20000

30000

40000

- 0.04 - 0.03 - 0.02 - 0.01 0 0.01

Rot at ion, θ (rad)

Mom

ent

(kN.

m)

Resistance Response

Mc=6569.2kN.m θ c=4.92x10- 5rad My=20895.1kN.m θ y=5.34x10- 4rad Mu=22276.6kN.m θ u=2.39x10- 3rad

PierP2

Mc,θ cMy,θ y

Mu,θ u

Mc,θ cMy,θ yMu,θ u

a) Case of stopper

- 40000

- 30000

- 20000

- 10000

0

10000

20000

30000

- 0.05 - 0.04 - 0.03 - 0.02 - 0.01 0 0.01


Mom

ent

(kN.

m)

Resistance Response

MuMy

Mc

Mc

My Mu


PierP1 -90000

-70000

-50000

-30000

-10000

10000

30000

-0.1 -0.08 -0.06 -0.04 -0.02 0 0.02

Rotation, θ (rad)

Mom

ent

(kN.

m)

Resistance Response

Mc=6569.2kN.m θ c=4.92x10-5rad My=20895.1kN.m θ y=5.34x10-4rad Mu=22276.6kN.m θ u=2.39x10-3rad

PierP2

McMyMu

McMyMu


Fig 12. Hysteretic response at the plastic hinge of pier under input wave 4.38m/s 2

Fig. 12 shows the hysteretic response at the plastic hinge of the pier P1 (rubber bearing)

and pier P2 (rigid pier and girder). The maximum rotations of P1, P2 are 3.15×10 -2rad, 3.67×10-

2rad and 4.06×10-2rad, 8.34×10-2rad for install stopper and non-stopper, respectively. The results

shows almost moment at the plastic hinge of pier overpass moment resistance of pier, thus crack

potential occurs in pier will happen.

VI. CONCLUSIONS

The dynamic behaviors of a multi-span highway bridge system under seismic excitations

are examined with various conditions. On the basis of the results and discussions of the current

study, the following conclusions can be made:

1. Although relative displacement superst ructure - substructure analysis and beam seat

length may be valuable in estimating seismic resistance of bridge in seismic zones, especially

for highway bridges located in a low to moderate seismic zone (i.e. acceleration coefficient,


A=0.09-0.29g) such as Vietnam. This bridge is safety during earthquake in Vietnam.

2. In conclusion, the continuous spans may remain in the elastic range without any

disruption to traffic due to the low seismic excitation such as Level 1, JRA -2002; while a partial

damage may occur due to the strong earthquake such as Level 2, JRA -2002. Thus evaluation

seismic by dynamic response analysis is very importance and needed.

3. The use of stoppers may decrease displacement of girder and prevent unseating fromsuperstructure

References

[1]. Specification of highway bridges, Part V seismic design, Japan Road Association, 2002.

[2]. 22TCN 272-05, Specification for bridge design, Ministry of Transport of Vietnam, 2005.

[3]. TCXDVN 375: 2006, Design of structures for earthquake resistance , Vietnam Ministry of

Construction, 2006.

[4]. Meterials design of bridge structure in Vietnam

[5]. AASHTO. Load and resistance factor design (LRFD) specifications for highway bridges. Washington

(DC): American Association of State Highway and Transportati on Officials (AASHTO), 1998.

[6]. AASHTO. AASHTO LRFD Bridge design specifications, 3rd Edition, Washington (DC): American

Association of State Highway and Transportation Officials, 2004.

[7]. Tongxiang An, Osamu Kiyomiya , Dynamic response analyses and mod el vibration tests on seismic

isolating foundation of bridge pier, Structural Eng./Earthquake Eng., JSCE, Vol. 23, No. 2, pp. 195s -214s,

2006.

[8]. Shigeru Miwa, Takaaki Ikeda , shear modulus and strain of liquefied ground and their application to

evaluation of the response of foundation structures, Structural Eng./Earthquake Eng., JSCE, Vol. 23, No.

1, pp. 167s-179s, 2006.

[9]. Yusuke Ogura, Shigeki Unjoh , Response characteristics of bridge abutments subjected to collision of

girder during an earthquake, Structural Eng./Earthquake Eng., JSCE, Vol. 23, No. 1, pp. 135s -141s, 2006.

[10]. Nasim K. Shattarat, Michael D . Symans, David I. McLean, William F. Cofer, Evaluation of

nonlinear static analysis methods and software tools for seismic analysis of highway bridges, Engineering

Structures, Vol. 30, Issue 5, pp. 1335-1345, 2008.

[11]. M. Ala Saadeghvaziri, A.R. Yazdani-Motlagh, Seismic behavior and capacity/demand analyses of

three multi-span simply supported bridges, Engineering Struct ures, Vol. 30, pp. 54-66, 2008


I. INTRODUCTION

During the last fifteen years, researchefforts in transportation have considerablyevolved at the international, European andnational level. The focus of the research in allfields of automotive industry and transport hasbeen the application of technological andoperational advances that will permit the safest,most comfortable and most cost-efficient possiblemobility of people and goods by private andpublic transport means, while at the same timerespecting the environment and natural resources.Relevant to that has been the development of anintegrated multimodal intelligent transportsystem, that will be efficient in terms of safety,effectiveness, cost and options provided to thepublic with respect their mobility.

The White Paper on Transport“European Transport Policy for 2010: Time todecide” and its mid-term review set outclearly those objectives to be addressed at apan-European level. The TechnologyPlatforms set up in the Transport sectors(ACARE for aeronautics and air transport,ERRAC for rail transport, ERTRAC for roadtransport, WATERBORNE for waterbornetransport, Hydrogen and Fuel cells) haveelaborated long-term visions and strategicresearch agendas which constitute usefulinputs to the approach and activ ities of theTransport theme and complement the needs ofpolicy makers and expectations of society.

In the 6th European research framework,the core objective of the activities carried outwas the promotion of road safety by means of

SUSTAINABLE TRAFFIC SAFETY POLICIES ANDRESEARCH PRIORITIES FOR SAFE AND

SECURE EUROPEAN ROADS

Dr. EVANGELOS BEKIARIS

Research Director of CERTH/HITForum for European Road TrafficSafety Institutes (FERSI) President

Summary: The European goal of reaching 50% reduction of road traffic accidents is

eluding us and efforts on national as well as European level need to be climaxed to even

coverage towards this goal. This paper performs a brief state of the art on recent (up to 6th

FP) efforts on European level to enhance road safety (including passive safety, active safety,

training and other measures) and then key priority areas towards the future (with reference

also to the 7th FP) are proposed along 5 axes: harmonised and complete traffic accident

database, passive safety systems, ITS and active safety systems, measures for dangerous

goods, simulation models and use of driving simulators. The paper concludes with detailed

research priorities, reflecting the author’s views and over 20 years of experience in TrafficSafety Research in Europe.


the use of new technologies (e-safety). Themain objectives of the 6th framework werereflected in the following research initiatives:

Creation of advanced vehicleapplications for accident prevention.

Creation of communication channelsamong the vehicles for the decentralisedmanagement of traffic.

“Intelligent” communicationbetween the vehicles and the infrastructure.

Cooperation between vehicles andmobile devices (PDA and mobile phones)aiming to support the driver or any other userin a seamless and dynamic way.

Development of a unified system forroad users charging and electronic tollcollection in Europe, via the use of satelliteGNSS technologies.

The industries in the automotive sector havebeen the main actors motivating the rapidprogress. Aiming at safer driving and at provisionof added value services to the driver, theautomotive industries have focused on thedevelopment of holistic systems for themanagement of the info provided in the vehicle,the in-vehicle navigation and the communication.

The intelligent transport systems (ITS)applications have started since 30 years ago,aiming, initially, to address the need for theefficient management of road infrastructureand, especially, of the urban network, i.e.roads and interchanges with traffic signs.Indicatively, one may refer to SCOOT (Split,Cycle and Offset Optimisation Technique)and SCATS (Sydney Coordinated AdaptiveTraffic System) applications.

Progressively, and as the telematics anddigital technologies were being improvedmore and more, the ITS applications wereexpanded in all transport means and addressed

a wide range of operational functions. Inspecific, the development of digitalframeworks for navigation, enabled thedevelopment of in-vehicle applications and ofother applications providing services out ofthe vehicle. The advanced localiSation andnavigation technologies, the wirelesstechnologies for mobile devices, the DSRC,RFID, DAB and RDS/TMS technologies havebeen considered the most significantlandmarks in the area. According to m arketestimates, in 2010, the demand for navigationdevices will be around 12 millions per year.

II. THE RESEARCH ROAD TOWARDSSAFER TRANSPORT

Over the last decade, the technologicaldevelopments addressed mainly the passivesafety systems, with regard to the human(mainly the driver), the vehicle and theenvironment. Concerning the vehicle passivesafety systems, the most considerable progresshas been made in relation to preventive carbodies, multiple airbags and advanced seat beltsystems. New structural frameworks (i.e.Honicomb) and materials (i.e. composites) havebeen developed for the front part of the vehicle(mostly of the passenger vehicles, semi-trucksand trucks) so as to be, among others, user -friendly to the vulnerable road users (e.g.motorcyclists, pedestrians, etc.), as well as forthe lateral part of the vehicle (mostly that one ofthe passenger vehicles), for the damping of themaximum possible energy during collision andthe reduction of the vehicle speed with theminimum possible deceleration.

The requirements of the crash tests havebeen further elaborated and are re-evaluatedand re-adjusted according to real accidents’results. The incompatibility among the severaltypes of vehicles, which is critical duringcollision (e.g. height difference in kinematicenergy absorption ranges during the collision


of a passenger vehicle with an off-road vehicle)has been addressed sufficiently, whereas avariety of research initiatives has dealt with thematerial fractural toughness and thereinforcement of the vehicle cabin for vehicleswith high centre of mass (i.e. trucks, buses, offroad vehicles, etc.) in roll-over cases, with thedevelopment and utilisation of moreadvantageous materials for the construction ofthe cabin, special seats, etc. In addition, thepotentials for the readjustment of the passivesafety systems in case the passenger is out ofposition (the original position upon which thesystem was designed and developed) have beeninvestigated. Great progress has been made inthe s/w programs that simulateincidents/accidents scenarios, providing thepotential for the new technologies in passivesafety to be applied and further investigated ina cost-efficient and less time consuming way(i.e. finite elements, “Multi Body Models”,etc.). Indeed, the vehicles produced during thelast decade increased passive safety incomparison to the older technology’s vehicles.As identified in the accident analysis realisedwithin the framework of the projectPENDANT of the 5th FW program, theseverity of the accidents, dealing actually withthe severity of injuries, in which the users ofthe passenger vehicles, constructed one yearafter 1998, were involved, was not as high asthat one corresponding to vehicles, constructedearlier than 1998.

Research in passive safety addressinghuman concerns mainly vehicles, where thepassenger is not surrounded by structures (i.e.cabin in passenger vehicles), and deals mostlywith the safety of bicyclists and motorcyclists.

More specifically, extended research hasbeen realised in these fields during the lastyears (ARPOSYS, TIP-CT-2004-506503),especially regarding the protection of the riderhead. Recently, the utilisation and evolution

of reinforced polymers (e.g. “CarbonReinforced Epoxy”, “SiC/Sic Ceramic MatrixComposites”, “GLARE”) has led to helmets,which are more resistant to collisions andfriction and also much lighter.

Furthermore, the utilisation of materialsthat absorb energy (while falling) for theprotection of knees, elbows, metatarsus,shoulders, pelvis and backbone has beeninvestigated and some minimum requirementsregarding quality and effectiveness have beenset by the EU.

Investigation has been also held for theparticipation and behaviour of the road sidefurniture in accidents (e.g. RISER project),either concerning those that aim to reduce theseverity of the consequences of an accident(i.e. safety islands), absorbing the greatestpossible kinematic energy during the collisionwith any type of vehicle (e.g. motorcycles,passenger vehicles, trucks, etc.) or those thataim to prevent accidents beforehand (e.g.traffic signs, light pillars, etc.).

In the last category, research has focusedon the detection of the most appropriate spotsfor their placement, by means of theinvestigation of the most common accidentscenarios taking into consideration the specificcharacteristics of the road/environment, as wellas on the structural framework of the object andthe materials used, aiming at the preventionfrom high instantaneous decelerations, that canresult in physical damages, and the preventionfrom the penetration of the road side furniture inthe passenger cabin. The above are evaluatedeither using feedback from accident statistics, or,after the implementation of the application, viacrash tests.

However, the most recent trends focus onactive safety systems, that will be able, notonly to provide the maximum possible safetyafter the accident, but furthermore act such as


to prevent it.

Typical applications of Advanced DriverAssistance Systems (ADAS), alsocommercialised nowadays, are the AdaptiveCruise Control systems (ACC) or theAdvanced Vehicle Control Systems (AVCS).ADAS function is based so far on theutilisation of the info that is logged by thesensors, with which the transport means (e.g.passenger vehicle, truck, etc.) are equipped. Inparallel, each year, more vehicles areequipped with various navigation systems,which make use of digital maps and vehiclepositioning. The capability of such systems tobe aware of the geometry and other propertiesof the road infrastructure, with the respectiveguidance and warning, reflects the sense ofcooperative systems and is estimated that mayhave significant positive impacts on trafficsafety and efficiency through the wholetransport network.

However, it must be pointed that thereare no ADAS systems today that take intoaccount the geometry of the road since thereexist no digital maps with information aboutthe curvature or other characteristics of theroad. One possible application is headlightsthat turn in advance according to the curvatureof the road, advanced warning to the driver ifa dangerous or abrupt turn is ahead, etc.Development of maps with detailedinformation about road characteristics ,geometry and condition of pavement is verycostly if carried out with traditional methods.An open research objective is thedevelopment of methods for the production ofadvanced maps that will be emi-automatic andbased on the analysis of paths of vehiclescollected via GPS with on-board units.

Given the fact that the possibility ofimplementing large infrastructure problems islimited (because of limited funding and

technical restrictions), the development ofintegrated intelligent transport systemsapplications is encouraged by the nationaland regional governments, infrastructureoperators and public authorities.

The use of sensors for trafficmeasurements, the detection ofincidents/accidents and the use of VariableMessage and Directional Signs (VMS/VDS)are already used in Europe in great extent.

Currently, the most dominant trend is theintegration of existing or underimplementation projects in an interoperableframework that will allow the cross -borderadoption of ITS.

The most innovative projects are thosedealing with electronic tolls collection(PISTA, MEDIA), e-ticketing and theexchange of traffic and other info amongseveral actors (cities, districts, etc.).

The major scope of ITS is the increase ofmobility of goods and people, in such a way,as to be in favor of all involved actors and theenvironment. This is especially applicable inthe field of dangerous goods transport, wherethe traffic incidents have multiple negativeimpacts with regard to safety (of the driversand the third party) as well as the fina ncialstatus and the marketing profile of thetransportation and the dangerous goodscompanies, but also to the environment.

For the above reasons, European projectsaim at the development of such technologiesand integrated services that will allow the s afeand more cost-efficient transport of dangerousgoods through the whole transport network.Most of them deal with the development ofthe technological framework that is requiredfor the dynamic management of the dangerousgoods fleet and the seamless pr ovision ofreliable info for the vehicle, the driver and the


cargo status, the time of delivery, the potentialof terrorism action (i.e. SAFESEANET ,TEMPO ARTS, MITRA, RIS, MVS, SHAFT,DETRACE, ULISSE, etc.), whereas in recentresearch initiatives, more integrated servicesare designed, including emergency services(Police and Ambulance reaction) and routeguidance with advanced Decision SupportSystems, that take into considerationoperational, financial and the environmentalinfo, real traffic conditions and riskassessments (i.e. GOOD ROUTE project).Common ontologies with regard to theclassification of the dangerous goodsaccording to ADR («Agreement concerningthe international carriage of Dangerous goodsby Road»), security and authenticationsystems for the privacy data protection (ofusers, companies, etc.), advanced navigationsystems and user interfaces for all involvedactors, reliable localisation, positioning andcommunication technologies, trafficmanagement information centres andimproved vehicle tracking technologiesconstitute only part of the existing and futuretechnological development in this area.

In parallel, some research initiatives (i.e.INFORMED project) have focused on the training ofthe professional drivers and their instructors,developing training programs that include training inadvanced techniques (i.e. anti roll-over, antiskid,defensive driving, etc.) of several types of vehicle,incorporating the use of multimedia software trainingtools, training with simulators and practical training(i.e. in test trucks), whereas have formulated a set ofpolicy recommendations for the improvement of therelevant European Directives dealing with trainingissues in this field. The need for the formulation of acommon European training, assessment andcertification framework for the professional and forall other types of drivers is emerging and obvious.

Before any other process, the appropriatecollection, reporting and in-depth analysis of

accidents, the reduction of whi ch is theobjective of each system under development,is the first mandatory step. The accidentanalysis may be performed in several scales,varying from the analysis at national level,where the total number of one countryaccidents is investigated and in ternationalcomparisons are further made, to the scale ofindividual accidents, where representation andin-depth assessment of the accident is realizedaiming at the identification of the root causesleading to that.

A series of research projects have b eenfunded for this purpose (i.e. STAIRS, EACS,PENDANT, SAFETYNET, TRACE, etc. ),whereas in several databases, accidents inEurope and, in some cases, in the rest of theworld, are reported (e.g. FACTS, NHTSA,MHIDAS, GES, etc.).

Simulation and modeling techniques ofthe vehicle and the traffic environment arealso considered to have a significantcontribution to preliminary research phase,during the last decades.

Traffic simulation models aredistinguished in microscopic (e.g.PARAMICS, RuTSIM, VISSIM, etc.),mesoscopic and macroscopic (e.g. VISUM,SATURN, etc.). Each of these categoriesdeals with different level in research. Forexample, the microscopic models are based onthe principles and the sense of vehiclesequence, analyse the individual behaviour ofeach vehicle, providing great accuracy bymeans of dynamic simulation and are usedmainly for the evaluation of the proposedstrategies and policies in middle and smallnetworks. The significant computing time thatis required makes the simulation of largenetworks unprofitable or even totallyunfeasible (e.g. in urban areas level, or for theassessment of several attributes, such as mean


speeds of vehicles, traffic distributionthroughput the network, etc.). The latestmentioned are addressed by macroscopicmodels that are static and are based on thebehaviour of the average vehicle population,following fluid dynamics theories. Suchmodels require less computing time; howevertheir accuracy is not so good. In addition, thetraffic simulation models are used for the newand innovative technologies (e.g. ADAS,AVCS) to perform impact assessment and toestimate the magnitude of variousenvironmental impacts (CO 2 emission, fuelconsumption, noise, etc.).

An indicative European project, withinthe framework of which, microscopic andmacrospopic models have been developed isthe ADVISORS project (GRD 1 1999 10047),whereas the ΙΝ-SAFETY project (FP6-2002-506716) aimed at the development andevaluation of microscopic and macroscopicmodels that assess the behaviour of users ofADAS/IVIS in several penetration rates, toenable the impact assessment in road safety.

Driver simulators simulate the vehicleoperation and the respective trafficenvironment. The accuracy of the simulation,their technical characteristics and their costmay differ significantly, depending on thepurpose of use. Driver simulators are used forseveral reasons, as for example, for thetraining of all drivers’ categories (e.g. novicedrivers, elderly drivers, professional drivers,etc.), for the assessment of their skills andtheir driving behaviour, for other researchpurposes like the design and development ofvehicles and parts of them (e.g. userinterfaces, ADAS, etc.), games andentertainment, etc.

The undergoing research in the areas ofdrivers’ simulators has provided evidencethat, within the different research and training

contexts of use and for the achievement of thedifferent goals each time, different simul ators,scenarios and environments are required,adjusted to the concrete needs of theapplication. Driver simulators may be singledisplay simulators, static, dynamic, semi -dynamic, virtual reality simulators, etc. Someof the best research driver simulato rs are thoseof VTT in Sweden, of Daimler Chrysler inGermany and of NADSin the U.S. Worktowards interoperable, multi -tasking and witha common reference architecture simulators iscurrently coordinated in TRAIN -ALLinitiative, involving most major simul atormanufacturers in Europe.

E-112 is a European directive requiringmobile and fixed operators to make availablethe location of every caller placing anemergency call. Mobile handsets are currentlylocated through the mobile operators whomake use of various techniques based on theknown locations of network antennas. Whileall European countries have in principleadopted the directive the system is not fullyoperational yet and is facing severe delays.

E-112 will form only part of e-call achain of actions that will bring rapidassistance to any motorist who experiences anaccident or mechanical failure on the road. E -call is a EU high priority initiative andsignificant research has been alreadyundertaken.

An e-call can be initiated manually by thevehicle occupants or triggered by an accidentand placed automatically by a black boxconnected with sensors that detect a collision.A voice and data connection is establishedwith the closest Public Service AnsweringPoint, which deploys and dispatchesassistance to the location communicated bythe call. Furthermore, it is envisioned thatrelevant data are transferred to a Service


Provider that provides additional servicessuch as towing or notification of next -of-kinfor which the caller subscribes. Initial ly, itwas forecasted that e-call would beoperational in Europe by 2010. Since there aresignificant delays to the deployment of E -112,it can be expected that the deployment of e -call will be further delayed.

The problems can be partly attributed tothe fact that the techniques used for thecalculation of location of mobile phones arenot very accurate or require significantinvestments from the mobile operators in orderto be acceptably accurate. Many operatorschoose the cell id method, which does notrequire additional equipment but provides theposition of a handset within a cell. Cells aresmall enough in urban areas, but cover bigareas outside of the cities resulting inpositioning with large margin of errors whereis needed more, e.g. in rural areas.

The increased availability of mobilehandsets with GPS capabilities (and laterGalileo) will solve the location problem. It isexpected that within the 3-5 next years allmobile handsets will be GNSS capable.

Significant research efforts should bemade in solving problems such as theprotection of privacy and developing servicesthat will make the whole concept of e -callcommercially viable. Since additional services(towing, notification) will be offered on asubscription basis, a whole bouquet of relat edservices must be on offer that will beattractive for the average driver and couldsustain business cases for the future ServiceProviders. Perhaps, initially Service Providerswould be insurance companies or roadsideassistance companies, but the real challenge isto develop and offer truly innovative servicesthat will create a new telematics -relatedindustry based on e-call. At the same time,

the benefits would be the reduced responsetime and the fast arrival of medical assistanceat the accident, which is known that highlyimproves the chance of survival.

III. THE ROADMAP TO THE FUTURE

Future research, already underimplementation in the 7 th FP of the EC, aimstowards a system that will enable theincorporation of the most recent evolutionsand achievements in the passive and activesafety fields, into the traffic safety arena. Thebenefited groups will be the society as awhole, the enterprises, the competitiveness ofwhich will rise in such a way, as to allow theirpenetration to the European and t heinternational arena, with the adequate capacityand know-how. More precisely the followingpriority research areas are correlated with highpotential impacts:

The complete recording and analysis oftraffic accidents, will result in theidentification of the major problems and needsthat will be targeted by the several systemsunder development.

Research around passive safetysystems and the respective implementationsare expected to:

Reduce the severity of the injuriesfrom collisions corresponding to passengersof vehicles and all other road users.

Motivate the further development ofthe technologies and sciences (finiteelements, etc.), which are used within theframework of the passive safety systems atnational level.

The focused research on ITS and therelevant implementations, especially thoseconcerning cooperative systems, areexpected to:

Improve the traffic flow and reduce


the negative environmental impacts in thetransportation sector (especially in an urbanenvironment) via the new combined mobilityservices.

Achieve considerable improvementswith regard to the safety, efficiency andcompetitiveness of the transport systems, thetraffic efficiency and in general the feasibilityin transport (in compliance with the Europeanvision for the reduction of fatal accidents by50% until year 2010 and by 100%, “0 fatalaccidents” in long term perspective ).

Demonstrate, qualitatively andquantitatively, via large scale Pilot trials, thepositive impacts of ITS in all aforementioned,encouraging the funding and the coordinationof all relevant initiatives on behalf of allinvolved actors.

Research with regard to thedevelopment of integrated services fordangerous goods fleet management and theimproved training, assessment andcertification of the drivers and theirinstructors are expected to:

Allow all dangerous goods vehiclesto be continuously tracked and monitored,providing the relevant notification,information or warning to all involved actorsautomatically, with no physical interventionand vehicle immobilisation and loss of timeand with no occurring problems and risks inthe traffic flow, which are very common incases, where heavy vehicles are put aside theroad.

Increase safety of the drivers andthird parties that are directly influenced(especially in urban areas), to face theterrorism in this area and the considerableenvironmental pollution due to the occurringaccidents.

The use of simulation models may lead

to significant savings in resources andincreased road safety since it is expected to:

Contribute towards the evaluation ofvarious transportation policies before they areapplied thus providing decision makers witha tool that permits them to perform “what -if”scenarios and permits them not only toestimate traffic loads congestion etc. but alsovarious environmental indicators (e.g. CO2emissions, etc.).

Permit the simulation of the impactsof new technologies (ADAS, IVIS) in theexisting road networks, before these areapplied.

Make feasible the construction ofnew and/or the improvement of existinginfrastructure, in the less expensive way,since the s/w for transport modelling mayprove to be especially effective tool in thecontext of RSA, RSI and black spotsmanagement.

The utilisation of driver simulators isexpected to:

Increase the safety of the drivers,mostly of the candidate and elderly ones, sincethe driver will have the opportunity toexperience a series of driving tasks and trafficenvironments, before s/he drives in real trafficconditions and also situations, which are verydifficult or totally unfeasible to be tried in realtraffic conditions (e.g. driving with fog, snow,collision with another vehicle, pedestrian, etc.).In this way, the training procedure and contentare also improved.

Make feasible the detection and theadoption of corrective measures for newsystems and infrastructures (being simulatedvia the proper s/w modeling), before these areapplied, by means of their evaluation from allaspects, including the investigation of the targetusers acceptance, leading to potential avoidance


of accidents and unsuccessful investments

IV. CONCLUSION AND RESEARCHPRIORITIES

The priorities outlined above areconsistent with those set by the EuropeanCommission. EC supports the implementationof such projects that comply with the identifiedpolitical priorities for the unification of thetranseuropean transport networks, as these areexpressed through a series of Directives, orstudies that are related, for instance, to: ThePan-European adoption of e-Call, the creationof a common service of road users charging,the plans for the GNSS technologies adoption(Directive 2004/52: Interoperability ofElectronic road toll systems), the creation ofcommon services for drivers and passengers,accessible by all (according to “Article 169 ofthe Treaty” and the priorities of the “AmbientAssisted Living” area and the “EuropeanStatement of Principles”), GALILEO adoption,the eSafety initiative, the activities of therecently developed ‘Agency for ITSimplementation’, etc.

A series of European TechnologicalPlatforms are related to the long-term goals ofall transport fields and modes (ACARE for airtransport, ERRAC for railway transport,ERTRAC for road transport, WATERBORNEfor seaways transport, etc.). ERTRA C is theone related mostly to Road Safety, whererelevant research priorities are defined.

According to the author, the mostrelevant research priorities follow below:

1. Sufficient collection of detailedaccident data, based on the current needs inresearch. Location of accidents should beregistered with detail preferably with GPS.Moreover, it has to be stressed out that

besides traffic accident data, exposure data arealso needed. Through analysis of trafficaccidents data will permit the pinpointing ofthe reasons they occur and result in theadoption of appropriate measures (both policyrelated as well improvements in theinfrastructure). Specifically the followingactivities should be supported:

Research for the development ofnational database that contains informationon all accidents that is regularly updated.

Development of a GIS database thatcould be used to analyze the occurrence ofaccidents taking also in account the location,the geometric characteristics of the road, etc.

2. Further research, development andevaluation of advanced passive safety systemsfor the vehicle, the driver and theenvironment, which will reduce considerablythe severity of accidents and will contribute totheir avoidance, as much as possible. Inspecific, at national level, the followingactivities should be supported:

Research for the development ofsafer road side furniture and passive safetysystems for the driver and the vehicle.

Research for the development ofnew structural frameworks and the utilisationof new material for the aforementioned andthe reduction of the incompatibility betweenseveral types of vehicles.

Evolution of the crash testsrequirements and validation of them (viaaccident analysis, s/w simulations and shortterm trials in real conditions).

Evolution and improvement of s/wprograms for simulation scenarios(accident/incident scenarios) and for the


structural analysis making use of the newcomposite materials.

3. The development of integrated ITSsolutions, that will result in a more eff icientand sustainable use of transportation and totraffic accidents reduction. This will beachieved by means of advanced warningstrategies and risk detection, a reliablenetwork of sensors, technological integrationof so far independent ADAS and inter actionof them with the user. The proposed solutionshave to be highly efficient, reliable andcontribute to the increase of safety andcomfort during driving and also friendly to theenvironment. Activities to be supportedinclude the development of:

European Strategic Research,Development, Implementation and Use ofITS, in order to meet national priorities and toachieve multiple benefits.

European Architecture, commonrequirements (Quality of Service andInteroperability) and ontologies (databasesand web services) for the seamless,interoperable and cost-efficient use of ITSanywhere, any time and from anyone.

Identification of critical thematicareas of ITS and the establishment of afocused development and applicationframework per area, via the adoption of“umbrella” projects.

Deployment of new sensors and theimprovement of already existing ones for themost reliable possible perception of theenvironment and the fulfillment of complexscenarios of use interfering withintersections, interchanges, tracking ofvulnerable road users under several trafficconditions (normal, adverse, low visibility

conditions, etc.).

Driver warning strategies, automaticcontrol of the vehicle and interactive userinterfaces (with haptic, acoustical, visualchannels) as well as of evaluationframeworks for their assessment in Pilot trialsand large-scale Field Operation Tests (FOTs) .

Smart parking management systemsthat permit trip makers to check availabilityof parking through a centralized parkingsystem and permits them to makereservations in advance, or in real timethrough a bidding process.

4. Especially, for the cooperative systemsarea, the development of integratedcooperative systems which will provideadvanced, reliable, fast and safe vehicle tovehicle communication and vehicle toinfrastructure communication in real time,aiming at the provision of information andwarning to the users in time and the automaticor semi-automatic ADAS activation, vialocalisation and positioning technologies andadvanced sensor networks. The proposedsolutions need to be financially feasible,aiming at the most limited possibleintervention in the existing nationalinfrastructures. In specific, at national level,the following activities should be supported:

The recording of the cooperativesystems applications, the identification ofdeficiencies and insufficient or complete lackof technological implementation (wheneverthis is considered necessary fromsocioeconomic aspects and also technicallyfeasible), the formulation of concreteproposals for the improvement or the fullimplementation of infrastructure and thedetermination of the respective short termand long term technological plan.


The development of a normalisedArchitecture, common requirements (Qualityof Service and Interoperability) andontologies (databases and web services) forthe seamless, interoperable and cost -efficientuse of ITS anywhere, any time and fromanyone.

According to the above ITSArchitecture, the development and interfacingof traffic and information managementcentres in urban, extra-urban, rural andinterurban networks.

The development of simulation toolsand evaluation platforms, which will allowthe technical and socio-economic evaluationof the proposed solutions.

5. The execution of large scale Pilots insimulators and test tracks for the evaluation ofITS applications (addressing also cooperativesystems), with regard to the reliability of theirperformance and their user friendliness andacceptance. In parallel, feasibility studies forthe proposed solutions will be conducted inthe national and European market. T hefollowing activities should be supported:

Evaluation frameworks and test plansfor Pilots (FOTs) impact assessment withregard to safety and the traffic environment ,which will include, among others, theexperimental and statistical planning and inadvance simulation, techniques for themeasurements gathering, addressing subjectiveand objective criteria, methodologies for theselection of the statistical sample, the scenariosof use and the timetables according to scientificmethods, methods for the trials conduct andfinally the drawing out of quantitative andqualitative conclusions dealing with theexpected impacts of the tested anddemonstrated applications.

The compatibility check of ITSapplications against the policies set in the areasof Transport and Environmental protection inEurope before their approval and funding.

Large scale ITS applications Pilotsand impact assessment. In specific, for thecooperative systems area, pilots for theevaluation of solutions dealing with vehicleto vehicle and vehicle to infrastructurecommunication and comparison of them toexisting solutions (via s/w for transportmicro/macro-modeling) should be encouraged.

In the European Research area, the aboveResearch priorities are already addressedwithin the relevant Research programmes ofDG INFSO, DG RESEARCH and DGTREN;whereas their implementation is under theumbrella of an EC Agency, that develops an‘Action Plan for the deployment of IntelligentRoad Transport Systems for more efficient,safer and cleaner transport’.

References

[1]. Communication from the Commission to the

European Parliament, the Council, the European

Economic and Social Committee and the

Committee of the Regions - Preparing Europe’sdigital future i2010 - Mid-term review

{SEC(2008) 470}.

[2]. ESafety Final Report of the eSafety Working

Group on Road Safety

(http://ec.europa.eu/information_society/activities/esafety/doc/esafety_library/esafety_wg_final_re po

rt_nov02.pdf), November 2002.

[3]. European Commission ‘Keep Europe moving’.Mid-term review of the 2001 transport White

Paper, 2006, ISBN 92-79-02312-8.

[4]. European Commission WHITE PAPER

‘European transport policy for 2010: time todecide’, 2001, ISBN 92-894-0341-1

http://ec.europa.eu/information_society/activities/


CT 2

I. INTRODUCTION

1.1 Backgrounds and problem statements

The available cost data on highway projects in developing countries are generally limitedin numbers, incoherent and contain a large amount of incomplete data. Limited financialresources and lack of systematic cost data collection are also the major problems which affectcost database quality. These factors negatively affect to cost estimators and planners to makeappropriate decisions. Accurate cost estimation in early project development is an impor tantissue where detailed information is not available and project costs are to be decided, in mostcases cost estimating relationship techniques are used. One of the challenges is accurate costestimation during pre-feasibility studies.

Most research studies on cost estimation early stage of project were conducted using datafrom developed countries (Healey, 1964; Sanders et al., 1992; Pearce et al., 1996; Smith et al.,

PRELIMINARY ROAD COST STUDIES IN

DEVELOPING COUNTRIES

Lecturer JAMSHID SODIKOVTashkent Automobile and Road Institute,20, street Movounnahr,Tashkent, 100020

Advisor EconomistZIYODULLO PARPIEVUNDP Uzbekistan Country Office

Abstract: In the early estimation there is compromise between the amounts of

information available and accuracy of estimation. We propose three levels of analysis such as

regional, country and project level for road cost models in order to provide efficient data

usage. The data for our research was obtained from the World Bank’s ROCKS database,which contains unit costs for road projects from over 80 developing countries. This paper

investigates the impact of road upgrading and improvement works on overland trade in 18 out

of 32 member countries of Asian Highway Network. The results indicated approximately 6.5

billion US dollars is required to upgrade roads and improve existing surface condition of the

selected sub-network with total length of 15,842 km. The gravity model approach was adopted

to quantitatively evaluate overland trade expansion taking into account road quality

improvements with two scenarios such as road quality increases up to 50% in the first

scenario in the second one up to 75%. The results suggests that in the first scenario total

intra-regional trade will increase about 20 percent to 48.7 billion US dollars annually, while

second scenario predicts that trade will increase by about 35 percent to 89.5 billion US

dollars annually.

Key words: Construction costs - Maintenance costs - Developing countries -Regression models

INTERNATIONAL COOPERATION ISSUE OF TRANSPOR TATION – Especial Issue – No.01 147

CT 2

1997; Hegazy et al., 1998; Adeli et al., 1998; Al -Tabtabai et. al., 1999; Morcous et al., 2001;Flyvbjerg et al., 2002, Emsley et al., 2002; Levinson et. al., 2003; Trost et al., 2003; Gwang -Heeet. al., 2004). Very limited number of research has been done in developing countries(Archondo-Callao et al., 2004; Buys et al., 2006). One of t he major reasons for a few numbersof researches is related to cost database availability in developing countries. Initial steps to buildcost database for developing countries was initiated by the World Bank’s Transportation Unit in1999 and ROCKS (Road Costs Knowledge System) was introduced. Cost data collected fromdeveloping countries all around the world and this system has large amount of incomplete datain some data items. This explains the need to examine and analyze the cost estimationtechniques and how to deal with missing data in developing countries context which is theprimary objective of this study. Additionally, due to nature of available data in the ROCKSdatabase efficient data usage has been introduced by level of analysis which is also i ncorporatedin this study.

Road agencies, contractors, consultants and financial institutions need road costsinformation, which usually is locally available, but in many case it is scattered and collected inunsystematic ways. These entities need to asse ss costs differences, but no framework tocompare road costs exists. In 1999, in response to this demand, the World Bank made the firstattempt to collect this information from 67 Implementation Completion Reports of Bank –financed projects that were implemented in the period 1995 – 1999. The study found that thelevel of detail provided in these types of documents was limited and that there is a worldwideneed for a framework to collect this type of information in 2000. Consequently, the Bankdecided to develop a simple system to collect road costs and to explore other sources ofinformation. This effort resulted on the ROCKS, which is being developed by the World Bank’sTransport Unit and is primarily based on the experience of Bank staff and the informat ioncontained in roads and highways projects in developing countries.

1.2 Concept of level of analysis

Concept of level of analysis was evolved from cost model application perspective andexisting data availability. As mentioned earlier that ROCKS database contains missing data, thetask is to utilize available data in efficient way which assists to develop better cost model withcertain application purpose. These application purposes can be defined as cost estimation studyin a given geographical region, cost study within a certain country and finally cost estimation ofa specific project with detailed information. It is generally known that detailed information islimited in regional cost studies especially during preliminary cost studies. This is due to e achcountry in particular region has certain amount of cost history data which vary among countries.In some countries there may be more project details are available in others limited projectdetails are available, in some extreme case there may be no cos t history database at all. In orderto balance available data amount (project details and number of projects) with cost modelapplication purposes, three level of analysis were proposed such as:

Regional level – cost model is developed based on limit ed project details with largenumber of observations


CT 2

Country level – cost model is developed based on relatively detailed project informationwith comparatively smaller number of observations

Project level – cost model is developed based on in depth project details with limitednumber of observations

Regional level of analysis – ROCKS database was divided into 6 regional subsets forregional level of analysis namely Africa, Asia, Caribbean -Central-Middle America, East Asia-Pacific Islands, Europe-Middle East, South America region. From figure 1 it can be observedthat large number of projects belong to Africa with 29%, Asia with 24%, and Europe -MiddleEast with 19%. Total number of projects consists of 1385 from 85 countries

Figure 1. Regional Level Project Distributions (World Data)

Average unit cost of work activities among these regions varies significantly. For instancenew construction projects in Africa region on average is about one million US dollars per kmbut in Europe-Middle East on average new project is about one and half million US dollars perkm.

Although number of observation is quite large but available data items are limited. Thesedata items are workactivity and pavement width in ROCKS. The other data items such as GDPper capita, annual mean precipitation, road network density, coastline divided by area arecollected from external sources to build cost model for regional level of analysis.

Country level of analysis include additional data items t o regional level of analysis fromROCKS such as rate of work per area which obtained by dividing work duration to pavementarea and contractor type. But number of observation has dropped from 1385 to 318. Additionaldata items are available only for Armeni a, Ethiopia, Ghana, Kyrgyz Republic, Lao PDR,

19%

12%

29%

24%4%

12%

Africa

Asia

Carib-Cent-Mid America

East Asia-Pacific Islands

Europe-Mid East

South America


CT 2

Nigeria, Poland, and Uganda. Distribution of projects in each country is shown in figure 2.Countries with large number of projects are Ghana, Uganda, Poland, and Lao PDR.

Figure 2. Country Level Project Distributions

Project level of analysis includes additional data items to country level of analysis fromROCKS such as terrain type, climate and surface thickness data. But number of observation hasdropped from 318 to 56. Additiona l data items are available only for Kyrgyz Republic, LaoPDR, and Nigeria. Distribution of projects in each country is shown in figure 3. The largestnumber of projects belongs to Lao PDR .

Figure 3. Project Distributions at Project Level

16%

20%

64%KyrgyzRepublicLaoPDRNigeria

1%3%

23%

12%

3%

12%3%

43%

Armenia

Ethiopia

GhanaKyrgyz Republic

Lao PDR

Nigeria

PolandUganda


CT 2

Data for project level of analysis is very limited in terms of number of observation butproject specific data such as terrain, climate data and surface thickness are available. Only threecountries have these data for bituminous reconstructio n and partial widening bituminous with 2lanes with reconstruction projects.

II. COST MODELS

2.1. Regression cost models

Regression models have been proven to be reliable and used for decades. There are someadvantages of regression models such as they can be defined by mathematical expression andexplain relationship between dependent variable and independent variables. There are also somedisadvantages in regression models such as multicollinearity, nonlinearity, heteroscedasticityand other issues which occur in regression model development. The details of these issues arewell described in literature (Lewis-Beck M. 1980, William D et al., 1985, Brikes D et al. 1993,Allison S et al., 1999, Miles J et al., 2001, Frank E. 2001). One of the powerful techni ques toovercome shortcoming of regression models is transformation dependent or independentresponse or both. Several types of data transformation were tested and log -log transformationwas chosen for our analysis. The advantage of log -log transformation lies on ease ofinterpretation (Carroll J et al., 1988).

At regional level we postulate that unit cost (UCij) is a function of country’s GDP (Gi),country’s road network density (RNDi), pavement width (PW), country’s annual meanprecipitation (APi), coastline divided by area of the country (DLi), project type (PTi) and region(RGi) defined as follows:

ijεj jPTjβ

j jRGjγiDL5αilogAP4αlogPW3αilogRND2αilogG1α0αijlogUC

where the symbols have the following values and meanings:

UCij = unit cost of project type j in country i (US $ 2004/km)

Gi = GDP per capita of country i(US $ 2004,PPP)

RNDi = road network density of country i(km per 1000 km2)

PW = pavement width (m)

APi = annual mean precipitation of country i(mm)

DLi = coastline divided by area of country i(km per 1000 km2)

PTi = dummy variable for project type

RGi = dummy variable for region

At country level we postulate that unit cost (UC ij) is a function of country’s GDP (G i),country’s annual mean precipitation (AP i), coastline divided by area of the country (DL i), rate ofwork per area (RW), project type (PT i) and contractor (CTR j) defined as follows:


CT 2

ijεj jCTRjγ

j jPTjβlogRW4αiDL3αilogAP2αilogG1α0αijlogUC



Gi = GDP per capita of country I (US $ 2004, PPP)

RW = Rate of work per area (Work duration divided by pavement area)

APi = annual mean precipitation of country i (mm)

DLi = coastline divided by area of country i(km per 1000 km 2)


CTRj = contractor type

At project level we postulate that unit cost (UC ij) is a function of country’s GDP (G i), rateof work per area (RW), surface thickness (PST), terrain (TR j), climate (CLj), project type (PT i)and contractor (CTR j) defined as follows:

ijεj jPTjj jCTRjλ

j jCLjγj jTRjβlogPST3αlogRW2αilogG1α0αijlogUC




RW = rate of work per area (Work duration divided by pavement area)

PST = pavement surface thickness (mm)

TRj = dummy variable for terrain type

CLj = dummy variable for climate



III. CASE STUDY

3.1. The asia highway network

The AHN covers routes which cross 32 member countries with approximate t otal length of140,000 km. It starts from Japan, Tokyo , and extends to Finland, Helsinki and Bulgaria, Sofia.The network passes mostly through existing roads in those countries. Asian Highway Designstandards comprise of Primary, Class I, Class II, and Cl ass III highway classifications which aredefined according to terrain classification, design speed, width (including right of way, lane,shoulder, median strip), minimum radii of horizontal curve, pavement and shoulder slope, type


CT 2

of pavement, maximum superelevation and grade, and structure loadings (AHNDS, 2004).Table 1 highlights recommended design standards for Asian Highway routes. The minimumrequirement which satisfies the Asian highway network standard is the Class III. According toAsian highway standards, it is suggested that the Class III should be applied only when thefunding for the construction or land for constructing road is limited. There are two priorities inwhich the Asian highway network member countries have to carry out, 1) To improv e roadconditions where it is necessary in Primary, Class I, Class II and Class III which cover 72% oftotal network, 2) to upgrade the rest 28% of network at least to the Class III but preferably toClass II.

Table 1. Asian highway standards (AHNDS, 2004)

Highwayclassification

Primary (4 or morelanes)

Class I (4 or morelanes)

Class II (2 lanes) Class III (2 lanes)

Terrainclassification

L R M S L RM

S L R M S L R M S

Design speed(km/h)

120

100 80 60 100 80 50 80 6050

40

6050

40

30

Width (m)

Right ofway

(50) (40) (40) (30)

Lane 3.50 3.50 3.50 3.00 (3.25)

Shoulder 3.00 2.50 3.00 2.50 2.50 2.00 1.5 (2.0)0.75(1.5)

Medianstrip

4.00 3.00 3.00 2.50 N/A N/A N/A N/A

Min. radii ofhorizontal curve(m)

520

350210

115

350210

80210

11580

50

11580

50

30

Pavement slope(%)

2 2 2 2 - 5

Shoulder slope(%)

3 – 6 3 – 6 3 – 6 3 - 6

Type of pavementAsphalt/cement

concreteAsphalt/cement


concreteDbl. bituminous

treatment

Max.superelevation(%)

10 10 10 10

Max. verticalgrade (%)

4 5 6 7 4 5 6 7 4 5 6 7 4 5 6 7

Structure loading(minimum)

HS20-44 HS20-44 HS20-44 HS20-44

The AHN database contains info rmation of the routes for the most of the countries, whichis about 18 countries out of 32, these information include road surface condition, pavementtype, terrain and other (AHND, 2004). Table 2 shows route condition and design standard ineach country. From this table it can be observed that about 15,842 km need to be improved orupgraded in order to provide good transportation communications.


CT 2

Table 2. Road surface condition and design standards in ESCAP member countries

No. Country Route No.AH Design Standard /

Surface Condition

TotalLength(km)

1 Armenia AH81, AH82, AH83 Class III or Higher / Bad 386

2 Bangladesh AH1, AH2, H41 Below Class III 450

3 Cambodia AH11 Below Class III 198

4 China AH3, AH32, AH42 Below Class III 542

5 Georgia AH81, AH82 Class III or Higher / Bad 55

6 India AH1, AH 2 Below Class III 75

7 IranAH1, AH8, AH70, AH72, AH75, AH78,

AH82 Class III or Higher / Bad 1084

8 Kazakhstan AH7, AH61, AH62, AH63, AH70 Below Class III 897

9 Kyrgyzstan AH7, AH61, AH65 Below Class III 370

10 Lao AH3 , AH11, AH12, AH13, AH15, AH16 Below Class III 656

11 Mongolia AH3, AH4, AH32 Below Class III 3486

12 Nepal AH 42 Below Class III/Bad 34

13 Pakistan AH2, AH4, AH7, AH 51 Below Class III / Bad 3144

14 RussiaAH4, AH6, AH7, AH8, AH30, AH31,

AH60/61/70 Below Class III / Bad 3640

15 Tajikistan AH7, AH65, AH66 Below Class III 343

16 Thailand AH1, AH15, AH16 Class III or Higher / Bad 68

17 Uzbekistan AH63 Below Class III 224

18 Vietnam AH14, AH15 Below Class III 190

Total 15842

Road surface condition and design standards in the following countries like Japan, South

Korea, Singapore, Malaysia, and Turkey are in good condition and satisfy Asian Highway

design standards. Data for countries like Democratic People’s Republic of Korea, Turkmeni stan,

Bhutan, Azerbaijan and Indonesia are found partially or not available therefore they were

dropped from analysis.

To estimate the cost of improvements and upgrading in above – mentioned table 2, we used

regression cost model at a regional level of an alysis. The results of predicted cost of these

upgrading and improvement works are displayed in Table 3.


CT 2

Table 3. Predicted cost of upgrading and improvement work

No. CountryPavementWidth (m) Road Upgrade/Improvements Expected Output

Total Length(km)

Total Cost(million US $

2002)1Armenia 6 - 7 Reconstruction Bituminous Improved Condition 138 23.3

7 -14 Reconstruction Bituminous Improved Condition 248 47.42Bangladesh < 4.5 Widening Adding Bituminous 2L and Recon Class II 100 57.5

4.5 – 6 Widening Adding Bituminous 1L and Recon Class II 350 76.03Cambodia 4.5 – 6 Widening Adding Bituminous 1L and Recon Class II 198 112.64China < 4.5 Upgrading Unsealed to Bituminous Class II 67 10.5

4.5 – 6 Upgrading Unsealed to Bituminous Class II 475 87.45Georgia 6 – 7 Reconstruction Bituminous Improved Condition 55 7.56India < 4.5 Widening Adding Bituminous 2L and Recon Class II 75 44.17Iran 7 – 14 Reconstruction Bituminous Improved Condition 1,042 199.7

6 – 7 Reconstruction Bituminous Improved Condition 42 7.18Kazakhstan 6 – 7 Upgrading Unsealed to Bituminous Class II 743 153.4

< 4.5 New Construction 2L Highway Class II 154 147.29Kyrgyzstan 7 – 14 Upgrading Unsealed to Bituminous Class I 370 91.4

10Lao 7 – 14 Reconstruction Bituminous Condition Improvement 244 42.46 – 7 Reconstruction Bituminous Condition Improvement 44 6.86 – 7 Upgrading Unsealed to Bituminous Class II 292 55.86 – 7 New Construction 2L Highway Class II 76 65.7

11Mongolia < 4.5 New Construction 2L Highway Class II 3,070 2,431.5< 4.5 Upgrading Unsealed to Bituminous Class II 416 57.1

12Nepal 4.5 – 6 Widening Adding Bituminous 1L and Recon Class II 26 5.26 – 7 Reconstruction Bituminous Condition Improvement 8 1.3

13Pakistan < 4.5 Widening Adding Bituminous 2L and Recon Class II 1,174 736.06 – 7 Reconstruction Bituminous Condition Improvement 1,042 196.57 – 14 Reconstruction Bituminous Condition Improvement 928 198.4

14Russia 7 -14 Upgrading Unsealed to Bituminous Class II 882 188.46 - 7 New Construction 2L Highway Class II 89 77.6< 4.5 New Construction 2L Highway Class II 876 764.37-14 Reconstruction Bituminous Condition Improvement 1,793 307.8

15Tajikistan < 4.5 New Construction 2L Highway Class II 48 46.26 - 7 Upgrading Unsealed to Bituminous Class II 278 57.87 -14 Upgrading Unsealed to Bituminous Class II 17 4.0

16Thailand > 14 Reconstruction Concrete Condition Improvement 40 7.66 -7 Reconstruction Bituminous Condition Improvement 18 2.3> 14 Reconstruction Bituminous Condition Improvement 10 1.7

17Uzbekistan 7 -14 Upgrading Unsealed to Bituminous Condition Improvement 224 56.518Vietnam 4.5 - 6 Widening Adding Bituminous 1L and Recon Class II 53 9.6

< 4.5 Widening Adding Bituminous 2L and Recon Class II 137 65.7Total 15,842 6,451.3

Total cost of upgrading and improvement works for 15,842 km in 18 countries would cost

about 6.4 billion US dollars. Figure 4 depicts the highway routes (thick red line) that are belong

to upgrading and road improvements in AHN. These routes play vital role in trade between Asia

and Europe.


CT 2

3.2 Intra – regional trade

Although 18 continental countries have enormously increased their overall trade over thepast several decades, intra-regional trade was still only 12 per cent of their total trade in 2005.


CT 2

Figure 5 shows the ratio of intraregional trade 1 to overall trade for countries in our sampl e. Theshare varies wildly, between 7.7 percent for China to 77 percent for Kyrgyzstan. There are twofacts worth noting from the graph. First, the share of intra -regional trade for small countries ismuch higher than for big countries. This is consistent with overall tendency for big countries tobe less open than small countries. Second, Former Soviet Union countries have much higherintra-regional trade on average. This can be attributed to the legacy of Soviet Union with itsclose trade and production networks among countries.

Figure 5. Share intra-regional trade in total trade, % (2005)

There are numerous reasons for relatively low levels of intra -regional trade in continentalAsia. Political and historical tensions certai nly have played a role, as well as the attractivenessof North American and European markets as a destination for exports products and source oftechnological imports.

But unfavorable geographical factors and low quality of transport infrastructure have al soimpaired intra-regional trade to a great extent. Vast and difficult terrain, especially in the innerpart of the Asian continent, has made overland trade among continental countries much lessprofitable. Due to the lack of adequate level of transport in frastructure, shipping goods from onecountry to another in the region through overland transport networks might be more expensivethan shipping from the region to North America and Europe through sea transport.

3.3. The gravity model approach

This paper adopts a gravity model approach to study the impact of AHN road upgrade ontrade. The gravity model originally stems from Newtonian physics, which simply states that theattraction between two physical objects is proportional to their masses, but inversely related to

1 Intra-regional trade is defined as trade among 18 countries in our sample.

0

10

20

30

40

50

60

70

80

Arm

enia

Ban

glad

esh

Cam

bodi

a

Chi

na

Geo

rgia

Indi

a

Iran

Kaz

akhs

tan

Kyr

gyzs

tan

Lao

Mon

goli

a

Nep

al

Paki

stan

Rus

sia

Taj

ikis

tan

Tha

ilan

d

Uzb

ekis

tan

Vie

tnam


CT 2

the distance between them. This paradigm has long been disregarded by economists due to itslack of theoretical foundation. However, due to successive works of various economists, it hasbeen gradually developed into a systematic economic model with a strong economic foundation.Anderson (1979) derived the gravity equation from monopolistic competition setting. Helpmanand Krugman (1985) showed that the basic gravity equation could be derived from thedifferentiated products trade. It is a theory that suggests that flows of goods depend on thedemand in the importing country and the supply of differentiated products from the exportingcountry. Deardorff (1995) showed that the gravity model is also consistent with Hecksher -Ohlininternational trade theory.

The gravity model has also been extensively utilized in empirical economic literature. Thusit was applied in estimation of bilateral trade flows, FDI flows, and equity flows. For example,Frankel (1997) used the gravity model approach to explain the factors affecting the formation oftrade blocs. In his standard gravity model, bilateral trade was explained by variables such asGNP, per capita GNP, distance, adjacency, language, and trading blocs. Gravity models havealso been used to explain determinants of FDI and equity flows. Kawai and Urata (1998) used agravity model to investigate the relationship between trade and FDI using Japanese data at theindustry level. FDI and trade were found to be generally complementary to each other. Portesand Rey (2000) also adopted a gravity model to study factors affecting equity flows among 14developed economies. Their empirical results demonstrated that market size, openness,efficiency of transactions, and distance are the most important determinants of b ilateral equityflows.

In recent years gravity model has increasingly been utilized in analyzing the impact ofinfrastructure on trade. Majority of studies show that transportation infrastructure quality hassignificant and robust impact on overall transpo rt costs. Notable examples include Redding andVenables (2004), Limao and Venables (2001), Coulibaly and Fontagné (2004), Martínez -Zarzoso and Nowak-Lehmann (2006), Buys, Deichmann and Wheeler (2006), Shepherd andWilson (2006) and others.

In particular, Redding and Venables (2001) use a ratio of roads to area as a proxy forquality of infrastructure and find that low infrastructure quality is the main factor behind lowtrade in Sub-Saharan Africa. Coulibaly and Fontagné (2004) study determinants of trade incountries belonging to the West African Economic and Monetary Union and find that paving allinter-state roads would increase trade by a factor of 3, and crossing a transit country reducesbilateral trade flows by 6%. Buys, Deichmann and Wheeler (2006) first estimate the costs ofinitial upgrading Sub-Saharan interstate road network as 20 billion dollars and 1 billion dollarsas cost of annual maintenance. Then they proceed to estimate the potential beneficial impact ofcontinental road network upgrading on overland trade as about $250 billion over 15 years.Limão and Venables (2001) estimate that poor infrastructure account for 40 percent of transportcosts for coastal countries and 60 per cent for landlocked countries.

In all of these studies the gravity model framework serves as a workhorse to estimate


CT 2

impact of infrastructure upgrading on trade. The attractiveness of the gravity model is that itallows us to address the specific questions in mind with regard to both the fundamentaleconomic and institutional determinants of trade in continental Asia. First, what are thefundamental determinants of trade? Are traditional variables of gravity models such aseconomic size, distance, tariff and common border significant explanatory variables? Second,does infrastructure quality matter in facilitating trade between among Asian countries? Thispaper offers some quantitative simulations to illustrate how the improvement of transportinfrastructure and reduction of tariff barriers can stimulate trade and econo mic development

3.4. Econometric specification and data description

Following the empirical literature, we specify a simple version of gravity model for totaltrade, exports and imports. In each specification GDP variable enters the trade regression in aproduct form. As a result, the gravity model for total trade takes the following form:

ijtj7i6j5i4ij3ij2jtit1ijt RRTarTarB)D(Ln)YY(LnLnT (2)

where Tijt indicates trade between country i and country j at time t, Y it and Yjt are realGDPs of country i and j, representing economic ma ss, Dij is distance between capital cities and

B is a common border dummy, ji TarandTar are tariff rates in country i and j, respectively.

Finally, ji RandR represent road quality indexes in country i and j, respectively.

Vast trade literature predicts expected signs and sometimes magnitudes of coefficients inequation (2). In particular, theory predicts that larger economic mass is associated with highervolumes of trade. Distance, as a proxy for transportation cost, is expected t o have a negativesign. Common border dummy is expected to have a positive sign. Trade is expected to have anegative relationship with tariff barriers and a positive relationship with road quality index.

Distance, calculated as a surface distance between capital cities according to latitude andlongitude (Wall, 1999; Raballand, 2003; Rose and Wincoop, 2001), is considered as proxy fortransportation cost in a borderless world. Border effect is expressed by inclusion of commonborder dummy (Rose and Wincoop, 2001; Rose, 2002; Breuss and Egger, 1999; Frankel andRose, 2001). The problem with simple great circle distance variable is that it does not fullycapture high transportation costs due to natural geographical location of landlocked and remotecountries. Transportation costs are usually affected by border delays (type of a non -tariffbarrier). To capture this peculiar feature of transportation cost we also include common borderdummy variable. Following Raballand (2003) we assumed for two coastal countri es there is aone border, and only for landlocked countries this variable is equal to one.

But even with distance and common border dummy variables, one cannot be sure that shetakes into account all complexities of transportation costs. One of the most i mportant factors foroverall transportation costs is the quality of transport infrastructure. Usually, the higher thequality of infrastructure the lower is the transportation costs and higher incentives for trade.Bougheas et al (1999) utilized stock of p ublic capital and length of motorway network and


CT 2

predicted a positive relationship between the level of infrastructure and the volume of trade.Limao and Venables (2001) developed unique infrastructure composite index, as a totalinfrastructure stock (roads, paved roads, telephones and railway networks) divided by the totalpopulation. But they excluded all transition economies in FSU and Europe due to missing datafor own and transit infrastructure. Unfortunately, lack of data for Asian countries madecalculation of composite index impossible. Instead, we utilized road quality index as anadditional.

The model is estimated for 19 Asian countries over the period 1995 -2004. Aggregatebilateral trade data are from International Monetary Fund’s Direction of Tr ade Statistics(DOTS) database. Data on GDP are taken from World bank’s World Development Indicators(WDI) database. Weighted average tariff rates are taken from Trade Analysis and InformationSystem (TRAINS) database, maintained by The United Nations Conf erence on Trade andDevelopment (UNCTAD). Except dummy variables, all variables are in logarithmical form.

3.5 Results and discussions

Table 4 shows estimation results. We consecutively estimate equations for trade, importsand exports. In the trade regression we follow Baldwin and Taglioni (2006) and use the productof real GDP.

Table 4. Gravity model estimations

Dependent variable Trade Exports Imports

Product of GDP 1.06 1.05 1.02[0.02]** [0.02]** [0.02]**

Distance -1.63 -1.66 -1.42[0.07]** [0.07]** [0.08]**

Common border dummy 1.40 1.36 1.62[0.12]** [0.14]** [0.13]**

Road quality index of country i -0.06 0.16 -0.16[0.08] [0.07]* [0.09]

Road quality index of country j 0.44 -0.06 0.79[0.08]** [0.09] [0.09]**

Average tariff rate of country i -0.23 -0.04 -0.37[0.05]** [0.05] [0.05]**

Average tariff rate of country j -0.20 -0.31 -0.15[0.04]** [0.04]** [0.05]**

Constant -35.64 -34.85 -37.04[0.77]** [0.78]** [0.90]**

Observations 2069 1920 1917R-squared 0.74 0.72 0.67Robust standard errors in brackets

* significant at 5%; ** significant at 1%


CT 2

The gravity model fits the data well and produces theoretically correct and economically

significant coefficients. As expected, the coefficients of real GDP are statistically significant,

slightly above 1 in all specifications. The coefficient of distance is negative, highly significant.

In elasticity terms it shows that 1 percent increase in distance is associated w ith a -1.65 percent

decline in trade and exports, and around 1.42 percent decrease in imports. Common border

dummy is also highly significant and positive. The point estimates of common border dummy

indicates that if countries share common border, they tra de with other 4 times more

(exp(1.4)=4.05). This effect is even stronger for imports. The coefficient of common border

dummy indicates that there is a lot of potential to increase overland trade, especially between

countries with common borders.

The coefficients of tariff rates are negative and statistically significant, and their magnitude

ranges from -0.15 to -0.37. Trade equation indicates that, say, 10 percent reduction in tariff rates

increase overall trade by about 2 percent. Taking into account tha t in most instances tariffs are

already quite low and they cannot be drastically decreased, it becomes clear that further

reductions of tariff rates among continental Asian countries have limited impact on trade.

On the other hand, the road quality index shows that a good transport infrastructure can

greatly facilitate trade. In particular, the positive coefficient of road quality index in trade

regression – 0.44, which is statistically highly significant, indicates that improvement of the

quality of overland roads can boost trade significantly. For example, if Nepal improves quality

of its roads index from 31 to 50 (48 percent improvement in logarithmic terms), it can expect its

overall trade increase by 21 percent (0.48 multiplied by 0.44), or by 285 mill ion US dollars

annually.

Based on the gravity model estimations, we can estimate the impact of road quality

upgrading on intra-regional trade. Table 5 shows the results of this exercise under two

scenarios: pessimistic and optimistic. Under pessimistic sc enario it is assumed that major road

improvement efforts will upgrade road quality index to 50 percent throughout the region. It is

pessimistic scenario because it assumes that major routes surface conditions will be improved

without any upgrading. Armenia , Georgia, Iran, and Thailand already have road quality grade of

50, so it is assumed that they will not benefit directly in terms of trade expansion due to road

improvement.

Scenario 2 assumes more ambitious criterion, namely, continental Asian count ries upgrade

their interstate road quality to 75. In terms of AHN classification, Table 5 shows the net impact

of road improvement on trade under two scenarios. Under Scenario 1 the total intra -regional

trade will increase about 20 percent to 48.7 bln US d ollars annually, while Scenario 2 predicts

that trade will increase by about 35 percent to 89.5 billion US dollars annually. The main

beneficiaries of the road improvement will be China, Russia, India, and Vietnam smaller

countries will also benefit from overall increase in trade due to the improvement in transport

infrastructure.


CT 2

Table 5. Net effect of road upgrading on trade

IV. CONCLUSION

Introduction level of analysis shed light on issue like efficient data usage in existingdatabase. It is important to mention that missing data is almost unavoidable part of datacollection process. So when some of the data times are missing then the common approach is touse available data. We recommended to separate data items in ROCKS in accordance to level ofanalysis such as regional, country and project level to utilize available data efficiently. Inregional level, the purpose was to select plausible data items with the largest number ofobservations which later were used for cost model development. In case of project level, the aimwas to select data items which represent project details in depth therefore number ofobservations were very limited. Country level analysis lies in between regional and project levelof analyses and contains some of regional level data items as well as project level data itemswhich lead to utilize more data with some project details in cost model development.

Regression cost models are widely used because they are easy and relatively fast toimplement, various well-documented procedures are available, and finally cost estimators preferto use regression models rather than analytical tools such neural networks because regressionmodels are well defined and mathematically explained whereas neural network works muchmore like black box.

Scenario 1(up to 50%)

Scenario 2 (upto 75%)

Armenia 50.0 530.4 - 94.6Bangladesh 25.0 4139.53 1,262.5 2,001.0Cambodia 25.0 819.74 250.0 396.3China 25.0 81683.26 24,912.1 39,484.8Georgia 50.0 881.01 - 157.2India 25.0 20066.49 6,120.0 9,699.9Iran 50.0 15022.54 - 2,680.1Kazakhstan 25.0 14964.24 4,563.9 7,233.6Kyrgyzstan 25.0 1436.8 438.2 694.5Lao 36.0 1022.67 148.1 330.6Mongolia 25.0 1042.75 318.0 504.1Nepal 30.9 1343.58 284.9 524.6Pakistan 40.7 4103.05 373.1 1,105.1Russia 37.3 41246.66 5,311.0 12,669.6Tajikistan 25.0 1021.1 311.4 493.6Thailand 50.0 24352.18 - 4,344.5Uzbekistan 25.0 3147.66 960.0 1,521.5Vietnam 25.0 11461.15 3,495.5 5,540.2

Total 228284.8 48,748.6 89,475.7

Net Effect of RoadUpgrading, mln US dollarsInter-regional

trade in 2004,mln US dollars

Road QualityIndexCountry


CT 2

It was estimated cost of upgrading and improvement works costs of sub -network of AHNusing World Bank’s ROCKS database. This provided initial perspec tive on the size of lumpyinvestment required to improve road condition in AHN. It was estimated that approximately 6.5billion US dollars is required to upgrade roads and improve existing surface condition of theselected sub-network with total length of 15,842 km of AHN. The gravity model approachexplained how big the trade expansion will increase. The net impact of road improvement ontrade under two scenarios was considered. In scenario 1, the total intra -regional trade willincrease about 20 percent to 48.7 bln US dollars annually, while Scenario 2 predicts that tradewill increase by about 35 percent to 89.5 billion US dollars annually. The main beneficiaries ofthe road improvement will be China, Russia, India, and Vietnam smaller countries will als obenefit from overall increase in trade due to the improvement in transport infrastructure. Priorityof road upgrading in each country is suggested to be carried out in a way that first roadcondition improvements need to be done after that upgrading to h igher class necessary to carryout. But it must also fit to each country’s network strategic plan. The results show that roadquality is positively associated with trade, while tariff rates are negatively correlated with trade.These results are consistent with transportation economics viewpoint that road upgradingdecreases transportation costs such as vehicle operation cost (fuel consumption, spare, etc) anduser cost (travel time). Higher traffic volumes allow the policymakers to take advantage thehigher trade volumes and decrease tariff rates further. Future research will focus on tradeexpansion among all AHN member countries taking into account other modes of transportation.

Reference

[1]. Adeli H. and Wu M. Regularization neural network for constru ction cost estimation, Journal ofConstruction Engineering and Management, ASCE 124(1), pp 18 -24, 1998

[2]. Allison S., Allison D. Multiple Regression: A Primer. Forge Press. 1999

[3]. Al-Tabtabai, H., Alex, A., and Tantash, M. Preliminary Cost Estimation of Highway ConstructionUsing Neural Networks, Cost Engineering, Association for the Advancement of Cost Engineering,Morgantown, WV. 41(3), 19-24, 1999

[4]. Archondo-Callao R., Bhandari A. Road Costs Knowledge System – ROCKS – Version 2.2, UserManual, World Bank, 2004

[5]. Anderson, J.E.A. 1979. A theoretical foundation for the gravity model. American Economic

Review, Vol. 69, pp. 106–16.

[6]. Andrew K. Rose & Eric van Wincoop , 2001. "National Money as a Barrier to International Trade: TheReal Case for Currency Union," American Economic Review, vol. 91(2), pages 386 -390, May.

[7]. Asian Highway Network Handbook, United Nation, New York, 2003.

[8]. Asian Highway Design Standard, UNESCAP, 2004

http://www.unescap.org/ttdw/common/tis/ah/AnnexII -E.pdf. Accessed: June 10, 2008


CT 2

[9]. Asian Highway Network Database (AHND), UNESCAP, 2004http://www.unescap.org/ttdw/common/tis/ah/Member%20countries.asp Accessed: June 10, 2008

[10]. Bunke, O and B. Droge . Bootstrap and cross-validation estimates of prediction error for linearregression models. Ann. Stat. 12, 1400 -1424, 1984

[11]. Baldwin, Richard, and Daria T aglioni (2006). "Gravity for Dummies and Dummies for GravityEquations," NBER Working Papers 12516

[12]. Bougheas, Spiros, Panicos O. Demetriades and Edgar L.W. Morgenroth (1999), “Infrastructure,Transport Costs and Trade”, Journal of International Econom ics, Vol. 47, 169-189.

[13]. Breuss, Fritz, and Peter Egger (1999), "How Reliable Are Estimations of East -West TradePotentials Based on Cross-Section Gravity Analyses?", Journal Empirica 26 (2), 81 -94

[14]. Brikes D, Birkes D, Dodge Y . Alternative Methods of Regression. Wiley-Interscience. 1993

[15]. Buys, Piet, Uwe Deichmann and David Wheeler (2006), Road Network Upgrading and OverlandTrade Expansion in Sub-Saharan Africa, World Bank Policy Research Working Paper 4097, December

[16]. Carroll J, Raymond J, Ruppert D. Transformation and Weighting in Regression. CRC Press. 1988.

[17]. Coulibaly and Fontagné (2005), South-South Trade: Geography Matters, Journal of AfricanEconomies, Vol. 15, Number 2, pp. 313 -341

[18]. Deardorff, A.V. (1998). Determinants of Bilateral Trade: Does Gravity Work in a NeoclassicalWorld?, in: J. Frankel (ed.), The Regionalization of the World Economy. University of Chicago Press.

[19]. Emsley M, Lowe D., Duff A, Harding A, Hickson A . Data Modeling and the application of a neu ralnetwork approach to the prediction of total construction cost. Journal of Construction ManagementEconomy, 20, pp.465-72, 2002

[20]. Frank H. Regression Modeling Strategies: with Applications to Linear Models, Logistic Regressionand Survival Analysis. Springer. 2001

[21]. Frankel, Jeffrey A. (1997). Regional Trading Blocs in the World Economic System, Institute forInternational Economics, Washington, D.C.

[22]. Frankel, Jeffrey, and Andrew Rose (2000). "An Estimate of the Effect of Currency Union s on Tradeand Growth," for a Conference on Currency Unions, organized by Alberto Alesina and Robert Barro,Hoover Institution, Stanford University, May 19 -20.

[23]. Flyvbjerg B. H., Mette S., and Buhl S . Underestimating costs in public works projects, er ror or lie,Journal of the American Planning Association, Vol. 68, No.3, pp 279 -292, 2002

[24]. Geisser, S. The predictive sample reuse method with applications. Journal of American StatisticsAssociation, 70, 320-328. 1975

[25]. Gwang-Hee K, Sang-Hoon A, Kyung-In K. Comparison of construction cost estimating modelsbased on regression analysis, neural networks, and case -based reasoning. Journal Building andEnvironment. Vol. 39, pp 1235-1242, 2004

[26]. Healey, J. M., “Errors in Project Cost Estimates,” Indian Economic Journal, Vol. 12, July- Sept. 1964.

[27]. Hegazy T. and Ayed A . Neural network model for parametric cost estimation of highway projects,Journal of Construction Engineering and Management, ASCE 124(3), pp 210 -218, 1998

http://www.unescap.org/ttdw/common/tis/ah/AnnexII-E.pdf

http://www.unescap.org/ttdw/common/tis/ah/Member%20countries.asp


CT 2

[28]. Helpman, E. and Krugman, P.R. (1985). Market structure and foreign trade. Increasing returns,imperfect competition, and the international economy, MIT Press, Cambridge, MA

[29]. Kawai, Masahiro and Shujiro Urata (1998), “Are Trade and Investment Substitutes orComplements? An Empirical Analysis of Japanese manufacturing Industries,” in Lee, Hiro, Roland andW. David, Economic Development and Cooperation in the Pacific Basin: Trade, Investment, andEnvironmental Issues, Cambridge; New York and Melbourne: Cambridge Univers ity Press, 1998.

[30]. Levinson, D. and Karamalaputi , Ramachandra. Predicting the Construction of New Highway Links.Journal of Transportation and Statistics Vol. 6(2/3) 81 –89, 2003

[31]. Lewis-Beck M. Applied Regression: An Introduction. Sage Publication Inc. 1980.

[32]. Miles J., Shevlin M. Applying Regression and Correlation: A Guide for Students and Researchers.Sage Publication Inc. 2001.

[33]. Limão, N. and Venables, A.J . (2001), "Infrastructure, Geographical Disadvantage, Transport Costsand Trade", World Bank Economic Review, 15, 451 -479.

[34]. Martínez-Zarzoso and Nowak-Lehmann (2006), Is Distance a Good Proxy for Transport Costs? TheCase of Competing Transport Modes, CeGE -Discussion Paper 54, May

[35]. Pearce A. R., Gregory, Rita A., and Willia ms L. Range Estimating for risk management usingartificial neural networks, Journal of Parametric Estimating, 14(1), pp 88 -93. 1996

[36]. Portes, Richard and Hélène Rey (2000). "The Determinants of Cross -Border Equity Flows: TheGeography of Information," Center for International and Development Economics Research (CIDER)Working Papers C00-111, University of California at Berkeley

[37]. Raballand, G. (2003). The determinants of the negative impact of land -lockedness on trade: anempirical investigation through the Central Asian case. ROSES, University of Paris 1, mimeo May 2003.

[38]. Redding, S., Venables, A. J . 2004. Economic geography and international inequality. Journal ofInternational Economics, 62(1): 53 -82.

[39]. ROCKS, Road Costs Knowledge S ystem (v. 2.2), (The World Bank, Transport Unit – TUDTR),2004. http://www.worldbank.org/transport/roads/rd_tools/rocks_main.htm . Accessed June 10, 2008

[40]. Rose, A.K. 2002. Do WTO members have more liberal trade policy? NBER Working Paper #9273

[41]. Sanders, S., and Maxwell, R., “Preliminary Estimating for Heavy Construction,” AACE Transactions,Association for the Advancement of Cost Engineering International, Morgantown , WV, 1992.

[42]. Stone, M. Cross-validatory choice and assessment of statistical predictions. Journal of the RoyalStatistical Society. Series B (Methodological), Vol. 36, No.2. pp. 111 -147. 1974

[43]. Shepherd, Ben, and John S. Wilson (2006), Road Infrastructure in Europe and Central Asia: DoesNetwork Quality Affect Trade? World Bank Policy Research Working Paper 4104, December 2006

[44]. Trost, S. , and Oberlender, G. Predicting accuracy of early cost estimates using factor analysis and multivariateregression. Journal of Construction Engineering Management, Vol. 129, No. 2, 2003, pp. 198 -204.

[45]. William D., Stanley F . Multiple Regression in Practice. Sage Publications Inc. 1985

[46]. Wall, Howard J. (1999). "Using the gravity model to estimate the costs of

protection,"Review, Federal Reserve Bank of St. Louis, issue Jan, pages 33 -40

http://www.worldbank.org/transport/roads/rd_tools/rocks_main.htm


CT 2

I. INTRODUCTION OF GLEM

1.1. Outline of Generalized LimitEquilibrium Method for Static Problems

Developing the limit equilibrium method(LEM), the authors have proposed thegeneralized limit equilibrium method, whichcan obtain a force field satisfying theequilibrium condition on every soil block andfailure condition both on the bottom plane andthe inter-block plane of the block (Figure. 1).Consequently, the GLEM can be consideredas an approximation method to obtain thenecessary condition of SLM. This method hasthe following features:

Quadrangle or triangle blocks as well asslices can be treated.

Safety factors are defined both on themain sliding surface and on inter -blockplanes.

Circular sliding surface as well as non -circular sliding surface can be treated.

All types of plasticity problems can beexpressed in a single formulation.

The GLEM can be applied to analyze all

types of static problems such as slope

stability, earth pressure, and bearing capacity.

For every static plastic problem, a num ber of

DEVELOPMENT OF GENERALIZED LIMIT EQUILIBRIUMMETHOD FOR THE FAILURE OF RETAINING WALLS

UNDER SEISMIC LOADINGS

XUAN BINH LUONG; VIKHONE SAYNHAVONGTHANH THUY HOANGDepartment of Civil EngineeringUniversity of Transport and Communications , VietnamMEIKETSU ENOKIDepartment of Civil EngineeringTottori University, Japan

Abstract: The classical theory of plasticity, represented by K ötter’s equation, has been

established for static problems [1]. Although this theory can be easily extended for dynamic

plasticity problems by introducing accelerations as inertia forces, up to now no researcher

has done this because of the difficulty in determining the acceleration distribution within the

body when the failure occurs. The objective of this research is to develop Generalized Limit

Equilibrium Method (GLEM) with the introduction of continuity condition of acceleration to

investigate the following cases of potential failures of retaining structures: active failure,

foundation-like failure, and slope-like failure, under seismic loadings, where the GLEM is one

of the limit equilibrium methods proposed by Enoki at al . The theoretical formulation of the

method, the illustrative examples, and the comparisons between the results of the proposed

method and other methods are demonstrated.

Keywords: Earth pressure, earthquake, limit equilibrium method, slope failure, rigid -plastic


CT 2

examples of the calculation were taken with

the use of GLEM and many other methods.

The comparisons showed that the results

obtained by GLEM agree with those obtained

by theoretical analysis. Another paper by the

authors (Enoki et al., 1991) [2] sho uld be

referred to for the details of the GLEM.

1.2. Outline of Generalized Limit EquilibriumMethod for Dynamic Problems

Newmark [3] proposed, for the first time,a displacement analysis method to evaluatethe effects of earthquakes on the stability ofslopes. The method was then developed byChang [4] and other researchers [5]. InNewmark’s method, the critical state in whichthe failure begins to occur is determined bythe pseudo-static analysis. When theearthquake-induced acceleration exceeds thecritical value, the failed soil mass isconsidered to slide along the slip surface as awhole rigid body. The residual displacementcan be determined by integrating the relativeacceleration.

Based on Newmark’s concept, theauthors have been developing a metho d toanalyze the motion of earth structures. In thismethod, the computing model of a foundationis shown in figure 1. When the earthquake-induced acceleration reaches a certain criticalvalue, incipient failure occurs and many slipplanes appear within the body. This criticalacceleration can be obtained by ordinarypseudo-static analysis in which theacceleration of every part of the structure isthe same with the input acceleration. Whenthe seismic acceleration exceeds the criticalvalue, the failed soil mass is then consideredas a rigid-plastic block system, in which thesurrounding surfaces of the block are just the

slip planes. The rigid blocks will moverelative to each other and to the base groundalong the slip planes. Across a slip plane(figure 2) the component of accelerationnormal to the slip plane is continuous, thecomponent parallel to the slip plane isdiscontinuous but the shear stress on the slipplane corresponding to the shear strength istransmitted. This continuity condition ofacceleration is combined with “GeneralizedLimit Equilibrium Method (GLEM) ” toanalyze the motion of the earth structures indynamic cases. The residual displacements ofevery block in both vertical and horizontaldirections can be computed by integratingtwice the relative acceleration of the block.

The proposed method permits theanalysis of all types of dynamic plasticproblems such as: bearing capacity andmotion of foundations, failure and motion ofslopes, earth pressure and motion of retainingwalls. Any types of sliding surface can betreated by dynamic GLEM.

Figure 1. A block system of a foundation

in earthquake motion and equilibrium

of the i - th block

Figure 1. A block system of a foundation inearthquake motion and equilibrium of

the i-thblock

i -Block numberi -Bottom plane numberi - Inter-block plane number

Foundation

12

i

n

12

i

i+1

n

n+1

1 2

i

n

n+1

v

h

vi’ hi’

v(n+1)’h(n+1)’

Bottom plane

Inter-block plane

B

i+1

Hi ViPiTi

i

i

i+1

i+1

Mig

Mihi’

Mivi’Vi+1

Si

Ri+1

Ri

Geometry and acting forces ofi-thblock


the i-thblock

i -Block numberi -Bottom plane numberii - Inter-block plane number

Foundation

12

i

n

12

i

i+1

n

n+1

1 2

i

n

n+1

v

h

vi’ hi’

v(n+1)’h(n+1)’

Bottom plane

Inter-block plane

B

i+1

Hi ViPiTi

i

i

i+1

i+1

Mig

Mihi’

Mivi’Vi+1

Si

Ri+1

Ri


i+1

Hi ViPiTi

i

i

i+1

i+1

Mig

Mihi’

Mivi’Vi+1

Si

Ri+1

Ri



CT 2

Figure 2. A sliding block model

II. FAILURE MECHANISMS OF RETAININGWALLS UNDER SEISMIC LOADINGS

Figures 3, 4 and 5 show the computingmodels of a retaining wall subjected toseismic loadings corresponding to three casesof failures: active failure (a), failure ofretaining wall as a foundation problem(foundation-like failure) (b), and failure of thewall as a slope problem (slope-like failure)(c). In mode (a), the wall is considered tomove outwards relative to sub -base, andcauses the active earth pressure. In mode (b),the base supporting the wall is fail ed and bothwall and sub-base slide outwards. In mode (c),both the sub-base and backfill are failed andthe system slides outwards. The failed soilmass is considered as a rigid-plastic blocksystem. Either triangular or quadrangularblocks can be used.

III. FORMULATION OF DYNAMIC GLEM FORTHE FAILURES OF RETAINING WALLS

Before the sliding occurs, theacceleration of every soil block is the same asthe acceleration of the sub-base. Theequilibrium equations of every block, thefailure conditions on the inter-block planesand bottom planes are used to obtain the forcefield as presented in the formulation ofdynamic GLEM, for the detail the [6] shouldbe refered to.

When the sliding occurs, the accelerations of

blocks are different from each other and from thesub-base. The equilibrium equations of everyblock, the failure conditions, and continuityconditions of acceleration on both inter -blockplanes and bottom planes are used. The numberof unknowns and the number of equations areshown in Table 1. The sliding acceleration ofthe wall is minimized to obtain the geometry ofthe sliding surface. The classical Newtonmethod is used herein to optimize thefunction value.

Figure 3. Active failure (a)

Figure 4. Foundation – like failure (b)

Figure 5. Slope – like failure (c)

1

H

TN

Bedrock

Sliding surface


n’’

’

n

Mg

Mv’Mh’

1

H

TN

Bedrock

Sliding surface


n’’

’

n

Mg

Mv’Mh’

1

H

TN

Bedrock

Sliding surface


n’’

’

n

Mg

Mv’Mh’

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’


Figure 4. Foundation-like failure (b)

i - Block numberi - Bottom plane numberi - Inter-block plane number

vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1

Figure 5. Slope-like failure (c)

v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’


vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’v

h

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1



vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1

i - Block numberi - Bottom plane numberii - Inter-block plane number

vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’


vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’v

h

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1



vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’


vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’v

h

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1



vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1


CT 2

IV. NUMERICAL EXAMPLE

A strip retaining wall with a mass of 25t,a height of 5m, and a width of 3m isconsidered. The frictional angle of backsurface of the wall is 11o. The backfill and soilbase have the parameters as = 32o, c =0.11tf/m2, = 1.6tf/m3. A sinusoidal wave isused as the input acceleration, which has thefrequency of 2Hz, the amplitude of horizontalcomponent is 6m/s2, the vertical componentequals zero. For the simplicity, the dilatancyangle is considered to be zero, and the surfaceof the backfill is horizontal. Four cases ofanalysis were carried out: case a1 - activefailure mode was taken with the frictionalangle of the bottom plane of the wall, , is17o; case a2 - active failure mode was takenwith the frictional angle of the bottom planeof the wall is 32o; case b - foundation-likefailure mode was taken; and case c - slope-like failure mode was taken.

The results of analyses are presented infigures 6 and 7. It can be seen from the figure6 that the solutions of the active failure modeare very differentwith the changeof . When =17o, the activefailure occursearlier thanfoundation-likefailure mode (caseb) and slope-likefailure (case c),and it is inopposite situationfor the case =32o. The graphalso indicates thatthe foundation-likefailure occurs laterthan slope-like

failure in this analysis. As stated in [6], the figure 7once again shows the comparison between theproposed method and Mononobe-Okabe method(M-O method) [7,8] for the dynamic earthpressures. It is clear to realize that, correspondingto the sliding process, the M-O method hasoverestimated the earth pressure.

V. EFFECT OF ROUGHNESS OF WALL-BOTTOM SURFACE

An investigation on the relation between thefrictional angle of the wall-bottom surface and thecritical acceleration, at which the sliding starts tooccur, was carried out. The analysis condition isthe same as the example above. The interrelation

between and the critical accelerationscorresponding to every failure mode is presented

in figure 8. In this analysis, when 25.47o, thefailure mode likely to happen is active failure.

When > 25.47o, the failure mode likely tohappen is slope-like failure. The wall seems to besafe with foundation-like failure.

Table 1. The number of equations and unknowns

Equations Unknowns

Equilibriumconditions

(a) (b) (c)On bottomplanes

(a) (b) (c)

In verticaldirection

n n+1 n+2 Normal forces n n n+1

In horizontaldirection

n n+1 n+2 Shear forces n n n+1

Failureconditions

On inter-blockplanes

On bottomplanes

n n n+1 Normal forces n-1 n+1 n+1

On inter-block planes

n-1 n+1 n+1 Shear forces n-1 n+1 n+1

ContinuityCondition ofAcceleration

Blockaccelerations

On bottomplanes

n n n+1 vi’ n n+1 n+2


n-1 n+1 n+1 hi’ n n+1 n+2

Total 6n-2 6n+4 6n+8 6n-2 6n+4 6n+8


CT 2

VI. EXAMPLE OF A SUPPOSED RETAININGWALL SUBJECTED TO THE NIIGATA KENCHUETSU EARTHQUAKE - 2004

A supposed retaining structure with

parameters as shown in Table 2 subjected to

the Niigata Ken Chuetsu Earthquake – 2004 is

considered. The cross-section of the structure

is assumed to oblique at an angle of about 23 o

to the E-W direction. So, the horizontal

component of the input acceleration is derived

from E-W and N-S accelerations. The vertical

component is just the U-D component of the

acceleration record.

It is supposed here that sliding occurs in

slope-like failure mechanism. Thus, the model

and block system for the problem are

presented in figure 9. The peak strength of

soil is applied to the duration from the

beginning of the shaking to the moment that

the first sliding finishes. Then the residual

strength is used for the remainder of the

shaking process. The wall bottom friction

angle is larger than 17o, therefore no relative

sliding occurs along this plane or the wall and

soil wedge right beneath the wall take the

same movement.

The analysis results including

accelerations of the wall and sub -base,

residual displacements of the wall, and

dynamic active earth pressure are presented in

figure 10. In order to see clearer the sliding

process, the acceleration data within the

duration from 19 s to 20 s are zoomed in. The

first critical state is reached at moment t =

19.19 s. The slip surface at this state is

obtained and considered as the actual slip

surface as shown in figure 10. This slip

surface is assumed to be unchanged during

sliding process. With the use of peak strength,

the horizontal critical acceleration for the first

sliding is greater than others. The sliding

occurs a number of times during shaking

process (figure 11a). Over the time of

shaking, the residual displacements of the

wall in horizontal and vertical directions are

computed and plotted in figure 11b.

Figure 11c shows the dynamic activeearth pressure obtained by the present methodand the M-O method. We can see once againhere that during sliding the M-O methodoverestimates the earth pressure. At somemoments, the overestimation of the M -Omethod is up to 130%.

Figure 6. Accelerations corresponding

to analyzed cases

Figure 7. Dynamic active earth pressure

- 6.5

- 4.5

- 2.5

- 0.5

1.5

3.5

5.5

0 0.1 0.2 0.3 0.4 0.5

Time (s)

Horiz

onta

l acc

eler

atio

n (m

/s2 )

- 8

- 5

- 2

1

4

7

0 0.1 0.2 0.3 0.4 0.5

Time (s)

Verti

cal a

ccel

erat

ion

(m/s

2 )Sliding start

Sliding end

case a1

case b

case a2

case c

input acceleration


case a1case a2

Mononobe-Okabe method

Figure 6. Accelerations corresponding toanalyzed cases

case a1

case a2

case b

case c

- 6.5

- 4.5

- 2.5

- 0.5

1.5

3.5

5.5

0 0.1 0.2 0.3 0.4 0.5

Time (s)

Horiz

onta

l acc

eler

atio

n (m

/s2 )

- 8

- 5

- 2

1

4

7

0 0.1 0.2 0.3 0.4 0.5

Time (s)

Verti

cal a

ccel

erat

ion

(m/s

2 )Sliding start

Sliding end

case a1

case b

case a2

case c

input acceleration

Sliding startSliding end

case a1

case b

case a2

case c

input acceleration


case a1case a2



case a1case a2



case a1case a2



case a1

case a2

case b

case c


case a1

case a2

case b

case c


case a1

case b

case a2

case c

input acceleration

0

50

100

150

200

250

0 0.1 0.2 0.3 0.4 0.5Time (s)

Act

ive

eart

h pr

essu

re (

kN/m

)


case a1case a2



case a1

case a2

case b

case c


case a1

case b

case a2

case c

input acceleration


case a1

case b

case a2

case c

input acceleration

0

50

100

150

200

250

0 0.1 0.2 0.3 0.4 0.5Time (s)

Act

ive

eart

h pr

essu

re (

kN/m

)


case a1case a2


0

50

100

150

200

250

0 0.1 0.2 0.3 0.4 0.5Time (s)

Act

ive

eart

h pr

essu

re (

kN/m

)


case a1case a2



case a1case a2



case a1

case a2

case b

case c


case a1

case a2

case b

case c


CT 2

-8.0-4.00.04.08.0

17 18 19 20 21 22 23 24 25 26 27

Horiz

ontal

acc.,

m/s

2

-4.0-2.00.02.04.0

17 18 19 20 21 22 23 24 25 26 27

Vertic

al ac

c., m

/2 Sub-base

Wall

1st Critical state

363 339 339

-34 -60 -0.2

No sliding No slidingSliding Sliding Sliding

No slidingNo sliding

Horiz

ontal

acc.,

102

gal

Vertic

al acc.

, 102

gal

Figure 8. Bottom surface roughness of the wall

and failure modes

Table 2. Analysis condition for motion of a retainingstructure using slope-like failure mechanism

Wall mass, M (t) 20.0Back surface friction of wall, (o) 10.0

Bottom surface friction of wall, (o) >17.0

Soil density, (t/m3) 2.0

Internal friction angle, (o) 25.0Cohesion, c (tf/m2) 1.0

Figure 9. Block system for analyzing

motion of a retaining structure with the use of

slope-like failure mechanism

Figure 10. Geometry of block system at the firstcritical state

a. Input accelerations and sliding accelerations of

the retaining wall

0

2

4

6

17 22 27 32Wall- bottom frictional angle (o)

Criti

cal a

ccel

erat

ion

(m/s

2 )

Figure 8. Bottom surface roughness of thewall and failure modes

foundation-like failure

slope-like failureactive failure

=25.47o

0

2

4

6


Criti

cal a

ccel

erat

ion

(m/s

2 )




=25.47o

12 3 4

5

6

H=5

m

B=3 m

Wall

Sub -base

4.9596 m

B=3 m

H=5

m

2.6519 m1.0957 m

-0.15

-0.1

-0.05

0

17 18 19 20 21 22 23 24 25 26 27

Time, s

Residu

al disp., m Horizontal disp.

Vertical disp.

Residua

ldisp

., 102

cm

Target = 10.85 cm

Target = 1.72 cm

b. Residual displacements of the retaining wall

Dynami

cPae

,tf

05

101520

17 18 19 20 21 22 23 24 25 26 27Time, s

Present methodM-O method

c. Dynamic earth pressures

Figure 11. Results of analysis for motion of a

retaining structure with the use of slope -like failure

mechanism

Time,s


CT 2

VII. CONCLUSIONS

The dynamic GLEM has been developed

to investigate the failures of a retaining wall

under seismic loadings with three mechanisms

such as active failure, foundation-like failure,

and slope-like failure. Not only the dynamic

earth pressure but the motion of the retaining

wall can be obtained also.

The foundation-like mechanism seems to

be unlikely to happen. When the bottom

surface roughness is low, the failure mode

likely to happen is slope-like failure. When

the roughness is high, the failure mode likely

to happen is active failure.

When sliding has not occurred, the active

earth pressures obtained by the proposed

method and the M-O method are almost the

same. During sliding process, the M-O method

seems to overestimate the active earth pressure

in comparison with the proposed method.

VIII. ACKNOWLEDGMENTS

The seismographic data used in this

research were obtained from the home page of

the Japanese National Research Institute for

Earth Science and Disaster Prevention (NIED)

[http://www.kik.bosai.go.jp].

Reference

[1]. V. V. Sokolovsky, “Static of soil media,”

Trans. Jones, D. H. and Schofield, A. N., Lond on

(Butterworth), 1956.

[2]. M. Enoki, N. Yagi, R. Yatabe, and E.

Ichimoto, “Generalized limit equilibrium method

and its relation to slip line method, ” Soils and

Foundations, Japanese Soc. of Soil Mech. and

Found. Engrg., vol. 31, no. 2, pp. 1–13, June 1991.

[3]. N. M. Newmark, “Effects of earthquakes on

dams and embankments, Fifth Rankine Lecture,”

Gétechnique, no. 2, pp. 139–160, 1965.

[4]. C. J. Chang, W. F. Chen, and J. T. P. Yao ,

“Seismic displacements in slopes by limit

analysis,” J. of Geotech. Engrg, ASCE, vol. 110,

no. 7, pp. 860–874, July 1984.

[5]. R. L. Michalowski and L. You,

“Displacements of reinforced slopes subjected to

seismic loads,” J. Geotech. And Geoenviron.

Engrg., vol. 126, no. 8, pp. 685–694, August 2000.

[6]. M. Enoki, B. X. Luong, N. Okabe and K. Itou,

“Dynamic Theory of Rigid-Plasticity,” J. Soil

Dynamics and Earthquake Engrng ., no. 25, pp.

635–647, 2005.

[7]. N. Monobe and H. Matsuo , “On the

determination of earth pressures during

earthquakes,” in Proc. World Engrg. Conf. , 9, 176,

1929.

[8]. S. Okabe, “General theory on earth pressure

and seismic stability of retaining wall and dam, ” J.

Japanese Soc. of Civ. Engnrs. , vol. 10-6, pp.

1277–1323, 1924

http://www.kik.bosai.go.jp

INTERNATIONAL COOPERATION ISSUE OF TRANSPORTATI ON – Especial Issue – No.01172

CT 2

I. OBJECTIVES

Transportation construction projects are

mainly conducted for the social and economicbenefits purpose. However, some of the

projects were not estimated and defined in a

unite way. The article presents a method todefine the benefits. This method can be used

as a reference with respect to the process of

planning and appraising of road projects.

II. CONTENT

The social and economic benefits of aproject are gains that the project brings tosociety and economics. These advantages arenormally estimated by making comparisonbetween a situation of having the project andthat of without the project.

A transportation construction project hasmajor benefits as the following: (1) gains ofreducing operation costs; (2) gains of savingpassenger’s time (and goods); (3) gains ofdecreasing number of accidents; (4) gains ofenvironmental pollution mitigation. Besides,when assessing the efficiency of projects,changing distance of transportation andmaintenance costs are also considered.

Profits of mitigating environmentalcontamination will be presented in section [1],whereas, the content of the article focuses onother types of benefits.

1. A method to define gains of reducingvehicle operation costs

Vehicle Operation Cost (VOC)comprises many costs of fuels and damages(motors, tires…). Of road constructions, thesementioned costs depend on road condition(geometrical structure, road surface…);activities of divers and traffic controlcapacity. VOC usually shows higher valuewith respect to sloping and rough roads. Amethod to determine VOC is presented insections [2; 3].

One of the most ultimate target whenconstructing road structures is to reduce thevalue of VOC. Benefits, contained bydecreasing vehicle operation costs accountedat t-th year, are calculated as the equationbelow.

)inewVOC

iold(VOCnew.L

itN

m

1i365.

1tB

(VND/year) (1)

A METHOD OF DETERMINING SOCIAL AND ECONOMICBENEFITS OF TRANSPORTATION CONSTRUCTION PROJECTS

Dr. BUI NGOC TOANUniversity of Transport and Communications

Summary: This paper presents an approach of determining basic social and economic

benefits of road projects


CT 2

Where, itN - annual average daily traffic

of the i-th type of vehicle in t -th year

(units/day)

m - number of the vehicle unit (inluding

good and passengers transportations) (units)

Lnew - transport length of a new orreconstructe road (km)

inew

iold VOC;VOC - vehilce operation

costs in two cases (having the project andwithout the project) (VND/vehicle.km).

2. A method to define gains ofdecreasing passengers travelling time

Time has its value and the human time

value can be measured. Transportation

construction projects aim to reduce time for

travelling of passengers. Many studiesdemonstrates that saved time valu e of

passengers largely depends on individuals

(purpose of the trip, attitude...). In a simplerand more precise view, that is, defining time

value of passengers bases on GDP of the

section surveyed and types of vehicle .

2.1. A case of having data of traff ic

volume

With respect to section surveyed, if thereis only data of traffic volume available

(without data of passengers transportation

capacity), the saving time value in t -th yearwill be calculated as the following equation.

ipac.G

it.i

avr.KitN

m

1i365.3

tB

(VND/year)

(2)

Where, ti - average saved number of

hours per passenger when using i -th type ofvehicle (hours)

ipacG - Value of a hour per passenger

when using i-th type of vehicle

(VND/person.hour)

iavrK - average number of passengers per

i-th type of vehicle

+ car: 2.5 – 3.0 (people)

+ bus: 15-35 (people)

+ motorbike: 1.0-1.5 (people)

2.2 A case of having numbers of

passengers transported due to vehicle types

ipac.G

it.Δ

m

1i

ipactQ3

tB (VND/year) (3)

where ipactQ - number of passengers

transported of i-th vehicle type in t-th year(people/year)

3. A method to measure advandges ofsaving time for goods transportation

Apprerance of a transportation

construction project can lead to a decrease of

travelling time. In other words, goods can bequicker used. This advandtage can be

estimated as a chance value due to the sooner

use of goods.

3.1 A case of having data of traffic

volume

In case of having data of traffic volume

only (lack data of goods tranportationcapacity), the saving time value for goods

transport can be accounted as below.

gds.Git.iavr.qi

tNm

1i365.4

tB

(VND/year)(4)

Where, iavrq - average load of the

vehicle transported i-th goods (tons/vehicle)


CT 2

ti - average saved number of hours ofvehicle carrying i-th goods (hours)

gdsG -average time saved value of 1 ton

of goods (VND/ton. hour)

3.2 A case of having data of goodstranported mass

gds.Git.m

1i

igdstQ4

tB (VND/year) (5)

where igdstQ - goods transported mass of

i-th vehicle type in t-th year (tons/year).

4. A method to define gains due to adecrease of number of accidents

Transportation construction projects

influece safety of passengers, goods and

vehicles by changing traffic volume or

tranportation condition. In other words, it can

decrease or increase number of accidents. For

instance, a new expressway upgrated for the

quality purpose might increase accidents if

there is no significant safety supplement.

Therefore, the influence should be estim ated.

To measure gains of reducing accidents,

it is necessary to experience 2 steps. The first

is to assess capability of decreasing collisions.

The second is to measure the advantages of

diminishing crashes.

In addition to the first step, a need is toapproximate or predict the number of crasheshappen on the road section considered. Thiscan base on data bank of road types and roadconditions after and before having the project[2; 3].

The number of accident decreased in t -th

year on a j-th road section jtA will be

collected after accomplishing step 1. Theadvantages mentioned in the second step canbe calculated as the following equation.

jtA.

jt.m

j acdC6tB (VND/year) (6)

Where, Cacd - average lost for anaccident. It can be defined by considering thedatabase of the section surveyed.

jtm -Coefficient considering effects of

situation of j-th road section in t-th yeartowards an accident.

III. CONCLUSION

This above mentioned approach is one ofmainly methods to approximate soci al andeconomic benefits of transportationconstruction projects. Besides, there are manyothers methods which largely depend onspecific condition (capital and database).

Reference

[1]. Dr. BUI NGOC TOAN. Environmental Issues

in Transportation construction projects. Scientific

studies Collection. 14-th Science and Technology

Reference – University of Transport and

Communications – 2000.

[2]. Dr. BUI NGOC TOAN. Planning and

Appraisal of Construction Projects. Culture

Publishing House-2002.

[3]. Prof. PhD. NGUYEN XUAN TRUC (chief

author): A Handbook of Road Design -Vol.1.

Education Publishing House – 2003.

[4]. Belli and other authors. Economic analysis of

investment activities – An analysis tool - A

realistic application. Culture Publishing House -

2002


CT 2

I. INTRODUCTION

Vietnam is located in the Southeast Asia Region and being neighbor of China, Laos and

Cambodia. The country has an area of 331.212 square kilometer and a population about 84 ,12

million (GSO 2008). During the last decades, the country economy grew by 7.5 % per year and

the poverty rate has been reduced from 51% in 1990 to 8% in 2005 (GSO 2007). In that period,

the transport sector in Vietnam has achieved significant improvements, which contribute

remarkably in the development of the country and region. While the trend is expected to

continue in the next decade, Vietnam’s transport sector faces a critical situation. As most of thetransport infrastructures are still being restored from the damages of war, lack of capital to

invest in new and high capacity infrastructure and services, sector performance level is still very

far from requirements of high capacity and quality to support the quick growth and foreign -

investment-driven economy.

In the followings, this report examines the conditions for transpor t development in and the

current situation of Vietnam’s transport sector. The information and data used in the analyses

TRANSPORT SECTOR IN VIETNAM:CURRENT ISSUES AND FUTURE AGENDA

DR.-ING. KHUAT VIET HUNGInstitute of Transport Planning and ManagementUniversity of Transport and Communications

Summary: This paper briefly reports about current situation and future perspectives of

transport system in Vietnam from a planning point-of-view. Firstly, it reports about the

external conditions for transport development in term of economy, demography, environment,

and technology and transport policies. Secondly it provides an overview about current

situation of transport sector in V ietnam. Finally, it drafts an agenda for transport in Vietnam

toward 2030.

Brief CV: Dr.-Ing. Khuat Viet Hung (University of Transport and Communication,

Vietnam) obtained his Doktor-Ingenieur degree from Darmstadt University of Technology (

+country name) in 2006. Immediately after that he has been reappointed as lecturer of the

Institute of Transport Planning and Management, University of Transport and Communication

(UTC). September 2007, he has been appointed as Director of the Consulting Center for

Transport Development (UTC). His work is integrated between teaching, doing research and

providing consultant service to transport sector in Vietnam. He is working and worked mainly

on highway planning and design as a research engineer. His recent major top ics of research

are performance-based urban transport planning, traffic management and traffic safety.


CT 2

had been collected mainly from up -to-dated reports and publications of Vietnamese and

international institutions and organizations. Cons equently, this report ends by a draft agenda for

Vietnam’s transport sector in the next decades.

II. CONDITIONS FOR TRANSPORT DEVELOPMENT IN VIETNAM

1. Economic conditions

The comprehensive innovation program, called Doi Moi, began in 1986 is the main fo rce

to drive the Vietnam Economy grow up c ontinuously at a quite high rate, about 7,6% per year.

The official statistical data indicates the Gross Domestics Product (GDP) in 2007 is about US$

68,3, but experts estimated an additional value, called undergro und economy, about 30% of the

official value should be accounted (Le 2008).

Figure 1. Stable growth of Vietnam’s GDP (1990-2006)

The most important achievement of Vietnam economic growth has been well known as thesharp reduction of poverty rate. The sha re of household living under poverty rate has beenreduced from 51% in 1990 to 8% in 2005 (World Bank 2006). With the new policy, Vietnam isintegrating actively into the world economy. Total import and export value in 2007 is countedfor US$ 111243,6 million, about 163% of the country’s GDP. The foreign investment inVietnam is also sharply increasing. At the end of 2007, the count ry attracted 9810 foreigninvestment projects with registered capital about US$ 99,6 billions. The government ofVietnam and most of economists has the same optimistic expectation of a continuous high rateof economic growth in the next decades. However, the main challenge of Vietnam’s economy inthe future is improvement of development efficiency. The change of Incremental Capital OutputRatio (ICOR) indicates clearly the reduction of investment efficiency in the last decade.According to the Ministry of Finance, the general ICOR of Vietnam increased from 3,39 in1995 to 5,9 in 2005 (Ministry of Finance 2006).

USD Billion

Source:Statistical Year Book of Vietnam 2007 – 1994’s Constant Price


CT 2

2. Demography and society

As similar to Germany after the Second World War, Vietnam has a period of baby boom

after the end of our national unification war in 1975. Among 85,2 million Vietnamese t oday,

about 60% of Vietnamese are less than 35 years old (GSO 2008). The young and dynamic

people are the most important resource for the fut ure development of Vietnam.

Together with economic development, Vietnam also achieves significant improvement in

social development in the last decade. According to United Nation Development Program, the

Human Development Index of Vietnam has been increase d from 0,620 in 1990 to 0,733 in 2005

(UNDP 2008). One of the most important achievement of human development in Vietnam are

relative high average life expectancy and adult literacy rate, about 73,7 years old and 90,3% in

2005. These indicate the country owns a healthy and educated labor force.

On the other hand, the high growth of economy always consists of some negative impacts,

for example the uncontrolled urbanization , environmental pollution or increase of social gaps

etc. It is necessary to emphasize that, the current political regime does not provide good

environment for social and political dialogues. Therefore, the risks of social crashes are quite

high within the society.

3. Natural Environment

As consequences of economic development, environment in Vietnam has been affected

negatively in all sectors, land and biodiver sity, water and ambient air. The forest cover is

restoring significantly after the end of Vietnam War, 1975, by huge forest planting programs.

On the other side, the natural forests are quickly reduced by official and non -official impacts.

The growth is coming together with wastes and increase of energy consumption , air pollutants

and global warming effects. The air quality in the main cities are declining significantly, the

traffic polluters and the fine particulate matter (PM10) are the main concern. A s stated by

UNDP (2008) the quantity of carbon dioxide emi ssions of Vietnam increased 25,3% per year

between 1990 and 2005.

4. Technology

In the last decade, Vietnam is one of the leading countries of growth in information

technology and telecommunication. The rate of telephone lines per 1000 people was counted a s

191, which has been drastically take -off from 1 only in 1990 (UNDP 2008). The same way of

booming is apparent in number internet users in this country. In 2000, there were only 200

thousands internet users in Vietnam, accounted for 0,3% of population. At the end of June 2008,

the number has been counted for about 20,16 million, about 23,6% of pollution (VNNIC

2008). Accounted as a key achievement in technology development in Vietnam, t he provision of

information and telecommunication technology infrastructure is a good basis to develop high -

tech applications in different industries and services, in cluding transport and logistics in this

country.


CT 2

5. Transport policies

As similar to many other developing countries , the main focus of transport policies in thelast two decades in Vietnam is road construction and rehabilitation. Lack of financial resourceand long-term vision are main obstacles of the inadequate consideration of railway and inlandwaterway, airports are also having the similar situation. Sea -ports have been announced as amain government focus in water transport , but lacking of development prioritization does notallow the country with more than 3000 km seacoast to have any regional competent port asHong Kong or Singapore.

In the vehicle aspect, Malaysian model of automotive industry development has beenapplied in Vietnam but in a smaller scale and poorer implementation. As stated in the primeminister decision number 177/2004/QD -TTg (Government of Vietnam 2004) , the target ofVietnam’s automotive industry in 2005 is 120.000 vehicles/year, but the total sale record of allVietnam automobile markers in 2006 was only 35.637 vehicles (VAMA 2007). At the sametime, the target of localization of automobile industry has not been achieved. However, thefailure of automobile industry does not make any hesitation to the decision of Vietnamesegovernment on its current focus on development of ship building industry with the aim to beone of the world leading countries in this sector. However, as similar to the beginning ofautomotive industry in 1990, Vietnam starts its dream by a huge number of unskilled labors anda government wish.

Source: (World Bank 2008)

Figure 2. Comparison between vehicle taxes and duties in selected countries

The semi-positive point of transport policy is the effort to keep high access price forindividual vehicles. This policy is ef fectively keeps the car-ownership in Vietnam at relativelow level. On the other hand, vehicle quality management has been ignored for long time. Therecent wake-up of Vietnamese government on this matter was only on paper, not in practice.

Transport services are on the process of decentralization, which has been completely (even

11200

3227 2917 2355600 282

2520

818

18182618

2864

3373

1345

504

736

366

1155245

1545

0

2000

4000

6000

8000

10000

12000

14000

16000

Vietnam Japan China Indonesia Thailand Germany UK France USA

Vehicle tax VAT Registration Vehicle Weight taxUS$


CT 2

extremely) practiced in road, inland water transport service, and recently in aviation. On theother hand, the process is much slower in railway industry.

For public bus transport in big cities, the government subsidy has been given to users viaoperators as the main effort to attract people to use this service in order to reduce trafficcongestion. According to the Department of Transport in Hanoi, number of bus riders has beenjumped from 16 million in 2001 to 350 million in 2007 . The same pace of increase is alsoobserved in bus service in Hochiminh City and other big cities in Vietnam. However, a bettermanagement system is needed in order to control the subsidy program, especially in Hochimimhcity, where the subsidy rate (VN Dong per passenger) is still very high in comparing with otherVietnamese cities.

Regarding fuel policies, the administrative measures are preferred in Vietnam. As the onlypositive point among fuel policies, leaded gasoline had been successful ly forbidden entire thecountry since July 2001. At the end of 2007, the government tried to give its control in fuelprice up, but very soon in early 2008, the tight control has already been resettled in order toreduce the inflation rate. In the area of alternative f uel, Vietnam achieved much lower progressin comparing with other ASEAN countries . Alternative fuel is almost absent in the marketalthough it has been addressed as an important content of the government policy papers.

III. EXISTING SITUATION OF TRANSPORT SECTOR IN VIETNAM

1. Transport demand and motorization

As consequence of economic growth, transport demand in Vietnam is growing intensively.In the Transport Development Strategy of Vietnam up to 2020, the Ministry of Transport (2007)projected an average growth rate of good transport demand abo ut 7.3% per year between 1990and 2030. The demand for passenger transport is growing even faster by 12% per year in thesame period. The explosion of demand presents a good opportunity for development of transportservice industries and also big challenge s for capacity of both infrastructure and service.

Source: MOT (2007)

Figure 3. Growing of freight and passenger transport demand in Vietnam

0

20.000

40.000

60.000

80.000

100.000

120.000

140.000

160.000

Railway Road InlandWaterway

Maritime Aviation

Mil. Ton-km Good Transport Demand (1990 - 2030)

1990 (1) 2005 (1) 2020 (2) 2030 (3)

0

100.000

200.000

300.000

400.000

500.000

600.000


Maritime(Coastal)

Aviation

Mil. Pax-km PassengerTransport Demand (1990 - 2030)

1990 (1) 2005 (1) 2020 (2) 2030 (3)


CT 2

Within 20 years, average income of Vietnamese people increases about 3 times, this isindicating in transport sector by the growth of vehicle ownership. As mentioned above, the tightcontrol of vehicle price is one of the main factors to keep the car ownership low and it drivespeople to enjoy the motorization process by motorcycle. According to the Vietnam Register, atthe end of 2006, total number of road motorized vehicle in Vietnam is about 18.830.000 units,94,9% of which are motorcycles. The statistical data also indicate a growth rate of motorcyclesin Vietnam is about 17,6% per year in the period fr om 1990 to 2006.

Source: National Traffic Safety Committee (2007)


The recent studies found that the Vietnam motorization curve has a similar sharp with thos ein Japan and Korea but at a lower level. If the trend would continue, Vietnam motorization willreach the Japanese and Korean level before 2030.

Source: Asia Pacific Energy Research Center (2005) and NTSC of Vietnam (2007)


Motorization in Vietnam (1990-2006)


CT 2

2. Infrastructure

Under high pressure of demand growth, Vietnam paid great effort to develop road transportinfrastructure and service. At the end of 2006, the country has 151.632 km road network, ofwhich 64.413 km have either asphaltic or concrete pavement, accounted for 42,5%. Thepercentage of paved national road has been increased from 60% in 1995 to about 92,5% in 2006(GSO 2008).

The railway has also someimprovement and rehabilitationduring the last decades, but thereis no new section of track hasbeen extended. At the end of2006, Vietnam railway has about2362 km narrow gauge tracklength and 300 locomotives.

Vietnam has about 80 ports,three of which are regionaltransport gates (Hai Phong, CaiLan, Hochiminh). The countryhas currently 20 airports, ofwhich three are internationalairport.

It was the most historictransport mode in Vietnam,inland waterway has quite highdensity network in the north andsouthern region of Vietnam witha total length of 9800 km.

However, at the moment,Vietnam is lacking of high-speedroad and railway transport links.

The country has so far nodeep-water sea port andcapacities of the internationalairports in Vietnam are very far

from which of its ASEAN neighbors, for example Suvarnabhumi Airport (Thailand) or ChangiAirport (Singapore). This problem is well aware by governments and industries as one of themain obstacle for Vietnam to have a golden economic era.

Source: Ministry of Transport (2007)

Figure 6. Strategic Transport Networks in Vietnam


CT 2

3. Expenditure and finance

As mentioned above, the increasing ICOR during last period indicating a large share ofinvestment went to infrastructure, in which transport is the focused sector. According to MOT(2007), average share of expenditure for transport infrastructure between 2000 and 2006 wasabout 2,8% of total GDP of the country. Other study found the significant growth of capitalexpenditure for new construction or major rehabilitation projects, from 71% in 1994 to 90% in2002, while the share of maintenance and operation expenditure was reduced sharply in thesame period (World Bank 2006).


Figure 7. Structure of expenditure for different transport modes in Vietnam

According to MOT (2007), road infrastructure has been intensively focused during the last

decade. About more than 90% of total central government transport expenditure has been spent

on the road. This presented a road-based transport development in Vietnam in the last decade.

Regarding the project implementation mechanism, most of the transport projects had been

awarded to the state-owned suppliers, which currently face serious problem of efficiency and

indebtness. In the recent report, the Office of State Audit of Vietnam emphasized that most ofstated-owned enterprises in transport sector having an inefficient management structure and

many of them are in serious financial imbalance. The report also referred another report of by

the Ministry of Finance, for example, the Transport Construction Engineering CorporationNumber 5 (CIENCO 5) has a debt per capita ratio about 40, the rate was about 22,5 in CIENCO

1 and about 20 in VINASHIN (SAV 2008). Another critical problem of the pro -state-owned

contractor attitude is corruption. The expenditure process has been done through a closed andnon-transparent system between government agencies and their son’s contractors.

As stated by Ministry of Transport in the Vietnam Transport Strategy up to 2020, the actual

annual expenditure for transport infrastructure in the period 2001 -2006 is counted at only 17%the planned annual fund requirement. However, this ministry made also a very ambitious

calculation for the period 2010-2010, in which the annual planned transport fund is accounted

Railway;6,1%

Inlandwaterway; 0,9%

Maritime; 3,8%

Road;88,7%

Air

transport; 0,0%

(a) 2000

Railway;3,2%


Maritime; 3,5%

Road;91,5%

Airtransport

; 0,2%

(b) 2006


CT 2

for about 12% of the annual GDP of Vietnam (MOT 2007). As similar to other developing

countries, the share of official development assistance (ODA) in transport expenditure isincreasing significantly, accounted for 42% total MOT’s expenditure in 2002, and theexpectation on ODA is also growing as indicated in the rep ort of many transport planning

studies. The Public-Private-Partnership and commercial loan accounted only 7 -8% and the rest

of expenditure came from government budget. However, the availability of low -interest rateODA for Vietnam is reducing and asking fo r a new and sustainable structure of transport

financing in the next period.

4. Transport performance

4.1. Enhancing development and poverty reduction

Regarding transport performance, it is necessary to emphasize the great contribution oftransport development in economic development and poverty reduction. Different studies aboutimpacts of transport infrastructure on development in Vietnam had been carried last decades andproved that developing large scale infrastructure in Vietnam helped to open up new businessopportunities and facilitated the spread of economic linkages between economic growth centersand its surrounding areas. For example, 90% of the investment along ha No –Hai Phongcorridor had been taken place due to the completion of National High way number 5 expansionproject (World Bank 2006).

4.2. Accessibility improvement

The road based transport development in the last decades made significant improvement ofaccessibility of households, of wh ich 80 percents are living the non -urbanized areas. Theaffordability to public transport and motorcycle, as the main individual transport mode, has beensignificantly improved and the cost of trucking service and fuel price has also been affordable incomparing with the neighboring countries (GTZ 2007).

Source: Household Living Standard Survey 1998, 2002

Figure 8. Impacts of road development in improving accessibility of rural households

Household access to road (rural are a)


CT 2

4.3. Traffic safety

Negatively, transport development in Vietnam is directly proportional to the gr owth oftraffic accidents and fatalities. With 15 road fatalities per 100.000 inhabitants, Vietnam roadnetwork is the most deadly one over the world.

Table 1. Traffic Accident Data in Selected Countries

CountriesNo. of

AccidentsNo. ofDeaths

No. ofInjuries

Death/accident Injuries/accident Injuries/death

(a) (b) (c) (d) (e) = c/b*100% (f) = d/b*100% (g) = (d) / (c)

Thailand (2000) 73,737 11,988 53,111 16.26 72.03 4.43

Malaysia (2000) 250,417 6,035 44,019 2.41 17.58 7.29

Japan (2005) 920,053 6,586 1,134,702 0.72 123.33 172.29

Vietnam (2007) 14,624 13,150 10,546 89.92 72.11 1.24

US (By car) 2005 33,041 2,494,000 75.48

Source: JICA & NTSC (2008)

4.4. Mobility

In general, the improvement of road accessibility and the affordability of motorcycle have

made the mobility level of most of V ietnamese people much better than before. However, the

growing demand will become over the current capacity of all main transport corridors in a very

near future and requiring prompt and effective solutions. In the urban area, current high

motorcycle ownership gives people a reasonable level of mobility, but the threat of congestion

is apparently in the next decade by booming of car use. In this regard, high pressure of WTO

commitment is one of the main factors to enhance car use in Vietnam.

4.5. Efficiency of System Operation

Regarding the efficiency, the road and air transport has achi eved significant improvement

while the others had got lower level of progress. Since last 1.2006, the regulated operating speed

in the national road has been increased from 50 -60 km per hour to 70 to 80 km per hour. As the

most focused investment transport mode in the last decade, the road transport carried about 54%

total good transport demand and 85,7% total passenger transport demand in Vietnam (MOT

2007). The air transport service has also achieved a significant improvement in both number of

passengers and diversifies of services. The total number of air tran sport passenger carried in

2007 was about 8,5 million, about double of the year 2004 figure, while the difference in

number of air-craft operations was only 1,4. The same level of improvement was observed in

the maritime transport, total cargo throughput v ia the ports of Vietnam increased 218% between

2007 and 2005. In contrast, the efficiency in railway and inland waterway has not been

improved due to poor quality, low capacity facilities and infrastructure and old model of

management.


CT 2

IV. AGENDA TOWARD THE FUTURE TRANSPORT DEVELOPMENT IN VIETNAM

The high expectation of economic development, the maturity in internationalization of theeconomy and the changes in social factors would result an explosion of transport demand inVietnam the next decades. On the other hand, the current situation of lacking infrastructure andservice capacity, ineffective traffic management, and inefficient planning and utilization ofresources would continue to be the main obstacles of Vietnam transport sector on the waytoward its future development. The following measures have been strongly recommended byboth foreign and Vietnamese experts and institutions in order to support the Vietnamesegovernment to develop its agenda for transport development in the next decades:

1. Enhancing efficiency in resource utilization and service deliveryThe first activity to enhance the efficiency of transport investment is to improve the

transparency of the expenditure process by the better control mechanisms and privatization ofstate-owned enterprises. The next urgent activity is to develop a comprehensive performance -based planning system and consequent procedures. The improvement of transport planning willhelp to avoid the un-attainable, biased and emotional focuses in transport development and todevelop proper priority list of activities in transport development .

2. Sustain financial structure for transport developmentAs stated in many consultant reports and advisories, t he first answers for the question of a

new financial structure is to emphasize again the importance of transport expenditure controlimprovement, which includes also an effective infrastructure pricing system This measurewould help the government to maximize the utilization of available resources. In order toovercome the financial shortage by the running-out of ODA age, suitable Public-Private-Participation models in transport investment and operation is strongly recommended. On theother hand, government should develop a transparent and applicable framework for the use ofgovernment bonds, which should be opened for competitive bidding. Finally, the developmentof a framework for municipal finance should be promoted as the key part of the on -goingdecentralization and administrative reformation process.

3. Facilitating compact and efficient urban growthAs mentioned shortly above, the development of Vietnam will contain a quick and lack of

control urbanization process. The indication of urban sprawls and ribbon urban growth requiresalso a new concept of compact and effici ent urbanization, which needs strengthened planning,and regulatory system, proper institutions, high capacity and efficiency transport infrastructuresand services.

4. Mitigating negative impacts of transportThe lack of high capacity requires a very car e-full examination and action on developing

new transport infrastructures and services. Regional infrastructures and services arerecommended to be in the first phase of capacity development in order to deal with transportproblems. In parallel, effective traffic management, traffic safety plan s and other demandmanagement tools are strongly recommended to be part of the solution package in mitigatingtransport congestion, accident and environmental impacts .

5. Institutional capacity and human resource deve lopmentTo realize all of the above requirements, institutional and human resource capacity in

transport sector must be correctly developed. A long term institutional capacity development


CT 2

frame-work is the first task to complete and following by the streng thened implementationcapacity. Implementation of these two measures asks to speed-up of governance reform process,especially to develop capacity of Vietnam Road Administration, which is responsible for largestamount of transport investment and propertie s. Improvements of other sectoral authorities arealso needed. Enforcement and compliance require also significant improvements. Finally, it isno doubt to address the needs to develop capacity local governments a nd contractors, which arethe most important stakeholders in transport development.

V. CONCLUSIONS

The study results indicated clearly that the development of transport sector is one of thekey factors of success in economic development of Vietnam in the last two decades of Doimoiera. The transport development contributed significantly to the poverty reduction, improvementhousehold accessibility and mobility of most of people, businesses and institutionscontributions in Vietnam. On the other hand, the limitations of low quality expenditure cont rol,planning and implementation, un -sustainable financing structure, transport problems, and thelow capacity of institution and human resource are the main obstacles for the development oftransport in particular and for the cause of industrialization an d modernization of Vietnam ingeneral. Overcome the obstacles is the main goal of the agenda for Vietnam’s transport sector inthe next decades.

Reference[1]. Government of Vietnam (2004), Master Plan of Automotive industry in Vietnam to 2010 and visionto 2020. 177/2004/QD-TTg.[2]. GSO (2007), Vietnam Statistical Year Book 2006 , Hanoi, General Statistics Office of Vietnam.GSO (2008), "Population and Population Density by province in 2006," 2008, fromhttp://www.gso.gov.vn.[3]. GSO (2008), Vietnam Statistical Year Book 2007 , Hanoi, General Statistics Office of Vietnam.[4]. GTZ (2007), International Fuel Prices 2007. T. P. A. Service, GTZ and Federal Ministry forEconomic Cooperation and Development.[5]. JICA & NTSC (2008), the Study on National Road Traffic Safety Master Plan in the SocialistRepublic of Vietnam: Interim Report. Hanoi, Japan International Cooperation Agency and NationalTraffic Safety Committee of Vietnam.[6]. Le, D. D. (2008), Is the underground economy of Vietnam accounted for 30%? VTC News. Hanoi,Vietnam Telecommunication Corporation.[7]. Ministry of Finance (2006), "Vietnam aims to a higher efficiency and sustainable development,"2008, from http://www.mof.gov.vn/Default.aspx?tabid=612&ItemID=31186 .[8]. MOT (2007), The Transport Development Strategy of Vietnam up to 2020, Ministry of Transport.[9]. SAV (2008), State Auditing Report 2007, Office of the State Audit of Vietnam.[10]. UNDP (2008), "Human Development Report 2007/2008."[11]. VAMA (2007), "VAMA sales record 2006," from http://vama.wordpress.com/2007/06/04/vama -sales-record-2006/.[12]. VNNIC (2008), "Report on internet statistics of Vietnam: June 2008," Retrieved 29 July 2008,from http://www.thongkeinternet.vn/jsp/trangchu/index.jsp .[13]. World Bank (2006), Infrastructure Strategy: Cross-sectoral issues, World Bank.[14]. World Bank (2006), Transport Strategy: Transition, Reform and Sustainable Management, WorldBank.[15]. World Bank (2008), Study in urban transport strategies for medium -sized cities in Vietnam, WorldBank, Hai Phong City People Committee, Ha Long City People Commit tee

http://www.gso.gov.vn

http://www.mof.gov.vn/Default.aspx

http://vama.wordpress.com/2007/06/04/vama-

http://www.thongkeinternet.vn/jsp/trangchu/index.jsp


I. INTRODUCTION

Recently, there are many projects in Vietnam that applied design seismic. Transportconstruction projects mainly apply Specification for Highway bridge 22TCN 272 -05 belongsMinistry of Transport (refered to AASHTO 1998). Building construct ion projects mainly applySpecification for Design of structures for earthquake resistance TCXDVN 375 -2006 (referred toEUROCODE 8). Both of Specifications use “Acceleration coefficient A” or “The map ofseismogenic zones and maximum seismic intensity” pub lished by Institute of Geophysicsbelong Vietnamese Academy of Science and Technology for seismic design. In addition, Japan,that has been suffered many earthquake damages, has much experience in seismic design.Question under investigation for Vietnamese engineer as well as Japanese engineer is ability ofapply Japanese Specification for Seismic design of construction projects in Vietnam. Thisarticle analyses the difference between AASHTO 1998 and Japanese Specification in seismicdesign; contribute to conclusion for applying of Japanese Specification for construction projectsin Vietnam.

II. COMPARISON OF SOME FACTORS EFFECT ON EARTHQUAKE LOAD INAASHTO 1998 AND JAPANESE SPECIFICATION

AASHTO 1998 Japanese Specification

Earthquake

loads

Seismic loads assumed to act in

any lateral direction.

Seismic load is inertia force that shall be

calculated in terms of the natural of each

design vibration unit.

COMPARISON BETWEEN JANPANESE SPECIFICATION AND

AASHTO 1998 SPECIFICATION IN SESMIC DESIGN

MSC. NGUYEN THI TUYET TRINHPh.D Candidate, University of Transportand Communications, VietnamDR. TAKEHIKO HIMENOKawaguchi Metal Industries, Japan

Abstract: Japan has much experience in seismic design. This article analyses the

difference between AASHTO 1998 and Japanese Specification in seismic design; contributes

to conclusion for applying of Japanese Specification for constructio n projects in Vietnam.

Keywords: Acceleration coefficient; Standard acceleration response spectrum; Elastic

response coefficients


Earthquake loads shall be taken

to be horizontal force effects on the

basis of the elastic response

coefficient, Csm and the equivalent

weight of the superstructure, and

adjusted by the response

modification factor, R.

The elastic seismic force effects

on each of the principal axes of a

component resulting from analyses in

the two perpendicular directions shall

be combined to form two load cases

as follows:

100 percent of the absolute

value of the force effects in one of

the perpendicular directions

combined with 30 percent of the

absolute value of the force effects in

the second perpendicular direction,

and


value of the force effects in the

second perpendicular direction



the first perpendicular direction.

Load combination in sesimic

design = Permanal loads + 1/2 Live

load + Water Pressure + Friction

Load + Earthquake effect

Inertia forces shall be generally

considered in two directions perpendicular to

each other. It can be assumed that the inertia

forces in the two orthogonal directions, i.e. the

longitudinal and transverse directions to the

bridge axis.

Inertia force shall be defined as the

horizontal force equal to the product of the

weight of a structure and the design horizontal

seismic coefficient and be considered acting on

the structure in the detection of the inertia

force of a design vibration unit.

Load combination in sesimic design =

Primary load + Earthquake effect (= Permanal

load + Water Pressure + Friction Load +

Earthquake effect)

- Earthquake effects (EQ):

(1) Inertia force due to an earthquake

(2) Earth pressure during earthquake

(3) Hydrodynamic pressure during earthquake

(4)Effect of liquefaction and liquefaction -

induced ground flow

(5) Ground displacement during earthquake

Calculation

formula for

earthquake

force by

statically

method

EQ=W.Csm/R

EQ: Earthquake force (kN)

W: Weigh of structure (kN)

Csm: Elastic response coefficient.

R: Response modification factor.

H=W.khco.Cz.Cs

H: Earthquake force (kN)


khco: Standard value of the design hori zontal

seismic coefficient.

CZ: Modification factor for zone.

CS: Force reduction factor.


Basic valueofearthquakeload

Acceleration coefficient A:Determined by the nationalearthquake ground motion map usedin the existing AASHTO provisions,that is a probabilistic map of peakground acceleration (PGA) on rockwhich was developed by the U.SGeological Survey (USGS, 1990).The map provides contours of PGAfor probability of exceedance (PE) of10% in 50 years, which isapproximately 15% PE in the 75years design life of tipycal highwaybridge.

Table 1 Acceleration coefficient

Accelerationcoefficient

Seismiczone

MSK - 64class

A 0.09 1 Class 6,5

0.09 < A 0.19

26,5 < Class

7,5

0.19 < A <0,29

37,5 < Class

8

A0,29 4 Class > 8

Standard acceleration responsespectrum S0 : obtained from strong motionrecords with 394 components observed at theground surface in Japan, with these resultsmodified to account for the characteristics ofpast earthquake damage, vibration propertiesof the ground, and other engineeringevaluation.

Table 2 Standard acceleration response

spectrum S0

Groundtype

SIIO (gal) with natural period T(s)

I

T<0,3

SIIO=4.436T2

/3

0,3≤T≤0,7

SIIO=2.000

0,7<T

SIIO=1.104/T5/

3

II

T<0,4

SIIO=3.224T2

/3

0,4≤T≤1,2

SIIO=1.750

1,2<T

SIIO=2.371/T5/

3

III

T<0,5

SIIO=2.381T2

/3

0,5≤T≤1,5

SIIO=1.500

1,5<T

SIIO=2.948/T5/

3

Factordepend onseismic zone

There is no modificationfactor for zone , accelerationcoefficient is classified according to

4 seismic zones, that are A 0.09;

0.09 < A 0.19; 0.19 < A < 0.29;

A0.29 corresponding to zone 1, 2,

3, 4

Modification factor for zone CZ is 1.0;0.85; 0.7 corresponding to zone A, B, C tocorrect the acceleration response spectrum S 0,that applied for the bridge in large scaleearthquake may happen

There are 3 zones following the regionalclassification map. The regional classificationof earthquake ground motion complied by theMinistry of Construction. This map has beenprepared by examining the results of studiedpublished so far concerning the seismic risk inJapan, to obtain practical applicable regionalcharacteristics of seismic risk and alsocomprehensively examining together withpractical applicable data on the earthquakeoccurring at inland active faults.


Factordependingon naturalperiod T andsoil profile

Elastic Seismic ResponseCoefficient Csm

A5,2T

AS2,1C

3/2m

sm

Tm: Period of vibration of them th mode (s)A: Acceleration coefficientS: Site coefficient specified

For soil profiles III and IV, andfor modes other than the fundamen talmode that have periods less than0.3s: Csm = A(0.8 + 4.0 Tm)

If the period of vibration forany mode exceeds 4.0 s:

3/4

3

m

smT

ASC

Figure 1. Accelerationresponse spectrum Csm

Standard value of the design horizontalseismic coefficient khco

khc0=f(T, S) as Table 3:Table 3. Standard value of the design

horizontal seismic coefficient k hco

Groun

d type

khco , value in term of natural

period T (s)

I

T<0,3

khcO=4.46T2/3

0,3≤T≤0,7khcO=2.0

0,7<T

khcO=1.24/T-4/3

IIT<0,4

khcO=3.22T2/3

0,4≤T≤1,2khcO=1.75

1,2<T

khcO=2.23/T-4/3

IIIT<0,5

khcO=2.38T2/3

0,5≤T≤1,5khcO=1.50

1,5<T

khcO=2.57/T-4/3

Figure 2. Standard value of the designhorizontal seismic coefficient khco

III. ANALYSIS CONSIDERATION METHOD OF DESIGN SEISMIC FORCE IN

AASHTO 1998 AND JAPANESE SPECIFICATION

3.1. Consideration method of design seismic force in Japanese Specification

Japanese Specification do not use acceleration coeffici ent A or PGA, then design by

acceleration response spectrum base on acceleration strong motion records actually obtained at

ground surface (obtained from earthquake happening in Japan such as Hyogo -ken Nanbu

earthquake of 1995 or disaster of large scale e arthquake in Kanto). The procedure of seismic

design is as follow:

a. Records actually obtained at ground surface

For example, during the Hyogo-ken Nanbu earthquake of 1995, the high acceleration was

0.0

0.5

1.0

1.5

2.0

2.5

0.0 1.0 2.0 3.0 4.0 5.0Acc

eler

atio

n re

spon

se s

pect

rum

x g

(m/s

2)

A=0.3g

A=0.4g

A=0.5g

A=0.7g

A=0.8g


recorded at Kobe Maritime Meteorological Observatory

Figure 3. Acceleration recorded during the Hyogo -ken Nanbu earthquake of 1995

b. Calculate acceleration response spectrum of structure

Act the obtained acceleration on structure having different natural period and establishacceleration response spectrum of the structure.

Fifure 4. Steps of establishment of acceleration response spectrum of the structure

c. Calculate acceleration response spectrum of structure for each ground type

Base on 3 ground types, establish acceleration response spectrum of structure for eachground type

5%

Natural period

Damping constant

(Damping constant: h1)

Speed response spectrum

Natural period (s)

Max

of

acce

lera

tion

resp

onse

Acc

eler

atio

n re

spon

se f

ollo

ws

tim

e

c. Response spectrumb. Response follows time historya.Damping constant not changed; Natural

period changed

Earthquake

Acc

eler

atio

n (g

al)

Time (s)

-1000

-500

0

500

1000

0 5 10

15

20

25

30

PGA =812cm/s2

Japanese does not use this valuedirectly


Figure 5. Acceleration response spectrum of struc ture for each ground type

d. Calculate standard value of design horizontal seismic coefficient

Standard value of design horizontal seismic coefficient for each natural period establishedby modification of acceleration response spectrum of ground motion t hrough damping constantfor each natural period

g

)h,T(Sk 0h

Therefore, acceleration response spectrum and standard value of design horizontal seismiccoefficient are little different as below figure 6:

Figure 6. Acceleration response spectrum and Standard value of design horizontal seismic coefficientNote: Upper line is Standard value of design horizontal seismic coefficientUnder line is Acceleration response spectrum

Natural period T (s)

Acc

eler

atio

n re

spon

se s

pect

rum

S (g

al)

-

Type IType IIType III

1 2 3 4 5

500

1000

1500

2000

Design seismic coefficient

Acceleration Response spectrum

Level 2 type II

0.0

5.0

10.0

15.0

20.0

25.0

0.0 1.0 2.0 3.0 4.0 5.0Natural period T (s)

Soil type I

Design spectrum (type I)

Solid type II

Design spectrum (type II)

Soil type III

Design spectrum type III

Acc

eler

atio

n re

spon

se s

pect

rum

S(m

/s2)


e. Actual calculation

In static design, calculate standard value of design horizontal seismic coefficient ( khco)

follow above steps, then multiply with below factors for getting design horizontal seismic

coefficient (khc) for design.

Factor for zone: Base on probability of earthquake occurring in z one Cz= 0,7~1,0

Factor for structure’s property: Base on plasticity of structure’s component12

1CS μ

with

μ allowable ductility ratio, about 0.45

Factor for damping: Base on damping method such as isolation bearing shoe C E=0.7~1.0

Factor for modification of dynamic: Base on relative difference between superstructure

and substructure Cm=1.2

3.2. Consideration method of design seismic force in AASHTO 1998

a. Decide acceleration coefficient

Acceleration coefficient in AASHTO 1998 is “peak of ground acceleration (PGA)” or

“maximum value of ground acceleration” considering return period or probability of exceedanc e

(PE), it looks seismic coefficient in seismic design

Decide acceleration coefficient from hazard map considering to zone’s properties and

return period or probability of exceedance

Table 5. Acceleration coefficient

Zone Acceleration coefficient

1 A0.09

2 0.09＜A0.19

3 0.19＜A＜0.29

4 0.29＜A


b. Calculate elastic seismic response coefficient Csm (or response acceleration of structure)

Response acceleration of structure for its natural period Tm established by modification of

acceleration coefficient through site coefficient S

A5.2T

AS2.1C 3/2

m

sm ， )T0.48.0(AC msm ,3/4

m

smT

AS3C

For soil type III, IV and T < 0.3 For T > 4.0

Figure 7. Acceleration response spectrum C sm

c. Actual calculation

In static design, calculate response acceleration of structure ( Csm)follow above steps, then

consider to below factors for getting C sm for design.

Factor for structure’s properties: Base on plasticity of structure’s component R= 0.8 ～5.0

Factor for damping: Base on damping method such as isolation bearing shoe B= 0.8～2.0

3.3. Basic different between ASHTO1998 and Japanese Specification in seismic design

Both of acceleration coefficient of AASHTO 1998 and standard value of design horizontal

seismic coefficient of Japanese Specification give similar result, however start point and

procedure to the result of both are different.

a. Start point and procedure to the result of both are different

Japanese Specification

0.0

0.5

1.0

1.5

2.0

2.5

0.0 1.0 2.0 3.0 4.0 5.0

Acc

eler

atio

n re

spon

se s

pect

rum

x g

(m

/s2) A=0.4g

A=0.5g

A=0.6g

A=0.7g

A=0.8g

A=0.3


A=0.16g


Obtained acceleration in the past → Acceleration response spectrum S → Consider to

modification factor for standard value of design horizontal seismic coefficient khco

AASHTO 1998

Acceleration coefficient A or PGA considering return period → Acceleration response

spectrum Csm → Consider to modifica tion factor for elastic seismic response coefficient C sm

b. Result of acceleration response spectrum

Both Specifications give curves of Acceleration response spectrum; the comment is given

base on graph:

Maximum value of acceleration response spectrum of b oth Specification concentrate to

similar value of natural period

Value of maximum acceleration response spectrum of level 2 earthquake of Japanese

Specification is similar to value of maximum acceleration response spectrum of acceleration

coefficient A=0.8 of AASHTO 1998

Reduction slope of acceleration response spectrum at long natural period of Japanese

Specification is more sloping than AASHTO 1998

Figure 8. Curves of acceleration response spectrum of AASHTO 1998 and Japanese Specificatio n

0.0

5.0

10.0

15.0

20.0

25.0

0.0

1.0

2.0

3.0

4.0 5.0

Japan Level 2

A=0.16g

A=0.5g

A=0.8g

Nhat tanBridgeA =0.16

Csm=0.40

Acc

eler

atio

n re

spon

se s

pect

rum

S (

m/s

2)

Earthquake level 2 of Japan is simi lar to max0.8g value of AASHTO

Reduction at long period is deference



IV. COMMENTS

Each Specification has own back ground and composes each design procedure by own

philosofy. Therefore, basically we cannot use two design Specifications with mixed way in one

project, namely each equation, table, etc. are not like subroutine s in Specification for aplying.

Comparison between two Specifications is necessary to more understand each back ground

and philosophy. Base on above comparisons, the comment is proposed that we cannot use

acceleration coefficient A of AASHTO 1998 for seis mic design according to Japanese

Specification. However in the case of applying the Japanese Specification to carry out seismic

design for projects in Vietnam, that have only acceleration coefficient, in limited range we can

use Elastic seismic response coefficient Csm of AASHTO 1998 replacing for factor of k hco.CZ for

seismic design according to Japanese Specification.

If possible, comparison between results of the existing bridges, which carried out by two

Specifications of seismic design, will give us m ore detail comment about the difference between

two Specifications in Seismic design

References

[1] American Association of State Highway and Transportation Officials (1998), AASHTO LRFD Bridge

Design Specifications

[2] Transport Ministry of Vietnam (2005), Specification for bridge design 22TCN -272-05

[3] Japan Road Association (2002), Specifications for Highway Bridges

[4] Multidisciplinary Center for Earthquake Engineering (2001), Recommended LRFD Guidelines for th e

Seismic Design of Highway Bridges

[5] American Association of State Highway and Transportation Officials (1999), Guidelines for

Specifications for Seismic Isolation Design

BOARD OF EDITORS - IN - CHIEF

Prof. V. Prikhodko; Prof.V. Silyanov; Prof. Wanming Zhai; Assoc.Prof. Tran Tuan Hiep

EDITORIAL COUNCIL

MADI’s Editorial WSJTU’s Editorial UTC’s Editorial

Prof. A. Buslaev Prof. Dr. Zhao Yong Assoc.Prof. Dr.Pham Van At

Prof. A. Chubukov Prof. Dr. Zhai Wanming Dr. Tran Van Dung

Prof. I. Demyanushko Prof. Dr. Qiu Yanjun Prof.Dr. Nghiem Van Dinh

Prof. I. Fedorov Prof. Dr. Li Qiao Msc. Vu Minh Duc

Prof. V. Gerami Prof. Dr. Gao Shibin Assoc.Prof. Dr. Tran Tuan Hiep

Prof. A. Ivanov Prof. Dr. Peng Qiyuan Assoc.Prof. Dr. Le Van Hoc

Prof. L. Makovskyi Prof. Dr. Pan Wei Assoc.Prof. Dr. Dao Quang Liem

Prof. A. Nikolayev Prof. Dr. Li Fu Assoc.Prof. Dr. Vu Duy Loc

Prof. V. Nosov Prof. Huang Nan Assoc.Prof. Dr. Tran Dac Su

Prof. P. Pospelov Prof. Dr. Liu Xueyi Assoc.Prof. Dr. Tu Si Sua

Prof. A. Rementzov Prof. Dr. Zheng Kaifeng Assoc.Prof..Dr.Nguyen Huy Thap

Prof.. M. Shatrov Prof. Dr. Zhang Jin Dr. Tran Quoc Thinh

Prof. Yu. Trofimenko Prof. Dr. Liu Dan Prof. Dr. Do Duc Tuan

Prof. M. Ulitskyi Prof. Dr. Yang Yiren Assoc.Prof. Dr. Tran Quang Vinh

Prof. V. Vlasov Prof. Liu Bin Assoc.Prof. Dr. Nguyen Van Vinh

Prof. V. Zhurakovskyi Prof. Dr. Song Jirong Dr. Nguyen Quynh Sang

Prof. V. Zorin

SECRETARY SECTION

MSc. V. Vinogradova Assoc. Prof. Lan Junsi MSc. Vu Minh Duc

MSc. V. Lipskaya

Dear researchers, colleagues and readers, Transportation is the means by which all people are connected,all human activities occur. Nowadays, in strong globalizing process,community activities have not been limited by countr ies’ borders;thus transportation becomes non - confrontiers. We, transportation makers, in this moment, have had a commonforum to together discuss, contribute, share and dedicate. First Especial Issue of international co -operating transportationscience journals of State Technical University (MADI) - Russia,Southwest Jiaotong University (SWJTU) - China and University ofTransportation and Communication (UTC) - Vietnam is published inspring – season of blooming and developments. We wish you and our transportation career were achieved,prosperous and fruitful. Science is non - limitation, Transportation is non - boder, Friendship is non - confrontiers,Aim toward the future, we will do our best to make transportation:

More intelligent and effective, Faster and safer, Cleaner and Greener,With that objective, by this forum, we together connect, endeavour,research, create, contribute, share and devote.

Moscow – Chengto - HanoiBoard of Editors - in - Chief


Pages 3 Prof. VAYCHESLAV PRIKHODKO

Prof. ALEXANDRVASILYEV

Prof. MIKHAIL NEMCHINOV

Prof. VLADIMIR NOSOV

Prof. PAVEL POSPELOV


Prof. VIKTOR USHAKOV

(State technical University -MADI)

Modern tendencies in development of highwaysin Russia

Pages 16 Assoc. Prof. Dr. TRAN TUAN HIEP

University of Communication andTransport, Vietnam

Researching on position of calculation slidecenter in computing the stability of road bed byslide circular arc method

Pages 24YANJUN QIU, A.M. ASCE

School of Civil Engineering,

Southwest Jiaotong University

Chengdu, 610031, China

Theoretical development and engineeringpractice of pavements in China

Pages 39 Dr. CESAR QUEIROZ

World Bank


State Technical University-MADI

Dr. ALEKSEY AKULOV


Launching public-private partnerships for highwaysin transition economies

Pages 50 Prof. DR. DO DUC TUAN

Meng. LE LANG VAN


Diesel engine diagnosis by vibration oil analysis

Pages 54Prof. VYACHESLAV M.

PRIKHODKO

Prof. VICTORIA D.GERAMI

Prof. ALEXANDER V. KOLIK

State Technical University -MADI,Moscow, Russia

Features of creation of logistic centers inconditions of siberia and far east of RussiaPages 63 Prof. MIKHAIL V. NEMCHINOV

Dr. ALEXEI S. MEN’SHOV State Technical University - Madi Dr. DMITRY M. NEMCHINOV The Association of Road Design

Institutions of Russia Dr. VERONIKA OSINOVSK AYA

Bryansk State TechnologicalAcademy, Bryansk, Russia

The environmental problems connected withhighway construction and maintenancePages 77 NGUYEN THANH SANG, Doctoral

studentPHAM DUY HUU, Professor

Institute of Science and Technology for Transport construction


An experimental research on sand concrete inMekong deltaPages 86 Prof. M.V. NEMCHINOV

Ph.D. student VU TUAN ANH

(State Technical University – MADI,Moscow, Russia)

Design of diversion ditch es for highway roadbedsPages 93 GHAZWAN AL-HAJI, ASP

KENNETH

Department of Science andTechnology (ITN), Linköping

University, 601 74 Norrköping,Sweden

The evolution of international road safety benchmarkingmodels: towards a road safety development in dex (rsdi)

TABLE OF CONTENT


Pages 109 WEIHUA MA, SHIHUI LUO

Traction Power State Key Laboratory,Southwest Jiaotong University,

Chengdu 610031, China

RONG-RONG SONG

College of computer science andtechnology, Southwest University

for nationalities,Chengdu

Sichuan 610041. China

Influence of track irregularity on ongitudinalvibration of wheelset and correlationperformancePages 125

TRAN VIET HUNG

Msc., Dept. of Civil Eng., Universityof Transport and Communication


NGUYEN VIET TRUN G

Dr. of Eng., Professor, Dept. of CivilEng., University of Transport and

Communication,


Seismic resistance of multi-spans pc bridgeunder earthquake occur in Vietnam

Pages 135 Dr. EVANGELOS BEKIARIS

Research Director of CERTH/HIT

Forum for European Road TrafficSafety Institutes (FERSI) President

Sustainable traffic safety policies and researchpriorities for safe and secure european roads

Pages 146 Lecturer JAMSHID SODIKOV

Tashkent Automobile and RoadInstitute,

20, street Movounnahr,Tashkent,100020

Advisor Economist

ZIYODULLO PARPIEVUNDP Uzbekistan Country Office

Preliminary road cost studies in developingcountries

Pages 165 XUAN BINH LUONG; VIKHONESAYNHAVONG

THANH THUY HOANG


University of Transport andCommunications, Vietnam

MEIKETSU ENOKI


Tottori University, Japan

Development of generalized limit equilibriummethod for the failure of retaining wallsunder seismic loadingsPages 172

Dr. BUI NGOC TOAN


A method of determining social and economic

benefits of transportation construction projects

Pages 175DR.-ING. KHUAT VIET HUNG

Institute of Transport Planning and

Management

University of Transport and

Communications

Transport sector in vietnam: current issues andfuture agendaPages 187

MSC. NGUYEN THI TUYET TRINH

Ph.D Candidate, University of

Transport

and Communications, Vietnam

DR. TAKEHIKO HIMENO

Kawaguchi Metal Industries, J apan

Comparison between janpanese specification

and aashto 1998 specification in sesmic design


I. CURRENT STATE AND TENDENCIESIN THE DEVELOPMENT OF ROADS INRUSSIA

In the last 15 years the rate ofmotorization in the country increased. Rapidmotorization in Russia will continue in theforeseeable future. According to thepredictions of the Ministry of Transport of theRussian Federation, in year 2010 the car park

will amount to approximately 36 -39 millionautomobiles (table 1). Traffic volume in the

road network has increased by 5%, andgrowth of the traffic intensity on the mainroads has reached 26.2%.

The motorization growth ratio predeterminesthe necessity to speed up the road networkmodernization development.

Table 1. The number of vehicles in Russia (at theend of the year)

1990 2001 2010

Lorries (including pick-ups andvans), thous. pcs.

2744 3329 4927…5321

Passenger cars, thous. pcs. 8964 21152 31299…33805

MODERN TENDENCIES IN DEVELOPMENTOF HIGHWAYS IN RUSSIA

Prof. VAYCHESLAV PRIKHODKOProf. ALEXANDR VASILYEVProf. MIKHAIL NEMCHINOVProf. VLADIMIR NOSOVProf. PAVEL POSPELOVProf. VALENTIN SILYANOVProf. VIKTOR USHAKOV(State technical University-MADI)

Abstracts: In the report the data on the level of motorization and the road network length in

the Russian Federation, the rates of its growth from 1995 till 2010 are presented. The necessity to

enhance the pavement is validated and the new approach to road pavement design and

construction are discussed. The current system for diagnosing and monitoring the road conditions

used in Russia is defined.

Key words: road network, state and growth prospects, pavement, new tendencies in design,

diagnostic system, road condition estimation


including owned by the citizens,thous. pcs.

8677 19984 29571…31938

Number of automobiles per 1000of citizens, pcs.

80 169 250…270*

Average increase in vehicle parkper annum, %

7,7 (from 1990 till2001)

5…6* (from 2002 till 2010)

Increase in the total number ofvehicles, thous. pcs.

12733 (from 1990till 2001)

11745…14645* (from 2002till 2010 )

Note. * - Forecast. (The proportion of the passenger cars in year 1990 ≈ 76,56%; in year 2001 ≈ 86,4%).

Over the period of 1995-2000 the length

of the road network increased from 519 to 584

thous. km including increase in the federal

road network from 41 thous. km to 46.3 thous.

km. Within these years 33.9 thous. km. of

roads were built and reconstructed, including

18 thous. km. of newly built roads, 183.2

thous. km. of reconstructed roads and 290 km

of bridges; the public road network was

replenished with 47 thous. km. of roads

supervised by farm producers.

In the Programme “Moderniza tion of the

Transport System in Russia (for the period of

2002-2010)” it is envisaged that by year 2010the public roads network will increase by 1.1

times. The length of the public roads network

in 2010 will total 670 thous. km including 50

thous. km of federal and 620 thous. km of

regional roads. The length of the roads with

capital type of pavement will amount to 428

thous. km, and those with the transition type –to 212 thous. km. The length of the roads with

four and more traffic lanes will increase fr om

4.3 to 8 thous. km, i.e. nearly twofold, that

will drastically reduce the possibility of traffic

jams.

Traffic capacity of the most congested

sections of the primary interregional and

international routes will increase 1.5 -2 times,

and on an average, by 10-12% within the

network. Level of congestion, characterizing

correspondence between the road network

technical level and its traffic volume, will on

average amount to 0.4-0.6 within the network

what is optimal according to comfort, safety

and efficiency of transportation.

Under these circumstances the road

network development and road service quality

management is the task of primary

importance, its solution exercises direct

influence over the pace of the socioeconomic

development of the country. Moderni zation of

the existent public roads is becoming of great

importance, i.e. bringing their application

properties and service state into line with the

requirements of the country’s car park andfactual traffic volume. The application

properties include speed, fluidity, safety and

comfort of driving, road capacity and the level

of congestion, the capacity to carry


automobiles and road-trains with the

permissible axial load, total weight and

clearance limit as well as environmental

safety.

The efficiency of road works can be

significantly enhanced if in the process of

planning one proves and optimizes the

management decisions based on the

estimation of the factual road situation. To

adequately utilize facilities and physical

resources meant for reconstruction, r epairs

and maintenance of roads, the Ministry of

Transport of the Russian Federation and State

Technical University-MADI have developed

and implemented the system of road situation

maintenance based on the results of the

systematical monitoring, diagnostic s and

evaluation of the real road service state. This

system is based on the complete, objective and

reliable information on parameters, characteristics

and conditions of roads and road buildings

operation, availability of faults and the terms of

their emerging, traffic flow characteristics that

can be obtained in the course of diagnosis,

inspection, acquisition, analysis and

organization of the data bank on road service

quality. The management system created and

the data bank make it possible to objectivel y

estimate and predict the road and road

buildings state in the process of further

operation. The complex index of the road

service conditions takes into account

influence of the following key elements,

parameters and characteristics: width of the

main reinforced surface of carriageway and

width of the bridge; road shoulder width and

condition; traffic volume and composition;

longitudinal grade and sight distance; curvature

in plan and superelevation slope; longitudinal

surface roughness; skid resistance coefficient;

pavement condition and durability; roughness in

transverse direction (rut depth); traffic safety;

engineering equipment and instruments; level of

maintenance management.

Over a period of years the federal road

repairs have been planned and carri ed out on

based on the results of the diagnostics. For the

first time in the road sector of the Russian

Federation an evidence based, objective and

reliable system of road service monitoring and

diagnostics have been developed and widely

used, it is necessary for effective management

of the application properties by developing

projects and performing works on road

reconstruction, total overhaul and maintenance.

The road application properties are

provided by real level of their maintenance,

geometrical parameters, technical

characteristics, engineering equipment and

instruments. Diagnostic makes it possible to

detect origin of faults, justify the road repair

and maintenance type and scope. The

diagnostics system generalizes (synthesizes)

of the main regulations requirements to the

road application properties and contributes to

the increase in the quality of service.

The diagnostic system makes it possible

to effectively control the operational condition

of the certain road sections, certain routes or

certain roads and the road network altogether .

II. NEW TENDENCIES IN DESIGN ANDCONSTRUCTION OF ROADPAVEMENTS

The following factors predeterminemodern tendencies in development ofpavement calculation and constructionmethods in Russia:


Growing demands of the currentroad traffic which consists of a significantportion of lorries and a high number (m orethan 65%) of passenger cars.

The sudden increase in the numberof cars (according to the Ministry of Transportof the Russian Federation over the periodfrom 1991 to 2025 – 4-5.5 times, including anincrease in the number of lorries – 4.8 times)created necessity in radical modernization ofthe road network and, in the first instance,modernization of the federal roads.

Analysis of the real life time ofpavement in the Russian Federation and othercountries shows that when put in the sameoperational conditions rigid pavement has thelife time of 1.6-2 times longer than that offlexible pavement.

Modern tendencies in improvementof road engineering, development of n ewroad-building materials such as modifiedhigh-quality cement concrete, fiber andpolymer concrete, high-strength compositesand so on; evolution of structural conceptsand techniques contribute to expansion ofrigid and composite pavement (bituminousconcrete pavement with cement concrete subbase).

The current road network has mostlyflexible pavements including bituminousconcrete pavement with the sub base made ofmaterials that are unsuitable for bendingtension (this is mostly broken stone); suchpaving materials don’t provide for thestandard life time of the pavement as in suchstructures only a relatively thin bituminousconcrete pavement works in bending; in otherwords, in the current operational conditionsflexible pavement of the capital type should

have a sub base of binding agent -treated andsteady working in bending materials that willlast till the pavement total overhaul.

In 2001 new designing standards offlexible pavements were introduced in theRussian Federation. The peculiarity of thenew method of flexible pavement strengthcalculation lies in the fact that the structurecalculation is based on the three failurecriteria (aggregate pavement elastic modulus,shear resistance of the constructional layersand subfoundation, bending tensi on resistanceof indistinguishable constructional layers) thatare determined considering replicationinfluence (aggregate computational number ofapplications) on the pavement life time. Theold method presumed stress calculating basedon the influence of the perspective (as at thelife time end) daily average number of stressapplications (reduced to the equivalent thrustload).

Synthesis of the 60-year-old practice ofrigid pavements application in the formerUSSR and the Russian Federation makes itpossible to consider that in the overwhelmingmajority of cases the imperfections causingtheir early aging are connected not with thedesigning and constructional miscalculations,but with the failure to comply with theregulations on constructional and opera tionalmethods of the rigid pavement, and in singlecases with failure to comply with theregulations on the materials used.

At the same time it is worth mentioningthat within the period stated we continued toimprove pavement designing, calculation andconstruction methods including moreadvanced rigid pavement construction.

In a number of countries the traditional


indistinguishable cement concrete pavement withmetal pegs in the contraction joints isgradually replaced by solid-drawn continuousor fiber reinforced pavement what makes itpossible to lengthen the life time 1.5 -2 timesas compared to the unreinforced pavement. Inthe 1970-s some control sections withcontinuous reinforced pavement were laid inMoscow (Profsoyuznaya Street, AltufyevskoeShosse) that are in effective operation evennowadays.

Continuous reinforced pavement wasalso laid at some airfields of the formerUSSR.

Continuous reinforced pavement andsubgrade calculation is much different fromthat of indistinguishable unreinforcedpavement and subgrade. The most significantdifferences lie in the validation theoreticalreinforcement percentage, identification of theopening width of the reinforced breaches,reinforcement durability calculation,engagement stop resistance calculation and soon.

Based on the results of the experimentalconstruction, building regulations fordesigning and construction of continuousreinforced pavement and subgrade weredeveloped.

At the moment the following aspects ofpavement design are of relevance in theRussian Federation:

Contraction crack resistance analysisfor the bituminous concrete pavement .

Analysis of cohesionless or slightlycohesive subgrade for rigid constructionallayers.

Analysis of unreinforced indistinguishable

joint-free subgrade made of low-modulus cementconcrete for rigid and flexible constructional layers.

Analysis of fiber-reinforcedindistinguishable joint-free pavement andsubgrade.

Analysis of bituminous concretepavement with indistinguishable cementconcrete subgrade including joi nt-freepavement made of low-modulus cementconcrete and so on.

One of the most important aspects ofpavement durability implementation andenhancement in the Russian Federation is theanalysis of cement concrete durability limitsunder the joint action of repeated dynamictraffic loading and alternate periodic freezingand throwing. Due to the fact that at themoment the theoretical bases for cementconcrete cold resistance analysis aredeveloped insufficiently, the problem statedshould be studied and solved on theexperimental basis. The experimentations willmake it possible to validate the rational andeffective strength reserves for the rigidpavement what provides for the necessarylevel of durability (i.e. the life time). Anotheracute problem lies in the enhancement of themethods for the objective technical andeconomic comparison of rigid and flexiblepavement that takes into consideration notonly expenses on construction and repairs, butalso on transportation under differentoperational conditions of the pavementscompared.

At the moment the construction expensesfor the top class roads with the rigid andflexible pavement differ insignificantly –within 5%. Nevertheless, the current methodsfor technical and economic comparison


usually underestimate the influence of thedifference in the life times of the rigid andflexible pavement on the transportation cost.

Comparative analysis of the domestic andforeign designing and construction decisionsconcerning pavement make it possible toconsider that the technical level of suchdecisions in our country is rather close tothose in the highly developed countriesabroad. The engineering decisions andmodern road machinery differ insignificantlyfrom the foreign analogs. The practice of rigidpavement construction showed that:

1. Rigid pavement has the highest liferatio and according to this criterion has nosubstantial alternative under the currentconditions in the Russian Federation.

2. In the process modernization of theprimary roads it is advisably to use on a largescale the following types of pavement:

Monolithic cement concretepavement.

Monolithic cement concretesubgrade for rigid and flexible pavement,providing stable bending tension resistance.

Continuous reinforced pavement andsubgrade.

Fiber reinforced pavement andsubgrade.

Rolled concrete subgrade.

Modified concrete pavement andsubgrade, including thin continuousreinforced layers to strengthen flexiblepavement.

Polymer concrete reinforcementproviding low traffic noise level.

3. At the moment we advise to applyrigid pavement in the following fields:

Primary multilane roads.

Federal and regional roads, class II-III.

Access ways and by-passes.

Toll roads.

Industrial roads.

Municipal roads.

4. To expand the scope of application ofthe new rigid pavement it is necessary toorganize and systematically performoperational monitoring, to carry out laboratoryand in situ field experiments concerningcement concrete durability limits under thejoint action of repeated dynamic trafficloading and alternate periodic freezing andthrowing.

III. THE CURRENT SYSTEM FORDIAGNOSING AND ESTIMATING THEROAD CONDITIONS

General provisions diagnostics and

estimation of the road conditions are two

interconnected steps in the process of road

conditions management.

Diagnostics is inspection, accumulation and

analysis of the information on the parameters ,

characteristics and conditions of road

operations, on imperfections and the reasons

of their emergence and other data necessary to

estimate and forecast the road conditions in

the course of the further operation.

Road conditions estimation is the

determination of the degree of compliance of

the real road characteristics and parameters to

the regulations satisfying traffic demands.


The scope and description of th e

diagnosing works depend on the method of

the road conditions evaluation.

Stages in the development of the

methods for road conditions diagnosing and

estimation

Starting with the first half of the last

century the road conditions diagnosing and

estimation was performed according to the

availability, nature and number of faults

(deformations, corrosions, departing from the

regulations), that characterized the road from

the position of durability, efficiency, life time

and so on,

In various evaluation method s the

number of such factors fluctuated between 10

and 40. Usually not the entire road conditions

but the conditions of the pavement was

estimated. The necessary information was

accumulated applying visual method.

The second generation of the methods for

road conditions diagnosing and evaluation is

represented by composite and complex

methods, the core of which lies in the fact that

the road is considered engineering

transportation construction meant for safety

traffic with specified speed and loads.

Road conditions are estimated not only

according to technical parameters and

characteristics but also according to transport

quality figures (TQF) that a road assures:

speed, safety, admission rate, axel weight

limit etc.

Herewith each parameter, characteristic and

value is evaluated separately. As a result for

every road section there are from 20 to 80

absolute or relative numbers with different

units of measurement illustrating compliance to

or deviations from regulations, here they help

to solve the problem of function and most

important maintenance actions.

Diagnostic data are collected both

visually and using measuring equipment and

laboratories.

Third generation of methods for

generalized or complex evaluation of road

conditions is based on the concept of pu rpose

of function of a road as a means for customer

service, consumers of road services that in

some way pay for these services.

From consumers’ point of view the mostimportant are the transport and operational

characteristics assured by a road: continuit y,

speed, convenience and traffic safet, traffic

capacity and a level of congestion, axel

weight limit and other figures that are relevant

to consumer characteristics of roads.

One of the first methods of this

generation is HDM-III (Highway Design and

Maintenance) and its further modification

HDM-IV where there is an extended

evaluation of parameters of plan, longitudinal

profile and pavement conditions according to

their influence on an average speed of

vehicular traffic.

Since 1990 in Russia is widely used a

method developed by Prof. A.P.Vasiliev, a

method of complex evaluation of road quality

and condition according to its consumer

characteristics based on this principle.

In some cases methods of the first and

the second generation are used.

The basic indices of an evaluation of


quality and condition of roads according to

consumer characteristics

An integral index that most fully reflects

main consumer characteristics is a traffic

speed evaluated through a coefficient of

provision of design speed rating:

rV

Va.maxCsr (1)

Csr – coefficient of provision of speed

rating; Va.max – actual maximal possible and

safe speed of a single passenger car assured

by road at its actual parameters and condition,

km/h; Vr – design speed, km/h.

For calculation convenience as a base of

speed rating is assumed speed that equals to

120 km/h. Then:

120a.maxV

srC (2)

As an additional index is taken a number

that shows permissible carrying capacity and

axel weight limits that were reduced to the

coefficient of provision of speed rating

reasoning from proportionality of speed and

carrying capacity influence on vehicle

productivity.

Assured by a road level of serviceability,

comfort and traffic safety are estimate d with

the help of the coefficient of engineering

structure and provision of the necessary

facilities (Cnf) and with the coefficient of the

maintenance degree (Cm).

For the generalized complex estimation of

road quality and its maintenance degree they

determine index of road quality and condition

(I) which includes a complex parameter of

transport quality and operation condition (CI),

a parameter of engineering structure and

provision of the necessary facilities (C nf) and a

the coefficient of the maintenanc e degree

(Cm):

I = CI . Cnf . Cm (3)

Procedure of estimation of transport

quality and operation condition of road

In the process of development of this

procedure the most complicated methodical

problem has been solved. The problem was to

find a mode of reduction of effect of various

parameters and figures on the consumer

characteristics to one quantity indicator

describing these characteristics.

Transport quality and operation condition

of each characteristic road section i s estimated

by a total coefficient of provision of speed

ratingtotalpciC which is assumed as a complex

index of road transport quality and operation

condition on a given road section

totalpciCiCI .

For deriving total TQF index they

calculate partial indices taking into account

the effect on the main index of following

parameters and characteristics: widths of

carriageway - Ccl1; width and condition of

shoulders - Ccl2; traffic volume and traffic

composition - Ccl3; longitudinal grades and

visibility of a road area - Ccl4; radiuses of

horizontal curves and superelevation - Ccl5;

roughness of pavement - Ccl6; skid resistance

coefficient the pavement - Ccl7; condition and

strengths of road base - Ccl8; depths of ruts -

Ccl9; traffic safety - Ccl10.


Maximum traffic speed is determined by

a computed method or by measuring the speed

on the road. Most often in use is a computed

method when partial indices of rated speed

provision are determined on the basis of data

about parameters and characteristics of the

road gained by direct measurements and road

conditions surveys.

During primary survey they collect data

on all parameters and characteristics of a road,

and during the following ones they collect

data only on variable parameters and

characteristics. All the information enters in a

line graph which later on is only adjusted.

The problem of an estimation of the

effect of each parameter on traffic speed is in

determining the physical sense and

mechanism of such effect, choosing a

calculation scheme and giving some

mathematical description that allows defining

a top speed of a design car. By dividing

maximal possible speed by a base design

speed they get a coefficient of rated speed

provision.

As an initial one is taken a scheme of

movement of a single or a first car in a group

of cars that goes along the lane with a

maximal possible speed which is determined

by the effect of parameters, characteristics and

a condition of road on interacting of the car

with the road, on driver’s psychophysiological condition and his perception of

the circumstances on the road, side effect of

the cars going on adjacent lane and possible

restrictions from cars going ahead on the lane

(longitudinal effect).

During measurements on the road they

accept as a maximal possible speed a speed of

85% of provision for a single car ahead of the

group of cars or traffic speed of 95% of

provision.

Explicit numbers of indices of rated

speed provision are first gained by calculation

on the base of known or again determined

interrelations, schemes and formulas where

the parameter to be estimated is an argument,

and function is a speed of a car, and then

indices are mustered experimentally by

measurement of actual speed of cars on the

roads.

It is determined that all the estimated

parameters of the road according to their

character of the effect on the scheme of car

movement can be divided into 4 groups:

1. Parameters that effect cars or driver

and through them they influence on traffic

scheme and first of all on speed not

interrelating with other parameters. Among

such parameters are:

Roughness and strength of

pavement.

Traffic flow and its composition on

the main lane (longitudinal restrictions).

Axel load (total permissible vehicle

weight).


or more others effect driver’s perception ofroad and decision about driving mode. Among

such combinations are:

Widths of carriageway and strength

edge strip.

Width and condition of shoulder .

Widths of carriageway and traffic


volume on adjacent lane.

Visibility and skid resistance

coefficient on a horizontal section etc.


or more others effect on interaction of a car

with the road and on speed. Among such

parameters are:

Longitudinal grade, rolling

resistance and skid resistance coefficient on

upgrade.

Radiuses of horizontal curve,

superelevation and transversal skid resistance

coefficient on a horizontal curve and others.

4. Parameters and their combination that

simultaneously affect on interaction of a car

with the road, driver’s perception of road andthrough these factors on the car moving.

Among such parameters are:

Longitudinal grade, skid resistance

coefficient and visibility of road surface on

the downgrade movement.

It is determined that a driver chooses

speed mode estimating the entire situation on

the road in total. But in difficult conditions

one single parameter or their combination

have here the most influence. That is why an

estimation of effect of each parameter in

different combinations was carried o ut and the

most unfavorable of them were determined.

On each characteristic road section they

get 10 partial indices of rated speed and on the

base of these indices was defined transport

quality and operation condition CI:

CI = f(Ccl1, Ccl2, …, Ccl10).

To solve the problem calculations in

three models were analyzed and compared:

The first model is a multiplicative one

where summarized values of CI are

determined by multiplying all partial indices.


conjecture that parameters of road render

cumulative effect on traffic speed according to

distributive law.

The second model is where the

summarized value of index of design speed

provision is obtained as the least value from

10 partial indices of design speed i.e. one of

the parameters of the road or a combination of

parameters unified in one partial index on this

section effects most of all on traffic speed or

safety.

The third model is an expansion of

function of summarized index of design speed

provision in a Taylor series limiting the series

to member that are not higher than third order.


conjecture that road parameters render

complex aggregate effect on traffic speed that

can be estimated through enumeration of their

different combinations with limits regarding

dual and triple interaction. And at the same

time on every section 3 coefficients that have

the least value are chosen from 10 partial

coefficients.

The functional test of models has been

carried out by comparison of the results of

calculation from all three models for the same

road sections to the results of measurements

of actual traffic speed on these sections. In

total there were chosen about 50 road sections

of different categories in various regions of

Russia.


The results of the test showed that the

first model gives significantly understated

values of summarized index of rated speed

provision especially in range of low values.

Difference of calculated values from actual

attains 40-50% and more.

The third model gives results that are

most closely congruent with the results of

actual measurements. The mathematical

expectation of discrepancies is 0,004, with a

variance of a discrepancy is 0,003, and a mean

square deviation of discrepancies is 0,057.

The second model gives resul ts that are

close enough to the results of actual

measurements. The mathematical expectation of

discrepancies is 0,045, with a variance of

discrepancies is 0,004, a mean square

deviation of discrepancies is 0,06. However

the second model in comparison with the third

has essential virtue which consists in its

elemental simplicity and accessibility. At the

same time results of determining the

summarized index of rated speed provision

coincide well with actual results. Therefore

the second model has been accep ted as a

working one.

Thus, value of the summarized

coefficient of design speed provision on each

road section is accepted equal to one of 10

partial coefficients that is the least in value.

A graph is drawn to visualize reduced

design and cutting of road, critical parameters

and characteristics, partial and summarized

values of coefficients of design speed

provision on each road section, and al so

superimpose lines of normative and maximum-

permissible least parameter of CI.

The analysis of this graph easily allows

to determine road sections that match and

mismatch the demands for consumer

characteristics of roads, to specify

quantitatively extent of misfit, to determine its

reasons and simultaneously to assign on each

section a complex of provisions on

elimination of all or some part of the

determined deficiencies and on adjustment of

road to complete or partial correspondence to

normative demands.

On the basis of stated above the

procedure of estimation of transport quality

and operation condition of road ne twork

maintained by a road agency, road network of

separate territory or a road network of the

country as a whole is developed.

Determining the index of engineeringequipment and provision of the necessaryfacilities. An index of provision of the


equipment (Ieq) is determined on the base of

the value of index of imperfection of

compliance of engineering equipment and

provision of the necessary facilities on the

road (Ic).

Imperfection of compliance here is

absence, insufficient amount or misfit to

normative demands to parameters,

constructions and to a feature placement of

engineering system and provision of the

necessary facilities.

The coefficient of imperfection of

compliance of engineering system and

provision of the necessary facilities is

determined by results of diagnostic study of

roads under formulas:


)sIi(I8

1eq.I (4)

721i iIiIiII (5)

Ii – partial index of imperfection of

compliance of places for rest which functional

influence is spread for significant road

sections; Ii1 … Ii7 – partial indices of

imperfection of compliance of elements of

engineering equipment which functional

influence is spread on short road sections

(crossings, junctions, entries and at -grade

intersections, bus stops, barriers, sidewalks

and foot - paths in settlements, road

carriageway marking, road lighting, traffic

signs).

The value of index of provision of the


equipment (Ieq) for each kilometer of road is

assumed depending on the value of Ic

according to Fig. 1 and place in a line graph

of motor road quality and condition

estimation.

Fig 1. Graph for determining the value of I eq: I, II,

III, IV, V – categories of roads

Determination the index ofmaintenance (Io).This index is determined on

the ground of the results of the road

conditions estimation performed by a

specially appointed commission in

compliance with the active regulations on the

road conditions estimation.

To estimate the road on the ground of the

visual examination the commission uses rates

5, 4, 3, 0 depending on the defects of the main

elements of the road revealed: the roadbed

and drainage system, pavement, engineering

structures, environment and engineering

machinery, landscaping and planting.

First, each element of the road on each

section is estimated, and then the aggregate

index of maintenance on each particular

section or kilometer is rated:

554321 РРРРРР

(6)

where Р1, Р2, Р3, Р4, Р5 – index of

maintenance of each road element.

The index of maintenance is then

rendered into the index of operational

maintenance (Fig. 2).

Fig 2. Diagram for determination of (Io)

Overall estimation of highway qualityand condition

Overall index of quality and condition


for each road (road section) is determined

from the formula 3.3.

The degree of compliance of factual

road application properties with the

regulations is estimated by the relative index

of road quality and condition:

rCI

rIrQ (7)

where Ir - overall index of road quality and

condition; CIr – regulatory requirements to

aggregate factor.

The road fully meets the regulatory

requirements if Qr ≥ 1.

Increase in the overall index of road

quality and condition is determined from the

following formula:

100%rI

brIe

rIrΔI

(8)

where erI , b

rI - overall indexes of road quality

and condition as at the beginning and the end

of the period under report.

On the ground of the road and network

quality and condition analysis main

tendencies in enhancement of traffic

operational characteristics, the order and

sequence of maintenance, repair and

rehabilitation works are planned.

The methods of repair and

reconstruction planning as well as the

methods of strategy validation under the

limited financial and material resources were

developed based on the results of the road

condition diagnostics and estimation.

There is software making it possible to

perform the condition estimation and

technical and economic analysis using PC.

IV. CONCLUSION


estimation of the road condition according to

the degree of compliance with the demands of

the road and service users was developed, the

end objective of the above method is to

enhance the road application properties.


estimation of highway quality and condition

according to the application properties was

approved as official regulation in Russia,

Belarus and Kazakhstan.

Alone on the territory of Russia

approximately 30-40 thous. km of roads are

yearly estimated applying this method. A

system of diagnostic centers a full set of

instruments, equipment, mobile laboratories

and computing, making it possible to

implement all the aspects of the method, was

set up. The practical work in the field of

diagnosing and estimation of federal highway

network condition is performed under

supervision of Road Research Institute of

Russia where the federal data bank on roads

was established. The annual plans for repair

and maintenance works are developed based

on the results of road conditions estimation.

References

[1]. Nadezhko A.A. etc. Road Science. ReferenceEncyclopedia. Vol. IV/A.P.Vasi lyev,V.D.Kazarnovskiy, etc. Science editorA.A.Nadezhko. –M.: Publ.House

“INFORMAVTODOR”, 2006. -393 p

16

CT 2


I. APPROACH

Investigating the slope stability of the road is a problem which has been researched sincelong time. There had existed many methods but the most popular in over the world isW.Fellenius’ proposal based on “slide circle arc” suppose [1],[2],[ 3],[4],...

A specially improtant issue when applying this method is determining position of the mostdangerous slide arc that is its center, hereby called “caculation slide center” and correspondentlyis “caculation slide arc”. However, it is so complicat e that up to now there have been many roadscientists proposing different methods to specify this :

According to W. Fellenius: “calculation slide center” bases on a line and he proposed themanner to specify it as fig. 01.

Some experts as Gonstein, D.W. Taylor, N.N. Maslov, N.A. Txytovitr, G.Pilot…established the monograps, tables or shew the lines on which calculation slide center existing.

In the past, that calculation based on experience, is complex and large quantity; it is limitedby computation facilities, thus the quantity of investigation had been not so enough. In otherhand there has not been instruments which has ability to specify whether minimum factor ofsafety (or calculation slide centre”) is correct or not?

Stability of road slope is influenced by many factors: road bed elevation, talus grade;height, width of banquette lateral; gravity unit, friction angle, cohesive force of soil; ponding

RESEARCHING ON POSITION OF CALCULATION SLIDE CENTERIN COMPUTING THE STABILITY OF ROAD BED

BY SLIDE CIRCULAR ARC METHOD

Assoc. Prof. Dr. TRAN TUAN HIEPUniversity of Communication and Transport, Vietnam

Abstract: Investigating the slope stability of the road is a problem which has been

researched since long time. There had existed many methods but the most popular in over the

world is W.Fellenius’ proposal based on “slide circle arc” suppose. A specially improtantissue when applying this method is determining position of the most dangerous slide arc tha t is

its center, hereby called “caculation slide center” and correspondently is “caculation slidearc”. However, it is so complicate that up to now there have been many road scientistsproposing different methods for that. This presents the result of resea rching to discover the

position of calculation slide center in computing the stability of road bed by slide circular arc

method.

Key Word: Road bed stability; Embankment stability; road slope stability


CT 2

situation…

It is problem that: situations of different embankments, essentially, how different stabil ityof road bed are?

Specially, key issue of this problem: how to properly determine the position of “calculationslide arc”?

Methodology:

Firstly, we establish utility software for automatically determining factor of safety of roadbed. This program is flexible instrument for computing and then being used to investigate thestability of various cases such as factor of safety, the position of calculation slide arccentre…according to the various parameters of embankment.

II. SOME MANNERS TO SPECIFY POSITION OF CALCULATION SLIDE CENTER

2.1. Proposal of W. Fellenius

According to W. Fellenius, position of calculation slide center is determined as below (refer tofig. 01):

Firstly, determining line EF; E is specified as figure with a depth H and a length 4.5H fr omtoe of talus; F is specified by the angles 1 and 2

Correspondingly with fill slope (referring to table 01)

Table 01. Values of 1 and 2

Talus grade 1 (Deg.) 2 (Deg.)

1:5

1:3

1:2

1:1,5

1:1

25

25

25

26

28

37

35

35

35

37

Fig 01. Determining Calculation slide center by W. Fellenius

18

CT 2


On line EF lengthening, to suppose the points as the centers of slide arcs: O1, O2, O3 ...Oi,.... On…(distance between them may be 0.25H -0.3H). Corresponding to center Oi, establishingthe slide arcs then specifying their factor of safety (K) and determining minimum value of them(Kmin). With n of supposed centers we will get n of Kmin . Using EF as coordination axle,based on the centers Oi to establish the graph of the values of Kmin then determine minimumvalue Min [Kmin(i)] and we will have center Omin1 on EF, correspondingly.

Basing on Omin1 to draw line CD being perpendicular with EF. Basing on CD to repeatgradually trial process as above, we will determine the point Omin2, center of the mostdangerous slide arc.

This method has not instructed the location of firstly supposed centers O1, O2… on EF,more over computation quantity is so large.

2.2. Some other proposals

To determine calculation slide arc (the arc which has minimum facto r of safety), there areothers manners being presented as: graphic, tabulate,… by D.W. Taylor, Gonstein, G. Pilot [1],[2], ... among them attention to two conclusion should be paid:

1. As proposal of G.Pilot, D.Taylor: calculation slide centers will ba se in vertical line MN whichis across middle point of fill slope and perpendicular with bottom line of road bed (Fig.02 ), [3], [4].

2. As proposals of other scientists, scope of calculation slide center is bisector PQ of angleNPK; PK is perpendicular with talus at middle point P (refer to Fig. 3)

Fig 02. Scope of calculation slide center by Pilot, Taylor

Fig 03. Scope of calculation slide center by experiment


CT 2

In spite of gradually trial determi nation of calculation slide arc, the manners above

mentioned helped to limit the scope of finding calculation center out and pointed out that scope

are the lines. However, they just only considered particular cases and the results were not

coincided. In fact, by our calculation many cases, it is proved that calculation centers have not

been in scopes pointed out by those methods.

III. ORIENTATION FOR INVESTIGATING POSITIONS OF CALCULATIONSLIDE CENTERS

Using Utility Program, after pilot trial calculating mo re than 200 problems of different road

beds which are various by dimension, fill soil mechanic, natural soil properties; we discovered that

there are many cases their calculation center position were relatively suitable with G. Pilot’sProposal, but so much other one, they are different. In other hand, should be noted that G. Pilot only

investigates with fill up soils being grain (C=0); and natural soil are merely cohesive soil ( =0); the

influence of flooding permeable water compressor also is neglected. T hus the result may not be

correct with general case (fill up soils and natural soils have ,C; and ponding water compressor…).

By experimental computation on PC, we also invent that the scope of Calculation slide

center can be determine definitely by bas ing on two orient lines (refer to Fig.4):

- First line (line I) is perpendicular with base line of embankment and across through

“equivalent talus”

- Second line (line II) is perpendicular with base line of embankment and pass through

middle point of banquette lateral

Position of lines I and II are specified by xI and xII

“Equivalent talus” can be understood as below:

In general form, embankment has banquette (with general talus form: MNEFH) and can be

converted in to form of homogeneous talus with out banquette (line IPQK). P and Q in turn are middle

point of sub taluses NE and FH; so we have two couples of triangle which are equal each other: S1=S2

and S3=S4. It is proved by Varinogn Law that two these forms are equivalent in slope stability.

Fig 04. Orient schema for investigating the position of calculation slide center

XI

XII

I II

20

CT 2


IV. EXPERIMENTAL COMPUTING TO DETERMINE THE POSITION OFCALCULATION SLIDE CENTER

After orientation of investigation scope, experimental computation and ste p 2 of research

are conducted. In this research, all parameters such as height of embankment; height and width

of banquette; talus grade; traffic load; gravity unit, internal friction angle, cohesive force unit of

fill soil and natural soil,… are in turn v ariable; then by using utility software, corresponding to

those cases, the investigation were implemented to specify the rule of the scope which will

cover all calculation slide centers.

Because the rule of position of calculation slide centers have been d etermined yet, so in this

stage we have investigate on very large grid of slide center to cover all points may be

calculation center.

The cases of experimental computing and investigating are:

1. Position of Calculation Slide Center (PCSC) is in cases o f various slope grades;

2. PCSC is in cases of various height of road bed (H);

3. PCSC is in cases of various height of banquette;

4. PCSC is in cases of various width of banquette;

5. PCSC is in cases of various cohesive unit of fill soil of road bed;

6. PCSC is in cases of various internal friction angle of fill soil;

7. PCSC is in cases of various cohesive unit of natural soil;

8. PCSC is in cases of various internal friction angle of natural soil;

Deriving from experiment computing results we discovered a rule:

It exist an Area of Calculation Slide Center (ACSC), in which, all positions belonging to

are ACSC have minimum factor of safety.

The coordination of ACSC is as below:

(xI - 1) O.x (xII + 1)

H O.y 2.5H

V. RELIABLE EVALUATION OF ESTIMATION

To evaluate the correct and reliable level of estimation, many computing problems are


CT 2

implemented and planned as below:

B = 10 20 (m)

H = 2 8 (m)

L = 0 20 (m)

h/H = 0 0.5

Tg = 0 1

hd = hh = 0

hd - hh = 0 0.4

H1/H = 0.5 2.0

Cdry = 2.0 5.0 ( T/m2 )

dry = 14 22 ( deg. )

Cn1 and Cn2 = 0.5 6.0 ( T/m2 )

n1 and n2 = 8 20 ( deg.)

Of which:

B - road bed width

H - road bed height

L - banquette width

h - banquette height

Tg- slope grade

hd and hh- highest and lowest water level

Cdry - cohesive force unit of dry fill soil

dry - internal friction angle of dry fill soil

Cn1 and Cn2 – cohesive force unit of natural soil lay 1 and lay 2

n1 and n2 – internal friction angle of natural soil lay 1 and lay 2

Criteria of natural soils of the layers are presented in table 02

22

CT 2


Table 02. Criteria of natural soilsStae of soil Criteria Clay Loamy Sandy

C 6.0 40 2.0Semi-hard0< B 0.25 20 23 28

C 4.0 2.5 1.5Plast-firm0.25< B < 0.5 18 21 26

C 2.0 1.5 1.0Plast-soft0.25< B <

0.75 14 17 24

C 1.0 1.0 0.5Plast- liquit0.75< B < 1.0 8 13 20

That planning can cover all problems existing in fact. The data are random selected.

Dimension of test samples is: n = 250; of which: 6 cases are out of estimated area withdeflection is 1m; thus error is less than 1%.

Error ratio is: f = 6/250 = 0.024.

Deducing from that: to gain the correctness of estimation: = 0.02 then reliability of

estimation () is as below:

2,07f)f(1

nεt

Referring to table of Laplast integral:

zz(t) = 0.4808

= 2 (t) = 0.962

Thus, the estimation of rule of ACSC above mentioned has error about 0.7% 4.7%; andreliability is: 96.2%.

VI. RECOMMENT AND CONCLUSION

Based on experiment research, investigating many cases we have t he conclusions:

1. Position of calculation slide center is various when geometric parameters of road bed,mechanic and physical criteria of soil (fill soil and natural soil) are changed. However almost ofthem is located in specific scope. Coordinati on of this area is specified as below:

Abscises are limited by two lines which are perpendicular with base of road bed and passthrough middle point of banquette and “equivalent talus”;

Ordinate is limited by height of road bed (H) and 2.5 H

To have high reliable this area need to be wider, thus it should be:

(xI - 1) O.x (xII + 1)

H O.y 2.5H


CT 2

Hn

4,5Hn

Hn1/1,5

1 601 2 3 4 5 6 7 8 9 10

7

8

9

10

11

12

13

14

15

16

1 87

2 12

2 53

2 87

2 98

3 07

3 20

3 33

3 92

1 47

1 54

1 66

1 84

2 12

2 39

2 68

2 72

2 89

2 99

1 43

1 45

1 49

1 56

1 66

1 84

2 11

2 29

2 56

2 66

1 42

1 41

1 44

1 48

1 53

1 62

1 74

1 86

2 83

2 28

1 44

1 44

1 44

1 47

1 50

1 56

1 64

1 73

1 82

1 92

1 48

1 47

1 46

1 47

1 50

1 54

1 61

1 67

1 76

1 84

1 52

1 50

1 49

1 51

1 52

1 55

1 59

1 66

1 71

1 80

35°

26°

Fig 05. Result of Computing case No.110

2. To review the position of calculation slide center proposed by W. Felenius and other autho rs:

Using utility software, by experiment computing many cases, drawing the grids withminimum factor of safety corresponding to their intersect -points; synchronously establishing thecalculation slide center by W. Fellenius for comparing.

It is shown in Fig. 05 as an example. Those experiment computations presented that:

W. Fellenius’ proposal has not given minimum factor of safety as request and has adeflection of position of calculation slide center . More over, Fellenius’ proposal is only appli edfor particular case of talus is line (with out banquette).

3. By the experiment computations we discovered that:

It is existed an area of calculation slide center in which any point has calculation factor ofsafety is similar and equal minimum factor of safety.

We called that area is Area of Calculation Slide Center (ACSC) and determined it asabove mentioned.

The ACSC is able to apply correctly for general case of road bed with banquette. In caseof talus is merely line, the proposal of Taylor and Pilot is similar.

References[1]. G.Pilot, M.Moreau, Remblais sur sols mous equipe de banquettes laterales, CPC, Paris, 1973 .[2]. D.W. Taylor, Fundamentals of soil mechanics, Newjork - London, 1954.[3]. Tran Tuan Hiep, Research to automize optimization design of road bed, Ph.Dr thesis . University ofCommunication and Transportation, Hanoi, 1993.[4]. Piere Lareal, Nguyen Thanh Long, Nguyen Quang Chieu, Vu Duc Luc, Le Ba Luong . Remblaisroutiers sur sols compressible dans les cond itions du VietNam, Insa de Lyon, 1989

Hn

4,5Hn

Hn


During last two decades, China haswitnessed rapid developments in pavementengineering with upgrading of transportationmarket and expansion of transportationinfrastructures including highways, city roadsand airports. Total length of highways inChina reached 1,930,500 km with 132,674 kmof National Trunk Road connecting majorcities at the end of 2005[1].Construction ofthe first freeway, Hu-Jia Freeway in MainlandChina started in 1984 was completed in 1988linking Shanghai municipal center to Jiadin g,a suburban district 20 km away. The “7918Program” [2], China’s National FreewaySystem as shown in Figure 1, was approvedby the State Council on Dec. 17, 2004 and hasa total mileage of 85,000 km. The “7918Program” was made to meet the nationalmodernization goal in the middle of 21st

century, and China will have a freeway of0.83 km/100 km2 by then, a standard roughlyequal to the current freeway system in US[]given in Figure 2. According to China ’s 2005highway statistics[1] issued by Ministry ofCommunications in May, 2006, total mileageof China’s freeway has reached 41,005 km,which is ranked the second longest freewaysystem in the world just after US with afreeway system of 91,285 km (56,699 miles)in 2005’s Highway Statistics[3] issued by USfederal Department of Transportation. Besidesthe huge task of completing domestic highwaynetwork, the Asia Highway Route[4] asillustrated in Fig. 3 was approved inNovember, 2003 connecting 32 countries inAsia with a mileage of over 140,000 km,among which China has 26,000 km in Fig. 4,almost one-fifth of the whole program.

THEORETICAL DEVELOPMENT AND ENGINEERINGPRACTICE OF PAVEMENTS IN CHINA

YANJUN QIU, A.M. ASCESchool of Civil Engineering,Southwest Jiaotong UniversityChengdu, 610031, China

Abstract: This paper presents a comprehensive review of historical theory development

and current construction practice of pavement engineering in C hina. Mechanical models,

design guides, construction techniques, evaluation methods and maintenance standards are

elaborated for PCC pavements and AC pavements. Differences in design methodology among

pavements of rural highways, urban roads and airport fields are discussed based on service

requirements. Lessons and experiences based on past 20 years ’ construction practice and

pavement performance are summarized. Current research areas in pavement engineering

associated with unconventional geological and/or landscaping in China ’s highway

construction and national strategic plan for pavement engineering are also covered in this

paper.

Key words: Pavement Engineering, Asphalt Concrete Pavement, Portland Cement

Concrete.


At the same time, pavements for urbanroads expanded to 247,000 km in China ’s 661cities by the end of 2005 with an average roaddensity of 10.93 m2 per capita[5]. Current cityresidents accumulated to 358,940,000 whilecity land covers 412,700 km2 in 2005[5].Routine maintenance, rehabilitation, resurfacing,reconstruction, and new construction of roadpavements will continue to grow with the fastdevelopment of urbanization andindustrialization in next decades. Urbanizationratio will be increased from current 41.7% to60% during the national program of “Socialist

New Village Program”[5], which obviouslyrequires more urban roads to be built.

In airfield sector, there are 142 certifiedairports with operating air routes connectingmajor cities in the world and/or majordomestic cities in mainland China by the endof 2005[6]. Among 133 airports, nearly 60 areclose to design capacity or oversaturated withrapid expansion of air traffic. Air passengersreached 241,935,000 and cargo transportationexceeded 5,526,00 tonnage with a platoon of863 commercial jetliners. Annul increase inair transportation is expected to be 14% in the

Figure 1. China’s National Freeway System Figure 2. Eisenhower Interstate System

Figure 3. Asian Highway (AH) Route Map Figure 4. China’s AH routes

Transportation


next 5 years in China, which leads to anincrease of commercial jetliners to 1580 anddozens of runways to be expanded,rehabilitated, and/or newly built[6]. A Boeingprojection in 2002[7] expected that 2320commercial jetliners will be required in 2022to support the second largest air transportationmarket in the world, right after US. Severalmetropolitan airports such as ShanghaiPudong International Airport, Beijing CapitalInternational Airport, Guangzhou BaiyunInternational Airport, and Chengdu ShuangliuInternational Airports are potential candidatesto develop hub-and-spoke system (HSS) inorder to become a key node in world air routenetwork.

I. DEVELOPMENT OF DESIGN THEORY

Portland cement concrete (PCC)

pavements and asphalt concrete (AC)

pavements are the two major pavement

categories in China’s pavement engineering.

With the decrease of pavement’s share in total

highway costs and understanding of life -cycle

cost analysis, AC pavements have gained

more and more applications in city roads and

airports as well as rural highways. Data from

highway statistics in China show that there are

532,697 km of high-type pavements with AC

pavement of 226,075km and PCC pavements

of 306,622 km, which account for 42.44% and

57.56%, respectively. All mid-type pavements

are asphalt roads with a total length of

461,901 km. PCC pavements used to be the

only type of paving of airfield including

runways, taxiways, and aprons in China ’sairports including civil transport airports and

military airports. However, the successful

application of AC pavements in runway in

Beijing Capital International Airport in 2001

proved an effective alternative in design of

pavement types in runway construction. AC

overlay became a competitive alternative in

airport upgrading since then due to its better

serviceability. With the development of

structural analysis of pavements, design

philosophy evolved from empirical methods

in early days into mechanistic -based

methodologies. Table 1 lists pavement design

guides of various sectors current used in

China, structural analysis theory of pavements

varying from CBR (California Bearing Ratio)

method to FEM (Finite Element Method)

method. It should be noted that highways in

China refers to roads in rural areas while

urban roads is used to address in cities.

Ministry of Communications of China’s State

Council and corresponding DOT (department

of transportation) authorities in provincial and

county governments are responsible for

“rural” highways including issuing design

specifications for highway pavements. At the

same time, Ministry of Construction and

corresponding municipal authorities are

responsible for “city” roads including

publishing design guides for urban roads. In

the early days, design guide of highway

pavements was also used for urban roads,

forest road, airport and other sectors of

industry. Between middle of 1980s and early

of 1990s, separate design guides for urban

road, forest road, industrial road and airport

were developed and most of them borrowed

ideas of highway guide, AASHTO guide [8],

FAA guide and guides of other countries.

Basically, AC pavements were analyzed based

on Burmister’s layer theory[9-12] while PCC

pavements were designed based on

Westergaard’s solution or elastic solid

foundation[13-21].


II. PCC PAVEMENTS

Five types of PCC pavements, namelyJPCP (jointed plain concrete pavement),JRCP(jointed reinforced concrete pavement),CRCP(continuous reinforced concretepavement), RCCP(rolling compacted concrete

pavement), and PRCP (prestressed reinforcedconcrete pavement) have been constructed inChina. JPCP is the most widely-used PCCpavement for highways, urban roads, andrunways. Design theory was based on elastictheory of Kirchhoff thick plate with a modulusof rigidity D, which is governed by Equation(1)[12,14,21]. Base layer and underlying

Table 1. List of pavement design guides of various sectors in China

Sector CategoryPavementCategory

Current Design Guides Code Updated Year

AC Specifications for design of highwayasphalt pavement

JTG D50-2006

2006

Highway (rural)

PCCSpecifications of cement concretepavement design for highway

JTG D40-2002

2002

AC Specifications of asphalt concretepavement design of civil airports MH 5011-1999

1999

Commercial airport

PCCSpecifications for cement concretepavement design for civil transportairports

MHJ 5004-95

1995

Military airport PCC Specifications of design of cementconcrete pavement for militaryairport

GJB 1278-91

1991

Urban road AC, PCC


CJJ 37-90 1990

Forest road AC, PCC

Design specifications of roadsurface for forest road

LYJ 131-92 1992

Industrial road AC, PCC


GBJ 22-87 1987

Note: (1) New design guide for AC highway pavement JTG D50-2006 implemented starting Jan. 01, 2007.

(2) Industrial roads refer to roads in factories and mines


subgrades can be assumed as elast ic half space(Hogg’s solution), Winkler foundation (denseliquid foundation, Westergaard ’s solution),Pasternak foundation, or multi -layer elasticfoundation, leading to different equations offoundation reaction, p(r) under axisymmetricloading of q(r) in Equation (1).

rprqrwD 4 (1)

Where

dr

d

rdr

d

dr

d

rdr

d 11 22224

Design guide for highway pavements wasfirst published in 1958 with 1954 version ofthen Soviet Union’s pavement design guide asblueprint[14]. The guide was revised in 1966(JT1004-66) and served as the only existingguide for pavements including AC and PCCpavements for highways, urban roads andairports. A separate PCC pavement designguide, Specifications of cement concretepavement design for highway JTJ012 -84, wasissued in 1984. Structural analysis was basedon theory of elastic think plate over elasticsolid foundations. Critical loading positionwas situated at the middle of transverse joints.Thickness calculation is solely based onloading stress without considering curlingstress. In 1994, the guide was revised asSpecifications of cement concrete pavementdesign for highway JTJ012-94 and curlingstress was included into thickness designconsideration. Critical loading position wasselected in the middle of longitudinal joints.In 2002, design guide was revised again usingreliability concept instead of factor of safety.Thickness is determined by limiting the

combination of loading fatigue stress pr and

curling fatigue stress tr to concrete strength

fr with a reliability coefficient of r inEquation (2).

r (pr + tr) ≤ fr (2)

Pavement maintenance has to meetstandards in “Technical specifications ofcement concrete pavement maintenance forhighway JTJ 073.1-2001”. RQI (ride qualityindex) in terms of IRI (international roughnessindex), TD (textural depth), PCI (pavementcondition index) and DBL (ratio of breakingplates), and representative deflection are usedto evaluate roughness, skid resistance, distressand strength of pavement structures,respectively. For freeway system, a separatespecification, “Expressway maintenancequality evaluation standards”, should be usedfor quality requirements in maintenance work.

Distress is evaluated based on type,amount and severity and different weighfactors are allocated to specific type ofdistress with certain amount and severity incalculating PCI. Major distresses found inPCC pavement include fatigue cracking,pumping, faulting, corner breaking, jointspalling, scaling, load-transfer deterioration,depression, durability cracking, pop-outs andreactive aggregate distress.

III. AC PAVEMENTS

Asphalt pavements used in China includeasphalt concrete (AC), asphalt macadam(AM), asphalt penetration, and asphalt surfacetreatment. Other types such as SMA (stonemastic asphalt), OGFC (open graded frictioncourse), cold mix, color mix, slurry seal,micro-surfacing, and recycled asphalt


pavement. Burmister’s layer foundationtheory[15-17] was used these days for asphaltpavement design for highways and urbanroads. Airport engineers usually use CBRmethod to design AC runway pavements inaccordance with MH 5011-1999.

Design guide for asphalt pavementswere included 1958 version of “Specificationof pavement design ” and revised version“Specification of highway pavementdesign(JT1004-66)”. In those two versions ofdesign guides, an elastic half space wasassumed and M.H.Yakunin’s solution in 1941based on Boussinesq’s theory was adopted tosolve structural responses in Equation (3).Vertical stress at a depth of z from half space

surface z under load intensity of p appliedover a single circular area of diameter z wasapproximated using a =1.5 in Equation (3).Practical application of design guide in 1950 ’sand 1960’s found that Yakunin’s solution wasnot sufficient to conduct structural analysis offlexible pavements.

2

D

zα1

pzσ

(3)

Field investigation, road tests, theoretical

analysis and laboratory research had been

conducted since 1968 and a new design guide

was issued for practical use, “Design

specification of flexible pavements for

highways (interim guide)”, in 1978. The

structural theory of 1978’s guide was based on

Burmister’s two-layer system. Surface

deflection in the middle of tandem tires was

used to check thickness. A nomograph was

provided in the guide. With the development

of computers and construction practice of

pavement in highways and urban roads, a

1986 version, “Design specifications of

flexible pavements for highways (014 -86)”,

was issued. Burmister’s three-layer system

was used to solve structural re sponses.

Allowable surface deflection, together with

bending stress at the bottom of asphalt layer

was used to calculate pavement thickness.

Different interlayer status could be assumed

based on construction process and specified

methods in analyzing pavements. In order to

accommodate tire application characteristics

in urban roads, a shear index at the pavement

surface was introduced with Mohr -Coulomb

strength criterion for AC surface layer. In

1997, another version of pavement design

guide, “Specifications for design of highway

asphalt pavement”, was published using

design software APDS (asphalt pavement

design software) based on Burmister ’s multi-

layer system. Design deflection instead of

allowable deflection was used to evaluate

pavement structural integrity. In this guide

which played a key role in the rapid expansion

of China’s freeway system, asphalt pavement

with semi-rigid base was emphasized to

reduce AC surface layer by increasing base

strength. In 2004, new AC design guide based

on engineering feedback, research inputs,


lessons and experience of pavement industry

abroad was drafted to seek reviews from

industry and institutes. This draft,

“Specifications for Design of Highway

Asphalt Pavement (JTG D50-2004)”, attracted

controversial discussions from various sides

based on different research backgrounds and

versifying engineering practices. Many new

ideas based on field investigations, research

findings and pavement performance in last 10

years were incorporated. Four base types,

semi-rigid base, flexible base, rigid base and

composite base are recommended. In

accordance with new construction guide,

“technical specifications for construction of

highway asphalt pavements (JTG F40 -2004)”,

gradation was adjusted compared to JTJ014 -

97. Finally, JTG D50-2006, was approved at

the end of 2006 by Ministry of Transportation

after long discussions and/or debates.

Maintenance work is governed by

“technical specifications for maintenance of

highway asphalt pavement JTJ 073.2-2001”

and PQI (pavement quality index) is used to

designate maintenance levels. PCI, RQI,

SSI(structural strength index), and SFC (side

friction coefficient) are calculated to evaluate

pavement condition, ride quality, structural

integrity, and traffic safety, respectively.

Automatic distress survey equipments, such

FWD (falling weight deflectometer) for

deflection and RTRRMS (response type road

roughness measuring system) roughometer or

profilometer for roughness, are recommended

for fast investigation of pavement condition.

Fatigue cracking, rutting, reflection cracking

and water damage are the primary four

distress types. Other distresses include low

temperature cracking, structural cracking,

bleeding, slippage cracking, and pumping.

IV. PAVEMENT RESEARCH IN CHINA

Major research topics in ear ly daysconcentrated in analytical solution ofpavement structures, based on theory ofelasticity of Boussinesq half space, Winklersystem, Burmister multilayer system, elasticsolid system. Influence charts, nomograph anddesign tables were published for designreference[6]. With the fast development offreeway system in China, asphalt pavementsgained more and more attention in researchfields as well as in industry. Mix design with aSHRP (Strategic Highway Research Program,of US) Superpave® understanding, pavementevaluation with PCI and PSI (present serviceabilityindex), transportation management system basedon Web-GIS (geographic information system),automatic pavement data acquisition, preventivemaintenance, and pavement recycling, all thesetopics became interests of research whichresulted in updating of specifications ofpavements with respect to policy, design,construction, supervision, acceptance, andmaintenance fields.

One special research interest in China isnon-conventional subgrade deformation andpavement responses in the context that Chinahas many mountainous areas, soft ground


zones, and existing highways with lowstandards. Literature reported research effortin embankments in sloped ground, cut -fillsubgrades, road widening over soft ground,crack propagation, and structured subgrades,see Figures 5-22 (note: All these illustrationswere cited from thesis-dissertation works ofMaster and PhD students of the author ’s).Figure 5 and Figure 6 show the difference indeformation characteristics of embankmentsover flat and sloped ground, which is acommon alignment result of roads inmountainous areas. It can be seen thatdeformation concentration at the downside toeof embankment constitutes a great concern inslope stability.


of flat ground


of sloped ground

Soft and weak ground is one of the

traditional geotechnical challenges. Residual

settlement after construction completion poses

significant influence on structural

performance of pavements over embankment

in soft ground. Step by step consolidation

prediction can be predicted as shown in Figure

7 and Figure 8 and can be a practical reference

in embankment construction monitoring.

Figure 7 and 8 are simulation results of

embankment with a top width of 34.5m in

Chengdu 4th Ring Road Freeway (Chengdu

Bypass Freeway).

Cut-fill transitional subgrade sections arecommon practice in highway design inmountainous areas due to geometriclimitations and geological conditions. Trafficloading in pavements over cut -fill subgradesin Figure 9 will produce additional structuralresponses as shown in Figure 10 which in turnwill result in shortened pavement service life.

Major problems in road widening projects

as illustrated in Figure 11 are differential

deformation between existing roads and newly-

added section, especially in soft and weak

ground. Additional structural responses produced

in pavements as a result of differential

deformation behavior of subgrades will influence

pavement performance. In figure 12, it can be

seen that different widening scenarios will cause

different deformation curve across transverse

section.


II. NỘI DUNG

Nhằm mục đích phân tích tài chính vàphân tích kinh tế - xã hội, có thể phân chia cácloại dự án theo tính chất của công tr ình mà nóxây dựng là (i) các dự án xây dựng công tr ìnhcông cộng và (ii) các dự án xây dựng côngtrình dân dụng và công nghiệp. Các loại dự ánnày có các đặc điểm khác nhau cần tính đếntrong quá trình phân tích.

-20

-15

-10

-5

0

5

10

0 20 40 60

Embankment bottom widthm

Gro

und

settl

emen

tcm

layer 1 layer 2 layer 3 layer 4layer 5 layer 6 layer 7 layer 8

- 7

- 6

- 5

- 4

- 3

- 2

- 1

00 5 10 15 20 25 30 35

Top embankment wi dt h（m）

Emba

nkme

nt s

urfa

ce s

ettl

emen

t（cm

）

Load st ep 1 l oad st ep 2l oad st ep 3 l oad st ep 4

Figure 7. Consolidation process during

embankment fill

Figure 8. Settlement development

under traffic

0 1 2 3 4 5 6

-1400

-1200

-1000

-800

-600

-400

-200

0

200

400

600Ho

riza

onta

l st

ress

top

of

subg

rade /

kPa

X /m

1# 2# 3# 4# 5# 6#

Figure 9. Traffic loading on cut-fill sectionFigure 10. Structural responses

of cut-fill section

h

1:m

Y

X

-4 -3 -2 -1 0 1 2 3 4 5 6

-2.8

-2.6

-2.4

-2.2

-2.0

-1.8

-1.6

-1.4

Surfa

ce D

efor

mat

ion

(cm

)

Transver di st ance f rom cent er l i ne (m)

one-si de wi deni ngt wo-si de asymmet r i cal wi deni ngt wo-si de symmet r i cal wi deni ng

Centerline of roadway

Figure 11. Pavement over road-widening section Figure 12. Structural responses

of widened roads

Cut – fill boundary

1#

2#

3#

4#

5#

6#


Semi-rigid base structure has beenmainstream type of asphalt pavements since1980s. Main idea of semi-rigid base is toprovide a strong base to bear traffic loadingwhile surface AC layer is in a compressivestress state, which is very different fromtraditional flexible AC pavement structureswhere tensile stress produced at the bottom ofsurface AC layer controls fatigue cracking[5].Major problems with semi-rigid base ACpavement structures are reflection cracking.Figure 13 shows that AC layer in flexiblestructure will produce tensile stress at lowdepth while only compressive stress existed insurface AC layer in semi-rigid base. Figure14 suggests that tensile stress existed inflexible AC bottom will move down to semi-rigid base and surface AC thickness can thusbe reduced to produce economical benefits.

0

0.05

0.1

0.15

-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4

Horizontal StressSX/MPa

Surf

ace

Lay

er T

hick

ness

/m

Flexible

Semi-Rigid

Figure 13. Maximum horizontal stress of AC layer

0

0.1

0.2

0.3

0.4

0.5

0.6

-0.60 -0.50 -0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20Horizontal Stress

Dep

th/m

Flexible

Semi-Rigid

Figure 14. Horizontal stress at deflection point

Figure 15 and Figure 16 shows differencein structural responses of AC pavements

between contact layer interface andcontinuous layer interface using Burmister ’slayer model. China’s AC pavement designguide use only continuous model since 1986.Surface deflection between these two modelsexhibit significant state as shown in F igure 15while vertical stress are relatively similar.

Figure 15. AC pavement surface deflection

Figure 16. AC pavement vertical stress

Influence of contact area on pavementresponses is given in Figure 17 and Figure 18.Circular loading has been used in China ’s ACpavement design guides since 1958, which isalso a common assumption in other countriesas a result of availability of analytical elasticsolution in earlier days. Rectangular loadingcan be used with numerical methods tocompare the difference of pavement responsesamong various assumption of contact areabetween tires and pavement surface.

Crack propagation simulation usingcomputer simulation has been o ne ofpavement research fronts in last decade withthe development of damage mechanics and

X – Distancem

X – Distancem

Y–

Smax

.Pri

ncip

al P

aY

– D

efle

ctio

n m


simulation program. Figure 19 and Figure 20illustrate HMA crack propagation processusing APCPPS2D program and computertechnique used in the program.

Figure 21 and Figure 22 illustrateapplication of image recognition technique inpavement cracking research. Using imageenhancing method, raw pavement crackingimage can be processed to produce highquality data for computer recognition.

Figure 17. Shearing stress of circular loading

Figure 18. Shearing stress of rectangular loading

Figure 19. Crack propagation of testing

HMA beam

Figure 20. Element technique of simulating crack tip

Figure 21. Image Enhancing technique

Figure 22. Image threshold in MATLAB

Figure 23 shows the scanning process of

asphalt samples using computer tomography

(CT) method. This new methodology proves

to be an effective tool in analyzing the

cracking propagation process of asphalt

concrete under load application. F igure 24 is

the scanned cross section of one Marshall

specimen.

X – Distance m

X – Distance mY–

Max

She

arin

g St

ress

Pa

Y–

Max

She

arin

g St

ress

Pa


Figure 23. CT process of pressured AC specimen

Figure 24 . CT scanned AC specimen

V. LESSONS AND EXPERIENCE

Major experience in pavementengineering learnt from last 20 years ’ rapidexpansion of China’s transportationinfrastructures can be summarized as follows.

1. Overloaded truck axles have been topone factor to cause premature failure ofpavements in highways. Orchestrated effortsfrom various government authorities andrelated industries must be taken to ensure thetrucks rolling in pavements not to exceedspecified axle load standards. WIM (Weigh -in-motion) and strict truck and transportationpolicy can help reduce heavy axles.

2. Drainage is the key factor to producequality subgrade and sound pavementstructures. Drainage facilities should besufficiently designed since later adding afterpavement completion and/or during operatingproves to be an inefficient alternative.

3. Structural analysis and material designshould be integrated to reach a balancedpavement quality. Thickness sufficiency canprevent pavement from premature failure instructural distresses such as fatigue cracking,while mix design plays a central role to avoidunexpected distresses such as AC shoveling,high-temperature rutting, water damage,potholes and other types of material -relateddistresses that cannot adequately addressed bystructural design, especially for ACpavements.

4. Construction techniques in ACpavements such as application of tack coatand prime coat have to be fully implementedto form a layer structural which is close to theabstracted mechanistic model in design guide,usually based on Burmister ’s multi-layerelastic theory, otherwise, constructiondeficiency compared to structural modelmight lead to premature failure.

5. Thickness of PCC plates governsexpected pavement life. A small increase inPCC thickness can lead to significant increasein pavement life. China’s new design guidefor PCC pavements, JTG D40-2003, reflectssuch understanding which is also based fieldinvestigation of PCC pavement performance.

6. Joint sealing in PCC is almost thelatest field work in construction of PCCpavement which often leads to negligence inconstruction supervision. However, jointsealing is one of governing factor to preventPCC from premature failure such as jointspalling, corner breaking, pumping andswelling. Care should be taken to ensure 100percent joint sealing to produce soundpavement structure.

7. Deformation characteristics andengineering behavior of subgrades have to beadequately understood in design process,especially for multi-lane freeway roads inmountainous areas where cut -fill transition

Scanned Data

Strain (%)

Stre

ss (

MPa

)


and sloped ground are frequently found.Structural subgrade, such as pile -supportedembankment, column-supported embankment,and other subgrade structures can limit soildeformation with the help of embeddedstructures.

8. Preventive maintenance, together withroutine maintenance, proves an effectivepractice to keep pavement quality over time.Pavement distress of light severity may easilyescalate to high severity during short period oftime without timely engineering measures.Preventive maintenance such as crack sealing,slurry treatment, and seal coat are more costeffective than corrective maintenance.

9. Stage construction in highways used tobe an engineering alternative in soft groundarea to produce adequate consolidation rateand avoid excessive settlement aftercompletion. However, experience in past twodecades proved that stage construction is not apractical choice based on life -cycle costeffectiveness consideration.

10. Road widening projects, especially inweak ground and/or sloped ground, needspecial considerations in differentialdeformation between existing subgrade andnewly-added section. Additional structuralresponses resulted from differentialdeformation will lead to unexpected distressor even premature failure. Geosynthetics, ifproperly designed and laid between new andold subgrades, proved to be effective materialto tackle subgrade problems.

11. Vegetation is the primarylandscaping type which promotes highwayquality, environmental friendliness, visualrelaxation, sight guidance, and drive safety.One concern that should be addressed is thatvegetation in medians and side slopes maycause drainage problems and weak subgrade.Therefore, effective drainage system forsubgrades and other pavement layers must by

fully designed, especially in vegetatedmedians.

12. Heavy trucks traveled at low speedsclimbing long grades pose a great challenge tosurface layer of pavements, especially for ACpavements in high temperature environment.In the summer of 2006 when temperatureloomed historical high for several weeks, deeprutting, shoveling, and slippage cracking tookplace in several routes with AC pavements.Material design can address part of this problem.However, shear resistance of AC surface alonglong grade section is recommended for check instructural design, which is part of the code forpavement design for urban roads.

13. Structural evaluation of geometricdesign is one of the most important aspects indesign stage and post-design stage. Analysisof highway location and geometric designbased on statics, kinetics and dynamics willlead to a consistent design with respect togeometrical outputs, structural responses, andservice life predictions.

14. It is not necessary to designpavement shoulders different from lanethickness. A reduced thickness of pavementstructures in shoulder does not producesignificant economical benefits compared tototal costs of highway investment. Shoulders,usually used as emergency lanes andmaintenance work space, can serve as trafficlanes when necessary. Furthermore, a singlethickness along traffic lanes and shoulderswill facilitate easy construction, especially forAC pavements.

15. Transitional sections of subgrades,such as bridge approaches, culvertapproaches, tunnel approaches, short subgradesection between road structures, high -fill-deep-cut and cut-fill transitions, are commondesign scenarios in mountainous areas anddeserve special consideration in construction


and supervision to form a pavement support ofconsistency along the route. Strict compactionstandards, reliable subgrade drainage systemand quality granular fills are effectivemeasures to construct subgrades in transitionalsections.

16. Temperature is the key factor inHMA mix production and paving. Climateprediction has to be tracked daily to assureadequate temperature at paving, compaction,as well as proof rolling. Sudden change inclimate, such as temperature drop, should beavoided in the whole process. Mix has to bediscarded if temperature cannot meet specifiedstandards at every stage of construction frommixing to completion.

17. Adequate compaction is required forHMA surface layer for the mix to form asurface layer with sufficient strength to limitstructural distresses such as fatigue crackingand rutting. However, over compaction is notrare due to inadequate understanding of ACpavement as well as HMA mix design.According to Superpave® mix designphilosophy, over compaction is not allowedand a 2% of air voids at the end of pavementservice life has to be expected. Overcompacted AC pavements do not havesufficient voids to deform under the repetitiveapplication of traffic loads and will exhibitrutting, shoveling and other types ofpremature distress.

18. Reflection cracking is the primarydistress in AC pavements with semi -rigidbase. It is crucial for all highway-relatedagencies responsible for design, construction,supervision, acceptance, maintenance, and/ormanagement to understand that adequate baseis sufficient to support pavement under trafficloading. “Over-strong” base, especially semi-

rigid pavement will crack easily and l ead toreflection cracking in surface layer. Flexiblebase, which constitute “flexible pavement”together with AC surface layer, isrecommended instead of semi-rigid base.Literature reported that micro-cracks in thebased layer produced by heavy rollercompaction before surface AC layerplacement prove to be an effectiveengineering practice to avoid reflectioncracking caused by cracks in succeedinglayers.

VI. CONCLUDING REMARKS

Pavement engineering in China is still abooming industry. Unpaved roads wit h amileage of 935,945 comprised of 48.48% outof 1,930,500-km highway networks at the endof 2005. National Freeway System of China isstill under halfway construction which isexpected to be completed in 2040 and at thesame time another 1,000,000 km newhighway will be added to China ’s highwaysystem. There are hundreds of runwaypavements and other airport pavements willbe newly built and rehabilitated during next20 years to meet the rapid expanding of airtransportation and hundreds of big cities wil lcontinue road upgrading work in the contextof China’s national policy of urbanization.Due to different environmental requirements,local industry and economy status, marketfluctuation of raw materials including asphaltcement and Portland cement, emerging of newmaintenance technologies, development ofdesign guides and better understanding ofroad engineers, PCC pavements and ACpavement will continue to be the two majorcategories in the future. AC pavements willfind more and more applications in ur banroads and airport runways. Modified asphaltincluding color asphalt will be tailored to meetdifferent transportation needs besides


engineering requirements. Internationalcooperation and exchange will play an evenmore important role in pavement indus trywhich will promote the continuousdevelopment of theoretical understanding andengineering practice of pavement with betterquality for roads and airports.

Acknowledgement Figures 5 to 24 areresearch illustrations of master and doctoralstudents of the author’s. It has been a greatpleasure to work with such smart youngtalents. The author pays special thanks tothese graduate students for providingmaterials in the completion of this manuscript.

Reference

[1]. Statistics of China’s Highway and WaterwayDevelopment, (in Chinese)http://www.moc.gov.cn/zhuzhan/tongjixinxi/, 2007-03-18

[2]. National High Speed Highway System PlanningProgram (in Chinese),http://www.moc.gov.cn/zhuzhan/jiaotongguihua/guojiaguihua/quanguojiaotong_HYGH/, 2007-03-18

[3]. Highway Statistics,http://www.fhwa.dot.gov/ohim/ohimstat.htm, 2007-03-18

[4]. Aisan Highway Program,http://www.unescap.org/TTDW/index.asp?MenuName=AsianHighway, 2007-03-18

[5]. Statistics of Municipal Construction(in Chinese),http://www.cein.gov.cn/news/show.asp?rec_no=14462,2007-03-18

[6]. Statistics of Civil Aviation Airports,http://www.caac.gov.cn/I1/, 2007-03-18

[7]. Boeing Projection on China’s Aviation Market,http://news.xinhuanet.com/fortune/2004-10/27/content_2144873.htm, 2007-03-18

[8]. AASHTO(1993), Guide for design of pavementstructures, American Association of State Highway andTransportation Officials.

[9]. Burmister. D.M.(1943), The theory of stresses anddisplacements in layered systems and applications to thedesign of airport runways, Proceedings of HighwayResearch Board, pp. 126-148.

[10]. Burmister. D.M.(1945), The general theory ofstresses and displacements in layered soil systems,Journal of Applied Physics, Vol. 6, No. 2, pp. 89-96, No.3, pp. 126-127; No. 5, pp. 296-302.

[11]. Burmister. D.M.(1945), Stress and displacementcharacteristics of a two-layer rigid base soil system: influencediagrams and practical applications, Proceedings of HighwayResearch Board, pp. 773-814.

[12]. Huang, Y.H. (2004), Pavement analysis anddesign, 2nd, Pearson Prentice Hall.

[13]. Ioannides, A M; Thompson, M R; Barenberg, E J,“WESTERGAARD SOLUTIONS RECONSIDERED” ,Transportation Research Record N1043,pp. 13-23.

[14]. Lin, X. (1988), Design methodology of flexiblepavement structures (in Chinese). Renming JiaotongPublishing House.

[15]. Westergaard. H.M.(1926), Analysis of Stresses inConcrete Pavements Due to Variations of Temperature,Proceedings, Highway Research Record, Vol. 6, pp201-215.

[16]. Westergaard. H.M.(1926), Stresses in ConcretePavements Computed by Theoretical Analysis PublicRoads, V. 7, No. 2, pp.25-35

[17]. Westergaard. H.M.(1927), Theory of ConcretePavement Design, Proceedings, Highway ResearchRecord, Part I, pp175-181.

[18]. Westergaard. H.M.(1933), Stresses in ConcreteRunways of Airports, Public Roads, V. 14, No. 10,pp.185-188

[19]. Westergaard. H.M.(1943), Stress concentrations in PlatesLoaded over Small Areas, ASCE Transactions, Vols. 108,pp.831-856

[20]. Westergaard. H.M.(1948), New Formulas forStresses in Concrete Pavements of Airfields, ASCETransactions, Vol.113, pp.425-444

[21]. Yoder & Witczak, Principles of PavementDesign, 2nd Edition by John Wiley & Sons, Inc.,

1975

http://www.moc.gov.cn/zhuzhan/tongjixinxi/

http://www.moc.gov.cn/zhuzhan/jiaotongguihua/guojiag

http://www.fhwa.dot.gov/ohim/ohimstat.htm

http://www.unescap.org/TTDW/index.asp

http://www.cein.gov.cn/news/show.asp

http://www.caac.gov.cn/I1/

http://news.xinhuanet.com/fortune/2004-


I. INTRODUCTION

Many governments do not have all the financial resources required to expand, maintain,and operate their country’s highway networks and other transport infrastructure. The overallresources needed are enormous. In the United States, for example, it is estimated that $55billion will be required annually over the next 20 years simply to maintain the highway andbridges in their current condition.

In many countries, the private sector has been involved in financing infrastructure throughconcessions under a public-private partnership (PPP) program. Broadly defined, a concession isa legal arrangement in which a firm obtains from the government the right to provide aparticular service (Kerf, 1998). PPP schemes, however, are somewhat under utilized in transitioneconomies, and there seems to be an enormous potential for more private sector involvement inthe financing and operation of highway assets in these countries.

With many countries increasingly interested in attracting private capit al to infrastructureprojects, institutions such as the World Bank can contribute through greater use of their

LAUNCHING PUBLIC-PRIVATE PARTNERSHIPS FORHIGHWAYS IN TRANSITION ECONOMIES

Dr. CESAR QUEIROZWorld BankProf. VALENTIN SILYANOVState Technical University-MADIDr. ALEKSEY AKULOV


Abstracts: In many countries, the private sector has been involved in fi nancing infrastructure

through concessions under a public-private partnership (PPP) program. This paper reviews

potential applications of partial risk guarantees, the required legal framework (e.g., concession

law) for attracting private capital for PPP sc hemes, possible steps for a country to launch a

program of private participation in highways, the concept of Greenfield and road maintenance

concession programs, and the treatment of unsolicited proposals. It also summarizes potential

applications of the World Bank Toolkit for PPP in Highways as an instrument to help decision -

makers and practitioners to define the best PPP approach for a specific country .

Key words: a public-private partnership (PPP) program; concession law; private capital; “build -own-operate-transfer” (BOOT) contract; “build-own-operate” (BOO) contract; “build-operate-transfer”(BOT) contract


guarantee power. Partial risk guarantees are particularly relevant in the context of seeking moreprivate involvement in the financing of road i nfrastructure.

This paper reviews potential applications of partial risk guarantees, the required legalframework (e.g., concession law) for attracting private capital for PPP schemes, possible stepsfor a country to launch a program of private participa tion in highways, the concept of Greenfieldand road maintenance concession programs, and the treatment of unsolicited proposals. It alsosummarizes potential applications of the World Bank Toolkit for PPP in Highways as aninstrument to help decision-makers and practitioners to define the best PPP approach for aspecific country.

II. CONCESSION LAWS

An appropriate concession law is fundamental for a country to establish an enablingenvironment for PPPs and it also serves as a possible marketing tool for i nvestors. It shouldapply to construction, expansion, rehabilitation and maintenance of assets providing a publicservice, aiming at improving the efficiency and modernization of public services. In general, aconcession law should include provisions for:

Definition of concepts and terms.

Transparent competitive bidding.

Allowing for bid evaluation on a net present value (NPV) basis .

Assurance of national treatment to foreign investors .

Assurance of compensation in the event of expropriation .

Assurance of access to international arbitration for foreign investors .

A general deference to the terms of specific contracts, which creates scope for flexibleapproaches between sectors and projects .

Public disclosure of concession agreements

A concession law can be kept relatively simple and general, while specific regulation withdetailed guidelines about the ways in which the procurement process will be conducted, criteria,contract award, selection committees, etc. should be documented in operational guidelines (ordecrees). A separation between law and regulation provides more flexibility for amendmentsduring the implementation of a PPP program.

It is usually beneficial to have a draft concession law reviewed by a law firm with a strong

international project finance practice and with a strong local knowledge base.

Public disclosure of concession agreements should be supported. In recent years a growing

number of countries have taken the step of publishing the concession agreements. This has

several benefits: (a) it provides a further check on corruption, which in addition to its direct


benefits can enhance the legitimacy of private sector involvement in often sensitive sectors; and

(b) when the concession agreement relates to the provision of services to the publ ic, it provides

consumers with a clearer sense of their rights and obligations, and can facilitate public

monitoring of concessionaire performance.

It is usual practice for concessions law to contemplate the concept of “negativeconcessions” in which the bidding criterion is minimum public contribution rather thanmaximum payment to the public authority. The law would make explicit the right of the publicauthorities to enter into multi -year contracts to pay the concessionaire the required stream ofpayments.

A concession law needs to link with other laws, such as:

Laws regulating the provision of public services

It is common to find aspects of utility services governed by sector -specific laws, some ofwhich establish specialist regulatory bodies. The relationship between those laws and bodiesand concession agreements needs to be spelt out, for example regarding tariffs and servicestandards.

Procurement laws

In order to provide a clear legal framework, the regime for bidding for concessions needs tobe clear vis-a-vis other procurement laws.

Laws governing foreign investment

The provisions of a concession law need to be clear relating to other laws that mightinclude restrictions of some kind on foreign participation. It is important that regardingconcessions there is no separate treatments for local and foreign investors.

Many countries distinguish between concessions for public works, concessions for thedelivery of public services, and concessions for the exploitation of natural resources. Aconcession law would need to reflect such distinctions. While concessions for public worksrequire investments, under many concessions for the delivery of public services the mainobligation of the concessionaire is to provide the service, rather than make specific investments.

There are situations in which one of the bidding criteria is based on minimum public

contribution (or “subsidies”) to construction or reconstruction costs. A Concession Agreement

may include: (i) re-build1 and/or build; (ii) operate; and (iii) maintain. It may cover the whole

spectrum of PPP, i.e., management and maintenance contracts, operation and maintenance

contracts, and Build-Operate-Transfer (BOT) concessions.

1 Re-build may include, for example, rehabilitation, modernization, refurbishing.


Other useful provisions in concession laws include (relevant definit ions can be found inrecent concession agreements):

The concept of “cannon” or “entry ticket fee,” which is a current practice worldwide.It is usually through the “cannon” that the concession’s high transaction costs are reimbursed bythe concessionaire.

The concept of periodic independent assessment of the concession assets to be carriedout by an expert acceptable to both parties and paid, preferably, by the concessionaire.

Amendments of concession agreements. International experience illustrates twogeneral approaches: (a) provide no special rights for the grantor to unilaterally amend orterminate, and so leave this to be determined by the parties by agreement; or (b) provide thegovernment with such special rights, but with carefully defined safeg uards for theconcessionaire.

Required land to provide the public services. Responsibilities of the grantingauthority may include providing adequate site condition, right of access, expropriation andacquisition of land, contingent environmental liabil ities, etc.

The concept of contract renegotiation, as it is better to be prepared to manage theprocess when renegotiation may be necessary. Concession laws should establish clearmechanisms for renegotiation and amendments (as a way to minimize contract distress andcancellation). The renegotiation of projects is not an unusual occurrence (Harris et al. 2003).

Provision for international arbitration.

Award of contracts through a two-step approach in which the qualitative requirements(e.g., experience, financial capability, management plan) and some of the quantitativerequirements (e.g., investment plans) are judged on a pass/fail basis. All bidders that pass thisstage are by definition qualified and step two judges the financial offers.

Exceptions to competitive bidding. For example, most countries permit sole sourcingin the case of very small contracts (where the costs of a tender would be disproportionate to thebenefits) and in emergency situations (where there is no reasonable time to conduct a tender ---which may be a particular concern when it relates to the delivery of public services.

III. UNSOLICITED PROPOSALS

Unsolicited proposals, which seem attractive to some governments in their wish toaccelerate road or motorway construction in the country, tend to be so controversial (usuallyinvolving allegations of corruption), that in fact they may take longer to negotiate than an open,competitive tender procedure. In theory, unsolicited proposals could generate beneficial ideas;in practice, there have been a number of unfavorable experiences, mostly as a result of exclusivenegotiations behind closed doors (in a recent case, a contract signed between a government anda private company included a clause that prohibits any leakage of the signed contract).


Several countries have adopted specific legislation to deal with such proposals, while somegovernments have simply forbidden unsolicited proposals to reduce public sector corruption andopportunistic behavior by private sector companies. The general experience with unsolicitedproposals is often negative, reflecting the fact that projects of this type have usually representedpoor value for money, were frequently incompatible with the actual development needs of thecountries, and their ability to pay. They also often elicit allegations of corruption. Corruptionhas been shown to be associated with the lack of adequate transport infrastructure in a country,as well as low economic development (Queiroz and Visser 2001). It is essential to eli minate orminimize the perception of corruption in PPP programs so that such programs can bestcontribute to a country’s economic development.

Some governments have adopted procedures to transform unsolicited proposals for privateinfrastructure projects into competitively tendered projects. Such countries include Chile, theRepublic of Korea, the Philippines, and South Africa (Hodges 2003).

IV. STEPS TO LAUNCH A PPP PROGRAM IN HIGHWAYS

A first step in launching a PPP program in highways in a country is to define the priorityprojects where the government envisages to elicit private investors financing of the total orpartial cost of the project. In the case of Russia, for example, several high priority projects forpotential PPP in highways have been desc ribed, such as Moscow-St. Petersburg motorway,outer Moscow ring road, Moscow -Minsk highway, access to Domodedovo airport, St.Petersburg bypass, bridge on Volga river at Volgograd.

Other steps to launch a PPP program would include (some of these steps ca n be done inparallel):

a. Enact relevant legislation (e.g., concession and toll road laws) .

b. Carry out feasibility study of priority projects. Employ reputable consultants, using wellprepared terms of reference (TOR). Identify/quantify social and ec onomic benefits; carry outfinancial assessment to help check the potential for attracting private capital (e.g., relativelyhigh overall financial rate of return and return on equity) .

c. Carry out environmental and social assessment, including mitigation plan and land

acquisition plan for the right of way

d. Prepare bid documents to select the concessionaire (or concessionaires) .

e. Assess the willingness of users to pay; review tolling / payment options (e.g., actual tolls,

shadow tolls, vignette system).

f. Draft concession and other agreements; define performance standard for the new

investment and the service standards during the operation period .

g. Carry out prequalification of potential concessionaires .


h. Invite prequalified firms/consortiums to su bmit bids.

i. Sign concession agreement with the best evaluated bidder .

j. Reach financial closure.

k. Monitor the performance of the concessionaire over the life of the concession .

International financial institutions (IFI) such as the World Bank can coo perate and assist inall of these steps. Forms of assistance may include:

a. Technical assistance to all required processing stages, including establishing a goodregulatory capability.

b. In case the project requires government subsidies (e.g., governmen t contribution to partof the construction cost), the IFI could consider financing a part of the subsidies .

c. The IFI could consider providing a partial risk guarantee (PRG) to support the selectedconcessionaire so it can borrow at lower interest rate an d longer maturity.

V. WORLD BANK PARTIAL RISK GUARANTEES

The World Bank through its guarantee instruments can help accelerate growth in transitionand developing countries by mobilizing private financing for infrastructure development andother projects of national importance.

By covering government performance risks that the market is not able to absorb ormitigate, the World Bank’s guarantee mobilizes new sources of financing at reduced financingcosts and extended maturities, thereby enabling commercial /private lenders to invest in projectsin transition and developing countries. Guarantees can mitigate a variety of critical sovereignrisks and effectively attract long-term commercial financing in sectors such as power, water,transport, telecom, oil and gas, and mining. Guarantees can also enhance private sector interestin public private partnerships. It can also help sovereign governments access the financialmarket.

The World Bank’s presence in transactions is seen by investors as a stabilizing factorbecause of the World Bank’s long term relationship with the countries and policy support itprovides to the governments. The World Bank Guarantees help catalyze private financing,

which leads to greater job and income opportunities for people, and thus co ntribute to the

achievement of the Millennium Development Goals’ overall challenge of reducing poverty. Awebsite dedicated to World Bank guarantees is available at:


XTGUARANTEES/0,,contentMDK:20267847~hlPK:545970~menuPK:64143502~pagePK:6 41

43534~piPK:64143448~theSitePK:411474,00.html



A World Bank operational policy regarding its guarantee program, OP 14.25, states that aguarantee objective is to mobilize private sector financing for development purposes. The Bankmay guarantee private loans with or without an associated Bank loan; the Bank does notguarantee equity investments. The Bank provides guarantees only to the extent necessary. Theoperational policy is available at:

http://wbln0018.worldbank.org/institutional/manuals/opmanual.nsf/toc1/A505EC4B4C9EB1658525672C007D0976?OpenDocument

Although guarantees may be structured in different ways, there are two basic kinds. Partialcredit guarantees cover debt service defaults on a specified portion of a loan, normally for apublic sector project. Partial risk guarantees cover debt service defaults on a loan, normally for aprivate sector project, when such defa ults are caused by a government's failure to meet itsobligations under project contracts to which it is a party. The nature and scope of governmentcontractual undertakings that the Bank backs vary depending on specific project, sector, andcountry circumstances. The Bank requires that the underlying contracts for partial riskguarantees contain appropriate dispute resolution procedures; if there is a dispute about thegovernment's obligations, the Bank's guarantee is triggered only after the government's liabilityhas been determined in accordance with such procedures. Both kinds of guarantees may coverscheduled interest as well as principal payments on a loan.

Both governments and the private sector benefit from a guarantee. Governments benefitbecause it:

Catalyzes private financing in infrastructure .

Provides access to capital markets .

Facilitates privatizations and public private partnerships .

Reduces government risk exposure by passing commercial risk to the private sector .

Improves impact of private sector participation on tariffs .

Encourages cofinancing.

The private sector benefits because it:

Reduces risk of private transactions in emerging countries .

Mitigates risk that the private sector does not control .

Opens new markets.

Lowers the cost of financing.

Improves project sustainability.

http://wbln0018.worldbank.org/institutional/manuals/opmanual.nsf/


The World Bank guarantee instruments have proved to be a powerful instrument in

catalyzing private financing to frontier markets. Twenty two guarantees with about US$ 1.4

billion exposure to the Bank have achieved a remarkable leverage by catalyzing more than US$

12 billion of private resources for projects worth US$ 24 billion (see examples in Figure 1).

Each dollar of guarantee financing has catalyzed close to 5 dollars of private finance.

Partial risk guarantees are particularly relevant in the context of seeking more private

involvement in the financing of road infrastructure. Such guarantees cover specific government

obligations spelled out in a support agreement with the project entity. Example of such

agreements include concession agreement, implementation agreement, build -own-operate-

transfer (BOOT) contract, build -own-operate (BOO) contract and, the most common form,

build-operate-transfer (BOT). Partial risk guarantees are appropriate for enhancing a pr oject’slimited recourse project financing, the most common method of financing concessions for

transport infrastructure. Figure 2 provides an illustration of how a partial risk guarantee can

apply to a highway concession contract (Queiroz 1999).

Figure 1. Examples of guarantees’ leverage in catalyzing private resources

0 500 1000 1500 2000

China Yangzhou

Philippines Leyte

PakistanHub

China Zheijiang

Jordan Telecom

China Ertan

Pakistan Uch

Lebanon Power

Morocco Jorf Lasfar

Thailand EGAT

Cote d'Ivoire Power

ArgentinaPBG

Colombia PBG

Bangladesh Haripur

Vietnam Phu My 2 -2

US$ Million

Guaranteed Amount

Private Capital Mobilized


Government

Obligations:

• Toll rate

• Permits/consents• Forex

• Change in law• Political events

• Termination

Private

Lenders

Concessionaire Loan Agreement

WBGuarantee

CounterGuarantee

Concession

Project company obligations: Construct and operate

highway; maintain and rehabilitate to keep up quality.

World

Bank

Figure 2. Structure of a Highway Concession Contract and World Bank Guarantee

VI. GREENFIELD AND ROAD MAINTENANCE CONCESSION PROGRAMS

Greenfield PPP projects include investment in new construction, usually on a newalignment, by the concessionaire, while in road maintenance/rehabilitation/operation (RM/R/O)concessions the concessionaire agrees to assume responsibility for an existing road or pa rt of aroad network. Several concession options are available and each country should select the mostappropriate for its prevailing conditions. Through the most typical forms of concession, acountry can transfer to the private sector the responsibilit y to: (i) build, operate and transfer(BOT) back to the public sector (at the end of the concession period) a road facility (e.g., amotorway, bridge, tunnel), or (ii) maintain, rehabilitate, operate (RM/R/O concessions) anexisting road or road links. Each concession can include individual links or a set of links in agiven area of the country (i.e., area -wide concessions).

Steps in the process of launching a road concession program include drafting of all relevantdocuments, a competitive selection of concessionaires, evaluation of proposals, as well as awardof the concession contract. When the main purpose of the concessions is to obtain extrabudgetary funding for roads, or release limited public funds for use on other roads (e.g.,secondary and rural roads), shadow-tolls (whereby payment to concessionaires are made out ofthe budget, based on traffic volumes and classification) would not be a feasible option.

All concessions require an institutional means, such as a unit, to monitor the private s ectorperformance, including compliance with the performance standards defined in the concessionagreement. The concession contract and public - private sector arrangements for a newinvestment concession and for a maintenance concession may be different , but a country canpursue both types of concessions at the same time.


VII. WORLD BANK TOOLKIT FOR PPP IN HIGHWAYS

The main objective of the World Bank Toolkit for PPP in Highways is to provide policy

makers from economies in transition with some guidance in the design and implementation of a

Public Private Partnership (PPP) in the highway sector. The Toolkit is structured under five

headings (or modules) and is navigated through a series of tree diagrams under each of these

headings. It also includes a library and interactive financial models. It is a multimedia product

available on a CD ROM and also available from the World Bank's web site at:


or


Using basic assumptions about a specific motorway project, the financial simulation tool of

the Toolkit is helpful to answer key questions on t he financial feasibility of the project. For

example, questions such as the ones below can be answered with minimum effort using the

Toolkit:

- What is the internal financial rate of return (IRR) of the project?

- In the absence of Government subsidies, c eteris paribus, what would be the return on

equity (ROE)?

- While subsidies may be paid by the Government during the construction period, it

recovers some of this payment through taxes during the operation period. What would be the

Government contribution to the proposed project that would lead to a financial balance for the

government throughout the concession period?

- In the absence of Government subsidies, ceteris paribus, what would be the required

initial toll rate to yield a return on equity (ROE) of 16%?

- Assuming that an initial average toll rate of US$0.06 per vehicle -km is the highest

acceptable by road users, an investment cost of US$3 million per km (typical for a four -lane

road on flat terrain), and an initial traffic volume (AADT) of 15,00 0 vehicles per day, what is

the minimum concession life that would generate a return on equity (ROE) likely to attract a

private sector concessionaire (say an ROE of 15% or higher)?

A recent update of the financial simulation tool is particularly appropri ate to answer the

above questions. The Excel file with the updated Tool is available on the World Bank website

at:

http://wbln0018.worldbank.org/ECA/Transport.nsf/ECADocByLink/01C97A272081983D

85256FD20061ECB8?Opendocument



http://wbln0018.worldbank.org/ECA/Transport.nsf/


VIII. CONCLUSIONS

This paper discussed potential applications of partial risk guarantees to assist countries intransition to seek more private involvement in the fi nancing of road infrastructure. It alsopresented a review of the required legal framework (e.g., concession law) for attracting privatecapital for PPP schemes, possible steps for a country to launch a program of privateparticipation in highways, the concept of greenfield and road maintenance concession programs,and the treatment of unsolicited proposals. The paper also summarized potential applications ofthe World Bank Toolkit for PPP in Highways as an instrument to help decision -makers andpractitioners to define the best PPP approach for a specific country

IX. DISCLAIMER

This paper reflects only the authors’ views, and should be used and cited accordingly.The findings, interpretations, and conclusions are the authors’ own. They should not beattributed to the World Bank, its Board of Directors, its management, or any of its membercountries.

References

[1]. Harris, C., Hodges, J., Schur M., and Shukla , P. 2003 “Infrastructure Projects: A Review of CanceledPrivate Projects” Public Policy for the P rivate Sector, Note No 252. Washington, D.C.: The World Bank

http://rru.worldbank.org/Documents/PublicPolicyJournal/252Harri -010303.pdf

[2]. Hodges, J. 2003 “Unsolicited Proposals - Competitive Solutions for Private Infrastructure” PublicPolicy for the Private Sector, Note No 258. Washington, D.C.: The World Bank


[3]. Hodges, J. 2003 “Unsolicited Proposals - The Issues for Private Infrastructure Projects” Public Policyfor the Private Sector, Note No 257. Washington, D.C.: The World Bank


[4]. Kerf, Michel, et.al. 1998. “Concessions for Infrastructure: A Guide to Their Design and Award.”World Bank Technical Paper No. 399. World Bank, Washington, D.C.

[5]. Queiroz, C. and Visser , A. 2001 "Corruption, Transport Infrastructure Stock and Economic

Development." Infrastructure and Poverty Briefing for the World Bank Infrastructure Forum, CD -ROM.

World Markets Research Centre Ltd. Washington, D.C.: The World Bank.

[6]. Queiroz, Cesar. 1999. “Highway Concessions and World Bank Guarantees.” International RoadFederation Regional Conference on European Transport and Roads. Lahti, Finland, 1 4-16 June 1999.

[7]. World Bank Toolkit for Public Private Partnerships (PPP) in Highways.


http://rru.worldbank.org/Documents/PublicPolicyJournal/252Harri-010303.pdf




CT 2

I. INTRODUCTION

Vibration and oil analysis can reveal a

great deal of information about machine’shealth. Therefore, vibration and oil analysis

are two key components for machine

condition monitoring. Oil analysis has been

used for at least fifty years in determining the

wear condition of machinery. Rails road

companies in the late 1940s and early 1950s

found that the metals in a sample of used oil

revealed the condition of the wearing parts in

their locomotive engines. Vibration and oil

analysis today are used to monitor the

condition of everything from aircraft jet

engines and helicopter gearboxes to

construction equipment industry, commercial

transportation, and industrial plants. Oil

analysis is taking place alongside vibration

monitoring as an indispensable and valuable

predictive maintenance tool in industry. This

paper presents a study of integrating vibration

and oil analysis for diesel engine condition

monitoring based on the result of oil and

vibration analysis at Hanoi Locomotive

Entreprise, Vietnam.

II. CONTAINS

2.1. Oil analysis applicationThe wearing parts of a machine such as

the gears, hydraulic pistons, bearings, andwear rings generate fine metal particles duringnormal operation. At the onset of a severewear mode the particle size increases and theappearance of the particles change.Knowledge of the particles and how theyrelate to the mode of wear permits a trainedanalyst to determine the wear status in amachine by measuring the fine and coarsemetal particles and then examining theparticles under a microscope. The testing forwear metals for condition monitoring andpredictive maintenance is testedpredominantly in spectrometric analysis or inwear debris analysis.

The advantages of oil debris monitoringcompared with other monitoring methodsinclude:

- The evidence in the oil is to be foundnowhere else.

- The cost benefit ratio is better thanother technique.

- The oil carries evidence of faults fromvarious parts.

2.2. Chemical identification of debris

Quantitative measurement i s oftenrequired for many machine condition

DIESEL ENGINE DIAGNOSISBY VIBRATION OIL ANALYSIS

Prof. DR. DO DUC TUANMEng. LE LANG VANUniversity of Transport and Communications

Abstract: This paper presents a research of oil and vibration analysis in machine

condition monitoring and evaluating the use of oil and particle analysis in practice. The paper

also mentions to capable of this method in diesel engine diagnostic based on the result of oil

and vibration analysis at Hanoi Locomotive Entreprise, Vietnam.


CT 2

monitoring applications. Quantifying debrisgives a feel for the likely wear that isoccurring in machines. The measured mass ofdebris is determined to know any change inthe trapped quantity, such as weight per ml,intensity per ml, or shape of the sizedistributions. Chemical identificationinstruments used in this research is Alloy Pro9388 Metal Analyser (Figure1).

Figure1. ALLOY PRO 9388 Metal Analyser

Figure 2. Chemical identificationof piston of D12E engine

2.3. Wear particle image analysisOil samples and vibration data which

were collected regularly at the HanoiLocomotive Entreprise over a period of 14months were carefully examined andcompared. A particle analyzer was used todetermine oil sample to assess the generaltrend of diesel engine conditions. Normalwear process from oil samples of locomotiveNo 660, 657 and 658: number of particles andparticle dimensions were small (from 0.2 to 5

micromet).

Figure 3. Particles in oil sample(Locomotive No 660)

Figure 4. Particles in oil sampleis increasing in numbers

(Oil sample on Locomotive No 658)Various types of wear process give

various types of particle shapes: spheres,fibers, slabs, curls, spirals and slivers, rolls,strands and fibers.

Figure 5. Sphere shape particle

(In oil samples of locomotive No 660)

Sphere shape: the presence of spheres in

2.298 μm

7.423 μm

13.64 μm

2.347 μm

2.381 μm

2.298 μm


CT 2

oils is quite frequent. Sometime spheres arefound in the new lubricating oil from acontainer.

Figure 6. Pebble shape particle


Figure 7. Chunks and slabs shape particle


The smooth sphere is often found in therunning-in period. The rough seems to involvemore severe wear. The other shapes include:distorted smooth ovoid/pebble shape, c hunksand slabs, curls, spirals and slivers, rolls,strands and fibers. The research process drawsthe following conclusion:

Rubbing wear and normal wear: are regularwear particles which are formed from betweenlubricated sliding surfaces. They would take theshape of ‘platelets’ up to 10 µm.

Cutting wear: These particles are formedby the metal parts digging into each other [2].

Rolling fatigue: Spherical particles

appear quite often. It would have to beassumed that the spheres come from fatiguecracks in bearings. Chunks of metal canappear from fatigue with the size up to 100µm. Another form of particle is the platelet,perhaps up to 50 µm.

Severe sliding wear: These are also largeparticles depending on the magnitude of thesliding action load and speed. Usually in theform of platelets with surface st ress acted onthe wearing surfaces. The higher the stresslevel, the larger the ratio of large particles tosmall particles [4].

2.4. Particle Dimension

Dimension of the particle changes fromseveral micromets to 100 or 300 micromets.Particle shape varies depends on wearprocces. A large dimension of particle informssevere wear process. The research indicates:

Normal wear procces produces particleswith sphere shape and their dimensions arebetween 5 and 10 micromet.

Cutting wear is caused when an abras iveparticle has imbedded itself in soft surface ofcopper alloy wear.

Fatigue wear occurs when cracks developin the component surface that leads togeneration of particles. Particle dimension isup to 100 micromet.

Sliding wear evolves during equipmentstress. The dimension of particle is more than10 micromet.

The particle dimensions from oil samplesof fault engines on locomotives were between100 micromets and 120 micromets.

2.5. Number of wear particles

Determining the number of wear particlesis one of demands for diagnosis procces. Thenumber of wear particles per millilitre gives

7.423 μm

3.347 μm


CT 2

an explicit, easy-to-understand format forcharacterising wear conditions of lubricatedtribosystem monitored. The numbers of wearparticles per millilitre counted from oil samplecollected from fault engines on locomotiveswere over 100 particles per millilitre.

Vibration Analysis

The equipment used for measurement andanalysis includes: vibration meters, analyzer andDasyLab software. The experiments werecarried on Hanoi Locomotive Entreprise overperiod of 14 months.





The vibrations are taken in the threeCartesian directions. In vibrationnomenclature, these are the vertical,horizontal and axial directions. This isnecessary due to the construction of machines– their defects can show up in any of three

directions and hence each should bemeasured. The data collector can collect andstore the data for comparision and trending.The database program stores vibration dataand makes comparisons between currentmeasurements, past measurements andpredefined alarm limits. Alarm limits forlocomotives D12E are 20mm/s and 7mm/s 2.

III. CONCLUSION

The diagnosis technique of combustionengines by oil and vibration analysis has beenresearched in several countries and has somesuccess. This research integrates vibration andoil analysis for diesel engines of K6S230DRused on D12E locomotives. When wearparticle dimensions exceeds 120 micromet,when concentration of metals (Cu, Fe, Cr) hasbeen increasing, shapes of oil debris areunnormal, vibration velocity and accelerationexceeds 20mm/s and 7 mm/s 2, engine faultsmay occur on piston, valves or piston rings.Oil analysis confirms the results of vibrationanalysis. Both wear debris and vibrationanalysis techniques were used to assess thediesel diagnose problems during this research.Oil debris analysis confirm the conclusionsabout the faults of diesel engine elementswhen had vibration alarms.

Reference[1]. Calder, N., Marine Diesel engines:Maintenance, troubleshooting, and repair,International Marine, 1992.[2]. Bowen, E.R. & Westcott, V.C. , Wear particleatlas, Naval Air Engineering Center, 1976.[3]. Doebelin, E. O., Measurement systems,McGraw- Hill Companies, 1990.[4]. Hunt, Trevor M., Handbook of wear debrisanalysis and particle detection, Elsevier AppliedScience, 1993.[5]. Rao, J.S., Vibratory condition monitoring ofmachines, Published by Addison -Wesley,America, 2000


Within many years in the economicplans and forecasts the regions of the Asianpart of Russia - Siberia and Far East - areconsidered as the most perspective.

The main precondition is their significantnatural-raw potential. In the Asian part of thecountry, where on 43 % of territory 6 percentsof the population live only, more than 60percents of ores of non-ferrous metals, almost100 percents of diamonds, third wood and twothird of fish resources are concentrated.

Just here there are strategic stock s ofcarbohydrates, due to which Russiastrengthens the global political and economicpositions.

One more feature of the Asian part ofRussia - direct boarder with China and Japan.This factor, in opinion of a number of theexperts, creates opportunities o f directregional integration in actively growing

economic systems of Asia-Pacific region.There is the possibility of construction ofcircuits of deliveries in a direction Asia -Europe on the basis of transit potential ofRussia, first of all – “Transsib Railway”.

The transport development of Siberia andFar East objectively requires a scientificsubstantiation, development and realizationunique paradigm. It should answer not onlypriorities of socio economic development ofthe Asian part of the country, but also strategyof development of Russian Federation as awhole.

Major element of this paradigm willbecome creation in Siberia and in Far East ofthe country the system of the logistic centers.

The concept of logistic centers grows outof searches of alternative ways ofdevelopment of transport system, which beactively ordered in the advanced countriessince the seventieth years.

FEATURES OF CREATION OF LOGISTIC CENTERS INCONDITIONS OF SIBERIA AND FAR EAST OF RUSSIA

Prof. VYACHESLAV M. PRIKHODKOProf. VICTORIA D.GERAMIProf. ALEXANDER V. KOLIKState Technical University-MADI, Moscow, Russia

Abstracts: In the paper the transportation problems of Siberia and Fa r East of Russia are

discussed. It was suggested to create a number of transport logistic centers in the region on the

basis of hub system (“Hub-Spoke”) to increase efficiency of transportation of goods.

Key Words: Transport of Russia, Siberia and Far East, logistic centers, hub system (“hub -

spoke”), logistic operation, articulated lorry, transit, Public Private Partnership (PPP)




will have a number of features.

First of all, logistic centers of the Asian

part of Russia should play a role intermodal

and unimodal transport hubs.

The organization of transportations in

system of hubs (differently, in system “hub

and spoke”, is effectively applied in sea

container business and in aircraft. This system

is used and in practice of automobile freight

traffic of a number of countries, for example,

American continent. Recently it all use in

system ground intermodal transport more

active.

In similar system a through servi ce

between items of a transport network are

replaced with a combination of transportations

between the allocated items - so-called hubs -

and traffic between points of origination

(destination) and hubs serving the appropriate

zone.

It is not enough for creation of system

“hub and spoke” only intermodal terminals.

The uniform system of organization of

transport process is necessary, at which the

disorder competition of the transport operators

is inadmissible. It should be replaced by

cooperation in conditions of deep functional

specialization and division of the market

between segments of main and regional

transportations.

The governing advantage, which is

provided by the system “hub -spoke” consists

in reduction of general number of transport

connections and due to it - concentration of

freight flows (Fig. 1). It allows to achieve

economy of scale and reduction of total costs.

For transport development of the Asian part of

Russia with its rather weak freight flows this

factor is represented critically impor tant.

The second feature of development of

creation of logistic centers in regions of

Siberia and Far East lies in accommodation

and specialization of these objects by an

essential image will determine development

of an infrastructure of different modes of

transport and distribution of cargo bases

between them.

By virtue of the marked above features ofa transport network of the Asian part ofRussia (fragmenting and weak developmentof a modal infrastructure) logistic centers willbecome natural units of joining existing andrecreate transport infrastructures. Thedecisions on accommodation and thesequences of input in build separate logisticcenters, in turn, will define(determine)priorities of selection and realization of theprojects of the transport communications.Differently, the network in considered regionshould develop in many respects not by aprinciple “from cities to city”, and by aprinciple “from logistic centers to logisticcenter”.


It is obvious, that in view of

characteristic for Siberia and Far East of

distances a key main kind in system of logistic

centers should be a railway transportation. In

view of it logistic centers should be created,

first of all, in those points, where is possible

and is necessary it effective joining with lines

of other modes of transport (Fig. 2).

Internal water transport, by keeping the

role of an alternative kind of communications

in the great Siberian rivers basins, will receive

reliable connection with ground transport

system through logistic centers placed in the

largest ports.

Cargo hubs of air transport, which

creation is actively discussed last years, first

of all, with reference to region of Siberia,

should by a natural image “blend with” in

system of logistic centers.

As to road transport, its task, first of all,

should become effective transport service of

zones of gravitation of logistic centers with

granting to clientele of maximal volume of

logistic services adding intermodal

transportation on a site supply- conveyance.

a)

Hub

Hub

b)

Hub

Hub


Fig 2. Transport service of the system of logistic centers

Regular rail transportation

Transportations by the internal water ways

Air Transportations

Supply – Deliver by road transport

Main road transportations

For logistic centers of Siberia and Far

East a zone of gravitation will be much more

extensive, than for similar objects in the

European part of Russia. ”First” and “last

mile”, as sometimes call supply - conveyance

of transportation abroad, can be stretched on

tens kilometers. Therefore separate important

task will become development and essential

increase of quality of a road network in zones

of gravitation of logistic centers - and it, in

turn, can become the essential factor of

regional development and tool of the decision

of a task of connection of the isolated today

occupied points with a basic transport network

of the country. In this case speech will go

about stable connection of the occupied points

not with a transport network in general, and

with national logistic system.

At the same time, in conditions of Siberiaand Far East the role of road transport can notbe limited to regional service. The mainautomobile transportations will becomenecessary on those directions, where logisticcenters for whatever reasons will not h aveamong themselves of regular railwaycommunication. In these conditions willbecome economically justified the realizationin Russia of the concept of trailers especiallyof large carrying capacity, which finds inworld practice rather wide application.

Now in European Union the idea ofincrease of allowable length and maximalcomplete weight of the trailers for theinternational transportations is activelydiscussed within the limits of the EU.


The most serious reason for the benefit of

this offer is the expected reduction of number

of vehicles on a road network, that critically

important for the overloaded European

highways. The efficiency and safety of

heavier trailers is proved by successful

practice of a number of the European

countries (Sweden, Finland and others), where

the national requirements already now allow

their application.

The idea of “road train” for a long time

and effectively is used also in a number of

states of USA, in Australia, Mexico, Brazil

and in other countries. The experience of

Australia, in particular, shows, that the

operation of similar vehicles is real at the

lowest axial loading of 6 tons.

In Russian Federation currently in use

road restrictions are established enough

arbitrary. Macroeconomic approach to

parameterization of the complex “road costs -

cost of a vehicle - cost of transportation” and

task solution of joint optimization of strength

properties of highways and such parameters of

lorries, as fully loaded mass and axial

loadings, undoubtedly, would allow to

achieve significant economic benefit in scales

of economy as a whole. However, not waiting

for statement and decision of this problem at a

national level, the introduction of the special

system of the road specifications for the

certain regions of the country, in particular -

for Siberia and Far East is represented quite

pertinent and useful. Primary factor of

efficiency should become significant

reduction of the cost price of automobile

transportations on those directions (including

between separate logistic centers), where the

vehicle is no alternative one. The

transportations by “road trains”, thus, will

become effective addition to system of main

rail transportation.

One more feature of logistic centers of

Siberia and Far East will be, obviously to

consist and that their creation will require a

new level of PPP in development of an

infrastructure.

The PPP idea is very popular both in

West, and in Russia. In our country the first

legal preconditions for its development,

including, for transport are created.

More often PPP assumes a role of the

private partner, first of all, as investor, which

considers the PPP project as an opportunity of

effective investments of free financial assets

at the certain guarantees and support on the

part of the state. But in the Asian part of

Russia it is required, probably, some other

approach to a choice of the partners in the

PPP projects.

First, the speech goes about the largest

enterprises of region, which will be interested

in creation of industrially logistic centers, first


of all, for maintenance of requirements of own

manufactures and chain of deliveries.

The prototype of such model can serve

logistic centers created in Germany in

partnership of authorities of city

Ludwigshafen and the chemical concern

BASF for transport service of new industrial

complex of the company. One figure -

processing of 300 thousand containers per one

year testifies only to capacity this largest in

the sort of object. But its efficiency is caused,

in many respects, that it is simultaneously

object of for general usage. Freight flows of

BASF are integrated on it with flows of other

users located in the given region.

Such model completely answers city -

forming function of the large enterprises,

typical for Siberia and Far East, and will

allow to effectively realize this function in

conditions of market economy.

The second group of the target partners

are large companies and groups of the

companies, which today independently

develop the marketing and transport networks

in the Asian part of Russia. The association of

their potential in frameworks of PPP will

create additional network effect both for all

participants of such partnership, and for

territories.

Characterizing the special role logistic

centers in transport system of the Asian part

of Russia, it is impossible to bypass a theme

Trans-Siberian transit.

The mention of huge unused potential of

Trans-Siberian transport bridge for a long

time has become a general place. The

development of transit within many years is

imperishable priority of the high level and is

considered as a major point of growth of

economy of Siberia and Far East.

Between that, from middle of the

ninetieth years, when this theme has become

actively discussed by the experts in the

transport markets there were changes, which

essentially have changed a ratio of the tariffs

on competing Asian - European routes not for

the benefit of Russia.

The sea container operators continued toincrease individual tonnage of linear ships,achieving a scale effect and stabilization - andin a number of cases, and reduction - tariffsfor sea container transportations.Simultaneously, during reforming the Russianrailways the internal cross subsidizing offreight traffic was liquidated which wasdistributed, in particular, to container transit.It objectively has resulted in increase of thetransit tariff on transportations of containersthrough Russia.

Thus, such advantages Trans-Siberianroute as, for example, shorter transit time,were appreciably shown on is not present byaction of the price factor. There are all basesto assume, that within the framework of thetraditional scheme of transit transportation atthe usual structure of costs and tariffs to


Import of goods from countries-exportersLogistic dealership

Export of goods to the countries of Europewith additional added costRegional distribution of goods withadditional added cost

Fig 3. Variant of Trans-Siberian transit “with additional added cost”

achieve essential growth of Trans -Siberiantransit it will be not possible. But the creati onin the Asian part of Russia logistic centersallows to realize a little bit other form of theAsian - European transit communication.

The cargoes addressed in the countries ofEurope, can be exposed at these centersadditional logistic processing. The speechgoes about such operations, as palleting,regrouping, packing, marks and others, whichnow are carried out by the countries inWestern Europe and consequently manage tothe importers extremely dearly. The inclusionof these operations in the Russian scheme ofdeliveries of cargoes from Asian -PacificRegion will add simple transit transportation

by much cheaper services in escalating theadded value of a final product (Fig. 3).

Other important function of logistic

centers of Siberia and Far East connected with

Asian freight flows, should become unloading

of Moscow and St.-Petersburg as “obligatory”

points of transshipment of the Asian import in

Russia. Thus the integration of import and

transit flows, and also logistic operations will

allow achieving additional economic benefit.

The enterprises and inhabitants of region

should not in addition pay transportation of

inward cargoes in the Moscow region and

Import of goods from countries - exporters

Logistic dealership

Regional distribution of goods withadditional added cost

Export of goods to the countries of Europe withadditional added cost


then again on East of Russia.

THE CONCLUSION

The economic development of Siberia

and Far East of Russia should carry advanced

character. For modern transport and logistic

technologies in this process the special place

should be assigned.

The transport development of the Asian

part of Russia has a number of features, which

do not allow to apply in this region model of

simple escalating of extent and increase of

density of transport networks. Paradigm of

“dotty” transport development of Siberia and

Far East should be based on creation of

system of regional logistic centers.

Logistic centers should be created as

compact technological objects, on which the

independent operators will carry out a

complex of the functions directed on

coordination and integration of logistic flows

and on increase their added cost. The

important factor of efficiency should become

optimum accommodation regional logistic

centers on a transport network and

organization of the stable communication

between them.



will have a number of features, in particular:

- Logistic centers of the Asian part of

Russia should play a role of transport hubs. It

will allow achieving a high degree of

concentration of freight flows and will raise

efficiency of transport process.

- Accommodation and specialization of

these objects in many respects will be defined

by priorities of selection and realization of the

projects of development of an infrastructure of

separate modes of transport and distribution of

cargo base between them. Thus the transport

network of the Asian part of the country will

develop by a principle “from logistic center to

logistic center”.

- The development and increase of

quality of a road network in zones of

gravitation logistic centers can become and

tool of the decision of a task of connection of

the isolated occupied points with a basic

transport network of the country and with

national logistic by system as a whole.

- The transportations between logistic

centers should be carried out, first of all, by

railway transportation. However on those

directions, where the regular railway

communication for whatever reasons will be

absent, the main automobile transportations

will be claimed. Thus the realization of the


concept of truck trains especially of large

carrying capacity is expedient .

- The creation logistic centers on the

basis of PPP in the Asian part of Russia is

expedient with attraction of the largest

enterprises which are carrying out city -

forming functions. The appropriate objects

(industrial logistic centers) will have the

“mixed” character, providing requirement of

large industrial complexes and working

simultaneously as logistic centers of general

usage.

- The creation in the Asian part of Russia

logistic centers will allow to realize the

modified variant Asian - European of transit

communication, at which the cargoes

addressed in the countries of Europe, will be

exposed additional logistic processing, and the

transit transportation will be complemented

by rather cheap services in escalating the

added value of a final product.

Reference

[1]. Materials of session of State Council of

Russian Federation and official documents.

October, 2003 - Moscow, State Council of Russian

Federation, 2003.- 514 p. (in Russian).

[2]. Prikhodko V.M., Kolik A.V., Gerami

V.D.Scientific The Profblems of Motor

Transportation Ensuring of National Logistic

System.- M.: Publ.House

“Technopoligraphcenter”, 2006. - 91 p. (in

Russian).

[3]. Prokofyeva T., Platonov S . Formation of

Transport-Logistic Infrastructure of Russia. –“Container Business”, № 1, 2005. - pp. 10-17. (in

Russian).

[4]. Belyaev V.M. Terminal Systems of

Transportations of Cargoes by Road Transport. -

M.: Transport Publ.House, 1987. - 282 p. (in

Russian).

[5]. Oreshin V.P. Planning of an Industrial

Infrastructure. The complex approach. - M.:

Economy Publ.House, 1986, -142 p. (in Russian).

[6]. Best Practice Handbook for Logistics Centres

in the Baltic Sea Region, 2003,


Handbook.pdf < http: //


Handbook.pdf >

[7]. Dirk Berendt, Importance of Harbors in

Logistic Chain. Tasks and opportunities, materials

of a conference “Transport and Logistic inInternational Trade”, Euroasian Transport Union(ЕАТС), 2004, www.eatu.org < http: //www.eatu.org >.

[8]. J.C.Johnson, D.F.Wood, D.L.Wardlow, Paul

R.Murphy. Contemporary Logistics. - Upper

Saddle River, NJ.: Prentice Hall, 1999. - 590 p.

[9]. Technological change and Multimodal Freight

Competition. By J.L.Courtney. Proceedings of the

Transportation Research Forum. - Boston, MA,

1984. - p.p. 116-121

[10]. White Paper - European Transport Policy for

2010: Time to decide. - Luxembourg, EC Official

Publications Department, 2001. - 150 p.p



www.eatu.org


CT 2 I. ENVIRONMENT PROBLEMS ON ROADS

The increase in the traffic volume (up to 40 -70 and more thous. cars per day at the largecities exits and 10-20 thous. on the majority of the federal roads), construction of new andreconstruction of the existing roads ha ve aggravated the problem of environmental protection.When considering the ecological problems in Russia considerable attention is traditionally anddeservedly paid to the automobiles. The result of it is the significant progress in the sphere ofengine-building accompanied by the sharp reduction in emission of harmful substances. In theyears coming Europe is planning to introduce standard EURO -5, Russia - standard EURO-2 andlater on - standard EURO-3. In the USA the President George Bush announced the d evelopmentof the new ecologically friendly car engine. But apart from the cars there are other factors thatcontribute to the environmental pollution. The most significant of them are highways affectingthe level of pollution coming from traffic. Moreover , the road itself has negative influence withthe roadside territory. Such influence is exerted by:

Road engineering elements: roadbed, bridge crossings and flyovers, water intakestructures and culverts.

Separate road engineering structures: pavement, roa dbed, shoulders.

Road infrastructure units: rest area, gasoline stations, food stations, public transportstations.

THE ENVIRONMENTAL PROBLEMS CONNECTED WITH

HIGHWAY CONSTRUCTION AND MAINTENANCE

Prof. MIKHAIL V. NEMCHINOVDr. ALEXEI S. MEN’SHOVState Technical University - MadiDr. DMITRY M. NEMCHINOVThe Association of Road Design Institutions of RussiaDr. VERONIKA OSINOVSKAYABryansk State Technological Academy, Bryansk, Russia

Abstracts: In the paper the environmental problems connected with road construction and

maintenance, trends and ways of their solving are formulated and presented. The problems with

deformation of the road embankment slopes are considered.

Key words: ecological safety of roads; paramet ers of ecological safety, road embankment,

slope of road embankment


CT 2

The above listed sources of highway influence over the environment affect all the natureelements: air, soil, water, biosphere. Air pollutio n over the highway is influenced by theroadbed, pavement surface material and texture, even interchanges design and peculiarities oftraffic. The roadbed in the form of a high embankment affects the thermal, humidity and windconditions of the roadside territory.

Roadway paving influences the quality and composition of the automobile exit gases, thequantity of wear debris of the automobile parts, including automobile tires, air pollution by thewear debris of the roadway covering, dust and garbage and it is an important factor of theformation of the level of traffic noise. Constructional features of the road crossing, means andmethods of organization and traffic control of automobiles also influence the quantity of the exitgases released by the automobi le engines.

The impact of automobile road on soil and water is no less diversified. The landscape ofthe area changes as a result of exemption of territory for the engineering constructions of theroads, careers, earth-deposits, construction sites, industr ial approaches. As a consequence ofdevelopment of the road network there occurs the fragmentation of the territory, change ofterrain and flora. The construction of grade level, bridges and crossovers is followed bydeformation of sub-base, development and strengthening of erosion processes. The regime ofrunoff of surface and ground offers is often broken, which is followed by drainage oroverdamping of territories, up to formation of marshes. It often leads to the erosion of the bed ofthe water streams, formation of ravines. The soil gets contaminated not only by the componentsof the exit gases of the automobiles but grade level erosion products, wear of roadway covering,by the materials used during the winter maintenance of roads (antiglaze reagents). Watercontamination of rivers and lakes occurs as the result of pollutant emission and impact oferosion products, wear of roadway coverings and automobile tires, dust, garbage, oil -productsand human wastes (in the locations of infrastructure facilities) .

Such kind of impact also has its consequences in biosphere: flora, fauna, including humans.The habitat of plants is limited as a result of the change of the regime of soil watering, drainageor underflooding of the territories, change of soil fertility as well as the presence ofcontaminative chemical agents. In the locations of rest areas occurs trampling and vegetationdamage, repacking of soil. The habitat of animals limits, natural migration ways change,acoustic environment becomes more complicated.

According to the law «On environment protection» in Russia the following items aresubject to protection from contamination, depletion, degeneration, damage, extermination andother negative impact of economic and other activity:

Land, the Earth's interior, soils.

Surface and underwater.

Forests and plants, animals and other organ isms and their genetic heritage.

Ambient air.


CT 2

Automobile roads belong to the objects of environmental threat. Depending on the level ofenvironmental threat they are divided in th ree classes. The first class are large objects, whichconsiderably influence the environment: federal and regional motorways and speedways of the Iand II technical categories with not less than four lanes and constructive works on them,separate bridges and crossovers with the length of more than 500 m. According to theInternational standards and Federal documents the construction of road objects of the first classbelongs to ecologically destructive activity categories. The second class represents the ob jects,which considerably influence the environment. They include the roads of II and III categorieswith predicted (perspective) rate of traffic more than 2000 veh. per day and the constructions ontheir surface, separate areas of other roads in populatio n centres and particularly protected areas.The third class is represented by the objects which have insignificantly influence, local actionon the environment: automobile roads with predicted traffic rate less than 2000 veh. per daytransport constructions on their surface, repair works.

Under environmental threat (safety, environmentally safe state) of the automobile roadthere is understood the ability of the road to provide the minimum of hazardous, formed byengineering constructions and constructions of the automobile road, impacts and pollution ofnature of the areas attached to the roads, their influence on the work of the road transport. Thelevel of environmental threat (safety) of the automobile road depends on its technical conditionand the technical state of the road buildings, the level of contamination of the naturalenvironment of the wayside, as well as influence of the technical condition of the road on thepollutant emission of the road transport.

With the purpose of quantitative estimatio n of the level of environmental security(environmentally safe state) of the road there are proposed special rates, which are divided intotwo groups – ecological and ecologically significant. The ecological ones include the rates,which characterize the level of air, water, soil pollution, bioenvironmental effect (human, flora,fauna) and reflect the cooperative effect on the nature of road transport as well as engineeringconstructions of automobile road. Ecologically significant rates include those, whic hcharacterize the technical condition of elements (constructions) of roads or maintenance works,which reflect the influence and environmental effect of the road and the effect of the latter onthe ecological rates of road transport. The level of ecologic al safety of the road is evaluated bycomparing factual and regulatory values of ecological and ecologically significant rates, statedin quantitative or qualitative form.

The state of road will be considered environmentally safe if:

There is no violation and pollution of the roadside territory, formed and caused byengineering constructions and road constructions, or they are as low as practicable with theexisting technologies and modern requirements.

There are created conditions, which provide the minima l possible (with the existingtechnologies and modern requirements) impact on nature from the side of road transport,which is at the road.


CT 2

Quantitative values and qualitative assessments of the environmentally safe state of road,its engineering facilities and constructions are represented in the branch regulatory document«Rates and norms of ecological safety of the road», prepared by the Road Agency of Ministry ofTransport of the Russian Federation and set in force since January 1, 2003.

With the purpose of environmental safety improvement of the roads in Russia there havebeen worked out the rules and norms of environmental design of road elements and roadsconstructions. The examples of such rules are the rules and standards of design and constructionof rest areas at the roads. The rules efficient up to the present moment both in Russia and abroadare made solely on the basis of requirements of road traffic safety ensuring. However the growthof traffic volume lead to the fact that rest areas are overl oaded by road transport, theenvironment of roadside territory cannot resist the excess human load. As a result rest areas donot carry the assigned functions – provide rest neither for the drivers nor for pedestrians and, bythis means, do not contribute to safety improving of road traffic.

The research of the recent years showed the significant impact of roadway coverings on thefuel consumption of automobile engines, and, in such a way, on the volume of exit gases.Beside the environment-oriented values the right choice of the material and texture options ofpavement also has an energy-conservative value. The research carried out, the results of whichare presented on the Fig. 1, revealed quite a complicated character of interrelation between thematerial and surface texture and fuel consumption in the whole actual speed ranges ofautomobile traffic. It is estimated that, on the road sections with the average speed of traffic of80 and more km/hour the minimum fuel consumption are observed on the cement -concretepavement in comparison with the asphalt -concrete. In case of traffic motion of less than 80 kmper hour there is observed quite an opposite picture.

Figure 1. Impact of the material and texture options of the roadway covering

on the motor car fuel consumption

Evenbituminousconcrete

Surfacetreatment

Cementconcrete

Fuel

con

sum

ptio

n, g

/km

Speed, km/h


CT 2

II. DEFORMATION OF THE ROAD EMBANKMENT SLOPES

The practice of construction and reconstruction of roads showed that the basic types ofdeformation of the earth embankments (roadbed), made from granular materials (sands, sandand gravel ground etc.) are surface erosion and local shear deformations in the form oflandslides, earthflows, caused by the impact of water on the ground. Such kind of deformationone may see in the regions with quite a cold climate, in the regions with snow falls, snowstormsand cold winter. This is the Northern and Central parts of the European territory of the RussianFederation, the whole territory of Siberia and Far East of the Russian Federation, Alaska (USA),high mountain areas of China (Tibet). It is confirm ed by the observations of numerous authors(Fig. 2) [1,2,3,4].

а) б)

в)


CT 2

г)

Figure 2. Local deformations of subgrade embankment:

а) road in Alaska [4]; б) Qinghai - Tibet railroads [3]; в) Russia - road «Yamal» [1]; г) Russia - road «Don» 104 km [1]

Deformations of grade level, caused by water erosion, are developed in the period, whenthe surface of formation is still not hardened and caused by considerable overspeeding of thewater flowing from the ground surface (usually dur ing the rains) of the standard (not eroding)speeds for the ground. The ways of prevention of such kind of deformations are well -known – itis timely embedment of the traffic way, waysides and slopes of grade level with the materials,which are highly resistant to the washaway.

The situation is more difficult with the deformations of the second type – shifts on theslopes. Local deformations of this type can be observed on the embankment slope of all types ofground. What is particularly interesting is the fact of appearance of shear deformations on thefill slopes from the cohesionless soil. Besides such deformations develop on the slopes of evenhigh fills (up to 8 and more m), with asphalt and cement -concrete pavements at the carriagewayand shoulders, with good grassy turfs at the slopes. Particularly often these deformations occurin the first 1-3 years of grade level service.

The possible schemes of local deformation developments are presented at the Fig. 3. In allcases there are slip lines of the defi nite coat massif of soil in the coating surface of the slope:

а) б)


CT 2

в)

Рис. 3. Development scheme of shear deformations on the fill slopes [6]:

а – due to identity element; б - inplane slip with uplift; в – destruction of the whole slope in the belt of

weathering on the circular cylindrical surface ; 1 - face of slope; 2 – capacity of the active zone ha; 3 –shift surface; 4 – assumed blocks; 5 - retaining prism in uplift zone

The condition of the slope stability is the balance or excess of restraining forces over theshear forces. Stability coefficient is:

tgαZitgψ

tgαZiγ

nСntgZiγзапR

ZiγnC

ntgZitg ψ

where γ - soil density; Zi - running coordinate of the active zone capacity of the slope

perpendicularly its surface; Zitgψ - coefficient of soil shift of the active zone h at depth Zi;

ntg , Cn - correspondingly calculated values of angle of repose and soil cohesion at depth Zi;

- rate of slope.

The analysis of the complex of restraining force showed that, the main role in the loss oflocal soil stability on the slopes is played by water, which causes decrease of angle of reposeand cohesion between the particulates and dynamically effects the soil grains.

Structural cohesion Сп in graded materials takes place only in case of high density and soilcompactness and predominantly in case of low homogeneity on grain -size classification and ispredetermined, mainly, by interlocking grain arrangement [9].

Table 1. Dependence of cohesion and angle of internal friction of soilfrom its porosity [9]

Cohesion С (МPа) and angle of internal friction (grade) withthe porosity factor Type of refuse stone

0,45 0,55 0,65 0,75

0,02 0,01 - -Gravel and coarse sand

43 40 38 -

0,03 0,02 0,01 -Sands of average coarseness

40 38 35 -

Fine sand 0,06 0,04 0,02 -

H


CT 2

38 36 32 28

0,08 0,06 0.04 0,02Dust sand

36 34 30 26

Note: Upper line - cohesion, lower - angle of repose.

The water gets into the soil on the slopes as a result of percolation in case of storm eventand snow melting.

In winter the soil of grade level freezes (after the temperature fall below -5°С). Isothermalcurve of zero temperature falls lower and lower from the surface of grade level. Temperaturedistribution in depth gives evidence of the character of the soil straight -freezing: maximal underthe roadway paving and lesser on the slopes of fi lls (Fig. 4 /8/).

Isotherm of February

Isotherm of March

Isotherm of April

Figure 4. Isotherms (°С) of the coat of grade level during the winter -spring months

(Moscow and Moscow region)


CT 2

In spring there is started a constant soil temperature rise i n the upper part of the grade level.Heat current changes its direction, moreover before the start of melting. Soil frost retreat startsfrom two sides: from above, from the surface of grade level, and from below, from the side ofthawed ground (in the mess or the ground of the grade level). The speed of frost retreat fromabove is more or less identical on all the areas and averages (for Moscow region) to 4 cm/day.Frost retreat from within averages to 0,6 -0,7 cm/day. On the whole the thickness of layer,melted from within, amounts - in relation to the whole thickness of frost -bound layer - to 7 up to34%.

After the start of snow melting the water from the upper coating of the snow cover, subjectto the forces of gravitation, passes through the snow to the soil slope. Under the influence ofmelt-water there is started gradual coat frost retreat. The part of melt -water gets into the pores ofthe unfrozen soil, the remaining part flows through the slope – through the face of slope, underthe snow cover. As the snow melts and the soil thaws the major part melt -water gets into the soilpores and the smallest part of it flows down the slope surface. At last there comes a moment,when the depth of the melted soil -work at the slope surface reaches the value, wherein the allamount of melt-water which enters the soil goes to the soil pores. The flow down the surface ofthe slope stops. Melt-water through the soil pores under the gravity forces reaches the surface ofthe soil still not melted. In case of quite a large openness there appears the water flow in thesoil. Gradually takes place the formation of seepage, which flows in the soil above the border ofthe section «thawed ground - frozen ground» (Fig. 5.)

Figure 5. Formation of the seepage on the slopes of the grade level during the snow melting

As a result there happens a considerable soil overwetting, followed by the decrease offorces, which secure the soil grains from the shift. In the zone of water filtration on soil grainsoperates the hydrodynamic head hB, which appears as the result of penetration of elementary

Snow cover

Water seepageDry ground

Frozen groundThawed ground


CT 2

rate of water flow q, and the following formation of seepage with the rate iqQ . Elementary

rate of flow qi is formed by water, which penetrates into the soil during the sno w melting on theslopes, and, in case of storm event, rainwater. The water flows through the surface of aquifuge –the surface of still frozen soil -work (in spring) or the surface of a more solid soil -work whichlies lower (in summer and in autumn).

The melting surface is not plain. That is why in some places, because of the outflowobstacle, the local additional body of water may occur, which increases weighing water impactand therefore decreases restraining forces.

The water of rains which fall during th e snow melting period accelerates and increases theprocess of snow melting, therefore leveling up the water flow in the coat. The rainwater itselfalso penetrates into the soil pores (because of the infiltration) increasing more the filtration flowand soil dampness. Because of the accelerated snow melting and soil frost retreat in the zone ofthe shelf of grade level there is possible a situation, when the water from the overdamping zoneunder the roadway paving through the unfrozen coat under the wayside and the upper part of fillslope comes into the filtration flow, which flows in the surficial belt of the slope.

At some time the soil overdamping reaches the level when, the shearing force exceedrestraining forces. So there happens a shift - local deformation in the form of slope gutter.

As regards sand the possibility of shift deformation is worsened by its tendency toattenuation in aqueous state. Attenuation often happens [5] under the influence of filtration flowon the sand structure, in particular in case of dynamic character of filtration forces. Recentlysettled refuse stone of earthworks is very sensitive to the dynamic forces. Dynamic effectsusually cause small shift of sand-grains, which cause sand fluidization.

In case of sand fluidization on the slopes, instead of vertical displacement of sand -grains inthe process of sand settlement, there occurs considerable relative flat and vertical displacementof values as a consequence of running ground dispersion. In case of sufficient surface slope t heburdens rush in the form flows to the lower areas, forming the covers, filling the cavities andhollows.

Deformation ratio depends on the rate of dynamic effects. Earthquakes can cause passingof sand into dilute state on the large area. The effects of explosions and vibration are causedonly by local fractures of area structure, quite close to the whence of dynamic effects. Veryoften the fluidization event happens in comparatively small scales, for example, in the event ofpeople walking or vehicle passing over the surface of loose water - saturated sands [5].

Fluidization is native to all quite loose granular soils of any grain size. However due to alarger permeability to water the retention time of coarse -grained soils in dilute state is less, thanthat of compact-grained and that is why the fluidization practically never occurs there.

The danger of fluidization for the resistibility and structural competence is defined not bythe fact of fluidization, but by the character of its flow. The dwelling tim e of sand in dilutecondition and toughness of burdens influences the possible construction displacement.


CT 2

The rightfulness of theoretical considerations concerning the reasons for formation of localdeformations on the slopes of grade level of roads was confirmed by the results of full -scalemeasurements of water content and soil (sand) density on the fill slopes of roads «Don» (km103-104), built from fine sand of the borrow pit «Martemianovo» of Tula region (filtrationcoefficient 1-3 m/day, gradation factor -1,67). Embankment height- from 1,5 to 8 m. Theresearch was carried out in the years 2001 -2005. Water content and soil density on the slopeswere defined at depth 0, 20, 40 and 60 cm. during different times of the year. The depth wascounted from the lower surface of top soil. The research was carried out in field and laboratoryconditions with the use of certified appliances. The character of coat moisture gradient of theslope part during the spring months is shown on Fig. 6.

Рис 6. Coat moisture gradient of the slope part of grade level in spring 2004 year. km 104 а/r «Don». 1 -5

- point on slope contour: 1 – on the edge, 5 - embankment foot, 2-3 - passing points

Table 2. The character of placement of thawed and frozen layers on the slope of the fill in the first

half of the day March, 2004, km. 104 а/r «Don», depth of fill 8 m, slope ratio 1:1,75. Slope orientation -

south, air temperature at night -10°С, day + 5°С

№ layer Soil state in the layer Layer height, m

1 Frozen ground 0,03-0,05

2 Hydromorphic soil 0,05-0,10

3 Fluidity soil 0,05-0,10

4 Frozen soil 0,15-0,20

5 Non frozen soil -

Embankment edge Center Embankment toe

Dep

th o

f so

il sa

mpl

ing,

cm


CT 2

The fact of water flow in the soil (filtration flow) was photographed (Fig. 7)

Fig 7. Water filtration on the border of frozen and unfrozen soil (km 104 а/ r «Don», 14.03.2004)

Density measurements, carried out simultaneously with the humidity estimation showedthat the soil on the slopes is in quite friable state (table 3).

Table 3. Soil density and humidity of the

fill slope on 104 km road «Don» (average rates). Slope orientation - south. 14.03.2004

№ measurement point Depth of measurement point, cm Coat density g/cm3 Coat humidity %

1

0-20-40-60

1,831,801,761,87

14,513,817,315,7

2

0-20-40-60

1,851,851,831,87

10,012,515,115,7

3

0-20-40-60

1,831,801,761,87

9,112,516,017,5

4

0-20-40-60

1,831,801,761,87

10,912,215,615,0

5

0-20-40-60

1,761,831,831,85

13,014,016,217,5

Note: Optimum density for this sand is 1,89 g/cm3, optimum humidity - 10,9%.


CT 2

Actual values of compacting factors (0,93 -0,95-0,97) during the first 2,5 years of slopework (0,93-0,95-0,97) turned out to be lower than regulatory value (min 0,99), which certifiessoil high porosity on the fill slopes.

The research of dynamic effects of automobile transport on the soil of the slope fill partswas carried out at 104 km а/r «Don» (depth of sand fill - 6-8 m) и9и!4km МКАR (depth of sandfill 2 m). On the а/r «Don» convulsion in coat generated by passage of single -unit truck of themass 22 t with the speed of 50, 60 and 80 km/h, at Moscow Ring Motorway there was moving areal traffic flow with the intensity (in one direction) 6480 veh/h (km 9) and 7200 veh./h (km14). The carriageway of the road «Don» has 4 lanes (two lanes in each direction), at MoscowRing Motorway - 4 lanes in each direction. Shoulder s at 1,0 m from the upper edge ofembankment are hardened by plant formation. In both cases the the road pavement made ofasphaltic concrete, the roadbase - of low cement content concrete, base – of sand.

Vibrational impact of automobiles on the soil of gr ade level was studied in dry weather, inJuly under the temperature of + 23°С and in November under the temperature of +4°С. Therewere registered mean square and peak heights (X, Y, Z) of vibration acceleration. Axle X isdirected perpendicularly to the road axle. The measurement time amounted to 5 to 10 minutesand included the automobile drive to the measurement point and automobile removal.

There were made measurements (the values of vidroaccelerations), processing of the resultsreceived according to the finite element method enabled to define, that the vibration impact onthe slopes for the conditions discussed in case of problem solving on normal stress amounted,average, from 0,1 up to 0,044 kg/cm2, tangentially - from 0,04 to 0,001 kg/cm2. Value pea ksfall within the upper and lower slope part, which suggests the increased load in these zones. Themovement of soil parts amounts to 0,6 up to 0,2 mm and on the whole uniformly decreases inproportion to the standing off pumping source (from the cover of the road) (Fig. 8).

III. CONCLUSION

The results of the research carried out enables to make a conclusion that local soil

deformation on the embankment slopes are determined by the combination of the range of

factors: low soil density in the slope surficia l belt, high soil moistening in spring period, the

presence of filtration stream of melted (and rain – in case of rain fall) water in the slope part of

the roadbed. Vibrations generated in the soil of the roadbed by the cars passing by contribute to

the disturbance of equilibrium of restraining and shearing forces.

Only one from the abovementioned factors is subject to control by the roads constructors -

soil density of slope parts of embankment. However at the present time the embankment

construction method implies that the slope soil is not compacted The technology of soil

compacting of slope of embankments still is not worked out. The recommendations concerning

the following overcutting of the unconsolidated slope soil, which one may find in references,

cannot be considered as rational due to many reasons. As a consequence, the tools for the works


CT 2

execution on slope soil stabilization are not available (one cannot view a small road roller,

which rolls down the slope as a major compaction tool).

Рис 8. Curve of the soil particles flow (mm) as a result of operation of single load. Summary constituent,

disturbing frequency 10 Hz

References

[1]. Evdokimov V.I. The nature of Yamal and roads: complicated interaction. The Journal «Russian Roadsof XXI century», № , 2005.

[2]. Kupachkin B., Radkevich А . Introduction to the new soil mechanics. The Journal «Russian roads ofthe XXI century», № 1, 2006

[3]. Kondratiev V.G. Qinghai-Tibetan railroads: new experience of the construction of grade level ondeep-frozen soil. The Journal «Transport construction», № 4, 2005.

[4]. G.Grondin, A. Guimond , G. Dore Impact of permafrost thaw on airfield and road infrastructures inNunavik - Quebec. «ROADS» (PIARC), № 332, 2006.

[5]. Ivanov P.L. Dilution of sandy ground. М., SEI, 1962

[6]. Nemchinov М.V., Men’shov А.S . Influence of vibration of the road transport on the local slopestability of road bed. The Journal «Science and Engineering for Roads», № 4, 2005.

[7]. Men’shov А.S. Provision of local slope stability of hi gh fills of roads made of granular soil. Ph.D.Thesis. Мoscow, МАDI(SТU), 2006.

[8]. Zolotar’ I.А., Puzakov N.А., Sidenko V.М . Aquatic- heating parameters of the grade level androadway paving. М., Publ. House “Transport”, 1971

[9]. SNiP 2.02.02-83. Base of buildings and constructions. М., 1985

[10]. Beliaev D.S., Yushkov B.S., Kychkin V.I., Rukavishnikova N.Е . Development and approbation of thevaluation method of technical condition of the grade level of the coats of transport works. The Journ al«Russian Roads of the XXI century», №3, 2005

Sand

1:1.75


CT 2I. INTRODUCTION

Many countries have, for many recently years, been establishing policies aimed to utilise as

much as possible the use of local materials in building construction. Sand concrete is a family of

cement concretes which can be used to overcome limitations about environmental or economic

problems in the use of coarse aggregate. Especially, some certain regions it has the depletion of

coarse aggregate deposits.

Back in 1869, concrete without coarse aggregate was used for buildings, the 52m high

lighthouse of Port-Said (Egypt), which is still in used, was built with sand only [1]. In Russia in

vast areas sand is the only building material to be used in concrete. Here, a lot of buildings have

been constructed with sand concrete since fifties. The und erground station in St.Petersburg was

built with precast arches of sand concrete. In Germany first investigations to increase the sand

content (more than 60 %) in a concrete mix design were made in 1971[2]. In France, since 1998

the national “SABLOCCRETE” project done with the cooperation of Russian, Algeria, Maroc

is a big project sand concrete [1].

In the Mekong Delta in Vietnam, there is an abundance of sand but lack of coarse

aggregates to produce traditional concrete. Therefore, the use of sand concre te to substitute for

AN EXPERIMENTAL RESEARCH ON SAND CONCRETEIN MEKONG DELTA

NGUYEN THANH SANG , Doctoral studentPHAM DUY HUU, ProfessorInstitute of Science and Technology for Transport constructionUniversity of Transport and Communications

Abstract: This paper presents an experimental research on mechanical properties of sand

concrete for possible use in the Meko ng Delta. The research has been car ried out at the

University of Transport and Communications, Hanoi, Vietnam. By applying experimental

method, several typical scenarios with different proportions of water, cement, sand, fillers,

and additives have been examined to determine the following m echanical properties of the

sand concrete: compressive strength, flexural strength, splitting strength, elastic modulus. The

obtained results showed that the sand concrete can be usable for different construction

projects in the Mekong Delta.


CT 2

traditional concrete is considered since transport of coarse aggregates from other regions to the

Mekong Delta is very costly.

Sand concrete is fine concrete consisting of a mixture of sand, cement, filler, and water.Besides these basic components, sand concrete typically includes one or more admixtures.When fine gravel is incorporated with sand and their ratio G/S remains below below 0.7 (withG= Gravel, S=Sand) the mix then can also be socalled as sand concrete. The sand concrete i sdistinguished from a traditional concrete by using high proportion of sand; with a smallproportion or without using fine gravel and the incorporation of filler. It is also distinguishedfrom the mortar by its composition (mortar generally contains high cement content) andespecially by its destination, as sand concrete are primarily intended for more traditional uses.

In order to apply this material into building purpose in Vietnam, several mechanicalproperties have been carried out. The result of this study will be characterized hereafter.

II. EXPERIMENTATION

2.1. Characterization of the used materials

2.1.1. Sand

Fine sand (FS) extracted from Vinh Long province near of Ho Chi Minh city and featuringby a maximum particle diameter of proximately 2.3 6mm. The proportion of grains (smaller than0.075mm) is 2%; organic impurities are lighter than standard colour.

Coarse sand (CS), from Tri An lake (region around Ho chi Minh city), presentscontinuous particle size distribution ranging from 0.075 to 4.7 5mm. However the fractionsmaller than 0.30mm remains very small; the proportion of grains smaller than 0.075mm is1.3%; organic impurities are lighter than standard colour.

Fine sand and coarse sand were mixed with a ratio FS/CS equals to 1.7 by mass. Theparticle size distributions of the various sands used are shown in Fig 1. Table 1 lists the set ofphysical characteristics for three types of sand. The modulus of fineness of mixed sand equals to2.75, and packing density is 60.5%.

Table 1. Physical properties of the various used sands

SandBulk density

(kg/m3)Specific density

(kg/m3)Finenessmodulus

Packing density(%)

Sandequivalent

FS

CS

FCS

1399

1510

1534

2500

2560

2536

1.74

3.30

2.75

56

59

60.5

84

87

86


CT 2

9.54.75

2.36

1.18

0.15 0.60.30.0

750

10

20

30

40

50

60

70

80

90

100

Grain diameter (mm)

Pass

sing

%

FS CS 'Ideal' curve CFS

Figure 1. Granular curve of the different sands

2.1.2 Cement

The cement used is Nghi Son PCB40 cement (similar to CEM I); it chemical analysis andcomposition are given in Table 2. The physical characteristics are following: specific density3100 kg/m3 and Blaine specific surface area 3690 (cm 2/g).

Table 2. Chemical analysis of the cement used

SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O free CaO

21.29 5.72 3.30 63.18 1.1 1.9 0.12 0.30 0.193

2.1.3 Fillers and admixture

The fillers used have been obtained by sifting (passing of 80 m sieve), from Hoa An

quarry (region around Ho Chi Minh City), and are mainly composed of limestone (98 mass % of

CaCO3). Its characteristics are following: specific mass 2740 kg/m3 and Blaine specific surface

area 3210 cm2/g.

The admixture used is a super plasticizer (a typical Sika product), with a dry matter content

of 1% of cement mass.

Drinking water is suitable for use in this concrete.

2.2. Preparation of sand concrete samples and testing

15 proportions of sand concrete mixes wich contains a mix of these two sands were used.

They have three various factors includes: water/cement ratio, filler content, and age of sand

Pass

sing

%

Grain diam eter (mm)

FS CS ‘Ideal’ curve


CT 2

concrete. The factors effected on mechanical propert ies of sand concrete were investigated.

The specimens produced have been cured in air at 27±2 oC for 24 hours. Then, they were

removed from the moulds and immesed in water until the day of testing.

Compressive strength, splitting strength, Young’s modulus o f elasticity were determined on

cylindrical specimens with 150mm diameter and 300mm height. Flexural strength was

determined (using three-points method) for each mix on three 100x100x400 mm prismatic

sample.

2.3. Experimental plan

In this experimental study, compressive strength (fc), flexural strength (f r), splitting strength

(fsp), and elastic modulus (E c) vary from water/cement ratio (w/c), filler content (f), and age of

sand concrete (t). The experimental design theory [11] is used to establish an opt imum

experimental procedure [8, 11] and to elaborate empirical models considering both experimental

parameters (Input: w/c, f, t) and results (Out put: f c, fr, fsp, Ec). True values (w/c, f, and t) and

normalize ones (x1, x2, x3) in [-1, 0, 1] interval is given in Table 3, 4. This transformation

enables to analyze in the same manner, both qualitative and quantifiable data [11]. In the

following, all capital letters (C: Cement, W: Water, S: Sand, F: Filler, A: additives) refer to

component weight per cubic meter of mix. When using experimental design, priority semi -

empirical models are built from mathematical expansion of the outputs as following:

y = a0+a1x1+a2x2+a3x3+a12x1x2+a12x1x2+a23x2x3+a11x12+a22x2

2+a33x3

These coefficients of the model are identified through regression analysis and all possible

parameters are ordered such as to keep only the most influent. In this case, 15 mixes are

required to establish the mathematical formulation.

Table 3. Field of parameters

Truevalues

Field of parameters Middle values Interval Normalize values

w/c 0.38 to 0.52 0.45 0.07 x1

f 100 to 150 125 25 x2

t 0.845 to 1.447 1.15 0.30 x3

The experiments were designed based on the orthogonal array technique and ac cording toStandard NF P 18 - 500, proportion of sand concre te are given in table 4.


CT 2

Table 4. Parameters their values

No Normalize values True values

x1 x2 x3 w/c f t

1 1 1 1 0.52 150 28

2 1 1 -1 0.52 150 7

3 1 -1 1 0.52 100 28

4 -1 1 1 0.38 150 28

5 1 -1 -1 0.52 100 7

6 -1 1 -1 0.38 150 7

7 -1 -1 1 0.38 100 28

8 -1 -1 -1 0.38 100 7

9 1.732 0 0 0.57 125 14

10 0 1.732 0 0.45 168 14

11 0 0 1.732 0.45 125 56

12 -1.732 0 0 0.33 125 14

13 0 -1.732 0 0.45 82 14

14 0 0 -1.732 0.45 125 4

15 0 0 0 0.45 125 14

Table 5. Sand concrete mix proportion for e xperiment

Compositions ( per 1m3)

Cement Water Sand Filler Additive TotalNo Ratio

(w/c)

Fillercontent

(f)

Age(t)

C (kg/) W (l) S (kg) F (kg) A (l) (kg)

1 0.52 150 28 413 215 1,507 150 4.55 2,290

2 0.52 150 7 404 210 1,507 150 4.44 2,275

3 0.52 100 28 413 215 1,532 100 4.13 2,264

4 0.38 150 28 513 195 1,444 150 5.64 2,308

5 0.52 100 7 404 210 1,532 100 4.04 2,250

6 0.38 150 7 513 195 1,444 150 5.64 2,308


CT 2

7 0.38 100 28 513 195 1,510 100 5.13 2,323

8 0.38 100 7 513 195 1,510 100 5.13 2,323

9 0.57 125 14 350 200 1,515 125 3.50 2,193

10 0.45 168 14 467 210 1,469 168 5.13 2,319

11 0.45 125 56 456 205 1,503 125 4.56 2,293

12 0.33 125 14 563 185 1,476 125 6.47 2,356

13 0.45 82 14 433 195 1,528 82 4.33 2,242

14 0.45 125 4 444 200 1,503 125 4.44 2,277

15 0.45 125 14 456 205 1,503 125 4.56 2,293

Each sand concrete proportion made in the laboratory has chosen from the experimentalplans. Concrete-mixer has revolving-paddle and 180 dm3 in volume and the rate of rotation is 37revolutions per minute. Mixing time is from 4 to 6 minutes. Making and curing concrete testspecimens, experimenting according to ASTM are given in Table 6. The results of theexperiment are given in Table 7.

Table 6. Measurement according to

No Testing Standard

1 Making and curing ASTM C192/C 192M-02

2 Slump of fresh concrete ASTM C143/ C143M-00

3 Compressive strength ASTM C39/C39-01

4 Flexural strength ASTM C78-02

5 Splitting strength ASTM C496-96

6 Young’s static modulus ASTM C469-02

Table 7. The results of experiment

S fc fr fsp EcNo x1 x2 x3

(cm) (MPa) (MPa) (MPa) (MPa)fc/fr fc/fsp fr/fsp

1 0.52 150 28 8.5 35.17 4.29 3.39 28,750 8.20 10.38 1.27

2 0.52 150 7 7.9 29.90 3.97 2.20 23,321 7.53 13.58 1.80

3 0.52 100 28 8.6 37.91 4.33 3.58 28,376 8.76 10.58 1.21

4 0.38 150 28 4.5 47.83 5.01 4.62 35,303 9.55 10.35 1.08

5 0.52 100 7 7.3 24.38 3.61 2.32 20,861 6.75 10.52 1.56


CT 2

6 0.38 150 7 8.8 37.21 4.46 3.20 29,202 8.34 11.65 1.40

7 0.38 100 28 5.5 55.42 5.32 4.63 34,255 10.42 11.96 1.15

8 0.38 100 7 12 37.54 4.37 3.59 28,647 8.59 10.46 1.22

9 0.57 125 14 8.5 28.60 3.55 2.87 22,184 8.06 9.95 1.24

10 0.45 168 14 8 42.28 4.07 3.04 32,270 10.39 13.93 1.34

11 0.45 125 56 9.4 44.02 4.36 4.04 36,423 10.10 10.89 1.08

12 0.33 125 14 3.5 48.33 5.39 3.83 35,803 8.97 12.62 1.41

13 0.45 82 14 4.5 37.24 3.65 2.87 32,557 10.20 12.97 1.27

14 0.45 125 4 5.5 29.26 3.81 2.52 23,840 7.68 11.61 1.51

15 0.45 125 14 3.5 39.57 4.12 3.02 32,239 9.60 13.09 1.36

2.4. Results and discussion

As the results in table 7, determination of regression model of mechanical properties is

completed by the aid of a computer with Maple software. This way is a quite simple and very

quickly to gives precise values of coefficients. After the regression models were determined.

Equation are verified as following [8]:

+ To verify the suitable results of experiment in accordance with the Cochran law.

+ To estimate coefficients to fit with by statistical method (at a 95% confidence level) in

accordance with the Student law.

+ To verify the optimal process quantities (factors) through the confirmation of

experiments the Fisher law.

Solving and fitting regression models are following:

Fitted model of compressive strength:


+ With true values:fc = 58.69 – 86.53 (w/c) + 17.29 lg(t)

Fitted model of flexural strength:



CT 2

+ With true values:fr = 5.99 – 6.27 (w/c) + 0.83 lg(t)

Fitted model of splitting strength:


+ With true values:fsp = 3.81 – 6.33 (w/c) + 1.80 lg(t)

Relation between compressive strength and elastic modulus was established:

Ec= 1808.fc0.767

From the obtained results in Table 7 and regression models several discussio ns can be

withdrawn:

In valid selective factors, types of sand concrete obtained compressive strength from

25MPa to 55MPa; flexural strength from 3.55 MPa to 5.93 MPa; splitting strength from 2.20

MPa to 4.6 MPa, and elastic modulus from 20861 MPa to 36423 MPa respectively. This results

point out the rate of an increase in compressive strength as fast as the rate of an increase in

flexural strength and splitting strength. The investigation of types of fracture shows : 69% cone

type (a), 22% cone and split type (b), 4% shear type (d), and 4% columnar type [ASTM C39].

Almost specimens of sand concrete have the type of fracture similary to the type of fracture of

traditional concrete. This manner availably says that “theories of mechanical properties of sand

concrete similar to those of traditional concrete”.

Regression models point out that the sensibility of w/c and t to the strength of sand

concrete is more than the sensibility of f .

The ratios of fc/fr, fc/fsp of sand concrete and traditional concrete which have compressive

strength from 15MPa to 55MPa are from 6.7 to 10.5, and from 9.9 to 13.9 in sand concrete;

from 9 to 12, and from 10 to 15 in traditional concrete, respectively [10]. Therefore, the rate of

an increase in flexural strength and splitting st rength of sand concrete is more fast than the

those of traditional concrete.

III. CONCLUTION

* The obtained results from this research, would hopefully be initial contributions to the

use of sand concrete in the Mekong Delta region in Vietnam (especially, aiming at enhancing

the reuse of local materials in some regions of Vietnam).


CT 2

* An experimental work has been performed to study how water/cement ratio (w/c), filler

(f) content, age of sand concrete (t) can modify some mechanical properties of sand concr ete.

* In this study, fitted models of experimental theory were simple equation to using to

design the proportion of sand concrete.

* Theories of mechanical properties of traditional concrete can be used for sand concrete.

* Mixing time is a half as long as compared to traditional concrete, and vibrating time is

shorter. As results, enery comsumption can be reduced.

Reference

[1]. Presses de l’Ecole Nationale des Ponts et Chaussées, Paris, ISBN: 2 -85978-221-4 (1994). Béton desable, caractéristiques et pratiques d’utilisation, Synthése du Projet National de Recherche etDéveloppement SABLOCRETE. Vol. 237 (in French).

[2]. Pilly, F. ; Eschke, K. Sandreicher Beton, Beton und Stahlbeton. Heft 12/71. p.298-302.

[3]. AFNOR Standard NF P 18-500 (1995). Bétons de sables. 12 p.

[4]. Nguyen Thanh Sang (2005) . Research into design of component and strength of powdered -sandconcrete; Transport and Communications Science Journal; No 12; pp 106 -112.

[5]. Nguyen, S.T, Pham, H.D (2007) , Study of the effect of limestone powder on plasticity and strength of

sand concrete in Vietnam, The Transport Journal, No 7, pp 30-32.

(http://www.cauduong.net/forum_posts.asp?TID=1940 ; http://www.moc.gov.vn

http://english.vista.gov.vn/english/st_documents_abstract/ )

[6]. NCS. Nguyen Thanh Sang, GS.TS. Ph ạm Duy Huu (2008) . The rerults of an experimental stud y on

Mechanical Properties of Sand Concrete In the Mekong Delta, The Transport Journal, No 05, pp 33-35.

[7]. J.J Chauvin, G. Grimaldi ; (1998); Les bétons de sable; Bull. Liaison Lab. Ponts Chaussées 157 (9 –15 (in French).

[8]. To cam Tu, Tran Van Dien, Nguyen Đinh Hien, Pham Chi Thanh; (1999) ; Experimental Design and

Statictical Analysis; Science Technology Publishing House.

[9]. Sinan H n sl olu; Osman Ünsal Bayrak (2004). Optimization of early flexural strength of pavement concrete with

silica fume and fly ash by the Taguchi method. Civil Engineering and Environmental Systems, Volume 21, Issue 2 June,

pages 79 – 90. http://www.informaworld.com/smpp/content~content=a713947294~db=all .

[10]. M.S. Shetty (2003), Concrete Technology (Theory and practice). RAM Narga, New Delhi -110 055.

[11]. G.Taguchi (1987). System of experimental design. Unipub/Kraus international Publication, 2 24p

http://www.cauduong.net/forum_posts.asp

http://www.moc.gov.vn

http://english.vista.gov.vn/english/st_documents_abstract/


CT 2

During construction of highway roadbeds at zero elevations, and of small embankments,the ditches are being constructed in excavations and they are the roadside diversion ditches. Thepurpose of these structures is to drain the roadside area and the roadb ed soil.

However the role the ditches play in order to facilitate the roadbed construction operation isinsufficiently studied. This statement can be confirmed by the fact that mostly often designingof roadside ditches is quite a formal process - shape and sizes of the roadside ditch crosssections are specified in compliance with standard plans, trapezoidal shape with 0.3m bottomwidth and 0.4m depth, with slope inclination of 1:1.5. Hydraulic calculations for ditches areusually not done. Currently valid standards and recommendations / 1 / suggest that hydrauliccalculation be done for the ditch lined with concrete slabs (fig. 1), and as a rule there are noditch liners of such kind on flatland highways and on those ones in semi -rough terrains.

Fig 1. Structural model of a roadside ditch

1 - concrete slab; 2 - sand-gravel bedding for slabs; 3 - longitudinal seams filled with mastic

The fact which confirms the above statement is that roadside ditches are shutoff when rampsare being installed, especially in inhabited localities. In these cases a culvert (usually its diameterdoes not exceed 0.5m) is placed in the ditch and the area around is filled with soil or concrete.The ditch cross-sectional area decreases by several times. As the result ther e appears a block for waterflow in the ditch which causes water accumulation on the upstream side. These ramps are also the spotswhere different litter accumulates too and it blocks the water flow in the ditch even more.

Another confirmation of insuff icient study of role the ditch plays in facilitating the

DESIGN OF DIVERSION DITCHES FOR

HIGHWAY ROADBEDS

Prof. M.V. NEMCHINOVPh.D. student VU TUAN ANH(State Technical University – MADI,Moscow, Russia)

WATER LEVEL


CT 2

highway roadbed operation and consequently of not understanding their significance is themaintenance of roadside diversion ditches in the highways operation period. In warm season ofa year ditches become overgrown with tall grass and bushes and are not being cleared out ofthem (the grass is not mowed in the ditches) (fig. 2). The ditches are the most choked upsections of a road's right-of-way.

Technogenic litter, scrapped automobile parts (old ti res, rags, etc.) are being accumulated there.

And as a result the ditches turn from being lotic systems to being water detention basins. Water isremoved from them only by infiltration into soil and into roadbed too and it is also removed byevaporation. In temperate climate conditions of middle Russia still water in a ditch in summer season of ayear may be observed up to 1 - 2 weeks (in rainy years even longer), in spring during snow melt periodand in autumn during rain period it can be observed up to se veral months. In the monsoonal climaticregions (the Far East of Russia, Vietnam, the south of China, the Indochinese countries) in the rainyperiod of the year water stagnation in the ditch system (fig. 3) may be observed during some months.

Fig 2. The ditches overgrown with grass and bushes. Roads of Vietnam Republic, Russia, Poland


CT 2

Fig 3. Slack-water in a ditch

Stagnation of water for a long time in a roadside ditch results in an increased humidity ofroadbed soil. By infiltration water penetrates into roadbed soil and under it ( fig. 4) overwettingit and thus decreasing roadbed stability and subbase load -carrying capacity under it.

Fig 4. Relative humidity of roadbed soil under the impact of ditch slack -water

at State Highway №1A, Lang Son, Vietnam.

Note: numbers on a cross profile show relative humidity of the earth roadbed

In figure 4 it is seen that in case of a long period of slack -water in a ditch, the water

gradually infiltrates and penetrates into the roadbed body over wetting its soil. Under the road

pavement relative humidity of soil has reached 60-65% gradually increasing with the depth and

reaching 80% at the depth of 30cm (from the pavement bottom), and over 80% at the depth of

70cm, i.e. already in the soil, at the embankment base it is over 80%. Relative humidity optimal

for ensuring maximum soil density is 5 0-60% on the road №1А

Figure 4 illustrates the case of overwetting the roadbed soil which had been constructed in

zero elevation or in excavation. When stagnating for a long time in a ditch the water gradually

reaches the bottom of a road pavement. By d efinite combinations of water level and duration of

its stagnation in a ditch, soil water permeability, relative humidity of soil may reach 100%.

Little by little in roadbed soil under the pavement a free water level is being formed whichis dependent on the ditch water level. Under certain conditions there might appear a situation


CT 2

when under the road pavement a certain water head is being formed. It is in upright position andis equal to difference in the heights of ditch water level and road pavement b ottom elevation.

When a roadbed is in the form of an embankment of small height, the scheme of soilwetting shown in Figure 4 is accompanied by soil wetting which is performed by an upwardcapillary rise of water. In this situation sufficiency of the em bankment height depends on soilwater permeability and its degree of compaction during the construction process. However apossibility of overwetting and decreasing of the embankment load -carrying capacity ispreserved.

In the above considered operating conditions of roadbed and ditches (in case of waterstagnation) the decisive factor for estimating the degree of soil wetting in the embankmentbody and under it becomes the factor of time during which water remains in a ditch. This factoris determined by quickness (velocity) of water evaporation.

In operating conditions, i.e. in a situation when water flows along the ditch, the water too isbeing infiltrated into soil. Water in ditches is observed during rains and some time after. This iscaused by the water runoff coming from water catchment areas adjacent to the road a nd also theone coming from carriageway surface. The depth of water stream depends on hydrometricparameters of rain and water catchment area. In most cases the rains intensity is much le ss thanthat of a designed one and the water stream depth in a ditch is not big (in observations made bythe author the depth varied from 2 - 4 up to 10 - 12 cm). However under these conditions wateris being infiltrated into soil and overwettens it. Figur e 5 represents soil humidity data, obtainedduring rainy period on the highways of Vietnam Republic. It is seen in the diagrams that anactive water infiltration into soil of the embankment body and its base is going on.

It appeared that when ditches are in operation, i.e. when they ensure rainwater runoff alongthem, soil humidity is much less inspite infiltration process, than in the case when waterstagnates in the ditches.

Fig 5. Relative humidity of roadbed soil


CT 2

Water flow duration in a dit ch beside the hydrometric parameters of rain and water

catchment area depends on conditions of water flow. If there exists any obstacle the water flow

velocity decreases and the flow depth increases.

Thus when designing the roadside diversion ditches it is necessary to take into account

their "operating" condition, i.e. the presence of any obstacles for water flow in ditches - things

that litter ditches both of man-made and of natural origin (grass and bushes vegetation). This

can be done by improving the hydraulic calculation of a ditch.

Hydraulic calculations for the ditch are being done on the basis of the classical hydraulic

equation: w.vQ

where "Q" is capacity (flow rate), "w" is flow area,

"v" is the flow velocity.

In this formula flow velocity appears to be the key factor.

According to the Chezy equation iRCv

where "i" is the longitudinal flow inclination, "C" is the coefficient, "R" is the hydraulic

radius the value of which is determined by cross-sectional parameters of water flow in a ditch:

w/χR

where χ is the wetted perimeter.

The hydraulic radius R is the overall index of geometrical sizes and shape of the ditch cross

section (in the limits of cross section of th e streamflow going along this ditch).

An important indicator reflecting conditions of water streamflow is the coefficient C in the

Chezy formula. The value of this coefficient in the present time is determined by a series of

empirical formulas. There exist up to 136 of such formulas (for different conditions of water

streamflow) but the main ones are:

For turbulent conditions of water movement (which is typical of the situation with water

flow in a ditch) the coefficient C may be determined by formulas of different types. For open

channels with absolutely rough walls and for open natural beds the most used formulas are:

- The Basen formula: )R/1n87/(1C

where n1 is the Basen's hydraulic roughness coefficient. The Basen formula is used when

calculating water diversion ditches.

- The N.N. Pavlovsky formula: 2nyRC

where n2 is the N.N. Pavlovsky's coefficient of hydraulic roughness.


CT 2

R0,12n0,750,132n2,5y

This formula is to be the main one at calculating beds with absolutely ro ugh sides.

- The Manning formula: 3n61

RC

where n3 is the Manning's coefficient of hydraulic roughness.

- The I.I. Agroskin formula: lgR)(k2g4C

where "k" is the coefficient which characterizes absolute roughness both qualita tively and

quantitatively, i.e. according to its type and sizes.

Values of the hydraulic roughness coefficient differ considerably based on data obtained by

the authors mentioned above. For instance, Basen's n 1 depending on channel sides and bottom

roughness is n1 = 0.50…4.00; N.N. Pavlovsky's and Manning's n 2 = n3 = 0.012…0.150; I.I.Agroskin's k = 3.15…1.90.

Research data of Basen, N.N. Pavlovsky, Manning, I.I. Agroskin and of other authors / 3 /

show that:

1. Bed and sides roughness (i.e. the degree of their overgrowing with grass and bushes,

littering with stones and etc.) has a great impact on water flow velocity (up to 7 -8 times and

more).

2. When designing the ditches and doing hydraulic calculations it is necessary to carefully

estimate the future operating condition of ditches and in connection with that to choose value of

the design coefficient of bed sides roughness.

Obstructions found in ditches and blocking water flow in there may be divided to two

groups (according to character and degree of im pact on water flow in a ditch). One group is the

grass growing on the bottom and on the slopes of a ditch. It might be considered to be the

roughness of bottom surface and bed sides and in this case its impact is to be estimated

according to the hydraulic roughness value of bed. The other group are things of technogenic

origin (automobile parts, tires, and other kind of litter) and bushes growing on the bottom and

on the walls of ditches. In this case, depending on location, quantity, character of things th at

choke up the bed and their impact on water flow may be considered when doing calculations as

for the case when streamflow comes over a threshold or also when it comes through one or

several holes in the wall and etc.

The most frequent (practically countrywide) thing which chokes up a ditch is the grass (it

has different height in various periods of warm season of a year). Technogenic litter is sooner or

later being cleared out of ditches by the road maintenance services. Usually this is performed


CT 2

once a year before road spring acceptance (except bushes which are cleared out very seldom;

but bushes grow not so often).

Taking into consideration all the facts mentioned above a conclusion should be made that

the improvement of ditch design method must in clude:

- An obligatory determination of runoff volume on the designed section of a roadside

diversion ditch (for runoff coming from terrains of natural water catchments the well -known

methods are to be used / 3 /, for runoff coming from carriageway surface s of highways the M.V.

Nemchinov formulas are to be used / 4 /);

- Hydraulic calculations for a ditch considering its maintenance peculiarities: overgrowing

with grass and bushes vegetation (in a varying degree) or in conditions of periodic mowing

when grass is not taller than 2 - 5cm and doesn't influence considerably the flow velocity of

water; when doing calculations the Basen formula is to be used as the one which reflects to the

fullest extent the conditions of streamflow in the roadside diversion di tches (it follows from the

fact that the Basen formula was in the first place obtained for conditions nearest to the roadside

diversion ditch conditions).

- Determination of cross-sectional parameters of a ditch with consideration of character and

degree of roadbed soil wetting;

- Specifying the operating mode of a roadside ditch - whether it should be a detention

structure (perhaps on road sections located in the rural inhabited localities where there exist

numerous access ramps to the adjacent agricultur al lands or absolutely no ditch maintenance

works are conducted) or it should be a structure for transferring the collected runoff to a

discharge point.

Reference

[1]. G.A.Fedotov, P.I. Pospelov and others . Encyclopedia for a Highway Engineer. Volume V. HighwayDesign M: Informavtodor, 2007 (in Russian) .

[2]. Research Data of Highways Roadbed Hydrothermal Conditions in Vietnam Academy ofTransportation and Technology. Hanoi, 2003

[3]. Guide to Hydraulic Calcu lations of Small Artificial Structures. Under the General Editorship of G.Y.Volchenkov. M., Transport, 1974. (in Russia)

[4]. M.V. Nemchinov Adhesive Qualities of Road Pavements and Automobile's Security of Service. M.,

Transport, 1985.(in Russian)


I. INTRODUCTION

The road safety situation is a complex issue and there are high number of accidents factorsand indicators involved. The characteristics between factors and road safety (e.g. exposureversus risk) have been studied at microscopic and macr oscopic levels in several articles andmodels (i.e. OECD, 1997). Several model have attempted to represent the complexity of roadsafety problem. For instance the Haddon Matrix (1972) focused on three factors: driver, vehicle,environment (road design) at three different time phases of the crash: pre -crash, crash, and post-crash. Rumar (1999) described the road safety problem as a function of three dimensions:exposure, accident risk and consequences.

Many countries recognise the importance of internationa l benchmarking to measure andcompare their own achievements and progress in road safety with other countries. This willallow countries to learn/improve based on existing practices and lessons in other countries.

Benchmarking can identify the strengths an d weaknesses in road safety performance fromcountry to another. This can increase the awareness of the problem among public and policymakers. This will also help policy makers to take appropriate actions to solve their countryproblems.

THE EVOLUTION OF INTERNATIONAL ROAD S AFETY

BENCHMARKING MODELS: TOWARDS A ROAD

SAFETY DEVELOPMENT INDEX (RSDI)

GHAZWAN AL-HAJI, ASP KENNETHDepartment of Science and Technology (ITN),Linköping University, 601 74 Norrköping, Sweden

Abstract: Since the publication of Smeed’s model in 1949, the research on road safetybenchmarking has progressed nationally and internationally. There are mainly three types of

benchmarking: product of safety, practices, and strategic benchmarking. They differ depe nding on

the type of indicators that model is trying to compare. It is not relevant to review all literature,

however, our focus in this paper is on the background to some of the main benchmarking models

that have been used in the past and very recent. In our brief review, we describe the evolution

process of benchmarking into four generations of development. The latest generation realised the

necessity of combining up all the three types of benchmarking into one index. The Road Safety

Development Index (RSDI), as an example, provides a broad picture compared to the traditional

(earlier) models in road safety.

Keywords: Road safety, RSDI, international comparisons, ranking, composite index, macro -

performance indicators


A number of benchmarking models are already being developed and they range fromrelatively simple models to highly complex depending on the number of indicators involved,details of data and complexity of methods used in calculations and analysis.

The model in general depends on what is designed for and how is it designed? In roadsafety benchmarking between countries, three types of models, which generally used :

1. Product Benchmarking is used to compare accident rates.

2. Practices Benchmarking is used to compare activities related to human-vehicle-roadperformance (e.g. seat belts use, crash helmets use, motorways level, etc.)

3. Strategic Benchmarking is used to compare National Road Safety Programme (NRSP),management and organisational framework .

The major obstacle in constructing any benchmarking model is the lack of data fromdifferent countries, especially in developing countries. To have meaningful benchmarking, it isnecessary to have reliable, valid, and available data.

There are several reasons for reducing the n umber of input variables in most earlybenchmarking models. These include the simplification needs in the model, reducing the errors,and also reducing the cost and time of the data collection and analysis.

In our perspective, we can describe the evolution of road safety benchmarking models intofour stages of generations, which is simplified in the following description and illustration of“generations” in figure 1:

The first generation is characterized with models that compare countries’ road safetyperformance in terms of risk and exposure indicators such as accident rates and

motorisation (Product Benchmarking). These models are cross -sectional models, which

observed at the same year.

The second generation takes the time into account. Theses models bench mark the road

safety product over time series. These models are useful to monitor the trends in road

safety in countries and indicate the direction of progress ahead.

The third generation is realized the need for increased integration between product

(accident rates) and other indicators in the same model (Product and Practices Benchmarking).

The fourth generation focuses on all three types of benchmarking: Product, Practices

and Strategic Benchmarking. One approach is RSDI (Al -Haji, 2005), which integrates

much of macro-performance indicators in road safety into a single value.

Most of the early models are still in use and being applied in different studies. However,

today computers are developing rapidly, which simplifies the work and analysis of a large

amount of road safety data that was not available before. This development has made the work

in the third and fourth generations become easier and closer to reality.



ProductBenchmarking




ProductBenchmarking

RSDI

The first, second and third generation The fourth generation Time

Figure 1. The Road Safety Benchmarking Evolution towards RSDI

In particular, it was reasonable to start with simplified benchmarking models. Picking upideas (i.e. performance indicators) from the first three generations was very useful in reachingthe fourth generation (i.e. RSDI).

The aim of this paper is to make a selected review of the main benchmarking models inroad safety that has been used in the past and very recent.

II. THE FIRST GENERATION: LINKING MOTORISATION, TRAFFIC RISK ANDPERSONAL RISK

An early study in 1949, R.J.Smeed compared twenty countries, mos tly European for theyear 1938, where he developed a regression model (log -linear model) and he found an inverse(or negative) relationship between the traffic risk (fatality per motor vehicle) and the level ofmotorisation (number of vehicles per inhabita nt). This regression represented the best estimatesof the mean values of traffic risk for each given value of motorisation (what is called leastsquare). This shows that with annually increasing traffic volume, fatalities per vehicle decrease(see Figure 2). Smeed concluded that fatalities (F) in any country in a given year are related tothe number of registered vehicles (V) and population (P) of that country by the followingequation:

F/V = α (V/P)- β (1)

Where F = number of fatalities in road accidents in the country

V = number of vehicles in the country

P = population


Fatalities Rate(per vehicle orkilometre driven) Developing

Countries

Less DevelopedCountries

Highly DevelopedCountries

α = 0.003, β = 2/3

This formula became popular and has been used in many studies. It is often called asSmeed's formula or equation despite some authors preferring to call it a law.

Motorisation

Figure 2. Motorisation and fatalities rate internationally (based on Smeed’s formula)

This nonlinear relationship can be translated to a linear one by taking the logarithms o f thetwo sides:

Log Y = log α + β log X (2)

Where Y is F/V and X is V/P

The number of fatalities can be derived from Smeed’s formula as: F = c.V α.Pβ, where c, α,β are parameters and they are estimated from data by using the least square method . For theSmeed data (year 1938) the formula was:

F = 0.0003 P2/3 V1/3 (3)

Personal Risk (fatalities per population) is obtained by multiplying both sides of Smeed’sequation (1) by V/P as follows:

F/P = a (V/P)1-b

or F/P = 0.0003(V/P)1/ (4)

Since 1949, many studies have discussed Smeed’s equation (1) or they made a reference tothis equation. Some authors followed the equation of estimating the regress ion parameters (α, β)of the data by calculating the country road safety performance in comparison to other countries;see Jacobs and Hutchinson (1973), Jacobs (1982), Mekky (1985). They found that Smeed’sformula can give a close estimation of the actual data and it can be applied to different samplesizes of countries and years with the use of different values of α and β. Jacob and Fouracre


(1)

Learning andsociety force

Engineering andeconomy force

Individuals’learning force

Low Medium High Motorisation

(1977) applied this formula to the same sample of countries used by Smeed for the years 1968 -1971 and they found that the formula remains stable. Jacobs and Hutchinson (1973) examinedthe data for 32 developing and developed countries from the year 1968. Mekky (1985) foundthat the equation significantly captures the relationship between motorisation and traffic risk; h eused cross sectional data for the Rich Developing Countries (RCDs). Al Haji (2001) compared26 countries around the world with different levels of development. The results from this studysupport Smeed’s view of the relation between motorisation and fata lity rates. The correlationwas high, 96% of the variations are explained for the low motorised countries and 93% for thehighly motorised countries.

Some authors have tried to develop Smeed’s formula and its accuracy further by includingseveral socio-economic variables in the model. Fieldwick (1987) has included speed limits inthe same model. The number of registered vehicles has been replaced by the total vehiclekilometre driven in many late studies (e.g. Silvak, 1983). This measure (vehicle kilometredriven) was not available at the time of Smeed’s study.

Nevertheless, some other studies have tried to explain why the curve of development(fatality rates) declines downwards as been noted in many countries and shown in Smeed’sformula. The studies have analysed the factors and measures that influence the development ofthe curve of road safety. A review of these studies is reported by (Elvik & Vaa, 2004) and(Hakim, 1991). Besides, Minter (1987) and Oppe (1991b) showed that Smeed’s law is a resultof a national learning process over time. The development in society at the national level is theresult from the developments at the local level. In other words, the individuals (road users) canlearn by experience in traffic where they improve their driving ski lls and knowledge, while thewhole society can learn by better national policy and action plans. The Figure shown here illustratesthese factors on the development curve of road safety.

Figure 3. The influencing factors on the development curve of road safety


At the same time, many studies have criticised Smeed’s model because it only concentrateson the motorisation level of country and ignores the impact of other variables, see (Broughton, 1988),

(Andreassen, 1985), (Adam, 1987), where according to Smeed’s model, population and vehicles are the

only country values, that influence the number of fatalities. This means that road safety

measures have no meaning because road fatalities can simply be predicted from population and

vehicle numbers in any country and any year. Andreassen (1985) criticised the model’s accuracybecause there would always be a decline in traffic risk for any increase in the number of

vehicles, but generally in non-linear way. Andreassen proposed relating fatalities to (V) B4 where

B4 is a parameter highly related to each particular country, even to countries with a similar

degree of motorisation. Furthermore, Smeed’s study analysed data for one year, it was a cross -

sectional analysis with no time series analysis (Adam, 1987). Smeed’s formula expected thedowntrend in fatalities rate but not the number of absolute fatalities, which has occurred in

almost most western countries in the seventies (Broughton, 1988). In other words, the trend

failed to fit and predict the same as the real figures in HDCs. Broughton has concluded that:

“Smeed’s formula has no generally validity”

In later years Smeed (in Oppe, 1991a) has commented on some of these remarks that:“…We must be guided by the data and not by our preconceived ideas...The numb er of fatalities

in any country is the number that the country is prepared to tolerate…”

Also, Haight (in Andreassen, 1985) has referred to Smeed’s equation that: “… When the

formula disagrees with the observations we tend to assume that the particular area underinvestigation is safer or less safe than it ought to be… ”

Regardless of whether one agrees or disagrees with Smeed’s model, the fact remains thatthe model gave a simplified and fairly good representation between traffic risk and motorisationof different parts of the world during the earlier stages of road safety development.

At the same time, there are many other curves developed and presented in different studies

in a simple way and with a small number of indicators (motorisation, personal risk an d traffic

risk), which can describe the development of road safety in different countries. For instance,

Koornstra & Oppe (1992) have suggested the model shown in (figure 4) to describe the long -

term development of the number of fatalities over time in hig hly developed countries (HDCs).

There is an increasing S-shaped curve with regard to the development of motorisation (referring

to the number of vehicle kilometres per year). There is a decreasing curve for the development

of the fatality rates per year (t raffic risk). Together, by multiplication the values of motorisations

and fatality rates, they result in the increase and decline of the number of fatalities that have

been noticed in HDCs in recent decades.


Motorisation

Time

Number offatalaties

Time

Traffic Risk

Time

Curve C:total fatalities

Curve B:Fatalaties per

population

Curve A:Fatalaties per unit travel

Figure 4. Road safety development in HDCs (Koornstra & Oppe, 1992)

Haight (1983) illustrated the development of road safety in developing countries as shownbelow. The total number of fatalities increases, the fatalities per unit of travel decreases, and thefatality per population remains al most stable or with some decline over time.

Time

Figure 5. Road safety development over time for developing countries (Haight, 1983)

The long-run trends which are shown in (Figure 4) and (Figure 5) based mainly on repeated

cross-section surveys from different countries for different years. The objective is to show

whether the change (development) of data varies over time.

The Timo model (1998) shows curves of number of fatalities and total national mileage by

time in many eastern and western Europe countries according to the development levels of

mobility (Figure 6). At the beginning of the growth of motorisation, total fatalities are very high,

but decline continuously at a declining rate when mobility increases. When the mobility reaches

the saturation level, the decrease in the number of fatalities has slightly stopped or fluctuated.


total mileage

total fatalities

Level of mobilityI II III IV

Less DevelopedCountriesFatalities per

inhabitantsDevelopingCountries

Figure 6. Total fatalities based on the development of mobility (Timo, 1998)

The correlation between traffic risk (fatalities per number of ve hicle kilometres) andpersonal risk (fatalities per number of population) is shown in Figure 7. With a growing numberof vehicles per population, countries move from the right to the left across the curve (Fred,2001). An early level of motorisation, first leads to a growing number of traffic -related deaths,but not necessarily with the same high growth in the number of population -related deaths.However, later at a medium level of motorisation, traffic and personal risks increase and bothvalues are high. At the third higher stage of motorisation, when a country is completelymotorised, traffic and personal risks decrease. The change between the three stages is due tobetter engineering of vehicles and roads and greater understanding of the system by the ro adusers.

Figure 7. Traffic risk and personal risk in different countries where countries move

from the right to the left across the curve (Adapted from work by Fred, 2001)

As we discussed earlier, the personal risk is a function of traffic risk and motorisation.

Navin (1994) has converted this function into the following equation (see also Figure 8):


Fatalaties perpopulation

Vehicles perpopulation

0MM

efTT

(5)

Where M0 is the value of motorisation at maximum personal risk,

Tf is the point where the exponential curve meets the T-axis,

T is the traffic risk, fatalities per number of vehicles, and

M is the motorisation, vehicles per population

Figure 8. Three-dimension model of motorisation and fatality rates (Navin, 1994)

The models previously mentioned are in some way based on regression models or multipleregression models or quadratic regression models. They employ more than variables to checkthe goodness of fit to data from different countries and to find the appropriate relatedequation(s).

III. THE SECOND GENERATION: LINKING TRAFFIC RISK, MOTORISATION ANDPERSONAL RISK WITH TIME

In this generation, many benchmarking models have been developed to describe andpredict safety development between countries on the basis of time serie s models and theories.They relate the variables to a function of time to determine the long run change in safety levelover time either in a monthly form or annually. These models attempt to find the smoothedcurves to the time series data.

Koornstra (1992) has shown that motorisation is considered to be dependent on time, andthe relationship between deaths and population should include time. To measure the correlationbetween the output and input variables, one should take into account the trends in the model. Hefound the following formula for approximating the number of fatalities for a country in aparticular year:


y

1c)ktV

maxV(w

ktVxtzVtF

(6)

Where Ft is the number of fatalities for a country in a year t,

Vt is the number of vehicle kilometres travelled in the year t,

Vmax is the maximum number of vehicle kilometres,

k is the time lag in years, and

x, w, z, y, and c are constants

Oppe (1989) assumes that fatality rates follow a negative exponential learning function inrelation to the number of vehicle kilometres and time. This method has been found to be mosteffective when the components describing the time series behave slowly over time as follows:

ln (Ft/Vt) = ln (Rt) = αt + β

Or, equivalently:

Rt = eαt+ β (7)

Where the ln function is the natural logarithm,

Ft is the number of fatalities for some country in a year t,

Vt is the number of vehicle kilometres travelled in that year,

Rt is Ft/Vt and

α, β are constants

This means that the logarithm of the fatality rate decreases (sign of improvement) if α isnegative proportional with time. This model is called the negative exponential learning model,where α is supposed to be less than zero. Both α and β are the parameters to fit.

Oppe (1991a) assumes that the amount of vehicle kilometres per year is related to time andit is assumed that traffic volume will develop over time by a logistic function of a saturationmodel. This assumption indicates that th e growth rate of traffic volume is a percentage of theratio between the traffic already existing and the remaining percentage of Vm as follows:

βαt)tVmV

tVln(

Or, equivalently:

β)(ααe1

mVtV (8)

Where Vt is the number of vehicle kilometres travelled in that year, and

Vm is the maximum number of vehicle kilometres


This formula shows that countries with a large α should have a fast growth in traffic. Thetraffic volume will increase quickly first and at the end it will reach its saturation level, which

differs from country to country.

Oppe has applied the two formul as (7 and 8) to data from six highly motorised countries

over the time period 1950-1985. He found that both models describe the data fairly well. He

concluded that the development in road safety is a result of the development (learning) of the

traffic system in the country, which is more or less similar to Smeed’s conclusions. However,Oppe’s theory in estimating the remaining growth of traffic is questionable, particularly whenwe know that many European countries are currently discussing the possibility t o stop or reduce

the increase rate of motorisation. It is uncertain whether the number of fatalities can be

predicted simply from the fitted curves or from the number of vehicle -kilometres. The question

is therefore whether this decreasing equation (7) ass umes that the fatality rate reduces to zero in

the end or not, and in this case what is the predicted year for one particular country according to

its current level of mobility? Besides, what will happen to the expected number of fatalities if

the country’s trend becomes fully motorised to 100%.

Adams (1987) has stated a similar relation between fatalities (F) and vehicle kilometres

(V), which was presented: Log (F/V) = a + b*y where y = year – 1985. Broughton (1988) has

tested this logarithmic model on data from Britain between 1950 and 1985 and the results fitted

well. In the same study Broughton applied the same model to data from four western countries:

U.S.A (1943-85), West Germany (1965-85), Norway (1947-85) and New Zealand (1948-83).

He found that this model describes the data pretty well.

(Broughton 1991) and (Oppe, 2001a) they developed another technique, the ‘singular valuedecomposition method’, in comparing road safety trends between different countries. Thistechnique investigates the similarit ies and dissimilarities between different groups of countries

regarding fatality trend. They compared various time series of data of countries jointly to

investigate the correlation between these series. This technique is useful in classifying the

countries that are similar (accidents patterns) to each other.

The more detailed time series data have led to advanced and sophisticated ways of fitting a

curve to data, especially with the current use of computer packages. For example, auto -

regressive integrated moving average (ARIMA) techniques are used to fit and forecast the time

series that are changing fairly quickly. ARIMA models should be stationary; otherwise we need

to transform the data to make them stationary. The first part of the model is the auto -regressive

(AR). This means that the Y factor is a relation of past values of Y. The second part is the

moving average (MA). This means that the Y factor is a function of past values of the errors;

see Frits et al. (2001). For instance Scott (1986) has appli ed this method to model the accidents

in England (seasonal and annual data). Oppe (2001b) has applied this method to a model that

predicts the accident data from Poland (1980 -2010).


IV. THE THIRD GENERATION: THE NEED FOR INCREASED INTEGRATION WITHMANY VARIABLES INVOLVED

Most early development efforts for international road safety development have focused on

one or a few indicators by means of risk indicators (accident rates), which are few and isolated.

The third generation has realised the need for increa sed integration between product (accident

rates) and other indicators in the same benchmarking model.

Page (2001) has compared safety situations and trends in the OECD countries from 1980 to

1994. He developed a statistical model using pooling cross -sectional time series. The model

gives a rough estimate of the safety performance of a country regarding some variables such as:

population levels, vehicle fleet per capita, percentage of young people, and alcohol

consumption. Based on this model, countries that are showing the best levels are Sweden, the

Netherlands and Norway.

Bester (2001) has developed a model by means of stepwise regression analysis. The criteria

will indicate the variables that should be added or removed in the model. The study used

collected data from different international sources and the variables used are: national

infrastructure and socio-economic factors (e.g. GDP per capita).

(Elvik & Vaa, 2004) has used techniques for evaluating the effectiveness of various road

safety measures (output) in different countries by using what is called the “before and afterstudy” evaluation technique. Similar techniques might be used to show the effectiveness of roadsafety measures that countries have taken.

(Asp & Rumar, 2001) developed the ‘Road Sa fety Profile (RSP)’. It includes all possiblequantitative and qualitative variables that may have been important in describing, explaining

and comparing road safety situations in different countries. This technique illustrates the

development in a country over time in a quick and easy illustration. RSP uses both types of

quantitative and qualitative indicators. The quantitative data obtained from international

sources. The qualitative indicators are derived from a survey of questionnaires to experts in eac h

country. Respondents’ countries were asked to answer questions regarding key road safetyissues. The answers are used to measure the RSP level for each country.

The countries were divided into three different groups of motorisation (low, medium and

high). The RSP technique includes more than 20 direct and indirect road safety indicators. Each

indicator is normalised on a scale from +2 to –2. Then the results are illustrated as a profile (see

Figure 9). This made the comparisons between countries simpler a nd easier. The Road Safety

Profile was seen as a successful tool for identifying the problems in the country where actions

are needed.


Level of Motorisation: Low Country 1 Country 2

Personal RiskTraffic Risk

Road safety statistics

Road safety trend

Road safety R&D

Road safety organisation

Road safety program

Road safety legislation

Traffic police

Driver education

Alcohol in traffic

Speed

Seat belts

Road standard

Paved roads

Road expenditure per total

Etc.......

-2

Country 3

Direct safety measures

Indirect safety measures

0 2 -2 0 2 -2 0 -2

1

2

2

0

2

-2

2

-1

1

1-2

1

2

1

0

2

0

-2

-2

0

2

0

1

1

0

0-1

-2

0

1

N.A

-1

2

-2

0

2

1

-1

1

-2

0

-1

0

0

0

-1-2

-2

N.A

-2

2

0

Figure 9. Illustration of Road Safety Profiles (Asp & Rumar, 2001)

The Globesafe database (Asp, 2004) is pre sently being constructed by means of IT andInternet. It facilitates the illustration of Road Safety Profile across countries.

V. THE FOURTH GENERATION: LINKING PRODUCT, PRACTICES AND STRATEGICBENCHMARKING- RSDI AS AN EXAMPLE

The latest generation realised the necessity of having a systematic way to add up all the

potential indicators of human, vehicle, road, environment, and regulation combined with

weights into one index. This will give a broad picture of benchmarking and not focus on one or

few particular aspects.

RSDI, as an example, combines all three types of benchmarking together: Product,

Practices and Strategic Benchmarking. This will be useful to tell success from failure in a

country. RSDI is capable to compare the road safety level and progres s across a large number of

countries and regions worldwide

Each benchmarking type is a sum of indicators and dimensions. There will be as balanced

and important indicators within each dimension as available of data as possible.

The following figure shows the main components involved in RSDI where each component

comprises a number of indicators:


Personal risk"deaths perpopulation"

Socioeconomicperformance

Organisationalperformance

Enforcementperformance

ProductBenchmarking



RSDI

Safer road users"behaviour"

Safer roadsperformance

Safer vehiclesperformance

Figure 10. RSDI conceptual framework (overall road safety performance)

The major steps used in the process of constructing RSDI are the following (Al -Haji,2005):

Finding the key indicators and dimension,

Normalising (standardising) the indicators,

Weighting the indicators,

Combining the chosen indicators into (RSDI) by using different techniques,

Applying RSDI for a sample of countries and performing an a nalysis of the results, andfinally

Testing the uncertainty and comparing the methods used with the obtained results.

The composite index of RSDI takes the form:

n

1i iw

iXn

1i iwRSDI

Where: Xi: normalised indicators for country i

wi: the weights of the Xi

n number of dimensions

The weights ranged from 0 to 1 and the sum of weights is one.

The RSDI ranged from 0 till 100. The higher values indicate a higher level of safety in thecountry. The lower values indicate the worst performance in country in term s of road safetylevel and vice versa. The target value of RSDI is 100 and it shows how far the country has to bedeveloped to provide safer roads.


VI. CONCLUSIONSInternational benchmarking models in road safety are in the interest of most countries and

international bodies since they will show the scale of the problem. This paper has reviewed inbrief the development of the benchmarking models and how they have been used.

Road safety is a complex issue and there are a high number of performance indicator s thatcan be used for benchmarking. However few of these indicators were generally included in theearlier modelling process (generation one and two).

There are mainly three types of benchmarking: product of safety, practices, and strategicbenchmarking. They differ depending on the type of indicators, which the models are trying tocompare. The literature review has shown the importance of having a large number of factorsinvolved in the benchmarking model. The road safety level in a country is a result of the wholedevelopment in society (e.g. health, education, enforcement, engineering, etc.).

The RSDI seems to provide a broader picture compared to the traditional early models inroad safety.

Today computers are developing rapidly, which simplifies th e work and analysis of cross-sectional data and time series data, which was not available before (e.g. to Smeed in 1949). Thisdevelopment has made the work in the third and fourth generations become easier and closer to reality.

AcknowledgementThe authors would like to express their appreciation to the projects work and teams of

Traffic Safety and Environment, TechTrnas project (Developing E -learning Applications andCourses in Road Safety to Russian Universities) at the Department of Science and Technol ogy,Linköping University in Sweden and State Technical University -MADI, Moscow, Russia.

References[1]. Adams, J., (1987). Smeed's law: some further thoughts. Traffic Engineering and Control 10 (7), pp. 70 -73.

[2]. Al-Haji, G., (2005). Towards a Road Sa fety Development Index (RSDI) - Development of anInternational Index to Measure Road Safety Performance. Licentiate Thesis No. 1174, ISBN 91 -85299-70-7, Linköping University, Sweden, pp. 15 -28.

[3]. Al-Haji, G., (2001). Traffic Safety in Developing Countri es- New Approaches in TechnologyTransfer by Using Distance Education Technique. Master’s thesis, LITH -ITN-EX-2001:156-SE,Linköping University, Sweden, pp. 7 -23.

[4]. Andreassen, D.C., (1985). Linking deaths with vehicles and population. Traffic Engineeri ng andControl, Vol. 26, pp. 547–549.

[5]. Asp, K., (2004). Globsafe database- A database for road safety analysis. Some parts of this database isavailable on-line: http://www.globesafe.org/ (last visited 2005 -05-16).

[6]. Asp, K., Rumar K., (2001). The Road Safety Profile. The RetsNet News “Regional Traffic Safety Network. No. 1”.[7]. Bester, C.J., (2001). Explaining national road fatalities. Accident Analysis and Prevention. Vol. 33, pp. 663 –672.

[8]. Broughton, J., (1991). Forecasting road accident cas ualties in Great Britain. Accident Analysis andPrevention 23 (5), pp. 353–362.

[9]. Broughton, J., (1988). Predictive Models of Road Accident Fatalities. Traffic Engineering andControl, May 1988, ISSN:0041-0683, pp. 296-300.

[10]. Elvik, R., Vaa, T., (2004). The Handbook of Road Safety Measures. Elsevier Amsterdam, ISBN: 0 -08-044091-6, pp.66 & pp. 676-803.

[11]. Evans, L., (1991). Traffic Safety and the Driver, New York: Van Nostrand Reinhold, P25 -43, pp. 60-95.

http://www.globesafe.org/


[12]. Fieldwick, R. and Brown R.J . (1987). The effect of speed limits on road casualties. TrafficEngineering and Control, Vol. 28, pp 635 -640.

[13]. Haddon, W. (1972). A logical framework for categorizing highway safety phenomena and activity.The Journal of Trauma, 12, pp. 193 -207

[14]. Haight, F., (1983). Traffic Safety in Developing Countries, Journal of Safety Research, Vol. 14, No.1, pp. 1-12.

[15]. Hakim, S., Shefer, D., Hakkert, A.S., Hocherman, I . (1991). A Critical Review of Macro Models forRoad Accidents. Accident Analysis & Prevention, Vol. 23, No. 5, pp. 379-400.

[16]. Jacobs, G.D., (1982). The potential for road accident reduction in developing countries. TransportReviews 2, pp. 213–224.

[17]. Jacobs, G.D., Fouracre, P.R ., (1977). Further research on road accident rate in developingcountries. TRRL report LR 270. Transport and Road Research Laboratory, Crowthorne, Berkshire.

[18]. Jacobs, G.D., Hutchinson, P., (1973). A study of accident rates in developing countries. TRRLreport LR 546. Transport and Road Research Laboratory, Crowtho rne, Berkshire.

[19]. Koornstra, M.J., (1992). The evolution of road safety and mobility. IATSS (InternationalAssociation of Traffic and Safety Sciences), Research, Vol.16, No.2, pp. 129 -148.

[20]. Koornstra, M.J., Oppe, S . (1992). Predictions of road saf ety in industrialised countries and EasternEurope; an analysis based on models for time series of fatality rates and motorised vehicle kilometres oramounts of passenger cars. Proceedings of International Conference ‘Road Safety in Europe’, VTI,Linköping, Sweden, pp.49-70.

[21]. Mekky, A., (1985). Effect of road increase in road motorisation levels on road fatality rates in somerich developing countries. Accident Analysis and Prevention 17 2, pp. 101 –109.

[22]. Minter, A. L., (1987). Road casualties- Improvements by Learning Processes. Journal of the TrafficEngineering and Control, Feb. 1987, pp. 74 -79.

[23]. Navin, F., Bergan, A., Qi , J., (1994) A Fundamental Relationship for Roadway Safety: A Model for GlobalComparisons. Transportation Research Board, Transportation Research Record 1441, Washington D.C., pp. 53 -60.

[24]. OECD ‘Organisation For Economic Cooperation and Development’, (1997). Road SafetyPrinciples and Models: Review Of Descriptive, Predictive, Risk and Accident Consequence Models.OCDE/GD(97)153, Paris.

[25]. Oppe, S., (2001b). Traffic Safety Development in Poland. SWOV, Leidschendam, Netherlands.

[26]. Oppe S., (2001a). International comparisons of road safety using Singular Value Decomposition.Leidschendam, SWOV Institute for Road Saf ety Research, The Netherlands, pp. 8 -18.

[27]. Oppe, S., (1991a). The Development of Traffic and Traffic Safety in Six Developed Countries.Accident Analysis and Prevention, Vol.23, No.5, pp. 401 -412.

[28]. Oppe S., (1991b). Development of traffic and traf fic safety: global trends and incidental fluctuations.Accident Analysis and Prevention; Vol.23: pp. 413 –22.

[29]. Oppe S., (1989). Macroscopic models for traffic and traffic safety. Accident Analysis andPrevention, Vol.21, pp. 225-232.

[30]. Page,Y., (2001). A statistical model to compare road mortality in OECD countries. AccidentAnalysis & Prevention, 33, pp. 371–385.

[31]. Rumar K. (1999). Transport safety visions, targets and strategies: beyond 2000. First EuropeanTransport Safety lecture, Brussels, European Transport Safety Council.

[32]. Smeed, R.J., (1949). Some statistical aspects of road safety research. Journal of Royal StatisticalSociety Series A 112, pp. 1–34.

[33]. Timo, E., (1998). Risks, exposures and accident data, VTI Conference 9A, part 6, nr 9A:6, s 85-96,

Linköping, Sweden


I. INTRODUCTION

Much attention has been focused onlateral performance of railway vehicles for along time in order to achieve higher speed andbetter comfort. But the longitudinal dynamicperformance was neglected except for thestudy on traction and braking performance.Since it has nothing to do with lateralperformance, forward velocity was normallyconsidered uniform. The results obtained inthis paper show that it may be an importantcause of wheel tread spalling and rail

corrugation with certain wavelength.

There are many puzzled enigma exist ingin rail vehicles, such as the rail vehicle mayhave some tremble in the course of speedup orat a not high speed; the ride index of railvehicle in some low speed might be worsethan that in the higher speed; the wheel treadspalls and the track wave wears with somewavelength; the vibration of rail vehicle wasquite bigger on some tracks. Thesephenomena indicate that the performance ofrail vehicle has some relation with the trackirregularity and the longitudinal dynamic

INFLUENCE OF TRACK IRREGULARITY ON ONGITUDINALVIBRATION OF WHEELSET AND CORRELATION

PERFORMANCE

WEIHUA MA, SHIHUI LUOTraction Power State Key Laboratory,Southwest Jiaotong University, Chengdu 610031, ChinaRONG-RONG SONGCollege of computer science andtechnology, Southwest Universityfor nationalities,ChengduSichuan 610041. China

Abstract: A longitudinal vibration of wheelset with respect to bogie frame often exist s with a

high acceleration magnitude and relative high frequency. At first a simplified model with a single

wheelset moving at a constant speed on a tangential track with irregularity is used to investigate

the longitudinal vibration dynamics. Results of the longitudinal vibration study indicate that the

longitudinal vibration frequency of the wheelset is most sensitive to the primary longitudinal

stiffness and the mass of the wheelset. As to the locomotive model, the longitudinal vibration was

concerned with cross-level irregularity and vertical profile irregularity. A method to estimate the

resonance speed is presented. Finally, the paper shows a possible solution to extend wheel -rail

service life by eliminating longitudinal vibration of the wheelset. The solution is simply arranging

the primary vertical damper with a forward angle, so that its damping component can be applied to

longitudinal direction.

Keywords: Track irregularity; Acceleration; Longitudinal vibration; Dynamics


behavior of the wheelset. In this paper, weinvestigate the relation between longitudinalvibration and the track irregularity from thelongitudinal dynamic point of view.

Lots of researches have been carried outin the field of wheel-rail contact fatiguemechanism in last ten years [1, 2, 3, 4].Wheel-rail wear is a necessary expense ofrailway transportation because of adhesion incontact patch between track and vehiclesystem. Total wear between wheel and rail isdue to stress generated by static, dynamic andthermal loadings [6]. The best way to reducewheel-rail wear is to reduce the unnecessarystress, which is caused by dynamic interactionbetween track and vehicle due to trackirregularity and vehicle vibration. But it isdifficult to determine which one is the mainfactor. According to the following analysis,wheelset longitudinal dynamic behavior maybe a possible source of the severe wheel-raildynamic interaction.

Since track irregularity is one of the mainreasons of rail vehicles vibration [13], a largenumber of researches have been made on theresponse of rail vehicles to the trackirregularity [10, 11]. They all aimed at thevertical or lateral vibration of rail vehicles, notreferred to longitudinal vibration of thewheelset. In this paper, we analyse theinfluence of tack irregularity on longitudinalvibrations.

II. THE TRACK IRREGULARITY

There always exists track irregularity inthe realistic track from the track setup and theinteraction wear between the railway vehiclesand the track. Researches show that the trackirregularity is a random process [11]. And it’sdifficult to decide what kind of the track

spectrum should be adopted in the vehicledynamic evaluation in our country.

Usually the spectrum density used todescribe track irregularity [5] can beexpressed as:

6ω3a4ω2a

2ω1a0a

6ω3b4ω2b

2ω1b0bS(ω(

(1)

In which, )3,2,1,0i(b,a ii is the

constant coefficient of the angular frequencyof space .

The American Federal Rail wayManagement Bureau (FRA) gets the trackspectrum according to a great deal ofmeasured data. It is divided into 6 Grades,among which Grade 1 is the worst and Grade6 is the best. Germany has formulated thespectrum of the high-speed trunk with lowinterference and high-speed trunk with highinterference. Our country has not formulatedthe standard track spectrum yet, but thedepartment concerned has carried out a fewresearch works, and offered some expressionsfor track spectrum according to the measureddata. But the acquired data in our country’sresearch is so few that it is unable to representthe statistical characteristic of trackirregularity in our country [11], so we analysethe problem by using the Germany trackirregularity.

According to the accumulativeexperience, aiming to the speedup track in ourcountry, we consider the vertical profileirregularity of the medium track iscorresponding to the vertical profileirregularity of Germany high-speed trunk withlow interference; the alignment irregularity ofthe medium track is worse than the alignmentirregularity of Germany high-speed trunk with


high interference; the cross-level irregularityof the medium track is correspond ing to thecross-level irregularity of Germany high-speed trunk with low interference. To make asimple and proper calculation , we use themedium track irregularity in this paper.

Time-domain excitation is used both inthe locomotive model and simpl ified model. Itis calculated by a polynomial express ion asfollowing:

2ω2aω1a0a

ω1b0bF(ω(

(2)

The coefficients of excitation used in themodel and the parallels with Germany high-speed trunk with high interference are listed intable 1

Table 1. Coefficients in polynomial expressiona0 a1 a2 b0 b1

Germanyalignment

0.016987 0.8452 1 0.001144 0

Germanyvertical

0.016987 0.8452 1 0.001519 0

Germanycross-level

0.00744 0.38718 1.283 0 0.00203

Modelused

alignment0.034 1.6904 1 0.0057 0

Modelused

vertical0.016987 0.8452 1 0.00093 0

Modelused

cross-level

0.00744 0.0774 1.283 0 0.0015

III. CALCULATION MODEL

3.1. Simplified model

The longitudinal vibration of wheelsetwas found in the locomotive model. It isnecessary but much time-consuming to makemore analysis for all kinds of railway vehiclesto study the phenomenon. In order to provethe result obtained by the locomotive modeland find the mechanism of longitudinalvibration, a simplified model with just onewheelset and minimum freedoms is set up

(see Fig.1).

Giv

en m

ass

v

Forwardvelocity

Cpx

Kpx,kpy

Wheelset

Rail

Irregularity

y x

z

Kpx-primary longitudinal stiffness; Kpy-the primary lateral stiffness; Cpx-primarylongitudinal damper;

Fig.1. The Simplified model with one wheelset

In the simplified model, mass ofwheelset: 2500 kg;

Roll mass moment of inertia of wheelset:500 kg/m2;

Pitch mass moment of inertia ofwheelset: 100 kg/m2;

Yaw mass moment of inertia of wheelset:500 kg/m2;

Mass of given mass: 19500 kg;

Degree of freedom of wheelset: 6;

Degree of freedom of given mass: 1;

Kpx: 1.2×107 N/m;

Kpy: 6.0×106 N/m;

Cpx: 1000 N.s/m.

Rail: 60kg/m rail;

Track gauge: 1435mm;

Rail cant: 1/40;

The lateral clearance between wheel andrail: 14 mm;

Tread profile: JM3 profile;

Wheelset

Irregularity

Rail


In the model, the friction coefficient μbetween wheel and rail is supposed to beconstant, and Kalker’s simplification theory(FASTSIM) is adopted to calculate the wheel -rail creepage-creep force. The equivalentconicity of the wheel tread within the scope of6 mm is about 0.12, and is about 0.2 withinthe scope of 6-8 mm, and will be greater than0.25 beyond the scope of 8 mm.

3.2. Locomotive Model

The analysed locomotive is designed tomeet the operation requirements of 200km/h.In order to reduce the axle load to 22 t, atrailer wheelset is considered besides twodriven wheelsets in a bogie (Fig.2). Eachdriven wheelset has a set of driving unit thatconsists of one traction motor with a gearboxconnected it rigidly. The driving unit can beconsidered mounted on bogie frame rigidly ormounted on it elastically in lateral direction inmodel.

CSYCSX

Trac

tion

Unit

KSZ

Trac

tion

Unit

TR

BOGIE

DW TW DW

TR

KSZ

CSY

CSX

DW—driven wheelset; TW—trailer wheelset;CSX —secondary suspension longitudinal damper;CSY—secondary suspension lateral damper;KSZ—secondary suspension vertical stiffness;TR—traction rod.

Fig.2. Model of A-1-A bogie

The model consists of car body, 2 bogies,

4 traction motors and 6 wheelsets. The

longitudinal damper and the lateral damper

are arranged in secondary suspension between

each bogie and the car body. Four coil springs

are arranged on each side of bogie as

secondary spring suspension to keep good

lateral running performance. The longitudinal

force will be transferred from bogie frame to

car body through a four-linking-rod

mechanism. It causes a rigid connecti on

between car body and front and rear bogies in

longitudinal direction. The primary

suspension is considered as a compact force

element at each axle box with st iffness in

three directions and vertical damping. T o

improve the speed of the calculation, t he axle

box itself will not be considered as a body

element, and it doesn’t influence the precision

of the result.

VI. ANALYSIS OF SIMPLIFIED MODEL

4.1. Longitudinal dynamic performance ofwheelset

In order to study longitudinal dynamic

performance in principal, an analysis is

conducted for the simplified model shown in

Fig.1. Longitudinal resonance vibration is

revealed on the track with irregularity at the

speed of 20km/h, and the longitudinal

vibration acceleration is quite large at the

frequency of 10.6Hz (see Fig.3). The

longitudinal creepage and adhesion coefficient

are shown in Fig.4. In Figs.3 and 4, we can

see that the longitudinal vibration resonance is

obvious and the adhesion coefficient always

approaches the presumed saturation value

0.25. So we can say that longitudinal

resonance vibration of the wheelset really

exits in the simplified model.

DWTWDW

BOGIE

Tra

ctio

n U

nit

Tra

ctio

n U

nit


0 5 10 15 20-150

-100

-50

0

50

100

150

Long

itudi

nal A

cc [

m/s2 ]

Time [ s ]

(a)In the time domain

0 5 10 15 20 2505

101520253035

Long

itudi

nal A

cc [

m/s2 ]

Frequency [ Hz ]


Fig 3. Longitudinal acceleration in the simplified

model at

0 5 10 15 20-0.75

-0.50

-0.25

0.00

0.25

0.50

0.75

Long

itudi

nal C

reep

age

Time [ s ]


0 5 10 15 200.00

0.05

0.10

0.15

0.20

0.25

0.30

Adhe

sion

Coef

ficien

t

Time [ s ]

(b) adhesion coefficient

Fig 4. Longitudinal creepage and adhesioncoefficient in the simplified model at 20km/h


10.62

yaw displacement

lateral displacement

longitudinalvibration

- 1. 25 - 1. 00 - 0. 75 - 0. 50 - 0. 25 0. 00 0. 25

40. 0

35. 0

30. 0

25. 0

20. 0

15. 0

10. 0

5. 0

0. 0

Root loci

Natural damping

Freq

uenc

y [H

z]

Fig 5. Root loci analysis of the simplified model

To find the inherent frequency of thesimplified model, the linear analysis of thesimplified model was carried out, and theresults are shown in Fig.5. The speed isranged from 1 km/h to 401 km/h and theincrement is 5 km/h. We can clearly see thatthe longitudinal vibration frequency nearlyunchanged with the speed, but the lateraldisplacement frequency and yaw displacementfrequency of the wheelset changed with thespeed. From the speed of 1km/h to 401km/h,the frequency of the wheelset longitudinalvibration is nearly 10.62Hz.

In the simplified model, the inherentfrequency is related to the system, and notvaried with the speed. But the frequency willchange according to the mass of the wheelsetand the primary longitudinal stiffness. Theinherent frequency could be estimated by thefollowing expression:

m

2

2

1f L

xk

(3)

Where m is the mass of wheelset and kxis the longitudinal stiffness of each axle box.Let kx = 1.2×107N/m, m = 5000kg, then


Hz03.11

5000

102.12

2

1f

7

L

If this vibration cannot be dampedeffectively, the rail corrugation with the

wavelength is developed on rail, and the

wavelength is:

m5.003.11

6.3/20

f

v

L

.

As the longitudinal resonance vibrationonly occurs at a given speed, so the

wavelength doesn’t increase with theincrease of the speed. The possible cause ofrail corrugation has been discussed in manypapers [8]. Here a new possible solution isproposed on the formation of rail corrugationwith certain wavelength from the longitudinaldynamic point of view.

The longitudinal vibration resonance ofwheelset doesn’t occur at other speeds. Forexample, at the speed of 10km/h or 30km/h,the longitudinal vibration resonance ofwheelset doesn’t happen (see Fig.6). So wecan draw the conclusion that the longitudinalvibration resonance is corresponding to thespeed; higher or lower than 20km/h thelongitudinal vibration resonance does nothappen in the simplified model, although thelongitudinal vibration is still very high.

0 5 10 15 20-15

-10

-5

0

5

10

15

Long

itudi

nal A

cc [

m/s

2 ]

T im e [ s ]

10 km /h

(a)In the time domain at 10 km/h

0 5 10 15 20 250.0

0.4

0.8

1.2

1.6

2.0 10 km/h

Long

itudin

al Ac

c [m

/s2 ]

Frequency [ Hz ]

(b)In the frequency domain at 10 km/h

0 5 10 15 20-40

-20

0

20

40

Long

itudin

al Ac

c [m

/s2 ]

Time [ s ]

30 km/h

(c)In the time domain at 30km/h

0 5 10 15 20 250.0

0.5

1.0

1.5

2.0

2.5

3.0 30 km/h

Long

itudin

al Ac

c [m

/s2 ]

Frequency [ Hz ]

(d)In the frequency domain at 30km/h

Fig 6. Longitudinal vibrations of wheelset under

different speeds

4.3. The speed of longitudinal vibrationresonance

Several frequencies can be recognized in

Figs. 3 and 6. A fixed frequency of about

10.6Hz is corresponding to longitudinalvibration the wheelset with respect to bogie

frame. A velocity depending on frequency is

the rolling angular velocity of the wheelset .


With the forward speed increasing from

10km/h to 20km/h and 30km/h, this valuevaried from 5.56rad/s to 11.1rad/s and

16.66rad/s respectively. Suppose nominal

radius of the wheelset is 500mm, then the

value of 2.0rad/s equals 1cycle/s of wheelrolling. So above three angular velocities are

corresponding to 2.78Hz, 5.55 Hz and 8.33

Hz of wheel rolling frequency. Dividing themby the frequency of 10.6Hz, we can get the

results of 3.81, 1.91 and 1.29. Thus a ratio is summarized as:

w

L

f

f , (4)

where fL is longitudinal vibration

frequency of wheelset according to bogieframe; fw is rolling frequency of wheelset

depending on forward speed and nominal

radius, which is expressed as:

s/circleR2

6.3/vf

ww (5)

As the vibration resonance only occurs

when the two vibration frequencies are closeor when one is an integer multiple of the

other, so a resonance vibration of wheelset in

longitudinal direction tends to be i nducedwhen the vibration frequency is an integer

multiple of wheel rolling speed, which means is an integer.

From Eqs.3-5, we can get :

NL

www

fR2.7R2f6.3v (6)

According to Eq.6, the resonance forward

speed for the simplified model is:

km/h12.192

62.105.02.7

fR2.7v L

w N

Suppose that the nominal radius of wheel

is 500 mm. The resonance vibration occurs in

20km/h considering influence of other factors

such as adhesion.

To further learn the influence of mass of

the wheelset and the primary longitudinal

stiffness on the longitudinal resonance

vibration of the wheelset, we carried out the

calculation of the speed at which resonance

vibration occurs with different mass and

primary longitudinal stiffness in the simplified

model. The results with changing mass and

constant primary stiffness are shown in table 2

and the results with changing primary

longitudinal stiffness and constant mass are

shown in table 3. In table 2-3, f1 means the

inherent frequency of the wheelset attained

through the root loci analysis; f2 means the

inherent frequency calculated through Eq.3;

v1 means the speed of longitudinal resonance

vibration calculated by incorporating f1 into

Eq.6; v2 means resonance speed calculated by

incorporating f2 into Eq.6; V means the

resonance speed gained by the simulation

calculation of the simplified model.

Table 2. The resonance speed gained by differentmethods with different wheelset mass

Mass

(kg)f1(Hz) f2(Hz)

v1

(km/h)

v2

(km/h)

V

(km/h)

1000 20.86 24.66 37.54 44.38 45

2000 15.93 17.43 28.67 31.38 32

3000 13.38 14.24 24.08 25.62 26

4000 11.76 12.33 21.17 22.19 23

5000 10.62 11.03 19.11 19.85 20

6000 9.75 10.07 17.55 18.12 19

7000 9.07 9.32 16.32 16.77 17

8000 8.52 8.72 15.32 15.69 16


Table 3. The resonance speed gained by differentmethods with different primary

longitudinal stiffness

Kpx

(N/m)f1(Hz) f2(Hz)

v1

(km/h)

v2

(km/h)

V

(km/h)

1.0e7 9.69 10.07 17.44 18.12 18

1.1e7 10.16 10.56 18.30 19.00 19

1.2e7 10.62 11.03 19.11 19.85 20

1.3e7 11.04 11.48 19.89 20.66 21

1.4e7 11.47 11.91 20.64 21.44 22

1.5e7 11.87 12.33 21.36 22.20 23

1.6e7 12.26 12.73 22.07 22.92 23

1.7e7 12.64 13.12 22.74 23.62 24

Through table 2 and table 3, we can seethat the softer the stiffness or the lighter thewheelset, the larger the difference between theapproximated value (v1,v2) and the practicalvalue V.

As the results of v2 and v are very close,so we can use Eqs.3 and 6 to quickly calculatethe resonance speed, and the result is quiteclose to that of the multi-body dynamiccalculation. If we replace the track irregularityused in the model with the track irregularity ofGermany high-speed trunk with highinterference or Germany high-speed trunkwith low interference, we can see that thelongitudinal resonance vibration at the sp eedof 20km/h, and the difference betweenresonance vibrations on different trunk is onlythe amplitude of the vibration.

For the simplified model that just has onewheelset, the longitudinal resonance vibrationwill happen on the smooth, level and tangen ttrack without irregularity as well. But theamplitude is smaller compared with theamplitude with track irregularity. It also

indicates that if the wheelset has occurredlongitudinal vibration resonance, thelongitudinal vibrations will not converge at aconstant speed.

V. ANALYSIS OF LOCOMOTIVE MODEL


The root loci analysis of the locomotivemodel is shown in Fig.7. The movement ofwheelset in front and rear bogies is out ofphase at the frequency of 19.2Hz, and theirinteraction makes this vibration not shifted tothe car body; whereas the movement ofwheelsets in front and rear bogies is in phaseat the frequency of 20.6Hz, and theirinteractions on car body is overlapping, whichleads to a strong longitudinal vibration of carbody. Moreover, as the gravity center of bogieframe doesn’t coincide with geometricalcenter, the longitudinal forces from axle boxesto bogie frames will cause nodding movementon both bogie frames, and a vertical excitationdue to the longitudinal force i s thus producedon both ends of car body which leads to astrongly nodding vibration of car body. Thefrequencies of the discussed model will notchange under an uniform running speed.

200km/h~400km/h, 41 points

longi. vib. of wheelsets w.r.tbogie frame in phase

longi. vib. of wheelsets w.r.tbogie frame against phase

-1.00 -0.75 -0.50 -0.25 0.00

0.00

10.0

20.0

30.0

40.0

50.0Root locii

Natural damping

Freq

uenc

y [H

z]

Fig 7. Root loci analysis of the locomotive model

Natural damping

Freq

uenc

y [H

z]


Substitute kx=2.0×107N/m; m=2500kg

into Eq.3

Hzf L 13.20

2500

1022

2

1 7

Suppose the nominal radius of wheelset

is 525mm. Then resonance speed for the

locomotive model is:

v=3.6× (2Rwfw) = 3.6 × [2Rw × (fL/N)] =

3.6 × [2× 0.525 × (20.6/1)]=77.87km/h

Further research showed that in the

locomotive model the nominal mass used to

calculate the inherent frequency should be

slightly smaller than the wheelset mass; and

the nominal stiffness used to calculate the

inherent frequency should be slightly bigger

than the primary longitudinal stiffness. In the

locomotive model, the wheelset longitudinal

inherent frequency is calculated just by the

mass of wheelset and the primary suspension

longitudinal stiffness, without consider ation

on the influence of the second suspension. So

the actual frequency would be bigger than that

calculated by Eq.3, and the actual speed

which is 100km/h would bigger than that

calculated by Eq.6.

5.2. Wheel-Rail Longitudinal Interaction

Vibration corresponding to above

eigenvalues is actually a small local rol ling

vibration of wheelset with respect to bogie

frame, either in phase or out of phase (see

Fig.8). Because of the track irregularit y, a

dynamic component is added to the nominal

forward speed for wheelset.

lo ca l ro lling v ib ra tion o f w h ee lse t

Fig 8. Local rolling vibration of wheelset

within a bogie

That means the reference velocity, whichis often considered constant for creepagecalculation is a variable. According to thedefinition of longitudinal creepage [6], it isexpressed as:

v/)cv(x (7)

where v is reference velocity and c is thecircumferential velocity of wheel at contactpoint. There are different definitions ofreference velocity [7], which lead to differentcalculation results of longitudinal creepage.

0 5 10 15 2027.2

27.4

27.6

27.8

28.0

28.2

28.4

Refe

renc

e Ve

locity

[m

/s]

Time [ s ]

(a) Reference velocity

0 5 10 15 2027.027.227.427.627.828.028.228.4

Circ

umfe

rent

ial V

eloc

ity [

m/s

]

Time [ s ]

(b) Circumferential velocity

Fig 9. Reference velocity and circumferential

velocity at contact point at 100km/h


For the analysed locomotive model , the

reference velocity and circumferential

velocity at contact point are given in Fig.9.

The Longitudinal creepage and adhesion

coefficent of wheelset are given in Fig.10.

0 5 10 15 20-5.0

-2.5

0.0

2.5

5.0

7.5

Long

itudin

al Cr

eepa

ge [

10-3]

Time [ s ]


0 5 10 15 200.00

0.05

0.10

0.15

0.20

0.25

0.30

Adhe

sion C

oeffic

ient

Time [ s ]

(b) Adhesion coefficient

Fig 10. Longitudinal creepage and adhesion

coefficient of wheelset

Longitudinal creepage of leading wheel

at contact patch for example is thus obtained

according to physical measurements in Figs.7

and 10, and at several points maximum

adhesion coefficient reaches saturation value

0.25, see figure 10(b).

5.3. Effect of wheelset longitudinal

vibration on carbody vertical vibration

0 5 10 15 20-7.5

-5.0

-2.5

0.0

2.5

5.0

7.5

Long

itudin

al Ac

c [m

/s2 ]

Time [ s ]


0 5 10 15 20 250.0

0.5

1.0

1.5

2.0

2.5

3.0

Long

itudin

al Ac

c [m

/s2 ]

Frequency [ Hz ]


Figure 11. Time history and frequency response

of longitudinal vibration acceleration of car body

The time history and the frequency

response of longitudinal vibration acceleration

of the car body were shown in Fig.11, in

which the longitudinal vibration acceleration

of the locomotive is 7.5m/s2, and the main

frequency is 20.6Hz.


0 5 10 15 20-75

-50

-25

0

25

50

75

Long

itudi

nal A

cc [

m/s

2 ]

Time [ s ]


0 5 10 15 20 250

5

10

15

20

Long

itudin

al Ac

c [m

/s2 ]

Frequency [ Hz ]


Fig 12. Time history and frequency response of

longitudinal vibration acceleration of wheelset

The time history and the frequencyresponse of longitudinal vibration accelerationof the first wheelset were shown in Fig.12, inwhich the amplitude of the longitudinalacceleration reached 70m/s2. It has two mainfrequencies that are very close and evenintense-coupling. The vibration whichfrequency is 20.6Hz is transferred to car body.

The maximum frequency of thelongitudinal vibration of car body equals thatof the wheelset. The vibration will not betransferred to the car body when thevibrations of two wheelsets in the same bogieare out of phase, as the vibrations of the twowheelsets within a bogie will be counteractedby the interaction. When the vibrations of thetwo wheelsets in the same bogie are in phase,it will cause high frequency vibration, which

would be transferred to car body, and producethe longitudinal oscillation and nod oscillationof the car body. From Figs.11and12, we cansee that the effect of wheelset longitudinalvibration on locomotive longitudinal vibrationis very large.

The time history and frequency respons eof vertical acceleration at the end of car bodyand the middle of car body were shown inFigs.13 and 14. The concerned resultindicated that the vertical vibration of thelocomotive was very intense, and clearlylongitudinal oscillation occured. Through theroot loci analysis, we can see that longitudinaland nodding vibration occurred actually.

0 5 10 15 20-4

-2

0

2

4Ve

rtica

l Acc

[m

/s2 ]

Time [ s ]

End of carbody


0 5 10 15 20 250.00

0.25

0.50

0.75

1.00

1.25

Vertic

al Ac

c [m

/s2 ]

Frequency [ Hz ]

End of carbody


Fig 13. Time history and frequency response of vertical

vibration acceleration at the end of car body


0 5 10 15 20-0.4

-0.2

0.0

0.2

0.4

Verti

cal A

cc [

m/s

2 ]

Time [ s ]

Middle of carbody


0 5 10 15 20 250.00

0.02

0.04

0.06

Vertic

al Ac

c [m

/s2 ]

Frequency [ Hz ]

Middle of carbody


Figs 14. Time history and frequency response of

vertical vibration acceleration at the middle of car body

The spectral analysis of the longitudinalvibration could find that a forced vibrationwith the frequency of 20.6Hz was applied atthe both ends of the car body. The spectralanalysis proved that the vibration frequenciesleading to up and down and nodding of the carbody are the values of 1.67 Hz and 1.37 Hzrespectively.

Although the longitudinal vibration andvertical vibration of car body are extremelybig, the amplitude of lateral acceleration isquite small (see Fig.15). It indicates that thecalculation is stable and the longitudinalresonance vibration of the wheelset is notcaused by the instability of the calculation.

0 5 10 15 20-1.0

-0.5

0.0

0.5

1.0

1.5

Later

al Ac

c [m

/s2 ]

Time [ s ]

End of carbody


0 5 10 15 20 250.00

0.03

0.06

0.09

0.12

0.15

Later

al Ac

c [m

/s2 ]

Frequency [ Hz ]

End of carbody


Fig 15. Time history and frequency response of Lateral

acceleration at the end of car body

5.4. The influence of wheelset longitudinalvibration on wheel treads spalling

Research results show that the strongwheel-rail dynamic force is the main reason

that causes the wheel treads contact fatigue

and spalling. Since the wheel-rail contact is athree-dimensional behavior on the contact

point, the lateral wheel-rail dynamic force can

be divided into three separate dimensionalforces that are lateral force, vertical force and

longitudinal force. On the condition of

reasonable control on the lateral force andvertical force, reducing the longitudinal force

can effectively reduce the wheel -rail dynamic

force. Many factors have influences on


wheelset longitudinal vibration. A strong

vibration, especially a resonance, may causeserious wheel-rail contact wear and shorten

service life. If the vibration is suppressed, one

of sources of wear on wheel-rail contact patch

will be eliminated. It is hopeful to improve thewheelset operation condition through reducing

the longitudinal force, and solve the wheel

tread spalling to some degree. The relevanttests are presently underway.

5.5. Influence of the track irregularity onwheelset longitudinal vibration in thelocomotive model

Through carrying out the linear vehicle

system analysis adopting the Germany high -speed trunks with high interference and low

interference and the medium irregularity as

the track irregularity separately, we get thesystem response depending on the speed [14].

The speed corresponding to the vibration

frequency that approximates to the inherentfrequency is the very speed at which

resonance vibration occurs.

The resonance frequencies with the three

irregularities have slightly differentamplitudes at the speed of 100km/h. So we

can say that the calculated resonance speed is

100km/h. The difference of the locomotivemodel and the simplified model is that the


locomotive wheelset will not occur on thetrack without irregularity.

In this paper, vertical profile irregularity,

alignment irregularity and cross -levelirregularity are considered, and we find the

phenomenon of longitudinal resonance

vibration of the wheelset clearly. Now disposeone of the three irregularities respectively to

find the source inducing longitudinal

vibrations.

First dispose vertical profile irregularityand reserve alignment irregularity and cross -

level irregularity, the result is shown in

Fig.16. It is shown that the longitudinal

acceleration of the wheelset and verticalacceleration of the car body is quite small,


wheelset does not happen. It indicates that ifthere is not vertical profile irregularity, the


wheelset will not take place.

0 5 10 15 20-0.4-0.20.00.20.40.60.81.0 Wheelset

Long

itudin

al Ac

c [m

/s2 ]

Time [ s ]

(a) At the wheelset

0 5 10 15 20-0.03-0.02-0.010.000.010.020.030.04

Verti

cal A

cc [

m/s

2 ]

Time [ s ]

End of carbody

(b)At the end of car body

Fig 16. Longitudinal acceleration without vertical

profile irregularity

Second dispose cross-level irregularityand reserve vertical profile irregularity andalignment irregularity, the result is shown inFig.17. In this case, longitudinal acceleration


of the wheelset and vertical acceleration of thecar body is much small, it doesn’t causelongitudinal resonance vibration too . It provesthat without the cross-level irregularity, thelongitudinal resonance vibration of thewheelset will not take place, and the influenceof cross-level irregularity is much bigger thanthat of the vertical profile irregularity.

0 5 10 15 20-8-6-4-20246 Wheelset

Long

itudi

nal A

cc [

m/s

2 ]

Time [ s ]

(a) At the wheelset

0 5 10 15 20-1.0

-0.5

0.0

0.5

1.0

1.5

Vertic

al Ac

c [m

/s2 ]

Time [ s ]

End of carbody


Fig 17 . Longitudinal acceleration

without cross - level irregularity

At last, dispose alignment irregularityand reserve vertical profile irregularity an dcross-level irregularity, the result is shown inFig.18. Obviously, with the vertical profileirregularity and cross-level irregularity, stronglongitudinal resonance vibration take s place

on the wheelset.

0 5 10 15 20-75

-50

-25

0

25

50

75

Long

itudi

nal A

cc [

m/s

2 ]

Time [ s ]

Wheelset

(a) At the wheelset

0 5 10 15 20-4

-2

0

2

4

Vertic

al Ac

c [m

/s2 ]

Time [ s ]

End of carbody


Fig 18 . Longitudinal acceleration without

the track alignment irregularity

The three above results show that the

longitudinal resonance vibration is the product

of interaction between vertical profileirregularity and the cross-level irregularity,

and the alignment irregularity has little

relation with the resonance. Single irregularitywill not cause longitudinal resonance

vibration of the wheelset at all.

If we replace vertical profile irregularityand cross-level irregularity with the

corresponding Germany irregularity, the


longitudinal resonance vibration will take

place at the same speed. But on a smoothtrack under the same speed, the longitudinal

vibration resonance will not taken place,

which also validate the above conclusion. As

the vertical profile irregularity and the cross-level irregularity always exis ts, the

longitudinal vibration of wheelset always

exists, and can even be developed into strongresonance within some ranges of the speed. If

the vehicle runs under the condition of

resonance vibration, wear of wheelset and railwill be very serious.

VI. A POSSIBLE SOLUTION FORLONGITUDINAL VIBRATION

The control method is reducing the

source causing the longitudinal resonance

vibration. There are two kinds of methods:

one is to improve the track quality, the

condition of the track and reduce the track

irregularity, but this method will cost a lot; the

other is to improve the locomotive

adaptability to the track irregularity, which

can be easily achieved through changing the

suspension parameter.

As the damper can reduce the vibrations,

we can apply longitudinal damp er to the

primary suspension of the locomotive.

Usually the locomotive do not have the

primary longitudinal damper and a small slant

angle is assigned to the damper in the vertical

direction.

. If the vertical damper has certain

inclinations to longitudinal direction, it will

produce a certain damping component in

longitudinal direction. Generally the slant

angle is 25°. Results for different damper

arrangements are shown in Fig.19.

0 5 10 15 20-75

-50

-25

0

25

50

75

Long

itudi

nal A

cc [

m/s2 ]

Time [ s ]

Vertical arrangement

(a) With a vertical angle

0 5 10 15 20-0.50

-0.25

0.00

0.25

Long

itudin

al Ac

c [m

/s2 ]

Time [ s ]

25 degree ahead incline

(b) With a slant angle of 25°

Fig 19. Longitudinal accelerations with different

damper arrangement

Result shows that the longitudinal

damper is very effective for reduc ing

longitudinal vibration. With the vertical

arrangement, an extremely high resonance

occurred, and its amplitude approached

70m/s2. With the slant angle of 25°, 28kNs/m

damping is produced in longitudinal direction

that eliminated resonance totally, and random

vibration with much smaller amplitude

become obvious. Comparing Fig.20 (right)

with Fig.20 (left), we can conclude that a

much longer service life of the wheel could

then be expected.


VII. CONCLUSION

Longitudinal vibration of the wheelsetwith respect to bogie frame always existswhile a vehicle operating on track. Theamplitude of the vibration is influenced bymany factors, such as vehicle structure,suspension and mass parameters, frictioncoefficient in wheel-rail contact patch andtrack irregularity, etc. In the worst case, astrong longitudinal resonance of the wheelsetmay be developed and cause severe wheel -railcontact fatigue, which lead to faster wheelspalling or rail corrugation with certainwavelength. In most cases, this vibration willnot be transferred to car body but still existwithin bogie with large amplitude and highfrequency, which makes contribution onreducing wheelset service life. Anapproximate approach is presented to estimatethe resonance speed according to the study onthe simplified model.

The longitudinal resonance vibration ofthe wheelset will worsen the vertical dynamicperformance and bring longitudinal shake ofthe locomotive. It relates to the cross-levelirregularity and the vertical profileirregularity, and will not take place on thesmooth track. Arranging the primary verticaldamper with a forward angle is an effectivesolution to eliminate longitudinal vibration.

References

[1]. Connon, D F, Pradier, H. Rail rolling contactfatigue research by the European Rail ResearchInstitute[J], Wear, 1996,191(1): 1-13.

[2]. Sun, J, Sawly, K J, et al. Progress in thereduction of wheel spalling[A]. in: Proceeding ofthe 12th International Congr ess on Wheelset[C],Qingdao, 1998: 18-29.

[3]. Sato, Y, Matsumoto, A. Review on rail

corrugation studies[A], in: Proceeding of the 5thInternational Conference on Contact Mechanics andWear of Wheel/Rail System[C], Tokyo, 2000: 74-80.

[4]. Jin X S, Shen Z Y. Rolling contact fatigue ofwheel/rail and its advanced research progress [J].Journal of the China Railway Society, 2001, 23(2):92-108.(in Chinese)

[5]. SIMPACK documentation, 2003, SIMPACK-Track Module, VIII-TE8.

[6]. Garg, V K, Dukkipati, R V. Dynamics of RailwayVehicle System [M]. Academic Press, 1984.

[7]. Knothe, K. A Contribution to thestandardization of definitions for contactphenomena in wheel-rail-systems [J], in: ZEV railGlas.Ann., 2003, 127: 204-211.

[8]. Grassie, S L, Kalousek, J. Rolling contactfatigue of rails: Characteristic, causes andtreatments [A]. in: Proceedings of the 6thInternational Heavy Railway Conference,Southafrica, 1997: 381-404.

[9]. Wickens, A H. Fundamental of Rail VehicleDynamics: Guidance and Stability [M] .Loughborough University, UK, 2003.

[10]. Zhai W M, Wang K Y , et al. Applications ofthe Theory of Vehicle-track Coupling Dynamics tothe Design of Modern Locomotives and RollingStocks [J]. Journal of the China Railway Society,2003, 26(4):24-30. (in Chinese)

[11]. Zhai W M. Vehicle-track Coupling Dynamics[M]. China Railway Publishing Company, BeiJing,2001.

[12]. Lin J H, Chen J Z, et al. Theory Analysis andTest Research of Chinese Main Track IrregularitiesPSD [J]. Chinese Journal of MechanicalEngineering, 2004, 40(1): 174-178. (in Chinese)

[13]. Zhang M. Dynamic of Response of aLocomotive to Random Lateral Rail Irregularities[J].Journal of Southwest Jiaotong University,1994, 29(1): 39-44. (in Chinese)

[14]. Qi F L, Luo L, et al. Application of ModalAnalysis Technology in the Study of Track SystemDynamics [J]. China Railway Sciences, 1999,

20(1): 1-8. (in Chinese)


I. INTRODUCTION

In present, many transport system constructions are now making rapidly progress accordingto social and economic development in Vietnam. Large and many bridges have beenconstructed, are underway and in planning. Vietnam does not have hist ory of large scaleearthquakes, however some small earthquake have been recorded by seismometers installed inVietnam. According to analysis of earthquake records obtained by the seismometers, Vietnam islocated moderated seismic activity area. Seismic des ign Specification in Vietnam (a part of22TCN 272-05) is established based on AASHTO LRFD 1998. Many bridges are design by thisSpecification.

Recently, some bridges have been constructed by Japanese economic support (ODA). Alsosome bridges are constructed in very soft ground or potentially liquefaction sand area.Evaluation by Japanese Specification is required for these conditions. It is necessary in order toconfirmation of seismic performance of the bridge and also to evaluate seismic performance bydynamic response analysis based on Japanese Specification. The dynamic response analysisconsidering material non-linearity can evaluate seismic performance on not only resistance ofmembers but also deformation, unseat, etc…In this paper, unseat of the girders is mainlyevaluated to secure of the existing bridge.

SEISMIC RESISTANCE OF MULTI-SPANS PC BRIDGE UNDER

EARTHQUAKE OCCUR IN VIETNAM

TRAN VIET HUNGMsc., Dept. of Civil Eng., University ofTransport and CommunicationCaugiay, Dongda, Hanoi, VietnamNGUYEN VIET TRUNGDr. of Eng., Professor, Dept. of Civil Eng.,University of Transport and Communication,Caugiay, Dongda, Hanoi, Vietnam

Abstracts: The response of bridges when subjected to seismic excitation can be evaluated

by dynamic response analysis methods. A preliminary seismic response analysis of a multi -

spans highway bridge in Vietnam was performed using d ynamic analysis procedures to

identify the potential for nonlinear response of bridge structure. In this study, a typical

concrete girder bridge (multiple frames in the longitudinal direction) in Vietnam is used to

seismic resistance evaluations. These eva luations are performed by performing nonlinear time

history analyses on FEA method. It is found that the typical bridge and structure

characteristic influence on response of the bridge during earthquake .

Keywords: Seismic analysis, non-linear analysis, bridge in Vietnam, earthquake

engineering, dynamic response analysis.


II. DYNAMIC RESPONSE ANALYSIS

The dynamic response of a structure depends on its mechanical characteristics and thenature of the induced excitation. Mechanical properties which are efficient to mi tigate thestructure’s response when subjected to certain inputs might have an undesirable effect duringother inputs. In a dynamic analysis, the number of displacements required to define thedisplaced positions of all the masses relative to their origina l positions is called the number ofdegrees of freedom (DOF). The equation of motion of an MDOF system is similar to the SDOFsystem, but the stiffness k, mass m, and damping c are matrices. The equation of motion to anMDOF system under ground motion can be written as

[M]{ u } + [C]{ u }+[K]{ u }= -[M]{B} gu (1)

The stiffness matrix [K] can be obtained from standard static displacement -based analysismodels and may have off-diagonal terms. The mass matrix [M] due to the negligible effect ofmass coupling can best be expressed in the form of tributary lumped masses to thecorresponding displacement degrees of freedom, resulting in a diagonal or uncoupled massmatrix. The damping matrix [C] accounts for all the energy-dissipating mechanisms in thestructure and may have off diagonal terms. The vector {B} is a displacement transformationvector that has values 0 and 1 to define degrees of freedom to which the earthquake loads are

applied. Value gu is ground acceleration.

For the purpose of analysis, energy absorbed by inelastic deformation in a structuralcomponent may be assumed to be concentrated in plastic hinges and yield lines. The location ofthese sections may be established by successive approximation to obtain a lower bound solutionfor the energy absorbed. For these sections, moment -rotation hysteresis curves may bedetermined by using verified analytic material models. In addition to a lin ear analysis, it iscommon practice to perform a capacity analysis associated with the desired inelastic response inwhich ductile flexural response occurs at selected plastic hinge regions within the structure. Theplastic hinge regions are detailed to en sure plastic behaviour while inhibiting nonductile failuremodes. The hysteresis properties were the Takeda hysteresis properties. The Takeda degradingstiffness (see Fig. 1) and bilinear elasto -plastic hysteresis (see Fig. 2) were considered to derivethe inelastic spectra.

Fig 1. Takeda model Fig 2. Elasto-plastic hysteresis


III. ANALYTICAL MODELING

The bridge is multi-span continue bridges, representative of typical bridges in Vietnam,evaluated under this study. For superstructure was a continuou s hollow slab beam bridge with 8spans, a total length of 250m. Pier columns and abutments have bored cast -in-place pilefootings. The bridge arrangements are 3 rigid frame piers (P2, P3, and P4) and 4 bent piers isinstalled by rubber bearings support (P0 , P1, P5 and P6). The pier columns are all circular withspiral or circular lateral reinforcements. Elevation for bridge is shown in Fig. 3. The soilconditions are almost medium sand, fine sand and gravelly sand.

Fig 3. Profile of bridge

G ird e r ( lin e a r b e a m e le m e n t)

B e a rin g s p rin gS to p p e r

L o n g .= 0 .1 2 m

S p a c in g

A c tin g fo rc e

D is p .

K = 1 0 k N /m

H o z i.= 0 .0 5 m

8

K = 1 0 k N /m-1

C o lu n m p ie rN o n lin e a r b e a m e le m e n t

N o n lin e a r ro ta tin gP la s tic h in g e s p rin g

B e a m e le m e n tF o o tin g

s p r in g e le m e n t

F o u n d a tio n g ro u n d s p rin g

P

D is p .

k

k 1

Fig 4. Modeling of a bridge pier

In the pier column, a linear rotating spring that modeled a plastic hinge, the column bodywas a nonlinear beam element. The rubber bearing used bilinear model is a nonlinear spring inhorizontal direction. The dynamic analysis was perf ormed using the Newmark β method andintegration time interval was 0.01 second. The nonlinear behavior of the columns is presentedby the Takeda model with the potential plastic hinge zone located at bottom of the column asshown in Fig. 4. The stress vs. strain relation of reinforcing bars is idealized by an elastic perfectplastic model.


Lp = plastic hinge length (Lp = 0.2h – 0.1D; 0.1D ≤Lp ≤ 0.5D) with h = height of thecolumn pier (2)

Table 1. Plastic hinge length of pier, Lp (m)P0, P6 P1, P5 P2, P4(*) P3(*)

h (m) 3.9 4.8 5.7 5.9Lp m) 0.5 0.7 0.4 0.4 0.5 0.5

(*) The piers P2, P3, P4 have 2 locations of the plastic hinge length at bottom and top of pier

The inelastic dynamic analysis is performed by incorporating the non -linear rotationalspring (Kx, Ky, Kθ). The dynamic characteristic value of the surface ground is 0.82s , i.e. type IIIground according to the seismic design specified in Japanese Specifications for Highway Bridge(Part V Seismic design). In the analysis, the damping model used was Rayleigh damping. Thedamping of a structure is related to the amount of energy dissipated during its motion. It couldbe assumed that a portion of the energy is lost due to the deformations, and thus damping couldbe idealized as proportional to the stiffness of t he structure. Another mechanism of energydissipation could be attributed to the mass of the structure, and thus damping idealized asproportional to the mass of the structure. In Rayleigh damping, it is assumed that the damping isproportional to the mass and stiffness of the structure.

[C]= a0 [M] + a1 [K] (3)

in which [C] = damping matrix of the physical system; [M] = mass matrix of thephysical system; [K] = stiffness matrix of the system; a0 and a1 are pre -defined constants. Thegeneralized damping of the nth mode is then given by:

Cn = a0 Mn + a1Kn (4)

Cn = a0Mn + a1 ωn2Mn (5)

nn

nn M2

C

(6)

n1

n

0n 2

a1

2

a

(7)

Fig 5. Rayleigh damping variation with natural frequency


Fig. 5 shows the Rayleigh damping variation with natural fre quency. The coefficients a0and a1 can be determined from specified damping ratios at two independent dominant modes(say, ith and jth modes). Expressing Eq. (8) for these two modes will lead to the followingequations:

i1

i

0i 2

a1

2

a

(8)

j1

j

0j 2

a1

2

a

(9)

When the damping ratio at both the i th and jth modes is the same and equals ξ, it can beshown that:

ji

ji0

2a

;

ji1

2a

(10)

It is important to note that the damping ratio at a mode between the ith and jth modes isless than ξ. And in practical problems, the specified damping ratios should be ch osen to ensurereasonable values in all the mode shapes that lie between the ith and jth modes shapes. In theanalysis, the damping model used was Rayleigh damping. The Rayleigh damping coefficientwas set based on the vibration mode of the structure. We s tudy in case of natural frequencies are2.337Hz and 7.305Hz for damping ratio are 0.0322 and 0.0568, respectively. The values area0 = 0.45829 and a1 = 0.00225.

Table 2. Damping ratio of the members

Member Damping ratio

Bridge column – pier

(nonlinear member)2%

Bridge column – pier, footing

(linear member)5%

Girder (linear member) 3%

Bearing spring 4%

Foundation spring 10%

This analysis used two ground motion records in Japan are acceleration records in theDorokyou Shihousho (Level 1 in Japa n code) and Kushirogawa-1994 (Level 2 in Japan code)with the peak ground acceleration of ground motion records are 1.41m/s 2 (following to lowearthquake occurs in Vietnam) and 4.38m/s 2, respectively. The main seismic attack on moststructures is the set of horizontal inertial forces acting on the structural masses, these forcesbeing generated as a result of horizontal ground accelerations. For most structures, verticalseismic loads are relatively unimportant in comparison with horizontal seismic loads. T herefore,in this study the structure is excited in the horizontal (longitudinal) direction. The records have avariety of peak ground acceleration as shown in Fig. 7.


0

5000

10000

15000

20000

25000

00.0

010.0

020.0

030.0

040.0

050.0

060.0

07

φ (1/ m)

Mom

ent

(kN.

m)

P1, P5P0, P6P2, P4P3

a) Moment vs. curvature relation

0

5000

10000

15000

20000

25000

00.0

005 0.001

0.0015 0.0

020.0

025 0.003

0.0035 0.0

040.0

045

θ (rad)

Mom

ent

(kN.

m)

P1, P5P0, P6P2, P4P3

b) Moment vs. rotation relation

at the plastic hinge

Fig 6. M - and M – θ relationships of column piers


- 1.5

- 1.0

- 0.5

0.0

0.5

1.0

1.5

2.0

0 10 20 30 40 50

Time (s)

Acce

lera

tion

(m/s

2 )

a) Horizontal acc. in the Dorokyou Shihousho



- 6.0

- 4.0

- 2.0

0.0

2.0

4.0

6.0

0 10 20 30 40 50 60

Time (s)

Acc.

(m/s

2 )

b) Horizontal acc. in the Kushirogawa, 1994


Fig 7. The ground motion records used in this analy sis

IV. RESPONSE OF BRIDGE STRUCTURE IN A LOW -MODERATE SEISMIC ZONE

The main objective is to determine relative superstructure -substructure displacementsobtained from an dynamic analysis for typical highway bridges located in a low to moderateseismic zone. Fig. 8 shows the response of the bridge for peak magnitude of the input wave is1.41m/s2 while install stopper and non-stopper. The accelerations of girder are 1.62m/s 2 and2.47m/s2 while maximum displacements of girder are 1.4cm and 2.2cm for install s topper andnon-stopper, respectively. The small displacement of girder can result in frame rigid structure atpier P2, P3, P4. With the results shows under a poor ground excitation, the bridge structure cannot damage of girder and bearing support. In this case, stopper and no-stopper weren’t influentto safety of structure when poor earthquake occurs. Fig. 9 shows the hysteretic response at theplastic hinge of the pier P1 (rubber bearing) and pier P2 (rigid pier and girder). The maximumrotations of P1, P2 are 1.072×10-4 rad, 4.149×10-5 rad and 3.134×10 -4 rad, 2.062×10-4 rad forinstall stopper and non-stopper, respectively. The difference can be caused by piercharacteristics and typical structure. However, the possibility of the undesired behavior, such asunseating failure of the superstructure, is still low since the absolute value of the relativedisplacement is very small.


- 3

- 2

- 1

0

1

2

3

0 10 20 30 40 50 60

Time (s)

Acce

larat

ion

(m/s

2 )


a) Acceleration of the girder at the end girder

(A1 side)

- 0.025

- 0.020

- 0.015

- 0.010

- 0.005

0.000

0.005

0.010

0.015

0.020

0.025

0 10 20 30 40 50 60

Time (s)

Disp

lacem

ent (

m)


b) Displacement of the end girder

(A1side)

0

0.005

0.01

0.015

0.02

0.025

P0 P1 P2 P3 P4 P5 P6

Horiz

onta

l Dis

plac

emen

t (m

) Non stopper Stopper

c) Displacement of top’s pier

Fig 8. Response of the bridge with wave 1.41m/s 2

- 8000

- 6000

- 4000

- 2000

0

2000

4000

6000

8000

- 0.0002 - 0.0001 0 0.0001 0.0002Rotat ion, θ (rad)

Mom

ent

(kN

.m)

Pier P1

- 8000

- 6000

- 4000

- 2000

0

2000

4000

6000

8000

- 0.0001 - 0.00005 0 0.00005 0.0001Rotat ion, θ (rad)

Mom

ent

(kN

.m)

Pier P2

Fig. 9a) Case of stopper

- 15000

- 10000

- 5000

0

5000

10000

15000

- 0.0004 - 0.0002 0 0.0002 0.0004Rotat ion, θ (rad)

Mom

ent

(kN

.m)

Pier P1

- 15000

- 10000

- 5000

0

5000

10000

15000

- 0.0004 - 0.0002 0 0.0002 0.0004Rotat ion, θ (rad)

Mom

ent

(kN

.m)

Pier P2


Fig 9. Hysteretic response at the plastic hinge of pier under input wave 1.41m/s2


V. RESPONSE OF THE BRIDGE WITH STRONG GROUND EXCITATIO N

Following to 22TCN 272-05, longitudinal stoppers shall be designed for a force calculated

as the acceleration coefficient times the permanent load of the lighter of the two adjoining spans

or parts of the structure. If the stopper is at a point where rel ative displacement of the sections of

superstructure is designed to occur during seismic motions, sufficient slack shall be allowed in

the stopper so that the restrainer does not start to act until the design displacement is exceeded.

Fig. 10 and Fig. 11 shows the response of the bridge for peak ground acceleration is 4.38m/s 2

when installed stopper and non-stopper. The accelerations of girder are 38.6m/s 2 and 4.52m/s2

while maximum displacements of girder are 10.13cm and 23.91cm for install stopper and non -

stopper, respectively. The acceleration of girder in case of install stopper larger non -stopper is

caused by the sudden exchange of velocity at the time pounding. Longitudinal displacement of

girder can be restrained by stopper in the top of pier.

- 0.05

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0 10 20 30 40 50 60 70

Time (s)Horiz

onta

l disp

lacem

ent (

m)

Non stopper Stopper

Fig 10. Displacement of the end girder (abutment A1) with input wave 4.38m/s 2

-0.06

-0.04

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0 10 20 30 40 50 60 70

Time (s)

Disp

lacem

ent (

m)


P0P3P2P1

a) Case of stopper

-0.05

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0 10 20 30 40 50 60 70

Time (s)

Disp

lacem

ent (

m)


P0

P1P2

P3


Fig 11. Displacement of top’s pier with input wave 4.38m/s2


- 40000

-30000

-20000

-10000

0

10000

20000

30000

-0.04 -0.03 -0.02 -0.01 0 0.01

Rotat ion, θ (rad)

Mom

ent

(kN.

m)

Resistance Response

Mc,θ c

Mu,θ u

Mc=6520.4kN.m θ c=8.54x10- 5rad My=20756.4kN.m θ y=9.29x10- 4rad Mu=22224.0kN.m θ u=4.18x10- 3rad Mc,θ c

My,θ y

Mu,θ u

My,θ y

PierP1

- 50000

- 40000

- 30000

- 20000

- 10000

0

10000

20000

30000

40000

- 0.04 - 0.03 - 0.02 - 0.01 0 0.01


Mom

ent

(kN.

m)

Resistance Response


PierP2

Mc,θ cMy,θ y

Mu,θ u

Mc,θ cMy,θ yMu,θ u

a) Case of stopper

- 40000

- 30000

- 20000

- 10000

0

10000

20000

30000

- 0.05 - 0.04 - 0.03 - 0.02 - 0.01 0 0.01


Mom

ent

(kN.

m)

Resistance Response

MuMy

Mc

Mc

My Mu


PierP1 -90000

-70000

-50000

-30000

-10000

10000

30000

-0.1 -0.08 -0.06 -0.04 -0.02 0 0.02

Rotation, θ (rad)

Mom

ent

(kN.

m)

Resistance Response

Mc=6569.2kN.m θ c=4.92x10-5rad My=20895.1kN.m θ y=5.34x10-4rad Mu=22276.6kN.m θ u=2.39x10-3rad

PierP2

McMyMu

McMyMu


Fig 12. Hysteretic response at the plastic hinge of pier under input wave 4.38m/s 2

Fig. 12 shows the hysteretic response at the plastic hinge of the pier P1 (rubber bearing)

and pier P2 (rigid pier and girder). The maximum rotations of P1, P2 are 3.15×10 -2rad, 3.67×10-

2rad and 4.06×10-2rad, 8.34×10-2rad for install stopper and non-stopper, respectively. The results

shows almost moment at the plastic hinge of pier overpass moment resistance of pier, thus crack

potential occurs in pier will happen.

VI. CONCLUSIONS

The dynamic behaviors of a multi-span highway bridge system under seismic excitations

are examined with various conditions. On the basis of the results and discussions of the current

study, the following conclusions can be made:

1. Although relative displacement superst ructure - substructure analysis and beam seat

length may be valuable in estimating seismic resistance of bridge in seismic zones, especially

for highway bridges located in a low to moderate seismic zone (i.e. acceleration coefficient,


A=0.09-0.29g) such as Vietnam. This bridge is safety during earthquake in Vietnam.

2. In conclusion, the continuous spans may remain in the elastic range without any

disruption to traffic due to the low seismic excitation such as Level 1, JRA -2002; while a partial

damage may occur due to the strong earthquake such as Level 2, JRA -2002. Thus evaluation

seismic by dynamic response analysis is very importance and needed.

3. The use of stoppers may decrease displacement of girder and prevent unseating fromsuperstructure

References

[1]. Specification of highway bridges, Part V seismic design, Japan Road Association, 2002.

[2]. 22TCN 272-05, Specification for bridge design, Ministry of Transport of Vietnam, 2005.

[3]. TCXDVN 375: 2006, Design of structures for earthquake resistance , Vietnam Ministry of

Construction, 2006.

[4]. Meterials design of bridge structure in Vietnam

[5]. AASHTO. Load and resistance factor design (LRFD) specifications for highway bridges. Washington

(DC): American Association of State Highway and Transportati on Officials (AASHTO), 1998.

[6]. AASHTO. AASHTO LRFD Bridge design specifications, 3rd Edition, Washington (DC): American

Association of State Highway and Transportation Officials, 2004.

[7]. Tongxiang An, Osamu Kiyomiya , Dynamic response analyses and mod el vibration tests on seismic

isolating foundation of bridge pier, Structural Eng./Earthquake Eng., JSCE, Vol. 23, No. 2, pp. 195s -214s,

2006.

[8]. Shigeru Miwa, Takaaki Ikeda , shear modulus and strain of liquefied ground and their application to

evaluation of the response of foundation structures, Structural Eng./Earthquake Eng., JSCE, Vol. 23, No.

1, pp. 167s-179s, 2006.

[9]. Yusuke Ogura, Shigeki Unjoh , Response characteristics of bridge abutments subjected to collision of

girder during an earthquake, Structural Eng./Earthquake Eng., JSCE, Vol. 23, No. 1, pp. 135s -141s, 2006.

[10]. Nasim K. Shattarat, Michael D . Symans, David I. McLean, William F. Cofer, Evaluation of

nonlinear static analysis methods and software tools for seismic analysis of highway bridges, Engineering

Structures, Vol. 30, Issue 5, pp. 1335-1345, 2008.

[11]. M. Ala Saadeghvaziri, A.R. Yazdani-Motlagh, Seismic behavior and capacity/demand analyses of

three multi-span simply supported bridges, Engineering Struct ures, Vol. 30, pp. 54-66, 2008


I. INTRODUCTION

During the last fifteen years, researchefforts in transportation have considerablyevolved at the international, European andnational level. The focus of the research in allfields of automotive industry and transport hasbeen the application of technological andoperational advances that will permit the safest,most comfortable and most cost-efficient possiblemobility of people and goods by private andpublic transport means, while at the same timerespecting the environment and natural resources.Relevant to that has been the development of anintegrated multimodal intelligent transportsystem, that will be efficient in terms of safety,effectiveness, cost and options provided to thepublic with respect their mobility.

The White Paper on Transport“European Transport Policy for 2010: Time todecide” and its mid-term review set outclearly those objectives to be addressed at apan-European level. The TechnologyPlatforms set up in the Transport sectors(ACARE for aeronautics and air transport,ERRAC for rail transport, ERTRAC for roadtransport, WATERBORNE for waterbornetransport, Hydrogen and Fuel cells) haveelaborated long-term visions and strategicresearch agendas which constitute usefulinputs to the approach and activ ities of theTransport theme and complement the needs ofpolicy makers and expectations of society.

In the 6th European research framework,the core objective of the activities carried outwas the promotion of road safety by means of

SUSTAINABLE TRAFFIC SAFETY POLICIES ANDRESEARCH PRIORITIES FOR SAFE AND

SECURE EUROPEAN ROADS

Dr. EVANGELOS BEKIARIS

Research Director of CERTH/HITForum for European Road TrafficSafety Institutes (FERSI) President

Summary: The European goal of reaching 50% reduction of road traffic accidents is

eluding us and efforts on national as well as European level need to be climaxed to even

coverage towards this goal. This paper performs a brief state of the art on recent (up to 6th

FP) efforts on European level to enhance road safety (including passive safety, active safety,

training and other measures) and then key priority areas towards the future (with reference

also to the 7th FP) are proposed along 5 axes: harmonised and complete traffic accident

database, passive safety systems, ITS and active safety systems, measures for dangerous

goods, simulation models and use of driving simulators. The paper concludes with detailed

research priorities, reflecting the author’s views and over 20 years of experience in TrafficSafety Research in Europe.


the use of new technologies (e-safety). Themain objectives of the 6th framework werereflected in the following research initiatives:

Creation of advanced vehicleapplications for accident prevention.

Creation of communication channelsamong the vehicles for the decentralisedmanagement of traffic.

“Intelligent” communicationbetween the vehicles and the infrastructure.

Cooperation between vehicles andmobile devices (PDA and mobile phones)aiming to support the driver or any other userin a seamless and dynamic way.

Development of a unified system forroad users charging and electronic tollcollection in Europe, via the use of satelliteGNSS technologies.

The industries in the automotive sector havebeen the main actors motivating the rapidprogress. Aiming at safer driving and at provisionof added value services to the driver, theautomotive industries have focused on thedevelopment of holistic systems for themanagement of the info provided in the vehicle,the in-vehicle navigation and the communication.

The intelligent transport systems (ITS)applications have started since 30 years ago,aiming, initially, to address the need for theefficient management of road infrastructureand, especially, of the urban network, i.e.roads and interchanges with traffic signs.Indicatively, one may refer to SCOOT (Split,Cycle and Offset Optimisation Technique)and SCATS (Sydney Coordinated AdaptiveTraffic System) applications.

Progressively, and as the telematics anddigital technologies were being improvedmore and more, the ITS applications wereexpanded in all transport means and addressed

a wide range of operational functions. Inspecific, the development of digitalframeworks for navigation, enabled thedevelopment of in-vehicle applications and ofother applications providing services out ofthe vehicle. The advanced localiSation andnavigation technologies, the wirelesstechnologies for mobile devices, the DSRC,RFID, DAB and RDS/TMS technologies havebeen considered the most significantlandmarks in the area. According to m arketestimates, in 2010, the demand for navigationdevices will be around 12 millions per year.

II. THE RESEARCH ROAD TOWARDSSAFER TRANSPORT

Over the last decade, the technologicaldevelopments addressed mainly the passivesafety systems, with regard to the human(mainly the driver), the vehicle and theenvironment. Concerning the vehicle passivesafety systems, the most considerable progresshas been made in relation to preventive carbodies, multiple airbags and advanced seat beltsystems. New structural frameworks (i.e.Honicomb) and materials (i.e. composites) havebeen developed for the front part of the vehicle(mostly of the passenger vehicles, semi-trucksand trucks) so as to be, among others, user -friendly to the vulnerable road users (e.g.motorcyclists, pedestrians, etc.), as well as forthe lateral part of the vehicle (mostly that one ofthe passenger vehicles), for the damping of themaximum possible energy during collision andthe reduction of the vehicle speed with theminimum possible deceleration.

The requirements of the crash tests havebeen further elaborated and are re-evaluatedand re-adjusted according to real accidents’results. The incompatibility among the severaltypes of vehicles, which is critical duringcollision (e.g. height difference in kinematicenergy absorption ranges during the collision


of a passenger vehicle with an off-road vehicle)has been addressed sufficiently, whereas avariety of research initiatives has dealt with thematerial fractural toughness and thereinforcement of the vehicle cabin for vehicleswith high centre of mass (i.e. trucks, buses, offroad vehicles, etc.) in roll-over cases, with thedevelopment and utilisation of moreadvantageous materials for the construction ofthe cabin, special seats, etc. In addition, thepotentials for the readjustment of the passivesafety systems in case the passenger is out ofposition (the original position upon which thesystem was designed and developed) have beeninvestigated. Great progress has been made inthe s/w programs that simulateincidents/accidents scenarios, providing thepotential for the new technologies in passivesafety to be applied and further investigated ina cost-efficient and less time consuming way(i.e. finite elements, “Multi Body Models”,etc.). Indeed, the vehicles produced during thelast decade increased passive safety incomparison to the older technology’s vehicles.As identified in the accident analysis realisedwithin the framework of the projectPENDANT of the 5th FW program, theseverity of the accidents, dealing actually withthe severity of injuries, in which the users ofthe passenger vehicles, constructed one yearafter 1998, were involved, was not as high asthat one corresponding to vehicles, constructedearlier than 1998.

Research in passive safety addressinghuman concerns mainly vehicles, where thepassenger is not surrounded by structures (i.e.cabin in passenger vehicles), and deals mostlywith the safety of bicyclists and motorcyclists.

More specifically, extended research hasbeen realised in these fields during the lastyears (ARPOSYS, TIP-CT-2004-506503),especially regarding the protection of the riderhead. Recently, the utilisation and evolution

of reinforced polymers (e.g. “CarbonReinforced Epoxy”, “SiC/Sic Ceramic MatrixComposites”, “GLARE”) has led to helmets,which are more resistant to collisions andfriction and also much lighter.

Furthermore, the utilisation of materialsthat absorb energy (while falling) for theprotection of knees, elbows, metatarsus,shoulders, pelvis and backbone has beeninvestigated and some minimum requirementsregarding quality and effectiveness have beenset by the EU.

Investigation has been also held for theparticipation and behaviour of the road sidefurniture in accidents (e.g. RISER project),either concerning those that aim to reduce theseverity of the consequences of an accident(i.e. safety islands), absorbing the greatestpossible kinematic energy during the collisionwith any type of vehicle (e.g. motorcycles,passenger vehicles, trucks, etc.) or those thataim to prevent accidents beforehand (e.g.traffic signs, light pillars, etc.).

In the last category, research has focusedon the detection of the most appropriate spotsfor their placement, by means of theinvestigation of the most common accidentscenarios taking into consideration the specificcharacteristics of the road/environment, as wellas on the structural framework of the object andthe materials used, aiming at the preventionfrom high instantaneous decelerations, that canresult in physical damages, and the preventionfrom the penetration of the road side furniture inthe passenger cabin. The above are evaluatedeither using feedback from accident statistics, or,after the implementation of the application, viacrash tests.

However, the most recent trends focus onactive safety systems, that will be able, notonly to provide the maximum possible safetyafter the accident, but furthermore act such as


to prevent it.

Typical applications of Advanced DriverAssistance Systems (ADAS), alsocommercialised nowadays, are the AdaptiveCruise Control systems (ACC) or theAdvanced Vehicle Control Systems (AVCS).ADAS function is based so far on theutilisation of the info that is logged by thesensors, with which the transport means (e.g.passenger vehicle, truck, etc.) are equipped. Inparallel, each year, more vehicles areequipped with various navigation systems,which make use of digital maps and vehiclepositioning. The capability of such systems tobe aware of the geometry and other propertiesof the road infrastructure, with the respectiveguidance and warning, reflects the sense ofcooperative systems and is estimated that mayhave significant positive impacts on trafficsafety and efficiency through the wholetransport network.

However, it must be pointed that thereare no ADAS systems today that take intoaccount the geometry of the road since thereexist no digital maps with information aboutthe curvature or other characteristics of theroad. One possible application is headlightsthat turn in advance according to the curvatureof the road, advanced warning to the driver ifa dangerous or abrupt turn is ahead, etc.Development of maps with detailedinformation about road characteristics ,geometry and condition of pavement is verycostly if carried out with traditional methods.An open research objective is thedevelopment of methods for the production ofadvanced maps that will be emi-automatic andbased on the analysis of paths of vehiclescollected via GPS with on-board units.

Given the fact that the possibility ofimplementing large infrastructure problems islimited (because of limited funding and

technical restrictions), the development ofintegrated intelligent transport systemsapplications is encouraged by the nationaland regional governments, infrastructureoperators and public authorities.

The use of sensors for trafficmeasurements, the detection ofincidents/accidents and the use of VariableMessage and Directional Signs (VMS/VDS)are already used in Europe in great extent.

Currently, the most dominant trend is theintegration of existing or underimplementation projects in an interoperableframework that will allow the cross -borderadoption of ITS.

The most innovative projects are thosedealing with electronic tolls collection(PISTA, MEDIA), e-ticketing and theexchange of traffic and other info amongseveral actors (cities, districts, etc.).

The major scope of ITS is the increase ofmobility of goods and people, in such a way,as to be in favor of all involved actors and theenvironment. This is especially applicable inthe field of dangerous goods transport, wherethe traffic incidents have multiple negativeimpacts with regard to safety (of the driversand the third party) as well as the fina ncialstatus and the marketing profile of thetransportation and the dangerous goodscompanies, but also to the environment.

For the above reasons, European projectsaim at the development of such technologiesand integrated services that will allow the s afeand more cost-efficient transport of dangerousgoods through the whole transport network.Most of them deal with the development ofthe technological framework that is requiredfor the dynamic management of the dangerousgoods fleet and the seamless pr ovision ofreliable info for the vehicle, the driver and the


cargo status, the time of delivery, the potentialof terrorism action (i.e. SAFESEANET ,TEMPO ARTS, MITRA, RIS, MVS, SHAFT,DETRACE, ULISSE, etc.), whereas in recentresearch initiatives, more integrated servicesare designed, including emergency services(Police and Ambulance reaction) and routeguidance with advanced Decision SupportSystems, that take into considerationoperational, financial and the environmentalinfo, real traffic conditions and riskassessments (i.e. GOOD ROUTE project).Common ontologies with regard to theclassification of the dangerous goodsaccording to ADR («Agreement concerningthe international carriage of Dangerous goodsby Road»), security and authenticationsystems for the privacy data protection (ofusers, companies, etc.), advanced navigationsystems and user interfaces for all involvedactors, reliable localisation, positioning andcommunication technologies, trafficmanagement information centres andimproved vehicle tracking technologiesconstitute only part of the existing and futuretechnological development in this area.

In parallel, some research initiatives (i.e.INFORMED project) have focused on the training ofthe professional drivers and their instructors,developing training programs that include training inadvanced techniques (i.e. anti roll-over, antiskid,defensive driving, etc.) of several types of vehicle,incorporating the use of multimedia software trainingtools, training with simulators and practical training(i.e. in test trucks), whereas have formulated a set ofpolicy recommendations for the improvement of therelevant European Directives dealing with trainingissues in this field. The need for the formulation of acommon European training, assessment andcertification framework for the professional and forall other types of drivers is emerging and obvious.

Before any other process, the appropriatecollection, reporting and in-depth analysis of

accidents, the reduction of whi ch is theobjective of each system under development,is the first mandatory step. The accidentanalysis may be performed in several scales,varying from the analysis at national level,where the total number of one countryaccidents is investigated and in ternationalcomparisons are further made, to the scale ofindividual accidents, where representation andin-depth assessment of the accident is realizedaiming at the identification of the root causesleading to that.

A series of research projects have b eenfunded for this purpose (i.e. STAIRS, EACS,PENDANT, SAFETYNET, TRACE, etc. ),whereas in several databases, accidents inEurope and, in some cases, in the rest of theworld, are reported (e.g. FACTS, NHTSA,MHIDAS, GES, etc.).

Simulation and modeling techniques ofthe vehicle and the traffic environment arealso considered to have a significantcontribution to preliminary research phase,during the last decades.

Traffic simulation models aredistinguished in microscopic (e.g.PARAMICS, RuTSIM, VISSIM, etc.),mesoscopic and macroscopic (e.g. VISUM,SATURN, etc.). Each of these categoriesdeals with different level in research. Forexample, the microscopic models are based onthe principles and the sense of vehiclesequence, analyse the individual behaviour ofeach vehicle, providing great accuracy bymeans of dynamic simulation and are usedmainly for the evaluation of the proposedstrategies and policies in middle and smallnetworks. The significant computing time thatis required makes the simulation of largenetworks unprofitable or even totallyunfeasible (e.g. in urban areas level, or for theassessment of several attributes, such as mean


speeds of vehicles, traffic distributionthroughput the network, etc.). The latestmentioned are addressed by macroscopicmodels that are static and are based on thebehaviour of the average vehicle population,following fluid dynamics theories. Suchmodels require less computing time; howevertheir accuracy is not so good. In addition, thetraffic simulation models are used for the newand innovative technologies (e.g. ADAS,AVCS) to perform impact assessment and toestimate the magnitude of variousenvironmental impacts (CO 2 emission, fuelconsumption, noise, etc.).

An indicative European project, withinthe framework of which, microscopic andmacrospopic models have been developed isthe ADVISORS project (GRD 1 1999 10047),whereas the ΙΝ-SAFETY project (FP6-2002-506716) aimed at the development andevaluation of microscopic and macroscopicmodels that assess the behaviour of users ofADAS/IVIS in several penetration rates, toenable the impact assessment in road safety.

Driver simulators simulate the vehicleoperation and the respective trafficenvironment. The accuracy of the simulation,their technical characteristics and their costmay differ significantly, depending on thepurpose of use. Driver simulators are used forseveral reasons, as for example, for thetraining of all drivers’ categories (e.g. novicedrivers, elderly drivers, professional drivers,etc.), for the assessment of their skills andtheir driving behaviour, for other researchpurposes like the design and development ofvehicles and parts of them (e.g. userinterfaces, ADAS, etc.), games andentertainment, etc.

The undergoing research in the areas ofdrivers’ simulators has provided evidencethat, within the different research and training

contexts of use and for the achievement of thedifferent goals each time, different simul ators,scenarios and environments are required,adjusted to the concrete needs of theapplication. Driver simulators may be singledisplay simulators, static, dynamic, semi -dynamic, virtual reality simulators, etc. Someof the best research driver simulato rs are thoseof VTT in Sweden, of Daimler Chrysler inGermany and of NADSin the U.S. Worktowards interoperable, multi -tasking and witha common reference architecture simulators iscurrently coordinated in TRAIN -ALLinitiative, involving most major simul atormanufacturers in Europe.

E-112 is a European directive requiringmobile and fixed operators to make availablethe location of every caller placing anemergency call. Mobile handsets are currentlylocated through the mobile operators whomake use of various techniques based on theknown locations of network antennas. Whileall European countries have in principleadopted the directive the system is not fullyoperational yet and is facing severe delays.

E-112 will form only part of e-call achain of actions that will bring rapidassistance to any motorist who experiences anaccident or mechanical failure on the road. E -call is a EU high priority initiative andsignificant research has been alreadyundertaken.

An e-call can be initiated manually by thevehicle occupants or triggered by an accidentand placed automatically by a black boxconnected with sensors that detect a collision.A voice and data connection is establishedwith the closest Public Service AnsweringPoint, which deploys and dispatchesassistance to the location communicated bythe call. Furthermore, it is envisioned thatrelevant data are transferred to a Service


Provider that provides additional servicessuch as towing or notification of next -of-kinfor which the caller subscribes. Initial ly, itwas forecasted that e-call would beoperational in Europe by 2010. Since there aresignificant delays to the deployment of E -112,it can be expected that the deployment of e -call will be further delayed.

The problems can be partly attributed tothe fact that the techniques used for thecalculation of location of mobile phones arenot very accurate or require significantinvestments from the mobile operators in orderto be acceptably accurate. Many operatorschoose the cell id method, which does notrequire additional equipment but provides theposition of a handset within a cell. Cells aresmall enough in urban areas, but cover bigareas outside of the cities resulting inpositioning with large margin of errors whereis needed more, e.g. in rural areas.

The increased availability of mobilehandsets with GPS capabilities (and laterGalileo) will solve the location problem. It isexpected that within the 3-5 next years allmobile handsets will be GNSS capable.

Significant research efforts should bemade in solving problems such as theprotection of privacy and developing servicesthat will make the whole concept of e -callcommercially viable. Since additional services(towing, notification) will be offered on asubscription basis, a whole bouquet of relat edservices must be on offer that will beattractive for the average driver and couldsustain business cases for the future ServiceProviders. Perhaps, initially Service Providerswould be insurance companies or roadsideassistance companies, but the real challenge isto develop and offer truly innovative servicesthat will create a new telematics -relatedindustry based on e-call. At the same time,

the benefits would be the reduced responsetime and the fast arrival of medical assistanceat the accident, which is known that highlyimproves the chance of survival.

III. THE ROADMAP TO THE FUTURE

Future research, already underimplementation in the 7 th FP of the EC, aimstowards a system that will enable theincorporation of the most recent evolutionsand achievements in the passive and activesafety fields, into the traffic safety arena. Thebenefited groups will be the society as awhole, the enterprises, the competitiveness ofwhich will rise in such a way, as to allow theirpenetration to the European and t heinternational arena, with the adequate capacityand know-how. More precisely the followingpriority research areas are correlated with highpotential impacts:

The complete recording and analysis oftraffic accidents, will result in theidentification of the major problems and needsthat will be targeted by the several systemsunder development.

Research around passive safetysystems and the respective implementationsare expected to:

Reduce the severity of the injuriesfrom collisions corresponding to passengersof vehicles and all other road users.

Motivate the further development ofthe technologies and sciences (finiteelements, etc.), which are used within theframework of the passive safety systems atnational level.

The focused research on ITS and therelevant implementations, especially thoseconcerning cooperative systems, areexpected to:

Improve the traffic flow and reduce


the negative environmental impacts in thetransportation sector (especially in an urbanenvironment) via the new combined mobilityservices.

Achieve considerable improvementswith regard to the safety, efficiency andcompetitiveness of the transport systems, thetraffic efficiency and in general the feasibilityin transport (in compliance with the Europeanvision for the reduction of fatal accidents by50% until year 2010 and by 100%, “0 fatalaccidents” in long term perspective ).

Demonstrate, qualitatively andquantitatively, via large scale Pilot trials, thepositive impacts of ITS in all aforementioned,encouraging the funding and the coordinationof all relevant initiatives on behalf of allinvolved actors.

Research with regard to thedevelopment of integrated services fordangerous goods fleet management and theimproved training, assessment andcertification of the drivers and theirinstructors are expected to:

Allow all dangerous goods vehiclesto be continuously tracked and monitored,providing the relevant notification,information or warning to all involved actorsautomatically, with no physical interventionand vehicle immobilisation and loss of timeand with no occurring problems and risks inthe traffic flow, which are very common incases, where heavy vehicles are put aside theroad.

Increase safety of the drivers andthird parties that are directly influenced(especially in urban areas), to face theterrorism in this area and the considerableenvironmental pollution due to the occurringaccidents.

The use of simulation models may lead

to significant savings in resources andincreased road safety since it is expected to:

Contribute towards the evaluation ofvarious transportation policies before they areapplied thus providing decision makers witha tool that permits them to perform “what -if”scenarios and permits them not only toestimate traffic loads congestion etc. but alsovarious environmental indicators (e.g. CO2emissions, etc.).

Permit the simulation of the impactsof new technologies (ADAS, IVIS) in theexisting road networks, before these areapplied.

Make feasible the construction ofnew and/or the improvement of existinginfrastructure, in the less expensive way,since the s/w for transport modelling mayprove to be especially effective tool in thecontext of RSA, RSI and black spotsmanagement.

The utilisation of driver simulators isexpected to:

Increase the safety of the drivers,mostly of the candidate and elderly ones, sincethe driver will have the opportunity toexperience a series of driving tasks and trafficenvironments, before s/he drives in real trafficconditions and also situations, which are verydifficult or totally unfeasible to be tried in realtraffic conditions (e.g. driving with fog, snow,collision with another vehicle, pedestrian, etc.).In this way, the training procedure and contentare also improved.

Make feasible the detection and theadoption of corrective measures for newsystems and infrastructures (being simulatedvia the proper s/w modeling), before these areapplied, by means of their evaluation from allaspects, including the investigation of the targetusers acceptance, leading to potential avoidance


of accidents and unsuccessful investments

IV. CONCLUSION AND RESEARCHPRIORITIES

The priorities outlined above areconsistent with those set by the EuropeanCommission. EC supports the implementationof such projects that comply with the identifiedpolitical priorities for the unification of thetranseuropean transport networks, as these areexpressed through a series of Directives, orstudies that are related, for instance, to: ThePan-European adoption of e-Call, the creationof a common service of road users charging,the plans for the GNSS technologies adoption(Directive 2004/52: Interoperability ofElectronic road toll systems), the creation ofcommon services for drivers and passengers,accessible by all (according to “Article 169 ofthe Treaty” and the priorities of the “AmbientAssisted Living” area and the “EuropeanStatement of Principles”), GALILEO adoption,the eSafety initiative, the activities of therecently developed ‘Agency for ITSimplementation’, etc.

A series of European TechnologicalPlatforms are related to the long-term goals ofall transport fields and modes (ACARE for airtransport, ERRAC for railway transport,ERTRAC for road transport, WATERBORNEfor seaways transport, etc.). ERTRA C is theone related mostly to Road Safety, whererelevant research priorities are defined.

According to the author, the mostrelevant research priorities follow below:

1. Sufficient collection of detailedaccident data, based on the current needs inresearch. Location of accidents should beregistered with detail preferably with GPS.Moreover, it has to be stressed out that

besides traffic accident data, exposure data arealso needed. Through analysis of trafficaccidents data will permit the pinpointing ofthe reasons they occur and result in theadoption of appropriate measures (both policyrelated as well improvements in theinfrastructure). Specifically the followingactivities should be supported:

Research for the development ofnational database that contains informationon all accidents that is regularly updated.

Development of a GIS database thatcould be used to analyze the occurrence ofaccidents taking also in account the location,the geometric characteristics of the road, etc.

2. Further research, development andevaluation of advanced passive safety systemsfor the vehicle, the driver and theenvironment, which will reduce considerablythe severity of accidents and will contribute totheir avoidance, as much as possible. Inspecific, at national level, the followingactivities should be supported:

Research for the development ofsafer road side furniture and passive safetysystems for the driver and the vehicle.

Research for the development ofnew structural frameworks and the utilisationof new material for the aforementioned andthe reduction of the incompatibility betweenseveral types of vehicles.

Evolution of the crash testsrequirements and validation of them (viaaccident analysis, s/w simulations and shortterm trials in real conditions).

Evolution and improvement of s/wprograms for simulation scenarios(accident/incident scenarios) and for the


structural analysis making use of the newcomposite materials.

3. The development of integrated ITSsolutions, that will result in a more eff icientand sustainable use of transportation and totraffic accidents reduction. This will beachieved by means of advanced warningstrategies and risk detection, a reliablenetwork of sensors, technological integrationof so far independent ADAS and inter actionof them with the user. The proposed solutionshave to be highly efficient, reliable andcontribute to the increase of safety andcomfort during driving and also friendly to theenvironment. Activities to be supportedinclude the development of:

European Strategic Research,Development, Implementation and Use ofITS, in order to meet national priorities and toachieve multiple benefits.

European Architecture, commonrequirements (Quality of Service andInteroperability) and ontologies (databasesand web services) for the seamless,interoperable and cost-efficient use of ITSanywhere, any time and from anyone.

Identification of critical thematicareas of ITS and the establishment of afocused development and applicationframework per area, via the adoption of“umbrella” projects.

Deployment of new sensors and theimprovement of already existing ones for themost reliable possible perception of theenvironment and the fulfillment of complexscenarios of use interfering withintersections, interchanges, tracking ofvulnerable road users under several trafficconditions (normal, adverse, low visibility

conditions, etc.).

Driver warning strategies, automaticcontrol of the vehicle and interactive userinterfaces (with haptic, acoustical, visualchannels) as well as of evaluationframeworks for their assessment in Pilot trialsand large-scale Field Operation Tests (FOTs) .

Smart parking management systemsthat permit trip makers to check availabilityof parking through a centralized parkingsystem and permits them to makereservations in advance, or in real timethrough a bidding process.

4. Especially, for the cooperative systemsarea, the development of integratedcooperative systems which will provideadvanced, reliable, fast and safe vehicle tovehicle communication and vehicle toinfrastructure communication in real time,aiming at the provision of information andwarning to the users in time and the automaticor semi-automatic ADAS activation, vialocalisation and positioning technologies andadvanced sensor networks. The proposedsolutions need to be financially feasible,aiming at the most limited possibleintervention in the existing nationalinfrastructures. In specific, at national level,the following activities should be supported:

The recording of the cooperativesystems applications, the identification ofdeficiencies and insufficient or complete lackof technological implementation (wheneverthis is considered necessary fromsocioeconomic aspects and also technicallyfeasible), the formulation of concreteproposals for the improvement or the fullimplementation of infrastructure and thedetermination of the respective short termand long term technological plan.


The development of a normalisedArchitecture, common requirements (Qualityof Service and Interoperability) andontologies (databases and web services) forthe seamless, interoperable and cost -efficientuse of ITS anywhere, any time and fromanyone.

According to the above ITSArchitecture, the development and interfacingof traffic and information managementcentres in urban, extra-urban, rural andinterurban networks.

The development of simulation toolsand evaluation platforms, which will allowthe technical and socio-economic evaluationof the proposed solutions.

5. The execution of large scale Pilots insimulators and test tracks for the evaluation ofITS applications (addressing also cooperativesystems), with regard to the reliability of theirperformance and their user friendliness andacceptance. In parallel, feasibility studies forthe proposed solutions will be conducted inthe national and European market. T hefollowing activities should be supported:

Evaluation frameworks and test plansfor Pilots (FOTs) impact assessment withregard to safety and the traffic environment ,which will include, among others, theexperimental and statistical planning and inadvance simulation, techniques for themeasurements gathering, addressing subjectiveand objective criteria, methodologies for theselection of the statistical sample, the scenariosof use and the timetables according to scientificmethods, methods for the trials conduct andfinally the drawing out of quantitative andqualitative conclusions dealing with theexpected impacts of the tested anddemonstrated applications.

The compatibility check of ITSapplications against the policies set in the areasof Transport and Environmental protection inEurope before their approval and funding.

Large scale ITS applications Pilotsand impact assessment. In specific, for thecooperative systems area, pilots for theevaluation of solutions dealing with vehicleto vehicle and vehicle to infrastructurecommunication and comparison of them toexisting solutions (via s/w for transportmicro/macro-modeling) should be encouraged.

In the European Research area, the aboveResearch priorities are already addressedwithin the relevant Research programmes ofDG INFSO, DG RESEARCH and DGTREN;whereas their implementation is under theumbrella of an EC Agency, that develops an‘Action Plan for the deployment of IntelligentRoad Transport Systems for more efficient,safer and cleaner transport’.

References

[1]. Communication from the Commission to the

European Parliament, the Council, the European

Economic and Social Committee and the

Committee of the Regions - Preparing Europe’sdigital future i2010 - Mid-term review

{SEC(2008) 470}.

[2]. ESafety Final Report of the eSafety Working

Group on Road Safety

(http://ec.europa.eu/information_society/activities/esafety/doc/esafety_library/esafety_wg_final_re po

rt_nov02.pdf), November 2002.

[3]. European Commission ‘Keep Europe moving’.Mid-term review of the 2001 transport White

Paper, 2006, ISBN 92-79-02312-8.

[4]. European Commission WHITE PAPER

‘European transport policy for 2010: time todecide’, 2001, ISBN 92-894-0341-1

http://ec.europa.eu/information_society/activities/


CT 2

I. INTRODUCTION

1.1 Backgrounds and problem statements

The available cost data on highway projects in developing countries are generally limitedin numbers, incoherent and contain a large amount of incomplete data. Limited financialresources and lack of systematic cost data collection are also the major problems which affectcost database quality. These factors negatively affect to cost estimators and planners to makeappropriate decisions. Accurate cost estimation in early project development is an impor tantissue where detailed information is not available and project costs are to be decided, in mostcases cost estimating relationship techniques are used. One of the challenges is accurate costestimation during pre-feasibility studies.

Most research studies on cost estimation early stage of project were conducted using datafrom developed countries (Healey, 1964; Sanders et al., 1992; Pearce et al., 1996; Smith et al.,

PRELIMINARY ROAD COST STUDIES IN

DEVELOPING COUNTRIES

Lecturer JAMSHID SODIKOVTashkent Automobile and Road Institute,20, street Movounnahr,Tashkent, 100020

Advisor EconomistZIYODULLO PARPIEVUNDP Uzbekistan Country Office

Abstract: In the early estimation there is compromise between the amounts of

information available and accuracy of estimation. We propose three levels of analysis such as

regional, country and project level for road cost models in order to provide efficient data

usage. The data for our research was obtained from the World Bank’s ROCKS database,which contains unit costs for road projects from over 80 developing countries. This paper

investigates the impact of road upgrading and improvement works on overland trade in 18 out

of 32 member countries of Asian Highway Network. The results indicated approximately 6.5

billion US dollars is required to upgrade roads and improve existing surface condition of the

selected sub-network with total length of 15,842 km. The gravity model approach was adopted

to quantitatively evaluate overland trade expansion taking into account road quality

improvements with two scenarios such as road quality increases up to 50% in the first

scenario in the second one up to 75%. The results suggests that in the first scenario total

intra-regional trade will increase about 20 percent to 48.7 billion US dollars annually, while

second scenario predicts that trade will increase by about 35 percent to 89.5 billion US

dollars annually.

Key words: Construction costs - Maintenance costs - Developing countries -Regression models


CT 2

1997; Hegazy et al., 1998; Adeli et al., 1998; Al -Tabtabai et. al., 1999; Morcous et al., 2001;Flyvbjerg et al., 2002, Emsley et al., 2002; Levinson et. al., 2003; Trost et al., 2003; Gwang -Heeet. al., 2004). Very limited number of research has been done in developing countries(Archondo-Callao et al., 2004; Buys et al., 2006). One of t he major reasons for a few numbersof researches is related to cost database availability in developing countries. Initial steps to buildcost database for developing countries was initiated by the World Bank’s Transportation Unit in1999 and ROCKS (Road Costs Knowledge System) was introduced. Cost data collected fromdeveloping countries all around the world and this system has large amount of incomplete datain some data items. This explains the need to examine and analyze the cost estimationtechniques and how to deal with missing data in developing countries context which is theprimary objective of this study. Additionally, due to nature of available data in the ROCKSdatabase efficient data usage has been introduced by level of analysis which is also i ncorporatedin this study.

Road agencies, contractors, consultants and financial institutions need road costsinformation, which usually is locally available, but in many case it is scattered and collected inunsystematic ways. These entities need to asse ss costs differences, but no framework tocompare road costs exists. In 1999, in response to this demand, the World Bank made the firstattempt to collect this information from 67 Implementation Completion Reports of Bank –financed projects that were implemented in the period 1995 – 1999. The study found that thelevel of detail provided in these types of documents was limited and that there is a worldwideneed for a framework to collect this type of information in 2000. Consequently, the Bankdecided to develop a simple system to collect road costs and to explore other sources ofinformation. This effort resulted on the ROCKS, which is being developed by the World Bank’sTransport Unit and is primarily based on the experience of Bank staff and the informat ioncontained in roads and highways projects in developing countries.

1.2 Concept of level of analysis

Concept of level of analysis was evolved from cost model application perspective andexisting data availability. As mentioned earlier that ROCKS database contains missing data, thetask is to utilize available data in efficient way which assists to develop better cost model withcertain application purpose. These application purposes can be defined as cost estimation studyin a given geographical region, cost study within a certain country and finally cost estimation ofa specific project with detailed information. It is generally known that detailed information islimited in regional cost studies especially during preliminary cost studies. This is due to e achcountry in particular region has certain amount of cost history data which vary among countries.In some countries there may be more project details are available in others limited projectdetails are available, in some extreme case there may be no cos t history database at all. In orderto balance available data amount (project details and number of projects) with cost modelapplication purposes, three level of analysis were proposed such as:

Regional level – cost model is developed based on limit ed project details with largenumber of observations


CT 2

Country level – cost model is developed based on relatively detailed project informationwith comparatively smaller number of observations

Project level – cost model is developed based on in depth project details with limitednumber of observations

Regional level of analysis – ROCKS database was divided into 6 regional subsets forregional level of analysis namely Africa, Asia, Caribbean -Central-Middle America, East Asia-Pacific Islands, Europe-Middle East, South America region. From figure 1 it can be observedthat large number of projects belong to Africa with 29%, Asia with 24%, and Europe -MiddleEast with 19%. Total number of projects consists of 1385 from 85 countries

Figure 1. Regional Level Project Distributions (World Data)

Average unit cost of work activities among these regions varies significantly. For instancenew construction projects in Africa region on average is about one million US dollars per kmbut in Europe-Middle East on average new project is about one and half million US dollars perkm.

Although number of observation is quite large but available data items are limited. Thesedata items are workactivity and pavement width in ROCKS. The other data items such as GDPper capita, annual mean precipitation, road network density, coastline divided by area arecollected from external sources to build cost model for regional level of analysis.

Country level of analysis include additional data items t o regional level of analysis fromROCKS such as rate of work per area which obtained by dividing work duration to pavementarea and contractor type. But number of observation has dropped from 1385 to 318. Additionaldata items are available only for Armeni a, Ethiopia, Ghana, Kyrgyz Republic, Lao PDR,

19%

12%

29%

24%4%

12%

Africa

Asia

Carib-Cent-Mid America

East Asia-Pacific Islands

Europe-Mid East

South America


CT 2

Nigeria, Poland, and Uganda. Distribution of projects in each country is shown in figure 2.Countries with large number of projects are Ghana, Uganda, Poland, and Lao PDR.

Figure 2. Country Level Project Distributions

Project level of analysis includes additional data items to country level of analysis fromROCKS such as terrain type, climate and surface thickness data. But number of observation hasdropped from 318 to 56. Additiona l data items are available only for Kyrgyz Republic, LaoPDR, and Nigeria. Distribution of projects in each country is shown in figure 3. The largestnumber of projects belongs to Lao PDR .

Figure 3. Project Distributions at Project Level

16%

20%

64%KyrgyzRepublicLaoPDRNigeria

1%3%

23%

12%

3%

12%3%

43%

Armenia

Ethiopia

GhanaKyrgyz Republic

Lao PDR

Nigeria

PolandUganda


CT 2

Data for project level of analysis is very limited in terms of number of observation butproject specific data such as terrain, climate data and surface thickness are available. Only threecountries have these data for bituminous reconstructio n and partial widening bituminous with 2lanes with reconstruction projects.

II. COST MODELS

2.1. Regression cost models

Regression models have been proven to be reliable and used for decades. There are someadvantages of regression models such as they can be defined by mathematical expression andexplain relationship between dependent variable and independent variables. There are also somedisadvantages in regression models such as multicollinearity, nonlinearity, heteroscedasticityand other issues which occur in regression model development. The details of these issues arewell described in literature (Lewis-Beck M. 1980, William D et al., 1985, Brikes D et al. 1993,Allison S et al., 1999, Miles J et al., 2001, Frank E. 2001). One of the powerful techni ques toovercome shortcoming of regression models is transformation dependent or independentresponse or both. Several types of data transformation were tested and log -log transformationwas chosen for our analysis. The advantage of log -log transformation lies on ease ofinterpretation (Carroll J et al., 1988).

At regional level we postulate that unit cost (UCij) is a function of country’s GDP (Gi),country’s road network density (RNDi), pavement width (PW), country’s annual meanprecipitation (APi), coastline divided by area of the country (DLi), project type (PTi) and region(RGi) defined as follows:

ijεj jPTjβ

j jRGjγiDL5αilogAP4αlogPW3αilogRND2αilogG1α0αijlogUC




RNDi = road network density of country i(km per 1000 km2)

PW = pavement width (m)

APi = annual mean precipitation of country i(mm)

DLi = coastline divided by area of country i(km per 1000 km2)


RGi = dummy variable for region

At country level we postulate that unit cost (UC ij) is a function of country’s GDP (G i),country’s annual mean precipitation (AP i), coastline divided by area of the country (DL i), rate ofwork per area (RW), project type (PT i) and contractor (CTR j) defined as follows:


CT 2

ijεj jCTRjγ

j jPTjβlogRW4αiDL3αilogAP2αilogG1α0αijlogUC



Gi = GDP per capita of country I (US $ 2004, PPP)

RW = Rate of work per area (Work duration divided by pavement area)

APi = annual mean precipitation of country i (mm)

DLi = coastline divided by area of country i(km per 1000 km 2)



At project level we postulate that unit cost (UC ij) is a function of country’s GDP (G i), rateof work per area (RW), surface thickness (PST), terrain (TR j), climate (CLj), project type (PT i)and contractor (CTR j) defined as follows:

ijεj jPTjj jCTRjλ

j jCLjγj jTRjβlogPST3αlogRW2αilogG1α0αijlogUC




RW = rate of work per area (Work duration divided by pavement area)

PST = pavement surface thickness (mm)

TRj = dummy variable for terrain type

CLj = dummy variable for climate



III. CASE STUDY

3.1. The asia highway network

The AHN covers routes which cross 32 member countries with approximate t otal length of140,000 km. It starts from Japan, Tokyo , and extends to Finland, Helsinki and Bulgaria, Sofia.The network passes mostly through existing roads in those countries. Asian Highway Designstandards comprise of Primary, Class I, Class II, and Cl ass III highway classifications which aredefined according to terrain classification, design speed, width (including right of way, lane,shoulder, median strip), minimum radii of horizontal curve, pavement and shoulder slope, type


CT 2

of pavement, maximum superelevation and grade, and structure loadings (AHNDS, 2004).Table 1 highlights recommended design standards for Asian Highway routes. The minimumrequirement which satisfies the Asian highway network standard is the Class III. According toAsian highway standards, it is suggested that the Class III should be applied only when thefunding for the construction or land for constructing road is limited. There are two priorities inwhich the Asian highway network member countries have to carry out, 1) To improv e roadconditions where it is necessary in Primary, Class I, Class II and Class III which cover 72% oftotal network, 2) to upgrade the rest 28% of network at least to the Class III but preferably toClass II.

Table 1. Asian highway standards (AHNDS, 2004)

Highwayclassification

Primary (4 or morelanes)

Class I (4 or morelanes)

Class II (2 lanes) Class III (2 lanes)

Terrainclassification

L R M S L RM

S L R M S L R M S

Design speed(km/h)

120

100 80 60 100 80 50 80 6050

40

6050

40

30

Width (m)

Right ofway

(50) (40) (40) (30)

Lane 3.50 3.50 3.50 3.00 (3.25)

Shoulder 3.00 2.50 3.00 2.50 2.50 2.00 1.5 (2.0)0.75(1.5)

Medianstrip

4.00 3.00 3.00 2.50 N/A N/A N/A N/A

Min. radii ofhorizontal curve(m)

520

350210

115

350210

80210

11580

50

11580

50

30

Pavement slope(%)

2 2 2 2 - 5

Shoulder slope(%)

3 – 6 3 – 6 3 – 6 3 - 6

Type of pavementAsphalt/cement



concreteDbl. bituminous

treatment

Max.superelevation(%)

10 10 10 10

Max. verticalgrade (%)

4 5 6 7 4 5 6 7 4 5 6 7 4 5 6 7

Structure loading(minimum)

HS20-44 HS20-44 HS20-44 HS20-44

The AHN database contains info rmation of the routes for the most of the countries, whichis about 18 countries out of 32, these information include road surface condition, pavementtype, terrain and other (AHND, 2004). Table 2 shows route condition and design standard ineach country. From this table it can be observed that about 15,842 km need to be improved orupgraded in order to provide good transportation communications.


CT 2

Table 2. Road surface condition and design standards in ESCAP member countries

No. Country Route No.AH Design Standard /

Surface Condition

TotalLength(km)

1 Armenia AH81, AH82, AH83 Class III or Higher / Bad 386

2 Bangladesh AH1, AH2, H41 Below Class III 450

3 Cambodia AH11 Below Class III 198

4 China AH3, AH32, AH42 Below Class III 542

5 Georgia AH81, AH82 Class III or Higher / Bad 55

6 India AH1, AH 2 Below Class III 75

7 IranAH1, AH8, AH70, AH72, AH75, AH78,

AH82 Class III or Higher / Bad 1084

8 Kazakhstan AH7, AH61, AH62, AH63, AH70 Below Class III 897

9 Kyrgyzstan AH7, AH61, AH65 Below Class III 370

10 Lao AH3 , AH11, AH12, AH13, AH15, AH16 Below Class III 656

11 Mongolia AH3, AH4, AH32 Below Class III 3486

12 Nepal AH 42 Below Class III/Bad 34

13 Pakistan AH2, AH4, AH7, AH 51 Below Class III / Bad 3144

14 RussiaAH4, AH6, AH7, AH8, AH30, AH31,

AH60/61/70 Below Class III / Bad 3640

15 Tajikistan AH7, AH65, AH66 Below Class III 343

16 Thailand AH1, AH15, AH16 Class III or Higher / Bad 68

17 Uzbekistan AH63 Below Class III 224

18 Vietnam AH14, AH15 Below Class III 190

Total 15842

Road surface condition and design standards in the following countries like Japan, South

Korea, Singapore, Malaysia, and Turkey are in good condition and satisfy Asian Highway

design standards. Data for countries like Democratic People’s Republic of Korea, Turkmeni stan,

Bhutan, Azerbaijan and Indonesia are found partially or not available therefore they were

dropped from analysis.

To estimate the cost of improvements and upgrading in above – mentioned table 2, we used

regression cost model at a regional level of an alysis. The results of predicted cost of these

upgrading and improvement works are displayed in Table 3.


CT 2

Table 3. Predicted cost of upgrading and improvement work

No. CountryPavementWidth (m) Road Upgrade/Improvements Expected Output

Total Length(km)

Total Cost(million US $

2002)1Armenia 6 - 7 Reconstruction Bituminous Improved Condition 138 23.3

7 -14 Reconstruction Bituminous Improved Condition 248 47.42Bangladesh < 4.5 Widening Adding Bituminous 2L and Recon Class II 100 57.5

4.5 – 6 Widening Adding Bituminous 1L and Recon Class II 350 76.03Cambodia 4.5 – 6 Widening Adding Bituminous 1L and Recon Class II 198 112.64China < 4.5 Upgrading Unsealed to Bituminous Class II 67 10.5

4.5 – 6 Upgrading Unsealed to Bituminous Class II 475 87.45Georgia 6 – 7 Reconstruction Bituminous Improved Condition 55 7.56India < 4.5 Widening Adding Bituminous 2L and Recon Class II 75 44.17Iran 7 – 14 Reconstruction Bituminous Improved Condition 1,042 199.7

6 – 7 Reconstruction Bituminous Improved Condition 42 7.18Kazakhstan 6 – 7 Upgrading Unsealed to Bituminous Class II 743 153.4

< 4.5 New Construction 2L Highway Class II 154 147.29Kyrgyzstan 7 – 14 Upgrading Unsealed to Bituminous Class I 370 91.4

10Lao 7 – 14 Reconstruction Bituminous Condition Improvement 244 42.46 – 7 Reconstruction Bituminous Condition Improvement 44 6.86 – 7 Upgrading Unsealed to Bituminous Class II 292 55.86 – 7 New Construction 2L Highway Class II 76 65.7

11Mongolia < 4.5 New Construction 2L Highway Class II 3,070 2,431.5< 4.5 Upgrading Unsealed to Bituminous Class II 416 57.1

12Nepal 4.5 – 6 Widening Adding Bituminous 1L and Recon Class II 26 5.26 – 7 Reconstruction Bituminous Condition Improvement 8 1.3

13Pakistan < 4.5 Widening Adding Bituminous 2L and Recon Class II 1,174 736.06 – 7 Reconstruction Bituminous Condition Improvement 1,042 196.57 – 14 Reconstruction Bituminous Condition Improvement 928 198.4

14Russia 7 -14 Upgrading Unsealed to Bituminous Class II 882 188.46 - 7 New Construction 2L Highway Class II 89 77.6< 4.5 New Construction 2L Highway Class II 876 764.37-14 Reconstruction Bituminous Condition Improvement 1,793 307.8

15Tajikistan < 4.5 New Construction 2L Highway Class II 48 46.26 - 7 Upgrading Unsealed to Bituminous Class II 278 57.87 -14 Upgrading Unsealed to Bituminous Class II 17 4.0

16Thailand > 14 Reconstruction Concrete Condition Improvement 40 7.66 -7 Reconstruction Bituminous Condition Improvement 18 2.3> 14 Reconstruction Bituminous Condition Improvement 10 1.7

17Uzbekistan 7 -14 Upgrading Unsealed to Bituminous Condition Improvement 224 56.518Vietnam 4.5 - 6 Widening Adding Bituminous 1L and Recon Class II 53 9.6

< 4.5 Widening Adding Bituminous 2L and Recon Class II 137 65.7Total 15,842 6,451.3

Total cost of upgrading and improvement works for 15,842 km in 18 countries would cost

about 6.4 billion US dollars. Figure 4 depicts the highway routes (thick red line) that are belong

to upgrading and road improvements in AHN. These routes play vital role in trade between Asia

and Europe.


CT 2

3.2 Intra – regional trade

Although 18 continental countries have enormously increased their overall trade over thepast several decades, intra-regional trade was still only 12 per cent of their total trade in 2005.


CT 2

Figure 5 shows the ratio of intraregional trade 1 to overall trade for countries in our sampl e. Theshare varies wildly, between 7.7 percent for China to 77 percent for Kyrgyzstan. There are twofacts worth noting from the graph. First, the share of intra -regional trade for small countries ismuch higher than for big countries. This is consistent with overall tendency for big countries tobe less open than small countries. Second, Former Soviet Union countries have much higherintra-regional trade on average. This can be attributed to the legacy of Soviet Union with itsclose trade and production networks among countries.

Figure 5. Share intra-regional trade in total trade, % (2005)

There are numerous reasons for relatively low levels of intra -regional trade in continentalAsia. Political and historical tensions certai nly have played a role, as well as the attractivenessof North American and European markets as a destination for exports products and source oftechnological imports.

But unfavorable geographical factors and low quality of transport infrastructure have al soimpaired intra-regional trade to a great extent. Vast and difficult terrain, especially in the innerpart of the Asian continent, has made overland trade among continental countries much lessprofitable. Due to the lack of adequate level of transport in frastructure, shipping goods from onecountry to another in the region through overland transport networks might be more expensivethan shipping from the region to North America and Europe through sea transport.

3.3. The gravity model approach

This paper adopts a gravity model approach to study the impact of AHN road upgrade ontrade. The gravity model originally stems from Newtonian physics, which simply states that theattraction between two physical objects is proportional to their masses, but inversely related to

1 Intra-regional trade is defined as trade among 18 countries in our sample.

0

10

20

30

40

50

60

70

80

Arm

enia

Ban

glad

esh

Cam

bodi

a

Chi

na

Geo

rgia

Indi

a

Iran

Kaz

akhs

tan

Kyr

gyzs

tan

Lao

Mon

goli

a

Nep

al

Paki

stan

Rus

sia

Taj

ikis

tan

Tha

ilan

d

Uzb

ekis

tan

Vie

tnam


CT 2

the distance between them. This paradigm has long been disregarded by economists due to itslack of theoretical foundation. However, due to successive works of various economists, it hasbeen gradually developed into a systematic economic model with a strong economic foundation.Anderson (1979) derived the gravity equation from monopolistic competition setting. Helpmanand Krugman (1985) showed that the basic gravity equation could be derived from thedifferentiated products trade. It is a theory that suggests that flows of goods depend on thedemand in the importing country and the supply of differentiated products from the exportingcountry. Deardorff (1995) showed that the gravity model is also consistent with Hecksher -Ohlininternational trade theory.

The gravity model has also been extensively utilized in empirical economic literature. Thusit was applied in estimation of bilateral trade flows, FDI flows, and equity flows. For example,Frankel (1997) used the gravity model approach to explain the factors affecting the formation oftrade blocs. In his standard gravity model, bilateral trade was explained by variables such asGNP, per capita GNP, distance, adjacency, language, and trading blocs. Gravity models havealso been used to explain determinants of FDI and equity flows. Kawai and Urata (1998) used agravity model to investigate the relationship between trade and FDI using Japanese data at theindustry level. FDI and trade were found to be generally complementary to each other. Portesand Rey (2000) also adopted a gravity model to study factors affecting equity flows among 14developed economies. Their empirical results demonstrated that market size, openness,efficiency of transactions, and distance are the most important determinants of b ilateral equityflows.

In recent years gravity model has increasingly been utilized in analyzing the impact ofinfrastructure on trade. Majority of studies show that transportation infrastructure quality hassignificant and robust impact on overall transpo rt costs. Notable examples include Redding andVenables (2004), Limao and Venables (2001), Coulibaly and Fontagné (2004), Martínez -Zarzoso and Nowak-Lehmann (2006), Buys, Deichmann and Wheeler (2006), Shepherd andWilson (2006) and others.

In particular, Redding and Venables (2001) use a ratio of roads to area as a proxy forquality of infrastructure and find that low infrastructure quality is the main factor behind lowtrade in Sub-Saharan Africa. Coulibaly and Fontagné (2004) study determinants of trade incountries belonging to the West African Economic and Monetary Union and find that paving allinter-state roads would increase trade by a factor of 3, and crossing a transit country reducesbilateral trade flows by 6%. Buys, Deichmann and Wheeler (2006) first estimate the costs ofinitial upgrading Sub-Saharan interstate road network as 20 billion dollars and 1 billion dollarsas cost of annual maintenance. Then they proceed to estimate the potential beneficial impact ofcontinental road network upgrading on overland trade as about $250 billion over 15 years.Limão and Venables (2001) estimate that poor infrastructure account for 40 percent of transportcosts for coastal countries and 60 per cent for landlocked countries.

In all of these studies the gravity model framework serves as a workhorse to estimate


CT 2

impact of infrastructure upgrading on trade. The attractiveness of the gravity model is that itallows us to address the specific questions in mind with regard to both the fundamentaleconomic and institutional determinants of trade in continental Asia. First, what are thefundamental determinants of trade? Are traditional variables of gravity models such aseconomic size, distance, tariff and common border significant explanatory variables? Second,does infrastructure quality matter in facilitating trade between among Asian countries? Thispaper offers some quantitative simulations to illustrate how the improvement of transportinfrastructure and reduction of tariff barriers can stimulate trade and econo mic development

3.4. Econometric specification and data description

Following the empirical literature, we specify a simple version of gravity model for totaltrade, exports and imports. In each specification GDP variable enters the trade regression in aproduct form. As a result, the gravity model for total trade takes the following form:

ijtj7i6j5i4ij3ij2jtit1ijt RRTarTarB)D(Ln)YY(LnLnT (2)

where Tijt indicates trade between country i and country j at time t, Y it and Yjt are realGDPs of country i and j, representing economic ma ss, Dij is distance between capital cities and

B is a common border dummy, ji TarandTar are tariff rates in country i and j, respectively.

Finally, ji RandR represent road quality indexes in country i and j, respectively.

Vast trade literature predicts expected signs and sometimes magnitudes of coefficients inequation (2). In particular, theory predicts that larger economic mass is associated with highervolumes of trade. Distance, as a proxy for transportation cost, is expected t o have a negativesign. Common border dummy is expected to have a positive sign. Trade is expected to have anegative relationship with tariff barriers and a positive relationship with road quality index.

Distance, calculated as a surface distance between capital cities according to latitude andlongitude (Wall, 1999; Raballand, 2003; Rose and Wincoop, 2001), is considered as proxy fortransportation cost in a borderless world. Border effect is expressed by inclusion of commonborder dummy (Rose and Wincoop, 2001; Rose, 2002; Breuss and Egger, 1999; Frankel andRose, 2001). The problem with simple great circle distance variable is that it does not fullycapture high transportation costs due to natural geographical location of landlocked and remotecountries. Transportation costs are usually affected by border delays (type of a non -tariffbarrier). To capture this peculiar feature of transportation cost we also include common borderdummy variable. Following Raballand (2003) we assumed for two coastal countri es there is aone border, and only for landlocked countries this variable is equal to one.

But even with distance and common border dummy variables, one cannot be sure that shetakes into account all complexities of transportation costs. One of the most i mportant factors foroverall transportation costs is the quality of transport infrastructure. Usually, the higher thequality of infrastructure the lower is the transportation costs and higher incentives for trade.Bougheas et al (1999) utilized stock of p ublic capital and length of motorway network and


CT 2

predicted a positive relationship between the level of infrastructure and the volume of trade.Limao and Venables (2001) developed unique infrastructure composite index, as a totalinfrastructure stock (roads, paved roads, telephones and railway networks) divided by the totalpopulation. But they excluded all transition economies in FSU and Europe due to missing datafor own and transit infrastructure. Unfortunately, lack of data for Asian countries madecalculation of composite index impossible. Instead, we utilized road quality index as anadditional.

The model is estimated for 19 Asian countries over the period 1995 -2004. Aggregatebilateral trade data are from International Monetary Fund’s Direction of Tr ade Statistics(DOTS) database. Data on GDP are taken from World bank’s World Development Indicators(WDI) database. Weighted average tariff rates are taken from Trade Analysis and InformationSystem (TRAINS) database, maintained by The United Nations Conf erence on Trade andDevelopment (UNCTAD). Except dummy variables, all variables are in logarithmical form.

3.5 Results and discussions

Table 4 shows estimation results. We consecutively estimate equations for trade, importsand exports. In the trade regression we follow Baldwin and Taglioni (2006) and use the productof real GDP.

Table 4. Gravity model estimations

Dependent variable Trade Exports Imports

Product of GDP 1.06 1.05 1.02[0.02]** [0.02]** [0.02]**

Distance -1.63 -1.66 -1.42[0.07]** [0.07]** [0.08]**

Common border dummy 1.40 1.36 1.62[0.12]** [0.14]** [0.13]**

Road quality index of country i -0.06 0.16 -0.16[0.08] [0.07]* [0.09]

Road quality index of country j 0.44 -0.06 0.79[0.08]** [0.09] [0.09]**

Average tariff rate of country i -0.23 -0.04 -0.37[0.05]** [0.05] [0.05]**

Average tariff rate of country j -0.20 -0.31 -0.15[0.04]** [0.04]** [0.05]**

Constant -35.64 -34.85 -37.04[0.77]** [0.78]** [0.90]**

Observations 2069 1920 1917R-squared 0.74 0.72 0.67Robust standard errors in brackets

* significant at 5%; ** significant at 1%


CT 2

The gravity model fits the data well and produces theoretically correct and economically

significant coefficients. As expected, the coefficients of real GDP are statistically significant,

slightly above 1 in all specifications. The coefficient of distance is negative, highly significant.

In elasticity terms it shows that 1 percent increase in distance is associated w ith a -1.65 percent

decline in trade and exports, and around 1.42 percent decrease in imports. Common border

dummy is also highly significant and positive. The point estimates of common border dummy

indicates that if countries share common border, they tra de with other 4 times more

(exp(1.4)=4.05). This effect is even stronger for imports. The coefficient of common border

dummy indicates that there is a lot of potential to increase overland trade, especially between

countries with common borders.

The coefficients of tariff rates are negative and statistically significant, and their magnitude

ranges from -0.15 to -0.37. Trade equation indicates that, say, 10 percent reduction in tariff rates

increase overall trade by about 2 percent. Taking into account tha t in most instances tariffs are

already quite low and they cannot be drastically decreased, it becomes clear that further

reductions of tariff rates among continental Asian countries have limited impact on trade.

On the other hand, the road quality index shows that a good transport infrastructure can

greatly facilitate trade. In particular, the positive coefficient of road quality index in trade

regression – 0.44, which is statistically highly significant, indicates that improvement of the

quality of overland roads can boost trade significantly. For example, if Nepal improves quality

of its roads index from 31 to 50 (48 percent improvement in logarithmic terms), it can expect its

overall trade increase by 21 percent (0.48 multiplied by 0.44), or by 285 mill ion US dollars

annually.

Based on the gravity model estimations, we can estimate the impact of road quality

upgrading on intra-regional trade. Table 5 shows the results of this exercise under two

scenarios: pessimistic and optimistic. Under pessimistic sc enario it is assumed that major road

improvement efforts will upgrade road quality index to 50 percent throughout the region. It is

pessimistic scenario because it assumes that major routes surface conditions will be improved

without any upgrading. Armenia , Georgia, Iran, and Thailand already have road quality grade of

50, so it is assumed that they will not benefit directly in terms of trade expansion due to road

improvement.

Scenario 2 assumes more ambitious criterion, namely, continental Asian count ries upgrade

their interstate road quality to 75. In terms of AHN classification, Table 5 shows the net impact

of road improvement on trade under two scenarios. Under Scenario 1 the total intra -regional

trade will increase about 20 percent to 48.7 bln US d ollars annually, while Scenario 2 predicts

that trade will increase by about 35 percent to 89.5 billion US dollars annually. The main

beneficiaries of the road improvement will be China, Russia, India, and Vietnam smaller

countries will also benefit from overall increase in trade due to the improvement in transport

infrastructure.


CT 2

Table 5. Net effect of road upgrading on trade

IV. CONCLUSION

Introduction level of analysis shed light on issue like efficient data usage in existingdatabase. It is important to mention that missing data is almost unavoidable part of datacollection process. So when some of the data times are missing then the common approach is touse available data. We recommended to separate data items in ROCKS in accordance to level ofanalysis such as regional, country and project level to utilize available data efficiently. Inregional level, the purpose was to select plausible data items with the largest number ofobservations which later were used for cost model development. In case of project level, the aimwas to select data items which represent project details in depth therefore number ofobservations were very limited. Country level analysis lies in between regional and project levelof analyses and contains some of regional level data items as well as project level data itemswhich lead to utilize more data with some project details in cost model development.

Regression cost models are widely used because they are easy and relatively fast toimplement, various well-documented procedures are available, and finally cost estimators preferto use regression models rather than analytical tools such neural networks because regressionmodels are well defined and mathematically explained whereas neural network works muchmore like black box.

Scenario 1(up to 50%)

Scenario 2 (upto 75%)

Armenia 50.0 530.4 - 94.6Bangladesh 25.0 4139.53 1,262.5 2,001.0Cambodia 25.0 819.74 250.0 396.3China 25.0 81683.26 24,912.1 39,484.8Georgia 50.0 881.01 - 157.2India 25.0 20066.49 6,120.0 9,699.9Iran 50.0 15022.54 - 2,680.1Kazakhstan 25.0 14964.24 4,563.9 7,233.6Kyrgyzstan 25.0 1436.8 438.2 694.5Lao 36.0 1022.67 148.1 330.6Mongolia 25.0 1042.75 318.0 504.1Nepal 30.9 1343.58 284.9 524.6Pakistan 40.7 4103.05 373.1 1,105.1Russia 37.3 41246.66 5,311.0 12,669.6Tajikistan 25.0 1021.1 311.4 493.6Thailand 50.0 24352.18 - 4,344.5Uzbekistan 25.0 3147.66 960.0 1,521.5Vietnam 25.0 11461.15 3,495.5 5,540.2

Total 228284.8 48,748.6 89,475.7

Net Effect of RoadUpgrading, mln US dollarsInter-regional

trade in 2004,mln US dollars

Road QualityIndexCountry


CT 2

It was estimated cost of upgrading and improvement works costs of sub -network of AHNusing World Bank’s ROCKS database. This provided initial perspec tive on the size of lumpyinvestment required to improve road condition in AHN. It was estimated that approximately 6.5billion US dollars is required to upgrade roads and improve existing surface condition of theselected sub-network with total length of 15,842 km of AHN. The gravity model approachexplained how big the trade expansion will increase. The net impact of road improvement ontrade under two scenarios was considered. In scenario 1, the total intra -regional trade willincrease about 20 percent to 48.7 bln US dollars annually, while Scenario 2 predicts that tradewill increase by about 35 percent to 89.5 billion US dollars annually. The main beneficiaries ofthe road improvement will be China, Russia, India, and Vietnam smaller countries will als obenefit from overall increase in trade due to the improvement in transport infrastructure. Priorityof road upgrading in each country is suggested to be carried out in a way that first roadcondition improvements need to be done after that upgrading to h igher class necessary to carryout. But it must also fit to each country’s network strategic plan. The results show that roadquality is positively associated with trade, while tariff rates are negatively correlated with trade.These results are consistent with transportation economics viewpoint that road upgradingdecreases transportation costs such as vehicle operation cost (fuel consumption, spare, etc) anduser cost (travel time). Higher traffic volumes allow the policymakers to take advantage thehigher trade volumes and decrease tariff rates further. Future research will focus on tradeexpansion among all AHN member countries taking into account other modes of transportation.

Reference

[1]. Adeli H. and Wu M. Regularization neural network for constru ction cost estimation, Journal ofConstruction Engineering and Management, ASCE 124(1), pp 18 -24, 1998

[2]. Allison S., Allison D. Multiple Regression: A Primer. Forge Press. 1999

[3]. Al-Tabtabai, H., Alex, A., and Tantash, M. Preliminary Cost Estimation of Highway ConstructionUsing Neural Networks, Cost Engineering, Association for the Advancement of Cost Engineering,Morgantown, WV. 41(3), 19-24, 1999

[4]. Archondo-Callao R., Bhandari A. Road Costs Knowledge System – ROCKS – Version 2.2, UserManual, World Bank, 2004

[5]. Anderson, J.E.A. 1979. A theoretical foundation for the gravity model. American Economic

Review, Vol. 69, pp. 106–16.

[6]. Andrew K. Rose & Eric van Wincoop , 2001. "National Money as a Barrier to International Trade: TheReal Case for Currency Union," American Economic Review, vol. 91(2), pages 386 -390, May.

[7]. Asian Highway Network Handbook, United Nation, New York, 2003.

[8]. Asian Highway Design Standard, UNESCAP, 2004

http://www.unescap.org/ttdw/common/tis/ah/AnnexII -E.pdf. Accessed: June 10, 2008


CT 2

[9]. Asian Highway Network Database (AHND), UNESCAP, 2004http://www.unescap.org/ttdw/common/tis/ah/Member%20countries.asp Accessed: June 10, 2008

[10]. Bunke, O and B. Droge . Bootstrap and cross-validation estimates of prediction error for linearregression models. Ann. Stat. 12, 1400 -1424, 1984

[11]. Baldwin, Richard, and Daria T aglioni (2006). "Gravity for Dummies and Dummies for GravityEquations," NBER Working Papers 12516

[12]. Bougheas, Spiros, Panicos O. Demetriades and Edgar L.W. Morgenroth (1999), “Infrastructure,Transport Costs and Trade”, Journal of International Econom ics, Vol. 47, 169-189.

[13]. Breuss, Fritz, and Peter Egger (1999), "How Reliable Are Estimations of East -West TradePotentials Based on Cross-Section Gravity Analyses?", Journal Empirica 26 (2), 81 -94

[14]. Brikes D, Birkes D, Dodge Y . Alternative Methods of Regression. Wiley-Interscience. 1993

[15]. Buys, Piet, Uwe Deichmann and David Wheeler (2006), Road Network Upgrading and OverlandTrade Expansion in Sub-Saharan Africa, World Bank Policy Research Working Paper 4097, December

[16]. Carroll J, Raymond J, Ruppert D. Transformation and Weighting in Regression. CRC Press. 1988.

[17]. Coulibaly and Fontagné (2005), South-South Trade: Geography Matters, Journal of AfricanEconomies, Vol. 15, Number 2, pp. 313 -341

[18]. Deardorff, A.V. (1998). Determinants of Bilateral Trade: Does Gravity Work in a NeoclassicalWorld?, in: J. Frankel (ed.), The Regionalization of the World Economy. University of Chicago Press.

[19]. Emsley M, Lowe D., Duff A, Harding A, Hickson A . Data Modeling and the application of a neu ralnetwork approach to the prediction of total construction cost. Journal of Construction ManagementEconomy, 20, pp.465-72, 2002

[20]. Frank H. Regression Modeling Strategies: with Applications to Linear Models, Logistic Regressionand Survival Analysis. Springer. 2001

[21]. Frankel, Jeffrey A. (1997). Regional Trading Blocs in the World Economic System, Institute forInternational Economics, Washington, D.C.

[22]. Frankel, Jeffrey, and Andrew Rose (2000). "An Estimate of the Effect of Currency Union s on Tradeand Growth," for a Conference on Currency Unions, organized by Alberto Alesina and Robert Barro,Hoover Institution, Stanford University, May 19 -20.

[23]. Flyvbjerg B. H., Mette S., and Buhl S . Underestimating costs in public works projects, er ror or lie,Journal of the American Planning Association, Vol. 68, No.3, pp 279 -292, 2002

[24]. Geisser, S. The predictive sample reuse method with applications. Journal of American StatisticsAssociation, 70, 320-328. 1975

[25]. Gwang-Hee K, Sang-Hoon A, Kyung-In K. Comparison of construction cost estimating modelsbased on regression analysis, neural networks, and case -based reasoning. Journal Building andEnvironment. Vol. 39, pp 1235-1242, 2004

[26]. Healey, J. M., “Errors in Project Cost Estimates,” Indian Economic Journal, Vol. 12, July- Sept. 1964.

[27]. Hegazy T. and Ayed A . Neural network model for parametric cost estimation of highway projects,Journal of Construction Engineering and Management, ASCE 124(3), pp 210 -218, 1998

http://www.unescap.org/ttdw/common/tis/ah/AnnexII-E.pdf

http://www.unescap.org/ttdw/common/tis/ah/Member%20countries.asp


CT 2

[28]. Helpman, E. and Krugman, P.R. (1985). Market structure and foreign trade. Increasing returns,imperfect competition, and the international economy, MIT Press, Cambridge, MA

[29]. Kawai, Masahiro and Shujiro Urata (1998), “Are Trade and Investment Substitutes orComplements? An Empirical Analysis of Japanese manufacturing Industries,” in Lee, Hiro, Roland andW. David, Economic Development and Cooperation in the Pacific Basin: Trade, Investment, andEnvironmental Issues, Cambridge; New York and Melbourne: Cambridge Univers ity Press, 1998.

[30]. Levinson, D. and Karamalaputi , Ramachandra. Predicting the Construction of New Highway Links.Journal of Transportation and Statistics Vol. 6(2/3) 81 –89, 2003

[31]. Lewis-Beck M. Applied Regression: An Introduction. Sage Publication Inc. 1980.

[32]. Miles J., Shevlin M. Applying Regression and Correlation: A Guide for Students and Researchers.Sage Publication Inc. 2001.

[33]. Limão, N. and Venables, A.J . (2001), "Infrastructure, Geographical Disadvantage, Transport Costsand Trade", World Bank Economic Review, 15, 451 -479.

[34]. Martínez-Zarzoso and Nowak-Lehmann (2006), Is Distance a Good Proxy for Transport Costs? TheCase of Competing Transport Modes, CeGE -Discussion Paper 54, May

[35]. Pearce A. R., Gregory, Rita A., and Willia ms L. Range Estimating for risk management usingartificial neural networks, Journal of Parametric Estimating, 14(1), pp 88 -93. 1996

[36]. Portes, Richard and Hélène Rey (2000). "The Determinants of Cross -Border Equity Flows: TheGeography of Information," Center for International and Development Economics Research (CIDER)Working Papers C00-111, University of California at Berkeley

[37]. Raballand, G. (2003). The determinants of the negative impact of land -lockedness on trade: anempirical investigation through the Central Asian case. ROSES, University of Paris 1, mimeo May 2003.

[38]. Redding, S., Venables, A. J . 2004. Economic geography and international inequality. Journal ofInternational Economics, 62(1): 53 -82.

[39]. ROCKS, Road Costs Knowledge S ystem (v. 2.2), (The World Bank, Transport Unit – TUDTR),2004. http://www.worldbank.org/transport/roads/rd_tools/rocks_main.htm . Accessed June 10, 2008

[40]. Rose, A.K. 2002. Do WTO members have more liberal trade policy? NBER Working Paper #9273

[41]. Sanders, S., and Maxwell, R., “Preliminary Estimating for Heavy Construction,” AACE Transactions,Association for the Advancement of Cost Engineering International, Morgantown , WV, 1992.

[42]. Stone, M. Cross-validatory choice and assessment of statistical predictions. Journal of the RoyalStatistical Society. Series B (Methodological), Vol. 36, No.2. pp. 111 -147. 1974

[43]. Shepherd, Ben, and John S. Wilson (2006), Road Infrastructure in Europe and Central Asia: DoesNetwork Quality Affect Trade? World Bank Policy Research Working Paper 4104, December 2006

[44]. Trost, S. , and Oberlender, G. Predicting accuracy of early cost estimates using factor analysis and multivariateregression. Journal of Construction Engineering Management, Vol. 129, No. 2, 2003, pp. 198 -204.

[45]. William D., Stanley F . Multiple Regression in Practice. Sage Publications Inc. 1985

[46]. Wall, Howard J. (1999). "Using the gravity model to estimate the costs of

protection,"Review, Federal Reserve Bank of St. Louis, issue Jan, pages 33 -40

http://www.worldbank.org/transport/roads/rd_tools/rocks_main.htm


CT 2

I. INTRODUCTION OF GLEM

1.1. Outline of Generalized LimitEquilibrium Method for Static Problems

Developing the limit equilibrium method(LEM), the authors have proposed thegeneralized limit equilibrium method, whichcan obtain a force field satisfying theequilibrium condition on every soil block andfailure condition both on the bottom plane andthe inter-block plane of the block (Figure. 1).Consequently, the GLEM can be consideredas an approximation method to obtain thenecessary condition of SLM. This method hasthe following features:

Quadrangle or triangle blocks as well asslices can be treated.

Safety factors are defined both on themain sliding surface and on inter -blockplanes.

Circular sliding surface as well as non -circular sliding surface can be treated.

All types of plasticity problems can beexpressed in a single formulation.

The GLEM can be applied to analyze all

types of static problems such as slope

stability, earth pressure, and bearing capacity.

For every static plastic problem, a num ber of

DEVELOPMENT OF GENERALIZED LIMIT EQUILIBRIUMMETHOD FOR THE FAILURE OF RETAINING WALLS

UNDER SEISMIC LOADINGS

XUAN BINH LUONG; VIKHONE SAYNHAVONGTHANH THUY HOANGDepartment of Civil EngineeringUniversity of Transport and Communications , VietnamMEIKETSU ENOKIDepartment of Civil EngineeringTottori University, Japan

Abstract: The classical theory of plasticity, represented by K ötter’s equation, has been

established for static problems [1]. Although this theory can be easily extended for dynamic

plasticity problems by introducing accelerations as inertia forces, up to now no researcher

has done this because of the difficulty in determining the acceleration distribution within the

body when the failure occurs. The objective of this research is to develop Generalized Limit

Equilibrium Method (GLEM) with the introduction of continuity condition of acceleration to

investigate the following cases of potential failures of retaining structures: active failure,

foundation-like failure, and slope-like failure, under seismic loadings, where the GLEM is one

of the limit equilibrium methods proposed by Enoki at al . The theoretical formulation of the

method, the illustrative examples, and the comparisons between the results of the proposed

method and other methods are demonstrated.

Keywords: Earth pressure, earthquake, limit equilibrium method, slope failure, rigid -plastic


CT 2

examples of the calculation were taken with

the use of GLEM and many other methods.

The comparisons showed that the results

obtained by GLEM agree with those obtained

by theoretical analysis. Another paper by the

authors (Enoki et al., 1991) [2] sho uld be

referred to for the details of the GLEM.

1.2. Outline of Generalized Limit EquilibriumMethod for Dynamic Problems

Newmark [3] proposed, for the first time,a displacement analysis method to evaluatethe effects of earthquakes on the stability ofslopes. The method was then developed byChang [4] and other researchers [5]. InNewmark’s method, the critical state in whichthe failure begins to occur is determined bythe pseudo-static analysis. When theearthquake-induced acceleration exceeds thecritical value, the failed soil mass isconsidered to slide along the slip surface as awhole rigid body. The residual displacementcan be determined by integrating the relativeacceleration.

Based on Newmark’s concept, theauthors have been developing a metho d toanalyze the motion of earth structures. In thismethod, the computing model of a foundationis shown in figure 1. When the earthquake-induced acceleration reaches a certain criticalvalue, incipient failure occurs and many slipplanes appear within the body. This criticalacceleration can be obtained by ordinarypseudo-static analysis in which theacceleration of every part of the structure isthe same with the input acceleration. Whenthe seismic acceleration exceeds the criticalvalue, the failed soil mass is then consideredas a rigid-plastic block system, in which thesurrounding surfaces of the block are just the

slip planes. The rigid blocks will moverelative to each other and to the base groundalong the slip planes. Across a slip plane(figure 2) the component of accelerationnormal to the slip plane is continuous, thecomponent parallel to the slip plane isdiscontinuous but the shear stress on the slipplane corresponding to the shear strength istransmitted. This continuity condition ofacceleration is combined with “GeneralizedLimit Equilibrium Method (GLEM) ” toanalyze the motion of the earth structures indynamic cases. The residual displacements ofevery block in both vertical and horizontaldirections can be computed by integratingtwice the relative acceleration of the block.

The proposed method permits theanalysis of all types of dynamic plasticproblems such as: bearing capacity andmotion of foundations, failure and motion ofslopes, earth pressure and motion of retainingwalls. Any types of sliding surface can betreated by dynamic GLEM.

Figure 1. A block system of a foundation

in earthquake motion and equilibrium

of the i - th block


the i-thblock

i -Block numberi -Bottom plane numberi - Inter-block plane number

Foundation

12

i

n

12

i

i+1

n

n+1

1 2

i

n

n+1

v

h

vi’ hi’

v(n+1)’h(n+1)’

Bottom plane

Inter-block plane

B

i+1

Hi ViPiTi

i

i

i+1

i+1

Mig

Mihi’

Mivi’Vi+1

Si

Ri+1

Ri



the i-thblock

i -Block numberi -Bottom plane numberii - Inter-block plane number

Foundation

12

i

n

12

i

i+1

n

n+1

1 2

i

n

n+1

v

h

vi’ hi’

v(n+1)’h(n+1)’

Bottom plane

Inter-block plane

B

i+1

Hi ViPiTi

i

i

i+1

i+1

Mig

Mihi’

Mivi’Vi+1

Si

Ri+1

Ri


i+1

Hi ViPiTi

i

i

i+1

i+1

Mig

Mihi’

Mivi’Vi+1

Si

Ri+1

Ri



CT 2


II. FAILURE MECHANISMS OF RETAININGWALLS UNDER SEISMIC LOADINGS

Figures 3, 4 and 5 show the computingmodels of a retaining wall subjected toseismic loadings corresponding to three casesof failures: active failure (a), failure ofretaining wall as a foundation problem(foundation-like failure) (b), and failure of thewall as a slope problem (slope-like failure)(c). In mode (a), the wall is considered tomove outwards relative to sub -base, andcauses the active earth pressure. In mode (b),the base supporting the wall is fail ed and bothwall and sub-base slide outwards. In mode (c),both the sub-base and backfill are failed andthe system slides outwards. The failed soilmass is considered as a rigid-plastic blocksystem. Either triangular or quadrangularblocks can be used.

III. FORMULATION OF DYNAMIC GLEM FORTHE FAILURES OF RETAINING WALLS

Before the sliding occurs, theacceleration of every soil block is the same asthe acceleration of the sub-base. Theequilibrium equations of every block, thefailure conditions on the inter-block planesand bottom planes are used to obtain the forcefield as presented in the formulation ofdynamic GLEM, for the detail the [6] shouldbe refered to.

When the sliding occurs, the accelerations of

blocks are different from each other and from thesub-base. The equilibrium equations of everyblock, the failure conditions, and continuityconditions of acceleration on both inter -blockplanes and bottom planes are used. The numberof unknowns and the number of equations areshown in Table 1. The sliding acceleration ofthe wall is minimized to obtain the geometry ofthe sliding surface. The classical Newtonmethod is used herein to optimize thefunction value.


Figure 4. Foundation – like failure (b)

Figure 5. Slope – like failure (c)

1

H

TN

Bedrock

Sliding surface


n’’

’

n

Mg

Mv’Mh’

1

H

TN

Bedrock

Sliding surface


n’’

’

n

Mg

Mv’Mh’

1

H

TN

Bedrock

Sliding surface


n’’

’

n

Mg

Mv’Mh’

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’


vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’v

h

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1



vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’


vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’v

h

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1



vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’


vh

21

i

n

1 2

i

n

12

i

n

vi’ hi’v

h

21

i

n

1 2

i

n

12

i

n

vi’ hi’




vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1



vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


vh

21

i

n

1i1 i n

vn’

hn’

n+2

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1


v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1

v

h2

1

i

n

1i1 i n

vn’

hn’

n+2

n+2

n+3

n+1

n

n+1

n+1


CT 2

IV. NUMERICAL EXAMPLE

A strip retaining wall with a mass of 25t,a height of 5m, and a width of 3m isconsidered. The frictional angle of backsurface of the wall is 11o. The backfill and soilbase have the parameters as = 32o, c =0.11tf/m2, = 1.6tf/m3. A sinusoidal wave isused as the input acceleration, which has thefrequency of 2Hz, the amplitude of horizontalcomponent is 6m/s2, the vertical componentequals zero. For the simplicity, the dilatancyangle is considered to be zero, and the surfaceof the backfill is horizontal. Four cases ofanalysis were carried out: case a1 - activefailure mode was taken with the frictionalangle of the bottom plane of the wall, , is17o; case a2 - active failure mode was takenwith the frictional angle of the bottom planeof the wall is 32o; case b - foundation-likefailure mode was taken; and case c - slope-like failure mode was taken.

The results of analyses are presented infigures 6 and 7. It can be seen from the figure6 that the solutions of the active failure modeare very differentwith the changeof . When =17o, the activefailure occursearlier thanfoundation-likefailure mode (caseb) and slope-likefailure (case c),and it is inopposite situationfor the case =32o. The graphalso indicates thatthe foundation-likefailure occurs laterthan slope-like

failure in this analysis. As stated in [6], the figure 7once again shows the comparison between theproposed method and Mononobe-Okabe method(M-O method) [7,8] for the dynamic earthpressures. It is clear to realize that, correspondingto the sliding process, the M-O method hasoverestimated the earth pressure.

V. EFFECT OF ROUGHNESS OF WALL-BOTTOM SURFACE

An investigation on the relation between thefrictional angle of the wall-bottom surface and thecritical acceleration, at which the sliding starts tooccur, was carried out. The analysis condition isthe same as the example above. The interrelation

between and the critical accelerationscorresponding to every failure mode is presented

in figure 8. In this analysis, when 25.47o, thefailure mode likely to happen is active failure.

When > 25.47o, the failure mode likely tohappen is slope-like failure. The wall seems to besafe with foundation-like failure.

Table 1. The number of equations and unknowns

Equations Unknowns

Equilibriumconditions

(a) (b) (c)On bottomplanes

(a) (b) (c)

In verticaldirection

n n+1 n+2 Normal forces n n n+1

In horizontaldirection

n n+1 n+2 Shear forces n n n+1

Failureconditions

On inter-blockplanes

On bottomplanes

n n n+1 Normal forces n-1 n+1 n+1


n-1 n+1 n+1 Shear forces n-1 n+1 n+1

ContinuityCondition ofAcceleration

Blockaccelerations

On bottomplanes

n n n+1 vi’ n n+1 n+2


n-1 n+1 n+1 hi’ n n+1 n+2

Total 6n-2 6n+4 6n+8 6n-2 6n+4 6n+8


CT 2

VI. EXAMPLE OF A SUPPOSED RETAININGWALL SUBJECTED TO THE NIIGATA KENCHUETSU EARTHQUAKE - 2004

A supposed retaining structure with

parameters as shown in Table 2 subjected to

the Niigata Ken Chuetsu Earthquake – 2004 is

considered. The cross-section of the structure

is assumed to oblique at an angle of about 23 o

to the E-W direction. So, the horizontal

component of the input acceleration is derived

from E-W and N-S accelerations. The vertical

component is just the U-D component of the

acceleration record.

It is supposed here that sliding occurs in

slope-like failure mechanism. Thus, the model

and block system for the problem are

presented in figure 9. The peak strength of

soil is applied to the duration from the

beginning of the shaking to the moment that

the first sliding finishes. Then the residual

strength is used for the remainder of the

shaking process. The wall bottom friction

angle is larger than 17o, therefore no relative

sliding occurs along this plane or the wall and

soil wedge right beneath the wall take the

same movement.

The analysis results including

accelerations of the wall and sub -base,

residual displacements of the wall, and

dynamic active earth pressure are presented in

figure 10. In order to see clearer the sliding

process, the acceleration data within the

duration from 19 s to 20 s are zoomed in. The

first critical state is reached at moment t =

19.19 s. The slip surface at this state is

obtained and considered as the actual slip

surface as shown in figure 10. This slip

surface is assumed to be unchanged during

sliding process. With the use of peak strength,

the horizontal critical acceleration for the first

sliding is greater than others. The sliding

occurs a number of times during shaking

process (figure 11a). Over the time of

shaking, the residual displacements of the

wall in horizontal and vertical directions are

computed and plotted in figure 11b.

Figure 11c shows the dynamic activeearth pressure obtained by the present methodand the M-O method. We can see once againhere that during sliding the M-O methodoverestimates the earth pressure. At somemoments, the overestimation of the M -Omethod is up to 130%.

Figure 6. Accelerations corresponding

to analyzed cases


- 6.5

- 4.5

- 2.5

- 0.5

1.5

3.5

5.5

0 0.1 0.2 0.3 0.4 0.5

Time (s)

Horiz

onta

l acc

eler

atio

n (m

/s2 )

- 8

- 5

- 2

1

4

7

0 0.1 0.2 0.3 0.4 0.5

Time (s)

Verti

cal a

ccel

erat

ion

(m/s

2 )Sliding start

Sliding end

case a1

case b

case a2

case c

input acceleration


case a1case a2



case a1

case a2

case b

case c

- 6.5

- 4.5

- 2.5

- 0.5

1.5

3.5

5.5

0 0.1 0.2 0.3 0.4 0.5

Time (s)

Horiz

onta

l acc

eler

atio

n (m

/s2 )

- 8

- 5

- 2

1

4

7

0 0.1 0.2 0.3 0.4 0.5

Time (s)

Verti

cal a

ccel

erat

ion

(m/s

2 )Sliding start

Sliding end

case a1

case b

case a2

case c

input acceleration


case a1

case b

case a2

case c

input acceleration


case a1case a2



case a1case a2



case a1case a2



case a1

case a2

case b

case c


case a1

case a2

case b

case c


case a1

case b

case a2

case c

input acceleration

0

50

100

150

200

250

0 0.1 0.2 0.3 0.4 0.5Time (s)

Act

ive

eart

h pr

essu

re (

kN/m

)


case a1case a2



case a1

case a2

case b

case c


case a1

case b

case a2

case c

input acceleration


case a1

case b

case a2

case c

input acceleration

0

50

100

150

200

250

0 0.1 0.2 0.3 0.4 0.5Time (s)

Act

ive

eart

h pr

essu

re (

kN/m

)


case a1case a2


0

50

100

150

200

250

0 0.1 0.2 0.3 0.4 0.5Time (s)

Act

ive

eart

h pr

essu

re (

kN/m

)


case a1case a2



case a1case a2



case a1

case a2

case b

case c


case a1

case a2

case b

case c


CT 2

-8.0-4.00.04.08.0

17 18 19 20 21 22 23 24 25 26 27

Horiz

ontal

acc.,

m/s

2

-4.0-2.00.02.04.0

17 18 19 20 21 22 23 24 25 26 27

Vertic

al ac

c., m

/2 Sub-base

Wall

1st Critical state

363 339 339

-34 -60 -0.2

No sliding No slidingSliding Sliding Sliding

No slidingNo sliding

Horiz

ontal

acc.,

102

gal

Vertic

al acc.

, 102

gal

Figure 8. Bottom surface roughness of the wall

and failure modes

Table 2. Analysis condition for motion of a retainingstructure using slope-like failure mechanism

Wall mass, M (t) 20.0Back surface friction of wall, (o) 10.0

Bottom surface friction of wall, (o) >17.0

Soil density, (t/m3) 2.0

Internal friction angle, (o) 25.0Cohesion, c (tf/m2) 1.0

Figure 9. Block system for analyzing

motion of a retaining structure with the use of

slope-like failure mechanism

Figure 10. Geometry of block system at the firstcritical state

a. Input accelerations and sliding accelerations of

the retaining wall

0

2

4

6


Criti

cal a

ccel

erat

ion

(m/s

2 )




=25.47o

0

2

4

6


Criti

cal a

ccel

erat

ion

(m/s

2 )




=25.47o

12 3 4

5

6

H=5

m

B=3 m

Wall

Sub -base

4.9596 m

B=3 m

H=5

m

2.6519 m1.0957 m

-0.15

-0.1

-0.05

0

17 18 19 20 21 22 23 24 25 26 27

Time, s

Residu

al disp., m Horizontal disp.

Vertical disp.

Residua

ldisp

., 102

cm

Target = 10.85 cm

Target = 1.72 cm

b. Residual displacements of the retaining wall

Dynami

cPae

,tf

05

101520

17 18 19 20 21 22 23 24 25 26 27Time, s

Present methodM-O method

c. Dynamic earth pressures

Figure 11. Results of analysis for motion of a

retaining structure with the use of slope -like failure

mechanism

Time,s


CT 2

VII. CONCLUSIONS

The dynamic GLEM has been developed

to investigate the failures of a retaining wall

under seismic loadings with three mechanisms

such as active failure, foundation-like failure,

and slope-like failure. Not only the dynamic

earth pressure but the motion of the retaining

wall can be obtained also.

The foundation-like mechanism seems to

be unlikely to happen. When the bottom

surface roughness is low, the failure mode

likely to happen is slope-like failure. When

the roughness is high, the failure mode likely

to happen is active failure.

When sliding has not occurred, the active

earth pressures obtained by the proposed

method and the M-O method are almost the

same. During sliding process, the M-O method

seems to overestimate the active earth pressure

in comparison with the proposed method.

VIII. ACKNOWLEDGMENTS

The seismographic data used in this

research were obtained from the home page of

the Japanese National Research Institute for

Earth Science and Disaster Prevention (NIED)

[http://www.kik.bosai.go.jp].

Reference

[1]. V. V. Sokolovsky, “Static of soil media,”

Trans. Jones, D. H. and Schofield, A. N., Lond on

(Butterworth), 1956.

[2]. M. Enoki, N. Yagi, R. Yatabe, and E.

Ichimoto, “Generalized limit equilibrium method

and its relation to slip line method, ” Soils and

Foundations, Japanese Soc. of Soil Mech. and

Found. Engrg., vol. 31, no. 2, pp. 1–13, June 1991.

[3]. N. M. Newmark, “Effects of earthquakes on

dams and embankments, Fifth Rankine Lecture,”

Gétechnique, no. 2, pp. 139–160, 1965.

[4]. C. J. Chang, W. F. Chen, and J. T. P. Yao ,

“Seismic displacements in slopes by limit

analysis,” J. of Geotech. Engrg, ASCE, vol. 110,

no. 7, pp. 860–874, July 1984.

[5]. R. L. Michalowski and L. You,

“Displacements of reinforced slopes subjected to

seismic loads,” J. Geotech. And Geoenviron.

Engrg., vol. 126, no. 8, pp. 685–694, August 2000.

[6]. M. Enoki, B. X. Luong, N. Okabe and K. Itou,

“Dynamic Theory of Rigid-Plasticity,” J. Soil

Dynamics and Earthquake Engrng ., no. 25, pp.

635–647, 2005.

[7]. N. Monobe and H. Matsuo , “On the

determination of earth pressures during

earthquakes,” in Proc. World Engrg. Conf. , 9, 176,

1929.

[8]. S. Okabe, “General theory on earth pressure

and seismic stability of retaining wall and dam, ” J.

Japanese Soc. of Civ. Engnrs. , vol. 10-6, pp.

1277–1323, 1924

http://www.kik.bosai.go.jp


CT 2

I. OBJECTIVES

Transportation construction projects are

mainly conducted for the social and economicbenefits purpose. However, some of the

projects were not estimated and defined in a

unite way. The article presents a method todefine the benefits. This method can be used

as a reference with respect to the process of

planning and appraising of road projects.

II. CONTENT

The social and economic benefits of aproject are gains that the project brings tosociety and economics. These advantages arenormally estimated by making comparisonbetween a situation of having the project andthat of without the project.

A transportation construction project hasmajor benefits as the following: (1) gains ofreducing operation costs; (2) gains of savingpassenger’s time (and goods); (3) gains ofdecreasing number of accidents; (4) gains ofenvironmental pollution mitigation. Besides,when assessing the efficiency of projects,changing distance of transportation andmaintenance costs are also considered.

Profits of mitigating environmentalcontamination will be presented in section [1],whereas, the content of the article focuses onother types of benefits.

1. A method to define gains of reducingvehicle operation costs

Vehicle Operation Cost (VOC)comprises many costs of fuels and damages(motors, tires…). Of road constructions, thesementioned costs depend on road condition(geometrical structure, road surface…);activities of divers and traffic controlcapacity. VOC usually shows higher valuewith respect to sloping and rough roads. Amethod to determine VOC is presented insections [2; 3].

One of the most ultimate target whenconstructing road structures is to reduce thevalue of VOC. Benefits, contained bydecreasing vehicle operation costs accountedat t-th year, are calculated as the equationbelow.

)inewVOC

iold(VOCnew.L

itN

m

1i365.

1tB

(VND/year) (1)

A METHOD OF DETERMINING SOCIAL AND ECONOMICBENEFITS OF TRANSPORTATION CONSTRUCTION PROJECTS

Dr. BUI NGOC TOANUniversity of Transport and Communications

Summary: This paper presents an approach of determining basic social and economic

benefits of road projects


CT 2

Where, itN - annual average daily traffic

of the i-th type of vehicle in t -th year

(units/day)

m - number of the vehicle unit (inluding

good and passengers transportations) (units)

Lnew - transport length of a new orreconstructe road (km)

inew

iold VOC;VOC - vehilce operation

costs in two cases (having the project andwithout the project) (VND/vehicle.km).

2. A method to define gains ofdecreasing passengers travelling time

Time has its value and the human time

value can be measured. Transportation

construction projects aim to reduce time for

travelling of passengers. Many studiesdemonstrates that saved time valu e of

passengers largely depends on individuals

(purpose of the trip, attitude...). In a simplerand more precise view, that is, defining time

value of passengers bases on GDP of the

section surveyed and types of vehicle .

2.1. A case of having data of traff ic

volume

With respect to section surveyed, if thereis only data of traffic volume available

(without data of passengers transportation

capacity), the saving time value in t -th yearwill be calculated as the following equation.

ipac.G

it.i

avr.KitN

m

1i365.3

tB

(VND/year)

(2)

Where, ti - average saved number of

hours per passenger when using i -th type ofvehicle (hours)

ipacG - Value of a hour per passenger

when using i-th type of vehicle

(VND/person.hour)

iavrK - average number of passengers per

i-th type of vehicle

+ car: 2.5 – 3.0 (people)

+ bus: 15-35 (people)

+ motorbike: 1.0-1.5 (people)

2.2 A case of having numbers of

passengers transported due to vehicle types

ipac.G

it.Δ

m

1i

ipactQ3

tB (VND/year) (3)

where ipactQ - number of passengers

transported of i-th vehicle type in t-th year(people/year)

3. A method to measure advandges ofsaving time for goods transportation

Apprerance of a transportation

construction project can lead to a decrease of

travelling time. In other words, goods can bequicker used. This advandtage can be

estimated as a chance value due to the sooner

use of goods.

3.1 A case of having data of traffic

volume

In case of having data of traffic volume

only (lack data of goods tranportationcapacity), the saving time value for goods

transport can be accounted as below.

gds.Git.iavr.qi

tNm

1i365.4

tB

(VND/year)(4)

Where, iavrq - average load of the

vehicle transported i-th goods (tons/vehicle)


CT 2

ti - average saved number of hours ofvehicle carrying i-th goods (hours)

gdsG -average time saved value of 1 ton

of goods (VND/ton. hour)

3.2 A case of having data of goodstranported mass

gds.Git.m

1i

igdstQ4

tB (VND/year) (5)

where igdstQ - goods transported mass of

i-th vehicle type in t-th year (tons/year).

4. A method to define gains due to adecrease of number of accidents

Transportation construction projects

influece safety of passengers, goods and

vehicles by changing traffic volume or

tranportation condition. In other words, it can

decrease or increase number of accidents. For

instance, a new expressway upgrated for the

quality purpose might increase accidents if

there is no significant safety supplement.

Therefore, the influence should be estim ated.

To measure gains of reducing accidents,

it is necessary to experience 2 steps. The first

is to assess capability of decreasing collisions.

The second is to measure the advantages of

diminishing crashes.

In addition to the first step, a need is toapproximate or predict the number of crasheshappen on the road section considered. Thiscan base on data bank of road types and roadconditions after and before having the project[2; 3].

The number of accident decreased in t -th

year on a j-th road section jtA will be

collected after accomplishing step 1. Theadvantages mentioned in the second step canbe calculated as the following equation.

jtA.

jt.m

j acdC6tB (VND/year) (6)

Where, Cacd - average lost for anaccident. It can be defined by considering thedatabase of the section surveyed.

jtm -Coefficient considering effects of

situation of j-th road section in t-th yeartowards an accident.

III. CONCLUSION

This above mentioned approach is one ofmainly methods to approximate soci al andeconomic benefits of transportationconstruction projects. Besides, there are manyothers methods which largely depend onspecific condition (capital and database).

Reference

[1]. Dr. BUI NGOC TOAN. Environmental Issues

in Transportation construction projects. Scientific

studies Collection. 14-th Science and Technology

Reference – University of Transport and

Communications – 2000.

[2]. Dr. BUI NGOC TOAN. Planning and

Appraisal of Construction Projects. Culture

Publishing House-2002.

[3]. Prof. PhD. NGUYEN XUAN TRUC (chief

author): A Handbook of Road Design -Vol.1.

Education Publishing House – 2003.

[4]. Belli and other authors. Economic analysis of

investment activities – An analysis tool - A

realistic application. Culture Publishing House -

2002


CT 2

I. INTRODUCTION

Vietnam is located in the Southeast Asia Region and being neighbor of China, Laos and

Cambodia. The country has an area of 331.212 square kilometer and a population about 84 ,12

million (GSO 2008). During the last decades, the country economy grew by 7.5 % per year and

the poverty rate has been reduced from 51% in 1990 to 8% in 2005 (GSO 2007). In that period,

the transport sector in Vietnam has achieved significant improvements, which contribute

remarkably in the development of the country and region. While the trend is expected to

continue in the next decade, Vietnam’s transport sector faces a critical situation. As most of thetransport infrastructures are still being restored from the damages of war, lack of capital to

invest in new and high capacity infrastructure and services, sector performance level is still very

far from requirements of high capacity and quality to support the quick growth and foreign -

investment-driven economy.

In the followings, this report examines the conditions for transpor t development in and the

current situation of Vietnam’s transport sector. The information and data used in the analyses

TRANSPORT SECTOR IN VIETNAM:CURRENT ISSUES AND FUTURE AGENDA

DR.-ING. KHUAT VIET HUNGInstitute of Transport Planning and ManagementUniversity of Transport and Communications

Summary: This paper briefly reports about current situation and future perspectives of

transport system in Vietnam from a planning point-of-view. Firstly, it reports about the

external conditions for transport development in term of economy, demography, environment,

and technology and transport policies. Secondly it provides an overview about current

situation of transport sector in V ietnam. Finally, it drafts an agenda for transport in Vietnam

toward 2030.

Brief CV: Dr.-Ing. Khuat Viet Hung (University of Transport and Communication,

Vietnam) obtained his Doktor-Ingenieur degree from Darmstadt University of Technology (

+country name) in 2006. Immediately after that he has been reappointed as lecturer of the

Institute of Transport Planning and Management, University of Transport and Communication

(UTC). September 2007, he has been appointed as Director of the Consulting Center for

Transport Development (UTC). His work is integrated between teaching, doing research and

providing consultant service to transport sector in Vietnam. He is working and worked mainly

on highway planning and design as a research engineer. His recent major top ics of research

are performance-based urban transport planning, traffic management and traffic safety.


CT 2

had been collected mainly from up -to-dated reports and publications of Vietnamese and

international institutions and organizations. Cons equently, this report ends by a draft agenda for

Vietnam’s transport sector in the next decades.

II. CONDITIONS FOR TRANSPORT DEVELOPMENT IN VIETNAM

1. Economic conditions

The comprehensive innovation program, called Doi Moi, began in 1986 is the main fo rce

to drive the Vietnam Economy grow up c ontinuously at a quite high rate, about 7,6% per year.

The official statistical data indicates the Gross Domestics Product (GDP) in 2007 is about US$

68,3, but experts estimated an additional value, called undergro und economy, about 30% of the

official value should be accounted (Le 2008).

Figure 1. Stable growth of Vietnam’s GDP (1990-2006)

The most important achievement of Vietnam economic growth has been well known as thesharp reduction of poverty rate. The sha re of household living under poverty rate has beenreduced from 51% in 1990 to 8% in 2005 (World Bank 2006). With the new policy, Vietnam isintegrating actively into the world economy. Total import and export value in 2007 is countedfor US$ 111243,6 million, about 163% of the country’s GDP. The foreign investment inVietnam is also sharply increasing. At the end of 2007, the count ry attracted 9810 foreigninvestment projects with registered capital about US$ 99,6 billions. The government ofVietnam and most of economists has the same optimistic expectation of a continuous high rateof economic growth in the next decades. However, the main challenge of Vietnam’s economy inthe future is improvement of development efficiency. The change of Incremental Capital OutputRatio (ICOR) indicates clearly the reduction of investment efficiency in the last decade.According to the Ministry of Finance, the general ICOR of Vietnam increased from 3,39 in1995 to 5,9 in 2005 (Ministry of Finance 2006).

USD Billion

Source:Statistical Year Book of Vietnam 2007 – 1994’s Constant Price


CT 2

2. Demography and society

As similar to Germany after the Second World War, Vietnam has a period of baby boom

after the end of our national unification war in 1975. Among 85,2 million Vietnamese t oday,

about 60% of Vietnamese are less than 35 years old (GSO 2008). The young and dynamic

people are the most important resource for the fut ure development of Vietnam.

Together with economic development, Vietnam also achieves significant improvement in

social development in the last decade. According to United Nation Development Program, the

Human Development Index of Vietnam has been increase d from 0,620 in 1990 to 0,733 in 2005

(UNDP 2008). One of the most important achievement of human development in Vietnam are

relative high average life expectancy and adult literacy rate, about 73,7 years old and 90,3% in

2005. These indicate the country owns a healthy and educated labor force.

On the other hand, the high growth of economy always consists of some negative impacts,

for example the uncontrolled urbanization , environmental pollution or increase of social gaps

etc. It is necessary to emphasize that, the current political regime does not provide good

environment for social and political dialogues. Therefore, the risks of social crashes are quite

high within the society.

3. Natural Environment

As consequences of economic development, environment in Vietnam has been affected

negatively in all sectors, land and biodiver sity, water and ambient air. The forest cover is

restoring significantly after the end of Vietnam War, 1975, by huge forest planting programs.

On the other side, the natural forests are quickly reduced by official and non -official impacts.

The growth is coming together with wastes and increase of energy consumption , air pollutants

and global warming effects. The air quality in the main cities are declining significantly, the

traffic polluters and the fine particulate matter (PM10) are the main concern. A s stated by

UNDP (2008) the quantity of carbon dioxide emi ssions of Vietnam increased 25,3% per year

between 1990 and 2005.

4. Technology

In the last decade, Vietnam is one of the leading countries of growth in information

technology and telecommunication. The rate of telephone lines per 1000 people was counted a s

191, which has been drastically take -off from 1 only in 1990 (UNDP 2008). The same way of

booming is apparent in number internet users in this country. In 2000, there were only 200

thousands internet users in Vietnam, accounted for 0,3% of population. At the end of June 2008,

the number has been counted for about 20,16 million, about 23,6% of pollution (VNNIC

2008). Accounted as a key achievement in technology development in Vietnam, t he provision of

information and telecommunication technology infrastructure is a good basis to develop high -

tech applications in different industries and services, in cluding transport and logistics in this

country.


CT 2

5. Transport policies

As similar to many other developing countries , the main focus of transport policies in thelast two decades in Vietnam is road construction and rehabilitation. Lack of financial resourceand long-term vision are main obstacles of the inadequate consideration of railway and inlandwaterway, airports are also having the similar situation. Sea -ports have been announced as amain government focus in water transport , but lacking of development prioritization does notallow the country with more than 3000 km seacoast to have any regional competent port asHong Kong or Singapore.

In the vehicle aspect, Malaysian model of automotive industry development has beenapplied in Vietnam but in a smaller scale and poorer implementation. As stated in the primeminister decision number 177/2004/QD -TTg (Government of Vietnam 2004) , the target ofVietnam’s automotive industry in 2005 is 120.000 vehicles/year, but the total sale record of allVietnam automobile markers in 2006 was only 35.637 vehicles (VAMA 2007). At the sametime, the target of localization of automobile industry has not been achieved. However, thefailure of automobile industry does not make any hesitation to the decision of Vietnamesegovernment on its current focus on development of ship building industry with the aim to beone of the world leading countries in this sector. However, as similar to the beginning ofautomotive industry in 1990, Vietnam starts its dream by a huge number of unskilled labors anda government wish.

Source: (World Bank 2008)

Figure 2. Comparison between vehicle taxes and duties in selected countries

The semi-positive point of transport policy is the effort to keep high access price forindividual vehicles. This policy is ef fectively keeps the car-ownership in Vietnam at relativelow level. On the other hand, vehicle quality management has been ignored for long time. Therecent wake-up of Vietnamese government on this matter was only on paper, not in practice.

Transport services are on the process of decentralization, which has been completely (even

11200

3227 2917 2355600 282

2520

818

18182618

2864

3373

1345

504

736

366

1155245

1545

0

2000

4000

6000

8000

10000

12000

14000

16000

Vietnam Japan China Indonesia Thailand Germany UK France USA

Vehicle tax VAT Registration Vehicle Weight taxUS$


CT 2

extremely) practiced in road, inland water transport service, and recently in aviation. On theother hand, the process is much slower in railway industry.

For public bus transport in big cities, the government subsidy has been given to users viaoperators as the main effort to attract people to use this service in order to reduce trafficcongestion. According to the Department of Transport in Hanoi, number of bus riders has beenjumped from 16 million in 2001 to 350 million in 2007 . The same pace of increase is alsoobserved in bus service in Hochiminh City and other big cities in Vietnam. However, a bettermanagement system is needed in order to control the subsidy program, especially in Hochimimhcity, where the subsidy rate (VN Dong per passenger) is still very high in comparing with otherVietnamese cities.

Regarding fuel policies, the administrative measures are preferred in Vietnam. As the onlypositive point among fuel policies, leaded gasoline had been successful ly forbidden entire thecountry since July 2001. At the end of 2007, the government tried to give its control in fuelprice up, but very soon in early 2008, the tight control has already been resettled in order toreduce the inflation rate. In the area of alternative f uel, Vietnam achieved much lower progressin comparing with other ASEAN countries . Alternative fuel is almost absent in the marketalthough it has been addressed as an important content of the government policy papers.

III. EXISTING SITUATION OF TRANSPORT SECTOR IN VIETNAM

1. Transport demand and motorization

As consequence of economic growth, transport demand in Vietnam is growing intensively.In the Transport Development Strategy of Vietnam up to 2020, the Ministry of Transport (2007)projected an average growth rate of good transport demand abo ut 7.3% per year between 1990and 2030. The demand for passenger transport is growing even faster by 12% per year in thesame period. The explosion of demand presents a good opportunity for development of transportservice industries and also big challenge s for capacity of both infrastructure and service.

Source: MOT (2007)


0

20.000

40.000

60.000

80.000

100.000

120.000

140.000

160.000


Maritime Aviation

Mil. Ton-km Good Transport Demand (1990 - 2030)

1990 (1) 2005 (1) 2020 (2) 2030 (3)

0

100.000

200.000

300.000

400.000

500.000

600.000


Maritime(Coastal)

Aviation

Mil. Pax-km PassengerTransport Demand (1990 - 2030)

1990 (1) 2005 (1) 2020 (2) 2030 (3)


CT 2

Within 20 years, average income of Vietnamese people increases about 3 times, this isindicating in transport sector by the growth of vehicle ownership. As mentioned above, the tightcontrol of vehicle price is one of the main factors to keep the car ownership low and it drivespeople to enjoy the motorization process by motorcycle. According to the Vietnam Register, atthe end of 2006, total number of road motorized vehicle in Vietnam is about 18.830.000 units,94,9% of which are motorcycles. The statistical data also indicate a growth rate of motorcyclesin Vietnam is about 17,6% per year in the period fr om 1990 to 2006.

Source: National Traffic Safety Committee (2007)


The recent studies found that the Vietnam motorization curve has a similar sharp with thos ein Japan and Korea but at a lower level. If the trend would continue, Vietnam motorization willreach the Japanese and Korean level before 2030.

Source: Asia Pacific Energy Research Center (2005) and NTSC of Vietnam (2007)


Motorization in Vietnam (1990-2006)


CT 2

2. Infrastructure

Under high pressure of demand growth, Vietnam paid great effort to develop road transportinfrastructure and service. At the end of 2006, the country has 151.632 km road network, ofwhich 64.413 km have either asphaltic or concrete pavement, accounted for 42,5%. Thepercentage of paved national road has been increased from 60% in 1995 to about 92,5% in 2006(GSO 2008).

The railway has also someimprovement and rehabilitationduring the last decades, but thereis no new section of track hasbeen extended. At the end of2006, Vietnam railway has about2362 km narrow gauge tracklength and 300 locomotives.

Vietnam has about 80 ports,three of which are regionaltransport gates (Hai Phong, CaiLan, Hochiminh). The countryhas currently 20 airports, ofwhich three are internationalairport.

It was the most historictransport mode in Vietnam,inland waterway has quite highdensity network in the north andsouthern region of Vietnam witha total length of 9800 km.

However, at the moment,Vietnam is lacking of high-speedroad and railway transport links.

The country has so far nodeep-water sea port andcapacities of the internationalairports in Vietnam are very far

from which of its ASEAN neighbors, for example Suvarnabhumi Airport (Thailand) or ChangiAirport (Singapore). This problem is well aware by governments and industries as one of themain obstacle for Vietnam to have a golden economic era.


Figure 6. Strategic Transport Networks in Vietnam


CT 2

3. Expenditure and finance

As mentioned above, the increasing ICOR during last period indicating a large share ofinvestment went to infrastructure, in which transport is the focused sector. According to MOT(2007), average share of expenditure for transport infrastructure between 2000 and 2006 wasabout 2,8% of total GDP of the country. Other study found the significant growth of capitalexpenditure for new construction or major rehabilitation projects, from 71% in 1994 to 90% in2002, while the share of maintenance and operation expenditure was reduced sharply in thesame period (World Bank 2006).


Figure 7. Structure of expenditure for different transport modes in Vietnam

According to MOT (2007), road infrastructure has been intensively focused during the last

decade. About more than 90% of total central government transport expenditure has been spent

on the road. This presented a road-based transport development in Vietnam in the last decade.

Regarding the project implementation mechanism, most of the transport projects had been

awarded to the state-owned suppliers, which currently face serious problem of efficiency and

indebtness. In the recent report, the Office of State Audit of Vietnam emphasized that most ofstated-owned enterprises in transport sector having an inefficient management structure and

many of them are in serious financial imbalance. The report also referred another report of by

the Ministry of Finance, for example, the Transport Construction Engineering CorporationNumber 5 (CIENCO 5) has a debt per capita ratio about 40, the rate was about 22,5 in CIENCO

1 and about 20 in VINASHIN (SAV 2008). Another critical problem of the pro -state-owned

contractor attitude is corruption. The expenditure process has been done through a closed andnon-transparent system between government agencies and their son’s contractors.

As stated by Ministry of Transport in the Vietnam Transport Strategy up to 2020, the actual

annual expenditure for transport infrastructure in the period 2001 -2006 is counted at only 17%the planned annual fund requirement. However, this ministry made also a very ambitious

calculation for the period 2010-2010, in which the annual planned transport fund is accounted

Railway;6,1%


Maritime; 3,8%

Road;88,7%

Air

transport; 0,0%

(a) 2000

Railway;3,2%


Maritime; 3,5%

Road;91,5%

Airtransport

; 0,2%

(b) 2006


CT 2

for about 12% of the annual GDP of Vietnam (MOT 2007). As similar to other developing

countries, the share of official development assistance (ODA) in transport expenditure isincreasing significantly, accounted for 42% total MOT’s expenditure in 2002, and theexpectation on ODA is also growing as indicated in the rep ort of many transport planning

studies. The Public-Private-Partnership and commercial loan accounted only 7 -8% and the rest

of expenditure came from government budget. However, the availability of low -interest rateODA for Vietnam is reducing and asking fo r a new and sustainable structure of transport

financing in the next period.

4. Transport performance

4.1. Enhancing development and poverty reduction

Regarding transport performance, it is necessary to emphasize the great contribution oftransport development in economic development and poverty reduction. Different studies aboutimpacts of transport infrastructure on development in Vietnam had been carried last decades andproved that developing large scale infrastructure in Vietnam helped to open up new businessopportunities and facilitated the spread of economic linkages between economic growth centersand its surrounding areas. For example, 90% of the investment along ha No –Hai Phongcorridor had been taken place due to the completion of National High way number 5 expansionproject (World Bank 2006).

4.2. Accessibility improvement

The road based transport development in the last decades made significant improvement ofaccessibility of households, of wh ich 80 percents are living the non -urbanized areas. Theaffordability to public transport and motorcycle, as the main individual transport mode, has beensignificantly improved and the cost of trucking service and fuel price has also been affordable incomparing with the neighboring countries (GTZ 2007).

Source: Household Living Standard Survey 1998, 2002

Figure 8. Impacts of road development in improving accessibility of rural households

Household access to road (rural are a)


CT 2

4.3. Traffic safety

Negatively, transport development in Vietnam is directly proportional to the gr owth oftraffic accidents and fatalities. With 15 road fatalities per 100.000 inhabitants, Vietnam roadnetwork is the most deadly one over the world.

Table 1. Traffic Accident Data in Selected Countries

CountriesNo. of

AccidentsNo. ofDeaths

No. ofInjuries

Death/accident Injuries/accident Injuries/death

(a) (b) (c) (d) (e) = c/b*100% (f) = d/b*100% (g) = (d) / (c)

Thailand (2000) 73,737 11,988 53,111 16.26 72.03 4.43

Malaysia (2000) 250,417 6,035 44,019 2.41 17.58 7.29

Japan (2005) 920,053 6,586 1,134,702 0.72 123.33 172.29

Vietnam (2007) 14,624 13,150 10,546 89.92 72.11 1.24

US (By car) 2005 33,041 2,494,000 75.48

Source: JICA & NTSC (2008)

4.4. Mobility

In general, the improvement of road accessibility and the affordability of motorcycle have

made the mobility level of most of V ietnamese people much better than before. However, the

growing demand will become over the current capacity of all main transport corridors in a very

near future and requiring prompt and effective solutions. In the urban area, current high

motorcycle ownership gives people a reasonable level of mobility, but the threat of congestion

is apparently in the next decade by booming of car use. In this regard, high pressure of WTO

commitment is one of the main factors to enhance car use in Vietnam.

4.5. Efficiency of System Operation

Regarding the efficiency, the road and air transport has achi eved significant improvement

while the others had got lower level of progress. Since last 1.2006, the regulated operating speed

in the national road has been increased from 50 -60 km per hour to 70 to 80 km per hour. As the

most focused investment transport mode in the last decade, the road transport carried about 54%

total good transport demand and 85,7% total passenger transport demand in Vietnam (MOT

2007). The air transport service has also achieved a significant improvement in both number of

passengers and diversifies of services. The total number of air tran sport passenger carried in

2007 was about 8,5 million, about double of the year 2004 figure, while the difference in

number of air-craft operations was only 1,4. The same level of improvement was observed in

the maritime transport, total cargo throughput v ia the ports of Vietnam increased 218% between

2007 and 2005. In contrast, the efficiency in railway and inland waterway has not been

improved due to poor quality, low capacity facilities and infrastructure and old model of

management.


CT 2

IV. AGENDA TOWARD THE FUTURE TRANSPORT DEVELOPMENT IN VIETNAM

The high expectation of economic development, the maturity in internationalization of theeconomy and the changes in social factors would result an explosion of transport demand inVietnam the next decades. On the other hand, the current situation of lacking infrastructure andservice capacity, ineffective traffic management, and inefficient planning and utilization ofresources would continue to be the main obstacles of Vietnam transport sector on the waytoward its future development. The following measures have been strongly recommended byboth foreign and Vietnamese experts and institutions in order to support the Vietnamesegovernment to develop its agenda for transport development in the next decades:

1. Enhancing efficiency in resource utilization and service deliveryThe first activity to enhance the efficiency of transport investment is to improve the

transparency of the expenditure process by the better control mechanisms and privatization ofstate-owned enterprises. The next urgent activity is to develop a comprehensive performance -based planning system and consequent procedures. The improvement of transport planning willhelp to avoid the un-attainable, biased and emotional focuses in transport development and todevelop proper priority list of activities in transport development .

2. Sustain financial structure for transport developmentAs stated in many consultant reports and advisories, t he first answers for the question of a

new financial structure is to emphasize again the importance of transport expenditure controlimprovement, which includes also an effective infrastructure pricing system This measurewould help the government to maximize the utilization of available resources. In order toovercome the financial shortage by the running-out of ODA age, suitable Public-Private-Participation models in transport investment and operation is strongly recommended. On theother hand, government should develop a transparent and applicable framework for the use ofgovernment bonds, which should be opened for competitive bidding. Finally, the developmentof a framework for municipal finance should be promoted as the key part of the on -goingdecentralization and administrative reformation process.

3. Facilitating compact and efficient urban growthAs mentioned shortly above, the development of Vietnam will contain a quick and lack of

control urbanization process. The indication of urban sprawls and ribbon urban growth requiresalso a new concept of compact and effici ent urbanization, which needs strengthened planning,and regulatory system, proper institutions, high capacity and efficiency transport infrastructuresand services.

4. Mitigating negative impacts of transportThe lack of high capacity requires a very car e-full examination and action on developing

new transport infrastructures and services. Regional infrastructures and services arerecommended to be in the first phase of capacity development in order to deal with transportproblems. In parallel, effective traffic management, traffic safety plan s and other demandmanagement tools are strongly recommended to be part of the solution package in mitigatingtransport congestion, accident and environmental impacts .

5. Institutional capacity and human resource deve lopmentTo realize all of the above requirements, institutional and human resource capacity in

transport sector must be correctly developed. A long term institutional capacity development


CT 2

frame-work is the first task to complete and following by the streng thened implementationcapacity. Implementation of these two measures asks to speed-up of governance reform process,especially to develop capacity of Vietnam Road Administration, which is responsible for largestamount of transport investment and propertie s. Improvements of other sectoral authorities arealso needed. Enforcement and compliance require also significant improvements. Finally, it isno doubt to address the needs to develop capacity local governments a nd contractors, which arethe most important stakeholders in transport development.

V. CONCLUSIONS

The study results indicated clearly that the development of transport sector is one of thekey factors of success in economic development of Vietnam in the last two decades of Doimoiera. The transport development contributed significantly to the poverty reduction, improvementhousehold accessibility and mobility of most of people, businesses and institutionscontributions in Vietnam. On the other hand, the limitations of low quality expenditure cont rol,planning and implementation, un -sustainable financing structure, transport problems, and thelow capacity of institution and human resource are the main obstacles for the development oftransport in particular and for the cause of industrialization an d modernization of Vietnam ingeneral. Overcome the obstacles is the main goal of the agenda for Vietnam’s transport sector inthe next decades.

Reference[1]. Government of Vietnam (2004), Master Plan of Automotive industry in Vietnam to 2010 and visionto 2020. 177/2004/QD-TTg.[2]. GSO (2007), Vietnam Statistical Year Book 2006 , Hanoi, General Statistics Office of Vietnam.GSO (2008), "Population and Population Density by province in 2006," 2008, fromhttp://www.gso.gov.vn.[3]. GSO (2008), Vietnam Statistical Year Book 2007 , Hanoi, General Statistics Office of Vietnam.[4]. GTZ (2007), International Fuel Prices 2007. T. P. A. Service, GTZ and Federal Ministry forEconomic Cooperation and Development.[5]. JICA & NTSC (2008), the Study on National Road Traffic Safety Master Plan in the SocialistRepublic of Vietnam: Interim Report. Hanoi, Japan International Cooperation Agency and NationalTraffic Safety Committee of Vietnam.[6]. Le, D. D. (2008), Is the underground economy of Vietnam accounted for 30%? VTC News. Hanoi,Vietnam Telecommunication Corporation.[7]. Ministry of Finance (2006), "Vietnam aims to a higher efficiency and sustainable development,"2008, from http://www.mof.gov.vn/Default.aspx?tabid=612&ItemID=31186 .[8]. MOT (2007), The Transport Development Strategy of Vietnam up to 2020, Ministry of Transport.[9]. SAV (2008), State Auditing Report 2007, Office of the State Audit of Vietnam.[10]. UNDP (2008), "Human Development Report 2007/2008."[11]. VAMA (2007), "VAMA sales record 2006," from http://vama.wordpress.com/2007/06/04/vama -sales-record-2006/.[12]. VNNIC (2008), "Report on internet statistics of Vietnam: June 2008," Retrieved 29 July 2008,from http://www.thongkeinternet.vn/jsp/trangchu/index.jsp .[13]. World Bank (2006), Infrastructure Strategy: Cross-sectoral issues, World Bank.[14]. World Bank (2006), Transport Strategy: Transition, Reform and Sustainable Management, WorldBank.[15]. World Bank (2008), Study in urban transport strategies for medium -sized cities in Vietnam, WorldBank, Hai Phong City People Committee, Ha Long City People Commit tee

http://www.gso.gov.vn

http://www.mof.gov.vn/Default.aspx

http://vama.wordpress.com/2007/06/04/vama-

http://www.thongkeinternet.vn/jsp/trangchu/index.jsp


I. INTRODUCTION

Recently, there are many projects in Vietnam that applied design seismic. Transportconstruction projects mainly apply Specification for Highway bridge 22TCN 272 -05 belongsMinistry of Transport (refered to AASHTO 1998). Building construct ion projects mainly applySpecification for Design of structures for earthquake resistance TCXDVN 375 -2006 (referred toEUROCODE 8). Both of Specifications use “Acceleration coefficient A” or “The map ofseismogenic zones and maximum seismic intensity” pub lished by Institute of Geophysicsbelong Vietnamese Academy of Science and Technology for seismic design. In addition, Japan,that has been suffered many earthquake damages, has much experience in seismic design.Question under investigation for Vietnamese engineer as well as Japanese engineer is ability ofapply Japanese Specification for Seismic design of construction projects in Vietnam. Thisarticle analyses the difference between AASHTO 1998 and Japanese Specification in seismicdesign; contribute to conclusion for applying of Japanese Specification for construction projectsin Vietnam.

II. COMPARISON OF SOME FACTORS EFFECT ON EARTHQUAKE LOAD INAASHTO 1998 AND JAPANESE SPECIFICATION

AASHTO 1998 Japanese Specification

Earthquake

loads

Seismic loads assumed to act in

any lateral direction.

Seismic load is inertia force that shall be

calculated in terms of the natural of each

design vibration unit.

COMPARISON BETWEEN JANPANESE SPECIFICATION AND

AASHTO 1998 SPECIFICATION IN SESMIC DESIGN

MSC. NGUYEN THI TUYET TRINHPh.D Candidate, University of Transportand Communications, VietnamDR. TAKEHIKO HIMENOKawaguchi Metal Industries, Japan

Abstract: Japan has much experience in seismic design. This article analyses the

difference between AASHTO 1998 and Japanese Specification in seismic design; contributes

to conclusion for applying of Japanese Specification for constructio n projects in Vietnam.

Keywords: Acceleration coefficient; Standard acceleration response spectrum; Elastic

response coefficients


Earthquake loads shall be taken

to be horizontal force effects on the

basis of the elastic response

coefficient, Csm and the equivalent

weight of the superstructure, and

adjusted by the response

modification factor, R.

The elastic seismic force effects

on each of the principal axes of a

component resulting from analyses in

the two perpendicular directions shall

be combined to form two load cases

as follows:


value of the force effects in one of

the perpendicular directions



the second perpendicular direction,

and


value of the force effects in the

second perpendicular direction



the first perpendicular direction.

Load combination in sesimic

design = Permanal loads + 1/2 Live

load + Water Pressure + Friction

Load + Earthquake effect

Inertia forces shall be generally

considered in two directions perpendicular to

each other. It can be assumed that the inertia

forces in the two orthogonal directions, i.e. the

longitudinal and transverse directions to the

bridge axis.

Inertia force shall be defined as the

horizontal force equal to the product of the

weight of a structure and the design horizontal

seismic coefficient and be considered acting on

the structure in the detection of the inertia

force of a design vibration unit.

Load combination in sesimic design =

Primary load + Earthquake effect (= Permanal

load + Water Pressure + Friction Load +

Earthquake effect)

- Earthquake effects (EQ):

(1) Inertia force due to an earthquake

(2) Earth pressure during earthquake

(3) Hydrodynamic pressure during earthquake

(4)Effect of liquefaction and liquefaction -

induced ground flow

(5) Ground displacement during earthquake

Calculation

formula for

earthquake

force by

statically

method

EQ=W.Csm/R

EQ: Earthquake force (kN)


Csm: Elastic response coefficient.

R: Response modification factor.

H=W.khco.Cz.Cs

H: Earthquake force (kN)


khco: Standard value of the design hori zontal

seismic coefficient.

CZ: Modification factor for zone.

CS: Force reduction factor.


Basic valueofearthquakeload

Acceleration coefficient A:Determined by the nationalearthquake ground motion map usedin the existing AASHTO provisions,that is a probabilistic map of peakground acceleration (PGA) on rockwhich was developed by the U.SGeological Survey (USGS, 1990).The map provides contours of PGAfor probability of exceedance (PE) of10% in 50 years, which isapproximately 15% PE in the 75years design life of tipycal highwaybridge.

Table 1 Acceleration coefficient

Accelerationcoefficient

Seismiczone

MSK - 64class

A 0.09 1 Class 6,5

0.09 < A 0.19

26,5 < Class

7,5

0.19 < A <0,29

37,5 < Class

8

A0,29 4 Class > 8

Standard acceleration responsespectrum S0 : obtained from strong motionrecords with 394 components observed at theground surface in Japan, with these resultsmodified to account for the characteristics ofpast earthquake damage, vibration propertiesof the ground, and other engineeringevaluation.

Table 2 Standard acceleration response

spectrum S0

Groundtype

SIIO (gal) with natural period T(s)

I

T<0,3

SIIO=4.436T2

/3

0,3≤T≤0,7

SIIO=2.000

0,7<T

SIIO=1.104/T5/

3

II

T<0,4

SIIO=3.224T2

/3

0,4≤T≤1,2

SIIO=1.750

1,2<T

SIIO=2.371/T5/

3

III

T<0,5

SIIO=2.381T2

/3

0,5≤T≤1,5

SIIO=1.500

1,5<T

SIIO=2.948/T5/

3

Factordepend onseismic zone

There is no modificationfactor for zone , accelerationcoefficient is classified according to

4 seismic zones, that are A 0.09;

0.09 < A 0.19; 0.19 < A < 0.29;

A0.29 corresponding to zone 1, 2,

3, 4

Modification factor for zone CZ is 1.0;0.85; 0.7 corresponding to zone A, B, C tocorrect the acceleration response spectrum S 0,that applied for the bridge in large scaleearthquake may happen

There are 3 zones following the regionalclassification map. The regional classificationof earthquake ground motion complied by theMinistry of Construction. This map has beenprepared by examining the results of studiedpublished so far concerning the seismic risk inJapan, to obtain practical applicable regionalcharacteristics of seismic risk and alsocomprehensively examining together withpractical applicable data on the earthquakeoccurring at inland active faults.


Factordependingon naturalperiod T andsoil profile

Elastic Seismic ResponseCoefficient Csm

A5,2T

AS2,1C

3/2m

sm

Tm: Period of vibration of them th mode (s)A: Acceleration coefficientS: Site coefficient specified

For soil profiles III and IV, andfor modes other than the fundamen talmode that have periods less than0.3s: Csm = A(0.8 + 4.0 Tm)

If the period of vibration forany mode exceeds 4.0 s:

3/4

3

m

smT

ASC

Figure 1. Accelerationresponse spectrum Csm

Standard value of the design horizontalseismic coefficient khco

khc0=f(T, S) as Table 3:Table 3. Standard value of the design

horizontal seismic coefficient k hco

Groun

d type

khco , value in term of natural

period T (s)

I

T<0,3

khcO=4.46T2/3

0,3≤T≤0,7khcO=2.0

0,7<T

khcO=1.24/T-4/3

IIT<0,4

khcO=3.22T2/3

0,4≤T≤1,2khcO=1.75

1,2<T

khcO=2.23/T-4/3

IIIT<0,5

khcO=2.38T2/3

0,5≤T≤1,5khcO=1.50

1,5<T

khcO=2.57/T-4/3

Figure 2. Standard value of the designhorizontal seismic coefficient khco

III. ANALYSIS CONSIDERATION METHOD OF DESIGN SEISMIC FORCE IN

AASHTO 1998 AND JAPANESE SPECIFICATION

3.1. Consideration method of design seismic force in Japanese Specification

Japanese Specification do not use acceleration coeffici ent A or PGA, then design by

acceleration response spectrum base on acceleration strong motion records actually obtained at

ground surface (obtained from earthquake happening in Japan such as Hyogo -ken Nanbu

earthquake of 1995 or disaster of large scale e arthquake in Kanto). The procedure of seismic

design is as follow:

a. Records actually obtained at ground surface

For example, during the Hyogo-ken Nanbu earthquake of 1995, the high acceleration was

0.0

0.5

1.0

1.5

2.0

2.5

0.0 1.0 2.0 3.0 4.0 5.0Acc

eler

atio

n re

spon

se s

pect

rum

x g

(m/s

2)

A=0.3g

A=0.4g

A=0.5g

A=0.7g

A=0.8g


recorded at Kobe Maritime Meteorological Observatory

Figure 3. Acceleration recorded during the Hyogo -ken Nanbu earthquake of 1995

b. Calculate acceleration response spectrum of structure

Act the obtained acceleration on structure having different natural period and establishacceleration response spectrum of the structure.

Fifure 4. Steps of establishment of acceleration response spectrum of the structure

c. Calculate acceleration response spectrum of structure for each ground type

Base on 3 ground types, establish acceleration response spectrum of structure for eachground type

5%

Natural period

Damping constant

(Damping constant: h1)

Speed response spectrum

Natural period (s)

Max

of

acce

lera

tion

resp

onse

Acc

eler

atio

n re

spon

se f

ollo

ws

tim

e

c. Response spectrumb. Response follows time historya.Damping constant not changed; Natural

period changed

Earthquake

Acc

eler

atio

n (g

al)

Time (s)

-1000

-500

0

500

1000

0 5 10

15

20

25

30

PGA =812cm/s2

Japanese does not use this valuedirectly


Figure 5. Acceleration response spectrum of struc ture for each ground type

d. Calculate standard value of design horizontal seismic coefficient

Standard value of design horizontal seismic coefficient for each natural period establishedby modification of acceleration response spectrum of ground motion t hrough damping constantfor each natural period

g

)h,T(Sk 0h

Therefore, acceleration response spectrum and standard value of design horizontal seismiccoefficient are little different as below figure 6:

Figure 6. Acceleration response spectrum and Standard value of design horizontal seismic coefficientNote: Upper line is Standard value of design horizontal seismic coefficientUnder line is Acceleration response spectrum


Acc

eler

atio

n re

spon

se s

pect

rum

S (g

al)

-

Type IType IIType III

1 2 3 4 5

500

1000

1500

2000

Design seismic coefficient

Acceleration Response spectrum

Level 2 type II

0.0

5.0

10.0

15.0

20.0

25.0

0.0 1.0 2.0 3.0 4.0 5.0Natural period T (s)

Soil type I

Design spectrum (type I)

Solid type II

Design spectrum (type II)

Soil type III

Design spectrum type III

Acc

eler

atio

n re

spon

se s

pect

rum

S(m

/s2)


e. Actual calculation

In static design, calculate standard value of design horizontal seismic coefficient ( khco)

follow above steps, then multiply with below factors for getting design horizontal seismic

coefficient (khc) for design.

Factor for zone: Base on probability of earthquake occurring in z one Cz= 0,7~1,0

Factor for structure’s property: Base on plasticity of structure’s component12

1CS μ

with

μ allowable ductility ratio, about 0.45

Factor for damping: Base on damping method such as isolation bearing shoe C E=0.7~1.0

Factor for modification of dynamic: Base on relative difference between superstructure

and substructure Cm=1.2

3.2. Consideration method of design seismic force in AASHTO 1998

a. Decide acceleration coefficient

Acceleration coefficient in AASHTO 1998 is “peak of ground acceleration (PGA)” or

“maximum value of ground acceleration” considering return period or probability of exceedanc e

(PE), it looks seismic coefficient in seismic design

Decide acceleration coefficient from hazard map considering to zone’s properties and

return period or probability of exceedance

Table 5. Acceleration coefficient

Zone Acceleration coefficient

1 A0.09

2 0.09＜A0.19

3 0.19＜A＜0.29

4 0.29＜A


b. Calculate elastic seismic response coefficient Csm (or response acceleration of structure)

Response acceleration of structure for its natural period Tm established by modification of

acceleration coefficient through site coefficient S

A5.2T

AS2.1C 3/2

m

sm ， )T0.48.0(AC msm ,3/4

m

smT

AS3C

For soil type III, IV and T < 0.3 For T > 4.0

Figure 7. Acceleration response spectrum C sm

c. Actual calculation

In static design, calculate response acceleration of structure ( Csm)follow above steps, then

consider to below factors for getting C sm for design.

Factor for structure’s properties: Base on plasticity of structure’s component R= 0.8 ～5.0

Factor for damping: Base on damping method such as isolation bearing shoe B= 0.8～2.0

3.3. Basic different between ASHTO1998 and Japanese Specification in seismic design

Both of acceleration coefficient of AASHTO 1998 and standard value of design horizontal

seismic coefficient of Japanese Specification give similar result, however start point and

procedure to the result of both are different.

a. Start point and procedure to the result of both are different

Japanese Specification

0.0

0.5

1.0

1.5

2.0

2.5

0.0 1.0 2.0 3.0 4.0 5.0

Acc

eler

atio

n re

spon

se s

pect

rum

x g

(m

/s2) A=0.4g

A=0.5g

A=0.6g

A=0.7g

A=0.8g

A=0.3


A=0.16g


Obtained acceleration in the past → Acceleration response spectrum S → Consider to

modification factor for standard value of design horizontal seismic coefficient khco

AASHTO 1998

Acceleration coefficient A or PGA considering return period → Acceleration response

spectrum Csm → Consider to modifica tion factor for elastic seismic response coefficient C sm

b. Result of acceleration response spectrum

Both Specifications give curves of Acceleration response spectrum; the comment is given

base on graph:

Maximum value of acceleration response spectrum of b oth Specification concentrate to

similar value of natural period

Value of maximum acceleration response spectrum of level 2 earthquake of Japanese

Specification is similar to value of maximum acceleration response spectrum of acceleration

coefficient A=0.8 of AASHTO 1998

Reduction slope of acceleration response spectrum at long natural period of Japanese

Specification is more sloping than AASHTO 1998

Figure 8. Curves of acceleration response spectrum of AASHTO 1998 and Japanese Specificatio n

0.0

5.0

10.0

15.0

20.0

25.0

0.0

1.0

2.0

3.0

4.0 5.0

Japan Level 2

A=0.16g

A=0.5g

A=0.8g

Nhat tanBridgeA =0.16

Csm=0.40

Acc

eler

atio

n re

spon

se s

pect

rum

S (

m/s

2)

Earthquake level 2 of Japan is simi lar to max0.8g value of AASHTO

Reduction at long period is deference



IV. COMMENTS

Each Specification has own back ground and composes each design procedure by own

philosofy. Therefore, basically we cannot use two design Specifications with mixed way in one

project, namely each equation, table, etc. are not like subroutine s in Specification for aplying.

Comparison between two Specifications is necessary to more understand each back ground

and philosophy. Base on above comparisons, the comment is proposed that we cannot use

acceleration coefficient A of AASHTO 1998 for seis mic design according to Japanese

Specification. However in the case of applying the Japanese Specification to carry out seismic

design for projects in Vietnam, that have only acceleration coefficient, in limited range we can

use Elastic seismic response coefficient Csm of AASHTO 1998 replacing for factor of k hco.CZ for

seismic design according to Japanese Specification.

If possible, comparison between results of the existing bridges, which carried out by two

Specifications of seismic design, will give us m ore detail comment about the difference between

two Specifications in Seismic design

References

[1] American Association of State Highway and Transportation Officials (1998), AASHTO LRFD Bridge

Design Specifications

[2] Transport Ministry of Vietnam (2005), Specification for bridge design 22TCN -272-05

[3] Japan Road Association (2002), Specifications for Highway Bridges

[4] Multidisciplinary Center for Earthquake Engineering (2001), Recommended LRFD Guidelines for th e

Seismic Design of Highway Bridges

[5] American Association of State Highway and Transportation Officials (1999), Guidelines for

Specifications for Seismic Isolation Design

BOARD OF EDITORS - IN - CHIEF

Prof. V. Prikhodko; Prof.V. Silyanov; Prof. Wanming Zhai; Assoc.Prof. Tran Tuan Hiep

EDITORIAL COUNCIL

MADI’s Editorial WSJTU’s Editorial UTC’s Editorial

Prof. A. Buslaev Prof. Dr. Zhao Yong Assoc.Prof. Dr.Pham Van At

Prof. A. Chubukov Prof. Dr. Zhai Wanming Dr. Tran Van Dung

Prof. I. Demyanushko Prof. Dr. Qiu Yanjun Prof.Dr. Nghiem Van Dinh

Prof. I. Fedorov Prof. Dr. Li Qiao Msc. Vu Minh Duc

Prof. V. Gerami Prof. Dr. Gao Shibin Assoc.Prof. Dr. Tran Tuan Hiep

Prof. A. Ivanov Prof. Dr. Peng Qiyuan Assoc.Prof. Dr. Le Van Hoc

Prof. L. Makovskyi Prof. Dr. Pan Wei Assoc.Prof. Dr. Dao Quang Liem

Prof. A. Nikolayev Prof. Dr. Li Fu Assoc.Prof. Dr. Vu Duy Loc

Prof. V. Nosov Prof. Huang Nan Assoc.Prof. Dr. Tran Dac Su

Prof. P. Pospelov Prof. Dr. Liu Xueyi Assoc.Prof. Dr. Tu Si Sua

Prof. A. Rementzov Prof. Dr. Zheng Kaifeng Assoc.Prof..Dr.Nguyen Huy Thap

Prof.. M. Shatrov Prof. Dr. Zhang Jin Dr. Tran Quoc Thinh

Prof. Yu. Trofimenko Prof. Dr. Liu Dan Prof. Dr. Do Duc Tuan

Prof. M. Ulitskyi Prof. Dr. Yang Yiren Assoc.Prof. Dr. Tran Quang Vinh

Prof. V. Vlasov Prof. Liu Bin Assoc.Prof. Dr. Nguyen Van Vinh

Prof. V. Zhurakovskyi Prof. Dr. Song Jirong Dr. Nguyen Quynh Sang

Prof. V. Zorin

SECRETARY SECTION

MSc. V. Vinogradova Assoc. Prof. Lan Junsi MSc. Vu Minh Duc

MSc. V. Lipskaya