UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Shear loading response of sand-silt mixtures Soysa, Achala Nishan 2021

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2021_may_soysa_achala.pdf [ 47.36MB ]
Metadata
JSON: 24-1.0395870.json
JSON-LD: 24-1.0395870-ld.json
RDF/XML (Pretty): 24-1.0395870-rdf.xml
RDF/JSON: 24-1.0395870-rdf.json
Turtle: 24-1.0395870-turtle.txt
N-Triples: 24-1.0395870-rdf-ntriples.txt
Original Record: 24-1.0395870-source.json
Full Text
24-1.0395870-fulltext.txt
Citation
24-1.0395870.ris

Full Text

Shear Loading Response of Sand-SiltMixturesbyAchala Nishan SoysaB.Sc.Eng.(Hons), University of Moratuwa, 2009M.Sc., University of Moratuwa, 2011M.A.Sc., The University of British Columbia, 2015A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Civil Engineering)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)February 2021© Achala Nishan Soysa 2021The following individuals certify that they have read, and recommend to the Faculty ofGraduate and Postdoctoral Studies for acceptance, the dissertation entitled:Shear loading response of sand-silt mixturessubmitted by Achala Nishan Soysa in partial fulfillment of the requirements for thedegree of Doctor of Philosophy in Civil EngineeringExamining Committee:Professor Dharma Wijewickreme, Department of Civil EngineeringSupervisorProfessor John Howie, Department of Civil EngineeringSupervisory Committee MemberProfessor Dirk van Zyl , Norman B. Keevil Institute of Mining EngineeringUniversity ExaminerProfessor Tony Yang, Department of Civil EngineeringUniversity ExaminerProfessor Christopher Baxter, University of Rhode Island, United States of AmericaExternal ExaminerAdditional Supervisory Committee Members:Professor Roger Beckie, Earth Ocean and Atmospheric SciencesSupervisory Committee MemberiiAbstractA comprehensive experimental research program was undertaken to investigate the shearloading response of sand-silt mixtures with the objective of addressing the inconsistenciesin the understanding of the mechanical behavior of these soils. Natural sand and siltsoriginating from the Fraser River Delta in British Columbia, Canada was used as parenttest materials to generate the full spectrum of mixtures needed for the study. The workincluded triaxial and direct simple shear tests on specimens reconstituted to achieve pre-selected sand-silt compositions. The reconstituted specimens were prepared using waterpluviation, or modified slurry deposition/consolidation techniques. New experimental pro-cedures to prepare sample, pour slurry, apply vacuum, and initially consolidate the slurrywere developed to equitably accommodate the full range of sand-silt mixture specimens.The effect of fines content was systematically investigated for both monotonic and cyclicshear loading cases. The undrained shear strength mobilized at phase transformation(Su−PT ) of the sand-silt mixtures decreased with increasing fines content (CF ) in triax-ial compression tests. In triaxial extension tests, the Su−PT increased with increasing CF ;however, no noticeable upward or downward trend on the effect of CF was noticeable inthe results from direct simple shear tests. The coarse-grain-based and fine-grain-based voidratios were observed to serve as better indices to examine the shear loading response ofsand-silt mixtures. The cyclic resistance ratio (CRR) decreased with increase in the coarse-grain-based void ratio for predominantly sandy specimens. Similarly, the value of CRRdecreased with increasing fine-grain-based void ratio for specimens having a silt dominantmatrix.The shapes of typical stress–strain loops in the initial stages of constant shear stress am-plitude cyclic loading tests, on both fine-and coarse-grained soils, were noted to be distinctlydifferent from those observable during the later stages. Considering the number of load-ing cycles corresponding to the commencement of the transition point in this stress-strainpattern change as the instance of unacceptable performance, a new shear stiffness–basedcriterion was developed to determine cyclic resistance ratio CRR from cyclic shear tests; thisprovides a more robust engineering basis to determine the CRR than the strain-amplitudebased criteria commonly used in current practice.iiiLay SummarySoil liquefaction (soil behave as a liquid) has caused severe damage to structures and life-lines during past earthquakes and has led to loss of life and disruption to living and workingconditions of citizens. Solid understanding of the mechanical behavior of soils is essentialto predict the response against the forces that will act on structures during both typicalusage and extreme cases (i.e., earthquakes), and in turn, to develop good engineering designpractices to minimize risk of unacceptable performance of structures. Much of the currentresearch focus has been on the liquefaction of sands whereas the seismic behavior of sand-siltmixtures has been investigated only on a limited scale. The present research comprises anadvanced laboratory experimental test program undertaken to improve the current knowl-edge on the behavior of sand-silt mixtures under earthquake loading, particularly, with itsoutcomes contributing to enhancing the seismic safety of built environment.ivPrefaceThis dissertation emerges from a research program conducted at the Department of CivilEngineering, University of British Columbia during the period from 2015 to 2020; and itcontains and details the research work that was undertaken by the author, Achala NishanSoysa, under the supervision of Professor Dharma Wijewickreme.I, Achala Nishan Soysa, was the lead investigator and principle contributor involved inreviewing the literature, developing the research proposal, forming the methodology, per-forming laboratory experiments, generating data, analyzing data and preparing dissertationmanuscript (including all the figures and tables). Dr. Wijewickreme was the supervisoryauthor, and he was involved in advising during the concept development, performance ofexperiments, organizing the content, and final editing of the manuscript.Some parts of this dissertation have already been published in the journals and confer-ences as detailed below:ˆ Wijewickreme, D. and Soysa, A., 2016. Stress-strain pattern-based criterion to assesscyclic shear resistance of soil from laboratory element tests. Canadian GeotechnicalJournal, 53 (9), 1460-1473 doi: 10.1139/cgj-2015-0499– This journal paper includes some parts in Section 2.4 and Chapter 3. Dr. Wijew-ickreme provided the concept and the initial formulation for the investigation. Thedata analysis and scrutinization of stress-strain curves was performed by me that re-sulted in the method for the identification of proposed stress-strain pattern change.This paper presents a new shear stiffness-based criterion, that is proposed to deter-mine cyclic shear resistance (i.e., unacceptable shear performance) in terms of theoccurrence of pattern change of stress-strain loops obtained from laboratory cyclicshear tests as oppose to the practicing criteria based on arbitrary selected thresholdshear strain magnitudes. I drafted the first version of the journal manuscript, and Dr.Wijewickreme reviewed the draft and assisted in addressing the comments from thereviewers.ˆ Wijewickreme D., Soysa, A. and Verma P., 2018. Response of Natural Fine-grainedSoils for Seismic Design Practice: A Collection of Research Findings from BritishColumbia, Canada., Soil Dynamics and Earthquake Engineering. 124, 280-296 doi:10.1016/j.soildyn. 2018.04.053– This journal paper includes some parts in Section 2.1. Priyesh Verma (a fellowPhD student), and I was involved in preparing the first draft of manuscript. Dr.Wijewickreme provided supervision of geotechnical research content and review of themanuscript from technical and editorial points of view.ˆ Soysa, A. and Wijewickreme, D., 2018. Initial observation on laboratory shear load-ing response of sand-silt mixtures. Geotechnical Earthquake Engineering and SoilDynamics V, Austin, Texas, USA. doi: 10.1061/9780784481486.031v– This conference paper includes some parts in Section 2.1, Chapter 5, and Ap-pendix G.1. I preformed the experimental laboratory tests, data analysis, and pre-pared the first draft. Dr. Wijewickreme provided supervision of geotechnical researchcontent and review of the manuscript from technical and editorial points of view. Idelivered the podium presentation at the conference in Texas, USA.ˆ Soysa, A. and Wijewickreme D., 2019. Observations on the effect of strain andloading rate on monotonic and cyclic direct simple shear response of reconstitutedFraser river silts, In proceedings of the 72nd Canadian Geotechnical Conference, Sep29 – Oct 2, 2019, St. John’s, Newfoundland and Labrador, Canada.– This conference paper includes some parts in Appendix G.2 and H.4. I preformedthe experimental laboratory tests, data analysis and prepared the first draft. Dr.Wijewickreme provided supervision of geotechnical research content and review ofthe manuscript from technical and editorial points of view. I delivered the podiumpresentation at the conference in St. John’s NL, Canada.ˆ Soysa, A. and Wijewickreme D., 2019. Monotonic shear loading response of sand-siltmixtures from direct simple shear tests, XVI Panamerican Conference on Soil Me-chanics and Geotechnical Engineering, Cancun, Mexico.– This conference paper includes some parts in Chapter 5. I preformed the experi-mental laboratory tests, data analysis and prepared the first draft. Dr. Wijewickremeprovided supervision of geotechnical research content and review of the manuscriptfrom technical and editorial points of view. Dr. Wijewickreme delivered the podiumpresentation at the conference in Cancun, Mexico.ˆ Soysa, A. and Wijewickreme D., 2020. Cyclic shear loading response of sand-siltmixtures from direct simple shear tests, 17th World Conference on Earthquake Engi-neering, Sendai, Japan.– This conference paper includes some parts in Chapter 6. I preformed the experi-mental laboratory tests, data analysis and prepared the first draft. Dr. Wijewickremeprovided supervision of geotechnical research content and review of the manuscriptfrom technical and editorial points of view.viTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiLay Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiiiList of Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiiiList of Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxviAcknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxviiDedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxviii1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Research Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3 Scope and Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.4 Thesis Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.1 Shear Loading Response of Sand-Silt Mixtures . . . . . . . . . . . . . . . . 52.1.1 Conflicting Findings on the Shear Loading Response of Sand-Silt Mix-tures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.1.2 Remarks on the Conflicting Findings . . . . . . . . . . . . . . . . . 122.2 Effects of Fines Content on the Void Ratio . . . . . . . . . . . . . . . . . . 122.3 Sand-Silt Mixture Specimen Preparation Methods for Laboratory Tests . . 202.3.1 Current Methods for Preparation of Reconstituted Specimens . . . 212.3.2 Challenges in Preparing Sand-Silt Mixture Specimens . . . . . . . . 232.4 Assessment of Cyclic Shear Resistance of Soil from Laboratory Tests . . . 252.4.1 General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252.4.2 Criteria or the Definition of Liquefaction or Cyclic Failure . . . . . 262.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30vii3 A New Criterion to Assess Cyclic Shear Resistance of Soil . . . . . . . 323.1 Assessment of the Potential for a New Stress-Strain Pattern Based-Approach 323.2 Effect of Data Recording Rate . . . . . . . . . . . . . . . . . . . . . . . . . 363.3 Cyclic Stress-Strain Pattern Development for Soil . . . . . . . . . . . . . . 373.3.1 Coarse-grained soils - sand . . . . . . . . . . . . . . . . . . . . . . . 373.3.2 Fine-grained soils - silt and clay . . . . . . . . . . . . . . . . . . . . 393.3.3 Fine-grained soils - mine tailings . . . . . . . . . . . . . . . . . . . . 403.4 Discussion on the Stress-Strain Pattern-Based Criterion . . . . . . . . . . . 443.5 Conclusions on the Stress-Strain Pattern-Based Criterion . . . . . . . . . . 474 Test Devices, Materials and Experimental Program . . . . . . . . . . . . 494.1 Test Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494.1.1 Direct Simple Shear Test Device . . . . . . . . . . . . . . . . . . . . 494.1.2 Triaxial Shear Test Device . . . . . . . . . . . . . . . . . . . . . . . 524.2 Tested Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554.2.1 Fraser River Sand . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554.2.2 Fraser River Silt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554.2.3 Mixtures of Fraser River Sand and Silt . . . . . . . . . . . . . . . . 574.2.4 Test Specimen Identification and Labeling Approach . . . . . . . . 574.3 Test Specimen Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . 574.4 Test Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 614.4.1 Direct Simple Shear Test Procedure . . . . . . . . . . . . . . . . . . 614.4.2 Triaxial Shear Test Procedure . . . . . . . . . . . . . . . . . . . . . 614.5 Experimental Test Program . . . . . . . . . . . . . . . . . . . . . . . . . . 625 Monotonic Shear Loading Response of Sand-Silt Mixtures . . . . . . . 645.1 Monotonic Test Program and Test Repeatability . . . . . . . . . . . . . . . 645.2 Constant-Volume Monotonic Direct Simple Shear Response . . . . . . . . . 745.2.1 Stress-strain response . . . . . . . . . . . . . . . . . . . . . . . . . . 745.2.2 Stress-path and excess pore-water pressure response . . . . . . . . . 765.3 Monotonic Triaxial Response . . . . . . . . . . . . . . . . . . . . . . . . . . 785.3.1 Undrained monotonic triaxial compression and extension response . 785.3.2 Drained monotonic triaxial compression response . . . . . . . . . . 835.4 Discussion and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 855.4.1 Undrained shear strength . . . . . . . . . . . . . . . . . . . . . . . . 915.4.2 Drained shear strength . . . . . . . . . . . . . . . . . . . . . . . . . 945.5 Summary and Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . 966 Cyclic Shear Loading Response of Sand-Silt Mixtures . . . . . . . . . . 986.1 Cyclic Test Program and Repeatability . . . . . . . . . . . . . . . . . . . . 986.2 Cyclic Stress-Strain and Stress-Path Response . . . . . . . . . . . . . . . . 1066.3 Pore-water Pressure Response . . . . . . . . . . . . . . . . . . . . . . . . . 1106.4 Cyclic Shear Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1166.5 Cyclic Pore Water Pressure Development and Shear Stiffness Degradation 1196.6 Discussion of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1346.7 Summary and Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . 139viii7 Conclusions and Recommendations . . . . . . . . . . . . . . . . . . . . . . 1417.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1417.2 Contributions to the Knowledge and Practice . . . . . . . . . . . . . . . . . 1457.3 Suggestions - Future Studies . . . . . . . . . . . . . . . . . . . . . . . . . . 145References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147AppendicesA Supplementary Summary of the Literature Review . . . . . . . . . . . . 166A.1 Previous Laboratory Experimental Studies on the Behavior of Sand-Silt Mix-tures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166A.2 Specimen Preparation Methods . . . . . . . . . . . . . . . . . . . . . . . . 170A.3 Criteria for the Definition of Liquefaction / Cyclic Failure . . . . . . . . . . 172B UBC Direct Simple Shear Test Device . . . . . . . . . . . . . . . . . . . . 179B.1 Loading Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179B.2 Strain Computation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180B.3 Data Acquisition and Control System . . . . . . . . . . . . . . . . . . . . . 180B.4 Device Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181C UBC Triaxial Shear Test Device . . . . . . . . . . . . . . . . . . . . . . . . 183C.1 Loading Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183C.2 Strain Computation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185C.3 Data Acquisition and Control System . . . . . . . . . . . . . . . . . . . . . 186C.4 Device Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187C.5 Membrane Compliance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191D Data Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195D.1 Direct Simple Shear Test Data Reduction . . . . . . . . . . . . . . . . . . . 195D.2 Triaxial Shear Test Data Reduction . . . . . . . . . . . . . . . . . . . . . . 197E Soil Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201E.1 Particle Size Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201E.2 Specific Gravity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203E.3 Soil Plasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203E.4 Material Phase Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204E.5 Scanning Electron Microscope Image Analysis . . . . . . . . . . . . . . . . 207F Uncertainty in Measurements and Derived Parameters . . . . . . . . . . 218F.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218F.2 Uncertainty in Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . 218F.3 Uncertainty in Derived Parameters . . . . . . . . . . . . . . . . . . . . . . 231G Specimen Preparation and Selection of Shear Loading Rate . . . . . . . 235G.1 Specimen Reconstitution Procedure . . . . . . . . . . . . . . . . . . . . . . 235G.2 Rate of Shearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242ixH Preliminary Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246H.1 Assessment of Slurry Deposition Specimen Preparation Method . . . . . . 246H.2 Test Repeatability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250H.3 Identification of Strain Level at Failure . . . . . . . . . . . . . . . . . . . . 253H.4 Selection of Appropriate Rate of Shear Loading . . . . . . . . . . . . . . . 255I Direct Simple Shear Test Results . . . . . . . . . . . . . . . . . . . . . . . . 263I.1 Preliminary Tests - Monotonic Strain Rates . . . . . . . . . . . . . . . . . . 264I.2 Preliminary Tests - Cyclic Loading Rates . . . . . . . . . . . . . . . . . . . 271I.3 100C0F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275I.4 95C05F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282I.5 85C15F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288I.6 75C25F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294I.7 63C37F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301I.8 50C50F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307I.9 40C60F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313I.10 30C70F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320I.11 20C80F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326I.12 10C90F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332I.13 0C100F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338J Triaxial Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345J.1 Preliminary Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346J.2 100C0F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355J.3 95C05F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364J.4 75C25F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371J.5 65C35F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378J.6 50C50F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386J.7 40C60F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395J.8 20C80F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399J.9 0C100F-Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407xList of Tables3.1 Material properties, test conditions, and parameters for coarse-grained ma-terials used in CDSS tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433.2 Material properties, test conditions, and parameters for fine-grained materi-als used in CDSS tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433.3 Material properties, test conditions, and parameters for mine tailings usedin CDSS tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434.1 Summary of the test program . . . . . . . . . . . . . . . . . . . . . . . . . . 635.1 Summary of the monotonic direct simple shear loading test results for thehomogeneous sand-silt mixtures . . . . . . . . . . . . . . . . . . . . . . . . . 665.2 Summary of the monotonic triaxial shear loading test results for the homo-geneous sand-silt mixtures - I . . . . . . . . . . . . . . . . . . . . . . . . . . 675.3 Summary of the monotonic triaxial shear loading test results for the homo-geneous sand-silt mixtures - II . . . . . . . . . . . . . . . . . . . . . . . . . . 685.4 Different types of void ratio definitions in the literature . . . . . . . . . . . 886.1 Summary of the cyclic direct simple shear loading test results for the homo-geneous sand-silt mixtures - I . . . . . . . . . . . . . . . . . . . . . . . . . . 1006.2 Summary of the cyclic direct simple shear loading test results for the homo-geneous sand-silt mixtures - II . . . . . . . . . . . . . . . . . . . . . . . . . . 1016.3 Summary of the cyclic triaxial shear loading test results for the homogeneoussand-silt mixtures - I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1026.4 Summary of the cyclic triaxial shear loading test results for the homogeneoussand-silt mixtures -II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103A.1 A summary of the key laboratory experimental studies on the response ofsand–silt mixtures identified from the literature review . . . . . . . . . . . . 167A.2 Summary of details on the specimen reconstitution methods . . . . . . . . . 170A.3 Summary of terminology / criteria for liquefaction or cyclic failure in literature172C.1 Summary of data set for device constants . . . . . . . . . . . . . . . . . . . 188C.2 Computation of expanded membrane internal diameter . . . . . . . . . . . . 190C.3 Device constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190C.4 Measurements set I for membrane modulus . . . . . . . . . . . . . . . . . . 192C.5 Measurements set II for membrane modulus . . . . . . . . . . . . . . . . . . 193C.6 Measurements set III for membrane modulus . . . . . . . . . . . . . . . . . 193E.1 Summary of particle distribution test on Fraser River silt . . . . . . . . . . 201E.2 Summary of particle distribution test on Fraser River sand . . . . . . . . . 202E.3 Data of specific gravity tests on Fraser River sand. . . . . . . . . . . . . . . 203xiE.4 Data of specific gravity tests on Fraser River silt. . . . . . . . . . . . . . . . 203E.5 Results of the quantitative analysis (weight percent) X-ray diffraction-Rietveldon Fraser River sand and silt. . . . . . . . . . . . . . . . . . . . . . . . . . . 204F.1 Summary of results for transducer calibrations in triaxial test device. . . . . 222F.2 Summary of results for transducer calibrations in direct simple shear testdevice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225F.3 Measurement resolutions and capacities of transducers in triaxial test device. 226F.4 Measurement resolutions and capacities of transducers in direct simple sheartest device. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228F.5 Estimated uncertainty of derived parameters in direct simple shear tests . . 232F.6 Estimated uncertainty of derived parameters in triaxial shear tests . . . . . 233H.1 Details of specimen preparation method SP-1d and SP-3d . . . . . . . . . . 247H.2 Summary of parameters in comparing specimen preparation methods SP-1dand SP-3d . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250H.3 Summary of parameters in comparing repeatability tests RI and RII . . . . 252H.4 Summary of parameters in comparing strain controlled and stress-controlledtests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255xiiList of Figures2.1 The relationships between liquefaction strength and fine/clay content fromthe test results by Yasuda et al. (1994) . . . . . . . . . . . . . . . . . . . . 72.2 Cyclic strength of Chile copper tailing sands with different silt content –Troncoso & Verdugo (1985) . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.3 Cyclic stress ratio versus sand skeleton void for sand with different silt content– Singh (1994) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.4 Static liquefaction potential increases as fines content and density increaseon Nevada 50/200 sand – Lade & Yamamuro (1997) . . . . . . . . . . . . . 92.5 Static liquefaction potential increases as fines content and density increaseon Nevada 50/80 sand – Lade & Yamamuro (1997) . . . . . . . . . . . . . . 92.6 Stress path and stress-strain curves for the Toyoyura sand with different siltcontent by Zlatovic & Ishihara (1995) . . . . . . . . . . . . . . . . . . . . . 102.7 Cyclic triaxial stress ratio versus number of cycles to initial liquefaction for[A] medium sand mixtures and [B] well-graded sand mixtures by Koester(1994) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102.8 Comparison of variations in normalized cyclic resistance from studies in whichcyclic resistance decreased and then increased with increasing silt content –Polito & Martin II (2003) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112.9 Schematic diagrams illustrating hypothesized particle structures for [I] loose,compressible state after deposition [II] after densification due to hearing in(a) sand with low silt content and (b) sand with low silt content—from Ya-mamuro & Lade (1997) and Yamamuro & Covert (2001) . . . . . . . . . . 132.10 Maximum, minimum, and quasi-natural void ratios for variations in finescontent on [A] Nevada 50/200 sand [B] Ottawa 50/200 sand—from Lade &Yamamuro (1997) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142.11 [A] The effect of the location and movement on the void ratio versus finescontent diagram on the particle structure between coarse-grains and fine-grains—from Lade & Yamamuro (1997) . . . . . . . . . . . . . . . . . . . . 142.12 Different types of particle structures at different fines contents along max-imum density line for binary spherical mixtures—from Lade et al. (1998). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.13 Intergranular soil mixture classification by Thevanayagam (2007a) . . . . . 162.14 Schematic illustrating volume and weight relationships for a saturated coarse-grained and fine-grained soil mixture . . . . . . . . . . . . . . . . . . . . . . 172.15 Typical liquefaction, limited liquefaction and cyclic mobility type responseduring undrained cyclic loading with associated stress-strain response, stress-path response and strain development – Vaid & Chern (1985) . . . . . . . 272.16 Hysteresis curves for cyclic loading test on loose and dense Sacramento RiverSand: Produced by extracting data from Seed & Lee (1966) . . . . . . . . . 28xiii3.1 Stress-strain relationship for the first cycle of loading (fine line) and thecycle at which -3 % axial strain is reached (thick line) for four specimen ofincreasing plasticity: Produced by extracting data from Bray & Sancio (2006),[LL: Liquid limit, PI: Plastic index, e: void ratio] . . . . . . . . . . . . . . . 333.2 Stress-strain behavior of Drammen Clay (Over consolidation ratio = 4) undersymmetrical direct simple shear loading: Produced by extracting data fromAndersen et al. (1988) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333.3 Cyclic shear stress-strain relationship of normally consolidated Fraser Riversilt (PI = 4) specimen with a cyclic stress ratio of 0.17, with an initial ver-tical effective stress of 100 kPa during a CDSS test - data from Sanin &Wijewickreme (2006) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353.4 ‘Incipient kink’ in a cyclic stress-strain loop . . . . . . . . . . . . . . . . . . 363.5 Effect of data recording rates in distinguishing the ‘incipient kink’ . . . . . 373.6 Cyclic stress-strain loops of sands in constant-volume CDSS tests . . . . . . 383.7 Cyclic stress-strain loops of silts and clay in constant-volume CDSS tests . 393.8 Cyclic stress-strain loops of mine-tailings in constant-volume CDSS tests . . 413.9 Comparison of the number of cycles for the occurrence of stress-strain pat-tern change in cyclic stress-strain loop and to reach 3.75 % shear strain for(A)Fraser River Sand, River Sand type I & II and Tailings Silty Sand; (B)Fraser River Silt, Serpentine River Sediments, Fraser River Deltaic Clay andKitimat Clay; and (C) Laterite, Copper, Red Mud Tailings and Gold Tailingstype I and II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423.10 Comparison of cyclic stress ratio vs. number of loading cycles derived fromthe proposed stress-strain pattern-based criterion and the shear strain crite-rion for CDSS tests performed on Fraser River Sand . . . . . . . . . . . . . 453.11 Comparison of cyclic stress ratio vs. number of loading cycles derived fromthe proposed stress-strain pattern-based criterion and the shear strain crite-rion for CDSS tests performed on Fraser River Silt . . . . . . . . . . . . . . 453.12 Comparison of cyclic stress ratio vs. number of loading cycles derived fromthe proposed stress-strain pattern-based criterion and the shear strain crite-rion for CDSS tests performed on Fraser River Deltaic Clay . . . . . . . . . 463.13 Comparison of cyclic stress ratio vs. number of loading cycles derived fromthe proposed stress-strain pattern-based criterion and the shear strain crite-rion for CDSS tests performed on Laterite Tailings . . . . . . . . . . . . . . 464.1 Typical distribution of shear stress and normal stresses distribution at themaximum shear stress of a soil specimen in DSS test device . . . . . . . . . 504.2 UBC Cyclic direct simple shear test device . . . . . . . . . . . . . . . . . . 514.3 UBC Cyclic triaxial test device . . . . . . . . . . . . . . . . . . . . . . . . . 544.4 Typical loading conditions and laboratory simulation models . . . . . . . . 554.5 Scanning Electron Microscopy Images of Fraser River Sand . . . . . . . . . 564.6 Scanning Electron Microscopy Images of Fraser River Silt . . . . . . . . . . 564.7 Particle size distribution of the sand, sand-silt mixtures and silt . . . . . . . 584.8 Uniformity assessment of the DSS specimens . . . . . . . . . . . . . . . . . 594.9 Uniformity assessment of the TRX specimens . . . . . . . . . . . . . . . . . 605.1 Repeatability analysis on the constant-volume DSS monotonic results I . . . 695.2 Repeatability analysis on the constant-volume DSS monotonic results II . . 70xiv5.3 Repeatability analysis on the triaxial shear test results I . . . . . . . . . . . 725.4 Repeatability analysis on the triaxial shear test results II . . . . . . . . . . 735.5 Normalized stress-strain response obtained from undrained monotonic directsimple shear tests on sand-silt mixtures . . . . . . . . . . . . . . . . . . . . 755.6 A comparison of normalized shear stress-strain responses of sand-silt mixturesresulted from constant-volume monotonic DSS tests . . . . . . . . . . . . . 765.7 Normalized stress-path response obtained from undrained monotonic directsimple shear tests on sand-silt mixtures . . . . . . . . . . . . . . . . . . . . 775.8 Excess pore-water pressure development response obtained from obtainedfrom undrained monotonic direct simple shear tests on sand-silt mixtures . 785.9 Normalized stress-strain response obtained from undrained monotonic triax-ial compression [U-MC –Test Series E] and undrained extension [U-ME –TestSeries F] tests on sand-silt mixtures . . . . . . . . . . . . . . . . . . . . . . 795.10 A comparison of normalized shear stress-strain responses of sand-silt mix-tures resulted from undrained monotonic triaxial compression [U-MC] andundrained extension [U-ME] tests . . . . . . . . . . . . . . . . . . . . . . . 805.11 Normalized stress-path response obtained from undrained monotonic triaxialcompression and extension tests on sand-silt mixtures . . . . . . . . . . . . 815.12 Excess pore-water pressure development response obtained from undrainedmonotonic triaxial compression and extension tests on sand-silt mixtures . . 825.13 Effective stress ratio versus axial strain response obtained from undrainedmonotonic triaxial compression and extension tests on sand-silt mixtures . . 825.14 Normalized stress-strain response obtained from drained monotonic triaxialcompression tests on sand-silt mixtures . . . . . . . . . . . . . . . . . . . . . 845.15 Volumetric strain development response obtained from drained monotonictriaxial compression tests on sand-silt mixtures . . . . . . . . . . . . . . . . 845.16 Effective stress ratio versus axial strain response obtained from drained mono-tonic triaxial compression tests on sand-silt mixtures . . . . . . . . . . . . . 855.17 Mass proportion and volume proportion of sand-silt mixture specimens testedat monotonic DSS tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865.18 Schematics illustrating coarse-grain and fine-grain matrix, fine-grain matrix,coarse-grain matrix and void matrix which are considered in defining Coarse-grain-based void ratio ecoarse and Fine-grain-based void ratio efine . . . . . 875.19 Variation of global void ratios e with fines content as a percent of soil weightCF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895.20 Global void ratios e, coarse-grain-based void ratio ecoarse and fine-grain-basedvoid ratio efine of DSS test specimens . . . . . . . . . . . . . . . . . . . . . 905.21 Range of applicability for ecoarse and efine with a comparison to e derivedfrom DSS test specimens . . . . . . . . . . . . . . . . . . . . . . . . . . . . 905.22 Variation of normalized Su−PT derived from triaxial monotonic compressionand extension (TRX-MC and TRX-ME), and DSS tests with respect to CF 925.23 Variation of normalized Su−PT derived with respect to e and ecoarse and efine— [A] triaxial undrained compression TRX-U-MC, [B] triaxial undrainedextension TRX-U-ME, and [C] constant-volume DSS. . . . . . . . . . . . . 935.24 Normalized Su−PT variation with respect to ecoarse and efine highlighting theeffects of ασ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945.25 Variation of φ′PT derived from triaxial monotonic compression and extension(TRX-MC and TRX-MC), and DSS tests with respect to CF . . . . . . . . 94xv5.26 Variation of normalized φ′PT derived with respect to e and ecoarse and efine— [A] triaxial undrained compression TRX-U-MC, [B] triaxial undrainedextension TRX-U-ME, and [C] constant-volume DSS. . . . . . . . . . . . . 956.1 Repeatability analysis on the constant-volume CDSS test results . . . . . . 1046.2 Repeatability analysis on the undrained CTX test results . . . . . . . . . . 1056.3 Comparison of the shear stress-strain responses of sand-silt mixtures derivedfrom the constant-volume CDSS test results . . . . . . . . . . . . . . . . . . 1066.4 Comparison of the stress-path responses of sand-silt mixtures derived fromthe constant-volume CDSS test results . . . . . . . . . . . . . . . . . . . . . 1076.5 Comparison of the shear stress-strain responses of sand-silt mixtures derivedfrom the constant-volume CTX test results . . . . . . . . . . . . . . . . . . 1086.6 Comparison of the stress-path responses of sand-silt mixtures derived fromthe constant-volume CTX test results . . . . . . . . . . . . . . . . . . . . . 1096.7 Comparison of the pore-water pressure development of sand-silt mixtures de-rived from the constant-volume CDSS tests with a CSR of 0.09 and undrainedCTX tests with a CSR of 0.11 . . . . . . . . . . . . . . . . . . . . . . . . . . 1116.8 Comparison of ru derived for sand-silt mixtures from the constant-volumeCDSS tests with different CSRs . . . . . . . . . . . . . . . . . . . . . . . . . 1126.9 Comparison of ru derived for sand-silt mixtures from the undrained CTXtests with different CSRs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1136.10 Comparison of change in ru with in each Loading cycle for sand-silt mixturesderived from the constant-volume CDSS test with different CSR . . . . . . 1146.11 Comparison of change in ru with in each Loading cycle for sand-silt mixturesderived from the undrained CTX test with different CSR . . . . . . . . . . 1156.12 Cyclic stress ratio [CSR] vs. number of loading cycles for stress-strain pattern-based criterion [Ncyc–PC ] from constant-volume CDSS tests . . . . . . . . . 1176.13 Cyclic stress ratio [CSR] vs. number of loading cycles for stress-strain pattern-based criterion [Ncyc–PC ] from undrained CTX tests . . . . . . . . . . . . . 1186.14 Comparison of cyclic stress ratio [CSR] vs. number of loading cycles forstress-strain pattern-based criterion [Ncyc–PC ] from constant-volume CDSStests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1196.15 Comparison of cyclic stress ratio [CSR] vs. number of loading cycles forstress-strain pattern-based criterion [Ncyc–PC ] from undrained CTX tests . 1196.16 Comparison of residual excess pore-water pressure ratios from undrainedCDSS tests with predictions from Green-Mitchell-Polito (GMP) – Green etal. (2000) model and Seed et al. (1975) model with α = 0.7 . . . . . . . . . 1226.17 Comparison of residual excess pore-water pressure ratios from undrainedCTX tests with predictions from Green-Mitchell-Polito (GMP) – Green etal. (2000) model and Seed et al. (1975) model with α = 0.7 . . . . . . . . . 1236.18 Derivation of cyclic secant shear modulus GLoop,N based on Stress-strainhysteresis loop in small strain and large strain magnitudes . . . . . . . . . . 1256.19 Comparison of the GLoop,N of sand-silt mixtures derived from constant-volume CDSS tests with a CSR of 0.09 and undrained CTX tests with aCSR of 0.11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1266.20 Comparison of the δD of sand-silt mixtures derived from the constant-volumeCDSS tests with a CSR of 0.09 and undrained CTX tests with a CSR of 0.11 1266.21 Degradation Index versus Number of loading cycles from CDSS tests . . . . 127xvi6.22 Degradation Index versus double amplitude cyclic shear strain from CDSStests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1286.23 Degradation Index versus Number of loading cycles from CTX tests . . . . 1296.24 Degradation Index versus double amplitude cyclic shear strain from CTX tests1306.25 Comparison of the Degradation Parameters of sand-silt mixtures derived fromthe constant-volume CDSS tests with a CSR of 0.09 . . . . . . . . . . . . . 1326.26 Comparison of the Degradation Parameters of sand-silt mixtures derived fromthe constant-volume CTX tests with a CSR of 0.11 . . . . . . . . . . . . . . 1336.27 ru at Ncyc = 1 and 10 for constant-volume CDSS tests with CSR = 0.09 andundrained CTX tests with CSR = 0.11 with respect to [A] CF , [B] ecoarse,and [C] efine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1356.28 Tends of cyclic resistance ratio [CRR] derived forNcyc–ref 15 for sand contactsdominant mixtures and for Ncyc–ref 23 for silt contacts dominant mixturesagainst ecoarse and efine respectively from CDSS tests . . . . . . . . . . . . 1366.29 Tends of cyclic resistance ratio [CRR] derived forNcyc–ref 15 for sand contactsdominant mixtures and for Ncyc–ref 25 for silt contacts dominant mixturesagainst ecoarse and efine respectively from CTX tests . . . . . . . . . . . . . 1366.30 The cyclic resistance ratio [CRR] derived for Ncyc–ref 15 for sand contactsdominant mixtures and for Ncyc–ref 25 for silt contacts dominant mixturesagainst global void ratio e from CDSS and CTX tests . . . . . . . . . . . . 1376.31 [A] CRR versus ecoarse and [B] CRR versus efine data points derived fromthis study in comparison with the results from different researchers . . . . . 138B.1 Load and displacement measurement in computing horizontal loading shaftfriction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182C.1 Positions of the Bellofram® rolling diaphragm in neutral floating, up-strokeand down-strike shown in a cross-section of the double acting piston. . . . . 184C.2 Load and displacement measurement in computing loading rod friction. . . 191C.3 Determination of Young’s modulus of rubber membrane. . . . . . . . . . . . 192E.1 Particle size distribution of Fraser River silt. . . . . . . . . . . . . . . . . . . 201E.2 Particle size distribution of Fraser River sand. . . . . . . . . . . . . . . . . . 202E.3 Rietveld refinement plot of Fraser River sand. . . . . . . . . . . . . . . . . . 205E.4 Rietveld refinement plot of Fraser River silt. . . . . . . . . . . . . . . . . . . 206E.5 Results – SEM Image Analysis : Fraser River sand Finer than #40 [425 µm]- Coarser than #60 [250 µm] . . . . . . . . . . . . . . . . . . . . . . . . . . 208E.6 Results – SEM Image Analysis : Fraser River sand Finer than #60 [250 µm]- Coarser than #80 [180 µm] . . . . . . . . . . . . . . . . . . . . . . . . . . 209E.7 Results – SEM Image Analysis : Fraser River sand Finer than #80 [180 µm]- Coarser than #100 [150 µm] . . . . . . . . . . . . . . . . . . . . . . . . . . 210E.8 Results – SEM Image Analysis : Fraser River sand Finer than #100 [150µm] - Coarser than #140 [106 µm] . . . . . . . . . . . . . . . . . . . . . . . 211E.9 Results – SEM Image Analysis : Fraser River sand Finer than #140 [106µm] - Coarser than #200 [75 µm] . . . . . . . . . . . . . . . . . . . . . . . . 212E.10 Results – SEM Image Analysis : Fraser River silt (Sample 1) Finer than#200 [75 µm] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213xviiE.11 Results – SEM Image Analysis : Fraser River silt (Sample 2) Finer than#200 [75 µm] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214E.12 Results – SEM Image Analysis : Fraser River silt (Sample 3) Finer than#200 [75 µm] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215E.13 Results – SEM Image Analysis : Fraser River silt (Sample 4) Finer than#200 [75 µm] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216E.14 Estimation of roundness and circularity of Fraser River sand . . . . . . . . . 217E.15 Estimation of roundness and circularity of Fraser River silt . . . . . . . . . 217F.1 Calibration regression curve - Load cell in triaxial test device (Stage 1). . . 220F.2 Calibration regression curve - Load cell in triaxial test device (Stage 2). . . 221F.3 Calibration regression curve - LVDT in triaxial test device. . . . . . . . . . 221F.4 Calibration regression curve - DPWPT in triaxial test device. . . . . . . . . 222F.5 Calibration regression curve - PWPT in triaxial test device. . . . . . . . . . 223F.6 Calibration regression curve - CPT in triaxial test device. . . . . . . . . . . 223F.7 Calibration regression curve - Vertical LC in direct simple shear test device. 224F.8 Calibration regression curve - Horizontal LC in direct simple shear test device.224F.9 Calibration regression curve - Vertical LVDT in direct simple shear test device.225F.10 Calibration regression curve - Horizontal LVDT in direct simple shear testdevice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225F.11 Stability of LC measurements. . . . . . . . . . . . . . . . . . . . . . . . . . . 226F.12 Stability of LVDT measurements. . . . . . . . . . . . . . . . . . . . . . . . . 227F.13 Stability of DPWPT measurements. . . . . . . . . . . . . . . . . . . . . . . 227F.14 Stability of PWPT measurements. . . . . . . . . . . . . . . . . . . . . . . . 227F.15 Stability of CPT measurements. . . . . . . . . . . . . . . . . . . . . . . . . . 228F.16 Stability of Vertical LC measurements. . . . . . . . . . . . . . . . . . . . . . 229F.17 Stability of Vertical LVDT measurements. . . . . . . . . . . . . . . . . . . . 229F.18 Stability of Horizontal LC measurements. . . . . . . . . . . . . . . . . . . . 229F.19 Stability of Horizontal LVDT measurements. . . . . . . . . . . . . . . . . . 230F.20 Specimen height and volume measurement stages in triaxial shear test. . . . 231G.1 Water pluviation preparation procedure . . . . . . . . . . . . . . . . . . . . 236G.2 Water pluviation in reconstituting a TRX specimen. . . . . . . . . . . . . . 237G.3 Water pluviation in reconstituting a DSS specimen. . . . . . . . . . . . . . . 238G.4 Slurry preparation procedure . . . . . . . . . . . . . . . . . . . . . . . . . . 239G.5 Saturated slurry deposition in reconstituting a TRX specimen. . . . . . . . 240G.6 Saturated slurry deposition in reconstituting a DSS specimen. . . . . . . . . 241G.7 Estimation of time for 100% consolidation [t100] (according to Bishop &Henkel (1957) method) from a typical consolidation curve for 0C100F spec-imen from 50kPa to 100kPa consolidation stress increment in the triaxialshear test device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243G.8 Estimation of time for 100% consolidation [t100] (according to Bishop &Henkel (1957) method) from a typical consolidation curve for 0C100F spec-imen from 50 kPa to 100 kPa consolidation stress increment in the directsimple shear test device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244H.1 Comparison of 1-D consolidation phase of slurry in specimen preparationmethod SP-1d and SP-3d. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248xviiiH.2 Comparison of isotropic consolidation phase of the specimens prepared fromSP-1d and SP-3d methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . 248H.3 Comparison of undrained compression phase of the specimens prepared fromSP-1d and SP-3d methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . 249H.4 Test repeatability comparison RI and RII in initial slurry consolidation phasewithin the split mold. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251H.5 Test repeatability comparison RI and RII in isotropic consolidation phase intriaxial tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251H.6 Test repeatability comparison RI and RII in undrained monotonic compres-sion triaxial tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252H.7 Comparison of strain controlled and stress controlled triaxial tests on FraserRiver silt. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254H.8 Comparison of shear stress-strain and stress path, development of pore-waterpressure and effective stress ratio response during the undrained monotonicTRX tests on 0C100F specimens with two different strain rates of 4% and25% per hour. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256H.9 Comparison of shear stress-strain and stress path, development of pore-waterpressure and initial excess pore-water pressure ratio response during theconstant-volume monotonic DSS tests on 0C100F specimens with a withthree different strain rates of 2%, 10% and 150% per hour. . . . . . . . . . . 258H.10 Comparison of shear stress-strain and stress path responses from undrainedCTX tests on 0C100F specimens with a CSR of about 0.18, at three differentloading periods of 30s, 100s, and 1000s. . . . . . . . . . . . . . . . . . . . . 259H.11 Comparison of development of pore-water pressure and accumulation of axialstrain responses from undrained CTX tests on 0C100F specimens with a CSRof about 0.18, at three different loading periods of 30s, 100s, and 1000s. . . 260H.12 Comparison of normalized shear stress-strain and normalized stress path re-sponses from constant-volume CDSS tests on 0C100F specimens with a CSRof about 0.14, at four different loading periods of 5s, 10s, 100s, and 1000s. 261H.13 Comparison of development of pore-water pressure and accumulation of axialstrain responses from constant-volume CDSS tests on 0C100F specimens witha CSR of about 0.14, at four different loading periods of 5s, 10s, 100s, and1000s. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262I.1 Results – DSS-i: 0C100F-M-SR-002: . . . . . . . . . . . . . . . . . . . . 264I.2 Results – DSS-ii: 0C100F-M-SR-010: . . . . . . . . . . . . . . . . . . . . 265I.3 Results – DSS-iii: 0C100F-M-SR-150: . . . . . . . . . . . . . . . . . . . 266I.4 Results – DSS-iv: 50C50F-M-SR-010: . . . . . . . . . . . . . . . . . . . 267I.5 Results – DSS-v: 50C50F-M-SR-150: . . . . . . . . . . . . . . . . . . . . 268I.6 Results – DSS-vi: 0C100F-M-SR-010: . . . . . . . . . . . . . . . . . . . 269I.7 Results – DSS-vii: 0C100F-M-SR-150: . . . . . . . . . . . . . . . . . . . 270I.8 Results – DSS-viii: 0C100F-14-05: . . . . . . . . . . . . . . . . . . . . . 271I.9 Results – DSS-ix: 0C100F-14-10: . . . . . . . . . . . . . . . . . . . . . . 272I.10 Results – DSS-x: 0C100F-14-100: . . . . . . . . . . . . . . . . . . . . . . 273I.11 Results – DSS-xi: 0C100F-14-1000: . . . . . . . . . . . . . . . . . . . . . 274I.12 Results – DSS-01: 100C0F-M: . . . . . . . . . . . . . . . . . . . . . . . . 275I.13 Results – DSS-01: 100C0F-M & DSS-12: 100C0F-MR: . . . . . . . . 276I.14 Results – DSS-26: 100C0F-13: . . . . . . . . . . . . . . . . . . . . . . . . 277xixI.15 Results – DSS-27: 100C0F-11: . . . . . . . . . . . . . . . . . . . . . . . . 278I.16 Results – DSS-28: 100C0F-10: . . . . . . . . . . . . . . . . . . . . . . . . 279I.17 Results – DSS-29: 100C0F-09: . . . . . . . . . . . . . . . . . . . . . . . . 280I.18 Results – DSS-26: 100C0F-13 & DSS-30: 100C0F-13R: . . . . . . . . 281I.19 Results – DSS-02: 95C05F-M: . . . . . . . . . . . . . . . . . . . . . . . . 282I.20 Results – DSS-02: 95C05F-M & DSS-13: 95C05F-MR: . . . . . . . . 283I.21 Results – DSS-31: 95C05F-13: . . . . . . . . . . . . . . . . . . . . . . . . 284I.22 Results – DSS-32: 95C05F-11: . . . . . . . . . . . . . . . . . . . . . . . . 285I.23 Results – DSS-33: 95C05F-10: . . . . . . . . . . . . . . . . . . . . . . . . 286I.24 Results – DSS-34: 95C05F-09: . . . . . . . . . . . . . . . . . . . . . . . . 287I.25 Results – DSS-03: 85C15F-M: . . . . . . . . . . . . . . . . . . . . . . . . 288I.26 Results – DSS-03: 85C15F-M & DSS-14: 85C15F-MR: . . . . . . . . 289I.27 Results – DSS-35: 85C15F-14: . . . . . . . . . . . . . . . . . . . . . . . . 290I.28 Results – DSS-36: 85C15F-13: . . . . . . . . . . . . . . . . . . . . . . . . 291I.29 Results – DSS-37: 85C15F-10: . . . . . . . . . . . . . . . . . . . . . . . . 292I.30 Results – DSS-38: 85C15F-09: . . . . . . . . . . . . . . . . . . . . . . . . 293I.31 Results – DSS-04: 75C25F-M: . . . . . . . . . . . . . . . . . . . . . . . . 294I.32 Results – DSS-04: 75C25F-M & DSS-15: 75C25F-MR: . . . . . . . . 295I.33 Results – DSS-39: 75C25F-13: . . . . . . . . . . . . . . . . . . . . . . . . 296I.34 Results – DSS-40: 75C25F-12: . . . . . . . . . . . . . . . . . . . . . . . . 297I.35 Results – DSS-41: 75C25F-10: . . . . . . . . . . . . . . . . . . . . . . . . 298I.36 Results – DSS-42: 75C25F-09: . . . . . . . . . . . . . . . . . . . . . . . . 299I.37 Results – DSS-40: 75C25F-12 & DSS-43: 75C25F-12R: . . . . . . . . 300I.38 Results – DSS-05: 63C37F-M: . . . . . . . . . . . . . . . . . . . . . . . . 301I.39 Results – DSS-05: 63C37F-M & DSS-16: 63C37F-MR: . . . . . . . . 302I.40 Results – DSS-44: 63C37F-14: . . . . . . . . . . . . . . . . . . . . . . . . 303I.41 Results – DSS-45: 63C37F-12: . . . . . . . . . . . . . . . . . . . . . . . . 304I.42 Results – DSS-46: 63C37F-10: . . . . . . . . . . . . . . . . . . . . . . . . 305I.43 Results – DSS-47: 63C37F-09: . . . . . . . . . . . . . . . . . . . . . . . . 306I.44 Results – DSS-06: 50C50F-M: . . . . . . . . . . . . . . . . . . . . . . . . 307I.45 Results – DSS-06: 50C50F-M & DSS-17 :50C50F-MR: . . . . . . . . 308I.46 Results – DSS-48: 50C50F-14: . . . . . . . . . . . . . . . . . . . . . . . . 309I.47 Results – DSS-49: 50C50F-12: . . . . . . . . . . . . . . . . . . . . . . . . 310I.48 Results – DSS-50: 50C50F-10: . . . . . . . . . . . . . . . . . . . . . . . . 311I.49 Results – DSS-51: 50C50F-09: . . . . . . . . . . . . . . . . . . . . . . . . 312I.50 Results – DSS-07: 40C60F-M: . . . . . . . . . . . . . . . . . . . . . . . . 313I.51 Results – DSS-07: 40C60F-M & DSS-18: 40C60F-MR: . . . . . . . . 314I.52 Results – DSS-52: 40C60F-14: . . . . . . . . . . . . . . . . . . . . . . . . 315I.53 Results – DSS-53: 40C60F-12: . . . . . . . . . . . . . . . . . . . . . . . . 316I.54 Results – DSS-54: 40C60F-10: . . . . . . . . . . . . . . . . . . . . . . . . 317I.55 Results – DSS-55: 40C60F-09: . . . . . . . . . . . . . . . . . . . . . . . . 318I.56 Results – DSS-52: 40C60F-14 & DSS-56: 40C60F-14R: . . . . . . . . 319I.57 Results – DSS-08: 30C70F-M: . . . . . . . . . . . . . . . . . . . . . . . . 320I.58 Results – DSS-08: 30C70F-M & DSS-19: 30C70F-MR: . . . . . . . . 321I.59 Results – DSS-57: 30C70F-14: . . . . . . . . . . . . . . . . . . . . . . . . 322I.60 Results – DSS-58: 30C70F-13: . . . . . . . . . . . . . . . . . . . . . . . . 323I.61 Results – DSS-59: 30C70F-10: . . . . . . . . . . . . . . . . . . . . . . . . 324I.62 Results – DSS-60: 30C70F-09: . . . . . . . . . . . . . . . . . . . . . . . . 325xxI.63 Results – DSS-09: 20C80F-M: . . . . . . . . . . . . . . . . . . . . . . . . 326I.64 Results – DSS-09: 20C80F-M & DSS-20: 20C80F-MR: . . . . . . . . 327I.65 Results – DSS-61: 20C80F-15: . . . . . . . . . . . . . . . . . . . . . . . . 328I.66 Results – DSS-62: 20C80F-13: . . . . . . . . . . . . . . . . . . . . . . . . 329I.67 Results – DSS-63: 20C80F-11: . . . . . . . . . . . . . . . . . . . . . . . . 330I.68 Results – DSS-64: 20C80F-09: . . . . . . . . . . . . . . . . . . . . . . . . 331I.69 Results – DSS-10: 10C90F-M: . . . . . . . . . . . . . . . . . . . . . . . . 332I.70 Results – DSS-10: 10C90F-M & DSS-21: 10C90F-MR: . . . . . . . . 333I.71 Results – DSS-65: 10C90F-15: . . . . . . . . . . . . . . . . . . . . . . . . 334I.72 Results – DSS-66: 10C90F-13: . . . . . . . . . . . . . . . . . . . . . . . . 335I.73 Results – DSS-67: 10C90F-11: . . . . . . . . . . . . . . . . . . . . . . . . 336I.74 Results – DSS-68: 10C90F-09: . . . . . . . . . . . . . . . . . . . . . . . . 337I.75 Results – DSS-11: 0C100F-M: . . . . . . . . . . . . . . . . . . . . . . . . 338I.76 Results – DSS-11: 0C100F-M & DSS-22: 0C100F-MR: . . . . . . . . 339I.77 Results – DSS-69: 0C100F-16: . . . . . . . . . . . . . . . . . . . . . . . . 340I.78 Results – DSS-70: 0C100F-14: . . . . . . . . . . . . . . . . . . . . . . . . 341I.79 Results – DSS-71: 0C100F-10: . . . . . . . . . . . . . . . . . . . . . . . . 342I.80 Results – DSS-72: 0C100F-09: . . . . . . . . . . . . . . . . . . . . . . . . 343I.81 Results – DSS-70: 0C100F-14 & DSS-73: 0C100F-14R: . . . . . . . . 344J.1 Results – TRX-i: 0C100F-U-MC-SP-1d: . . . . . . . . . . . . . . . . . 346J.2 Results – TRX-ii: 0C100F-U-MC-SP-3d: . . . . . . . . . . . . . . . . . 347J.3 Results – TRX-iii: 0C100F-U-MC-SP-3d-R: . . . . . . . . . . . . . . . 348J.4 Results – TRX-iv: 0C100F-U-MC-0.15kPa σa/s: . . . . . . . . . . . . 349J.5 Results – TRX-v: 0C100F-U-MC-4% a/h: . . . . . . . . . . . . . . . . 350J.6 Results – TRX-vi: 0C100F-U-MC-25% a/h: . . . . . . . . . . . . . . . 351J.7 Results – TRX-vii: 0C100F-18-1000s: . . . . . . . . . . . . . . . . . . . 352J.8 Results – TRX-viii: 0C100F-18-100s: . . . . . . . . . . . . . . . . . . . . 353J.9 Results – TRX-ix: 0C100F-18-30s: . . . . . . . . . . . . . . . . . . . . . 354J.10 Results – TRX-01: 100C0F-U-MC: . . . . . . . . . . . . . . . . . . . . . 355J.11 Results – TRX-02: 100C0F-U-ME: . . . . . . . . . . . . . . . . . . . . . 356J.12 Results – TRX-03: 100C0F-D-MC: . . . . . . . . . . . . . . . . . . . . . 357J.13 Results – TRX-04: 100C0F-U-MC-R: . . . . . . . . . . . . . . . . . . . 358J.14 Results – TRX-33: 100C0F-16: . . . . . . . . . . . . . . . . . . . . . . . 359J.15 Results – TRX-34: 100C0F-14: . . . . . . . . . . . . . . . . . . . . . . . 360J.16 Results – TRX-35: 100C0F-11: . . . . . . . . . . . . . . . . . . . . . . . 361J.17 Results – TRX-36: 100C0F-09: . . . . . . . . . . . . . . . . . . . . . . . 362J.18 Results – TRX-33: 100C0F-16 & TRX-37: 100C0F-16R: . . . . . . . 363J.19 Results – TRX-05: 95C05F-U-MC: . . . . . . . . . . . . . . . . . . . . . 364J.20 Results – TRX-06: 95C05F-U-ME: . . . . . . . . . . . . . . . . . . . . . 365J.21 Results – TRX-07: 95C05F-D-MC: . . . . . . . . . . . . . . . . . . . . . 366J.22 Results – TRX-38: 95C05F-17: . . . . . . . . . . . . . . . . . . . . . . . 367J.23 Results – TRX-39: 95C05F-15: . . . . . . . . . . . . . . . . . . . . . . . 368J.24 Results – TRX-40: 95C05F-13: . . . . . . . . . . . . . . . . . . . . . . . 369J.25 Results – TRX-41: 95C05F-11: . . . . . . . . . . . . . . . . . . . . . . . 370J.26 Results – TRX-08: 75C25F-U-MC: . . . . . . . . . . . . . . . . . . . . . 371J.27 Results – TRX-09: 75C25F-U-ME: . . . . . . . . . . . . . . . . . . . . . 372J.28 Results – TRX-10: 75C25F-D-MC: . . . . . . . . . . . . . . . . . . . . . 373xxiJ.29 Results – TRX-42: 75C25F-17: . . . . . . . . . . . . . . . . . . . . . . . 374J.30 Results – TRX-43: 75C25F-14: . . . . . . . . . . . . . . . . . . . . . . . 375J.31 Results – TRX-44: 75C25F-13: . . . . . . . . . . . . . . . . . . . . . . . 376J.32 Results – TRX-45: 75C25F-11: . . . . . . . . . . . . . . . . . . . . . . . 377J.33 Results – TRX-11: 65C35F-U-MC: . . . . . . . . . . . . . . . . . . . . . 378J.34 Results – TRX-12: 65C35F-U-ME: . . . . . . . . . . . . . . . . . . . . . 379J.35 Results – TRX-13: 65C35F-D-MC: . . . . . . . . . . . . . . . . . . . . . 380J.36 Results – TRX-13: 65C35F-D-MC-R: . . . . . . . . . . . . . . . . . . . 381J.37 Results – TRX-46: 65C35F-15: . . . . . . . . . . . . . . . . . . . . . . . 382J.38 Results – TRX-47: 65C35F-13: . . . . . . . . . . . . . . . . . . . . . . . 383J.39 Results – TRX-48: 65C35F-11: . . . . . . . . . . . . . . . . . . . . . . . 384J.40 Results – TRX-49: 65C35F-09: . . . . . . . . . . . . . . . . . . . . . . . 385J.41 Results – TRX-15: 50C50F-U-MC: . . . . . . . . . . . . . . . . . . . . . 386J.42 Results – TRX-16: 50C50F-U-ME: . . . . . . . . . . . . . . . . . . . . . 387J.43 Results – TRX-17: 50C50F-D-MC: . . . . . . . . . . . . . . . . . . . . . 388J.44 Results – TRX-18: 50C50F-U-MC-R: . . . . . . . . . . . . . . . . . . . 389J.45 Results – TRX-50: 50C50F-17: . . . . . . . . . . . . . . . . . . . . . . . 390J.46 Results – TRX-51: 50C50F-16: . . . . . . . . . . . . . . . . . . . . . . . 391J.47 Results – TRX-52: 50C50F-13: . . . . . . . . . . . . . . . . . . . . . . . 392J.48 Results – TRX-53: 50C50F-11: . . . . . . . . . . . . . . . . . . . . . . . 393J.49 Results – TRX-50: 50C50F-17 & TRX-54: 50C50F-17R: . . . . . . . 394J.50 Results – TRX-19: 40C60F-U-MC: . . . . . . . . . . . . . . . . . . . . . 395J.51 Results – TRX-20: 40C60F-U-ME: . . . . . . . . . . . . . . . . . . . . . 396J.52 Results – TRX-21: 40C60F-D-MC: . . . . . . . . . . . . . . . . . . . . . 397J.53 Results – TRX-22: 40C60F-U-MC-R: . . . . . . . . . . . . . . . . . . . 398J.54 Results – TRX-23: 20C80F-U-MC: . . . . . . . . . . . . . . . . . . . . . 399J.55 Results – TRX-24: 20C80F-U-ME: . . . . . . . . . . . . . . . . . . . . . 400J.56 Results – TRX-25: 20C80F-D-MC: . . . . . . . . . . . . . . . . . . . . . 401J.57 Results – TRX-26: 20C80F-U-MC-R: . . . . . . . . . . . . . . . . . . . 402J.58 Results – TRX-55: 20C80F-19: . . . . . . . . . . . . . . . . . . . . . . . 403J.59 Results – TRX-56: 20C80F-17: . . . . . . . . . . . . . . . . . . . . . . . 404J.60 Results – TRX-57: 20C80F-15: . . . . . . . . . . . . . . . . . . . . . . . 405J.61 Results – TRX-58: 20C80F-11: . . . . . . . . . . . . . . . . . . . . . . . 406J.62 Results – TRX-27: 0C100F-U-MC: . . . . . . . . . . . . . . . . . . . . . 407J.63 Results – TRX-28: 0C100F-U-ME: . . . . . . . . . . . . . . . . . . . . . 408J.64 Results – TRX-29: 0C100F-D-MC: . . . . . . . . . . . . . . . . . . . . . 409J.65 Results – TRX-59: 0C100F-20: . . . . . . . . . . . . . . . . . . . . . . . 410J.66 Results – TRX-60: 0C100F-18: . . . . . . . . . . . . . . . . . . . . . . . 411J.67 Results – TRX-61: 0C100F-15: . . . . . . . . . . . . . . . . . . . . . . . 412J.68 Results – TRX-62: 0C100F-11: . . . . . . . . . . . . . . . . . . . . . . . 413J.69 Results – TRX-60: 0C100F-18 & TRX-63: 0C100F-18R: . . . . . . . 414xxiiList of SymbolsB Skempton’s coefficient BCc Coefficient of curvatureCF Fines content—as a percent of soil weightCu Coefficient of uniformityd Degradation parameter—by Idriss et al. (1978)d∗ Degradation parameter—by Sharma & Fahey (2003)D Equivalent viscous damping factor derived from hys-teresis in a stress-strain loopDn Particle size corresponding to n% finerDr Relative densitye Global void ratioec Post consolidation global void ratioecoarse Coarse-grain-based void ratioefine Fine-grain-based void ratioESR Effective stress ratio – axial to radial stress ratioESRmax Maximum effective stress ratioESRmin Minimum effective stress ratioESRPT Effective stress ratio at phase transformationGLoop,N Secant shear modulus for whole stress-strain loop cor-responding to N thcycleGs Specific gravityHc Post-consolidated specimen heightNcyc Number of loading cyclesNcyc–ref Equivalent Ncyc corresponding to a 7.5 magnitudeearthquakeNcyc[=2.5%] Number of loading cycles when axial strain equals to2.5 % in cyclic triaxial shear testsNcyc[γ=3.75%] Number of loading cycles when shear strain equals to3.75 % in cyclic direct simple shear testsNcyc[γ=10%] Number of loading cycles when shear strain equals to10 % in cyclic direct simple shear testsNcyc–PC Number of loading cycles when the stress-strain looppattern changes - i.e. incipient kinkru Excess pore-water pressure ratior∗u Residual excess pore-water pressure ratioru−max Maximum excess pore-water pressure ratioru−∆N Change of excess pore-water pressure ratio at Nth cyclewith respect to N-1th cyclexxiiiru[=2.5%] Excess pore-water pressure ratio when axial strainequals to 2.5 % in cyclic triaxial shear testsru[γ=3.75%] Excess pore-water pressure ratio when shear strainequals to 3.75 % in direct simple shear testsru[pat.chg] Excess pore-water pressure ratio when the stress-strainloop pattern changes - i.e. incipient kinkRτ,γ Repeatability Index - based on stress-strain curve in adirect simple shear testRτ,σ Repeatability Index - based on stress-path curve in adirect simple shear testRτcyc,s Repeatability Index - based on cyclic stress-path curvein a triaxial shear testRτcyc,σ Repeatability Index - based on cyclic stress-path curvein a direct simple shear testRσ, Repeatability Index - based on stress-strain curve in atriaxial shear testRs,t Repeatability Index - based on stress-path curve in atriaxial shear testSu Mobilized undrained shear strength (Maximum valueduring the test up to 20% shear strain or 10 % axialstrain)Su−peak Mobilized peak undrained shear strength (prior to dila-tive tendency in response or strain softening type re-sponse)Su−PT Mobilized undrained shear strength at phase transfor-mations′Mean effective confining stress in MIT s-t spaces′o Initial mean effective confining stress during triaxialshearing phaseTcyc Period of the loading cyclet′Effective deviator stress in MIT s-t spaceWdry Dry weight of the soil specimenVc Post-consolidated specimen volumeασ Direction of major principal stress with respect to thedeposition directionδD Degradation Indexγ Shear strainγmax Maximum Shear strain during cyclic direct simple sheartestγ˙ Shear strain ratea Axial straina[max] Maximum Axial strain during cyclic triaxial testa[ru−max] Axial strain corresponding to maximum excess pore-water pressurexxiva[v−max] Axial strain corresponding to maximum volumetricstrainq Shear strainr Radial strainv Volumetric strainv−max Maximum volumetric strain in drained triaxial comres-sion test∆U Excess pore-water pressureηPT Stress ratio of q/p′or τ/σ′vc at phase transformationσ′a Effective axial stressσ′r Effective radial stressσ′vc Effective vertical confining stressσ′vco Initial effective vertical confining stress during directsimple shearing phaseτ Shear stressφ′MO Mobilized effective stress friction angle at maximumobliquityφ′PT Mobilized effective stress friction angle at phase trans-formationxxvList of AcronymsCDSS Cyclic Direct Simple ShearCPT Cone Penetration Test or Cell Pressure TransducerCRR Cyclic Resistance RatioCSR Cyclic Stress Ratio amplitudeCTX Cyclic Triaxial Shear TestDA Double AmplitudeDSS Direct Simple ShearDPWPT Differential Pre Water Pressure TransducerEPR Electro-Pneumatic RegulatorsLC Load CellLL Liquid LimitLVDT Linear Variable Differential TransfomerNGI Norwegian Geotechnical InstituteNP Non-PlasticPI Plastic IndexPL Plastic LimitPWPT Pore-Water Pressure TransducerSA Single AmplitudeSEM Scanning Electron MicroscopeSPT Standard Penetration TestTRX Triaxial ShearUBC University of British ColumbiaxxviAcknowledgementsThe research program would not be a success if not for the great financial support frommany sources. The research was conducted with the financial support provided by the Nat-ural Sciences and Engineering Research Council of Canada (NSERC) Discovery/AcceleratorSupplement Grant. Faculty of Applied Science Graduate Award (2016), Thurber Engineer-ing Graduate Scholarship in Civil Engineering (2016), Four Year Fellowship (2016-2020)which provided the great financial support during my research program, are gratefully ac-knowledged.My sincere and heartfelt gratitude goes out to Dr. Dharma Wijewickreme for his excel-lent supervision and immense support during the research. His excellent guidance, usefuladvice, kind support and valuable feedback truely shaped my research career. I would alsolike to thank Dr. Nemkumar Banthia, Dr. Erik Eberhardt, Dr. Jonathan Fannin, Dr. LiamFinn, Dr. John Howie, Dr. Loretta Li and Dr. Mahdi Taiebat for facilitating an uniqueand excellent academic experience for me at UBC. I am specially grateful to Dr. Yogi Vaid,for his valuable advice during the conductance of laboratory experimental test program.Members of the comprehensive examination committee, Dr. Roger Beckie, Dr. AliKhalili and Dr. Sumi Siddiqua, are kindly acknowledged for the interesting and thought-provoking questions on scientific and engineering principles, followed with intrigue discus-sions by bringing some aspects form different view points towards the research topic. Thecomments, feedback and suggestions on the research matter provided by the supervisorycommittee members, Dr. Roger Beckie and Dr. John Howie, are deeply acknowledged. Thetimely directions and kind supports that I had from Dr. Sadana Gamage and Dr. PrajaktaJadhav are also thankfully reminded.Dr. Elisabetta Pani and Jacob Karbel are kindly acknowledged for the effort, kind assis-tance and technical support during the quantitative phase analysis and producing scanningelectron microscope images on the test materials used in this study that were carried outat the Department of Earth Ocean and Atmospheric Sciences and Department of MaterialsEngineering, respectively, in the University of British Columbia. Thanks are due to HansMuten (undergraduate student research assistant) for the assistance in some of the datareduction and image analysis.The contributions of technical advices and support received from Doug Hudniuk, ScottJackson, Simon Lee, Bill Leung, Paula Parkinson, Sylvain Picard, Harald Schrempp, JohnWong and UBC Civil Engineering Workshop personnel are thankfully appreciated. Specialthanks are owed to my colleagues (Ruslan Amarasinghe, Anays Antunes, Sara Ataii, DanielBarnes, Andres Barrero, Antone Dabeet, Vinoth Ganapathiraman, Jeremy Groves, MichaelHuber, Thushara Jayasinghe, Jenifer Liu, Vincent McClelland, Judy Mei, IlaibibakamOmunguye, Francesca Palmieri, Santiago Quinteros, Andres Reyes, Drian Roos, AinurSeidalinova, Gaziz Seidalinov, Paul Slangen, Ana Valverde, Priyesh Verma, and MichelleWesolowski) for the friendly support and fruitful discussions throughout the research pro-gram and most importantly for the eventful time that we spent together in the GraduateGeotechnical Laboratory, many of the seminars, conferences, symposiums and workshops.xxviiThis dissertation is dedicated toNaturethat somehow constantly manage to raisemy curiosity to explore, research and understand herxxviiiChapter 1Introduction1.1 Research MotivationSoil is a multi-phase (comprising solids, liquids and gases) and particulate (consisting par-ticles that are of different sizes and shapes) material. The stiffness and strength of soilare generated from inter-particle friction, leading to a complex mechanical behavior thatis density and stress-level dependent, and inelastic with coupling between shear and vol-umetric strain arising due to rearrangement of soil grains/voids. Particularly in terms ofthe observed significant differences in engineering performance under shear loading, the me-chanical behavior soil is usually examined and described on the basis of its particle size –viz., coarse-grained (sandy soils) and fine-grained (silty and clayey soils).Over the past century, significant research work has been performed to understand themechanical behavior of coarse-grained and fine-grained soils, whereas the studies on theshear loading behavior of sand-silt or sand-clay mixtures are limited in spite of the factthat most geotechnical engineering situations involve soil deposits comprising a mixture ofcoarse and fine particles. The evidence of ground failures due to past earthquakes thathave occurred in areas underlain by sand-silt in mixed form and/or as inter-bedded layersequences of coarse- and fine-grained soils highlights the need to further understand thebehavior of such soils.The findings and conclusions derived from the studies on shear loading response ofsand and silt mixtures still remain conflicting despite many researchers have investigatedthe topic in the last three decades. For instance, a significant number of studies havereported that the presence of, and increase in, the silt content in a sand would tend toincrease the resistance of sand to liquefaction (Seed et al. 1983; Robertson & Campanella1985; and Ishihara 1996). The basis for these studies primarily comprises data gatheredfrom laboratory experimental investigations, in situ testing, and field observations duringand aftermath of seismic induced ground motions. In contrast to those findings, someinvestigations based on laboratory experimental studies have indicated that the presence,and in particular the increase of silt, in a sandy soil mass would cause its liquefactionresistance to decrease (Troncoso & Verdugo 1985; Verdugo & Ishihara 1996; and Lade &Yamamuro 1997). In addition to the above-mentioned conflicting findings, other findingsin the available literature also suggests that the increasing fines content, up to a thresholdvalue, would initially decrease the shear resistance of sand, and beyond that threshold theshear resistance would increase with further increase of fines content (Zlatovic & Ishihara1995; Law & Ling 1992; and Polito & Martin II 2001).It is also noteworthy that the differences in the basis of the comparison used (i.e. finescontent, relative density, global void ratio, or sand skeleton void ratio) have been iden-tified as reasons for the above apparent discrepancies in the understanding of the trends(increasing or decreasing) in soil resistance and strength with respect to changes in siltcontent. It has been reported that the liquefaction resistance of sand-non-plastic fines mix-tures may either decrease or increase with increasing fines content, when compared at the1same global void ratio; thus, effect of threshold values of fines in the soil matrix needs to beexplored. Thevanayagam (2007a,b) developed a framework for inter-grain contact densityindices based on the primary and secondary contacts of a simplified two-sized spherical par-ticle system to produce an enhancement in characterizing response of soil mixtures throughidentifying the soil type that dominant in the primary contacts. It provided a simplifiedtwo-sized spherical binary mixture approximation to depict the contact density variation asthe fines content increased in the binary mixture; however, the approach does not accountfor the factors such as soil fabric, plasticity, and aging effect. In the case of coarse-finemixtures (silty sands, sandy silts, sand-silt mixtures, clayey sand, sandy clay, and sand-clay mixtures), an effective governing parameter(s), capable of adequately capturing theresponse against both monotonic and cyclic shear loading is/are yet to be identified.Geotechnical laboratory testing plays an important role in advancing our understandingof the mechanical behavior of soil and complements the essential components from in–situtesting combined with the knowledge from case histories in characterizing the geo-materials.With this background, there is a need to undertake comprehensive laboratory experimentaltesting to study and address the knowledge gaps with respect to the shear loading responseof coarse-fine mixtures.1.2 Problem StatementThe current approaches to assess the monotonic and/or cyclic failure (liquefaction) of soilrelies on frameworks that consider the mechanical behavior of geomaterial as either sand-like, or clay-like. On the other hand, the mechanical behavior of mixtures of coarse-grainedand fine-grained soil (e.g., sand-silt or sand-clay) is complex, and our understanding on thisfront is limited. As such, the shear loading response of coarse-fine mixtures of soil needs tobe further investigated to advance the current state of knowledge and to propose/developmethods for more accurate assessments of liquefaction and cyclic softening potential, thus,leading to the scope of work and main objectives of this thesis as descried below.1.3 Scope and ObjectivesThe scope of the research work presented in this thesis focuses on a comprehensive labo-ratory element testing program to investigate the shear loading response of reconstitutedmixtures of sand and silt. The components of the study from a high-level point of view are:ˆ Undertaking a comprehensive laboratory test program to generate an extensive ex-perimental data set to understand the mechanical response of sand and silt mixtures.ˆ Analysis/interpretation of the data acquired from the experimental work to under-stand the effect of sand-silt composition on the shear loading response of sand-siltmixturesˆ Identify approaches and governing/controlling parameters that would effectively de-scribe the observed response of such mixtures.ˆ Explore alternate criteria for the assessment of liquefaction triggering of fine-grainedsoils - recognizing the lack of robustness of the currently used shear strain criteria(mainly developed for coarse-grained soils).2The specific research objectives of the current study can be listed as:i To summarize the current state of knowledge and understanding on the shear loadingresponse of sand-silt (sand-fine) mixtures and their resistance against liquefaction orcyclic softening with a focus of field based in-situ tests and laboratory experimentaltests data. This work provided the rationale for the work undertaken.ii To form a more fundamentally robust and more effective indicator for performancedeterioration from an engineering performance point of view to serve the purposes ofassessment of failure for the whole range of fines content in sand-silt mixtures duringshear loading as oppose to the widely used criteria based on excess pore-water pressureratio for coarse-grained soils and an arbitrary strain threshold for the fine-grained soils.iii To establish a specimen preparation method(s) for reconstituting sand, sand-silt mix-tures and silt specimens for the laboratory experimental element tests, paying specialattention on uniformity, repeatability and fully saturated conditions of the preparedspecimens, and also to achieve specimens that are considered to mimic the soil depo-sition conditions in natural fluvial sand-silt deposits.iv To investigate the monotonic and cyclic shear loading response of sand-silt mixturesand to determine the effects of fines content, loading type, and loading mode onthe observed responses during the experiments. In this, monotonic and cyclic shearloading tests on sand-silt mixture specimens using triaxial and direct simple sheardevices are required to generate experimental data. Experiments are intended tocover a spectrum of sand-silt mixture compositions (ranging from 100% sand to 100%silt)v Examine potential new void ratio definitions for sand-silt mixtures with due attentionpaid to ‘coarse-grained dominant’ or ‘fine-grained dominant’ particle matrices, andderive behavioral trends based on such void ratio definitions to effectively characterizethe cyclic and monotonic shear response of silt-sand mixtures.1.4 Thesis OutlineThe study is presented in seven chapters and ten appendices in the thesis that are outlinedas below:ˆ Chapter 1 introduces the research motivation, presents the problem statement, iden-tifies the scope, objectives and finally outlines the thesis arrangement.ˆ Chapter 2 reviews the current state of knowledge with focuses on (i) shear loadingresponse of sand-silt mixtures, (ii) sand-silt mixture specimen preparation and (iii)assessment of cyclic shear resistance of soil from laboratory tests, and summarizesknowledge gaps and bases for the identified scope and objectives of the current study.ˆ Chapter 3 introduces a new criterion—based on stress-strain and stiffness responseof soil rather than the criteria based on excess pore-water pressure development orarbitrary strain level threshold—to assess the shear resistance of soil from laboratorytests.3ˆ Chapter 4 presents details of the test devices (direct simple shear test device andtriaxial shear test device), test materials (Fraser River sand, Fraser River silt andthe sand-silt mixtures produced from the parent materials) with a summary of testspecimen preparation and the test program along with rationale.ˆ Chapter 5 presents the direct observations in comparing the monotonic shear testresults with respect to the fine-grained soil content in the tested sand-silt mixtures.Then the analysis of the test results and interpretation are detailed.ˆ Chapter 6 initially describes the stress-strain, stress-path, pore-water pressure de-velopment responses observed in cyclic shear tests on sand-silt mixtures. The cyclicshear resistance and cyclic shear stiffness degradation characteristics derived from thetest results are subsequently detailed and are followed by a discussion of the resultsand interpretations.ˆ Chapter 7 lists the decisive observations and main conclusions derived from the pre-sented experimental research, highlighting the novel contributions and their potentiallimitations. It also includes a note on the future research direction and recommenda-tions on characterizing the shear loading response of sand-silt mixtures.The supplementary material on the matters that were discussed in the Chapter 2 arelisted in table format under the Appendix A. A comprehensive description of on the testdevices utilized in the current study is presented in Appendix B and Appendix C fordirect simple shear test device and triaxial shear test device including their loading mecha-nisms, strain computations, device constants and additional information. The data obtainedfrom the laboratory experiments performed using direct simple shear device and triaxial testdevice were processed and reduced; then engineering parameters were computed. The pro-cess of data reduction and equations used for the computation of engineering parametersare detailed in Appendix D. The characteristic information of the Fraser River sand andFraser River silt tested in the study—particle size distribution, specific gravity, soil plas-ticity, material phase analysis, Scanning Electron Microscope Image analysis and shapedescriptors—are presented in Appendix E. Results, derived from measurements based onlaboratory experimental tests, are required to be assessed for the uncertainty for properinterpretation and accurate quantification of the test results. Hence, uncertainty in mea-surements and derived parameters in this experimental study are detailed in Appendix F.The water pluviation and saturated slurry deposition methods were used to reconstitutethe test specimen during the research study. Appendix G intends to detail the specimenreconstitute procedures for those two methods with respect to specimens in both direct sim-ple shear test and triaxial shear test. Further, it also includes a description on the initialinvestigations in selecting appropriate strain rates and loading rates for the tests perform inthis study. The four sections in Appendix H describe the effect of specimen preparation,test repeatability, identification of strain level at failure and selection of appropriate shearstrain rates and shear loading rates. The results obtained from all the monotonic and cyclicshear tests perform in the test program under the current research were presented in theforms of plots showing stress-strain, stress-path, strain accumulation and excess pore-waterdevelopment response during direct simple shear tests—Appendix I, and triaxial sheartest—Appendix J.4Chapter 2Literature ReviewMechanical loading response of soil is complex, and the soil element behavior is governedby several aspects including: (i) inherent properties of soil; (ii) characteristics of loading;and (iii) boundary conditions both in terms space coordinates as well as time. For exam-ple, grain size, packing density, friction angle, microstructure, and minerology are some ofthe soils’ inherent properties, whereas confining stress, stress history, load amplitude, du-ration and frequency of loading, and initial static shear stress can be considered as some ofthe loading characteristics. Boundary conditions include stress and/or strain constraints,drainage conditions including level of saturation, effect of soil layering, as well as times for(and after) consolidation.Laboratory element testing of soils plays a critical role in enhancing the understandingof the mechanical response of soil under monotonic/static and cyclic loading, and studiesinvolving the three considerations as described above have formed the basis for developingexploratory research frameworks in this regard.With the above backdrop and considering the need to study the response of sand-silt mixtures as the main focus of this thesis, this chapter summarizes the current stateof knowledge. The understanding of the shear loading response of sand-silt mixtures isreviewed in Section 2.1, particular from the point of view of the research objective i –identified in Section 1.3. The influence of fines content on void ratio of the sand-silt mixturesare discussed in 2.2 with a focus on the shear loading response. The methods available forpreparing specimens of sand-silt mixtures for laboratory testing is addressed in Section 2.3,while Section 2.4 summarizes the evolution of criteria, those have been used in laboratoryexperiments to assess the cyclic shear resistance. Important aspects and details those havepaved the way to the present research in this dissertation are summarized in Section 2.5.2.1 Shear Loading Response of Sand-Silt Mixtures1The shear loading response of sand and silt mixtures has been investigated by many re-searchers in the last thirty years; however, findings and conclusions derived from thosestudies still remain conflicting. Some of the studies have concluded that the presence of finesin sand would increase the material shear loading resistance against liquefaction; whereas,some others have determined that the presence of fines would reduce the shear resistance ofsand-silt mixtures. In some of the cases, it has been reported that increasing fines content,up to a threshold value, would initially decrease the shear resistance and then increase withfurther increase of fines content.1A version of this section is published in Soysa, A. & Wijewickreme, D. 2018. Initial observation on labora-tory shear loading response of sand-silt mixtures. Geotechnical Earthquake Engineering and Soil DynamicsV, Austin, Texas doi: 10.1061/9780784481486.031Wijewickreme D.,Soysa, A. and Verma P., 2019. Response of Natural Fine-grained Soils for Seismic De-sign Practice: A Collection of Research Findings from British Columbia, Canada. Soil Dynamics andEarthquake Engineering. 124, 280-296 doi: 10.1016/j.soildyn. 2018.04.0535These varying and conflicting findings suggest that there is a need to undertake moreadvance research – in particular, to delineate the effects of soil grain size, distribution ofgrain size, and packing arrangement in the soil matrix to characterize the shear loadingresponse of sand-silt mixtures.A summary of the key laboratory experimental studies on the response of sand-silt mix-tures, sandy soils and silty soils identified through a comprehensive review on the availableliterature is presented in Table A.1. Relevant findings derived from some selected previousstudies (including both in-situ filed tests and laboratory experimental test) are detailed inthe section below from the viewpoint of presenting the current status of knowledge andidentifying the knowledge gaps/deficiencies for effective characterization of the mechanicalresponse of sand-silt mixtures.2.1.1 Conflicting Findings on the Shear Loading Response of Sand-SiltMixturesIncrease in shear resistance with increasing fines contentMany studies have indicated that that the presence and increasing silt content of sandhas been reported to increase the resistance of soils to liquefaction [Kaufman & Chang1982; Tokimatsu & Yoshimi 1983; Tokimatsu & Yoshimi 1984; Seed et al. 1983; Robertson& Campanella 1985; Koester & Tsuchida 1988; Pitman et al. 1994; Yasuda et al. 1994;Ishihara 1996; and Amini & Qi 2000]. The basis for these studies primarily comprises datagathered from laboratory experimental investigations, in situ testing, and field observationsduring and aftermath of seismic shaking.From the test results derived from a series of triaxial tests performed on sand-silt mix-tures with pre-selected gradation, Kaufman & Chang (1982) reported that as the silt contentincreases in the sand-silt mixtures, the cyclic shear resistance increases over that of parentsand. They also reported the differences in permeabilities and fabric of the soil, when siltis introduced to the sand, possibly resulted the different pore-water pressure generationcharacteristics. When the silt content increases beyond 60%, the rate of increase in cyclicshear strength with respect to silt content was observed to be greatly decreased. Tokimatsu& Yoshimi (1983) and Tokimatsu & Yoshimi (1984) developed criteria and correlationson liquefaction susceptibility of sand with fines content based on Standard PenetrationTest [SPT] results, and they concluded that sand with higher fines content indicated muchgreater liquefaction resistance than clean sands (when both clean sand and sand with finesshow the similar penetration resistance), and this tendency was observed to increase withincrease in fines content.Similarly, Seed et al. (1983), Robertson & Campanella (1985), and Shibata & Teparaksa(1988) recognized that, in assessing liquefaction susceptibility, silty sand shows greater cyclicresistance ratio than clean sand when compared under similar exhibited cone penetrationresistance values. Such expressions generally presented as corrected blow count in SPTversus cyclic shear resistance ratio [CRR] overlain on data from liquefaction case histories,where increase of CRR with increasing fines content could be observed for a given pene-tration resistance. Youd et al. (2001) clarified that, although Seed et al. (1985) noted anapparent increase of CRR with increased fines content from penetration resistance data,it is not clear whether an increase of liquefaction resistance or a decrease of penetrationresistance would be the driver of the observed results.Koester & Tsuchida (1988) summarized and reviewed the Japanese research findings6(i.e. Kondoh et al. 19871; Matsumoto et al. 19882; and Sasaki et al. 19883) and reportedgeneral increase in cyclic resistance in sand with increasing fines content. Further, Koester& Tsuchida (1988) concluded that percentage of fines alone is not an adequate criterion todistinguish deposits susceptible to liquefaction. Yasuda et al. (1994) performed a series oftests on silty sand collected from liquefied sites and artificially filled sites and concludedthat the liquefaction resistance would slightly increase as the fines content increase (withfairly large scatter in the results); however, a lesser amount of scatter in the results wasobserved when the liquefaction resistance was compared with respect to the clay contentsof the mixture (see Figure 2.1).Average00.10.20.30.40 20 40 60 80 100Cyclic Stress Ratio causing LiquefactionFine Content finer than 75μm %σ′o = 49 kPaDA = 5%Diameter 5 cmDiameter 7.5 cmAverage00.10.20.30.40 10 20 30 40 50Cyclic Stress Ratio causing LiquefactionClay Content finer than 5μm %σ′o = 49 kPaDA = 5%Diameter 5 cmDiameter 7.5 cmFigure 2.1: The relationships between liquefaction strength and fine/clay content from thetest results by Yasuda et al. (1994)By conducting undrained triaxial compression tests on specimens of Ottawa sand mixedwith silt, Pitman et al. (1994) concluded that the presence of silt would decrease the struc-tural instability of the mixture. Ishihara (1996) commented on the degree of liquefiabilityof sand containing “more or less cohesive” fines – he concluded that the cyclic resistanceof sand tends to increase to a certain extent under its normally consolidated state withincreasing content of fines, but to a greater extent if it is over consolidated. Investigatingthe liquefaction resistance of silty sands for both uniform and layered soil conditions, Amini& Qi (2000) concluded that increase in silt content would increase the soil resistance toliquefaction.Decrease shear resistance with increasing fines contentIn contrast to findings presented earlier, some investigators [Troncoso & Verdugo 1985; Singh1994; Verdugo & Ishihara 1996; and Lade & Yamamuro 1997] seem to conclude that thepresence, and in particular the increase, of silt in sand would cause its liquefaction resistance1Kondoh, M., Sasaki, Y., & Matsumoto, H. (1987). Effect of Fines Contents on Soil Liquefaction Strength(Part I), Proceedings of the Annual Meeting of the Japanese Society of Soil Mechanics and FoundationEngineering (In Japanese), Public Works Research Institute, Ministry of Construction, Tsukuba, Japan.2Matsumoto, H., Sasaki, Y., & Kondoh, M. (1988). Effect of Fines Contents on Soil Liquefaction Strength(Part 2), Proceedings of the Annual Meeting of Japanese Society of Soil Mechanics and Foundation Engi-neering (In Japanese), Public Works Research Institute, Ministry of Construction, Tsukuba, Japan.3Sasaki, Y., Matsumoto, H., & Kondoh, M. (1988). Liquefaction Strength Based on Laboratory Soil Tests- Effects of Fines Content, Technical Report No. 2590, (In Japanese), Public Works Research Institute,Ministry of Construction, Tsukuba, Japan.7to decrease. Based on the results obtained from cyclic triaxial tests on homogeneous sand-silt mixtures prepared from processed copper tailings, Troncoso & Verdugo (1985) reportedthat a moderate increase in fines content could cause a substantial increase in liquefactionsusceptibility as the shown cyclic strength derived from CTX tests in Figure 2.2. Troncoso& Verdugo (1985) postulated that when small silt particles fill within the irregular voidsskeleton produced from angular, sharp and large sand, it would reduce the interlockingeffects, and in turn, result in significant loss of cyclic resistance. The difference in the derivedcyclic shear resistance of silty sand with different silt contents seem to be more noticeablein the range from 0% to 15% of silt contents than that for the range from 22% to 30%.Troncoso & Verdugo (1985) suggested that greater compressibility and lesser permeabilitydue to silt presence in the sand could be the reasons for the reported substantial increasein the liquefaction susceptibility of silty sands with moderate increase of silt content in thesoil.Singh (1994) performed cyclic triaxial tests on mixtures prepared from Flint Shot No.4 sand and silt portion of Pasadena fine sand, and he concluded that for a given relativedensity, sands containing 10%, 20%, or 30% of silt by weight would have lesser resistanceto liquefaction than sand with no silt. Verdugo & Ishihara (1996) assessed the inherentliquefaction vulnerability of Toyoura sand and silt through an index defined as relative con-tractiveness; they reported that, when the presence of low plastic fines content is significant,soil has a greater opportunity to exist in the field in a contractive state that, in turn, couldlead to liquefaction and/or flow failure.00.20.40.61 10 100Cyclic Stress RatioNumber of Cyclese = 0.850% F5% F10% F15% F22% F30% FFigure 2.2: Cyclic strength of Chile cop-per tailing sands with different silt con-tent – Troncoso & Verdugo (1985)100% sand10% silt20% silt30% silt0.10.20.30.40.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2Cyclic Stress Ratio for failure at 20 cyclesSand Skeleton void ratioRelative Density of each sample = 50%Figure 2.3: Cyclic stress ratio versus sandskeleton void for sand with different siltcontent – Singh (1994)Lade & Yamamuro (1997) performed undrained triaxial compression tests on reconsti-tuted samples of Nevada and Ottawa sands with systematic variations in the non-plasticfines content and reported that increasing the fines content in sands greatly increased theliquefaction potential. The effective stress paths derived for the tests on Nevada 50/200sand with increasing fines content by Lade & Yamamuro (1997) shown in Figure 2.4 indi-cate decreasing values for maximum stress difference, while stress-strain plots indicate thatincreased fines content caused the sand to liquefy at lower values of axial strain. Except forthe case of response observed for the Nevada 50/80 sand specimen, similar response shownin Figure 2.4 can be seen in Figure 2.5 for the specimens with increasing fines content inNevada 50/80 sand. From the response observed for the cases of increasing fines contentin both Nevada 50/200 sand and Nevada 50/80 sand, Lade & Yamamuro (1997) reported8that despite the relative and absolute densities increase with the inclusion of fines in to thesand, the potential for static liquefaction increases, contrary to the typical anticipated soilresponse with respect to increasing density. Based on those observations, Lade & Yama-muro (1997) concluded that neither relative density nor void ratio alone could be consideredas satisfactory indicators of liquefaction potential in silty sands.01020304050600 10 20 30 40 50 60Stress Difference (kPa)Effetive Mean Normal Stress (kPa)Nevada 50/200 SandDr = 22% (0.812)0% FinesDr = 20% (0.790)10% FinesDr = 26% (0.770)20% FinesDr = 33% (0.752)30% Fines051015200 2 4 6 8 10Stress Difference (kPa)Axial Strain (%)Nevada 50/200 SandIncreasing Fines ContentIncreasing Fines ContentFigure 2.4: Static liquefaction potentialincreases as fines content and density in-crease on Nevada 50/200 sand – Lade &Yamamuro (1997)01020304050600 10 20 30 40 50 60Stress Difference (kPa)Effetive Mean Normal Stress (kPa)Nevada 50/80 SandDr = 11% (0.828)0% FinesDr = 19% (0.764)10% FinesDr = 20% (0.729)20% FinesDr = 25% (0.709)30% FinesDr = 42% (0.706)50% Fines051015200 2 4 6 8 10Stress Difference (kPa)Axial Strain (%)Nevada 50/80 SandIncreasing Fines ContentIncreasing Fines ContentFigure 2.5: Static liquefaction potentialincreases as fines content and density in-crease on Nevada 50/80 sand – Lade &Yamamuro (1997)Decrease shear resistance to a threshold and recover with increasing finescontentIn addition to the above conflicting conclusions of the occurrence of either increase ordecrease of shear resistance of sand-silt mixtures, as the silt content is increased, anotherslightly different set of deductions can also be found from the available literature. Thispertains to the observation that the shear resistance would initially decrease with increasingsilt content until some minimum resistance is reached; then, the shear resistance wouldincrease as the silt content continues to increase.Zlatovic & Ishihara (1995) performed undrained triaxial tests on Toyoura sand and siltmixtures specimens having a fines content of 10%, 15%, 25%, 30%, and 40% preparedfrom water sedimentation methods. The increasing silt contents up to 30% resulted in9corresponding increments in contractiveness and decrements in peak and residual strengthas depicted in Figure 2.6. They also reported that with further increase in silt content, thepeak and residual strengths increased along with decreasing contractiveness.10%15%30%40%00.20.40.60.811.21.40 0.2 0.4 0.6 0.8 1 1.2 1.4Normalized deviator stressNormalized mean effective stressToyoura sand with silt10%15%30%40%00.20.40.60.811.21.40 5 10 15 20Normalized deviator stressAxial strain (%)Toyoura sand with siltFigure 2.6: Stress path and stress-strain curves for the Toyoyura sand with different siltcontent by Zlatovic & Ishihara (1995)[A] [B]0.10.20.30.40.50.60.1 1 10 100 1000Cyclic Triaxial Stress Ratio Number of loading cyclesUCD cyclic triaxial tests - medium sandec = 0.558, PI =4%, σ′3 = 15 psivarying fine content %0512.52045600.10.20.30.40.50.60.1 1 10 100 1000Cyclic Triaxial Stress Ratio Number of loading cyclesUCD cyclic triaxial testswell-graded sandec = 0.48, PI =4%, σ′3 = 15 psivarying fine content %0512.5204560Figure 2.7: Cyclic triaxial stress ratio versus number of cycles to initial liquefaction for [A]medium sand mixtures and [B] well-graded sand mixtures by Koester (1994)Based on the results obtained from cyclic triaxial tests and hollow cylinder torsionalsimple shear tests performed on various sand-silt mixtures, Koester (1994) noted that thestrength in sand, silt, and clay mixtures progressively decreased to a lower-bound valuewhen fines content was increased up to 24 % ∼ 30 % of dry weight; then, further increasein fines content caused increase in cyclic strength. Further, Koester (1994) concluded that10fines content has greater influence on cyclic shear resistance of soils containing fines at agiven void ratio than the plasticity index of the fines fraction of the coarse-fine mixture. Heconcluded that cyclic strength of soils may not be characterized on the basis of gradationalone, and reported that presence of fines does not necessarily assure increased resistanceto development of residual excess pore-water pressures leading to liquefaction.Law & Ling (1992) and Polito & Martin II (2001) are among those who observed that theresistance would initially decrease with increasing silt content, and thereafter increase withfurther increase in silt content. Law & Ling (1992) reported the difficulties in quantifying theliquefaction resistance using a single physical parameter such as void ratio, fines content, andplasticity index; they emphasized that liquefaction resistance is affected by the combinedinfluence from all these parameters. As such, attempts to relate liquefaction resistance toone parameter in isolation would not lead conclusive findings. Polito & Martin II (2001)concluded that the cyclic resistance of a silty soil is dependent not on the actual silt contentof the soil, but upon whether it is above or below the limiting silt content for the sand.Hence, Polito & Martin II (2001) suggested that no general statements as to the liquefactionsusceptibility of a soil at a specific silt content (e.g., 10% or 30 % silt) can be made withoutknowing the limiting silt content of the soil.00.10.20.30.40.50.60.70.80.910 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1Normalized Cyclic ResistanceFine-grained ContentChalk sand &  Jackfish silt: Law and Ling (1992)Medium Sand - Koester (1994)Monterey No. 0/30: Polito and Martin (2001)Yatesville sand and silt: Polito and Martin (2001)Figure 2.8: Comparison of variations in normalized cyclic resistance from studies in whichcyclic resistance decreased and then increased with increasing silt content – Polito & MartinII (2003)112.1.2 Remarks on the Conflicting FindingsThe review of literature presented in Section 2.1.1 indicates that the field testing-based cor-relations, developed from seismic induced liquefaction events, on liquefaction susceptibilityhave typically predicted more resistance to liquefaction when the fines content of the soils isincreasing. However, most of the laboratory experimental studies contradict the field-basedcorrelation and suggest initial increase of fines content of sand makes the soil more suscepti-ble to liquefaction. As opposed to the experimental results on reconstituted specimens, thefield tests on the soils that have undergone seismic-induced liquefaction or stress conditionshave the possibility of including stress-history, natural soil fabric and aging effects, and sucheffects can be considered as a plausible explanation for the above contradictory results.Vaid (1994) pointed out that depending on the basis of the comparison (i.e. relative den-sity, void ratio, or sand skeleton void ratio), different trends of increase (or decrease) in cyclicresponse would emerge with respect to change in silt content. Similarly, Finn et al. (1994)stated that the presence of fines would cause cyclic strength either to increase, decrease, ornot influenced when the comparison is made under identical penetration resistance, grossvoid ratios, and sand skeleton void ratios, respectively. Xenaki & Athanasopoulos (2003)also concluded that the liquefaction resistance of sand-non-plastic fines mixtures may eitherdecrease or increase with increasing fines content, when compared at the same global voidratio; they suggested that the effect of threshold values of fines in the soil matrix needs tobe explored.2.2 Effects of Fines Content on the Void RatioLade & Yamamuro (1997), Yamamuro & Lade (1997) and Lade et al. (1998) explored theparticle structure of clean sand when it is mixed with increasing amounts of non-plastic silt.They hypothesized that sands with low silt content tend to form a particle structure wherethe sand grains would provide the load-bearing skeleton, and the silt particles would occupythe void spaces between the sand particles; since these voids are relatively unoccupied (dueto the low amount of fines), the overall soil mass could result in a meta-stable structure.They also hypothesized that fine particles could locate in between the sand particles atcontact points of sand particles when they are deposited in to a relatively loose configurationas schematized in Figure 2.9[Ia] thus increasing the instability of the particle structure.When a structure is formed from a loosely deposited state, when compressed and sheared,the silt particles that were initially separating the larger grains could be forced into the voidspaces (see Figure 2.9[IIa]) causing significant volume reduction both during consolidationand shearing. Lade & Yamamuro (1997) and Yamamuro & Lade (1997) mentioned that thishypothesis can be attributed to their observed contractive response with respect to loosesilty sands. They further mentioned that at higher confining stress, this collapse wouldassist the silt particles to locate in the void spaces in sand skeleton, and move the sandparticles into better contact formations amongst them, resulting in increased dilation.The sand with high silt content leads to a hypothesized particle structure [as per Yama-muro & Covert (2001)] that sand particles are further dispersed due to the high quantity ofloose silt particles as schematically shown in Figure 2.9[Ib]. As the stresses develop duringconsolidation and shearing phases, as illustrated in Figure 2.9[IIb], it would result in no-table compressibility of the soil matrix, largely due to the volume reduction in the particlestructure of loosely placed silt that occupies the space in between sand particles. Yamamuro& Covert (2001) mentioned that even at larger strain levels, contact between sand particles12may not be established fully, hence, silt particles would possess the load-bearing skeletonresulting continuous contractive type response. For these reasons, it was postulated thatthe expected response of the sand with low silt content with respect to that of sand withhigh silt content could be very different.[I] Silty sand as deposited [II] Silty sand compressed and shearedSand grains displaced apart by silt grainsSilt grains at contact pointsSand grains displaced far apart by large quantity of silt grainsSilt grains in significant quantityLarge volume reduction during initial shearingLarge grains move into better contact as shearing continues resulting in increasing dilatancy(a) (b)Large grains move into better contact only when sheared to large strains resulting in slight  increases in dilatancy(a) (b) Large volume reduction during initial and continued shearingFigure 2.9: Schematic diagrams illustrating hypothesized particle structures for [I] loose,compressible state after deposition [II] after densification due to hearing in (a) sand withlow silt content and (b) sand with low silt content—from Yamamuro & Lade (1997) andYamamuro & Covert (2001)The maximum and minimum void ratios, determined for the two different base sandswith varying fines contents presented by Lade & Yamamuro (1997) are presented in Fig-ure 2.10. They reported that the maximum and minimum void ratio lines derived underdifferent fines content appear to be approximately parallel; an initial decrease can be ob-served in the void ratio up until the fines content increases to a threshold value and, there-after, the void ratio would increase with further increase in fines content. The gradient ofthe curves can be dependent on the ratio of the mean diameters and shape of the parentmaterials. Some of the other investigators such as Kuerbis & Vaid (1988) and Pitman etal. (1994) have also identified and presented similar trends for the variation of void ratiowith respect to the fines content; subsequently, Thevanayagam (1998) and Cubrinovski &Ishihara (2002) have also confirmed such trends for the variation of void ratio with finescontent.The broadly practiced methods outlined in ASTM (2016a) and ASTM (2016b) areapplicable to soils with non-cohesive fines only up to 15%, by dry mass. Nevertheless, Lade& Yamamuro (1997) reported that the soil mixtures were energetically densified till themaximum possible densities were achieved to obtain the minimum void ratio line for the soilpresented in Figure 2.10. For the derivation of maximum void ratio line, soil mixtures wereprovided with minimum amount of input energy during the deposition process – dry funneldeposition with no drop height. Lade & Yamamuro (1997) examined the characteristicspresented in Figure 2.10 within the context of soil behavioral aspects which led to theoutcome presented in the Figure 2.11.13Maximum Void RatioMinimum Void Ratio0.40.50.60.70.80.911.11.20 20 40 60 80 100Void RatioFines Content (%)Nevada 50/200 SandMaximum Void RatioMinimum Void Ratio0.40.50.60.70.80.911.11.20 20 40 60 80 100Void RatioFines Content (%)Ottawa 50/200 Sand[A] [B]Figure 2.10: Maximum, minimum, and quasi-natural void ratios for variations in finescontent on [A] Nevada 50/200 sand [B] Ottawa 50/200 sand—from Lade & Yamamuro(1997)Maximum Void Ratio LineMinimum Void Ratio Line0.40.60.811.21.40 10 20 30 40 50 60 70 80 90 100Void RatioFines Content (%)Greater decrease in void ratio, large grains tightly packed; fines in void spacesDecreasing effect of particle structuredue to increasing density, which forces fines in to void spaces and larger particles closerFigure 2.11: [A] The effect of the location and movement on the void ratio versus finescontent diagram on the particle structure between coarse-grains and fine-grains—from Lade& Yamamuro (1997)It depicts the relatively greater reduction in minimum void ratios when fine-grains are14introduced in the coarse-grain matrix, with respect to that of maximum void ratios. Bothminimum and maximum void ratios indicated decreasing trend with initial increase of finescontent in the region where the coarse-fine particle structure is dominant. Further increasein fines content resulted in greater rate of increase of maximum void ratios (with respect tofines content) in comparison to that rate for minimum void ratios.Lade et al. (1998) also attempted to interpret the variation of void ratios by simplifyingthe particles to be spherical in a binary mixture in the Figure 2.12. The point A in the plotrefers to the void ratio for the case involving larger spherical particles alone. As smallerspherical particles fill the voids formed by the matrix of larger spherical particles, densitytends to increase, thus the void ratio decreases till to point B. Further increase of smallerspherical particles proportion in the binary mixture would disperse the larger sphericalparticles far apart, the increasing void ratio eventually reaches point C when a particlestructure formed only be smaller spherical particles.ABCDe2e1emin% Fines by volume% Coarse particles by volume01001000Void ratio Coarse particlesVoid ratio Fine particlesPorosity of coarse particlesPacking = 𝑒11 + 𝑒1Figure 2.12: Different types of particle structures at different fines contents along maximumdensity line for binary spherical mixtures—from Lade et al. (1998)Identifying the deficiencies in global void ratio as an index of contact density of coarse-fine mixtures, Thevanayagam (2007a) developed a frame work for inter-grain contact densityindices based on the primary and secondary contacts of a simplified two-sized sphericalparticle system as shown in Figure 2.13. The scenario of fine-grains are fully confined withthe voids formed by the coarse-grain stricture and fine-grains would not contribute to theload bearing/transferring mechanism is identified in ‘Case i’ by Thevanayagam (2007a) andpresented the inter-granular void ratio as the grain contact density index for that case. Forthe ‘Case ii’ and ‘Case iii’ where primary grain contact is still dominated by the inter-coarse15grain contact and fine-grains either confined and partially in contact with coarse-grains orconfined in yet separate the coarse-grains, Thevanayagam (2007a) introduced equivalentinter-granular void ratio. When the primary grain contact is switched from inter-coarsegrain contacts to inter-fine grain contacts, as the fines contents of the mixture increases,Thevanayagam (2007a) defined two categories based on the fact whether coarse-grains arefully dispersed (‘Case iv-1’ ) or partially dispersed (‘Case iv-2’ ). The inter-fine void ratioand equivalent inter-fine void ratio are defined as the grain contact density indices for thosetwo cases respectively. In Figure 2.13, primary grain contact in cases of coarse-grain soilmix and fine-grain soil mix is distinguished through the use of the threshold fines content.Coarse-grain soil mix Fine-grain soil mix Soil mixture Primary grain contactInter-coarse grain contact dominantInter-fine grain contact dominantSecondary grain contactMicrostructure Grain contact density indexCoarse grain fully dispersed– Case iv-1Fine grain confined and partially contact with coarse grains  – Case iiFine grain confined and separator of coarse grains– Case iiiCoarse grain partially dispersed– Case iv-2𝑒𝑐 =(𝑒 + 𝑐𝑓)(1 − 𝑐𝑓)(𝑒𝑐)𝑒𝑞 =[𝑒+(1−𝑏)𝑐𝑓][1−(1−𝑏)𝑐𝑓]𝑒𝑓 =𝑒𝑐𝑓(𝑒𝑓)𝑒𝑞 =𝑒𝑐𝑓 +(1−𝑐𝑓)𝑅𝑑𝑚Fine grain fully confined within voids – Case iFigure 2.13: Intergranular soil mixture classification by Thevanayagam (2007a)As per above, by introducing a set of void ratio parameters in the context of: equivalentinter-granular, equivalent inter-fine, inter-granular, and inter-fine (as opposed to the globalvoid ratio), Thevanayagam (2007a,b) proposed a framework to characterize the shear load-ing response coarse-fine mixtures. Although this approach does not account for the factorssuch as soil fabric, plasticity, and aging effect, it provided a simplified two-sized sphericalbinary mixture approximation to depict the contact density variation as the fines contentincreased in the binary mixture. Following this framework, different investigators havereported different threshold values for the fines content that would change the materialresistance from a decreasing to an increasing trend. However, the determination of the por-tion of the fine grains that contribute to the active inter-grain contact [expressed through a“b parameter” by Thevanayagam (2007a)] becomes challenging when applied with respectto real life coarse-fine-grained soil mixtures, as the proposed method for the estimation ofthe b parameter is based on reanalysis of published binary packing studies.16The framework of Critical State Soil Mechanics (CSSM) by Roscoe et al. (1958); Schofield& Wroth (1968) recognizes void ratio and confining stress of soil as two key parameters thatcan be used to postulate the dilative (dense of critical soil) and contractive (loose of critical)responses. Further, Been & Jefferies (1985) introduced the concept of the state parameter(Ψ) to identify the difference between void ratios at a given state and at the critical stateon the same effective stress, and it became to be a prominent parameter in describing thebehavior of granular material due to its ability to eliminate the effects of confining stressand density, but also to acknowledge the combined effects due to confining stress, density.Deploying the frame work presented by Thevanayagam (2007a,b) with the use of stateparameter concept by Been & Jefferies (1985), several researchers have consistently triedto evaluate the shear resistance of sand-silt mixtures. Some of those studies can be listedas Baki et al. (2012); Belkhatir et al. (2014); Bensoula et al. (2014); Missoum et al. (2013);and Rahman & Lo (2014). Most of the conclusions deriving from above studies are based onthe assumption of the existence of unique equivalent granular steady state line (Equivalentgranular state parameter and instability stress ratios approximately follows a single lineirrespective of the fines content of the soil mixture). Additionally, the relationships of shearresistance and equivalent granular state parameter proposed in above studies have beenderived from limited number of laboratory test data. The review of the findings on themechanical responses of sand-silt mixtures indicate the necessity of further testing on suchmixtures and to enhance the current understanding in characterization of sand-silt mixtures.Inter-granular void ratioA review of the literature indicates that a number of researchers have investigated or usedthe concept of inter-granular void ratio, and also used different terminologies to refer intheir work. Prior to present those details, it is important to identify the volume and weightrelationship in a saturated soil mixture as presented in Figure 2.14. The total volume ofthe specimen is the summation of volume of voids (in saturated condition, it would bethe volume of water) and volume of solids (both fine-grained soil and coarse-grained soilare considered as solids). Similarly, soil-water weight-based relationships can be formed aillustrated in Figure 2.14.Coarse-grained soilFine-grained soilWaterVT = VC+VF+VWWT = WC+WF+WW𝑊𝑤 = 𝑤𝑊𝑠WS = WC+WF𝑊𝐹 = 𝐶𝐹𝑊𝑠𝑊𝐶 = 1− 𝐶𝐹 𝑊𝑠w = 𝑊𝑤/𝑊𝑠𝐶𝐹 = 𝑊𝐹/𝑊𝑠𝑉𝐶 =1 − 𝐶𝐹 𝑊𝑠𝐺𝑠𝐶𝛾𝑤𝑉𝐹 =𝐶𝐹𝑊𝑠𝐺𝑠𝐹𝛾𝑤𝑉𝑤 =𝑤𝑊𝑠𝛾𝑤VT – Total volume, VC – Volume of coarse-grained soil, VF – Volume of fine-grained soil, Vw – Volume ofwater, WT – Total weight, WC – Weight of coarse-grained soil, WF – Weight of fine-grained soil,WS – Weight of soil, Ww – Weight of water, w – water content, CF – Fines content, γw – Unit weight ofwater, GsC – Specific gravity of coarse-grains, and GsF – Specific gravity of fine-grainsFigure 2.14: Schematic illustrating volume and weight relationships for a saturated coarse-grained and fine-grained soil mixture17The work by Mitchell (1976) can be considered as the earliest attempt to introduce theconcept of ‘Void ratio in the granular phase’. Considering the weight and volume of eachphases of the granular soils with clay, Mitchell (1976) deduced the following expression:w100+C100GSC=(1− C100)eGGSC(2.1)where w = water content %, C = percent clay by weight, GSC = specific gravity of clay particles and eG= void ratio of granular phaseMitchell (1976) then noted that at least one-third of solids in soil mass needed to be clay,to result in a clay-dominated soil behavior – i.e., by the clay particles preventing directinter-particle contact of the coarser particles. This value has later identified to be in thesame range as the value for threshold fines content in accordance with the approach byThevanayagam (2007a).Shen et al. (1977) investigated the effect of fines on the response of Ottawa sand, withfixed dry density for all the specimens with different fines content. They reported that the‘Void ratio of the Sand structure’ varies with the fines content, however, did not present anexpression for the term they defined as ‘Void ratio of the Sand structure’.Lupini et al. (1981) investigated the residual strength of cohesive soil and identified theeffects of rotund particles on it as comparison to effects of platy particles, therefore theyimplied that interference between the rotund particles present in the soil should be reflectedmore directly by the ‘Void ratio of the granular phase of the soil’ represented by the symboleg, as per the Equation 2.2 below.eg =volume of platy particles and watervolume of rotund particles(2.2)While assessing a comparison of stress-strain and stress-path responses observed forspecimens prepared using mixtures of Brenda Mine tailings sand with Kamloops silt, Kuer-bis et al. (1988) introduce the ‘sand skeleton void ratio’, eskeletoneskeleton =[VTGsρw − (M −Msilt)]M −Msilt (2.3)where VT = total volume of the specimen, Gs = specific gravity of sand, ρw = density of water, M = totalmass of the specimen and Msilt = mass of silt in the specimenGeorgiannou et al. (1990) presented following relationship in defining ‘Granular void ra-tio’, eg, in an evaluation of a series of triaxial compression and extension tests.eg =volume of voids + volume of clayvolume of granular phase(2.4)Georgiannou et al. (1990) reported a higher granular void ratio for clayey sand than thevoid ratio of a clean sand. He postulated that it modifies the fabric of the sand, as contactconditions eventually causing changes in undrained brittleness at constant granular voidratio.As indicated earlier, with the assumption that the specific gravity of coarse and fineparticles are equal, Thevanayagam (1998) presented the below expression in defining ‘Inter-granular void ratio’. Thevanayagam (1998) suggested that inter-granular void ratio as an18index to represent the active coarser-granular frictional contacts that sustain the normaland shear forces in a coarse-fine mixture.es =(e+ FC100)(1− FC100) (2.5)where es = inter-granular void ratio, e = global void ratio, FC = fines content as a percentage of the totalweight of solidsMonkul & Yamamuro (2011) presented a general expression of inter-granular void ratiofor the cases for differences in specific gravity values for -coarse-grained and fine-grainedsoils as below.es =e+ (G/Gf )(FC/100)1− (G/Gf )(FC/100) (2.6)where e = overall (global) void ratio, G = specific gravity of the overall soil (weighted average of sand andsilt constituents), Gf = specific gravity of fines, FC = percentage of fines by total weight of dry soilThreshold fines contentAs elaborated earlier, it has been identified that the fines content up to certain value,the behavior of a coarse-fine mixture would be governed by the coarse-grain structure orcoarse-grain skeleton, and in those cases, the fines are considered to be inactively placed inthe structure without exerting any contribution to the loading bearing/transferring mecha-nisms. When the fines content surpasses a critical value, fine particles begin to dominate ingoverning the response. Threshold fines content denotes the change of soil behavior beinggoverned by the coarse particles to being dominated by the fine particles.If only particle arrangement in coarse-grains and fine-grains is considered without fo-cusing on the soil response against loading, increasing the fines content in a mixture wouldinitially decrease the maximum void ratio or minimum void ratio to a minimum value, thenfurther increase of fines content causes increase of those void ratios. The correspondingfines content that the decreasing trend void ratios switches to increasing trend can be iden-tified as threshold fines content. From the experimental observations, the minimum value ofeither minimum or maximum void ratio has been obtained when the fines content is some-where between 20% to 40%, and most commonly at 30% (Cubrinovski & Ishihara 2002;Thevanayagam 1998). The following text briefly presents the expressions which have beendeveloped by previous researchers in identifying or predicting the threshold fines content.The investigation on the packing density of spherical particles and results from the pack-ing arrangement derived for several of the single-sphere-size materials and binary, ternary,quaternary packing arrangements by McGeary (1961) paved the way for further investiga-tions by Lade et al. (1998) to identify the variation of void ratios of coarse-fine mixturesas the fines content varies. Although the shape of grain-size distribution curve and finescontent are studied on the effects of the values of minimum and maximum void ratios, Ladeet al. (1998) recognized the need of exploring the grain size and shape and also the methodof deposition in accounting for their effects on the packing arrangements in the mixture.Hazirbaba (2005) termed the transition point above which there are enough fines thatthe sand grains are no longer in contact as the limiting fines content, and they presented19the following relationship to compute the limiting fines content.Limiting Fines Content =WfinesWsand +Wfines=GsfesGsfes +Gss(1 + ef )(2.7)where Wfines = fines solid weight, Wsand = sand solid weight, Gsf = specific gravity of the fines, Gss =specific gravity of the sand, ef = void ratio of the fines, and es = maximum index void ratio of the sand.Thevanayagam (2007a) presented the following relationship for the upper bound of thresh-old fines content, CFth.CFth ≤ 100ec1 + ec + emax,HF% =100eemax,HF% (2.8)where ec = inter-granular void ratio, emax,HF = maximum void ratio of host fine-grain soilBased on calibrations with the data available for different sand–fines mixtures, Rahman& Lo (2008) developed the following expression to predict the Threshold Fines Content,TFC:TFC = A(11 + eα−βχ+11 + χ)(2.9)where A = 0.4 is an asymptotic value,α and β are calibrated values through curve fittings and have yield tobe 0.5 and 0.13 respectively, χ = D10/d50 particle size ratio of sand particle at 10% finer and fines particlediameter at 50% finerThe continuous effort to incorporating grain size, distributions and shape to enhancethe accuracy in predicting the threshold fines content has been further advanced by newermodels suggested by C. Chang et al. (2015), Zuo & Baudet (2015), C. Chang et al. (2016)Sarkar et al. (n.d.) to suggest novel models.2.3 Sand-Silt Mixture Specimen Preparation Methods forLaboratory TestsThe use of reconstituted soil specimens is a commonly adopted alternate approach ingeotechnical laboratory element testing primarily due to the difficulties in retrieving undis-turbed soil samples. In addition, reconstituted specimens are considered preferable for con-trolled studies of fundamental soil behavior because of the ability to prepare specimens withalmost identical particle fabric/structure through standardized reconstitution techniques.Considering the need to undertake the research proposed in this thesis through con-trolled testing of mixtures of sand and silt, specimen preparation technique(s) that lead torepeatable granular structure of such mixtures was considered of prime importance. Withthis objective, the Section 2.3 focuses on the available methods in current practice for recon-stituting of soils for direct simple shear and triaxial shear tests. Based on this, some of themethods considered for the present research are described in Section 2.3.1, and summarizedin Table A.2 in Appendix A. The challenges associated with the methods in preparation ofsand-silt mixtures are briefly presented in Section 2.3.2.202.3.1 Current Methods for Preparation of Reconstituted SpecimensAir pluviationSand can be air pluviated by using a sand rainer (a combination of tube, shutter and adiffuser) or else using spooning sand over a dispensing screen. The free-fall height, sizeand arrangement of the holes in the diffuser are influencing parameters of the density anddepositional energy of the sand specimen. Miyura & Toki (1982), Rad & Tumay (1987),Vaid & Negussey (1988), Vaid et al. (1999), and Wijewickreme, Sriskandakumar, & Byrne(2005) used the changes of fall-height to achieve different densities and reported that slowerpouring/raining rate increases the density, whereas faster rates may promote arching ofparticles resulting them to be locked into a looser state, thus yielding lower densities. Vaid& Negussey (1988) and Vaid et al. (1999) mentioned the importance of keeping the fallheight relative to the top of the specimen constant during the sand raining/pouring pro-cess as the specimen height increases to achieve uniform density. Air pluviation method isgenerally considered as an appropriate technique to produce sand specimens with a rangeof repeatable densities. However, it may produce non-uniform specimens for sand with siltsand silty sands, as visible segregation has been observed by F. Wood et al. (2008).Dry funnel depositionSands or sand-silt mixtures which are already placed into a funnel (the sprout of the funnelkept at the bottom of the mold) are deposited into the mold by slowly and vertically raisingthe funnel, allowing soil despoliation with essentially zero drop-height. Silty sand specimenswere prepared by this method by the investigations of Ishihara (1993), Lade & Yamamuro(1997), Yamamuro & Lade (1997), and Yamamuro & Covert (2001). The variations of thismethod include the gentle tapping on the funnel or rapid raising of the funnel to increasethe density of the deposited soil.Water pluviation/semdientationSands or sandy soil are gravity deposited in the water medium. To ensure the saturation,soils are thoroughly mixed with de-aired water, then boiled, let cool down to the roomtemperature and kept under vacuum in preparation for the pluviation. Finn et al. (1971),Mullins et al. (1977), Vaid & Thomas (1995), and Vaid et al. (1999) are some examplesof research work on specimens prepared through water pluviation. A change in fall-heightwould not influence the density of the deposited soil, as in most cases, the particles reachterminal velocity before they get deposited. Though gentle tapping or vibrating on the baseof the specimen could be done, after the completion of the pluviation, in attempt to achievehigher density, careful attention may be required to avoid possible sand boils specially insilty sands.Moist TampingSoils with some moisture are placed (preferably in many layers of specified thickness) inthe mold followed by desired levels of tamping. The placement and compaction that gen-erally takes place in construction practice is considered to be replicated in moist tamping.Castro (1969), Ladd (1974), Ladd (1978), Mullins et al. (1977), Been & Jefferies (1985),Pitman et al. (1994), Vaid & Thomas (1995), and Høeg et al. (2000) have practiced moisttamping method to prepare silty soil and sand-silt mixture specimens. Bradshaw & Baxter(2007) developed a modified moist tamping and reported three main response for the mod-ifications as: (i) quick and easy preparation of samples, (ii) the uniform compactive effort21is ensured throughout specimen, and (iii) ability to scale-up the compactive effort for thepreparation of larger samples for calibration chamber testing. However, the moist tampingmethod inherits criticism and challenges due to inevitable layering induced non-uniformityconcerns, undesirable volume changes during saturation process, and unrealistic (“honeycomb”) soil structure with non-uniform void ratios, compared to naturally deposited soils(Frost & Park, 2003; Vaid & Sivathayalan, 2000).Slurry DepositionInitially Kuerbis & Vaid (1988) presented slurry deposition method for the preparation ofsilty sands and sands with silts specimens. The de-aired water mixed with soil are boiledthen kept under vacuum to make a saturated solution (similar to the case of water pluvi-ation); then, the slurry was thoroughly mixed in a sealed mixing tube/container throughrotational shaking, before placing the soil in the test mold. The water content of the slurryhas been identified as a key variable in ensuring saturation and homogeneity of the preparedspecimens. With some variations in the mixing and placement, several versions of slurry de-position method have been practiced in the preparation of sandy silt and silt – for examplesCarraro & Prezzi (2008), Sanin (2010), Castelbaum et al. (2011), Tastan & Carraro (2013),Ahmadi-Naghadeh & Toker (2019) and Krage et al. (2020). For example, Castelbaum etal. (2011) have utilized a mixing apparatus and a method with an auger rotation and slurryinjection, inspired from the in-situ soil mixing. To produce homogeneous specimens across arange of plasticity indexes and particle gradations while preventing the particle segregation,Krage et al. (2020) used a rotational mixing chamber for thorough mixing under vacuum.Slurry DisplacementIn this method, a saturated slurry formed, by thorough mixing of only the fine-grainedsoils with de-aired water, is initially poured into the specimen mold. Another portion ofprepared homogeneous slurry can be gently mixed with coarse-grained particles to form auniformly mixed paste ensuring minimum air entrapment, before they are deposited slowlyby a spoon into the slurry bath making them to settle in the slurry while the slurry getsdisplaced. Ishihara et al. (1978) introduced this method in preparing coarser-grained soilhaving high amount of fines. Using a similar slurry displacement concept, Khalili & Wijew-ickreme (2008) developed an approach to prepare specimens of highly gap-graded mixturesof fine-grained mine waste and rock (mixture called “paste rock”), and they illustrated thatthe method would lead to specimens with very good uniformity and degree of saturation.They reported that the mix ratio between coarse and finer fractions of soil and density ofthe specimens could be controlled by changing the water content of the slurry.Slurry ConsolidationSheeran & Krizek (1971) introduced a slurry consolidation method to produce large boxsamples. A slurry with a water content of twice or more of the liquid limit of soil is placedand consolidated to a desired stress level in a one-dimensional large consolidometer. Theobtained sample can be then trimmed to make the test specimens. The slurry consolida-tion for specimen preparation is initially identified as a method that more suitable for highplasticity, fine-grained soils. To minimize the considerable amount of time for the consol-idation and the effects of trimming, Wang et al. (2011) consolidated the slurry inside themembrane-lined split mold used for triaxial testing. In this manner, the specimen mem-brane was stretched to match the walls of the mold by vacuum application; it also was notedto minimize the possible friction induced stress non-uniformity between the soil during the22consolidation process.2.3.2 Challenges in Preparing Sand-Silt Mixture SpecimensHomogeneity and uniformity of test specimens, in terms of soil matrix as well as stress-strainconditions, are critical considerations for obtaining high quality and meaningful data fromlaboratory element testing. Kuerbis & Vaid (1988) identified that reconstitution techniquesshould essentially possess the ability to produce specimens with the following attributes: (i)uniformity in terms of void ratio; (ii) fully saturated conditions (e.g., undrained testing); (iii)homogeneity (i.e., well mixed without particle segregation for fines-content uniformity); and(iv) mode of soil deposition similar to the application under study (e.g., gravity-depositedconditions); (v) meet the density range expected for the application under study (e.g.,density of an in-situ soil deposit). The specimen preparation methods, described in Sec-tion 2.3.1 and summarized in Table A.2, clearly show the need to exercise care in selectingthe method of reconstitution meet the requirements of a given application - as no singlemethod would be able to match all the needed requirements for the subject case.Vaid et al. (1999) compared the response of ‘undisturbed sand specimens’ (that werethought to be the best possible samples) retrieved by in-situ ground freezing with the corre-sponding reconstituted counterparts at identical initial states in terms of stresses and voidratios, and reported that the soil fabric resulted by the water pluviation closely simulatesthat of the natural alluvial and hydraulic fill sands. Although, the moist tamping methodcould be used to reconstitute soils spanning across very loose to dense void ratios, Vaid etal. (1999) and Frost & Park (2003) emphasized that the non-uniformity of the specimensprepared from moist tamping could lead to questionable test results. Ghionna & Porcino(2006) assessed the cyclic liquefaction resistance obtained from specimens prepared usingwater sedimentation and reported that they closely approximate response derived by test-ing of undisturbed samples from a marine water environment (in both isotropically andanisotropically consolidated tests).In summary, water pluviation/sedimentation method can be identified as a suitabletechnique, while acknowledging that there may be limitations in achieving target densitiesand possible segregation as the fines content increases – for example, in preparing sandspecimens when it needs to represent the in-situ soil fabric in a fluvial deposit.The behavioral differences and fabric alterations of undisturbed and reconstituted moist-tamped and slurry deposited specimens of Witwatersrand gold tailings, sandy silt, sandyclay were compared by N. Chang et al. (2011); they observed that neither moist tampingnor slurry deposition techniques could fully replicate the observed shearing response ofundisturbed samples. N. Chang et al. (2011) further reported that specimens prepared fromslurry deposition method generally replicated the fabric and behavior of the undisturbedsample better than those from moist-tamped specimens.Accordingly, the reconstitution of specimens of silty sands and sand with through slurrydeposition method can be considered a better option than those from moist compactingmethod, considering the homogeneity of the specimen and also the attractiveness in theability of the former to simulate the soil fabric found in fluvial deposits. Additionally,saturated conditions can be relatively easily achieved with slurry deposition. It is, however,to be noted that appropriate control of water content remains a significant concern, whenslurry deposition is used to prepare specimens to explore the effect of a wide range of fines onthe soil behavior. Thicker slurry (low water content) or “paste” raises issues of saturationof the specimen due to entrapped air both during mixing and placing of the slurry into23the split mold. On the other hand, thinner slurry (high water content) could entrap a lotof small air bubbles when it is poured into the specimen mold and also could promotesthe segregation of fine-grains from coarse-grains. Furthermore, since thinner slurry wouldundergo significant amount of volume strain during the consolidation process, it may alsoconsume longer periods for the initial slurry consolidation.Hyde et al. (2006) consolidated a silt slurry of twice the liquid limit of water content,using a piston in a steel cylinder, then extruded and sealed in a thin-walled sample tubes forlater assembly in the triaxial cell. The slurry consolidation in the forms of batch style (typ-ically achieved through a large consolidometer, where multiple specimens may be preparedfrom one batch) leads to use extrusion tubing or trimming the material to prepare TRX orDSS specimens where some degree of disturbance is inevitable. However, for low plastic siltsand sandy silts, slurry consolidation in batch style raises concern of significant disturbanceincluding the transferring specimen on to the base pedestal in the test device. In address-ing those issues, Wang et al. (2011) presented a new slurry consolidation technique, wherespecimens can be prepared directly on the base platen and then saturation, consolidation,and shearing can then be completed with the specimen in the same position. This methodconsume time for the preparation and consolidation, hence, the base pedestal of the testdevice would be occupied for the whole duration. To expedite the preparation process,Wang et al. (2011) developed a handling method in which the specimen could be preparedon a special experimental setup, secured by the split mold and then can be transferred intobase pedestal with minimum disturbance. Therefore, one specimen can be prepared whileanother one is being tested in the test device simultaneously. Wang et al. (2011) used vac-uum to assist achieving uniform stresses along the full height of the specimen during initialslurry consolidation (prior to the triaxial testing), reducing shear stresses at the specimenboundary. Barnes (2015) reconstituted Kamloops silt specimens deploying the techniqueby Wang et al. (2011) with slight modifications: (i) placing an extension collar on top ofthe split mold during slurry placement; (ii) no vacuum application during the initial slurryconsolidation; and (iii) removal of extension collar upon the completion of initial slurryconsolidation to trim the excess soil protruded beyond the split mold height. Barnes (2015)deployed an extension collar to provide additional height during slurry consolidation thatis required due to the larger volume of slurry at the loosest state of deposition. To securethe saturation of the slurry and workability during slurry mixing, higher water content maybe required depending on the soil type, thus extension collar provides the support for theexcess volume beyond the split mold height.Based on the above literature review, it appears that the slurry deposition methodand a version of slurry consolidation method would be the prudent candidates for thepreparation of silts with sands, sandy silt, and silt (specimens towards the silt end of thesand-silt mixtures spectrum). Once this approach is adopted, initial trials are required toidentify the appropriate water content of the slurry to optimize the slurry preparation withfocusing of saturation and segregation. Furthermore, initial trials would give insights onthe required volume to produce specimen of desired height or to identify the amount ofexcess material protruded beyond the split mold height. To minimize the disturbance inconsolidated slurry during the handling process, the specimen preparation directly on thetriaxial base (as opposed to transferring from another tube, etc.) was considered the bestviable option for the present study, although it was recognized that relatively long time willlikely be required to prepare the specimen to a level that is ready to be tested.242.4 Assessment of Cyclic Shear Resistance of Soil fromLaboratory Tests12.4.1 General RemarksThe response of soils under cyclic shear loading is a subject that has received wide attentionparticularly due to the well observed occurrence of liquefaction and associated geotechnicalhazards in loose saturated soils. The review of past work suggests that the definition of liq-uefaction from an engineering point of view is not straightforward. In describing the failureof Calaveras dam in California, Hazen (1920) introduced the term “liquefied”, and it is con-sidered as the origin of the term liquefaction. Specially after the observations made duringand after the 1960 Chile earthquake, 1964 Niigata earthquake, and 1964 Alaska earthquake,the term liquefaction triggering is now often associated with earthquakes (Davis et al., 1988).When investigating the soil behavior after Niigata earthquake 1964, Ohsaki (1966) describedthat the soil attains a liquid-like state due to decrease of effective grain to grain contactpressure (arising as a result of rise in excess pore-water pressure) - i.e., liquefaction due toearthquake shaking. Early investigations on earthquake induced soil liquefaction by Seed &Lee (1966), Yoshimi (1967), Seed & Lee (1967a,b), Terzaghi et al. (1968), Peacock & Seed(1968) , Kishida (1969) and Finn et al. (1970) used broad definitions as initial liquefac-tion, partial liquefaction, complete liquefaction to describe the phenomenon of liquefaction.Later, when extensive laboratory experiments [Seed & Peacock 1971; Finn et al. 1971; Shi-bata et al. 1972; Ishihara & Yasuda 1972; Yoshimi & Kuwabara 1973; Ishibashi & Sherif1974; Lee & Albaisa 1974; Lee et al. 1975; Castro 1975; Wong et al. 1975; De Alba et al.1976; Silver et al. 1976; Mullins et al. 1977; Martin et al. 1978; Vaid & Finn 1979; Andersenet al. 1980; Vaid & Chern 1983; Whitman 1985; Tatsuoka et al. 1986; Toki et al. 1986, ;Whitman 1987] on liquefaction were performed, it was necessary to define a criterion forliquefaction in order to identify, compare and analyze the strength or resistant of soil againstliquefaction. Many variations of broad definitions of liquefaction could already be seen inthe literature by late seventies, which led to considerable of inconsistencies in comparing theoutcomes from experiments undertaken by different researchers. For example, Marcuson IIIet al. (1978) highlighted that the terminology used to describe the liquefaction phenomenonand its effects created incongruities, ambiguities, and discrepancies. In an attempt to bringunity and consistency in usage, Marcuson III et al. (1978) (with the Geotechnical Division’sCommittee on Soil Dynamics in American Society of Civil Engineers) suggested a list ofrecommended definitions on terminology such as liquefaction, cyclic strain softening, porepressure ratio, peak pore pressure ratio, 100% pore pressure ratio, shear strength, cyclicshear resistance, limited flow strain, unlimited flow strain, ground failure, lateral spreading,sand boil, and flow failure.With the above definitions used as the basis, a number of criteria were proposed to definethe occurrence of liquefaction; these were mainly based on excess pore-water pressure ratioru [= ratio of excess pore-water pressure, (∆U) to initial effective vertical confining pressure,(σ′vco)], or axial/shear strain levels developed during the cyclic loading process [e.g., doubleamplitude (DA) axial strain in cyclic triaxial tests (CTX)]. The excess pore-water pressureratio (ru) equals to 100% was considered as the initial liquefaction or onset of liquefaction.As the experimental work progressed, it became customary to define liquefaction based on1A version of this section was published in Wijewickreme, D. & Soysa, A. 2016. Stress-strain pattern-basedcriterion to assess cyclic shear resistance of soil from laboratory element tests. Canadian GeotechnicalJournal, 53 (9), 1460-1473 doi: 10.1139/cgj-2015-049925the attainment of 5% DA axial strain in CTX (Poulos et al., 1985). Use of 5% DA axialstrain in a CTX [or ±2.5% single amplitude (SA) axial strain] as a criterion for assessing soilresistance against liquefaction became an attractive approach since it provided a convenientway of comparing the response observed from cyclic shear testing under different appliedcyclic loadings. For example, NRC (1985) used the results of the studies from Seed (1976)and Yoshimi et al. (1984) where ±2.5% SA and 5% DA axial strain were used as criteriato evaluate number of loading cycles in CTXs to establish cyclic resistance of certain soils.Ishihara (1996) stated that “the occurrence of 5% double amplitude axial strain in the cyclictriaxial will be taken up as a criterion to coherently define the state of cyclic softeningor liquefaction of the soils covering from clean sands to fines containing sands.” Whenusing cyclic direct simple shear (CDSS) test for liquefaction assessments, SA shear strainof ±3.75% in CDSS tests is equivalent to reaching a ±2.5% single amplitude strain inCTXs, simply based on the strain Mohr circle considerations and undrained conditions; assuch, this strain criteria has been used as “triggering point of liquefaction” by a number ofresearchers [e.g. Vaid & Sivathayalan 1996; Wijewickreme, Sanin, & Greenway 2005, andPorcino et al. 2012].In order to provide a comprehensive background, the currently available literature de-scribing the efforts to define “unacceptable performance” (i.e., definition of liquefaction orcyclic failure) are described detail in the following section and detailed description on theevolution of the criteria used in laboratory tests are briefed in Table A.3.2.4.2 Criteria or the Definition of Liquefaction or Cyclic FailureBasic cyclic shear response of soilsThree types of stress-strain responses have been observed from the data available fromlaboratory undrained (or constant-volume) element testing: liquefaction, cyclic mobilitywith limited liquefaction, and cyclic mobility without limited liquefaction. For example, inorder to distinguish between the seismic ground failures causing limited movements (e.g.,lateral spreading) and those involving large movements (e.g., flow slides), Seed & Peacock(1971) suggested the term “cyclic mobility”. Castro (1975) pointed out that liquefactiondevelops only in loose sand, but cyclic mobility can be induced in the laboratory even indense sand; in this definition of “liquefaction” type response, the soil experiences contractivedeformation until the steady state is reached – i.e., sand behaves in a contractive strain-softening manner with abrupt loss of stiffness and/or strength with development of largestrains. In the “cyclic mobility without limited liquefaction”, the shear strains increasegradual progressive manner with increasing number of load cycles – gradual buildup ofpore-water pressure although there is no strain-softening (Castro 1969; Vaid & Chern 1985).This behavior was termed “cyclic mobility with limited liquefaction” lies essentially betweenthe above two types: i.e., some limited strain softening occurs during the part of the loadcycles leading the state of phase transformation (Vaid & Chern, 1985) then the soil starts tobehave in a dilative manner during increasing parts of the loading cycle, and in subsequentunloading parts of the cycles, larger excess pore-water pressures are developed producinga state of zero effective stress as illustrated in Figure 2.15. The loading cycles that followwould cause a dilative response leading to a strain-hardening tendency. Research on thecyclic shear behavior of fine-grained soils (Boulanger & Idriss 2006; Bray et al. 2004; andSanin & Wijewickreme 2006) has indicated that some natural silts and tailings display cyclicmobility type strain development during cyclic loading with no liquefaction in the form of26strain softening accompanied by loss of shear strength.0510152025300Axial Strain, ε aNumber of cyclesm-15-10-505101520253035400Deviator Stress , (σd)Axial Strain, εaSteady Statem-15-10-5051015202530350(σ' 1-σ' 3) / 2(σ'1 + σ'3) / 2n-15-10-5051015202530350Deviator Stress , (σd)Axial Strain, εan-20-10010203040500(σ' 1-σ' 3) / 2(σ'1 + σ'3) / 20510152025300Axial Strain, ε aNumber of cyclesCyclicmobilityLimited liquefaction0510152025300Axial Strain, ε aNumber of cycles-15-10-505101520253035400(σ' 1-σ' 3) / 2(σ'1 + σ'3) / 2-10-5051015202530350Deviator Stress , (σd)Axial Strain, εaLiquefaction                 Limited Liquefaction            Cyclic MobilityStrain Development           Stress-path response               Stress-strain ResponseFigure 2.15: Typical liquefaction, limited liquefaction and cyclic mobility type responseduring undrained cyclic loading with associated stress-strain response, stress-path responseand strain development – Vaid & Chern (1985)Available criteria to define liquefaction or cyclic failureThe criteria that were developed over almost a half-century by past researchers to definethe onset of “liquefaction” or cyclic failure were investigated through an extensive literaturereview herein; detailed results of that literature review are presented in Table A.3, and onlythose criteria considered as most relevant are discussed below.Initially, the phenomena of liquefaction were identified and categorized based on the typeof mechanical response of soil observed during monotonic shearing. Many researchers haveused the attainment of 100% of ru, which is defined as the ratio of shear-loading-inducedexcess pore pressure to the initial effective confining stress, as the criteria for liquefactionsince soil has no effective stress under that condition. Based on the observations made27on cyclic loading tests for sand, Seed & Lee (1966) defined initial/partial liquefaction asthe stage where pore-water pressure equals to the initial effective confining stress for someshort time during the loading cycle; however, the failure of the specimen was defined asthe attainment of 20% DA axial strain in CTXs. In accordance with the definition of Seed& Lee (1966), in reported results from element tests, loose sand [Dr (relative density) =38%] “liquefied” in 9 cycles whereas dense sand [Dr = 76%] underwent 867 cycles priorto liquefaction indicating gradual accumulation of cyclic strain as shown in Figure 2.16.Consideration of the criterion that pore-water pressure develops and equals to the effectiveconfining stress as liquefaction was identified as a possible temporary state by Martin etal. (1975) based on laboratory testing with crystal silica sand, where they noted that thecondition of ru = 100% coincides with the point where the applied cyclic stress is zero inloading cycles. Additionally, Poulos et al. (1985) considered that the minimum undrainedstrength that remains after occurrence of ru = 100% is the undrained steady-state strength,which was noted to solely be a function of void ratio. Emphasizing that ru = 100% cannotbe used to estimate strength, Poulos et al. (1985) suggested that consideration of a definitionof liquefaction based on momentary ru = 100% condition is misleading and fundamentallyunsound. Assessing the liquefaction characteristics of silts and sandy silty with a comparisonto those of sand, Singh (1996) mentioned the issue in establishing criteria for liquefactionof undisturbed silts with respect to the combine consideration of pore pressure and straindevelopment. Singh (1996) further explained that although pore pressure criteria can beused to describe the initial liquefaction in sands and laboratory-prepared samples of silts,the criterion is not possible to use on most undisturbed silts as the 100% pore pressureincrease is not achieved during the testing of undisturbed silts.Axial Strain (%) Deviator Stress kg / cm2 -2.5-2-1.5-1-0.500.511.522.5Test No.209  e = 0.87  Dr = 38%  Cycle 1 Test No.209  Cycle 8 & 9 Cycle 9 Cycle 8 Test No.209  Cycle 10 Sample liquefaction on cycle 9 (a) Loose Sand, σ3 = 5 kg/cm2 (b) Dense Sand, σ3 = 5 kg/cm2 -2.5-2-1.5-1-0.500.511.522.5-0.2 -0.1 0 0.1 0.2Test No.206  e = 0.71  Dr = 78%  Cycle 1 -0.6 -0.4 -0.2 0 0.2Test No.206 Cycle 840 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6Test No.206 Cycle 867 Sample liquefaction on cycle 866 Figure 2.16: Hysteresis curves for cyclic loading test on loose and dense Sacramento RiverSand: Produced by extracting data from Seed & Lee (1966)Although Seed & Lee (1966) defined liquefaction based on the excess pore-water pres-28sure, failure criteria that they used for assessing their specimen in CTX was 20% DA axialstrain. Later, Gordon et al. (1974) considered 5% of SA strain as a criterion for failure,whereas Lee & Albaisa (1974) used a 5% DA, or ±2.5% SA strain as a failure criterion.In spite of this, studies by both Gordon et al. (1974) and Lee et al. (1975) considered ex-cess pore-water pressure equals to confining stress as the definition for liquefaction. Castro(1975) identified that substantial loss of soil shear strength that causes the soil mass toactually flow is the right condition of liquefaction, while gradual increase of strain withoutentailing a loss in shear strength is cyclic mobility; in this, DA strain of 5% was consideredto be cyclic failure. Since then, many studies have used axial strain criteria such as 2.5%,5%, 10% and 20% to define cyclic failure (refer Table A.3), and in turn, to estimate cyclicshear resistance . Seed & Idriss (1982) highlighted that for dense samples of sand, the onsetof liquefaction or cyclic mobility is not so abrupt and a critical condition is normally con-sidered to develop when the pore pressure ratio builds up to a value of 100% and the cyclicshear strain around that instance is about ∼=±5%. However, no evidence or fundamentalreasons for selecting a particular strain limit to define unacceptable performance have beenmentioned anywhere in the literature, in spite of the persistent use of a certain shear strainlimit as the criterion to define this performance level.When the available literature is examined, it can be noted that the single amplitudestrain levels of: 2.5% SA (Mullins et al., 1977), 3% to 5% SA (Lefebvre & Pfendler, 1996),1.4% to 2% SA (Boulanger & Idriss, 2004) and 3% SA (Bray & Sancio, 2006) have beennoted as approximately corresponding to initial liquefaction – which is defined as that statewhere the pore pressure would be equal to the effective confining stress. However, it is alsoof relevance to state that it has become customary to use the term liquefaction when 5%DA strain in a CTX has been achieved as noted by Poulos et al. (1985). Considering this5% DA CTX strain level criterion and using the Mohr Circle of strain and constant-volumecondition, Vaid & Sivathayalan (1996), Sriskandakumar (2004), Sanin (2005), Wijewick-reme, Sanin, & Greenway (2005) and Porcino et al. (2012) have noted that the 2.5% SAstrain in CTX is equivalent to achieving a SA shear strain of 3.75% in a CDSS test.A few other criteria for liquefaction can also be found in the literature. Gratchev &Sassa (2009) considered that sand specimens in ring shear tests as liquefied when ru =0.95. Development of the pore pressure and its characteristics during cyclic loading hasalso been used as criteria for liquefaction. Based on the observation of pore-water pressuredevelopments during CTXs for Toyoura sand, Tanimoto & Suga (1970) and Shibata etal. (1972) used rate of changing pore-water pressure during loading and unloading phasesduring unloading cycles to identify the onset of liquefaction. Similarly, abrupt increase ofpore pressure or shear strain was used to identify the initial liquefaction during ring torsionalshear tests conducted for saturated sand (Yoshimi & Oh-oka, 1975). In another approach,first turnover point in effective stress-path of octahedral stress during torsional simple sheartests on sand was considered as the initial liquefaction by Ishibashi & Sherif (1974). Chern(1985) defined true liquefaction and limited liquefaction emphasizing the strain softeningand limited/unlimited strain potential.It is clear that, although much focus has been on the development of pore pressure andshear strain accumulation during cyclic loading in defining “unacceptable performance”,there has been no sound evidence or fundamental reasons for selecting a particular ru valueor strain limit to define such performance. The criterion of ru ∼= 100% was developed basedon the observation of the response of loose saturated sands. The response of fine-grainedsoils has revealed that excess pore water pressure ratio would not reach 100%, despitethe gradual strain accumulation occurs during cyclic shear tests. It seems that axial strain29levels, which corresponds to the excess pore-water pressure ratio equals 100% that have beenobserved for the loose saturated sands, have been simply selected to develop a strain-basedcriterion. Despite the basis solely arising from tests on sands, the same strain-based criteriahave also been used to assess the cyclic shear resistance of fine-grained soils while usingdifferent arbitrary selected threshold strain values for strain-based criterion. Moreover, nocriteria have been developed for this purpose considering the stress-strain response (i.e.,stiffness), which is a more representative parameter in assessing engineering performancethan the excess pore-water pressure ratio alone or shear strain alone. This highlighted theneed and relevance to examine this aspect, which led to one of the key investigation themesof this thesis.2.5 SummaryShear loading response of Sand-Silt mixturesThe complexity in the understanding (or lack of) of the effect of fines on the liquefactionresistance is evident from the widely varying findings as systematically presented in Sec-tion 2.1. It is quite possible that the conflicting results with respect to the effect of finescontent in the performance of silt and mixtures may be arising from the manner in which thetest specimens were prepared. The disparities of the conclusions on the response of sand-siltmixtures made by previous researchers that were summarized earlier can be identified aslikely due to:ˆ Problems in defining the packing density - e.g., maximum and minimum void ratiosare based on methods using dry soil. Electrostatic forces make dry silt to adhere tosand grains; hence, maximum and minimum void ratios cannot be robustly definedfor silt/sand mixtures;ˆ Use of moist tamping to prepare silt-sand specimens – i.e., undesirable method knownto produce non-uniform and far-from-natural fabric;ˆ Lack of attention paid to seismic loading mode – i.e., use of the triaxial device thatdoes not address principal stress rotational effects;ˆ Not systematically covering the complete fines content spectrum in a given study toassess the overall response types and derive robust conclusionsThese conflicting findings suggest that there is a need to conduct more advance researchwhile systematically addressing the above considerations. With this background, an ad-vanced research program was undertaken to study the shear loading response of sand-siltmixtures forming the main scope of this thesis.Sand-Silt mixture specimen preparation methods for laboratory testsEach specimen reconstitution methods—such as air pluviation, dry funnel deposition, waterpluviation/sedimentation, moist tamping, slurry deposition slurry displacement, and slurryconsolidation—detailed in Section 2.3.1 is best suited for a selected type of soils and wouldnot work for all types of soils. Each method has inherent advantages and disadvantages.30The water pluviation and slurry deposition methods have been recognized to producespecimen that possess similar soil fabric and structure to those in the in-situ natural depo-sition of the sand and silt in a river deltaic, fluvial environment. Although water pluviationmethod and slurry consolidation method are considered the best suited for sand and siltrespectively, preparation of silty sand, sand-silt mixtures and sandy silt types of specimensneeds further attention and careful practices to ensure the saturation and homogeneity ofthe reconstituted specimens.It appears that the slurry deposition method and a version of slurry consolidationmethod would be the prudent candidates for the preparation of mixture of sand and silt forthis study, and it was decided to refine and further develop the methods as an integral partof the scope of this thesis.Assessment of cyclic shear resistance of soil from laboratory testsAs may be noted from the review of literature presented in Section 2.4, the liquefaction orcyclic failure are commonly identified or assessed based on the excess pore-water pressuredevelopment and/or the amount of accumulated axial or shear strain during cyclic loadingin laboratory experimental studies. Except that strain accumulation is generally commen-surate with excess pore-water pressure development, definition of onset of “liquefaction”or “cyclic failure” (unacceptable performance) based purely on an arbitrary cyclic strain isneither fundamental nor robust. Both pore-water pressure and strain-based criteria weredeveloped based on the observation made on the response of loose saturated sands, butlater those criteria have been used to assess fine-grained soils as well.Other than the criteria based on strain alone and/or excess pore-water pressure develop-ment, no criteria have been developed for defining unacceptable performance via consideringthe stress-strain response (or stiffness), which is a more significant parameter in address-ing engineering performance than the excess pore-water pressure ratio or shear strain alone.Therefore, it would be logical to examine signs of distinguishable changes in the overall cyclicstress-strain response as a way of finding the occurrence of unacceptable performance. Inparticular, the occurrence of significant change in stiffness (i.e., shape and pattern of stress-strain curve) would serve as a more effective indicator of performance deterioration froman engineering performance point of view; moreover, it is of relevance to note that signifi-cant rise of pore-water pressure and accumulation of shear strain are both expected to betypically reflected in stiffness changes.A criterion that is based on the stress-strain or stiffness characteristics of the soil wouldnot only have the advantage of reflecting the mechanical performance in an overall manner,but it also would be able to effectively applicable to the whole spectrum of soils from gravel,sand, silt to clay. With this thinking, it was considered important to examine this topicfurther and explore the possibility of developing a stress-strain-based criterion as a part ofthis thesis.31Chapter 3A New Criterion to Assess CyclicShear Resistance of Soil1The cyclic shear response of soils is commonly examined using undrained (or constant-volume) laboratory element tests conducted using triaxial shear (TRX) test and directsimple shear (DSS) test devices. The CRR from these tests is expressed in terms of thenumber of cycles of loading to reach an unacceptable level of performance that is definedin terms of the attainment of a certain excess pore-water pressure and (or) strain level.While strain accumulation is generally commensurate with excess pore-water pressure, thedefinition of unacceptable performance in laboratory tests based purely on cyclic straincriteria is not robust as shown in the review of available literature that is discussed inSection 2.4. The shear stiffness is a more fundamental parameter in describing engineeringperformance than the excess pore-water pressure alone or shear strain alone; so far, nocriterion has considered shear stiffness to determine CRR.With this objective in mind, initially, a number of observations on the undrained cyclicshear behavior of soils reported in the literature are assessed to establish the potential fora new stress-strain pattern-based approach for delineating unacceptable performance undercyclic shear loading; and presented in Section 3.1. The data recording intervals govern theidentification of the inherent shape of the stress-strain loop of soil during cyclic shearing,hence, the effect of data recording rate on the shape of stress-strain loops is examined inSection 3.2 since this is an important factor that will affect the pattern change observations.Then, an extensive database containing the stress-strain relationships obtained from CDSStests on sand, silt, clay, and mine tailings conducted at the University of British Columbia(UBC) are examined in Section 3.3 to verify the suitability of the new stress-strain pattern-based approach. The new approach is also compared with the commonly used shear strainbased criteria (i.e., attainment of SA shear strain of ±3.75% for unacceptable performanceduring cyclic loading), and the results are presented in Section 3.4. Moreover, conclusionswith respect to the new criterion to assess the cyclic shear resistance of soil are summarizedin Section 3.5.3.1 Assessment of the Potential for a New Stress-StrainPattern Based-ApproachThe observed stress-strain response of loose and dense Sacramento River sand during cyclictriaxial (CTX) tests by Seed & Lee (1966) is shown in Figure 2.16. As may be noted,the loose sand specimen experiences significant loss of stiffness in an abrupt manner whenmoving from the 8th cycle of loading to the 9th cycle, whereas dense sand survives the loading1A version of this chapter is published in Wijewickreme, D. & Soysa, A. 2016. Stress-strain pattern-basedcriterion to assess cyclic shear resistance of soil from laboratory element tests. Canadian GeotechnicalJournal, 53 (9), 1460-1473 doi: 10.1139/cgj-2015-049932well with only gradual degradation of stiffness for more than eight hundred cycles beforethe excess pore-water pressure becomes equal to confining pressure, which was consideredas liquefaction during this study. Similar abrupt loss in stiffness for contractive sand hasalso been observed by Wijewickreme, Sriskandakumar, & Byrne (2005) using data fromCDSS tests on Fraser River sand. To further assess the shape of stress-strain curves, thecyclic stress-strain response for the first cycle of loading and for the cycle at which -3%axial strain is reached, during undrained CTXs on initially hydrostatically consolidatedspecimens of fine-grained soils having four different plasticity indices, as reported by Bray& Sancio (2006) are shown in Figure 3.1. Furthermore, stress-strain loops derived from aCDSS test on over-consolidated Drammen clay by Andersen et al. (1988) shown in Figure3.2 provides the opportunity to identify the shape of the stress-strain loop at initial stageof cyclic shearing (i.e. 1st loading cycle - approximately 0.5% shear strain) with respectto that at later stage of cyclic shearing (i.e. 884th loading cycle - approximately 4% shearstrain)Axial Strain, εa (%) -40-30-20-10010203040-5 -4 -3 -2 -1 0 1 2 3 4 5D5-P2A LL = 25 PI = 0 e = 0.83 Cycle 11   Cycle 1   -5 -4 -3 -2 -1 0 1 2 3 4 5J5-P3A LL = 27 PI = 7 e = 0.75     Cycle 13   Cycle 1 -5 -4 -3 -2 -1 0 1 2 3 4 5A6-P6A LL = 38 PI = 11 e = 0.94 Cycle 15 Cycle 1 -5 -4 -3 -2 -1 0 1 2 3 4 5A6-P10A LL = 44 PI = 18 e = 1.09 Cycle 139 Cycle 1 Deviator Stress, q (kPa) Figure 3.1: Stress-strain relationship for the first cycle of loading (fine line) and the cycleat which -3 % axial strain is reached (thick line) for four specimen of increasing plasticity:Produced by extracting data from Bray & Sancio (2006), [LL: Liquid limit, PI: Plastic index,e: void ratio]Cycle 1 Cycle 700 Cycle 884 -40-30-20-10010203040-5 -4 -3 -2 -1 0 1 2 3 4 5Shear Stress τ (kPa) Shear Strain γ % -5 0 5Cycle 884 -1 0 1Cycle 1 Figure 3.2: Stress-strain behavior of Drammen Clay (Over consolidation ratio = 4) undersymmetrical direct simple shear loading: Produced by extracting data from Andersen et al.(1988)33Careful observation of Figures 2.16, 3.1, and 3.2 indicates that the stress-strain responsechanges from one initial pattern (termed Type X herein) to another (termed Type Y herein)with increasing number of loading cycles as follows:(i) Type X: Initially, the stress-strain response of the first cycle (small strain range) issmooth and incremental shear stiffness is maximum at the starting point of the loading;(ii) Type Y: At later stages of cyclic loading, the incremental shear stiffness is relativelylow during the starting potion of loading cycle and then, the incremental shear stiffnessincreases with increasing strain – i.e., strain hardening type response.For example, the stress-strain loops for cycle 1 in loose sand and dense sand testsin Figure 2.16 would represent the above described Type X pattern, whereas cycle 10 inloose sand test and cycle 867 in dense sand test would represent the Type Y pattern. InFigure 3.1, stress-strain loops depicted in thin lines would correspond to the Type X pattern;alternatively, the stress-strain loops depicted in thick lines demonstrate the characteristicsof Type Y pattern. In a general sense, Type X pattern is a relatively smooth loop; on theother hand, there is a clear noticeable ‘kink’ in the Type Y pattern of stress-strain loops.The transition from the early sign of kink in stress-strain response to the Type Y stress-strain response seems to take place relatively rapidly (in terms of the number of cycles)for soils exhibiting significantly contractive tendency; for example, the behavior presentedin Figure 2.16(a) for loose sand shows the transition to Type Y occurs in a matter of oneloading cycle after the signs of a kink has been observed. With materials exhibiting dilativetendency, this transformation would take place in a gradual manner (requiring several tomany loading cycles for the transition); for example, for the dense sand test shown inFigure 2.16(b), after reaching the signs of a kink, it took 25 additional loading cycles forthe stress-strain loops to gradually exhibit Type Y pattern.To further demonstrate this aspect for dilative soils, the results from a CDSS testsconducted by Sanin & Wijewickreme (2006) on a normally consolidated low plastic siltspecimen [PI=4] are illustrated in Figure 3.3. In addition to having all the load cycles inone graph, for the purpose of clarity, the cyclic stress-strain loops for the first eleven loadingcycles are also shown separately in Figure 3.3.The curve segment between positions (A) to (B) shown on the stress-strain loops inFigure 3.3 represents the first quarter of the loading cycle, while the curve segment betweenpositions (B) to (C) corresponds to the second and third quarters and finally, the positions(C) to (D) shows the fourth quarter of the loading cycle. By carefully scrutinizing the cyclicstress-strain loops in Figure 3.3, it can be seen that in first few cycles (up to 6th cycle), theincremental shear stiffness is highest at the starting point of loading (A), then it keeps ondecreasing till the stress increases to its maximum at (B). A similar trend continues fromsection (B) to (C) and section (C) to (D). However, in the 6th loading cycle, the trenddeviates from continuous reduction of incremental shear stiffness with increasing strain tomaintenance of almost same stiffness with increasing strain. In the 7th cycle [from (A) to (B)segment] and 8th cycle [both in (A) to (B) section and (B) to (C) section], ‘kinks’ can beobserved in the stress-strain loops as noted in Figure 3.3. The Type X pattern identified inthe initial loading cycles (greatest incremental shear stiffness at the beginning and gradualcontinuous reduction of incremental shear stiffness until the shear stress increases to itsmaximum) is gradually changed to a case with the visible ‘kinks’ in 7th and 8th loadingcycles. As may be noted, the ‘kinks’ gradually manifest in a more prominent manner withnext successive loading cycles as in cycle 11 showing the shape of fully-developed Type Ypattern described earlier. For the purpose of terminology, hereinafter, this initial or first34-2 0 2Cycle # 5 -20-15-10-505101520-2 -1 0 1Cycle # 4 -2 -1 0 1Cycle # 3 -1 0 1Cycle # 2 -1 -0.5 0 0.5 1Cycle # 1 -20-15-10-505101520-10 -5 0 5 10Shear Stress Shear Strain -10 -5 0 5 10Cycle # 11 -10 -5 0 5Cycle # 10 -10 -5 0 5Cycle # 9 -20-15-10-505101520-5 0 5Cycle # 8 -4 -2 0 2 4Cycle # 7 -4 -2 0 2Cycle # 6 A B C D A B C D A B C D A B C D “Incipient Kink” “Incipient Kink” Figure 3.3: Cyclic shear stress-strain relationship of normally consolidated Fraser River silt(PI = 4) specimen with a cyclic stress ratio of 0.17, with an initial vertical effective stressof 100 kPa during a CDSS test - data from Sanin & Wijewickreme (2006)sign of the development of a distinguishable kink (such as that shown in 7th cycle in Figure3.3), is defined as the occurrence of ‘incipient kink’.With this background, a new stress-strain pattern-based approach for delineating unac-ceptable performance under cyclic shear loading is proposed below as a criterion to definecyclic shear resistance from laboratory tests. The method uses a stiffness-based criterionto determine unacceptable performance of a given material during constant-volume cyclicshear loading, and it involves observing the evolution of the strain-stain loop pattern changeas the number of constant-amplitude sinusoidal shear stress loading cycles applied to thespecimen is increased.Knowing that the stress-strain loops at the beginning of cyclic loading are Type X andthose at the end of cyclic loading are Type Y, the key is to look for the point at which thestress-strain pattern begins to transition from Type X to Type Y. In most instances, theoccurrence of such incipient kink serves as a ‘flag’ that indicates the first sign of deviatingfrom Type X behavior. Herein, it is proposed that the load cycle number corresponding tothat point of initiation of transition be considered as the number of cycles for unaccept-able performance for that applied cyclic stress ratio. In an overall sense, the process foridentification of incipient kink from a given cyclic shear test comprises two aspects:(i) examination of the stiffness variation within a given loop; and;(ii) observation on the evolution of the pattern change between strain-strain loops as35the strain development takes place.The recognition of the formation of an ‘incipient kinks’ in the stress-strain loop as theload cycles progress as illustrated in Figure 3.4 assists defining unacceptable performance(cyclic failure); it is, however, to be noted that localized changes in stiffness in a givenstress-strain loop that do not contribute to changing the overall loop-over-loop stress-strainpattern should not be counted in determining the ‘incipient kink’. It is also relevant to notethat seeking for the ‘incipient kinks’ is to assist the process of demarcating a transitionpoint from Type X to Type Y behaviors, as opposed to the definition of a stage of failure.The suitability of the approach for both fine-grained and coarse-grained material is ex-amined and verified in the sections below using the extensive database derived from cyclicshear testing performed at UBC. The database contains the cyclic stress-strain relation-ships for different materials such as sand, silt / clay, and mine tailings obtained fromstress-controlled constant-volume CDSS tests with loading frequency of 0.1 Hz conductedat UBC including those by Sriskandakumar (2004), Sanin (2005, 2010), James et al. (2011),Seidalinova (2014), and Soysa (2015).Stress-strain loop without a kink Stress-strain loop with visible "IncipientKink" Stress-strain loop with a kink Figure 3.4: ‘Incipient kink’ in a cyclic stress-strain loop3.2 Effect of Data Recording RateIn laboratory experimentation, the data recording interval/frequency plays an importantrole in gathering high quality data to capture the response of the test material, and therebyobtaining meaningful and accurate engineering parameters. As such, it is reasonable toexpect that the number of data points used to plot the shear stress-strain loops from thecyclic shear tests has the potential to impact the assessment of the incipient kink throughvisual observations, and this aspect is investigated below.In all tests that were considered for the study, the data were collected using a high-speed data acquisition system. The data collected at a high frequency (some 1000 pointsper second) was then averaged so that 10 data points collected per second was availablefor plotting the shear stress-strain response. Since all the cyclic direct simple shear testswere performed with constant-amplitude sinusoidal loading applied at 0.1 Hz (or with a 10-second period), 100 data points were available for plotting the response over a given loadcycle. It was considered prudent to assess the adequacy of this rate to accurately capturethe instance of incipient kink; with this in mind, the stress-strain response for a given testwas plotted in Figure 3.5. representing three cases of data recording rates slower than 10036per loading cycle (i.e., 50, 20, and 12 data points per loading cycles).100 Data Points 50 Data Points 20 Data Points 12 Data Points Kink Kink Kink Kink Kink Kink Figure 3.5: Effect of data recording rates in distinguishing the ‘incipient kink’As may be visually noted, the depicted stress-strain loop is not affected up until thedata recording rate has been reduced to about 20 data points per a loading cycle. Assuch, the rate of data recording rate of 100 data points per cycle used in the UBC testingconsidered herein was considered quite appropriate for delineating the incipient kink.3.3 Cyclic Stress-Strain Pattern Development for Soil3.3.1 Coarse-grained soils - sandThe cyclic stress-strain loops derived from constant-volume CDSS tests conducted at UBCfor Fraser River sand, two other river sands (called Sand Type I and Sand Type II) and atailings silty sand were visually observed to examine the occurrence of incipient kinks forcoarse-grained soils. The specimens for the tests conducted on Fraser River sand were air-pluviated, Sand Types I and II were water pluviated, whereas the tailings sand was preparedusing a slurry deposition process. Typical results from the tests showing the stress-strainloops immediately prior (Type X pattern), during the initial transition point (or incipientkink occurrence), and subsequent to the transition (Type Y pattern) are shown in Figure3.6.It can be noted that for these relatively contractive sandy soils, transformation fromType X stress-strain loop pattern to Type Y occurs relatively abruptly – i.e., note thatthere is only two cycle difference between the 1st and 3rd column graphs in Figure 3.6 forall rows. Therefore, it becomes straightforward to declare that the loading cycle shownon the graphs in the 2nd column of Figure 3.6 would correspond with the occurrence ofincipient kink (or change of pattern in stress-strain loop occurs from Type X to Type Y).The stress-strain loops prior to the occurrence of incipient kink (graphs on the 1st columnof Figure 3.6) clearly show the reduction of incremental shear stiffness with increasing shearstrain in both loading and unloading cases for all considered sand types. In Figure 3.6,it is also notable that the specimens undergo significant shear strain during the stress-strain loop with the incipient kink (graphs on the 2nd column), or immediately after theoccurrence of incipient kink (graphs on the 3rd column). Cyclic stress-strain responses ofsand obtained CDSS tests which were used for the study are summarized in Table 3.1 withthe details of the material properties and the CDSS tests parameters such as Cyclic StressRatio amplitude (CSR) and σ′vco. It is of interest to note that once the stress-strain loophaving the initial transition point (or incipient kink) is reached, the shear strain of 3.75% is37Fraser River Sand CSR* =0.08 σ′vco = 100 kPa Dr = 40 %     River Sand Type I CSR* =0.15 σ′vco = 100 kPa Dr = 50 %      River Sand Type II CSR* =0.11 σ′vco = 100 kPa Dr = 51 %      Tailings Silty Sand CSR* =0.08 σ′vco = 100 kPa ec = 0.708 -10-50510-0.4 -0.2 0 0.2 0.4Shear Stress (kPa) Shear Strain % Cycle # 16 -2 -1 0 1Cycle # 17 -10 -5 0 5 10Cycle # 18 -20-1001020-1 -0.5 0 0.5Shear Stress (kPa) Shear Strain % Cycle # 2 -4 -2 0 2Cycle # 3 -10 -5 0 5Cycle # 4 -12-60612-0.5 -0.3 -0.1 0.1Shear Stress (kPa) Shear Strain % Cycle # 2 -3 -2 -1 0 1Cycle # 3 -10 -5 0 5Cycle # 4 -10-50510-1.5 -1 -0.5 0 0.5Shear Stress (kPa) Shear Strain % Cycle # 34 -4 -2 0 2Cycle # 35 -10 -5 0 5Cycle # 36 Incipient Kink Incipient Kink Incipient Kink Incipient Kink Figure 3.6: Cyclic stress-strain loops of sands in constant-volume CDSS testsreached almost simultaneously or within the immediate following loading cycle (can be seenin Figure 3.6), indicating that the numbers of loading cycles required for the occurrenceof stress-strain pattern change and for the attainment of ±3.75% shear strain are verysimilar for the tested sand. This is well illustrated in Figure 3.9(A) where the number ofloading cycles for the occurrence of stress-strain pattern change is directly compared withthe number of loading cycles to reach ±3.75% shear strain. This provides further evidenceto the postulate presented herein that the loading cycle that is associated with the changeof stress-strain pattern (or the occurrence of the incipient kink) is the true beginning of thedeterioration of the cyclic stress-strain behavior of the coarse-grained materials, and alsoindirectly supports the rationality of previous selection of 2.5% SA axial strain in a CTX,or 3.75% SA shear strain in a CDSS test, as a criterion for the triggering of liquefaction insands.383.3.2 Fine-grained soils - silt and clayThe stress-strain responses obtained at UBC by CDSS tests on relatively undisturbed nat-ural samples of Fraser River silt, Serpentine River sediments, Fraser River deltaic clay, andKitimat clay—retrieved using thin-walled tube samplers—were examined with respect to theproposed stress-strain pattern-based criterion. Some results from testing over-consolidatedand reconstituted material are also included in the database considered herein (i.e., over-consolidated Fraser river silt and reconstituted Serpentine River sediments, Fraser Riverdeltaic clay and Kitimat clay) and Table 3.2 summarizes the details of the material prop-erties and the CDSS tests parameters of those fine-grained soil.-15-10-5051015-1 0 1Shear Stress (kPa) Shear Strain % Cycle # 1 -2 -1 0 1Cycle # 19 -4 -2 0 2Cycle # 39 -40-2002040-1.5 -0.5 0.5 1.5Shear Stress (kPa) Shear Strain % Cycle # 2 -4 -2 0 2Cycle # 8 -10 -5 0 5Cycle # 16 -50-2502550-3 -2 -1 0 1 2Shear Stress (kPa) Shear Strain % Cycle # 1 -4 -2 0 2Cycle # 8 -8 -4 0 4Cycle # 17 -20-1001020-2 -1 0 1 2 3Shear Stress (kPa) Shear Strain % Cycle # 1 -3 -1 1 3Cycle # 4 -5 0 5Cycle # 10 Fraser River Silt (Normally Consolidated) CSR* =0.15 σ′vco = 85 kPa ec = 0.94  Serpentine River Sediments CSR* =0.17 σ′vco = 201 kPa ec = 0.88   Fraser River Deltaic Clay CSR* =0.26 σ′vco = 152 kPa ec = 1.69    Kitimat Clay CSR* =0.24 σ′vco = 80 kPa ec = 1.01   Incipient Kink Incipient Kink Incipient Kink Incipient Kink Figure 3.7: Cyclic stress-strain loops of silts and clay in constant-volume CDSS testsThe typical cyclic stress-strain loops derived from the testing of these natural fine-grained soils are shown in Figure 3.7. Visually identified incipient kinks are also presentedin Figure 3.7 with the stress-strain loops corresponding to reaching a shear strain limit of39±3.75%. It is of value to note the difference observed between contractive sands and fine-grained material with respect to the number of loading cycles needed to transition froma stress-strain loop with incipient kink to stress-strain loop with a fully developed kink.As shown earlier in Figure 3.6, in relatively contractive sands, this transition seem to takeplace within one to two loading cycles from the occurrence of the incipient kink; whereas infine-grained soils (as in Figure 3.7), the fully developed kink has been observed to manifestgradually over several number of loading cycles since the incipient kink.The number of loading cycles to observe stress-strain pattern change stage is comparedwith those to reach ±3.75% SA strain limit in CDSS tests for natural silt and clay in Figure3.9(B). Unlike for the sands, it is apparent that the number of cycles to reach the incipientkink for fine-grained material is less than those required to reach a shear strain limit of±3.75%.3.3.3 Fine-grained soils - mine tailingsThe cyclic response of a range of fine-grained undisturbed laterite tailings, copper tailings,red mud tailings, and reconstituted gold tailings during constant-volume CDSS tests werealso examined. Both the laterite tailings and copper tailings were normally consolidatedprior to the cyclic loading, whereas, the results for gold tailings arise from tests conductedon soil specimens that were initially normally consolidated and over-consolidated as shownin Table 3.3. Furthermore, the test series for red-mud tailings were conducted after initialnormal consolidation followed by an application of pre-determined initial static shear stressbias (i.e., to simulate sloping ground conditions in the field). Regardless of the initialstatic shear stress bias and consolidation condition, the pattern changes in the stress-strainresponses in terms of the incipient kink were identifiable as shown in Figure 3.8. Similarto the incipient kinks observed for the natural fine-grained soils (as per Figure 3.7), theemergence of the fully developed kinks for tailings also takes place over several loadingcycles.Again, the number of loading cycles to observe stress-strain pattern change in stress-strain loops of the mine tailings is compared with those to reach ±3.75% SA strain limitin CDSS tests in Figure 3.9(C). Similar to those observed for fine-grained natural soils inFigure 3.9(B), the number of loading cycles required for stress-strain pattern change for themine tailings is smaller than those required to reach a ±3.75% SA strain limit.40Laterite Tailings CSR* =0.234 σ′vco = 100 kPa ec = 1.32  OCR = 1  Copper Tailings CSR* =0.2 σ′vco = 94 kPa ec = 0.987 OCR = 1  Red-mud Tailings CSR* =0.2 σ′vco = 100 kPa ec = 1.981 OCR = 1 τh/σvco = 0.077    Gold Tailings Type I CSR* =0.48 σ′vco = 51 kPa ec = 0.73 OCR = 8    Gold Tailings Type II CSR* =0.10 σ′vco = 102 kPa Dr = 0.62 OCR = 1 -20-1001020-2 -1 0 1Shear Stress  (kPa) Shear Strain % Cycle # 3 -4 -2 0 2Cycle # 5 -6 -4 -2 0 2Cycle # 8 -30-1501530-1 -0.5 0 0.5 1Shear Stress (kPa) Shear Strain % Cycle # 1 -2 -1 0 1 2Cycle # 3 -5 0 5Cycle # 6 Incipient Kink Incipient Kink -30-1501530-1 -0.5 0 0.5 1Shear Stress (kPa) Shear Strain % Cycle # 5 -2 -1 0 1 2Cycle # 10 -5 0 5Cycle # 14 -15-551525-0.5 0.5 1.5 2.5Shear Stress (kPa) Shear Strain % Cycle # 5 0 1 2Cycle # 8 0 2 4Cycle # 25 Incipient Kink Incipient Kink -10-50510-3 -1.5 0 1.5 3Shear Stress (kPa) Shear Strain % Cycle # 16 -4 -2 0 2Cycle # 17 -10 -5 0 5Cycle # 18 Incipient Kink Figure 3.8: Cyclic stress-strain loops of mine-tailings in constant-volume CDSS tests411101001 10 100Ncyc γ = 3.75 % Ncyc Stress-strain pattern change Fraser River SandRiver Sand Type IRiver Sand Type IITailings Silty Sand11010010001 10 100 1000Ncyc γ = 3.75 % Ncyc Stress-strain pattern change  Fraser River SiltSerpentine River SedimentsFraser River Deltaic ClayKitimat Clay1101001 10 100Ncyc γ = 3.75 % Ncyc Stress-strain pattern change Lateriate TailingsCopper TailingRed Mud TailingGold Tailings IGold Tailings II(A) (B) (C) Figure 3.9: Comparison of the number of cycles for the occurrence of stress-strain pattern change in cyclic stress-strain loop and to reach3.75 % shear strain for (A)Fraser River Sand, River Sand type I & II and Tailings Silty Sand; (B) Fraser River Silt, Serpentine RiverSediments, Fraser River Deltaic Clay and Kitimat Clay; and (C) Laterite, Copper, Red Mud Tailings and Gold Tailings type I and II42Table 3.1: Material properties, test conditions, and parameters for coarse-grained materials used in CDSS tests.MaterialComposition (%)CuD50Dr ec σ′vcoCSR No. ofGravel Sand Silt Clay mm (τcyc/σ′vco) testsFraser River Sand - 100 - - 1.6 0.26 0.40–0.44 100–200 0.08–0.15 7River Sand Type I 1 96 3 - 2.3 0.225 0.49–0.70 100–150 0.11–0.23 6River Sand Type II - 99 1 - 2.2 0.224 0.47–0.51 100–150 0.09–0.20 4Tailings Silty Sand - 70 23 7 15 0.114 0.70–0.75 50–200 0.08–0.15 9Table 3.2: Material properties, test conditions, and parameters for fine-grained materials used in CDSS tests.MaterialComposition (%)PI Specimen condition ec σ′vcoCSR No. ofSand Silt Clay (τcyc/σ′vco) testsFraser River Silt 13 77 10 4 Normally conoli-dated0.82-0.97 85–400 0.12–0.29 18Overconsolidated* 0.85-0.91 100 0.20–0.21 5Serpentine River Sediments 16 56 28 7 Undisturbed 0.88-1.33 100–200 0.12–0.20 14Reconstituted 0.66-0.68 200 0.10–0.16 3River Sand deltaic clay - 14 86 34 Undisturbed 1.66-1.84 75–150 0.2–0.33 7Reconstituted 1.21-1.41 75 0.15–0.25 3Kitimat clay 1 54 45 17 Undisturbed 0.96-1.01 80 0.17–0.26 3Reconstituted 0.71-0.77 80 0.16–0.24 3*OCR = 1.3-2.1Table 3.3: Material properties, test conditions, and parameters for mine tailings used in CDSS tests.MaterialComposition (%)Gs ec Dr σ′vcoCSR No. ofSand Silt Clay (τcyc/σ′vco) testsLaterite tailings - 65 35 4.1 1.37-1.45 - 100 0.11–0.25 57 91 2 4.1 1.12-1.39 - 200 0.15–0.25 4Copper tailings 24 58 18 2.8 0.99-1.09 - 100 0.20–0.27 2Red-mud tailings* 2 81 17 3 1.52-1.98 - 100-200 0.20–0.27 4Gold tailings type I † 29 44 27 2.9 0.53-0.73 - 50 0.16–0.65 9Gold tailings type II 34 55 11 3.8 - 0.58-0.69 100-400 0.07–0.16 20*α (Initial static shear bias ratio) = 0.077; † OCR = 1-8433.4 Discussion on the Stress-Strain Pattern-Based CriterionPerformance of soils under cyclic shear loading is commonly assessed on the basis of excesspore-water pressure and/or the level of shear strain development during cyclic loading.Since significant rise of pore-water pressure and accumulation of shear strain are expectedto reflect in stiffness changes, it was considered appropriate to evaluate the cyclic resistancein terms of changes in incremental stiffness, along with the variations in the patterns ofstress-strain loops, during cyclic loading.From the observations from cyclic shear test data, it becomes clear that there is asignificant difference in the soil stress-strain pattern between the initial and later stages ofload cycles during undrained (or constant-volume) cyclic loading (i.e., Type X versus Type Ybehavioral patterns as distinguished for identification purpose). The soil materials exhibita relatively smooth shear stress-strain loop patterns in the early stages of cyclic loading(Type X); at large strain levels, the stress-strain loop pattern would typically display a fullydeveloped kink. The method proposed herein is focused on determining the load cyclenumber corresponding to the point of initiation of transition from Type X to Type Y andthen considering that point as the number of cycles for unacceptable performance for theapplied cyclic stress ratio during the subject laboratory test.As noted in the previous sections (see Figure 3.6), in relatively contractive sands, thetransition of the stress-strain loop from Type X to Type Y stress-strain loop pattern occursrelatively abruptly – i.e., taking only about one or two loading cycle to change from TypeX to Type Y patterns. As such, in such soils, the identification of the loading cycle for thatchange of pattern in stress-strain loop (or the incipient kink) becomes relatively straightfor-ward. In contrast, for the case of fine-grained soils (natural soils and some mine tailings asshown in Figure 3.7 and 3.8), several number of loading cycles are needed to transition fromType X to Type Y. In such situations, it is required to examine the stiffness variation withina given stress-strain loop with the objective of seeking the first sign of deviating from TypeX behavior towards Type Y; in this, the initial occurrence of a potential ‘incipient kinks’(as illustrated in Figure 3.4) in the stress-strain loop well signifies that point of transitionfrom Type X to Type Y behavioral pattern.It is of value to compare the similarities and differences between the cyclic resistancederived from the proposed new stress-strain pattern change approach and the commonlyused shear-strain-based approach. In view of this, the CRR and number of loading cycleplots estimated using the above two approaches, considering selected test series from dif-ferent soils considered herein, are presented separately in Figures 3.10, 3.11, 3.12, and 3.13.Namely: Fraser River Sand at vc = 100 and 200 kPa, Fraser River Silt at σ′vco = 100 and200 kPa, Fraser River Deltaic Clay at σ′vco = 75 and 150 kPa, and Laterite Tailings at σ′vco= 100 and 200 kPa were considered; only two test series per soil was selected for the brevity.The arrows on the figures are used to pictorially indicate when the number of loading cyclesderived from stress-strain pattern change based criterion is observably different to thosefrom shear strain based criterion for fine grained soils.As shown in Figure 3.10, in the case of sands, the CRR curves developed from thetwo approaches are essentially similar. This is as expected since the observations for theindividual points from all tests [see Figure 3.9(A)] indicated that the number of loadingcycles for the pattern change of stress-strain loops with the occurrence of incipient kinks inthe stress-strain loops are almost equal to the number of loading cycles to reach a±3.75% SAshear strain level (i.e., currently well-established shear strain-based liquefaction assessmentcriteria).4400.050.10.150.21 10 100Cyclic  Stress Ratio, CSR  Number of loading cycles Stress-strain pattern change basedStrain based (ϒ=±3.75%) Fraser River Sand σ'vc = 100 kPa 1 10 100Number of loading cycles Stress-strain pattern change basedStrain based (ϒ=±3.75%) Fraser River Sand σ'vc = 200 kPa Figure 3.10: Comparison of cyclic stress ratio vs. number of loading cycles derived fromthe proposed stress-strain pattern-based criterion and the shear strain criterion for CDSStests performed on Fraser River Sand00.050.10.150.20.251 10 100Cyclic  Stress Ratio, CSR  Number of loading cycles Stress-strain pattern change basedStrain based (ϒ=±3.75%) Fraser River Silt σ'vc = 100 kPa 1 10 100Number of loading cycles Stress-strain pattern change basedStrain based (ϒ=±3.75%) Fraser River Silt σ'vc = 200 kPa Figure 3.11: Comparison of cyclic stress ratio vs. number of loading cycles derived fromthe proposed stress-strain pattern-based criterion and the shear strain criterion for CDSStests performed on Fraser River Silt4500.050.10.150.20.250.30.351 10 100Cyclic  Stress Ratio, CSR  Number of loading cycles Stress-strain pattern change basedStrain based (ϒ=±3.75%) Fraser River Deltaic Clay σ'vc = 75 kPa 1 10 100Number of loading cycles Stress-strain pattern change basedStrain based (ϒ=±3.75%) Fraser River Deltaic Clay σ'vc = 150 kPa Figure 3.12: Comparison of cyclic stress ratio vs. number of loading cycles derived fromthe proposed stress-strain pattern-based criterion and the shear strain criterion for CDSStests performed on Fraser River Deltaic Clay00.050.10.150.20.250.31 10 100Cyclic  Stress Ratio, CSR Number of loading cycles Stress-strain pattern change basedStrain based (ϒ=±3.75%) Laterite Tailings σ'vc = 100 kPa 1 10 100Number of loading cycles Stress-strain pattern change basedStrain based (ϒ=±3.75%) Laterite Tailings σ'vc = 200 kPa Figure 3.13: Comparison of cyclic stress ratio vs. number of loading cycles derived fromthe proposed stress-strain pattern-based criterion and the shear strain criterion for CDSStests performed on Laterite TailingsOn the other hand, the CRR curves developed for the fine-grained materials such asnatural silt (Figure 3.11 for Fraser River silt) and clay (Figure 3.12 for Fraser River deltaicclay) as well as mine tailings (Figure 3.13 for Laterite tailings) indicate that the strain-basedcriterion gives rise to a relatively higher CRR in comparison with those developed based onthe stress-strain pattern change with the occurrence of incipient kink in cyclic stress-strainloop. The CRR curves derived for the other fine-grained soils used for the study also follow46that trend.From an engineering performance point of view, it can be argued that the proposednew method is robust as changes in shear stiffness is a more effective indicator of materialperformance, than a simple value of a shear strain or excess pore-water pressure. In thiscontext, the CRR values derived using the new method deserves attention since those valuesseem to be lower than the CRR estimated for the results derived from fine-grained naturaland tailings materials considered herein.3.5 Conclusions on the Stress-Strain Pattern-BasedCriterionA new shear stiffness-based criterion is proposed to determine cyclic shear resistance (i.e.,unacceptable shear performance) in terms of the occurrence of pattern change of stress-strain loops obtained from laboratory cyclic shear tests. The proposed new criterion isassessed with the results of current practice of strain-based criterion using cyclic stress-strain responses obtained from 126 cyclic direct simple shear tests [Sand (26), silt and clay(56), tailings (44)] conducted at the University of British Columbia, Canada. The details ofdevelopment, assessment, verification and limitations that needs to address in future workrelated to the new shear stiffness-based criterion are summarized with conclusions below.ˆ The method involves following the evolution of the strain-stain loop pattern changeas the number of shear stress loading cycles applied to the specimen is increased. Theapproach became feasible since it was observed that consistent inherent differencesdo exist in the stress-strain loop patterns between initial and later stages of a givencyclic loading process for many soil types ranging from coarse-grained to fine-grainedmaterials – i.e., the soil materials typically exhibit a relatively smooth shear stress-strain loop patterns in the early stages of cyclic loading (called Type X pattern);at large strain levels, the stress-strain loop pattern would typically display a fullydeveloped kink (called Type Y pattern).ˆ The method, through careful visual examination, determines the point of initiation oftransition from Type X type to Type Y (from relatively ‘smooth’ stress-strain loopswithout ‘kink’ to stress-strain loops with ‘kink’ ). The load cycle number correspond-ing to the point of initiation of transition is considered as the number of cycles forunacceptable performance for that applied cyclic stress ratio. Since changes in shearstiffness is a more effective indicator of material performance, than a simple valueof a shear strain or excess pore-water pressure, the occurrence of pattern change instress-strain loops also provides a robust and meaningful basis to determine cyclicshear resistance (or unacceptable performance) in laboratory cyclic shear tests.ˆ In the case of relatively contractive sands, the transition from Type X to Type Ystress-strain loops patterns occurred in a rapid manner, and it was found that theCRR developed from the new approach is essentially similar to that derived fromthe commonly used strain-based criteria (e.g., based on the attainment of ±3.75%threshold shear strain).ˆ In the case of fine-grained materials, the transition from Type X to Type Y patternswould manifest gradually over several loading cycles in relatively less contractive soils.However, it was found that the stress-strain pattern-change based criteria consistently47estimates a lower CRR in comparison to that from the shear strain-based criteriawith threshold shear strain limit of ±3.75%. In other words, the new method seemsto always produce a “lower bound” for the CRR values when compared with thoseobtained from the common shear strain based criterion. These factors suggest thatthe new stress-strain pattern-change based approach provides a good indication of thecommencement of the deterioration of the cyclic stress-strain behavior, and as such,it would serves as a robust and prudent candidate criterion for the determination ofunacceptable performance under cyclic loading.ˆ The above conclusions have been developed mainly based on the information availablefrom the laboratory CDSS tests database at UBC. However, it is important to notethat the stress-strain plots derived through cyclic triaxial testing of sand and fine-grained materials including tailings by others (Romero 1995, Bray & Sancio 2006;Zergoun & Vaid 1994) have also shown the visually identifiable stress-strain patternchange through incipient kinks in cyclic stress-strain plots. Therefore, it is judgedthat the proposed stress-strain pattern change based criterion would be applicable forboth CTX and CDSS test results in order to define liquefaction or cyclic failure forthe test specimen.ˆ The newly proposed criterion was based mainly on the available laboratory data on thecyclic shear performance of some specific natural sands, low plastic silts, and minetailings materials that are considered a significant concern during seismic inducedground shaking. Additional studies will be required to assess the applicability of theapproach to the performance of other sands, fine-grained soils, including those withrelatively high plasticity (i.e., clay).48Chapter 4Test Devices, Materials andExperimental ProgramThis chapter details the experimental aspects of the research work presented in the disser-tation. The details of the test devices—triaxial shear test device and direct simple sheartest device—used in this study are presented in Section 4.1. The geo-materials and theindex properties related details are then presented in Section 4.2. Next, the methods usedfor specimen preparations and test procedures are described in Sections 4.3 and 4.4. Thefinal part of the chapter details the experimental test program in Section 4.5.4.1 Test DevicesThe shear strength and deformation characteristics of soil is generally assessed by varioustypes of laboratory testing devices – e.g., direct shear, direct simple shear, directional shearcell, ring shear, triaxial, hollow cylinder torsional and hollow cylinder triaxial test devices.However, direct simple shear (DSS) and triaxial shear (TRX) tests can be identified as thewidely used test devices in laboratory experimental research studies to determine stress-strain-strength characteristics of soil. The Sections 4.1.1 and 4.1.2 present overviews of thedirect simple shear test and triaxial shear test which are then followed by details of theUBC-DSS (see Figure 4.2) and UBC-TRX (see Figure 4.3) device set-ups at the GraduateGeotechnical Laboratory in UBC.4.1.1 Direct Simple Shear Test DeviceThe DSS test evolved after the concerns due to shortcomings of basic direct shear testdevices (usually called as shear box tests) specially the stress and strain non-uniformityalong the soil specimen mainly due to effects of progressive shear failure from the ends to themiddle of the shear box. In an attempt to develop more uniform strain distribution withinthe soil specimen by modifying the loading mechanism, Swedish Geotechnical Institute(SGI) employed a heavy rubber tube with a set of aluminum rings—which are closelyplaced, but separated—to laterally confine the soil specimen (Kjellman, 1951). Thoughthe introduction of heavy rubber tube provides a fully airtight enclosure to enhance thedrainage and evaporation issues in the direct shear box, the compression of the rubber tubebetween soil and aluminum rings cause unintended yet inevitable deformations (Sowers,1964). Addressing the identified inability of vertical walls covering the circular specimen toimpose vertical shear stresses, Roscoe (1953) developed another simple shear test apparatusat Cambridge University. The Cambridge type DSS device has a rectangular prismaticspecimen that contained in a box with rigid side boundaries to impose simple shear strainto a soil specimen by a combination of shear stresses imposed on the top and bottomboundaries and a varying normal stress distribution upon all side of the rectangular box(Roscoe, 1953). Tackling the challenges in placing an undisturbed clay samples in to the49DSS device and further improving the concept of Kjellman (1951) for a DSS device with acylindrical specimen, Bjerrum & Landva (1966) at Norwegian Geotechnical Institute (NGI)produced a practically improved DSS device in which cylindrical specimen was confined bya wire-reinforced rubber membrane replacing the rigid aluminum rings enclosing the rubbertube. The NGI type DSS device got praised for its improved facilitation for testing on‘undisturbed specimens’ obtained from field tube samplers.Saada & Townsend (1981) commented that Roscoe (1953) had identified that any typeof DSS device would not be able to impose uniform pure shear to the specimen due to thedifficulty in applying complementary shear stresses that need to be imposed on the lateralfaces, thus Roscoe (1953) provided necessary boundary conditions through rigid lateralfaces to impose simple shear strain on the rectangular specimen. Both the Cambridgetype and the Norwegian Geotechnical Institute type DSS test devices claim plane straincondition (Saada & Townsend, 1981); however, both device types have been criticized forthe non-uniformities of stress distributions (Airey & Wood, 1987; Budhu, 1984; Prevost &Høeg, 1976; Sowers, 1964; D. Wood et al., 1979). The typical distribution of shear stressand normal stresses at the maximum shear stress of a soil specimen in DSS test device isillustrated in Figure 4.1.Initial and Deformed shapesShear Stress distribution at maximum shear stressNormal stress distribution at maximum shear stressFigure 4.1: Typical distribution of shear stress and normal stresses distribution at themaximum shear stress of a soil specimen in DSS test deviceThe effects of partial differential boundary slippage at the interface between the soilspecimen and the top and bottom loading plates (Prevost & Høeg, 1976), the less restraintfrom the flexible circular-vertical boundary allowing possible out-of-plane movements tooccur (Budhu, 1984), additional normal confining pressures of the radial and hoop typesacting on vertical planes due to the roughness of the loading plates (Saada & Townsend,1981) and the unknown intermediate stress conditions leading to challenging determinationof stress state that ultimately contributes to limited interpretation of the test results—onlyshear stress and normal stress on the horizontal plane are known—(Wijewickreme et al.,2013; D. Wood et al., 1979) are some of the major criticism on the DSS test device thatcould be found in the literature.A photographic analysis of the DSS specimen during shearing by Youd (1972) has shownthat deformed profiles for larger displacement of the specimen with wire-reinforced mem-brane are approximately quite close to the true simple shear condition, thereby suggestedthat at smaller strain excursions, the approximation is even better. Airey & Wood (1984,1987) performed DSS tests on soils with different particle sizes on a specially instrumentedDSS test device—capable of measuring the vertical stress distribution at the horizontalboundary; determining deformations and strains through radio-graphic techniques and con-cluded that use of DSS test to determine the stress-strain-strength behavior is better suited50for fine-grained soils. Three-dimensional finite element analysis by Lucks et al. (1972) re-vealed that approximately 70% of the specimen had a remarkably uniform stress condition.Finn & Vaid (1977) conducted constant-volume cyclic loading tests that were free frommany difficult and time-consuming features of undrain cyclic loading test and reported thatit is quickly to perform while indicating higher re-productively in results.Vertical Load CellHorizontal Load CellVertical LVDTHorizontal LVDTElectro-Pneumatic RegulatorABCDEABCD EFigure 4.2: UBC Cyclic direct simple shear test device51In spite of the limitations with respect to knowledge of the stress state in the test speci-men, the DSS test has been effectively used to obtain peak shear strength, undrained shearstrength, post-cyclic shear strengths through monotonic tests, and to derive cyclic shearresistance through cyclic simple shear tests. In particular, the simplicity and practicality intesting undisturbed soil specimens of the DSS test and the continuous rotation of principalstresses during the DSS loading which mimic the field conditions during seismic inducedloading have attracted wide attention in researchers and practitioners to collect extensiveamount of data from DSS tests to characterize the soil behavior.Dyvik et al. (1987) used an NGI-type DSS test device with a pressure chamber to performtruly undrained tests with pore-water pressure measurements and compared the results withthose from a conventional constant-volume tests and showed that the test results obtainedby the two methods are equivalent for all practical purposes. Thus, in constant-volume DSStests, Dyvik et al. (1987) showed that the increments (or decrements) in vertical stress areequal to the measured excess pore pressures in an undrained DSS test.Over the recent past, many additional improvements to DSS test devices, such as loadingmechanisms to address specific research intentions have been reported by a number ofresearchers (Boulanger et al., 1993; DeGroot et al., 1991, 1996; Doroudian & Vucetic, 1995;Ishihara & Nagase, 1988; Ishihara & Yamazaki, 1980).With the above considerations in mind, the DSS test device at UBC (shown in Fig-ure 4.2) was used for the experimental program herein. The UBC-DSS device was used forthe DSS testing; the device is a modified Marshall-Silver-NGI-type DSS device (Silver &Seed, 1971) and it essentially follows the simple shear testing methodology of the NGI-typeDSS apparatus described by Bjerrum & Landva (1966). As a part of the master’s thesisof the author (Soysa, 2015), the UBC-DSS test device was upgraded with a new data ac-quisition and control system. The UBC-DSS test device is detailed in Appendix B; datareduction details in Appendix D.1; quantification of uncertainty in measured parametersin Appendix F; and estimated uncertainties in the derived parameters form DSS tests inTable F.54.1.2 Triaxial Shear Test DeviceIn summarizing the history of triaxial shear (TRX) tests and its’ evolution, Dea´k et al.(2012) reported the first confirmed experiments using triaxial test on rock material by Kick(1892)1 and mentioned those tests were qualitative experimental work, and more detailedand quantitative experimental research work was followed by von Ka¯rma¯n (1910, 1911)2yet again on rock cores subjected to different confining pressures. These tests—unlike tothe unconfined tests—showed that brittle rock material becomes plastic due to the increasein confined hydro-static pressure. Later, triaxial tests on soil attracted wide attractions,and U.S. Corps of Engineers – Taylor (1947) summarized results of triaxial tests performed1Kick F. (1892) Die Principien der mechanischen Technologie und die Festigkeitslehr, (The Principles ofMechanical Technology and Strength) Zweite Abhandlung, Zeitschrift des Vereines Deutscher Ingenieure[Journal of the Association of German Engineers] in German. 36, 919–933.2von Ka¯rma¯n, T. (1910) Mito˝l fu¨gg az anyag ige´nybeve´tele? [What determines the stress-strain behavior ofmatter?] Magyar Me´rno¨k e´s E´p´ıte´sz Egylet Ko¨zlo¨nye [Journal of the Society of Hungarian Engineers andArchitects] in Hungarian. 44 (10): 212–226von Ka¯rma¯n, T. (1911). Festigkeitsversuche unter allseitigem Druck [Strength tests under pressure from allsides]. Zeitschrift des Vereines Deutscher Ingenieure [Journal of the Association of German Engineers] inGerman. 55, 1749–1757.52by Casagrande (1947)1 and Taylor (1947)2 with different test types called as: (a) SlowTest – consolidated drained test, (b) Consolidated-quick Test – consolidated undrainedtest, and (c) Quick Test – unconsolidated undrained test. The strength of cohesive soildetermined from TRX tests were presented by Skempton (1948) and the procedure forestimations of cohesion and internal angle of friction were detailed with the construction ofMohr’s circles. The ability to determine the exact stress state of the TRX specimen—unliketo the case of DSS test specimen—was very useful in characterizing the soil strength.The considerations with respect to the degree of saturation and measurement of pore-water pressure (Bishop & Eldin, 1950), rate of shearing and strength correction for mem-brane and drains (Bishop & Henkel, 1957), membrane compliance (Lade & Hernandez,1977; Vaid & Negussey, 1984), soil anisotropy, disturbance from tube sampling, nonuniformstresses and strains caused by frictional end caps (Germaine & Ladd, 1988) are notewor-thy in shaping and developing the TRX tests on soils to its’ current state. In additionto the conventional CD – consolidated drained test, CU – consolidated undrained test, andUU – unconsolidated undrained test; many special varieties of TRX tests—such as extensiontests, anisotropic consolidated test, tests with no lateral strain [K0-Tests], constant-volumetest, constant mean confine pressure [Constant P ] test were developed to investigate spe-cific behavior of soils. The constant improvements and enhancement in preparation oftest specimen, instrumentation, measurements and control of laboratory test devices withrespect to TRX tests are comprehensively presented in Lade (2016). With many of theadvantages—such as controllable drainage conditions; measurement of volume change orpore-water pressure; determination of moduli based on the measured loads and deforma-tions; and ability to perform tests with many varieties of stress and strain paths which canoccur in many of the field situations—and few limitations—identified as stress concentra-tions and strain non-uniformities at the end plates; limited to simulate axi-symmetric stressconditions—of TRX tests play a major role in today’s geotechnical investigations speciallyin characterizing soil response.Accordingly, the TRX test device at UBC (shown in Figure 4.3) was also used for theexperimental program herein. The UBC-TRX test device is detailed in Appendix C; datareduction details in Appendix D.2; quantification of uncertainty in measured parametersin Appendix F and estimated uncertainties in the derived parameters form TRX tests inTable F.6Figure 4.4 schematically shows typical loading conditions underneath an embankmentat a field conditions and laboratory element test models to simulate these types of condi-tions. Therefore, a combination of TRX-compression, TRX-extension and DSS tests wereconsidered for this laboratory experimental study in characterizing soil response under bothmonotonic and cyclic shear loading.1Casagrande, A. (1947) Summary of a Harvard Report, published in Triaxial Shear Research. U.S. Corps ofEngineers.2Taylor, D.W. (1947) Summary of a MIT Report, published in Triaxial Shear Research. U.S. Corps ofEngineers.53Load CellLinear Variable Differential TransformerDifferential Pore Water Pressure TransducerPore Water PressureTransducer Cell Pressure TransducerElectro-Pneumatic RegulatorABCDEFABCDEF FFigure 4.3: UBC Cyclic triaxial test device54Compression TestExtension Test Direct Simple Shear TestBACBʹAʹCʹFigure 4.4: Typical loading conditions and laboratory simulation models4.2 Tested MaterialsFraser River sand and Fraser River silt were selected as test materials for the research studyand details of these test materials and the tests conducted are provided in the followingsections. Additional information related to test materials are presented in Appendix E.4.2.1 Fraser River SandThis material was obtained from a local supplier in bulk form, and the sand has been dredgedfrom the Fraser River in the Lower Mainland of British Columbia, Canada. The particlesize distribution assessment on sand (presented in Appendix E.1) showed that the sand hasan average particle size D50 of 0.29 mm, uniformity coefficient Cu of 1.59, coefficient ofcurvature Cc of 1.21. The specific gravity Gs was found to be 2.72. The material phaseanalysis performed on a sample from the Fraser River sand (refer Appendix E.4) revealedthat it is composed of ∼55% Quartz and ∼33% Feldspar. Scanning Electron Microscope[SEM] images of Fraser River sand (Figure 4.5) were analyzed for the shape descriptors(details are presented in Appendix E.5) and the sand grains generally seems to be sub-rounded to rounded in shape with some sub-angular particles.4.2.2 Fraser River SiltThe Fraser River Silt—a natural deposit of fine-grained soil—located at the south bankof the Fraser River adjacent to the Port Mann bridge in British Columbia was excavatedusing a commercial excavator and transported to UBC in 1000 kg bags. The natural silt inbulk form had had a notable organics content (mostly pieces of wood and roots) that wereremoved by wet-sieving prior to the use in both index tests and the tests performed in thisstudy.55#40 - #60 [425 µm - 250 µm] #80 - #100 [180 µm - 150 µm]#60 - #80 [250 µm - 180 µm] #100 - #140 [150 µm - 106 µm] #140 - #200 [106 µm - 75 µm]Figure 4.5: Scanning Electron Microscopy Images of Fraser River Sand#200 - #230 [75 µm - 63 µm] #230 - #325 [63 µm - 44 µm] #335 - #500 [44 µm - 25 µm] Finer than #500 [25 µm]Figure 4.6: Scanning Electron Microscopy Images of Fraser River Silt56The silt has a fine-grit feel to the touch with the occasional coarse-gritty texture due tothe presence of ∼7% fine-sand particles (refer Table E.1). The particle size distribution ofthe silt is presented in Figure E.1 and it showed that the silt consist of ∼18% of clay-sizeparticles. Index tests carried out on this material indicated a natural moisture contentof 19.3%, and a specific gravity, Gs of 2.72. The material was classified as non-plasticafter consecutive attempts to perform Atterberg Limit tests were failed, resulting in thenumber of blows required to close the groove is always less than 25. The Fraser River siltcomprises of ∼46% Feldspar, ∼36% Quartz, ∼7% Actinolite and ∼6% Clinochlore (referAppendix E.4). Few selected SEM images of Fraser River silt are shown in Figure 4.6) werealso analyzed for the shape descriptors and details are presented in Appexndix E.5.4.2.3 Mixtures of Fraser River Sand and SiltAforementioned sand and silts obtained from Fraser River Delta were processed to obtain thebasic sand and silt components required for the experimental study. The sand componentwas generated by wet sieving to remove the finer portion (finer than 75 µm). With respectto the silt component, the Fraser River silt was oven dried and then sieved to remove thefine sand potion (coarser than 75 µm). The sand and silt were further processed to obtainingredient sand and silt components required to arrive at particle size distributions for aspecific soil mixture for testing; in this regard, sand retained on sieve sizes #30, #40, #50,#60, #80, #100, #120, #150 and #200, and silt retained on sieve sizes #230, #270, #325and panwere collected in separate storage containers. The calculated portions of sand andsilt required for the specific gradations were taken from the respective storage containers,weighed and mixed in order to control the specific gradation of the soil specimens.4.2.4 Test Specimen Identification and Labeling ApproachFor the convenience of reference, the test specimens of sand and silt mixtures used in thisthesis were identified using a labeling approach as indicated below. Sample naming/labelingwas designated based on their coarse and fines content; for example, 100C0F refers to 100%processed Fraser River sand (particles finer than 75 µm were removed), whereas 0C100Fdenotes to 100% processed Fraser River silt (particles coarser than 75 µm were removed).20C80F stands for the specimen with 20% of 100C0F and 80% of 0C100F fraction expressedbased on dry weight.The 100C0F and 0C100F materials essentially comprise the natural gradation of FraserRiver sand and silt respectively, after the removal of finer and coarser fraction as mentionedearlier. The intent herein was to achieve a sand-silt mixture that arises from ‘parent-materials’ that have a natural gradation of a soil—see Figure 4.7.4.3 Test Specimen PreparationThe 100C0F and 95C05F specimens were prepared by water pluviation method; whereasother specimens (with finer fraction > 5%) were prepared by the method of slurry deposition(see Appendix G.1 for deposition methods). Initial trials on specimen preparation of sand-silt mixtures by water pluviation method revealed that the fine-grained soils would segregatein the specimen when the fines content is greater than 10%. As such, saturated slurrydeposition method was used for the preparation of specimens with other sand-silt mixturesand the 100% silt. Suitable water contents were selected based on visual observations and57experience-based judgment when preparing the sand-silt mixtures for slurry preparation, inorder to ensure minimum amount of segregation with fully saturated specimens.0%20%40%60%80%100%0.01 0.1 1Percentage FinerGrain Size (mm)100C0F95C05F85C15F75C25F63C37F50C50F40C60F30C70F20C80F10C90F0C100FFigure 4.7: Particle size distribution of the sand, sand-silt mixtures and siltDetails on the reconstitution of specimen in both DSS and TRX test devices are pre-sented in Appendix G.1. Uniformity of the specimen was evaluated by the comparativeassessment of particle size distribution curves derived from two equal slices (top and bot-tom) of a DSS specimen and three equal slices (top, middle and bottom) of a TRX specimenand those particle size distributions results are shown in Figure 4.8 and 4.9 for DSS andTRX specimens respectively. Although, a visual inspection of particle size distributioncurves of the slices of both DSS and TRX specimens indicated that they possess an ac-ceptable uniformity, as an attempt to quantify the uniformity of the specimens, averagedeviation of the particle size distribution curves was computed.In Figure 4.8, particle size distribution curves of the top slice of the DSS specimensare depicted in red line with cross marks, while blue line with diamonds refer for thoseof bottom slices. The absolute differences between the gradation curves derived for topand bottom slices with respect to the gradation curve of the specimen are averaged andadded up for the number of point under consideration to define the ‘Average Deviation’ andshown in Figure 4.8. The maximum average deviation of 3.2% was observed for a 75C25Fspecimen, and maximum deviation of 7% can be seen for a 95C05F specimen, though theaverage deviation was 2.5%. The particle size distribution curves of the top, middle andbottom slices of the TRX specimens are depicted in red line with cross marks, green linewith hollow box, blue line with solid circles respectively in Figure 4.9. The deviation wascomputed as the summation of absolute differences of percentage of finer by mass in top,middle and bottom slices with respect to that of the specimen. The 40C60F specimenshowed the maximum average deviation of 6.6%.580%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)100C0FAD - 0.4 %0%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)95C05FAD - 2.5 %0%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)85C15FAD - 1.3 %0%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)75C25FAD - 3.2 %0%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)63C37FAD - 0.5 %0%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)50C50FAD - 1.6 %0%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)40C60FAD - 0.7 %0%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)30C70FAD - 0.6 %0%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)20C80FAD - 0.4 %0%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)10C90FAD - 0.2 %0%25%50%75%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)BOTTOM TOP0C100FAD - 0.4 %AD - Average Deviation𝐴𝐷 =1𝑛෍1𝑛𝐹% 𝑇 − 𝐹% 𝑆+ 𝐹% 𝐵 − 𝐹% 𝑆(F%)T , (F%)B , (F%)S - Finer percentage in the top, bottom  slices and the specimenFigure 4.8: Uniformity assessment of the DSS specimens590%20%40%60%80%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)100C0FAD - 0.6%TOPMIDDLEBOTTOM0%20%40%60%80%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)95C05FAD - 3.1%0%20%40%60%80%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)75C25FAD - 3%0%20%40%60%80%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)65C35FAD - 4.5%0%20%40%60%80%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)50C50FAD - 0.8%0%20%40%60%80%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)40C60FAD - 6.6%0%20%40%60%80%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)20C80FAD - 2.6%0%20%40%60%80%100%0.01 0.1 1Percent Finer by MassParticle Size (mm)0C100FAD - 1%AD - Average Deviation=1𝑛෍1𝑛ሼሽ𝐹% 𝑇 − 𝐹% 𝑆 + 𝐹% 𝑀 − 𝐹% 𝑆+ | 𝐹% 𝐵 − 𝐹% 𝑆|(F%)T , (F%)M ,(F%)B, (F%)S= Finer percentage in the top, middle, bottom slices, and the specimenFigure 4.9: Uniformity assessment of the TRX specimens604.4 Test Procedure4.4.1 Direct Simple Shear Test ProcedureOnce the specimens were prepared and placed in the DSS test device, they were consolidatedto a vertical effective stress of 100kPa. Lateral restrains induced by the wire-reinforcedrubber membrane results in essentially zero lateral strain; hence, close to one-dimensionalconsolidation occurs in DSS device.The DSS specimens were kept at the consolidation test phase for a duration of 1,800 sec-onds (30 minutes) to 10,800 seconds (3 hours), depending on the time duration required forthe test specimen to gain approximately constant vertical strain with respect to time for theapplied consolidation stress (i.e. sufficient time for the completion of primary consolidationof the test specimens).Two types of shear loading phases—monotonic or cyclic shear loading—were conductedunder constant-volume condition in DSS device. The monotonic loading was applied instain-controlled manner, whereas cyclic loading was stress-controlled. A shear strain rateof approximately 10% shear strain per hour was deployed. Monotonic shear loading testswere conducted until the specimen suffered approximately 20% of shear strain.The frequency of the constant amplitude sinusoidal wave was 0.1 Hz (Period = 10 s)for the stress controlled cyclic shear loading. As constant-volume condition was enforcedon the test specimen by controlling its external boundaries (controlling the geometry ofthe soil matrix volume), there is no shear-induced pore-water development during this DSSshearing. Instead, the vertical force measured by the load cell at the vertical stress boundaryprovides a direct reflection of the vertical effective stress experienced by the specimen atany given time. As such, the equalization of pore-water pressure during cyclic shearingbecomes a non-issue in constant-volume DSS testing.4.4.2 Triaxial Shear Test ProcedureAfter placement and set-up, the specimens in the TRX device were hydrostatically consol-idated to the desired vertical effective stress levels. For TRX specimens, the duration ofconsolidation ranged from 6,000 seconds (∼1.5 hours) to 18,000 seconds (5 hours).For monotonic loading, the TRX tests were performed in a strain-controlled manner,and they comprised of conventional drained and undrained triaxial compression (D-MCand U-MC) and conventional undrained triaxial extension (U-ME) tests. During drainedtests, loading process should be such that it would allow time for shear-induced excess pore-water pressure (∆U) to sufficiently dissipate from the drainage provided at the end(s) ofthe specimen. In a similar fashion, during undrained shear tests, the rate of loading shouldbe sufficiently slow to allow equalization of ∆U within the specimen so that measuredpore-water would be representative of that within the specimen.Following the guidelines and recommendations by Bishop & Henkel (1957) and Germaine& Ladd (1988)—detailed in Appendix G.2, and based on the observations on the resultsof preliminary tests on rate of shearing—detailed in Appendix G.2, it was determined thatthe performance of: (a) drained TRX tests at an axial strain rate of 2% per hour wouldallow the needed dissipation of ∆U ; and (b) undrained TRX tests were performed at anaxial strain rate of 4% per hour would allow the needed equalization of ∆U .For stress controlled cyclic triaxial shear tests (CTX), sinusoidal wave forms with afrequency of 0.01 Hz (Period = 100 s) was deployed. Though the frequency of 0.01 Hz islower than those typically encountered during a seismic induced ground motions, a slower61frequency allows for better measurement of pore-water pressure in CTX tests. Zergoun& Vaid (1994), based on their studies on the behavior of clay, has clearly noted that theresults from undrained cyclic triaxial shear tests cannot be confidently interpreted in ‘fast ’loading tests due to inadequate time for equalization of throughout specimen. With this inmind, preliminary tests were conducted on 0C100F material with three different frequenciesnamely 0.001, 0.01, 0.033 Hz (1000, 100, 30 s periods) as detailed in Appendix H.4. Basedon the experience-based judgment on the observations arising from these preliminary tests,a frequency of 0.01 Hz was selected as suitable for the CTX tests for this research.The raw data obtained from the experimental transducers in DSS and TRX deviceswere processed to derive engineering parameters such as stresses, strain, etc. The processof data reduction and equations used for the computation of engineering parameters areseparately detailed in Appendix D. Further, the results—derived from measurements basedon laboratory experimental tests—are required to be assessed for the uncertainty with afocus of possible experimental errors for proper interpretation and accurate quantificationof the test results. The uncertainty in measurements and derived engineering parametersare also detailed in Appendix F.4.5 Experimental Test ProgramAs identified earlier in scope and objectives for this research (in Chapter 1.3), the systematicexperimental laboratory study presented herein focuses on characterizing the mechanicalbehavior of sand-silt mixtures, covering the complete range of sand to silt compositions inthe mixtures using the DSS and TRX test devices. It should be noted that all shear testsunder this test program were performed at a consolidation stress of 100 kPa. The mainrationales behind the decision on restricting the tests to a single selected stress level of100 kPa are (i) large number of tests required to study the effects of fines content even ata single effective confining stress; (ii) effective confining stress has been well studied andunderstood for both sand as well as silty sands; (iii) effective confining stress of 100 kPawas considered reasonable since most of the geotechnical parameters have been developedconsidering 100 kPa (which is also very closely equal to the average atmospheric pressure)as the benchmark and normalizing stress level.The experimental test program undertaken comprises of 81 DSS tests and 70 TRX testsand can be summarized (see Table 4.1) and expressed with respect to the following facetsthat were addressed:ˆ The preliminary DSS tests [Test Series A and I] and TRX tests [Test Series D andK] to select appropriate shear loading rate for both strain-controlled monotonic testsand stress-controlled cyclic tests.ˆ Constant-volume, strain-controlled monotonic DSS tests on uniform sand-silt mixtures[Test Series B] and repeatability tests [Test Series C]ˆ The preliminary TRX tests to assess the slurry deposition preparation method andto identify the strain level at failure. [Test Series D]ˆ Strain-controlled undrained TRX monotonic compression [Test Series E], extension[Test Series F] tests and drained TRX monotonic compression [Test Series G] onuniform sand-silt mixtures and repeatability tests [Test Series H]62ˆ Constant-volume, stress-controlled CDSS tests and repeatability tests [Test Series J]on uniform sand-silt mixturesˆ Strain-controlled undrained CTX tests and repeatability tests [Test Series L] on uni-form sand-silt mixturesTable 4.1: Summary of the test programTest Series Material DetailsNo. of TestsDSS -uniformA Uniform - 0C100F, 50C50F, 100C0F Preliminary Tests 7B Uniform - 100C0F, 95C05F, 85C15F, 75C25F, 63C37F, 50C50F, 40C60F, 30C70F, 20C80F, 10C90F, 0C100FDSS Tests 11C Repeatability Tests 11TRX -uniform D Uniform - 0C100F Preliminary Tests 6E100C0F, 95C05F, 75C25F, 65C25F, 50C50F, 40C60F, 20C80F, 0C100FUndrained Compression 8F Undrained Extension 8G Drained Compression 8H 100C0F, 65C25F, 50C50F, 40C60F, 20C80F Repeatability Tests 5CDSS -UniformI Uniform - 0C100F Preliminary Tests 4JUniform - 100C0F, 95C05F, 85C15F, 75C25F, 63C37F, 50C50F, 40C60F, 30C70F, 20C80F, 10C90F, 0C100FCDSS Tests with Repeatability Tests48CTX -UniformK Uniform - 0C100F Preliminary Tests 4L100C0F, 95C05F, 75C25F, 65C25F, 50C50F, 20C80F, 0C100FCTX Tests with Repeatability Tests31Total Number of Tests 15163Chapter 5Monotonic Shear LoadingResponse of Sand-Silt Mixtures1The Chapter 5 presents a detailed examination of the results arising from the constant-volume monotonic DSS tests, undrained monotonic compression/extension TRX tests, anddrained compression TRX tests which were focused on characterizing the monotonic shearloading response of sand-silt mixtures produced from Fraser River sand and silt. The exper-imental observations and findings obtained from 29 number of DSS tests and 35 number ofTRX tests are compared and assessed for the characterization of monotonic shear loadingresponse of sand-silt mixtures. For the purpose of brevity, except for selected typical andsummary results, presentation of detailed test results—such as stress-strain and stress-pathresponses, pore-water pressure development, and variation of effective stress ratio—for eachindividual test are included in Appendix I and Appendix J.The details of the monotonic test program and the assessment on the monotonic testrepeatability are presented in Section 5.1. In particular, typical stress-strain and stres