Académique Documents
Professionnel Documents
Culture Documents
Geot 1996 46 3 383
Geot 1996 46 3 383
3, 383±426
Application of soil mechanics principles to the L'application des principes de la meÂcanique des
design of pavement foundations, the design of sols aÁ la conception des couches de fondation et
complete pavements and to their structural de chausseÂes entieÁres, ainsi qu'aÁ l'eÂvaluation
evaluation `in-service' has lagged some way structurale des chausseÂes en service a un certain
behind knowledge accumulated through re- retard sur l'eÂtat actuel des connaissances. Les
search. Present design methods are generally meÂthodes de conception actuelles tendent aÁ eÃtre
empirical and often based on use of the empiriques et reposent souvent sur l'essai de
California Bearing Ratio test, which was aban- portance californien, que la Californie elle-
doned in California some ®fty years ago. meÃme n'utilise plus depuis une cinquantaine
The soil mechanics problem is one of under- d'anneÂes. Le de® qui se pose aÁ la meÂcanique des
standing how soils and granular materials sols est de comprendre la reÂaction des sols et
respond to repeated loading and applying this des mateÂriaux granulaires aÁ des charges reÂpeÂ-
knowledge to pavement design with the aid of teÂes et d'appliquer les lecËons qu'on en tire aÁ la
appropriate theoretical analysis and an under- conception des chausseÂes en s'appuyant sur une
standing of failure mechanisms. Non-linear analyse theÂorique judicieuse et sur la compreÂ-
stress-strain characteristics are a particular hension des meÂcanismes de rupture. Un aspect
feature of the problem and have to be catered particulier de cette question est les caracteÂris-
for in design and evaluation. Various `tools' are tiques de tension-deÂformation non lineÂaires,
available to assist the pavement engineer. These dont il faut tenir compte dans la conception et
include theoretical analysis, laboratory testing l'eÂvaluation. L'ingeÂnieur des chausseÂes dispose aÁ
apparatus, ®eld testing and full-scale trials with cette ®n d'une panoplie d'outils: analyse theÂori-
appropriate instrumentation. que, essais en laboratoire, essais sur le terrain
The resilient and permanent strain response et essais en vraie grandeur sur chausseÂes in-
of clays and granular materials is reviewed in strumenteÂes. On examine ici les deÂformations
the context of the requirements for design. The eÂlastiques et permanentes d'argiles et de mateÂ-
essentially empirical UK Highways Agency riaux granulaires dans le contexte des criteÁres
method of design and its evolution are discussed de conception. On analyse aussi la meÂthode de
in the light of current soil mechanics knowledge. conception, essentiellement empirique, utiliseÂe
By contrast, the development of mechanistically par l'administration routieÁre du Royaume-Uni
based approaches is outlined, together with (UK Highways Agency), ainsi que son eÂvolution,
suggestions for the implementation of present aÁ la lumieÁre de ce que l'on sait actuellement sur
knowledge in a practical method of design for la meÂcanique des sols. Par opposition, l'exposeÂ
pavement foundations. Parallels are drawn deÂcrit brieÁvement les meÂthodes meÂcanistes qui
between road and rail track design and devel- ont eÂte formuleÂes et preÂconise l'application des
opments relating to the latter are also reviewed. connaissances actuelles dans une meÂthode pra-
A forward look suggests that further research to tique de conception des couches de fondation. Il
improve understanding of the effective stress compare la conception des routes et celle des
state below pavements and the application of voies ferreÂes et examine les progreÁs reÂaliseÂs
recent theoretical work on partially saturated dans ce dernier domaine. Pour ce qui est de
soils could form the basis for improved pave- l'avenir, le geÂnie routier aurait fort aÁ gagner de
ment engineering in the future. recherches plus pousseÂes sur les tensions ef®-
caces dans le sous-sol, ainsi que de l'application
KEYWORDS: pavements and roads; design; repeated de reÂcents travaux theÂoriques sur des sols
loading; clays; ®eld instrumentation; deformation; partiellement satureÂs.
laboratory tests.
INTRODUCTION
Pavements are civil engineering structures used for
Professor of Civil Engineering, University of Notting- the purpose of operating wheeled vehicles safely
ham, UK. and economically. There is a very wide range of
383
Downloaded by [] on [12/02/21]. Copyright © ICE Publishing, all rights reserved.
384 BROWN
pavement structures depending on the nature of the proper use of geosynthetics for drainage, ®ltration,
vehicles to be accommodated, the wheel loads separation and reinforcement is also important, as
involved and the numbers of such loads to be are the uses of other ground improvement tech-
carried over a given time period. Fig. 1 shows niques such as stabilization and the effects of frost.
cross-sections for a number of different pavement Increasingly, the profession is becoming con-
types ranging from unsurfaced `gravel' roads cerned with the evaluation and maintenance of
commonly found in developing countries, to heavy pavements rather than the design and construction
duty ¯exible bituminous or rigid concrete pave- of new works. Structural evaluation and the design
ments used for the motorway systems of indus- of remedial treatment are rapidly expanding
trialized countries. Railway track is included as activities in which the role of soil mechanics,
another specialist type of pavement in which while not as signi®cant as for new construction, is
the method of transmitting load to the soil differs nonetheless of considerable importance. This paper
from a highway or airport pavement but for which does not attempt to cover all aspects of pavement
the essential principles of soil mechanics equally engineering or to consider all pavement types.
apply. Rather, it concentrates on the role of soil mech-
It is clear from the structures in Fig. 1 that the anics and, hence, on the pavement foundation,
scope of pavement engineering is rather wide. An which can broadly be de®ned as one or more
essential ingredient is soil mechanics since all the layers of compacted unbound granular material
structures are in intimate contact with the ground placed over the subgrade soil (Fig. 2). The soil
and most combine one or more layers of unbound may be either undisturbed, in cuttings, or re-
granular material. In addition, the complete pave- moulded, on embankments. Since the interaction
ment engineer needs to understand the principles between the foundation and the bituminous or
of asphalt mechanics, of concrete technology and concrete construction placed over it is central to
of stabilization as well as the more complex pavement design and to structural evaluation
structural con®gurations used in railway track. procedures, some discussion of bound materials
The engineer needs to be concerned with vehicle and of failure mechanisms is required. This allows
loading, whether from trucks, aircraft, container the proper application of soil mechanics principles
terminal traf®c or railway rolling stock and with for the foundation to be put in its correct context.
the in¯uence of the environment (notably tempera- In order to do this, only bituminous construction
ture and water) on the pavement structure. The will be considered. Reference is made to rail-track
construction but this topic is adequately covered by
Bitumen seal
Selig & Waters (1994).
The paper presents the design problem, explain-
Granular Granular ing the background to current, essentially empiri-
Soil Soil
cal, practice and presenting the most signi®cant
(a) Gravel road
results of research carried out since the mid-1950s
(b) Sealed gravel road
with emphasis on recent developments. Discus-
sion covers design philosophy, theoretical analysis,
material properties, laboratory tests, ®eld testing,
Asphaltic Concrete pilot-scale experiments and extensions of present
Granular Granular knowledge to engineering practice.
Soil Soil
Although pavement engineering procedures vary
(c) Asphalt pavement
somewhat around the world, the essential features
(d) Concrete pavement
of present practice are generally common and
rather empirical. In looking at the details and how
Asphaltic Concrete
Cement treated or concrete Cement treated
Granular Granular Moving wheel
Load Load
Soil Soil
Base Ballast
Concrete or brick
blocks on sand Rail on sleepers
Sub-base Sub-ballast
Asphaltic or cement treated
Foundation
Ballast (granular) Subgrade
Subgrade
Granular
Sub-Ballast (granular)
Soil (a) (b)
Soil
(g) Block pavement (h) Railway
Fig. 2. De®nition of pavement foundation: (a) road;
Fig. 1. Cross-sections of various pavement types (b) rail track
is fewer but the stress levels are much higher ments into two parts rather than to apply a
as shown in Fig. 3(b). These data were ob- single elasto±plastic (or elasto±visco±plastic)
tained from a haul road experiment at the analysis as is common for monotonic loading
Bothkennar soft clay site in Scotland (see problems in geotechnics.
p. 392).
(d) Under a single application of a moving wheel Other branches of soil mechanics which have
load, a pavement responds in an essentially something in common with pavements include
resilient manner. However, irrecoverable plastic earthquake engineering, off-shore foundation engi-
and viscous strains can accumulate under neering and machine foundation design. In all
repeated loading. This presents the opportunity cases, there is the common theme of repeated or
to separate the theoretical analysis of pave- cyclic loading. It is signi®cant to note that the late
Professor Harry Seed, who contributed so much to
20 understanding soil mechanics in the context of
earthquake loading, began his interests in repeated
loading with the pavement problem (Seed et al.,
1955). Furthermore, when, the ®rst major struc-
tures were being designed for North Sea oil
15 exploitation in the 1970s, the experience of re-
peated loading of subgrades was used to evolve
research programmes for clays of relevance to
wave loading (Andersen et al., 1976).
Vertical stress: kPa
80
Space has prohibited any discussion in this
paper of several important subjects including drain-
60
age, application of geosynthetics, stabilization,
frost effects and the special problems associated
40 with aircraft pavements.
20
PAVEMENT ENGINEERING TOOLS
0 Theory
0 0.2 0.4 0.6 0.8 1.0 There has been extensive application of the
Time: s theory of elasticity to the analysis of layered
(b) pavement systems. Burmister (1943) developed
Fig. 3. In situ vertical stress measurements in sub- the essential equations, and, following early sets
grades: (a) below 165 mm asphalt construction at of tabulated solutions, (e.g. Acum & Fox, 1951)
Wake®eld; (b) below 350 mm granular layer at various computer programs were developed to
Bothkennar assist in obtaining results in a convenient form.
Time: s
8 7 6 5 4 3 2 1 0
Deviator
stress
Axial
deformation Elastic
deformation
Fig. 7. Response of overconsolidated silty clay to bursts of undrained repeated loading (after Hyde, 1974)
Pressure
600 kPa
0
Asphalt γ 5 23 kN/m3
0.05 Linear elastic E 5 2000 MPa
ν 5 0.35
0.10
Sub-base
0.20 γ 5 21kN/m3
s 5 3 kPa
Ko 5 1.0
0.30 K-θ model, equation (22)
K1 5 8000
K2 5 0.70
(σ in kPa)
ν 5 0.30
0.50
0.70
1.00
Subgrade
γ 5 20 kN/m3
1.30
Brown's model, equation (6)
K 5 50 MPa
n 5 0.40
ν 5 0.45
1.60
2.00
z: m
Half-space with E 5 200 MPa and ν 5 0.45
Rail
Sleeper
Wheel
load
Chord Modulus
Ballast
Subballast
Overburden
Subgrade layer 1
Subgrade layer 2
Secant moduli
Bedrock
Strain
Fig. 11. Geotrack model for railtrack (after Selig &
Fig. 9. De®nition of resilient strain in ®nite element Waters, 1994)
computations
included much instrumentation, but have relied
τ on super®cial observations and measurements. The
Original stresses (σ*, τ*) TRL's experiments on the A1 at Alconbury Hill
(σ*r 2 τ*) (Lee & Croney, 1962) and the AASHO Road Test
(σr 2 τ) Corrected stresses (σ, τ)
(Liddle, 1962) in Illinois are classic examples of
this type of experiment, the data from which have
σ*3
formed the basis for the empirical pavement design
σ*1 = σ1 σ
methods widely used today.
α α
pavement constructions are built for testing at that Brown & Brodrick (1981b). The most useful
location. instruments have proved to be a pressure cell
Most of the data obtained from these acceler- incorporating a simple strain gauged diaphragm
ated loading devices have related to failure and inductance strain coils. For full-scale experi-
mechanisms, principally in the bound pavement ments carried out in South Africa and Australia,
layers. Consequently, little soil mechanics know- the multi-depth de¯ectometer has also proved
ledge has been accumulated. One exception con- extremely useful for measuring de¯ections at
cerns the South African experiments involving various points within the depth of a pavement
injection of water into the granular layer and (Basson et al., 1981).
observations of the associated build-up of perma- A promising low cost technique for measuring
nent deformation described by Freeme & Servas water content is currently being used in the USA
(1985). Fig. 12 illustrates typical data for pavements in pavement test sections forming part of the
with different qualities of granular layer. The Federal Highway Administration's Long Term
advantages of using good quality dense crushed Pavement Performance experiments. The Time
rock, type G1 (NITRR, 1985b) are apparent. The Domain Re¯ectometry (TDR) principle is utilized
in¯uence of effective drainage can also be noted. by inserting a three-pronged probe into the soil.
At Nottingham, a pilot scale facility (Brown & The transmission and re¯ection of an electro-
Brodrick, 1981a) has been used for over 20 years magnetic pulse allow an apparent length of the
to study a range of pavement problems under probe to be determined. This is a function of the
controlled conditions. Loads up to 1´7 t can be soil dielectric constant which is directly related to
applied and speeds up to 16 km/h on pavements the volumetric water content using equations
constructed in a 1´5 m deep test pit, 4´8 m long devised by Topp et al., (1984). The usual water
and 2´4 m wide. content by mass can then be calculated knowing
the dry density of the soil.
Pavement instrumentation. For experiments in-
volving in situ instrumentation, several transducers The Bothkennar pavement experiments. In 1987,
have proved effective in monitoring both response the Science and Engineering Research Council
to individual wheel loads and the changes in stress, purchased a soft clay site at Bothkennar near the
strain and de¯ection with repeated loading. In Forth Estuary and established it for full-scale
addition to these three parameters, temperatures experimental geotechnical research. Full details of
and pore pressures have also been measured, the the site and results of extensive tests on the clay are
former with great success, using simple thermo- described in Greenwood et al., (1992).
couples, the latter with more limited success. This An unsurfaced pavement loop was constructed
is unfortunate, since stress determination in and in 1989 incorporating 16 test sections with the
below pavements involves total stress, so without principal objective of studying the reinforcing
suction or pore pressure measurements, the effec- effects of various geosynthetics placed at the
tive stress state is not reliably known. interface between a granular layer and the clay
Appropriate instrumentation is reviewed in (Little, 1993). The data which were gathered on
Brown (1978) and that applied in pilot and full- the performance of these sections under repeated
scale experiments at Nottingham is described by truck loading were used to assess design methods
for both reinforced and unreinforced haul roads.
Over 400 instruments were installed, nearly all
Material type
sections being involved. The instrumentation layout
G4 was designed to measure speci®c effects related to
20 G3 assumptions in the design methods. Inductance
Permanent deformation: mm
0.5
60 (a)
50 Vertical stress
Stress
Horizontal stress
Vertical stress: kPa
40
Shear stress when
wheel moves in
30 opposite direction
20
Time
10
Shear stress
0
0 0.4 0.8 1.2 1.6 2.0
Time: s (b)
(b)
Fig. 14. Stress conditions under a moving wheel load:
Fig. 13. In situ measurements from Bothkennar haul (a) stresses on pavement element; (b) variation of
road experiments (after Little, 1993): (a) vertical stresses with time
strain at formation level; (b) vertical stress at
formation level
Clearly, for pavements this would demand complex
facilities. A close match to ®eld conditions can be
a pressure cell at the same depth is presented in obtained by use of a Hollow Cylinder Apparatus
Fig. 13(b). (HCA). This allows control of both normal and
shear stress in a manner which can match the in
situ case as shown in Fig. 16. Repeated load
Laboratory testing HCAs have been developed at the University of
Figure 14 illustrates the general stress regime California, Berkeley (Alavi, 1992) principally to
experienced by an element of material in or below test bituminous materials, and at Nottingham for
a pavement structure as a result of a moving wheel granular materials (Chan & Brown, 1994).
load within the plane of the wheel track, that is, Interest in the HCA was partly stimulated by
the longitudinal plane. There are pulses of vertical the problems associated with the Simple Shear
and horizontal stress accompanied by a double Apparatus (SSA), which also has the potential
pulse of shear stress with a sign reversal on the to reproduce the in situ stress regime. The most
vertical and horizontal planes. Fig. 15 shows the serious dif®culties with the SSA were applications
associated pattern of principal stresses illustrating of uniform stress conditions and accurate measure-
the rotation of principal planes which takes place. ment of stresses and deformations on the specimen
For elements of material in the lower part of a under repeated loading conditions. Shaw & Brown
bituminous or concrete layer, the horizontal stres- (1986) describe an SSA in which both the vertical
ses are tensile, elsewhere they are compressive. and shear stresses can be applied cyclically, but for
One approach to laboratory testing is to select which the problems of stress measurement were
the equipment which reproduces the ®eld situation. identi®ed.
LVDT
Actuator
Servo
hydraulic
supply Piston
(b) To
computer
Fig. 15. Stresses on a pavement element: (a) principal
stresses ± element rotates; (b) no rotation ± shear
stress reversal Strain gauged
diaphragm
Load
cell
W
LVDTs
MT
Test
specimen
PO θ
Pi σz
r
τz θ
σθ
(a)
σr Transducer
To
computer
300 mm
section only
measured by cast epoxy strain hoops ®tted with
foil strain gauges. Similar apparatus has been used
in France (Paute et al., 1993), while 400 mm dia.
75 mm
specimens were adopted in the Dutch equipment LVDT
developed by Sweere (1990). He demonstrated that
the large diameter was needed to accommodate Rod attached
particle sizes up to 40 mm since inaccurate results to location stud
were obtained when using smaller specimens.
Equally, if the grading is scaled down or the
larger fractions omitted when using smaller speci-
mens, unrepresentative data are obtained.
In recent times, simpli®ed pieces of apparatus (b)
have been developed for use in engineering Fig. 18. Equipment for repeated load triaxial testing
practice, derived from the more complex facilities of granular materials (after Brown et al., 1989): (a)
used in research studies. For bituminous mixtures, diagram of loading equipment; (b) position of strain
the Nottingham Asphalt Tester (Cooper & Brown, transducers
1989), has emerged as a practical and reliable tool
for measuring the key mechanical properties of
bituminous materials. The development of compar- (AASHTO, 1986). Because soils testing is inher-
able, simpli®ed testing facilities for soils and ently more dif®cult to perform than asphalt testing,
granular materials has not progressed as far as simpli®ed techniques present more of a challenge.
for bituminous mixtures. This is somewhat surpris- One of the principal problems is that of preparing
ing since repeated load triaxial testing has been reliable and representative specimens. However,
used in the USA since the 1950s (Seed et al., recent work at Nottingham, in conjunction with
1955). The US Transportation Research Board TRL, has endeavoured to develop practical tests for
issued a detailed guide for such testing in 1975 soils and granular materials which could be adopted
(Transportation Research Board, 1975) and there for design purposes.
is an AASHTO speci®cation for such tests A pneumatically operated repeated load triaxial
Specimen
are for reading small resilient strains under
82 mm
repeated loading and can resolve to 28 microstrain. Strain
An external LVDT is used to measure the larger gauges Aluminium
block with
plastic strains and total strains in monotonic tests cup fitting
to failure.
For radial strain, the same proximity transducers
are used as for the cyclic loading facility (Fig.
17(b)). The `target' for each transducer is a 30 mm
square of aluminium foil placed inside the latex
membrane. Radial strains can be resolved to 10 Adjustable
Cruciform
vane
microstrain. No provision is made for pore pres- fixing
sure measurement, since specimens will, in gen-
eral, be partially saturated and independent soil
suction determination is recommended. 40 mm
The axial load is measured by a load cell
formed from strain gauging a narrowed section of Fig. 20. Axial deformation measuring system for soils
(after Cheung, 1994)
the loading rod inside the triaxial cell, a technique
used earlier by Austin (1979). The applied load is is air with a maximum value of 200 kPa but with
controlled by an electro-pneumatic regulator ®tted no facilities for cyclic application. A computerized
near the actuator. Deviator stresses up to 200 kPa system is used to control experiments and to
can be applied. The con®ning pressure medium monitor the data. A Windows environment with
Fig. 19. Simpli®ed repeated load triaxial system for soils (after Cheung, 1994)
Fig. 21. Simpli®ed repeated load triaxial system for Fig. 22. Deformation measuring system for granular
granular materials (after Cheung, 1994) materials (after Cheung, 1994)
30 23
Wheel position
20 cm right-hand side
80 20
140 19
700 Rutting
Tensile strain: microstrain
Subgrade
Sub-ballast 0.4 Ballast 10
Ballast
Settlement
Tamping
0.2 Subgrade 5
Sub-ballast
0.0 0
5 10 15 20 25
Traffic: MGT
Swelling line
Deviator stress, q
C
Deviator stress, q
Pavement
Compression construction
F P
Swelling
D Lowering
water
E table
P′
O P A G B
Mean normal effective stress, p ′
P′ Removal of overburden
Effects of construction
Fig. 35. Effects of construction operations on stress
conditions in `cut'
CSL (failure)
This discussion assumes that the soil remains in
a saturated state throughout the sequence of events.
While this may well be true for the cutting, on an
C embankment, the situation is less certain. Farrer
B X (1979) reported pore pressures below the pavement
Deviator stress, q
C
Wheel loading
δq
F X
qr
Q (Fill)
δq Wheel loading δu
Time
E T
qr
TSP
ESP
50 0 15 30
stress of 1´3 times the value of static yield over the Keuper Marl
permanent strain: kPa
for subgrades to prevent any signi®cant contribu- dif®cult. However, a pragmatic approach is sug-
tion to permanent deformation in the pavement. gested in the ®nal section of this paper. If the
This would involve ensuring that the ratio of actual value of accumulated plastic shear strain
deviator stress to mean normal effective stress or after N cycles is required, Cheung proposed the
soil suction was kept below a critical value. Brown following relationship based on testing up to 1000
& Dawson (1992) used this approach for design cycles
and suggested a ratio of 2 for pavement founda- q b
tions, recognizing that some plastic strain in the åp (N) A r (log N B) (5)
subgrade at this stage in the construction is s
permissible. They also noted that the reconstituted where A, b and B can be de®ned for the particular
soil specimens (Brown et al., 1987) had been soil. Although equation (5) is only valid for
tested at a higher frequency than the compacted relatively few load applications, this could still
specimens (Loach, 1987), which will have in¯u- be of use in pavement foundation design where
enced the result in view of the noted viscous the number of construction traf®c movements is
behaviour. limited.
Later, more extensive testing by Cheung (1994) Much more research has been devoted to the
on compacted clays using the apparatus shown in measurement of resilient soil properties under
Fig. 19, produced data such as those shown in repeated loading. The parameter, resilient modulus,
Fig. 40. These resulted from tests involving 1000 was introduced by Seed et al. (1962) and de®ned
cycle bursts of repeated deviator stress at 2 Hz on as repeated deviator stress divided by recoverable
compacted, uncon®ned specimens of Keuper Marl (resilient) axial strain in the triaxial test. They
and London Clay, two of the soils tested by Loach demonstrated that it varied with the magnitude of
(1987). The suction for the specimen featured in the repeated deviator stress, as shown in Fig. 41.
Fig. 40 was 44 kPa leading to a threshold deviator Later work by Dehlen & Monismith (1970) showed
stress, according to Loach, of 22 kPa. This point is that suction also had an important in¯uence. The
seen to coincide with the sharp change in slope of ®rst attempt to relate resilient modulus to the
the line in Fig. 40. However, not all of Cheung's effective stress was reported by Brown et al.
data demonstrated this clear change in slope. (1975) who, working with reconstituted silty clay,
Cheung used an alternative approach to design obtained the data in Fig. 42 for a range of initial
suggesting that the plastic strain after 1000 cycles speci®c volumes, overconsolidation ratios and
should be limited to 1%. The deviator stress initial effective stresses. These data were used to
causing this (qt ) was related to soil suction, deduce the empirical relationship:
yielding ratios of qt /s of 0´8 for Keuper Marl, n
p90
(wL 33´7%, wp 17´6%), 0´4 for Bothkennar Er K (6)
qr
clay (wL 54´3%, wp 25´1%) and 0´5 for Lon-
don Clay (wL 76%, wp 25´2%). The range of where K and n depend on the soil type, p90 is the
soil suction for Cheung's specimens was 20±80 kPa. initial mean normal effective stress and qr is the
These various triaxial test results suggest that
the allowable transient deviator stress is a function 16000
of the effective stress state of the soil. Since the
initial stress state, particularly for compacted soil, 14000
is uncertain, precise application of these data is
12000
Resilient modulus: psi
1.0 10000
Permanent axial strain: %
0.8
8000
0.6
6000
0.4
4000
0.2
2000
0
0 10 20 30 40 50 0
0 5 10 15 20 25 30 35 40
Repeated deviator stress: kPa Deviator stress: psi
Fig. 40. Plastic strain after 1000 cycles against re- Fig. 41. Relationship between resilient modulus (after
peated deviator stress for compacted silty clay (after 105 repetitions) and repeated deviator stress for a silty
Cheung, 1994) clay (after Seed et al., 1962)
4 190
10 76
200 20 38 Granular Case A
4 )
10 ) 40
)
1
150 20
Subgrade
100
Case B
Depth: m
50
2
60
This was developed for (qr / p90 ) values between 76
shear modulus was applicable with p90 replaced by Repeated deviator stress: kPa
suction (Brown et al. 1987).
This model re¯ects the non-linear resilient Fig. 44. Comparison of `model' and data for resilient
modulus of London Clay (after Dawson & Correia,
behaviour of clays and, when applied to pavement 1993)
analysis, demonstrates a sharp increase in stiffness
with depth as shown in Fig. 43 taken from Brown
et al., (1987). It is also able to model the in¯uence Gomes Correia (1985), they were able to suggest an
of water table position. Its main shortcoming is approximate general equation to estimate resilient
that unrealistically high values of resilient modulus modulus from the stress conditions and the plastic
are predicted at low deviator stresses. Dawson and limit wp as follows:
Gomes Correia (1993) suggested an improved Er 49 200 950p90 ÿ 370qr ÿ 2400wp (9)
expression for resilient modulus as follows
Er D Ep90 ÿ Fqr (8) in which Er , p90 and qr are in kPa and wp is a
percentage. The equation is compared with experi-
in which D, E and F can be determined from the mental data in Fig. 45 and applies only for resilient
test data. Fig. 44 indicates the quality of ®t for a set modulus values up to 80 MPa, which covers the
of tests on compacted London Clay. By examining practical range for clay subgrades in the UK. In the
Brown et al.'s data for Keuper Marl, London Clay case of compacted clays, p90 is taken as the soil
and Gault Clay, together with data on Kaolin from suction.
500
1
400
0.9
Normalized modulus: G/Gmax
300
0.8
qr /p ′e
0.6
nesses.
0.4 Prediction Although Boyce's model satis®es the laws of
po′ = 33 kPa, Ro = 18
0.2 thermodynamics, it does not ®t experimental data
as convincingly as design engineers might require.
0
0.001 0.01 0.1 1
Jouve & Elhannani (1993) proposed a more
Resilient shear strain: %
general model derived from Boyce's work, in
which the elastic potential implied by the expres-
Fig. 48. Comparison of Loach model for resilient sion for â was replaced, since its use did not
modulus with data from elsewhere conform with experimental evidence. However, the
Shear strain where H and d are constants for the material, or, in
(microstrain)
its more popular version, known as the K±è model
E
200
UR
Er K 1 èK 2 (22)
IL
160
FA
400
where K1 and K2 are the constants and è 3 p9.
130 This simple approach cannot be realistically
300 used to compute stress conditions in a granular
layer or a sand subgrade. However, it is of use for
Deviator stress q : kPa
100
modelling a granular layer when effects in the
bituminous material or the soil below are of
200
70
interest. Equations (21) and (22) could be regarded
as a lower bound to data of the type shown in
Fig. 50 where the effects of both deviator stress
100 40
and mean normal effective stress are illustrated.
If resilient properties are de®ned in terms of
the parameter Er , then a corresponding value of
100 200 300 Poisson's ratio is required which should itself be
stress-dependent as noted by Sweere (1990), al-
(−15,−6) Normal stress p : kPa though, in practice, a constant value, usually 0´3, is
(a) used.
An important aspect of the contours in Fig. 49
500 is the dependence on stress ratio (ç q/ p9) which
E
(microstrain)
1200
400
100
800
1400
300
based on a pragmatic approach to pavement design
in which the peak to peak values of stress and
200 strain are considered. They ignore the detailed
relationships within individual cycles. This is
justi®ed on the basis that the pavement problem
400
100
is one in which very large numbers of cycles are
200
800
different for Kr and Gr and by removing the
relationship between â and the other parameters. 600
This increases the number of parameters to be ‘Pessimum
determined from three to ®ve but gives a better ®t 400
line of fit’
to experimental data.
Because the modelling of resilient behaviour 200
for granular materials is complex, simpli®ed ap-
proaches have been adopted for design. The most 0
trivial involves constant Poisson's ratio, usually 0´3, 0 50 100 150
and a resilient modulus given by Peak mean normal effective stress: kPa
to try to model the strain accumulation under the Stress paths for resilient strain testing
large numbers of cycles relevant to pavement Stress paths for plastic strain testing
loading. Pappin's results did not produce a corre-
sponding expression for volumetric plastic strain.
Paute et al., (1993) describe the procedures
derived from Pappin's work which are used in
France to characterize permanent strains for granu- 200
lar materials. Only axial permanent strain is
0.5
accumulated strain, become very large.
0.4
Although empirical models have been developed
to match the measured data for repeated load 0.3
triaxial tests on granular materials, in particular
cases, testing is still needed to determine the 0.2
Permanent
various parameters. The models therefore only strain
0.1
provide a framework within which experience
suggests that the data may be interpreted. Thom 0
& Brown (1988) proposed a series of stress paths 5 10 25 50
Number of load cycles
that could be applied to evaluate routinely both
resilient and plastic strain characteristics. These are Fig. 53. In¯uence of shear stress reversal on accumu-
illustrated in Fig. 52 and show 19 stress paths to lation of plastic strain in a dry crushed rock (after
deal with resilient response, all of which involve Chan & Brown, 1994)
peak values below the threshold, and a single,
20th, path to characterize plastic strain. Finally, Brown & Chan (1996) have shown that there is a
unless failure has developed under repeated load- difference between unidirectional and bidirectional
ing, a monotonic test can follow to measure shear shear reversal (representing one-way and two-way
strength. About 20 cycles on each of the paths wheel loading respectively), the former leading to
for resilient strain are adequate, while the more lower strains than the latter and hence smaller rut
damaging paths for plastic strain could be applied depths. Their work was based on both HCA and
for 104 ±105 cycles. A frequency of 1 Hz is ap- wheel tracking tests.
propriate.
Chan's (1990) experiments with the Hollow
Cylinder Apparatus demonstrated that shear rever- PAVEMENT DESIGN DEVELOPMENTS
sal (rotating principal planes) does in¯uence plastic The CBR method
strain accumulation under repeated loading. This is Background. The most in¯uential early work on
illustrated by the data in Fig. 53 from Chan & pavement design and associated soil testing was
Brown (1994) showing the increased rate of strain carried out by the California Division of High-
when shear reversal is introduced to a specimen ways. Porter (1938) presented early recommenda-
initially subjected to triaxial stress conditions. tions for pavement layer thicknesses, based on
an empirical equation reported by Powell et al., below in an attempt to put matters into perspective
(1984) in relation to the soil mechanics of pavements
outlined in the previous section.
E 176CBR064 MPa (27)
In view of the earlier observations, quoted UK subgrade research. Between 1943 and 1961,
above, to the effect that CBR relates to shear Dr Croney and his colleagues at RRL conducted
strength, con®rmed in some detail by Black extensive research on pavement subgrades. Their
(1962), it is surprising that is should be used to work was aimed at the determination of water
estimate what is effectively the resilient modulus content pro®les beneath pavements and they used
Er of the soil. Brown et al., (1990) demonstrated soil suction as a major parameter taking, as their
that Er was not a simple function of CBR but starting point, research done by agronomists,
depended on soil type and the applied deviator notably Scho®eld (1935). He proposed the pF scale
stress level. Their results are summarized in Fig. for measuring soil suction, where the pF value is
55. In Fig. 55(a), the dependence of the relation- log10 (suction in cm of water). The `p' comes from
ship on deviator stress for Keuper Marl is clear. In an analogy with the logarithmic scale of acidity
Fig. 55(b) results for three soils at a deviator stress (pH) and the `F' stands for `Free energy'. Croney's
of 40 kPa are compared with equation (27) and the de®nition of suction (s) differs from that generally
simpler, Mr 10 CBR (MPa), frequently used in accepted in studies of partially saturated soils and
design. Sweere (1990) could ®nd no correlation known as matrix suction
between the CBR and resilient modulus for a range s ua ÿ uw (28)
of granular materials.
The research on subgrades which forms the where ua is pore air pressure and uw is pore water
background to current UK practice is reviewed pressure.
Croney assumed ua 0 and de®ned suction as
ÿuw under zero external stress. The effect of
applied total stress p was taken into account by
300
20 use of a `compressibility coef®cient' á as follows
u s áp (29)
in which s is a negative quantity.
Resilient modulus: MPa
200
It is important to note that p was de®ned by the
experiments of Black et al., (1958) as a hydrostatic
40 pressure whereas, in the practical application of the
100 ideas to predictions of water content pro®les in the
60
80
®eld, Croney & Coleman (1952) had regarded it
100 simply as the vertical component of stress.
There has been dif®culty over the years in
0
0 2 4 6 8 10
understanding the correct de®nitions and applica-
CBR: % tion of the á parameter. A useful discussion can be
(a)
found in Jones (1979) and Black (1979), though a
study of Croney (1952) and Croney & Coleman
(1952) provide fuller explanations.
120 It is worth noting that, for saturated clays, á 1
arl
er M and hence
Resilient modulus: at deviator
p
Keu R
CB
stress of 40 kPA: MPa
10 s u ÿ p ÿ(p ÿ u) (30)
80 M=
R
0. 64 where p is the mean normal total stress.
7. 6 CB As s is a negative quantity, it follows that it is
Mr = 1
London Clay equal to the mean normal effective stress p9 since
40
Gault Clay p9 p ÿ u (31)
This has allowed some bridges to be built
0
0 2 4 6 8 10 between conventional Terzhagi soil mechanics and
CBR: % the soil physics approach of Croney and his
(b) colleagues. (Croney, 1977). This link is also
Fig. 55. Relationships between CBR and resilient
apparent in the expression for resilient modulus
modulus for clays (after Brown et al., 1987): (a) equation (7). Croney & Coleman (1948) regarded
Keuper marl; (b) Three soils compared with empirical the pavement as a two-layer elastic system and
predictions at deviator stress of 40 kPa stated:
0.75
behaviour of pavements under traf®c. To provide
a satisfactory subgrade, the soil should operate at 0.5
stress levels within the `elastic range' . . .. The
pavement engineer is, therefore, more concerned 0.25
with the elastic modulus of the soil and the
behaviour under repeated loading'. 0
0 1 2 3 4 5 6 7 8
Penetration: mm
Grainger & Lister (1962) developed repeated
load triaxial test facilities at RRL to measure Fig. 56. Repeated load CBR tests on reconstituted silty
`elastic moduli' but their early work was curtailed. clay (after Loach, 1987)
subgrade soil mechanics reveals the need for shear to normal stress at failure and, hence, relates
improved test methods to quantify the resilient strongly to the angle of shearing resistance.
modulus. Such equipment has been available for Hveem & Carmany (1948), recognized that the
many years and has emerged from being a research `dynamic modulus of elasticity' for the subgrade
tool to a practical test for design application. This was a parameter of relevance to understanding the
matter is further discussed in the ®nal section. fatigue cracking of asphalt surfaces and that
monotonic loading would not be adequate for its
determination. Hveem (1955) in his classical paper
The rational approach to the Highway Research Board developed the
While the US Army Corps of Engineers were theme of resilient behaviour of pavements. He
enthusiastically embracing the CBR approach to devised a repeated load version of the Stabilometer
pavement design, Hveem and his colleagues in test known as the Resiliometer to characterize
California were devising new tests and developing subgrades. There was an appreciation of the fact
theories as to how pavements developed failure that a low stabilometer value (strength) does not
conditions. Fig. 25 is the classical cross-section always correlate with high resilience. By collating
they drew to demonstrate cracking and deforma- ®eld measurements of surface de¯ection with the
tion. They invented the Stabilometer test to incidence of cracking, he was able to suggest
measure the frictional characteristics of unbound allowable de¯ection values for different pavement
materials and the Cohesiometer to determine constructions. This led to the worldwide adoption
tensile strength of surfacing. The Stabilometer, of surface de¯ection measurement as a routine
shown in Fig. 57, is still used today and is a form method for assessing the structural condition of
of triaxial test from which a resistance value R is pavements.
determined using the equation The K-mould devised by Handy & Fox (1987)
1 ÿ (óh =óv ) and evolved by Semmelink & de Beer (1993,
R (32) 1995) has similarities to the Resiliometer. Lateral
100 stress is mobilized by an elastic support system
where óh is measured for a typical value of óv with a stiffness which can be varied between
1100 kPa. 15±60 MPa, designed to simulate in situ conditions.
Hveem & Carmany (1948) state that R is a This provides a state somewhere between K0 (zero
measure of the ratio of maximum shear stress to lateral strain) and uncon®ned. Preliminary results
major principal stress (ô/ó1 ) as follows have been obtained for resilient properties and for
100R 1 ÿ (óh =óv ) (ô=ó1 ) (33) permanent deformation characteristics of granular
materials.
However, simple analysis of the Mohr's circle The pioneering work of Hveem and his col-
reveals that leagues combined theoretical concepts, ®eld ob-
1 ÿ (óh =óv ) q=ó1 2ô=ó 1 (34) servations and the development of innovative
laboratory tests in a co-ordinated manner which
where ó1 óv and q óv ÿ óh 2ô is still a model for good research and development.
Notwithstanding the disputed factor of two, R is Professor Seed and his colleagues at the
seen to be a reasonably fundamental measure of University of California at Berkeley followed the
lead established by Hveem. They developed the
Head of testing machine repeated load triaxial test and introduced the term
`modulus of resilience' (Seed et al., 1955). This
was later changed to `resilient modulus' (Seed et
Piston for
al., 1962) which is de®ned on p. 406 where the
applying stress dependence of resilient modulus is illustrated
load to specimen
by Fig. 41. Seed et al., (1962, 1965) demonstrated
Load Dial
gauge
Displacement
pump
that de¯ection in repeated plate loading tests could
be predicted using their laboratory data and
Pressure appropriate elastic theory.
gauge Test
specimen The early work on fatigue testing of bituminous
mixtures was also pursued at Berkeley under
Liquid
Professor Carl Monismith (Monismith et al.,
1961) and, quite independently, at Nottingham by
Flexible
Adjustable
stage diagram Professor Peter Pell (Pell, 1962).
Platen of An understanding of the non-linear properties of
testing machine
soils and granular materials evolved simultaneously
Fig. 57. The Hveem stabilometer (after Hveem & at Berkeley and Nottingham in the 1960s. Hicks &
Carmany, 1948) Monismith (1971) reported comprehensive data for
Depth: m
methods have generally used two design criteria, 1.0
the tensile elastic strain of the bottom of the
bituminous layer, to deal with fatigue cracking,
and the vertical elastic strain at the top of the Surface deflection
under centre of load = shaded area
subgrade. This latter was derived from back-analy-
sis of pavements of known performance with
1.5
respect to rut development and is a semi-empirical
attempt to deal with this failure mechanism using a
linear elastic based theory for design. (Brown &
Brunton, 1984).
Little serious soil mechanics is incorporated in
most of the available design methods based on a 2.0
mechanistic approach although the effects of non-
linear behaviour in soil and granular materials are
taken into account either directly or indirectly by
most of them.
Fig. 58. Typical variation of resilient vertical strain
The validity of theoretical models for pavement with depth through an asphalt pavement (after Brown
systems has been quite widely studied with com- et al., 1987)
parisons between computed and measured values of
the key design parameters generally giving reason-
able con®dence for use of the theory in design. resulting resilient moduli of its layers can then be
The use of back-analysis to determine resilient used with some con®dence in design calculations
properties has been widely implemented in practice to assess residual life and determine structural
(Brunton et al., 1992). It is a good example of the requirements for future life. This may take the
sensible use of analytical techniques and only gives form of a strengthening overlay or of partial re-
plausible results when the non-linear properties of construction.
the subgrade and, in certain cases, the sub-base are The technique has recently been used in
taken into account. The programs LEAD and FEAD conjunction with a case study to compare new
developed by d'Almeida et al., (1993) allow the ideas on pavement foundation design with current
two parameters in each of equations (6) and (22) empirical practice. Two sections of the A564
to be determined through an iterative procedure. Derby Southern By-Pass were tested at various
The bituminous or cement treated layers are stages of construction. One section was in cut and
assumed to be linear elastic and effective stiff- the other on an embankment. The soil was a silty
nesses are computed for these. The effective clay in both cases, the capping was locally won
stiffness accounts for the reduction in stiffness sand and gravel, the sub-base was crushed lime-
caused by cracking and general deterioration stone in the cutting and granite on the embank-
relative to the `as built' value for the undamaged ment and the road base was a dense bituminous
layer. material. The FWD was used to obtain de¯ection
Fig. 58 shows a typical pro®le of vertical elastic data and the results were back-analysed to deter-
strain through a pavement structure. Summation of mine effective resilient moduli for pavement layers
this strain with depth yields the surface de¯ec- as construction proceeded. The data in Table 1 for
tion. The large contribution from the subgrade is the section in cut show how the capping and
apparent, demonstrating the importance of proper subgrade mobilized higher resilient moduli when
characterization of resilient properties for the soil. covered by sub-base and road base owing to
An attraction of using theory in back analysis is increased con®ning stress and decreased deviator
that the model is calibrated to the structure through stress in keeping with the non-linear stress±strain
the matching of surface de¯ection pro®les. The relationships.
Sub-base ± 240
Capping 90 200 Design foundation
Subgrade 70 200
Prepare subgrade
Table 2 shows values of equivalent foundation
stiffness computed using equation (1) and the Check in situ performance
central de¯ection measurement only. The values
increased as tests were conducted successively on NO Performance satisfactory?
the subgrade, capping and sub-base of the em-
YES
bankment. This shows the effect of building up the
foundation on the stiffness of support provided to Place aggregate?
Rut depth: mm
20
Ash
number of load applications N. This strain level is 30
Sand &
regarded as a tentative suggestion at present. Full 40 Gravel
details of the wedge model are described by Thom 50
Granodiorite
predicted
et al. (1993). An iterative computation is used to 60
measured
Limestone
70
obtain a solution. 80
For the situation where the rutting is entirely 1 10 100 1000
contributed by the subgrade and for the determina- Number of passes
(a)
tion of equivalent foundation resilient modulus,
linear elastic layered system analysis is used. The
0
non-linear resilient properties of the materials are 10
accounted for by using the iterative approach
Rut depth: mm
20
described on p. 389. This allows compatibility of 30
Granite
(550 mm)
stresses and resilient moduli to be established. For 40
the subgrade rutting criterion, the deviator stress at 50
Gravel
formation is calculated and equation (35) used to 60 predicted (400 mm)
measured
check whether the allowable strain of 0´6% is 70
80
exceeded. 10 100 1000 10 000
The equivalent foundation stiffness is deter- (b)
mined using the computed surface de¯ection for
the layered system and calculating the resilient Fig. 61. Predicted rut depths compared with measure-
modulus for an equivalent semi-in®nite elastic half ments at full-scale (after Dawson et al., 1993) (a)
space using equation (1). The analysis described Loughborough trials; (b) Bothkennar trials
2r
C Rcb θc θa
A
Rba
h
φ
Fcb
B
Rb Fba
Fb
Ra
Fig. 60. Proposed `wedge model' for calculating rutting in pavement foundations (after Thom et al., 1993)
kennar the model overpredicted rutting. The labora- tion design has been outlined which, with a little
tory test data to support the Bothkennar calculations further development, could provide the basis to
were less complete than for the Loughborough tests. upgrade present practice. One feature of this is
The method clearly needs to be used more ex- the simpli®ed test procedures for materials which
tensively for a wider range of materials and condi- could be made more user-friendly in the manner
tions and to be re®ned. It is, however, considered achieved for bituminous materials using the Not-
to provide the basis for improved design of tingham Asphalt Tester (Cooper & Brown, 1989). A
pavement foundations in the future. wide range of soils and granular materials, under a
variety of conditions, needs to be tested.
Further research is needed, notably to obtain a
Railtrack design better understanding of the effective stress state
Li (1994) presents a good summary of proce- in pavement foundations. This will require good
dures for the determination of granular layer quality, long-term ®eld monitoring using moisture
thicknesses in railtrack. He describes a method and suction measuring techniques for which there
developed at the University of Massachusetts have been recent advances. The signi®cant im-
which uses the GEOTRACK program for modelling provement in understanding of partially saturated
response to wheel loading. The design criteria are soils through the work of Wheeler & Karube
vertical plastic strain and vertical permanent (1995), Alonso et al. (1990) and Fredlund &
deformation at formation level. The former is to Rahardjo (1993) provides a sound basis to improve
prevent plastic ¯ow, which leads to progressive the models used for pavement design. These could
failure of the top of the subgrade, while the latter be combined with elasto±visco±plastic models for
relates to the overall deformation of the subgrade. bituminous materials to provide a comprehensive
Varying the thickness of the granular layers theoretical framework for ¯exible pavements in the
in¯uences the transient deviator stress level in future.
the subgrade and, hence, the plastic strain åp . The Little or nothing has been reported herein on the
plastic strain after N cycles is computed by the vital matter of pavement drainage. Suf®ce it to say
following equation, derived from repeated load that improvements in the effectiveness and relia-
triaxial tests on clays bility of drainage so that water conditions below
å1p A(qr =cu )m N b (36) pavements remain stable and water contents at
reasonable levels continue to be a desirable aim.
in which A, m and b are experimentally deter-
mined, N is the number of load applications, cu is
the undrained shear strength and qr is the repeated ACKNOWLEDGEMENTS
deviator stress. In preparing this Rankine Lecture, the Author
has drawn liberally on the research carried out by
his colleagues at Nottingham University over many
CONCLUSIONS years. The successive research staff who have
This wide ranging review of soil mechanics in contributed new ideas and brought a variety of
pavement engineering has outlined the techniques skills to bear on both experimental and theoretical
which are available to study the problem using work have provided a stimulating environment and
theory, laboratory testing and ®eld experiments. It a large source of information. Particular thanks are
has primarily been concerned with bituminous expressed to: Ahmed Lashine, Adrian Hyde, John
pavements for highways but other pavement types Boyce, Jack Pappin, Simon Loach, Peter Ansell,
obey similar principles. The particular character- Phil Shaw, Joao d'Almeida, Bob Overy, Mike
istics of railtrack have also been described. O'Reilly, Matthew Raybould, Alex Tam and Peter
It was considered helpful to place present Little, all of whom have moved on to other places,
knowledge and practice in a historical perspective, and to present colleagues at Nottingham and SWK
since pavement soil mechanics has developed, to Pavement Engineering Ltd; Andrew Dawson, Nick
some extent, in isolation from mainstream geo- Thom, Ron Jones, Ian Richardson, Barry Brodrick,
technics. It is ironic that much of the research on Francis Chan, Lam Wah Cheung and John Nutt.
repeated loading of soil and granular materials has Grateful thanks are also extended to Carolyn
been quite sophisticated and comparable in quality Parkinson and Hayley Drabble, who dealt with
with that developed in other ®elds of soil mech- the text and ®gures respectively.
anics, but procedures used in current practice Without the work of all these people, this
remain empirical and backward. Much the same lecture would not have been possible.
can be said of theoretical modelling. Permission was granted by the Highways
The background knowledge accumulated from Agency to conduct tests on the A564 Derby
research presents an opportunity for improving Southern By-Pass and particular thanks are ex-
current practice. A simpli®ed approach to founda- pressed to Tarmac Construction Ltd. and to Scott
Instn Civ. Engrs, 79, Pt. 2, 1±31. of granular layers in pavements, PhD thesis, Uni-
Brown, S. F. & Bush, D. I. (1972). Dynamic response of versity of Nottingham.
model pavement structure, J. Transp. Engng, Am. Soc. Chan, F. W. K. & Brown, S. F. (1994). Signi®cance of
Civ. Engrs, 98, TE4, 1005±1022. principal stress rotation in pavements, Proc. 13th Int.
Brown, S. F., O'Reilly, M. P. & Pappin, J. W. (1989). A Conf. Soil Mech. Foundn Engng, Delhi, 4, 1823±
repeated load triaxial apparatus for granular materials, 1826.
in Unbound Aggregates in Roads, London, Butter- Chang, C. S., Adegoke, C. W. & Selig, E. T. (1979). A
worth, 143±158. study of analytical models for track support systems,
Brown, S. F. & Chan, F. W. K. (1996) Reduced rutting in Transp. Res. Record No. 733, Transportation Research
unbound granular pavement layers through improved Board, Washington, DC, 12±19.
grading design, Proc. Instn Civ. Engrs Transport, 117, Cheung, L. W. (1994). Laboratory assessment of pave-
40±49. ment foundation materials, PhD thesis, University of
Brown, S. F. & Dawson, A. R. (1992). Two-stage Nottingham.
approach to asphalt pavement design, Proc. 7th Int. Claessen, A. I. M., Edwards, J. M., Sommer, P. & UgeÂ, P.
Conf. Asphalt Pavements, Nottingham, 1, 16±34. (1977). Asphalt pavement design ± the Shell Method,
Brown, S. F., Lashine, A. K. F. & Hyde, A. F. L. (1975). Proc. 4th Int. Conf. Structural Design of Asphalt
Repeated load triaxial testing of a silty clay, GeÂo- Pavements, Ann Arbor, Michigan, 1, 39±74.
technique 25, No. 1, 95±114. Cooling, L. F. et al. (1961), Discussion on Session 3:
Brown, S. F., Loach, S. C. & O'Reilly, M. P. (1987). Roads and Runways and Agriculture in Pore pressure
Repeated loading of ®ne grained soils, Contractor and suction in soils, London, Butterworth, 143±151.
Report 72, Transportation Research Laboratory. Cooper, K. E. & Brown, S. F. (1989). Development of a
Brown, S. F., O'Reilly, M. P. & Loach, S. C. (1990). The simple apparatus for the measurement of the mech-
relationship between California Bearing Ratio and anical properties of asphalt mixes, Proc. Eurobitume
elastic stiffness for compacted clays, Ground Engng, Symposium, Madrid, 494±498.
23, No. 8, 27±31. Cox, B. (1980). Reconstruction, including design plan-
Brown, S. F. & Pappin, J. W. (1981). Analysis of ning and execution from the engineer's point of view,
pavements with granular bases, Transp. Res. Record Highway Engr., 27, No. 12, 13±15.
No. 810, Transportation Research Board, Washington, Croney, D. (1952). The movement and distribution of
DC, 17±22. water in soils, GeÂotechnique 5, No. 1, 1±16.
Brown, S. F. & Pell, P. S. (1967). An experimental Croney, D. (1977). The design and performance of road
investigation of the stresses, strains and de¯ections pavements, London, HMSO.
in a layered pavement structure subjected to dynamic Croney, D. & Coleman, J. D. (1948). Soil thermo-
loads, Proc. 2nd Int. Conf. Structural Design of dynamics applied to the movement of moisture in
Asphalt Pavements, Ann Arbor, Michigan, 487±504. road foundations, Proc. 7th Int. Conf. Appl. Mech.,
Brown, S. F. & Selig, E. T. (1991). The design of London, 3, 163±177.
pavement and rail track foundations, Chapter 6 in Croney, D. & Coleman, J. D. (1952). The estimation of
Cyclic loading of soils: from theory to design, the vertical moisture distribution with depth in
Blackie, Glasgow and London, 249±305. unsaturated cohesive soils, Road Note No. 1709,
Brown, S. F., Tam, W. S. & Brunton, J. M. (1986). DSIR, HMSO.
Development of an analytical method for the struc- Croney, D. & Coleman, J. D. (1954). Soil structure in
tural evaluation of pavements, Proc. 2nd Int. Conf. relation to soil suction (pF), J. Soil Sci. 5, No. 1,
Bearing Capacity of Roads and Air®elds, Plymouth, 75±84.
1, 267±276. d'Almeida, J. C. G. R. (1993). Analytical techniques for
Brown, S. F., Tam, W. S. & Brunton, J. M. (1987). the structural evaluation of pavements, PhD thesis,
Structural evaluation and overlay design: analysis and University of Nottingham.
implementation, Proc. 6th Int. Conf. Structural De- d'Almeida, J. R., Brown, S. F. & Thom, N. H. (1994). A
sign of Asphalt Pavements, Ann Arbor, Michigan, 1, pavement evaluation procedure incorporating material
1013±1028. non-linearity. Non-destructive testing and back-calcu-
Brunton, J. M. & Akroyde, P. M. (1990). Monitoring the lation of moduli. (2nd Vol.), American Society for
performance of a full-scale experimental pavement, Testing and Materials, STP 1198, 218±232.
Proc. 3rd Int. Conf. Bearing Capacity of Roads and Davis, E. H. (1949). The California Bearing Ratio
Air®elds, Trondheim, 585±594. method for the design of ¯exible roads and runways,
Brunton, J. M. & d'Almeida, J. R. (1992). Modeling GeÂotechnique 1, No. 4, 249±263.
material non-linearity in a pavement backcalculation Dawson, A. R., Brown, S. F., Thom, N. H. & Cheung, L.
procedure, Transp. Res. Record No. 1377, Transporta- W. (1993). Improvements to road foundation design:
tion Research Board, Washington, DC, 99±106. laboratory tests and design method, Report PRG
Brunton, J. M., Armitage, R. J. & Brown, S. F. (1992). 93023, University of Nottingham.
Seven years' experience of pavement evaluation, Dawson, A. R. & Plaistow, L. (1993). Parametric study ±
Proc. 7th Int. Conf. on Asphalt Pavements, Notting- ¯exible pavements, Flexible Pavements, Balkema,
ham, 3, 17±30. 229±237.
Burmister, D. M. (1943). Theory of stresses and Dawson, A. R. & Gomes Correia, A. (1993). The effects
displacements in layered systems and application to of subgrade clay condition on the structural behaviour
the design of airport runways, Proc. Highway Res. of road pavements, Flexible Pavements, Balkema,
Board, 23, Washington, DC, 126±148. 113±119.
Chan, F. W. K. (1990). Permanent deformation resistance Dawson, A. R. & Thom, N. H. (1994). Software manual
TRH 14, CSIR, Pretoria. Testing and Materials, STP 1213, 134±161.
Nutt, J. L. (1982). Finite element modelling of partially Robnett, Q. L., Thompson, M. R., Knutson, R. M. &
saturated soils for a study of the California Bearing Tayabji, S. D. (1975). Development of a structural
Ratio test, MSc thesis, Imperial College, University of model and materials evaluation procedures, University
London. of Illinois, Report No. DoT-Fl-30038.
O'Reilly, M. P. (1985). Mechanical properties of granular Rowe, G. M., Brown, S. F., Sharrock, M. J. & Bouldin,
materials for use in thermal energy stores, PhD M. G. (1995). Visco-elastic analysis of hot mix
thesis, University of Nottingham. asphalt pavement structures, Transp. Res. Record No.
O'Reilly, M. P., Brown, S. F. & Overy, R. F. (1989). 1482, Transportation Research Board, Washington,
Viscous effects observed in tests on an anisotropically DC, 44±51.
normally consolidated silty clay, GeÂotechnique 39, Scho®eld, R. K. (1935). The pF of the water in soil,
No. 1, 153±158. Trans. 3rd Int. Congress on Soil Science, 2, Oxford,
Pappin, J. W. (1979). Characteristics of a granular 37±48.
material for pavement design, PhD thesis, University Scho®eld, A. N. & Wroth, C. P. (1968). Critical state soil
of Nottingham. mechanics, McGraw-Hill, London.
Pappin, J. W. & Brown, S. F. (1980). Resilient stress- Seed, H. B., Chan, C. K. & Lee, C. E. (1962). Resilience
strain behaviour of a crushed rock, Proc. Int. Symp. characteristics of subgrade soils and their relation to
Soils under cyclic and transient loading, Swansea, 1, fatigue failures, Proc. Int. Conf. Structural Design of
169±177. Asphalt Pavements, Ann Arbor, Michigan, 611±636.
Pappin, J. W., Brown, S. F. & O'Reilly, M. P. (1992). Seed, H. B., Chan, C. K. & Monismith, C. L. (1955).
Effective stress behaviour of saturated and partially Effects of repeated loading on the strength and
saturated granular material subjected to repeated deformation of compacted clay, Proc. High. Res.
loading, GeÂotechnique 42, No. 3, 485±497. Board, 34, Washington, DC, 541±558.
Paute, J-L., Hornych, P. & Beraben, J.-P. (1993). Selig, E. T. & Waters, J. M. (1994). Track geotechnology
Repeated load triaxial testing of granular materials and substructure management, Thomas Telford, Lon-
on the French network of Laboratoires des Ponts et don.
ChausseÂes, Flexible Roads, Balkema, 53±64. Semmelink, C. J. & de Beer, M. (1993). Development of
Pell, P. S. (1962). Fatigue characteristics of bitumen and a dynamic DRTT K-mould system, Res. Report DPVT
bituminous mixes, Proc. Int. Conf. Structural Design 216, Roads and Transp. Tech., CSIR, Pretoria.
of Asphalt Pavements, Ann Arbor, Michigan, 310±323. Semmelink, C. J. & de Beer, M. (1995). Rapid determi-
Pell, P. S. & Taylor, I. F. (1969). Asphaltic road materials nation of elastic and shear properties of road building
in fatigue, Proc. Assoc. Asphalt Paving Tech., 38, materials with the K-mould, in Unbound aggregates
371±422. in roads, University of Nottingham, 151±162.
Porter, O. J. (1938). The preparation of subgrades, Proc. Shaw, P. & Brown, S. F. (1986). Cyclic simple shear
Highway Res. Board, 18, No. 2, Washington, DC, testing of granular materials, Geotech. Testing J.
324±331. American Society for Testing and Materials 9, No. 4,
Porter, J. (1950). Development of the original method for 213±220.
highway design, Development of CBR ¯exible pave- Shaw, P. & Brown, S. F. (1988). The behaviour of dry
ment design method for air®elds ± a symposium, granular materials under repeated load biaxial and
Trans. Am. Soc. Civ. Engrs, 115, 461±467. triaxial stress conditions, GeÂotechnique 38, No. 4,
Powell, W. D., Potter, J. F., Mayhew, H. C. & Nunn, M. 627±634.
E. (1984). The structural design of bituminous roads, Sorenson, A. & Hayven, M. (1982). The Dynatest 8000
Lab. Report 1132, TRRL. Falling Weight De¯ectometer test system, Proc. Int.
Ramsamooj, D. V., Majidzadeh, K. & Kauffmann, E. M. Symp. Bearing Capacity of Roads and Air®elds,
(1972). The analysis and design of the ¯exibility of Trondheim, 1, 464±470.
pavements, Proc. 3rd Int. Conf. Structural Design of Sousa, J. B. & Chan, C. K. (1991). Computer applica-
Asphalt Pavements, London, 692±704. tions in the geotechnical laboratories of the University
Raybould, M. R. (1992). The response of silt-clay of California at Berkeley, Proc. Geotech. Eng. Conf.,
mixtures to cyclic loading, PhD thesis, University of Boulder, Colorado, American Society of Civil Engi-
Nottingham. neers.
Raybould, M. J. & Brown, S. F. (1993). Experimental Stewart, H. E., Selig, E. T. & Norman-Gregory, G. M.
methods in determining cyclic response of soils with (1985). Failure criteria and lateral stresses in track
particular reference to triaxial tests on Hostun sand, foundations, Transp. Res. Record No. 1027, Trans-
Proc. Conf. Experimental Methods in Earthquake portation Research Board, Washington, DC, 59±64.
Eng. and Struct. Dynamics, St.-Remy-leÁs-Chevreuse, Sun, J. I., Golesorkhi, R. & Seed, H. B. (1988). Dynamic
France, 156±168. Moduli and Damping Ratios for Cohesive Soils.
Ridley, A. M. (1995). Strength±suction±moisture content Report No. UBC/EERC-88/15, Earthquake Engineer-
relationships for kaolin under normal atmospheric ing Research Center, 42.
conditions, Proc. 1st Int. Conf. on Unsaturated Soils, Sweere, G. T. H. (1990). Unbound granular bases for
Paris, 645±651. roads, Doctoral thesis, Delft University of Technology.
Roblee, C. J., Xiang-Song, L., Chan, C. K., Idriss, I. M., Thom, N. H. & Brown, S. F. (1988). The effect of
Wang, G., Herrmann, L. R. & Jackura, K. A. (1994). grading and density on the mechanical properties of a
Feasibility of a tool for in situ measurement of crushed dolomitic limestone, Proc. Aust. Road Res.
material properties of clays over a wide strain range, Board, 14, Pt. 7, 94±100.
Dynamic Geot. Testing II, American Society for Thom, N. H., Dawson, A. R. & Brown, S. F. (1993).
practical designs which could lead to a more ing. It was an ambitious lecture presented with a
theoretical approach than is currently used in the light, deft, touch. When he started, pavement engi-
UK. Some empiricism would remain, but at a neering may have been a Cinderella but our
lower level in the hierarchy. eighteenth home Lecturer has ensured that it has
We have been privileged to hear an acknow- come of age.
ledged master of his subject demonstrate clearly I ask you to join with me in showing your
the role of soil mechanics in pavement engineer- appreciation and thanks by acclamation.