BORANG PENGESAHAN STATUS TESIS

Transcription

1 1 PSZ 19:16 (Pind. 1/97) UNIVERSITI TEKNOLOGI MALAYSIA BORANG PENGESAHAN STATUS TESIS JUDUL: UPLIFT CAPACITY OF ENLARGED BASE PILE IN SAND SESI PENGAJIAN: 25/26 Saya TIEO GEE HWU (HURUF BESAR) mengaku membenarkan tesis (PSM/Sarjana/Doktor Falsafah)* ini disimpan di Perpustakaan Universiti Teknologi Malaysia dengan syarat-syarat kegunaan seperti berikut: 1. Tesis adalah hakmilik Universiti Teknologi Malaysia. 2. Perpustakaan Universiti Teknologi Malaysia dibenarkan membuat salinan untuk tujuan pengajian sahaja. 3. Perpustakaan dibenarkan membuat salinan tesis ini sebagai bahan pertukaran antara institusi pengajian tinggi. 4. **Sila tandakan ( ) SULIT TERHAD (Mengandungi maklumat yang berdarjah keselamatan atau kepentingan malaysia seperti yang termaktub di dalam AKTA RAHSIA RASMI 1972) (Mengandungi maklumat TERHAD yang telah ditentukan oleh organisasi/badan di mana penyelidikan dijalankan) TIDAK TERHAD Disahkan oleh (TANDATANGAN PENULIS) Alamat Tetap: 249, TAMAN KANGKAR BARU, 844 SUNGAI MATI, MUAR, JOHOR. (TANDATANGAN PENYELIA) PM DR. RAMLI BIN NAZIR Nama Penyelia Tarikh: 21 APRIL 26 Tarikh: 21 APRIL 26 CATATAN: * Potong yang tidak berkenaan. ** Jika tesis ini SULIT atau TERHAD, sila lampirkan surat daripada pihak berkuasa/organisasi berkenaan dengan menyatakan sekali sebab dan tempoh tesis ini perlu dikelaskan sebagai SULIT atau TERHAD. Tesis dimaksudkan sebagai tesis bagi Ijazah Doktor Falsafah dan Sarjana secara penyelidikan, atau disertasi bagi pengajian secara kerja kursus dan penyelidikan, atau Laporan Projek Sarjana Muda (PSM).

2 I hereby declare that I have read this project report and in my opinion this project report is sufficient in terms of scope and quality for the award of the degree of Bachelor of Civil Engineering. Signature :.. Name of Supervisor : PM Dr. Ramli bin Nazir Date : 21 April 26

3 i UPLIFT CAPACITY OF ENLARGED BASE PILE IN SAND TIEO GEE HWU This thesis is submitted as a partial fulfilment of the requirement for the award of the degree of Bachelor of Civil Engineering Faculty of Civil Engineering Universiti Teknologi Malaysia APRIL, 26

4 ii I declare that this project report entitled Uplift Capacity of Enlarged Base Pile in Sand is the result of my own research except as cited in the references. The project report has not been accepted for any degree and is not concurrently submitted in candidature of any other degree. Signature :.. Name : TIEO GEE HWU Date : 21 April 26

5 To my beloved mother and father iii

6 iv ACKNOWLEDGEMENT I wish to express my sincere appreciation to my thesis supervisor, PM Dr. Ramli bin Nazir, for encouragement, guidance and constructive critics conveyed to me in preparing this thesis. Without his continued support and interest, this thesis would not have been the same as presented here. Staffs of Geotechnical Laboratory, Civil Engineering Faculty, also deserve special thanks for their assistance and guidance in preparation of apparatus, preparation of specimen and test procedure. My sincere appreciation also extends to all my friends who have provided assistance during experimental set-up. I cannot even start off the tests without their help. Their opinions and advices are useful indeed. Unfortunately, I am unable to list down all of them due to limited space. I am grateful to all my family members for their endless supports.

7 v ABSTRACT Enlarged Base Piles (EBP) are required to support structures subjected to overturning loads imposed by wind or wave. Uplift tests investigating the influences of foundation geometry on the uplift behaviour of piles with enlarged bases were carried out in dry loose sand and dense sand with unit weight of 14.8kN/m 3 and 16.9kN/m 3. The EBP models with different base diameter, shaft diameter or base angle were tested at embedment ratio ranging from 1 to 5. The results indicated that uplift capacity of EBP increases with an increase in sand density, base diameter and embedment depth. In contrast, increase in the base angle result in a decrease in uplift capacity. It is apparent that base angle, together with embedment ratio, base diameter and sand density, should be taken into account in the design of EBP. Failure mechanism studies conducted on EBP in dense sand show a slightly curve failure surface while the same studies in loose sand indicates an interface mechanism dominated the behaviour of EBP. Six empirical equations derived from the uplift tests give reasonable values on uplift capacity for EBP with base angle of 3 o, 45 o and 6 o in the case of dense sand and loose sand. A number of theories originated from small scale model tests on anchor give higher breakout factor values than those from the empirical equations.

8 vi ABSTRAK Tiang sambut bertapak besar adalah diperlukan untuk menanggung bangunan yang dikenakan beban angkat naik yang disebabkan angin atau ombak. Ujikaji angkat naik dijalankan untuk mengkaji pengaruh faktor-faktor geometri ke atas keupayaan angkat naik yang dimiliki oleh tiang sambut bertapak besar yang ditanam dalam pasir padat and pasir longgar. Pasir yang digunakan adalah pasir kering dengan berat unit γ d =14.8 kn/m 3 untuk pasir longgar dan γ d =16.9 kn/m 3 untuk pasir padat. Ujikaji-ujikaji ini menunjukkan keupayaan angkat naik meningkat apabila ketumpatan pasir, garis pusat tapak dan nisbah kedalaman bertambah. Sebaliknya, penambahan dalam sudut tapak dan garis pusat aci menyebabkan keupayaan angkat naik berkurang. Pemerhatian dari mekanisme kegagalan pasir di sekitar model tiang di dalam pasir padat menunjukkan kewujudan permukaan kegagalan yang bercondong sedikit sedangkan pemerhatian ke atas model tiang di dalam pasir longgar menunjukkan perlakuan angkat naik dikawal oleh mekanisme permukaan. Kebanyakan kaedah reka bentuk yang wujud sekarang adalah berdasarkan keputusan yang diperoleh daripada ujikaji skala kecil ke atas asas jenis sauh. Teori-teori ini didapati telah menganggar lebih keupayaan angkat naik yang dimiliki tiang sambut bertapak besar di dalam pasir padat dan pasir longgar. Enam persamaan empirikal yang diperoleh daripada ujikaji angkat naik memberi nilai keupayaan angkat naik yang munasabah ke atas tiang sambut bertapak besar yang mempunyai sudut tapak sebanyak 3 o, 45 o dan 6 o di dalam pasir padat dan pasir longgar.

9 vii CONTENT CHAPTER TITLE PAGE DECLARATION DEDICATION ACKNOWLEDGEMENT ABSTRACT ABSTRAK CONTENT LIST OF TABLES LIST OF FIGURES LIST OF SYMBOLS ii iii iv v vi vii x xii xvii LIST OF APPENDIX xix 1 INTRODUCTION General Introduction Problem Statement Objectives Scope of Study Significant of Research 4

10 viii 2 LITERATURE REVIEW Researches carried out on uplift capacity of pile Failure Mechanism Lateral Earth Pressure Coefficients Equations for Estimating Uplift Capacity Balla s Method for Estimating Uplift Capacity Meyerhof and Adams s Method for Estimating 14 Uplift Capacity 2.7 Breakout Factor Influence of Foundation Geometry 19 on Breakout Factor 2.8 Buckingham Pi Theorem 22 3 METHODOLOGY Scope Preparation of Apparatus Method Used to Obtain Loose Sand Packing Method Used to Obtain Dense Sand Packing Uplift Test Procedures Failure Mechanism Test Procedures Determination of Test Parameters 33 4 RESULTS AND ANALYSIS Failure Mechanism Studies Failure Mechanism Studies for Enlarged 35 Base Pile in Loose Sand

11 ix Failure Mechanism Studies for Enlarged 38 Base Pile in Dense Sand 4.2 The Influence of Embedment Ratio on Uplift 41 Capacity 4.3 The Influence of Shaft Diameter and Diameter 44 Ratio on Uplift Capacity 4.4 The Influence of Base Diameter on Uplift 48 Capacity 4.5 The Influence of Base Angle on Uplift Capacity The Influence of Sand Density on Uplift Capacity Empirical Equations Derived from Present Test Comparison between Existing Design Methods 63 and Experimental Work 5 CONCLUSIONS AND RECOMMENDATIONS Conclusions Recommendations 7 REFERENCES 71 Appendices A C 73-8

12 x LIST OF TABLES TABLE NO. TITLE PAGE 2.1 Assumption made in design methods 9 summarized by Ilamparuthi and Dickin (21) 2.2 Values of coefficient of shape factor, m base on soil 11 friction angle, φ (Meyerhof and Adams, 1968) 2.3 Relationship between critical embedment ratio, 11 (L/D b ) cr and soil friction angle, φ (Meyerhof and Adams, 1968) 2.4 The coefficient m and shape factors base on angle of 15 friction of soil, φ (Meyerhof, 1968) 2.5 Existing design methods summarized by Dickin and 16 Leung (199) 2.6 Comparison between theoretical uplift breakout 18 factors, N U for piles in sand with bell diameter of 1 m and with angle of friction of 4 o 3.1 Summary of uplift tests conducted for the various 23 combinations of parameters 3.2 Physical quantities assumed to be involved in 34 establishing the limiting uplift carrying capacity of uplift loaded enlarged base pile

13 xi 4.1 Comparison between breakout factors from 64 existing design methods and empirical equation for piles in loose sand with φ = 35 o 4.2 Comparison between breakout factors from existing 64 design methods and empirical equation in dense sand with φ = 44.5 o

14 xii LIST OF FIGURES FIGURE NO. TITLE PAGE 2.1 Assumed failure mechanisms for belled piers subject 8 to uplift loads classified by Dickin (1988). 2.2 Variation of uplift factor, B q with embedment ratio, 12 (L/Db) and soil friction angle, φ (Braja M.Das, 1984). 2.3 Coefficients of Breaking out resistance (Balla, 1961) (a) Shallow pile 14 (b) Deep pile 2.5 Variation of breakout factor N u with bell angle α 2 in centrifuge uplift tests (Dickin, 1988). 2.6 Design curve for belled piers in sand in centrifuge 2 uplift tests (Dickin and Leung, 1992). 2.7 Variation of breakout factor N u with diameter ratio, 21 b s /b b in centrifuge uplift tests (Dickin, 1988). 2.8 Variation of breakout factor N u with embedment 21 ratio L/b b for 1 m diameter belled piers in sand. (Dickin and Leung, 199). 3.1 Hand-held compactor used for compacting sand From left to right : load cell, linear variable 25 displacement transducer (LVDT)

15 xiii 3.3 Datalogger used for recording and printing out 26 experimental data 3.4 Control box used for adjusting the speed of motor Sand rainer used in sand-rain technique Enlarged base pier model pile with base angle, 27 α = 3, α =45 andα = Variations of density index and dry unit weight with 28 falling sand height for sand passing 2.36mm sieve size (Chosun, 1999) 3.8 Variation of density index and dry unit weight with 29 duration of mechanical compaction of sand passing 2.36mm sieve size (Chosun, 1999) 3.9 Schematic diagram for uplift tests Schematic diagram for failure mechanism tests Uplift test being carried out on model pile Failure mechanism test set-up in the laboratory Front view of semi-cylindrical model pile before 36 subjected to uplift load 4.2 Early movement of sand mass at the toe of the pile 37 Formed void at the bottom of the pile buried in loose Sand 4.3 Further movement of the pile pushed the sand mass 37 upward 4.4 Localize failure for enlarged base pile in loose sand Front view of semicylindrical model pile before 39 subjected to uplift load 4.6 Early movement of sand mass at the toe of the pile 4 formed void at the bottom of the pile buried in dense sand

16 xiv 4.7 Vertical uplift of the pile pushed the sand mass upward 4 further and formed rupture planes above the base 4.8 Failure mechanism for enlarged base pile in dense sand Variation of uplift load with vertical displacement for 42 Pile with Db = 1mm and Ds = 3mm in dense sand 4.1 Variation of uplift load with vertical displacement for 43 pile with Db = 1mm and Ds = 3mm in loose sand 4.11 Variation of uplift capacity with embedment ratio for 43 piles with Db = 1mm and Ds = 3mm in loose and dense sand 4.12 Variation of uplift load with vertical displacement for 45 piles with Db = 1 mm and L/Db = 3 in loose sand 4.13 Variation of uplift load with vertical displacement for 45 Piles with Db = 1 mm and L/Db = 3 in dense sand 4.14 Variation of uplift capacity with embedment ratio for 46 piles with Db = 1 mm in dense sand 4.15 Variation of uplift capacity with embedment ratio for 46 piles with Db = 1 mm in loose sand 4.16 Variation of uplift capacity with diameter ratio for 47 piles with Db = 1 mm in dense sand 4.17 Variation of uplift capacity with diameter ratio for 47 piles with Db = 1 mm in loose sand 4.18 Variation of uplift load with vertical displacement for 49 piles with Ds = 3 mm and L/Db = 3 in loose sand 4.19 Variation of uplift load with vertical displacement for 49 piles with Ds = 3 mm and L/Db = 3 in dense sand 4.2 Variation of uplift capacity with embedment ratio for 5 piles with Ds = 3 mm in loose sand 4.21 Variation of uplift capacity with embedment ratio for 5 piles with Ds = 3 mm in dense sand

17 xv 4.22 Variation of uplift capacity with base diameter for 51 piles with Ds = 3 mm in dense and loose sand 4.23 Variation of uplift load with vertical displacement for 52 piles with Ds = 3 mm, Db = 1mm and L/Db = 2 in loose sand 4.24 Variation of uplift load with vertical displacement for 53 piles with Ds = 3 mm, Db = 1mm and L/Db = 2 in dense sand 4.25 Variation of uplift capacity with embedment ratio for 53 piles with Ds = 3 mm and Db = 1 mm in loose sand 4.26 Variation of uplift capacity with embedment ratio for 54 piles with Ds = 3 mm and Db = 1 mm in dense sand 4.27 Variation of uplift load with vertical displacement for 55 piles with Ds = 3 mm, Db = 1mm and L/Db = 1 in dense sand and loose sand 4.28 Variation of uplift load with vertical displacement for 56 piles with Ds = 3 mm, Db = 1mm and L/Db = 5 in dense sand and loose sand 4.29 Variation of uplift capacity with embedment ratio for 56 piles with Db = 75mm and Ds = 4mm in loose and dense sand 4.3 Variation of breakout factor with embedment ratio 59 for piles with base angle of 3 o in loose sand 4.31 Variation of breakout factor with embedment ratio 59 for piles with base angle of 45 o in loose sand 4.32 Variation of breakout factor with embedment ratio 6 for piles with base angle of 6 o in loose sand 4.33 Variation of breakout factor with embedment ratio 6 for piles with base angle of 3 o in dense sand

18 xvi 4.34 Variation of breakout factor with embedment ratio 61 for piles with base angle of 45 o in dense sand 4.35 Variation of breakout factor with embedment 61 ratio for piles with base angle of 6 o in dense sand 4.36 Variation of breakout factor with embedment 62 ratio for piles with different base angle in loose sand 4.37 Variation of breakout factor with embedment ratio 62 for piles with different base angle in dense sand 4.38 Comparison between uplift tests and theoretical 65 predictions for enlarged base piles in loose sand 4.39 Comparison between uplift tests and recent 66 theoretical predictions for enlarged base piles in loose sand 4.4 Comparison between uplift tests and theoretical 67 predictions for enlarged base piles in dense sand 4.41 Comparison between uplift tests and recent 67 theoretical predictions for enlarged base piles in dense sand

19 xvii LIST OF SYMBOLS V - Breaking Out Resistance G 1 - Bulk Weight of Soil G 2 - Weight of Foundation Slab T v - Shear Reaction to Skin Friction D b - Base Diameter D s - Shaft Diameter L - Embedment Depth N - Breakout Factor α - Base Angle γ - Unit Weight of Soil Q - Uplift Capacity c u - Average Undrained Shear Strength of Soil φ Angle of Friction of Soil I d - Density Index P u - Uplift Load ϕ - Peak Friction Angle ψ - Dilatancy Angle Α - Plan Area of Pile Bell S f - Shape Factor K - Coefficient of Lateral Earth Pressure H - Maximum Height of Rupture Surface BBq - Uplift Factor

20 xviii K u - Uplift Coefficient m - Coefficient of Shape Factor Q un - Net Uplift Capacity Q ug - Gross Uplift Capacity W - Effective Weight of Pile θ - Cone Angle with Vertical

21 xix LIST OF APPENDIX APPENDIX TITLE PAGE A Dimensional Analysis 73 B Variation of Log N with Log L/Db for 75 Enlarged Base Piles in Loose Sand C Variation of Log N with Log L/Db for 78 Enlarged Base Piles in Dense Sand

22 1 CHAPTER I INTRODUCTION 1.1 General Introduction Pile foundations, also known as substructure, are used to carry and transfer the load of the structure to the bearing ground located at some depth below ground surface. The main components of the foundation are the pile cap and the piles. Piles are long and slender members which transfer the load through the shallow soil of lower bearing capacity to deeper soil or rock of high bearing capacity. They are also used in normal soil condition to resist uplift forces or in poor soil condition to resist lateral forces. Most of the piles are made from three main types of material which are wood, steel and concrete. Piles are normally driven, drilled or jacked into the ground and connected to pile caps. Pile foundations are generally used to support compressive loads from superstructure. Some structures such as tall chimneys, transmission towers and jetty structures are subjected to overturning loads imposed by wind. In such cases, piles are required to resist uplift forces which are much greater than the weight of the structure itself. In addition to wind, marine structures and transmission line towers are hit by wave forces and line tension respectively. Straight shafted piles obtain their resistance to uplift through friction between the pile shaft and the surrounding soil. It may be increased by under-reaming or belling out the bottom of the piles. The uplift capacity of a buried enlarged base or belled pile essentially come from the weight of soil within the failure zone above the base, the frictional

23 2 resistance along the failure surface and the self-weight of the foundation. The required pullout resistance can be achieved by increasing sand density, base diameter of pile and the depth of embedment. The influence of these parameters on the uplift capacity in sand has been investigated by many researchers. However, studies regarding the influence of base angle on the uplift behaviour are limited. According to Joseph, E. Bowles, the ratio of the base diameter to the shaft diameter (D b /D s ) should be no greater than 3. Enlarged base piles are proven to be economical foundation type to provide resistance to both compressive and uplift load. The cost advantages of enlarged base piles are due to the reduced pile shaft diameter which results in less concrete needed to replace the excavated material. However, the construction of enlarged base pile exposes the construction workers to the danger of collapse of the excavation. Contractors build enlarged base or bell out the bottom of the pile by using special belling buckets. The enlarged base is usually cut at 45 or 6 degree angle with maximum diameter of being not more than three times the diameter of the shaft. 1.2 Problem Statement Deep foundations with enlarged bases such as enlarged base pile or belled pile are often used to support a variety of land structures such as guyed lattice towers, transmission towers, tension cables for suspension bridges and tent-type roofs and marine structures such as floating platforms, tension leg platforms and guyed towers. These structures are often subjected to wind loading which cause pullout forces much greater than the weight of the structure itself. In addition to wind, marine structures are also hit by wave forces. Thus, enlarged base pile, which gains higher upward capacity than straight shaft pile through bearing on top of the base, is required to resolve this problem.

24 3 1.3 Objective To study the failure mechanism around an enlarged base pile subjected to uplift in dense sand and loose sand. To define the uplift capacity of enlarged base piles embedded in sand. To make comparison between uplift behaviour for enlarged base piles embedded in dense sand and in loose sand. To define the influence of foundation geometry on the uplift behaviour of piles with enlarged bases. To compare breakout factor derived from present tests with those from existing design methods. 1.4 Scope of Study The experimental work is devoted to laboratory small size model test. There are two tests that would be carried out, failure mechanism test and uplift test. Failure mechanism tests are conducted to study the behaviour of failure mechanism around the model pile subjected to uplift. On the other hand, uplift tests are conducted to define the influence of geometric factors on the uplift capacity of piles with enlarged bases in dry Kota Tinggi sand. Enlarged base piles with bell diameter, D b = 75mm, 1mm, 125mm, 15mm, shaft diameter, D s = 3mm, 4mm, 5mm and base angle α = 3, 45, 6 will be tested for uplift capacity. Both tests are carried out in loose sand and dense sand having dry unit weight of 14.8 kn/m 3 and 16.9 kn/m 3 respectively. Comparison is made between uplift behaviour for enlarged base piles embedded in dense sand and in loose sand. Six breakout factor equations formulated from the tests based on base angle are compared with those observed in model tests carried out by other researchers.

25 4 1.5 Significance of Research The experiment is conducted to gain a better understanding of uplift behaviour of enlarged base piles with different size in both loose sand and dense sand. The research carried out take into account the influence of base angle on uplift capacity of enlarged base pile in addition to the existing parameters of base diameter, embedment depth and soil unit weight. This can help to develop an equation provides more accurate estimation of uplift behaviour for enlarged base pile in sand. This fundamental research will increase the possibility and confidence of building piles with enlarged base in sand to resist pullout forces.

26 5 CHAPTER II LITERATURE REVIEW 2.1 Previous Researches carried out on uplift behaviour of pile Large-scale field tests on anchors for transmission line towers conducted by Giffels et al. (196), Ireland (1963) and Adams and Hayes (1967) were early investigations regarding the pullout resistance. In contrast, Majer (1955), Balla (1961), Downs and Chieurzzi (1966), Baker and Kondner (1966), Meyerhof and Adams (1968), Hanna and Carr (1971), Hanna and Sparks (1973), Das and Seeley (1975a, b), Clemence and Veesaert (1977), Andreadis et al. (1981), Sutherland et al. (1982), Murray and Geddes (1987), Ghaly et al. (1991a, b) employed Conventional laboratory model studies at unit gravity to gain a better understanding of their behaviour. On the other hand, Ovesen (1981), Tagaya et al. (1983, 1988), Dickin (1988), Dickin and Leung (199,1992) managed to provide data on anchors and belled piles subjected to stress levels experienced by full scale prototypes based on the development of the centrifugal modelling technique. Theoretical analyses include those of Vesic (1971) based on cavity expansion theory, limit equilibrium analyses of Chattopadhyay and Pise (1986), Saran et al. (1986) and elasto-plastic finite element analyses of Rowe and Davis (1982). Inevitably, these researches have resulted in many different design methods.

27 6 Andreadis and Harvey (1981), who conducted a series of laboratory model and field tests on anchors, developed design charts to determine the uplift resistance of anchors. Meanwhile, Ovesen (1981) suggest a relationship for the uplift capacity of plate anchors based on centrifugal model tests. A series of laboratory model tests on anchors were conducted by Sutherland. From these tests, he concluded that the angle of inclination of the inverted cone slip surface with the vertical is a function of soil friction angle and relative density. By using finite-element analysis and elastic-plastic model, Rowe and Davis (1982) managed to determine the uplift capacity of horizontal strip anchors. Vermeer and Sutjiadi (1985) noticed that the angle of the inverted cone slip surface with the vertical is equal to the dilatancy angle ψ of the soil. In contrast of that, Murray and Geddes (1987) stated that the angle is equal to the soil friction angle. According to Dickin and Leung (199,1992), increase in the angle of base and in diameter ratio result in a decrease in net uplift capacity and failure displacement. 2.2 Failure Mechanism Many of the existing design methods of uplift capacity of enlarged base pile are based on results obtained from small laboratory model test on anchor foundation. Dickin (1988) have summarized a number of existing design method giving the uplift capacity of horizontal anchors in sand. He classified the assumed failure mechanism of most design methods into three categories which are vertical slip-surface model, inverted truncated cone model and curved slip-surface model. These three categories are shown in Figure 2.1. Majer (1955), one of the earliest researchers, assumed the failure mechanism is a vertical slip surface above the anchor. According to him, the uplift capacity is basically the total of weight of soil above the anchor and shear resistance along the perimeter of the vertical slip surface. After that, Balla (1961) come up with different assumption that the slip surface above small model anchors was a tangential curve. According to Balla, the failure surface for shallow footings embedded in dense sand was nearly circular in

28 7 elevation and the tangent to the surface of ground contact was at an angle of approximately 45 -φ /2 to the horizontal. He obtained a reasonable correlation between theory and the results of full scale tests on shallow footings by assuming a circular failure path. Mac Donald (1963), who had done model test in sand, showed that the failure surface was approximately parabolic for shallow depths and the failure plane was approximately vertical for greater depths. The diameter of the cylinder formed was about 1.75 times the base diameter of the footing. The theory developed by Mac Donald (1963) assumed failure for the shallow case is conical with angle of inclination equal to half the angle of internal friction. Meanwhile, failure for the deep case was assumed to be cylindrical with a cylinder diameter of 1.75 times the base diameter. A relationship between the ratio of unit uplift resistance to overburden pressure and the ratio of footing depth to width was demonstrated by Sutherland (1965). This dimensionless empirical relation helped Sutherland to predict full scale behaviour. Spence, B.E. (1965) examined a theory in which shear was mobilized on a cylindrical surface extending only partially to ground level. He realized that the ratio of the cylinder height to base diameter was constant with the ratio of the depth to base diameter by taking into account the full suction and soil weight. Downs and Chieurzzi (1966), who had done field test observations on belled piers, hold on to the concept that the uplift capacity is derived from the weight of the soil in an inverted cone above the bell plus the self weight of the pier. Then, Meyerhof and Adams (1968) suggested a pyramidal shaped slip surface above the anchor based on his observation in laboratory model tests. According to Clemence and Veesaert (1977), the angle θ in the inverted cone slip-surface model is equal to φ / 2 to the vertical, where φ is the angle of friction.

29 8 Figure 2.1 : Assumed failure mechanisms for belled piers subject to uplift loads classified by Dickin (1988) 2.3 Lateral Earth Pressure Coefficients The lateral earth pressure coefficients in uplift recommended by various investigators are presented in Table 2.1. Due to the various assumptions made about the failure mechanisms, the coefficients values differ in case such theories are compared and shear stresses acting in the soil.

30 9 Table 2.1. Assumption made in design methods summarized by Ilamparuthi and Dickin (21) Source Recommended φ (Degree) Value of Remarks Coefficient K K Meyerhof and K u = K varies linearly Adams (1968) with L/D ratio Clemence and K o K = 1 used in Veesaert calculations (1977) Surterland et K o al. (1982) Kulhawy K a to K o Loose - Used stress (1985) K o to 1 Medium - modification 1 to K p Dense - factor to adjust for construction influences Chattopadhyay K a and Pise (1986) Bobbitt and S f K u % of Meyerhof Clemence and Adams (1987) Ghaly et al. (1991) K p = ( ( ) sin δ sin δ ) (1968) values δ =.6-.7 φ for shallow anchors Half-cut model tests (e.g. Dickin and Leung, Ilamparuthi) show strong evidence that failure surfaces above shallow anchors or belled piles in dense sand are slightly curved and extend outwards from the corner of the foundation to the soil

31 1 surface at an angle close to φ/2 to the vertical. Similar findings have been reported by a number of other investigators. Despite these findings, Meyerhof and Adams theory adopts an equivalent K value acting on a vertical plane extending above the anchor. It is similar to the simple vertical slip surface model developed by Majer (1955). It may be noted from Table 2 that the K values for a cylindrical failure surface are higher than those for the truncated/curved failure wedges. For a cylindrical failure wedge, K value which is greater than unity and less than the K p value have been used. On the other hand, K o value has been adopted for the curved failure wedge. 2.4 Existing Uplift Capacity Analysis Theories of designing uplift capacity of enlarged base pile are derived from researches conducted by Majer (1955), Balla (1961), Meyerhof and Adams (1968), Ovesen (1981), Dickin and Leung (199,1992) and so on. Amendment on the design method has been made from time to time by taking into account more factors to produce more complete and accurate design. Generally, enlarged base pile gain uplift capacity through skin resistance between the pile shaft and the soil and bearing on top of the base. The net ultimate resistance of a pile subjected to uplift forces: Q un = Q ug - W Where Q un = Net uplift capacity Q ug = Gross uplift capacity W = Effective weight of pile The net uplift capacity of enlarged base pile can be calculated from formula generated by Meyerhof and Adams (1968) and Das Seeley (1975). Q un = B q A q γl Where B q = uplift factor

32 11 A q = Area of bottom of enlarged base γ = Unit weight of soil above the bottom of the bell L = length of the pile Uplift factor, B q = 2(L/D b ) K u tan φ [ (m(l/d b ) Where K u = Uplift coefficient (.9 for soil friction angle from 3 o to 45 o ) φ = angle of friction of soil m = coefficient of shape factor ] Table 2.2. Values of coefficient of shape factor, m base on soil friction angle, φ (Meyerhof and Adams, 1968) φ m Values of uplift factor, B q are dependent on embedment ratio, (L/D b ). Values of B q become constant when critical embedment ratio, (L/D b ) cr is reached while (L/D b ) cr increases with increased soil friction angle, φ. Figure 2.2 show an relationship between uplift factor B q and embedment ratio (L/D b ) based on angle of friction of soil. Table 2.3. Relationship between critical embedment ratio, (L/D b ) cr and soil friction angle, φ (Meyerhof and Adams, 1968) φ (L/D b ) cr

33 12 Figure 2.2 : Variation of uplift factor, B q with embedment ratio,(l/db) and soil friction angle, φ (Braja M.Das, 1984) 2.5 Balla s Method for Estimating Uplift Capacity A circular failure surface was noted by Balla (1961) in the breaking out resistance of mushroom foundation for pylons. He proposed a solution for the value of the pullout resistance for mushroom foundation. V = G 1 + T v + G 2 Where V = The breaking out resistance G 1 = The bulk weight of soil G 2 = The weight of foundation slab T v = Shear reaction to skin friction Formula to obtain G 1, G 1 = (D-v) 3 γ F 1 (ϕ, λ)

34 13 Where F 1 (ϕ, λ) = factor based on angle of friction of angle of soil and on a coefficient characteristic for the shape of the foundation body and given by λ = ( D-v ) / B γ = bulk unit weight of soil Formula to obtain G 2, G 2 = R 2 π (D-v-m)( γ e - γ) + 1/3 π m (R 2 + R R a + R a 2 ) + R 2 πvγ e Formula to obtain Tv, T v = (D-v) 3 [ (c/v)(1/d v)f 2 (ϕ,γ) + F 3 (ϕ, λ) ] Where F 2 ( φ, λ ) and F 3 ( φ, λ ) are factor depending on the angle of friction and the formal coefficient. Sum of coefficients F1 and F3 Angle of internal friction: φ Figure 2.3 : Coefficients of Breaking out resistance (Balla, 1961)

35 Meyerhof and Adams s Method for Estimating Uplift Capacity resistance, P u. Meyerhof and Adams (1968) proposed a solution for the value of the pullout For a shallow pile: π 2 P u = πβc u D + Sf BγD K u tanφ + W 2 For a deep pile: π 2 P u = πβc u H + Sf Bγ ( 2D H ) HK u tanφ + W (a) (b) Figure 2.4 : (a) Shallow pile (b) Deep pile Where B = diameter of enlarged base c u = average undrained shear strength of soil φ = angle of friction of soil γ = bulk unit weight of soil W = weight of soil and pile in cylinder of diameter B and height D S f = a shape factor K u = coefficient of lateral earth pressure H = maximum height of rupture surface

36 15 be as follows: According to Meyerhof and Adams, values for K u and S f may be assumed to K u = K p tan.67 φ ( where K p = coefficient of passive earth pressure ) S f = 1 + md mh ( with a maximum value, for deep piles, of 1 + ) B B Table 2.4. The coefficient m and shape factors base on angle of friction of soil, φ (Meyerhof, 1968) φ H/B m S f Breakout Factor Simple expressions and design chart were provided by researchers in order to facilitate the determination of the uplift capacity. The net ultimate uplift resistance, Qu is defined as the gross uplift capacity of the piles minus the weight difference between the piles and soil displaced by the piles. It can be expressed in terms of a dimensionless breakout factor. Breakout factors, N U for piles embedded in sand were calculated by using the following definition: N u Qu = γ AL Where Q U = net ultimate uplift resistance A = plan area of pile bell L = depth of embedment. γ = unit weight of soil

37 16 Formulation of the breakout factor proposed by a number of researchers is summarized by Dickin and Leung (199) in Table 2.5. From the table, it is noted that most of the researchers suggest breakout factor as a function of embedment ratio D/B b and soil friction angle. Besides, some methods are mainly for the design of rectangular anchors. To use it in case of circular anchor, diameter b b is taken as equal to that of a square anchor with equivalent width B e having the same area as the circular anchor. Hence, B e = 2 πb b 4. Unique relationship between effective breakout factor and effective embedment ratio is derived for a given sand density. Dickin and Leung (199) also made a comparison between theoretical uplift breakout factors, N U for piers in sand with bell diameter of 1 m and with friction angle of 4 o in Table 2.6. The comparison showed a large variation that exists between calculations from different design methods. Obviously, difficulties arise when it come to deciding which theory gives a reasonable estimation of uplift capacity of enlarged base pile. Table 2.5. Existing design methods summarized by Dickin and Leung (199) Researcher Majer (1955) Balla (1961) Downs and Chieurzzi (1966) Method of Analysis Vertical slip surface Tangentialcurve slip surface Inverted cone slip surface; cone angle with vertical = ϕ Formulation N u = 1+ 2K (L/b b ) tan ϕ where K is the coefficient of lateral stress in soil N u = (F1 + F3) 4 π L b b Where F1 and F3 are dependent on ϕ and γ and obtained from chart provided by author N u = 1+ 2 L b b 2 4 tanϕ + 3 L b b 2 tan 2 b ϕ + b s b 2

38 17 Meyerhof and Adams (1968) Clemence and Veesaert (1977) Andreadis and Harvey (1981) Ovesen (1981) Rowe and Davis (1982) Surterland et al. (1982) Vermeer and Sutjiadi (1985) Murray and Geddes (1987) Pyramidalshaped slip surface Inverted cone slip surface; cone angle with vertical = φ / 2 Derived from model and field tests on anchor. Derived from centrifugal model tests on horizontal anchor plates Finite-element analysis giving uplift capacity of strip anchors Inverted cone slip surface; cone angle is a function of ϕ Inverted cone slip surface; cone angle = ψ Inverted cone slip surface; cone angle = ϕ L L Nu = 2 K tan + u ϕ m bb bb Where K u =.9 for 3 o < φ <45 o and m is the shape factor, dependent on ϕ N u = 2 2 L φ 2 ϕ 1 L 1 L ϕ 1 + tan + 4K + o tanϕ cos tan b b 3 b 2 where K o is the coefficient of lateral earth pressure. Chart of N u in terms of ϕ and L/b b provided by authors L N u = 1 + (4.32 tan ϕ ) B e 2 BBe = ( π / 4) b b 1.5 where N u = F γ R ψ R K R R where F γ is a function of ϕ and L/bb, Rψ is a function of ψ and L/bb, R k and R R may be taken as unity; charts are provided by the authors. 2 N u = 8 L L tan 2 4 tan α + α bb b b 2 2 where α =.25 [ I ( 1 + cos ϕ) + 1+ sin ϕ]ϕ L N u = B tanϕ cosϕ e critical state friction angle N u = 1 + L B D cv π tanϕ 2+ 3 e B e where ϕ cv is the L tanϕ

39 18 Notes: N u = breakout factor; L = embedment depth; b b = diameter of bell; ϕ = peak friction angle; ψ = dilatancy angle; γ = soil unit weight; I D = soil density index. Table 2.6. Comparison between theoretical uplift breakout factors, N U for piles in sand with bell diameter of 1 m and with angle of friction of 4 o Research L/b b Remarks source Majer (1955) K =K =.4 Balla (1961) Assume F 1 + F 3 =.6 For L/b b = 5, 7 Downs and Assume b s = Chieurzzi (1966) Meyerhof and K u =.9; m =.35 Adams (1968) Clemence and K =.4 Veesaert (1977) Ovesen (1981) BBe =.886 m Andreadis and Harvey (1981) Rowe and Davis (1982) Surterland et al. (1982) Vermeer and Sutjiadi (1985) Murray and Geddes (1987) Extrapolated results For L/b b = F γ =1.8 =1.5 R ψ S f = F γ =3.2 =1.16 R ψ S f =2.5 γ 18.7 F =4.8 =1.3 S f =3 R ψ R k = R R = 1 ψ = 17 o BBe =.886 m I D = BBe =.886 m φ cv = 25.6 o (ψ = 17 o ) BBe =.886 m

40 Influence of Foundation Geometry on Breakout Factor Dickin and Leung (199) conducted a laboratory study in the centrifuge simulating the uplift behaviour of belled pier in dry sand. The centrifugal model tests were employed to investigate the influence of foundation geometry on the uplift capacity of enlarged base piles in sand. The geometry of an enlarged base pile include shaft diameter b s, bell diameter b b, embedment ratio L/b b and bell angle α. Besides, the effects of sand properties such as soil unit weight and density index were also investigated. The centrifuge tests conducted by Dickin and Leung (199) suggest that the breakout factor decreases when bell angle increase for 1 m diameter belled piers in dense and loose sand. The influence of bell angle on breakout factor is small for α values less than 62 o but it decreases obviously for steeper angles as seen in Figure 2.5. This observation match the finding of Dewaikar (1985) who suggested a negligible influence of bell angle for α values between 15 and 6 o from his finiteelement analysis. The unit weight of sands may differ due to a variety of grain shape, size and grading. Thus, an alternative relationship based on density index I d shown in Figure 2.6 is more useful for design purposes. Figure 2.7 shows the influence of diameter ratio, b s /b b on breakout factor for enlarged base pile embedded at embedment ratio, L/b b = 4 with a 45 o bell angle. It was observed that the breakout factor decreases steadily with increased diameter ratio. The variation of breakout factor with embedment ratio for 1 m diameter belled piers in dense (γ = 16 kn/m 3 ) and loose (γ = 14.3 kn/m 3 ) sand is given in Figure 2.8.

41 2 Figure 2.5 : Variation of breakout factor N u with bell angle α in centrifuge uplift tests (Dickin, 1988) Figure 2.6 : Design curve for belled piers in sand in centrifuge uplift tests (Dickin and Leung, 1992)

42 21 Figure 2.7 : Variation of breakout factor N u with diameter ratio, b s /b b in centrifuge uplift tests (Dickin, 1988) Figure 2.8 : Variation of breakout factor N u with embedment ratio L/bb for 1 m diameter belled piers in sand (Dickin and Leung, 199)

43 Buckingham Pi Theorem Buckingham ' s Pi theorem states that if there are n variables in a problem and these variables contain m primary dimensions (for example M, L, T), the equation relating all the variables will have (n-m) dimensionless groups. In other words, if an equation involve n variables is dimensionally homogeneous, it can be reduced to a relationship among (n-m) independent dimensionless groups where m is the minimum number of reference dimensions required to describe the variables. Buckingham referred to these groups as π groups. In order to perform dimensional analysis, one must first list all of the variables that are defining a problem. These variables include dimensional and non dimensional constants. They must be expressed in terms of basic or primary dimensions such as mass, length, and time or force, length, and time. Finally, the pi groups can be written in relation to one another to describe the problem in the form of π l = f(π 2, π 3,.. π n-m ). In this case, Π1 would contain the dependent variable. The actual functional relationship, f, must be determined by experiment. If a relation subsists among any number of physical quantities of n different kinds with Q 1,Q 2, Q n represent one quantity of each kind and the remaining quantities of each kind are specified by their ratios r, r,, and so on, the relationship can be described in the form of π l = f(π 2, π 3, π n-m, r, r ).

44 23 CHAPTER III METHODOLOGY 3.1 Scope A semi-cylindrical model pile with a 1mm diameter base, a 5 mm diameter shaft, a 45 o base angle and a length of 3 mm is embedded in both loose and dense sand. The model is buried in sand to a depth of 3 mm to give an embedment ratio of 3. Failure mechanism around the pile when it is subjected to successive uplift load will be observed. Enlarged base model piles having shaft diameter, D s = 3mm, 4mm and 5mm, base diameter, D b = 75mm, 1mm, 125mm and 15mm and base angle, α = 3, 45 and 6 are used for uplift test. The pile models are tested at a range of embedment ratio from 1 to 5. Uplift tests for these pile models are conducted in b oth loose sand packing and dense sand packing. The results of test are categorized into three categories base on base angle. A summary of the test programme divided into three categories is shown below in Table 3.1. Table 3.1. Summary of uplift tests conducted for the various combinations of parameters Bell diameter α = 3 o α = 45 o α = 6 o 75 mm, 1 mm, 125 mm, 15 mm Shaft diameter 3 mm, 4 mm, 5 mm Embedment ratio 1, 2, 3, 4, 5

45 24 Sand Packing Loose, Dense 3.2 Preparation of Apparatus 1. Sand Container (Uplift Test) An empty tank with side glass walls is used as a container for sand. The tank has an area of 6 mm x 6 mm and a depth of 1 mm. Scale with vertical intervals of 1 cm is marked on the tank. 2. Glass box (Failure Mechanism Test) 6 mm in length, 25 mm in width and 4 mm in depth Scale with vertical intervals of 2 mm is marked on the tank. 2. Model Piles A concrete mix of cement, sand and coarse aggregate is prepared at proportion of 1:2:4 according to the weight of the materials. The concrete was then poured into the moulds of enlarged base pile. The concrete was compacted. The model piles were cured for 28 days. 3. Tools Compactor LVDT (Linear Variable Displacement Transducer) Low speed Motor Data logger Control Box Sand Rainer Load Cell (5 kn)

46 25 Figure 3.1 : Hand-held compactor used for compacting sand Figure 3.2 : From left to right : load cell, linear variable displacement transducer (LVDT)

47 26 Figure 3.3 : Datalogger used for recording and printing out experimental data Figure 3.4 : Control box used for adjusting the speed of motor

48 27 Figure 3.5 : Sand rainer used in sand-rain technique Figure 3.6 : Enlarged base pier model pile with base angle, α = 3, α =45 and α = 6

49 Method Used to Obtain Loose Sand Packing Loose sand packing is obtained using sand-rain technique. The sand is deposited into tank from a height of 33 mm through a sand rainer. Figure 3.7 shows the variation of density index and dry unit weight with falling sand height. At a height of 33 mm, an average unit weight of γ d =14.8kN/m 3 is achieved. Sand Passing Sieve 2.36mm sieve size Density Index, I d Density Index Dry unit weight of soil Density Index Dry unit weight of soil, (kn/m 3 ) Height, H (cm) Figure 3.7 : Variations of density index and dry unit weight with falling sand height for sand passing 2.36mm sieve size (Chosun, 1999) 3.4 Method Used to Obtain Dense Sand Packing Dense sand packing is achieved through compaction of each 75 mm layer of sand using hand held compactor. Each layer of sand is compacted for 1 minutes to attain an average unit weight of γ d =16.9kN/m 3. Figure 3.8 show the variation of density index and dry unit weight with duration of mechanical compaction for each layer.

50 29 Sand Passing Sieve 2.36mm Sieve Size Density Index Dry unit weight of soil 17.1 Index Density, I d Dry unit weight of soil, (kn/m 3 ) Time, T (minutes) 16.6 Figure 3.8 : Variation of density index and dry unit weight with duration of mechanical compaction of sand passing 2.36mm sieve size (Chosun, 1999) 3.5 Uplift Test Procedures 1. The model pile is gripped firmly by steel ring connected to the pulling arm. 2. The model is lowered into the tank slowly and held vertically with the aid of spirit level. 3. The sand is deposited into the tank following sand placement technique of loose sand packing. 4. After the required depth of embedment is achieved, LVDT is placed at the top of the pile to measure the vertical displacement. 5. Load cell is mounted on the pulling arm and connected to data logger. 6. Motor and data logger is turned on. 7. Readings are taken at interval of 1 seconds and printed out by the data logger. 8. Pile is assumed to have encountered failure when peak value of uplift

51 3 load has reached. 9. The steps from 1 to 8 are repeated but sand placement technique of dense sand packing is used instead of loose sand packing. 1. The steps from 1 to 9 are repeated by using model piles with different parameter. Low Speed Motor Steel Rod Load Cell Data logger LVDT Pulling Arm Sand Pile Model Control Box Figure 3.9 : Schematic diagram for uplift tests

52 Failure Mechanism Test Procedures 1. The semicylindrical model pile is attached to a strand connected to motor winch. 2. The model is lowered into the tank slowly and kept in close contact with the glass wall of the tank. 3. It is held vertically with the aid of spirit level. 4. The sand is poured into the tank following sand placement technique of loose sand packing mm-thick horizontal layers of dyed sand are formed at 2 mm vertical intervals to enable failure mechanism in the sand mass around the model to be examined. 6. The model is pulled out at a constant rate of.6 mm/min. 7. The successive failure pattern is observed and photographed. 8. The steps from 1 to 7 are repeated by using sand placement technique of dense sand packing. Pulley 4mm 5mm Sand surface Steel angle 3mm dyed sand layer at an interval of 2mm semi-cylindrical model pile Loading cable Control box Motor 6mm 2m 2mm Figure 3.1 : Schematic diagram for failure mechanism tests

53 32 Figure 3.11 : Uplift test being carried out on model pile Figure 3.12 : Failure mechanism test set-up in the laboratory

54 Determination of Test Parameters The variables involved in the experiment are Q u, D b, D s, L, γ, α and I d. Fundamental dimensions are force and length. The variables measured by fundamental dimensions are Q u, D b, D s, L and γ. Meanwhile, α is dimensionless and I d is the ratio of Δ γ which is (γ -γmin) / (γ max - γ min ). Table 3.2 shows the physical and fundamental quantities involve in the relationship. In this case, there is m=2 fundamental dimensions required for measuring n = 6 kinds of quantity. According to G.A. Vignaux and J.L. Scott, different fundamental dimensions can be ascribed to different directions of length labelled as x, y, and z by using further dimensional analysis called ``vectorised'' or ``directional'' DA. Therefore, embedment length, L would have dimension Lz. Base Diameter, D b has an overall dimension of a single length, L but it shares dimensions along the x and y axis equally since it is measured horizontally. Thus, it is given a dimensional formula of L.5 x L.5 y. Volume, v would have dimensions L z L x L y rather than L 3. The equation relating all the variables have n-m = 4 dimensionless groups. Buckingham referred to these groups as π groups. Since π is a dimensionless group, the dimensions of each π are F L. The final equation obtained is in the form of : f (π 1, π 2, π 3, π 4 ) =. D b, L and γ have been chosen among the available variables, as the repeating variables. Total number of variables in each π is (3 + 1) = 4. Base on Buckingham s pi theorem, four sets of dimensionless products were developed using simulation of appropriate physical quantities. π 1 = D b -2 γ -1 L -1 Q u = Q γd u 2 b L π 2 = D b -1 D s = D D s b π 3 = α π 4 = I d Qu D s Hence, f (, 2, α, Id ) = γd L D b b

55 34 Q γd u 2 b L D s = f (,α, Id ) D b Q u = D 2 D s b Lγ f (,α, Id ) D b Where : Q u = net ultimate uplift capacity D b = diameter of the base D s = diameter of the shaft L = length of pile embedment γ = effective unit weight of sand α = angle of base I d = density index of sand Table 3.2. Physical quantities assumed to be involved in establishing the limiting uplift carrying capacity of uplift loaded enlarged base pile Material properties Pile Symbol Units Fundamentals Quantities Shaft diameter D s mm L Base diameter D b mm L Base angle α Length L mm L Uplift load Q n kn F Sand Bulk Unit Weight γ kn/m 3 FL -3 Density Index I d - - o -

56 35 CHAPTER IV RESULTS AND ANALYSIS 4.1 Failure Mechanism Studies The failure mechanism tests for enlarged base pile embedded in loose and dense sand were carried out in the glass box with 6 mm length, 25 mm width and 4 mm depth. A semi-cylindrical model pile having 1mm diameter base, 5 mm diameter shaft and 45 o base angle was buried in sand to a depth of 3 mm to achieve an embedment ratio of 3. Failure mechanism around the pile when it is subjected to successive uplift load will be observed. Figure 4.1 to Figure 4.4 illustrate failure mechanism developed around the pile in loose sand. On the other hand, sand movement due to vertical displacement of pile in dense sand can be observed in Figure 4.5 until Figure Failure Mechanism Studies for Enlarged Base Pile in Loose Sand Failure mechanism around the model in loose sand is not evident because the movement is only restricted to the pile-sand interface. The front view of semicylindrical model pile in loose sand before subjected to uplift load is illustrated in Figure 4.1. Figure 4.2 shows the little movement of sand at the toe of pile corresponding to the successive vertical displacement of the model. Once the running motor started to pull out the pile, passive zone was formed above the

57 36 enlarged base due to compression of sand. Arrow 1, 2 and 3 shows upward movement of dyed sand above the base. A void was formed at the bottom of the enlarged base with subsequent uplift movement. Further uplift movement of the pile pushed the sand mass upward as shown in Figure 4.3. The sand around the base of the pile collapsed to fill the void created underneath the pile. This can be seen in Figure 4.4. It is clear that pullout resistance is more governed by skin friction rather than shear resistance of sand. However, the height of skin friction affected is only about 2-3 times the shaft diameter above the bell. Due to loose condition, the low density of sand reduces the uplift resistance contributed by the weight of soil within the failure zone above the base. As a result, the peak uplift capacity of enlarged base pile in loose sand is lower than that in dense sand. Figure 4.1 : Front view of semi-cylindrical model pile before subjected to uplift load

58 Figure 4.2 : Early movement of sand mass at the toe of the pile formed void at the bottom of the pile buried in loose sand Figure 4.3 : Further movement of the pile pushed the sand mass upward

59 38 Figure 4.4 : Localize failure for enlarged base pile in loose sand Failure Mechanism Studies for Enlarged Base Pile in Dense Sand The front view of semi-cylindrical model pile in dense sand before subjected to uplift load is illustrated in Figure 4.5. Figure 4.6 illustrates early movement of sand mass at the toe of the pile that formed void at the bottom of the pile buried in dense sand. While the pile was continuously pulled out at a constant rate of.6 mm/min, the rupture planes are formed gradually above the base. The failure surfaces were seen much clearer in Figure 4.7 with further uplift movement of the pile. Arrow 1 and 2 shows upward movement of dyed sand above the base. The movements of sand involved are relatively higher than those associated with peak uplift load in uplift tests due to larger displacement. Figure 4.8 shows failure mechanism around the pile at a displacement of approximately 6 mm where sand around the base of pile collapsed to fill the void created. Active zone and passive zone were formed as shown in the figure. Failure surface above model pile

60 39 in dense sand was slightly curved and extend outwards from the corner of the foundation to the soil surface at an angle close to φ/2 to the vertical. This is similar to that reported by several other researchers such as Meyerhof and Adams (1968) and Rowe and Davis (1982). Balla (1961) also observed the similar failure mechanism for semi-cylindrical models of mushroom foundations for pylons. The failure surface in dense sand tends to diverge in the form of rupture plane. In contrast, failure zone in loose condition is more localized since most of the uplift resistance is contributed by skin friction. The pullout resistance of pile in dense sand is not governed by skin friction as in the case of loose sand. The collapse of sand below the base in dense condition when void was created due to continuous vertical displacement is not as evident as what it did in loose condition. This is because of the close contact and smaller void between sand particles. The model pile in dense sand was subjected to higher uplift load compared to loose sand and took longer time to have failure mechanism developed around it. Figure 4.5 : Front view of semi-cylindrical model pile before subjected to uplift load.

61 4 1 1 Figure 4.6 : Early movement of sand mass at the toe of the pile formed void at the bottom of the pile buried in dense sand 2 2 Figure 4.7 : Vertical uplift of the pile pushed the sand mass upward further and formed rupture planes above the base

62 41 A A B B A : Active Zone B : Passive Zone Figure 4.8 : Failure mechanism for enlarged base pile in dense sand 4.2 The Influence of Embedment Ratio on Uplift Capacity Embedment ratio is the most important factor that influences the uplift capacity of enlarged base pile. This can explain why almost all the existing design methods take into account the effect of embedment ratio. The uplift resistance of enlarged base pile increase with an increase of embedment ratio. Variation of uplift load with vertical displacement for pile with 1mm base diameter, 3mm shaft diameter and 3 o base angle buried in dense sand and loose sand in a range of L/Db = 1 to L/Db = 5 is given in Figure 4.9 and Figure 4.1 respectively. All enlarged base piles embedded in sand ranging from L/Db = 1 to L/Db = 5 reach their own maximum uplift load almost at the same vertical displacement but it is evident that uplift load corresponding to vertical displacement varies a lot based on the embedment ratio. Uplift load required to pull out the enlarged base pile increase with an increase of embedment ratio.