Deep basements & cut & cover - 3

8/19/2019 Deep basements & cut & cover - 3

1/86

National University of SingaporeDepartment of Civil Engineering

CE 5112

Structural design and construction of

deep basements &cut & cover structures

Lecture 3

1


2/86

Words of wisdom

Rules for the game of engineering:

1. Engineering is a noble sport which calls for good sportsmanship.Occasional blundering is part of the game. Let it be your ambition to

be the first one to discover and announce your blunders. If somebodyelse gets ahead of you, take it with a smile and thank him for hisinterest. Once you begin to feel tempted to deny your blunders in theface of reasonable evidence, you have ceased to be a good sport. Youare already a crank or a grouch.

2. The worst habit you can possibly acquire is to become uncriticaltowards your own concepts and at the same time skeptical towardsthose of others. Once you arrive at that state, you are in the grip of senility, regardless of your age.

3. When you commit one of your ideas to print, emphasize everycontroversial aspect of your thesis which you can perceive. Thus, youwin the respect of your readers and are kept aware of the possibilitiesfor further improvement. A departure from this rule is the safest wayto wreck your reputation and to paralyze your mental activities.

4. Very few people are either so dumb or so dishonest that you could notlearn anything from them. Karl Terzaghi

2


3/86

Practical Design Considerations

1) Introduction – sharing of structural engineer perspectives

2) General requirements – clients, builders & designers

3) Ground, soil profile & gases

4) Concept of effective stress vis-à-vis total stress5) Groundwater control

6) Movements caused by excavation activities

7) Methods of construction8) Types of earth retaining system

9) Influence of foundations type adopted

10) Site Investigation

11) Geotechnical & structural analysis, soil-structure interaction

12) Protective measures

13) Durability and waterproofing

14) Safety, legal and contractual issues & risk communications3


4/86

Design and analysis of retaining system

Design of retaining structures has traditionally been carried out usingsimplified analyses or empirical approaches. Methods have been developedfor free-standing gravity walls, embedded cantilever walls (fixed earthsupport) or embedded walls with a single prop (free earth support). Theseare described in BS 8002 and CIRIA Report C580. Statically indeterminate,

multiple propped retaining systems have often been dealt with usingempirical approaches.

Suitable factors of safety have been applied to cater for uncertainties aboutsoil properties, to allow for the often-approximate nature of the calculationmodel and to ensure that retaining wall displacements are acceptable.

Development of these factors has been based on experience, often as a result of trial and error.

The rise of sophisticated FEM programs has led to considerable advances inthe analysis and design of retaining structures. Designers are able to modelthe behavior of walls in service and investigating the mechanisms of soil-

structure interaction. Giving designers the ability to predict service loadsand wall movements with more confidence. Allowing designers greater

understanding of wall behavior and identifying of major influences and keyareas affecting the design of retaining systems. (Refinement & SensitivityAnalyses)

4


5/86

Stability considerations

Limiting earth pressures

The soil pressures to be resisted by an earth retaining structurevery much depend on the magnitude of strains permissible in

the ground. The pressures of the ground at active and passivefailure define the lower and upper limits of these forces andrelated strains. The lower (active failure) or upper (passivefailure) limits are reached when the soil is allowed respectively

to extend or compress laterally to permit full mobilization of the soil’s shear strength. These two extremes are usuallyexpressed by the coefficients of active and passive earth pressure, K a , and K p respectively. These coefficients give the

ratio between lateral and vertical effective pressures at activeand passive failure. They are calculated from the soil strength,the angle of wall friction and the geometry of the wall and thesoil surface.

5


6/86


Relationship between earth pressure coefficient, K a,p and wall movement

6


7/86

Time considerations

Earth pressure coefficient is between K a and K o

7


8/86


Water pressures and the effects of seepage

The forces exerted by groundwater are often greater than those from the soil. Careful consideration should

be given to variation of water levels & pressures oneach side of the wall.

Even more significant can be the effects of seepage of

water around the base of the wall and into the basement area. This will tend to reduce water pressures below hydrostatic on the outside of the walland increase water pressure above hydrostatic on the

excavation side. The higher pressures inside willresult in lower vertical effective stresses and thus passive earth pressures. Seepage effects need to be properly accounted for in assessing stability and wall

performance.8


9/86


Gravity wall systems

Active pressures are assessed and applied to the retaining wall

and, passive pressures are assumed in front of the wall, Water

pressure are added based on the drainage and seepage regimearound the wall. Base friction is considered. The resulting

force R is then calculated and stability is checked for:

Conventional bearing capacity

Sliding & rotational stabilities

Slip circle

9


10/86


Cantilever & Singly-propped wallsCantilever walls: The mode of failure of the wall is by rotation

about a point near the toe applying the resulting active and

passive pressures. This is a statically determinate system and,

for any given active and passive pressure limits, there is a unique solution for the depth of wall.

Singly-propped walls: The failure mode is by rotation of wall

about the prop level. This is also a statically determinatestructure.

10


11/86

Stability considerationsMulti-propped walls

Overall instability is unlikely to arise in the cases of multi-

propped/anchored walls because of the redundancy of the structure system.

However, local instability may arise as the result of local overstressing and

the formation of hinges.

The amount by which the toe of a wall extends below excavation level is

controlled by stability - kick in requirement or to limit seepage. As the clay

softens, movement will occur towards the excavation. Movement in the

soil at the sides of the excavation would have detrimental effects on the

foundations of adjacent structures or nearby services.No generally accepted methods for analyzing such failure by way of hand

calculations.

Multi-propped wall stability

11


12/86


Increase the strength of the toeof the wall especially where it

connects to the base slab (a).

Do away with the void under

the base slab. This may result ina build-up of pressure on the

base slab, which must be

accounted for in the slab design

(b).

Increase the vertical effectivestress in the soil immediately in

front of the toe of the wall.

This can be achieved by

installing pin piles (c) or by

using a partial soil-bearing baseslab (d).

Extend walls deeper into

stronger soil if such soil is

present (e).

12


13/86

Stability considerationsCircular Basements

Circular basement plan may provide an economical solution, as a circular wall

structure can sustain hoop compressive stresses caused by radial earth pressure. The

required basement must fit within the circular plan without excessive waste of

space. Most important, uniform hoop compression will occur only where ground

and groundwater conditions are uniform around the circular structure.Soil mix piles, bored piles and diaphragm wall are used in circular basements, they

are designed to span vertically between circular walings or internal lining walls. At

Heathrow airport, a large circular basement (or cofferdam) using secant piles was

used for a large circular excavation, of 30m in diameter & 30m deep, was installed

through disturbed ground following a tunnel collapse. An internal continuousreinforced concrete lining, cast progressively with excavation to “prop” the piles,

with the lining acting in hoop compression.

When diaphragm walls are used for circular wall construction, it is formed by

straight panels. These walls are designed to span vertically between circular walingsor the walls themselves are allowed to act in hoop compression where continuity of

reinforcement was provided through vertical joints in the diaphragm wall to ensure

development of hoop stresses. One such large circular diaphragm walls was built for

the basement of the new world library, the Bibliotheca in Alexandria, Egypt. It was

l50m in diameter and 35m deep, designed to also resist seismic forces. (Singapore – 100m with walings for Marina IR Sand)

13


14/86

Stability considerationsFactors of safety

At present, there are several ways in which the factor of safety can be applied for

wall stability: Partial factors, factored soil strength or embedment depth or lumped

factor applied to some combination of the active and passive earth pressures. In

order to arrive at the same wall design, each approach requires a different numerical

factor of safety. Therefore the adopted factor of safety need to be consistent withthe approach.

The only consistent approach for all types of wall is the use of factored soil

strength. This is the approach proposed by CIRIA Report C580 which also offers

guidance on the appropriate choice of strength parameters. The factored soil

strength, or ‘allowable mobilized strength’, can be used consistently in calculationsof both bearing capacity and wall stability. Factors of safety can be increased to a

magnitude sufficient to limit movements to an acceptable level basing on

experience. This method should be used only when displacements are not a critical

concern.

Determining soil and wall movements is difficult and is likely to remain only

approximate until further numerical analyses are calibrated against field experience.

Consequently, the recommended factors of safety used for stability analyses are

often large enough to limit movements to an acceptable level.

14


15/86

Temporary works design

Temporary props/anchors are replaced by permanent support

in the form of floor slabs of the finished structure. Basement

structures may therefore be subjected to two different sets of

loading and support conditions during construction &

permanent as-built. Both must be considered carefully

including lock-in stresses.

Clayey soil behavior can be time-dependent, with differentcharacteristics under short- (undrained) and long-tern

(drained) loading. Depending on the past stress history of the

clay, i.e., normally or over-consolidated, & the method of

construction, the long-term strength may be higher or lower

than in short-term.

15


16/86

Temporary works design

For deep basement construction, the soil is likely to be weaker

in the long term. While it is possible to estimate long- and

short-term soil strength, it is difficult to predict the length of

time for the change in strength. This is a sensitive issue because any underestimate of the strength reduction could lead

to an unsafe situation. While an overestimate lead to expensive

temporary work design.

Unless there are good reasons to the contrary, analyses should

be undertaken to show that the factor of safety using effective

strength calculations based on long-term conditions is greater

than one. This is particularly relevant if circumstances mightlead to a temporary stage of excavation being delayed beyond

the anticipated period.

16


17/86

Earth pressures

In designing a retaining structure the magnitude and

distribution of the stresses and movements for both the

temporary and permanent works stages of construction should

be ascertained. The earth pressures will depend on the initialin situ soil stresses, wall construction method and its stiffness,

and the number & stiffness of supports.

Backfilled wallsWith gravity retaining walls, if soil is backfilled behind the

wall, the compaction process will induce both transient and

residual horizontal pressures on the wall. The amount of these

pressures depends on the type of fill, state of compaction and

flexibility of the wall and its supports.

17


18/86

Earth pressures

Initial earth pressures & coefficient of earth pressure at rest K o

In its initial natural state, the horizontal effective stresses in

the ground will be somewhere between those associated withactive failure and passive failure. K o is defined as the ratio

between the horizontal and vertical effective stresses initially.

The magnitude and variation with depth of the initial ‘at rest’horizontal effective stresses depend on the loading history of

the soil.

18


19/86


20/86

Earth pressures

Estimates of Ko in such deposits can be obtained from a

knowledge of the over-consolidation ratio (OCR) using the

expression:

This equation is not applicable if the deposit has been

subsequently reloaded, as the effective stresses then follow the path BC and tend towards the initial loading path

Thus for a given soil deposit, K o can vary from location to

location depending on the stress history at each. Its value must

lie between Ka and Kp and its relative position between these

limits will govern the amount of movement required to

mobilize either.

'sin

OCR K K onco

20


21/86

Earth pressures

In normally consolidated soil, K o is slightly larger than K a .Little horizontal movement will therefore be necessary to

mobilize active earth pressure conditions, whereas significant

movements will be needed to mobilize passive conditions. In

a heavily over-consolidated soil, K o is larger & slightly less

than K p. The vice versa is then true.

21


22/86

Earth pressures

Earth pressure coefficients are dependent on

’ the effective angle of shearing resistance,

c’ the apparent cohesion, or c u the undrained shear strength,

the angle of wall friction,

cw the wall adhesion.

The values can usually be chosen by reference to the

borehole logs, standard penetration tests, and laboratorytests and descriptions.

The value of c’ is is influenced by the stress level of the test,

the rate of strain, the degree of weathering and the amount

of swelling experienced before the test. The values are onlyof the order of 0-10 kN/m². Unless the engineer is

confident about the value of c’ it is recommended that it be

taken as zero, as it can have a significant effect on the

design.22


23/86

Earth pressures

The values of and cw, however, is usually estimated by the engineer, andthe values chosen will have a significant effect on the earth pressure

coefficients, particularly for the passive case. For temporary steel sheet pile

cofferdams it is recommended that the maximum values of these

parameters should not exceed:

Note: c’ normally taken as zero

Where the toe of the wall penetrates into hard rock the above valuesshould be reduced by 50% for any overlying dense granular material or

over-consolidated clay. For overlying loose granular material the values

should be taken as zero. For anchor walls which have the freedom to move

upwards on mobilization of the passive pressure then zero values should be

taken.23


24/86

Earth pressures - at rest

In undisturbed ground the ‘at rest’ pressure is

K ov’

where:

K o = The ‘at rest’ pressure coefficient

& v’ = Effective vertical stress

Normally consolidated ground K o = 1-sin

’

24


25/86

The limiting active and passive pressures acting on the wall at any depth z aregiven by:

pa = K a v - K acc u Total horizontal pressure active

p p = K pv + K pcc u Total horizontal pressure passive

wherev = z + q Total vertical stress

= Bulk density

q = Any uniform surcharge on ground surface

K a = Active pressure coefficient (taken as 1.0 in this case)K p = Passive pressure coefficient (taken as 1.0 in this case)

c u = Undrained shear strength

cw = Wall adhesion

Earth pressures - Short-term, undrained, total stress analysis

2 1 w

ac

u

c K

c 2 1 w pc

u

c K

c

25


26/86

Active pressure cannot be less than hydrostatic water pressure, i.e.,tension crack will fill up with water. The depth of the tension crack will

be (2c u-q)/. The active water pressure will be wz where z = depth from

the surface.

Where tension cracks will not fill with water, and to take account of any softening of the clay, the total active pressure at any level should be

assumed to be not less than 5z kN/m². This is known as the ‘Minimum

Equivalent Fluid Pressure’ with the density of being 5 kN/m³.

For passive pressures it is recommended that a reduction factor isapplied to the undrained shear strength, c u, to allow for any general

softening of the clay during the period of the temporary work. In

addition c u should be taken as zero at excavation level and increased

linearly to its reduced value at a depth of one metre.

Earth pressures - Short-term, undrained, total stress analysis

26


27/86


28/86

It is recommended that c’ (& cw/c’) be taken as zero unlessthere is confident of the values.

The “pore” water pressure must be added to the effective

horizontal pressure to give the total horizontal pressureacting on the wall.

i.e. pa = pa ’ + u p p = p p’ + u

Earth pressures - Long-term, drained, effective stress analysis

28


29/86

Earth pressures - Mixed total and effective stress analysis

Mixed total and effective stress design can be appropriate for temporary works

design in stiff clays. Effective stresses with full water pressures are used for theactive pressures at the back of the wall, at least over the zone of potential tension

cracks, i.e. to a depth of (2c u-q)/.

Total stresses can be used below this zone provided that there is an active pressure

of not less than the equivalent minimum fluid pressure, 5z kN/m². Total stresses are

used for the passive pressures in front of the wall using appropriate undrained shear

strength values, c u, making due allowance for general softening of the clay. In

addition a reduction of c u to zero over in the first metre of ground below excavation

level should be made.

Pressure diagram for mixed

total & effective stress design

29


30/86

Design of wall members

Temporary works where traditional construction techniques

is adopted, only the long-term or permanent conditions need

be checked.

However, for basements excavation where the wall structural

member is subject to differing temporary and permanent

conditions, both cases must be considered.

30


31/86


1. Soil and groundwater pressures acting behind and in front of the wall

together with surcharge pressures from construction load, buildings or

roads.2. Reactions from the support systems, both temporary and permanent.

These forces may induce axial compressions or tensions due to

inclination of anchors or struts. CIRIA Report C517 provides design

guidance for temporary props based on extensive field measurementsfor a wide range of props & ground conditions.

3. Abnormal loadings, higher groundwater levels caused by flooding or

water-filled tension cracks, accident or construction surcharges. For

longer-term, it is only necessary to consider average or ambientloading conditions

4. Building permanent loads such as floor slabs and columns. These

forces may be eccentric to the wall and generate bending moments.

Applied wall forces are derived from the following:

31


32/86


BS 8110 & 5400 require load factors to be applied to service conditions @

1.4 and 1.5 respectively to obtain ULS values. BS 8110 states that this

factor can be reduced if the loads are derived from an elastic analysis. If

soil strengths are factored in order to derive the loads, in accordance withthe recommendations of CIRIA 580 & BS 8002, the requirements in the

structural codes are not appropriate. This somewhat confusing situation

is discussed more fully in Section A8.2.7 of CIRIA 580.

BS 8002 suggests that soil structure interaction calculations, modeling theSLS, can also be used to estimate the structural loads, and implies that

these loads are not factored to provide a ULS value. CIRIA 580,

recommends that the SLS values be multiplied by 1.35. It also

recommends to check both SLS and limit equilibrium, and that the value

used for the ULS structural design is the greater of the two values:

Bending moments and shear forces

the values derived using the factored soil strengths

the SLS values multiplied by 1.35.

32


33/86


When subjected to the complex loading from soil,

groundwater and structure, the wall structural member will,to a greater or lesser extent, deform. Wall stiffness often has

little influence on the total deformations, which are governed

primarily by soil conditions, the method and sequence of

construction and the wall support system.

All wall members will crack. However, as it is often the

primary defense against groundwater ingress, crack control is

necessary - BS 8007. Long-term durability also depends onthe severity of cracking.

Wall movements and cracking

33


34/86

DESIGN & DURATION OF LOADING

Temporary works may have to function for periods of one to two years. Thisis particularly significant for the design of temporary retaining system in

over-consolidated clay soils as they have a high short-term (undrained)

strength that reduces with time.

The strength of normally consolidated clays in the temporary condition isoften taken as undrained. But there are circumstances in which the drained

strength may be both appropriate and more critical; e.g. the calculation of

soil passive resistance beneath the base of an excavation.

For over-consolidated clay CIRIA 104 recommends that where the drainedstrength of the clay is taken a lower factor of safety should be used (as the

drained strength is lower & better defined). For temporary conditions a

mixed drained/undrained approach is suggested, where the retained clay is

treated as drained, but a reduced undrained strength (to allow for softening

of the soil immediately below the excavation) is used to calculate passive

resistance. This approach should only be used where there is considerable

experience of excavations within the clay stratum concerned or where the

temporary conditions are of short duration.

34


35/86

Computer programs for designing retaining systems

Analysis involves simplifications and idealizations. An

appropriate analysis is one that adequately models the

dominant effects without being overly complex.

One of the dangers of computer programs is that they areeasy to use without the user necessarily having an

understanding of the principles and idealizations on which

they are based. We must understand their principles and

limitations. Simple programs are often adequate for analyzing bending moments and shear forces in a wall, but

are likely to be inadequate for modeling ground movements.

Traditional methods which are based on a prescribeddistribution of earth pressures.

Deformation methods in which the earth pressure

distribution is computed as a function of ground and

structural stiffness35


36/86

Use of Computer Program

The use of a computer program is encouraged subject to the

following:

• the program should be validation in detail before general use

• the output must include all the input data necessary to carry

out an independent check

• use of the program must be by or under the direct supervision

of an experienced engineer.

Use of a computer program saves time. This also enables a number of

analyses to be made in a relatively short time, and thus the sensitivity of the

structure to changes in ground conditions, water levels and prop levels can beassessed. It is advisable to carry out an independent manual check on the

final computer analysis. It is recommended that true scale pressure diagrams

showing ground strata, water, excavation and strut levels are drawn by hand

if they are not part of the output from the program.

Last Lecture

36


37/86


Multi-prop walls

Analysis of walls with more than one level of props is complicated by the

complex soil/structure interaction arising from the construction sequence.

The ground is usually excavated in stages and the props are introduced at

each level. This modifies the behavior of the surrounding ground so that theclassical active state cannot develop.

The major concern is to assess the strut loads at each level of props.

Experience has shown that most failures occur due to overload in the props,

often accompanied by local web failures of the walings if they are notadequately stiffened. Failures by bending of the walls or walings are rare.

The positions of struts are usually selected to prevent excessive deflections

during construction of the cofferdam and to suit the construction sequence

for the works within it. We also need to check the overall stability of the

cofferdam and its base.

Last Lecture

37

U f C t P


38/86


Props are the most vulnerable elements in the retaining system

and are usually over-designed. Failures are extremely rare and

are generally caused by poor detailing, misjudgment of ground

conditions or accidents.

The failure of a prop could have serious consequences and might

lead to progressive collapse of the excavation. Buckling failures

also tend to be sudden.The cost of the propping system is usually small in comparison

with the cost of the retaining wall. While efficient design of the

propping system is to be encouraged, this is not an area in

which a major reduction in overall construction costs should be

expected.

38

D i f ll b


39/86


These programs are based on the simplest form of analysis. A

limiting condition is assumed and equilibrium applied to

obtain a solution. Programs are available to analyze gravity

walls, embedded cantilever and singly-propped walls. Active

and passive soil conditions are usually assumed with various

types of safety factor.

For multi-propped walls, empirically derived soil pressuredistributions are sometimes employed. These programs are

best used to obtain basic wall dimensions such as embedment

& estimate structural service loads. They give approximate

but unreliable solution to the complex deep basements

excavation, and therefore not recommended. Such programs

do not account for soil-structure interaction and cannot

estimate wall and/or soil movements.

Limit equilibrium programs – Traditional

39

D i f ll b


40/86


When considering the stability of gravity retaining walls or

anchored embedded walls, it is necessary to calculate slope

stability. Generally such analysis programs allow for both

circular and non-circular slip surfaces and using factored, or

mobilized soil strength.

It must be emphasized that all of the above computation

methods should be calibrated against case histories. Manyassumptions are required for the input parameters and even

finite-element analyses cannot be relied upon to give sensible

results without some calibration.

For simpler programs, calibration is even more important as

they are more limited than FEM in their ability to

extrapolate from a situation where the results are known to

another.

Limit equilibrium programs

40



41/86

Design of wall membersLimit equilibrium method for a of cantilever wall

The pressures at the toe of the pile have been replaced by a resultant force F3 at C some distance

above the toe. The forces F1 and F3 act through the centres of gravity of their respective areas.

The depth BC is found by assuming a level for C and calculating the moments for the forces F1and F3 about level C. This is repeated until the moments are in balance.

To correct the error caused by the use of the ‘simplified method’ the depth BC should be

increased by 20% to give the design penetration BD.

The maximum bending moment occurs at the point of zero shear at level X-X.

41



42/86

Design of wall membersNormalized depths of embedment at failure (after Bica and Clayton, 1998)

Dry sand øps is the peak value of the

plane strain angle of shearing resistance

øps degree

42



43/86

Design of wall membersLimit equilibrium method for a propped wall with free earth support

T the prop load, F1 and F2 act through the centres of gravity of their respective

areas. To calculate the penetration BD the depth of d is assumed and moments of

the forces F1 and F2 are calculated about the level of the prop T. This is repeated

until the moments balance.

The prop load T can then be found by balancing the forces, i.e.:

T = F1 - F2

& the maximum bending moment in the pile will be at the level of zero shear X-X.

43



44/86

Design of wall membersLimit equilibrium method for a propped wall with fixed earth support

Simplify ‘equivalent beam method’

The forces at the toe of the pile are replaced by a resultant force F3 some distance above the toe. Then

assume the point of zero bending moment (level Y-Y) occurs at the level where the active and passive

pressures balance i.e. the net pressure is zero.

T the prop load and F1, F2 & F3 acting through the centres of gravity of their respective areas.

The simplified pressure diagram show force F3 acts at the level of C. The prop load T can then be found bytaking moments about and above level Y-Y (assumed point of zero bending moment). Finally the depth of

C can be found by assuming its level and calculating the moments about and above this level. This is

repeated until the moments balance.

To correct the error caused by the use of the simplified method the depth BC is increased by 20% to give

the design penetration BD. The maximum bending moment in the pile occurs at level X-X, the point of

zero shear.

44

Design of Prop members


45/86

Design of Prop members

Limit equilibrium method

The design of the props will depend upon the analysis method adopted in

the calculations for the design of the wall. Prop loads calculated from

limit equilibrium calculations may be unconservative, as the effects of

soil-structure interaction are not included. In such circumstances, thecalculated prop loads should be increased by 85% to allow for the effects

of stress redistribution & arching behind the wall. Soil-structure

interaction methods that allow stress redistribution to model more

realistically the non-linear pressure profile behind the wall should providecalculated values of prop loads that better represent the particular project

circumstances modeled. Irrespective of the type of analysis undertaken,

the calculated prop loads should be checked for adequacy by comparing

them with those derived from comparable experience, Wherever possible,

this should be based on reliable field measurements from ease history data

in comparable conditions.

45

Type of analysisAdvantages Limitations


46/86

/software Advantages Limitations

Limitequilibrium

e.g. STAWAL,

ReWaRD

Needs only thesoil strength

Simple &

straightforward

Does not model soil-structureinteraction, wall flexibility &

construction sequence

Does not calculate deformations.

Hand calculations of deformations

possible by relating mobilized

strength, soil shear strain & wall

rotation (rarely done); or through

empirical databases

Statically indeterminate systems(e.g. multi-propped waits), non-

uniform surcharges & berms require

considerable idealization

Can model only drained (effectivestress) or undrained (total stress)

conditions

2-D only

Results take no account of pre-

excavation stress state46

Design of prop members


47/86


Traditional methods are quite simple to use and errorsleading to failure are very rare. One of the most widely used

methods, that proposed by Terzaghi and Peck (1967), is

compared with prop load measurements on site, and in most

cases the calculated loads are higher than those measured.

Will be discussed later in more details

47



48/86


With current computing power, more complex methods of analysis have been developed and are widely available. These

methods are collectively known as “deformation methods”.

In these methods the internal wall and support forces for the

temporary support system are calculated. The computed

support forces are often greater than those obtained from the

more traditional methods.

beam on springs

beam on elastic continuum

finite difference methods boundary element methods

finite element methods

48

Prop loads


49/86

Prop loads

In some cases deformation methods of calculation predictmuch larger prop loads than those which arise in practice.

Field measurements of 46 props from four sites are compared

with values calculated by deformation methods before theexcavations were made (Stroud et al, 1994).

The calculated values are generally 1.4 - 10 times the

measured values. There are a few props that fall outside thisrange. Five of these attracted little or no load at all and there

were six props for which the calculated load was 0.7 - 1.3

times the measured load.

49

Prop loads


50/86

All computation methods/programs should be calibrated against case histories

Prop loads

50



51/86


Retaining wall is represented by a beam & the soil and props

as a series of springs.

For embedded walls, a more realistic estimate is oftenneeded, and the calculation should take soil-structure

interaction into account. The simplest of these represent the

wall as a structural member usually employing a finite-

difference or finite-element approximation, with the soil as a series of unconnected springs.

The construction sequence is simulated by adding and

subtracting loads from the wall. Both structural stresses andwall movements are calculated. While such programs

represent a significant improvement over the simpler limit

equilibrium approaches, they still have severe imitations. For

example:

Beam-on-spring model programs

51



52/86


It is difficult to select appropriate spring stiffness to

represent the soil – require judgment & experience.

It is difficult to include the effects of any soil berm.

By representing the soil with a set of independent springs,

it is difficult/impossible to reproduce the observed stress

redistribution arising from wall flexibility.Generally do not allow for the influence of the release in

vertical stress caused by the process of excavation. Thus,

deep-seated movements arising from this process are not

included.

Only the wall movements are computed, difficult to

estimate the movements of adjacent Structures


52



53/86


These programs compute prop forces, earth pressures,

internal bending moments, shear forces and wall movements.

In comparison with field data, the computed values are oftenconservative, particularly in the case of prop loads.

The corresponding movements cannot be found with

sufficient accuracy from a simple elastic model of this type

and operators may manipulate the input in order to influence

the calculated deflections.

Movements predicted from this type of analysis should be

regarded with caution.


53



54/86

g

In this method the soil is represented as an elastic continuum

generated by interpolation of a pre-stored library of results

from finite element computations.

Beam on e1astic continuum programs

54

Type of analysisAd Li i i


55/86

Type of analysis


Subgrade

reaction/beam

on springs

e.g. WALLAP

Full soil-structure

interaction analyosis is

possible, modelling

construction sequence,

etc.

Soil modelled as a bed

of elastic springs

Soil-structure

interaction taken into

account

Wall movements are

calculated

Relatively

straightforward

Results take account of

excavation stress state

Idealisation of soil

behavior is likely to be

crude

Subgrade moduli can be

difficult to assess

2D only

Berms and certain

structural connectionsare difficult to model

Global effects not

modeled explicitly

Ground movements

around wall are not

calculated

55



56/86

In these programs the soil to each side of the wall is

represented by a boundary element. These programs

overcome most of the difficulties listed earlier apart from theestimation of the movements of adjacent structures. They

also involve many assumptions and simplifying idealizations:

Can give a good understanding of how the overall system behaves and which parameters are likely to control the

designs

May not give realistic displacement predictions.

Boundary element programs

56

T f l i


57/86

Type of analysis

/software

Advantages Limitations

Pseudo-finite

element

e.g. PREW

WALLAP

Full soil-structure

interaction analysis is

possible, modeling

construction sequence, etc

Soil modeled as an elastic

solid with soil stiffness

matrices calculated using a

finite element program

Soil-structure interaction

takes into account

Wall movements are

calculatedRelatively straightforward

Takes account of pre-


2D only

Limited to linear

elastic soil model, with

active & passive limits

Berms and certain

structural connections

are difficult to model

Global effects not

modeled explicitly

Ground movements

around wall are not

calculated

57



58/86

FEM in which the soil and wall are modeled as a mesh of

elements, In this way increasingly complex soil behaviour can

be modelled, e.g. as linear elastic/plastic and bi-linear elastic/plastic materials, Highly nonlinear behavior of soil

can also be included by more complex methods such as the

brick model.

FEA are complicated and must be used by experienced

engineers. It is also necessary to estimate the initial stresses

in the ground after wall installation. Wall installation tends

to reduce significantly the in-situ earth pressure coefficientK o, which is itself difficult to establish for heavily over-

consolidated soils.

Numerical analysis programs

58



59/86

At one particular site, the widely adopted approach, where

the installation effects are ignored and elastic/perfectly

plastic soil behavior is assumed, gave good predictions of movement, but prop loads were over-estimated. The

prediction of prop load improved with the effect of wall

installation included.

FEM is complex and cannot be regarded as infallible. Goodresults come from experienced engineers in well researched

soils. The selection of input parameters and the

interpretation of the results require care and experience.

FEM is of greatest benefit in the design of support systems

for excavations which, by reason of their size or complexity,

fall outside the range of available case histories.


59



60/86

Usually based on FEM and it is in principle possible to

analyze the complete 3D construction process from

temporary to permanent works, current limitations on

computing resources usually restrict analyses to 2D plane

strain or axis-symmetric sections. With such an approach, it

is possible to simulate the construction process and include

all significant structural members. Stresses, retrains andmovements both in the soil & structure can be predicted. The

effects on adjacent structures such as tunnels, sewers and

buildings can also be assessed.

The method is more expensive & requires detailed

information on soil properties, etc. Full FEA have become

more widely used. Nicoll Highway Incident, Fort Canning

Tunnel by NUS.


60



61/86

For design purposes a compromise approach is tocarry out a limited number of full numerical

analyses in combination with simpler computations.

The FEM analyses are then used to calibrate the

simpler approaches, which is then used to assess the

effects of design modifications.

Once the design is finalized, it may be necessary to

carry out a few additional full numerical analyses to

check the adopted solution.

In Singapore, critical excavation is always by FEM

as computation resource only limits 3D analysis. So

it is 3D to calibrate 2D FEA.61

Design of wall & prop members


62/86

Deformation methods as well as traditional

methods, should be calibrated against field

measurements. The case histories summarized in

CIRIA 517 provide a frame of reference against

which analytical results can be judged.

Where calculated prop loads obtained analytically

differ significantly from those predicted from case

histories by the empirical method given. The

physical reasons for the differences should be

identified and explained as part of the design report

which includes the analyses.

62

Type of analysis



63/86

Finite element &

finite difference

e.g. SAFE (2D FE)

PLAXIS (2D & 3D

FE)

CRISP (2D & 3DFE)

FLAC (3D & 3D

FD)

ABAQUS (3D FE)

DYNA (3D FE)

Full soil-structure interaction

analysis is possible, modeling

construction sequence etc

Complex soil models can

represent variation of

stiffness with strain &

anisotropyTakes account of pre-


Can model complex wall and

excavation geometry

including structural &support details

Wall and ground movements

are computed

Potentially good

representation of pore water

response

Can model consolidation as

soil moves from undrained to

drained conditions

Can carry out 2D or 3D

analyses

Can be time-consuming to set up

& difficult to model certain

aspects, eg wall installation

Quality of results dependent on

availability of appropriate stress

strain models for the ground

Extensive high-quality data (eg pre-excavation lateral stresses as

welt as soil stiffness and strength)

needed to obtain most

representative results

Simple (linear elastic) soil modelmay give unrealistic ground

movements

Structural characterization of

many geotechnical finite element

and finite difference packages

may be crude

Significant software-specific

experience required by user

Basic representation of pore

water response

63

TEMPERATURE EFFECTS


64/86

Designers noma1ly ignore temperature effects in

props for flexible walls, but account for them in

props for stiff walls

Where the effects of temperature are explicitly

considered, it is usual to follow the guidance given

in BS5400: Part 2 also referenced in Bridge Design

Standard BD42/94. Designers either choose a small

temperature range, e.g. 10 C, but assume the prop to

be completely restrained, or they adopt a larger

temperature range, e.g. 30

C, but assume thetemperature effect to be only 50% cent of the fully

restrained value.

64

TEMPERATURE EFFECTS


65/86

Reductions in temperature below prop installation

temperature will normally remove the effect of any pre-load,

to zero in some instance.

Envelopes produced by Peck (1969a) are based on max.

measured loads with some temperature effects.It is not usual for deformation methods of analysis to include

temperature effects, although this is possible. Temperature

effects are normally added to the predicted prop loads after

the analysis is complete.

Safety factor is lower when temperature effect is to be

considered (1.2). Load caused by effect of temperature range

are passive action, as the load increase in caused by prop pushing against the ground, unlike persistent acting earth

pressure. Any yielding of strut system will release such load.

65

Prop members – Temperature effects on propsChange in temperature of a prop from its installation temperature will cause


66/86

g p p p p

it to expand or contract according to the relationship: L = t L

where:

L = change in prop length

= thermal coefficient of expansion for the prop material t = change in prop temperature from the installation temperature

L = prop length.

If prop is restricted from expanding freely, additional load is generated in

the prop. The magnitude of this additional load is:

Ptemp = × t×E×A×(/100)

where:

E = Young’s modulus of the prop material

A = cross-sectional area of the prop

= percentage degree of restraint of the prop. (70% for stiff walls in stiff

ground & 40% for flexible walls in stiff ground).

Select the appropriate value of

to suit particular project circumstances &

on the basis of comparable experience.66

Prop members – TEMPERATURE MEASUREMENTS in S’pore


67/86

Measured Strut Loads and Temperature Change with Time

MRT North East Line Dhoby Ghaut station

J. X. Niu et al

67

Prop members – Temperature effects on props


68/86

E, Young’s modulus of steel, is equal to 205x106 kN/m2 and

Thermal coefficient of expansion for steel and is equal to 12x10-6/oC

* Rigidity of strut system should also be considered. We should consider effective E*

Load Increase in Struts of Various Sizes in Temperature Change of 10oC.

68

Prop members – TEMPERATURE MEASUREMENTS in S’pore


69/86

Measured Wall Movements and Temperature Change with Time

MRT North East Line Dhoby Ghaut stationJ. X. Niu et al

69



70/86

For multipropped walls, as it is difficult to predictthe pressure distribution acting on the back of a

wall, empirical methods were developed, eg

Terzaghi and Peck (1967). These methods are basedon the results of actual field measurements and take

account of the method of construction

70

Apparent Earth Pressure Diagrams


71/86

Sand Soft/medium clay Stiff clay

Pressure envelope method — Terzaghi and Peck

Last Lecture

71



72/86

Last Lecture

72

Design of prop members – CIRIA C517 CLASSIFICATION


73/86

Excavation classified on the basis of the type of ground:Class A- normally and slightly overconsolidated clay soils (soft and firm

clays)

Class B - heavily overconsolidated clay soils (stiff and very stiff clays)

Class C - granular/cohesionless soilsClass D - walls retaining both cohesive and cohesionless soils (mixed soils).

For firm & soft clays (Class A) and stiff clays (Class B) these have been sub-divided

according to wall type based on the wall stiffness:

Flexible (F) walls - timber sheet pile and soldier pile/king post walls

Stiff (S) walls - contiguous, secant and diaphragm concrete walls.

Flexible walls retaining soft clay soil (Class AF) have been further sub-divided

according to base stability conditions into “stable” and “enhanced stability” cases.

Walls in granular soils (Class C) are sub-divided into “dry” and “submerged” cases.

The classification is denoted by its reference number. For example, BF3 is a case

history for an excavation in an over-consolidated clay (B) supported by a flexible wall

(F). Within each classification the number increases with excavation depth, e.g. AF1

is the shallowest excavation and AF28 the deepest.73



74/86

Variations of total prop load within an excavation (after Flaate and Peck, 1973)

74

Design of prop members – CIRIA C517 CLASSIFICATION


75/86

Variations between the loads in props within the sameexcavation are very significant, 30%-60%. It follows that:

1. Designer of a propping system should expect large

variations in load within the temporary propping system.The entire system should redistribute load from any

overstressed element. The eventual collapse mechanism

should be ductile (yielding under constant Load) rather than

brittle.

2. Maximum load at a prop level is much higher than the

average prop load within the excavation at that prop level.

3. Measurement of loads in only one or two of the props

within an excavation may markedly under-predict the actual

maximum prop load.

75

Prop members – TEMPERATURE MEASUREMENTS

Temperature changes have little effect on prop loads, at least for props


76/86

supporting flexible walls & soft ground.But for props supporting stiff walls in stiff (Class B and C) soils,

temperature effects are normally considered in design.

It is concluded that temperature effects should be allowed for by making

simple checks on the structural members selected as props, but that anincrease in characteristic prop loads may not be required, e.g. safety factor is

normally lower if temperature range ( 5-7%) is considered.

Prop temperature is similar to the ambient air temperature.

76

Prop members – PRELOADING

Props can be preloaded either by jacking between the prop and the waling. It


77/86

Props can be preloaded either by jacking between the prop and the waling. It

is often applied to flexible king post and lagging systems to take up the

initial slack following installation. Preloading is a “significant” additional

cost.

The preload applied to take up slack in the support system is typically

around 10% of the design (working) load. In some cases additional preloadof 50-80% of the design load was applied to stiffen the support system in an

attempt to reduce wall deflections. Preloads of that order will result in the

wall being pushed back into the ground when they are applied.

Preloading does not reduce the variation in loading between individual props. Temperature changes can vary the prop load, & in some cases, fall in

temperature can completely eliminate the effect of preloading.

Preloading a prop with a higher load than is needed to take up slack in the

support system can result in a higher prop load than would otherwise berequired. Most engineers consider that there is little benefit in introducing

additional load in this way.

77

Prop members – PRELOADING


78/86

Loss of preload due to a reduction in prop temperature

78

Analysis using the distributed prop load methodProp loads are currently determined either by traditional methods or by


79/86

Method for calculating the distributed prop load

deformation methods, but the reliability of any method can only beevaluated by comparison with reliable field measurements.

Methods of predicting prop loads directly from field measurements were

1st developed by Terzaghi and Peck (1967).

79

Analysis using the distributed prop load method


80/86

Some modifications are proposed and predictions are nowgiven in terms of characteristic values used in limit state

codes, such as BS8110, BS5400, Eurocodes and BS5950. In

limit state terminology the characteristic value is a cautious

estimate and in statistical terms is the one which has only a 5% probability of being exceeded.

There are a number of factors that require further

consideration:

1. Duration of loading

2. Temperature effects3. Base stability (for excavations in soft clays).

80

REVIEW OF PECKS RECOMMENDATIONS

The principal findings are:


81/86

1. The envelopes originally proposed by Peck did not enclose all

the data on which they were based, but covered within the

factor of safety of the prop.

2. Loads from the apparent earth pressure envelopes is not thecharacteristic loads in a limit state calculation.

3. Case histories available confirm that Peck’s tentative

recommendations for stiff walls in stiff clays are not

conservative.4. The widely used practice of taking the buoyant weight of the

soil along with a careful estimate of the possible water pressure

was found to give reasonable agreement with the measured

loads & a modified envelope for granular soils is put forward onthat basis.

5. No simple conclusion was reached in respect of Class D (mixed

soils) sites.

81

Class B, C & D soilsFor Class B soils (heavily overconsolidated clay) the negative pore


82/86

water pressures decrease towards equilibrium values with time & theeffective stresses, hence soil strength reduces.

Pore pressure changes in cohesionless soils (Class C) occur rapidly.

They would not gradual variations in load with time. In certain

circumstances the creep of granular soils can be sufficient to cause build-up of load with time.

Mixed soils profiles (Class D) would only be expected to have load

variations with time if there were sufficiently thick layers of clay

that would not drain quickly. Cases D14a to c include the presenceof a 4 m thick stratum of what is predominantly clay within a

generally cohesionless soil profile, but there were no significant prop

load increases over a 12-month period.

Observations indicate that significant increases in prop load can

occur with time in excavations within Class B and Class D soils, but

that this is not always observed and that increases in load may be

minimal.

82

Normalized DPL diagrams according to the classification


83/86

are as follows:

Soft clays & flexible walls with enhanced base stability

(Class AF)

Soft clays & flexible walls with stable bases (Class AF)

Firm clays & flexible walls (Class AF)

Soft clays & stiff walls (Class AS)

Stiff clays & flexible walls (Class BF)

Stiff clays & stiff walls (Class BS)Granular soils (Class C)

Layered cohesive and granular soils (Class D).

83

Characteristic distributed prop load diagrams for Class

A, Class B & Class C soils


84/86

Class A soils (soft to firm clays)

For Class AS, tentatively as Class AF

84




85/86

Class B soils (stiff to very stiff clays)

85




86/86

Class C soils (granular Soils)

86

Deep basements & cut & cover - 3

Documents