IGC. 50 th INDIAN GEOTECHNICAL CONFERENCE STUDY ON SHORT PILE GROUP CONNECTED TO RIGID RAFT UNDER HORIZONTAL LOAD

Transcription

1 50 th IGC 50 th INDIAN GEOTECHNICAL CONFERENCE 17 th 19 th DECEMBER 015, Pune, Maharashtra, India Venue: College of Engineering (Estd. 1854), Pune, India STUDY ON SHORT PILE GROUP CONNECTED TO RIGID RAFT UNDER HORIZONTAL LOAD Soumya Roy 1, Bikash Chandra Chattopadhyay, Ramendu Bikash Sahu 3 ABSTRACT Quick growth of many cities in last two decades all over the world led to rapid increase in height and number of buildings even in unfavorable ground conditions. Since the 80 s a new foundation technique called as piled raft foundation have been developed and used extensively to reduce the total, differential settlements along with the tilts of structures subjected to different loading conditions using short piles as settlement reducers Burdland et al., In piled raft foundations piles are extensively used to reduce settlements and deflections to permissible limits rather than to support weight of structures. Although a good number of studies have been made on piled raft foundations for vertical loads, works involving study on piled raft behavior under lateral loads are limited. Lateral capacity of piled raft foundation is basically derived from the lateral capacities of piles. Piles subjected to lateral loading are commonly analyzed using semi-empirical method known as the p-y analysis (Matlock and Ripperger 1958; Matlock 1970; Reese et al. 1975). However this is mainly adopted for analyzing long piles. In the analysis of these data, the soil pressures acting on the pile are modeled as discrete independent nonlinear springs, i.e., soil continuity is disregarded. An alternative approach, elastic analysis, which makes use of boundary integral equations and treats soil as an elastic continuum is sometimes used (Banerjee and Davies 1978; Poulos 1971). This approach, however, does not account for soil yielding or pile yielding and it is, therefore, only suitable for the prediction of the load-deformation responses of laterally loaded piles at small strain levels. However, the selection of appropriate secant stiffnesses is difficult, and such analyses cannot be used to predict the 1 Roy_Soumya, Civil Engineering Department, M S I T, Kolkata, West Bengal, India, roy.shoummo@gmail.com Chattopadhyay_Chandra Bikash, Ex. Prof. & Ex Head, Civil Department, IIEST, Shibpur, Howrah,ccbikash@yahoo.com 3 Sahu_Bikash Ramendu, HoD, Civil Department, J.U., Kolkata, India, address@coep.ac.in

2 Authors names separated by comma / &, limited to one line on all even pages (Times New Roman 8 italics, aligned left) effect of overload in any rational manner. Further, elastic analyses tend to underestimate the values of maximum bending moment, regardless of the choice of soil stiffness. Other several methods have been published for predicting the ultimate lateral resistance to piles in cohesionless soils (Brinch Hansen 1961; Broms 1964; Reese et al. 1974; Poulos and Davis 1980; Fleming et al. 199). However, these methods often produce significantly different ultimate resistance values. Even standard Indian codes provide design guidelines for long piles only. This makes it difficult for practicing engineers to effectively select the appropriate method when designing laterally loaded piles in cohesionless soils. Because the problem of determining the ultimate resistance to a laterally loaded pile is three dimensional and nonlinear, finding a rigorous solution is highly unlikely. Thus existing solutions for ultimate lateral resistance are either of a semi-empirical nature or employ approximate analysis which often involves considerable simplifications (Jamiolkowski and Garassino 1977). These approximations may account for the significantly different ultimate resistance values obtained from the different methods. This paper first reviews the existing methods for predicting the ultimate lateral resistance to piles in cohesion less soils. Extensive model tests were conducted on raft supported on short piles. The lateral soil resistance and deflection characteristics of the short piles are observed through lateral load test on pile groups subjected ultimate failure condition. The obtained data are compared with various lateral load carrying capacities of piles. To check the accuracy of the available methods, the tests are carried out to selected displacement limits first followed by horizontal loading till ultimate failure condition. Finally, a method for calculation of lateral resistance rafts supported by short pile group is proposed which gives good agreement with 1g model tests. Keywords: Short Pile, Pile Group, Lateral Load, Pile- Raft Foundation.

3 50 th IGC 50 th INDIAN GEOTECHNICAL CONFERENCE 17 th 19 th DECEMBER 015, Pune, Maharashtra, India Venue: College of Engineering (Estd. 1854), Pune, India STUDY ON SHORT PILE GROUP CONNECTED TO RIGID RAFT UNDER HORIZONTAL LOAD Soumya Roy, PhD candidate, Jadavpur University, Kolkata, West Bengal. Dr. Bikash Chandra Chattopadhyay, Ex. Head & Ex. Professor, Civil Engineering Department, IIEST, Howrah, West Bengal, Dr. Ramendu Bikash Sahu, HoD, Civil Engineering Department, J.U., Kolkata, West Bengal, ABSTRACT: Investigation of the lateral load resistance of piled raft foundation is made by laboratory model test on piled raft foundation supported on single and group piles. The experiments are carried out with varying size, spacing of piles in group and length to diameter ratio (L/d) of the piles. Lateral resistance of pile is a function of shape, size, spacing and length to diameter ratio (L/d) of the pile. In this study, model pile is single pile, and group piles having configurations are of (x1, x) which satisfy the Meyerhof s relative stiffness limit of pile for flexible pile. For model pile embedded length to diameter ratio (L/d) are 0, 30, 35 and spacing are S = 3d, 4.5d, 6d. The load-displacement responses, ultimate resistance, group efficiency of piles with different spacing and number of piles in group have been qualitatively and quantitatively studied in the present work. From the load-displacement curve, ultimate lateral resistances are obtained by double tangent method. The results are compared with some available analytical methods and conclusions are drawn for deriving lateral resistance of pile group in cohesion less deposits. INTRODUCTION There are several approaches available in the literature to estimate lateral load resistance of pile in sand and clay by Broms [1]; Meyerhof et al. []; Meyerhof and Ranjan [3]; Meyerhof and Sastry [4]; Meyerhof and Yalcin [5]; Sastry [6]; Sastry and Meyerhof [7]; Zhang et al. [8]. Also some simplified methods are proposed by Meyerhof et al. [9] and Patra & Pise [10] for cohesionless soil. However, these methods often produce significantly different ultimate lateral load resistance value as Zhang et al. [8]. This makes it difficult to select appropriate method when designing laterally loaded pile in cohesionless soil. A comparative study has been made in this paper between Mayerhof and Patra & Pise methods and experimental results with the intension that it could add some value on the understanding of practicing engineers. Also a parametric study is done on ultimate lateral load resistance. Because the problem of determining the ultimate resistance to a laterally loaded pile is three dimensional and nonlinear, finding a rigorous solution is highly unlikely. Thus existing solutions for ultimate lateral resistance are either of a semi empirical nature or employ approximate analysis. This paper first reviews the existing methods for predicting the ultimate lateral resistance to piles in cohesionless soils. By analyzing the lateral soil resistance distribution along the width of the pile and based on test results of eight model rigid piles in cohesionless soils from the published literature, a simple method is developed for calculating the ultimate lateral resistance (including frontal soil resistance and side shear resistance) to piles in cohesionless soils. Because the proposed method considers the ultimate resistance of the soil (not of the pile), it is applicable to both flexible and rigid piles. To check the accuracy of the proposed method, it is first used to calculate the ultimate lateral resistance to flexible model test piles in a centrifuge. The calculated values agree well with those obtained from the tests. Predicting the lateral load capacity of rigid test piles in cohesionless

4 Authors names separated by comma / &, limited to one line on all even pages (Times New Roman 8 italics, aligned left) soils using the proposed method also yields satisfactory results. LITERATURE REVIEW In general, analysis of laterally loaded pile can be divided in to two groups: (1) Short pile and () Long pile. The lateral load on long pile can be analyze on the concept of sub grade modulus or considering soil is an elastic medium. There are a large number of analytical and field and laboratory investigation are available for predicting lateral strength of piles. According to Winkler model, an elastic medium (soil) can be replaced by a series of infinitely close, independent elastic spring. Based on this assumption, Matlock and Reese [11] gave a general formula of determining moment and displacement of a vertical pile which is subjected to lateral load and moment. Brom s [1] developed a simplified solution for laterally loaded piles based on the assumption of: (a) Shear failure on soil, which is the case of short pile and (b) Bending of the pile governed by plastic yield resistance of pile section, which is applicable on long pile. Meyerhof and Ranjan [3] conducted model tests on rigid single pile and pile groups under central inclined loads in homogeneous sand and developed a semi-empirical interaction relationship assuming elliptical variation of ultimate capacity under axial to lateral load. Chattopadhyay and Pise [13] suggested graphical methods to estimate the ultimate lateral resistance from the load-ground line deflection diagram of single pile. They found that the estimate values by the log load-log deflection method are in closer agreement with the experiment results and applicable to a pile of any length. Liu and Meyerhof [14] carried out a nonlinear analysis and showed that Le/L depends not only on the pile stiffness K rs but also L/d and other factors. McVay et al. [15] conducted centrifuge test on free head single and 3x3 pile groups at 3d and 5d pile spacing. Results of the test showed that group efficiency is independent of soil density. The behavior of the pile group was analyzed by the 3D elastoplastic analysis soil behavior with no tension characteristic as well as frictional elements of slippage on the pile soil interface. The pile group efficiency was estimated to be when the displacement reached 0.1d. It was found that for circular pile the pressure distribution across the diameter is not uniform, is maximum at center, and is almost zero at the two edges. A method is proposed to predict the soil pressure distribution and lateral resistance and is found to be in closed agreement with various field and laboratory data. Analytical Methods There are various methods of analysis the lateral loads on pile foundation. Following are the major methods of analysis the lateral loads of pile foundation: Elastic Method A general method for determining moments and displacements of a vertical pile embedded in a granular soil and subjected to the lateral load and moment at the ground was given by Matlock and Reese [11]. Consider a pile of length L subjected to a lateral force Qg and a moment Mg at the ground surface (z = 0) as shown in the Fig. 1.The figure shows the general deflected shape of the pile and the soil resistance caused by the applied load and the moment. According to a Simpler, Winkler s model, and elastic medium can be replaced by a series of infinitely closed independent elastic springs. Based on this assumption, p k (1) x Where, k = modulus of subgrade reaction, p = pressure on soil and x = deflection. The sub-grade modulus for granular soils at a depth z is defined as k nhxz () Where, n h = constant of modulus horizontal subgrade reaction. Using the theory of Brom s on an elastic oundation we can write

5 50 th IGC 50 th INDIAN GEOTECHNICAL CONFERENCE 17 th 19 th DECEMBER 015, Pune, Maharashtra, India Venue: College of Engineering (Estd. 1854), Pune, India E pi p K r = relative stiffness of pile = EsL E s = average lateral elastic modulus of soil L Fig. 1 Elastic deflection of a vertical pile. Where, E p = Modulus of elasticity in the pile material, I p = Moment of inertia of the pile section. Based on Winker s model p = - kx The sign in the above equation is negative because the soil reaction is in the direction opposite to the pile deflection, The solution of the above expression results in the following expressions Pile deflection at any depth Slope of pile at any depth Moment in pile at any depth Meyerhof s Method More recently, Meyerhof (1995) provided the solutions for laterally loaded rigid and flexible pile, which are summarized below. According to Meyerhof s method, a pile can be defined as flexible if Short piles in sand Ultimate lateral capacity of short piles in sand is given by Q 0.1 dl k 0. pdl ug b 4 Where b = unit wt. of soil, k b = resultant soil pressure co-efficient, p = limit pressure. The limit pressure can be given as p = 40 N q tan (KPa) and p = 60 N q tan (KPa) Where N q = bearing capacity factor. The maximum moment, M max in the pile due to the lateral load Q u(g) applied on pile is given by M 0. Q L M max 35 u( g ) y For long (flexible) piles in sand, the ultimate lateral load, Q u(g), can be estimated from the above equation by substituting an effective length (L e ) for L where, Le kr 1 L The maximum moment in a flexible pile due to a working lateral load Q g applied at the ground surface is 0. M 0.3k Q L 0. Q L max r g 3 Patra and Pise s Method Single Pile: Patra and Pise [10] modified Meyerhof s equations on the following assumptions a) The active earth pressures on the rear sides of the piles and the vertical tip resistance is neglected. b) The passive earth pressure at failure of a pile was taken as for the wall by multiplying it by a constant shape factor 3. Thus the ultimate lateral resistance QL s for a rigid free head fully embedded pile is g

6 Authors names separated by comma / &, limited to one line on all even pages (Times New Roman 8 italics, aligned left) Q 0.36 dl Ls k b Where, k b = coefficient of passive earth pressure on the wall. For a laterally loaded flexible pile relative stiffness of pile K rs , it was suggested by Meyerhof and Yalcin (1984) that in the absence of structural failure the ultimate lateral load for soil failure can be estimated for the above equation by using a corresponding effective embedment depth Le instead of L. Thus the ultimate lateral resistance Q Ls for a flexible, free head, fully embedded pile is Q 3(0.1 d ) L k Ls e b Pile group of X1 The ultimate lateral resistance of line pile group (x1) could be found out by extending the analysis for a single pile and considering the forces acting on piles and soil mass within the pile group. Active earth pressure develops at the back face of the front pile on the soil mass within the pile group and passive pressure at the front face of the rear pile on the soil mass within the pile group. This will develop the frictional resistance between the soil mass within the pile group and the soil mass along the side of the piles on the vertical plane. Again, the edges of the piles along the sides of the piles generate frictional resistance between pile and soil mass. The general configuration of line pile groups acted upon by the different forces considered as below: Q F Lg P p Where, F = frictional resistance on the vertical plane along the side of the pile group and P p = passive earth pressure for the front pile. Fig. Two pile group configuration Frictional resistance along the side of the pile group could be approximately found out as F L / kss Where, K s = co-efficient of earth pressure along the side of the pile group; L= embedded length of pile S = spacing of the pile group The ultimate lateral resistance of a rigid two pile group is thus given by Q L ( K s 0.36dk ) Lg Pile group of X s b The ultimate lateral resistance of square pile group (x) could be found-out by considering that passive earth resistance develops on the front piles along with side resistance. For a rigid square group it is expressed as Q L ( K s 0.7dk ) Lg s b Group Efficiency Variation of ultimate lateral capacity of pile group is generally expressed by group efficiency, and it is expressed as QLg n mqls Where, Q Lg = Ultimate lateral capacity of pile group, Q Ls = ultimate lateral capacity of single pile, m = No. of rows in a pile group and n = No. of columns in a pile group EXPERIMENTAL SETUP AND TEST PROGRAM Foundation Bed Medium dense sand obtained from the Hoogly river bed was used as foundation medium. The size of the model tank was 1mX1mX1m. Sand was placed at a placement density of 13kN/m 3 and the angle of internal friction is = 36º. Specific gravity of the sand used in the model tank is.4.

7 50 th IGC 50 th INDIAN GEOTECHNICAL CONFERENCE 17 th 19 th DECEMBER 015, Pune, Maharashtra, India Venue: College of Engineering (Estd. 1854), Pune, India Pile and Raft Mild steel tube of 0-mm outer diameter and 1mm wall thicknesses were used as model pile. For ensuring the pile-soil friction, the shaft of the piles was made rough by carefully machining it in lathe. The average outside diameter for rough pile was 0mm. The embedment length-to-diameter was 0, 30, and 35. Steel plate of 5mm thickness was taken as raft. The piles were fixed with the raft by heavy duty nuts and bolts. Model Test Piles Singe pile of embedment length-to-diameter (L/d) 0, 30, and 35 were tested. Group pile of configuration (X1), (X) of embedment length - to-diameter (L/d) 0, 30, and 35 were tested. Test Setup In this experiment, piles were subjected to lateral load and for this purpose an experimental setup were made to apply lateral load on pile. At first the model pile or piles were placed in the model tank. Then the soil medium (sand) placed in the tank from a certain height (0.5m) for maintaining fairly uniform placement density. For applying lateral load in the pile, a inextensible flexible weir was attached in the pile cap and a vertical stand with frictionless pulley was used to change the direction of vertical load to lateral load. Two dial gauges are attached to the pile cap to measure the lateral and vertical deflection of the pile. EXPERIMENTAL PROCEDURE The arrangement of test setup is shown in Fig. 3. Loading arrangement are made in such a way that it will act laterally to pile cap. After placing piles with the pile cap in the tank, sand was placed in the tank by falling from certain height of about 0.5m. The lateral load was applied to the pile cap through a pulley arrangement with flexible weir attached to the pile cap. The other end was attached to the loading apron. Load was applied by dead weight over the loading pan starting from the smallest with gradual increase in stages. Same loading sequences are maintained for all model pile. The loading sequences are 0.50, 1.0, 1.50,.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10, 11, 1, 15, 17, 0, 3, 6, 8, 30, 3, 34 Kg etc. Dial gauges having sensitivity of 0.01 mm were used for measuring the lateral and vertical deflection. When load was placed in the loading pan, it moves down and due to the pulley it act laterally to the pile cap. On application of lateral load, the piles are deflected and the deflections are recorded through the dial gauges. Deflections for corresponding load were noted. By plotting the noted value, load versus deflection curve was obtain which non-linear in nature. Ultimate lateral resistance of the pile was obtain from the curve by double tangent method (DTM) or the point where the curve show a greater deflection without further increasing any load. TEST RESULTS AND DISCUSSION 8 Fig. 3 Loading Arrangement The ultimate lateral resistance of the pile was found out by plotting lateral load versus displacement diagram in the plain graph paper. The curve was non-linear in nature. At the ultimate resistance, pile showed some deflection without any increase in load and this load was taken as ultimate load for the pile. Load displacement diagrams Curves of applied horizontal load versus lateral displacement for single pile, x1 and x pile group s are shown in Fig: 4 to Fig: 8.The load displacement curves are in general, similar and

8 Authors names separated by comma / &, limited to one line on all even pages (Times New Roman 8 italics, aligned left) non-linear. It is observed that the axial displacement as compared to the lateral displacements is negligible. They are shown (Fig: 10) for record and observation purpose only. From the lateral pull versus lateral displacement diagrams, for a particular value of lateral movement of pile, the magnitude of pull increases with increase in spacing. Lateral failure occurred at a pile head displacement from 4 to 8 mm (0.d to 0.4d) for L/d=0. However, for L/d=30, the lateral failure occurred at a pile head displacement of 6 to 10 mm (0.3d to 0.5d). Ultimate resistance for different cases has been estimated from the lateral pull-displacement diagrams by the double-tangent methods (DTM). The minimum load obtained from those diagrams is considered as the ultimate resistance for a particular condition. It is observed that the failure load is the point at which the curve exhibits a pick or maintains increase in displacement with no further increase in lateral resistance. Fig. 4 Ultimate lateral resistance for various L/d ratios Fig. 7 Ultimate lateral resistance of x pile raft for L/d =0 3 Fig. 5 Ultimate lateral resistance for x 1 pile raft for L/d = 0. Fig. 8 Ultimate lateral resistance of x pile raft for L/d =30 Fig. 6 Ultimate lateral resistance for x 1 pile raft for L/d = 30. Ultimate resistance Fig. 9 Ultimate lateral resistance of x pile raft for L/d =30

9 50 th IGC 50 th INDIAN GEOTECHNICAL CONFERENCE 17 th 19 th DECEMBER 015, Pune, Maharashtra, India Venue: College of Engineering (Estd. 1854), Pune, India to rigid raft. Following estimated conclusions can be made from the present study. Table 3 Comparison ultimate lateral resistance of pile raft on (x1) pile group with Patra & Pise method to observed experimental values Fig. 10. Variation efficiency of pile raft supported on x and x1 pile group. COMPARISON Following are the table which gives a comparison between the observed ultimate lateral resistances of the piled raft foundation supported on short pile groups to the theoretical value obtained from different methods: Table 1 Comparisons of ultimate lateral resistance of single pile by different methods with experimental values Table Comparison of ultimate lateral resistance of pile raft on (x) pile group with Patra & Pise method to observed experimental values CONCLUSIONS A detailed result of the experimental work has been given in this present work on short piles connected The ultimate lateral capacity of pile group depends on the length to diameter ratio of pile, pile friction angle, pile group geometry, spacing of piles in a group and sand placement density. The quantitative and qualitative influence of those parameters has been investigated. The load displacement curves are nonlinear. Lateral failure occurred at a pile head displacement from 4 to 8 mm (0.d to 0.4d) for L/d=0. However, for L/d=30, the lateral failure occurred at a pile head displacement of 6 to 10 mm (0.3d to 0.5d). Group efficiency of pile increases with an increase in pile spacing. It has seen that the efficiency at 3d spacing is less than that of 4.5d and 6d spacing. ACKNOWLEDGEMENT Special thanks go to the Department of Civil Engineering, Meghnad Saha Institute of Technology, Kolkata, West Bengal for their humble support in performing all the lengthy experimental works in their Soil Mechanics Laboratory and to the test tank fabricator Mr. Sanjoy who has tremendously helped in making the experimental setup to its perfection. REFERENCES 1. Brom s, B Lateral resistance of pile in cohesion less soil. J. Soil Mech. And Found. Div., ASCE, Vol. 90 (3), pp Meyerhof, G. G., Mathur, S. K., and Valsangkar, A. J Lateral Resistance and Deflection of Rigid Walls and Piles in Layered

10 Authors names separated by comma / &, limited to one line on all even pages (Times New Roman 8 italics, aligned left) Soils. Canadian Geotechnical Journal, 18(): Meyerhof, G. G., and Ranjan, G The bearing capacity of rigid piles under inclined load in sand. I. Vertical Piles. Canadian Geotechnical Journal, 9, Bro B Design of laterally loaded pile. Journal of the Soil Mechanics and Foundation Division, ASCE, Vol. 91 (3), pp Chattopadhyay, B.C., & Pise, P.J Ultimate lateral resistance of piles Int. J.Struct., Roorkee, India, pp Meyerhof, G. G., and Sastry, V. V. R. N Bearing capacity of rigid piles under eccentric and inclined loads. Canadian Geotechnical Journal,, Meyerhof, G. G., and Yalcin, A. S Pile capacity for eccentric inclined load in clay. Canadian Geotechnical Journal, 1, Sastry, V. V. R. N Bearing capacity of piles in layered soil, PhD Thesis, Technical University of Nova Scotia, Halifax, NS. 14. Liu, Q. F., & Meyerhof, G.G New methods of nonlinear analysis of laterally loaded flexible piles. Comp. and Geomech., Elsevier applied science, England. 4(3), pp Mc Vay. M., Casper, R., & Shang. T. 1995, Lateral response of three row groups in loose and dense sands at 3D and 5D pile spacing. J. Geotech. Engrg., ASCE. Vol. 11(5).pp Sastry, V. V. R. N., and Meyerhof, G. G Lateral soil pressures and displacements of rigid piles in homogeneous soils under eccentric and inclined loads. Canadian Geotechnical Journal, 3, Zhang, L., Silva, F., and Grismala, R Ultimate Lateral resistance to piles in cohesion less soils. Journal of Geotechnical and Geoenvironmental Engineering, 131(1): Meyerhof, G. G., Sastry, V., and Yalcin, A. S Lateral Resistance and Deflection of Flexible Piles. Canadian Geotechnical Journal, 5(3): Patra, N. R., and Pise, P. J Ultimate lateral resistance of pile groups in sand. Journal of Geotechnical and Geoenvironmental Engineering, 17(6): Matlock, H., Reese, L.C Generalized Solution for Laterally Loaded Piles. Journal of the Soil Mechanics & Foundations Division, American Society of Civil Engineers. Vol. 86, No.SM5, Part 1, PP