First published 2017 in Great Britain and the United States by ISTE Ltd and John Wiley & Sons, Inc.

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms and licenses issued by the CLA. Enquiries concerning reproduction outside these terms should be sent to the publishers at the undermentioned address:

ISTE Ltd

27-37 St George’s Road

London SW19 4EU

www.iste.co.uk

John Wiley & Sons, Inc.

111 River Street

Hoboken, NJ 07030

USA

www.wiley.com

The rights of Alain Puech and Jacques Garnier to be identified as the authors of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act 1988.

Library of Congress Control Number: 2017948079

British Library Cataloguing-in-Publication Data

A CIP record for this book is available from the British Library

ISBN 978-1-78630-198-7

Table of Contents

Cover

Title

Foreword

Preface

List of Symbols

1 SOLCYP Project

1.1. Motivations
1.2. The SOLCYP project
1.3. Content and nature of this book
1.4. Regulatory context
1.5. Bibliography

2 Scope and Field of Application of Recommendations

2.1. Variable loading and cyclic loading
2.2. Structures to which this discussion pertains
2.3. Effects of cyclic loading on the foundations
2.4. Types of piles
2.5. Types of soils
2.6. Bibliography

3 Cyclic Loading

3.1. General
3.2. Characterization of cyclic loads
3.3. Taking account of real cyclic loading in the design process
3.4. Bibliography

4 Introduction to Cyclic Degradation

4.1. Introduction
4.2. Cyclic degradation of soil properties
4.3. Cyclic degradation of soil–pile interfaces
4.4. Cyclic degradation of pile response
4.5. Appendices
4.6. Bibliography

5 SOLCYP Design Strategy

5.1. General methodology
5.2. Knowledge of loads
5.3. Analysis of regulatory loads
5.4. Criteria of cyclic severity for axial loads
5.5. Cyclic severity criteria for transverse loading
5.6. Detailed characterization of the cyclic loads
5.7. Cyclic pile design methods
5.8. Obtaining the parameters
5.9. Bibliography

6 Behavior of Piles Subject to Cyclic Axial Loading

6.1. Introduction
6.2. Large international programs
6.3. Tests in clay soils
6.4. Tests in sands
6.5. About the static load-bearing capacity
6.6. Summary
6.7. Appendix: cyclic loading tests on piles at the Merville site
6.8. Bibliography

7 Design of Piles Subjected to Cyclic Axial Loading

7.1. Introduction
7.2. General principles
7.3. The NGI approach
7.4. The ICL approach
7.5. The RATZ-CYCLOPS suite of programs
7.6. The SCARP program
7.7. Finite Element Method approaches
7.8. The SOLCYP approach for non-cohesive soils
7.9. Bibliography

8 Behavior of Piles Subject to Cyclic Lateral Loading

8.1. Soil–pile interaction under lateral loading
8.2. Main experimental data
8.3. Available data on the effect of the cycles
8.4. Contribution of the SOLCYP project
8.5. Data obtained on the effect of cycles
8.6. Final overview of the data on the effect of cycles
8.7. Bibliography

9 Design of Piles Subject to Cyclic Lateral Loading

9.1. Recap of the current rules
9.2. Methodology to take account of cyclic loads
9.3. Taking account of the cycles by the global method SOLCYP-G
9.4. Taking account of cycles by a local method SOLCYP-L
9.5. Domains of validity and example of application
9.6. Conclusion
9.7. Bibliography

10 Determination of Cyclic Parameters for Pile Design

10.1. Introduction
10.2. Parameters for the design of piles subjected to cyclic loads
10.3. Obtaining the parameters for the design of piles subjected to cyclic loading
10.4. Bibliography

11 Recommendations for Testing Piles Under Cyclic Loading

11.1. Introduction
11.2. Reasons for the tests
11.3. The different test methods
11.4. Contribution of calibration chamber tests: Axial loading
11.5. Contribution of centrifuge tests: axial or transverse loading
11.6. In situ axial loading tests
11.7. Transverse loading tests in situ
11.8. Appendix 1: Recap on scaling effects
11.9. Appendix 2: In situ axial loading
11.10. Appendix 3: Transverse loading in situ
11.11. Bibliography

Index

End User License Agreement

List of Tables

4 Introduction to Cyclic Degradation

Table 4.1. Relative stiffness of test piles (see references in Chapter 6)
Table 4.2. Relative stiffness of typical bored piles
Table 4.3. Relative stiffness of typical driven piles
Table 4.4. Direct sand–structure shear tests. SOLCYP project

5 SOLCYP Design Strategy

Table 5.1. Distribution of waves in a North Sea storm for varying durations [HAN 80]

6 Behavior of Piles Subject to Cyclic Axial Loading

Table 6.1. Main international programs of cyclic loading in situ tests on piles
Table 6.2. Susceptibility of piles to cyclic degradation

7 Design of Piles Subjected to Cyclic Axial Loading

Table 7.1. Input to the SOLCYP-DEG module (cycle 1) and output from the SOLCYP-DEG module to the control elements

8 Behavior of Piles Subject to Cyclic Lateral Loading

Table 8.1. Theoretical boundaries between rigid piles and flexible piles according to different authors
Table 8.2. Main testing programs conducted recently on the behavior of piles subject to cyclic lateral loading
Table 8.3. Overview of values put forward for the coefficient in the case of sandy soils
Table 8.4. Values of the coefficient α for a flexible pile in accordance with expression [8.5], as proposed by [LIN 99]
Table 8.5. Coefficient m of the power law (for studies conducted on sandy soils)
Table 8.6. Factors affecting the value of the degradation parameters α or m (case of sandy soils)
Table 8.7. Properties of prototype piles simulated in centrifuge model tests
Table 8.8. Conventional limit loads H_lim of the different piles tested
Table 8.9. Values obtained for the coefficient Np by different authors for SOC saturated clay
Table 8.10. Main data concerning the factors which may affect the values of the coefficients α or m in expressions [8.33]

9 Design of Piles Subject to Cyclic Lateral Loading

Table 9.1. Examples of determination of the limit load H_lim
Table 9.2. Values of the stiffness coefficient CR involved in expression [9.10]
Table 9.3. Case of sands of medium or high density. Expression of the coefficients r_c (P-multipliers) to be applied to the reaction P of the static P–y curves (valid if H_max < H_lim)
Table 9.4. Domain of application of the global method SOLCYP-G for sands
Table 9.5. Domain of application of the local method SOLCYP-L for sands
Table 9.6. Domain of application of the global method SOLCYP-G for clays
Table 9.7. Characteristics of the pile discussed in the example
Table 9.8. Thickness and depth of the different layers of soil
Table 9.9. Equivalent cyclic loading sequences
Table 9.10. Values of the P-multipliers r_c with the data of the example discussed
Table 9.11. Comparison of the results on the pile response under H_max = 3500kN using the different approaches (SOLCYP-G and SOLCYP-L)

10 Determination of Cyclic Parameters for Pile Design

Table 10.1. Soil parameters for the design of piles under monotonic and cyclic axial loading
Table 10.2. Soil parameters for the design of piles under monotonic and cyclic lateral loading
Table 10.3. Lab tests to determine the parameters needed for cyclic design of piles

11 Recommendations for Testing Piles Under Cyclic Loading

Table 11.1. Characteristics of the different methods for testing piles subjected to cyclic loading

List of Illustrations

1 SOLCYP Project

Figure 1.1. Installation of driven piles and bored piles on the Merville site
Figure 1.2. Setup for two-way cyclic tests
Figure 1.3. Pile test under axial loading in the IFSTTAR centrifuge
Figure 1.4. Calibration chamber at the 3S-R Laboratory
Figure 1.5. Calibration chamber at the Laboratoire Navier-Cermès

2 Scope and Field of Application of Recommendations

Figure 2.1. Periods and numbers of cycles characteristic of typical cyclic events

3 Cyclic Loading

Figure 3.1. Definitions for regular cyclic loading
Figure 3.2. Example of load induced on top of a monopile by a wind turbine in operation (calculated)
Figure 3.3. Extreme axial loads transferred to the head of the three foundation piles of an offshore wind turbine (tripod structure). Waves with a return period of 50 years
Figure 3.4. How to take account of cyclic loading when designing foundations
Figure 3.5. Definitions for real cyclic loading
Figure 3.6. Example of a raw signal and the same signal processed by the “Cascade” tool
Figure 3.7. Wöhler’s qualitative curve, or S–N curve, for a given material, and application of Miner’s rule for a series of cycles (S is the stress and N_f the number of cycles corresponding to failure)
Figure 3.8. Equivalent uniform cycles method
Figure 3.9. Contour diagram shear strains in Drammen clay [AND 88]
Figure 3.10. Example of application of the procedure of accumulation of pore pressure (from [AND 88])
Figure 3.11. Law of superposition of strains
Figure 3.12. Tests by [KAG 91] on a carbonate sand; a) volumetric strains; b) residual shear stresses as a function of the number of cycles for different sequences
Figure 3.13. [WIC 05] tests on silica sand; a) sequences of loading applied in triaxial tests; b) overview of cumulative axial strains; c) an example of a “deviatoric stress–axial strain” curve obtained
Figure 3.14. Envelope curves of the measured rotation of the pile under lateral loading with two sequences of loads: increasing (A → B → C) and decreasing (C→ B → A) [LEB 09]
Figure 3.15. Load–displacement curves measured at the head of a pile subjected to lateral loading for sequence L1 and sequence L2 [PER 10]
Figure 3.16. Normalized cyclic displacement as a function of the number of cycles for a pile in sand under axial loading (legend: L: Large, M: Medium and S: Small). Centrifuge tests on scale models in the SOLCYP project

4 Introduction to Cyclic Degradation

Figure 4.1. Typical result of a monotonic simple shear test on Loon-plage sand (SOLCYP, GeM)
Figure 4.2. Triaxial tests on Fontainebleau sand NE 34 (SOLCYP, CERMES); a) Mohr–Coulomb failure line and characteristic threshold, in the case of dense sand; b) effect of density index on the internal friction angle and the characteristic angle
Figure 4.3. Result of a typical undrained cyclic triaxial shear test on Fontainebleau sand NE34 (SOLCYP, CERMES)
Figure 4.4. Cyclic shear resistance curves for Fontainebleau sand N34 for two density indices. N_f = number of cycles to cause liquefaction, defined for 5% axial strain of the sample (SOLCYP, CERMES)
Figure 4.5. Principle of construction of a distortion contour diagram based on two-way simple shear tests
Figure 4.6. Construction of a distortion contour diagram on reconsolidated Merville clay (OCR = 4, two-way triaxial test, SOLCYP, GeM)
Figure 4.7. Principle of construction of an excess pore pressure contour diagram
Figure 4.8. Development of strains and excess pore pressures in a non-symmetrical cyclic shear test
Figure 4.9. Effect of mean shear stress on the cyclic response of a carbonate silty clay (CSS tests). Figure taken from COFS document. In this example, τ max = τ_m + τ_cy = 42kPa)
Figure 4.10. Effect of mean deviator on the cyclic response of Fontainebleau sand (Triaxial tests, I_D = 65%) (SOLCYP, CERMES)
Figure 4.11. Contour diagram for Drammen clay (OCR=1) obtained with the DSS tests [AND 09]
Figure 4.12. Construction of distortion contour diagrams (data from NGI)
Figure 4.13. Iso-distortion diagrams under simple shear for three values of the number of cycles N, Drammen clay
Figure 4.14. Comparison of the contour diagrams for two clays (direct simple shear) (from [JEA 06])
Figure 4.15. Definition of shear moduli
Figure 4.16. Evolution of the cyclic shear modulus Gcy (according to [O’RE 91]
Figure 4.17. Typical evolution of the normalized cyclic shear modulus G_cy/G_max,_N, where G_max,_N is the shear modulus G_max at the cycle N, for saturated soils (according to [VUC 94])
Figure 4.18. Result of a “storm” triaxial test on two normally consolidated clays [HIC 79]; a) illite; b) bentonite
Figure 4.19. Direct shear box at 3SR lab, modified for CNL, CV and CNS tests. Top to bottom: a) assembly diagram; b) overall photo and measurement of forces; c) measurement of soil–rough plate relative displacements
Figure 4.20. Evolution of the shear stress a) and normal displacement b) for a CNL shear test with imposed tangential relative displacement thresholds at constant normal loading of 100kPa (dense sand ID = 90%); [u] and [w] are the normal and tangential relative displacements [SHA 97]
Figure 4.21. Evolution of the shear stress a) and of the normal loading b) for a CV cyclic test with imposed tangential relative displacement thresholds with zero normal displacement after application of the initial normal loading (σ_n= 100kPa) (dense sand, I_D = 90%) [SHA 97]
Figure 4.22. Cyclic direct shear test (CNS type): a) friction degradation; b) drop in normal stress; dense silica sand I_D = 85%; σ_n0 = 150kPa; kn = 500kPa/mm [MOR 02]
Figure 4.23 Diagram of the principle behind the calibration chamber test device at CERMES, showing the instrumented probe-pile, installed in the soil mass
Figure 4.24. Visual representation of the degradation of local friction in a mass of dry NE34 Fontainebleau sand over the first 300 loading cycles (controlled displacement test, [TAL 11])
Figure 4.25. Visual representation of the degradation of local friction in a mass of saturated Speswhite kaolinite over the first 50 loading cycles (controlled displacement test, [MUH 15])
Figure 4.26. Evolution of the stresses and relative displacements over the course of cycles: current and mean cyclic values
Figure 4.27. Position of current cycle in the stress plane, and situation in relation to the classical physical boundaries of the interface (characteristic, critical and peak states), in terms of ratio of the normalized shear by the vertical normal stress (η): a) CNL test, fixed cycle; b) CNS test, inclined cycle and mobile toward the physical boundaries
Figure 4.28. Cyclic CNL tests, σ_ncm0 = 310kPa, Δ_τ = 10kPa, I_D0 = 30% (left) and 90% (right), for different levels of mean cyclic shearing (η_nc0 ≈ 0; 0.5 η_pic ; 0.9 η_pic)
Figure 4.29. High-density CNS test, σ_ncm0= 310kPa, Δ_τ = 10kPa, η_nc0 = 0.30, I_D0 = 90%. k_n = 1000kPa/mm. Failure (critical state) at 2035 cycles
Figure 4.30. Two CNS tests with the same initial normal stress (η_nc0 = 100kPa), same low density (I_D0 = 30%) and same shear amplitude (Δ_τ = 10kPa). Left: high initial normalized shear (η_nc0 = 0.30) and low normal stiffness (k_n = 1000kPa/mm). Right: low initial normalized shear (η_cm0 = 0.15) and high normal stiffness (k_n = 5000kPa/mm). In both cases, failure (critical state) occurs after two cycles
Figure 4.31. Overall response of the soil–pile system for different configurations
Figure 4.32. Typical pile head displacements corresponding to zones A (stable), B (metastable) and C (unstable) in the cyclic stability diagram. These are permanent (irreversible) displacements: z_m = 1/2 (z_max + z_min)
Figure 4.33. Plancoët pile (sands and silts) [PUE 81]
Figure 4.34. NGI pile test on Haga clay (from [KAR 83])
Figure 4.35. Illustration of the principle of a cyclic stability diagram [POU 82]
Figure 4.36. Distribution of axial transfer stiffness K_axial = V/w in the pile (from [MUR 75])
Figure 4.37. Real distribution of soil reactions and simplified model
Figure 4.38. Static P-y curves at different depths (from [ROS 04])
Figure 4.39. a) Examples of cyclic P-y curves observed in a dense sand (I_D = 86%), pile penetration 12m, H_max = 960kN, H_c = 240kN (Rosquoët, 2004); b) examples of cyclic P-y curves observed in a slightly overconsolidated saturated clay, pile penetration 16m, H_max = 350kN, H_c = 200kN [KHE 12])
Figure 4.40. a) Cyclic stiffness at the head of a pile with 16m penetration in a slightly overconsolidated saturated clay [KHE 12]. Definition of the loading and unloading stiffness; b) Cyclic stiffness at the head of a pile with 16m penetration in a slightly overconsolidated saturated clay [KHE 12]. Evolution of cyclic stiffness values with the number of cycles (H_m = 200kN, H_c = 100kN)
Figure 4.41. Evolution of cyclic stiffness at cycle N normalized by the stiffness at the first cycle (from [KHE 12])
Figure 4.42. Pile subject to lateral loading. Effect of relative stiffness of the pile on the pile head displacement, and boundaries between a stiff (short) pile and a flexible (long) pile. From [PER 10]
Figure 4.43. Evolution of [u]_cm and [w]_cm as a function of the number of cycles (N). Cyclic CNL tests with σ’_{n cm} = 60, 120 and 310kPa, Δτ = 10kPa, η_cm ≈1/2η_peak. Fontainebleau sand: a), b) loose sand (I_D0 ≈ 30%); c), d) dense sand (I_D0 ≈ 90%)
Figure 4.44. Evolution of [u]_cm and [w]_cm as a function of the number of cycles (N). Cyclic CNL tests with σ’_{n cm} = 60, 120 and 310kPa, ∆τ = 10kPa, η_cm ≈ 9/10η_peak. Fontainebleau sand (I_D0 ≈ 30% and I_D0 ≈ 90%)
Figure 4.45. Evolution of [u]_cm and [w]_cm as a function of the number of cycles (N). Cyclic CNL tests where σ’_{n cm} = 310kPa and ∆τ = 10kPa, with three levels of η_cm. Fontainebleau sand: a), b) loose sand (I_D0 ≈ 30%); c), d) dense sand (I_D0 ≈ 90%)
Figure 4.46. Influence of the amplitude (∆τ = 10 and 20kPa) on the intensity of [u]_cm and [w]_cm as a function of the number of cycles N. CNL cyclic tests in which σ’_{n cm} = 310kPa, with three levels of η_cm. Fontainebleau sand: a), b) loose sand (I_D0 ≈ 30%); c), d) dense sand (I_D0 ≈ 90%)
Figure 4.47. Influence of the amplitude (∆τ = 20 and 40kPa) on the intensity of [u]_cm and [w]_cm as a function of the number of cycles N. CNL cyclic tests where σ’_{n cm} = 310kPa, η_cm ≈ 1/2η_peak and ηcm ≈ 9/10 η_peak. Fontainebleau sand: a), b) loose sand (I_D0 ≈ 30%); c), d) dense sand (I_D0 ≈ 90%)
Figure 4.48. Influence of the imposed normal stiffness (kn) in CNS interface shear tests where σ’_{n cm0} = 100kPa. Dense Fontainebleau sand (I_D0 ≈ 90%): a) degradation of the normal stress as a function of N; b) evolution of η_cm as a function of N
Figure 4.49. Influence of the imposed normal stiffness (k_n), and of Δτ in CNS interface shear tests where σ’_{n cm0} = 100kPa. Loose Fontainebleau sand (I_D0 ≈ 30%): a) degradation of the normal stress as a function of N; b) evolution of ηcm as a function of N
Figure 4.50. Diagram of separately homogeneous zones in the soil sample (thickness h) placed in the shear box: in contact with the structure, the interface (thickness t); above it, the oedometric “buffer” (thickness h-t)

5 SOLCYP Design Strategy

Figure 5.1. SOLCYP procedure f or design of piles subject to axial cyclic loading
Figure 5.2 SOLCYP procedure for design of piles subject to transverse cyclic loading
Figure 5.3. SOLCYP procedure for analysis of loads in the case of axial or transverse loading
Figure 5.4. Zero-damage lines (here failure line for N_f = 1000) and failure lines at N_f = 100; traction tests; Dunkirk [JAR 12], Fontainebleau [RIM 13], Haga [KAR 85], Loon-Plage and Merville [PUE 13]
Figure 5.5. Example of the use of a fictitious cyclic stability diagram to judge the severity of cyclic loads [JAR 12]. The overall safety factor is equal to 1.5; the isocontour N_f =100 is from the test at Haga; the zero-damage line is drawn from the Dunkirk tests
Figure 5.6. Illustrative cyclic stability diagram
Figure 5.7. Degrees of cyclic severity for a number of foundations
Figure 5.8. Limit values of N or H_c/H_max for a given ratio y_N/y_l (case of a pile in sand)
Figure 5.9. Stability diagrams for a flexible pile in slightly overconsolidated saturated clay [KHE 12]
Figure 5.10. Limit values of N or H_c/H_max for a given ratio y_N/y₁, (case of a pile in a slightly overconsolidated saturated clay)
Figure 5.11. Limit values of N or H_c/H_max for a ratio y_N/y₁

6 Behavior of Piles Subject to Cyclic Axial Loading

Figure 6.1. Haga; number of cycles to failure N_f as a function of V_m and V_c normalized by the ultimate static tension capacity V_ut [KAR 85]
Figure 6.2. Comparison of the axial cyclic capacities of piles at Onsøy and Lierstranda with that of the piles at Haga [KAR 92b]
Figure 6.3. Schematic representation of the variations in undrained shear resistance and shear distortion around the test piles in Haga [KAR 92b]
Figure 6.4. Static and cyclic creep loads for the Cran pile. ż₀ = pile head displacement rate under constant static loading; ż_max = pile head displacement rate under maximum cyclic loading ([PUE 12], personal communication)
Figure 6.5. Cyclic tests conducted by the BRE in glacial clay at Cowden. Criterion of cyclic failure, potentially equivalent to an isovalue line N_f = 1000 [OVE 86]
Figure 6.6. Merville tests. Load–pile head displacement curves on the bored F2 pile. One-way compression loading [BEN 13, PUE 13a]. Only the CC3, CC8, CC9 and CC10 tests generate significant cumulative pile head displacement
Figure 6.7. Merville tests. Cyclic interaction diagram for bored piles under compression; one-way and two-way tests
Figure 6.8. Merville tests. Cyclic interaction diagram for screw piles; one-way compression and tension tests
Figure 6.9. Merville tests. Cyclic interaction diagram for driven piles; one-way tension tests
Figure 6.10. ICP tests in London clay (Canon’s Park). Pile installed by rapid jacking. Effective stress paths on the shaft at three levels, including a cycle of unloading and reloading (ABCDE); from [BON 95]
Figure 6.11. ICP tests on clay from different sites. Normalized stress paths. Drawn from [LEH 92], and updated by [CHO 97]
Figure 6.12. Comparison of the cyclic axial capacities of the piles at Tilbrook Grange and Haga [NOW 92]
Figure 6.13. Overview of interaction diagrams available for clays
Figure 6.14. Plancoët pile (sands and silts). Pile head displacements during a one-way cyclic test [PUE 81]
Figure 6.15. Plancoët pile. Comparison of the creep loads under static and cyclic loading. ż₀ = pile head displacement rate under constant static loading, ż_max = pile head displacement rate under the maximum constant cyclic loading [PUE 82a]
Figure 6.16. Interaction diagram with local failure on the shaft. ICP test at Labenne (dune sand, very loose to loose) and Dunkirk (marine sand of medium to high density). [LEH 92, CHO 97] (ICP = Imperial College Pile)
Figure 6.17. Tests in the large calibration chamber, 3SR, Grenoble. Cyclic interaction diagram of ICP mini-piles installed in NE34 Fontainebleau sand of medium density under a vertical confinement pressure of 150kPa [RIM 12]
Figure 6.18. Tests in the large calibration chamber, 3SR, Grenoble. ICP mini-piles installed in NE34 Fontainebleau sand of medium density. Effective stress paths on the shaft at three levels. Stable zone. 1000 cycles [RIM 12]
Figure 6.19. Tests in the large calibration chamber, 3SR, Grenoble. ICP mini-piles installed in NE34 Fontainebleau sand of medium density. Effective stress paths on the shaft at three levels. Metastable zone. 1000 cycles [RIM 12]
Figure 6.20. Tests in the large calibration chamber, 3SR, Grenoble. ICP mini-piles installed in NE34 Fontainebleau sand of medium density. Effective stress paths on the shaft at three levels. Unstable zone. [RIM 12]
Figure 6.21. Dunkirk tests. Interaction diagram obtained on metallic pipe piles driven into dense sands. Drawn from [JAR 00]
Figure 6.22. Tests at Loon-Plage. Load–pile head displacement relationships. F2 bored pile; one-way tension tests [BEN 13]
Figure 6.23. Tests at Loon-Plage. Load–pile head displacement relationships. F5 bored pile; one-way compression test [BEN 13, PUE 13a]
Figure 6.24. Tests at Loon-Plage: load–pile head displacement relationships; Bored pile F1; one-way tension test [BEN 13, PUE 13a]
Figure 6.25. Loon-Plage tests: load–pile head displacement relationships; Bored pile F4; one-way compression test [BEN 13, PUE 13a]
Figure 6.26. Loon-Plage tests; interaction diagram for bored piles under one-way compression [BEN 13b, PUE 13a]
Figure 6.27. IFSTTAR centrifuge tests on dense N34 Fontainebleau sand; Loon-Plage; Interaction diagram for cast in place piles under one-way compression [GUE 12, PUE 13a]
Figure 6.28. Stub of a cyclic stability diagram for model bored piles, cemented in a shelly sand (tests by [NAU 85])
Figure 6.29. Cyclic stability diagram for bored piles in calcarenite [LEE 91]
Figure 6.30. Metallic pipes driven into dense sand at Dunkirk; a) Effect of time on the tension capacity for fresh loading (IAC “Intact Ageing Characteristics”) and multiple loads (O to E and O’ to D’ paths¹). The capacities are normalized with respect to predictions using the ICP method [JAR 06]; b) comparison with the data from two other sites in sands [RIM 15]
Figure 6.31. Combined effect of time and static preshear on the axial static capacity of piles at Haga [KAR 85]
Figure 6.32. Merville tests. Comparison of load–pile head displacement curves obtained in conventional static tests on B1 (driven), S1 (screwed) and F1 (bored) piles [BEN 13]
Figure 6.33. Merville tests. Load–pile-head displacement curves on the B4 driven pile. One-way tension loading. [BEN 13, PUE 13a]
Figure 6.34. Merville tests. Load–pile head displacement curves on the driven B1 pile. One-way compression loading, [BEN 13a, BEN 13c, PUE 13a]
Figure 6.35. Merville tests. Load–pile head displacement curves on the driven B4 pile. One-way tension loading [BEN 13a, BEN 13c, PUE 13a]
Figure 6.36. Merville tests. Load–pile head displacement curves on the F2 bored pile. One-way compression loading, [BEN 13a, BEN 13c, PUE 13a]
Figure 6.37. Merville tests. Load–pile head displacement curves on the bored screw pile S1. One-way compression loading, [BEN 13a, BEN 13c, PUE 13a]
Figure 6.38. Merville tests. Load–pile head displacement curves on the bored screw pile S2. One-way tension loading, [BEN 13a, BEN 13c, PUE 13a]

7 Design of Piles Subjected to Cyclic Axial Loading

Figure 7.1. Diagram of stresses under axial loading around the shaft of a pile [JAR 91]
Figure 7.2. Cross-section of a fine layer of clay (15 × 40mm), sticking to the wall of the ICP pile at Pentre [CHO 96]. Several residual vertical shear surfaces are visible
Figure 7.3. Contour diagrams showing the mean and cyclic distortions γ_m and γ_cv as a function of the mean and cyclic shear stresses τ_a and τ_cy for a given number of cycles. Drammen clay; a) N=1; b) N=10; c) N=100 cycles [AND 04]
Figure 7.4. Cyclic distortion as a function of the cyclic stress and the number of cycles [AND 04]
Figure 7.5. Response measured and calculated by PAXCY of a test pile in [HAG 90]
Figure 7.6. Degradation in capacity calculated by the ABC method for different penetration depths of the pile
Figure 7.7. Interaction diagram for the pile tests in Dunkirk. Comparison between experimental data and predictions after calibrationof the parameters A, B and C (from [JAR 12a])
Figure 7.8. Scheme of t–z load-transfer relationships in RATZ
Figure 7.9. Fatigue curves calculated by RATZ, compared to lab test data on carbonate sediments [RAN 03]
Figure 7.10. Pile head displacements, CYCLOPS [ERB 10]
Figure 7.11. Some t–z curves at different levels, CYCLOPS [ERB 10]
Figure 7.12. Interface model used in SCARP
Figure 7.13. Degradation of the friction capacity of model piles in sand [POU 88]
Figure 7.14. GOPAL R6 pile. Calibration of the load–pile head displacement curve under monotonic loading, produced by SCARP [DEL 13]
Figure 7.15. F4 pile at Loon-Plage. Calibration of the distributions of forces in the pile subjected to monotonic loading, produced by SCARP [DEL 13]
Figure 7.16. Matlock and Foo’s degradation model, calibrated on the GOPAL and Loon-Plage piles [DEL 13]
Figure 7.17. Poulos’ degradation model calibrated on the GOPAL and Loon-Plage piles [DEL 13]
Figure 7.18. GOPAL R5 pile: degradation of the bearing capacity as a function of the cycles in the SCARP simulations [DEL 13]
Figure 7.19. SCARP simulations: distributions of forces in the GOPAL R5 pile before and after the cyclic test CY2 [DEL 13]
Figure 7.20. SCARP simulations: degradation of friction along the GOPAL R5 pile [DEL 13]
Figure 7.21. SCARP simulations: comparisons between the measured and calculated permanent displacements at the head of GOPAL R5 pile; a) and F2 Loon-Plage pile; b) [DEL 13]
Figure 7.22. Principle of [WIC 04] mixed procedure for design of piles subjected to cyclic loading
Figure 7.23. Interface friction angle as a function of D₅₀ [JAR 05b]
Figure 7.24. Example of a chart of the drop in mean normal cyclic stress usable for the design of piles under axial cyclic loading in non-cohesive soils
Figure 7.25. Evolution of the degradation factor F_deg (F_deg = 1- T_rN/T_ri) as a function of the depth and number of cycles; GOPAL-R5 pile (driven)
Figure 7.26. Evolution of the t–z curves as a function of the depth and of the number of cycles. R5.CY2 test. Depth a) 3.3m and b) 13.8m
Figure 7.27. Principles of construction of the t–z cyclic law TZC
Figure 7.28. Evolution of the cyclic degradation in accordance with the SOLCYP method
Figure 7.29. Evolution of the degradation factor as a function of the depth and of the number of cycles. R5 pile (driven)
Figure 7.30. Evolution of degradation factor as a function of the depth and of the number of cycles. F2 pile
Figure 7.31. Evolution of tension resistance as a function of the number of cycles. R5 pile (driven)
Figure 7.32. Example of a cyclic stability diagram obtained with the SOLCYP method and the TZC law
Figure 7.33. Geometry and FEM mesh
Figure 7.34. F1 test: force–pile head displacement curves, in situ measurements and model
Figure 7.35. F2 TC1 test: force–pile head displacement curve for the first cycle
Figure 7.36. F2 TC1 test: force–pile head displacement, 367 cycles before final pull-out
Figure 7.37. Modeling of the F2 TC1 test: force–pile head displacement, 367 cycles before final pull-out
Figure 7.38. F2 TC1 test: pile head displacement as a function of the number of cycles
Figure 7.39. Distribution of the a) normal stresses and b) shear stresses along the shaft of the pile, at cycles 1 and 367, under the mean nominal load (280kN)

8 Behavior of Piles Subject to Cyclic Lateral Loading

Figure 8.1. Notations pertaining to the components of the cyclic lateral loading
Figure 8.2. Effect of relative stiffness on the pile head displacement and boundaries between rigid (short) pile and flexible (long) pile [PER 10]
Figure 8.3. Real distribution of the soil reactions and simplified model
Figure 8.4. Mechanisms of soil deformation around the pile [MUR 93]
Figure 8.5. Repeated phenomena of detachment and erosion under cyclic loading [BEA 79]
Figure 8.6. Values of the coefficient as a function of the depth coefficient [LIN 99]
Figure 8.7. Pile head displacement as a function of the number of cycles N. Tests at Plancoët on LPCs site [HAD 02]
Figure 8.8. Evolution of the relative displacement y_N/y₁ as a function of the number of cycles N and comparison with a logarithmic law [VER 03]
Figure 8.9. Displacement y_N/y₁ as a function of the number of cycles [KLI 12]
Figure 8.10. Displacement y/B of the pile under lateral loads H applied at different heights above the soil, varying between 2.5B and 17.25B [KLI 14]
Figure 8.11. Effect of a cyclic load on the bending moment for numbers of cycles N = 1, 5 and 20 [REE 75]
Figure 8.12. Evolution of the moment profiles with cycles ([MEI 86] cited by [RAK 09])
Figure 8.13. Evolution of the moment profiles with cycles up to N = 10,000 [HAD 02]
Figure 8.14. Evolution of the moment profiles with cycles up to N = 50 [VER 03]
Figure 8.15. Experimental P–y curves in a stiff OC clay [REE 75]
Figure 8.16. Principle of construction of cyclic P–y curves ([REE 01], cited by [PER 10a])
Figure 8.17. Profiles of mechanical characteristics of the soils determined by in-flight static CPT tests
Figure 8.18. Sensor to measure the force applied to the head and linking devices with the pile
Figure 8.19. Domains of the cyclic lateral loads studied (H_max = H_m + H_c)
Figure 8.20. Lateral force as a function of the pile head displacement for 1000 cycles [RAK 09]
Figure 8.21. Incremental displacement as a function of the number of cycles N [RAK 09]
Figure 8.22. Relation between the coefficient α in relation [8.16] and the ratio H_c/H_max [ROS 04]
Figure 8.23. Evolution of the pile head displacement as a function of the number of cycles N (up to N = 75,000) on a semi-log scale (test H_c/H_max = 1/3)
Figure 8.24. Evolution of the moment profiles with the cycles up to N = 75,000 [RAK 09]
Figure 8.25. Example of cyclic P–y curves at different depths for H_max = 960kN and H_c = 240kN in a dry sand [ROS 04]
Figure 8.26. Examples of upper envelope curve of cyclic P–y curves in a dry sand [ROS 04]
Figure 8.27. Cyclic P–y envelope curves obtained for H_max = 240kN, 480kN and 720kN where H_c/H_m = 0.5 in a dry sand [RAK 09]
Figure 8.28. Mechanisms observed on the surface in the case of SOC saturated clay [KHE 12]
Figure 8.29. Mechanisms observed at the surface in unsaturated OC clay [KHE 12]
Figure 8.30. Different responses of piles in SOC saturated clay [KHE 12]
Figure 8.31. Evolution of accumulation rate δ₁₀₀ as a function of the ratio H_c/H_max in SOC clay [KHE 12]
Figure 8.32. Evolution of the normalized cyclic displacement y_N/y₁ as a function of the number of cycles in SOC clay [KHE 12]
Figure 8.33. Stability diagram for a flexible pile in SOC saturated clay [KHE 12]
Figure 8.34. Evolution of the parameters (k, m) as a function of H_c/H_max for the quasi-stable domain [KHE 12]
Figure 8.35. Evolution of moment profiles with the number of cycles (SOC clay) [KHE 12]
Figure 8.36. Evolution of the parameters μ and p as a function of H_c /H_max (SOC clay) [KHE 12]
Figure 8.37. Evolution of the parameters and q with the number of cycles N (SOC clay) [KHE 12]
Figure 8.38. Static P–y curves of SOC saturated clay [KHE 12]
Figure 8.39. Reaction curves under 40 cycles in SOC clay at H_m = 150kN and H_c = 200kN [KHE 12]
Figure 8.40. Examples of cyclic P-y curves at two different depths in SOC clay [KHE 12]
Figure 8.41. Comparison of the logarithmic and power laws for the highest values of α and m observed in sands: α = 0.24 [LIN 99] and m = 0.143 [LON 94]
Figure 8.42. Comparison of the laws obtained by Rosquoët and Rakotonindriana. Pile with 12m penetration depth and 0.72m diameter in a sand of medium density (case of H_max = 720kN and H_c = 240kN)
Figure 8.43. Evolution of the pile displacement with the cycles (case of piles driven into dense sands and for loading of the type H_c/H_max = 0.5)
Figure 8.44. Comparison of the effects of the ratio H_c/H_max on the coefficients α or m (degradation parameters)

9 Design of Piles Subject to Cyclic Lateral Loading

Figure 9.1. Coefficient of reduction in soil reaction P caused by cyclic loading in accordance with the API recommendations in the case of sands [PER 10]
Figure 9.2. Construction of the cyclic P–y curves (soft clay) according to the API recommendations
Figure 9.3. Examples of determination of the conventional limit load H_lim for piles in a dense sand
Figure 9.4. Pile head displacement y/B under the static limit load H_lim for different soil types, pile types and pile head conditions (free or embedded)
Figure 9.5. Example of the effect of pile stiffness E_pI_p on the relative displacement y/B of the head for different values of the lateral load H applied to the head
Figure 9.6. Variation of y’/B as a function of E_pI_p/(E_pI_p)_ref relative to the example in Figure 9.5 (case H = H_lim = 1410kN)
Figure 9.7. Static P–y curve to be chosen for designing a pile under static loading (pressuremeter method) in reference to the French norm NF P 94-262
Figure 9.8. Comparison of the displacement y₂₀₀₀/B (under N = 2000 cycles) obtained by the global method and by the local method for different values of the bending stiffness of the pile (between 0.5(E_pI_p)_fl and 50(E_pI_p)_fl)
Figure 9.9. Comparison of the maximum moment M_max,2000 (under N = 2000 cycles) obtained by the global method and by the local method for different values of the bending stiffness of the pile (between 0.5(E_pI_p)_fl and 50(E_pI_p)_fl)
Figure 9.10. Values of the coefficients r_c as a function of the number of cycles N for sands of medium to high density (case H_c/H_max = 0.5 and H_max/H_lim = 0.5)
Figure 9.11. Limit values of the number of cycles N so that the P-multiplier r_c remains positive (including in the layer z/B < 1.5)
Figure 9.12. Static P–y curves according to NF P 94-262
Figure 9.13. Idealized cyclic loading (maximum load and number of cycles per sequence)
Figure 9.14. Loading curve for the pile with 30m penetration depth and 1.067m diameter
Figure 9.15. Effect of pile stiffness EI on the displacement y₁/B of the head under load H_lim
Figure 9.16. Behavior of pile under the static load H_max = 3500kN
Figure 9.17. Application of cyclic severity criterion for N_rep = 2000 and y_2000/y1 < 1.5
Figure 9.18. Equivalent “cyclic” P–y curves
Figure 9.19. Behavior of the pile under the load H_max = 3500kN after the cyclic sequence
Figure 9.20. Static P–y curves prescribed by API (submerged sand, φ’ = 38° and γ’ = 6kN/m³)
Figure 9.21. Loading curve obtained with the API static P–y curves
Figure 9.22. Effect of stiffness E_pI_p on the pile head displacement under a lateral load of 2740kN (conventional limit load H_lim)
Figure 9.23. Static and cyclic ultimate reactions applicable according to the API recommendations (submerged sand, φ’ = 38° and γ’ = 6kN/m³)
Figure 9.24. Displacement of the pile y_N/B (%) under H_max = 2000kN for N cycles according to the SOLCYP-G method (H_c/H_max = 0.5 or 0.25 and CR = 1.15)

10 Determination of Cyclic Parameters for Pile Design

Figure 10.1. Comparison of the evolution of the cyclic strain during two-way loading tests on different clays, OCR = 4 [HAN 14]
Figure 10.2. Effect of shear velocity on a deepwate clay from the Gulf of Guinea [COL 10]
Figure 10.3. Evolution of the soil–pile interface angle in a ring shear test. Fontainebleau sand. Rough steel
Figure 10.4. Path of consolidation and preconditioning of a sand sample to simulate installation by driving [AGH 15]
Figure 10.5. Effect of the preconsolidation and preconditioning path on the response of a sand sample subjected to 4500 cycles. The paths ABCD and AD are shown in Figure 10.4 (with 48 hours of creep, no cyclic preshearing at point C and CSR = 0.25 for the 4500 cycles) [AGH 15]
Figure 10.6. The different phases of the CNS cyclic direct soil–pile shear test: 1) and 2): application of the initial values of the mean normal stress (σ′_{n cm 0}) and of the mean shear stress (τ_cm); application of the normal stiffness k_n to the interface; 3) application of a series of N cycles of fixed amplitude (τ_cy); if failure is reached (large, uncontrollable displacements): the test is stopped at N_f; 4) if failure is not reached: a final cycle of extremely high amplitude is applied, before the sample is allowed to return to a state of zero stress
Figure 10.7. Different protocols of cyclic tests
Figure 10.8. Pressure–volume curve from a cyclic test carried out on the Loon-Plage site in the SP2 borehole at 8.5m
Figure 10.9. CPT-MPFA cyclic tests; soft clay site, Burswood; Friction on sleeve situated 25 diameters above the tip [HEB 05]
Figure 10.10. CPT-MPFA cyclic tests; carbonate sand site, Ledge Point; Degradation of friction on a sleeve situated 25 diameters above the tip as a function of the number of cycles [HEB 05]

11 Recommendations for Testing Piles Under Cyclic Loading

Figure 11.1. Design support tests. Axial loads. Principle diagram
Figure 11.2. Design support tests. Axial loading. Principle of calibration of the cyclic stability diagram
Figure 11.3. Examples of equipment capable of carrying out tests on reduced-scale models
Figure 11.4. Photo of the setup used in a two-way cyclic loading test
Figure 11.5. In situ axial loading test: a) jack, load cell and four displacement sensors at the head; b) reaction system
Figure 11.6. Instrumentation with strain gages fixed on to: a) the external wall of a driven pile; b) the reinforcement bars in a bored pile
Figure 11.7. Example of instrumentation for determining local laws: removable extensometers for a concrete pile
Figure 11.8. Instrumentation with optical fibers: optical fiber fixed to the casings and poured in concrete (yellow and black wire)
Figure 11.9. In situ transverse test. Setup comprising fixed base, jack, load cell and displacement sensors
Figure 11.10. PISA project: principle of application of the load 8m above the soil to simulate the combined action of swell and wind on a monopile
Figure 11.11. PISA project: bird’s eye view of the setup used to apply cyclic forces of 700 metric tons 8m above the soil; Dunkirk site (photo credit: PISA project, EDF-EN)

Foreword

With the design of new types of structures or infrastructures, linked to the recent developments in the domains of energy, transport and industrial engineering, the insufficiency of the usual methods of design of pile-based foundations, subject to cyclic loading, is deeply felt in the profession. The case of wind turbines, both onand off-shore, is a representative example. The recommendations offered in this book are intended to partly fill this technical and regulatory gap, proposing a methodology and calculation methods to take account of these effects of variable loads in the design of pile foundations. They are based on experiments, both in a laboratory environment and at real scale, and simulations conducted as part of the SOLCYP project.

SOLCYP, in fact, is the natural continuation of the past 40+ years of efforts in numerous research works on the same topic of cyclic loading. This work was initiated and driven forward by Professor Jean Biarez, at the soil mechanics laboratory at the Ecole centrale de Paris, which, in the 1970s, became the first institution in France to implement a vast research program on the cyclic behavior of soils. Two main topics were developed: soil behavior when subjected to seismic loading, mainly in collaboration with EDF, and behavior over very large numbers of cycles, in collaboration with Elf Aquitaine and the Institut français du pétrole (IFP – French Petroleum Institute).

Around the end of the decade, the oil industry’s turn toward exploitation in deep waters (at the time, over 400m of depth) led to the emergence of new platform designs, including TLPs (Tension Leg Platforms), which raised the issue of the piles being able to withstand tension forces – particularly cyclic forces. The tests conducted for the IFP, from 1978 onward on the Plancoët site by the Laboratoire régional des ponts et chaussées (LRPC – Regional Bridge and Roadway Laboratory) in Saint-Brieuc, were the first in situ pile tests to be published. Next came the tests conducted by the BRE (Building Research Establishment, UK), and the Norwegian Geotechnical Institute (NGI)’s first experiments. The Plancoët tests were followed by tests at Cran in soft clay and later by tests at Plouasne in cemented carbonate sands. The tests conducted in France were conducted according to the same framework: they were run by the ARGEMA (Association de recherche en géotechnique marine – Marine Geotechnical Research Association), including oil companies (ELF, TOTAL), contractors and engineering firms in the offshore oil and gas sector (CG DORIS, SEA TANK Co, CFEM, ETPM, etc.), research institutes (IFP and IFREMER, formerly CNEXO). They were financed by partners with a significant public contribution through the FSH (Fonds de soutien des hydrocarbures – Hydrocarbon Support Fund). The CLAROM (Club pour les actions de recherche sur les ouvrages en mer – Club for Research on Marine Structures) later replaced ARGEMA, with the involvement of new actors including TECHNIP, SUBSEA 7, Doris Engineering, SAIPEM and Fugro.

In parallel, research actions were carried out between CLAROM (or the IFP directly) and certain university labs (such as the Ecole centrale de Paris and the 3SR Lab in Grenoble). These actions involved laboratory tests or numerical developments.

Next, we note the creation, in January 1986, of the GRECO (coordinated research group) “Geomaterials”, the “outside the walls” laboratory set up in 1986 by the CNRS and the Ministry in charge of research, which constituted a significant starting point for the structuring and coordination, in France, of theoretical and applied research on geomaterials (soils, rocks and concrete) and geotechnical works. It encapsulated most specialized skills, with personnel from public and private labs, contractors, engineering companies, consultants and owners. The work of the GRECO, which covers a much larger area than piles alone, includes topics such as “soil–structure interfaces” and “modeling of skin friction on piles”, which are at the heart of the SOLCYP project’s subject. It should be noted that the geomaterials GRECO has since become the European lab network “ALERT Geomaterials” (ALERT stands for Alliance of Laboratories in Europe for Education, Research and Technology), which continues to federate and promote scientific exchanges on geomaterials, structures and the environment, both in Europe and beyond. The richness of this academic production is due, in no small part, to the collaboration which was set up between the universities, technical centers, large organizations and design offices.

In light of this brief historical overview, which is by no means intended to be exhaustive, it must be recognized that the SOLCYP project is a natural continuation of the research which has gone before. This continuity is assured not only by the scientific topics dealt with, but also (thus proving the success of the previous programs) by the participation of SOLCYP’s operator partners, all of whom were involved in these previous actions, in terms of tests or laboratory work:

– IFSTTAR and CERMES (LCPC/LRPC), which notably conducted tests in a calibration chamber and in situ tests;
– 3SR, which conducted laboratory tests and calibration chamber tests, and developed theoretical models;
– Ecoles centrales in Nantes and Lyon: laboratory tests and numerical developments;
– Lille University: development of theoretical models.

We must not overlook the continued presence of Alain Puech, the technical director of the SOLCYP project, who, thanks to his professional affiliations (IFP, Geodia and now Fugro), has been a stakeholder in all these works of research.

Comité français de mécanique des sols et de géotechnique

It would be remiss of me to close this foreword without extending grateful recognition to all the actors involved in the project, for the quality of their work and the richness of their exchanges, which it has been my privilege to witness. My gratitude goes first to Alain Puech who, with skill and patience, headed up and brought together a large number of contributors from various backgrounds, who could have had conflicting interests, to complete a vast project, coherent and useful for the profession. My thanks also go to all the partners and organizations having financed the project, who are explicitly named in the preface and Chapter 1.

Alain PECKER
Chairman of the SOLCYP project board

Design of Piles Under Cyclic Loading

SOLCYP Recommendations

Foreword