Early Age Mechanical Behavior and Stiffness Development of … · 2013-09-27 · ii Early Age...

Early Age Mechanical Behavior and Stiffness Development of

Cemented Paste Backfill with Sand

by

Abdullah Muhammad Galaa Muhammad Idris Abdelaal

A thesis submitted in conformity with the requirements

for the degree of Doctor of Philosophy

Department of Civil Engineering

University of Toronto

© Copyright by Abdullah M. G. M. I. Abdelaal (2011)

ii

Early Age Mechanical Behavior and Stiffness Development of

Cemented Paste Backfill with Sand

Abdullah M. G. M. I. Abdelaal

Doctor of Philosophy

Department of Civil Engineering

University of Toronto

2011

Abstract

Rapid delivery of backfill to support underground openings attracted many mines to adopt paste

backfilling methods. As a precaution to prevent liquefaction and to improve the mechanical

performance of backfills, a small portion of a binder is added to the paste to form the cemented

paste backfill (CPB). Recently, adding sand to mine tailings (MT) in CPB mixes has attracted

attention since it enhances the flow and mechanical characteristics of the pastefill. This thesis

investigates the effects of adding sand to CPB on the undrained mechanical behavior of the

mixture (CPBS) under monotonic and cyclic loads. Liquefaction investigations took place at the

earliest practically possible age. Beyond this age, the present research focused on characterizing

the evolution of stiffness and obtaining the values of the stiffness parameters that could be useful

for designing and modeling backfilling systems.

The liquefaction investigation involved monotonic compression and extension triaxial tests.

Neither flow nor temporary liquefaction was observed for all cemented and uncemented

specimens under monotonic compression, while temporary liquefaction was observed for all

specimens under monotonic extension. The addition of binder and sand to MT was found to

slightly strengthen the pastefill in compression while weakening it in extension. Under cyclic

iii

loading, the addition of sand negatively impacted the cyclic resistance. However, binder was

found to be more effective in the presence of sand. All specimens exhibited a cyclic mobility

type of response.

The evolution of effective stiffness parameters for two CPB-sand mixtures was monitored in a

non-destructive triaxial test for five days. Self-desiccation was found to not be influential on the

development of early age stiffness. Moreover, a framework is suggested to predict the undrained

stiffness at degrees of saturation representative of the field. The credibility of the proposed test in

providing stiffness parameters at representative strain levels of the field was verified.

iv

Acknowledgments

I am, first and foremost, grateful to God for having good people in my life to help in all

capacities until I reached this stage. The author is grateful to Prof. Murray Grabinsky, Professor of

Civil Engineering at the University of Toronto, and Prof. Will Bawden, Pierre Lassonde Chair in

Mining Engineering, University of Toronto, for their support over the past five years and for giving

me the opportunity to pursue my graduate studies at U of T. Their trust and encouragement to opt for

a fast-track PhD program meant a lot to me.

I‟m also appreciative of the financial support I received throughout my program from the

OGSST/Robert Smith fund and the William Trow Scholarship.

I‟m sincerely grateful to my fellow colleagues from the Geotechnical Lab for their help and

advice. Special thanks for Dr. Ben Thompson, Dr. Abdolreza Saebimoghaddam, and Dr.

Dragana Simon for their help and training, and for the invaluable discussions from which I

learned a lot. I also value the well-made equipment and sound advice from the very busy

Giovanni Buzzeo in the machine shop.

I wish to deeply thank my parents, my wife, and my parents-in-law for their love, support,

encouragement, and the warm cross-Atlantic family feeling they are always keen to provide, no

words can express my gratitude.

Finally, I would like to extend my gratitude to the Housing and Building Research Center in

Cairo, Egypt, for granting me a leave of absence to complete my program at the University of

Toronto.

v

Table of Contents

Acknowledgments .......................................................................................................................... iv

Table of Contents ............................................................................................................................ v

List of Tables ................................................................................................................................. ix

List of Figures ................................................................................................................................. x

List of Appendices ...................................................................................................................... xvii

List of Abbreviations ................................................................................................................. xviii

Chapter 1 Introduction ........................................................................................................ 1

1.1 Problem Statement .............................................................................................................. 1

1.2 Research Objectives ............................................................................................................ 2

1.3 Thesis Overview ................................................................................................................. 3

Chapter 2 Mechanical Response of Early Age Cemented Paste Backfills with

Sand. Part I: Monotonic Shear Response ................................................................................... 5

Abstract ...................................................................................................................................... 5

2.1 Introduction ......................................................................................................................... 6

2.2 Background ......................................................................................................................... 6

2.2.1 Liquefaction of Cemented Paste Backfills .............................................................. 6

2.2.2 Liquefaction of silty sands, sandy silts, and gap grade mixtures ............................ 8

2.2.2.1 At the minimum void ratio ....................................................................... 9

2.2.2.2 At the maximum void ratio ....................................................................... 9

2.2.3 State lines .............................................................................................................. 10

2.3 Materials Tested ................................................................................................................ 11

2.4 Sample Preparation ........................................................................................................... 13

2.5 Testing Program ................................................................................................................ 14

2.6 Consolidation Response .................................................................................................... 15

2.7 Undrained Response to Monotonic Compression Loading .............................................. 16

2.7.1 Uncemented mine tailings ..................................................................................... 16

2.7.2 Uncemented MT-Sand mixtures ........................................................................... 17

2.7.3 Cemented paste backfills with sand ...................................................................... 18

2.7.4 Effect of strain rate on uncemented specimens ..................................................... 19

2.8 Undrained Response to Monotonic Extension Loading ................................................... 20

vi

2.9 Drained Response ............................................................................................................. 21

2.10 Discussion ......................................................................................................................... 22

2.11 Conclusions ....................................................................................................................... 24

2.12 References ......................................................................................................................... 27

Tables and Figures ................................................................................................................... 30

Chapter 3 Mechanical Response of Early Age Cemented Paste Backfills with

Sand. Part II: Cyclic Shear Response ....................................................................................... 58

Abstract .................................................................................................................................... 58

3.1 Introduction ....................................................................................................................... 59

3.2 Background ....................................................................................................................... 59

3.2.1 Cyclic response of cemented paste backfills ........................................................ 59

3.2.2 Role of non-plastic fines ....................................................................................... 61

3.2.2.1 Global or gross void ratio (e) .................................................................. 63

3.2.2.2 Sand skeleton void ratio (ec) ................................................................... 64

3.2.2.3 Relative density (Dr) ............................................................................... 64

3.3 Materials Tested ................................................................................................................ 65

3.4 Sample Preparation ........................................................................................................... 66

3.5 Testing Program ................................................................................................................ 67

3.6 Test Results ....................................................................................................................... 68

3.6.1 Failure criteria ....................................................................................................... 68

3.6.2 Cyclic response of uncemented MT specimens. ................................................... 68

3.6.3 Cyclic response after adding sand to MT. ............................................................ 69

3.6.4 Cyclic response after adding 4.5% binder to MT. ................................................ 70

3.6.5 Cyclic response after adding 4.5% binder to the MT-sand mixture ..................... 70

3.6.6 Effect of curing age on cyclic resistance .............................................................. 71

3.7 Discussion ......................................................................................................................... 72

3.7.1 Cyclic response ..................................................................................................... 72

3.7.2 Strain anisotropy ................................................................................................... 73

3.7.3 Evaluation of the cyclic resistance ........................................................................ 73

3.8 Conclusions ....................................................................................................................... 75

3.9 References ......................................................................................................................... 77

Tables and Figures ................................................................................................................... 80

vii

Chapter 4 Characterizing Stiffness Development in Early Age Cemented Paste

Backfills with Sand in a non-Destructive Triaxial Test ........................................................... 91

Abstract .................................................................................................................................... 91

4.1 Introduction ....................................................................................................................... 92

4.2 Background ....................................................................................................................... 93

4.2.1 Stiffness of geomaterials at pre-failure strains ...................................................... 93

4.2.1.1 Stiffness of particulate media ................................................................. 93

4.2.1.2 Strain dependency of stiffness moduli ................................................... 94

4.2.2 Measuring stiffness parameters ............................................................................. 96

4.2.2.1 Methods for measuring stiffness............................................................. 96

4.2.2.2 Triaxial tests for soil stiffness parameters .............................................. 97

4.2.3 Stiffness development in CPB .............................................................................. 99

4.3 Experimental Procedure and Setup ................................................................................. 100

4.3.1 Materials composition and sample preparation .................................................. 100

4.3.2 Experimental program ........................................................................................ 101

4.3.2.1 The triaxial apparatus and test description ........................................... 101

4.3.2.2 Verifying experimental parameters ...................................................... 103

4.4 Results ............................................................................................................................. 105

4.4.1 Effective stiffness parameters ............................................................................. 105

4.4.2 Hydration-induced changes ................................................................................ 108

4.4.2.1 Volume change (shrinkage) .................................................................. 108

4.4.2.2 Self-desiccation suction ........................................................................ 108

4.5 Discussion ....................................................................................................................... 109

4.5.1 Effective stiffness parameters ............................................................................. 109

4.5.2 Predicting the undrained stiffness parameters .................................................... 110

4.5.3 Analyzing the hydration-induced changes .......................................................... 112

4.6 Conclusions ..................................................................................................................... 114

4.7 References ....................................................................................................................... 116

Tables and Figures ................................................................................................................. 120

Chapter 5 Conclusions and Recommendations ............................................................. 135

5.1 Summary of Findings ...................................................................................................... 135

viii

5.1.1 Conclusions related to monitoring the consolidation response of the different

mixtures ............................................................................................................... 135

5.1.2 Conclusions stemming from studying the mechanical response of paste

mixtures tested under monotonic loading ........................................................... 136

5.1.3 Conclusions drawn from the cyclic triaxial tests ................................................ 138

5.1.4 Conclusions related to stiffness measurements ................................................... 138

5.2 Contributions and Industrial Implications ...................................................................... 140

5.3 Recommendations for Future Work ................................................................................ 141

References (Combined) ............................................................................................................ 143

ix

List of Tables

Table 2-1 Chemical compositions of the tailings and binder received from the mine. ............... 31

Table 2-2 Gradients of the best-fit lines indicating v/a values. ............................................. 32

Table 2-3 Compression indices for the tested MT, MTS, and CPBS specimens. ........................ 33

Table 2-4 Summary of the key parameters for specimens tested under udrained monotonic (a)

compression at 2%/min, and (b) extension at -2%/min. ............................................ 34

Table 2-5 Slopes of the UFL, PTL and TIL, in (a) the s´-t space and (b) the Mohr stress space,

from undrained monotonic compression and extension triaxial testing at 2%/min. . 35

Table 3-1 Summary of test parameters and results ....................................................................... 81

x

List of Figures

Figure 2-1 Typical relationships between void ratio (e) and the fines content (FC) at maximum

and minimum compaction. .......................................................................................... 36

Figure 2-2 A schematic diagram showing different arrangements of soil particles in a blend for

(a) loosely packed particles and lack of contact between coarser grains; and (b)

densely packed particles where coarser grains maintain contact. ............................... 37

Figure 2-3 State lines and state points for materials exhibiting stable and temporary liquefaction

behaviors. .................................................................................................................... 38

Figure 2-4 Particle size distributions of the tested materials and mixtures. ................................. 39

Figure 2-5 Temporal response of the electric conductivity (EC) for different mixes (after

Thottarath, 2010) ......................................................................................................... 40

Figure 2-6 The relationship between the global void ratio and FC at different Sand-MT mixtures

with the specimens tested at different FC, e, and 'c shown in black markers ........... 41

Figure 2-7 The observed volumetric strain versus axial strain at the end of the hydrostatic

consolidation phase for (a) uncemnted specimens, and (b) cemented specimens ...... 42

Figure 2-8 e-log ′c plots for the tested specimens in terms of (a) global void ratio, e, (b)

intergranular void ratio, ec, and (c) and the interfine void ratio (ef) ........................... 43

Figure 2-9 Monotonic response of 100% mine tailings (MT). (a) Stress path, (b) Pore pressure

ratio, ru, versus axial strain (c) Stress-strain behavior, (d) stress obliquity versus axial

strain. ........................................................................................................................... 44

Figure 2-10 Monotonic response of MTS45. (a) Stress path, (b) Pore pressure ratio, ru, versus

axial strain (c) Stress-strain behaviour, (d) stress obliquity versus axial strain. ......... 45

Figure 2-11 Monotonic response of uncemented MTS55. (a) Stress path, (b) Pore pressure ratio,

ru, versus axial strain (c) Stress-strain behaviour, (d) stress obliquity versus axial

strain. ........................................................................................................................... 46

xi

Figure 2-12 Monotonic response of CPBS45/2.2 cured for 4 hours. (a) Stress path, (b) Pore

pressure ratio, ru, versus axial strain (c) Stress-strain behavior, (d) stress obliquity

versus axial strain. ....................................................................................................... 47










Figure 2-16 Monotonic response of MTS55 at 200 kPa effective confining pressure under

different strain rates. (a) Stress path, (b) Pore pressure ratio, ru, versus axial strain (c)

Stress-strain behaviour, (d) stress obliquity versus axial strain. ................................. 51

Figure 2-17 Monotonic response of MT at 200 kPa effective confining pressure under different

strain rates. (a) Stress path, (b) Pore pressure ratio, ru, versus axial strain (c) Stress-

strain behaviour, (d) stress obliquity versus axial strain. ............................................ 52

Figure 2-18 Drained response of MT in compression. (a) Stress path, (b) Volumetric strain

versus axial strain (c) Stress-strain behavior, (d) stress obliquity versus axial strain. 53

Figure 2-19 Drained response of MTS55 in compression. (a) Stress path, (b) Volumetric strain

versus axial strain (c) Stress-strain behavior, (d) stress obliquity versus axial strain. 54

Figure 2-20 e-log p' plots showing (a) UFL in compression, (b) UFL in extension, (c) PTL in

compression, and (b) PTL in extension for the tested materials. ................................ 55

Figure 2-21 SEM images of (a) MT tested in this study and (b) gold tailings tested by

Saebimoghaddam (2010). ........................................................................................... 56

xii

Figure 2-22 Summary of the state parameters ´UFL indicating the changes in strength in response

to the addition of sand and binder. .............................................................................. 57

Figure 3-1 Typical response of cyclically loaded MT specimens ('c=100 kPa and CSR=0.10):

(a) Stress paths with failure lines obtained from monotonic testing, (b) stress strain

behavior, (c) excess pore water pressure ratio (ru) versus number of cycles, and (d)

axial strain versus number of cycles. .......................................................................... 82

Figure 3-2 Typical response of cyclically loaded MT specimens ('c=100 kPa and CSR=0.20):




Figure 3-3 Cyclic resistance chart for uncemented mine tailings (MT) and MT-sand mixture

(MTS55). ..................................................................................................................... 84

Figure 3-4 Typical response of cyclically loaded MTS55 specimen ('c=100 kPa and CSR=0.09):




Figure 3-5 Typical response of cyclically loaded CPB/4.5 specimen ('c=100 kPa and

CSR=0.09): (a) Stress paths with failure lines obtained from monotonic testing, (b)

stress strain behavior, (c) excess pore water pressure ratio (ru) versus number of

cycles, and (d) axial strain versus number of cycles. .................................................. 86

Figure 3-6 Cyclic resistance chart for cemented specimens (CPB/4.5 and CPBS55/4.5) at four

and seven hours of curing time in comparison with the uncemented specimens. ...... 87

Figure 3-7 Typical response of cyclically loaded CPBS55/4.5 specimen ('c=100 kPa and



cycles, and (d) axial strain versus number of cycles. .................................................. 88

xiii

Figure 3-8 Comparing cyclic resistance MT and MTS55 with those of uncemented tailings in

literature ...................................................................................................................... 89

Figure 3-9 Variation of cyclic resistance with (a) relative density, Dr, and (b) fines content, FC%.

..................................................................................................................................... 90

Figure 4-1Stiffness degradation with strain and approximate strain levels for soil structures (after

(Mair 1993)). ............................................................................................................. 121

Figure 4-2 The triaxial setup for stiffness, volume change, and suction measurements(a) A

schematic of the triaxial setup, (b) A picture of the triaxial cell and VCD, and (c) A

picture of the modified triaxial base. ........................................................................ 122

Figure 4-3 The T5x miniature tensiometer (a) main components and properties (after UMS

GmbH München 2008), and (b) a schematic showing the dimensions. ................... 123

Figure 4-4 Examining volumetric changes due to leakage and compressibility of cell fluid. .... 124

Figure 4-5 The resulting behavior from the pilot test. (a) Stress-Strain loops at four hours, (a)

comparing stress-strain loops at four hours and one week, (c) volume change

(shrinkage) measured over the testing period, and (d) suction. ................................ 125

Figure 4-6 Axial stress-strain loops for: (a) CPBS55/4.5_A at four hours, (b) CPBS55/4.5_B at

four hours, (c) CPBS55/4.5_A at 122 hours, and (d) CPBS55/4.5_B at 126 hours. 126

Figure 4-7 Axial stress-strain loops for: (a) CPBS55/2.2_A at four hours, (b) CPBS55/2.2_B at

four hours, (c) CPBS55/2.2_A at 120 hours, and (d) CPBS55/2.2_B at 120 hours. 127

Figure 4-8 Effective stiffness parameters measured over five days for CPBS55/4.5 and

CPBS55/2.2. (a) Measured Young's modulus, E, (b) measured Poission's ratio, , (c)

calculated shear modulus, G, and (d) calculated bulk modulus, K. .......................... 128

Figure 4-9 Volumetric changes (shrinkage) of CPBS specimens over the test period. .............. 129

xiv

Figure 4-10 Development of self-desiccation suction in CPBS55/4.5, CPBS55/2.2 and the pilot

specimen over the test period. ................................................................................... 130

Figure 4-11 Volume change during the 20 minutes of stiffness loading at four hours of curing is

affected by the overall shrinkage rate of the specimen. Example shown above is for

specimen CPBS55/4.5_B. ......................................................................................... 131

Figure 4-12 Predicted undrained stiffness parameters measured over five days for

CPBS55/4.5_B. (a) Undrained Young's modulus, Eu, (b) Undrained Poission's ratio,

u, (c) Undrained shear modulus, Gu, and (d) Undrained bulk modulus, Ku. ........... 132

Figure 4-13 A schematic showing the difference between (a) the sealed tensiometer shaft in the

triaxial cell, and (b) the unsealed tensiometer shaft for separate suction measurements.

................................................................................................................................... 133

Figure 4-14 Suction development in specimens tested separately out of the cell and specimens

tested by Witteman (personal communication). ....................................................... 134

Figure A- 1 Typical response of cyclically loaded MT specimens ('c=100 kPa and CSR=0.12):



axial strain versus number of cycles. ........................................................................ 151

Figure A- 2 Typical response of cyclically loaded MT specimens ('c=100 kPa and CSR=0.15):



axial strain versus number of cycles. ........................................................................ 152

Figure A- 3 Typical response of cyclically loaded MTS55 specimens ('c=100 kPa and



cycles, and (d) axial strain versus number of cycles. ................................................ 153

xv









Figure A- 6 Typical response of cyclically loaded CPB/4.5 specimens ('c=100 kPa and












Figure A- 9 Typical response of cyclically loaded CPBS55/4.5 specimens ('c=100 kPa and








xvi









Figure A- 13 Typical response of cyclically loaded CPB/4.5 specimens at seven hours of curing

('c=100 kPa and CSR=0.20): (a) Stress paths with failure lines obtained from

monotonic testing, (b) stress strain behavior, (c) excess pore water pressure ratio (ru)

versus number of cycles, and (d) axial strain versus number of cycles. ................... 163

Figure A- 14 Typical response of cyclically loaded CPBS55/4.5 specimens at seven hours of

curing ('c=100 kPa and CSR=0.25): (a) Stress paths with failure lines obtained from

monotonic testing, (b) stress strain behavior, (c) excess pore water pressure ratio (ru)

versus number of cycles, and (d) axial strain versus number of cycles. ................... 164

Figure C- 1 Predicted undrained stiffness parameters measured over five days for

CPBS55/4.5_A. (a) Undrained Young's modulus, Eu, (b) Undrained Poission's ratio,



CPBS55/2.2_A. (a) Undrained Young's modulus, Eu, (b) Undrained Poission's ratio,



CPBS55/2.2_B. (a) Undrained Young's modulus, Eu, (b) Undrained Poission's ratio,


xvii

List of Appendices

Appendix A Cyclic Testing Results .......................................................................................... 150

Appendix B Axial Stress-Strain Loops for Stiffness Measurements ........................................ 165

Appendix C Predicted Undrained Stiffness Parameters ........................................................... 180

xviii

List of Abbreviations

BFS Blast Furnace Slag.

CD Consolidated Drained.

CPB Cemented Paste Backfill.

CPBS Cemented Paste Backfill with Sand.

CPT Cone Penetration Test.

CRR Cyclic Resistance Ratio.

CSR Critical Stress Ratio.

CSR Cyclic Stress Ratio.

CU Consolidated Undrained.

EC Electric Conductivity.

FC Fines Content.

FCth Threshold Fines Content.

FL Failure Line.

LDT Linear Deformation Transducer.

LL Liquid Limit.

LSC Limiting Silt Content.

LVDT Linear Variable Differential Transformers.

MSO Maximum Stress Obliquity.

MSOP Maximum Stress Obliquity Points.

MT Mine Tailings.

MTS Mine Tailings with Sand.

NPF Non-Plastic Fines.

PC Portland Cement.

PI Plasticity Index.

PL Plastic Limit.

PTL Phase Transformation Line.

PTP Phase Transformation Points.

PTS Phase Transformation State.

PWP Pore Water Pressure.

SAR Strain Anisotropy Ratio.

SEM Scanning Electron Microscopy.

xix

SPT Standard Penetration Test.

SS Steady State.

SSL Steady State Line.

TIL Temporary Instability Line.

TIS Temporary Instability State.

UCS Unconfined Compressive Strength.

UFL Ultimate Failure Line.

UFS Ultimate Failure State.

VCD Volume Change Device.

XRF X-Ray Fluorescence.

Chapter 1 Abdullah M.G.M.I. Abdelaal, Doctor of Philosophy, 2011 1

Chapter 1

Introduction

1.1 Problem Statement

Rapid delivery of backfill to support underground openings has attracted many mines to adopt

paste backfilling methods. As a precaution to increase liquefaction resistance and to improve the

mechanical performance of backfills, a small portion of a cementitious material (binder) is added

to the paste to form the cemented paste backfill (CPB). Reducing the filling time or opting for a

continuous pour could add to the cost of the barricade and raises the risk of liquefaction as huge

volumes of the placed CPB would still be in the early hydration stages. Therefore, the prime

aspects to be carefully investigated during early curing ages of CPB are liquefaction

susceptibility and stiffness development.

The first signs of strength development due to the action of binder correspond to the acceleration

phase in the hydration process. This phase usually commences several hours after mixing.

Therefore, the behavior of the material before reaching the acceleration phase would be primarily

controlled by the effective stress. Consequently, liquefaction susceptibility can represent a major

concern. A few researchers have investigated the liquefaction potential of CPB at very early

ages, such as le Roux (2004) and Saebimoghaddam (2010) who conducted studies on silt-sized

CPB at ages as early as three hours. However, the risk of liquefaction at early ages when CPB is

mixed with sand still requires further investigation, especially when the active hydration process

is delayed under the effect of supplementary cementing materials. At Kidd Creek mine, operated

by Xstrata Copper Canada, silt-sized mine tailings (MT) is mixed with glacial sand and slag

based binder to obtain the unconventional CPB; referred to as CPBS hereafter.

While “early age” in liquefaction problems refers more to concerns about the first few hours

after placement, it may extend to several days until the CPB fully transitions from a non-


Newtonian fluid to a solid. Characterizing the evolution of stiffness of CPB during such a

transitional period is paramount as it will assist in developing: 1) more representative modeling

processes, 2) safer and more economic barricade designs, and 3) significant impact on the

scheduling of mining activities. Significant contributions have been made by many researchers to

study the mechanical behavior of CPB after significant curing times. Characterizing the early age

stiffness of CPB, however, has not received as much attention in the literature. Therefore, most

designs are based on experience rather than scientific rationale.

A few attempts have been made to characterize stiffness development during the first few days

by measuring the velocity of mechanical wave propagation through CPB. Helinski et al. (2007)

measured S-Wave velocity to obtain the shear modulus and assumed a constant value for

Poisson‟s ratio to compute the other elastic stiffness parameters. Galaa et al. (2011) conducted

combined measurements for P-wave and S-wave velocity to obtain all stiffness parameters while

avoiding the need to assume any of the parameters. However, the obtained stiffness parameters

from wave velocity measurements correspond to smaller strain levels than most of the

anticipated strains in the field. Furthermore, under the ongoing binder hydration process, all

stiffness parameters are continuously changing. Therefore, fixing a value for one parameter to

obtain the rest can be largely misleading. Thus, experimental determination of the elastic

stiffness parameters at higher representative strain levels is considered essential.

1.2 Research Objectives

In this thesis, liquefaction susceptibility and stiffness development of CPB-sand blends are the

prime focus of interest, and, hence, three objectives are set out as given below with a brief

description of the approach adopted to achieve each objective.

1. Investigate the influence of mixing CPB with high proportions of sand on liquefaction

susceptibility and the undrained mechanical behavior under monotonic loading regimes.

This objective is achieved by comparing the behavior of paste before and after adding

sand. The binder was added to the sand-containing paste to investigate the combined

influence of sand and binder. Deeper insight into the mechanical behavior of CPBS is

attempted by integrating the undrained and drained behaviors.


2. Examine the separate and combined effects of adding sand and binder on the undrained

cyclic resistance of the paste. To achieve this objective, conventional geotechnical

earthquake engineering concepts and frameworks describing the arrangement of particles

in granular mixes are applied to analyze the results of the cyclic testing. Light is also shed

on the effect of curing age on the cyclic resistance prior to reaching the initial setting time

of the paste.

3. Monitor the stiffness development of CPBS and quantify the elastic stiffness parameters

at an intermediate strain level over the first few days after mixing. To achieve this

objective, the triaxial apparatus is utilized to measure the stiffness parameters with

satisfactory resolution coupled with monitoring hydration-induced volume and pore

water pressure changes in order to understand the mechanisms contributing to the

evolution of stiffness during the early curing ages.

1.3 Thesis Overview

The research effort in the thesis is documented in three manuscripts prepared for submission to

academic journals. Each of the three objectives mentioned above is addressed in one of the

manuscripts, for which a chapter of the thesis is dedicated. Some background material is

reiterated in Chapters 2 to 4.

Chapter 2 reviews the literature on mechanical behavior of silt-sand mixtures and CPB under

undrained monotonic loading. In addition, the chapter presents the monotonic undrained testing

program undertaken on cemented and uncemented specimens with different sand contents. The

results are presented for monotonic compression and extension loading. The response of

uncemented MT and MT-sand mixtures to varying the strain rate is also investigated. The

behavior under drained loading conditions is presented and compared to the undrained behavior.

Implications on the field performance are discussed.

Chapter 3 reviews the literature on the cyclic behavior of silt-sand mixture, silt-size mine

tailings, and CPB and describes the undrained cyclic triaxial tests performed to study the single

and combined effects of sand and binder on the cyclic resistance of the paste. The effect of

curing ages, earlier than the acceleration phase, on the cyclic resistance of cemented specimens is

discussed.


Chapter 4 reviews the relationship between the stiffness of geomaterials and strain and briefly

describes the different methods used to measure stiffness parameters at the different strain levels.

In addition, the advantages and limitations of the triaxial tests in measuring stiffness parameters

are reviewed. A non-destructive triaxial test procedure and setup are introduced to measure the

stiffness parameters of CPBS for several days after mixing. The modifications introduced to the

triaxial apparatus to enable measuring the hydration-induced changes are presented.

Furthermore, the chapter discusses the checks performed to assure a high accuracy level for the

measured parameters. The obtained effective stiffness parameters are presented and the

mechanisms contributing to stiffness evolution are discussed. Additionally, a framework to

predict the undrained stiffness parameters is suggested for CPBS at saturation levels similar to

that of the field.

Chapter 5 presents the overall conclusions, contributions, and suggestions for future research.

Finally, it is worth emphasizing that although the current research is not intended to provide a

constitutive law for the behavior of CPBS, it attempts to put forward a significant contribution in

three main directions. First, it provides sound understanding of the continuously evolving

stiffness of CPB/CPBS during early curing ages. Second, it sets out an experimental framework

for reasonably accurate determination of the stiffness parameters at representative strain levels.

Third, it provides deeper insight into the undrained mechanical behavior of a complex cemented

paste backfill-sand mixture under static and dynamic loads. It is hoped that this research provides

a step-forward towards a rational design of cemented paste backfilling systems.


Chapter 2

Mechanical Response of Early Age Cemented Paste

Backfills with Sand. Part I: Monotonic Shear

Response

Abstract

Adding sand to mine tailings in cemented paste backfill (CPB) mixes has attracted attention

since it helps to overcome pipe flow problems, reduces water consumption, and enhances binder

efficiency. Liquefaction potential of early age CPB has been an increasing concern due to the

potential consequences to mine safety and economy. In this chapter, the effect of adding sand to

mine tailings (MT) and that of adding binder to MT-sand mixtures on liquefaction susceptibility

are investigated under monotonic compression and extension loads. A monotonic undrained

testing program was undertaken on specimens with different sand contents while varying the

binder content for each MT-sand mixture. Specimens were prepared to replicate the placement

method at one of Ontario‟s mines and were tested at effective confining stresses up to 400kPa,

which covers the overburden stress range measured at the mine. The inter-coarse and inter-fine

void ratios suggest the scarcity of contacts or communication between particles of the coarse

matrix in the blend. The influence of adding binder on the undrained behavior, despite the

associated increase in void ratio and compressibility, is found negligible for specimens tested at

4 hrs after mixing. Strain rate dependence is significantly reduced for sand-containing

specimens. Several drained tests were undertaken to support the state lines acquired by the

undrained tests. The results show that when adding 45% and 55% Sand to MT with or without

binder, the behavior of the blend is dominated by that of the silt-sized MT.


2.1 Introduction

During the past two decades, cemented paste backfill (CPB) has been gaining popularity in the

mining industry. A major advantage of CPB is that only a small portion of a cementitious

material (binder) is required to provide significant enhancements in its mechanical

characteristics. Adding binder to paste backfills has been shown to provide liquefaction

resistance. However, at early ages, before the binder undergoes enough hydration to strengthen

the paste, liquefaction susceptibility of CPB is still questionable (le Roux, 2004). Particle size

distribution of tailings, process water and tailings chemistries, binder type and content,

environmental conditions, and many other geometrical and process related factors, can

significantly vary from mine to mine or even within the same mine. Accordingly, the undrained

mechanical behavior of CPB needs more research to gain more understanding of how the

material might behave under the imposed circumstances by the aforementioned factors. One of

the main concerns of the present work is to assess the risk of liquefaction when CPB is mixed

with sand (CPBS), especially when the active hydration process is delayed under the effect of

supplementary cementing materials. The following sections will review the findings of studies

conducted to assess the liquefaction potential of silt-sized CPB under monotonic loads, followed

by those conducted on silty sands and gap graded mixtures.

2.2 Background

2.2.1 Liquefaction of Cemented Paste Backfills

Since the 1980‟s, backfill designers have considered cemented backfill materials as liquefaction

resistant if a 100 kPa unconfined compressive strength (UCS) is reached. Clough et al. (1989)

suggested this guideline based on a study conducted on round, clean Monterey sand #0/30.

Relying on this guideline or “rule of thumb” may be misleading as CPBs in most mining

operations have significantly different particle size distributions, particle shapes and chemistries,

which may result in completely different binder hydration mechanisms and efficiencies (le Roux,

2004). Moreover, the experimental study conducted by Clough et al. (1989) was directed towards

those sands that are cemented at points of contact between sand grains. Such materials are

termed “contact bound” materials and were found to exhibited flow liquefaction type of failure

even at high UCS values. In the case of CPB and CPBS, the void volume is filled with fine

grains and is cemented. Therefore, their behavior is expected to be different from the behavior


observed by Clough et al. (1989). Several studies were conducted to assess the liquefaction

potential of silt-sized CPB, e.g. Aref (1989), Pierce et al. (1998), Broomfield (2000), and Been et

al. (2002). In all of these studies, CPB showed dilative behavior and no significant pore pressure

development under monotonic compression loading in consolidated-undrained (CU) triaxial

tests. Therefore, they concluded that the materials tested have low liquefaction potential.

However, the earliest curing ages of specimens tested in these studies were 6 days as tested by

Aref (1989), 28 days as reported by Pierce et al. (1998), 2 days in Broomfield (2000), and

several months of curing as mentioned in Been et al. (2002).

Rapid rise rates in narrow stopes and the desire to continuously pour, when feasible, pose more

risk of liquefaction at earlier ages than those in the aforementioned studies. Cases where the

liquefaction potential of CPB is investigated at very early ages are very limited. Only le Roux

(2004) and Saebimoghaddam (2010) investigated the liquefaction resistance of CPB at ages as

early as three hours. The cemented silt-sized gold tailings tested by le Roux (2004) under

monotonic compression loading showed that despite the contractive behavior it exhibited,

generated pore water pressures (PWP) were small and the specimens were not susceptible to

liquefaction. le Roux (2004) attributed this behavior to the intensive fabric of the material.

Saebimoghaddam (2010) tested cemented and uncemented silt-sized gold tailings in compression

and in extension and found that in compression, both cemented and uncemented tailings

exhibited a stable behavior that transforms from contractive to dilative as strain progresses. In

extension, cemented specimens went through temporary instability (limited liquefaction) before

going into the dilative phase. Uncemented silt-sized tailings were tested by Crowder (2004) and

Al Tarhouni (2008) under monotonic compression. Crowder (2004) observed similar behavior to

that observed by Saebimoghaddam (2010) but the dilation phase was significantly milder.

However, Al Tarhouni (2008) observed limited liquefaction and flow liquefaction at

substantially lower void ratios than those in the previous studies. In addition, Fourie and

Papageorgiou (2001) and Fourie et al. (2001) reported lab and field evidences of flow

liquefaction of tailings containing 40% particles larger than 0.075 mm.

At this point, it is useful to discuss the undrained behavior of tailings and natural soils composed

of mixtures of fine and course grained soils as per the following sections.


2.2.2 Liquefaction of silty sands, sandy silts, and gap grade mixtures

Evidence in the literature of the role of nonplastic fines (NPF) on the liquefaction behavior of

sands shows little consensus. Kuerbis et al. (1988) showed that sand with silt content up to 20%

exhibited less contraction than cleaner sands under compression and extension loadings. Also,

Pitman (1994) observed higher stability with the increase of fines content (FC). On the other

hand, Sladen et al. (1985), Ishihara (1993), Lade and Yamamuro (1997), Yamamuro and Lade

(1997 and 1998), Fourie and Papageorgiou (2001), Yamamuro and Covert (2001) have reported

an increasing tendency to contraction with increasing FC. More recently, however, Khalili et al.

(2010) reported that although the behavior of paste rock (in that particular study it was composed

of a highly gap-graded mixture containing around 18% fines) was still controlled by the skeleton

of the coarser particles, a reduction in strength was observed when fines were added. Fourie and

Papageorgiou (2001) attributed the contradiction in the effect of NPF to the lack of a common

basis for comparison. Some studies based their conclusions on whether the steady state line

(SSL) moves up or down, and others, such as Yamamuro and Lade (1997), looked at the effect of

NPF on liquefaction susceptibility through the relative density. Thevanayagam (1998),

Thevanayagam and Mohan (2000), and Thevanayagam et al. (2002) presented a different

approach based on the arrangement of particles. These studies suggested defining a threshold FC

(FCth) that distinguishes between whether the coarse (intergranular) or the fine (interfine) matrix

will be dominating the behavior of the mixture. Consequently, either the actual or an equivalent

void ratio of the dominant matrix leads to assessing the liquefaction susceptibility of the whole

mixture. Wickland et al. (2006) reported that the factors affecting the density of a mixture, and

consequently its hydraulic and mechanical behaviors, include particle size ratio, mixture ratio,

and the density of individual components.

It appears from the above mentioned findings that particle arrangement has a major influence on

the behavior of a mixture. The following sections provide further insights into the changes in the

particles arrangement corresponding to changing the FC and the packing density. The void ratio

(e) is an index for the packing density of soil particles. However, it does not always describe the

state of intergranular contact. Understanding the variation of e by changing FC at the same

compaction energy helps in predicting the resulting particle structure that will later help in

understanding the behavior.


Figure 2-1 shows the typical e-FC relationships at minimum and maximum packing densities

(emax and emin). The relationships in Figure 2-1 are explained in detail in sections 2.2.2.1 and

2.2.2.2 in accordance to Lade and Yamamuro (1997) and Thevanayagam et al. (2002).

2.2.2.1 At the minimum void ratio

Energy is applied to the soil to acquire the maximum possible density. The relationship between

e and FC at maximum compaction or emin line, shown in Figure 2-1, starts at a granular matrix

with no fines. When fines are introduced to this matrix, finer particles are accommodated within

the void spaces created between the coarse grains (the host soil). This mechanism results in a

decrease in the void ratio of the whole mixture (global void ratio, e). Particles of the host soil

tend to keep contact with each other. Further increase of the FC adds finer grains to the

intergranular void space until it is full with densely packed fine particles (see Figure 2-2b). At

this point, the maximum density (the minimum void ratio) of all gradations is reached. Usually

the minimum void ratio of silty sands is achieved at FC ranging between 20% and 30%. Further

introduction of fines forces the coarser grains to lose contact and be displaced apart of each other

by a densely packed fine matrix. This results in a decrease in the overall density and increase in

the void ratio. Any further addition of fines is associated with an increase in the minimum void

ratio. Spacing between coarse grains increases by increasing the FC and the grains become

sporadically isolated in the interfine matrix, which will dominate the mechanical behavior of the

mixture.

2.2.2.2 At the maximum void ratio

A minimum amount of energy is applied to the soil to acquire the minimum possible density.

When fines are initially added to the coarse-grained matrix, not all the fine particles move into

the intergranular void space as the applied compaction energy is not sufficient (see Figure 2-2a).

Some fine particles separate the coarser grains which cause smaller decrease in the void ratio

compared to that obtained in the case of the emin line. This arrangement leaves the overall soil

behavior influenced by the presence of fines. Adding more fines to the void space leads to more

particles separating the coarser grains until they are widely displaced from each other. Therefore,

the void ratio continues to increase until FC reaches 100%.


Lade and Yamamuro (1997) pointed out the effect of sand gradation (uniformity) on the amount

of decrease of the void ratio for the maximum and minimum void ratio lines. Uniform sands

have larger void spaces that can accommodate more fines resulting in more decrease of the void

ratio with the increase in FC. On the other hand, graded sands will initially have smaller sand

particles to fill the void space between the larger ones and, therefore, less decrease of the void

ratio is observed. For example, the mine tailings tested by Fourie and Papageorgiou (2001)

showed no initial decrease in emax.

2.2.3 State lines

Defining the state of failure is a concern that arose from flow slides and earthquake failures.

States of failure or flow can be defined at points that form unique state lines for a given material.

State lines for specimens showing totally contractive, or strain softening, behavior can be easily

determined. When the material exhibits constant strength with increased strain, the steady state is

defined. In dilative materials, where specimens show a strain hardening behavior, defining the

steady state (SS) is difficult (Fourie and Papageorgiou, 2001). In addition, Yamamuro and Lade

(1998) reported that the steady state line (SSL) may not always exist for silty sands. Therefore,

other states are defined on the basis of uniqueness for a given material. The following discussion

focuses on the states associated with dilative behaviors that do not reach a clear steady state.

The critical stress ratio (CSR) is defined when a dilative material undergoes limited liquefaction.

The effective stress ratio at the peak deviator stress in the effective stress path represents the

CSR points as shown in Figure 2-3 (Chern, 1985; Kuerbis et al., 1988; Thomas, 1992; Al

Tarhouni, 2008). Kramer (1996) referred to the unique line that can be plotted as the locus of

CSR points as the flow liquefaction surface (FLS) or the temporary instability line (TIL) (see

Figure 2-3). The phase transformation state (PTS) is another state defined at the point when

contraction terminates and dilative behavior follows, as shown in Figure 2-3. Phase

transformation points (PTP) at different densities fall on the same unique line which is referred

to as the phase transformation line (PTL) (Thomas, 1992; and Al Tarhouni, 2008). Chern (1985)

reported that the angle of the PTL determined from the undrained triaxial testing is equal to that

of the SSL determined from drained loading. Following the PT state, the ultimate failure state

(UFS) is reached at the maximum stress obliquity (MSO). The unique line plotted as the locus of


the maximum stress obliquity points (MSOP) is referred to as the ultimate failure envelope/line

(UFL) (Kuerbis et al., 1988).

In the light of this background, an extensive laboratory triaxial testing program was conducted to

investigate the influence of mixing CPB with high proportions of sand on the undrained triaxial

response at very early curing ages. The shear response under monotonic compression and

extension loads is investigated at different binder contents. Also, a series of consolidated drained

(CD) tests are performed. The results of this program will help in understating how the added

coarse grains may contribute to the behavior at the early age. The response of the mixture to

cyclic loading is presented in Chapter 3.

2.3 Materials Tested

The basic materials used in this testing program involved silica tailings, glacial sand, and binder.

All those components are supplied by Kidd Creek mine in sealed pails. The tailings used at the

mine are excavated from local mine tailings impoundments. Figure 2-4 shows the particle size

distributions determined by undertaking sieve analysis, in accordance to ASTM C136-06, for

particles coarser than 75 microns then the hydrometer test, in accordance to ASTM D422-63, for

the percentage finer than 75 microns. The percentage finer than 20 microns in the tailings are

around 40%, which is larger than the minimum of 15% suggested to avoid particle settlement

and segregation during paste transport and placement. Atterberg limits were determined in

accordance to ASTM D4318-05. The liquid limit (LL) of the tailings is 23%, the plastic limit

(PL) is 18%, and the plasticity index (PI) is 5%. Therefore, the existing fines are considered non-

plastic or “sand-like” fines according to Boulanger and Idriss (2004) where PI of 7% is

suggested as the boundary between “sand-like” and “clay-like” fine grained soils. The specific

gravity of the tailings is 2.79 determined in accordance to ASTM D854-06. X-Ray Fluorescence

Spectroscopy (XRF) was performed to determine the chemical composition of the tailings. The

summary of the chemical composition is shown in Table 2-1. The composition presented

indicates that the material is rich in silica while alumina can also be considered as a major

component. Minor amounts of iron, magnesium, calcium oxides and sulphur oxides were also

found. Such low sulphur content is not considered troublesome to the hardening process of CPB

as emphasized in many studies on different types of binder (Benzaazoua et al., 1999, 2002, and

2004).


The tailings had previously been stored for a long time on ground surface and might have been

subjected to oxidation and alteration. However, all the tests were done on tailings in their current

state as used in the CPBS mixture.

The glacial sand used at the mine was obtained from native eskers. Glacial sand is naturally

silica rich and has not undergone any mineral processing. Therefore, no chemical composition

tests were performed as sand is considered inert towards any chemical reaction taking place

during binder hydration. The particle size distribution of sand is shown in Figure 2-4. The

specific gravity of sand is 2.70, obtained in accordance to ASTM D854-06 and ASTM C127-07.

The mine used a pre-blended binder provided by Lafarge and comprised of 90% blast furnace

slag (BFS) and 10% Type 10 Portland cement (PC). In the construction industry, BFS can

compose up to 70% of the blended cement which is significantly lower than the proportion of

BFS in the binder used in this study. BFS improves the workability of the fresh mix, reduces the

water demand due to the presence of high glassy content, lowers the heat of hydration, and

alleviates the deleterious effects of an alkali-silica reaction (Simon, 2005). The chemical

composition of the binder obtained via XRF is presented in Table 2-1. The binder primarily

contains calcium and silica with minor amounts of magnesium and alumina. The binder is mixed

with tailings and sand at 2.2% and 4.5% of the dry weight of solids. Water is added to reach a

water content of 28% to generate a paste with a consistency of wet concrete, and a slump of

between 140 and 190 mm (5.5 and 7.5 inches).

Specimens tested in this study were composed of uncemented 100% mine tailings (MT), an

uncemented mixture of 55% MT and 45% sand (MTS45), and an uncemented mixture of 45%

MT and 55% sand (MTS55). The particle size distribution curves of the uncemented mixtures

are shown in Figure 2-4. According to USCS, the glacial sand is classified as poorly graded sand

with some silt, the mine tailings as sandy silt, and the mixtures as poorly graded (moderately

gap-graded) silty sands. From Figure 2-4, both MTS45 and MTS55 contain 38% and 29% finer

than 20 microns; thus, segregation of coarser particles is not expected. Cemented specimens of

the same mixtures were also tested at binder contents of 2.2% and 4.5% and will be denoted as

CPBS45/2.2, CPBS55/2.2, CPBS45/4.5, and CPBS55/4.5.

Thottarath (2010) investigated the setting characteristics of the cemented mixes by measuring the

electric conductivity (EC) evolution in both CPBS55/2.2 and CPBS55/4.5 and the results are


shown in Figure 2-5. The peak of the EC response represents the start of the acceleration phase

in binder hydration. The acceleration phase marks an increase in ionic consumption and solid

particle formation. This peak usually coincides with the initial set values obtained by Vicat

needle penetration tests (Saebimoghaddam, 2010). As it appears in Figure 2-5, EC response in

CPBS55/2.2 and CPBS55/4.5 reaches the peak at 3200 mins (≈53 hours) and 700 mins (≈12

hours), respectively.

2.4 Sample Preparation

To prepare cemented specimens for a triaxial test, standard preparation methods such as moist

placement, dry deposition, and water sedimentation are not suitable as the components had to be

mixed with cement and water in advance of placement. In this testing program, specimens of 70

mm diameter are prepared according to the method suggested by Crowder (2004), le Roux

(2004) and Saebimoghaddam (2010) for creating triaxial CPB specimens. Binder, sand, and MT

are mixed at the desired moisture content (28%). In this procedure, the constituents are mixed

continuously for 10 minutes and no signs of segregation were observed. Moisture content is then

measured before casting into the split mould. The mixture is then cast into the mould on three

lifts. A glass rod is used to spear each lift around 20 times to remove the large entrained air

voids. This “rodding” technique is essentially modified moist tamping (le Roux, 2004). The

specimen‟s height is measured prior to a one dimensional consolidation procedure under a 5 kg

dead weight (equivalent to about 12.5 kPa) that lasts for one hour. This pressure is equivalent to

that exerted by a 0.6 m column of such material with a bulk density of 20 kN/m3 after mixing.

Knowing that the rise rate in the stope is 0.25 m/hr, such pressure could be reached after about

three hours of filling. After the dead weight consolidation process, the split mould was removed

while the specimens self-stood. The final height and collected water are measured for void ratio

calculations. Back saturation followed according to ASTM D4767-04. A minimum B-Value of

0.96 was reached at back pressures ranging between 200 and 250 kPa. The specimen was then

consolidated for one hour under an isotropic pressure to reach the desired effective confining

pressure. The whole process takes four hours from mixing to loading. Tests done to monitor

hydration showed that the initial set is not earlier than 12 hours. Therefore, at four hours age, the

binder contribution to liquefaction resistance is considered negligible.


To obtain uniform laboratory specimens that result in a representative behaviour of the soil

skeleton, oversized particles in the glacial sand were removed prior to mixing with other

constituents. This process is known as scalping. It should be noted that the mechanical and

hydrological behaviour of the specimen is not affected by the scalping process as long as the

removed particles float in the finer matrix (Khalili et al., 2010). For the glacial sand used in this

study, the cut-off particle size was 6.3 mm which results in a diameter to maximum grain size

ratio of around 11. The minimum diameter to maximum grain size ratio for a consolidated

undrained triaxial test specimen is 6 as recommended by the ASTM D4767-04 while

Jamiolkowski et al (2005) argued that 8 is an ideal ratio. In the case presented herein, removed

particles form < 6% of the glacial sand and drops to < 3% of the total weight of the mixtures.

Therefore, this small fraction is expected to be isolated within the finer soil matrix.

2.5 Testing Program

The focus of this experimental program is to assess how adding different proportions of sand or

sand and binder to mine tailings affect the liquefaction potential and the mechanical undrained

response of the paste. A series of monotonic compression and extension triaxial tests were

performed. CU tests were conducted on the uncemented specimens of MT, MTS45 and MTS55

and cemented specimens of CPBS45/2.2, CPBS55/2.2, CPBS45/4.5, and CPBS55/4.5.

The experiments were performed at effective confining pressures ranging from 25 kPa to 400

kPa and void ratios varying according to the percentage of sand or sand and binder in the

mixture. The axial loading stage was strain controlled for all experiments at an axial strain rate of

2%/min in compression and of -2%/min in extension. In addition, the response of MT and

MTS55 at different strain rates in compression was investigated to examine the validity of using

2%/min, which is higher than the strain rates applied to silty sands and silts.

For a deeper insight on the behaviour of the tested materials and for a more accurate

determination of the state lines, CD triaxial compression tests were also conducted. Only the

uncemented MT and MTS55 were tested under drained conditions. The axial strain rate to ensure

drainage is calculated based on the coefficient of consolidation (Cv) which varies, therefore the

calculated axial strain rate for drained tests varied from 0.045 %/min to 0.2 %/min for both MT

and MTS55. A strain rate of 0.05 %/min was selected to be close to the lower calculated value.


The results of CD tests will complement the CU testing program and give a better understanding

of how the material behaves.

2.6 Consolidation Response

The lines of emax and emin are defined for different mixtures of sand and MT as shown in Figure

2-6. The global void ratios at monotonic loading for specimens tested at different effective

confining stresses (′c) with different FC are also plotted in black markers on the same chart in

Figure 2-6. The initial global void ratios resulting from the previously mentioned preparation

method are relatively low. Consequently, the resulting e after isotropic consolidation is close to

the emin line (see Figure 2-6). At the end of the hydrostatic consolidation stage, the observed

response of volumetric strain (v) versus axial strain (a) for the triaxial specimens is shown in

Figure 2-7. The gradients of the best-fit lines represent v/a values that range from 3.18 to

3.68 depending on the mixture as summarized in Table 2-2. Isotropic response should result in

gradients of 3.0 under perfect hydrostatic loading. The gradients of cemented and uncemented

mixtures compare well with those presented by Khalili et al. (2010).

Compressibility response of cemented and uncemented specimens consolidated under stresses

ranging from 25 kPa to 400 kPa is shown in Figure 2-8 in the e-log ′c space, where it is clear

that the logarithmic trend lines fit the data points with coefficients of determination (R2) > 0.92.

Figure 2-8a shows the relationship between the global void ratio (e) and log ′c, while Figure 2-

8b and Figure 2-8c show the same relationship in terms of the intergranular void ratio (ec) and

the interfine void ratio (ef), respectively. Those void ratios are calculated according the following

equations (after Thevanayagam et al., 2002):

FC

FCeec

1 [2- 1]

and

FC

ee f [2- 2]

Also, the corresponding compression indices Cc, Ccc and Ccf are calculated from the fitted lines

based on e, ec, and ef and values are summarized in Table 2-3. Adding sand or sand and binder

drastically changed the compression index. When only sand is added to MT, Cc, Ccc, and Ccf


decreased around 70% to 50%. When the binder is added to the MT-Sand mixtures, Cc, Ccc, and

Ccf increased around 100%, thus bringing the compressibility back close to that of MT. Although

the compressibility of the MTS45 was found too close to that MTS55, cemented specimens with

higher sand content (CPBS55) had lower compressibility than that of the lower sand content

(CPBS45) and increasing the binder content has significantly decreased the compressibility in

comparison to the lower sand content. The question that might be answered in the coming

sections is whether such mixtures with different compressibilities behave differently under

monotonic loading. The compression index of uncemented MT specimens tested was less than

half the values reported by Crowder (2004) and Saebimoghaddam (2010) for two different silt-

size gold tailings. Such lower compressibility of MT can be due to the higher fraction of course

grained soil that was initially present than in the previous studies.

2.7 Undrained Response to Monotonic Compression Loading

The stress path results in this research work are presented in the s´-t space. The stress path and

stress-axial strain plots use the variables normal stress (s´) and the shear stress (t), which are

defined as follows.

2'

'' 31 s

, 2

'' 31 t

[2- 3]

In Mohr stress space, s´ and t variables represent the top point of the Mohr circle representing the

stress state at any moment during the test. The state line has an inclination of ´ in Mohr stress

space, while it has an inclination of ´ in the s´-t space. The relation between the inclinations of

the same state line in both Mohr and s´-t stress spaces can be correlated by tan ´ = sin ´. Many

studies plot their stress paths in the p´-q space, where p´= (´1+ 2´3)/3 and q= (´1-´3).

2.7.1 Uncemented mine tailings

This section presents the monotonic test results for five MT specimens at five different effective

confining pressures ranging from 25 kPa to 400 kPa as shown in Figure 2-9. The void ratios, FC

and ′c for the specimens are summarized in Table 2-4a. Figure 2-9a shows the undrained stress

paths of MT specimens. Although the void ratios vary from 0.680 to 0.555, all the specimens

experienced the same initial contractive behavior that was followed by a dilative behavior.


Specimens tended to contract at the initial stages of axial straining and this tendency to

contraction was reflected by increase in pore water pressure generation and observed deviation

from the hypothetical drained path. The tendency to contraction terminated when the pore water

pressure reached a maximum value (umax) at PTP, beyond which the specimens tended to dilate

as shown by the decrease in the pore water pressure. It was observed that dilation of specimens

decreased with the increase in the effective confining pressure. Figure 2-9b shows the pore

pressure ratio ru (u/ć) versus strain. A good statistical fit of PTL was obtained by defining a

unique PTP in the s´-t space for each specimen. The PTL has an angle ´PT= 29.2o in the s´-t

space and a corresponding angle ´PT= 34.0o in the Mohr stress space.

All specimens exhibited strain hardening behavior. As shown in Figure 2-9c, neither flow

liquefaction nor temporary liquefaction was observed at all effective confining pressures.

Furthuremore, it is clear that the normal stress (s′) increased indefinitely without reaching a

plateau (Figure 2-9c) and the excess pore water pressure decreased asymptotically (Figure 2-9b)

indicating that no steady state was reached until the tests were stopped at approximately 20%

strain. For such silty material, however, UFL can be defined at maximum stress obliquity points

(MSOP), i.e., maximum t/s´= (´1-´3)/(´1+´3)) as in Figure 2-9d. This state was temporarily

reached and, then, the ratio t/s´ deviated from the horizontal and this deviation decreased by

increasing the effective confining pressure. A good statistical fit of UFL to the maximum t/s´

points has an angle ÚFL= 31.0o in the s´-t space and a corresponding angle ÚFL= 36.9

o in the

Mohr stress space. A summary for the slopes of UFL and PTL of all the mixtures is given in

Table 2-5.

2.7.2 Uncemented MT-Sand mixtures

The monotonic test results for five MTS45 specimens at different effective confining pressures

are shown in Figure 2-10. The void ratios, FC and ′c of each specimen are presented in Table 2-

4a. The entire set of specimens exhibited similar behavior to that of the MT specimens. The

angles of UFL and PTL in the s´-t space are only within the order of two degrees higher than

those of MT (see Table 2-5). The specimen tested at ć=50 kPa shows distinctive t/s´ peak more

than the other specimens (Figure 2-10d).


The monotonic test results for three MTS55 specimens at different effective confining pressures

are shown in Figure 2-11. Void ratios and FC are presented in Table 2-4. All MTS55 specimens

exhibited similar behavior to that of MT and MTS45 specimens. Although MTS55 has a larger

percentage of course grained particles, MTS55 specimens showed more contraction (PWP

generation) and less dilation than what is observed in the case of MT and MTS45.

The angles of UFL and PTL in the s´-t space are only within the order of one degree less than

those of MTS45 despite the larger sand content that MTS55 contains (see Table 2-5). The

specimen tested at ´c=25 kPa exhibited more distinctive stress obliquity peak than the other two

specimens (Figure 2-11d).

For the sake of comparison, specimens composed of glacial sand only were tested and were

found to exhibit much less contraction and considerably higher rates of dilation than that of all

the above-mentioned specimens. Moreover, sand specimens exhibited typical dense sand

behavior and showed more distinctive strain hardening behavior than MT and MT-sand mixtures

and reached a steady state at an axial strain of approximately 12%. Consequently, it is clear that

the behavior of MT-sand mixtures was far from being influenced by sand and was dominated by

the behavior of MT.

2.7.3 Cemented paste backfills with sand

Monotonic test results for the specimens of CPBS45/2.2, CPBS55/2.2, CPBS45/4.5, and

CPBS55/4.5 are shown in Figure 2-12 to Figure 2-15. The void ratios and ′c for the specimens

are summarized in Table 2-4a. Nonplastic fines contents are also listed in Table 2-4a. However,

it is worth mentioning that binder content contributes to the FC calculated for each of the

cemented specimens. The behavior of the cemented specimens was essentially similar behavior

to that of MT-Sand mixtures. At both sand contents, the rate of dilation was milder than that of

their uncemented counterparts. This is more pronounced in the case of 4.5% binder. The angles

of UFL and PTL in s´-t space are only within the order of 1.5o higher than those of their

uncemented counterparts (see Table 2-5). There is no specific relationship observed between the

slopes of UFL and PTL and the binder content.


2.7.4 Effect of strain rate on uncemented specimens

Strain rate values in strain controlled triaxial testing were determined to ensure pore water

pressure equilibration during the test. In the present thesis, the strain rate (2%/min) under which

the specimens were loaded was higher than the common values used for materials of similar

hydraulic conductivity. However, the reason behind working at a relatively high strain rate is to

minimize the effect of the ongoing hydration process on the response of the cemented blends

during the experiment. Uncemented specimens were tested at the same strain rates to allow

comparison. Knowing that the strain rate influences the behavior of materials, it is important to

investigate the possible variations in the behavior in response to changing the strain rate and to

examine the validity of using such a relatively high strain rate. MTS55 specimens were selected

for testing at different strain rates because it is the most commonly used mixture at the mine after

adding binder, while MT specimens were also tested as control specimens to investigate how

adding sand may affect the strain rate dependence of the paste mixture. Consistent void ratios

were obtained when specimens were hydrostatically consolidated under 200 kPa effective

confining pressure. Therefore, monotonic compression tests were conducted at this stress level.

Four specimens of MTS55 were tested at 200 kPa effective confining pressure under axial strain

rates of 2%, 1%, 0.1%, and 0.02%/min. The pre-shearing void ratio was 0.320 for specimens

tested at 2%, 0.1%, and 0.02%/min and was 0.315 for the specimen tested under 1%/min. The

difference is within the ±0.005 margin of error in calculating the void ratio.

The stress paths of the specimens are shown in Figure 2-16a, where it appears that all the

specimens exhibited the same behavior as that of the 2%/min specimen. Although PWP

generated at lower strain rates was slightly higher than that at 2%/min, it did not show significant

changes in either the behavior or the location of the state lines if plotted for each specimen

individually. PTL has an angle ´PT= 31.6 o

in the s´-t space and a corresponding angle ´PT=

38.0o in the Mohr stress space. Whereas, UFL has an angle ´UFL= 32.3

o in the s´-t space and a

corresponding angle ´UFL= 39.2o in the Mohr stress space.

Three MT specimens were tested at 200 kPa effective confining pressure under axial strain rates

of 2%, 0.05%, 0.02%/min. The pre-shearing void ratios were 0.61±0.005. In this case, where no

sand was added, the behavior at lower strain rates is substantially different from that of the

2%/min specimen (see Figure 2-17). PWP generated at 0.05% and 0.02%/min is significantly


higher than that generated at 2%/min as can be seen in Figure 2-17b. Although the higher PWP

brought the specimen towards more instability, the specimen did not exhibit an explicit limited

liquefaction behavior. However, the higher PWP generated at the lower strain rates could be a

result of the more uniform PWP distribution across the specimen in comparison to the case of

higher strain rates. In contrast to the rest of the specimens tested in this research, the specimen

tested at 0.05%/min exhibited strain softening behavior and the specimen tested at 0.02%/ min

exhibited very mild strain hardening behavior (Figure 2-17c). PTL has an angle ´PT= 30.9 o

in

the s´-t space and a corresponding angle ´PT= 36.8o in the Mohr stress space. Whereas, UFL has

an angle ´UFL= 31.9 o

in the s´-t space and a corresponding angle ´UFL= 38.5o in the Mohr stress

space.

2.8 Undrained Response to Monotonic Extension Loading

MT, MTS, and CPBS specimens were subjected to undrained extension loading at a strain rate of

-2%/min. The behavior of MT, MTS, and CPBS specimens was found to be generally the same.

The material started as contractive and transfered to dilative at a unique point until it reached

UFL. However, and in contrast to compression, all specimens exhibited temporary instability

before going through phase transformation. Extension results are shown in Figure 2-9 to Figure

2-15 and the pre-shearing void ratio of each specimen are listed in Table 2-4b. Necking was

observed for all samples tested at 100 and 200 kPa. In the shear stress (t) versus axial strain plots

in Figure 2-9 to Figure 2-15, necking occurred at the second peak after the UFS was already

reached. It was also observed that necking occurred at smaller strains as ‟c increased.

A good statistical fit of the temporary instability line (TIL) was obtained by defining a unique

temporary instability point (TIP) in the s´-t space for each specimen. The angles of UFL, PTL,

and TIL in the s´-t space are listed in Table 2-5a. Angles of UFL and PTL for all specimens (MT

and mixtures) in extension are lower than those in compression which may be attributed to

anisotropy. In the case of MT specimens, the difference between the angles of these state lines in

compression and extension is about 3o for UFL and is about 6.5

o for PTL, all expressed in the s´-t

space. By adding sand, this difference increased indicating higher anisotropy, reaching the order

of 5o for UFL and approximately 10

o for PTL; for both sand contents. Adding binder to the MT-

sand mixtures was observed to cause increase in the above mentioned difference more than for

their uncemented counterparts indicating much higher anisotropy in the case of CPBS. The


difference is in the order of 9o for UFL and is about 13

o for PTL for both binder contents. This

difference did not show any particular pattern of change by changing the binder content.

Generally, when comparing uncemented MT-sand mixtures with MT, the angles of UFL and

PTL in the s´-t space in extension are only within the order of 2o less than those of MT.

Moreover, when comparing cemented MT-sand mixtures (CPBS) with their uncemented

counterparts, the angles of UFL and PTL in the s´-t space in extension are within the order of 3o

less.

In contrast to the behavior in compression, the effective confining pressure caused significant

influence on PWP generation. For all cemented and uncemented specimens, the generated PWP

considerably decreased by increasing the effective confining pressures indicating higher stability.

This agrees with the results reported by Saebimoghaddam (2010). This influence is less

pronounced in the cemented specimens.

2.9 Drained Response

CD triaxial tests were conducted on specimens of MT and MTS55 under 0.05%/min axial strain

rate in compression. Drained test results for three MT specimens at three different effective

confining pressures of 50 kPa, 100 kPa and 200 kPa are shown in Figure 2-18. The void ratios

prior to shearing for specimens tested at 50, 100, and 200 kPa are 0.670, 0.640, and 0.600,

respectively. In contrast to the undrained specimens, volume change measurements for drained

specimens tested at ‟c = 100 kPa and 200 kPa show contractive behavior with no transition to

dilation, while the specimen tested at ‟c = 50 kPa exhibited slight dilation (see Figure 2-18b).

Nevertheless, a steady state was not reached and, instead, the UFS was defined. Moreover, PTL

was not determined for drained MT specimens. The UFL has an angle ´UFL= 30.6o in the s´-t

space which is almost the same as that obtained from the undrained test (´UFL undrained= 31.0o).

Drained test results for four MTS55 specimens at different effective confining pressures of 25

kPa, 50 kPa, 100 kPa, and 200 kPa are shown in Figure 2-19. The void ratios prior to shearing

for specimens tested at 25 kPa, 50 kPa, 100 kPa, and 200 kPa are 0.360, 0.355, 0.350 and 0.350,

respectively. All drained specimens showed strain hardening behavior. Volume change

measurements in Figure 2-19b show initial contraction followed by a dilative behavior but

volume did not fully recover until the end of the test. Saebimoghaddam (2010) observed the


same behavior for silt-sized gold tailings. The UFL has an angle ÚFL= 31.4o in the s´-t space

which is the same as that obtained from the undrained test (ÚFL undrained= 31.5o). Similarly, PTL

has an angle ÚFL= 30.9o in the s´-t space which is almost the same as that obtained from the

undrained test (´PTL undrained= 30.6o).

2.10 Discussion

Neither flow liquefaction nor temporary instability was reached by all MT, MTS and CPBS

specimens tested in compression within the range of effective stresses (25 kPa to 400 kPa) and

the testing axial strains. For a better understanding of the materials‟ behavior, the state lines are

plotted in the pre-shearing global void ratio versus the mean effective stress (e-log p') space in

Figure 2-20. UFL and PTL obtained from the monotonic compression testing are shown in

Figure 2-20a and Figure 2-20c, respectively. From these plots, it can be seen that UFL and PTL

of each of the cemented mixtures are not affected by the binder content and their position

remained almost the same. PTPs and MSOPs obtained from CD tests on MT and MTS55 are

plotted on the same charts in solid black markers (see Figure 2-20a and c). The points show good

fit to the corresponding state lines obtained from CU tests. Since MT under drained conditions

did not undergo a PTS, MSOPs are plotted in both e-log p' charts. The MSOPs from CD tests on

MT show a better fit to the PTL obtained from CU tests on MT. Although this cannot be asserted

as a final conclusion as more data will be needed, this might agree with the conclusion reported

by Chern (1985) that the angle of PTL determined from the undrained triaxial testing on

saturated sands is equal to that of SSL determined from drained loading.

UFL and PTL obtained from monotonic extension tests are shown Figure 2-20b and Figure 2-

20d, respectively. Similar to compression, UFL and PTL of each of the cemented mixtures were

not affected by the binder content and their positions remained almost the same. Compression

and extension state lines are plotted on the same scale for the sake of comparison. The positions

of UFLs and PTLs obtained from extension tests, compared with those obtained from

compression tests, are almost identical for all materials except for MT, which showed a slight

difference between the slopes in the e-log p' space. This observation seems to contradict the

previous observation of unequal slopes of the state lines, in the s'-t' space, obtained from

compression and extension. Most studies suggest that this inequality in the s'-t' space is a result

of the anisotropy of the soil fabric (for example, Vaid and Thomas, 1995; Khalili et al., 2010;


and Saebimoghaddam, 2010). However, Riemer and Seed (1997) obtained three different SSL by

testing isotropic fine sand specimens in triaxial compression, triaxial extension, and simple

shear. They attributed the differences to different stress paths followed in the different tests. The

materials tested in the present research underwent a one dimensional consolidation stage, which

makes the anisotropy of the fabric a more plausible reason for the inequality of the state

parameters in the s'-t' space.

Saebimoghaddam (2010) conducted a similar research, using the same apparatus, on a different

mine tailings material that contained only 10% particles > 75 micron. He observed anisotropy

but not to same extent even when compared with MT specimens tested in the present research.

Knowing that Saebimoghaddam (2010) followed the same preparation method as in this thesis,

the higher anisotropy in the present work can be attributed to the presence of higher coarse

grained fraction in MT specimens relative to what Saebimoghaddam (2010) tested and also to the

presence of much larger portions of sand in the MT-Sand mixtures and CPBS. Furthermore,

significant differences in particle fabrics are observed from the SEM images shown in Figure 2-

21. In the current work, the tested MT has more flakey particles and a substantial particle size

variation, as shown in Figure 2-21a. On the other hand, particles of William‟s tailings tested by

Saebimoghaddam (2010) are more round and uniform. The more flakey particles might be the

reason for a more anisotropic fabric when such preparation method is followed. The flakey

particles could have resulted in a week particle configuration under extension loads. However, it

should be noted that the observed anisotropy correspond to large strains only. Such anisotropy in

the behaviour may not be observed at smaller strains as will be shown in Chapter 3 and Chapter

4. Since this preparation method was developed based on the placement method in field, the

behaviour of CPBS in the field is expected to be the same to that observed in compression and

extension tests depending on the closest stress path.

The limiting FC or the limiting silt content (LSC), beyond which sand particles are no longer

contiguous, is determined to be 22% for the MT-sand mixtures tested in the present work. LSC is

calculated according to the method presented in Chapter 3. The calculated LSC asserts that the

fine matrix controlled the behaviour of all cemented and uncemented paste-sand mixtures.

Another observation that can be made from the state lines plots is that the state lines of the

cemented mixtures (CPBS) are parallel to those of MT and not to those of their uncemented


counterparts (MTS). Also, the values of Cc for CPBS are closer to that of the MT rather than that

of their uncemented counterparts. This could be an indicator that although the increase in FC due

to the addition of binder is very small, the addition of binder to MT-sand mixtures increases the

dominance of the finer matrix to the behaviour of the overall mixture. Bearing in minds the

above findings, it can be asserted that adding binder had only positive impact on the material‟s

behaviour in compression, while it has a negative impact on the behaviour in extension after four

hours of curing.

To summarize the results and shed some light on their engineering significance, the state

parameter ´UFL, as a strength indicator, is plotted against the coarse grains content in Figure 2-

22 for the cemented and uncemented specimens tested in this research work. Figure 2-22 shows

that despite the obtained increase in strength upon adding sand and binder to MT, which varied

with changing the sand content, a corresponding decrease of the strength in extension was more

significant. Moreover, the strength of the cemented mixtures in extension showed considerable

deterioration upon adding binder. This is considered alarming if stopes backfilled with CPBS are

expected to undergo extension load paths during early curing ages.

2.11 Conclusions

- An extensive triaxial testing program was performed to investigate the effect of adding

sand or sand and binder to mine tailings on the liquefaction potential and mechanical

behavior. Generally, flow liquefaction was not observed for MT, MTS or CPBS (cured

for four hours) under monotonic compression loading. All cemented and uncemented

specimens exhibited limited instability under monotonic extension loading and the

effective confining pressure had a large influence on PWP generation, in contrast to the

compression specimens.

- The compressibility of MT tested in the present research was considerably lower than

that of the silt-sized mine tailings tested in previous studies, which is attributed to the

initially present fraction of course grained particles, which is higher than those reported

in previous studies.

- The addition of sand to MT reduced the compressibility of the mixture. The slopes of the

state lines in the s'-t' space increased in compression while decreased in extension.


Specimens exhibited higher anisotropy when sand was added to MT when comparing

extension and compression state lines.

- The addition of binder to MT-sand mixtures increased the compressibility of the mixture

making it close to those of MT. The slopes of the state lines in the s'-t' space increased in

compression while significantly decreased in extension compared to their uncemented

counterparts. Cemented specimens exhibited higher anisotropy by comparing extension

and compression state lines. This is considered alarming if stopes backfilled with CPBS

are expected to undergo extension load paths during early curing ages. Generally, the

behavior did not change by adding binder. However, less dilation was observed

especially at the higher sand content.

- The behavior of MT dramatically changed at lower strain rates showing a higher

tendency to instability. However, no temporary instability was observed and MT

specimens continued through PT as at high strain rates. On the other hand, MTS55

exhibited no significant changes by changing the strain rate. Therefore, adding 55% sand

reduced the strain rate dependence observed with MT.

- Drained tests on MT specimens showed completely different behavior from that of the

undrained tests. The drained behavior was found to be contractive; however, the slope of

the UFL was the same. The response of MTS55 under drained loading was initially

contractive and then demonstrated dilation without recovering the initial volume. The

state lines obtained from drained and undrained tests for MTS55 were found to be

essentially the same.

- Analyzing the state lines in the e-log p' space did not distinguish the anisotropy observed

in the s'-t' space. Also, the state lines in the e-log p' and Cc values showed that the

behavior of CPBS is more influenced by the finer fraction at both sand contents.

Generally, the behavior of all cemented and uncemented specimens appear to be

dominated by the fine-grained matrix. The extent to which the fine matrix dominated the

behavior varied depending on the mixture.

One of the main implications of the present work on backfill design and practice is that the

designers of paste backfilling systems should account for the anisotropy at large strains that


increases by adding sand and binder. The strength that CPB or CPBS exhibits under an

increasing overburden pressure exerted by an ongoing pour (monotonic compression) is expected

to be different from what it exhibits under rock wall closure. The stress path due to rock wall

closure can be too close to that of the extension tests, under which the material exhibited a

weaker response and temporary liquefaction. The preparation method followed in this research is

believed to replicate the field placement technique. However, further research is needed to obtain

CPB specimens at higher void ratios. This will help in cases of narrow stopes where rise rates are

much higher.


2.12 References

Al Tarhouni, M. 2008. Liquefaction and post-liquefaction behaviour of gold mine tailings under

simple shear loading. M.A.Sc., Carleton University, Canada.

Aref, K. 1989. A study of the geotechnical characteristics and liquefaction potential of paste

backfill. Ph.D., McGill University, Montreal, Canada.

ASTM C136 2006. Standard test method for sieve analysis of fine and coarse aggregates. The

Annual Book of ASTM Standards, Sect 4, Vol. 04.02.

ASTM D854 2006. Standard test methods for specific gravity of soil solids by water pycnometer.

The Annual Book of ASTM Standards, Sect 4, Vol. 04-08.

ASTM D422–63 2007. Standard test method for particle size analysis of soils. The Annual Book

of ASTM Standards, Sect 4, Vol. 04-08.

ASTM D4318–05 2007. Standard test methods for liquid limit, plastic limit, and plasticity index

of soils. The Annual Book of ASTM Standards. Sect 4, Vol. 04-08.

ASTM C127 2007. Standard Test Method for Density, Relative Density (Specific Gravity) and

Absorption of Coarse Aggregate, The Annual Book of ASTM Standards, Sect 4, Vol. 04-

08

ASTM D4767 2004. Standard test method for consolidated undrained triaxial compression test

for cohesive soils. The Annual Book of ASTM Standards, Sect 4, Vol. 04-08.

Been, K., Brown, E.T., and Hepworth, N. 2002. Liquefaction potential of paste fill at Neves

Corvo mine, Portugal. Institution of Mining and Metallurgy Transactions. Section A:

Mining Technology, 111(JAN/APR): A47-A58.

Benzaazoua, M., Fall, M., and Belem, T. 2004. A contribution to understanding the hardening

process of cemented pastefill. Minerals Engineering, 17(2): 141-152.

Benzaazoua, M., Belem, T., and Bussière, B. 2002. Chemical factors that influence the

performance of mine sulphidic paste backfill. Cement and Concrete Research, 32(7):

1133-1144.

Benzaazoua, M., Ouellet, J., Servant, S., Newman, P., and Verburg, R. 1999. Cementitious

backfill with high sulfur content physical, chemical, and mineralogical characterization.

Cement and Concrete Research, 29(5): 719-725.

Boulanger, R.W. and Idriss, I.M., 2004. Evaluating the potential for liquefaction or cyclic failure

of silts and clays. Report No. UCD/CGM-04/01, Center for Geotechnical Modeling,

University of California, Davis, CA, 130 p.

Broomfield, D. 2000. Liquefaction potential of paste backfill. M.Sc., Queen‟s University,

Canada.

Chern, J. 1985. Undrained response of saturated sands with emphasis on liquefaction and cyclic

mobility. PhD, University of British Columbia, Canada.

Clough, G.W., Iwabuchi, J., Shafii Rad, N., and Kuppusamy, T. 1989. Influence of cementation

on liquefaction of sands. Journal of Geotechnical Engineering, 115(8): 1102-1117.


Crowder, J.J. 2004. Deposition, consolidation, and strength of a non-plastic tailings paste for

surface disposal. PhD, University of Toronto, Canada.

Fourie, A.B. and Papageorgiou, G. 2001. Defining an appropriate steady state line for

merriespruit gold tailings. Canadian Geotechnical Journal, 38(4): 695-636.

Fourie, A.B., Blight, G.E., and Papageorgiou, G. 2001. Static liquefaction as a possible

explanation for the merriespruit tailings dam failure. Canadian Geotechnical Journal,

38(4): 707-719.

Ishihara, K. 1993. Liquefaction and flow failure during earthquakes. Geotechnique, 43(3): 351-

415.

Jamiolkowski, M., Kongsukprasert, L., and Lo Presti, D. 2005. Characterization of gravelly

geomaterials. In Proceedings of The Fifth International Geotechnical Conference, Cairo,

Egypt, 12 January 2005. pp. 1–27.

Khalili, A., Wijewickreme, D., and Wilson, G. 2010. Mechanical response of highly gap-graded

mixtures of waste rock and tailings. part I: Monotonic shear response. Canadian

Geotechnical Journal, 47(5): 552-565.

Kramer, S.L. 1996. Geotechnical earthquake engineering. Prentice Hall, Upper Saddle River,

N.J.

Kuerbis, R., Negussey, D., and Vaid, Y.P. 1988. Effect of gradation and fines content on the

undrained response of sand. Hydraulic Fill Structures, pp. 330-345.

Lade, P.V. and Yamamuro, J.A. 1997. Effects of nonplastic fines on static liquefaction of sands.

Canadian Geotechnical Journal, 34(6): 918-928.

le Roux, K. 2004. In situ properties and liquefaction potential of cemented paste backfill. PhD,

University of Toronto, Canada

Pierce, M., Bawden, W.F., and Paynter, J.T. (1998) Laboratory testing and stability analysis of

paste backfill at the golden giant mine. In Minefill ‟98: Proceedings of the 6th

International Symposium on Mining with Backfill, Brisbane, pp. 159–165.

Pitman, T.D. 1994. Influence of fines on the collapse of loose sands. Canadian Geotechnical

Journal, 31(5): 728-739.

Riemer, M. and Seed, R., 1997. Factors affecting apparent position of steady-state line. Journal

of Geotechnical and Geoenvironmental Engineering, 123: 281-8.

Saebimoghaddam, A. 2010. Liquefaction of early age cemented paste backfill. PhD, University

of Toronto, Canada.

Simon, D. 2005. Microscale analysis of cemented paste backfill. PhD, University of Toronto,

Canada.

Sladen, J.A., D'Hollander, R.D., Krahn, J., and Mitchell, D.E. 1985. Back analysis of the Nerlerk

berm liquefaction slides. Canadian geotechnical journal, 22(4): 579-588.

Thevanayagam, S. 1998. Effects of fines and confining stress on undrained shear strength of silty

sands. Journal of Geotechnical and Geoenvironmental Engineering, 124(6): 479.

Thevanayagam, S. , Mohan, S. 2000. Intergranular state variables and stress-strain behaviour of

silty sands. Géotechnique, 50(1): 1-23.


Thevanayagam, S., Shenthan, T., Mohan, S., and Liang, J. 2002. Undrained fragility of clean

sands, silty sands, and sandy silts. Journal of Geotechnical and Geoenvironmental

Engineering, 128(10): 849-859.

Thomas, J. 1992. Static, cyclic and post liquefaction undrained behaviour of fraser river sand.

MASc, The University of British Columbia, Canada.

Thottarath, S. 2010. Electromagnetic Characterization of Cemented Paste Backfill in the Field

and Laboratory. MASc, The University of Toronto, Canada.

Vaid, Y.P. and Thomas, J. 1995. Liquefaction and postliquefaction behavior of sand.

International Journal of Rock Mechanics and Mining Sciences & Geomechanics

Abstracts, 32(8, pp. 379A-379A): December.

Wickland, B.E., Wilson, G.W., Wijewickreme, D. and Klein, B. 2006. Design and Evaluation of

Mixtures of Mine Waste Rock and Tailings. Canadian Geotechnical Journal. Vol 43(9)

pp. 928-945.

Yamamuro, J.A. and Lade, P.V. 1997. Static liquefaction of very loose sands. Canadian

geotechnical journal, 34(6): 905-917.

Yamamuro, J.A. and Lade, P.V. 1998. Steady-state concepts and static liquefaction of silty

sands. Journal of Geotechnical and Geoenvironmental Engineering, 124(9): 868-877.

Yamamuro, J.A. and Covert, K.M. 2001. Monotonic and cyclic liquefaction of very loose sands

with high silt content. Journal of Geotechnical and Geoenvironmental Engineering,

127(4): 314-324.


Tables and Figures


Table 2-1 Chemical compositions of the tailings and binder received from the mine.

Oxide Tailings (%) Binder (%)

SiO2 47.1 30.5

Al2O3 13.1 7.3

Fe2O3 6.7 0.7

MgO 4.8 11.1

CaO 6.4 47.4

Na2O 1.7 0.4

Ba 0.1 0.1

SO3 4.4 1.1

K2O 2.7 0.5

others 1 1.1


Table 2-2 Gradients of the best-fit lines indicating v/a values.

Specimen v/a

MT 3.18

MTS45 3.48

MTS55 3.62

CPBS45/2.2 3.37

CPBS45/4.5 3.32

CPBS55/2.2 3.68

CPBS55/4.5 3.55


Table 2-3 Compression indices for the tested MT, MTS, and CPBS specimens.

Material Cc Ccc Ccf

MT 0.048 0.138 0.073

MTS45 0.016 0.03 0.036

MTS55 0.017 0.027 0.044

CPBS45/2.2 0.036 0.07 0.075

CPBS45/4.5 0.035 0.068 0.068

CPBS55/2.2 0.03 0.045 0.067

CPBS55/4.5 0.025 0.04 0.054

Sand 0.017 0.018


Table 2-4 Summary of the key parameters for specimens tested under udrained monotonic (a) compression at

2%/min, and (b) extension at -2%/min.

(a) (b)

Material FC (%)

Specimen ID

'c (kPa)

e Material FC (%)

Specimen ID

'c (kPa)

e

MT 70.0

1 25 0.680

MT 70.0

6 50 0.660

2 50 0.665 7 100 0.605

3 100 0.655 8 200 0.605

4 200 0.590

MTS45 45.8

6 50 0.380

5 400 0.555 7 100 0.370

MTS45 45.8

1 25 0.385 8 200 0.350

2 50 0.380

MTS55 38.2

4 50 0.345

3 100 0.370 5 100 0.335

4 200 0.650 6 200 0.320

5 400 0.340

CPBS45/2.2 48.0

6 50 0.455

MTS55 38.2

1 25 0.360 7 100 0.445

2 200 0.325 8 200 0.415

3 400 0.315

CPBS45/4.5 50.3

4 50 0.465

CPBS45/2.2 48.0

1 25 0.495 5 100 0.460

2 50 0.470 6 200 0.415

3 100 0.440

CPBS55/2.2 40.4

4 50 0.415

4 200 0.425 5 100 0.360

5 400 0.395 6 200 0.370

CPBS45/4.5 50.3

1 50 0.470

CPBS55/4.5 42.7

4 50 0.400

2 100 0.445 5 100 0.385

3 200 0.425 6 200 0.355

CPBS55/2.2 40.4

1 25 0.425

2 200 0.385

3 400 0.300

CPBS55/4.5 42.7

1 25 0.415

2 200 0.360

3 400 0.355


Table 2-5 Slopes of the UFL, PTL and TIL, in (a) the s´-t space and (b) the Mohr stress space, from

undrained monotonic compression and extension triaxial testing at 2%/min.

(a)

Compression Extension

UFL PT UFL PT TIL

MT 31.0 29.2 28.0 22.7 13.1

MTS45 32.3 31.3 27.6 22.8 13.2

MTS55 31.6 30.6 26.1 19.3 11.3

CPBS45/2.2 32.3 31.6 23.6 17.3 10.9

CPBS45/4.5 33.7 33.2 24.5 19.9 11.4

CPBS55/2.2 32.8 32.3 24.0 19.5 11.8

CPBS55/4.5 32.8 31.7 24.1 18.4 10.8

Sand only 32.4 28.3 31.1 22.7 --

(b)


UFL PT UFL PT TIL

MT 36.9 34.0 32.1 24.7 13.5

MTS45 39.2 37.4 31.5 24.8 13.6

MTS55 37.9 36.2 29.4 20.5 11.5

CPBS45/2.2 39.2 37.9 25.9 18.2 10.2

CPBS45/4.5 41.8 40.9 27.1 21.2 11.6

CPBS55/2.2 40.2 39.2 26.5 20.8 12.1

CPBS55/4.5 40.1 38.2 26.6 19.5 10.9

Sand only 39.4 32.5 37.0 24.7 --


0 20 40 60 80 100

FC (%)

e

Maximum void ratio (emax)

Minimum void ratio (emin)

Figure 2-1 Typical relationships between void ratio (e) and the fines content (FC) at maximum and minimum

compaction.


(a) (b)

Figure 2-2 A schematic diagram showing different arrangements of soil particles in a blend for (a) loosely

packed particles and lack of contact between coarser grains; and (b) densely packed particles where coarser

grains maintain contact.


Figure 2-3 State lines and state points for materials exhibiting stable and temporary liquefaction behaviors.


Figure 2-4 Particle size distributions of the tested materials and mixtures.


Figure 2-5 Temporal response of the electric conductivity (EC) for different mixes (after Thottarath, 2010)


0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

0 10 20 30 40 50 60 70 80 90 100

e

FC (%)

Maximum void ratio (emax)

Minimum void ratio (emin)

Figure 2-6 The relationship between the global void ratio and FC at different Sand-MT mixtures with the

specimens tested at different FC, e, and 'c shown in black markers


0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0 0.005 0.01 0.015 0.02 0.025 0.03

Vo

lum

etr

ic S

train

Axial Strain

MT

MTS45

MTS55

isotropic

(a)

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0 0.005 0.01 0.015 0.02 0.025 0.03

Vo

lum

etr

ic S

train

Axial Strain

CPBS45/2.2

CPBS45/5.5

CPBS55/2.2

CPBS55/4.5

isotropic

(b)

Figure 2-7 The observed volumetric strain versus axial strain at the end of the hydrostatic consolidation

phase for (a) uncemnted specimens, and (b) cemented specimens


0.3

0.35

0.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

10 100 1000

e

effective confining stress ( 'c)

e3

MT

CPBS45/2.2CPBS45/4.5

CPBS55/2.2

MTS45

MTS55

Sand

CPBS55/4.5

(a)

0.3

0.8

1.3

1.8

2.3

2.8

3.3

3.8

4.3

10 100 1000

e


ec3

MT

CPBS45/4.5

CPBS45/2.2

MTS45CPBS55/2.2

MTS55

Sand

CPBS55/4.5

(b)

0.7

0.75

0.8

0.85

0.9

0.95

1

1.05

1.1

10 100 1000

e


ef3

MTCPBS45/2.2

CPBS55/2.2

MTS55 CPBS45/4.5

MTS45

CPBS55/4.5

(c)

Figure 2-8 e-log ′c plots for the tested specimens in terms of (a) global void ratio, e, (b) intergranular void ratio, ec, and (c) and the interfine void ratio

(ef)


-150

-100

-50

0

50

100

150

200

250

300

350

400

0 100 200 300 400 500 600 700

t =

(

1-

3)/

2,

(kP

a)

s' = (1+3)/2, (kPa)

Phase Transformation

Maximum Stress Obliquity

Temporary Instability

Hypothetical drained path

(a)

1

2 34

5

67 8

UFL

PTL

UFLPTL

TILCompression Extension

UFL PT UFL PT TIL31.0 29.2 28.0 22.7 13.1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

-22 -18 -14 -10 -6 -2 2 6 10 14 18 22 26

r u

Axial Strain (%)



(b)

1

2

3

4

5

8

7

6

-100

-50

0

50

100

150

200

250

300

350

-22 -18 -14 -10 -6 -2 2 6 10 14 18 22 26

t =

(

1-

3)/

2,

(kP

a)

Axial Strain (%)




`

(c)

1

2

3

4

5

67

8

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

-22 -18 -14 -10 -6 -2 2 6 10 14 18 22 26

Str

es

s O

bli

qu

ity,

t/s

'

Axial Strain (%)



`

(d)

1 3

24

5

6

78

Figure 2-9 Monotonic response of 100% mine tailings (MT). (a) Stress path, (b) Pore pressure ratio, ru, versus axial strain (c) Stress-strain behavior, (d)

stress obliquity versus axial strain.


-150

-100

-50

0

50

100

150

200

250

300

350

400

0 100 200 300 400 500 600 700

t =

(

1-

3)/

2,

(kP

a)

s' = (1+3)/2, (kPa)




(a)


UFL

PTL

TIL

UFL

PTL


UFL PT UFL PT TIL32.3 31.3 27.6 22.8 13.2

12

43

5

67 8

-1.5

-1

-0.5

0

0.5

1

-20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

r u

Axial Strain (%)



(b)

2

7

6

45

3

1

8

-100

-50

0

50

100

150

200

250

300

350

-20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

t =

(

1-

3)/

2,

(kP

a)

Axial Strain (%)




(c)

4

1

3

2

5

7

8

6

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

Str

es

s O

bli

qu

ity,

t/s

'

Axial Strain (%)

Phase Transformation Points

Maximum Stress Obliquity Points

(d)

7

8

3

2

4

5

1

6

Figure 2-10 Monotonic response of MTS45. (a) Stress path, (b) Pore pressure ratio, ru, versus axial strain (c) Stress-strain behaviour, (d) stress obliquity

versus axial strain.


-150

-100

-50

0

50

100

150

200

250

300

350

400

0 100 200 300 400 500 600 700

t =

(

1-

3)/

2,

(kP

a)

s' = (1+3)/2, (kPa)




(a)


UFL

PTL

UFL

PTL

TILCompression Extension

UFL PT UFL PT TIL31.6 30.6 26.1 19.3 11.3

651

2

3

4

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

r u

Axial Strain (%)



(b)

3

2

1

5

6

4

-100

-50

0

50

100

150

200

250

300

350

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22

t =

(

1-

3)/

2,

(kP

a)

Axial Strain (%)




(c)

6

54

1

2

3

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

r u

Axial Strain (%)



(b)

3

2

1

5

6

4

Figure 2-11 Monotonic response of uncemented MTS55. (a) Stress path, (b) Pore pressure ratio, ru, versus axial strain (c) Stress-strain behaviour, (d)

stress obliquity versus axial strain.


-150

-100

-50

0

50

100

150

200

250

300

350

400

0 100 200 300 400 500 600 700

t =

(

1-

3)/

2,

(kP

a)

s' = (1+3)/2, (kPa)




(a)


UFL

PTL

UFL

PTL

TIL

76

12

34

5

8Compression Extension

UFL PT UFL PT TIL32.3 31.6 23.6 17.3 10.9

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

-22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

r u

Axial Strain (%)



(b)

76

8

5

43

1

2

-100

-50

0

50

100

150

200

250

300

350

-22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

t =

(

1-

3)/

2,

(kP

a)

Axial Strain (%)




(c)

8

76

5

4

3

2

1

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

Str

es

s O

bli

qu

ity,

t/s

'

Axial Strain (%)



(d)

6

7

8

4

53

2

1

Figure 2-12 Monotonic response of CPBS45/2.2 cured for 4 hours. (a) Stress path, (b) Pore pressure ratio, ru, versus axial strain (c) Stress-strain

behavior, (d) stress obliquity versus axial strain.


-150

-100

-50

0

50

100

150

200

250

300

350

400

0 100 200 300 400 500 600 700

t =

(

1-

3)/

2,

(kP

a)

s' = (1+3)/2, (kPa)




(a)


UFL

PTL

UFL

PTL

TIL


UFL PT UFL PT TIL

33.7 33.2 24.5 19.9 11.4

65

4

3

21

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

r u

Axial Strain (%)



(b)

56

4

32

1

-100

-50

0

50

100

150

200

250

300

350

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

t =

(

1-

3)/

2,

(kP

a)

Axial Strain (%)




(c)

6

54

3

2

1

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

Str

es

s O

bli

qu

ity,

t/s

'

Axial Strain (%)



(d)

5

64

2 3

1




-150

-100

-50

0

50

100

150

200

250

300

350

400

0 100 200 300 400 500 600 700

t =

(

1-

3)/

2,

(kP

a)

s' = (1+3)/2, (kPa)




(a)

Hypothetical drained path3

5 6

1

2

4

UFL

PTL

UFLPTL

TIL


UFL PT UFL PT TIL32.8 32.3 24.0 19.5 11.8

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

r u

Axial Strain (%)



(b)

6

5

4

2

3

1

-100

-50

0

50

100

150

200

250

300

350

-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

t =

(

1-

3)/

2,

(kP

a)

Axial Strain (%)




(c)

2

1

45

6

3

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

Str

es

s O

bli

qu

ity,

t/s

'

Axial Strain (%)



(d)

6

5

4

32

1




-150

-100

-50

0

50

100

150

200

250

300

350

0 100 200 300 400 500 600 700

t =

(

' 1-

' 3)/

2,

(kP

a)

s' = ('1+'3)/2, (kPa)




(a)


3

UFLPTL

UFL

PTL

TIL

45

6

1

2


UFL PT UFL PT TIL32.8 31.7 24.1 18.4 10.8

-0.2

0

0.2

0.4

0.6

0.8

1

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

r u

Axial Strain (%)



(b)

5

4

6

3

12

-100

-50

0

50

100

150

200

250

300

350

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

t =

(

' 1-

' 3)/

2,

(kP

a)

Axial Strain (%)




(c)

45

6

3

2

1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

Str

es

s O

bli

qu

ity,

t/s

'

Axial Strain (%)



(d)

4

5 6

3 2

1




y = 0.615xR² = 0.9698

y = 0.6324xR² = 0.9835

0

20

40

60

80

100

120

140

160

180

200

0 50 100 150 200 250 300 350

t =

(

' 1-

' 3)/

2, (

kP

a)

s' = ('1+'3)/2, (kPa)

MTS55_2%/min

MTS55_1%/min

MTS55_0.1%/min

MTS55_0.02%/min



(a)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

-2 0 2 4 6 8 10 12 14 16 18 20 22

r u

Axial Strain (%)

MTS55_2%/min

MTS55_1%/min

MTS55_0.1%/min

MTS55_0.02%/min



(b)

0

20

40

60

80

100

120

140

160

180

200

-2 0 2 4 6 8 10 12 14 16 18 20 22

t =

(

'1-

'3)/

2, (

kP

a)

Axial Strain (%)

MTS55_2%/min

MTS55_1%/min

MTS55_0.1%/min

MTS55_0.02%/min



(c)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

-2 0 2 4 6 8 10 12 14 16 18 20 22

Str

es

s O

bliq

uit

y, t

/s'

Axial Strain (%)

MTS55_2%/min

MTS55_1%/min

MTS55_0.1%/min

MTS55_0.02%/min



(d)

Figure 2-16 Monotonic response of MTS55 at 200 kPa effective confining pressure under different strain rates. (a) Stress path, (b) Pore pressure ratio,

ru, versus axial strain (c) Stress-strain behaviour, (d) stress obliquity versus axial strain.


y = 0.5977xR² = 0.5637

y = 0.6225xR² = 0.9885

0

50

100

150

200

250

300

0 100 200 300 400 500

t =

(

' 1-

' 3)/

2, (

kP

a)

s' = ('1+'3)/2, (kPa)

MT_2%/min

MT_0.05%/min

MT_0.02%/min




(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

-2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

r u

Axial Strain (%)

MT_2%/min

MT_0.05%/min

MT_0.02%/min



(b)

0

50

100

150

200

250

300

-2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

t =

(

'1-

'3)/

2, (

kP

a)

Axial Strain (%)

MT_2%/min

MT_0.05%/min

MT_0.02%/min



(c)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

-2 0 2 4 6 8 10 12 14 16 18 20 22 24 26

Str

es

s O

bliq

uit

y, t/

s'

Axial Strain (%)

MT_2%/min

MT_0.05%/min

MT_0.02%/min



(d)

Figure 2-17 Monotonic response of MT at 200 kPa effective confining pressure under different strain rates. (a) Stress path, (b) Pore pressure ratio, ru,

versus axial strain (c) Stress-strain behaviour, (d) stress obliquity versus axial strain.


y = 0.591x

R2 = 0.9998

0

50

100

150

200

250

300

350

400

0 100 200 300 400 500 600 700

s', (s1+s3)/2 (kPa)

t, (

s1- s

3)/

2 (

kP

a)

e=0.670

e=0.640 e=0.600

-7

-6

-5

-4

-3

-2

-1

0

0 5 10 15 20 25

Axial Strain (%)

Vo

lum

etr

ic s

tra

in (

%)

PTP

MSOP

e=0.600

'c=200kPa

e=0.640

'c=100kPa

e=0.670

'c=50kPa

0

50

100

150

200

250

300

350

0 5 10 15 20 25

Axial Strain (%)

t, (

s1- s

3)/

2 (

kP

a)

PTP

MSOP

e=0.600

'c=200kPa

e=0.640

'c=100kPa

e=0.670

'c=50kPa

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 5 10 15 20 25

Axial Strain (%)

t/s'

PTP

MSOP

e=0.600

'c=200kPa

e=0.640

'c=100kPa

e=0.670

'c=50kPa

Figure 2-18 Drained response of MT in compression. (a) Stress path, (b) Volumetric strain versus axial strain (c) Stress-strain behavior, (d) stress

obliquity versus axial strain.

(a)

(b)

(c) (d)


y = 0.6096x

R2 = 1

y = 0.5991x

R2 = 0.9999

0

50

100

150

200

250

300

350

400

450

0 100 200 300 400 500 600 700

s', (1+3)/2 (kPa)

t, (

1-

3)/

2 (

kP

a)

PTP

MSOP

e=0.360

e=0.355

e=0.350 e=0.32

-2.5

-2

-1.5

-1

-0.5

0

0 5 10 15 20 25

Axial Strain (%)

Vo

lum

etr

ic s

tra

in (

%)

PTP

MSOP

e=0.320

'c=200kPa

e=0.350

'c=100kPa

e=0.355

'c=50kPa

e=0.360

'c=25kPa

0

50

100

150

200

250

300

350

0 5 10 15 20 25

Axial Strain (%)

t, (

1-

3)/

2 (

kP

a)

PTP

MSOPe=0.320

'c=200kPa

e=0.350

'c=100kPa

e=0.355

'c=50kPa

e=0.360

'c=25kPa

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 5 10 15 20 25

Axial Strain (%)

t/s'

PTP

MSOP

e=0.320

'c=200kPa

e=0.350

'c=100kPa

e=0.355

'c=50kPa

e=0.360

'c=25kPa

Figure 2-19 Drained response of MTS55 in compression. (a) Stress path, (b) Volumetric strain versus axial strain (c) Stress-strain behavior, (d) stress

obliquity versus axial strain.

(a) (b)

(c) (d)


0.25

0.3

0.35

0.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

10 100 1000

Vo

id r

atio

, e

mean effective stress, p'

State lines at MSO

MT

MTS45

MTS55

CPBS45/4.5

CPBS45/2.2

CPBS55/4.5

CPBS55/2.2

Sand

MT Drained

MTS55 Drained

MT_low strain rates

MTS55_low strain rates

Log. (MT)

Log. (MTS45)

Log. (MTS55)

Log. (CPBS45/4.5)

Log. (CPBS45/2.2)

Log. (CPBS55/4.5)

Log. (CPBS55/2.2)

Log. (Sand)

(a)

0.25

0.3

0.35

0.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

10 100 1000

Vo

id r

atio

, e


State lines at MSO

MT

MTS45

MTS55

CPBS45/4.5

CPBS45/2.2

CPBS55/4.5

CPBS55/2.2

Sand

Log. (MT)

Log. (MTS45)

Log. (MTS55)

Log. (CPBS45/4.5)

Log. (CPBS45/2.2)

Log. (CPBS55/4.5)

Log. (CPBS55/2.2)

Log. (Sand)

(b)

0.25

0.3

0.35

0.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

10 100 1000

Vo

id r

atio

, e


State lines at PT

MT

MTS45

MTS55

CPBS45/4.5

CPBS45/2.2

CPBS55/4.5

CPBS55/2.2

Sand

MT Drained (MSOP)

MTS55 Drained

MT_low strain rates

MTS55_low strain rates

Log. (MT)

Log. (MTS45)

Log. (MTS55)

Log. (CPBS45/4.5)

Log. (CPBS45/2.2)

Log. (CPBS55/4.5)

Log. (CPBS55/2.2)

Log. (Sand)

(c)

0.25

0.3

0.35

0.4

0.45

0.5

0.55

0.6

0.65

0.7

0.75

10 100 1000

Vo

id r

atio

, e


State lines at PT

MT

MTS45

MTS55

CPBS45/4.5

CPBS45/2.2

CPBS55/4.5

CPBS55/2.2

Sand

Log. (MT)

Log. (MTS45)

Log. (MTS55)

Log. (CPBS45/4.5)

Log. (CPBS45/2.2)

Log. (CPBS55/4.5)

Log. (CPBS55/2.2)

Log. (Sand)

(d)

Figure 2-20 e-log p' plots showing (a) UFL in compression, (b) UFL in extension, (c) PTL in compression, and (b) PTL in extension for the tested

materials.


Figure 2-21 SEM images of (a) MT tested in this study and (b) gold tailings tested by Saebimoghaddam

(2010).

(b) SE Image of Williams MT

(a) SE Image of MT, Kidd mine


18.0

20.0

22.0

24.0

26.0

28.0

30.0

32.0

34.0

36.0

25 35 45 55 65

U

FL

Coarse grains content

Cemented_EXT

Uncemented_COMP

Uncemented_EXT

Cemented_COMP

MT

MTS

45

MTS

55

CP

BS5

5/2

.2

CP

BS5

5/4

.5

CP

BS4

5/2

.2

CP

BS4

5/4

.5

Figure 2-22 Summary of the state parameters ´UFL indicating the changes in strength in response to the

addition of sand and binder.


Chapter 3

Mechanical Response of Early Age Cemented Paste

Backfills with Sand. Part II: Cyclic Shear Response

Abstract

Investigating the cyclic response of cemented paste backfills (CPB) at the very early curing age

is attracting the attention of many researchers. Several studies proved that the addition of binder

enhances the cyclic resistance of CPB even before the initial setting time of the mixture. The

present research investigates the cyclic response when CPB is mixed with sand (CPBS) as

practiced by some mines to achieve better strength and flow characteristics. A series of

undrained cyclic triaxial tests was performed to investigate the combined effect of adding sand

and binder on the cyclic response of pastefill. Besides, the separate effect of adding either sand

or binder was investigated. The cyclic response of the tested mixtures was found to be generally

controlled by the fine matrix. Sand particles in all cemented and uncemented mixtures remained

isolated within the fine matrix and did not mechanically participate to the force chain.

Nevertheless, the presence of sand increased the effectiveness of the binder, which was reflected

on the cyclic resistance of CPBS. Cemented specimens exhibited higher anisotropy than their

uncemented counterparts, which in general agrees with the monotonic testing results.


3.1 Introduction

In addition to natural seismic activities, cemented paste backfills (CPB) can be subjected to

dynamic loading resulting from routine blasting and mining induced seismic events such as rock

bursts and local ground movements. After Clough et al. (1989), it has been taken as common

practice to consider CPB resistant to liquefaction when the unconfined compressive strength

(UCS) of 100 kPa is reached. However, the early age period, before binder contributes to the

strength of CPB, is the most critical age and is worthy of serious investigation. Saebimoghaddam

(2010) suggested that the dynamic response of CPB systems to far-field seismic events (rock

bursts and production blasts) can be investigated using the conventional geotechnical earthquake

engineering framework. Within this framework, several studies investigated the cyclic

liquefaction potential of CPB at various ages. A limited number of liquefaction studies were

conducted on early age CPB, all of which tested silt-sized cemented mine tailings. This research

work is geared towards investigating liquefaction potential when CPB is mixed with sand

(CPBS), especially when the active hydration process is delayed under the effect of

supplementary cementing materials. The following sections review the undrained behavior of

silt-sized CPB under cyclic loads. Next, the literature on the behavior of sand-silt mixtures and

the effect of varying the fraction of non-plastic fines (NPF) is discussed.

3.2 Background

3.2.1 Cyclic response of cemented paste backfills

The dynamic response of cemented, round, clean sand was investigated by Clough et al. (1989),

who suggested that when the UCS of cemented sand reaches 100 kPa, the material can be

considered resistant to liquefaction under cyclic loading conditions. Since then, pastefill

designers took it as common practice or a “rule of thumb” to consider CPB as not liquefiable

when reaching a UCS of 100 kPa. The experimental study conducted by Clough et al. (1989) was

directed towards those sands that are cemented at points of contact between sand grains. Such

materials are termed “contact bound” materials and were found to exhibited flow liquefaction

type of failure even at high UCS values. In the case of CPB and CPBS, the void volume is filled

with fine grains and is cemented. Therefore, their behavior is expected to be different from the

behavior observed by Clough et al. (1989). Being aware of the drastic difference between round,


clean sand tested by Clough et al. (1989) and silt-size mine tailings constituting the major

fraction of CPB, it is clear that the cyclic resistance of CPB is worthy of further investigation.

Aref (1989) performed in-situ CPT and laboratory tests on CPB after six days of curing and

showed that material is unsaturated, highly frictional and dilative, with increasing resistance as

binder content increases. Also, Been et al. (2002) performed in-situ CPT and laboratory tests on

CPB several months after placement and found that CPB with binder contents below 1% are

highly susceptible to liquefaction. Binder contents higher than 1% showed an increasing

resistance to liquefaction. On the contrary to the rule of thumb, Broomfield (2000) found that

CPB tested at two days was resistant to liquefaction although the UCS was lower than 100 kPa.

The apparent contradiction about when CPB becomes liquefaction resistant is understandable as

each CPB tested in the aforementioned studies had a unique combination of particle size

distribution, binder type, binder content, and curing conditions. However, more concerns

surround the cyclic resistance of CPB at the very early ages before the binder adds significant

strength to CPB. le Roux (2004) is one of the limited studies that investigated the cyclic

behaviour of CPB after three hours of curing. le Roux (2004) specified 0.16 as the maximum

stable cyclic stress ratio (CSR) below which CPB (with 5% binder) becomes liquefaction

resistant. After 12 hours of curing, CPB sustained a relatively high CSR (0.30). Also,

Saebimoghaddam (2010) performed cyclic triaxial tests on CPB after four hours of curing. He

showed that by adding 3% binder to mine tailings, the number of cycles required to reach the

state of cyclic mobility increased by an order of magnitude compared to uncemented mine

tailings. After 12 hours of curing, CPB was resistant to cyclic mobility.

Pastefill designers at some mines add sand to the paste mix to improve its pipe-flow

characteristics and to achieve higher strength. However, adding sand to silt-size CPB may raise

more concerns of its liquefaction susceptibility. So far, no cases in the reviewed literature have

addressed the cyclic response of CPB-sand mixtures (denoted as CPBS herein). In this Chapter,

the role of sand in the cyclic response of the cemented mixture is investigated. Therefore, the

behavior of uncemented sand-silt mixtures presented in previous studies is reviewed in the

following paragraphs.


3.2.2 Role of non-plastic fines

The dynamic response of clean sands has been extensively studied during the past five decades.

However, many incidents have shown that silty sands liquefy under seismic loads. The mining

industry has motivated numerous studies on nonplastic silts, being the basic component of mine

tailings, which has been found susceptible to liquefaction (Dobry and Alvarez, 1967; Okusa et

al., 1980; Garga and McKay, 1984; Fourie and Papageorgiou, 2001; Crowder, 2004; le Roux,

2004; Al Tarhouni, 2008; Saebimoghaddam, 2010). Several attempts have been made to

understand the effect of introducing fine-grained materials to clean sands. The outcome of those

attempts led to contradictory conclusions. Several studies have reported that increasing the NPF

content of sand increases the cyclic resistance of the mixture (Chang et al., 1982; Dezfulian,

1984; Kuerbis et al., 1988; Amini and Qi, 2000), while other studies found that increasing the

NPF negatively impacts the cyclic resistance (Troncoso and Verdugo, 1985; Finn et al., 1994;

Vaid, 1994). Furthermore, Singh (1994) tested silty sands and Wijewickreme et al. (2010) tested

paste-rock (mixture of crushed rocks and mine tailings) and both found that the mixtures exhibit

lower resistance than either that of the host soil or the fine-grained filler.

Other studies reported that the cyclic resistance of sand initially decreases as the NPF fraction

increases until a minimum resistance is reached, after which any increase in the NPF content is

associated with increase in the cyclic resistance (Koester, 1994; Singh, 1994; Polito and Martin,

2001; and El-Mamlouk et al., 2008). This apparent contradiction in the trends reported in the

literature is ascribed to the different measures of relative density on which the cyclic resistance is

evaluated. These measures are namely the gross or global void ratio of the mixture, the sand

skeleton void ratio, and the relative density of the whole mixture. Comprehensive studies

conducted by Polito and Martin (2001) and Thevanayagam et al. (2002) evaluated the cyclic

resistance in terms of each of the relative density measures used in previous studies in an attempt

to answer the question of when each of those measures is appropriate to be used to assess the

effect of NPF on the liquefaction potential of a mixture. The following paragraphs review the

outcomes of these comprehensive evaluations.

The relation between the maximum and minimum void ratios (emax and emin) and the NPF content

has been discussed in details in Chapter 2. In summary, as NPF content increases, emax and emin

initially decrease until their lowest value is reached at a critical NPF content and as NPF content


continues to increase, emax and emin increases. This relation was the basis upon which Polito and

Martin (2001) and Thevanayagam et al. (2002) presented new conceptual frameworks to the

behavior of mixtures.

Polito and Martin (2001) attributed the transition in the behavior of sand-silt mixtures to the

transition from sand dominated soil structure to silt (or fines) dominated soil structure. They

marked this transition by the limiting silt content (LSC) which is defined as the maximum

amount of silt that can be contained in the void space while maintaining a contiguous sand

skeleton. Beyond LSC, sand grains are suspended in a silt matrix with rare or no sand grain to

sand grain contact. Nevertheless, this framework does not account for the case when sand grains

are not in active contact while they can still engage and develop inter-particle stresses. LSC can

be obtained from the following equation:

)1( max,

max,

HFss

HSsf

eG

eGLSC

[3- 1]

where emax,HS and emax,HF are the maximum void ratio values of clean sand and pure silt,

respectively, and Gss and Gsf are specific gravity values for clean sand and pure silt, respectively.

Polito and Martin (2001) reported that LSC for most sand-silt mixtures ranges between 25% and

45%.

Different approach to the conceptual framework has been introduced by Thevanayagam et al.

(2002) to incorporate the level of engagement of each matrix in the mixture based on the fines

content. In this approach a threshold value for fines content (FCth) defines the limit between two

main categories of microstructure: (A) primarily the coarse grains are in contact, and (B)

primarily the fine grains are in contact with each other. FCth is defined by equation (3-2). Under

category A, where FC<FCth, three cases (i, ii, and iii) are identified and while two cases (iv-1 and

iv-2) are indentified under B, where FC>FCth.

HF

the

eFC

max,

[3- 2]

In case (i), the fine grains are confined within the void space between coarse grains with minor

role in supporting the coarse grain skeleton. In this case, the coarse grain skeleton void ratio (ec)

is suggested to be the index of reference when investigating the liquefaction potential. In case


(ii), the fine grains are still confined within the void space but partially support the coarse grain

skeleton. In case (iii), fine grains partially separate the coarse grains. In cases (ii) and (iii),

confined fine grains partially support the coarse grains while the separating fine grains actively

participate in the force chain. Therefore, an equivalent coarse grain skeleton void ratio (ec)eq is

introduced to account for the contribution of fines in cases (ii) and (iii).

In case (iv-1), the coarse grains are fully dispersed (isolated) in the fine grain matrix. In this case,

the fine grain skeleton void ratio (ef) is suggested to be the index of reference as the behavior is

entirely governed by the fine grains. In case (iv-2), FC is lesser and the coarse grains are closer

than in case (iv-1). In this case, coarse grains play a secondary role and therefore, an equivalent

fine grain skeleton void ratio (ef)eq is introduced to account for the contribution of the coarse

grains. It is now apparent that the global void ratio “e” is not an appropriate relative density

measure to characterize the mechanical response of a soil mixture. However, the proposed FCth

approach by Thevanayagam et al. (2002) incorporated the global void ratio of the prepared

mixture (equation 3-2). On the contrary, the limiting silt content approach proposed by Polito

and Martin (2001) considers only emax values of clean sand and pure silt; therefore, LSC is a

constant value regardless of the void ratio of the mixture (El-Mamlouk et al., 2008).

A review of cyclic resistance trends evaluated in terms the different measures of relative density

is given hereinafter.

3.2.2.1 Global or gross void ratio (e)

On the basis of holding the global void ratio constant while altering the silt content, Troncoso

and Verdugo (1985) and Finn et al. (1994) reported a decrease in the cyclic resistance with

increase in the silt content up to 30%. On the other hand, Chang et al. (1982) reported an

increase in the cyclic resistance with increase in the silt content for a given void ratio. On the

same void ratio basis, Ishihara (1980) reported minor changes in the behavior with increasing silt

content. Koester (1994), Polito and Martin (2001), and El-Mamlouk et al. (2008) tested a wider

spectrum of silt contents and reported that the cyclic resistance of sand initially decreases as the

silt content increases until a minimum resistance is reached, usually around 24% to 35%, after

which any increase in the silt content is associated with an increase in the cyclic resistance.

Furthermore, when cyclic resistance was plotted against the global void ratio, no particular trend

was obtained. In conclusion, there is no unique relationship between silt content, void ratio, and


cyclic resistance for sand-silt mixtures, thus, the global void ratio is not an appropriate measure

to evaluate the cyclic resistance.

3.2.2.2 Sand skeleton void ratio (ec)

Finn et al. (1994) and Vaid (1994) reported that only minor changes in the cyclic resistance are

observed when silt content is increased from 0% to 20% while maintaining a constant sand

skeleton void ratio. On the same basis of constant ec (or the relative density of parent sand),

Koester (1994) tested samples with silt contents ranging from 5% to 60% and reported that the

cyclic resistance decreases as silt content increases until a minimum resistance is reached, after

which the cyclic resistance increases as the silt content continues to increase. Kuerbis et al.

(1988) and El-Mamlouk et al. (2008) showed an increase in cyclic resistance with the increase in

silt content for a given ec. Whereas, Polito and Martin (2001) tested two types of sand-silt

mixture, one type of which is reported to agree with such observation, and for the other no

change in cyclic resistance was reported. Polito and Martin (2001) concluded that the sand

skeleton void ratio is considered as an appropriate measure to evaluate the cyclic resistance as

long as the silt content is lower than the LSC (i.e. sand particles are still contiguous).

3.2.2.3 Relative density (Dr)

Amini and Qi (2000) reported an increase in the cyclic resistance with the increase in silt

content, from 10% to 50%, at a given relative density. Polito and Martin (2001) and El-Mamlouk

et al. (2008) concluded that for silty sands below LSC the cyclic resistance increases with the

increase in relative density. Also, for soils above LSC the cyclic resistance is controlled by the

relative density of the soil. However, lower values of the cyclic resistance are observed at similar

relative densities to those below LSC. Additionally, the increase in cyclic resistance associated

with the increase in relative density occurs at a slower rate.

It can be concluded from the above discussion that out of the density measures used to evaluate

the role of NPF in the cyclic resistance of soil mixtures, the relative density (Dr) can be

considered the most reliable measure to describe the trends whether the behavior of the mixture

is dominated by the course grain matrix or the fine grain matrix.


3.3 Materials Tested

The basic components for the mixtures tested in this testing program are similar to those in

Chapter 2. Those components are silica tailings, glacial sand, and BFS binder. The percentage

finer than 20 microns in the mine tailings are around 40%. The liquid limit (LL) of the tailings is

23%, the plastic limit (PL) is 18%, and the plasticity index (PI) is 5%. Therefore, the existing

fines are considered non-plastic or “sand-like” fines according to Boulanger and Idriss (2004).

The specific gravity of the tailings is 2.79. X-Ray Fluorescence Spectroscopy (XRF) was

performed to determine the chemical composition of the tailings which indicates that the material

is rich in silica while alumina can also be considered as a major component. Minor amounts of

iron, magnesium and equivalents of calcium and sulphur were also found in their oxide forms.

The present low sulphur content is not considered troublesome to the hardening process of CPB

as emphasized in many previous research performed on different types of binder (e.g.,

Benzaazoua et al., 2002). The tailings had previously been stored for a long time on surface and

might have been subjected to oxidation and alteration. However, all the tests were done on

tailings in its current state as used in the CPB mixture.

The used glacial sand is naturally silica rich, and is considered inert towards any chemical

reaction that takes place during binder hydration. The specific gravity of sand is 2.70. The mine

used a pre-blended binder provided by Lafarge and comprised of 90% blast furnace slag (BFS)

and 10% Type 10 Portland cement (PC). The BFS fraction in the binder is significantly higher

than that used in construction industry. XRF results show that the binder primarily contains

calcium and silica with minor amounts of magnesium and alumina. The binder is mixed with

tailings and sand at 4.5% of the dry weight of solids. Water is added to reach a water content of

28%.

Specimens tested in the present thesis were composed of uncemented 100% mine tailings (MT),

an uncemented mixture of 45% MT and 55% sand (MTS55). Cemented specimens of the same

mixtures were also tested at a binder content of 4.5%. Cemented mine tailings and cemented

MT-Sand mixtures are denoted as CPB/4.5 and CPBS55/4.5, respectively.

Thottarath (2010) investigated the setting characteristics of the cemented mixes by measuring the

electric conductivity (EC) evolution in both CPB/4.5 and CPBS55/4.5. The EC response in

CPB/4.5 and CPBS55/4.5, which peaks with the occurrence of the initial setting of binders,


reaches the peak at around 1200 mins (≈25 hours) and 700 mins (≈12 hours), respectively.

Therefore, triaxial testing took place at four hours assuming that the binder has not yet reached

the initial set and has not significantly contributed to the strength of the mixture.

3.4 Sample Preparation

To prepare cemented specimens for a triaxial test, standard preparation methods such as moist

placement, dry deposition, and water sedimentation are not suitable as the components had to be

mixed with cement and water in advance of placement. In this testing program, specimens of 70

mm diameter are prepared according to the method suggested by Crowder (2004), le Roux

(2004) and Saebimoghaddam (2010) for creating triaxial CPB specimens. The procedure is

detailed in Chapter 2. Binder, sand, and MT are mixed continuously for 10 minutes at the desired

moisture content, 28%. The material was cast in a split mould then underwent a one dimensional

consolidation procedure under a 5 kg dead weight (equivalent to about 12.5 kPa) that lasts for

one hour. Saturation process then followed according to ASTM D4767-04. A minimum B-Value

of 0.96 was reached at back pressures ranging between 200 kPa and 250 kPa. The specimen is

then consolidated for one hour under an isotropic pressure to reach an effective confining

pressure of 100 kPa. At the end of the hydrostatic consolidation stage, volumetric strain (v) and

axial strain (a) were measured to ensure the specimens responded isotropically to the applied

hydrostatic stress. The values of v/a did not exceed 3.6 (NB: 3 is the value for perfect

hydrostatic loading) which is acceptable (Khalili et al., 2010). The whole process takes four

hours from mixing to loading. This preparation method resulted in consistent pre-shearing

relative densities where the relative densities of all mixtures ranged between 73% and 79%. The

pre-shearing characteristics of the tested specimens and the key test parameters are summarized

in Table 3-1.

To obtain uniform laboratory specimens that result in a representative behaviour of the soil

skeleton, oversized particles in the glacial sand were removed prior to mixing with other

constituents. This process is known as scalping. The mechanical and hydrological behaviour of

the specimen is not affected by the scalping process as long as the removed particles float in the

finer matrix (Khalili et al., 2010). For the glacial sand used in this study, the cut-off particle size


was 6.3 mm. The diameter to maximum grain size ratio after scalping was around 11 which

fulfill the requirements of ASTM D4767-04. Therefore, behaviour of the specimen is not

expected to be affected by the removal of the oversized particles.

3.5 Testing Program

A series of cyclic triaxial tests were performed using a servo-hydraulic triaxial machine designed

by Geotechnical Consulting and Testing Systems (GCTS). An internal (submersible) load cell

was used to overcome the unfavourable effect of friction developing between the loading ram

and the sealing and guiding connection. The scope of this experimental program is to assess how

adding sand and/or binder to mine tailings may affect the cyclic resistance. Cyclic triaxial tests

were conducted on the uncemented specimens of MT and MTS55 and cemented specimens of

CPB/4.5 and CPBS55/4.5. Cemented specimens were tested after four hours of curing and some

specimens were tested after seven hours of curing which is still earlier than the initial set time as

mentioned in a previous section.

All cyclic liquefaction tests were stress controlled in which the degree of loading is described by

the targeted cyclic stress ratio (CSR). The CSR is defined as the ratio between the applied shear

stress (=d/2) to the pre-shearing effective confining stress ('c) (i.e. CSR=d/(2'c)). In this

thesis, specimens are tested only under effective confining stress of 100 kPa. At a given density,

the ratio between the cyclic resistance at any confining stress to the cyclic resistance at 100 kPa

is defined by K. This ratio is determined experimentally to indicate the sensitivity of the cyclic

resistance of a given material to varying the pre-shearing effective confining stress. Vaid and

Thomas (1995), Ishihara (1996), Vaid et al. (2001), and Wijewickreme et al. (2010) reported that

K is almost unity for sands and paste-rocks tested under effective confining stresses below 500

kPa. Moreover, Saebimoghaddam (2010) did not report significant changes in the cyclic

resistance of uncemented silt-size tailings and CPB in response to changing the pre-shearing

effective confining stress from 50 kPa to 100 kPa. In situ stresses were measured at the stopes

where the same CPBS55/4.5 was used (Thompson et al., 2009). The maximum pressure reached

after a week of filling was within the range of 500 kPa. Therefore, the cyclic resistance under

effective confining pressures within the range measured in situ is expected to be close to that

under 100 kPa.


Uniform stress cycles were applied at a frequency of 0.10 Hz in this testing program to achieve

the best control of the loading system at 'c =100 kPa. This frequency falls at the lower bound of

the typical frequency range applied for triaxial testing (e.g., Hyde et al., 2006; and

Saebimoghaddam, 2010). The effect of frequency on the number of cycles to liquefaction is

insignificant in stress controlled cyclic loading (e.g., Riemer et al., 1994; and Xenaki and

Athanasopoulos, 2003).

3.6 Test Results

3.6.1 Failure criteria

Liquefaction and cyclic mobility is conceived in this study according to the terms defined by

Vaid and Chern (1985). Liquefaction is essentially used in the literature to describe the

accumulation of pore water pressure (PWP) and the associated strains under undrained loading

conditions in the laboratory or field. The effective stress gradually diminishes and the shear

strength consequently decreases. The complete loss of shear strength leads to a flow failure

which is usually exhibited by loose soils. Flow failure is also referred to as flow or initial

liquefaction. On the other hand, denser soils may exhibit partial loss of strength leading to

progression in shear strains. These limited deformations under cyclic loading are commonly

described as cyclic mobility. In the present study, as long as flow liquefaction is not reached, the

criterion to identify the state of cyclic mobility is when the specimen induces 5% double

amplitude (DA) axial strain. As well, the cyclic resistance ratio (CRR) is taken as the magnitude

of CSR required to produce 5% DA axial strain in 20 uniform load cycles (Ishihara, 1996). The

testing machine was very sensitive to the changes in the material composition and CSR.

Therefore, new gain (control) parameters had to be defined for each test. The test was only

considered successful if the resulting stress cycles were uniform.

3.6.2 Cyclic response of uncemented MT specimens.

Specimens composed of mine tailings only (MT) were tested under 'c=100 kPa and CSR

ranging from 0.10 to 0.20. The results of four successful tests at different CSR values are

summarized in Table 3-1. The pre-shearing void ratios were around 0.645±0.005 at an overall

relative density around 75%. Fine particles composed 67% of the MT. Figure 3-1 and Figure 3-2

show the behavior of MT specimens at the lowest and highest CSR values. The response of MT


at the other CSR values was essentially the same (see details in Table 3-1 and Appendix A).

Generally, all specimens initially exhibited excess PWP (u) that progressively developed with

the number of loading cycles (see Figure 3-1c and Figure 3-2c). As PWP accumulates, the stress

paths in Figure 3-1a and Figure 3-2a move from the initial effective stress (100 kPa) towards the

origin. The failure envelope obtained from the monotonic testing in compression and extension,

presented in Chapter 2, are superposed on the stress path plot in Figure 3-1a and Figure 3-2a.

The cyclic stress path is well encompassed by the failure line (FL) in extension while in

compression some sort of residual strength was exhibited. At the lower CSRs the changes in

axial strain were not significant and considerable accumulation of axial strains only happened a

few cycles prior to reaching cyclic mobility state (at 5% DA axial strain) (see Figure 3-1d and

Figure 3-2d). It is also clear that more strains developed under the extension half of the cycle

than in the compression half. The development in PWP is plotted in terms of the excess pore

pressure ratio (ru), where ru=u/'c. Failure was reached before the effective stress reached zero

as ru approached unity. The value of ru at failure ranged between 0.91 and 0.95 except for

specimen MT-4, tested at CSR = 0.2, the cyclic mobility was reached at ru of 0.85. None of the

tested specimens exhibited flow liquefaction type of failure which is consistent with the dilative

behavior shown in Chapter 2. The cyclic resistance chart is obtained by plotting the cyclic stress

ratio and the number of cycles required to reach 5% DA axial strain (N). As shown in Figure 3-3,

N decreases as CSR increases. The cyclic resistance ratio (CRR) for MT is 0.11.

3.6.3 Cyclic response after adding sand to MT.

Specimens with 55% sand added to mine tailings (MTS55) were tested under 'c=100 kPa and

CSR ranging from 0.09 to 0.18. The results of four successful tests at different CSR values are

summarized in Table 3-1. The pre-shearing void ratios were around 0.355±0.005 at an overall

relative density around 79%. Fine particles composed 38% of the uncemented MT-sand mixture.

MTS55 specimens exhibited essentially the same behavior as that of MT (see Figure 3-4). Also,

the stress path in compression crossed the FL while was well encompassed in extension. More

strains developed in extension than in compression and failure was reached before the effective

stress reached zero as ru approached unity. The value of ru at failure ranged between 0.95 and

0.97 except for specimen MTS55-4, tested at CSR = 0.2, the cyclic mobility was reached at ru of

0.87. None of the tested specimens exhibited flow liquefaction type of failure which is consistent

with the monotonic behavior presented in Chapter 2. The cyclic resistance ratio (CRR) for


MTS55 is 0.10 as shown in Figure 3-3. Although this indicates only a slight decrease (9%) in

CRR after adding sand, the difference in CRR is larger than the scatter of the results of each

mixture and is considered statistically significant.

3.6.4 Cyclic response after adding 4.5% binder to MT.

Specimens with 4.5% binder added to mine tailings (CPB/4.5) were tested after four hours of

curing under 'c=100 kPa and CSR ranging from 0.09 to 0.2. The results of four successful tests

at different CSR values are summarized in Table 3-1. The pre-shearing void ratios were around

0.740±0.005 at an overall relative density around 73%. Fine particles composed 71.5% of the

cemented MT mixture. The behavior of CPB/4.5 specimens did not show significant difference

from that of the uncemented MT specimens. More strains developed in extension than in

compression and failure was reached before the effective stress reached zero as ru approached

unity. The value of ru at failure was consistent around 0.96 except for specimen CPB/4.5-4,

tested at CSR = 0.2, the cyclic mobility was reached at ru of 0.64.

Specimen CPB/4.5-1, tested under CSR=0.09, showed a significantly higher resistance to

liquefaction than its uncemented counterpart as shown in Figure 3-5. Specimen CPB/4.5-1

reached failure at 261 cycles compared to 37 cycles for the uncemented MT specimen at the

same CSR. However, as shown in Figure 3-6, at the higher CSR cemented specimens exhibited a

slight increase in cyclic resistance compared to their uncemented counterparts. From Figure 3-6,

CRR of 0.125 was obtained for CPB/4.5, which is 14% higher than that of MT. The result of

specimen CPB/4.5-1 was plotted separately as it failed after more than one hour of testing.

Within this period, properties of the pore fluid are assumed to have changed and consequently

this result was excluded when fitting the line to the results of the other CPB/4.5 specimens.

3.6.5 Cyclic response after adding 4.5% binder to the MT-sand mixture

Specimens with 4.5% binder added to MT-sand mixture (CPBS55/4.5) were tested after four

hours of curing under 'c=100 kPa and CSR ranging from 0.11 to 0.25. The results of five

successful tests at different CSR values are summarized in Table 3-1. The pre-shearing void

ratios were around 0.420±0.005 at an overall relative density around 74%. Fine particles

composed 42.7% of the cemented MT-sand mixture. The behavior of CPBS55/4.5 specimens did

not show difference from that of the other cemented and uncemented specimens (see Figure 3-7).


This time, the stress path in compression as well as in extension crossed the FL as shown in

Figure 3-7a. More strains developed in extension than in compression and failure was reached

before the effective stress reached zero as ru approached unity. The value of ru at failure was

consistent around 0.96 except for specimen CPBS55/4.5-4, tested at CSR = 0.25, where the

cyclic mobility was reached at ru of 0.85. From Figure 3-6, it can be observed that adding binder

in the case of MT-sand mixtures significantly enhanced the cyclic resistance compared to adding

binder to MT only. At the same CSR, the number of cycles required to reach 5% DA axial strain

for CPBS55/4.5 is about four to five times that of their uncemented counterparts. In addition,

CRR for CPBS55/4.5 is 0.16 compared to 0.10 for their uncemented counterparts (MTS55).

Moreover, the specimen tested at CSR = 0.11 did not fail and no tangible strains have

accumulated until the test was terminated after 326 cycles (see Appendix A).

3.6.6 Effect of curing age on cyclic resistance

Specimens CPB/4.5-5 and CPBS55/4.5-6 were tested after seven hours of curing under CSR of

0.20 and 0.25, respectively, and the results are plotted in Figure 3-6. Both CPB/4.5-5 and

CPBS55/4.5-6 exhibited the same type of behavior as their cemented and uncemented

counterparts. Specimen CPB/4.5-5 reached failure at 4 cycles which is twice the number of

cycles needed to reach failure for their counterparts tested after four hours of curing under the

same CSR. The value of ru at failure was 0.90 for specimen CPB/4.5-5. On the other hand,

specimen CPBS55/4.5-6 reached failure at 39 cycles which is about 20 times the number of

cycles taken to reach failure for their counterparts tested after four hours of curing under the

same CSR. The value of ru at failure was 0.95 for specimen CPBS55/4.5-6. le Roux (2004) and

Saebimoghaddam (2010) also reported an increase in the cyclic resistance when testing CPB

specimens after 12 hours of curing relative to the cyclic resistance at four hours.

Saebimoghaddam (2010) reported that after 12 hours initial setting had already been observed

for the used CPB specimens. In the present research, however, seven hours is still earlier than the

initial set time.


3.7 Discussion

3.7.1 Cyclic response

The stress path as specimens approached cyclic mobility crossed the failure lines (FL)

determined from monotonic testing. This discrepancy was observed in compression for the

uncemented specimens, while it is observed in both compression and extension for the cemented

specimens of CPBS. Similar observations are reported by Saebimoghaddam (2010) for silt-size

uncemented tailings and CPB. The results presented by Wijewickreme et al. (2010) showed

similar discrepancy for paste-rock samples and considered that this discrepancy was insignificant

and that FL still bounded the cyclic stress path. Furthermore, this discrepancy was not

encountered in their results for tailings-only, rock-only, or for paste-rock cyclically loaded with

static shear bias.

Boulanger and Truman (1996) conducted cyclic triaxial tests on anisotropically consolidated

specimens of loose sand and found that the mobilized friction angle is higher than the critical

state friction angle in compression and extension. However, when those specimens were

reloaded at an initial static shear stress (shear bias) of 40 kPa, the cyclic stress path complied

well with critical state lines. Boukpeti and Drescher (2000) and Boukpeti et al. (2002) reported

that compression and extension stress paths of many soils exhibiting strain hardening behaviors

(including soils going through temporary liquefaction) may cross the steady state line (SSL) and

then move back to terminate lying on the SSL. They added that this is observed for isotropically

consolidated soils and analyzed this observation as part of the hardening behavior. They

introduced a model that accounted for a hardening parameter which mathematically and

physically allows the stress path to cross the SSL. The outcome of the model matched the

experimental results published by Ishihara (1980) fairly well.

The monotonic results presented in Chapter 2 showed that both cemented and uncemented

specimens exhibited a strain hardening behavior and the discrepancies between the behavior in

compression and extension was ascribed to the intrinsic anisotropy of the fabric. This anisotropy

increased by adding sand to MT and increased further by adding binder to the MT-sand mixture.

This is consistent with the results presented herein, as the stress path of the cemented specimens

progressed further beyond the failure lines in compression and extension than in the case of


uncemented specimens. In summary, both anisotropy and hardening justifications for the stress

path-FL inconsistency are valid for the results of this research work.

3.7.2 Strain anisotropy

Cemented and uncemented specimens in this thesis developed more strains in the extension half

cycle than in the compression half. This observation agrees with results of the gap-graded paste-

rock and the rock-only specimens tested by Wijewickreme et al. (2010) and the results of the

CPB tested by Saebimoghaddam (2010). In order to quantify this observation in the present

work, the ratio between the magnitudes of the negative (extension) axial strain to the magnitude

of the positive (compression) axial strain at the cycle which 5% DA axial strain is reached is to

be termed the strain anisotropy ratio (SAR). The values of SAR for each test are listed in Table

3-1. Specimens of MT and CPB/4.5 showed a decrease in SAR with the increase in CSR. The

CSR dependence was not observed for the specimens containing 55% sand (MTS55 and

CPBS55/4.5). However, SAR substantially increased for CPB/4.5 specimens compared to their

uncemented counterparts (MT). Similarly, CPBS55/4.5 specimens had higher SAR than their

uncemented counterparts (MTS55). Nevertheless, it was found lower than that of the CPB/4.5

specimens. This increasing anisotropy in the case of binder could be attributed to the change in

the configuration of particles in response to the addition of binder. These results agree with the

monotonic results presented in the counterpart study in Chapter 2 and the cyclic results presented

by Saebimoghaddam (2010). In summary, adding binder to the paste increases the anisotropy in

the cyclic behavior of CPB (or CPBS) compared to the uncemented case. More importantly, the

observed anisotropy occurs only at large strains by approaching cyclic mobility, while the small

strains generated at early load cycles did not exhibit such anisotropy. This corroborates the

analysis made in Chapter 2 that all paste mixes exhibited anisotropy only at large strains.

3.7.3 Evaluation of the cyclic resistance

The cyclic mobility type of response for MT specimens agrees with the results reported for

uncemented mine tailings by Wijewickreme et al. (2005) and Saebimoghaddam (2010), and for

natural silty soils by Bray and Sancio (2006) and Sanin and Wijewickreme (2006). The cyclic

resistance of MT specimens is significantly lower than that of the uncemented tailings tested by

Ishihara (1980), Crowder (2004), and Saebimoghaddam (2010) as shown Figure 3-8. Ishihara

(1996) reported that smaller test specimens tend to exhibit higher resistance under cyclic loading.


This can be the reason for the lower cyclic resistance of the specimens tested in the present work

as they have larger dimensions than in the studies compared in Figure 3-8.

According to Polito and Martin (2001) and El-Mamlouk et al. (2008), at a given relative density

the cyclic resistance of sand-silt mixtures remains constant in response to increasing FC up to

LSC (i.e. sand controlled behavior). The cyclic resistance drastically drops by getting higher than

LSC (i.e. switching to silt controlled behavior) to reach a constant value with increasing FC.

Relative densities of all specimens tested in this research work range between 73% and 79%. The

variations in relative densities from one mixture to the other fall within a very narrow spectrum

and can be considered insignificant to the behavior. To affirm this assumption, the cyclic

resistance is plotted against the relative density in Figure 3-9a. Only a minute difference in CRR

is observed corresponding to the difference between the relative densities of MT and MTS55.

Although CPB/4.5 and CPBS55/4.5 specimens fall within the narrow relative density spectrum,

the higher CRR exhibited is due to the action of the binder. Figure 3-9b shows that MT and

MTS55 specimens complies with the findings of Polito and Martin (2001) and El-Mamlouk et al.

(2008) as CRR variation with FC is insignificant. Again, CPB/4.5 and CPBS55/4.5 specimens

exhibited higher resistance to liquefaction in Figure 3-9b despite having FC close to MT and

MTS55, respectively.

The LSC for the MT-sand mixtures is calculated as 22% which is slightly lower than the

common LSC values of silty sands reported in the literature (24% to 35%). Since all specimens

had higher FC than LSC, it can be postulated that the fine matrix was prevailing and dominated

the behavior even in the cemented cases. Also, all mixtures can be classified as case (iv-1)

according to Thevanayagam et al. (2002), where sand grains are dispersed and isolated in the

fine grained matrix and do not contribute to the behavior of the mixture. However, the

combination of adding sand and binder to MT significantly increased the cyclic resistance. This

can be attributed to the reduced specific surface area and contact points after adding sand which

increased the effectiveness of the binder.

In summary, it is important to point out that through the followed conceptual framework and the

observations of this work, sand particles in all cemented and uncemented mixtures remained

isolated within the fine matrix and did not mechanically participate to the force chain.

Nevertheless, it can be said that the presence of sand increased the effectiveness of the binder.


3.8 Conclusions

A series of cyclic triaxial tests was performed to investigate the combined effect of adding sand

and binder on the cyclic response of pastefill. Also, the separate effect of adding either sand or

binder was investigated. The outcome of this experimental study showed that uncemented and

cemented mine tailings and MT-sand mixtures exhibited a cyclic mobility type of response. For

materials exhibiting a cyclic mobility type of response, the possibility of a catastrophic failure

due to flow at small triggering strains is very low. In particular, it is unlikely that the material

tested in this research work will completely lose its strength at low strains, which would

otherwise lead to dramatic pressure increases and a breach of the barricade with subsequent

outflow into the mine infrastructure. Moreover, the possibility of developing large strains, at

which cyclic mobility occurs, in a stope is also very low due to the displacement constraints

imposed by the surrounding rock mass. Furthermore, even if such large strains are developed and

caused a barricade failure, only limited movements of the pastefill mass into the undercut are

expected rather than a catastrophic flow.

Analyzing the induced strains under cyclic loading, all combinations showed weaker response in

extension than in compression, thus, indicating anisotropy. The strain anisotropy ratio was

introduced to quantify the anisotropic behavior for the sake of comparing the different pastefill

compositions. The observed anisotropy, when failure (cyclic mobility) is approached, increased

after adding binder to MT or MTS55. In this regard, the results generally agree with the

monotonic results presented in Chapter 2. However, the cyclic stress paths of uncemented

pastefill specimens (MT and MTS55) crossed the failure lines obtained from monotonic results

in compression. Cemented specimens (CPBS55/4.5) crossed the failure lines in compression and

extension. This discrepancy is described in the literature either as a feature of hardening

materials or as an anisotropic behavior, where both descriptions apply to the materials tested in

this study.

Altering the fines content by adding sand and/or binder did not affect the relative density of the

resulting mixture under the preparation method followed for all specimens. As a result, adding

sand to MT caused only slight (9%) reduction in the cyclic resistance. The addition of 4.5%

binder increased the cyclic resistance for specimens tested at four hours of curing. Specimens

tested at seven hours of curing showed that the cyclic resistance continues to increase as time


elapses before the initial set time of the paste. This confirms that, before the onset of the

acceleration phase, the binder affects the resistance and the behavior of the material as a result of

its cementation reactions and not only due to altering the fines content of the mixture.

The behavior was controlled by the fine matrix in MT, MTS, CPB and CPBS. Sand particles in

all cemented and uncemented mixtures remained isolated within the fine matrix and did not

mechanically participate to the force chain. Nevertheless, the presence of sand increased the

effectiveness of the binder which has been reflected on the cyclic resistance of CPBS.


References



Amini, F. and Qi, G.Z. 2000. Liquefaction testing of stratified silty sands. Journal of

Geotechnical and Geoenvironmental Engineering, 126(3): 208-217.






Corvo mine, Portugal. Institution of Mining and Metallurgy.Transactions.Section A:




1133-1144.

Boukpeti, N. and Drescher, A. 2000. Triaxial behavior of refined superior sand model.

Computers and Geotechnics, 26(1): 65-81.

Boukpeti, N., Mróz, Z., and Drescher, A. 2002. A model for static liquefaction in triaxial

compression and extension. Canadian Geotechnical Journal, 39(6): 1243-1253.

Boulanger, R.W. and Truman, S.P. 1996. Void redistribution in sand under post-earthquake

loading. Canadian Geotechnical Journal, 33(5): 829-834.




Bray, J. and Sancio, R. 2006. Assessment of the liquefaction susceptibility of fine-grained soils.

Journal of Geotechnical and Geoenvironmental Engineering, 132(9): 1165-1177.


Canada.

Chang, N., Yeh, S., and Kaufman, L. 1982. Liquefaction potential of clean and silty sands. In

Proceedings of Third International Earthquake Microzonation Conference , Seattle, USA,

Vol. 2, pp. 1017-1032.





Dezfulian, H. 1984. Effects of silt content on dynamic properties of sandy silts. In Proc., 8th

World Conf. on Earthquake Engineering, Prentice-Hall, Englewood Cliffs, N.J., pp. 63-

70.

Dobry, R. and Alvarez, L. 1967. Seismic failures of Chilean tailings dams. J. Soil Mech.and

Found.Div., ASCE, 93(6): 237-260.


El-Mamlouk, H.H., Hassan, A.M., and Hussein, A.K. 2008. Undrained cyclic behavior of

nonplastic silty sand. Journal of Engineering and Applied Science, 55(5): 379-399.

Finn, W.D.L., Ledbetter, R.H., and and Guoxi, W. 1994. Liquefaction in silty soils: Design and

analysis. Ground Failures Under Seismic Conditions, Geotechnical Special Publication,

No. 44, ASCE. pp. 51-76.

Fourie, A.B. and and Papageorgiou, G. 2001. Defining an appropriate steady state line for


Garga, V.K. and and McKay, L.D. 1984. Cyclic triaxial strength of mine tailings. Journal of

Geotechnical Engineering, 110(8): 1091-1105.

Hyde, A.F.L., Higuchi, T., and Yasuhara, K. 2006. Liquefacation, cyclic mobility, and failure of

silt. Journal of Geotechnical and Geoenvironmental Engineering, 132(6): 716-735.

Ishihara, K. 1980. Cyclic strength characteristics of tailings materials. Soils and Foundations,

20(4): 127-142.

Ishihara, K. 1996. Soil behaviour in earthquake geotechnics. Clarendon Press, Oxford.




Koester, J. 1994. Influence of fines type and content on cyclic strength. Ground Failures Under

Seismic Conditions, Geotechnical Special Publication, No. 44, ASCE. pp. 17-33.


undrained response of sand. In Hydraulic fill structures, Geotechnical. Special

Publication No. 21, ASCE. pp. 330-345.


University of Toronto, Canada.

Okusa, S., Anma, S., and and Maikuma, H. 1980. Liquefaction of mine tailings in the 1978 izu-

oshima-kinkai earthquake, central Japan. In Proc., 7th World Conf. on Earthquake

Engrg., 3: 89-96.

Polito, C.P. and Martin, J.R. 2001. Effects of nonplastic fines on the liquefaction resistance of


Riemer, M.F., Gookin, W.B., Bray, J.D., and Arango, I. 1994. Effects of loading frequency and

control on the liquefaction behavior of clean sands. Geotechnical Engineering Rep. No.

UCB/GT/94-07, Univ. of California, Berkeley, Calif.


of Toronto, Canada.

Sanin, M.V. and Wijewickreme, D. 2006. Cyclic shear response of channel-fill fraser river delta

silt. Soil Dynamics and Earthquake Engineering, 26(9): 854-869.

Singh, S. 1994. Liquefaction characteristics of silts. Ground Failures Under Seismic Conditions,

Geotechnical Special Publication, No. 44, ASCE. pp. 105-116.




Engineering, 128(10): 849-859.

Thompson, B. D., Grabinsky, M. W., Bawden, W. F. and Counter, D. B., 2009, In-situ

measurements of paste backfill in long-hole stopes, ROCKENG09: Proceedings of

3rd CANUS Rock Mechanics Symposium, Toronto, May. 10 pp.

Thottarath, S. 2010. Electromagnetic Characterization of Cemented Paste Backfill in the Field

and Laboratory. MASc, The University of Toronto, Canada.

Troncoso, J.H. and Verdugo, R. 1985. Silt content and dynamic behavior of tailing sands. In

Proc., XI Int. Conf. on Soil Mechanics and Foundation Engineering, Vol. 2, pp. 1311–

1314.

Vaid, Y.P. 1994. Liquefaction of silty soils. Ground Failures Under Seismic Conditions,

Geotechnical Special Publication No. 44, ASCE. pp. 1-16.

Vaid, Y.P. and Thomas, J. 1995. Liquefaction and postliquefaction behavior of sand. Journal of


Vaid, Y.P. and Chern, J.C. 1985. Cyclic and monotonic undrained response of saturated sands.

Advances in the Art of Testing Soils Under Cyclic Conditions: 120-147.

Vaid, Y.P., Stedman, J.D., and Sivathayalan, S. 2001. Confining stress and static shear effects in

cyclic liquefaction. Canadian Geotechnical Journal, 38(3): 580-591.

Wijewickreme, D., Khalili, A., and Wilson, G. 2010. Mechanical response of highly gap-graded

mixtures of waste rock and tailings. part II: Undrained cyclic and post-cyclic shear

response. Canadian Geotechnical Journal, 47(5): 566-582.

Wijewickreme, D., Sanin, M.V., and Greenaway, G.R. 2005. Cyclic shear response of fine-

grained mine tailings. Canadian Geotechnical Journal, 42(5): 1408-1421.

Xenaki, V.C., Athanasopoulos, G.A. 2003. Liquefaction resistance of sand-silt mixtures: An

experimental investigation of the effect of fines. Soil Dynamics and Earthquake

Engineering, 23 (3): 183-194.

http://refworks.scholarsportal.info/refworks/~2~


http://www.scopus.com.myaccess.library.utoronto.ca/record/display.url?eid=2-s2.0-0037392897&origin=reflist&sort=plf-f&cite=2-s2.0-0037392897&src=s&imp=t&sid=NRluY3R6TNqHmPdfQOSDMQx%3a60&sot=cite&sdt=a&sl=0



Tables and Figures


Table 3-1 Summary of test parameters and results

Specimen ID Curing time

(hrs) CSR

aN

bSAR ru FC%

ce

dec

eef

fDr%

MT-1 - 0.10 37 9.0 0.95 67 0.645 3.985 0.963 75

MT-2 - 0.12 14 4.4 0.92 67 0.640 3.970 0.955 76

MT-3 - 0.15 5 2.6 0.91 67 0.645 3.985 0.963 75

MT-4 - 0.20 2 2.9 0.85 67 0.645 3.985 0.963 75

CPB/4.5-1 4 0.09 261 11.7 0.96 71.5 0.740 5.105 1.035 73

CPB/4.5-2 4 0.13 15 7.3 0.96 71.5 0.740 5.105 1.035 73

CPB/4.5-3 4 0.15 6 29.0 0.96 71.5 0.735 5.088 1.028 74

CPB/4.5-4 4 0.20 2 4.5 0.64 71.5 0.740 5.105 1.035 73

CPB/4.5-5 7 0.20 4 3.9 0.9 71.5 0.735 5.088 1.028 73

MTS55-1 - 0.09 25 3.2 0.97 38.2 0.355 1.193 0.929 79

MTS55-2 - 0.12 10 3.1 0.95 38.2 0.350 1.184 0.916 79

MTS55-3 - 0.14 6 3.2 0.97 38.2 0.355 1.193 0.929 79

MTS55-4 - 0.18 2 3.3 0.87 38.2 0.360 1.201 0.942 78

CPBS55/4.5-1 4 0.11 No liquefaction - - 42.7 0.420 1.478 0.984 74

CPBS55/4.5-2 4 0.16 23 4.3 0.96 42.7 0.420 1.478 0.984 74

CPBS55/4.5-3 4 0.19 8 3.9 0.96 42.7 0.420 1.478 0.984 74

CPBS55/4.5-4 4 0.22 4 3.6 0.95 42.7 0.425 1.487 0.995 73

CPBS55/4.5-5 4 0.25 2 3.4 0.85 42.7 0.420 1.478 0.984 74

CPBS55/4.5-6 7 0.25 39 3.0 0.95 42.7 0.425 1.487 0.995 74

a Number of cycles at 5% DA axial strain

b Strain anisotropy ration (SAR)

c Global or gross void ratio

d Void ratio of sand skeleton

e Void ratio of fine matrix

f Relative density based on gross void ratio


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 5 10 15 20 25 30 35 40

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-30

-20

-10

0

10

20

30

-5 -4 -3 -2 -1 0 1 2 3 4 5

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-6

-4

-2

0

2

4

0 5 10 15 20 25 30 35 40

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure 3-1 Typical response of cyclically loaded MT specimens ('c=100 kPa and CSR=0.10): (a) Stress paths with failure lines obtained from

monotonic testing, (b) stress strain behavior, (c) excess pore water pressure ratio (ru) versus number of cycles, and (d) axial strain versus number of

cycles.


-50

-40

-30

-20

-10

0

10

20

30

40

50

0 20 40 60 80 100 120

t =

(

1-

3)/

2,

(kP

a)

s' = (1+3)/2, (kPa)

Failure line (FL) from monotonic undrained tests

= 31o

(a)

FL, = 28o

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 0.5 1 1.5 2 2.5 3

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-50

-40

-30

-20

-10

0

10

20

30

40

50

-12 -10 -8 -6 -4 -2 0 2 4 6 8

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-12

-10

-8

-6

-4

-2

0

2

4

6

8

0 0.5 1 1.5 2 2.5 3

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure 3-2 Typical response of cyclically loaded MT specimens ('c=100 kPa and CSR=0.20): (a) Stress paths with failure lines obtained from


cycles.


0

0.05

0.1

0.15

0.2

0.25

1 10 100

CS

R

Number of cycles to reach 5% DA axial strain

MT_100kPa

MTS55_100kPa

Figure 3-3 Cyclic resistance chart for uncemented mine tailings (MT) and MT-sand mixture (MTS55).


-50

-40

-30

-20

-10

0

10

20

30

40

50

0 20 40 60 80 100 120

t =

(

1-

3)/

2,

(kP

a)

s' = (1+3)/2, (kPa)

Failure line (FL) from monotonic undrained tests= 31.6o

FL

= 26.1o

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 5 10 15 20 25 30

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-25

-20

-15

-10

-5

0

5

10

15

20

25

-5 -4 -3 -2 -1 0 1 2 3

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-6

-4

-2

0

2

4

0 5 10 15 20 25 30

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure 3-4 Typical response of cyclically loaded MTS55 specimen ('c=100 kPa and CSR=0.09): (a) Stress paths with failure lines obtained from


cycles.


-20

-10

0

10

20

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 50 100 150 200 250 300

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-30

-20

-10

0

10

20

30

-7 -6 -5 -4 -3 -2 -1 0 1 2

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-8

-6

-4

-2

0

2

0 50 100 150 200 250 300

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure 3-5 Typical response of cyclically loaded CPB/4.5 specimen ('c=100 kPa and CSR=0.09): (a) Stress paths with failure lines obtained from


cycles.


0

0.05

0.1

0.15

0.2

0.25

0.3

1 10 100 1000

CS

R


MT_100kPa

CPB/4.5_100kPa

CPB/4.5_100kPa+1hr

MTS55_100kPa

CPBS55/4.5_100kPa

CPB/4.5_100kPa_7hrs

CPBS55/4.5_100kPa_7hrs

CPBS55/4.5_100kPa_no liquefaction

Figure 3-6 Cyclic resistance chart for cemented specimens (CPB/4.5 and CPBS55/4.5) at four and seven hours

of curing time in comparison with the uncemented specimens.


-50

-40

-30

-20

-10

0

10

20

30

40

50

0 20 40 60 80 100 120

t =

(

1-

3)/

2,

(kP

a)

s' = (1+3)/2, (kPa)

(a)Failure line (FL) from monotonic undrained tests

= 33.7o

FL

= 24.5o

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 5 10 15 20 25

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-40

-30

-20

-10

0

10

20

30

40

-6 -5 -4 -3 -2 -1 0 1 2

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-6

-4

-2

0

2

0 5 10 15 20 25

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure 3-7 Typical response of cyclically loaded CPBS55/4.5 specimen ('c=100 kPa and CSR=0.16): (a) Stress paths with failure lines obtained from


cycles.


0

0.05

0.1

0.15

0.2

0.25

1 10 100

CS

R


MT_100kPa

MTS55_100kPa

Saebimoghaddam 2010

Crowder 2004

Ishihara et al. 1980

Figure 3-8 Comparing cyclic resistance MT and MTS55 with those of uncemented tailings in literature


0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

72 73 74 75 76 77 78 79

CR

R

Dr (%)

MT_100kPa

MTS55_100kPa

CPB/4.5_100kPa

CPBS55/4.5_100kPa

(a)

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0 10 20 30 40 50 60 70 80

CR

R

FC (%)

MT_100kPa

MTS55_100kPa

CPB/4.5_100kPa

CPBS55/4.5_100kPa

(b)

Figure 3-9 Variation of cyclic resistance with (a) relative density, Dr, and (b) fines content, FC%.


Chapter 4

Characterizing Stiffness Development in Early Age

Cemented Paste Backfills with Sand in a non-

Destructive Triaxial Test

Abstract

Obtaining early age stiffness parameters for cemented paste backfill (CPB) is important for

designing and modeling backfilling systems. Previous studies used geophysical methods to

obtain stiffness parameters, which correspond to very small strain levels. A wide range of field

strains are misrepresented by small strain parameters. In the present research work, a non-

destructive triaxial testing procedure is proposed to obtain intermediate strain stiffness

parameters. In addition, the volume and pore water pressure changes due to the ongoing cement

hydration process are measured to understand the mechanisms contributing to stiffness

development. The evolution of effective (skeleton) stiffness parameters is monitored for five

days and the results are presented for the plug and main pour CPB-sand mixes. A framework is

suggested to predict the undrained stiffness at degrees of saturation representative of the field

conditions. Very low suction values were measured, which indicates that self-desiccation is not

influential on the development of early age stiffness. The study concludes that the proposed test

setup and procedure in providing stiffness parameters at representative strain levels of the field is

a credible laboratory method.


4.1 Introduction

Cemented paste backfill (CPB) has been gaining popularity in the mining industry during the

past two decades. Confidence in cemented paste backfilling systems has been increasing as it has

the advantage of easy transportation to underground openings, rapid rates of backfilling,

increased ore recovery, less water drainage concerns, and the environmental benefits of diverting

tailings from surface storage. A comprehensive understanding of the strength and stiffness

development of CPB during early binder hydration stages has a great impact on the safety and

the economy of mines. In particular, it helps to achieve safe and economic barricade designs and

to optimize the backfilling schedule. However, the adopted backfill design strategies are

generally considered overly conservative due to lack of comprehensive understanding of the

properties and behavior of CPB in the early age stages.

During the early curing ages of CPB, stope closure due to adjacent mining activities as well as

elastic rock wall relaxation can cause intermediate strain levels in CPB (e.g., Milne et al., 2004),

which range from 0.001% to 0.1% according to Ishihara (1996). Therefore, the provision of

stiffness parameters at strain levels representative of field strains is imperative for modeling and

design purposes. Characterization of an isotropic elastic material requires the determination of at

least two material parameters out of four possible measurements (i.e. Young‟s modulus E and

Poisson‟s ratio or shear modulus G and bulk modulus K). In the limited number of studies

performed to acquire the early age stiffness of CPB, bender elements were used to characterize

its shear stiffness. In some of these studies, a constant value was assumed for another parameter

in addition to the shear modulus so that the other material parameters could be calculated. There

is a reservation on assuming a constant value for any stiffness parameter of a cemented material

undergoing hydration and transitioning from a non-Newtonian fluid to a solid. Recent effort by

Galaa et al. (2011) obtained a complete set of stiffness parameters for early age CPB through

combined measurements of P-wave and S-wave velocities. The parameters obtained by

measuring the velocities of body waves correspond to very small strains of less than 0.001%

(Addo and Robertson, 1992; and Leong et al., 2004). Such strain levels are not representative of

a wide range of strains encountered in the field. However, measuring body wave velocities is still

a successful tool to monitor the pattern of stiffness evolution. Furthermore, all studies assumed


full saturation of the freshly mixed CPB while field evidence confirmed the presence of air in the

form of occluded bubbles (le Roux et al., 2005).

To provide backfill properties at strain levels representative of field strains, there is a need to

conduct a non-destructive test on CPB to obtain the stiffness parameters and monitor their

evolution during the first few days of curing at intermediate strain ranges. In this chapter, a non-

destructive triaxial test is used to obtain the stiffness parameters at an intermediate strain, while

also measuring the volume and pore water pressure (PWP) changes due to the ongoing cement

hydration process. The effective (drained or solid skeleton) stiffness parameters are obtained and

a framework is suggested to predict the undrained (overall) stiffness at saturation conditions

representative of the field conditions.

4.2 Background

4.2.1 Stiffness of geomaterials at pre-failure strains

4.2.1.1 Stiffness of particulate media

Models relating stresses to deformations were initially developed for continuums and, then,

either adopted as originally developed (at certain strain levels) or modified for geomaterials. The

fact that most geomaterials are composed of at least two and sometimes four phases poses more

complexity in understanding its behavior. According to Fredlund (1993), those phases are:

a) The solid skeleton, which is the assembly of particles exhibiting strength through contact

friction and often cementation.

b) The pore fluid, which may be only water in the case of full saturation or air-water mixture

in case of partial saturation at high degrees of saturation (Sr). Air exists in the form of

occluded bubbles in air-water mixtures with Sr greater than approximately 85%.

c) The continuous air phase, which usually exists at Sr values lower than around 85%.

Matric suction can be obtained by separate measurements of the pore air and pore water

pressures.


d) The air-water interface or contractile skin, which is considered a separate phase as it has

different properties from those of the contiguous water phase. It exists as a membrane

under tension and is interwoven throughout the soil structure. This phase provides

influence on the mechanical behavior when the air phase is continuous.

When soil particles are uncemented, the effective stress is the predominant contributor to the

skeletal stiffness. Therefore, the overall stiffness of a saturated soil depends on the drainage

conditions of loading. When the soil is loaded under undrained conditions, pore water is assumed

incompressible and will develop pore water pressures under the applied stresses causing

reduction in the effective stress. While under drained conditions, all excess pore water pressures

are assumed to fully dissipate and soil deforms according to the stiffness of the skeleton. On the

other hand, the overall stiffness of an unsaturated soil with a continuous air phase is affected by

the matric suction and the effective stress will be function of the pore air and pore water

pressures developed under undrained loading conditions. Under drained loading conditions,

unsaturated soils deform according to the stiffness of the skeleton. In the case of undrained

loading of unsaturated soils, where air is in the form of occluded bubbles, the overall undrained

bulk stiffness is a combination of the stiffness of the solid skeleton and that of the air-water

mixture. Schuurman (1966), Fredlund (1976), Santamarina et al. (2001), and Tsukamuto et al.

(2002) proposed formulations to acquire the overall bulk stiffness knowing the amount of air in

the air-water mixture. Those formulations are discussed later in this chapter.

In summary, the values of the undrained stiffness parameters change according to the saturation

conditions of the soil. Whereas, under drained conditions, the parameters have the same value

and are termed effective stiffness parameters. The shear modulus obtained from drained loading

should be the same as that obtained from undrained loading at the same effective stress, since the

pore fluid does not sustain shear stresses.

4.2.1.2 Strain dependency of stiffness moduli

Stiffness moduli of materials degrade as the strain level they are measured at increases. There is

a consensus that 0.001% strain marks the strain threshold beyond which continuous increase of

strain entails considerable degradation of the stiffness of most soils and weak rocks (see Figure

4-1). Moduli measured at, or below, this strain level are taken as the reference small strain

stiffness moduli (Eo or Go) at which the linear elastic model is accurately justified (Ishihara,


1996). Seed and Idriss (1970) reported that only the void ratio influences the variation in the

stiffness of a sandy soil measured within such small strain ranges. However, Kokusho (1980)

reported that the reference shear modulus (Go) is proportional to square root of the effective

confining pressure. Clayton and Heymann (2001) and Gasparre et al. (2007) found that even the

stress history and loading path do not affect the reference moduli of cohesive soils and that

stiffness measured at small strains is independent of the strain rate and thus can be determined

from cyclic (dynamic) or static tests.

Soil behavior is considered to be within the intermediate strain range if the problem is associated

with strains below 0.1%. Within such strain levels, the stiffness decreases as the strain increases.

However, no destructuring or damage to the tested soil is expected. Ishihara (1996) reported that

soils loaded under intermediate strains absorb and dissipate energy causing a hysteretic nature in

the stress-strain relationship. Nevertheless, repetitive loading does not cause any change

(degradation) in the stiffness or damping properties. According to Ishihara (1996), this behavior

is termed “non-degraded hysteresis type” and soil characteristics can be accurately represented

using the linear visco-elastic theory. In summary, soil stiffness at the intermediate strain ranges is

strain-dependent but cycle-independent.

Another factor that affects the values and the rate of degradation of the stiffness with increasing

strain is the effective confining stress (Seed and Idriss, 1970; and Clayton, 2011). Higher

stiffness is expected at higher effective confining stresses. In addition, the stress path was found

to affect the stiffness and its rate of degradation with strain for cohesive soils tested within the

intermediate strain range (Clayton and Heymann, 2001). Strain rate becomes an important factor

that affects stiffness values at intermediate strains (Tatsuoka and Shibuya, 1992; and Sorensen et

al., 2007) in contrast to the case of small strains.

For strains > 0.1%, stiffness and damping properties change not only with strain, but also with

load repetition. Ishihara (1996) referred to this kind of behavior as “degraded hysteresis type”. At

this level, considerable destructuring of the material takes place. The stress history, stress path,

effective confining pressure, and strain rate affect the manner of degradation of stiffness with

strain and load repetition. The initial rates of stiffness degradation at intermediate strain levels

are higher than the rates at larger strain levels (see Figure 4-1). The three strain levels defined in


the previous paragraphs are assumed to occur prior the failure/yield of the material and are

termed “pre-failure deformations”.

4.2.2 Measuring stiffness parameters

4.2.2.1 Methods for measuring stiffness

The variability of engineering problems and technology tools has offered numerous techniques

for measuring the stiffness parameters of geomaterials. Stiffness measurements can be conducted

in the field or laboratory and can be obtained via direct measurements or indirectly through

correlations with other parameters. The choice between field and laboratory testing, or often a

combination of both, is made under constraints pertaining to the required parameters, the

availability of equipment, the expected quality of the data, the nature of the project/ground. The

methods for stiffness measurement can be classified into:

a) Large strain field measurements: in which stiffness parameters are obtained via

correlations based on the accumulation of data and case histories. Such methods include

the simple SPT and CPT. High strain characteristics, such as soil strength, are measured

and their results are usually correlated to small strain soil properties (Kramer, 1996).

b) Indirect field measurements: in which geophysical techniques are utilized to obtain the

stiffness parameters via measuring the velocities of either body or surface waves. Such

techniques include continuous surface wave testing, spectral analysis of surface wave,

down-hole and cross-hole testing, seismic CPT, and seismic dilatometer. Stiffness

parameters, usually Go, obtained from these methods correspond to very low strain levels.

c) Direct field measurements: in which stress-strain relationships are obtained in situ at a

certain depth. This category includes the dilatometer and pressuremeter families. These

tests are suitable for obtaining intermediate to large strain stiffness parameters (Ng et al.

2000).

d) Indirect laboratory measurements: in which piezoelectric crystals and/or bender elements

are used to measure body wave velocities from which stiffness parameters are inferred

assuming an isotropic elastic medium. Bender elements and other piezoelectric crystals

can be embedded in the caps and pedestals of the triaxial apparatus (Schultheiss, 1981;


Bates, 1989; Leong et al., 2004), the direct simple shear device (Dyvik and Olsen, 1989),

and the resonant column device (Ferreira et al., 2006) for complementary stiffness

measurements. Stiffness parameters measured from these tests correspond to very small

strain levels. In addition, this category includes tests where the soil is forced to vibrate to

determine its resonant frequencies and its response at different frequencies. Soil moduli

and damping properties can be determined by analyzing that response. Resonant column

tests apply this concept in torsion, axially, and in flexure. These tests can provide

stiffness parameters, commonly G, at small to intermediate strains starting at strains as

low as 0.00001% (Clayton, 2011).

e) Direct laboratory measurements: this category includes tests where the specimen is

loaded in different modes while measuring the associated deformations. The most

commonly practiced tests are the triaxial tests, simple shear tests, and torsional shear tests

(Seed and Idriss, 1970; and Kramer, 1996). Normally, triaxial testing is capable of

measuring intermediate to large strain stiffness parameters while it faces accuracy issues

when testing at small strains, which is more pronounced for stiff soils. However,

advancements in deformation and load measurements can extend its capability to

measuring stiffness parameters at strain levels as low as 0.0001% (e.g. Tatsuoka et al.,

1994).

4.2.2.2 Triaxial tests for soil stiffness parameters

This research is concerned more about triaxial testing. Therefore, this section reviews the

limitations of achieving high accuracy in strain and load measurements in triaxial apparatus.

Moreover, the advancements that extended its capabilities towards measuring stiffness

parameters for wide strain ranges, even for very small strain levels, are discussed. The triaxial

apparatus has been more commonly used in routine testing than other laboratory testing devices.

Indeed it has been distinguished because of its simple mechanism, ease of handling, inexpensive

use, wide availability, ease of access to practicing engineers, and its ability to execute various

loading patterns and stress paths, compared with other testing devices. Nevertheless, routine

triaxial tests are less credible when determining stiffness and damping properties at strain levels

smaller than 0.01% (Kokusho, 1980; and Tatsuoka et al., 1994). Hence, for such small strain

ranges the resonant column device is more advantageous than the triaxial apparatus. The


resonant column device usually provides the shear modulus and damping properties at very small

strains down to 0.00001%. However, it does not provide direct measurements of the elemental

behavior of the tested soil (Jardine et al., 1984). In other words, the deformations required to

simultaneously obtain more than one stiffness parameter cannot be measured in the resonant

column test.

The lack of accurate measurements of soil deformations in routine triaxial testing is ascribed to

measuring displacements externally, which include many extraneous movements. Soil

deformations can be masked by the movements which ensue from the compliances of the load

cell, the loading ram, and the loading frame. In addition, specimen setting errors due to bedding

between soil and platens, trapped air around top cap, and ram misalignment contribute to the

noise in the measured displacements. In addition to displacement errors, friction between the

loading ram and the ram bushings skews load measurements and consequently biases the

measured stiffness. The load measurement problem has been solved by using submersible load

cells which are mounted inside the triaxial cell between the ram and the top cap.

During the past four decades, many researchers attempted to offer means to circumvent the noise

which hampers the acquisition of credible small strain stiffness parameters from triaxial devices.

Several approaches were, therefore, developed to obtain the net local deformations of the test

specimens. One approach is to measure the relative displacement between two or more reference

points along the central length of the specimen. This approach has been implemented via

attaching linear variable differential transformers (LVDT) axially and radially along peripheries

of the specimen (e.g. Brown and Snaith, 1974). Goto et al. (1991) and Tatsuoka et al. (1994)

adopted the same approach via mounting linear deformation transducer (LDT) along with an

array of proximeters. Moreover, Burland and Symes (1982) and Jardine et al. (1984) used

electrolytic level device to measure local axial strains and axial tilt of the specimen enabling

axial strain measurements with resolution up to 0.001%.

Other studies opted for measuring local deformations using optical techniques. The advent of

high resolution cameras and the advancements in image processing tools and computational

power of computers have been utilized to track the movement of a number of reference points on

the specimen in real time (Macari et al., 1997; and Gachet et al., 2007). Nevertheless, this

technique still suffers from limited accuracy at small strain levels. Heymann et al. (2005)


managed to measure ultra-small deformations using the laser-optic interferometry technique

which provides contactless displacement measurements with resolutions up to 0.0006 m.

The aforementioned developments have led to significant increase in the availability of high-

quality small strain stiffness data from advanced triaxial testing. The choice depends on the cost

and availability of the equipment used to attain high resolution measurements.

4.2.3 Stiffness development in CPB

In CPB, the presence of cement introduces stiffness to the soil skeleton as two major

mechanisms take place. One of these mechanisms is the cement hydration reaction which results

in the formation of solid products that fill the pore space and create bonds between the particles

of mine tailings. The second mechanism is self-desiccation, which also has a considerable effect

on the early age stiffness of CPB, when the effective stress predominates, and is discussed later

in this section.

A number of studies have been conducted to monitor the hydration process for different

compositions of CPB as well as under different environmental and chemical conditions (e.g.

Benzaazoua et al., 2004; Simon, 2005; and Thottarath, 2010). However, only a few studies have

monitored the evolution of the material stiffness during early curing ages and even fewer studies

have quantified the stiffness parameters. Klein and Simon (2006) conducted a study to monitor

the development of early age stiffness so as to compare different material compositions. To carry

out this comparison, they measured the shear wave velocity using bender elements and

performed Vicat needle tests to mark the initial and final set of the material. Klein and Simon

(2006) targeted monitoring a stiffness indicator, such as shear wave velocity, rather than

quantifying the stiffness parameters.

In a first attempt to quantify the early stiffness parameters of hydrating CPB, Helinski et al.

(2007) measured the shear wave velocity in a triaxial test using embedded bender elements.

However, bender elements only permit the measurement of the shear stiffness. Inferring, for

instance, the bulk stiffness of the material while monitoring only the development of the shear

stiffness of CPB over time requires the assumption of a value for the Poisson‟s ratio () (e.g.,

Helinski et al., 2007). However, assuming a constant can be problematic. For example,

D'Angelo et al. (1995) and Boumiz et al. (1996) showed that for cemented pastes and mortars,


starts close to 0.5 for the fresh material in a fluid state and decreases to reach 0.2 after 24 hours.

Therefore, combined S-wave and P-wave measurements are required in order to experimentally

quantify both bulk and shear stiffness parameters and the related elastic parameters. In a recent

study conducted by Galaa et al. (2011), combined measurements of P-wave and S-wave

velocities were performed to obtain a complete set of stiffness parameters for early age CPB

without the need to assume any of them. With regards to field stiffness measurements, le Roux et

al. (2005) measured the shear stiffness of CPB in the stope using a self-boring pressuremeter

(SBP) at intermediate to large strain levels. Although the test was carried out after curing ages of

5 and 11 months, the technique is valid at earlier curing ages.

The self-desiccation mechanism is associated with hydration by binding water during the

hydration reaction. The net volume of the hydration products is less than that of the reactants.

Therefore, the pore volume is occupied with less volumes and as the hydration process continues

more water is consumed. This causes the water-gas interface to sink into the pores pulling the

solid particles together by surface tension (Kim and Lee, 1999; and Acker. 2004). Several studies

quantified the suction development ensuing from self-desiccation in CPB. Grabinsky and Simms

(2005), Helinski et al. (2007), and Simms and Grabinsky (2009) studied the self-desiccation

developing in CPB. The resulting matric suction was directly measured by Grabinsky and Simms

(2005) for 5% CPB using tensiometers. Suction was immediately observed, and increased at a

rate of 40 kPa/day between the first and the second day to exceed 100 kPa after about 6 days

when the tensiometer cavitated. Helinski et al. (2007) measured the developing suction by

monitoring the reduction in the applied back pressure. The cumulative reduction reached 800 kPa

after 10 days. Such magnitudes of suction significantly pull particles together and will

consequently cause an increase in the skeleton stiffness. The testing program in the present

research work was planned to measure the stiffness and capture the contribution of the self-

desiccation mechanism to the measured stiffness.

4.3 Experimental Procedure and Setup

4.3.1 Materials composition and sample preparation

The basic components for the mixtures tested in this research are similar to those used in Chapter

2 and 3. These components are silica tailings (MT), glacial sand, and binder (10% Type 10 PC


and 90% BFS). X-Ray Fluorescence Spectroscopy (XRF) was performed to determine the

chemical composition of the tailings and binder and the results were shown earlier in Chapter 2.

Specimens tested in this experimental program were composed of 45% MT and 55% sand in

addition to the binder. Stiffness testing took place on specimens with binder contents of 2.2%

and 4.5%, and will be denoted as CPBS55/2.2 and CPBS55/4.5, respectively. In the field, the

initial plug is composed of CPBS55/4.5 and the main pour is composed of CPBS55/2.2. As

previously mentioned in Chapter 2, the peaks of the electric conductivity (EC) measurements

conducted by Thottarath (2010), which indicates the initial setting for CPBS55/2.2 and

CPBS55/4.5, occurred at around 3200 min (≈53 hours) and 700 min (≈12 hours), respectively.

The triaxial specimens were prepared according to the same procedure followed in Chapters 2

and 3 with slight modifications. After casting CPBS, the specimen underwent dead weight

consolidation under 12.5 kPa after which the cell was filled with de-aired water. The used water

had been de-aired thoroughly for eight hours under vacuum so as to minimize/eliminate the

compressibility of the cell fluid. This helps to increase the accuracy of measuring the overall

volume change of the specimen through measuring the volume change of the cell fluid. The

specimen was then hydrostatically consolidated at 30 kPa while omitting the back saturation

stage to keep the specimen under similar conditions to the field. The hydrostatic consolidation

lasted for two hours, after which the specimen, at the age of four hours, was ready for the first

stiffness loading stage. The stiffness loading procedure is explained in the next section.

4.3.2 Experimental program

4.3.2.1 The triaxial apparatus and test description

A servo-hydraulic triaxial machine designed by GCTS was used to apply prescribed small

displacement cycles of 0.1 mm peak-to-peak amplitude. Specimens had diameters of 70 mm

each, and heights ranging from 145 mm to 150 mm. The strains resulting from the displacement

cycles were around 0.033% for such specimens‟ heights, which fall within the intermediate strain

range. The displacement cycles were applied at different ages throughout the testing period,

which lasted for five days after mixing. The frequency of cycling was 0.005 Hz and the

specimens were allowed to drain. The resulting axial strain rate was 0.04%/min which is lower

than the strain rate requirements to ensure drainage (see testing program section in Chapter 2).

Since perfect drainage is expected, the parameters measured in this study are the effective


stiffness parameters. Also, suction and volume changes resulting from self-desiccation and

cement hydration processes were measured throughout the testing period. A new base for the

triaxial cell was machined to accommodate the tensiometer used for direct suction

measurements. Figure 4-2a shows a schematic of the triaxial setup, locations of the sensors, and

the modified base. In addition, Figure 4-2b shows the triaxial cell and the volume change device

(VCD). Figure 4-2c shows a picture of the modified base with the tensiometer tip protruding out

of the pedestal with a sufficient length to account for the thickness of the porous stone.

The accuracy of measurements was tied to four main sensors in the triaxial setup, namely, the

load cell, VCD, the external displacement transducer, and the tensiometer. A submersible load

cell of 500 lbs capacity was used to eliminate the effects of the ram-bushing friction as shown in

Figure 4-2b. The load cell enabled stress measurements with a resolution of one kilopascal. VCD

used in this study is a frictionless rolling diaphragm type designed by GCTS with 0.01 ml

resolution (see Figure 4-2b). VCD resolution enabled volumetric strain measurements to the

nearest 0.002%. The specifications of VCD state that it is hysteresis free. However, hysteresis

was found to diminish under pressures > 25 kPa. Hysteresis free measurements might have

needed this pressure to fully expand the rolling diaphragm around the inner piston. Therefore,

specimens were consolidated at 30 kPa and the cell pressure was kept at this value throughout

the test period. Constant pressure was maintained at 30 kPa by the digital system controller,

which ensures that the ram movements in and out of the cell do not cause any cell pressure

change and consequently the triaxial cell walls do not deform during the strain cycles. Volume

change due to ram movement was calculated and the measured overall volume changes during

the strain cycles were corrected accordingly. VCD was fully saturated prior to use and all lines

and VCD compartments were flushed with deaired water to remove air bubbles prior to use. This

ensures that air does not introduce error in the volume change measurement either by its

compressibility or by dissolving in water during the test (Diez d'Aux, 2008).

As for axial displacements, they were measured using an external LVDT with a resolution of

0.005 mm, which is sufficient for the strain levels of interest in the present research.

T5x miniature tensiometer was used for direct measurements of suction. This transducer is

capable of measuring up to 200 kPa negative PWP and 85kPa positive PWP. The device is

composed of two main components; the sensor body and the shaft (see Figure 4-3). The shaft and


the ceramic tip have small dimensions to allow minimal soil disturbance and fast response. The

tensiometer was connected to a data logger controlled by DaisyLab software that enabled

continuous measurements during the test period. The capacity and response time of the T5x

tensiometer is dependent on the quality and efficiency of the preparation procedure or what is

termed “servicing”. The measurement range was temporarily extended by subjecting the

tensiometer to cycles of negative and positive PWP to remove cavitation nuclei (Guan and

Fredlund, 1997; Tarantino and Mongiovi, 2001; Take and Bolton, 2004; and Simms and

Grabinsky, 2009). After servicing, the tensiometer was inserted into the modified base of the

triaxial cell which allows sealing and securing the tensiometer so as to remain in contact with the

specimen and prevent the escape of water through leakage or evaporation. CPBS was poured in

the split mould over the tip of the tensiometer which is the only part that protrudes out of the

base as can be seen in Figure 4-2a and Figure 4-2c.

Testing was planned on two specimens of each composition (CPBS55/2.2 and CPBS55/4.5).

However, several checks were made prior to testing CPBS to assure the accuracy of the

measurements as shown in the following section.

4.3.2.2 Verifying experimental parameters

Several sources of noise had to be examined to ensure that the accuracy of measuring stiffness

parameters at the intermediate strain range was not compromised. Furthermore, the

compressibility of the cell fluid had to be examined to determine the deaeration period that

provides virtually incompressible water. Therefore, dearation time was varied until the volume of

the cell water did not change under pressure. Figure 4-4 shows the volumetric changes that

resulted upon the application of 30 kPa on water deaired for four and seven hours (Water#1 and

Water#2, respectively). Water#1 exhibited 3.8 ml compression during 21 hours, whilst Water#2

showed almost no change in volume. However, minor fluctuations in the volume of Water#2,

which did not exceed ±0.15 ml, are attributed to fluctuations in room temperature over the 21

hours of measurements. These fluctuations occurred on the long term and are not expected to

affect volume change measurements during the application of the strain cycles which takes only

20 minutes. During the strain cycles, volume change was measured rather than directly

measuring radial strains in order to avoid mounting local displacement transducers in contact

with CPBS specimens before shearing. At such early ages, CPBS specimens are very soft and


may experience pre-shearing deformations while sticking or mounting transducers. Also, the

result of Water#2 in Figure 4-4 shows that the triaxial cell is leakage-proof.

In addition, noise due to system compliance was examined by testing a dummy steel specimen

under the same experimental setup. The recorded displacement is assumed to sum frame

deformations, ram deflection, and load cell compliance, while deformations of the dummy

specimen are assumed negligible. A stress of 200 kPa was applied and the maximum recorded

displacement reached 0.004 mm, which is beyond the resolution of the LVDT. CPBS specimens

are expected to develop stresses even lower than 200 kPa by the end of the test period and the

system should exhibit less than the recorded deformations when a dummy specimen was used.

Therefore, the system was considered of sufficient stiffness and the overall system compliances

were not considered to mask the measured displacements within the intermediate strain level.

Finally, seating, bedding, and tilting errors are not worrisome as CPBS specimens were cast and

consolidated in place prior to loading and usually such errors are associated with specimens

prepared and shaped before being mounted in the cell.

A pilot test was conducted on a CPB material known to be softer (from UCS data) than the one

of interest in the present research. The pilot test aimed at ensuring that the material behaves

elastically under the proposed test and that the resulting deformations are captured with adequate

accuracy with the available devices and transducers. Confirming the elastic behavior within the

proposed strain level validates the application of the relationships between the measured and the

calculated elastic stiffness parameters. Moreover, obtaining an elastic recoverable behavior

ensures that the assumption of not destructuring the material is not violated. The amplitude of the

displacement cycles applied in the pilot test was twice the planned value for the main testing

program. Figure 4-5a shows the resulting stress-strain loops at the softest age (four hours). The

loops show that the material did not undergo any plastic deformations and the behavior was

elastic. A secant modulus can be easily defined from the shown loops. Figure 4-5b compares the

stress-strain loops obtained after one week with the four hour loop. The behavior at the earliest

and the latest tested ages fall under the “non-degraded hysteresis type” as defined by Ishihara

(1996).

The recorded volume change (shrinkage) of the pilot CPB specimen over the test period (one

week) is shown in Figure 4-5c. The specimen shrank approximately 11 ml (approximately 2%


volumetric strain) in one week, which indicates that the fluctuations in volume change (±0.15

ml) due to room temperature changes might not be a significant source of noise for long term

volume change measurements. The results of suction measurements are shown in Figure 4-5d. It

should be noted that suction measurements for the pilot test were carried out on a separate

specimen where the tensiometer‟s hole was not sealed, in contrast to in the modified triaxial cell

base. The separate specimen had a smaller volume (diameter of 50 mm and height of 100 mm)

than the one in the triaxial cell. The measured suction exceeded 100 kPa by the end of a week of

monitoring and showed good agreement with the self-desiccation suction measured using similar

tensiometer for the same material by Witteman and Simms (2010).

The combined error in stiffness measurements was estimated to be ±8% if the material‟s reaction

to the applied 0.1 mm P-P strain reaches ±200 kPa. The error decreases as the reaction of the

specimen decreases. The maximum reaction exhibited by the pilot CPB specimen was

approximately 100 kPa after one week of curing. The error at such stress level is estimated to be

±3.5%. At four hours age, the estimated error was ±3.0%. The results of the pilot test were

satisfactory and indicated that credible results can be obtained using the proposed test setup for

the main testing program.

4.4 Results

4.4.1 Effective stiffness parameters

This section presents the behavior of CPBS55/4.5 and CPBS55/2.2 under the prescribed strain

cycles for stiffness measurements at different curing ages. Two specimens of CPBS55/4.5 (A

and B) were tested to ensure the reproducibility of the results. Figure 4-6 shows the axial stress-

strain loops for both CPBS55/4.5_A and CPBS55/4.5_B resulting from the strain cycles

(stiffness loading) at four hours and after five days of curing. Figure 4-6a and Figure 4-6b shows

that the behaviour of both specimens at four hours was recoverable exhibiting linear hysteresis.

The secant Young‟s modulus Esec obtained from the loops was found to initially increase as

strain cycles progressed. Upon reaching the fourth cycle, and sometimes the third cycle, the

increase in the Esec ceased and the cycles coincided. Therefore, Esec at each age was picked from

the fifth cycle, which is highlighted in red in all the shown loops in Figure 4-6. The problem of

modulus change with cycles occurred only for the specimens tested at four hours. After this age,


all cycles were almost identical as shown for specimens tested from eight hours until the last

stiffness loading after five days (see Appendix B). Stress levels reached around 125 kPa at which

the estimated error in the measured modulus reached ±5%. At earlier ages, the error due system

compliance decreased, while the error due to load cell resolution increased resulting in a

combined error of approximately ±5%.

Upon reaching the setting time of the mixture, estimated as 12 hours for CPBS55/4.5, the stress-

strain loops experienced a sort of nonlinearity during the extension part of the strain cycle.

Figure 4-6c and Figure 4-6d show example of the observed nonlinearity at the age of five days.

Despite the observed nonlinearity, stress-strain loops were totally recoverable from the first to

the last cycle. Therefore, it is believed that the resulting behaviour was not a result of any plastic

deformations, but rather a result of losing contact between the top cap and the specimen. In other

words, as the material became stiffer it exhibited less deformation under the extension half of the

strain cycles and the ram-top cap assembly tended to separate from the top surface of the

specimen and the recorded stresses decreased. This problem could have been avoided if the top

cap was capable of providing a grip so as to record the actual material behaviour in extension.

Since such grip was not available, the linear Esec was calculated assuming that the stress-strain

loop behaves linearly as if the compression half was mirrored in extension. The assumed

extension half of the loop is plotted in dotted red lines in Figure 4-6c and Figure 4-6d.

At the lower binder content, specimens of CPBS55/2.2 were weaker but exhibited the same

behaviour in terms of the axial stress-strain loops (see Figure 4-7). Figure 4-7a and Figure 4-7b

shows that the behaviour of CPBS55/2.2_A and CPBS55/2.2_B at the age of four hours was

recoverable exhibiting linear hysteresis. Initial increase of Esec as loading progressed until the

fourth or third cycle was also observed for the specimen tested at four hours. By the fifth day of

testing, stress levels reached around 70 kPa at which the estimated error in the measured

modulus reached ±4.5%. At four hours, the error due to load cell resolution brought the

cumulative error to approximately ±7% as the generated stresses were around ±10 kPa.

However, the error after this age continued to be within the range of ±5%. The nonlinearity of

stress-strain loops was observed only after the fourth day (see Figure 4-7c and Figure 4-7d and

Appendix B) although the setting time for CPBS55/2.2 was estimated as 53 hours. Figure 4-7c

and Figure 4-7d show examples of the observed nonlinearity at the age of five days where Esec

was obtained similarly as in the case of CPBS55/4.5.


The measured Young‟s moduli for specimens of CPBS55/4.5 and CPBS55/2.2 over the five days

of testing are shown in Figure 4-8a. The stiffness of specimens A and B of each composition

showed close match, especially at the lower binder content, which indicates the repeatability of

the test and gives credibility for the measured parameters. Specimens with 4.5% binder showed

an expected superiority especially during the first two days where the rate of stiffness gain was

much higher than that of the specimens with 2.2% binder.

Having been able to measure the axial strain (a) and volumetric strain (v), the value of

Poisson‟s ratio () can be directly calculated from the following relationship:

a

v

1

2

1 [4- 1]

The results of the measured at different ages during the test period are shown in Figure 4-8b

for CPBS55/4.5 and CPBS55/2.2. The values of for all specimens and binder contents varied

from approximately 0.40 at the curing age of four hours to reach approximately 0.25 at 72 hours

and continued around this value until the end of the test. Both binder contents followed the same

trend of decrease of over curing time. The obtained results support the reservation mentioned

earlier on the assumption of a constant stiffness parameter for a cemented material undergoing

hydration.

Knowing the effective E and at a certain age, the effective G and K can be calculated at this

age using the following relationships assuming an elastic isotropic material (Clayton, 2011):

12

EG [4- 2]

213

EK [4- 3]

The obtained G and K over time are shown in Figure 4-8c and Figure 4-8d, respectively, for

CPBS55/4.5 and CPBS55/2.2. The evolution of G of all specimens followed the same trend as

the evolution of E. However, the evolution of K was very sensitive to the change in which

resulted in some deviations from the trends of evolution of E and G. This is more pronounced in

the bigger difference between the rates of the bulk stiffness gain of CPBS55/4.5_A and

CPBS55/4.5_B. In addition, CPBS55/4.5_A experienced a decrease in K after the first day in


response to a slight change in the rate of decrease of for the same specimen. Following this

decrease, K continued to increase until the end of the test.

4.4.2 Hydration-induced changes

4.4.2.1 Volume change (shrinkage)

Since the volume of hydration products is lesser than the volume of the reactants, cemented

materials shrink as the hydration process progresses. Water to cement ratio (W/C) is one of the

major factors affecting the amount of shrinkage. The measured shrinkage profiles for

CPBS55/4.5 and CPBS55/2.2 are shown in Figure 4-9. As expected, specimens with the higher

binder content shrank more than those with the lower binder. Specimens with the same binder

content showed acceptable agreement in the values and trends of shrinkage over the test period.

It was observed that stiffness loading at different ages caused permanent disturbance in the

measured volumes as can be seen in Figure 4-9 for CPBS55/4.5_A. However, this was not

observed for all specimens. Moreover, specimen CPBS55/2.2_A showed more fluctuations in

volume changes than the other specimens, which can be ascribed to more fluctuations in the

room temperature during the test. Heights and volumes used in calculating stiffness parameters

were corrected according to the monitored volumetric changes for each specimen.

4.4.2.2 Self-desiccation suction

The results of suction measurements for CPBS55/4.5 and CPBS55/2.2 are shown in comparison

with that of the pilot CPB specimen in Figure 4-10. The generated suctions in CPBS55/4.5_A

and CPBS55/4.5_B did not exceed 7 kPa throughout the test which is very low compared to the

generated suction in the pilot specimen. CPBS55/4.5 exhibited an initial increase in PWP

(decrease in suction) up to approximately 30 hours followed by increasing suctions. Suction

generated in CPBS55/2.2_A throughout the test did not exceed 3 kPa and the suction profile did

not show an initial decrease as observed in CPBS55/4.5 specimens. Communication with the

tensiometer in CPBS55/2.2_B was accidentally lost during the test. The available suction

measurements from CPBS55/2.2_A show that suction was disturbed at each stiffness loading

after which it took, on average, five hours to equilibrate and return to its value before loading.

Values of the drops in suction that occurred at each stiffness measurement were accumulated and

the resulting suction profile is shown in brown in Figure 4-10 and labelled as


“CPBS55/2.2_A_cumulative suction”. The maximum cumulative suction for the CPBS55/2.2_A

reached approximately 14 kPa after 120 hours.

To eliminate the effect of stiffness loading on the developing suctions, only suction was

measured (i.e. no stiffness loading during the test period) for one specimen in the triaxial cell.

This specimen was labelled “CPBS55/4.5_no loading” in Figure 4-10. Suction exhibited a slight

decrease during the second day, then continued to increase until it reached approximately 20 kPa

at the end of the test period. The maximum suction measured was around threefold the measured

suctions in the loaded specimens indicating that the stiffness loading affected the generation of

suction. However, 20 kPa is still significantly lower than the values measured for the pilot CPB

specimen and also lower than suctions measured by (Simms and Grabinsky 2009) using a similar

tensiometer.

4.5 Discussion

4.5.1 Effective stiffness parameters

The measured Esec suffered low accuracy when the material was very soft at four hours due to the

limited resolution of the load cell relative to the measured low stresses, as well as to the

constantly increasing modulus during stiffness loading up to the fourth cycle. The rate of

shrinkage at four hours was very high and the recorded volume changes during the 20 minutes of

stiffness loading were oscillating around a continuously changing volume as shown in Figure 4-

11. For a specimen undergoing shrinkage, particles are brought closer and the material becomes

stiffer. This could be the reason for the continuous increase in Esec during stiffness loading at

four hours.

The formulations applied to obtain the effective G and K knowing E and assume that the

material is elastic and isotropic. This assumption is considered valid in the present study despite

the anisotropy observed at large strains in Chapter 2 and 3. The validity of the assumption is

based on the isotropic behaviour observed at small strains under the early load cycles as shown

in Chapter 3 and confirmed in the present chapter.

The evolution of stiffness (E, G, and K) of CPBS55/4.5 specimens started at a higher rate up to

the age of approximately 18 hours, after which stiffness gain dropped to less than half its initial


rate (see Figure 4-8). The change in the rate of stiffness gain is believed to occur only after the

initial setting time, which was estimated as 12 hours by Thottarath (2010). The rate of stiffness

gain of CPBS55/2.2 exhibited a slight break at 48 hours. However, the break was towards a

slightly higher rate. The initial setting time for CPBS55/2.2 was estimated at 53 hours, which is

close to the age at which the rate of stiffness gain changed. Thompson et al. (2009) conducted

field pressure measurements in a stope filled by the same pastefill material. Pore pressure

measurements for CPBS55/4.5, which was used as plug, was found equal to the total stress until

approximately 14 hours after the sensor was reached by the paste. The deviation of PWP implies

that the solid skeleton exhibited sufficient stiffness to carry part of the overburden pressure. The

age at which the PWP deviated from the total stresses falls within the same ballpark of the initial

setting time and the change in the rate of stiffness evolution.

4.5.2 Predicting the undrained stiffness parameters

The degree of saturation of block and core specimens collected from the field by le Roux et al.

(2005) was found to range from 80% to 100%. Within this range air is likely to exist in the form

of occluded air bubbles as mentioned earlier in the background section. In addition, lots of small

air bubbles were visually detected in CPBS block specimens collected from the field. This

section focuses on predicting the undrained stiffness parameters for CPBS within a saturation

range representative of the field conditions. The overall stiffness is a combination of the skeleton

stiffness, which has already been determined from the drained triaxial testing, and the

compressibility of the air-water mixture.

Several frameworks had been proposed to incorporate the compressibility of air-water mixtures

when air is in bubble form. Schuurman (1966) proposed a framework based on Kelvin‟s equation

for a single capillary tube. This approach lacked practicality as it incorporated immeasurable

quantities such as the number and size of air bubbles in the pore fluid. Another framework

proposed by Fredlund (1976) was based on the compressibility law (Boyle‟s Law) and the

solubility law of gases in water (Henry‟s Law). Fredlund (1976) introduced a formulation to

compute the compressibility or the bulk modulus of an air-water mixture (Kaw) that can be

rewritten as:


a

rr

w

raw

K

hSS

K

SK

.)1(

1

[4- 4]

Where, Kw and Ka are the bulk moduli of water and air, respectively, and „h‟ is the volumetric

coefficient of solubility defined as the total volume of each gas that can be dissolved in water

under atmospheric pressure. By knowing Kaw at any degree of saturation (Sr), the overall

undrained bulk modulus (Ku) and the overall undrained Poisson‟s ratio (u) can be calculated

from the following formulations (after Tsukamuto et al., 2002):

n

nKKK saw

u

[4- 5]

B

B

s

ssu

)21(3

)21(3

[4- 6]

Where, n is the porosity of the specimen, s is the effective (skeleton) Poisson‟s ratio, and B is

the pore pressure parameter introduced by Skempton (1954), which is defined as:

s

aw

nK

KB

1

1 [4- 7]

Obtaining u requires substituting equation [4-7] into [4-6]. At this point Ku and u are known

and the Eu and Gu (Gu = Gs) can be obtained from the following equations:

)21(3 uuu KE [4- 8]

)1(2

)21(3

u

uuu

KG

[4- 9]

In the present research, Kaw was calculated by equation [4-4] at different values of Sr (Sr = 0.99,

0.95, 0.90 and 0.85), and n was corrected according to the recorded volume changes. Since the

heat of hydration profile was not available, the temperature in the specimen was assumed

constant and the value of h was taken as 0.01868, which corresponds to room temperature. The


undrained stiffness parameters were calculated from equations [4-5] to [4-9] knowing the

effective stiffness parameters from the drained triaxial testing, which takes the subscription „s‟ in

this section, and knowing Kaw at the aforementioned degrees of saturation. The predicted

undrained parameters for specimen CPBS55/4.5_B are shown in Figure 4-12 as an example. The

values of Eu at 99% saturation were slightly larger than the corresponding Es while the values of

Gu were the same as Gs, as expected. On the other hand, the values of Ku at 99% saturation were

10% to 20% larger than Ks while u values were about 5% larger than s. The values of Ku and u

dramatically decreased with the decrease of the degree of saturation from 99% to 95% and

continued to decrease to approach the values of the effective parameter at 85% as shown in

Figure 4-12. The predicted undrained parameters for the rest of the specimens are presented in

Appendix C.

Although the presented frameworks (relationships) were applied to S=85%, it should be noted

that as Sr decreases the accuracy of those relationships decreases for the following reason. The

possibility is higher to encounter large air bubbles as Sr decreases. Large air bubble could be

significantly larger than the surrounding particles and will affect the structure of the solid matrix

(Wheeler 1986). The case of interaction between the individual air bubbles and the solid matrix

is very complicated. Therefore, the approach presented by Fredlund (1976) is considered

practically sufficient in the present research.

4.5.3 Analyzing the hydration-induced changes

To analyse suction measurements, some differences between the pilot CPB specimen and the

CPBS measured in the present work should be pointed out. First, the binder used in the pilot

specimen was Portland cement (PC) while the binder used in CPBS specimens had 90% BFS. In

many cement and concrete research studies, slag based cements were found to exhibit larger

shrinkage than PC (e.g., Collins and Sanjayan 2001). Therefore, having lower suctions

developing in CPBS specimens disagrees with previous concrete studies. Second, in the pilot

test, suction was measured in a separate specimen tested on a similar modified base but not

enclosed in a triaxial cell. Besides, the specimen had a smaller volume (diameter of 50 mm and

height of 100 mm) than the one in the triaxial cell. Therefore, the generated heat of hydration

will be lost quickly rendering the specimen at room temperature, while specimens tested in the

triaxial cell are considered to hydrate in a relatively closed system that may keep the heat more


efficiently. Having heat preserved, relatively, in the specimen will cause the expansion of the

pore fluid and reduce the generated suctions. Moreover, Barcelo et al. (2005) reported initial

swelling of cement mortars when PC was blended with BFS, which may help explain the

reduction in suction values observed for CPBS55/4.5 specimens. However, volume change

measurements for CPBS55/4.5 and CPBS55/2.2 showed continuous shrinkage and did not reflect

any swelling or expansion. The continuous shrinkage could be a result of the creep of the

specimens under the cell pressure. However, such interpretation is excluded as it does not

explain why specimens with the higher binder content exhibited larger shrinkage (see Figure 4-9)

while it should be more resistant to creep deformations. Finally, although the separate pilot

specimen was sealed in a rubber membrane, the hole created in the base to accommodate the

tensiometer was not sealed. Figure 4-13a shows a schematic diagram to demonstrate how sealing

blocks the air path that could lead to the bottom of the specimen and prevents the formation of

localized dry zones. Figure 4-13b shows the unsealed case where air leaks into the specimen

through a tiny conduit formed at the shaft-pedestal interface (the blue path). As a result, a skin of

dryer material will be formed in vicinity of the ceramic tip and bias the measured suction.

To narrow down the potential explanations to why only limited levels of suction were measured

in the perfectly sealed triaxial cell, two specimens of CPBS55/4.5 and CPBS55/2.2 were tested

under conditions similar to those of the separate pilot CPB specimen. Suction development in

those specimens, labelled as “CPBS55/4.5_separate” and “CPBS55/2.2_separate”, is shown in

Figure 4-14. These specimens exhibited higher levels of suction, which compare well to the

results obtained by Witteman (personal communication 2011) for the same materials. Again,

CPBS55/4.5_separate and CPBS55/2.2_separate as well as specimens tested by Witteman were

not tested in a triaxial cell and the tensiometer hole was not sealed. Therefore, and by excluding

specimen expansion as it was not reflected on volume changes, it is believed that suction

measured out of the triaxial cell was biased by the skin of dryer material formed around the

ceramic tip. In addition, in order to gain a better understanding of the developing suctions and

the corresponding volumetric changes, future work may focus on measuring the total suction in

order to account for the osmotic suctions that may have developed but not captured using the

tensiometer, which records only matric suctions.


4.6 Conclusions

A non-destructive triaxial testing procedure and setup were introduced in order to achieve direct

measurement of the effective stiffness parameters of hydrating CPBS specimens at different

curing ages. In addition, the hydration-induced volumetric changes and self-desiccation suction

were continuously monitored during the testing period in order to understand the mechanisms

contributing to the stiffness development. Several checks were conducted to support the

credibility of the proposed test procedure and setup. The measured stiffness parameters

correspond to an intermediate strain level that is considered representative of a wide range of the

field strains. The stress-strain results for CPBS tested at two binder contents showed that the

material behaved elastically from the very early age at four hours until the test was terminated

after five days. Two specimens of each material composition were tested and their results proved

the reproducibility of the test results. Evolution of stiffness moduli and Poisson‟s ratio confirmed

the reservation on assuming constant values for any stiffness parameter during the early curing

age of CPB. The rate of stiffness gain was found to change at an age that is approximately equal

to the setting time obtained from other tests. Poisson‟s ratios for all specimens followed the same

decreasing pattern regardless of the binder content. In addition, the continuous increase in

stiffness before reaching the onset of the acceleration phase confirms that the binder affects the

resistance and the behavior of materials due cementation reactions during early curing ages

beside altering the fines content, as concluded in Chapter 3.

A framework was introduced to predict the undrained stiffness parameters knowing the effective

stiffness parameter and the degree of saturation. The framework is valid for high degrees of

saturation (> 85%) as the air phase is expected to be in the form of occluded bubbles. Field

evidence for the presence of air as bubbles was provided after examining block specimens

obtained from the stope. The values of Ku and u at 99% saturation were found greater than the

corresponding drained parameters and dramatically decreased as the degree of saturation was

reduced until they approached the values of the effective parameters at about 85% saturation.

The values of Eu showed slight difference from their corresponding effective values, while Gu did

not show any change by changing the degree of saturation.

Although volume changes (shrinkage) were observed for all CPBS specimens over the test

period, insignificant magnitudes of self-desiccation suctions were measured. Therefore, stiffness


gain was considered to be controlled by the bonding hydration reactions. Stiffness loading cycles

were found to disturb and reduce the developing suctions. In addition, the generated heat of

binder hydration might have altered the developing suctions. However, this was not confirmed as

such reductions, or variations, were not reflected on the measured volume changes. Suction

measured without sealing the shaft of the tensiometer was biased by the drying skin of CPBS

formed around the ceramic tip of the tensiometer.

It is believed that this research enables the provision of stiffness parameters at conditions as

close as possible to field conditions. However, further research is required to obtain the complete

set of stiffness parameters at the small strain levels and capture their rate of degradation as strain

increases.


4.7 References

Acker, P. 2004. Swelling, shrinkage and creep: A mechanical approach to cement hydration.

Materials and Structures, 37(268): 237-243.

Addo, K.O. and Robertson, P.K. 1992. Shear-wave velocity measurement of soils using Rayleigh

waves. Canadian Geotechnical Journal, 29(4): 558-568.

Barcelo, L., Moranville, M., and Clavaud, B. 2005. Autogenous shrinkage of concrete: A balance

between autogenous swelling and self-desiccation. Cement and Concrete Research,

35(1): 177-183.

Bates, C.R. 1989. Dynamic soil property measurements during triaxial testing. Geotechnique,

39(4): 721-726.



Boumiz, A., Vernet, C., and Tenoudji, F.C. 1996. Mechanical properties of cement pastes and

mortars at early ages: Evolution with time and degree of hydration. Advanced Cement

Based Materials, 3(3-4): 94-106.

Brown, S.F. and Snaith, M.S. 1974. The measurement of recoverable and irrecoverable

deformations in the repeated load triaxial test. Géotechnique, 24(2): 255-259.

Burland, J.B. and Symes, M. 1982. Simple axial displacement gauge for use in the triaxial

apparatus. Géotechnique, 32: 62-5.

Clayton, C.R.I. 2011. Stiffness at small strain: Research and practice. Geotechnique, 61(1): 5-37.

Clayton, C.R.I. and Heymann, G. 2001. Stiffness of geomaterials at very small strains.

Geotechnique, 51(3): 245-255.

Collins, F. and Sanjayan, J.G. 2001. Microcracking and strength development of alkali activated

slag concrete. Cement and Concrete Composites, 23(4-5): 345-352.

D'Angelo, R., Plona, T.J., Schwartz, L.M., and Coveney, P. 1995. Ultrasonic measurements on

hydrating cement slurries: Onset of shear wave propagation. Advanced Cement Based

Materials, 2(1): 8-14.

Diez d'Aux, M. 2008. Ultrasonic wave measurement through cement paste backfill. MASc.,


Dyvik, R. and Olsen, T.S. 1989. Gmax measured in oedometer and DSS tests using bender

elements. In Proceedings of the International Conference on Soil Mechanics and

Foundation Engineering, Vol. 1, pp. 39-42.

Ferreira, C., da Fonseca, A.V., and Santos, J.A. 2006. Comparison of simultaneous bender

elements and resonant column tests on porto residual soil. In Proceedings of the

Geotechnical Symposium in Rome on Soil Stress-Strain Behavior: Measurement,

Modeling and Analysis, Rome, pp. 523–535.

Fredlund, D.G. 1993. Soil mechanics for unsaturated soils. Wiley, New York.


Fredlund, D.G. 1976. Density and compressibility characteristics of air-water mixtures. Canadian


Gachet, P., Geiser, F., Laloui, L., and Vulliet, L. 2007. Automated digital image processing for

volume change measurement in triaxial cells. Geotechnical Testing Journal, 30(2): 98-

103.

Galaa, A.M., Thompson, B.D., Grabinsky, M.W., and Bawden, W.F. (2011). Characterizing

stiffness development in hydrating mine backfill using ultrasonic wave measurements.

Canadian Geotechnical Journal, 48 (8): 1174-1187

Gasparre, A., Nishimura, S., Minh, N.A., Coop, M.R., and Jardine, R.J. 2007. The stiffness of

natural london clay. Geotechnique, 57(1): 33-47.

Goto, S., Tatsuoka, F., Shibuya, S., Kim, Y., and Sato, T. 1991. Simple gauge for local small

strain measurements in the laboratory. Soils and Foundations, 31(1): 169-180.

Grabinsky, M.W. and and Simms, P. 2005. Matric suction generated by self-desiccation in

cemented paste backfill. In Proceedings of the 9th International Seminar on Paste and

Thickened Tailings, Limerick, 3-7 April 2006, Ireland.

Guan, Y. and Fredlund, D.G. 1997. Use of the tensile strength of water for the direct

measurement of high soil suction. Canadian Geotechnical Journal, 34: 604-614.

Helinski, M., Fourie, A., and Fahey, M. 2007. Assessment of the self-desiccation process in

cemented mine backfills. Canadian Geotechnical Journal, 44(10): 1148-1156.

Heymann, G., Clayton, C.R.I., and Reed, G.T. 2005. Triaxial ultra-small strain measurements

using laser interferometry. Geotechnical Testing Journal, 28(6): 544-552.


Jardine, R.J., Symes, M.J., and Burland, J.B. 1984. Measurement of soil stiffness in the triaxial

apparatus. Geotechnique, 34(3): 323-340.

Kim, J. and Lee, C. 1999. Moisture diffusion of concrete considering self-desiccation at early

ages. Cement and Concrete Research, 29(12): 1921-1927.

Klein, K. and Simon, D. 2006. Effect of specimen composition on the strength development in

cemented paste backfill. Canadian Geotechnical Journal, 43(3): 310-324.

Kokusho, T. 1980. Cyclic triaxial test of dynamic soil properties for wide strain range. Soils and

Foundations, 20(2): 45-60.


N.J.

le Roux, K., Bawden, W.F., and Grabinsky, M.F. 2005. Field properties of cemented paste

backfill at the golden giant mine. Transactions of the Institution of Mining and

Metallurgy, Section A: Mining Technology, 114(2): A65.

Leong, E., Yeo, S., and Rahardjo, H. 2004. Measurement of wave velocities and attenuation

using an ultrasonic test system. Canadian Geotechnical Journal, 41(5): 844-860.

Macari, E.J., Parker, J.K., and Costes, N.C. 1997. Measurement of volume changes in triaxial

tests using digital imaging techniques. Geotechnical Testing Journal, 20(1): 103-109.


Mair, R.J. 1993. Unwin memorial lecture 1992: Developments in geotechnical engineering

research: Application to tunnels and deep excavations. Proceedings - ICE: Civil

Engineering, 97(1): 27-41.

Milne, D., Pakalnis, R., Grant, D., and Sharma, J. 2004. Interpreting hanging wall deformation in

mines. International Journal of Rock Mechanics and Mining Sciences, 41(7): 1139-1151.

Ng, C.W.W., Pun, W.K., and Pang, R.P.L. 2000. Small strain stiffness of natural granitic

saprolite in hong kong. Journal of Geotechnical and Geoenvironmental Engineering,

126(9): 819-833.

Santamarina, J.C., Klein, K.A., and Fam, M.A. 2001. Soils and waves. J. Wiley & Sons, New

York.

Schultheiss, P.J. 1981. Simultaneous measurement of P & S wave velocities during conventional

laboratory soil testing procedures. Marine Geotechnology, 4(4): 343-367.

Schuurman, I.E. 1966. Compressibility of an Air/water mixture and a theoretical relation

between air and water pressures. Geotechnique, 16(4): 269-281.

Seed, H.B. and Idriss, I.M. 1970. Soil moduli and damping factors for dynamic response

analysis. Report EERC 70-10, University of California Berkeley, Berkeley, California.

Simms, P. and Grabinsky, M. 2009. Direct measurement of matric suction in triaxial tests on

early-age cemented paste backfill. Canadian Geotechnical Journal, 46(1): 93.


Canada.

Skempton, A.W. 1954. The pore-pressure coefficients A and B. Geotechnique, 4(4): 143-147.

Sorensen, K.K., Baudet, B.A., and Simpson, B. 2007. Influence of structure on the time-

dependent behaviour of a stiff sedimentary clay. Geotechnique, 57(1): 113-124.

Take, W.A. and Bolton, M.D. 2004. Tensiometer saturation and the reliable measurement of soil

suction. Geotechnique, 54(3): 229-232.

Tarantino, A. and Mongiovi, L. 2001. Experimental procedures and cavitation mechanisms in

tensiometer measurements; unsaturated soil concepts and their application in

geotechnical practice. Geotechnical and Geological Engineering, 19(3-4): 189-210.

Tatsuoka, F. and Shibuya, S. 1992. Deformation characteristics of soils and rocks from field and

laboratory tests. In Proceedings of the 9th Asian Regional Conference on Soil Mechanics

Foundation Engineering, 2: 101-170.

Tatsuoka, F., Teachavorasinskun, S., Dong, J., Kohata, Y., and Sato, T. 1994. Importance of

measuring local strains in cyclic triaxial tests on granular materials. In ASTM Special

Technical Publication, pp. 288-302.


measurements of paste backfill in long-hole stopes. ROCKENG09: Proceedings of


Thottarath, S. 2010. Electromagnetic characterization of cemented paste backfill in the field and

laboratory. MASc, The University of Toronto, Canada.


Tsukamuto, Y., Ishahara, K., Nakazawa, H., Kamada, K., and Huang, Y. 2002. Resistance of

partly saturated sand to liquefaction with reference to longitudinal and shear wave

velocities. Soils and Foundations, 42(6): 93-104.

UMS GmbH München. 2008. T5,T5x pressure transducer tensiometer: User manual. UMS,

Munchen, Germany.

Wheeler, S.J. 1986. The stress-strain behaviour of soils containing gas bubbles. PhD., The

University of Oxford, England, UK.

Witteman, M. and Simms, P. 2010. Hydraulic response in cemented paste backfill during and

after hydration. In Proceedings of the 13th International Conference on Paste and

Thickened Tailings, Toronto, Ontario, May 3rd to 6th 2010, pp. 199-208.

Witteman, M. 2011 (personal communication)


Tables and Figures


Figure 4-1 Stiffness degradation with strain and approximate strain levels for soil structures (after (Mair

1993)).


0.1mm

Load Cell

Tensiometer

O-ring seal

(a)

VCD

Submersible Load Cell

(b)

Modified Triaxial Base

Tensiometer(c)

Figure 4-2 The triaxial setup for stiffness, volume change, and suction measurements. (a) A schematic of the

triaxial setup, (b) A picture of the triaxial cell and VCD, and (c) A picture of the modified triaxial base.


(a)

(b)

Figure 4-3 The T5x miniature tensiometer (a) main components and properties (after UMS GmbH München 2008), and (b) a schematic showing the

dimensions.


-4

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 5 10 15 20 25

vo

lum

e c

han

ge (m

l)

time (hrs)

water#1, deiared for 4 hrs

water#2, deaired for 7 hrs

Figure 4-4 Examining volumetric changes due to leakage and compressibility of cell fluid.


-40

-30

-20

-10

0

10

20

30

40

50

-0.07 -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07

Net

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

(a)

4 hrs

-150

-100

-50

0

50

100

150

-0.08 -0.07 -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

Ne

t De

via

tor

Str

es

s (k

Pa

)

Axial Strain (%)

144 hrs

4 hrs

(b)

Figure 4-5 The resulting behavior from the pilot test. (a) Stress-Strain loops at four hours, (b) comparing stress-strain loops at four hours and one week,

(c) volume change (shrinkage) measured over the testing period, and (d) suction.


-20

-15

-10

-5

0

5

10

15

20

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

(a)

Esec

1

4 Hrs

-20

-15

-10

-5

0

5

10

15

20

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

(b)

4 Hrs

-150

-100

-50

0

50

100

150

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

(c)

122 Hrs

-150

-100

-50

0

50

100

150

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

(d)

126 Hrs

Figure 4-6 Axial stress-strain loops for: (a) CPBS55/4.5_A at four hours, (b) CPBS55/4.5_B at four hours, (c) CPBS55/4.5_A at 122 hours, and (d)

CPBS55/4.5_B at 126 hours.


-15

-10

-5

0

5

10

15

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

(a)

4 Hrs

-15

-10

-5

0

5

10

15

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

(b)

4 Hrs

-80

-60

-40

-20

0

20

40

60

80

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

(c)

120 Hrs

-80

-60

-40

-20

0

20

40

60

80

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

(d)

120 Hrs

Figure 4-7 Axial stress-strain loops for: (a) CPBS55/2.2_A at four hours, (b) CPBS55/2.2_B at four hours, (c) CPBS55/2.2_A at 120 hours, and (d)

CPBS55/2.2_B at 120 hours.


0

50

100

150

200

250

300

350

0 20 40 60 80 100 120 140 160

Eff

ec

tiv

e Y

ou

ng

's M

od

ulu

s, E

(M

Pa

)

Time (hrs)

CPBS55/4.5_A

CPBS55/4.5_B

CPBS55/2.2_A

CPBS55/2.2_B

(a)Onset of accelaration phase for CPBS55/4.5

Onset of accelaration phase for CPBS55/2.2

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40 60 80 100 120 140 160

Eff

ec

tiv

e P

ois

so

n's

ra

tio

(M

Pa

)

Time (hrs)

CPBS55/4.5_A

CPBS55/4.5_B

CPBS55/2.2_A

CPBS55/2.2_B

(b)

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140 160

Eff

ec

tiv

e S

he

ar

mo

du

lus

, G

(M

Pa

)

Time (hrs)

CPBS55/4.5_A

CPBS55/4.5_B

CPBS55/2.2_A

CPBS55/2.2_B

(c)Onset of accelaration phase for CPBS55/2.2


0

50

100

150

200

250

0 20 40 60 80 100 120 140 160

Eff

ec

tiv

e B

ulk

Mo

du

lus

, K

(M

Pa

)

Time (hrs)

CPBS55/4.5_A

CPBS55/4.5_B

CPBS55/2.2_A

CPBS55/2.2_B

(d)Onset of accelaration phase for CPBS55/2.2


Figure 4-8 Effective stiffness parameters measured over five days for CPBS55/4.5 and CPBS55/2.2. (a) Measured Young's modulus, E, (b) measured

Poission's ratio, , (c) calculated shear modulus, G, and (d) calculated bulk modulus, K.


0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

0 20 40 60 80 100 120 140

Sh

rin

ka

ge

(-

V/V

), (

%)

Time (hrs)

CPBS55/4.5_A

CPBS55/4.5_B

CPBS55/2.2_A

CPBS55/2.2_B

stiffness loading



Figure 4-9 Volumetric changes (shrinkage) of CPBS specimens over the test period.


Figure 4-10 Development of self-desiccation suction in CPBS55/4.5, CPBS55/2.2 and the pilot specimen over

the test period.


-3.00E-02

-2.00E-02

-1.00E-02

0.00E+00

1.00E-02

2.00E-02

3.00E-02

4.00E-02

0 200 400 600 800 1000 1200 1400Vo

lum

e S

tra

in (%

)

Time (seconds)

Fifth cycle

Figure 4-11 Volume change during the 20 minutes of stiffness loading at four hours of curing is affected by

the overall shrinkage rate of the specimen. Example shown above is for specimen CPBS55/4.5_B.


0

50

100

150

200

250

300

0 20 40 60 80 100 120

Un

dra

ine

d Y

ou

ng

's M

od

ulu

s (M

Pa

)

Time (hrs)

Es

E (S=0.99)

E (S=0.95)

E (S=0.90)

E (S=0.85)

(a)

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0 20 40 60 80 100 120

Un

dra

ine

d P

ois

so

n's

ra

tio

Time (hrs)

ns

n (S=0.99)

n (S=0.95)

n (S=0.90)

n (S=0.85)

(b)

0

20

40

60

80

100

120

0 20 40 60 80 100 120

Un

dra

ine

d S

he

ar M

od

ulu

s (M

Pa

)

Time (hrs)

Gs

G (S=0.99)

G (S=0.95)

G (S=0.90)

G (S=0.85)

(c)

0

20

40

60

80

100

120

140

160

180

200

0 20 40 60 80 100 120

Un

dra

ine

d B

ulk

Mo

du

lus

(M

Pa

)

Time (hrs)

Ks

K (S=0.99)

K (S=0.95)

K (S=0.90)

K (S=0.85)

(d)

Figure 4-12 Predicted undrained stiffness parameters measured over five days for CPBS55/4.5_B. (a) Undrained Young's modulus, Eu, (b) Undrained

Poission's ratio, u, (c) Undrained shear modulus, Gu, and (d) Undrained bulk modulus, Ku.


O-ring seal

Porous stone

Membrane

CPBS CPBS

Air in Air in

Air path

Drying skin

(a) Sealed Case (b) Unsealed Case

Air path Air path

Figure 4-13 A schematic showing the difference between (a) the sealed tensiometer shaft in the triaxial cell,

and (b) the unsealed tensiometer shaft for separate suction measurements.


Figure 4-14 Suction development in specimens tested separately out of the cell and specimens tested by

Witteman (personal communication).


Chapter 5

Conclusions and Recommendations

The research presented in this thesis focuses on investigating the undrained mechanical behavior

of CPB-sand mixtures under monotonic (static) and cyclic (dynamic) loads. The mechanical

behavior is investigated at the earliest possible age of four hours. After this age, focus is directed

to stiffness gain, aiming at characterizing the evolution of stiffness and obtaining the stiffness

parameters useful for designing and modeling backfilling systems. It is believed that the findings

constitute original contributions to the science and engineering of CPB, particularly when the

mixture contains high proportions of sand. This offers enhanced understanding for mines

adopting cemented paste backfills and consequently leads to safer and more economic designs of

backfilling systems.

5.1 Summary of Findings

5.1.1 Conclusions related to monitoring the consolidation response of the

different mixtures

- The compression index of the tested uncemented MT specimens was less than half the

values reported by Crowder (2004) and Saebimoghaddam (2010) for two different silt-

size gold tailings. This low compressibility of MT is attributed to the original sand

fraction that is higher than that reported in the previous studies.

- The addition of sand caused around 50% to 70% decrease in the compression index

compared with that of MT.

- The addition of a small fraction of binder (≤ 4.5%) caused around 100% increase in the

compression index compared with that of their uncemented sand-containing counterparts.

This conclusion contradicts the reported decrease in compressibility that occurred as a


result for adding binder to silt-size gold tailings in Saebimoghaddam (2010). Such

contradiction could be due the difference in binder type used in each study.

5.1.2 Conclusions stemming from studying the mechanical response of paste

mixtures tested under monotonic loading

- Neither flow nor temporary liquefaction was identified for any of the tested specimens of

MT, MTS and CPBS (cured at four hours) under monotonic compression loading. All

cemented and uncemented specimens exhibited the same type of behavior, which

initiated as contractive and then transformed into dilative upon reaching the phase

transformation state. Despite the substantial difference in compressibility and the initial

void ratio caused by adding sand and/or binder, the type of behavior was not altered.

Moreover, changing the effective confining pressure did not change the behavior.

- Temporary liquefaction was observed for all cemented and uncemented specimens tested

under monotonic extension loading. In contrast to compression, the generated PWP

decreased with the increase in the effective confining pressure indicating higher stability.

- No steady state was identified under monotonic compression and extension loading;

instead, the ultimate failure state was identified at maximum stress obliquity.

- In compression, the addition of sand to MT caused an increase within the order of 2o in

the slope of the failure line in the s'-t' space. Moreover, less dilation was observed at the

higher sand content implying less pipe flow problems.

- In extension, the addition of sand to MT caused a decrease within the order of 2o in the

slope of the failure line in the s'-t' space.

- In compression, the addition of binder caused an increase within the order of 1.5o in the

angles of the failure lines the s´-t space compared with that of their uncemented

counterparts. In addition, no specific relationship was observed between the slopes of the

state lines and the binder content. Furthermore, the rate of dilation for the cemented

specimens was less than that of their uncemented counterparts.


- In extension, the addition of binder caused a decrease in the order of 3o in the angles of

the failure lines in the s´-t space compared with that of their uncemented counterparts.

- The strength of MT in compression was higher than in extension indicating undrained

anisotropy at large strains, which increased by adding sand and increased further by

adding binder. Moreover, the addition of binder showed that regardless of the slight

increase in the obtained strength in compression, the corresponding decrease of strength

in extension was more significant. Considering only the positive impact of binder in

compression can pose serious risk as the obtained ´UFL in extension was around 9o less

than that obtained in compression for cemented specimens.

- The behavior of MT in compression dramatically changed towards a more unstable

behavior in response to reducing the axial strain rate from 2%/min to 0.02%/min.

However, no temporary instability was observed. On the other hand, MTS55 exhibited

negligible change with changing the strain rate. Therefore, adding 55% sand is

considered to have reduced the strain rate dependence observed in the case of MT.

- The drained behavior of MT specimens was different from their undrained behavior

where the drained behavior was totally contractive. Nevertheless, the slope of the failure

line was the same. The response of MTS55 under drained loading was initially

contractive followed by dilative behavior and did not recover the initial volume. The state

lines obtained from drained and undrained tests for MTS55 were essentially the same.

- Comparing the state lines obtained in compression with those obtained in extension in the

e-log p' space did not emphasize the anisotropy observed in the s'-t' space.

- The behavior of all cemented and uncemented specimens was dominated by the fine-

grained matrix. The extent to which the finer matrix dominated the behavior varied

depending on the mixture. The slopes of the state lines in the e-log p’ space and Cc values

for cemented specimens (CPBS) were closer to that of MT, which indicated that the

influence of the finer fraction in the cemented specimens is stronger than in their

uncemented counterparts.


5.1.3 Conclusions drawn from the cyclic triaxial tests

- Cyclic mobility type of response was identified for MT, MT-sand mixtures, and their

cemented counterparts with 4.5% binder. This agrees with the behavior under monotonic

undrained testing where flow liquefaction was not reached for any of the tested

specimens.

- The strain response to stress cycles in extension was identical to that in compression at

the small strain corresponding to initial cycles. As cyclic mobility was approached, larger

strains were generated in extension than those in compression; indicating a weaker

response in extension.

- The strain anisotropy ratio (SAR) was introduced to quantify the anisotropic behavior for

comparing the different pastefill compositions. SAR increased in response to adding

binder to MT or MTS55, which generally agrees with the monotonic testing results.

- The cyclic stress paths of uncemented specimens (MT and MTS55) crossed the failure

lines obtained from monotonic results in compression, whereas cemented specimens

(CPBS55/4.5) crossed the failure lines in compression and extension. This observation

agrees with those reported in the literature for strain hardening and anisotropic materials.

- The addition of sand to MT caused only a slight (9%) reduction in the cyclic resistance.

On the other hand, the addition of 4.5% binder to MT caused 14% increase in the cyclic

resistance for specimens tested after four hours of curing, while the addition of 4.5%

binder to MT-sand mixture caused 60% increase in the cyclic resistance. Therefore, the

binder is considered more effective when added to the pastefill in the presence of sand.

- Specimens tested at seven hours of curing showed that the cyclic resistance continues to

increase as time elapses before reaching the initial set time of the paste, which was

estimated to take place after at least 12 hours of curing.

5.1.4 Conclusions related to stiffness measurements

As a priori, it is to be noted that a non-destructive triaxial testing procedure and setup was

introduced to directly measure the effective stiffness parameters of hydrating CPBS specimens at


different curing ages. In addition, the hydration-induced volumetric changes and self-desiccation

suction were continuously monitored during the testing period to understand the mechanisms

contributing to stiffness development. The main findings are given below.

- The results of several checks and the consistent results support the credibility of the

proposed test procedure and setup in measuring the stiffness parameters corresponding to

an intermediate strain level, which is considered representative of field strains.

- The stress-strain results for CPBS tested at two binder contents demonstrated that the

material behaved elastically from the very early age of four hours until the test was

terminated after five days.

- The stiffness parameters were measured with satisfactory accuracy and the

reproducibility of the results was verified.

- Evolution of stiffness moduli and Poisson‟s ratio confirmed the reservation on assuming

constant values for any stiffness parameter during the early curing of CPB.

- The rate of stiffness gain changed at an age approximately equal to the setting time

obtained from other tests. Poisson‟s ratios for all specimens followed the same decreasing

pattern regardless of the binder content.

- Stiffness of CPBS55/4.5 continuously increased before reaching the initial setting time,

which explains the substantial increase observed in its cyclic resistance at seven hours

compared with that observed at four hours. This also supports the hypothesis that the

binder at very early age affects the resistance and the behaviour of mixtures due its

cementing properties and not only due to altering the fines content.

- A framework was introduced to predict the undrained stiffness parameters knowing the

effective stiffness parameter and the degree of saturation. The framework was based on

the approach proposed by Fredlund (1976) to calculate the compressibility of air-water

mixtures. This framework is valid for high degrees of saturation (> 85%) where the air

phase is likely to exist in the form of occluded bubbles as confirmed by field

observations.


- The values of the undrained parameters Ku and u at 99% saturation were greater than the

corresponding drained parameters and dramatically decreased with the decrease in the

degree of saturation until they approached the values of the effective parameters upon

reaching 85% saturation. Values of Eu showed slight difference from their corresponding

effective values, while Gu did not show any change.

- Finally, it is concluded that stiffness gain during the first few days was considered to be

controlled by the bonding hydration reactions because insignificant magnitudes of self-

desiccation suctions were measured.

5.2 Contributions and Industrial Implications

It is believed that the current thesis constitutes significant contribution in three main directions.

First, it provides deeper insights into the undrained mechanical behavior of a complex cemented

paste backfill-sand mixture under static and dynamic loads. Furthermore, it provides the first

attempt to perform a detailed investigation of the behavior of CPB-sand mixtures. This

investigation demonstrated that the addition of sand rendered the mechanical behavior of the

mixture more sensitive to the change of stress path under monotonic loading than the behavior of

silt-sized CPB and mine tailings tested in previous studies. It was also found that the addition of

binder increased this sensitivity further, which appeared as anisotropy at larger strains. These

findings have important implications on backfilling practices as the strength that CPBS exhibits

under an increasing overburden pressure exerted by an ongoing pour, i.e., monotonic

compression loading, would be different from the strength it exhibits under rock wall closure.

The stress path due to rock wall closure can be close to that of the extension tests. Consequently,

designers of paste backfilling systems should account for the anisotropy at large strains that

increases with the addition of sand and binder. Generally, neither under monotonic loading nor

under cyclic loading has any of the tested specimens exhibited flow liquefaction and the finer

matrix is considered to control the behavior of all specimens even when adding 55% sand to the

mixture. In addition, cyclic testing results showed that the binder is more effective in the

presence of sand which was attributed to the reduced specific surface area and contact points

after adding sand. It was also observed that binder was more effective in increasing the resistance

to cyclic loading than to monotonic loading.


The second major contribution is the proposed experimental framework to determine the

stiffness parameters at representative strains in a non-destructive triaxial test. The proposed

experimental setup represents the first attempt to quantify the complete set of elastic stiffness

parameters for CPB or CPBS, during the first few days of curing. Moreover, the need to assume

any of the stiffness parameters to facilitate the calculations is avoided. The triaxial setup

additionally allows contemporaneous monitoring of the developing suctions and volumetric

changes during the early hydration stages.

The third major contribution is the provision of an advanced understanding of the continuous

evolution of the stiffness of CPB/CPBS during early curing ages. Using the proposed non-

destructive triaxial test, the effective (skeleton) stiffness parameters for CPBS used at Kidd

Creek mine were measured during the first five days of curing. This is a very useful aid in design

and modeling of engineering problems incorporating the strain level at which the parameters

were measured. In addition, a theoretical framework is proposed to predict the undrained

stiffness parameters at similar saturation conditions to the field where the air phase exists as

occluded air bubbles. This research represents a step advance towards a rational design of

cemented paste backfilling systems.

5.3 Recommendations for Future Work

Some recommendations follow from the findings of the current research due to the need to

address some limitations and potential follow-ups. This section presents the recommended future

research work.

Although the specimen preparation procedure followed in the current research replicated the in

situ placement method, the resulting void ratios were lower than those obtained from field

samples. Laboratory specimens had a maximum void ratio of 0.495, while the void ratios

obtained for block CPBS specimens collected from the field were around 0.600. Block

specimens were obtained after demolishing the fill fence and collecting the blocks from behind

it. Further research should be conducted to assess the level of disturbance of the block specimens

collected in this manner and investigate possible reasons behind the difference in the void ratios

obtained from the field and laboratory specimens.


As mentioned earlier in Chapter 4, self-desiccation is an important contributor to the stiffness

gain of CPB. However, there is no standard procedure to measure self-desiccation suction, which

has led to confusion in explaining the obtained suction values. The procedure proposed in the

current research can be standardized for such a purpose after conducting a comparative study

between different combinations of binders and mine tailings. Future work may also focus on

measuring the total suction in order to account for the osmotic suctions that may have developed

but not captured using the tensiometer, which records only matric suctions

Minor improvements can be applied to the triaxial device to extend the strain range at which

stiffness parameters are measured. Obtaining the reference (small strain) stiffness moduli and the

pattern of stiffness degradation with strain from the triaxial test will provide designers with a

complete set of stiffness parameters at different stain levels. Accomplishing this helps in the

selection of appropriate parameters depending on the expected strains for any design problem. In

addition, stiffness parameters obtained from the triaxial test can be compared with those obtained

using other non-destructive tests.

References Abdullah M.G.M.I. Abdelaal, Doctor of Philosophy, 2011 143

References (Combined)

Acker, P. 2004. Swelling, shrinkage and creep: A mechanical approach to cement hydration.

Materials and Structures, 37(268): 237-243.

Addo, K.O. and Robertson, P.K. 1992. Shear-wave velocity measurement of soils using Rayleigh

waves. Canadian Geotechnical Journal, 29(4): 558-568.

Amini, F. and Qi, G.Z. 2000. Liquefaction testing of stratified silty sands. Journal of

Geotechnical and Geoenvironmental Engineering, 126(3): 208-217.





ASTM C136 2006. Standard test method for sieve analysis of fine and coarse aggregates. The

Annual Book of ASTM Standards, Sect 4, Vol. 04.02.

ASTM D854 2006. Standard test methods for specific gravity of soil solids by water pycnometer.

The Annual Book of ASTM Standards, Sect 4, Vol. 04-08.

ASTM D422–63 2007. Standard test method for particle size analysis of soils. The Annual Book

of ASTM Standards, Sect 4, Vol. 04-08.

ASTM D4318–05 2007. Standard test methods for liquid limit, plastic limit, and plasticity index

of soils. The Annual Book of ASTM Standards. Sect 4, Vol. 04-08.

ASTM C127 2007. Standard Test Method for Density, Relative Density (Specific Gravity) and

Absorption of Coarse Aggregate, The Annual Book of ASTM Standards, Sect 4, Vol. 04-

08



Barcelo, L., Moranville, M., and Clavaud, B. 2005. Autogenous shrinkage of concrete: A balance

between autogenous swelling and self-desiccation. Cement and Concrete Research,

35(1): 177-183.

Bates, C.R. 1989. Dynamic soil property measurements during triaxial testing. Geotechnique,

39(4): 721-726.


Corvo mine, Portugal. Institution of Mining and Metallurgy Transactions. Section A:






1133-1144.


Benzaazoua, M., Ouellet, J., Servant, S., Newman, P., and Verburg, R. 1999. Cementitious

backfill with high sulfur content physical, chemical, and mineralogical characterization.

Cement and Concrete Research, 29(5): 719-725.

Boukpeti, N. and Drescher, A. 2000. Triaxial behavior of refined superior sand model.

Computers and Geotechnics, 26(1): 65-81.

Boukpeti, N., Mróz, Z., and Drescher, A. 2002. A model for static liquefaction in triaxial

compression and extension. Canadian Geotechnical Journal, 39(6): 1243-1253.

Boulanger, R.W. and Truman, S.P. 1996. Void redistribution in sand under post-earthquake

loading. Canadian Geotechnical Journal, 33(5): 829-834.




Boumiz, A., Vernet, C., and Tenoudji, F.C. 1996. Mechanical properties of cement pastes and

mortars at early ages: Evolution with time and degree of hydration. Advanced Cement

Based Materials, 3(3-4): 94-106.

Bray, J. and Sancio, R. 2006. Assessment of the liquefaction susceptibility of fine-grained soils.

Journal of Geotechnical and Geoenvironmental Engineering, 132(9): 1165-1177.


Canada.

Brown, S.F. and Snaith, M.S. 1974. The measurement of recoverable and irrecoverable

deformations in the repeated load triaxial test. Géotechnique, 24(2): 255-259.

Burland, J.B. and Symes, M. 1982. Simple axial displacement gauge for use in the triaxial

apparatus. Géotechnique, 32: 62-5.

Chang, N., Yeh, S., and Kaufman, L. 1982. Liquefaction potential of clean and silty sands. In

Proceedings of Third International Earthquake Microzonation Conference , Seattle, USA,

Vol. 2, pp. 1017-1032.

Chern, J. 1985. Undrained response of saturated sands with emphasis on liquefaction and cyclic

mobility. PhD, University of British Columbia, Canada.

Clayton, C.R.I. 2011. Stiffness at small strain: Research and practice. Geotechnique, 61(1): 5-37.

Clayton, C.R.I. and Heymann, G. 2001. Stiffness of geomaterials at very small strains.

Geotechnique, 51(3): 245-255.





Collins, F. and Sanjayan, J.G. 2001. Microcracking and strength development of alkali activated

slag concrete. Cement and Concrete Composites, 23(4-5): 345-352.

D'Angelo, R., Plona, T.J., Schwartz, L.M., and Coveney, P. 1995. Ultrasonic measurements on

hydrating cement slurries: Onset of shear wave propagation. Advanced Cement Based

Materials, 2(1): 8-14.


Dezfulian, H. 1984. Effects of silt content on dynamic properties of sandy silts. In Proc., 8th

World Conf. on Earthquake Engineering, Prentice-Hall, Englewood Cliffs, N.J., pp. 63-

70.

Diez d'Aux, M. 2008. Ultrasonic wave measurement through cement paste backfill. MASc.,


Dobry, R. and Alvarez, L. 1967. Seismic failures of Chilean tailings dams. J. Soil Mech.and

Found.Div., ASCE, 93(6): 237-260.

El-Mamlouk, H.H., Hassan, A.M., and Hussein, A.K. 2008. Undrained cyclic behavior of

nonplastic silty sand. Journal of Engineering and Applied Science, 55(5): 379-399.

Dyvik, R. and Olsen, T.S. 1989. Gmax measured in oedometer and DSS tests using bender

elements. In Proceedings of the International Conference on Soil Mechanics and

Foundation Engineering, Vol. 1, pp. 39-42.

Ferreira, C., da Fonseca, A.V., and Santos, J.A. 2006. Comparison of simultaneous bender

elements and resonant column tests on porto residual soil. In Proceedings of the

Geotechnical Symposium in Rome on Soil Stress-Strain Behavior: Measurement,

Modeling and Analysis, Rome, pp. 523–535.

Finn, W.D.L., Ledbetter, R.H., and and Guoxi, W. 1994. Liquefaction in silty soils: Design and

analysis. Ground Failures Under Seismic Conditions, Geotechnical Special Publication,

No. 44, ASCE. pp. 51-76.

Fourie, A.B. and Papageorgiou, G. 2001. Defining an appropriate steady state line for


Fourie, A.B., Blight, G.E., and Papageorgiou, G. 2001. Static liquefaction as a possible

explanation for the merriespruit tailings dam failure. Canadian Geotechnical Journal,

38(4): 707-719.

Fredlund, D.G. 1993. Soil mechanics for unsaturated soils. Wiley, New York.

Fredlund, D.G. 1976. Density and compressibility characteristics of air-water mixtures. Canadian


Gachet, P., Geiser, F., Laloui, L., and Vulliet, L. 2007. Automated digital image processing for

volume change measurement in triaxial cells. Geotechnical Testing Journal, 30(2): 98-

103.

Galaa, A.M., Thompson, B.D., Grabinsky, M.W., and Bawden, W.F. (2011). Characterizing

stiffness development in hydrating mine backfill using ultrasonic wave measurements.

Canadian Geotechnical Journal,48 (8): 1174-1187.

Gasparre, A., Nishimura, S., Minh, N.A., Coop, M.R., and Jardine, R.J. 2007. The stiffness of

natural london clay. Geotechnique, 57(1): 33-47.

Goto, S., Tatsuoka, F., Shibuya, S., Kim, Y., and Sato, T. 1991. Simple gauge for local small

strain measurements in the laboratory. Soils and Foundations, 31(1): 169-180.

Grabinsky, M.W. and and Simms, P. 2005. Matric suction generated by self-desiccation in

cemented paste backfill. In Proceedings of the 9th International Seminar on Paste and

Thickened Tailings, Limerick, 3-7 April 2006, Ireland.


Guan, Y. and Fredlund, D.G. 1997. Use of the tensile strength of water for the direct

measurement of high soil suction. Canadian Geotechnical Journal, 34: 604-614.

Garga, V.K. and and McKay, L.D. 1984. Cyclic triaxial strength of mine tailings. Journal of


Helinski, M., Fourie, A., and Fahey, M. 2007. Assessment of the self-desiccation process in

cemented mine backfills. Canadian Geotechnical Journal, 44(10): 1148-1156.

Heymann, G., Clayton, C.R.I., and Reed, G.T. 2005. Triaxial ultra-small strain measurements

using laser interferometry. Geotechnical Testing Journal, 28(6): 544-552.

Hyde, A.F.L., Higuchi, T., and Yasuhara, K. 2006. Liquefacation, cyclic mobility, and failure of

silt. Journal of Geotechnical and Geoenvironmental Engineering, 132(6): 716-735.

Ishihara, K. 1980. Cyclic strength characteristics of tailings materials. Soils and Foundations,

20(4): 127-142.

Ishihara, K. 1993. Liquefaction and flow failure during earthquakes. Geotechnique, 43(3): 351-

415.


Jamiolkowski, M., Kongsukprasert, L., and Lo Presti, D. 2005. Characterization of gravelly

geomaterials. In Proceedings of The Fifth International Geotechnical Conference, Cairo,

Egypt, 12 January 2005. pp. 1–27.

Jardine, R.J., Symes, M.J., and Burland, J.B. 1984. Measurement of soil stiffness in the triaxial

apparatus. Geotechnique, 34(3): 323-340.




Kim, J. and Lee, C. 1999. Moisture diffusion of concrete considering self-desiccation at early

ages. Cement and Concrete Research, 29(12): 1921-1927.

Klein, K. and Simon, D. 2006. Effect of specimen composition on the strength development in

cemented paste backfill. Canadian Geotechnical Journal, 43(3): 310-324.

Koester, J. 1994. Influence of fines type and content on cyclic strength. Ground Failures Under

Seismic Conditions, Geotechnical Special Publication, No. 44, ASCE. pp. 17-33.

Kokusho, T. 1980. Cyclic triaxial test of dynamic soil properties for wide strain range. Soils and

Foundations, 20(2): 45-60.


N.J.


undrained response of sand. In Hydraulic fill structures, Geotechnical. Special

Publication No. 21, ASCE. pp. 330-345.

Lade, P.V. and Yamamuro, J.A. 1997. Effects of nonplastic fines on static liquefaction of sands.

Canadian Geotechnical Journal, 34(6): 918-928.




le Roux, K., Bawden, W.F., and Grabinsky, M.F. 2005. Field properties of cemented paste

backfill at the golden giant mine. Transactions of the Institution of Mining and

Metallurgy, Section A: Mining Technology, 114(2): A65.

Leong, E., Yeo, S., and Rahardjo, H. 2004. Measurement of wave velocities and attenuation

using an ultrasonic test system. Canadian Geotechnical Journal, 41(5): 844-860.

Macari, E.J., Parker, J.K., and Costes, N.C. 1997. Measurement of volume changes in triaxial

tests using digital imaging techniques. Geotechnical Testing Journal, 20(1): 103-109.

Mair, R.J. 1993. Unwin memorial lecture 1992: Developments in geotechnical engineering

research: Application to tunnels and deep excavations. Proceedings - ICE: Civil

Engineering, 97(1): 27-41.

Milne, D., Pakalnis, R., Grant, D., and Sharma, J. 2004. Interpreting hanging wall deformation in

mines. International Journal of Rock Mechanics and Mining Sciences, 41(7): 1139-1151.

Ng, C.W.W., Pun, W.K., and Pang, R.P.L. 2000. Small strain stiffness of natural granitic

saprolite in hong kong. Journal of Geotechnical and Geoenvironmental Engineering,

126(9): 819-833.

Okusa, S., Anma, S., and and Maikuma, H. 1980. Liquefaction of mine tailings in the 1978 izu-

oshima-kinkai earthquake, central Japan. In Proc., 7th World Conf. on Earthquake

Engrg., 3: 89-96.

Pierce, M., Bawden, W.F., and Paynter, J.T. (1998) Laboratory testing and stability analysis of

paste backfill at the golden giant mine. In Minefill ‟98: Proceedings of the 6th

International Symposium on Mining with Backfill, Brisbane, pp. 159–165.

Pitman, T.D. 1994. Influence of fines on the collapse of loose sands. Canadian Geotechnical

Journal, 31(5): 728-739. Polito, C.P. and Martin, J.R. 2001. Effects of nonplastic fines on

the liquefaction resistance of sands. Journal of Geotechnical and Geoenvironmental

Engineering, 127(5): 408-415.

Riemer, M.F., Gookin, W.B., Bray, J.D., and Arango, I. 1994. Effects of loading frequency and

control on the liquefaction behavior of clean sands. Geotechnical Engineering Rep. No.

UCB/GT/94-07, Univ. of California, Berkeley, Calif.

Riemer, M. and Seed, R., 1997. Factors affecting apparent position of steady-state line. Journal

of Geotechnical and Geoenvironmental Engineering, 123: 281-8.


of Toronto, Canada.

Sanin, M.V. and Wijewickreme, D. 2006. Cyclic shear response of channel-fill fraser river delta

silt. Soil Dynamics and Earthquake Engineering, 26(9): 854-869.

Santamarina, J.C., Klein, K.A., and Fam, M.A. 2001. Soils and waves. J. Wiley & Sons, New

York.

Schultheiss, P.J. 1981. Simultaneous measurement of P & S wave velocities during conventional

laboratory soil testing procedures. Marine Geotechnology, 4(4): 343-367.

Schuurman, I.E. 1966. Compressibility of an Air/water mixture and a theoretical relation

between air and water pressures. Geotechnique, 16(4): 269-281.


Seed, H.B. and Idriss, I.M. 1970. Soil moduli and damping factors for dynamic response

analysis. Report EERC 70-10, University of California Berkeley, Berkeley, California.

Simms, P. and Grabinsky, M. 2009. Direct measurement of matric suction in triaxial tests on

early-age cemented paste backfill. Canadian Geotechnical Journal, 46(1): 93.


Canada.

Singh, S. 1994. Liquefaction characteristics of silts. Ground Failures Under Seismic Conditions,

Geotechnical Special Publication, No. 44, ASCE. pp. 105-116.

Skempton, A.W. 1954. The pore-pressure coefficients A and B. Geotechnique, 4(4): 143-147.

Sladen, J.A., D'Hollander, R.D., Krahn, J., and Mitchell, D.E. 1985. Back analysis of the Nerlerk

berm liquefaction slides. Canadian geotechnical journal, 22(4): 579-588.

Sorensen, K.K., Baudet, B.A., and Simpson, B. 2007. Influence of structure on the time-

dependent behaviour of a stiff sedimentary clay. Geotechnique, 57(1): 113-124.

Take, W.A. and Bolton, M.D. 2004. Tensiometer saturation and the reliable measurement of soil

suction. Geotechnique, 54(3): 229-232.

Tarantino, A. and Mongiovi, L. 2001. Experimental procedures and cavitation mechanisms in

tensiometer measurements; unsaturated soil concepts and their application in

geotechnical practice. Geotechnical and Geological Engineering, 19(3-4): 189-210.

Tatsuoka, F. and Shibuya, S. 1992. Deformation characteristics of soils and rocks from field and

laboratory tests. In Proceedings of the 9th Asian Regional Conference on Soil Mechanics

Foundation Engineering, 2: 101-170.

Tatsuoka, F., Teachavorasinskun, S., Dong, J., Kohata, Y., and Sato, T. 1994. Importance of

measuring local strains in cyclic triaxial tests on granular materials. In ASTM Special

Technical Publication, pp. 288-302.

Thevanayagam, S. 1998. Effects of fines and confining stress on undrained shear strength of silty

sands. Journal of Geotechnical and Geoenvironmental Engineering, 124(6): 479.

Thevanayagam, S. , Mohan, S. 2000. Intergranular state variables and stress-strain behaviour of

silty sands. Géotechnique, 50(1): 1-23.



Engineering, 128(10): 849-859.

Thomas, J. 1992. Static, cyclic and post liquefaction undrained behaviour of fraser river sand.

MASc, The University of British Columbia, Canada.


measurements of paste backfill in long-hole stopes. ROCKENG09: Proceedings of


Thottarath, S. 2010. Electromagnetic characterization of cemented paste backfill in the field and

laboratory. MASc, The University of Toronto, Canada.


Troncoso, J.H. and Verdugo, R. 1985. Silt content and dynamic behavior of tailing sands. In

Proc., XI Int. Conf. on Soil Mechanics and Foundation Engineering, Vol. 2, pp. 1311–

1314.

Tsukamuto, Y., Ishahara, K., Nakazawa, H., Kamada, K., and Huang, Y. 2002. Resistance of

partly saturated sand to liquefaction with reference to longitudinal and shear wave

velocities. Soils and Foundations, 42(6): 93-104.

UMS GmbH München. 2008. T5,T5x pressure transducer tensiometer: User manual. UMS,

Munchen, Germany.

Vaid, Y.P. 1994. Liquefaction of silty soils. Ground Failures Under Seismic Conditions,

Geotechnical Special Publication No. 44, ASCE. pp. 1-16.

Vaid, Y.P. and Thomas, J. 1995. Liquefaction and postliquefaction behavior of sand. Journal of


Vaid, Y.P. and Chern, J.C. 1985. Cyclic and monotonic undrained response of saturated sands.

Advances in the Art of Testing Soils Under Cyclic Conditions: 120-147.

Vaid, Y.P., Stedman, J.D., and Sivathayalan, S. 2001. Confining stress and static shear effects in

cyclic liquefaction. Canadian Geotechnical Journal, 38(3): 580-591.

Wheeler, S.J. 1986. The stress-strain behaviour of soils containing gas bubbles. PhD., The

University of Oxford, England, UK.

Wickland, B.E., Wilson, G.W., Wijewickreme, D. and Klein, B. 2006. Design and Evaluation of

Mixtures of Mine Waste Rock and Tailings. Canadian Geotechnical Journal. Vol 43(9)

pp. 928-945.

Wijewickreme, D., Khalili, A., and Ward Wilson, G. 2010. Mechanical response of highly gap-

graded mixtures of waste rock and tailings. part II: Undrained cyclic and post-cyclic

shear response. Canadian Geotechnical Journal, 47(5): 566-582.

Wijewickreme, D., Sanin, M.V., and Greenaway, G.R. 2005. Cyclic shear response of fine-

grained mine tailings. Canadian Geotechnical Journal, 42(5): 1408-1421.

Witteman, M. and Simms, P. 2010. Hydraulic response in cemented paste backfill during and

after hydration. In Proceedings of the 13th International Conference on Paste and

Thickened Tailings, Toronto, Ontario, May 3rd to 6th 2010, pp. 199-208.

Witteman, M. 2011 (personal communication).

Xenaki, V.C., Athanasopoulos, G.A. 2003. Liquefaction resistance of sand-silt mixtures: An

experimental investigation of the effect of fines. Soil Dynamics and Earthquake

Engineering, 23 (3): 183-194

Yamamuro, J.A. and Lade, P.V. 1997. Static liquefaction of very loose sands. Canadian

geotechnical journal, 34(6): 905-917.

Yamamuro, J.A. and Lade, P.V. 1998. Steady-state concepts and static liquefaction of silty


Yamamuro, J.A. and Covert, K.M. 2001. Monotonic and cyclic liquefaction of very loose sands

with high silt content. Journal of Geotechnical and Geoenvironmental Engineering,

127(4): 314-324.





Appendix A Abdullah M.G.M.I. Abdelaal, Doctor of Philosophy, 2011 150

Appendix A

Cyclic Testing Results


-20

-10

0

10

20

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 2 4 6 8 10 12 14 16 18

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-30

-20

-10

0

10

20

30

-12 -10 -8 -6 -4 -2 0 2 4 6 8

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-12

-10

-8

-6

-4

-2

0

2

4

6

8

0 2 4 6 8 10 12 14 16 18

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure A- 1 Typical response of cyclically loaded MT specimens ('c=100 kPa and CSR=0.12): (a) Stress paths with failure lines obtained from


cycles.


-20

-10

0

10

20

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4 5 6 7 8 9

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-40

-30

-20

-10

0

10

20

30

40

-15 -10 -5 0 5 10

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-14

-12

-10

-8

-6

-4

-2

0

2

4

6

8

10

0 1 2 3 4 5 6 7 8 9

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure A- 2 Typical response of cyclically loaded MT specimens ('c=100 kPa and CSR=0.15): (a) Stress paths with failure lines obtained from


cycles.


-20

-10

0

10

20

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 2 4 6 8 10 12 14

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-30

-20

-10

0

10

20

30

-8 -6 -4 -2 0 2 4 6

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-8

-6

-4

-2

0

2

4

6

0 2 4 6 8 10 12 14

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure A- 3 Typical response of cyclically loaded MTS55 specimens ('c=100 kPa and CSR=0.12): (a) Stress paths with failure lines obtained from


cycles.


-20

-10

0

10

20

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4 5 6 7

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-40

-30

-20

-10

0

10

20

30

40

-7 -6 -5 -4 -3 -2 -1 0 1 2 3 4

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-8

-6

-4

-2

0

2

4

0 1 2 3 4 5 6 7

Ax

ial s

tra

in,

%

Number of cycles

(d)



cycles.


-20

-10

0

10

20

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 0.5 1 1.5 2 2.5 3 3.5 4

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-50

-40

-30

-20

-10

0

10

20

30

40

50

-10 -8 -6 -4 -2 0 2 4 6

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-10

-8

-6

-4

-2

0

2

4

6

0 0.5 1 1.5 2 2.5 3 3.5 4

Ax

ial s

tra

in,

%

Number of cycles

(d)



cycles.


-20

-10

0

10

20

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 2 4 6 8 10 12 14 16 18

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-30

-20

-10

0

10

20

30

-12 -10 -8 -6 -4 -2 0 2 4

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-12

-10

-8

-6

-4

-2

0

2

4

0 2 4 6 8 10 12 14 16 18

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure A- 6 Typical response of cyclically loaded CPB/4.5 specimens ('c=100 kPa and CSR=0.13): (a) Stress paths with failure lines obtained from


cycles.


-20

-10

0

10

20

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4 5 6 7 8

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-40

-30

-20

-10

0

10

20

30

40

-10 -8 -6 -4 -2 0 2 4

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-40

-30

-20

-10

0

10

20

30

40

-10 -8 -6 -4 -2 0 2 4

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)



cycles.


-20

-10

0

10

20

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 0.5 1 1.5 2 2.5 3 3.5 4

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-50

-40

-30

-20

-10

0

10

20

30

40

50

-12 -10 -8 -6 -4 -2 0 2 4 6 8

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-12

-10

-8

-6

-4

-2

0

2

4

6

8

0 0.5 1 1.5 2 2.5 3 3.5 4

Ax

ial s

tra

in,

%

Number of cycles

(d)



cycles.


-20

-10

0

10

20

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 50 100 150 200 250 300 350

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-30

-20

-10

0

10

20

30

-0.07 -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-2

0

2

0 50 100 150 200 250 300 350

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure A- 9 Typical response of cyclically loaded CPBS55/4.5 specimens ('c=100 kPa and CSR=0.11): (a) Stress paths with failure lines obtained from


cycles.


-30

-20

-10

0

10

20

30

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4 5 6 7 8 9 10

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-50

-40

-30

-20

-10

0

10

20

30

40

50

-6 -5 -4 -3 -2 -1 0 1 2 3

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-6

-4

-2

0

2

4

0 1 2 3 4 5 6 7 8 9 10

Ax

ial s

tra

in,

%

Number of cycles

(d)



cycles.


-30

-20

-10

0

10

20

30

0 20 40 60 80 100 120 140

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4 5 6

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-50

-40

-30

-20

-10

0

10

20

30

40

50

-8 -6 -4 -2 0 2 4

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-8

-6

-4

-2

0

2

4

0 1 2 3 4 5 6

Ax

ial s

tra

in,

%

Number of cycles

(d)



cycles.


-30

-20

-10

0

10

20

30

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4 5 6

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-60

-50

-40

-30

-20

-10

0

10

20

30

40

50

60

-10 -8 -6 -4 -2 0 2 4 6

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-10

-8

-6

-4

-2

0

2

4

6

0 1 2 3 4 5 6

Ax

ial s

tra

in,

%

Number of cycles

(d)



cycles.


-30

-20

-10

0

10

20

30

0 20 40 60 80 100 120

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4 5 6 7

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-50

-40

-30

-20

-10

0

10

20

30

40

50

-15 -10 -5 0 5 10

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-14

-12

-10

-8

-6

-4

-2

0

2

4

6

8

0 1 2 3 4 5 6 7

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure A- 13 Typical response of cyclically loaded CPB/4.5 specimens at seven hours of curing ('c=100 kPa and CSR=0.20): (a) Stress paths with

failure lines obtained from monotonic testing, (b) stress strain behavior, (c) excess pore water pressure ratio (ru) versus number of cycles, and (d) axial

strain versus number of cycles.


-40

-30

-20

-10

0

10

20

30

40

0 20 40 60 80 100 120 140

(σ' 1

-σ' 3

)/2

, kP

a

(σ'1+σ'3)/2, kPa

(a)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 5 10 15 20 25 30 35 40 45 50

Po

re p

res

su

re r

ati

o,

(∆u

/σ' 3

)

Number of cycles

(c)

-60

-50

-40

-30

-20

-10

0

10

20

30

40

50

60

-6 -5 -4 -3 -2 -1 0 1 2 3 4

(σ' 1

-σ' 3

), k

Pa

Axial strain. %

(b)

-6

-4

-2

0

2

4

0 5 10 15 20 25 30 35 40 45 50

Ax

ial s

tra

in,

%

Number of cycles

(d)

Figure A- 14 Typical response of cyclically loaded CPBS55/4.5 specimens at seven hours of curing ('c=100 kPa and CSR=0.25): (a) Stress paths with

failure lines obtained from monotonic testing, (b) stress strain behavior, (c) excess pore water pressure ratio (ru) versus number of cycles, and (d) axial

strain versus number of cycles.

Appendix B Abdullah M.G.M.I. Abdelaal, Doctor of Philosophy, 2011 165

Appendix B

Axial Stress-Strain Loops for Stiffness

Measurements


Axial stress-strain loops for specimen CPBS55/4.5_A

-40

-30

-20

-10

0

10

20

30

40

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

8hrs: CPBS55/4.5_A

-50

-40

-30

-20

-10

0

10

20

30

40

50

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

12hrs: CPBS55/4.5_A


-80

-60

-40

-20

0

20

40

60

80

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

18.5hrs: CPBS55/4.5_A

-100

-80

-60

-40

-20

0

20

40

60

80

100

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

26.5hrs: CPBS55/4.5_A


-100

-80

-60

-40

-20

0

20

40

60

80

100

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

42hrs: CPBS55/4.5_A

-150

-100

-50

0

50

100

150

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

72hrs: CPBS55/4.5_A


-150

-100

-50

0

50

100

150

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

100hrs: CPBS55/4.5_A

Axial stress-strain loops for specimen CPBS55/4.5_B

-30

-20

-10

0

10

20

30

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

8hrs: CPBS55/4.5_B


-50

-40

-30

-20

-10

0

10

20

30

40

50

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

14.5hrs: CPBS55/4.5_B

-60

-40

-20

0

20

40

60

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

21.5hrs: CPBS55/4.5_B


-100

-80

-60

-40

-20

0

20

40

60

80

100

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

31.5hrs: CPBS55/4.5_B

-100

-80

-60

-40

-20

0

20

40

60

80

100

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

50hrs: CPBS55/4.5_B


-100

-80

-60

-40

-20

0

20

40

60

80

100

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

72hrs: CPBS55/4.5_B

-120

-100

-80

-60

-40

-20

0

20

40

60

80

100

120

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

99hrs: CPBS55/4.5_B


Axial stress-strain loops for specimen CPBS55/2.2_A

-30

-20

-10

0

10

20

30

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

8hrs: CPBS55/4.5_B

-20

-15

-10

-5

0

5

10

15

20

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

13hrs: CPBS55/2.2_A


-25

-20

-15

-10

-5

0

5

10

15

20

25

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

22hrs: CPBS55/2.2_A

-25

-20

-15

-10

-5

0

5

10

15

20

25

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

31hrs: CPBS55/2.2_A


-30

-20

-10

0

10

20

30

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

46hrs: CPBS55/2.2_A

-50

-40

-30

-20

-10

0

10

20

30

40

50

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

70hrs: CPBS55/2.2_A


-80

-60

-40

-20

0

20

40

60

80

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

96hrs: CPBS55/2.2_A

Axial stress-strain loops for specimen CPBS55/2.2_B

-20

-15

-10

-5

0

5

10

15

20

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

8hrs: CPBS55/2.2_B


-25

-20

-15

-10

-5

0

5

10

15

20

25

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

13.5hrs: CPBS55/2.2_B

-25

-20

-15

-10

-5

0

5

10

15

20

25

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

24hrs: CPBS55/2.2_B


-35

-25

-15

-5

5

15

25

35

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

32hrs: CPBS55/2.2_B

-40

-30

-20

-10

0

10

20

30

40

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

48hrs: CPBS55/2.2_B


-60

-40

-20

0

20

40

60

-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

74.5hrs: CPBS55/2.2_B

-80

-60

-40

-20

0

20

40

60

80

-0.05 -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04

Devia

tor

Str

ess (

kP

a)

Axial Strain (%)

97hrs: CPBS55/2.2_B

Appendix C Abdullah M.G.M.I. Abdelaal, Doctor of Philosophy, 2011 180

Appendix C

Predicted Undrained Stiffness Parameters


0

50

100

150

200

250

300

350

0 20 40 60 80 100 120

Un

dra

ine

d Y

ou

ng

's M

od

ulu

s (M

Pa

)

Time (hrs)

Es

E (S=0.99)

E (S=0.95)

E (S=0.90)

E (S=0.85)

(a)

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40 60 80 100 120

Un

dra

ine

d P

ois

so

n's

ra

tio

Time (hrs)

ns

n (S=0.99)

n (S=0.95)

n (S=0.90)

n (S=0.85)

(b)


0

20

40

60

80

100

120

140

0 20 40 60 80 100 120

Un

dra

ine

d S

he

ar M

od

ulu

s (M

Pa

)

Time (hrs)

Gs

G (S=0.99)

G (S=0.95)

G (S=0.90)

G (S=0.85)

(c)

0

50

100

150

200

250

0 20 40 60 80 100 120

Un

dra

ine

d B

ulk

Mo

du

lus

(M

Pa

)

Time (hrs)

Ks

K (S=0.99)

K (S=0.95)

K (S=0.90)

K (S=0.85)

(d)

Figure C- 1 Predicted undrained stiffness parameters measured over five days for CPBS55/4.5_A. (a)

Undrained Young's modulus, Eu, (b) Undrained Poission's ratio, u, (c) Undrained shear modulus, Gu, and (d)

Undrained bulk modulus, Ku.


0

20

40

60

80

100

120

140

160

0 20 40 60 80 100 120

Un

dra

ine

d Y

ou

ng

's M

od

ulu

s (M

Pa

)

Time (hrs)

Es

E (S=0.99)

E (S=0.95)

E (S=0.90)

E (S=0.85)

(a)

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0 20 40 60 80 100 120

Un

dra

ine

d P

ois

so

n's

ra

tio

Time (hrs)

ns

n (S=0.99)

n (S=0.95)

n (S=0.90)

n (S=0.85)

(b)


0

10

20

30

40

50

60

0 20 40 60 80 100 120

Un

dra

ine

d S

he

ar M

od

ulu

s (M

Pa

)

Time (hrs)

Gs

G (S=0.99)

G (S=0.95)

G (S=0.90)

G (S=0.85)

(c)

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120

Un

dra

ine

d B

ulk

Mo

du

lus

(M

Pa

)

Time (hrs)

Ks

K (S=0.99)

K (S=0.95)

K (S=0.90)

K (S=0.85)

(d)

Figure C- 2 Predicted undrained stiffness parameters measured over five days for CPBS55/2.2_A. (a)




0

20

40

60

80

100

120

140

160

0 20 40 60 80 100 120

Un

dra

ine

d Y

ou

ng

's M

od

ulu

s (M

Pa

)

Time (hrs)

Es

E (S=0.99)

E (S=0.95)

E (S=0.90)

E (S=0.85)

(a)

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 20 40 60 80 100 120

Un

dra

ine

d P

ois

so

n's

ra

tio

Time (hrs)

ns

n (S=0.99)

n (S=0.95)

n (S=0.90)

n (S=0.85)

(b)


0

10

20

30

40

50

60

0 20 40 60 80 100 120

Un

dra

ine

d S

he

ar M

od

ulu

s (M

Pa

)

Time (hrs)

Gs

G (S=0.99)

G (S=0.95)

G (S=0.90)

G (S=0.85)

(c)

0

20

40

60

80

100

120

0 20 40 60 80 100 120

Un

dra

ine

d B

ulk

Mo

du

lus

(M

Pa

)

Time (hrs)

Ks

K (S=0.99)

K (S=0.95)

K (S=0.90)

K (S=0.85)

(d)

Figure C- 3 Predicted undrained stiffness parameters measured over five days for CPBS55/2.2_B. (a)



Early Age Mechanical Behavior and Stiffness Development of … · 2013-09-27 · ii Early Age...

Documents

Transcript of Early Age Mechanical Behavior and Stiffness Development of … · 2013-09-27 · ii Early Age...