- Email: [email protected]
Two dimensional experimental study for the behaviour of surface footings on unreinforced and reinforced sand beds overlying soft pockets
Geotextiles and Geomembranes 28 (2010) 589e596
Contents lists available at ScienceDirect
Geotextiles and Geomembranes journal homepage: www.elsevier.com/locate/geotexmem
Technical Note
Two dimensional experimental study for the behaviour of surface footings on unreinforced and reinforced sand beds overlying soft pockets M.H.A. Mohamed* School of Engineering, Design and Technology, University of Bradford, Bradford, West Yorkshire BD7 1DP, UK
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 October 2009 Received in revised form 25 May 2010 Accepted 30 May 2010 Available online 26 June 2010
This paper presents results of a comprehensive investigation undertaken to quantify the efficiency of using reinforcement layers in order to enhance the bearing capacity of soils that are characterised by the existence of localised soft pockets. Small-scale model experiments using two dimensional tank were conducted with beds created from well graded sand with mean particle size of 300 mm but prepared with different dry densities. A relatively softer material was embedded at predetermined locations within the sand beds so as to represent localised soft pockets. Various arrangements of soil reinforcement were tested and compared against comparable tests but without reinforcement. In total 42 tests were carried out in order to study the effect of the width and depth of the soft pocket, the depth of one reinforcing layer and the length and number of reinforcing layers on the soil bearing capacity. The results show clearly that the ultimate bearing capacity reduces by up to 70% due to the presence of a soft pocket. It was also noted that the proximity of the soft pocket also influenced the bearing capacity. Reinforcing the soil with a single layer or increasing the length of reinforcement is not as effective as was anticipated based on previous studies. However, bearing capacity increased significantly (up to 4 times) to that of unreinforced sand when four layers of reinforcement were embedded. The results suggest that rupture of the bottom reinforcement layer is imminent in heavily reinforced sand beds overlying soft pockets and therefore its tensile strength is critical for successful reinforcement. Ó 2010 Elsevier Ltd. All rights reserved.
Keywords: Bearing capacity Ground variability Soft pockets Soil reinforcement
1. Introduction and background The bearing capacity of weak soils is often inadequate to support imposed building loads. To minimise induced stresses within the weak soil to an acceptable level, it is common practice to reinforce the surface soil stratum with layers of reinforcement and use granular material to fill in between reinforcing layers. Reinforcement of weak grounds is considered as an effective and relatively economical approach for enhancing bearing capacity and reducing associated settlement. Several experimental, analytical and numerical research studies have been undertaken to investigate the impact of various configurations of reinforcement on the overall behaviour of reinforced soils under surface loads (e.g., Yetimoglu et al., 1994; Huang and Meng, 1997; Adams and Colin, 1997; Alawaji, 2001; Yamamoto and Otani, 2002; Patra et al., 2006; Basudhar et al., 2007; El Sawwaf, 2007, Ghazavi and Lavasan, 2008, Zhou and Wen, 2008, Chen et al., 2009, Sharma et al., 2009; Latha and Somwanshi, 2009). The experimentally based
* Tel.: þ44 1274 233856; fax: þ44 1274 234111. E-mail address: [email protected] 0266-1144/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.geotexmem.2010.06.001
studies were undertaken for various footing shapes (e.g. strip, rectangular, square and circular) and using different types of reinforcement (e.g. geogrid, geotextile and metal strips, sheets or rods). This variation has led to different recommendations for the optimum configuration of the reinforcement layers. In addition, to enhance the shearing resistance between reinforcing layers and surrounding soil, it was found to be beneficial to use a frictional material to fill in between reinforcing layers (e.g. Saleh, 2001), particularly coarse sand (Phanikumar et al., 2009). Shearing resistance was also enhanced by using geogrid layers (e.g., Guido et al., 1986; Yetimoglu et al., 1994; Patra et al., 2005). Improving the shearing resistance might require less anchorage length and hence reduce the overall cost. Jewel et al. (1984) pointed out that the mechanism for mobilization of friction resistance in geogrid-reinforced soils is different from that of geotextile based. The failure mechanisms and deformation patterns of reinforced soil were also subject to several investigations. Binquet and Lee (1975) identified three possible failure mechanisms which might occur in reinforced soils as a function of the reinforcement arrangement and strength of layers. Those were failure above the reinforced zone and failure due to upper ties break and ties pull out. Various deformation patterns of reinforced sand beds by a single
590
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
layer at failure were studied by Michalowski and Shi (2003). They suggested that some alterations occur to the failure mechanism due to the existence of the reinforcing layer. It was also noted that the deformation pattern of reinforced sand maintains many of the failure characteristics of unreinforced sand when one long layer is used. Haung and Tatsouka (1990) concluded that the length of reinforcing layers is a major factor affecting the deformation pattern of reinforced sand beds. They clearly distinguished between a “deep footing effect” which occurs as a result of short reinforcement and “wide slab effect” which is associated with the use of long high-strength reinforcing layers. Recently Sharma et al. (2009) investigated the occurrence of a failure mechanism within the reinforced zone and developed an analytical model to estimate the soil ultimate bearing capacity. The results of the proposed model were in good agreement with those obtained from physical testing. Most of the previous studies investigated the performance of reinforced sand, reinforced cohesive soil or reinforced sand overlying a clay layer of uniform thickness. However, less attention has been given for more practical soils in which local changes in the ground conditions occur e.g. due to the inclusion of localised soft pockets and/voids. They occur as a result of natural and/or manmade activities and are often encountered in construction sites. For example, poor supervision and implementation of engineering works would result in poorly filled voids and trenches (BRE, 2004) as well as a significant variation in the ground strength in abandoned landfill sites. Furthermore, soft pockets/zones can also be encountered in many virgin soils in the form of e.g. soft clay plugs within sandy meander belts and in tropical areas due to leaching and deposition of fine clay particles by the infiltration of water (Prothero and Schwab, 2004). The size and location of a soft pocket can vary widely and its existence can cause a significant ground distress and might result in intolerable ground movement if appropriate remedial measures are not undertaken. Despite the early efforts by Binquet and Lee (1975) to understand the behaviour of reinforced soils overlying a soft pocket, there is a lack of detailed investigations to quantify the effect of the local variability of ground conditions. In particular, the influence of localised soft pockets on the behaviour of subsurface soils under foundation loadings is poorly understood and has hindered the development of a comprehensive analytical model. Unlike soft pockets, studies for reinforced ground overlying voids received more consideration (see for example, Giroud et al., 1990; Poorooshasb, 1991; Das and Khing, 1994; Wang et al., 1996; Alexiew, 1998; Sireesh et al., 2009). This paper therefore aims to investigate the effects of; (i) width and depth of a localised soft pocket on the ultimate bearing capacity of unreinforced sand beds (ii) reinforcement configuration including; the depth of a single reinforcement layer, number of reinforcing layers, and length of reinforcement on the load carrying capacity of the reinforced system. These factors will be investigated at two different states of soil packing so as to study the influence of soil density and hence soil strength on the behaviour of the reinforced system. In addition, the failure mechanisms of unreinforced and reinforced soils above localised soft pocket will be described and discussed. 2. Laboratory techniques and materials An experimental set-up utilising a fully automated Instron loading machine that is controlled using computer software was developed to study the behaviour of unreinforced and reinforced sand beds with the inclusion of soft pockets. A tank with length of 500 mm, height of 500 mm and width of 200 mm was manufactured for this study. The front side of the tank was made of 20 mm thick plexiglass to enable visual inspection and the remaining sides
were made of wood. It should be noted that the tank was also braced to eliminate its deformation at high values of footing pressure. A rough rigid model footing with a width of 100 mm and covering the whole width of the tank was used to simulate plane strain conditions. The applied load and settlement of the footing were measured and recorded electronically through the data acquisition system every 1 s. Dry silica sand with a narrow range of particle sizes between 600 mm and 75 mm was used. It was found that the median diameter of the sand is 300 mm, D10 ¼ 180 mm, D30 ¼ 240 mm and D60 ¼ 330 mm. The uniformity and curvature coefficients are determined and found to be 1.18 and 1.0 respectively. According to the British Standards BS5930, the sand used can be classified as poorly graded sand. In order to enable preparation of sand beds with a range of densities, a sand deposition technique has been developed, calibrated and utilised in this investigation. The sand deposition technique has been successfully used in previous studies (e.g., Gottardi et al., 1999) and proved to construct sand samples of uniform density. The experiments were carried out at two different densities which are 1540 kg/m3 and 1660 kg/m3 so as to represent two practical soil states (loose and dense respectively). Hereafter sand beds will be referred to as either loose or dense sand. A large shear box with internal dimensions of 300 mm 300 mm was used for determination of the angle shearing resistance for loose and dense samples. The angle of shearing resistance was found to be 30.5 and 35.4 for loose and dense samples respectively. Several materials have been tested under unconfined compression in order to determine if it would make an appropriate soft material. Arma-foam sound 240 with Young’s modulus of 124 kPa was used as a soft pocket. The Arma-foam sound 240 is a porous material manufactured by controlled extrusion (Khan et al., 2005). It should be noted that no lateral deformation was observed during the loading and therefore the measured vertical deformation is directly related to the applied vertical compressive stress. Various types of reinforcement with a range of strengths have been used in previous studies. In this investigation a decision was taken to use weak reinforcement so as to potential modes of failure can be observed. Sheets of Guilt paper which has a relatively low tensile strength are available. The tensile strength of the proposed reinforcing materials was determined experimentally and found to be 2.47 N/m2 at an extensible stain of 3%. In addition the ultimate tensile strength at failure is 3.5 N/m2. The angles of friction between Guilt paper and loose sand and between Guilt paper and dense sand were 19.6 and 26 respectively. 3. Testing programme and procedure Fig. 1 illustrates all parameters that have been considered in this investigation. An experimental programme was designed to test the influence of these parameters. Table 1 presents the values of the parameters used in the 40 experimental tests. Initial tests were carried out on unreinforced sand beds without the inclusion of soft pockets to determine the bearing pressure-settlement behaviour of loose and dense sands. Then two batches of testing were carried out to investigate the impact of localised soft pocket width and depth on the behaviour of the unreinforced loose and dense sand beds under surface loading. Subsequently a number of experiments were conducted to investigate each parameter as stated in Table 1 and depicted in Fig. 1. Sand beds were prepared using the sand raining technique. The process of pouring sand was continued until the predetermined level of the soft pocket was reached, then the sand surface was carefully levelled off and the softer material was placed in accordance with the desired test conditions. The sand deposition process was resumed until the sand level reached the level of the bottom
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
591
Fig. 1. Illustration of the problem and parameters considered.
reinforcing layer. The reinforcing sheet was placed horizontally and symmetrically around the centreline of the soft pocket. The process was repeated for all layers of reinforcement. Finally, the sand surface was levelled off and a model footing was placed on the surface ensuring that the model footing is located symmetrically around the vertical axis of the reinforcing layers and the localised soft pocket. The model footing was then loaded in such way so that a settlement rate of 1 mm/min was obtained in all tests. Load was measured with an accuracy of 1 105 N. The tests were stopped either automatically when a failure occurred which is indicated by severe loss in bearing resistance or manually when a 30 mm of vertical settlement was reached. 4. Results and discussion It should be noted that all measured values for the vertical settlement (S) are presented as a function of the footing width (B). In addition a characteristic parameter “bearing pressure ratio, (BPR)” is introduced and defined as:
BPR ¼
tests (Batch 0) on sand beds without the inclusion of a soft pocket indicated that classical load-settlement relations were obtained and the ultimate bearing capacity was found to be 57 kPa and 114 kPa for loose and dense sand beds respectively at S/B of 10%. 4.1. Batch I: effect of a localised soft pocket width In total eight experiments were undertaken to quantify the effect of width of a localised soft pocket on the bearing capacity of unreinforced loose and dense sand beds as illustrated in Table 1 (series B and C). Fig. 2a and b shows the results of the bearing pressure against settlement ratio for different soft pocket widths. Also shown are the results of the load-settlement relations for sand beds without the inclusion of soft pockets. The results clearly demonstrate that the load carrying capacity of a sand bed reduces significantly with the increase in the width of a soft pocket irrespective of the state of the soil packing. In other words, applying a similar level of bearing pressure would result in unacceptable foundation settlement if the soft pocket was not identified and
qu ðunreinforced=reinforced systems with inclusion of a soft pocketÞ qu ðunreinforced sand without inclusion of a soft pocketÞ
where qu is the ultimate bearing capacity (kN/m2) of the unreinforced and reinforced sand beds with and without the inclusion of soft pockets. The characteristic parameter (BPR) reflects the loss in the value of the bearing capacity due to the existence of a soft pocket as well as taking into account the benefit of adding reinforcement layers. In this paper all values for the ultimate bearing pressure values were determined at a vertical settlement of 10% of the footing width (S/B ¼ 10%). This settlement value was selected because (i) it would indicate practically a failure due to excessive settlement and (ii) careful inspection of the laboratory data clearly indicates that failure occurred in almost all the experiments at S=B 10%. It should be noted that the observed increase in resistance at large settlement is impractical and due to (i) the build up of overburden pressure above the foundation level due to bulging of sand and (ii) an increased stiffness of the soft pocket. Results of loading
(1)
considered. The results suggested that the local shear failure experienced with the loose sand deposits remains a characteristic feature for loose sand beds with a localised soft pocket. Nevertheless the failure pattern in dense sand beds containing a localised soft pocket showed a direct dependence on the width of the soft pocket (see, Fig. 2b). For a relatively small soft pocket, local shear failure would occur whereas as when the width of the soft pocket increases it is more likely that a punching shear failure will be dominant. Localised soft pockets beneath the footing intercept the failure zone in the sand bed and cause significant transformation in stresses and hence the observed failure pattern. With further increase in the footing load, soil arch fails and subsequently a significant portion of the stresses is transferred to the soft pocket leading to a substantial settlement and punching shear failure. To aid the discussion, the BPR is determined using Equation (1) and
592
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
Table 1 Testing programme. Series
Variable parameter
0 I
A B
H
Density Bs ¼ 25, 50, 75 and 100 mm Bs ¼ 25, 50, 75 and 100 mm Ds ¼ 50, 100, 150 and 250 mm Ds ¼ 50, 100, 150 and 250 mm u ¼ 25, 50, 75, 60 and 100 mm u ¼ 25, 50, 75, 60 and 100 mm N ¼ 1, 2, 3 and 4
I
N ¼ 1, 2, 3 and 4
J
br ¼ 200, 250 and 300 mm br ¼ 200, 250 and 300 mm
C II
D E
III
F G
IV
V
K
Fixed parameters Hs ¼ 50 mm, Ds ¼ 100 mm, Loose Hs ¼ 50 mm, Ds ¼ 100 mm, Dense Hs ¼ 50 mm, Bs ¼ 50 mm, Loose Hs ¼ 50 mm, Bs ¼ 50 mm, Dense br ¼ 200 mm, Loose br ¼ 200 mm, Dense br ¼ 200 mm, u ¼ h ¼ 25 mm, Loose br ¼ 200 mm, u ¼ h ¼ 25 mm, Dense N ¼ 2, h ¼ 25 mm, u ¼ 25 mm, Loose N ¼ 2, h ¼ 25 mm, u ¼ 25 mm, Dense
plotted in Fig. 3 as a function of the soft pocket width (Bs) for loose and dense sands. It can be seen that there is a gradual reduction in the load carrying capacity as the width of soft pocket increases. The results also show that the effect of the soft pocket is more pronounced in dense sand beds. There is a loss of 70% of the load carrying capacity of dense sand when a localised soft pocket with a width of 0.5B (50 mm) exists at a depth of B (100 mm).
a 20
40
60
80
100
120
140
160
4.2. Batch II: effect of the localised soft pocket depth In this batch, the depth of a localised soft pocket was varied in accordance with the test conditions stated in Table 1 (series D and E). Data generated from eight tests for loose and dense sand beds with a soft pocket at different depths are presented in Fig. 4a and b. It can be seen that there is a direct loss in the load carrying capacity as the soft pocket gets closer to the foundation in the loose sand beds. In other words, the deeper the soft pocket the lower its influence on the measured values of the bearing pressure and
a
Bearing pressure, kPa 0
Fig. 3. Variation of BPR as a function of the soft pocket width.
180
0
0
Bearing pressure, kPa 0
20
40
60
80
100
120
Bs = 0.50 B (50 mm)
10
S /B, %
Bs = 1.00 B (100 mm) 15
15
Loose sand - no soft pocket
20
20 B
25
10
Ds = 2.50 B (250 mm) Loose sand - no soft pocket
15
15
20
20 B Ds = variable
25
Hs = 0.5B
30
10
25
Ds = 1.0B
5
Ds = 1.00 B (100 mm) Ds = 1.50 B (150 mm)
S/B, %
10
180 0
5
Bs = 0.75 B (75 mm)
160
Ds = 0.50 B (50 mm)
5
Settlement , mm
5
140
0
Bs = 0.25 B (25 mm)
30
25
Hs = 0.5B 30
30
Bs = variable
Bs = 0.5B
35
35 35
35
b
Set t lement , mm
Batch no
Bearing pressure, kPa 0
20
40
60
80
100
120
140
160
b
180
0
0 Bs = 0.25 B (25 mm) Bs = 0.50 B (50 mm)
5
5
Bearing pressure, kPa 0
20
40
60
80
100
120
140
160
180
0
0
5
5
10
10
Dense sand - no soft pocket
15
20
25
10
15
20 B Ds = 1.0B
25
S/B, %
Bs = 1.00 B (100 mm)
Settlement , mm
S /B, %
10
B
15
15
Ds = variable 20 Ds = 0.50 B (50 mm) 25
Ds = 1.00 B (100 mm)
20
Hs = 0.5B
Set t lement, mm
Bs = 0.75 B (75 mm)
Bs = 0.5B 25
Ds = 1.50 B (150 mm)
30
30
Hs = 0.5B
30
Bs = variable 35
Ds = 2.50 B (250 mm)
30
Dense sand - no soft pocket
35
Fig. 2. Variation of bearing pressure with S/B for different widths of soft pocket (a) unreinforced loose sand, (b) unreinforced dense sand.
35
35
Fig. 4. Variation of bearing pressure with S/B for different soft pocket depths (a) unreinforced loose sand, (b) unreinforced dense sand.
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
associated settlement. Values of the BPR determined at 10% S/B as a function of the soft pocket depth (Ds) suggest that the bearing capacity reduces by 30% when a soft pocket of width (B) is inserted within a loose sand bed at a depth of (B). Local shear failure is observed in all tests involving loose sand beds independent of the depth of the soft pocket. The data presented indicate that soil arching is not a major factor governing the behaviour of loose sand beds since considerable settlement is observed shortly after applying a relatively small loading increment. The results indicated that there is a limited negative impact of the soft pocket and this occurs when it is placed at a depth equal to 2.5B (250 mm). Unlike the loose sand deposits, data for the dense sand deposits show larger reduction in the value of the bearing capacity due to the existence of a soft pocket. This highlights the influence of the initial relative stiffness between the soil mass and the softer zone. For shallow soft pockets that are within a depth of B (100 mm), a loss of w55% in the load carrying capacity of dense sand bed is obtained. In this case induced stresses underneath the footing are strongly affected by the presence of the weak pocket. Careful inspection of the data presented in Fig. 4b shows that for all cases of soft pocket, there is an abrupt loss in the soil resistance at some stage of the loading which could be explained by the failure of the soil arch. This means that a redistribution of the stresses occurs and helps in bridging over the weak pocket until a final failure point. Furthermore, the observed subsequent gain in the strength at large settlement is not practical and primarily because of the overburden pressure created by the significant settlement. In these tests, a punching shear failure occurs and is evident by the excessive settlement at a particular bearing pressure. Some anomalies are encountered in the load-settlement curves for soft pockets that are placed at a depth of 2.5B. These experiments were repeated couple of times and almost identical curves are obtained. The irregularity in the ultimate bearing capacity could be due to the initial compression of the soft pocket under the impact of sand raining and overburden pressure exerted afterwards. However, no clear explanation can be drawn out and further testing is underway. Combining results of Batch I and Batch II, it seems reasonable to conclude that the presence of soft pockets of width >0.25B (25 mm) and within a depth of 1.5B (150 mm) would strongly affect the ultimate bearing capacity. This highlights that serious considerations should be given to the effects of ground variability on the load carrying capacity of the whole soil mass. 4.3. Batch III: effect of depth of one reinforcement layer This batch of experiments was conducted using one sheet of reinforcement layer that is placed at different depths to determine precisely whether one reinforcement layer is sufficient to enhance the load carrying capacity of the sand beds with the inclusion of a soft pocket as well as to determine the optimal depth of a reinforcement layer. The test conditions of Batch III are illustrated in series F and G in Table 1. Fig. 5a and b shows the variation of the bearing pressure with settlement ratio (S/B) for sand reinforced with one layer and with the inclusion of a soft pocket. Also shown are the curves for unreinforced sand beds with and without the inclusion of soft pocket. The results obtained for loose sand beds demonstrate that generally a marginal improvement is achieved by the use of one reinforcement layer irrespective of its depth. In addition, a slightly higher bearing capacity can be obtained when the layer of reinforcement is placed at a depth of 0.75B (75 mm). In comparison, the dense sand beds showed considerable improvement by the use of one layer of reinforcement as shown in Fig. 5b. The data illustrate that generally for shallow reinforcements of <0.5B (50 mm) as well as for deep reinforcement at 1B (100 mm), a slight improvement could be achieved. In addition, deep
593
Fig. 5. Variation of bearing pressure with S/B for different u (a) reinforced loose sand, (b) reinforced dense sand.
reinforcement results in better improvements if a settlement ratio of >7% can be tolerated. However, best enhancement is obtained when the reinforcement layer is placed at an intermediate level e.g. 0.6e0.75 B (60e75 mm) which improves the ultimate bearing capacity by almost 100% to that of the unreinforced sand bed but it does not still recover the full strength of the dense sand without the soft pocket. This gives an indication that maximum shear stress would develop at that depth so that the reinforcing layer will take up a significant portion of the shear stress causing further confinement and hence leading to an increased resistance to failure. Upon the completion of each test, sand was carefully excavated and the shape of the deformed reinforcement sheet was recorded by imaging. Fig. 6 shows a presentation for the deformed shapes against the depth of one layer of reinforcement. Fig. 6 indicates that the overall deformation is related to the location of the reinforcement layer with respect to the foundation and the soft pocket. A bowl shape is observed for reinforcement layers at shallow depths. Furthermore the shape is flattened out at deeper levels and even might be inverted to those previously observed when a reinforcement layer is directly placed on top of the soft pocket. The latter suggests that movement of the sand along the sides of the soft pocket was inward towards the soft pocket which is opposite to the well reported deformation pattern for reinforced sand without the inclusion of soft pockets (Binquet and Lee, 1975; Michalowski and Shi, 2003). In addition, reinforcement layers at intermediate levels deformed differently. Fig. 6 illustrates that more settlement is observed away from the centreline which suggests that higher shear bands developed and concentrated underneath the edges of
594
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
b
a
Fig. 6. Deformation pattern of one layer of reinforcement at different depths (a) reinforced loose sand beds (b) reinforced dense sand beds.
that the reinforced dense sand above the localised soft pocket resists the footing load up to a S/B value of 5% irrespective of the number of reinforcing layers. Over this range of settlement, the whole ground behaves like an elastic material transferring the induced stresses to the sides bridging over the localised weaker zone until the critical bearing pressure is reached. Furthermore, upon passing the critical bearing pressure, an abrupt increase in the measured settlement occurred indicating the commencement of
a 50
100
150
200
250
300
350
400 0
0 B
5
5 Ds = 1.0 B
S/B, %
10
u = h = 0.25 B
10 N = variable
Hs = 0.5B 15
15
Bs = 0.5B
20
20 N= 4
Sett lement , mm
4.4. Batch IV: effect of number of reinforcing layers
25
25
30
Unreinforced loose sand -with soft pocket
35
N= 2
N= 1
N= 3
30
Unreinforced loose sand no soft pocket
35
Bearing pressure, kPa
b
S/B, %
The number of reinforcing layers increased gradually from one layer up to four layers whilst keeping the vertical spacing between layers at 0.25B (25 mm) and length of each layer to be 2B (200 mm) to quantify the efficacy of the increased number of reinforcing layers. Two series of testing (H and I) are conducted for achieving the objectives of this batch and their conditions are stated in Table 1. Fig. 7a and b shows the bearing pressure against S/B for loose and dense sand respectively. The data shows that there is a significant improvement in the load carrying capacity as the number of layers increases. This can be attributed to (i) higher stiffness and confinement for the reinforced region beneath the foundation as a result of the addition of more reinforcement layers and (ii) better frictional resistance in the case of dense sand. In both sands, the bearing capacity increased by around four-fold to that of the unreinforced sand bed measured at 10% settlement and almost twice that of deposits without a localised soft pocket when four layers are used. This indicates that increasing the number of reinforcing layers is the most effective way in enhancing the load carrying capacity. Comparing data for loose sand beds presented in Fig. 7a with those generated for dense sand beds in Fig. 7b illustrates that different characteristics for the load-settlement curves are obtained. In loose sand beds very low yielding values followed by a gradual improvement in the bearing resistance with further settlement is experienced up to a S/B value of about 12%. Subsequently the gradient of the load-settlement curve flattens which could be due to the effect of the compressed localised soft pocket and overburden pressure above the foundation level. In contrast, the dense sand behaved in a different manner. The results indicate
Bearing pressure, kPa 0
0
50
100
150
200
250
300
350
400
0
0
5
5
10
10
15
15 N= 4
20
20 N= 3
25
30
35
Unreinforced loose sand -with soft pocket
N= 1
Set t lement , mm
the footings. Based on visual observation, it appears that the reinforcement layer performs as a membrane in tension. Of note no breakage/tear is observed in any of the reinforcement sheets in all experiments of this batch. This suggests that failure of the reinforced soil is due to the insufficient anchorage and bond with the surrounding soil. Thus, one layer of reinforcement is insufficient to improve the load-settlement relation to its full potential without the existence of the soft pocket. These data oppose the early recommendation stated by Alexiew (1998) that a layer of reinforcement is sufficient to bridge over a void. Furthermore, placing the reinforcing layer immediately above the soft pocket or void will not produce the best improvements as recommended by (BRE, 2004; Alexiew, 1998).
25
N= 2 30 Unreinforced loose sand no soft pocket
35
Fig. 7. Variation of bearing pressure with S/B for different N (a) reinforced loose sand, (b) reinforced dense sand.
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
a
Bearing pressure, kPa 0
50
100
150
200
250
300 0
0 Unreinforced loose sand - no soft pocket Unreinforced loose sand - with soft pocket 5
br = 2.5 B (250 mm)
10
br = 3.0 B (300 mm)
B 15
15
u = h = 0.25 B br = variable N= 2
Ds = 1.0 B
25
Bs = 0.5B 25
30
35
30
b
20
Hs = 0.5B
20
Set t lement, mm
S/B, %
5
br = 2.0 B (200 mm)
10
Bearing pressure, kPa 0
50
100
150
200
250
300 0
0 Unreinforced dense sand - no soft pocket 5
Unreinforced dense sand - with soft pocket
5
br= 2.0 B (200 mm)
15
br = 2.5 B (250 mm) br = 3.0 B (300 mm)
10
15
20
20
25
25
30
30
35
35
Sett lement, mm
10
S/B, %
a punching shear failure. To aid the discussion, the BPR measured at 10% settlement is plotted as a function of the number of layers (N) and shown in Fig. 8. It is clear that there is an exponential increase in the bearing capacity as the number of layers increases. The results indicate that there is no optimal number for the reinforcing layers and increasing the stiffness of the soil by increasing the number of reinforcing layers is key to achieve the best possible improvement. Visual inspection of the reinforcing layers after the completion of experiments in loose sand beds indicates that there is no tear whatsoever developed in cases of using 1, 2 and 3 reinforcing layers. For a 4-layer reinforcing system, the bottom layer failed exactly along the centreline which highlights that the tensile strength of the bottom layer controls the behaviour of the reinforced ground. In this case the reinforced region behaves as a flexible reinforced slab. Thus the failure of the bottom reinforcing layer occurs when the tensile stress induced by bending exceeds the maximum tensile strength of the reinforcing layer. It should be noted that this contradicts the earlier assumption that for reinforced ground the stresses are higher at shallower depths and as a result top layers fail first (Binquet and Lee, 1975). No failure is observed for 1 and 2 layer reinforcing systems in dense sand beds containing a soft pocket. However, two tears symmetrically around the centreline occurred in cases of 3 and 4-layer systems. This can be explained by the good grip between the reinforcing layers and surrounding soil.
595
4.5. Batch V: effect of length of reinforcing layers In order to investigate the influence of increasing the length of the reinforcing layers on the bearing capacity, six experiments were carried out with different lengths of reinforcement in loose and dense sand beds. In each experiment two layers of reinforcement were placed at a vertical spacing of 0.25B (25 mm) underneath the model footing. Data for the effect of length of reinforcement are presented in Fig. 9a and b. Results for the loose sand beds suggest that almost identical load-settlement curves are obtained. This indicates that increasing the length of reinforcement in a loose sand deposit does not provide significant extra enhancement for the soil load carrying capacity. This is primarily due to the weak bond between reinforcing layers and surrounding soil. However, for dense sand, the load carrying capacity improved slightly with the increase in the length of reinforcing layers. There has been an increase of 20% in the bearing capacity measured at a settlement of 10% when the length of reinforcement is increased from 2B (200 mm) to 3B (300 mm). The beneficial effect of the length of
2 1.8 1.6
Dense sand Loose sand
1.4
BP R
1.2 1 0.8 0.6 0.4 0.2 0 0
1
2
3
No of layers, N Fig. 8. Variation of BPR as a function of N.
4
5
Fig. 9. Variation of bearing pressure with S/B for different br (a) reinforced loose sand, (b) reinforced dense sand.
reinforcement is present but it is not pronounced as with the effect of the number of reinforcing layers. In these experiments, no fracture of the reinforcement layers occurred which is consistent with previous observations for 2-layer systems in the previous group of experiments.
5. Conclusions A small-scale laboratory experimental programme was carried out systematically in a 500 mm 500 mm 200 mm tank to investigate the effect of width and depth of a soft pocket on the ultimate bearing capacity of unreinforced loose and dense sand beds. Various geometrical parameters describing the reinforcement arrangement were studied in a systematic manner. The main conclusions from the experimental investigation are: The existence of soft pockets has a major impact in reducing the capacity of the soil to resist surface loads. Their impact is directly related to the relative strength of the soft pocket and surrounding soil. With the increase in the width of soft pockets the bearing capacity reduces dramatically. It was found that for a soft pocket of similar footing width placed at a depth of B, the bearing capacity reduced by 55% in loose sand and 70% in dense sand. Soft pockets within a depth of 1.5B below the footing interfere with the failure zone underneath the foundation and result in significant loss to the carrying capacity.
596
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
Failure of the reinforced loose sand beds containing a soft pocket is mainly due to local shear failure. However, for reinforced dense sand beds with the inclusion of a soft pocket, various modes of failure can occur with punching shear failure as a dominant mode of failure. Minor improvement can be expected with the addition of one reinforcing layer irrespective of its depth. However, better results are obtained with dense sand but it does not fully mitigate the negative impact of the soft pocket existence. It is apparent that the key parameter in improving sand beds overlying soft pockets is the number of reinforcing layers. The bearing capacity increased by almost four times when the region underneath the footing was reinforced by four layers. This improvement is equivalent to almost doubling the bearing capacity of unreinforced dense sand bed without the inclusion of soft pocket. Failure of the reinforcing layers in loose sand beds was largely due to low frictional resistance between reinforcing layers and surrounding soil. However, fracture of the deeper reinforcing layers occurs in dense sand beds which suggest that the reinforced region behaves as a flexible reinforced slab. Acknowledgment The author is grateful to Mr T. Kroh, Ms L. Charles and Mr S. Perng for their help with the experiments. The author is also grateful to Prof. Simon Tait for his comments to improve the paper. References Adams, M.T., Colin, J.G., 1997. Large model spread footing load tests on geosynthetic reinforced soil foundations. Journal of Geotechnical and Geoenvironmental Engineering, ASCE 123 (1), 66e72. Alawaji, H.A., 2001. Settlement and bearing capacity of geogrid-reinforced sand over collapsible soil. Geotextile and Geomembranes 19, 75e88. Alexiew, D.A., 1998. Overspanning a void by specially produced geogrids. In: Proceedings of the Geosynthetics Asia’97, 3e10, Bangalore, India. Basudhar, P.K., Saha, S., Deb, L., 2007. Circular footings resting on geotextile reinforced sand bed. Geotextiles and Geomembranes 25 (6), 377e384. Binquet, J., Lee, K.L., 1975. Bearing capacity test on reinforced earth slabs. Journal of Geotechnical Engineering, ASCE 101 (12), 1241e1255. British Standards Institute, BS 5930, 1999. Code of practice for site investigations. Building Research Establishment, BRE, 2004. Working Platforms for Tracked Plant Good Practice Guide to the Design, Installation, Maintenance and Repair of Ground Supported Working Platforms. BRE Bookshop, Watford, UK. Chen, Q., Abu-Farskh, M., Sharma, R., 2009. Experimental and analytical studies of reinforced crushed limestone. Geotextiles and Geomembranes 27 (5), 357e367. Das,B.M.,Khing,K.H.,1994.Foundationonlayeredsoilwithgeogridreinforcementeeffect ofavoid.GeotextilesandGeomembranes13(8),545e553.
El Sawwaf, M., 2007. Behavior of strip footing on geogrid-reinforced sand over a soft clay slope. Geotextiles and Geomembranes 25 (1), 50e60. Ghazavi, M., Lavasan, A.A., 2008. Interference effect of shallow foundations constructed on sand reinforced with geosynthetics. Geotextiles and Geomembranes 26 (5), 404e415. Giroud, J.P., Bonaparte, R., Beech, J.F., Gross, B.A., 1990. Design of soil layer-geosynthetic systems overlying voids. Geotextiles and Geomembranes 9 (1), 11e50. Gottardi, G., Houlsby, G.T., Butterfield, R., 1999. Plastic response of circular footings on sand under general planar loading. Geotechnique 49 (4), 453e469. Guido, V.A., Dong, K.G., Sweeny, A. 1986. Bearing capacity of shallow foundations reinforced with geogrids and geoexiles. In: 2nd Canadian Symposium on Geotextiles and Geomembranes, Edmonton, Alberta, Canada. Huang, C.C., Meng, F.Y., 1997. Deep-footing and wide-salb effects in reinforced sand ground. Journal of Geotechnical and Geoenvironmental Engineering, ASCE 123 (1), 30e36. Huang, C.C., Tatsuoka, F., 1990. Bearing capacity of reinforced horizontal sandy ground. Geotextiles and Geomembranes 9, 51e82. Jewel, R.A., Milligan, G.W.E., Sarsby, R.W., Dubois, D., 1984. Interaction between soil and geogrids. In: Symp. Polymer Grid Reinforcement in Civil Engineering, London, England, pp. 18e30. Khan, A., Horoshenkov, K.V., Benkreira, H., Patel, R., 2005. Controlled extrusion of porous media for acoustic applications. In: Proceedings of the 1st Symposium on the Acoustics of Poroelastic Media (SAPEM), Lyon, France. Latha, G.M., Somwanshi, A., 2009. Bearing capacity of square footings on geosynthetic reinforced sand. Geotextiles and Geomembranes 27 (4), 281e294. Michalowski, R.L., Shi, L., 2003. Deformation patterns of reinforced foundation sand at failure. Journal of Geotechnical and Geoenvironmental Engineering 129 (6), 439e449. Patra, C.R., Das, B.M., Atalar, C., 2005. Bearing capacity of embedded strip foundation on geogrid reinforced sand. Geotextiles and Geomembranes 23, 454e462. Patra, C.R., Das, B.M., Bhoi, M., Shin, E.C., 2006. Eccentrically loaded strip foundation on geogrid-reinforced sand. Geotextiles and Geomembranes 24 (4), 254e259. Phanikumar, B.R., Prasad, R., Singh, A., 2009. Compressive load response of geogridreinforced fine, medium and coarse sands. Geotextiles and Geomembranes 27 (3), 183e186. Poorooshasb, H.B., 1991. Load settlement response of a compacted fill layer supported by a geosynthetic overlying a void. Geotextiles and Geomembranes 10, 179e201. Prothero, D.R., Schwab, F., 2004. Sedimentary Geology: an Introduction to Sedimentary Rocks and Stratigraphy. Freeman and Company, USA. Saleh, N.M., 2001. Experimental evaluation of soilegeotextile interface friction properties. Journal of Engineering and Applied Science 48 (3), 419e435. Sharma, R., Chen, Q., Abu-Farskh, M., Yoon, S., 2009. Analytical modelling of geogrid reinforced soil foundation. Geotextiles and Geomembranes 27 (1), 63e72. Sireesh, S., Sitharam, T.G., Dash, S.K., 2009. Bearing capacity of circular footing on geocell-sand mattress overlying clay bed with void. Geotextiles and Geomembranes 27 (2), 89e98. Wang, M.C., Feng, Y.X., Jao, M., 1996. Stability of geosynthetic-reinforced soil above a cavity. Geotextiles and Geomembranes 14, 95e109. Yamamoto, K., Otani, J., 2002. Bearing capacity and failure mechanism of reinforced foundations based on rigid-plastic finite element formulation. Geotextile and Geomembranes 20, 367e393. Yetimoglu, T., Jonathan, T.H., Saglamer, A., 1994. Bearing capacity of rectangular footings on geogrid-reinforced sand. Journal of Geotechnical and Geoenvironmental Engineering, ASCE 120, 2083e2099. Zhou, H., Wen, X., 2008. Model studies on geogrid- or geocell-reinforced sand cushion on soft soil. Geotextiles and Geomembranes 26 (3), 231e238.
Contents lists available at ScienceDirect
Geotextiles and Geomembranes journal homepage: www.elsevier.com/locate/geotexmem
Technical Note
Two dimensional experimental study for the behaviour of surface footings on unreinforced and reinforced sand beds overlying soft pockets M.H.A. Mohamed* School of Engineering, Design and Technology, University of Bradford, Bradford, West Yorkshire BD7 1DP, UK
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 October 2009 Received in revised form 25 May 2010 Accepted 30 May 2010 Available online 26 June 2010
This paper presents results of a comprehensive investigation undertaken to quantify the efficiency of using reinforcement layers in order to enhance the bearing capacity of soils that are characterised by the existence of localised soft pockets. Small-scale model experiments using two dimensional tank were conducted with beds created from well graded sand with mean particle size of 300 mm but prepared with different dry densities. A relatively softer material was embedded at predetermined locations within the sand beds so as to represent localised soft pockets. Various arrangements of soil reinforcement were tested and compared against comparable tests but without reinforcement. In total 42 tests were carried out in order to study the effect of the width and depth of the soft pocket, the depth of one reinforcing layer and the length and number of reinforcing layers on the soil bearing capacity. The results show clearly that the ultimate bearing capacity reduces by up to 70% due to the presence of a soft pocket. It was also noted that the proximity of the soft pocket also influenced the bearing capacity. Reinforcing the soil with a single layer or increasing the length of reinforcement is not as effective as was anticipated based on previous studies. However, bearing capacity increased significantly (up to 4 times) to that of unreinforced sand when four layers of reinforcement were embedded. The results suggest that rupture of the bottom reinforcement layer is imminent in heavily reinforced sand beds overlying soft pockets and therefore its tensile strength is critical for successful reinforcement. Ó 2010 Elsevier Ltd. All rights reserved.
Keywords: Bearing capacity Ground variability Soft pockets Soil reinforcement
1. Introduction and background The bearing capacity of weak soils is often inadequate to support imposed building loads. To minimise induced stresses within the weak soil to an acceptable level, it is common practice to reinforce the surface soil stratum with layers of reinforcement and use granular material to fill in between reinforcing layers. Reinforcement of weak grounds is considered as an effective and relatively economical approach for enhancing bearing capacity and reducing associated settlement. Several experimental, analytical and numerical research studies have been undertaken to investigate the impact of various configurations of reinforcement on the overall behaviour of reinforced soils under surface loads (e.g., Yetimoglu et al., 1994; Huang and Meng, 1997; Adams and Colin, 1997; Alawaji, 2001; Yamamoto and Otani, 2002; Patra et al., 2006; Basudhar et al., 2007; El Sawwaf, 2007, Ghazavi and Lavasan, 2008, Zhou and Wen, 2008, Chen et al., 2009, Sharma et al., 2009; Latha and Somwanshi, 2009). The experimentally based
* Tel.: þ44 1274 233856; fax: þ44 1274 234111. E-mail address: [email protected] 0266-1144/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.geotexmem.2010.06.001
studies were undertaken for various footing shapes (e.g. strip, rectangular, square and circular) and using different types of reinforcement (e.g. geogrid, geotextile and metal strips, sheets or rods). This variation has led to different recommendations for the optimum configuration of the reinforcement layers. In addition, to enhance the shearing resistance between reinforcing layers and surrounding soil, it was found to be beneficial to use a frictional material to fill in between reinforcing layers (e.g. Saleh, 2001), particularly coarse sand (Phanikumar et al., 2009). Shearing resistance was also enhanced by using geogrid layers (e.g., Guido et al., 1986; Yetimoglu et al., 1994; Patra et al., 2005). Improving the shearing resistance might require less anchorage length and hence reduce the overall cost. Jewel et al. (1984) pointed out that the mechanism for mobilization of friction resistance in geogrid-reinforced soils is different from that of geotextile based. The failure mechanisms and deformation patterns of reinforced soil were also subject to several investigations. Binquet and Lee (1975) identified three possible failure mechanisms which might occur in reinforced soils as a function of the reinforcement arrangement and strength of layers. Those were failure above the reinforced zone and failure due to upper ties break and ties pull out. Various deformation patterns of reinforced sand beds by a single
590
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
layer at failure were studied by Michalowski and Shi (2003). They suggested that some alterations occur to the failure mechanism due to the existence of the reinforcing layer. It was also noted that the deformation pattern of reinforced sand maintains many of the failure characteristics of unreinforced sand when one long layer is used. Haung and Tatsouka (1990) concluded that the length of reinforcing layers is a major factor affecting the deformation pattern of reinforced sand beds. They clearly distinguished between a “deep footing effect” which occurs as a result of short reinforcement and “wide slab effect” which is associated with the use of long high-strength reinforcing layers. Recently Sharma et al. (2009) investigated the occurrence of a failure mechanism within the reinforced zone and developed an analytical model to estimate the soil ultimate bearing capacity. The results of the proposed model were in good agreement with those obtained from physical testing. Most of the previous studies investigated the performance of reinforced sand, reinforced cohesive soil or reinforced sand overlying a clay layer of uniform thickness. However, less attention has been given for more practical soils in which local changes in the ground conditions occur e.g. due to the inclusion of localised soft pockets and/voids. They occur as a result of natural and/or manmade activities and are often encountered in construction sites. For example, poor supervision and implementation of engineering works would result in poorly filled voids and trenches (BRE, 2004) as well as a significant variation in the ground strength in abandoned landfill sites. Furthermore, soft pockets/zones can also be encountered in many virgin soils in the form of e.g. soft clay plugs within sandy meander belts and in tropical areas due to leaching and deposition of fine clay particles by the infiltration of water (Prothero and Schwab, 2004). The size and location of a soft pocket can vary widely and its existence can cause a significant ground distress and might result in intolerable ground movement if appropriate remedial measures are not undertaken. Despite the early efforts by Binquet and Lee (1975) to understand the behaviour of reinforced soils overlying a soft pocket, there is a lack of detailed investigations to quantify the effect of the local variability of ground conditions. In particular, the influence of localised soft pockets on the behaviour of subsurface soils under foundation loadings is poorly understood and has hindered the development of a comprehensive analytical model. Unlike soft pockets, studies for reinforced ground overlying voids received more consideration (see for example, Giroud et al., 1990; Poorooshasb, 1991; Das and Khing, 1994; Wang et al., 1996; Alexiew, 1998; Sireesh et al., 2009). This paper therefore aims to investigate the effects of; (i) width and depth of a localised soft pocket on the ultimate bearing capacity of unreinforced sand beds (ii) reinforcement configuration including; the depth of a single reinforcement layer, number of reinforcing layers, and length of reinforcement on the load carrying capacity of the reinforced system. These factors will be investigated at two different states of soil packing so as to study the influence of soil density and hence soil strength on the behaviour of the reinforced system. In addition, the failure mechanisms of unreinforced and reinforced soils above localised soft pocket will be described and discussed. 2. Laboratory techniques and materials An experimental set-up utilising a fully automated Instron loading machine that is controlled using computer software was developed to study the behaviour of unreinforced and reinforced sand beds with the inclusion of soft pockets. A tank with length of 500 mm, height of 500 mm and width of 200 mm was manufactured for this study. The front side of the tank was made of 20 mm thick plexiglass to enable visual inspection and the remaining sides
were made of wood. It should be noted that the tank was also braced to eliminate its deformation at high values of footing pressure. A rough rigid model footing with a width of 100 mm and covering the whole width of the tank was used to simulate plane strain conditions. The applied load and settlement of the footing were measured and recorded electronically through the data acquisition system every 1 s. Dry silica sand with a narrow range of particle sizes between 600 mm and 75 mm was used. It was found that the median diameter of the sand is 300 mm, D10 ¼ 180 mm, D30 ¼ 240 mm and D60 ¼ 330 mm. The uniformity and curvature coefficients are determined and found to be 1.18 and 1.0 respectively. According to the British Standards BS5930, the sand used can be classified as poorly graded sand. In order to enable preparation of sand beds with a range of densities, a sand deposition technique has been developed, calibrated and utilised in this investigation. The sand deposition technique has been successfully used in previous studies (e.g., Gottardi et al., 1999) and proved to construct sand samples of uniform density. The experiments were carried out at two different densities which are 1540 kg/m3 and 1660 kg/m3 so as to represent two practical soil states (loose and dense respectively). Hereafter sand beds will be referred to as either loose or dense sand. A large shear box with internal dimensions of 300 mm 300 mm was used for determination of the angle shearing resistance for loose and dense samples. The angle of shearing resistance was found to be 30.5 and 35.4 for loose and dense samples respectively. Several materials have been tested under unconfined compression in order to determine if it would make an appropriate soft material. Arma-foam sound 240 with Young’s modulus of 124 kPa was used as a soft pocket. The Arma-foam sound 240 is a porous material manufactured by controlled extrusion (Khan et al., 2005). It should be noted that no lateral deformation was observed during the loading and therefore the measured vertical deformation is directly related to the applied vertical compressive stress. Various types of reinforcement with a range of strengths have been used in previous studies. In this investigation a decision was taken to use weak reinforcement so as to potential modes of failure can be observed. Sheets of Guilt paper which has a relatively low tensile strength are available. The tensile strength of the proposed reinforcing materials was determined experimentally and found to be 2.47 N/m2 at an extensible stain of 3%. In addition the ultimate tensile strength at failure is 3.5 N/m2. The angles of friction between Guilt paper and loose sand and between Guilt paper and dense sand were 19.6 and 26 respectively. 3. Testing programme and procedure Fig. 1 illustrates all parameters that have been considered in this investigation. An experimental programme was designed to test the influence of these parameters. Table 1 presents the values of the parameters used in the 40 experimental tests. Initial tests were carried out on unreinforced sand beds without the inclusion of soft pockets to determine the bearing pressure-settlement behaviour of loose and dense sands. Then two batches of testing were carried out to investigate the impact of localised soft pocket width and depth on the behaviour of the unreinforced loose and dense sand beds under surface loading. Subsequently a number of experiments were conducted to investigate each parameter as stated in Table 1 and depicted in Fig. 1. Sand beds were prepared using the sand raining technique. The process of pouring sand was continued until the predetermined level of the soft pocket was reached, then the sand surface was carefully levelled off and the softer material was placed in accordance with the desired test conditions. The sand deposition process was resumed until the sand level reached the level of the bottom
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
591
Fig. 1. Illustration of the problem and parameters considered.
reinforcing layer. The reinforcing sheet was placed horizontally and symmetrically around the centreline of the soft pocket. The process was repeated for all layers of reinforcement. Finally, the sand surface was levelled off and a model footing was placed on the surface ensuring that the model footing is located symmetrically around the vertical axis of the reinforcing layers and the localised soft pocket. The model footing was then loaded in such way so that a settlement rate of 1 mm/min was obtained in all tests. Load was measured with an accuracy of 1 105 N. The tests were stopped either automatically when a failure occurred which is indicated by severe loss in bearing resistance or manually when a 30 mm of vertical settlement was reached. 4. Results and discussion It should be noted that all measured values for the vertical settlement (S) are presented as a function of the footing width (B). In addition a characteristic parameter “bearing pressure ratio, (BPR)” is introduced and defined as:
BPR ¼
tests (Batch 0) on sand beds without the inclusion of a soft pocket indicated that classical load-settlement relations were obtained and the ultimate bearing capacity was found to be 57 kPa and 114 kPa for loose and dense sand beds respectively at S/B of 10%. 4.1. Batch I: effect of a localised soft pocket width In total eight experiments were undertaken to quantify the effect of width of a localised soft pocket on the bearing capacity of unreinforced loose and dense sand beds as illustrated in Table 1 (series B and C). Fig. 2a and b shows the results of the bearing pressure against settlement ratio for different soft pocket widths. Also shown are the results of the load-settlement relations for sand beds without the inclusion of soft pockets. The results clearly demonstrate that the load carrying capacity of a sand bed reduces significantly with the increase in the width of a soft pocket irrespective of the state of the soil packing. In other words, applying a similar level of bearing pressure would result in unacceptable foundation settlement if the soft pocket was not identified and
qu ðunreinforced=reinforced systems with inclusion of a soft pocketÞ qu ðunreinforced sand without inclusion of a soft pocketÞ
where qu is the ultimate bearing capacity (kN/m2) of the unreinforced and reinforced sand beds with and without the inclusion of soft pockets. The characteristic parameter (BPR) reflects the loss in the value of the bearing capacity due to the existence of a soft pocket as well as taking into account the benefit of adding reinforcement layers. In this paper all values for the ultimate bearing pressure values were determined at a vertical settlement of 10% of the footing width (S/B ¼ 10%). This settlement value was selected because (i) it would indicate practically a failure due to excessive settlement and (ii) careful inspection of the laboratory data clearly indicates that failure occurred in almost all the experiments at S=B 10%. It should be noted that the observed increase in resistance at large settlement is impractical and due to (i) the build up of overburden pressure above the foundation level due to bulging of sand and (ii) an increased stiffness of the soft pocket. Results of loading
(1)
considered. The results suggested that the local shear failure experienced with the loose sand deposits remains a characteristic feature for loose sand beds with a localised soft pocket. Nevertheless the failure pattern in dense sand beds containing a localised soft pocket showed a direct dependence on the width of the soft pocket (see, Fig. 2b). For a relatively small soft pocket, local shear failure would occur whereas as when the width of the soft pocket increases it is more likely that a punching shear failure will be dominant. Localised soft pockets beneath the footing intercept the failure zone in the sand bed and cause significant transformation in stresses and hence the observed failure pattern. With further increase in the footing load, soil arch fails and subsequently a significant portion of the stresses is transferred to the soft pocket leading to a substantial settlement and punching shear failure. To aid the discussion, the BPR is determined using Equation (1) and
592
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
Table 1 Testing programme. Series
Variable parameter
0 I
A B
H
Density Bs ¼ 25, 50, 75 and 100 mm Bs ¼ 25, 50, 75 and 100 mm Ds ¼ 50, 100, 150 and 250 mm Ds ¼ 50, 100, 150 and 250 mm u ¼ 25, 50, 75, 60 and 100 mm u ¼ 25, 50, 75, 60 and 100 mm N ¼ 1, 2, 3 and 4
I
N ¼ 1, 2, 3 and 4
J
br ¼ 200, 250 and 300 mm br ¼ 200, 250 and 300 mm
C II
D E
III
F G
IV
V
K
Fixed parameters Hs ¼ 50 mm, Ds ¼ 100 mm, Loose Hs ¼ 50 mm, Ds ¼ 100 mm, Dense Hs ¼ 50 mm, Bs ¼ 50 mm, Loose Hs ¼ 50 mm, Bs ¼ 50 mm, Dense br ¼ 200 mm, Loose br ¼ 200 mm, Dense br ¼ 200 mm, u ¼ h ¼ 25 mm, Loose br ¼ 200 mm, u ¼ h ¼ 25 mm, Dense N ¼ 2, h ¼ 25 mm, u ¼ 25 mm, Loose N ¼ 2, h ¼ 25 mm, u ¼ 25 mm, Dense
plotted in Fig. 3 as a function of the soft pocket width (Bs) for loose and dense sands. It can be seen that there is a gradual reduction in the load carrying capacity as the width of soft pocket increases. The results also show that the effect of the soft pocket is more pronounced in dense sand beds. There is a loss of 70% of the load carrying capacity of dense sand when a localised soft pocket with a width of 0.5B (50 mm) exists at a depth of B (100 mm).
a 20
40
60
80
100
120
140
160
4.2. Batch II: effect of the localised soft pocket depth In this batch, the depth of a localised soft pocket was varied in accordance with the test conditions stated in Table 1 (series D and E). Data generated from eight tests for loose and dense sand beds with a soft pocket at different depths are presented in Fig. 4a and b. It can be seen that there is a direct loss in the load carrying capacity as the soft pocket gets closer to the foundation in the loose sand beds. In other words, the deeper the soft pocket the lower its influence on the measured values of the bearing pressure and
a
Bearing pressure, kPa 0
Fig. 3. Variation of BPR as a function of the soft pocket width.
180
0
0
Bearing pressure, kPa 0
20
40
60
80
100
120
Bs = 0.50 B (50 mm)
10
S /B, %
Bs = 1.00 B (100 mm) 15
15
Loose sand - no soft pocket
20
20 B
25
10
Ds = 2.50 B (250 mm) Loose sand - no soft pocket
15
15
20
20 B Ds = variable
25
Hs = 0.5B
30
10
25
Ds = 1.0B
5
Ds = 1.00 B (100 mm) Ds = 1.50 B (150 mm)
S/B, %
10
180 0
5
Bs = 0.75 B (75 mm)
160
Ds = 0.50 B (50 mm)
5
Settlement , mm
5
140
0
Bs = 0.25 B (25 mm)
30
25
Hs = 0.5B 30
30
Bs = variable
Bs = 0.5B
35
35 35
35
b
Set t lement , mm
Batch no
Bearing pressure, kPa 0
20
40
60
80
100
120
140
160
b
180
0
0 Bs = 0.25 B (25 mm) Bs = 0.50 B (50 mm)
5
5
Bearing pressure, kPa 0
20
40
60
80
100
120
140
160
180
0
0
5
5
10
10
Dense sand - no soft pocket
15
20
25
10
15
20 B Ds = 1.0B
25
S/B, %
Bs = 1.00 B (100 mm)
Settlement , mm
S /B, %
10
B
15
15
Ds = variable 20 Ds = 0.50 B (50 mm) 25
Ds = 1.00 B (100 mm)
20
Hs = 0.5B
Set t lement, mm
Bs = 0.75 B (75 mm)
Bs = 0.5B 25
Ds = 1.50 B (150 mm)
30
30
Hs = 0.5B
30
Bs = variable 35
Ds = 2.50 B (250 mm)
30
Dense sand - no soft pocket
35
Fig. 2. Variation of bearing pressure with S/B for different widths of soft pocket (a) unreinforced loose sand, (b) unreinforced dense sand.
35
35
Fig. 4. Variation of bearing pressure with S/B for different soft pocket depths (a) unreinforced loose sand, (b) unreinforced dense sand.
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
associated settlement. Values of the BPR determined at 10% S/B as a function of the soft pocket depth (Ds) suggest that the bearing capacity reduces by 30% when a soft pocket of width (B) is inserted within a loose sand bed at a depth of (B). Local shear failure is observed in all tests involving loose sand beds independent of the depth of the soft pocket. The data presented indicate that soil arching is not a major factor governing the behaviour of loose sand beds since considerable settlement is observed shortly after applying a relatively small loading increment. The results indicated that there is a limited negative impact of the soft pocket and this occurs when it is placed at a depth equal to 2.5B (250 mm). Unlike the loose sand deposits, data for the dense sand deposits show larger reduction in the value of the bearing capacity due to the existence of a soft pocket. This highlights the influence of the initial relative stiffness between the soil mass and the softer zone. For shallow soft pockets that are within a depth of B (100 mm), a loss of w55% in the load carrying capacity of dense sand bed is obtained. In this case induced stresses underneath the footing are strongly affected by the presence of the weak pocket. Careful inspection of the data presented in Fig. 4b shows that for all cases of soft pocket, there is an abrupt loss in the soil resistance at some stage of the loading which could be explained by the failure of the soil arch. This means that a redistribution of the stresses occurs and helps in bridging over the weak pocket until a final failure point. Furthermore, the observed subsequent gain in the strength at large settlement is not practical and primarily because of the overburden pressure created by the significant settlement. In these tests, a punching shear failure occurs and is evident by the excessive settlement at a particular bearing pressure. Some anomalies are encountered in the load-settlement curves for soft pockets that are placed at a depth of 2.5B. These experiments were repeated couple of times and almost identical curves are obtained. The irregularity in the ultimate bearing capacity could be due to the initial compression of the soft pocket under the impact of sand raining and overburden pressure exerted afterwards. However, no clear explanation can be drawn out and further testing is underway. Combining results of Batch I and Batch II, it seems reasonable to conclude that the presence of soft pockets of width >0.25B (25 mm) and within a depth of 1.5B (150 mm) would strongly affect the ultimate bearing capacity. This highlights that serious considerations should be given to the effects of ground variability on the load carrying capacity of the whole soil mass. 4.3. Batch III: effect of depth of one reinforcement layer This batch of experiments was conducted using one sheet of reinforcement layer that is placed at different depths to determine precisely whether one reinforcement layer is sufficient to enhance the load carrying capacity of the sand beds with the inclusion of a soft pocket as well as to determine the optimal depth of a reinforcement layer. The test conditions of Batch III are illustrated in series F and G in Table 1. Fig. 5a and b shows the variation of the bearing pressure with settlement ratio (S/B) for sand reinforced with one layer and with the inclusion of a soft pocket. Also shown are the curves for unreinforced sand beds with and without the inclusion of soft pocket. The results obtained for loose sand beds demonstrate that generally a marginal improvement is achieved by the use of one reinforcement layer irrespective of its depth. In addition, a slightly higher bearing capacity can be obtained when the layer of reinforcement is placed at a depth of 0.75B (75 mm). In comparison, the dense sand beds showed considerable improvement by the use of one layer of reinforcement as shown in Fig. 5b. The data illustrate that generally for shallow reinforcements of <0.5B (50 mm) as well as for deep reinforcement at 1B (100 mm), a slight improvement could be achieved. In addition, deep
593
Fig. 5. Variation of bearing pressure with S/B for different u (a) reinforced loose sand, (b) reinforced dense sand.
reinforcement results in better improvements if a settlement ratio of >7% can be tolerated. However, best enhancement is obtained when the reinforcement layer is placed at an intermediate level e.g. 0.6e0.75 B (60e75 mm) which improves the ultimate bearing capacity by almost 100% to that of the unreinforced sand bed but it does not still recover the full strength of the dense sand without the soft pocket. This gives an indication that maximum shear stress would develop at that depth so that the reinforcing layer will take up a significant portion of the shear stress causing further confinement and hence leading to an increased resistance to failure. Upon the completion of each test, sand was carefully excavated and the shape of the deformed reinforcement sheet was recorded by imaging. Fig. 6 shows a presentation for the deformed shapes against the depth of one layer of reinforcement. Fig. 6 indicates that the overall deformation is related to the location of the reinforcement layer with respect to the foundation and the soft pocket. A bowl shape is observed for reinforcement layers at shallow depths. Furthermore the shape is flattened out at deeper levels and even might be inverted to those previously observed when a reinforcement layer is directly placed on top of the soft pocket. The latter suggests that movement of the sand along the sides of the soft pocket was inward towards the soft pocket which is opposite to the well reported deformation pattern for reinforced sand without the inclusion of soft pockets (Binquet and Lee, 1975; Michalowski and Shi, 2003). In addition, reinforcement layers at intermediate levels deformed differently. Fig. 6 illustrates that more settlement is observed away from the centreline which suggests that higher shear bands developed and concentrated underneath the edges of
594
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
b
a
Fig. 6. Deformation pattern of one layer of reinforcement at different depths (a) reinforced loose sand beds (b) reinforced dense sand beds.
that the reinforced dense sand above the localised soft pocket resists the footing load up to a S/B value of 5% irrespective of the number of reinforcing layers. Over this range of settlement, the whole ground behaves like an elastic material transferring the induced stresses to the sides bridging over the localised weaker zone until the critical bearing pressure is reached. Furthermore, upon passing the critical bearing pressure, an abrupt increase in the measured settlement occurred indicating the commencement of
a 50
100
150
200
250
300
350
400 0
0 B
5
5 Ds = 1.0 B
S/B, %
10
u = h = 0.25 B
10 N = variable
Hs = 0.5B 15
15
Bs = 0.5B
20
20 N= 4
Sett lement , mm
4.4. Batch IV: effect of number of reinforcing layers
25
25
30
Unreinforced loose sand -with soft pocket
35
N= 2
N= 1
N= 3
30
Unreinforced loose sand no soft pocket
35
Bearing pressure, kPa
b
S/B, %
The number of reinforcing layers increased gradually from one layer up to four layers whilst keeping the vertical spacing between layers at 0.25B (25 mm) and length of each layer to be 2B (200 mm) to quantify the efficacy of the increased number of reinforcing layers. Two series of testing (H and I) are conducted for achieving the objectives of this batch and their conditions are stated in Table 1. Fig. 7a and b shows the bearing pressure against S/B for loose and dense sand respectively. The data shows that there is a significant improvement in the load carrying capacity as the number of layers increases. This can be attributed to (i) higher stiffness and confinement for the reinforced region beneath the foundation as a result of the addition of more reinforcement layers and (ii) better frictional resistance in the case of dense sand. In both sands, the bearing capacity increased by around four-fold to that of the unreinforced sand bed measured at 10% settlement and almost twice that of deposits without a localised soft pocket when four layers are used. This indicates that increasing the number of reinforcing layers is the most effective way in enhancing the load carrying capacity. Comparing data for loose sand beds presented in Fig. 7a with those generated for dense sand beds in Fig. 7b illustrates that different characteristics for the load-settlement curves are obtained. In loose sand beds very low yielding values followed by a gradual improvement in the bearing resistance with further settlement is experienced up to a S/B value of about 12%. Subsequently the gradient of the load-settlement curve flattens which could be due to the effect of the compressed localised soft pocket and overburden pressure above the foundation level. In contrast, the dense sand behaved in a different manner. The results indicate
Bearing pressure, kPa 0
0
50
100
150
200
250
300
350
400
0
0
5
5
10
10
15
15 N= 4
20
20 N= 3
25
30
35
Unreinforced loose sand -with soft pocket
N= 1
Set t lement , mm
the footings. Based on visual observation, it appears that the reinforcement layer performs as a membrane in tension. Of note no breakage/tear is observed in any of the reinforcement sheets in all experiments of this batch. This suggests that failure of the reinforced soil is due to the insufficient anchorage and bond with the surrounding soil. Thus, one layer of reinforcement is insufficient to improve the load-settlement relation to its full potential without the existence of the soft pocket. These data oppose the early recommendation stated by Alexiew (1998) that a layer of reinforcement is sufficient to bridge over a void. Furthermore, placing the reinforcing layer immediately above the soft pocket or void will not produce the best improvements as recommended by (BRE, 2004; Alexiew, 1998).
25
N= 2 30 Unreinforced loose sand no soft pocket
35
Fig. 7. Variation of bearing pressure with S/B for different N (a) reinforced loose sand, (b) reinforced dense sand.
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
a
Bearing pressure, kPa 0
50
100
150
200
250
300 0
0 Unreinforced loose sand - no soft pocket Unreinforced loose sand - with soft pocket 5
br = 2.5 B (250 mm)
10
br = 3.0 B (300 mm)
B 15
15
u = h = 0.25 B br = variable N= 2
Ds = 1.0 B
25
Bs = 0.5B 25
30
35
30
b
20
Hs = 0.5B
20
Set t lement, mm
S/B, %
5
br = 2.0 B (200 mm)
10
Bearing pressure, kPa 0
50
100
150
200
250
300 0
0 Unreinforced dense sand - no soft pocket 5
Unreinforced dense sand - with soft pocket
5
br= 2.0 B (200 mm)
15
br = 2.5 B (250 mm) br = 3.0 B (300 mm)
10
15
20
20
25
25
30
30
35
35
Sett lement, mm
10
S/B, %
a punching shear failure. To aid the discussion, the BPR measured at 10% settlement is plotted as a function of the number of layers (N) and shown in Fig. 8. It is clear that there is an exponential increase in the bearing capacity as the number of layers increases. The results indicate that there is no optimal number for the reinforcing layers and increasing the stiffness of the soil by increasing the number of reinforcing layers is key to achieve the best possible improvement. Visual inspection of the reinforcing layers after the completion of experiments in loose sand beds indicates that there is no tear whatsoever developed in cases of using 1, 2 and 3 reinforcing layers. For a 4-layer reinforcing system, the bottom layer failed exactly along the centreline which highlights that the tensile strength of the bottom layer controls the behaviour of the reinforced ground. In this case the reinforced region behaves as a flexible reinforced slab. Thus the failure of the bottom reinforcing layer occurs when the tensile stress induced by bending exceeds the maximum tensile strength of the reinforcing layer. It should be noted that this contradicts the earlier assumption that for reinforced ground the stresses are higher at shallower depths and as a result top layers fail first (Binquet and Lee, 1975). No failure is observed for 1 and 2 layer reinforcing systems in dense sand beds containing a soft pocket. However, two tears symmetrically around the centreline occurred in cases of 3 and 4-layer systems. This can be explained by the good grip between the reinforcing layers and surrounding soil.
595
4.5. Batch V: effect of length of reinforcing layers In order to investigate the influence of increasing the length of the reinforcing layers on the bearing capacity, six experiments were carried out with different lengths of reinforcement in loose and dense sand beds. In each experiment two layers of reinforcement were placed at a vertical spacing of 0.25B (25 mm) underneath the model footing. Data for the effect of length of reinforcement are presented in Fig. 9a and b. Results for the loose sand beds suggest that almost identical load-settlement curves are obtained. This indicates that increasing the length of reinforcement in a loose sand deposit does not provide significant extra enhancement for the soil load carrying capacity. This is primarily due to the weak bond between reinforcing layers and surrounding soil. However, for dense sand, the load carrying capacity improved slightly with the increase in the length of reinforcing layers. There has been an increase of 20% in the bearing capacity measured at a settlement of 10% when the length of reinforcement is increased from 2B (200 mm) to 3B (300 mm). The beneficial effect of the length of
2 1.8 1.6
Dense sand Loose sand
1.4
BP R
1.2 1 0.8 0.6 0.4 0.2 0 0
1
2
3
No of layers, N Fig. 8. Variation of BPR as a function of N.
4
5
Fig. 9. Variation of bearing pressure with S/B for different br (a) reinforced loose sand, (b) reinforced dense sand.
reinforcement is present but it is not pronounced as with the effect of the number of reinforcing layers. In these experiments, no fracture of the reinforcement layers occurred which is consistent with previous observations for 2-layer systems in the previous group of experiments.
5. Conclusions A small-scale laboratory experimental programme was carried out systematically in a 500 mm 500 mm 200 mm tank to investigate the effect of width and depth of a soft pocket on the ultimate bearing capacity of unreinforced loose and dense sand beds. Various geometrical parameters describing the reinforcement arrangement were studied in a systematic manner. The main conclusions from the experimental investigation are: The existence of soft pockets has a major impact in reducing the capacity of the soil to resist surface loads. Their impact is directly related to the relative strength of the soft pocket and surrounding soil. With the increase in the width of soft pockets the bearing capacity reduces dramatically. It was found that for a soft pocket of similar footing width placed at a depth of B, the bearing capacity reduced by 55% in loose sand and 70% in dense sand. Soft pockets within a depth of 1.5B below the footing interfere with the failure zone underneath the foundation and result in significant loss to the carrying capacity.
596
M.H.A. Mohamed / Geotextiles and Geomembranes 28 (2010) 589e596
Failure of the reinforced loose sand beds containing a soft pocket is mainly due to local shear failure. However, for reinforced dense sand beds with the inclusion of a soft pocket, various modes of failure can occur with punching shear failure as a dominant mode of failure. Minor improvement can be expected with the addition of one reinforcing layer irrespective of its depth. However, better results are obtained with dense sand but it does not fully mitigate the negative impact of the soft pocket existence. It is apparent that the key parameter in improving sand beds overlying soft pockets is the number of reinforcing layers. The bearing capacity increased by almost four times when the region underneath the footing was reinforced by four layers. This improvement is equivalent to almost doubling the bearing capacity of unreinforced dense sand bed without the inclusion of soft pocket. Failure of the reinforcing layers in loose sand beds was largely due to low frictional resistance between reinforcing layers and surrounding soil. However, fracture of the deeper reinforcing layers occurs in dense sand beds which suggest that the reinforced region behaves as a flexible reinforced slab. Acknowledgment The author is grateful to Mr T. Kroh, Ms L. Charles and Mr S. Perng for their help with the experiments. The author is also grateful to Prof. Simon Tait for his comments to improve the paper. References Adams, M.T., Colin, J.G., 1997. Large model spread footing load tests on geosynthetic reinforced soil foundations. Journal of Geotechnical and Geoenvironmental Engineering, ASCE 123 (1), 66e72. Alawaji, H.A., 2001. Settlement and bearing capacity of geogrid-reinforced sand over collapsible soil. Geotextile and Geomembranes 19, 75e88. Alexiew, D.A., 1998. Overspanning a void by specially produced geogrids. In: Proceedings of the Geosynthetics Asia’97, 3e10, Bangalore, India. Basudhar, P.K., Saha, S., Deb, L., 2007. Circular footings resting on geotextile reinforced sand bed. Geotextiles and Geomembranes 25 (6), 377e384. Binquet, J., Lee, K.L., 1975. Bearing capacity test on reinforced earth slabs. Journal of Geotechnical Engineering, ASCE 101 (12), 1241e1255. British Standards Institute, BS 5930, 1999. Code of practice for site investigations. Building Research Establishment, BRE, 2004. Working Platforms for Tracked Plant Good Practice Guide to the Design, Installation, Maintenance and Repair of Ground Supported Working Platforms. BRE Bookshop, Watford, UK. Chen, Q., Abu-Farskh, M., Sharma, R., 2009. Experimental and analytical studies of reinforced crushed limestone. Geotextiles and Geomembranes 27 (5), 357e367. Das,B.M.,Khing,K.H.,1994.Foundationonlayeredsoilwithgeogridreinforcementeeffect ofavoid.GeotextilesandGeomembranes13(8),545e553.
El Sawwaf, M., 2007. Behavior of strip footing on geogrid-reinforced sand over a soft clay slope. Geotextiles and Geomembranes 25 (1), 50e60. Ghazavi, M., Lavasan, A.A., 2008. Interference effect of shallow foundations constructed on sand reinforced with geosynthetics. Geotextiles and Geomembranes 26 (5), 404e415. Giroud, J.P., Bonaparte, R., Beech, J.F., Gross, B.A., 1990. Design of soil layer-geosynthetic systems overlying voids. Geotextiles and Geomembranes 9 (1), 11e50. Gottardi, G., Houlsby, G.T., Butterfield, R., 1999. Plastic response of circular footings on sand under general planar loading. Geotechnique 49 (4), 453e469. Guido, V.A., Dong, K.G., Sweeny, A. 1986. Bearing capacity of shallow foundations reinforced with geogrids and geoexiles. In: 2nd Canadian Symposium on Geotextiles and Geomembranes, Edmonton, Alberta, Canada. Huang, C.C., Meng, F.Y., 1997. Deep-footing and wide-salb effects in reinforced sand ground. Journal of Geotechnical and Geoenvironmental Engineering, ASCE 123 (1), 30e36. Huang, C.C., Tatsuoka, F., 1990. Bearing capacity of reinforced horizontal sandy ground. Geotextiles and Geomembranes 9, 51e82. Jewel, R.A., Milligan, G.W.E., Sarsby, R.W., Dubois, D., 1984. Interaction between soil and geogrids. In: Symp. Polymer Grid Reinforcement in Civil Engineering, London, England, pp. 18e30. Khan, A., Horoshenkov, K.V., Benkreira, H., Patel, R., 2005. Controlled extrusion of porous media for acoustic applications. In: Proceedings of the 1st Symposium on the Acoustics of Poroelastic Media (SAPEM), Lyon, France. Latha, G.M., Somwanshi, A., 2009. Bearing capacity of square footings on geosynthetic reinforced sand. Geotextiles and Geomembranes 27 (4), 281e294. Michalowski, R.L., Shi, L., 2003. Deformation patterns of reinforced foundation sand at failure. Journal of Geotechnical and Geoenvironmental Engineering 129 (6), 439e449. Patra, C.R., Das, B.M., Atalar, C., 2005. Bearing capacity of embedded strip foundation on geogrid reinforced sand. Geotextiles and Geomembranes 23, 454e462. Patra, C.R., Das, B.M., Bhoi, M., Shin, E.C., 2006. Eccentrically loaded strip foundation on geogrid-reinforced sand. Geotextiles and Geomembranes 24 (4), 254e259. Phanikumar, B.R., Prasad, R., Singh, A., 2009. Compressive load response of geogridreinforced fine, medium and coarse sands. Geotextiles and Geomembranes 27 (3), 183e186. Poorooshasb, H.B., 1991. Load settlement response of a compacted fill layer supported by a geosynthetic overlying a void. Geotextiles and Geomembranes 10, 179e201. Prothero, D.R., Schwab, F., 2004. Sedimentary Geology: an Introduction to Sedimentary Rocks and Stratigraphy. Freeman and Company, USA. Saleh, N.M., 2001. Experimental evaluation of soilegeotextile interface friction properties. Journal of Engineering and Applied Science 48 (3), 419e435. Sharma, R., Chen, Q., Abu-Farskh, M., Yoon, S., 2009. Analytical modelling of geogrid reinforced soil foundation. Geotextiles and Geomembranes 27 (1), 63e72. Sireesh, S., Sitharam, T.G., Dash, S.K., 2009. Bearing capacity of circular footing on geocell-sand mattress overlying clay bed with void. Geotextiles and Geomembranes 27 (2), 89e98. Wang, M.C., Feng, Y.X., Jao, M., 1996. Stability of geosynthetic-reinforced soil above a cavity. Geotextiles and Geomembranes 14, 95e109. Yamamoto, K., Otani, J., 2002. Bearing capacity and failure mechanism of reinforced foundations based on rigid-plastic finite element formulation. Geotextile and Geomembranes 20, 367e393. Yetimoglu, T., Jonathan, T.H., Saglamer, A., 1994. Bearing capacity of rectangular footings on geogrid-reinforced sand. Journal of Geotechnical and Geoenvironmental Engineering, ASCE 120, 2083e2099. Zhou, H., Wen, X., 2008. Model studies on geogrid- or geocell-reinforced sand cushion on soft soil. Geotextiles and Geomembranes 26 (3), 231e238.