78

Active pressure on gravity walls supporting purely

frictional soils
D. Loukidis and R. Salgado

Abstract: The active earth pressure used in the design of gravity walls is calculated based on the internal friction angle of
the retained soil or backfill. However, the friction angle of a soil changes during the deformation process. For drained load-
ing, the mobilized friction angle varies between the peak and critical-state friction angles, depending on the level of shear
strain in the retained soil. Consequently, there is not a single value of friction angle for the retained soil mass, and the active
earth pressure coefficient changes as the wall moves away from the backfill and plastic shear strains in the backfill increase.
In this paper, the finite element method is used to study the evolution of the active earth pressure behind a gravity retaining
wall, as well as the shear patterns developing in the backfill and foundation soil. The analyses relied on use of a two-surface
plasticity constitutive model for sands, which is based on critical-state soil mechanics.
Key words: finite elements, plasticity, retaining walls, sands.
Résumé : La pression active des terres utilisée dans la conception des murs gravitaires est calculée à partir de l’angle de
friction interne du sol ou du remblai retenu. Cependant, l’angle de friction d’un sol change durant le processus de déforma-
tion. Dans le cas d’un chargement drainé, l’angle de friction mobilisé varie entre l’angle de friction au pic et celui à l’état
critique, dépendant du niveau de déformation en cisaillement dans le sol retenu. En conséquence, il n’y a pas de valeur
unique d’angle de friction pour une masse de sol retenue, et le coefficient de pression active des terres varie à mesure que
le mur se sépare du remblai et que les déformations plastiques en cisaillement augmentent dans le remblai. Dans cet article,
la méthode par éléments finis est utilisée pour étudier l’évolution de la pression active des terres derrière un mur de soutène-
ment gravitaire, ainsi que les patrons de cisaillement qui se développent dans le remblai et dans le sol de fondation. Les ana-
lyses sont réalisées à l’aide d’un modèle constitutif de plasticité à deux surfaces pour des sables, qui est basé sur la
mécanique de l’état critique des sols.
Mots‐clés : éléments finis, plasticité, murs de soutènement, sables.
[Traduit par la Rédaction]

Introduction Caquot and Kerisel (1948) produced solutions in tabulated

form, assuming slip surfaces with logarithmic spiral shape.
The active earth pressure acting on the back of a retaining
More recently, Paik and Salgado (2003) estimated the active
wall controls its design. The active earth pressure is ex-
earth pressure behind rigid walls by improving the formula-
pressed as the product of the vertical effective stress s v0 in
tion of Handy (1985), which considers soil arching concepts.
the retained soil mass or backfill1 and the active earth pres-
Limit analysis has also been used to study the active earth
sure coefficient KA. The earliest and simplest methods for the
pressure problem. Rigorous upper bound values for KA estab-
calculation of the active earth pressure for purely frictional
lished by Chen (1975) and Soubra and Macuh (2002) using
backfills are those based on the Coulomb and Rankine theo-
limit analysis are in very close agreement with the values of Ca-
ries. For a backfill with horizontal surface, the Rankine solu-
quot and Kerisel (1948). Sokolovskiĭ (1965) solved the problem
tion is mathematically exact for a vertical and smooth wall
of active and passive earth pressure using the method of charac-
backface. Coulomb’s solution assumes a planar slip surface
teristics. More recently, Lancellotta (2002) provided a rigorous
and is equivalent to an upper bound solution. For a horizon-
lower-bound solution for active pressures in closed form:
tal backfill and a vertical wall backface, Coulomb’s solution

yields cosd
½2 KA ¼ ðcosd
1 þ sinf
½1 KA ¼
cos 2 f pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2  sin 2 f  sin 2 dÞ e½darcsin ðsind=sinfÞ tanf
cosdf1 þ ½sin ðf þ dÞ sinf=cosdg

Received 21 March 2011. Accepted 26 September 2011. Published at www.nrcresearchpress.com/cgj on 20 December 2011.
D. Loukidis. Department of Civil and Environmental Engineering, University of Cyprus, Nicosia 1678, Cyprus.
R. Salgado. School of Civil Engineering, Purdue University, W. Lafayette, IN 47907-1284, USA.
Corresponding author: D. Loukidis (e-mail: loukidis@ucy.ac.cy).
1The paper is not restricted to backfilled walls. To call attention to applicability of the discussion to walls supporting natural ground as

well as completely backfilled walls, the terms “retained soil mass” and “backfill” are used interchangeably throughout.

Can. Geotech. J. 49: 78–97 (2012) doi:10.1139/T11-087 Published by NRC Research Press
96
References
Abbo, A.J., and Sloan, S.W. 2000. SNAC. User manual. Version 2.0
[computer program]. Department of Civil, Surveying and Environ-
mental Engineering, University of Newcastle, Callaghan, Australia.
Canadian Geotechnical Society. 1992. Foundation engineering manual.
Canadian Geotechnical Society (CGS), Bitech, Vancouver, B.C.
Caquot, A., and Kerisel, J. 1948. Tables for the calculation of passive
pressure, active pressure and bearing capacity of foundations.
Gauthier Villars, Paris, France.
Carraro, J.A.H. 2004. Mechanical behavior of silty and clayey sands.
Ph.D. thesis, Purdue University, West Lafayette, Ind.
Carraro, J.A.H., Bandini, P., and Salgado, R. 2003. Liquefaction
resistance of clean and nonplastic silty sands based on cone
penetration resistance. Journal of Geotechnical and Geoenviron-
mental Engineering, 129(11): 965–976. doi:10.1061/(ASCE)1090-
0241(2003)129:11(965).
Chen, W.F. 1975. Limit analysis and soil plasticity. Elsevier Science
Publishers, B.V., Amsterdam, the Netherlands.
Dafalias, Y.F., Papadimitriou, A.G., and Li, X.S. 2004. Sand
plasticity model accounting for inherent fabric anisotropy. Journal
of Engineering Mechanics, 130(11): 1319–1333. doi:10.1061/
(ASCE)0733-9399(2004)130:11(1319).
Day, R.A., and Potts, D.M. 1998. The effect of interface properties on
retaining wall behaviour. International Journal for Numerical and
Analytical Methods in Geomechanics, 22(12): 1021–1033. doi:10.
1002/(SICI)1096-9853(199812)22:12<1021::AID-NAG953>3.0.
CO;2-M.
Fang, Y.-S., and Ishibashi, I. 1986. Static earth pressures with various
wall movements. Journal of Geotechnical Engineering, 112(3):
317–333. doi:10.1061/(ASCE)0733-9410(1986)112:3(317).
Fukushima, S., and Tatsuoka, F. 1984. Strength and deformation
characteristics of saturated sand at extremely low pressure. Soils
and Foundations, 24(4): 30–48. doi:10.3208/sandf1972.24.4_30.
Gudehus, G., and Nübel, K. 2004. Evolution of shear bands in sand.
Géotechnique, 54(3): 187–201. doi:10.1680/geot.2004.54.3.187.
Handy, R.L. 1985. The arch in soil arching. Journal of Geotechnical
Engineering, 111(3): 302–318. doi:10.1061/(ASCE)0733-9410
(1985)111:3(302).
Ichihara, M., and Matsuzawa, H. 1973. Earth pressure during
earthquake. Soils and Foundations, 13(4): 75–86. doi:10.3208/
sandf1972.13.4_75.
Iwasaki, T., Tatsuoka, F., and Takagi, Y. 1978. Shear moduli of sands
under cyclic torsional shear loading. Soils and Foundations, 18(1):
39–56. doi:10.3208/sandf1972.18.39.
Lam, W.-K., and Tatsuoka, F. 1988. Effects of initial anisotropic fabric
and s2 on strength and deformation characteristics of sand. Soils
and Foundations, 28(1): 89–106. doi:10.3208/sandf1972.28.89.
Lancellotta, R. 2002. Analytical solution of passive earth pressure.
Géotechnique, 52(8): 617–619. doi:10.1680/geot.2002.52.8.617.
Leśniewska, D., and Mróz, Z. 2001. Study of evolution of shear band
systems in sand retained by flexible wall. International Journal for
Numerical and Analytical Methods in Geomechanics, 25: 909–
932. doi:10.1002/nag.160.
Li, X.S., and Dafalias, Y.F. 2000. Dilatancy of cohesionless soils.
Géotechnique, 50(4): 449–460. doi:10.1680/geot.2000.50.4.449.
Loukidis, D. 2006. Advanced constitutive modeling of sands and
applications to foundation engineering. Ph.D. thesis, Purdue
University, West Lafayette, Ind.
Can. Geotech. J. Vol. 49, 2012
Loukidis, D., and Salgado, R. 2008. Analysis of the shaft resistance
of non-displacement piles in sand. Géotechnique, 58(4): 283–296.
doi:10.1680/geot.2008.58.4.283.
Loukidis, D., and Salgado, R. 2009. Modeling sand response using
two-surface plasticity. Computers and Geotechnics, 36(1–2): 166–
186. doi:10.1016/j.compgeo.2008.02.009.
Loukidis, D., Chakraborty, T., and Salgado, R. 2008. Bearing
capacity of strip footings on purely frictional soil under eccentric
and inclined loads. Canadian Geotechnical Journal, 45(6): 768–
787. doi:10.1139/T08-015.
Manzari, M.T., and Dafalias, Y.F. 1997. A critical state two-surface
plasticity model for sands. Géotechnique, 47(2): 255–272. doi:10.
1680/geot.1997.47.2.255.
Milligan, G.W.E. 1974. The behaviour of rigid and flexible retaining
walls in sand. Ph.D. thesis, University of Cambridge, Cambridge, UK.
Murthy, T.G. 2006. Undrained static response of sands with fines. Ph.
D. thesis, Purdue University, West Lafayette, Ind.
Murthy, T.G., Loukidis, D., Carraro, J.A.H., Prezzi, M., and
Salgado, R. 2007. Undrained monotonic response of clean and
silty sands. Géotechnique, 57(3): 273–288. doi:10.1680/geot.
2007.57.3.273.
Nemat-Nasser, S., and Okada, N. 2001. Radiographic and micro-
scopic observation of shear bands in granular materials. Géotech-
nique, 51(9): 753–765. doi:10.1680/geot.2001.51.9.753.
Paik, K.H., and Salgado, R. 2003. Estimation of active earth pressure
against rigid retaining walls considering arching effects. Géotech-
nique, 53(7): 643–653. doi:10.1680/geot.2003.53.7.643.
Papadimitriou, A.G., and Bouckovalas, G.D. 2002. Plasticity model
for sand under small and large cyclic strains: a multiaxial
formulation. Soil Dynamics and Earthquake Engineering, 22(3):
191–204. doi:10.1016/S0267-7261(02)00009-X.
Potts, D.M., and Fourie, A.B. 1986. A numerical study of the effects
of wall deformation on earth pressures. International Journal for
Numerical and Analytical Methods in Geomechanics, 10(4): 383–
405. doi:10.1002/nag.1610100404.
Salgado, R. 2008. The engineering of foundations. McGraw-Hill,
New York.
Sokolovskiĭ, V.V. 1965. Statics of granular media. Pergamon, Oxford,
UK.
Soubra, A.-H., and Macuh, B. 2002. Active and passive earth
pressure coefficients by a kinematical approach. Proceedings of
the Institution of Civil Engineers — Geotechnical Engineering,
155(2): 119–131. doi:10.1680/geng.2002.155.2.119.
Uesugi, M., Kishida, H., and Tsubakihara, Y. 1988. Behavior of sand
particles in sand–steel friction. Soils and Foundations, 28(1): 107–
118. doi:10.3208/sandf1972.28.107.
Vardoulakis, I. 1980. Shear band inclination and shear modulus of
sand in biaxial tests. International Journal for Numerical and
Analytical Methods in Geomechanics, 4(2): 103–119. doi:10.
1002/nag.1610040202.
Vardoulakis, I., and Sulem, J. 1995. Bifurcation analysis in geome-
chanics. Blackie Academic and Professional, Glasgow, Scotland.
Wolf, H., König, D., and Triandafyllidis, Th. 2005. Centrifuge model
tests on sand specimen under extensional load. International
Journal for Numerical and Analytical Methods in Geomechanics,
29(1): 25–47. doi:10.1002/nag.403.
Yoshimine, M., Ishihara, K., and Vargas, W. 1998. Effects of
principal stress direction and intermediate stress on undrained
shear behavior of sand. Soils and Foundations, 38(3): 179–188.
Published by NRC Research Press
Loukidis and Salgado 97

Appendix A
Table A1. Values of constitutive model parameters for Toyoura and clean Ottawa sands.

Parameter value
Parameter Clean
Parameter symbol Toyoura sand Ottawa sand Comment
Small-strain (“elastic”) n 0.15 0.15* Poisson’s ratio
parameters Cg 900 611 Parameter controlling the magnitude of the maximum shear
modulus Gmax
ng 0.400 0.437 Exponent controlling the rate of increase of Gmax with effec-
tive confining stress
g1 0.0010 0.000 65 Parameters controlling the decrease of “para-elastic” shear
a1 0.40 0.47 modulus G with shear strain
CS Gc 0.934 0.780 Intercept of CS line in e–p′ space
l 0.019 0.081 Parameter controlling inclination of CS line in e–p′ space
x 0.70 0.196 Parameter controlling curvature of CS line in e–p′ space
Mcc 1.27 1.21 Critical stress ratio in triaxial (TX) compression conditions
Bounding surface kb 1.5 1.9 Parameter controlling the increase of friction angle with
sand density
Dilatancy Do 0.90 1.31 Inclination of the stress–dilatancy curve
kd 2.8 2.2 Parameter controlling the stress ratio at phase transformation
Plastic modulus h1 1.62 2.20 Parameters controlling the magnitude of plastic modulus
h2 0.254 0.240
elim 1.00 0.81 Upper limit for void ratio for which the plastic modulus be-
comes zero
m 2.0 1.2 Parameter controlling stress ratio in undrained instability
state
Stress-induced c1 0.72 0.71 Ratio of the critical stress ratio in TX extension to that in
anisotropy TX compression
c2 0.78 0.78* Parameter controlling the value of the magnitude of inter-
mediate principal stress relative to the two other principal
stresses under plain-strain conditions
ns 0.35 0.35* Parameter controlling the magnitude of the friction angle in
plane-strain conditions relative to the friction angle in TX
compression
Inherent anisotropy a 0.29 0.31 Parameter controlling the intercept of CS line in e–p′ space
under conditions other than TX compression
kh 0.11 0.39 Parameter controlling the variation of plastic stiffness, with
the direction of loading relative to the axis of sand de-
position
m 0.05 0.05 Radius of conical yield (loading) surface
*Assumed.

Published by NRC Research Press

Cite this article Research Article Keywords: design methods & aids/
Lees AS (2020) Paper 1800116 foundations/temporary works
The bearing capacity of a granular layer on clay. Received 06/07/2018;
Proceedings of the Institution of Civil Engineers – Geotechnical Engineering 173(1): 13–20, Accepted 16/04/2019;
https://doi.org/10.1680/jgeen.18.00116 Published online 28/05/2019

ICE Publishing: All rights reserved

Geotechnical Engineering

The bearing capacity of a granular

layer on clay
Andrew S. Lees PhD, CEng, MICE
Senior Application Technology Manager, Tensar International, Nicosia,
Cyprus (alees@tensar.co.uk) (Orcid:0000-0003-2417-7532)

Granular layers are often placed over weaker clay soils to improve the bearing capacity of working platforms and
spread foundations. Their design requires the calculation of a two-layer bearing capacity. The commonly used existing
calculation models are quite empirical with imprecise input parameters and can err on the non-conservative side in
some situations. Other proposed methods tend to involve multiple design charts and are suited only to either strip
or circular foundations. In this paper a new and highly practical design method is proposed whose input parameters
are derived directly from the shear strengths of the two layers without the need for empirical-based charts.
The output can be obtained either from a single chart or a few equations that are straightforward to implement
into a spreadsheet. It can be applied quite generally to both surface and shallow embedded foundations, circular and
rectangular and with dry or saturated granular layers, and also to working platform design for tracked plant.
The method was validated using three independent numerical studies and centrifuge testing. The results showed
that, although errors can increase for foundation shapes towards square or circular and with increasing overburden,
the errors are smaller than those of the commonly used existing methods and are also conservative.

Notation support the load with an adequate margin of safety against

B foundation width or diameter bearing capacity failure. Consequently, in the design of such
D foundation embedment depth applications, there is a need to calculate the bearing capacity
g gravitational acceleration of a granular layer overlying a weaker clay soil. Typically, the
H granular layer thickness between foundation base loads are predominately vertical and assumed uniform over
and clay horizontal rectangular or circular areas. The geometry of such
ID relative density index cases is illustrated in Figure 1.
k linear increase of su with depth
L foundation length Bearing capacity failure involves punching shear through the
Nc, Nq, Nγ bearing capacity factor granular layer and a bearing capacity mechanism in the
p0′ effective vertical stress at base of granular layer underlying clay, unless the granular layer is sufficiently thick
pf gross bearing capacity of granular layer on clay for bearing capacity failure to occur entirely within the upper
Q constant layer, which is described later in this section. Two commonly
qg bearing capacity of granular layer of used calculation models are the semi-empirical Meyerhof
infinite depth (1974) or Hanna and Meyerhof (1980) method and the load
qs surface bearing capacity of clay spread or projected area method. In the former, the granular
qu net bearing capacity of granular layer on clay layer is assumed to be in a state of passive failure along
sc, sq, sγ shape factor vertical planes beneath the footing edges. Limit equilibrium
su undrained shear strength of clay analysis was used to derive design equations with a punching
su0 surface value of su shear coefficient derived empirically from model footing tests
T granular layer load transfer efficiency at 1g and presented in charts. A drawback of this method is
W weight of granular block below foundation that the charts were not presented in non-dimensional form so
within assumed load spread are appropriate only for the granular layer density and thick-
α load spread angle ness used in their preparation (Burd and Frydman, 1997).
γ weight density of granular layer
γsat saturated weight density of granular layer In the load spread method (Equation 1), the granular layer is
ϕ′ internal friction angle of granular layer assumed to spread load uniformly to the underlying clay and
′
ϕcv critical-state friction angle the shear strength contribution of the granular layer is ignored
(Terzaghi and Peck, 1948; Yamaguchi, 1963).
1. Introduction
 
Granular layers are often used in working platforms and 2H W
1: qu ¼ 1þ tan α ðsu Nc þ p00 Þ 
beneath spread foundations to distribute concentrated loads to B B þ 2H tan α
underlying weaker clay soils such that the weaker soil can

13
Geotechnical Engineering The bearing capacity of a granular
Volume 173 Issue 1 layer on clay
Lees

Gross bearing capacity, pf

(qu = pf – γ D)

D Overburden
stress, γ D

Granular layer B
Punching shear
α
H
Typical pre-failure
Self-weight vertical
load spread
effective stress, p'0,
including overburden

Bearing capacity
failure

Clay

Figure 1. Geometry and terminology of the problem

The angle α of load spread to the vertical is assumed the same
as the angle of straight shear planes in the granular layer.
Many values for α have been proposed, ranging from 18·4°
(i.e. a 3:1 load-spread) (Young and Focht, 1981) to 45°
(Myslivec and Kysela, 1978), equal to ϕ′ (Baglioni et al., 1982)
or dependent on the ratio of the bearing capacities of the sand
and clay alone (Jacobsen et al., 1977), as summarised by
Craig and Chua (1990). Brocklehurst (1993) and Ballard et al.
(2011) showed that α is also influenced by the undrained
shear strength of the clay (su). The term ‘load spread angle’
is misleading because some designers have attempted to
measure the load spread in tests with earth pressure cells at
relatively small strains and then adopted the value in bearing
capacity design even though the eventual punching shear angle
may be significantly smaller (Madhav and Sharma, 1991), as
illustrated in Figure 1, and the load spread also changes as
strains increase towards failure. The appropriate α angle could
be neither the actual load spread nor punching shear angle
but more an expression of the load transfer efficiency of the
granular layer that depends on the shear strengths of the soils.
Limit analysis has also been applied to the problem, including
upper bound analysis (Michalowski and Shi, 1995) and finite-
element limit analysis (FELA) for upper and lower bounds of
strip footings by Shiau et al. (2003) and lower bound analysis
of circular footings by Kumar and Chakraborty (2015). In the
first two studies a critical depth of the granular layer was
identified above which the failure mechanism occurred entirely
within the granular layer and bearing capacity became indepe-
ndent of clay strength. This is the H value at which qu (the net
bearing capacity of a granular layer on clay) equals qg (the
bearing capacity of a granular layer of infinite depth) and
depends on a complex interaction between the granular layer
depth and density, foundation geometry and the shear
strengths of the granular layer and the subgrade. Shiau et al.
(2003) found that the Brinch Hansen (1970) bearing capacity
factor (Nγ) provided reasonably accurate predictions of qg.
Burd and Frydman (1997) performed an extensive parametric
study of dimensionless groups using finite-difference and finite-
element analysis (FEA) of surface strip footings on granular
layers with different ϕ′ values overlying clay subgrades of
different su/p0′ values with different H/B ratios. They also ident-
ified a critical depth that matched well with that obtained
using Nγ for qg.
Michalowski and Shi (1995), Kumar and Chakraborty (2015)
and Burd and Frydman (1997) all provided design charts as
outcomes from their work but each required multiple charts
to cover the typical range of cases, making them difficult to
use. In addition, they are suitable for either strip or circular
footings, and sometimes surface footings, only. Consequently,
they do not see common use in design.
This paper combines these previous investigations with pub-
lished centrifuge test data to propose a new, simple method to
characterise the bearing capacity performance of granular layers
on clay soils that can be used for vertically loaded surface and
shallow foundations of both circular and rectangular shapes.
2. Physical study
Okamura et al. (1997, 1998) performed 60 centrifuge model tests
on circular and strip footings loaded to failure on Toyoura fine to
medium sand overlying clay subgrades of different strengths. The
strip footings applied surface load but the circular footing was
used to apply both surface and embedded loadings; the H/B ratio
was also varied. In the centrifuge tests, su increased with depth,
which needs to be converted into an equivalent uniform value.

14
Geotechnical Engineering The bearing capacity of a granular
Volume 173 Issue 1 layer on clay
Lees

Tani and Craig (1995) characterised the increased bearing 6

capacity of circular foundations on clay using a dimensionless su: kPa
ratio kB/su0 where k is the linear increase of su with depth from 5 23
a surface value su0. However, Shiau et al. (2003) studied the
specific effect of k on the bearing capacity of an overlying
4
granular layer and found that it was tempered by the overlying T

qu/qs
granular layer and even more so as H/B increased due to
3
the proportionally higher influence of the granular layer on
the overall bearing capacity, eventually to the point where the
bearing capacity mechanism occurred entirely within the 2
granular layer. Furthermore, the limit on shear stresses
generated at the interface between the layers of differing stiff- 1
ness was found to be governed by su at the surface of the clay 0 1 2 3
rather than at depth. k values ranged between about 0·5 and H/B
4 kPa/m in the centrifuge models and corrections in line with (a)
those determined by Shiau et al. (2003) were applied to the 4
surface su value. Typically, a factor of only about 1·05 was su: kPa
applied to su0 to obtain the equivalent uniform value. In design, 10
it would appear conservative and not particularly inaccurate 26
3 T
to adopt the surface su value in design except in cases with
51
(qu/qs)½

large B and k values, in which case the study by Shiau et al.

(2003) or other further analysis may prove useful.
When the bearing capacity obtained in centrifuge tests is
expressed as a dimensionless ratio of the clay surface bearing
capacity of the same foundation shape (Nc taken as 5·14 for
the strip and 6·2 for the circular foundation) and plotted
against H/B as shown in Figure 2, a linear relationship is
apparent for the range of H/B values tested whose slope is
dependent on the clay su, according to the circular footing
data. Even the tests with non-zero overburden pressure
(denoted ‘ob’ in Figure 2 and ranging between 23 and 37 kPa)
followed the same behaviour as the surface footings when the
overburden pressure γD was subtracted from the failure bearing
stress to obtain qu (qs was still calculated for a surface footing).
A more linear plot was obtained for the circular footings when
plotting the square root of qu/qs, which can be explained by
the load spread through the granular layer in two horizontal
directions instead of just one in the case of a strip footing.
The linear relationship can be expressed as shown in
Equations 2 and 3 for the strip and circular footings respect-
ively, where T is the slope of the linear relationship and
represents the load transfer efficiency of the granular layer.
qu H qg
2: ¼1þT  ðstrip footingÞ
qs B qs
 
qu H 2 qg
3: ¼ 1þT  ðsquare or circular footingÞ
qs B qs
If the granular layer self-weight stresses in Equation 1 are
assumed to cancel each other out, Equations 2 and 3 are
35 ob
2
T
1
0 1 2
H/B
(b)
Figure 2. Plots of bearing capacity ratio against H/B from
centrifuge testing reported by Okamura et al. (1997): (a) strip
footings; (b) circular footings
obtained where T = 2tanα but avoiding explicit use of the
misleading term ‘load spread angle’. Instead, the efficiency of
the granular layer for increasing bearing capacity ratio is
expressed as a T value. The ratio qu/qs starts at 1 for H/B = 0
because the foundation is directly on the subgrade but, as H/B
increases, so does the contribution of the granular layer to the
bearing capacity according to T. It is higher for low-strength
clays because shear stresses are relatively low in the granular
layer and it can maintain higher load spread. For higher clay
su values, shear stresses in the granular layer are higher at
bearing capacity failure and the load spread narrows, causing
the granular layer to appear less efficient to the extent,
possibly, where T becomes negative because the load becomes
more concentrated with depth and the punching shear planes
become inclined inwards, as demonstrated by discrete layout
optimisation (DLO) analysis conducted by Ballard et al.
(2011). The inequality is needed for those cases where
H exceeds the critical depth and shear failure occurs entirely
within the granular layer. This bearing capacity is denoted as
qg and is recommended to be calculated using the Brinch

15
Geotechnical Engineering The bearing capacity of a granular
Volume 173 Issue 1 layer on clay
Lees

Hansen (1970) equation for Nγ, which was found to give
reasonably accurate predictions of critical depth in the studies
carried out by Shiau et al. (2003) and Burd and Frydman
(1997). The lines shown in Figure 2 would eventually reach the
critical depth as H/B increases, above which failure entirely
within the granular layer rather than punching shear failure
becomes critical. The critical depth itself can be determined as
the H value when qu = qg.
The variation of T with su in the centrifuge tests is plotted
(for individual tests rather than obtaining a single T value for
each su value by linear regression from the data in Figure 2)
in Figure 3, with su expressed as the dimensionless ratio su/p0′
where p0′ is the vertical effective stress at the base of the
granular layer with zero foundation load. The granular layers
were submerged with water in the centrifuge models so the use
of effective stress is an important distinction while the use of
total vertical stress was found to give less correlation between
T and su. It is apparent that the circular footing data for all the
different clay strengths, overburdens and H/B ratios plot close
to the curve shown determined for a granular layer ϕ′ value of
40° later in this paper, except for those cases where H exceeded
the critical depth and failure was observed to occur entirely
within the granular layer. This figure shows more clearly the
dependency of the T value on su. The small number of data
points for the strip footing plot well above the curve, which is
likely due to the higher ϕ′ value in plane strain compared with
the triaxial compression conditions of the circular footing.
Indeed, in seven surface circular footing centrifuge tests on
overburden stress. The FELA study carried out by Kumar and
Chakraborty (2015) was on circular footings in axisymmetry
and included variable overburden stress. All of them assumed a
rigid footing and a rough interface with the granular layer.
Outputs from the three studies are compared with each other
and the Okamura et al. (1997) centrifuge data in Figure 4 for
a granular layer with ϕ′ = 40°, which was included in all three
numerical studies. The cases where H exceeded the critical
depth are omitted in Figure 4. The Burd and Frydman (1997)
output is denoted as FEA and the Shiau et al. (2003) and
Kumar and Chakraborty (2015) studies as FELA for strip and
circular footing cases, respectively. The centrifuge data is
denoted C. The peak ϕ′ values were determined according to
Equation 4 to take account of the stress dependency of ϕ′
(where Q is a constant taken as 10 for silica sands) and ranged
between the values shown in brackets in Figure 4. The FEA
and FELA outputs for the strip footings with zero overburden
(D/B = 0) in Figure 4(a) compare very well with each other and
lie on the same curve drawn in Figure 4. With the addition
of overburden, the T value increased slightly, despite the nor-
malisation of su with p0′ . The T values in the circular footing
5
4
3 FEA D/B = 0 FELA D/B = 0 FELA D/B = 1
2
T

Toyoura sand, Okamura et al. (1997) obtained Nγ values

between 127 and 160 but in two tests with strip footings, Nγ
1
values of 179 and 252 were obtained.
0
3. Numerical studies 0 2 4 6 8
The results of three numerical studies are now presented. –1 su /p'0
The finite-difference and FEA study conducted by Burd and (a)
Frydman (1997) and the FELA study of Shiau et al. (2003) were
on strip footings in plane strain with the latter including variable 5

C D = 0 (35–41) C D > 0 (34–38)

4
2·5 FELA D/B = 0 FELA D/B = 0·25
Strip Circular Circular overburden > 0 FELA D/B = 0·5
3
2·0

1·5 2
T
T

1·0 1

0·5
0
0 2 4 6 8
0
0 1 2 3 4 –1 su /p'0
su /p'0 Failure entirely (b)
within granular layer
Figure 4. Numerical and physical study of T values compared for
Figure 3. Plot of T against normalised clay shear strength from ϕ0 = 40° (centrifuge ϕ0 = 34–41°): (a) strip footing; (b) circular
centrifuge testing reported by Okamura et al. (1997) footing

16
Geotechnical Engineering The bearing capacity of a granular
Volume 173 Issue 1 layer on clay
Lees

case (Figure 4(b)) are slightly higher than the same curve
drawn and, in a similar fashion to the strip footing, increase
further with the addition of overburden stress. The results also
compare well with the centrifuge data, which generally show
slightly lower ϕ′ values. According to Lee et al. (2013)
4: ϕ0 ¼ ϕcv
0
þ 265½ID ðQ  ln qu Þ  1
Burd and Frydman (1997), Shiau et al. (2003) and Kumar and
Chakraborty (2015) also considered other ϕ′ values for the gran-
ular layer and the outputs from these (all with zero overburden)
are plotted for the strip (B/L= 0) and square (B/L= 1) footing
cases in Figures 5(a) and 5(b), respectively. The square footing
T values were calculated using the B value giving the same
foundation area as the circular footing simulated in the FELA.
This resulted in a slightly lower T value (by a factor of √(π/4))
for the square footing compared with the equivalent area circu-
lar footing. The numbers shown in the legend of Figure 5
denote the ϕ′ value in degrees. In addition to the ϕ′ = 40° curve
shown in Figure 4, further curves are shown for ϕ′ = 45, 35 and
30°. It can be seen that these curves fit the strip footing data
well. However, the outputs for the square footing cases (B/L = 1)
fit the curves less well and were more variable, suggesting some
dependency on other variables. However, for the intended practi-
cal design use, the curves provide an appropriate lower bound.
The centrifuge data are again plotted on the graphs to show
that the strip footing values compare well with the numerical
studies and the peak ϕ′ values were determined using Equation
4 multiplied by 1·2 to take approximate account of the plane
strain conditions. Additional centrifuge data (shown by the
open circle symbols) are added to Figure 5(b). These data were
taken from the work of Lee et al. (2013) who performed circular
footing penetration tests on a fine sand layer overlying clay of su
around 17 kPa in a drum centrifuge.
The curves were plotted according to the power-law expression
shown in Equation 5 and linear relationships between A and B
and ϕ′ were determined as shown in Equations 6 and 7 (with
ϕ′ in radians) to fit the output data.
 A
su
5: T ¼ 14 þB
p00

3·5 6: A ¼ 041ϕ0  018

3·0
FEA 32 FEA 40 FEA 48
2·5
FELA 30 FELA 40 FELA 50 7: B ¼ 42ϕ0  34
2·0 C D = 0 (44–48)

1·5
T

1·0 These allow direct determination of T from the ϕ′ of the

φ'
0·5 granular layer and are considered applicable for circular and
45°
40° rectangular foundations with horizontal bases and vertical
0
35° loading across the range of values shown in Figure 5 for which
–0·5 30°
0 1 2 3 4 5 6
there are numerical analysis outputs from independent sources.
su /p'0 Peak ϕ′ values are only appropriate where the strain level
(a) required to cause bearing capacity failure in the clay does not
exceed the peak failure strain in the granular layer. With soft
3·5 clay subgrades, shear strains in the granular layer at bearing
C D = 0 (35–41) C D > 0 (34–38) capacity failure could well exceed around 10%, which can be
3·0
FELA 30 FELA 35
2·5 higher than the peak failure strains in dense granular soils.
FELA 40 FELA 45
Consequently, post-peak shear strengths are usually more
2·0 C D = 0 (37–40)
appropriate for the granular layer.
1·5
T

1·0 The curves plotted in Figure 5 and defined in Equations 5–7 can
φ' be used to determine T values which, in turn, using Equations 2
0·5
45° and 3, can be used to determine the bearing capacity of a granu-
0 40°
35° lar layer overlying clay in the design of foundations and working
–0·5 30° platforms. This is demonstrated by examples in Section 5. For
0 1 2 3 4 5 6
B/L ratios between 1 and 0, the bearing capacity can be interp-
su /p'0
olated linearly, as verified by Lees (2017) in a study of different
(b) B/L ratios using three-dimensional FEA.
Figure 5. Variation of T with su and ϕ0 : (a) strip footing;
(b) square footing Key advantages of this approach are the simple derivation of
the bearing capacity directly from the shear strength of the

17
Geotechnical Engineering The bearing capacity of a granular
Volume 173 Issue 1 layer on clay
Lees

individual soil layers and the straightforward visualisation of
the contribution of the individual strengths to overall bearing
capacity, as in Figure 5 for example. The relationships also
appear to be valid, or else err on the conservative side, for
foundations with overburden pressure, for all shape ratios and
when the granular layer becomes saturated.
T values become very sensitive to changes in su/p0′ at low su/p0′ ,
so sensitivity analysis would be particularly recommended in
such cases and the design values of clay strength selected care-
fully. The uncertainty in such cases is increased by the difficulty
of obtaining reliable su values in soft clays and the susceptibility
of su to change due to variations in environmental conditions.
4. Comparison with existing methods
The Meyerhof (1974) method is recommended for routine
working platform design in BRE (2004). The T values
calculated from the Meyerhof (1974) method for a range of su
and H/B values with ϕ′ values of 35, 40 and 45° for strip and
circular footings are plotted in comparison with the newly
derived curves in Figures 6(a) and 6(b) for an overburden
stress (γD) of zero and 20 kPa, respectively. The punching
5 et al. (1998) back-analysed their centrifuge test results to
4 45° strip 40° strip 35° strip
45° circ. 40° circ. 35° circ.
shear coefficient Kptan δ was calculated according to Meyerhof
(1974), taking δ as 0·67ϕ′. The T values obtained agree reason-
ably well with the new curves but with some notable excep-
tions. The Meyerhof (1974) method appears to overpredict the
bearing capacity for strip footings with ϕ′ = 45° and low su/p0′ ,
which was addressed somewhat with the empirical factors of
Hanna and Meyerhof (1980) but not in a dimensionless way.
There is also a large difference between the T values of strip
and circular footings, perhaps due to the constant shape factor
of 2 that was applied to obtain the circular footing bearing
capacity. Furthermore, due to the assumption of a vertical
punching shear plane, the T values are constrained to be
positive and approach zero for all ϕ′ values as su/p0′ increases,
although the numerical modelling discussed in this paper as
well as the DLO studies of Ballard et al. (2011) show that
negative T values can be obtained for cases with relatively high
su/p0′ but H below the critical depth. As shown in Figure 6(b)
and noted by Michalowski and Shi (1995), with the addition
of overburden stress, the Meyerhof (1974) and Hanna
and Meyerhof (1980) methods can significantly overpredict
the bearing capacity, particularly at high ϕ′ values. This is
due to the assumption that strength interaction parameters are
uninfluenced by the overburden stress; however, Okamura
demonstrate that the punching shear coefficient Ks is influ-
enced by overburden stress.

3
A similar comparison with the load spread method is shown
2 for strip footings with assumed load spread angles of 26·6°
T

(2:1) and 45° (1:1) in Figure 7. The corresponding T values

1 φ' were found to be 1 and 2, respectively, but with an increase at
45° low su/p0′ values due to the increasing effect of the granular
0 40°
35° layer self-weight correction in Equation 1 as su decreases.
Clearly, each load spread angle is applicable for only a very
–1
0 2 4 6 8 10 narrow range of parameter combinations and adoption of a
su /p'0 single value across different designs could result in significant
(a) over- or underprediction of bearing capacity.

5
5
45° strip 40° strip 35° strip
4
45° circ. 40° circ. 35° circ. 2:1
4
3 1:1
3
2
T

2
T

1 φ'
45° 1 φ'
0 40°
35° 45°
0 40°
–1 35°
0 2 4 6 8 10
–1
su /p'0 0 2 4 6 8 10
(b) su /p'0

Figure 6. Comparison of the Meyerhof (1974) method and new Figure 7. Comparison of load spread method and new design
design curves: (a) surface footing; (b) 20 kPa overburden stress curves

18
Geotechnical Engineering The bearing capacity of a granular
Volume 173 Issue 1 layer on clay
Lees

5. Examples & It is versatile: the method is applicable to all rectangular

Two worked examples are shown in Tables 1 and 2 for a
working platform and an embedded foundation, respectively.
At the end of each calculation, the outputs are compared with
those of the commonly used Meyerhof (1974) and load spread
methods to help illustrate the deficiencies of these methods.
6. Conclusions
A new practical method to calculate the bearing capacity of a
granular layer overlying clay simply from the shear strengths of
the two soils has been proposed.
The important advantages of this new approach include the
following.
& It is simple: just one design chart can be used for all cases,
instead of the multiple charts in more recently published
design methods.
shape ratios, embedded and surface, dry and saturated
granular layers, and to a wide range of clay and granular
layer strengths.
& All the relationships are dimensionless rather than derived
from a narrow set of test conditions.
& It is understandable: the influence of both the granular
layer and clay strength on the bearing capacity can be
visualised easily in one chart.
& It is programmable: simple design equations allow
straightforward spreadsheet calculation.
& It has been validated based on a review of multiple and
independent numerical and physical modelling studies.
Comparisons with numerical analyses show that errors can
increase for foundation shapes towards square or circular
and with increasing overburden; however, these errors are con-
servative and smaller than those commonly found in existing

Table 1. Calculation example 1

Example 1: Track loading on working platform
Main points: rectangular surface load, dry granular layer

Input data: B = 0·6 m, L = 3·8 m, su = 7 kPa, ϕ0 = 43°, γ = 18 kN/m3, H = 0·75 m

p00 = 0·75  18 = 13·5 kPa
su/p00 = 7/13·5 = 0·519
T = 1·68 (from Figure 5(a) or Equations 5–7)
B/L = 1: B/L = 0:
qs = scNcsu = 1·2  5·14  7 = 43·2 kPa qs = scNcsu = 1·0  5·14  7 = 36·0 kPa
Equation 3: qu/qs = [1 + T(H/B)]2 = [1 + 1·68(0·75/0·6)]2 = 9·61 Equation 2: qu/qs = [1 + T(H/B)] = [1 + 1·68(0·75/0·6)] = 3·10
qu = 43·2  9·61 = 415·2 kPa qu = 36·0  3·10 = 111·6 kPa
B/L = 0·6/3·8 = 0·158
Interpolation between B/L = 1 and B/L = 0:
qu = 159·6 kPa
qg = γDNqsq + 0·5BγNγsγ = 0 + 0·5  0·6  18  165  0·94 = 832 kPa
qu ≤ qg? Yes
qu taken as 159 kPa (Meyerhof (1974) method: 213 kPa; 2:1 load spread method: 90 kPa)

Table 2. Calculation example 2

Example 2: Shallow spread foundation
Main points: rectangular load, embedded foundation, saturated granular layer

Input data: B = 1·2 m, L = 2·0 m, su = 25 kPa, ϕ0 = 45°, γsat = 20 kN/m3, H = 0·60 m, D = 1·0 m, groundwater level at ground surface
γD = 20  1·0 = 20 kPa
p00 = (1·6  20)−(1·6  10) = 16 kPa
su/p00 = 25/16·0 = 1·56
T = 1·02 (from Figure 5(a) or Equations 5–7)
B/L = 1: B/L = 0:
qs = scNcsu = 1·2  5·14  25 = 154·2 kPa qs = scNcsu = 1·0  5·14  25 = 128·5 kPa
Equation 3: qu/qs = [1 + T(H/B)]2 = [1 + 1·02(0·6/1·2)]2 = 2·28 Equation 2: qu/qs = [1 + T(H/B)] = [1 + 1·02(0·6/1·2)] = 1·51
qu = 154·2  2·28 = 351·6 kPa qu = 128·5  1·51 = 194·0 kPa
B/L = 1·2/2·0 = 0·6
Interpolation between B/L = 1 and B/L = 0:
qu = 288·6 kPa
qg = γDNqsq + 0·5BγNγsγ = (20  135  1·12) + (0·5  1·2  10  241  0·76) = 3024 + 1099 = 4123 kPa
qu ≤ qg? Yes
qu taken as 288 kPa (Meyerhof (1974) method: 388 kPa)
Gross bearing capacity = 269 + 9·5 = 278 kPa

19
Geotechnical Engineering
Volume 173 Issue 1
methods. The Meyerhof (1974) and load spread methods can
result in an unsafe overprediction of bearing capacity in some
cases.
Other observations made during this study can be summarised
as follows.
& In cases where su increases with depth, it is conservative
and not particularly inaccurate to adopt the surface su
value in design except in cases with large foundation
widths and a rapid increase in su with depth when more
detailed study may prove economical.
& A check on shear failure entirely within the granular layer
is needed and studies show the Brinch Hansen (1970)
equation for Nγ is the most appropriate.
& Shear strain levels at the onset of bearing capacity failure
in weaker clay typically exceed peak failure values in dense
granular layers, so post-peak granular layer strengths are
generally more appropriate in design.
& For circular foundations, the T value should be multiplied
by 1·13.
REFERENCES
Baglioni VP, Chow GS and Endley SN (1982) Jack-up foundation stability
in stratified soil profiles. Proceedings of the 14th Offshore
Technology Conference, Houston, TX, USA. Offshore Technology
Conference, Houston, TX, USA, pp. 363–383, https://doi.org/
10.4043/4409-MS.
Ballard JC, Delvosal P, Yonatan P, Holeyman A and Kay S (2011)
Simplified VH equations for foundation punch-through
sand into clay. Frontiers in Offshore Geotechnics II (Gourvenec S
and White D (eds)). CRC Press, Boca Raton, FL, USA,
pp. 655–660.
BRE (Building Research Establishment) (2004) Working Platforms for
Tracked Plant, BR 470. BRE, Watford, UK.
Brinch Hansen J (1970) A revised and extended formula for
bearing capacity. Bulletin of the Danish Geotechnical Institute
28: 5–11.
Brocklehurst CJ (1993) Finite Element Studies of Reinforced and
Unreinforced Two-Layer Soil Systems. DPhil thesis, University of
Oxford, Oxford, UK.
Burd HJ and Frydman S (1997) Bearing capacity of plane-strain
footings on layered soils. Canadian Geotechnical Journal 34(2):
241–253.
Craig WH and Chua K (1990) Deep penetration of spud-can foundations
on sand and clay. Géotechnique 40(4): 541–556, https://doi.org/
10.1680/geot.1990.40.4.541.
Hanna AM and Meyerhof GG (1980) Design charts for ultimate bearing
capacity of foundations on sand overlying soft clay. Canadian
Geotechnical Journal 17(2): 300–303.
Jacobsen M, Christensen KV and Sorensen CS (1977) Gennemlokning af
tynde sandlag. Svenska Vagoch Vatten Byggares Riksforbund,
Stockholm, Sweden, pp. 23–25 (in Danish).
Kumar J and Chakraborty M (2015) Bearing capacity of a circular
foundation on layered sand-clay media, Soils and Foundations
55(5): 1058–1068.
Lee KK, Cassidy MJ and Randolph MF (2013) Bearing capacity on sand
overlying clay soils: experimental and finite-element investigation
of potential punch-through failure. Géotechnique 63(15):
1271–1284, https://doi.org/10.1680/geot.12.P.175.
20
The bearing capacity of a granular
layer on clay
Lees
Lees AS (2017) Bearing capacity of a stabilised granular layer on clay
subgrade. In Bearing Capacity of Roads, Railways and Airfields
(Loizos A, Al-Qadi I and Scarpas T (eds)). Taylor & Francis,
London, UK, pp. 1135–1142.
Madhav M and Sharma J (1991) Bearing capacity of clay overlain by
stiff soil. Journal of Geotechnical Engineering 117(12): 1941–1948.
Meyerhof GG (1974) Ultimate bearing capacity of footings on sand
layer overlying clay. Canadian Geotechnical Journal 11(2): 223–229.
Michalowski RL and Shi L (1995) Bearing capacity of footings over
two-layer foundation soils. Journal of Geotechnical Engineering
121(5): 421–428.
Myslivec A and Kysela Z (1978) The Bearing Capacity of Building
Foundations. Elsevier, Amsterdam, the Netherlands.
Okamura M, Takemura J and Kimura T (1997) Centrifuge model tests on
bearing capacity and deformation of sand layer overlying clay.
Soils and Foundations 37(1): 73–88.
Okamura M, Takemura J and Kimura T (1998) Bearing capacity
predictions of sand overlying clay based on limit equilibrium
methods. Soils and Foundations 38(1): 181–194.
Shiau JS, Lyamin AV and Sloan SW (2003) Bearing capacity of a sand
layer on clay by finite element limit analysis. Canadian
Geotechnical Journal 40(5): 900–915.
Tani K and Craig WH (1995) Bearing capacity of circular foundations on
soft clay of strength increasing with depth, Soils and Foundations
35(4): 21–35.
Terzaghi K and Peck RB (1948) Soil Mechanics in Engineering Practice,
1st edn. Wiley, New York, NY, USA.
Yamaguchi H (1963) Practical formula of bearing value for two layered
ground. Proceedings of the 2nd Asian Regional Conference of
Soil Mechanics and Foundation Engineering, Tokyo, Japan
(Kogakkai D (ed.)) Kenkyusha, Tokyo, Japan, pp. 99–105.
Young AG and Focht JA (1981) Sub-surface hazards affect mobile
jack-up rig operations. Sounding, McClelland Engineers 3(2): 4–9.
How can you contribute?
To discuss this paper, please email up to 500 words to the
editor at journals@ice.org.uk. Your contribution will be
forwarded to the author(s) for a reply and, if considered
appropriate by the editorial board, it will be published as
discussion in a future issue of the journal.
Proceedings journals rely entirely on contributions from the
civil engineering profession (and allied disciplines).
Information about how to submit your paper online
is available at www.icevirtuallibrary.com/page/authors,
where you will also find detailed author guidelines.
 C A N . G E O T E C H J . V O L . 1 1 , 1974

-i 0 -- and circles. The corresponding value of qu is given

by Eq. [4] after multiplying the first term of the right
hand side ( 1 +0 . 2 B / L ) and the second term by
(1 + B / L ) , with a maximum interpolated between
Eqs. [5]and [7].
The above analyses indicate that the critical thick-
ness ratio H / B of the sand layer for the maximum
value of q, depends on the bearing capacity ratio
9 >>9 q d q t , friction angle @, the shape and depth/width
( 0 ) THIN LAYER (b) THICK LAYER ratio D / B of the footing. This critical ratio H / B is
FIG. 1. Failure of soil below footing on dense found to vary theoretically from the thickness ratio
sand layer above soft clay. Hr/B of the failure surface for a homogeneous thick
bed of sand when q J q t = 1 (Fig. l b ) to about 2 to
3 times that value for q d q t = 0.
Further, The ultimate bearing capacity of a footing group
is the smaller value of either the sum of the ultimate
loads of the individual footings or the bearing capacity
where K , = coefficient of passive earth pressure. of an equivalent raft foundation consisting of the
The angle 6 decreases from about @ near the footings and enclosed sand mass, on which the
footing edge to about zero at the clay surface. Trial punching shear resistance acts, only on the perimeter
calculations using a logarithmic spiral failure surface of the equivalent raft.
in the sand have shown that the average value of 6
is generally in the range of about @ / 2to 3 @ / 4so that Tests
an approximate value of 2@/3 may be suggested, as Model tests on rough strip and circular foot-
was found in the uplift case. In practice it is con-
venient to use a coefficient K , of punching shearing ings of 3 in. width were made on the surface
resistance on the vertical plane through the footing and at shallow depth in dense sand layers over-
edge so that lying clay of various strengths at the Nova
[31 K , tan @ = K , tan 6 Scotia Technical College (Commissiong 1968;
Ho (1973). The inorganic clay of medium plas-
Substituting Eqs. [2]and [3]into [ I ] ticity used was made more brittle by the addi-
[4] q. = c N , + yH2(1 + 2 D / H ) K Btan @ / B+ yD tion of 2% of lime. After packing the clay into
with a maximum of the large footing test box it was allowed to cure
for a week before the layer of well-graded
clean sand was placed on the clay. The footings
where N, and N , are bearing capacity factors. were loaded to failure within a few minutes to
The punching shear coefficients K , have been de- minimize moisture migration, and the ultimate
termined from the corresponding earth pressure bearing capacity occurred at a settlement of
coefficients K , (Caquot and Kerisel 1949) and are
shown in Fig. 2 for various friction angles @ of the 10-20% of the footing width. Typical test
sand. It is of interest to note that K , increases rapidly results (Fig. 3) show the rapid increaee of the
with @ from about one to two times the Rankine ultimate bearing capacity with thickness of the
value of tan2(45" + @ / 2 ) . sand layer from a minimum for a footing on
The analysis for strip footings can be extended to clay to a maximum for a thick sand deposit.
circular footings by determining the passive resistance
P, inclined at 6 on a vertical cylindrical surface In order to compare the results of the present
through the footing edge (Fig. l a ) . Thus, Eq. [4] and some earlier tests (Tcheng 1957; Yama-
becomes guchi 1963) with the proposed theory, the
[61 g ~= 1.2cNC+ 2yH2(1+ 2 D / H ) s K , tan @ / B+ yD experimental data have been analyzed to de-
termine the punching shear coefficient K, in
with a maximum of Eq. [4] and sK, in Eq. [6] for strip and circu-
lar footings, respectively, on sand layers of
where s = shape factor governing the passive earth thicknesses smaller thanthe critical value. The
pressure on a cylindrical wall. Although the corre- test results after a small correction for the
sponding value of s can be estimated from approxi- effect of footing settlement at failure (Fig. 4)
mate earth pressure theories, it may be conservatively show that for footings on the surface of dense
taken as unity, especially for small ratios of H / B .
The ultimate bearing capacity of rectangular foot- sand in particular, the deduced coefficients K,
ings of width B and length L can be obtained by ixrease as the thickness of the sand layer de-
interpolation between the bearing capacities of strips creases, which may be explained by an increase
 MEYERHOF: ULTIMATE BEARING CAPACITY

Ln 30 I I I
5 Model Tests (see Fig.4):
od
X Str~p 0 x
xLn2 5 - o Circle !Y 50
X d Diepth ~ r e s l e n ~t e i t s 8: 1 3 i n . ( ~ ' = 4 7 ' )

.
W
[r
4 Field Records : :: I Width \ x Strip o Circle
20 - Brown&Poterson(l964)
0
Dovisson (1972)
-
Z
1 15 -
2
3
,
a Theory

wu 4- Y o r n o g u c h i (1963):B=Z-4 1 n l 9 =36')
0
LL
W
0

ANGLE OF INTERNAL FRICTION

THICKNESS OF LAYER BELOW FOOTING H
WIDTH OF FOOTING I3
FIG. 2. Coefficients of punching shearing resis- FIG. 4. Experimental puncyng shear coefficients
tance. for model tests.

agree fairly well with the proposed theory,

-.-
"i Width
LAYER THICKNESS BELOW FOOTING
WIDTH OF FOOTING
FIG.3. Typical results of model footing tests on
dense sand overlying clay.
of the friction angle + at small overburden
pressures.
The minimum values of K, and sK, of foot-
ings at the surface and shallow depth in dense
sand are roughly the same and indicate a shape
factor of approxirnatcly unity. Using these
results the deduced coefficients K , are found to
which is someyhat con's&vative for large fric-
tion angles (Fig. 2 ) .
Field Observations
In the Eastern Townships of Quebec a stor-
age tank of 70 ft diameter was resting on 12 ft
of thick loose sand fill overlying a 35 ft thick
stratum of highly plastic sensitive Leda Clay,
and the foundation failed by punching into the
clay during first loading at a base pressure of
2390 p.s.f. (Brown and Paterson 1964). Using
a 2 to 1 spread of the load through the fill, an
average undraincd shear strength of 530 p.s.f.
for the clay, and an average surcharge of
280 p.s.f., the authors obtained a factor of
safety of 1.1 8 on the bearing capacity of the
clay at failure.
An alternative estimate can be made by
using the present analysis to deduce the co-
H efficient of punching shear K, in conjunction
B with a reduccd shear strength of the clay.
Analysis of short term failures on highly plastic
clays (Bjerrum 1972) has shown that the
cffects of anisotropy, progressive failure, and
rate of loading may reduce the undrained shear
~trengthby about 15% for the present plas-
ticity index of 35% for the clay. Substituting
the above data into Eq. [6] 2390 110 X 12 +
= 0.85 x 530 x 6.05 +
2 x 110 x 12'sK.
tan 4/70 + 280 gives sK, tan 4 = 1.55 at
failure. For an estimated friction angle 4 of the
 226 C A N . GEOTECH J . VOL. 1 1 , 1974
loose fill of about 30' and a shape factor
s = 1 , the deduced K, would be about 2.7,
which appears reasonable.
The second case record refers to a silo in the
St. Lawrence River Valley (Davisson 1972).
The structure is 150 ft wide by 210 ft long
and rests on a piled raft foundation at a depth
of 5 ft in 7 0 ft of thick loose to compact sand
f ll ovellying a 140 ft thick Leda Clay stratum.
The clay has an average plasticity index of
40% and an undrained shear strength in-
creasing from about 1600 p.s.f. at the surface
to about 2500 p.s.f. at a depth of about 80 ft.
Thc rapid settlement and tilting about the
major axis of the structure on first loading
and subsequcnt dredging of the dock area in-
dicated at least local bearing capacity failure.
Using thc total structural and fill loading of
9120 p.s.f. at dredge level and the surface
shear strength of the clay, the author con-
cluded that base failure had occurred.
If the piled raft foundation is considered as
an equivalent pier at dredge level and the
average shcar strength of the clay in the failure
zone is used with a 15% reduction as above,
the ultimate bearing capacity of the rectangu-
lar foundation is estimatcd to be q,, = ( 1 + not yet available. However, on the simplifying as-
0.2 X 150/210) 0.85 X 2100 X 5 = 10200
p.s.f. which is greater than the applied load.
Accordingly it may be suggested that local
edge failure of the clay has led to the tilting of
the silo by a portion of the equivalent pier
punching through the 28 ft thick layer of com-
pact submerged sand between dredge level and
the Leda Clay. The corresponding coefficient
K, can be deduced for various effective re-
duced widths of the foundation to give the
maximum value of K,, which is found for a
strip footing of 55 ft width or about one-third
of the base width of the equivalent pier.
For this type of edge failure the punching
shearing resistance is mobilized on one side
only of the equivalent footing, and substituting
this data and corresponding undrained shear
strength of 1800 p.s.f. of the clay into the
modified Eq. [4] 9120 = 0.85 X 1800 X 5.14
+ 55 x 28'K, tan + / 2 x 55 gives K , tan + =
3.2 at failure. Using an estimated + of about
35" for the compact sand, the deduced K ,
is about 4.5. This value of K, as well as that
deduced from the other field record support
the corresponding theoretical coefficients, as
shown in Fig. 2.
Loose Sand on Stiff Clay
Theory
I f the ultimate bearing capacity of a thick bed of
sand ( q t ) is much less than that o f the clay (q,),
the bearing capacity o f a footing on a thin sand
layer may be estimated on the simplifying assumption
that the layer rests on a rigid base. A solution o f this
limiting case has been obtained for a rough strip
footing (Mandel and Salencon 1972) and the cor-
responding failure zones are shown in Fig. 5a for a
rough base. The sand mass beneath the footing fails
laterally by squeezing, and the ultimate bearing
capacity may be expressed by
with a maximum o f
In Eq. [8] N,' and N,,' are modified bearing ca-
pacity factors, which dependkon $, the ratio o f H/B,
and the degree o f roughness o f the rigid base. The
theoretical factors N,' add N,,' for a rough base are
shown in Fig. 6. These factors increase rapidly with
$ from the lower limits of N , and N,,, respectively,
and as the ratio o f H / B decreases from the depth
ratio o f H f / B o f the failure surface in a thick bed of
sand. The values o f H f / B are given in the lower part
o f Fig. 6 for the factors N , and N,,.
A solution for the bearing capacity o f a circular
footing on a thin sand layer resting on a rigid base is
sumptions that in radial planes the stresses and shear
zones are identical to those in transverse planes o f a
corresponding strip footing (Fig. 5a), the contact
pressure distribution beneath the circle may be as-
sumed to be similar to that for a strip (Meyerhof and
Chaplin 1953). On this basis an approximate estimate
of the ultimate bearing capacity o f a rough circular
footing can be obtained from the average contact
pressure at failure and may be expressed by
with a maximum o f
In Eq. [lo] s,' and s,,' are modified shape factors
for circular footings, and the approximate values
which have been determined by the proposed analysis
are shown in Fig. 7 . These factors decrease rapidly
FIG. 5. Failure of soil below footing on loose
sand layer above stiffclay.
 MEYERHOF: ULTIMATE BEARING CAPACITY
LAYER THICKNESS !-! footings (Meyerhof 1951). Thus, Eq. [8] can be
FOOTING WIDTH B modified by multiplying the first term o f the right
10000 I I hand side by [ l - ( 1 - s,')B/L] and the second term by
.-G- I /0.2 [ I - ( 1 - s,~')B / L ] to obtain the ultimate bearing
Z 5000
/ i/ capacity o f a rectangle, with a maximum value inter-
polated between Eqs. [9]and [I 11.
As the bearing capacity o f the sand ( q t ) ap-
proaches that o f the underlying clay (q,,),the failure
surface beneath a footing resting on a thin layer o f
sand extends into the clay and the shear zones be-
come discontinuous at the interface (Fig. 5b). The
corresponding ultimate bearing capacity may be esti-
mated by using an empirical parabolic interaction
relationship o f

with a maximum for H / B = 0 when q,, = qs given

by Eqs. [9] and [ I l l , and a minimum for H / B 2
Ht/B when q,, = qt given by Eqs. [5] and [7] for
strip and circular footings, respectiv~ly.
I f a uniform load is carried by'a flexible founda-
tion, the possibility o f a local edge failure has to be
considered by using an effectiveheduced width o f the
foundation, as shown above.
I I I
I ' I
Tests
20' 25' 30' 35' 40' 45'
Model tests as mentioned before were car-
ANGLE OF INTERNAL FRICTION + ried out on strip and circular footings resting
FIG. 6. Theoretical modified bearing capacity fac- on compact sand layers overlying firm clay
tors for strip footings. (C2missiong 1968). The test results (Fig. 8 )
LAYER THICKNESS 5 show a decrease of the ultimate bearing capa-
FOOTING WIDTH B city with greater sand layer thickness from a
maximum for a footing on clay to a minimum
for a thick sand deposit and the observations
support the relationship of Eq. [12]. Figure 8
also shows a special test in which the bearing

X Strip Cloy Cohesion cu =4.0 p.s.i.

0 Circle Sand F r i c t i o n A n g l e $
50 Width B = 3 in.
G-

ANGLE OF INTERNAL FRICTION 4

FIG. 7. Theoretical modified shape factors for
circular footings.
from the conservative upper limits o f s, = 0.6 and
s,,= 1.0 for a thick bed of sand, as @ increases and
the ratio H / B decreases, and both factors tend to
similar values for small ratios o f H / B .
Similarly, the ultimate bearing capacity o f a rec-
tangular footing on a thin sand layer resting on a LAYER THICKNESS BELOW FOOTING 5
rigid base can be estimated by assuming that below
WIDTH OF FOOTING B
the central part of the rectangle the contact pressure
is the same as for a strip and that at the ends the FIG.8. Results o f model footing tests on compact
contact pressures are similar to those below circular sand overlying clay.
228 C A N . GEOTECH. J . VOL. l I , 1974

capacity ratio q,,/q, was unity. In this case u .

z 2000
Present Tests x N:
I o
failure of the footing appeared to be a com- N;
oZI
-23 '\ 'I M I ~ O V ITourmerK71).
C~ i-N (B=&)
bination of punching and sequeezing of the
sand layer. The corresponding bearing capa-
Z
m
1000 '"I
" heor or^ - -Ni
[li
city increased with greater sand layer thickness
from about H / B = 1 to a maximum at about
H / B = 1.5 and then decreased to the bearing >
capacity of thick sand deposit at a ratio of about t
H / B = 2.5.
4
3
200---
In order to investigate the limiting case of U

the bearing capacity of sand layers on a rigid

base, a number of strip and circular footing 4
tests were made on the surface and at shallow
depth in sand layers resting on a rough rigid
baze. Typical test results are shown in Fig. 9.
The rapid increase of the bearing capacity with
smaller sand layer thickness led to some crush-
ing of the sand in the central zone beneath the
footing and the corresponding friction angle + LAYER THICKNESS BELOW FOOTING H
WIDTH OF FOOTING €3
was expected to decrease from the initial value.
The present and some earlier tests (Milovic FIG. 10. Comparison of theoretical and experi-
mental modified bearing capacity factors.
and Tournier 197 1 ) can be compared with the
propozed theory by analyzing the experimental deduced from the tests on sand layers resting
results to deduce the modified bearing capacity on a rigid base are shown in Fig. 10. Compari-
factors N,' and N,,' and the modified shape son of these factors for footings at various
factors s,' and s,,' in Eqs. [8] and [lo]. The ratios of H / B at failure indicates reasonable
modified factors N,' and N,,' which have been agreement with the theoretical values, es-
pecially in view of the difficulty in obtaining
representative friction angles + from conven-
-
.-
400.. "
For Legend see Fig. 3 tional shear tests. The experimental modified
VI (+=3a0) shape factors (Fig. 11) also agree fairly well
& 200 I I
with the proposed analysis, although the factors
', tend to be smaller and the factors s,,' tend
s

I I I I I
0 0.5 1.0 1.5 2-0 00

LAYER THICKNESS BELOW FOOTING H LAYER THICKNESS BELOW FOOTING H

WIDTH OF FOOTING 8 WIDTH OF FOOTING B
FIG. 9. Typical results of model footing tests on FIG. 11. Comparison of theoretical and experi-
sand layer on rigid base. mental modified shape factors.
300 CAN. GEOTECH. J. VOL. 17, 1980

Design charts for ultimate bearing capacity of

foundations on sand overlying soft clay

Departtnerzt of Civil Etzgitleeritzg, Concordia Utliversity, Morltreal, P.Q., Carlada H3G I M 8

AND
G. G. MEYERHOF
Departt??etltof Civil Etlgirzeerit~g,Nova Scoria Teclitrical College, Halifax, N.S., Carlnda B3J 2x4
Received August 15, 1979
Accepted January 4, 1980

It is often the case that the base of a footing rests in a foundation material consisting of more
than one layer. Problems of this type have been under investigation by the authors for the last
few years, and the results of these studies were reported for footings on two layers of soil and
for footings on three layers of sand. This paper is an attempt to extend the authors' previous
theory to cover the case of footings resting on a subsoil consisting of a dense layer of sand over-
lying a soft clay deposit. The results of this analysis are presented in the form of design charts.

I1 est frCquent qu'une fondation superficielle repose sur un matCriau stratifie. Des problkmes
de ce type ont CtC CtudiCs par les auteurs depuis quelques annCes et les rCsultats de ces Ctudes
ont CtC rapportis pour le cas des semelles sur fondations bi-couches et pour des semelles sur un
tri-couches de sable. Cet article tente de gCnCraliser la thCorie antCrieure des auteurs au cas de
semelles reposant sur une fondation formCe d'une couche de sable dense au dessus d'un dCpBt
d'argile molle. Les resultats de cette analyse sont prCsentCs sous forme d'abaques de dimen-
sionnement.
[Traduit par la revue]
Can. Geotech. J., 17, 30C-303 (1980)

Dense Sand Overlying Soft Clay be taken as the total passive earth pressure P,, in-
The assumption involved in predicting the theoreti- clined at an average angle 6,acting upwards (Fig. 1).
cal ultimate bearing capacity from the punching Thus, for a strip footing of width B and depth D in
theory is that, at the ultimate load, a soil mass in the the upper sand layer, the ultimate bearing capacity is
upper sand layer of roughly truncated pyramidal approximately given by
shape (friction angle is pushed into the lower
layer (cohesion C2) (Meyerhof 1974). The forces on
[I] q, = qb + (2/B)(Pp sin 6) - TIHI qt
the assumed vertical punching failure surfaces in the where q ~and
, qt are the ultimate bearing capacities of
upper layer (of thickness of H below the footing) can the strip footing on a very thick bed of the lower soft
0008-3674/80/020300-04!$01.00/0
$j! 1980 National Research Council of Canada/Conseil national de recherches du Canada

97 Dense
Soft Clay
(C.i
(a) STRESS DIAGRAM (b) FAILURE O F SOIL
FIG.1. Strip footing under vertical load on layered soil
clay layer and the upper sand layer, respectively
(Meyerhof 1955); and y l is the unit weight of the
upper sand layer.
The values of P , depend to a large extent on the
value of the average mobilized angle of shearing
resistance 6 on the assumed failure planes, and the
following arguments can be introduced in evaluating
its values:
(1) If the analysis is made on the real curved planes
of failure, the angle of friction 6 will be equal to cbl.
If, however, the analysis is made on the assumed
vertical planes, the angle of friction 6 mobilized must
be less than +l as failure has not taken place on the
assumed planes.
(2) Based on the fact that the failure strain of the
upper sand layer is less than that of the lower soft
clay layer, simultaneous occurrence of the shearing
failure in both layers could not take place and more
strain is required in the upper layer to reach the lower
layer failure strain value. Thus, the mobilized angle
of shearing resistance of the sand layer could be less
than the peak value and could approach the residual
value.
(3) The mobilized passive earth pressure on the
Assumed
Failure Planes -
FIG.2. Method of determining the passive earth pressure on the assumed planes of failures.
NOTES 301
assumed vertical failure planes in the sand layer
decreases with a decrease in the lower soft clay layer
strength. This can be explained by the fact that with
decreasing lower layer strength, the vertical displace-
ment of the sand punching column increases and the
lateral movements decrease, resulting in a decrease in
the passive pressure. This lateral movement may not
be sufficient for the maximum mobilization of the
Sand
passive pressure that would be generated by the full
value of the angle of shearing resistance
A mathematical verification for arguments (1) and
(2) is difficult at best, if not impossible. Also, it is
difficult to separate these effects in evaluating the
average mobilized angle of shearing resistance 6 and,
consequently, the mobilized passive pressure on the
assumed failure planes. However, these difficulties
may be overcome by expressing the angle 6 in the
dimensionless form (6/+1).
In order to study the reduction in the passive
pressure due to the existence of the weak lower layer,
a sliding surface was assumed, consisting of an arc of
a circle in the clay layer (bd) and a straight part in the
sand layer (de) behind a rough vertical wall (Fig. 2).
It was shown (Meyerhof 1974) that the passive
pressure
+
[2] P, = 0.5ylH2(1 2D/H)K,/cos 6
where K, = coefficient of passive earth pressure;
setting
[3] K, tan = K, tan 6
where K, = coefficient of punching shear, and sub-
stituting [2] and [3] into [I] gives
[41 q, = qb + Y Jf2(1 + 2D/H)K, tan +1/B
- y1H I qt
The theoretical study was conducted using the
same experimental data of (0, and Ce, where very
good agreement was achieved with the results of strip
footing tests on a dense sand layer overlying a soft
clay deposit (Table 1; Fig. 3). The theoretical study
- - X"'
Line Dense Sand
 CAN. GEOTECH. J. VOL. 17, 1980

BEARING CAPACITY RATIO

q,/s,
FIG.3. Punching shear parameter.

was then extended to cover wide ranges of the angle

of internal friction of the upper sand layer and the
cohesion C z of the lower clay layer. The results of
this analysis are presented in the form of two design
charts (Figs. 3, 4) for the case of a strip footing on a
sand layer averlying a soft clay deposit. From Fig. 3
the punching shear parameter 6/41 can be determined
knowing and the ratio of q2/qI where: ql =
0.5ylBN, (for homogeneous upper sand) and q2 =
CzNc for homo,geneous lower soft clay.
consequently,- the punching shear coefficient K,
can be found from Fig. 4 so that [4] can be used to
determine the ultimate bearing capacity of the strip
footing on the layered soil.
Equation [4] for strip footings can be extended to
circular footings as follows

where qt, and qt are the ultimate bearing capacities of

the circular footing on a very thick bed of the lower
soft clay layer and the upper sand layer, respectively;
and S, is a shape factor for punching shear resistance
on a cylindrical surface. The results of model tests of
circular footings on a dense sand layer overlying a
soft clay layer gave a shape factor S, of 1.1-1.27. For
a conservative design, S, may be taken as unity
(Meyerhof and Hanna 1978).
For footings under inclined loads on a dense sand UNDRAINED SHEAR STRENGTH OF CLAY (KPa)
layer overlying a clay deposit, an inclination factor i, 4. ~ ~ ~ f fof i punching
~ i ~ shear:
~ t ~ = 500;
can be introduced to [4] and [5] for strip and circular (b) cp, = 450;(c) cp, = 40".
 NOTES

TABLE1. Analysis of surface strip footing tests on dense sand overlying clay

Test 4 1 Cz Ratio of Observed q, Ratio of Calculated qb Deduced

No. (deg) (kPa) H/B qz/ql 6/$1

footings, respectively (Meyerhof and Hanna 1978),
as follows
[61 q,, = qb, + ylH2(1 + 2 0 cos a/H)Ksis
tan +l/B - ylH 5 qtv
and
where q,, is the vertical component of the ultimate
bearing capacity q, in the direction of the load; qb,,
and qt, are vertical components of the ultimate bear-
ing capacity under inclined loads qb and qt on thick
beds of the lower and the upper soil, respectively
(Meyerhof 1953); a is the load inclination with the
vertical; and is is the inclination factor given by
Meyerhof and Hanna (1978).
For footings on two sand layers overlying a clay
deposit, [4] and [5] can be written as follows (Hanna
and Meyerhof 1979)
and the ratio of qz/ql; Ks2 is the punching shear
coefficient for the middle laver and can be determined
from the charts provided kere (Figs. 3, 4) knowing
the values of $ Z and the ratio of q3/q2;and f is the
average of y l and y 2 of the upper and middle layers,
respectively.
Conclusions
The design charts presented in this paper, together
with the punching theory previously developed by the
authors, can be utilized to predict the ultimate bear-
ing capacity of footings on a dense sand layer over-
lying a soft clay deposit.
Acknowledgement
The financial support from the Committee in Aid
of Scholarly 'Activity of Concordia University and
the Natural Sciences and Engineering Research
Council Canada is acknowledged.

[8] qu = qb + Ks1- y'H12

B
tan +, + K.2 7
?Hz2 HANNA,A. M., and MEYERHOF, G. G. 1979. Ultimate bearing
capacity of foundations on a three-layer soil, with special
reference to layered sand. Canadian Geotechnical Journal,
16, pp. 412-414.
MEYERHOF, G. G. 1953. The bearing capacity of foundations
and under eccentric and inclined loads. Proceedings, 3rd Inter-
national Conference on Soil Mechanics and Foundation
Engineering, Zurich, Vol. 1, pp. 440-445.
1955. Influence of roughness of base and ground-water
conditions on the ultimate bearing capacity of foundations.
GCotechnique, 5, pp. 227-242.
1974. Ultimate bearing capacity of footings on sand
layer overlying clay. Canadian Geotechnical Journal, 11,
pp. 223-229.
where KS1is the punching shear coefficient for the MEYERHOF, G. G., and HANNA, A. M. 1978. Ultimate bearing
upper layer and can be determined from Fig. 2 of capacity of foundations on layered soils under inclined load.
Meyerhof and Hanna (1978) knowing the values of Canadian Geotechnical Journal, 15, pp. 565-572.