Download as pdf
Download as pdf
You are on page 1of 34
Load carrying capacity of foundations Prof. M. F. Randolph, University of Western Australia, Australia Prof. M. B. Jamiolkowski, Technical University of Turin, Italy Dr L. Zdravkovié, Imperial College London, UK Abstract This paper addresses developments over the last 50 years, since Skempton’s early contributions on the bearing capacity of shallow and embedded foundations, and focuses in particular on the gradual replacement of semi-empirical design approaches by more rigorous analytical or numerical solutions. Bearing capacity factors for shallow foundations on clay and sand are summarised from recent publications, taking account of foundation geometry and embedment, and variation of strength with depth. The effects on bearing capacity of other aspects of soil response are discussed, including strength anisotropy, consolidation under preload and compressibility. Traditional ways of evaluating foundation capacity under inclined or eccentric loading are being replaced by the use of interaction diagrams in vertical, moment and horizontal loading space; examples of these are presented. While the capacity of pile foundations is outside the scope of the paper, the trend towards the use of a small number of ‘settlement reducing’ piles beneath a raft foundation is discussed. Introduction One of the most widely quoted of Skempton’s papers is his classic “The Bearing Capacity of Clays” (Skempton, 1951), in which he documents the range of bearing capacity factors for strip and circular foundations at different embedded depths. The paper gathers together theoretical results based on plasticity solutions and cavity expansion analogies, compares these with results from field and model footing tests and arrives at proposed depth and shape factors that have withstood the test of time. Indeed, while more precise solutions have since been developed, Skempton’s original values were uncannily accurate and are of sufficient accuracy from a practical viewpoint. His paper also considers the load-displacement response of foundations, and provides insightful guidance on likely settlements at different factors of safety against failure. It therefore seems fitting at this conference to review theoretical solutions for bearing capacity of foundations in clay, taking account of geometry, embedment and strength profile, and also of the effects of strength anisotropy and partial consolidation. Complementary solutions for bearing capacity of foundations on sand are also summarised, with particular focus on solutions for Ny (a topic addressed by several papers over the last decade), the critical question regarding choice of friction angle, and also the effect of soil compressibility. The traditional bearing capacity approach incorporates a series of adjustment factors to cover factors such as those listed above, but also horizontal (or inclined) and moment (or eccentric) loading. This approach is rapidly being replaced by the use of interaction envelopes in vertical, moment, horizontal load space, and recent work in that area will be summarised. Detailed consideration of pile capacity is outside the scope of this paper, but some discussion is included of the bearing response of combined pile and raft foundations, and the suggested design approach of locating the pile support over the central part of the raft foundation. Advances in Geotechnical Engineering: The Skempton Conference, 2004, Thomas Telford, London, 208 ADVANCES IN GEOTECHNICAL ENGINEERING Shallow Foundations in Clay ‘The capacity of shallow foundations on clay is fundamental to much of civil engineering, and also provides one of the building blocks at undergraduate level for teaching about geotechnical failure. A review is presented here that summarises some of the key developments in recent years, where classical solutions for strip footings on homogeneous soil have been extended to cover practical aspects of soil behaviour and more general foundation geometries. Inevitably, design codes tend to lag behind such developments, but one of the aims here has been to document areas where semi-empirical approaches such as Skempton’s original correction factors for embedment, or Meyerhof’s (1953) approach for eccentric loading, may now be replaced by more rigorous analytical or numerical solutions, In addition, the effects of strength anisotropy, compressibility and consolidation under preload are also discussed. Vertical Bearing Capacity Calculation of the vertical bearing capacity of shallow foundations on clay builds on the classical solution of Prandt! (1921) for a strip load at the surface of a homogeneous rigid- plastic medium with shear strength, sy. The basic solution gave the average pressure, qu, at failure as (2 + x) times the shear strength of the medium. That solution was subsequently modified to account firstly for geometry (rectangular and circular instead of a strip) and secondly for the depth of embedment. These modifying factors led to a general form of bearing capacity adopted by Skempton (1951) in the form du = de8eNeoSu ) where s. and d. are respectively shape and depth correction factors to the basic bearing capacity factor of Ne = 5.14. Skempton proposed rather simple and conservative expressions for these factors, given by 14+0.2h/BS1.5 and s.=1+0.2B/L Q where h is the embedment of a footing of width, B, and length, L (& B). The depth factor, and suggested rounded values for N. of 5 and 6 (=1.2 x 5) for surface strip and circular foundations, are significantly lower than the curves of N. that he plotted for both foundation shapes, where the initial depth gradient is approximately double (0.4 rather than 0.2). Skempton did not attempt to address how to deal with soil profiles where the strength varies appreciably with depth, although he hypothesised that it was sufficient to adopt an average shear strength over the depth 2B/3 below the foundation, provided the variation within that zone was no more than about +50 %. More rigorous treatment of soil profiles where the strength increased linearly with depth was provided by Davis and Booker (1973) for surface strip foundations, Unfortunately a tendency in design practice has arisen whereby depth and shape factors derived originally for strip foundations on homogeneous soil are used inappropriately in situations where a significant strength gradient exists. For surface foundations, Houlsby and Wroth (1983) and Salengon and Matar (1982) pointed out how the shape factor decreases as the strength gradient increases. For a linear strength variation expressed as Sy =Syo +kz (3) where Sy is the strength at foundation level and k strength gradient with depth, z, the shape factor for rough-based foundations reduces below unity for KB/s. > 2, and in the extreme case Of So tending to zero approaches a value of 2/3. ‘A number of solutions have been published over the last 20 years addressing the bearing capacity of foundations in clay in a rigorous manner, and for a wide range of geometries and RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIC 209 strength profiles, and it is high time that the rather crude approach of depth and shape factors (and indeed other modifying factors) is replaced by more appropriate solutions for the problem at hand, These issues are addressed below, starting with plasticity solutions based on simple rigid-plastic soil response, but then exploring the effects of soil compressibility, anisotropy and consolidation. Plasticity Solutions Plasticity solutions, which are exact within the accuracy of numerical integration, are available for surface strip and circular foundations (Davis and Booker, 1973; Martin; 2001) for soil where the strength varies linearly with depth. In general, the strength intercept at the ground surface may be written as Sym, While the strength at foundation level is Sy (Figure 1), The bearing capacity factor, Nc, for a surface foundation (with h = 0, and therefore Sy = Sim) is then a function of the non-dimensional ratio x = KB/Sy oF kD/syo for strip and circular foundations respectively. These factors are summarised for a practical range of « in Table 1, together with the shape function giving the ratio between the bearing capacity of circular and strip foundations. As noted above, the shape factor reduces below unity for x > 2, reaching 0.9 for x= 10 Figure 1 Linear strength profile Table 1 — Bearing capacity factor, Ne for rough-based surface foundations = KB/s oF KD/Sva oi 2 6 10 Davis & Booker (1973) _| 5.14 | 6.61" | 7.60 10.42" | 12.66" Martin (2001) 6.05 | 6.95 | 7.63 [9.69 | 11.37 Shape factor, Scirte 1.18 | 1.05 [1.00 | 0. 0.93 | 0.90 "Exact values supplied by Martin (private communication) For square foundations, the bearing capacity should be slightly lower, by 2 to 3.% (Levin, 1955), than that for a circular foundation of equivalent area. Hence the customary shape factor of 1.2, for homogeneous soil, errs somewhat on the optimistic side. However, the error small, and to a first approximation it seems reasonable to adopt a shape factor for rectangular foundations that is an extension of that proposed by Skempton (1951), given by 8¢ ©1+(Sciscte —)B/L (4) where Scicte Corresponds to the values in Table 1. Some support for this approach is provided by solutions for the capacity of rectangular plate anchors (Merifield et al., 2003). They gave a lower bound bearing capacity of 11.9 for a deeply embedded (smooth) square plate, compared with the exact solution of 12.42 for a smooth circular plate (4% higher), and they also suggested bearing capacity factors for rectangular plate anchors that vary approximately linearly with the aspect ratio, B/L, of the plate. The effect of embedment has been explored by Martin (2001), who presented lower and upper bound solutions for embedded circular foundations in soil with linearly varying strength 210 ADVANCES IN GEOTECHNICAL ENGINEERING profiles, Attention here is restricted to foundations modelled as having a rough base (a likely condition in reality), but smooth sides (conservative). For embedded foundations, no exact plasticity solutions have been found, and the bracket between lower and upper bounds widens rapidly as the relative embedment increases. Lower bound solutions are presented in Figure 2, in terms of Ne = qu/Sy. As the embedment ratio, h/D, increases the bearing capacity factor tends towards ~9.3 for all values of «* = KD/Sum. KDisen = 10 0 0.5 1 15 2 Embedment, /D Figure 2 Lower bound bearing capacity for embedded circular foundations (Martin, 2001) x——" [Din La Depth ail Skeinpton (1951) 0 05 1 15 2 Embedment, t/D Figure 3 Equivalent depth factors for embedded circular foundations in non- homogeneous soil (after Martin, 2001) ‘An alternative way of presenting the results in Figure 2 is in terms of depth factors to be applied to the bearing capacity factor for a surface foundation with the same value of X= kD/Sw (noting the distinction between x and «*). These are shown in Figure 3, along with the original depth factors implied by Skempton’s (1951) Fig. 2 (noting that these are greater than his conservative design suggestion given in Equation 2). ‘Two points are clear: firstly the lower bound depth factors are all greater than Skempton’s original design curve, although not RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIG 211 by very much; and secondly the depth factors tend to decrease slightly as the degree of strength non-homogeneity (or x value) increases. This latter effect might be anticipated, since the contribution to the bearing capacity from the soils above the foundation level will decrease as. the strength gradient increases. Allowance for Finite Soil Stiffness The plasticity solutions described above are based on a rigid plastic idealisation of the soil, which is not an appropriate assumption for embedments greater than about twice the width of the foundation. For homogeneous soil, the mechanism of failure changes from a flow mechanism extending back to the free surface, to a confined mechanism analogous to cavity expansion, for an embedment ratio between I and 2 (Hu et al., 1999). Figure 4 compares plasticity solutions with finite element results for pre-embedded circular foundations at different depths in homogeneous soil. The finite element results for higher embedment ratios are for foundation displacements of 0.3D, since the load displacement curves for h/D greater than unity did not reach a plateau. A large deformation analysis showed that the steady state bearing capacity for a foundation pre-embedded at h/D = 4 was not reached until a displacement of 4 diameters. The final bearing capacity factor was 12.7, which is similar to theoretical cone factors for the given soil conditions. 147 Upper bound Bearing|2 Martin (2001) capacity factor 10 =f ‘N N 8 in Siowersans fit 6 Martin (2001) a 4 2 0 0 1 2 3 4 7 Embedment depth, h/D Figure 4 Comparison of finite element and plasticity solutions Similar analysis in soils with a high strength gradient showed that the transition depth from surface failure mechanism to confined failure occurred at an embedment ratio between 2 and 3, and that the bearing capacity factor mobilised at a displacement of 0.1D ranged between 9.3 and 9.9 at high embedment ratios (Hu and Randolph, 2002). For both homogeneous soil and soil with a high strength gradient, it appears that results from finite element analysis show good agreement with lower bound plasticity solutions, with bearing capacity factors of around 9 at deep embedment, for practical displacement limits. Overall, these results demonstrate the remarkable accuracy of Skempton’s original hypothesis, based on quite limited test data, of the bearing capacity limit of 9s, for deeply embedded circular foundations, and the comparable value of 7.5s, for strip foundations (see Wang and Carter, 2002). ‘The bearing capacity of foundations on a two-layer medium has been explored using numerical plasticity solutions by Merifield et al. (1999) and through large deformation finite element analysis by Wang and Carter (2002). ‘These papers have shown that, for the common situation where a stronger layer of clay overlies weaker material, the simple approach of Meyerhof and 212 ADVANCES IN GEOTECHNICAL ENGINEERING Hanna (1978), based on punching shear through the upper layer and a conventional bearing capacity calculation in the lower layer, tends to be conservative for high strength ratios. The degree of conservatism can reach around 20 % for strength ratios in excess of 5. Pre-failure displacement of the foundation plays an important role in the failure mechanism for layered profiles, since there is a geometric non-linearity due to gradual thinning of the upper layer (Wang and Carter, 2002). Load-displacement responses from large deformation analyses can show a peak followed by softening (due to punching through the upper layer) if the soil weight is ignored (See Figure 5). However, if the soil weight is included then the increase in capacity due to penetration and the resulting additional surcharge more than offsets any decrease due to punch-through. Wang and Carter’s results were for caisson-type foundations, with smooth vertical sides, and the same conclusion would not necessarily be true for a plate like foundation, such as the spudcans used offshore for mobile drilling units, where soil is free to flow back on top of the plate. Large deformation eal 22m | 8) 7 0.6(2+2)) 6 ‘Small deformation cxej=0.8 or 1 a 219(2+2%) | eee ar gy ae stat ; eyo 850 2 0.212 82)] 0.1(2+2%)/ 0 50 100 150 200 250 300 350 Gys/Be,=6,s/Bc, Figure 5 Load displacement responses for strip foundations on weightless two-layer soil with strengths c; and cz (Wang and Carter, 2002) Effects of Preloading Preloading of a foundation at some proportion of its ultimate capacity will lead to partial consolidation of the underlying soil, and thus an increase in bearing capacity. In soil which is originally homogeneous or with strength increasing with depth, partial consolidation will give rise to a crust of stronger soil immediately below the footing. The time-scale for consolidation can be estimated, from published solutions (Booker and Small, 1986), with Ts. and Tso values (where T= c,t/D*, with c, the consolidation coefficient) of about 0.1 and 5 (see Figure 6). The effect on bearing capacity has been investigated recently by Zdravkovie et al, (2003) for strip and circular foundations. They considered different strength profiles ranging from a typical normally consolidated profile with a 2 m crust (Figure 7a) to linear profiles with overconsolidation ratios from I to 25 (Figure 7b). ‘The magnitude of bearing capacity following preloading by different proportions of the initial bearing capacity (both expressed as percentage) is shown in Figure 8 for strip and circular foundations with B = D = 2m. It may be seen that for a typical preload level of 30 to 50 % of the initial bearing capacity, ignificant increase due to preloading occurs only for the normally consolidated strength profiles (with or without the crust). For OCR of 2 or more, the increase is below 10 % unless the preload level is increased. RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIC 213 ‘Surface 2=0 impermeable era oa permeable rea =-— Surface 2=0 completely permeable (Chiarella and Booker 1975) 1:0, woul L ones sul au. 0.007 0.07 Oy i] 0 100 sect Figure 6 Consolidation curve for rigid circular foundations (Booker and Small, 1986) ock ce Sena ek: ck Ocr=2s Fitted profile «Data fom triaxial ‘compression tests ee i wo 2 % 40 ° too 200~—«300~—«400 Undrained strength, S, (kPa) Undrained strength, S, (kPa) (a) Typical soft clay profile with 2m crust __(b) Linear strength profile with varying OCR Figure 7 Strength profiles considered in preloading study (Zdravkovic et al., 2003) Gain in pesng capacty (sone! L Ginn pec apacty (asa ita coming cacy) ° i er er) ° m4 88D Lave preioae Love of predoad (as a6 of rt Soaring copecy) (asa fina own capacty) Figure 8 Increase in bearing capacity due to preloading 214 ADVANCES IN GEOTECHNICAL ENGINEERING Results from a field preloading test at Bothkennar have been presented by Lehane and Jardine (2003) for a 2.4m square foundation that was left for 11 years under a preload of 65 % (marked as ‘Test B’ in Figure 9) of the initial bearing capacity (marked as ‘Test A’ in Figure 9). The average vane shear strength in the upper 6 m was 25 kPa (Jardine et al., 1995) with an estimated overe: lation (or yield stress) ratio of about 2. Taking a field cy of 30 m’/year, the time-scale for primary consolidation may be estimated as too ~ 5*2.4/30 ~ I year, so the I1 year period is well in excess of that needed for primary consolidation. During the preload period, the mean settlement increased from around 50 min to 230 mm. When the footing was then loaded (“Test C’ in Figure 9), the response was very stiff up to an average pressure of 160 kPa (115 % of the original bearing capacity) and failure was ~205 kPa (see Figure 9). The observed 48 % increase in bearing capacity compares with a predicted increase of around 25 % (Figure 8a). It seems likely that secondary consolidation effects play a significant role in improving the bearing response, which would explain the conservative prediction of the increase in capacity. Settlement in Test, s(mm) 220 y 190 Sut of2001 Tew | a= Applied Stess, 4» 5 - ° 100 200 300 400 00 Mean footing settlement, § (mm) Figure 9 Loading response of footing before, during and after preload (Lehane and Jardine, 2003) A further example of the effects of preloading is shown in Figure 10, from Randolph et al (1999), for a model offshore foundation with skirts of 0.4D in normally consolidated kaolin. In ‘two separate tests the foundation was preloaded to 40 % of the initial bearing capacity, left for periods estimated to represent 50 % and 100% primary consolidation and then loaded to failure. In both cases a very high (but brittle) bearing capacity increase by over 50% was observed, probably associated with increased side friction, but the underlying increase in capacity was around 25% as seen in the right-had diagram (ratio of bearing resistance at a given depth to the original ‘undisturbed’ bearing resistance). This is in excellent agreement with the increase predicted in Figure &b. RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIC 215 Ratio of bearing resistances 0 10 20 113) 16 19 — 100% consolidation ~ 50% consolidation Figure 10 Increase in bearing capacity due to consolidation Interaction Diagrams for Horizontal and Moment Loading The effect of horizontal (or inclined) and moment (or eccentric) loading has traditionally been handled by additional multiplicative factors to the bearing capacity equation (Equation 1). These factors are based on theoretical solutions, such as Green (1954) for inclined loading, or the concept of effective footing width (Meyethof, 1953). In recent years, driven by the more onerous lateral loading environment in offshore applications, this approach is being replaced gradually through the use of interaction diagrams or failure envelopes in V, M and H loading space. The analytical basis for the bearing capacity of strip foundations under combined loading was addressed by Houlsby and Puzrin (1999) for homogeneous conditions, with particular attention to conditions where the foundation starts to lift away from the soil (assuming no tensile contact stresses between foundation and soil), and the ‘scaling’ of foundations where lift-off occurs to an equivalent foundation in full contact with the soil. A more extensive study for soils with strength increasing with depth, with and without lift-off has been published by Ukritchon et al. (1998). In the interaction envelopes discussed below, the reference point for moment loading is the centre of the interface between foundation and soil. For circular foundations in full (tensile) contact with the soil, Taiebat and Carter (2000) proposed an equation of the form (*) [eine (2) Vo) “LMC "Hy “hi where the subscript u denotes the capacity under the corresponding uniaxial loading. The parameter 0 reflects the skewed nature of the envelope in the M-H plane, whereby the maximum moment capacity is obtained for a positive (following the sign convention of Butterfield et al., 1997) horizontal load, typically 80 to 90 % of the ultimate value. A value for 1 of 0.3 was suggested for homogeneous soil conditions, although the gradient of the failure envelope in the M-H plane will be affected by the parameter x = kD/si (Bransby and Randolph, 1998; Gourvenec and Randolph, 2003). (5) ‘A. common assumption in design is that, while the value of uniaxial load capacities, Vy, My and H, will be affected by aspects such as geometry (strip, circle, rectangle), embedment and strength gradient factor, x, the normalised shape of the failure envelope will be affected to a much smaller degree. ‘This is illustrated in Figure 11 in the three planes M=0, H=0 and V=0, where the barred parameters v, m and h represent the loads normalised by their 216 ADVANCES IN GEOTECHNICAL ENGINEERING corresponding uniaxial capacities, as tabulated in Table 1 (for V) and Table 2 (for M), with the assumption that Fl, = Bsy) Or Asyo for strip and circular foundations respectively (with A being the area of the circular foundation). It may be seen that the shapes of the failure envelopes are relatively similar for strip and circular foundations, although there is a more si of the strength parameter, x ificant effect Table 2— Moment capacity factor, Mu/B*Su. of Mu/ADSy for surface foundations (after Gourvenec and Randolph, 2003) { | k= KDisye 0 1] 2 3 10 Strip _| Finite element 0.72 | 087 ) 1.01 | 1.15 Upper bound ~ [0.69 | 0.83 [0.96 | 1.08 1 Cirele | Finite element 0.69 | 0.83 1.04 Upper bound 0.67 | 0.78 | 0.88 | 0.97 | | FT v “4 10 0.9 | snoresing os o7 os 10 08 08 04 -02 n& Figure 11 Failure envelopes with loads normalised by ultimate values (Gourvenec and Randolph, 2003) For surface foundations where lift-off can occur under the trailing edge of the foundation, the shape of the failure envelope changes so that it passes through the V = M = H = 0 point, with maximum moment or horizontal capacity developed at a vertical load of half the ultimate value. The general shape may be expressed in a simplified form as |e sens an SRE RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIC ce bum? un? ) num Gl) Ga) Gee) 6 where m, ~ 0.1, hy ~ 0.125 and the skew parameter, a~ 0.3, although more complete treatment is given by Martin and Houlsby (2001), with the skew parameter varying as the vertical load increases. The results of Gourvenec and Randolph (2003) also suggest that the value of a will reduce as the homogeneity parameter, x, increases. The sliding capacity of shallow foundations on clay is influenced to a much greater extent by Partial consolidation than are the vertical or moment capacities. Bransby (2002) has presented results of preloading on the inclined capacity of a strip foundation resting on normally consolidated clay. His results for vertical capacity are broadly consistent with those shown in Figure 8b, although giving slightly greater increase in capacity for a given level of preload. However, the horizontal capacity was found to increase by as much as factor of 3, for a preload level of 60 % of the original ultimate bearing capacity, with an approximately. linear relationship with the level of preloading, Rotational Stability The rotational stability of strip and circular foundations has been investigated by Taiebat and Carter (2002), who have confirmed the general form of Equation 6, and also demonstrated reasonable agreement with the effective width principle of Meyerhof (1953). While this approach is appropriate for situations where the moment loading arises from horizontal forces such as wind or wave loading, a different form of instability can arise in the case of eccentric self-weight loading. This has been investigated in detail with respect to the Leaning Tower of Pisa (Potts and Burland, 2000; Potts, 2003). A rotational buckling instability can occur in situations where the rotational stiffness is insufficient to compensate for any increase in applied moment due to rotation of the foundation and the supported structure. Figure 12 shows vectors of incremental displacements and the plastic zone at failure for a plane strain simulation of the Tower of Pisa and foundations. Potts and Burland (2000) show that the tower weight at which instability occurred reduced by a factor of 2 as the rigidity index, G/s, for the soil was reduced from 1000 (where failure occurred by conventional vertical bearing capacity) to 10 (leaning instability). Figure 12 Vectors of incremental displacements and plastic zone at failure for low soil stiffness (Potts and Burland, 2000) 218 ADVANCES IN GEOTECHNICAL ENGINEERING : W tt ERO a Figure 13 Simplified model of leaning instability Feaning instability of the Tower of Pisa was also suggested by Lancelotta (1993), who developed a design chart using a simple Winkler model for the soil support. ‘The form of the plastic zone in Figure 12 suggests a simplified (and conservative) model, as illustrated in Figure 13, where the vertical load leads to localised failure over a width, s, at the leading edge of the foundation, while the remaining (elastic) zone of width, B~ s, provides the rotational stiffness. This model leads to a failure load that, for a strip foundation, may be expressed as +f- fA) ” Where f is related to the rotational stiffness, Ko, of the full width foundation, the relative height of the centroid, b/B, and the ultimate bearing pressure, qu, by Mien (8) For the Pisa example, where h/B ~ 1.5 and v = 0.49, the reductions in bearing capacity for G/sy values of 1000, 100 and 10 are respectively 0.095, 0.27 and 0.62. ‘These are broadly consistent With the finite element results presented by Potts and Burland (2000), although slightly Sgraaivative. A’ similar calculation can be made for circular or rectangular foundations adjusting the rotational stiffness appropriately. Strength Anisotropy Seely the solutions for bearing capacity have been expressed in terms of a simple isotropic Skat strength, si, which has tended historically to be evaluated through triaxial compression tests, or field vane tests, However, foundation problems impose a range of different strece pats on the soil, and these lead 10 different shear strengths, In offshore engineering, it is (Cmmon practice to take account of strength anisotropy when calculating ultimate capacities {{auritzsen and Schjetne, 1976), or at least to base the design shear strength on the sverage {rom triaxial compression, simple shear and triaxial extension strengths. For most soils. the cegree of strength anisotropy increases as the plasticity index reduces and the average strength is typically close to that measured in simple shear (Ladd, 1991; Jardine and Menkil-1999) Zdravkovie et al. (2001) presented results from finite element analyses of the pull-out capacity of shallow skirted foundations, using an anisotropic soil model based on the MIT-E3 saci Figure 14a presents stress paths from the hollow cylinder experiments on silt (Zdravkovie & Jardine, 2000), in which identical K. consolidated samples were sheared with inclinations of the major principal stress varying from 0° to 90° to the vertical direction. The undremed strength anisotropy observed ini these experiments was simulated using the MIT-E3 model and RANDOLPH, JAMIOLKOWSK! AND ZDRAVKOVIG 279 the predicted available undrained strengths, with depth, in triaxial compression, extension and simple shear are presented in Figure 14b. =“) Triaxial compression + Phase Transformation = 15) = as em € z 3 & o 20 = 75 a so a 25, al ol. Pe leet intemal ete 0% 40 6) 8) 400 ¥20 ¥40 ¥60 180 200 220 Inne oolneran ee en es co voc an ’ (KPa) Undrained strength, S, (kPa) (a) Stress paths from hollow cylinder tests _(b) Strength profiles from MIT-E3 model Figure 14 Anisotropic strengths measured and modelled by Zdravkovic et al. (2001) 0790° «Isotropic MIT-E3 model To —Fitedelipses 4 Anisotropic MIT-ES model Rgaeg 88sbs Ultimate horizontal force, Hy, (MPa) 2 od ° 50 100 150 200 Ultimate vertical force, Vy (MPa) Figure 15 Failure envelopes for skirted foundation (after Zdravkovic et al., 2001) ‘The degree of anisotropy may be expressed in terms of the ratios of strengths in triaxial extension and compression, and in simple shear and triaxial compression, which in this example were 0.52 and 0.66 respectively. The average strength is 0.73 times the triaxial compression strength. Figure 15 shows the failure envelope in V-H space for a skirted foundation of diameter 17 m, s of 12 m, and loaded at different angles, 0, through the centre of the foundation top. computed using the anisotropic strength model are some 22 % lower than those ing an isotropic model with the same triaxial compression strength. Hence if design calculations were based on the average shear strength in the three modes plotted in Figure 14, the resulting design would be conservative by 0.73/0.78, or about 6 %. 220 ADVANCES IN GEOTECHNICAL ENGINEERING Potts and Zdravkovic (2001) have described application of the same anisotropic strength model for the analysis of surface strip and circular foundations under combined V and H loading, and shown that the capacities were reduced, on average, by 20 % and 30 % respectively, relative to capacities using an isotropic triaxial compression strength. Again, from a design perspective, calculations using isotropic strength based on the average of triaxial compression, simple shear and triaxial extension would yield reasonable estimates of capacity, even if up to 9% conservative for the plane strain’case. The different effects of anisotropy for plane strain and axisymmetric problems have an important consequence for the shape factors adopted in bearing capacity calculations, For example, for homogeneous soil conditions where a shape factor of 1.2 would normally be used, the above calculations would suggest that anisotropic effects would reduce the capacity of a circular foundation by 10 % more than that for a strip foundation, implying a shape factor close of 1.03 (comparing 0.7*6.05 with 0.8*5.14). A similar, but more pronounced, effect has been noted in comparing the penetration resistance of plane strain cylindrical (T-bar), and axisymmetric spherical (ball) penetrometers. While conventional plasticity solutions show that the resistance of a ball penetrometer should be some 22% higher than that of a T-bar, field results indicate virtually identical penetration resistance in soft clays (see Figure 16), and resistances for ball and plate penetrometers that are even lower than the T-bar in a varved silty clay (DeJong et al., 2004). The implication is that shape factors for bearing capacity problems may be 20 % lower than implied by standard plasticity solutions. Undrained shear strength, sy (kPa) 0 10 20 30 40 50 Vane: peak © Vane: remoukied Plate Depth (m) 6-1 Note: Bearing factor 18 {o/s = 10.5 used for all penetrometers 2 Figure 16 Shear strengths from different shaped penetrometers (Chung and Randolph, 2004) Shallow Foundations on Sand The general form of bearing capacity calculation for shallow foundations on sand involves the superposition of the components due to surcharge and due to increase in effective stresses below foundation level due to the self-weight of the soil. Traditionally both of these components are then modified to allow for the geometry of the foundation, or inclined or eccentric loading. Bearing capacity factors for strip and circular foundations may not be obtained to very high accuracy, for a given friction angle for the soil, using numerical implementations of the method of characteristics, However, the greatest source of uncertainty lies in an appropriate choice of friction angle, allowing for the stress-dependent nature of this, quantity. This aspect of foundation design is discussed in detail later. As for foundations on clay, failure of foundations under combined vertical, moment and horizontal loading may best be treated using interaction diagrams, rather than correction factors to the standard bearing capacity equation. Examples of these will be presented. nS RANDOLPH, JAMIOLKOWSKI AND. ZDRAVKOVIC 227 Vertical Bearing Capacity ‘The vertical bearing capacity of shallow foundations in sand may be expressed as fi strip YB Gu = 5qNG™ py +5,NH%? a 2% ects Peis the effective surcharge at foundation level and shape factors s, and S; are applied to pee ,surcharge ‘bearing capacity factor, Ne and the self-welght factor, No. For a circular foundation, the footing width, B, is replaced by the foundation diameter, B, Voluts the Provo) ATeh a eguation 9 are generally based on publications in the 1970s such ae Lenses (1970), Meherhot (1973) or Vesic (1973), although analytical solutions are availabe only for 2 yor a strip foundation. Other solutions derived using the method of characrericeee are Founclatised below, giving Nq and Ny as functions of the friction angle, 6, for strip and eneulae foundations. A further imporiant consideration is the choice of friction angle, which enue take into account the curvature of the strength envelope exhibited by sands, Plasticity Solutions Classical solutions for bearing capacity of foundations on sand have used the (Sokolvskii, Hoon David and Booker, 1971; Salengon and Matar, 1982). There have also toon cove} Micha ans addressing the variation of N, with friction angle ¢* (Bolton and Lau, 1993; Pichalowski, 1997; Ukritchon et al, 2003; Cassidy and Houlsby, 2002). With the exception of Uiaichon et al, who used finite element formulations of the lower and upper Hous ap ProarTss.t0 bracket Nj, and Michalowski’s upper bound solutions, the remaining approceies are still based on the method of characteristics, although in most cases incorporating simplifying assumptions. With modem computing power, itis possible to obtain high accuracy using the method of characteristics, for example using freely available softwar such as the sion, Martin (2004) has compared calculations of Ny, for a strip footing And Showed virtually identical results with those of Sokolovskii (1965) ‘and Salengon and Matar (1982), His values are given in Table 3, together with corresponding values of Nq and Ny for a circular foundation (assuming a fully rough foundation base) iting values of shape factors sq and s, are also tabulated, to illustrate that no simple relationship exiees for rece, Table 3 — Bearing capacity factors, Ny and N, for rough-based surface foundations (using ABC software of Martin, 2003) Friction ‘Strip foundation | Circular foundation | Derived shape factors Langle, ¢' ee - —— ~ [degrees [Ni [enema Nz _[W, % 5 137 0.113 W7i 0.081 1.09 071 10 2.47 0.433 2.96 0.32 1.20 0.74 15 3.94 Lis [5.25 0.93 1.33 0.79 20 6.40 [2.84 [9.61 2.42 150 | _0.85 Ear 10.7 [6.49 | 18.4 6.07 _173 0.94 | 30 18.4 14:6 ee 3 7.28 | 515 se D0 1.05 S5mbee | eeees253 34.5 60:8 2eea a 119 | 2/49 7 [40 64.2 | 85.6 193* [124 3.01 1.45 45 135 [234 321% 418 3.86 | 1.78 “* Note, these results contain crossing characteristics, which may affect the accuracy of the solutions 222 ADVANCES IN GEOTECHNICAL ENGINEERING It may be ascertained that the relationship between N, and Ng for the two types of foundation approximates to Ny, =LS5(Nq -1)tan@’ (strip) and Ny #0.75(Nq ~1)tang' (circle) (10) for friction angles in the range 15 to 40°, For rectangular foundations, it is necessary to interpolate between values of bearing capacity factor for a strip (as B/L tends to zero) and a circle (for a square foundation with B/L = 1), The bearing capacity factors in Table 3 are for soil obeying an associated flow rule, with dilation angle equal to the friction angle. In practice, the angle of dilation for soil is significantly lower than the friction angle, but it has been found that the bearing capacity is relatively insensitive to the angle of dilation (Potts and Zdravkovie, 2001). The roughness of the foundation has a much greater effect, particularly in respect of Ny, where corresponding values for smooth-based foundations would be approximately 50 % of those in Table 3. Choice of Friction Angle It is well recognized that, with the exception of loose materials, the shear strength envelope for coarse-grained soils is curvilinear, owing to the stress dependent nature of dilatancy. As illustrated in Figure 17, if the normal effective stress, o'g, at failure on the failure plane is smaller than some critical value, (o'r)eri then the peak angle of shearing resistance, ¢', decreases as o's increases. However, as the normal stress increases, dilatancy is progressively inhibited so that beyond the critical value even dense soils exhibit a contractive behaviour. The shear strength envelope then becomes linear, characterized by a constant angle of shearing resistance at critical state, dev. This linear strength envelope also holds for loose contractive soils at all values of o's (Since essentially (G'n)eit Feduces to zero). The critical state friction angle, cv, provides a lower bound for 6' for a given soil. Peak strength envelope ‘Strength envelope at eritical state A 9 oe we Ina given material ey represents the lower bound of possible values of g! Shear stress, T (S'w erie Dilative response € | -> Contractive response Nonmal effective stress, 6° Figure 17 Shear strength envelope for coarse-grained soils The experimental evidence shows that in the range of o's (o'er the value of 6' varies uniquely, in accordance with Rowe's (1962) stress-dilatancy theory, in relation to the dilatancy rate at failure (dey?/de)")max Where ¢, and €; are the vohimetric and major principal strains respectively and the superscript p denotes the plastic component. The dilatancy rate depends on the combined influence of the initial void ratio, ¢, and mean effective stress, p'., Moreover, in a given soil at a given state (de,"/de.")nax depends on the inclination (8), of the major principal effective stress at failure, o's, with respect to the bedding plane, reflecting the anisotropic nature of @. RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIG 223 The pressure level dependency of $' is well reflected in the empirical strength-dilatancy relationship proposed by Bolton (1986): $'~ Gey =m, where I, = Dp (Q~én(pF)) (10) where Dx is the actual relative density, 1, is a dilation indicator with m taken as either 3 for triaxial (or general) stress conditions and 5 for plane strain. The quantity Q may be taken as the natural logarithm of the grain crushing strength (expressed in kPa) and p'ris the mean effective stress at failure (also in kPa), Limits of 0 and 4 were suggested for I. ‘The m values suggested by Bolton were obtained analysing the results of triaxial compression tests (8 = 0) performed on 17 siliceous and quartz sands. However, due to the dependency of the maximum dilation rate on 8, values of m ate also susceptible to varying with the principal stresses orientation, as shown by Pradham et al. (1988), Sasitharan (1989) and Tatsuoka et al. (1997), For given values of dey and Dg for the in situ soil, an appropriate value for the stress dependent friction angle, 4° can be estimated using the following empirical formula to assess the average mean effective stress at failure: (a) Shallow Foundations (De Beer, 1967): : 1 I qu +3040 Pe =——— ay 1+ tan? g' 1 +sin¢! 4 where o',, is the effective overburden stress at the foundation depth. (b) Deep Foundations (Fleming et al, 1992) Pr = Vaus\0 = Naoto (2) In both of these formulae, an iterative approach is required to evaluate an appropriate value of ¢' and hence of the ultimate bearing pressure. Values of Q and d2y are provided in Table 4 for a number of geotechnically well-characterized sands, Table 4 — Values of Q and ¢c, derived from triaxial compression tests Sand | Mineralo; o da) Reference Ticino ___|Siliceous 10.8 33.5 Jamiolkowski et all Toyoura ~ [Quartz [98 32_ | (2003) Hokksund Ticeous 92 Samia _| Mol Quartz, 10 31.6 | Yoon (1991) Ottawa Fines 0% 9.8 30 (Quartz sand with Fi 10.9 32.3 varying fines Fines 10 % 10.8 32.9 Salgado et al. (2000) | content) Fin 10. 33.1 Fin 99 33.5 ‘Antwerpian Quartz & Glauconite [7.810 8.5 | 31.5 Yoon (1591) “Kenya Calcareous | B.S 40.2 | Jamiolkowski etal. Quiou Caleareous 75 EW (2003) J Containing comparable amounis of quartz and feldspar grains 224 ADVANCES IN GEOTECHNICAL ENGINEERING ‘A number of formulations for the convex failure envelope observed with sands have been presented, (Yaroshenko, 1964; Berezantzev and Kovalec, 1968; Baligh, 1975; Maksimovic, 1989). Of these, the formulation of Baligh (1975), is most consistent with the current practice in presentation of strength test data, with the shear stress at failure, tr expressed as tr a o'q| tangs + tno to 2 HH (13) Pa where @', and a are angles defining the strength envelope and p, is atmospheric pressure (~100 kPa). This expression represents a curvilinear strength envelope that holds in the range G'r< (G'n)ert beyond which the linear Mohr-Coulomb relationship applies, with 4's = dey and 0 ‘The failure criterion in Eq. determined as tang, = = tangy, + tna te 2) 4) o’ 13 allows the secant peak angle of shearing resistance to be and the tangent peak angle of shearing resistance as tan oi a Se = tan - tana: toe 2) (15) ofr Pa In the above formulae, 6', corresponds respectively to ', at o'7= 2.72p, and to $ at o'r = Pa Figure 18 shows the values of angle o: for different sands as obtained from triaxial compression tests. This angle is close to 0° for low relative densities corresponding to very loose soils and increases with increasing relative density, approaching values close to 10° or even higher in very dense siliceous materials. In a given soil and for a given density, ot is mostly controlled by the'crushability of the grains, as illustrated in Table 5. ay T Tus pres een loan na 29°. 2 Siliceous sand a @ Caleareous s eG ¢ 1g0{. 8 Glauconitic 2 5 8 a 28 S10 Range for ap 2 siliceous sand. 7 < sf | ah 4 ooLi& 1 o 20 40 60 80 100 Relative density, Dr % Figure 18 Shear strength envelope for coarse-grained soils As a consequence of the dependency of the dilatancy rate on the inclination of o'r at failure to the bedding planes, characterised by the angle 8, the value of a may be different for compression and extension loading, For example, Sasitharan (1989) quotes corresponding c RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIC 225 values from triaxial extension tests for Erksak sand of 1.5, 2.1 and 3.1 for the three relative densities in Table 5. In order to arrive at the design values of secant or tangent friction angles, it is necessary to estimate, at least approximately, the magnitude of o'r. This can be attempted with the help of Eqs. (11) and (12), taking into account the Mohr’s circle relationship between pieand o'e. Table 5 — Values of angle, o,, for four sands (from triaxial compression tests) Sand_ ‘References Sasitharan (1989) | Erksak (Quartz) Ticino (Siliceous) Jamiolkowski (1990) jolkowski Quiou GBioclastie + (unpublished) calcareous) es | Kenya (Oolithic Jamiolkowski calcareous) (unpublished) Effect of Soil Compressibility Unlike for clays, where volume change is restricted by lack of drainage, the bulk compressibility of sands can have a significant effect on practical limits for bearing capacity, Depending on the size of the foundation, design settlement limits of 5 or 10 % of the foundation diameter or width are commonly adopted, and such limits may lead to reduced foundation ‘capacity’ compared with conventional bearing capacity formulae derived on the basis of rigid-plastic response. Two aspects are discussed below. The first is the influence of dilation angle on the bearing capacity and load-settlement response, and the second is the influence of pre-yield elastic compression on the design capacity. Interestingly, this is a topic where Skempton made an early and significant contribution, using cavity expansion theory to establish a theoretical basis for design bearing capacities in sand (Skempton et al., 1953). Recent numerical studies of the effect of the dilation angle on the ultimate bearing capacity have been presented by Frydman and Burd (1997) for strip foundations, and Erickson and Drescher (2002) for circular foundations. Broadly, these studies indicated that the dilation angle has little effect for low friction angles, but starts to become significant for friction angles of 35° or more. However, the values of N, for strip foundations recommended by Frydman and Burd, based on dilation angles consistent with Bolton (1986), all lie above the more rigorous plasticity values quoted in Table 3 for fully associated flow. For the circular geometry, the values of N, in Table 3 correspond to Erickson and Drescher’s values for dilation angles marginally ‘greater than $'/2 (Which exceeds the expected dilation angle). Similarly, values of Nq derived from Erickson and Drescher’s N'_ values also fall below those quoted in Table 3 for friction angles of 35° and above. It is clear that accurate numerical values for bearing capacity factors for rough circular foundations are extremely difficult to obtain, as commented by the various authors who have attempted this, and are also increasingly affected by dilation angle at high friction angles. As such, the values of Ny and Ny in Table 3 for circular foundations should be used with caution for high friction angles. Indeed, as discussed below, the effect of soil compressibility starts to dominate for high friction angles, and the values of bearing capacity factors tabulated for friction angles in excess of 35° are unlikely to be achievable in practice A further effect of dilation angle is its influence on the settlement required to mobilise the ultimate capacity. Potts and Zdravkovie (2001) have given examples of this for 6! = 25 °, for strip and ciroular foundations, varying the dilation angle between zero and 25°, At this moderate level of friction angle the computed ultimate capacities are little affected by the 226 ADVANCES IN GEOTECHNICAL ENGINEERING dilation angle. However, for the particular parameters chosen, the settlement to mobilise that Fuhacity is more than doubled for a dilation angle of zero, compared with the case of associnneg flow. This may therefore influence the design value of capacity adopted ina given application, The influence of compressibility on the foundation capacity that may be mobilised at limited euagrments has been studied in an extensive series of calibration chamber experimen inverving cone penetration and plate loading tests. These tests have led (a) to conclations cf ‘lative density and cone resistance with in situ stress levels, and (b) to ranges ioe the Proportion of the cone resistance that can be mobilised for seitlements of 3 ot 10 Meat the plate diameter. The correlation of relative density, Dg, (expressed as a fraction) with cone resistance, qe, and in situ mean effective stress, p', may be expressed as (Jamiolkowski et al., 2003) 1 fae) : Cee “ Where q. and p'y are both in kPa, and the various parameters ai Ci = 0.46; C2= 2.96. This relationship is shown in Figure 19. Da Co = 300 (dimensional); Waa se Sure, ranging from full scale tests to model tests on embedded circular plates Within a calibration chamber, and also numerical analysis. There isa strong level oF agreement among the various recommendations, lending confidence ia their application, Cone resistance (MPa) 0 10 20 3040 50 60 Dr= 100% 100 200 300 Mean effective stress (kPa) Figure 19 Relationship between cone resistance, relative density and mean effective stress for coarse-grained soils (Jamiolkowski et al 2003) Adopting an upper limit for qu/q. of 0.2 from Table 6 and combining this with the cone resistance values shown in Figure 19 (or given by Eq. 16), upper limits on Ng for embedded foundations may be obtained. Even for very dense sand (Dx = 0.9 to 0.95) tears Ny values for foundation design would be around 150 for a mean effective stress of so kPa, and RANDOLPH, JAMIOLKOWSK!I AND ZDRAVKOVIC 227 would reduce at higher in situ stress levels, Comparing this with the values in Table 3 suggest that N, values for friction angles of greater than 35 ° are unlikely to be achievable in practice. Table 6 - Bearing capacity of embedded foundations as proportion of cone resistance {Source ‘Soil type [ Ratio": q/qe at References __ _ [wi 25% [w= 10% Full scale tests | Sands and Not 0.10 to 0.16] Bustamente & Gianeseili (1982), gravels available _| Franke (1991), Alsamman (1995) jeep plate tests in Ticino sand 0.10 to 0.16] 0.13 to0.21| Ghionna et al, (1994) calib.chamber cous) | Deep plate tests i) Quiou sand 0.13 O18 jolkowskT and Sarrt (2000) ib. chamber __| (calcareous) E Pile Toad test i T sand 0,10 to 0.14] 0.14 to 0.17| Fioravante (1994), Fioravante calib. chamber__| (siliceous) _| al. (1995) ‘Numerical Ticino sand 0.09 to 0.10] 0.13 to 0.21) Salgado and Lee (1999) _ analysis _ (siliceous) [0.10 to 0.14[ 0.13 to 0.21] Sami (2001) "Lower limit corresponds to medium density (Dx ~ 35 %), upper limit to Dy ~ 90% For foundations on the soil surface, or of limited embedment, account must also be taken of the settlements needed to mobilise the theoretical bearing capacity of 0.5BN, (or 0.SDN,). From centrifuge model tests on spudcan foundations, Murff et al. (1992) proposed a modified bearing capacity calculation for flat spudcans of Nyl-e aD] Nay 1) qu where w is the penetration from seabed level and N, and N, are conventional bearing capacity factors for a circular foundation. The exponential adjustment term, with proposed ot values ranging from 4 for loose sand (§' ~ 33 °) to 26 for dense sand (' ~ 39°), was to allow for sand compression prior to mobilisation of the bearing capacity. For a settlement of 5 to 10 % of the foundation diameter, this would lead to a reduction factor on the N, term of 0.2 to 0.3 for loose sand, and 0.7 to 0.9 for dense sand. ‘The approach proposed by Murff et al. is an alternative to the empirical suggestion of Terzaghi (1943), of adopting full N, and Ng values, but with a friction angle of 2/3 the true value. A more extreme case of the effect of soil compression was illustrated by Finnie and Randolph (1994) for surface foundations on calcareous silts and sands. They found that for foundations diameters ranging from 1 to 15 m the bearing pressure varied linearly with penetration depth, w, from the surface, and could be expressed as qu = My'w (18) where the ‘bearing modulus’, M, was around 35 for calcareous silt, and 35 for calcareous sand (see Figure 20). “The bearing modulus also showed a transition’ from drained to undrained conditions, as the rate of penetration was increased, with transition points of vD/cy < 0.01 for ed conditions and vD/c, > 10 for undrained conditions, where cy is the consolidation coefficient. Essentially, the bearing modulus, M, reflects the Na term in the bearing capacity equation (with Ny being zero) but the value of M is lower than theoretical N, values for the calcareous soil friction angle of 38 to 40° The effect of soil compressibility on practical bearing capacity factors for foundations in sand is an area that is well overdue for detailed study, With improvements in the ability to simulate the compression response of natural sands (e.g. Pestana and Whittle, 1995), modified bearing capacity factors for practical settlement limits of 5 to 10 % of the foundation width need to be developed. 228 ADVANCES IN GEOTECHNICAL ENGINEERING eo : ¢ Tests inst Fao x Tests in end ? Draned \ #@ 20 j & i ; of te oeccoo: coor oot ~~«00—«+0000 Penetration velocity x dlameter / coefficient of consolidation Figure 20 Effect of loading rate on non-dimensional bearing modulus (Finnie and Randolph, 1994) Interaction Diagrams for Horizontal and Moment Loading Just as for foundations on clay, the traditional treatment of the effects of horizontal and moment loading by means of multiplicative inclination and effective width factors is gradually being replaced by the use of yield envelopes in V, M, H load space. This approach has been spearheaded by Butterfield and his co-researchers (Butterfield and Ticof, 1979; Nova and Montrasio, 1991; Butterfield and Gottardi, 1995; Gottardi and Butterfield, 1993, 1995) The general form of yield surface is similar to that for clay, although it may be expressed in a very elegant form by defining normalised moments and horizontal load as (Gottardi et al., 1999) as MI/DY, . 4v(i—v)" H/Vo v = tte. where v= 19) 4v(l-v) were Vo a with the yield function then expressed simply as fant 2 os ) (mn, h, hm, ) A subscript o has been used rather than u, to emphasise the work-hardening aspect of foundations in sand, with m,Vo, and hVe representing the maximum moment and maximum horizontal load to cause yield, for the current (uniaxial) yield load, Vo. Experimental studies on circular foundations led to parameter values of m, = 0.09, he = 0.12 and a= -0.22, which are very similar to those reported by Gottardi and Butterfield (1993) for a rectangular footing with B/L = 0.2. It should be noted that the sense of the skewed yield envelope in the M-H plane (for a given vertical load, V) is the reverse of that for foundations on clay, since positive horizontal load leads to a reduction in moment capacity compared with the case of H = 0 -1=0 (20) Results from finite element analyses of strip footings on sand under combined (H-V-M) loading have been published recently by Zdravkovie et al. (2002), and these confirm the accuracy of the yield function given above. A comparison between the experimental data RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIG 229 presented in Gottardi & Butterfield (1993), their proposed parabolic and elliptical yield functions and finite element results is shown in Figure 2 06> os fae O45 04 005 0058s MeV, (©) Projection of failure loads in H-M/B plane Figure 21 Projection of failure loads for strip foundations on sand (after Zdravkovic et al., 2002) Performance of Pile-Supported Shallow Foundations ‘The latter part of the Skempton (1951) paper addressed the issue of working settlements for shallow foundations on clay, where he plotted measured settlements, normalised by the width of the foundation, as a function of the proportion of ultimate bearing resistance mobilised. He showed that both short-term and long-term settlements were approximately linear functions of these quantities, At typical factors of safety against failure, the foundation response may be evaluated sufficiently using a simple elastic approach, either through analytical solutions or by integrating vertical strains within the soil. The non-linear response of soil needs to be considered in assessing an appropriate (average) soil modulus and, iff the settlement pattern 230 ADVANCES IN GEOTECHNICAL ENGINEERING surrounding the foundation is required, the effect this non-linearity has on the distribution of strains and resulting settlements must also be considered (Jardine et al., 1986). Damage due to foundation movements is mainly associated with differential settlements and distortion of the superstructure (Burland and Wroth, 1974). However, there is a close correlation between distortion, B (ratio of differential settlement, Aw, to the spatial separation, ) and the absolute settlement. This is illustrated in Figure 22, from Ricceri and Soranzo (1985) who evaluated data from 69 case studies in Italy. They commented that a critical (maximum) settlement of 80 mm might be appropriate as a limit to avoid damage. 1.0E-01 q 1.0E-02 6 >= o “ 1.0E-03 a : oo a 4 q De oo" 1.0E-04 1-08 ° full points: damaged structures L.0E-05 + oot I 10 100 1000 Wms [or] Figure 22 Correlation between maximum distortion and settlement from case studies (Ricceri and Soranzo, 1985) Mandolini (2003) used this limit, together with simple clastic relationships, to divide foundations into different classes where design is governed respectively by an ultimate limit state (foundations smaller than a critical dimension, with low factor of safety against failure) or a serviceability limit state (larger foundations with a high factor of safety against bearing failure). For example, for a circular or square foundation on a deep homogeneous soil deposit, the average settlement (Wavg) may be expressed approximately as (Skempton , 1951) Gog 8 pee Sa 21 a. G @) where qave/qy represents the degree of mobilisation of the ultimate bearing capacity and G/sy is the rigidity index. For a typical mobilisation ratio of ~0,3 (factor of safety of ~3), and assuming a rigidity index of 100, the absolute magnitude of settlement will exceed 80 mm for foundations of dimensions greater than about 20 m. Mandolini (2003) quotes a corresponding critical dimension of 5 to 7m for foundations on sand, above which serviceability (or settlement) will be more critical than bearing capacity. RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIC 231 For foundations that are larger than the above critical dimensions, pile support may prove necessary in order to reduce the differential or maximum settlement to an acceptable level. Russo and Viggiani (1998) and Viggiani (2001) followed a similar argument in respect of pile groups, dividing them into small (width less than the pile length) and large (width greater than the pile length) categories, with the design dominated by capacity in the former group, and settlement considerations in the latter group. They noted in particular that for large pile groups, the pile cap itself will often provide sufficient factor of safety against collapse, and thus the role of the piles is in reducing settlement or modifying the pattern of settlement. The concept of viewing piles as ‘settlement reducers’ rather than as the primary means of ithstanding the loads from the superstructure was first suggested by Burland ef al. (1977). is an important distinction, since it allows the piles to be designed to carry loads close to their ultimate geotechnical capacity without compromising the safety of the complete foundation, which is provided by the raft (or pile cap). The critical design questions for composite, or piled raft, foundations may be simplified inte two broad categories, as suggested by Russo and Viggiani. For ‘small’ pile groups (B/Lp < where Ly is the length of the piles), the focus is on bearing capacity. ‘The pile cap may stiti contribute significantly to the capacity of the complete foundation, but it is necessary tc consider whether it is appropriate merely to superimpose the separate capacities of pile gro\ and pile cap in evaluating the overall capacity. For ‘large’ pile groups (B/L., > 1), settlemer considerations start to dominate and the question then is how many piles are needed (and where should they be located) in order to satisfy the serviceability criteria, Bearing Capacity of Combined Foundation Traditionally, the capacity of pile groups with a ground-contacting cap was estimated by lering the capacity of a cuboidal block of soil enclosing the piles and plan arca of the cap. was validated by observations from model tests (Whitaker, 1961) that groynd-contacting cap forced the transition from individual pile failure to block failure even at moderately large pile spacings. Cooke (1986) demonstrated the accuracy of this approach through a further series of model tests. However, with the trend towards piled rafts where tiie pile group is concentrated in the central region of the raft, or where the pile lengths vary through the group, a potentially more useful approach is to consider superposition of the capacities of the separate components as proposed by De Sanctis and Mandolini (2003), They expressed the capacity of the piled raft (PR) in terms of those of the unpiled raft (UR) and the pile group (G), as Qe Gyn Que +%6Qe (22) Results from an extensive parametric study, undertaken by means of three-dimensional finite clement analysis, suggested that it was appropriate to adopt an efficiency factor of unity for cic, with the piles beneath the raft achieving the same capacity as for a free-standing pile group This is consistent with results from centrifuge model tests, which suggest that, if anything, the presence of a ground-contacting pile cap increases the ultimate capacity of the piles (Horikoshi and Randolph, 1996; Conte et al., 2003) By contrast, De Sanctis and Mandolini (2003) found that the proportion of the raft capacity that could be mobilised at a displacement of 10 % of the raft width fell in the range 0.4 to 1. with an average value of around 0.75 and a trend of decreasing efficiency for closely spaced piles ocoupying the full raft area (see Figure 23 — where BR, L, s and d are the raft widih, and pile embedment, spacing and diameter respectively, and Ag/A is the proportion of total raft area, A, occupied by the central pile group). ‘The authors also concluded, however, that at a settlement of 2.5 % of the raft width, the factor of safety of the complete foundation was typically about 2.2 and could be estimated within reasonable accuracy by merely summing the factors of safety for each component, This suggests that it may be sufficient to adopt efficiency factors of unity in Equation 22, accepting that slightly greater displacements than 0.1B migt essary in order to mobilise the estimated overall bearing capacity. 232 ADVANCES IN GEOTECHNICAL ENGINEERING 0 Fr 4 6 8 sid FH— Rid = 28, Lid s+ DIBR/d = 28, Lid = 40 BRI = 20, Lis = 20 | mBRI = 20, Lid = 40 ABRId = 12, Lid = 20 ABRId = 12, Lid = 40 0.4 0.2 Figure 23 Deduced reduction factor, aux, on unpiled raft capacity (De Sanetis and Mandolini, 2003) Design of Pile Support ‘The potential for reducing the amount of pile support beneath raft foundations was well illustrated by Cooke (1986), who showed that the average settlement from a number of actual piled rafts Was around 0.1 % of the foundation dimension, B. This implies a factor of safety of 10 or more for a raft foundation from Equation 21 (taking G/s, ~ 100), which is consistent with the range 6 to 14 estimated by Cooke. For large foundations, with widths of 30 to 100 m, a settlement of 0.1 % of the foundation width corresponds to 30 to 100 mm. At the upper end starts to exceed the limit of 80 mm suggested by Ricceri and Soranzo (1985) and if fewer piles are used, with consequential reduction of the factor of safety against bearing failure, this limit would be reached for moderate sized foundations. The risk of damage associated with maximum settlements in excess of 80 mm can be reduced by concentrating the pile support over the central region of the raft, thereby suppressing differential settlements (Randolph, 1994). Optimising the location, as well as the number, of piles beneath the raft foundation has been investigated extensively over the last decade (Horikoshi and Randolph, 1998; Viggiani, 2001; Reul and Randolph, 2004). For a piled raft, of width B and atea Ap, assume that the pile support is provided by n piles distributed over the central area, Ac, of the raft. Design of the pile support must consider the ratio Ap/Apa, the relative pile length, Ly/B, and the total pile capacity, Qo, in comparison with the total applied load, Ppr. Horikoshi and Randolph suggested that the differential settlement between the centre and mid-sides of the raft, under uniform loading could be largely eliminated for a set of optimal conditions where total pile capacity is 40 to $0 % of the applied load and the area of pile support is 20 to 30 % of the total raft area. As shown in Figure 24, for that load range, the piles are loaded to about 80 % of their ultimate capacity (PG/Qa ~ 0.8) and the piles carry about 40 % of the total applied load (Po/Por ~ 0.4).. 0 etn RE RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIG 233 0.6 Proportion of 0.5 load carried by piles, Po/Pp_ 0.4 03 oe Ag/App = 0.09 0.2 oe L/B~ 1s 0 02 04 06 08 1 Ratio of total pile capacity to total applied load, Qe/Pp Figure 24 Optimised conditions for minimal differential settlement (Horikoshi and Randolph, 1998) Broadly similar conclusions were reached in the extensive parametric study presented by Viggiani (2001) where the total length of pile support (nL,) to minimise the differentict settlement was evaluated for different pile group geometries, Fe showed that a few long piles Were more efficient (ie, lower total length piles) at reducing differential settlement thant large number of shorter piles. Even where a significant proportion of the load is concentrated at tre edges of the raft, central pile support is still effective at reducing differential settlements and the absolute magnitude of raft bending moments (Reul and Randolph, 2004; Randolph, 20030), The above studies have focused on somewhat simplified conditions, such as uniform loading and soil stiffness that is relatively constant with depth. For practical situations, the optimal conditions will vary and will need to be evaluated through numerical analysis using one of the many software tools now available (Poulos, 2001). Even where the soil conditions are relatively soft at foundation level, the general principle of minimising differential settlements by the use of much shorter (if any) piles towards the edges of the foundation and longer piles in the centre, will still hold. An excellent example of this was published recently by Liew et al 2002), for a 20m diameter oil storage tank sitting on soft clay, where the shear strength increased approximately linearly from 5 kPa at foundation level to 65 kPa at a depth of 35 m- The operating load for the tank was around 25 MN, corresponding to an applied pressure of 80 kPa on the soil, which is well in excess of the bearing capacity at the soil surface. ‘The foundations were designed with a very thin (flexible) raft, and piles that varied in length from 74m at the edge of the tank to 36 m in the centre, as shown in Figure 25a. The tank was proof loaded to just under 30 MN and the settlement profile measured using horizontal inclinometers installed beneath the tank. ‘The resulting settlement profiles are shown in Figure 25b. As may be seen, in spite of the shorter piles towards the edges of the foundation, the settlement profile still shows a pronounced concavity towards the centre of the tank. Settlement profiles computed from the pile group analysis software Piglet (Randolph, 2003a) are also shown in Figure 25b, These have been matched to the maximum’ measured settlements for each load level, and show reasonable agreement, but without the asymmetry evident in the measured profiles at high load levels, For comparison, a profile is also shown for the 25 MN load, based on uniform pile lengths of 30 m (as opposed to the varying pile lengths of the actual design). That leads to differential settlements of 29 % of the average, compared with 22 % for the actual design, confirming the effectiveness of concentrating pile support beneath the centre of the tank. 234 ADVANCES IN GEOTECHNICAL ENGINEERING Tank (20 m dia.) zy Paar 0.5 m thick sand pad > 0.5 m thick concrete rat ~ 0.35 m diameter spun concrete pipe piles (a) Geometry of pile group Position across tank (m) 4 82 0 2 4 6 & 10 PIGLET (actual pile kengths) Settlement (mm) (b) Measured and computed settlement profiles beneath tank Figure 25 Pile group and settlements during proof loading of 20 m diameter oil tank (Liew et al., 2002) RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIC 235 Conclusions One of the main purposes of this paper has been to assemble some of the theoretical advances made over the last few decades in the treatment of bearing capacity of foundations. The design codes in most countries are still based on the use of a variety of multiplicative adjustment factors for the bearing capacity of shallow foundations on clay or sand. However, in many cases rigorous solutions, which can be computed to high accuracy within a few seconds, are now available. ‘This should allow ‘basic’ bearing capacity calculations, based on simple rigid plastic models of soil response, to be conducted consistently and unequivocally by every engineer without the need for semi-empirical correlations or adjustment factors. Attention can then be focused, in a logical manner, on those aspects of real soil behaviour that require adjustment of the basic solutions ~ for example strength anisotropy, layered soil profiles, stress-dependent strength, compressibility and so forth. For uniaxial loading, the bearing capacity factors given in Tables 1, 2 and 3 are believed to be accurate to three significant figures. For combined loading, interaction diagrams should replace the use of corrections for inclined and eccentric loading and some examples of these are given, either analytically (Equations 5, 6, 20) or in diagram form (Figures 11, 21). ‘The challenge now is to bring these into our design codes, together with supporting commentary on how they should be modified as discussed above. For clays, determination of the shear strength profile is of paramount importance, which can be achieved through field (penetration or vane) tests, or by careful sampling followed by laboratory testing. The former approach leads to some form of average shear strength for the different modes of shearing undergone in the test, while the latter approach allows separate evaluation of shear strengths in these different modes, Whereas traditionally the triaxial compression test was the standard for evaluating strength profiles, other forms of test generally give lower strengths, and may be more suitable in arriving at an appropriate average shear strength for bearing capacity assessment. In sands, the stress-level dependency of the friction angle and the effects of compressibility are critical factors for consideration, and are reflected in the cone resistance. However, it is not possible to separate out these (and other) effects from the single measurement of cone resistance, and correlations between allowable bearing capacity, cone resistance, and some independent measure of soil compressibility, perhaps from reconstituted samples, may prove the most reliable design approach. Skempton made extensive contributions on many aspects of the design of shallow and deep foundations and the present paper has only addressed a small part of this topic. In particular, it has not been possible to consider design parameters for pile foundations here, although discussion has been included on composite piled-raft foundations, ‘The principle of designing for ultimate limit state using the bearing capacity of a shallow foundation, possibly enhanced by relatively shallow ground treatment, and using piles or some other form of deep reinforcement of the soil to ensure serviceability limit states are satisfied, is an important and growing trend within foundation engineering, At this stage, the concept of fully optimised piled raft foundations has been argued as a concept and validated through numerical analysis — but this design approach has yet to become accepted in practice and remains as a challenge for the coming decade, References Alsamman, O.M. (1995). The Use of CPT for Calculating Axial Capacity of Drilled Shafis. PhD Thesis, University of Illinois of Urbana-Campaign, IL. Baligh, MM. (1975). Theory of Deep Site Static Cone Penetration Resistance. Dept of Civil Engineering. MIT, Cambridge, Mass, Publ. N R75-56. Berezantsev, V.G. amd Kovalec, LV. (1968). Consideration of the curvilinear of the shear graph when conducting tests on model foundations. Translated from Osnovanija, Fundamenty I Mekhanika Gruntov, \(Jan-Feb), 1-4 236 ADVANCES IN GEOTECHNICAL ENGINEERING Bolton, M.D. (1986). The strength and dilatancy of sands. Géotechnique, 36 (1), 65-78. Bolton, M.D. and Lau, C.K. (1993). Vertical bearing capacity factors for circular and strip footings on Mohr-Coulomb soil. Can. Geotech. J. 30, 1024-1033. Booker, J.R. and Small, J.C. (1986). The behaviour of an impermeable flexible raft on a deep layer of consolidating soil. Int.. J. for Num. and Anal. Methods in Geomech,, 10, 311-327. Bransby, M.F. (2002). The undrained inclined load capacity of shallow foundations after consolidation under vertical loads. Proc. Int. Conf. on Numerical Methods in Geomech., Rome, 431-437. Bransby, M.F. and Randolph, M.F. (1998). Combined loading of skirted foundations, Géotechnique, 48(5), 637-655. Burland, J.B., Broms, B.B., & De Mello, V.F.B (1977). Behaviour of foundations and structures. Proc. 9" Int. Conf. on Soil Mech. and Found. Engrg., Tokyo, 2, 495-546. Burland, J.B. and Wroth, C.P. (1974). Settlement of buildings and associated damage: State of the art review. Proc. Conf on Settlement of Structures, Cambridge, 611-654, Bustamante, M. and Gianeselli, L. (1982). Pile bearing capacity prediction by means of static penetrometer. Proc. ESOPT II, Amsterdam. Butterfield, R., and Ticof, J. (1979). The use of physical models in design. Discussion, Proc. 7" Ewrop. Conf on Soil Mech. and Found. Eng., 4, 259-261. Butterfield, R. and Gottardi, G. (1994). A complete three-dimensional failure envelope for shallow foundations on sand. Géotechnique, 44(1), 181-184, Butterfield, R., Houlsby, G.T. and Gottardi, G. (1997). Standardised sign conventions and notation for generally loaded foundations. Géotechnique, 47(5), 1051-1054, idy, M.J. and Houlsby, G.T. (2002). Vertical bearing capacity factors for conical footings on sand, Géotechnique, 52(9), 687-692. Chung, S.F, and Randolph, M.F. (2004). Penetration resistance in soft clay for different shaped penetrometers. Proc. 2nd Int. Conf. on Site Characterisation, Porto, Cooke, R.W. (1986). Piled raft foundations on stiff clays: a contribution to design philosophy. Géotechnique, 36(2), 169-203. Conte, G., Mandolini, A. and Randolph, M.F. (2003). Centrifuge modelling to investigate the performance of piled rafts. Proc. 4"" Int. Geotech. Sem. on Deep Found. on Bored & Auger Piles - BAP IV, Ghent, 355-366. Davis, E.H. and Booker, J.R. (1971). The bearing capacity of strip footings from the standpoint of plasticity theory. Proc, Ist Australia-New Zealand Conf. on Geomech., Melbourne, 276-282. s, EH. and Booker, J.R. (1973). The effect of increasing strength with depth on the bearing capacity of clays. Géotechnique, 23(4), 551-563, De Beer, E. (1967). Bearing capacity and settlement of shallow foundations on sand, Proc, Symp. on Bearing Capacity and Settlement of Foundations. Dept of Civil Engineering, Duke Univ. Durham, NC, Edit. A. Vesic, 15-43, Delong, J.T., Yafrate, N.J., DeGroot, D.J., and Jakubowski, J. (2004). Evaluation of the undrained shear strength profile in soft layered clay using full-flow probes. Proc. 2nd Int. Conf, on Site Characterisation, Porto. De Sanetis, L. and Mandolini, A. (2003). On the ultimate vertical load of piled rafts on soft clay soils. Proc. Int. Symp. On Deep Foundations on Bored and Augered Piles, BAP IV, Ghent, 379-386. Erickson, H.L. and Drescher, A. (2002). Bearing capacity of circular footings. J. Geotech. Geoenviron. Eng, 128(1), 38-43 Finnie, M.S. and Randolph, M.F. (1994). Punch-through and liquefaction induced failure of shallow foundations on calcareous sediments, Proc. Int. Conf. on Behaviour of Offshore Structures, BOSS '94, Boston, 217-230. Cc Da\ (arr RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIC 237 Fioravante, V. (1994). Centrifuge modelling of axially loaded piles in sand. Proc Workshop on Pile Foundations Experimental Investigations. Analyses and Design, Napoli, 125-163, Floravante, V., Ghionna, V.N., Jamiolkowski, M.B. and Pedroni, $, (1995). Load carry 2 capacity of large diameter bored piles in sand and gravel. Proc. 10” ARCSMFE, 2, 3-13 Fleming, W.G.K., Weltman, A.J, Randolph, M. F. and Elson, W.K, (1992) Piling Engineering. Blackie Academic and Professional, Glasgow, 2nd Edit. Franke, E. (1991). Collection of load settlement curves, Part 2: Piles soils. Internal Report, University of Dortmund. Frydman, S. and Burd, H.J. (1997). Numerical studies of bearing-capacity factor Ny. uJ Geotech. Geoenviron, Eng., 123(1), 20-29. Ghionna, V.N., Jamiolkowski, M.B., Pedroni. S, and Sangado, R. et al (1994). Tip displacement of drilled shafts in sands. in Vertical and Horizontal Deformations of Foundations and Embankments, Ed. A.T. Yeung and G.Y. Felio, ASCE , GSP40, New York, 2, 1039-1057, Cottardi, G. and Butterfield, R. (1993). On the bearing capacity of surface footings on sand under general planar loads. Soils and Foundations, 33(3), 68-79. Gottardi, G. and Butterfield, R, (1995). The displacement of a model rigid surface footing on dense sand under general planar loading, Soils and foundations, 35(3), 71-82. Gottardi, G., Houlsby, G.T. and Butterfield, R. (1999). Plastic response of circular footings on sand under general planar loading. Géotechnique, 49(4), 453-469, Gourvenec, S.M. and Randolph, M.F. (2003). Effect of strength non-homogeneity on the shape of failure envelopes for combined loading of strip and circular foundations on clay, Géotechnique, 53(6), 575-586. Green, AP. 1954. The plastic yielding of metal junctions due to combined shear and pressure J. Mech, Phys. Solids, 2: 197-211. Hansen, J.B. (1970). A revised and extended formula for bearing capacity. Bulletin No. 28, Danish Geotechnical Institute. Horikoshi, K. and Randolph, M.F. (1996). Centrifuge modelling of piled raft foundations on clay. Géotechnique, 46(4), 741-752. Horikoshi, K. and Randolph, M.F. (1998). Optimum design of piled rafts. Géotechnique, 48(3), 301-317. Houlsby, G. T. and Martin, C. M. (2003). Undrained bearing capacity factors for conical footings on clay. Géotechnique. 53(5), 513-520. Houlsby, G.T. & A.M. Puzrin 1999. The bearing capacity of a strip footing on clay under combined loading. Proc. R. Soc. London (Ser. A), 455: 893-916, Houlsby, G.T. and Wroth, C.P. (1983), Calculation of stresses on shallow penetrometers and footings. Proc. IUTAMIUGG Seabed Mechanics, Newcastle, 107-112, Hu, Y., Randolph, M.F, and Watson, P.G. (1999), Bearing response of skirted foundations on non-homogeneous soil, J. Geot. Eng. Div., ASCE, 125(11), 924-935 Hu, Y. and Randolph, M.F. (2002). Bearing capacity of caisson foundations on normally consolidated clay. Soils and Foundations, 42(5), 71-77. Jamiolkowski, M, (1990). Shear strength of cohesionless soils from CPT. De Mello Volume 191-204, Ed. Edgar Blucker, San Paulo. Jamiolkowski, M.B., Lo Presti, D.C.F, and Manassero, M. (2003). Evaluation of relative density and sheer strength of sands from cone penetration test (CPT) and flat dilatometer (DMT), Soil Behaviour and Soft Ground Construction, Eds. 1.7. Germain, T.C. Sheahan and R.V. Whitman, ASCE, GSP 119, 201-238. Jamiolkowski, M.B. and Sarri, H. (2000). On axial load capacity of drilled shafts in coarse grained soils. Libro Homenaje on José Antonio Jiménes Salas. Ministero de Fomento, Cedex, Madrid, 111-128 tip in cohesionless 238 ADVANCES IN GEOTECHNICAL ENGINEERING Jardine, R.J., Lehane, B.M., Smith, P.R. and Gildea, P.A. (1995). Vertical loading experiments on rigid pad foundations at Bothkennar, Géotechnique, 45(4), 373-597. Jardine, R.J. and Menkiti, C.O. (1999). The undrained anisotropy of Ko consolidated sediments. Proc. 12th Eur. Conf. Soil Mech. and Geot. Eng., Amsterdam, 1101-1108. Jardine, R.J., Potts, D.M., Fourie, A.B, and Burland, J.B. (1986). Studies of the influence of non-linear stress strain characteristics in soil-structure interaction. Géotechnique, 36(3), 377-396. Ladd, C. (1991). Stability evaluation during staged construction. J. Geor, Eng., ASCE, 117(4), 540-615 Lancellota, R. (1993). Stability of a rigid column with non-linear restraint. Géotechnique, 43(2), 331-332. Lauritzsen, R. and Schjetne, K. (1976). Stability calculations for offshore gravity structures. Proc. 8th Offshore Tech. Conf., Houston, 75-82. Lee, J.H. and Salgado, R. (1999). Determination of pile base resistance in sands. J. of Geor. and Geoenv, Eng., 125(8), 673-683. Lehane, B.M. and Jardine, R.J. (2003). Effects of long-term pre-loading on the performance of a footing on clay, Géotechnique, 53(8), 689-695. Levin, A., 1955. Indentation pressure of a smooth circular punch. Quart. Appl. Math., 13, 381-389 Liew, S.S., Gue, S.S. and Tan, Y.C, (2002), Design and instrumentation results of a reinforcement concrete piled raft supporting 2500 ton oil storage tank on very soft alluvium deposits. Proc. 9” Int. Con, on Piling and Deep Foundations, Nice. Maksimovic, M. (1989). Non-linear failure envelope for coarse-grained soils. Proc. 12th ICSMEE, Rio De Janeiro, Brazil, 731-734. Mandolini, A. (2003). Design of piled raft foundations: practice and development. Proc. Int Symp. On Deep Foundations on Bored and Augered Piles, BAP IV, Ghent, 59-80. Martin, C.M. (2001). Vertical bearing capacity of skirted circular foundations on Tresca soil. Proc, 15th ICSMGE, Istanbul, 1, 743-746. Martin, C.M. (2003). New software for rigorous bearing capacity calculations. Proc. Int. ‘Symp. on Foundations: Innovations, Observations, Design and Practice. Dundee, 381-592. Martin, C.M. (2004). Discussion of Ukritchon et al. (2003), submitted to J. of Geot. and Geoenv. Eng., ASCE Martin, C.M. and Houlsby, G.T. (2001). Combined loading of spudcan foundations on clay: numerical modeling. Géotechnique, 51(8), 687-699. Martin, C.M. and Randolph, M.F. (2001). Applications of the lower and upper bound theorems of plasticity to collapse of circular foundations, Proc. 10th IACMAG, Tucson, 2, 1417-1428, Merifield, R.S., Lyamin, A.V., Sloan, SW. and Yu, H.S. (2003). Three-dimensional lower bound solutions for stability of plate anchors in clay. J. of Geot. and Geoeny. Eng., 129(3), 243-253. Merifield, R.S., Sloan, $.W. and Yu, H.S. (1999). Rigorous plasticity solutions for the bearing capacity of two-layered clays. Géotechnique, 49(4), 471-490. Meyerhof, G.G. (1953). The bearing capacity of foundations under eccentric and inclined loads, Proc. 3rd Int. Conf, on Soil Mech, and Found. Eng., Zurich, 1, 440-445. Meyerhof, G.G. (1973). Some recent research on the bearing capacity of foundations. Canadian Geotechnical Journal, \(1), 16-26. Meyerhof, G.G. and Hanna, A.M. (1978). Ultimate bearing capacity of foundations on layered soils under inclined load, Canadian Geotechnical Journal, 15, 565-572. Michalowski, R.L. (1997). An estimate of the influence of soil weight on bearing capacity using limit analysis. Soils and Foundations, 37(4), 57-64. RANDOLPH, JAMIOLKOWSKI AND ZDRAVKOVIC 239 Murff, J.D., Prins, M.D., Dean, E.T.R., James, R.G, and Schofield, A.N. (1992), Jackup rig foundation modelling, Proc. 24th Annual Offshore Technology Conf, Houston, OTC 6807. Nova, R. and Montrasio, L. (1991). Settlements of shallow foundations on sand. Géotechnique, 41(2), 243-256. Pestana, J.M. and Whittle, A.J. (1995). Compression model for cohesionless soils. Géotechnique, 45(4), 611-631. Potts, D.M. (2003). Numerical analysis: a virtual dream or practical reality. 42nd Rankine Lecture, Géotechnique, 53(6), 533-573. Potts, D.M. and Burland, J.R. (2000). Development and application of a numerical model for simulating the stabilisation of the Leaning Tower of Pisa. Proc. Booker Memorial ‘Symposium: Developments in Theoretical Geomechanics, Sydney, 737-758. Potts, D.M. and Zdravkovic, L. (2001). Finite Element Analysis In Geotechnical Engineering: Application, Thomas Telford, London. Poulos, H.G. (2001). Piled raft foundations: design and applications. Géotechnique, 51(2), 95- 113, Pradhan, T.B.S., Tatsuoka, F., Horii, N. (1988). Simple shear testing on sands in a torsional shear apparatus. Soils and Foundations, 28(2), 95-112. Prandtl, L. (1921). Eindringungsfestigkeit und festigkeit von schneiden, Zeit. angew. Math. Mech 1(1), 15-20. Randolph, M.F. (1994). Design methods for pile groups and piled rafts. Proc. 13" Int. Conf. on Soil Mech. and Found. Eng., New Delhi, 5, 61-82. Randolph, M.F. (2003a). PIGLET: Analysis and Design of Pile Groups, Version 5.0. Users’ Manual, Perth, Randolph, M.F. (2003b). 43 Rankine Lecture: Science and empiricism in pile foundation design. Géotechnique, 53(10), 847-875. Randolph, M. F., Watson, P. G. and Fahey, M. (1999). Site characterisation and foundation design in soft sediments, Proc. Int. Conf. on Offshore and Nearshore Geotechnical Engineering, Navi Mumbai, 35-47. Reul, O. and Randolph, MF. (2004). Design strategies for piled rafts subjected to non- uniform vertical loading. J. Geotech. and Geoenv. Eng. Div, ASCE, 130(1), 1-13 Ricceri, G. & Soranzo M. 1985. An analysis of allowable settlements of structures. Rivista Naliana di Geoteenica, 19(4), 177-188. Rowe, P.W. (1962). The stress-dilatancy relation for static equilil particles in contact. Proc. Royal Society London, 269A, 500-527. Russo, G. and Viggiani, C. (1998). Factors controlling soil-structure interaction for piled rafts. Proc. Int. Conf. on Soil-Structure Interaction in Urban Civil Eng., Darmstadt, 297-322 Salengon, J. and Matar, M. (1982). Capacité portante des foundations superficielles circulaires. Journal de Mécanique Théorique et Appliquée, \(2), 237-267. Salgado, R., Bandini, P, and Karim, A. (2000), Shear strength and stiff Geotech and GeoEnviron Eng., ASCE, 126(5), 451-462 Sarri, H. (2001). Pali Trivellati in Sabbia Soggetti a Carico Assiale: Modellazione Fisica e Numerica. PhD Thesis, University of Torino, Sasistharan, S. (1989). Stress path dependency of dilatancy and stress-strain response of sand. M. Sc. Thesis, Dept. of Civil Engineering, University of British Columbia, Vancouver. Skempton, A. W. (1951). The bearing capacity of clays. Building Research Congress, London, 1, 180-189, Skempton, A. W., Yassin, A.S. and Gibson, R.E. (1953). Théorie de la force portante des pieux. Annales de l'Institut Technique du Batiment et des Traveaux Publics, 6, 285-290. Sokolovskii, V.V. (1965) Statics of Granular Media, Pergamon, New York. rium of an assembly of of silty sand. J. of 240 ADVANCES IN GEOTECHNICAL ENGINEERING Taiebat, H.A. and Carter, J.P, (2000). Numerical studies of the bearing capacity of shallow foundations on cohesive soil subjected to combined loading. Géotechnique, 50(4), 409- 418. ‘Taiebat, H.A. and Carter, J.P. (2002). Bearing capacity of strip and circular foundations on undrained clay subjected to eccentric loads, Géotechnique, 52(1), 61-64. Tatsuoka, F., Jardine, RJ. Lo Presti, D.C.F., Di Benedetto, H., Kohata, T. (1997). Characterizing the pre-failure deformation properties of geomaterials”. Proc 14th ICSMFE, Hamburg, 4, 2129-2164. Terzaghi, K. (1943). Theoretical Soil Mechanics, Wiley, New York. Ukritchon, B. Whittle, A.J. and Klangvijit, C. (2003). Calculations of bearing capacity factor N, using numerical limit analyses. J. Geot, and Geoeny. Eng., ASCE, 129(6), 468-474, Ukritchon, B. Whittle, A.J. and Sloan, S.W. (1998). Undrained limit analysis for combined loading of strip footings on clay. J. Geot. and Geoeny, Eng., ASCE 124(3): 265-276. Vesic, A.S. (1973). Analysis of ultimate loads of shallow foundations. J. Soil Mech. and Found. Eng. Div., ASCE, 99(SM1), 45-73. Viggiani, C. (2001). Analysis and design of piled foundations, First Arrigo Croce Lecture, Rivista taliana di Geotecnica, 35(1), 47-75. Wang, C.X. and Carter, J.P. (2002), Deep penetration of strip and circular footings into layered clays. Int. J. of Geomechanics, 2(2), 205-232. Whitaker, T. (1961). Some experiments on model piled foundations in clay. Proc. Symp. on Pile Foundations, 6" Congr. Int. Assoc. Bridge and Struct. Eng., Stockholm, 124-139. Yaroshenko, V.A. (1964). Interpretation of the results of static penetration of sands. Fundamety Projekt, n. 3. Yoon, Y. (1991). Static and Dynamic Behaviour of Crushable and Non-Crushable Sands Ph.D. Thesis, Ghent University. Zdravkovic, L. and Jardine, RJ. (2000). Undrained anisotropy of K, consolidated silt, Canadian Geotechnical J., 37, 178-200. Zdravkovic, L., Ng, P.M. and Potts, D.M. (2002). Bearing capacity of surface foundations on sand subjected to combined loading. Proc. Int. Conf. on Num. Methods in Geotech. Eng,, NUMGE V, 323-330, Zdravkovic, L., Potts, D.M. and Jackson, C. (2003), Numerical study of the effect of preloading on undrained bearing capacity. Int. J. of Geomechanics, 3, 1-10. Zaravkovie, L., Potts, D.M. and Jardine, R.J. (2001). A parametric study of the pull-out capacity of bucket foundations in soft clay. Géotechnique, 51(1), 55-67.

You might also like