Você está na página 1de 13

KSCE Journal of Civil Engineering (2013) 17(4):699-711 DOI 10.

1007/s12205-013-0406-x

Geotechnical Engineering

www.springer.com/12205

Model Studies of Bearing Capacity of Strip Footing on Sand Slope


M. Salih Keskin* and Mustafa Laman**
Received December 16, 2011/Accepted August 27, 2012

Abstract
An experimental investigation into the ultimate bearing capacity of strip footing on sand slope is reported. The parameters investigated are the effect of setback distance of the footing to the slope crest, slope angle, relative density of sand and footing width on the ultimate bearing capacity of strip footings. A series of finite element analyses was additionally performed on a prototype slope to ascertain the validity of the findings from the laboratory model tests and to supplement the results of the model tests. The agreement between observed and computed results is found to be reasonably well in terms of load-settlement and general trend of behavior. The results show that the ultimate bearing capacity increases with increase in setback distance, relative density of sand, footing width and decrease in slope angle. At a setback distance of five times of the width of the footing, bearing capacity remains constant like that of a footing on level ground. Keywords: slope, bearing capacity, shallow foundation, laboratory test, finite element method

1. Introduction
The bearing capacity of the foundations is a primary concern in the field of geotechnical engineering. Design of foundations on a horizontal ground surface depends on the mechanical characteristics of the soil such as unit weight, shear strength etc., and the physical properties of the foundation such as depth, width, and shape. There are two considerations to decide the allowable bearing pressures for shallow foundations; the safety factor against ultimate shear failure must be adequate and the settlements under allowable bearing pressures should not exceed tolerable values. Several methods may correctly predict the bearing capacity of foundations resting on or in level grounds (Terzaghi, 1943; Meyerhof, 1963; Hansen, 1970; Vesic, 1975). They are either based on laboratory or in-situ test results. However, there are many circumstances where foundations must be built on or near a slope. Due to the land limitation, architectural and economical purposes, structures are generally placed on the slope crest or at a setback distance from the slope crest. Examples include bridge piers supported on approach embankments, foundations on electrical transmission towers and some buildings. The stability of the slope and the bearing capacity of a foundation constructed near to the edge of a slope are important factors in the performance of the structure built near a slope. When a foundation is constructed on sloping ground, one side of the foundation is exposed to the sloping surface. Thus, as the foundation soil approaches limit state the plastic region of failure is very limited and it significantly affect the mechanical stability

of slope and thereby the bearing capacity of the foundation. Consequently, the bearing capacity of a foundation near a sloped fill is less than that on a flat ground. Therefore, in many situations, the conventional types of shallow foundations are inadequate even though they are very economical and easier to construct. Hence, researchers have paid attention to investigate the bearing capacity and settlement behavior of footings on slopes. The methods to predict the bearing capacity of footings on or in level grounds are well developed. However the bearing capacity of footings near or on slopes, much still remains to be investigated. The theoretical methods on the bearing capacity of footing on a slope (Meyerhof, 1957; Hansen, 1970; Vesic, 1975; Saran et al., 1989) are generally developed based on the bearing capacity equation proposed by Terzaghi (1943) and valid for a limited range of footing location and embedded depth. Meyerhof (1957) proposed a theoretical solution to determine the ultimate bearing capacity of a shallow foundation located on the face of a slope. According to Meyerhof, the ultimate bearing capacity can be expressed as: 1 - BNq qu = cNcq + -2 where, B = Width of footing c = Cohesion of soil Ncq, Nq = Bearing capacity factors = Unit weight of soil (1)

*Assistant Professor, Dept. of Civil Engineering, University of Dicle, Diyarbakir 21280, Turkey (Corresponding Author, E-mail: mskeskin@dicle.edu.tr, mskeskin21@gmail.com) **Professor, Dept. of Civil Engineering, University of Cukurova, Adana 01330, Turkey (E-mail: mlaman@cu.edu.tr) 699

M. Salih Keskin and Mustafa Laman

For the special case of a footing resting on the surface of a saturated, clean, granular material (apparent cohesion, c = 0), the first term of Eq. (1) become zero and Meyerhofs equation reduces to 1 - BN q qu = -2 (2)

Brinkgreeve and Vermeer, 2002) to verify the model test results.

2. Experimental Study
2.1 Test Set-up A series of laboratory model tests were performed in a test box made of a steel frame with inside dimensions of 1.140 m (length), 0.475 m (width) and 0.500 m (depth) as shown in Fig. 1. The bottom and vertical edges of the box were stiffened using angle sections to avoid lateral yielding during soil placement and loading of the model footing. The two sidewalls of the test box were made of 20 mm thick glass to see the sand sample during preparation and observe the sand particle deformations during the tests. The box was enough rigid to provide plane strain conditions for all model tests. Static vertical loads were applied to the model footings by a motor-controlled hydraulic jack system. The system attached to the loading frame located above the test box has a loading rate of 0.5 mm/min. An electronic 15 kN capacity load cell was used to measure applied loads. Settlements of the footing were measured using two Linear Variable Displacement Transducers (LVDTs) located at the two corners of the model footing. The load cell and the displacement transducers were connected to data acquisition system (ADU) for recording and data handling.

Graham et al. (1987) provided an analytical solution for the bearing capacity of a shallow footing on the top of a cohesionless slope based on the method of stress characteristics. Gemperline (1988) and Shields et al. (1990) developed empirical equations for the ultimate bearing capacity factors for a footing on a slope based on centrifuge tests. Narita and Yamaguchi (1990) used a method of slices to determine the bearing capacity of a three dimensional footing located on top of a slope. Buhan and Garnier (1998) used yield design theory to evaluate the ultimate bearing capacity of a shallow rectangular footing located on top of a slope. Furthermore, experimental studies on the bearing capacity and settlement behavior of footings on slopes are relatively limited (Shields et al., 1977; Garnier et al., 1984; Gemperline, 1988). Shields et al. (1977) have carried out series of experiments to obtain the bearing capacity factor Nq for a footing on cohesionless slope. They showed that the theory of Meyerhof (1957) overestimates the magnitude of the bearing capacity. However, at shallow depths close to the edge of the slope the theory is closer to the experimental values. Garnier et al. (1984) presented an experimental study on strip footing near a slope to evaluate the coefficient of reduction of bearing capacity due to slope effect. The tests were performed using three slope models and loads were applied on the model footing at different distances from the edge of slope. For different slope models, it was found that the bearing capacity of the footing was not practically different from the value of distance/width ratio (b/B) greater than 6 due to the effect of slope. Research in the area of bearing capacity of footings on sloping ground is very much in demand because of the significant effect of the slope on the bearing capacity. Physical modeling is one of the best approaches to overcome the limitations of analytical methods (Wood, 2004). It is common practice in geotechnical engineering that the results gained from a physical model are used to validate analytical and numerical models. Although the laboratory model tests have several drawbacks such as the scale effect, model tests do provide reasonable understanding of the bearing capacity of shallow foundations. The object of this study is to investigate the ultimate bearing capacity of strip footings located on top of a slope with laboratory model tests. In the study, the relationship between the footing response and the variable parameters including, edge distance between the footing and the crest of slope, angle of the slope inclination, relative density of sand and width of the footing were investigated. And also, numerical analyses on a prototype footing-slope system were conducted using a commercial finite element program PLAXIS (professional version 8,

Fig. 1. Schematic View of the Experimental Set-up: (a) Side view, (b) Plan view
KSCE Journal of Civil Engineering

700

Model Studies of Bearing Capacity of Strip Footing on Sand Slope

Loading tests were carried out on two model strip footings in order to investigate the effect of footing width. The model footings were 70 mm and 50 mm in width, 465 mm in length and 20 mm in thickness and fabricated from mild steel with a hole at its center to accommodate a ball bearing. The footings were located on the sand. The lengths of the footings were made almost equal to the width of the test box of the tank to maintain plane strain conditions. The load was applied to the model footing through a ball bearing which was placed between the model footing and the proving ring and allowed the footing to rotate freely as it approached failure and eliminated any potential moment transfer from the loading fixture. 2.2 Model Ground The soil used for the model tests was uniform, clean and fine sand obtained from Cakit River bed. The sand was washed, dried and sorted by particle size. The particle size distribution was determined using the dry sieving method and the results are shown in Fig. 2. Using the Unified Soil Classification System, the material was determined to be poorly graded sand (SP). The specific gravity of the soil particles was determined by picnometer test. The maximum and minimum dry densities of the sand were measured and corresponding values of the minimum and maximum void ratios were calculated. Table 1

summarizes the general physical characteristics of the sand. To obtain a reasonably homogeneous sand bed throughout the experimental study, the same compaction procedure was used to deposit sand in 50 mm thick layers into the model box. In this method the quantity of sand for each layer, which was required to produce a specific relative density, was first weighed and placed in the box and compacted by a hand-held vibratory compactor until achieving the required layer height. The experimental tests were conducted on samples prepared with average unit weights of 16.5, 17.0 and 17.5 kN/m3. Corresponding relative densities of the samples were 45, 65 and 85%, respectively. The estimated internal friction angles of the sand were 40.6, 41.8 and 43.5o, respectively. 2.3 Preparation of Sand Slope Model sand slopes with slope angles () of 20, 25 and 30 were prepared by using the same compaction procedure in layers of 50 mm thick sand. The inner surfaces of the test box were

Fig. 2. Grain Size Distribution of the Model Sand

Table 1. Properties of Sand Bed Property Coarse sand fraction (%) Medium sand fraction (%) Fine sand fraction (%) D10 (mm) D30 (mm) D60 (mm) Uniformity coefficient, Cu Coefficient of curvature, Cc Specific gravity (kN/m3) Maximum dry unit weight (kN/m3) Minimum dry unit weight (kN/m3) Maximum void ratio Minimum void ratio Classification (USCS)
Vol. 17, No. 4 / May 2013

Value 00.0 46.4 53.6 00.18 00.30 00.50 02.78 01.00 02.68 17.9 15.5 00.729 00.497 SP

Fig. 3. Procedure for Construction of Sand Slope: (a) Soil Compacted Until Level Ground, (b) Apparatus Placed on Level Ground, (c) Compacting the First Sloped Layer, (d) Compacting the Second Sloped Layer, (e) Compacting the Third Sloped Layer, (f) Compacting the Last Sloped Layer, (g) Removing the Apparatus, (h) Footing Placed at the Surface
701

M. Salih Keskin and Mustafa Laman

Table 2. Model Test Program Series 1 2 3 4 5 6 7 8 Constant parameters Tests on level ground, =0, B=70 mm Test on level ground, =0, B=50 mm, Dr = 65% =30, B=70 mm, Dr = 65% =25, B=70 mm, Dr = 65% =20, B=70 mm, Dr = 65% =30, B=70 mm, Dr = 45% =30, B=70 mm, Dr = 85% =30, B=50 mm, Dr = 65% Variable parameters Dr = 45-65-85% b/B=0-1-2-3-4-5 b/B=0-1-2-3-4-5 b/B=0-1-2-3-4-5 b/B=0-1-2-3-4-5 b/B=0-1-2-3-4-5 b/B=0-1-2-3-4-5

marked at 50 mm intervals to make easy the preparation of the sand bed in layers and the geometry of the slope was marked on the glass walls for reference. The sand was compacted in layers up to slope toe and then a special adjustable apparatus to get the predetermined slope angles developed in this study was placed to obtain the sloping surface. The process continued layer by layer until the height of the slope was reached. Great care was given to level the slope face using special apparatus so that the relative density of the top surface was not affected (Fig. 3). In this method there is no need for excavation and seems to be useful to form a sloping surface with a desired angle and allows compact the sand uniformly. The model strip footing was then placed on the surface of the compacted sand and finally the load was applied until reaching failure. 2.4 Test Program An experimental program was carried out to investigate the effects of the variable parameters including, edge distance of the footing to the slope crest (b), angle of the slope inclination (), relative density of sand (Dr), and width of the footing (B) on the bearing capacity of strip footing on a sand slope. Model loading tests were performed in seven test programs. Table 2 summarizes all the tests programs with constant and variable parameters used. Some tests were repeated at least twice to verify the consistency of the test data.

a decreasing stiffness and simultaneously irreversible plastic strains develop. The observed relationship between the pressure and axial strain can be well approximated by a hyperbola as used in the variable elastic, hyperbolic model (Duncan and Chang, 1970). However the HSM is far superior to the hyperbolic model, being capable of simulating non-linear, inelastic, stress dependent material behaviour. Limiting states of stress described by means of the friction angle (), the cohesion (c), and the dilatancy angle (). In addition, the increase in soil stiffness with pressure is accounted for in all three stiffness used, i.e., the triaxial loading stiffness E50, the triaxial unloading/reloading stiffness Eur and the oedometer loading stiffness Eoed (Dickin and Laman, 2007). The model strip footing was modeled as elastic beam elements based on Mindlins theory with flexural rigidity of EI=163 kNm/m and normal stiffness of EA=3.4 105 kN/m. The analyses were carried out using a plane strain model in sand with three different densities as in the tests. During the generation of the mesh, 15-node triangular elements were selected in preference to the alternative 6-noded versions in order to provide greater accuracy in the determination of stresses. PLAXIS incorporates a fully automatic mesh generation procedure, in which the geometry is divided into elements of the basic element type, and compatible structural elements. In the finite element modeling, as the slope surface is not horizontal, the initial stress condition of the slope was established first by applying the gravity force due to soil. A prescribed footing load was then applied in increments accompanied by iterative analysis up to failure. PLAXIS generates full fixity at the base of the geometry and smooth conditions at the vertical sides. Values of soil parameters used in the numerical investigation are shown
Table 3. Values of Soil Parameters used in PLAXIS Analyses Parameter Unit weight, (kN/m3) Primary loading stiffness, E 50 (kN/m2) Initial stiffness, Eoed (kN/m2) Unloading/reloading stiffness, Eur (kN/m2) Cohesion, c (kN/m2) Friction angle, () Dilatancy angle, () Poissons ratio, Earth pressure coefficient at rest, K0
ref

16.5 20000 20000 60000 0.10 40.6 10.6 0.25 0.35

Value 17.0 28000 28000 84000 0.10 41.8 11.8 0.25 0.33

17.5 40000 40000 120000 0.10 43.5 13.5 0.25 0.31

3. Numerical Modeling
A series of two-dimensional Finite Element Analyses (FEA) on a prototype footing-slope system was carried out in order to validating the results of the laboratory model tests and providing insights into the deformation behavior within the soil mass. The finite element analysis was performed using the commercial program PLAXIS (Brinkgreeve and Vermeer, 2002). The geometry of the prototype footing-slope system was assumed to be the same as the laboratory model. The same angle of slope inclination (=20o, 25o, 30o) and the material of steel plate for footing and sand were used in the numerical study. An elasto-plastic hyperbolic model described as the Hardening Soil Model (HSM) was used from those available in PLAXIS to describe the non-linear sand behavior in this study. When subjected to primary deviatoric loading, cohesionless soil shows

Fig. 4. Prototype Slope Geometry, Finite Element Mesh, and Boundary Conditions
KSCE Journal of Civil Engineering

702

Model Studies of Bearing Capacity of Strip Footing on Sand Slope

in Table 3. Shear strength and stiffness parameters representing sand conditions derived from series of drained triaxial compression tests. The prototype slope geometry, finite element mesh, and the boundary conditions are shown in Fig. 4.

4. Results and Discussions


A total of 40 model tests were conducted on model strip footing on sand slope. The effect of the edge distance of the footing to the slope crest, the slope angle, the relative density of sand, and footing width on the ultimate bearing capacity was obtained and discussed. Numerical study was also carried out using the finite element model using PLAXIS to verify the model tests. The bearing capacity behavior of the footing on sand slope, is represented using a non-dimensional factor, called bearing capacity reduction factor, i. This factor is defined as the ratio of the ultimate bearing capacity of footing on slope (quslope) to the ultimate bearing capacity of footing on level ground (qu). quslope i = -----------qu (3)

Fig. 5. Variations of q with s/B for Model Tests with Different Ratios of b/B

The footing settlement (s) is also expressed in non-dimensional form in terms of the footing width, B as the ratio s/B (%). The ultimate bearing capacities for the model are determined from the load-displacement curves as the pronounced peaks, after which the footing collapses and the load decreases. In loaddisplacement curves which did not exhibit a definite failure point, choosing a single value of may be extremely subjective. In this study, 0.1B method, which takes the ultimate bearing capacity at a settlement of 10% of the footing width, was used. Although this method is completely arbitrary, it may actually be close to the average soil strain at failure, forces a fixed value at qu, and treats the displacement of all footing sizes the same (Cerato, 2005). 4.1 Effect of Footing Distance to the Slope Crest A series of tests was performed on strip footing resting on a sand slope, in order to investigate the effect of distance of the footing to the slope crest (b/B). Tests were conducted for b/B ratios of 0.0, 1.0, 2.0, 3.0, 4.0 and 5.0. Test was also conducted with the footing on a level surface (=0o) for the purpose of comparison. In these test series, the slope angle was =30o, the relative density of sand was Dr=65% and the footing width was B=70 mm. Load settlement curves for five different b/B ratios obtained from model tests and FE analyses are presented in Figs. 5 and 6. The results are also summarized in Table 4. The results indicate that, the ultimate bearing capacity increases with increasing setback distance of the footing from the slope crest and the finite element results provide a reasonable fit with the experimental results. When the footing is moved away from the slope crest (b/B=0) to the setback distance of b/B=1.0, there is a serious increase in bearing capacity (an average value of
Vol. 17, No. 4 / May 2013

Fig. 6. Variations of q with s/B for FE Analyses with Different Ratios of b/B Table 4. Results of Footing Located at Different Locations from the Slope Crest (=30o, Dr=65%, B=70 mm) b/B 0 1 2 3 4 5 Level ground qu (kN/m2) Test Analyses 31.50 36.27 56.70 58.64 72.86 74.27 90.27 89.88 109.11 107.12 124.40 123.83 130.40 134.74 i Test 0.24 0.43 0.56 0.69 0.84 0.95 1.00 Analyses 0.27 0.44 0.55 0.67 0.80 0.92 1.00

70%). However, the rate of increase in bearing capacity decreases with increasing distance of setback until b/B=5.0 where the ultimate bearing capacity of the footing on slope approaches that of a footing on level ground. The effect of slope is minimized when the footing is placed at an edge distance beyond five times width of the footing. This change in bearing capacity of the footing with its location relative to slope crest can be attributed to soil passive resistance from the slope side. When, the footing is placed far away of the slope, the passive resistance from the slope side to the failure wedge under the footing increases. Also,

703

M. Salih Keskin and Mustafa Laman

Fig. 7. Variations of i with b/B

as b/B increases a much greater force is required for the failure surface to reach the slope. Hence, the bearing capacity of the footing increases. The variations of bearing capacity reduction factor, i, at different edge distance of the footing width (b/B) for slope angle of = 30o obtained from model tests and FE analyses are shown in Fig. 7. The figure clearly shows that the general trend of FE analyses is agree fairly well with this of the model tests. The results show that the values of i increases almost linearly with an increase in setback distance. i values obtained from model tests and FE analysis for the strip footing located on the setback distance of 5.0B are 0.95 and 0.92, respectively. Hence, it is inferred that the ultimate bearing capacity for a footing at a setback distance larger than 5.0B may be close enough to that in the level ground case and there would be no benefit from moving the footing any further from the slope. 4.2 Effect of Slope Angle In addition to the series of model tests for footing located on slope with slope angle of = 30o, another series of model tests were carried out for footing on slope with slope angles of = 25o and 20o to examine the influence of the slope angle on the bearing capacity of the strip footing. For each slope angle, the footing was placed at five different setback distances such that the setback distance to footing width ratio b/B = 0.0, 1.0, 2.0, 3.0, 4.0 and 5.0. Dr = 65% and B = 70 mm were constant for all model tests. Table 5 summarizes the results of model tests and FE analysis. The variations of qu and i with slope angle for different setback distances obtained from load-displacement curves both experimentally and numerically are shown in Figs. 8 and 9, respectively. As seen from Figs. 8 and 9, the agreement between experimental and numerical results is reasonably well. Fig. 8 clearly shows that, the ultimate bearing capacity decreases with an increase in slope angle. The bearing capacity also increases with increasing setback distance and at setback distance of 5.0B; is not influenced by the slope gradient. A similar trend was observed for variations of i with slope angle for different setback

Fig. 8. Variations of qu with Slope Angle for Different Setback Distances

Fig. 9. Variations of i with Slope Angle for Different Setback Distances Table 5. Results of Footing on Slope with Different Slope Angles (Dr=65%, B=70 mm)

25

20

b/B 0 1 2 3 4 5 0 1 2 3 4 5

qu (kN/m2) Test Analyses 44.45 48.81 67.81 74.08 84.43 90.76 100.41 106.42 114.80 118.10 125.27 127.30 60.80 63.53 81.19 81.94 97.82 100.84 107.25 112.84 120.02 122.49 127.78 133.59

i Test 0.34 0.52 0.65 0.77 0.88 0.96 0.47 0.62 0.75 0.82 0.92 0.98 Analyses 0.36 0.55 0.67 0.79 0.88 0.94 0.47 0.61 0.75 0.84 0.91 0.99

distances as seen in Fig. 9. The value of i increases with decrease of slope angle and increase of setback distance.
KSCE Journal of Civil Engineering

704

Model Studies of Bearing Capacity of Strip Footing on Sand Slope

It is known that when a slope angle is increased, e.g. by excavation, it becomes less stable. This is evident in that the bearing capacity is significantly reduced as the slope angle increases. As the slope angle is increased the area of the slope is decreased therefore the failure zone is decreased resulting in a much smaller bearing capacity. To overcome this reduction Fig. 8 shows that while increasing the slope angle but moving the footing further from the slope a greater bearing capacity of footing is achieved. For example a 20 degree slope with b/B=0.0 has approximately the same bearing capacity as a 30 degree slope with b/B=1.0. Also, it can be concluded that the effect of setback distance on the bearing capacity of footing is regardless the value of slope angles ranged between 20o and 30o and at setback distance of b/B=5 the slope has no influence on the bearing capacity and the problem is considered as a level ground problem. 4.3 Effect of Relative Density of Sand In order to study the effect of relative density, in addition to the series of model tests for footing on sand slope with relative density of Dr = 65%, two series of tests were carried out on model footing located on different setback distances of sand slope with the densities of Dr = 45% and 85%. b = 30o and B = 70 mm were constant for all model tests of this series. The results are summarized in Table 6. The variations of qu and i with relative density of Dr for different setback distances are shown in Figs. 10, 11 and 12. Table 6 and Fig. 10 clearly show that the bearing capacity of the footing on a slope is significantly increases with an increase of relative density of sand and at setback distance of 5.0B the effect of slope is minimized. A good agreement was observed between the results of model tests and FE analysis. The data trends in Fig. 10 show that as relative density of sand increases the ultimate bearing capacity of the footing increases linearly. This increase in bearing capacity of footing with relative density of sand can be attributed with soil-footing interaction. As the

Fig. 10. Variations of qu with Relative Density for Different Setback Distances

relative density increases, the angle of friction of the sand increases and hence the adhesion and friction between soil and footing increases. This causes a larger failure surface and leading to greater bearing capacity. Figures 11 and 12 show the variations of i with relative density of sand at different setback distances of footing obtained from model tests and FE analysis, respectively. The results of model tests and FE analyses show that the values of bearing capacity reduction factor, i increases almost linearly with increase of setback distance ratio of b/B. However, the relative density of sand has not considerable effect on the behavior of i. As mentioned before the bearing capacity factor, i is the ratio of the ultimate bearing capacity of footing on slope to the ultimate bearing capacity of footing on level ground. When the relative density of sand increases the ultimate bearing capacity of footing on slope and level ground increase almost equally. Hence, there are slight differences between i values and it can be concluded that the relative density of sand has negligible effect on the

Table 6. Results of Footing on Slope with Different Relative Densities of Sand (=30o, B=70 mm) Dr b/B 0 1 2 3 4 5 0 1 2 3 4 5 qu (kN/m2) Test Analyses 27.61 26.01 41.22 43.70 50.31 54.32 62.77 67.14 73.05 76.68 82.41 88.32 44.40 45.83 76.62 74.85 94.51 95.89 118.14 115.87 143.67 138.19 163.34 153.53 i Test 0.32 0.48 0.59 0.74 0.86 0.96 0.26 0.45 0.56 0.70 0.85 0.97 Analyses 0.27 0.46 0.57 0.70 0.80 0.92 0.28 0.46 0.59 0.71 0.85 0.94

45%

85%

Fig. 11. Variations of i with Relative Density for Different Setback Distances (Model Test)

Vol. 17, No. 4 / May 2013

705

M. Salih Keskin and Mustafa Laman

Fig. 12. Variations of i with Relative Density for Different Setback Distances (FEA)

Fig. 13. Variations of qu with Footing width for Different Setback Distances

Table 7. Results of Footing on Slope with Different Footing width (=30o, Dr=65%) B b/B 0 1 2 3 4 5 0 1 2 3 4 5 qu (kN/m2) Test Analyses 15.54 29.20 40.02 47.33 51.27 58.71 69.00 72.02 80.50 84.12 89.88 102.42 31.50 36.27 56.70 58.64 72.86 74.27 90.27 89.88 109.11 107.12 124.40 123.83 i Test 0.17 0.43 0.55 0.73 0.86 0.96 0.24 0.43 0.56 0.69 0.84 0.95 Analyses 0.26 0.42 0.52 0.64 0.75 0.91 0.27 0.44 0.55 0.67 0.80 0.92

50 mm

70 mm

Fig. 14.Variations of i with Footing Width for Different Setback Distances (Model Test)

bearing capacity reduction factor, i. 4.4 Effect of Footing Width A series of test were carried out on a model footing (B = 50 mm) located on different setback distances of sand slope in order to investigate the effect of footing width on the bearing capacity. The other parameters were = 30o and Dr = 65% in model tests. Table 7 summarizes the results of model tests and FE analyses. The variations of qu and i with footing width of B at different setback distances are shown in Figs. 13, 14 and 15. As seen from Table 7 and Fig. 13, although the ultimate bearing capacity values obtained from numerical analyses do not fit completely with the experimental results and give greater bearing capacity, especially for footing of B = 50 mm, but the agreement is reasonably well. This discrepancy may be related to the model and foundation parameters chosen in numerical and experimental model. Based on the numerical and experimental results, it is clear that, as the footing width increases the ultimate bearing capacity of footing on a slope increases and effect of

Fig. 15. Variations of i with Footing Width for Different Setback Distances (FEA)

slope is minimized at setback distance of b/B = 5.0. This increase in bearing capacity can be attributed to the larger contact area between soil and footing. Larger displacements and vertical stresses occur below the larger footing. Therefore, the failure
KSCE Journal of Civil Engineering

706

Model Studies of Bearing Capacity of Strip Footing on Sand Slope

wedge becomes larger and the frictional resistance on failure planes becomes greater and bearing capacity of the footing increases. Figures 14 and 15 show the variations of i with footing width at different setback distances obtained from model tests and FE analysis, respectively. The results of model tests and FE analyses show that the values of bearing capacity reduction factor, i increases with increase of setback distance ratio of b/B. The behavior of i is similar to that behavior observed in test series on the effect of relative density of sand and seems to be footing width has not considerable effect on the bearing capacity reduction factor, i. 4.5 Parametric Study 4.5.1 Effect of Footing Width In order to evaluate the effect of footing width on the bearing capacity factor Nq, finite element analyses were performed on strip footings with widths of 0.05, 0.06, 0.07, 0.08, 0.09, 0.10, 0.11, 0.12 and 0.13 m. The strip footings were located on two different distances from the edge of the slope (b/B=0 and 1) having slope angle of =30 and relative density of Dr=65%. The ultimate bearing capacity, qu and bearing capacity factor, Nq values obtained from the analyses are summarized in Figs. 16 and 17, respectively. A significant effect of footing width may be seen in the results. The qu values increase and Nq values decrease with an increase in B. Also, it is clear that qu and Nq values increase with an increase in setback distance ratio of b/B. Fig. 16 shows that the value of qu can be expressed as a linear function of B, as follow: qu = x1 + x2 B (4)

Fig. 17. Variation of Nq with B

506.33 for b/B=1. Fig. 17 shows that curve-fitting using Eq. (5) generates good values of correlation coefficient. Nq = x1 + x2 B + x3 B
2

(5)

in which, x1=14.507, x2=310.55 for b/B=0 and x1=22.09, x2=

in which, x1=96.631, x2=-664.34, x3=2334.6 for b/B=0 and x1= 179.16, x2=-1662.1, x3=7048 for b/B=1. For the evaluation of bearing capacity factor (Nq) limited footing widths were used in the analysis. This part of study aims to only show the effect of the footing width on the bearing capacity factor Nq. Therefore, the derived equations (Eqs. 4 and 5) are valid within the content of current study. These equations should be considered as designating the relation between footing width and Nq. Hence, when these equations compared with the other studies larger values up to 2 times are obtained. 4.5.2 Effect of Footing Rigidity In this study, the strip footing was used and the footing was considered as a flexible body. A series of finite element analysis were performed considering the footing as a rigid body. One way of rigid body modeling is to assign prescribed uniform displacements to the footing which is the case for the rigid body behavior. In the analysis, the same geometry was considered and the settlement of the footing is simulated by means of a uniform

Fig. 16. Variation of qu with B


Vol. 17, No. 4 / May 2013 707

Fig. 18. Rigid Model

M. Salih Keskin and Mustafa Laman

Fig. 21. Mesh and Boundary Conditions for 3D Model

Fig. 19. Variations of qu with b/B

Fig. 22. Variations of qu with b/B

Fig. 20. Variations of i with b/B

indentation at the top of the sand layer instead of modeling the footing itself as seen in Fig. 18. The prescribed displacement was taken as s = 0.1B = 0.0070 m. In the analysis the slope angle was = 30, the relative density was Dr = 45%, the footing width was B = 0.070 m, and the distance of the footing to the slope crest were varied from b = 0 to 5B. Figures 19 and 20 show the variations between qu-b/B and i-b/ B values, respectively. When comparing the qu results obtained by flexible and rigid assumptions, it can be noticed that for the same displacement the rigid footing, exhibited more bearing capacity (about 8-10%) than that from flexible footing. 4.5.3 Effect of 3D Modeling A series of 3D finite element analysis were performed using PLAXIS 3D to take care of 3D effects on the test results. In the analysis the same geometry of the experimental set-up was used. b/B ratios were 0, 1, 2, 3, 4 and 5, footing width was B=70 mm, slope angle was =30 and the relative density of sand was Dr = 65%. The used mesh and boundary conditions for 3D model is seen in Fig. 21. Figs. 22 and 23 show the variations between the qu and i values obtained from experimental, 2D and 3D analysis with b/B ratios, respectively. As seen from Fig. 22, the qu values obtained from 3D analysis

Fig. 23. Variations of i with b/B

are somewhat greater (about 2%) than the values obtained from 2D plane strain analysis. However, the i values obtained from 3D and 2D analysis which shown in Fig. 23 are very close.

5. Failure Mechanism
For soil slopes, the results of previous studies showed that the failure model could be categorized as perfect plastic failure with a circular slip surface. In common, the failure region was divided into active zone, passive zone, and radial shear zone; however, the radial shear zone was in global shear failure and bounded by a logarithmic spiral line. This section is devoted to discuss the
KSCE Journal of Civil Engineering

708

Model Studies of Bearing Capacity of Strip Footing on Sand Slope

Fig. 24. Output for = 20, b/B = 0: (a) Deformed Mesh, (b) Total Displacements, (c) Total Displacement Contours, (d) Shear Strains

Fig. 25. Output for = 25, b/B = 0: (a) Deformed Mesh, (b) Total Displacements, (c) Total Displacement Contours, (d) Shear Strains

failure mechanism observed in experimental and numerical studies of bearing capacity of strip footings adjacent to the crest of a sand slope. The form of failure plane below the footing can illustrate the influence of the slope angle and the setback distance from the slope crest on the bearing capacity. In order to clearly demonstrate the differences in the failure plane due to effect of slope angle , and b/B ratio, the case of soil with = 20, 25 and 30 with different b/B ratios of 0, 1, 2, 3, 4 and 5 are chosen for comparison. The failure mechanisms obtained from numerical analyses for different slope angles are shown in Figs. 24 through 26. Fig. 27(a) and 27(b) show the picture of the model slope at the peak footing load for = 30 with setback distance ratio of b/B = 0 and 1, respectively. At the ultimate stage, the bearing pressure reaches the peak load, and begins to decrease suddenly with increasing settlement. For all slope angles, the failure initialing from the left edge of the footing propagated into the slope surface and resulted in the total loss of passive resistance. As a result, the footing collapses and the soil deform laterally toward the sloping side. Once a fully connected failure surface was created, the soil begins to push toward the sloping side. As mentioned before, from the comparison of the results of model tests and numerical analyses, the ultimate bearing capacity consistently decreases as the slope angle increases. Unlike the level ground case, the failure mode was not symmetrical and only developed in the sloping side. According to figures, it appears that the failure process for the 20, 25 and 30 slopes are very similar. A starting at the left edge of the footing propagated with increasing load. It can be noted that if the footing is located close to the edge of the slope, the soil below the footing tends to move toward the slope, since it has less shear resistance. The form of failure surfaces are affected by the setback distance from the slope crest. With increasing setback distance, the failure zone is larger and will provide a higher bearing capacity. In Fig. 28(a), the

Fig. 26. Output for = 30, b/B = 0: (a) Deformed Mesh, (b) Total Displacements, (c) Total Displacement Contours, (d) Shear Strains

Fig. 27. Pictures of the Model Sand at the Peak Footing Load: (a) = 30o, b/B = 1, (b) = 30o, b/B=0
Vol. 17, No. 4 / May 2013

Fig. 28. Contours of Shear Strain ( = 30): (a) b/B = 0, (b) b/B = 1, (c) b/B = 2, (d) b/B = 3, (e) b/B = 4, (f) b/B = 5

709

M. Salih Keskin and Mustafa Laman

Fig. 29. Geometry of Model used in LE Analysis

Fig. 30. Slip Surface Obtained from LE Analysis

influence area concentrates mainly on the side of the slope. When the footing is located away from the slope crest, the bearing capacity of the footing increases as expected. It can be said that, the degree of confinement on the side of the slope increases and part of the stress due to the footing begins being governed by soil on the side of the ground surface. According to the Figs. 28(a-f), the displacement of the soil occurs on the side of ground surface and the influence of the slope is vanishing by locating the footing further away from the slope crest. At the ratio of b/B = 5.0, the failure plane becomes symmetric, which illustrates that the stress spread on both side of the footing. In order to analyze the failure mechanism based on Limit Equilibrium (LE) method a series of analyses was performed. The limit equilibrium analysis was carried out using the commercial program GeoStudio (GeoStudio, 2012. Geo-Slope Int. Ltd.). The geometry of the model was taken to be the same as the laboratory model (Fig. 29). Using limit equilibrium, GeoStudio can model heterogeneous soil types, complex slip surface geometry, and variable pore-water pressure conditions. The slip surface obtained from limit equilibrium analysis is shown in Fig. 30. It can be seen from Fig. 30 that a circular surface propagates towards the slope until the sloping ground is reached. It should also be noted that when the factor of safety was equal to 1 the ultimate load was about 60 kN/m2 which is reasonable close to the ultimate bearing capacity value of 56.70 kN/m2 obtained from experimental study.

6. Conclusions
A series of model tests and numerical analyses has been carried

out to investigate the bearing capacity of a strip footing resting on sand slopes. The study primarily aimed at determining the effect of setback distance of the footing to the slope crest and its behavior with various s slope angles, relative density of sand and width of the footing. Based on the experimental and numerical studies, the following main conclusions are made: 1. The results show that the bearing capacity and bearing capacity reduction factor i, increase almost linearly with an increase in setback distance up to setback distance to footing width ratio of b/B = 5.0. Beyond this value, the ultimate bearing capacity remains constant like that of a footing located on a level ground. 2. The bearing capacity of strip footing on sand slope is significantly dependent on the slope angle, relative density of sand, and width of the footing. 3. The results clearly show that as the slope angle increases or slope becomes steeper the ultimate bearing capacity of footing decreases. Moving the footing further from the slope overcomes this reduction due to the increase in slope angle. 4. The bearing capacity of the footing on a slope is significantly increases with an increase of relative density of sand. However, the relative density of sand has not considerable effect on the behavior of i. 5. Based on the numerical and experimental results, as the footing width increases the ultimate bearing capacity increases and bearing capacity factor decreases. Furthermore, it seems to be footing width has a negligible effect on the bearing capacity reduction factor, i. 6. The results show that the value of i characterizing the linear relationship between the effect of setback distance of the footing to footing width ratio (b/B) and the slope angle (). It can be expressed as functions of b/B and , independent from relative density of sand (Dr) and footing width (B). 7. From series of numerical studies (using finite element analyses) with the variation in the footing size, empirical equations have been derived between the bearing capacity characteristics (qu and Nq) and footing width for the strip footings located on two different distances from the edge of the slope (b/B=0 and 1). A significant effect has been seen in the results. The qu values increase linearly and Nq values decrease parabolically with an increase in B. Also, the values of qu and Nq increase with an increase in setback distance ratio of b/B. 8. The results of the parametric study show that for the same displacement the rigid footing, exhibited more bearing capacity than that from flexible footing. Also, 3D analysis gives somewhat greater bearing capacity values than the values obtained from 2D plane strain analysis. 9. A close agreement between the experimental and numerical results on general trend of behavior is observed. However, the ultimate bearing capacity values obtained from FEA appears to be greater than that obtained from the model tests.
KSCE Journal of Civil Engineering

710

Model Studies of Bearing Capacity of Strip Footing on Sand Slope

Acknowledgements
The authors thank the Scientific Research Project Directorate of Cukurova University for supporting this study (Project no: MMF2006D1).

References
Cerato, A. B. (2005). Scale effect of shallow foundation bearing capacity on granular materials, PhD Dissertation, Univ. of Massachusetts, Amherst. De Buhan, P. and Gaernier, D. (1988). Three dimensional bearing capacity analysis of a foundation near a slope. Soils Found., Vol. 38, No. 3, pp. 153-163. Dickin, E. A. and Laman, M. (2007). Uplift response of strip anchor in cohesionless soil. Adv. Eng. Softw., Vol. 38, Nos. 8-9, pp. 618-625. Duncan, M. and Chang, C. Y. (1970). Nonlinear analysis of stress and strain in soils. J. Soil Mech. Found. Div., ASCE, Vol. 96, No. 5, pp. 1629-1653. Garnier, J., Canepa, Y., Corte, J. F., and Bakir, N. E. (1994). Etude dela portance de foundations en bord de talus. Proceedings of the 13th International Conference on Soil Mechanics and Foundation Engineering, New Delhi, India, pp. 705-708. Gemperline, M. C. (1988). Centrifuge modeling of shallow foundations. In Proc., ASCE Spring Convention, pp. 45-70. GeoStudio, (2012). Geo-Slope Int. Ltd. Graham, J., Andrews, M., and Shields, D. H. (1988). Stress characteristics for shallow footings in cohesionless slopes. Can. Geotech. J., Vol.

25, No. 2, pp. 238-249. Hansen, J. B. (1970). A revised and extended formula for bearing capacity. Danish Geotech. Inst., Bulletin 28, pp. 5-11. Meyerhof, G. G. (1957). The ultimate bearing capacity of foundations on slopes. In Proc., IV Int. Conf. Soil Mech. Found. Eng., London, England, Vol. 1, 384-387. Meyerhof, G. G. (1963). Some recent research on the bearing capacity of foundations. Can. Geotech. J., Vol. 1, No. 1, pp. 16-26. Narita, K. and Yamaguchi, H. (1990). Bearing capacity analysis of foundations on slopes by use of log-spiral sliding surfaces. Soils Found., Vol. 30, No. 3, pp. 144-152. Plaxis (2002). User manual, 2D version8 (Edited by Brinkgreeve, R.J.B.), Delft University of Technology & PLAXIS b.v., The Netherlands. Saran, S., Sud, V. K., and Handa, S. C. (1989). Bearing capacity of footings adjacent to slopes. J. Geotech. Eng., ASCE, Vol. 115, No. 4, pp. 553-573. Shields, D., Chandler, N., and Garnier, J. (1990). Bearing capacity of foundations in slopes. J. Geotech. Eng., ASCE, Vol. 116, No. 3, pp. 528-537. Shields, D. H., Scott, J. D., Bauer, G. E., Deschemes, J. H., and Barsvary, A. K. (1977). Bearing capacity of foundations near slopes. Proceedings of the 9th International Conference on Soil Mechanics and Foundation Engineering, Vol. 1, pp. 715-720. Terzaghi, K. (1943). Theoretical soil mechanics, John Wiley & Sons, New York. Vesic, A. S. (1975). Bearing capacity of shallow foundations. In Foundation Engineering Hand-Book, Winterkorn, H.F., and Fang, H.Y., Eds., Van Nostrant Reinhold Co., pp. 121-147. Wood, D. M. (2004). Geotechnical modelling, Spon Press.

Vol. 17, No. 4 / May 2013

711

Você também pode gostar