SUBGRADE CRITERIA FOR AIRPORT FLEXIBLE PAVEMENT DESIGN By Dr. Marshall R. Thompson 1 , Member, ASCE and Manuel O. Bejarano 2 , Student Member, ASCE ABSTRACT Current mechanistic-empirical airport pavement design procedures use Elastic Layer Programs (ELP) to predict pavement responses (deflections, stresses, strains) generated by the gear load. The procedures incorporate subgrade strain criteria for controlling pavement rutting. WESICE/F.AA and Al vertical compressive strain criteria were developed from ELP analyses of pavement sections per the revised CBR equation. The limited scope of pavement test sections and performance data required extrapolation to other subgrade, loading and climatic conditions. The large and varying rutting criteria used to interpret the test section performance data are not consistent with the more rigorous criteria generally associated with high type airport pavements. A subgrade stress ratio (SSR = repeated deviator stress /soil strength) approach is presented in this paper. The University of Illinois (U of IL) SSR criteria ensure the pavement exhibits - stable" subgrade permanent deformation performance_ Subgrade rutting is 1 Professor Emeritus, Dept of Civil Eng, Univ. Of Illionis at Urbana-Champaign, Newmark Civil Engineering Lab, 205 N. Mathews Ave., Urbana, IL 61801 22 Graduate Research Assistant, Dept of Civil Eng, , Univ. of Illinois at Urbana-Champaign, Newmark Civil Eng. Lab , 205 N Mathews Ave , Urbana, IL 61801
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
SUBGRADE CRITERIA FOR AIRPORT FLEXIBLE PAVEMENT DESIGN
By Dr. Marshall R. Thompson1, Member, ASCE and Manuel O. Bejarano2, Student Member,
ASCE
ABSTRACT
Current mechanistic-empirical airport pavement design procedures use Elastic
Layer Programs (ELP) to predict pavement responses (deflections, stresses,
strains) generated by the gear load. The procedures incorporate subgrade strain
criteria for controlling pavement rutting. WESICE/F.AA and Al vertical
compressive strain criteria were developed from ELP analyses of pavement
sections per the revised CBR equation. The limited scope of pavement test
sections and performance data required extrapolation to other subgrade, loading
and climatic conditions. The large and varying rutting criteria used to interpret the
test section performance data are not consistent with the more rigorous criteria
generally associated with high type airport pavements.
A subgrade stress ratio (SSR = repeated deviator stress /soil strength) approach
is presented in this paper. The University of Illinois (U of IL) SSR criteria ensure
the pavement exhibits -stable" subgrade permanent deformation performance_
Subgrade rutting is controlled by limiting SSR to acceptable levels, depending on
traffic The WES/CE/FAA strain criteria expressed in SSR terms are below about
0.4 These SSRs are very conservative and result in increased pavement thickness.
Permissible SSRs for airport subgrades are probably in the range of 0.5 to 0.7.
Load pulse characteristics (stress level and duration) of multiple-wheel gear
configurations and stress history effects (sequence of stress level applications) on
subgrade permanent deformation accumulation need to be further considered in
implementing the SSR concept.
1 Professor Emeritus, Dept of Civil Eng, Univ. Of Illionis at Urbana-Champaign, Newmark Civil Engineering Lab, 205 N. Mathews Ave., Urbana, IL 6180122 Graduate Research Assistant, Dept of Civil Eng, , Univ. of Illinois at Urbana-Champaign, Newmark Civil Eng. Lab , 205 N Mathews Ave , Urbana, IL 61801
INTRODUCTION
Current mechanistic-based airfield pavement design procedures such as
the US Army WES (Barker and Brabston, 1975), Asphalt Institute (1987),
Army/Air Force Technical Manual (Departments of the Army, and the Air Force,
1989), and LEDFAA (FAA, 1995) use Elastic Layer Programs (ELP) to predict
the structural response of flexible pavements to applied aircraft gear loads. These
pavement responses (strain, stress, deflection) are related to pavement
performance (asphalt concrete (AC) fatigue cracking and pavement rutting) using
transfer functions. The most common existing transfer function relates pavement
life (number of load repetitions) to subgrade vertical compressive strain. Subgrade
strain is the controlling factor for most WES/CE'LEDFAA designs. However,
pavement sections designed using these criteria tend to be conservative and
limited insight is gained concerning the mechanisms of pavement distress
development. It is necessary to improve these subgrade criteria to produce more
functional, reliable and cost-effective airfield flexible pavements. This paper
presents and analyzes current subgrade strain criteria for airfields, introduces the
Suberade Stress Ratio (SSR) as an alternative, and demonstrates the influence of
new aircraft large gear configurations (e.g., the Boeing 777 tridem) on the
repeated loading behavior of cohesive soils.
CURRENT SUBGRADE DESIGN CRITERIA
The WES Subgrade Design Criteria
In the early 1970's the US Army Waterways Experimental Station (WES)
began the development of a pavement design procedure based on elastic layered
theory. Barker and Brabston's report (1975) presents the procedure. The WES
procedure was implemented as a design manual for the Departments of the Army
and the Air Force (1989). The Federal Aviation Administration (FAA) computer
program LEDFAA (FAA, 1995) is a modification of the WES program and is the
design procedure for pavements serving the Boeing 777. In general, the WES
procedure incorporates subgrade strain criteria for controlling pavement rutting. It
assumes that surface rutting is mainly due to subgrade shear deformation implying
that negligible permanent deformation accumulates in the pavement layers above
the subgrade. Subgrade rutting is controlled by limiting the vertical compressive
strain at the top of the subgrade. The subgrade strain criteria were developed from
ELP analyses of idealized conventional pavement sections subjected to typical
aircraft gear loads. In the development of the subgrade design criteria Barker and
Brabston (1975) indicated that:
".., it was desired to use a group of pavement sections which covered a range
of design conditions. The design parameters which were to be varied were the
subgrade modulus, the design aircraft and the number of load repetitions. The
variations and number of pavement sections required preclude the direct use of
test section data However, since the present CE and FAA thickness design criteria
represent a statistical treatment of test section data, it was possible to use the CE
and FAA procedures to generate idealized pavement sections. For various
loadings of aircraft with single-wheel, dual-wheel, or dual-tandem gears, this
procedure was used to generate pavement sections which would perform
satisfactorily at 1,200, 6,000, and 25,000 annual departures on 3-, 6-, 10-, and 20-
CBR subgrades."
The WES development of the design criteria included the following
assumptions:
Traffic is represented in number of operations of the fully loaded design
aircraft.
Loads are essentially static, and the load in each tire is circular and uniform.
The pavement is a linear elastic layered system with full friction between
interfaces.
The bottom layer is of infinite thickness.
The deformation characteristics of the pavement materials are represented by
the modulus of elasticity and Poisson's ratio as determined in a repeated load
test.
The AC modulus and Poisson's ratio were selected as 1380 MPa (200 ksi) and
0.5 respectively.
Granular base and subbase layer were subdivided in sublayers. The modulus
of each sublayer was determined based on the sublayer thickness and the
modulus of the material below the sublayer.
The subgrade modulus is related to CBR by the equation Es (NiPa) = 10.3
CBR (Es (psi)=1500 CBR).
Strains were computed using the elastic layer program CHEVIT (Barker &
Brabston, 1975). The WES report does not indicate how many pavement sections
and aircraft gear loadings were analyzed. The WES strain criteria are presented in
Figure 1 and are algebraically expressed by the equation:
N = 1000 x ( 0 . 000247+0 .000245 x log10( Es)
εv )0. 0658 x Es 0 .559
where N = Allowable repetitions
ev = Vertical strain at the top of the subgrade
Es = Modulus of the subgrade, psi
The design criteria were validated using WES test data presented by
Ahlvin et al., (1971) The Multiple Wheel Heavy Gear Load (NIWHGL) tests
weere conducted on pavement sections with an AC layer 7.5 cm. (3 in.) thick and
a crushed stone base 15 cm. (6 in.) thick. The subbase layer was a non plastic
gravel-sand with thicknesses ranging from 15 to 83 cm. (6 to 33 in.) placed over a
4-CBR subgrade. The subgrade was 91 cm (36 in.) of a heavy clay (known as
Vicksburg Buckshot Clay -- VBC) placed on top of the natural soil (a lean clay).
Traffic loadings included single-wheel, dual tandem (i.e., B-747) and 12-wheel (i
e, C-5) gear loads.
Figure 1. WES/CB/FAA Subgrade Strain Criteria
The failure criteria for the WES-N1WHGL tests were one inch heave above
the pavement surface or severe surface cracking. Failure criteria established in this
manner may be indirectly or not related with subgrade shear structural failure.
One inch heave may be due to large subgrade shear displacement and severe
disruption of the different pavement layers (Lane et al., 1993). Surface rutting was
not considered to be a critical factor in judging pavement failure. However,
measured surface rutting at failure in the WES-MWHGL test sections were not
constant and varied with pavement thickness and gear load (Chou, 1977). Rut
depths at failure were in the range of 13 to 90 mm. (0.5 to 3.5 in.) (Ahlvin et al.,
1971). Results of the full-scale tests are presented in Table 1. From these data, the
relationship in Figure 2 was developed. Barker and Brabston (1975) noted that
none of the test data extended to higher traffic levels. They concluded that the test
data do not represent a complete verification of the subgrade criteria but that an
extrapolation of the criteria to higher traffic levels is justifiable.
Table 1 Summary of MWHGL Test Section Data (Barker & Brabston, 1975)
Figure 2. Verification of the WES/CET:VA Subgrade Strain Criteria
Pavement failure (as related to subgrade strain criteria) is based on cumulative
damage according to Miner's hypothesis. In the Army/Air Force Design manual
(19S9) and the LEDFAA program (FAA, 1995) the damage factor is calculated as
the ratio of applied traffic repetitions (n) of a single aircraft to allowable load
repetitions to failure (N) of the aircraft. The subgrade cumulative damage factor
(CDF) is the sum of the damage factors for the various aircraft, thus the final
pavement design represents variations in applied loads (mixed traffic conditions).
Failure is predicted when the CDF reaches a value of one. (Barker & Brabston,
1975; Barker & Gonzalez, 1991; FAA, 1995).
The Asphalt Institute (AI) Subgrade Design Criterion
Witczak (1972) proposed vertical subgrade strain criteria for the design of
Full-Depth Asphalt Pavements. Witczak's failure criteria are to limit the vertical
strain on top of the subgrade (evaluated at a critical temperature of the asphalt
concrete layer) for a given number of strain applications. The theoretical study
was developed from an analysis of the revised CE CBR design thickness equation
(Ahlvin et al , 1971) for flexible pavements subjected to MWHGL aircraft traffic.
Witczak indicated that the revised CBR equation translates into allowable strains
(greater than those calculated by the earlier CBR equation) approaching
asymptotic values for high levels of traffic. Therefore, for multiple wheel aircraft,
flexible pavement thickness requirements approach a constant value for a large
number of load repetitions
A DC-8-63F aircraft was used to analyze the effect of subgrade type and
repetition level on permissible maximum vertical subgrade strain. It was
concluded that for all practical purposes aircraft type and subgrade modulus
yielded insignificant differences in the permissible maximum vertical strain –
strain repetition relationship.
In Witczak's criteria, a permissible vertical strain of 1460 microstrain is used
at lx106 repetitions at a limiting AC modulus of 690 NIPa (100 ksi) Suggested
thickness reduction percentages of 95, 85, 70 and 50 approximate thickness
requirements for repetitions of 105, 104, 10' and 102, respectively. Environmental
effects are considered by a thickness adjustment factor: "a factor of 1.0 is applied
for high annual air temperature conditions and 0.9 for cooler environments."
SUBGRADE STRAIN CRITERIA COMMENTS
In highway flexible pavement mechanistic-based design procedures, it has
been customary to express subgrade design criteria in the form:
εv = L(1/N)"
where N = Permissible number of ESAL's
εv = Subgrade compressive vertical strain
L, m = Empirically developed parameters
Typical highway criteria (National Cooperative Highway Research Program
Report, 1990) show that the "L" parameter ranges from 1.0x 10-2 to 2 8x 10.2 and
in some procedures "L" varies with design reliability. The "m" parameter varies
between 0.22 and 0.28. The philosophy of the subgrade vertical strain criteria is to
control pavement rutting (in the range of 10 to 20 mm. (0.4 to 0.6 in.)) by limiting
subgrade resilient strain for a specific number of load repetitions
Laboratory test data (Townsend & Chisolm, 1976) on the %/BC subgrade
(presented in Figure 3 for 1000 load repetitions) show that subgrade resilient
strains are not uniquely related to permanent strain accumulation for a given
number of load repetitions. For a given resilient strain, low strength soils
experience larger permanent deformation than higher strength soils. In the WES
criteria, the allowable number of strain repetitions is a function of the resilient