0% found this document useful (0 votes)
13 views399 pages

Theses1 1608

The thesis by Sina Anzanpour investigates the load transfer mechanisms of ground tendons under static and dynamic loading conditions, focusing on cable bolts used in underground excavations. It includes experimental and numerical studies to assess the performance of different cable bolts under various loading scenarios, revealing significant differences in behavior between static and dynamic conditions. The research aims to enhance the understanding of tendon behavior, which is crucial for designing effective reinforcement systems in civil and mining engineering applications.

Uploaded by

Mahmoud Essam
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
13 views399 pages

Theses1 1608

The thesis by Sina Anzanpour investigates the load transfer mechanisms of ground tendons under static and dynamic loading conditions, focusing on cable bolts used in underground excavations. It includes experimental and numerical studies to assess the performance of different cable bolts under various loading scenarios, revealing significant differences in behavior between static and dynamic conditions. The research aims to enhance the understanding of tendon behavior, which is crucial for designing effective reinforcement systems in civil and mining engineering applications.

Uploaded by

Mahmoud Essam
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 399

University of Wollongong - Research Online

Thesis Collection

Title: The Load Transfer Mechanism of Ground Tendons Subjected To Static and Dynamic Loading
Conditions

Author: Sina Anzanpour

Year: 2022

Repository DOI:

Copyright Warning
You may print or download ONE copy of this document for the purpose of your own research or study. The
University does not authorise you to copy, communicate or otherwise make available electronically to any
other person any copyright material contained on this site.
You are reminded of the following: This work is copyright. Apart from any use permitted under the Copyright
Act 1968, no part of this work may be reproduced by any process, nor may any other exclusive right be
exercised, without the permission of the author. Copyright owners are entitled to take legal action against
persons who infringe their copyright. A reproduction of material that is protected by copyright may be a
copyright infringement. A court may impose penalties and award damages in relation to offences and
infringements relating to copyright material.
Higher penalties may apply, and higher damages may be awarded, for offences and infringements involving
the conversion of material into digital or electronic form.

Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily
represent the views of the University of Wollongong.

Research Online is the open access repository for the University of Wollongong. For further information
contact the UOW Library: research-pubs@uow.edu.au
University of Wollongong
Research Online

University of Wollongong Thesis Collection University of Wollongong Thesis Collections


2017+

2022

The Load Transfer Mechanism of Ground Tendons Subjected To Static and


Dynamic Loading Conditions
Sina Anzanpour
University of Wollongong

Follow this and additional works at: https://ro.uow.edu.au/theses1

University of Wollongong
Copyright Warning
You may print or download ONE copy of this document for the purpose of your own research or study. The University
does not authorise you to copy, communicate or otherwise make available electronically to any other person any
copyright material contained on this site.
You are reminded of the following: This work is copyright. Apart from any use permitted under the Copyright Act
1968, no part of this work may be reproduced by any process, nor may any other exclusive right be exercised,
without the permission of the author. Copyright owners are entitled to take legal action against persons who infringe
their copyright. A reproduction of material that is protected by copyright may be a copyright infringement. A court
may impose penalties and award damages in relation to offences and infringements relating to copyright material.
Higher penalties may apply, and higher damages may be awarded, for offences and infringements involving the
conversion of material into digital or electronic form.

Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily
represent the views of the University of Wollongong.

Recommended Citation
Anzanpour, Sina, The Load Transfer Mechanism of Ground Tendons Subjected To Static and Dynamic
Loading Conditions, Doctor of Philosophy thesis, School of Civil, Mining and Environmental Engineering,
University of Wollongong, 2022. https://ro.uow.edu.au/theses1/1608

Research Online is the open access institutional repository for the University of Wollongong. For further information
contact the UOW Library: research-pubs@uow.edu.au
The Load Transfer Mechanism of Ground Tendons Subjected To
Static and Dynamic Loading Conditions

Sina Anzanpour

Supervisors:
Prof. Naj Aziz
A/Prof. Ali Mirzaghorbanali
Dr. Jan Nemcik

This thesis is presented in fulfillment of the requirement for the conferral of the degree:
Doctor of Philosophy

University of Wollongong
School of Civil, Mining, and Environmental Engineering

December 2022
Certification

I, Sina Anzanpour, declare that this thesis submitted in fulfillment of the requirements
for the conferral of the degree of Doctor of Philosophy, from the University of
Wollongong is wholly my own work unless otherwise referenced or acknowledged.
This document has not been submitted for qualifications at any other academic
institution.

Sina Anzanpour
27st of November, 2022
List of Publications

Anzanpour, S., Aziz, N. Mirzaghorbanali, A., Nemcik, J., & Rastegarmanesh, A. (2023).
The behaviour of cable bolts under axial and shear loading conditions. World Mining
Congress, June 2023, Australia.

Khaleghparast, S., Aziz, N., Remennikov, A., & Anzanpour, S. (2023). An experimental
study on shear behaviour of fully grouted rock bolt under static and dynamic loading
conditions. Tunnelling and Underground Space Technology, 132, 104915.
https://doi.org/10.1016/j.tust.2022.104915

Aziz, N., Mirzaghorbanali, A., Anzanpour, S., Rastegarmanesh, A., Khaleghparat, S.,
Nemcik, J., Oh, J., & Si, G. (2022). Angle Shear Testing of 15.2 mm Seven Wire Cable
Bolt. Rock Mechanics and Rock Engineering, 55(7), 3919–3937.
https://doi.org/10.1007/s00603-022-02847-2

Anzanpour, S., Aziz, N., Mirzaghorbanali, A., Nemcik, J. & Rastegarmanesh, A. (2023).
The behaviour of cable bolts under axial and shear loading conditions. ISRM International
Symposium- Eurock 2022, September 2022, Finland.

Anzanpour, S., Aziz, N. Nemcik, J. Remennikov, A. Mirzaghorbanali, A., & Wallace, J.


(2022). Static and dynamic tendon pullout test research at University of Wollongong.
Resource Operators Conference (ROC), P182-189, February 2022, Australia.
https://ro.uow.edu.au/coal/844/

Rastegarmanesh, A., Mirzaghorbanali, A., McDougall, K., Aziz, N., Anzanpour, S.,
Nourizadeh, H., & Moosavi, M. (2022). Axial Performance of Cementitious Grouted Cable
Bolts Under Rotation Constraint Scenarios. Rock Mechanics and Rock Engineering, 55(9),
5773–5788. https://doi.org/10.1007/s00603-022-02950-4

Anzanpour, S., Aziz, N., Nemcik, J., Mirzaghorbanali, A., & Wallace, J. (2021).
Introduction to new methods of static and dynamic pull testing of rock bolts and cable bolts.
Resource Operators Conference (ROC), February 2021, 210–217.
https://ro.uow.edu.au/coal/811/
Aziz, N., Anzanpour, S., Khaleghparat, S., Rastegarmanesh, A., Mirzaghorbanali, A.,
Remmenikov, A., Nemcik, J., Oh, J., & Si, G. (2021). Angled shear testing of 15.2 mm
seven wire cable bolt. Resource Operators Conference, February 2019, 18–20.
https://ro.uow.edu.au/coal/814/

Khaleghparast, S., Anzanpour, S., Aziz, N., Remennikov, A., & Mirzaghorbanali, A.
(2020). Static and dynamic testing of tendons. ISRM International Symposium-Eurock
2020, Jun 2020, Norway. https://onepetro.org/ISRMEUROCK/proceedings-
abstract/EUROCK20/All-EUROCK20/ISRM-EUROCK-2020-214/451174
Acknowledgments

“I can no other answer make, but thanks, and thanks.”– William Shakespeare

I would like to express my deepest appreciation to my supervisors. Words cannot express my


appreciation to Professor Naj Aziz, who has always been not just a supervisor, but a close
friend, a father like figure, and a wise team leader with a wealth of knowledge on tendon
technology. Studying under his supervision and in his research team, was my main motivation
for finishing my master's degree in engineering in 2012. I would also like to extend my deepest
gratitude to my second supervisor, A/prof. Ali Mirzaghorbanali, who was a source of never-
ending technological support throughout this project. I must also thank Dr. Jan Nemcik and
Prof. Alex Remennikov for generously lending me their knowledge and supervision. I am
extremely grateful to a brilliant intellectual and scholar, Dr. Ameen Topa, whose mentorship
and unparalleled knowledge of numerical modeling was crucial to the making of this
dissertation.

I would also like to thank the technical officers of the Department Civil, Mining,
Environmental and Architectural Engineering, Faculty of Engineering and Information
Sciences, UOW, the engines of the faculty who are always behind the high-quality generated
laboratory data. The safety of the students, testing preparation, manufacturing, and
maintenance of the testing facilities are only a few examples of their benevolent, full-time
support. Among the team, I would like to give my special thanks to Mr. Travis Marshall, Mr.
Jordan Wallace, Mr. Cameron Neilson and Mr. Duncan Best. I also wish to thank a colleague
and best friend, Dr. Saman Khaleghparast, for his support through his insightful suggestions
and technical assistance. Finally, I would like to acknowledge a colleague at the University of
Southern Queensland, Dr. Ashkan Rastegarmanesh, for a great and memorable four-year
cooperation and friendship.

Thanks should also go to my brother, Mr. Hamed Anzanpour, for guidance and counseling in
every step of my life. To my beautiful sister, Hediyeh Anzanpour, for all the emotional support
through this journey, and my friend for life, Ms. Solmaz Niagansafi, for cheering me on and
encouraging me to see life after graduation.

Finally, yet importantly, I would like to express my sincere appreciation to my beloved parents
for their unconditional love, support, and encouragement. My lovely mother, Taji Ghiassedin,
and the greatest father ever, Shoja Anzanpour, all these could not be even initiated without
your warm support. Herby, I would love to dedicate this thesis to you two beautiful souls
because I would love to see the endless smile on your faces.
COVID-19 Impact Statement

(2020-2022)

Most of the thesis was based on manufacturing new laboratory equipment and implementing
laboratory experiments. Shipping of the equipment parts was significantly delayed due to
shipping restrictions induced by COVID-19. Limited access to the workshop, restrictions on
the working hours of technical officers and researchers, and limited access to the laboratory
facilities for the experiments during the COVID-19 directly affected the timing and number of
experiments.
Abstract

Underground excavations such as tunnels, ground slopes, unstable rock walls, and cliffs are
some examples where rock bolts and cable bolts can be used to safeguard human life and
equipment. Rock bolts and cable bolts (so-called tendons) are tension members that pull the
loose parts of the ground together and anchor them to the nearest intact and stable ground.
Reinforcement of sagging bedding layers and strata, supporting the immediate roof of a tunnel
in a discontinuous rock, and stabilizing the rock slopes that have the potential to topple can be
some of the routine applications of ground anchors. However, in practice, either tensile or
shear, or the combination of both tensile and shear loads, might be applied to the tendons. The
performance of the ground system is highly dependent on the ground conditions as well as
installation properties. Thus, it is paramount to understand the exact loading conditions of
tendons and propose a proper design for each application according to the requirements of each
specific application.

A variety of analytical, experimental, and numerical research have been carried out to illustrate
the failure mechanism of tendons under axial and shear loads. Axial loading capacity and
failure modes of tendons are examined by in-field or laboratory pullout tests. Laboratory
single-shear or double-shear tests also aim to characterise the behaviour of tendons under shear
loading conditions. Angled shear tests are for cases where both axial and shear loads are applied
simultaneously. While the number of studies on axial loading is significantly high, studies on
angled and pure shear are limited. This stems from the fact that shear studies can be merely
undertaken in the laboratory and require special powerful testing facilities, while pullout tests
can be done both in the field and in the laboratory.

Although most ground anchors are designed for long-term loading conditions such as
gravitational settlement and deformation, local earthquakes such as blasting in mines,
rockburst, coal burst, and global earthquakes apply significantly high dynamic load to the
tendons in very short periods of time. Hence, attention must be paid to the charachterisation of
the tendons' behaviour under dynamic loading as well as static loading. Field measurements of
the dynamic load applied on tendons, the development of mathematical and numerical models,
and laboratory tests are some of the study methods for dynamic loading conditions. Various
laboratory technologies have been developed for dynamic pullout testing of the tendon in the
laboratory. This is while dynamic shear loading tests are not as common as axial loading. This
even becomes less common with dynamic shear testing of tendons at angles. Besides, a
comparative study of both static and dynamic loading in axial and shear loading conditions
there was little else be seen in the literature review of this topic. It is also worth mentioning
that the technology of tendons has developed and is yet to develop according to the needs of
the industry. The number of studies on rock bolts is considerably larger as it is a common
reinforcement system among both civil and mining engineers. This is while studies on cable
bolts are relatively limited. Therefore, continuous research and examination of new tendons
are inevitable.

This thesis aims to investigate different possible loading conditions applied to cable bolts. To
this, two different cable bolts (15.2 mm seven-wire and 63 t nine-wire Sumo cable bolts) were
subjected to different loading conditions, including static and dynamic perpendicular shear,
static angled shear, and static and dynamic pullout tests. The thesis is divided into two main
sections.

The first section covers the testing procedure and results of shear tests. Static perpendicular
and angled tests were undertaken in the completion of former studies carried out by the same
research team at the University of Wollongong. While dynamic perpendicular shear testing of
cable bolts was done for the first time in this study. Two different testing setups, including
Cylindrical Double-Shear Box (also named MK-IV) and an angled double-shear setup, were
employed for this study. Test results showed that the great flexibility of cable bolts causes the
bending of the tendons and creation of hinge points in the shear surfaces before failure. Results
indicated that tensile failure of wires is more common than shear failure even though the
installation was perpendicular to the loading direction. More propensity for the tensile failure
of wires could be observed in the angled shear samples. The ultimate shear load of the cable
bolt was roughly 40% higher in the angled cable bolts compared with those installed
perpendicularly to the loading direction; however, the allowed displacement before failure in
the angled shear sample was significantly smaller. In other words, the energy absorption of the
cable in static loading is constant. The ultimate shear load and absorbed energy of the cable
bolt in the dynamic shear test were around 50% higher than in the perpendicular static tests.

The second section commences with the introduction, test sample preparation, and verification
of the new testing facility for static and dynamic pullout tests. The Static and Dynamic Pullout
test Apparatus (SDPA) was designed and manufactured as part of this thesis in order to provide
the opportunity for a comparative study of tendons under different loading conditions. In other
words, a comprehensive study of the tendon under different loading conditions could be
undertaken under one roof, which makes the comparisons more precise and reliable. The
performance of the SDPA was verified by Steel Split Setup, which is used for pullout testing
of short-encapsulated tendons. Results of both static and dynamic pullout tests showed that,
unlike shear tests, the pullout resistance of the cable bolt in dynamic tests was approximately
50% lower than in the static tests.

A numerical model was generated using ANSYS/LS-DYNA to simulate the exact contribution
of cable bolts in the static pullout tests. The model was verified by the experimental results and
later used for the parametric studies and sensitivity analysis. Also, a simplified version of the
model was proposed by adjusting the mechanical parameters to reduce the runtime while
preserving the quality of the results. Parameters such as embedment length were studied
through numerical simulation. This is while implementation of this parameter in the laboratory
was impractical, inefficient, and expensive.

All in all, it is worth mentioning that all the test results have been obtained in the laboratory
and under idealized testing conditions with the required simplifications; however, the outcome
of this study can assist mining and geotechnical industries with a better understanding of the
behaviour of tendons under different loading conditions.
List of Names or Abbreviations

Acronym Complete word


ACARP Australian Coal Industry’s Research Program
ACI American Concrete Institute
ANSYS Analysis Systems
AMSJ Australian Mine Safety Journal
AS Australian Standards
ASTM American Standard for Testing Materials
AUD Australian Dollar
B&W Barrel and Wedge
B/G Bulbed Cable Encapsulated with Cementitious Grout
B/R Bulbed Cable Encapsulated with Resin
BEM Boundary Element Method
BS British Standard
BSM Bond Strength Model
CANMET Canada Centre for Mineral and Energy Technology
CMC Continuously Mechanically Coupled
CME Civil, Mining, and Environment
DEM Discrete Element Method
FDM Finite Different Method
FEM Finite Element Method
FLAC Fast Lagrangian Analysis of Continua
IGES Initial Graphic Exchange Specification
LIVM Impedance Voltage Mode
P/G Plain Cable Encapsulated with Cementitious Grout
P/R Plain Cable Encapsulated with Resin
PFC Particle Flow Code
SDPA Static and Dynamic Pullout Testing Apparatus
SSS Steel Split Sets
UCS Uniaxial Compressive Strength
UDEC Universal Distinct Element Code
UNSW University of New South Wales
UOW University of Wollongong
USQ University of Southern Queensland
UTS Ultimate Tensile Strength
VDC Volts of Direct Current
W/C ratio Water Cement ratio
WASM Western Australia School of mines
Table of Contents

Introduction .................................................................................................... 1
Rationale...................................................................................................................... 1

Background of the Study ............................................................................................. 2

Motivations of the Study ............................................................................................. 4

Key Objectives ............................................................................................................ 5

Methodology ............................................................................................................... 5

Thesis Structure ........................................................................................................... 6

Scope ........................................................................................................................... 8

Literature Review of Axial and Shear Loading of Tendons ...................... 9


Introduction ................................................................................................................. 9

Ground Anchor Components .................................................................................... 15

2.2.1 Rock Bolts and Cable Bolts ............................................................................... 15

2.2.2 Grout .................................................................................................................. 19

Development of Laboratory Pullout Testing Technologies over Time..................... 20

2.3.1 Previous Static Pullout Testing Apparatus ........................................................ 20

2.3.2 Previous Dynamic Pullout Testing Apparatus ................................................... 34

Pullout Mechanism.................................................................................................... 43

Analytical Studies ..................................................................................................... 50

Numerical Studies ..................................................................................................... 63

Parametric Studies ..................................................................................................... 69

2.7.1 Tendons Properties (Cable bolt and rock bolt) .................................................. 69

2.7.2 Confinement Medium ........................................................................................ 75

2.7.3 Embedment Length ............................................................................................ 81

2.7.4 Grout Properties (cement or resin grouts) .......................................................... 85

2.7.5 Loading Condition ............................................................................................. 89

Review of the Recent Improvements in Shear Testing of Tendons .......................... 92


Summary ................................................................................................................. 101

Laboratory Study on Static and Dynamic Shear Loading of cable bolts


104
Introduction ............................................................................................................. 104

Mechanisms and Preparation of Perpendicular Double Shear Test ........................ 106

3.2.1 Sample Preparation for Perpendicular Double Shear Test .............................. 111

Mechanism and Preparation of Angled Shear Test ................................................. 115

3.3.1 Sample Preparation for Angled Double Shear Test ......................................... 116

Selection of Testing Materials ................................................................................ 120

3.4.1 Cementitious Grout .......................................................................................... 120

3.4.2 Mastic Resin..................................................................................................... 129

3.4.3 Concrete Blocks ............................................................................................... 131

3.4.4 Cable Bolts ....................................................................................................... 134

Summary ................................................................................................................. 136

Results and Analysis of Static and Dynamic Double Shear Tests ......... 139
Introduction ............................................................................................................. 139

Experimental Plan ................................................................................................... 140

Perpendicular Static Shear Test Results .................................................................. 141

4.3.1 Tests Observations ........................................................................................... 144

Angled Static Shear Test Results ............................................................................ 145

4.4.1 Tests Observations ........................................................................................... 148

Dynamic Shear Test Results ................................................................................... 150

4.5.1 Tests Observations ........................................................................................... 154

4.5.2 Trial Dynamic Shear Test on High-Strength Cable Bolts ............................... 157

Summary ................................................................................................................. 159

Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)


161
Introduction ............................................................................................................. 161
Description of Static and Dynamic Pullout Test Apparatus ................................... 162

Rig Assembly and Experimental Procedures .......................................................... 169

Laboratory Testing Machines, Facilities, and Measurement Tools ........................ 175

5.4.1 Displacement and Strain Measurements .......................................................... 177

5.4.2 Accelerometer .................................................................................................. 178

5.4.3 High-Speed Camera ......................................................................................... 178

5.4.4 Dynamic Drop Hammer ................................................................................... 179

Apparatus Modifications ......................................................................................... 181

5.5.1 Concrete Confinement ..................................................................................... 181

5.5.2 Anti-Rotation System....................................................................................... 185

SDPA Performance Verification ............................................................................. 187

Summary ................................................................................................................. 190

Results and Analysis of Axial Static and Dynamic Pullout tests ........... 192
Introduction ............................................................................................................. 192

Experimental Program............................................................................................. 192

Static Test Results of 15.2 mm Cable Bolts ............................................................ 195

6.3.1 Tests Observations ........................................................................................... 198

Static Test Results of 63 t Sumo Cable Bolts ......................................................... 201

Tests Observations .................................................................................................. 205

Dynamic Pullout Test Results ................................................................................. 212

6.6.1 Tests Observations ........................................................................................... 215

Summary ................................................................................................................. 221

Numerical Modeling for Pullout Loading of Cable Bolts ....................... 223


Introduction ............................................................................................................. 223

Proposing of the Detailed Pullout Test Model ........................................................ 225

7.2.1 Solver Selection ............................................................................................... 225

7.2.2 Geometry Generation ....................................................................................... 227

7.2.3 Mesh Generation .............................................................................................. 228


7.2.4 Boundary Conditions ....................................................................................... 229

7.2.5 Contacts and Interactions ................................................................................. 230

7.2.6 Mass Scaling .................................................................................................... 233

7.2.7 Material Properties ........................................................................................... 233

7.2.8 Outputs ............................................................................................................. 241

Numerical Results ................................................................................................... 242

Proposing a Simplified Pullout Test Model ............................................................ 245

Summary ................................................................................................................. 249

Discussion on Axial and Shear Loading of Cable Bolts .......................... 250


Introduction ............................................................................................................. 250

Shear Tests Analysis ............................................................................................... 251

8.2.1 Failure Mechanism of Wires in a Shear Test ................................................... 251

8.2.2 Effect of Cable Geometry on the Shear Behaviour ......................................... 253

8.2.3 Effect of Grout Strength on Shear Load Capacity ........................................... 253

8.2.4 Effect of Static and Dynamic Loading Rate on Shear Load Capacity ............. 254

8.2.5 Effect of Angle of Installation on Shear Load Capacity .................................. 256

Pullout Tests Analysis ............................................................................................. 258

8.3.1 Failure Mechanism of Encapsulation Material ................................................ 259

8.3.2 Parametric Study .............................................................................................. 270

Combination of Shear and Pullout Loading ............................................................ 280

Summary ................................................................................................................. 282

Conclusions and Recommendations ......................................................... 286


Introduction ............................................................................................................. 286

Behaviour of the tendons in shear tests ................................................................... 287

Behaviour of the tendon in pullout (axial) tests ...................................................... 287

Numerical analysis .................................................................................................. 289

Recommendations ................................................................................................... 291

References ............................................................................................................................. 292


A: Appendices .................................................................................................................... 318
A.1 Appendices of Chapter 4 ......................................................................................... 318

A.1.1 A: Static perpendicular shear test results ......................................................... 319

A.1.2 B: Static angled shear test results..................................................................... 322

A.1.3 C: Dynamic perpendicular shear test results .................................................... 323

A.2 Appendices of Chapter 5 ......................................................................................... 326

A.2.1 Detailed drawings of the designed Static and Dynamic Pullout test Apparatus
326

A.3 Appendices of Chapter 6 ......................................................................................... 341

A.3.1 A: Static pullout test results of 15.2 mm cable bolt ......................................... 341

A.3.2 B: Static pullout test results of 63 t Sumo cable bolt ....................................... 350

A.3.3 C: Dynamic pullout test results of 63 t Sumo cable bolt ................................. 355

A.4 Appendices of Chapter 7 ......................................................................................... 361

A.4.1 Input parameter of materials and contacts ....................................................... 361

A.4.2 Verification of grout material .......................................................................... 362

A.4.3 Simulation of 15.2 mm cable bolt tensile test (using ANSYS) ....................... 362

A.4.4 Detailed pullout test model results ................................................................... 363

A.4.5 Low friction-deformable cable ........................................................................ 364

A.4.6 High friction- deformable cable ....................................................................... 364

A.4.7 Proposed simplified model .............................................................................. 365

A.5 Appendices of Chapter 8 ......................................................................................... 365


List of Tables

Table 2-1: Summary of apparatus and test methodologies used in the literature to study the
axial performance of bolts ....................................................................................................... 22
Table 2-2: Summary of apparatus and test methodologies used in the literature to study the
axial performance of bolts under dynamic loads ..................................................................... 36
Table 2-3: Effect of birdcage and epoxy coating on pullout load............................................ 70
Table 2-4: Cementitious grout with different additives for pullout test, after (Benmokrane et
al., 1995) .................................................................................................................................. 88
Table 3-1: Grout properties used in the numerical model ..................................................... 129
Table 3-2: Mechanical properties of Sandstone and Shale in Australia, after (Bertuzzi & Pells,
2002) ...................................................................................................................................... 132
Table 3-3: Results of UCS tests of concrete samples ............................................................ 134
Table 3-4: Properties of cable bolts used in this study .......................................................... 135
Table 4-1: Experimental test plan for static and dynamic shear tests .................................... 141
Table 4-2: Static perpendicular shear test results of 15.2 mm cable bolt .............................. 142
Table 4-3: Static perpendicular shear test results of 63 t Sumo cable bolt ............................ 142
Table 4-4: Static angled shear test results .............................................................................. 145
Table 4-5: Dynamic perpendicular shear test results ............................................................. 151
Table 5-1: Pullout test result carried out by SPDA and SSS ................................................. 189
Table 6-1: Experimental test plan for static and dynamic pullout tests ................................. 194
Table 6-2: Static pullout test results of 15.2 mm cable bolt .................................................. 195
Table 6-3: Static pullout test result of 63 t Sumo cable bolt encapsulated in cement grout and
resin ........................................................................................................................................ 202
Table 6-4: Maximum deformation of the B&W in different loading conditions. ................. 209
Table 6-5: Dynamic pullout test results of 63 t Sumo cable bolt........................................... 213
Table 7-1: Implicit and explicit solver boundary ................................................................... 226
Table 7-2: Properties of contact between cable bolt and encapsulation material .................. 233
Table 7-3: Features of the material cards for concrete (Adopted from (LS-DYNA, 2018)) . 234
Table 7-4: Input parameters of material keyword for grout and resin ................................... 239
Table 7-5: Input parameters of elastic material for steel ....................................................... 241
Table 7-6: Input parameters of contact keyword for grout .................................................... 247
Table 8-1: recorded load and deformation of B&W in static and dynamic tests ................... 267
Table 8-2: Cable bolt elongation measurements in static and dynamic pullout tests ............ 272
Table A-1: Experimental test plan for static and dynamic shear tests ................................... 318
Table A-2: Static pullout test results of 15.2 mm cable bolt ................................................. 341
Table A-3: Static pullout test result of 63 t Sumo cable bolt encapsulated in cement grout and
resin ........................................................................................................................................ 350
Table A-4: Dynamic pullout test results of 15.2 mm cable bolt ............................................ 355
List of Figures

Figure 1-1: Different loading modes on ground tendons (Barley & Windsor, 2000)................ 3
Figure 1-2: Classification of studies around ground tendons .................................................... 4
Figure 2-1 Ruckburst in Australian mine................................................................................. 10
Figure 2-2: General classification of pullout test experiments ................................................ 12
Figure 2-3: Different types of specimens used in debonding test (after (C. Yan, 1992) ......... 13
Figure 2-4: Bonded Anchor Failure Modes (adapted from (Droesch, 2015)) ......................... 13
Figure 2-5 Possible failure modes of the cable bolting system (Jeremic and Delaire, 1983) .. 14
Figure 2-6: Ground support system with rock bolts(S. Yan et al., 2019) ................................ 16
Figure 2-7: Main components of cable bolts ........................................................................... 16
Figure 2-8: Common modifications on cable bolt profile, adopted from (Barley & Windsor,
2000; C. Windsor, 1992).......................................................................................................... 17
Figure 2-9: Some of the registered patents for modification of cable bolts (Eaton et al., 1998;
Hedrick, 2005; Stankus, 2001)................................................................................................. 18
Figure 2-10: Different bolt types in terms of physical features, left to right: Rebar, Bulbed
cable, plain cable, hot-dip galvanized steel cable, epoxy coated steel cable (Satola & Hakala,
2001) ........................................................................................................................................ 18
Figure 2-11: Effect of w/c ratio on UCS strength (J. Chen, 2016) ......................................... 19
Figure 2-12: Calculated load and strain gauges on the king-wire of the 15.2 mm cable bolt (L.
A. Martin et al., 2000) .............................................................................................................. 44
Figure 2-13: Pullout mechanism of 15.2 mm cable bolts, Left: plain cable bolt and cement grout
(Hyett et al., 1995), Right: P=plain, B=Bulbed cable bolt and cement grout (Aoki et al., 2002)
.................................................................................................................................................. 45
Figure 2-14: Typical pullout mechanism of FRP rock bolts (Thenevin et al., 2017) .............. 45
Figure 2-15: Typical pullout mechanism of Reflex cable bolts with different embedment
lengths (Left) and confinement pressure (Right) (Thenevin et al., 2017) ............................... 45
Figure 2-16: Comparison of different failure modes (X. Li, 2019) ......................................... 46
Figure 2-17: Pullout mechanism of uncoated POLYSTAL (Mah, 1994) ................................ 47
Figure 2-18: Performance of MW9 cable bolt in pullout test (J. Chen, 2016) ........................ 48
Figure 2-19: Pullout behaviour of FRP bars and cable bolts encapsulated in resin (L. B. Martin,
2012) ........................................................................................................................................ 49
Figure 2-20: Idealized bond-slip model (Benmokrane et al., 1995) ........................................ 52
Figure 2-21: (left) Relationship between shear stress and strain of soils, (Right) Schematic
diagram of shear failure phases (Xiao & Chen, 2008)............................................................. 53
Figure 2-22: Tri-linear law with characteristic parameters (Z. Wu et al., 2010) ..................... 54
Figure 2-23: Slippage model of (L. B. Martin, 2012).............................................................. 54
Figure 2-24: Bilinear local bond-slip model (Yuan et al., 2004) ............................................. 55
Figure 2-25: Typical full-range theoretical load-displacement curve...................................... 56
Figure 2-26: Trilinear bond-slip model (Ren et al., 2010) ....................................................... 57
Figure 2-27: Evolution of shear stress. (a, b) Elastic stage; (c, d) elastic–softening stage; (e, f)
elastic– softening–debonding stage; (g) softening–debonding stage; (h, i) debonding stage; I,
II and III represent elastic, softening and debonding stress state, respectively (Ren et al., 2010)
.................................................................................................................................................. 58
Figure 2-28: Five stages to debonding (J. Chen et al., 2015) .................................................. 59
Figure 2-29: Distribution of shear stress along a fully grouted rock bolt subjected to an axial
load (C. C. Li & Stillborg, 1999) ............................................................................................. 61
Figure 2-30: 2D FEM model for pullout test of short bolts (Meloni et al., 2013) ................... 64
Figure 2-31: Rock bolt element with only shear coupling spring (Ma et al., 2016) ................ 65
Figure 2-32: Comparison between numerical and laboratory results on plain cable installed in
(left) weak confining medium, (right) strong confining medium (J. Chen, 2016) .................. 66
Figure 2-33: Free-body diagram of the model (Mirarco et al., 2018; St-pierre, 2007) ........... 66
Figure 2-34: Numerical and experimental model of anchor bolt (Pitrakkos et al., 2010) ....... 68
Figure 2-35: Full-scale FEM model of dynamic pullout test (Tahmasebinia et al., 2021) ...... 69
Figure 2-36: Effect of tendon properties on pullout load capacity (Satola, 2007) .................. 70
Figure 2-37: Effect of cable surface on pullout load (Aziz et al., 2016) ................................. 71
Figure 2-38: Effect of indentation on pullout load capacity (X. Li, 2019) .............................. 72
Figure 2-39: Comparison between birdcaged and standard cable bolts (Hutchins et al., 1990)
.................................................................................................................................................. 72
Figure 2-40: Effect of cable properties on pullout test results (Hagan & Li, 2017) ................ 73
Figure 2-41: Comparison of strain in bulbed and plain cables during the pullout test (Forbes &
Vlachopoulos, 2016) ................................................................................................................ 75
Figure 2-42: Effect of confinement size on pullout test results (Rajaie, 1990) ....................... 77
Figure 2-43: Effect of sample size on pullout load (J. Chen, 2016) ........................................ 77
Figure 2-44: Effect of confinement pressure on pullout load tests using two different w/c ratios
for grout (left: w/c: 0.3, right: w/c: 0.4) (MacSporran, 1993) ................................................. 78
Figure 2-45: Effect of confining pressure on the bond capacity (Moosavi et al., 2005) ......... 78
Figure 2-46: Effect of confinement pressure on pullout test (Moosavi et al., 2005) ............... 79
Figure 2-47: Effect of active and passive confinement on the test results (Thomas, 2012) .... 80
Figure 2-48: Effect of top confinement on pullout test results (Hagan, 2004) ........................ 81
Figure 2-49: Effect of embedment length on the pullout behaviour of cable bolt (left), and rock
bolt (right) (Benmokrane et al., 1995) ..................................................................................... 81
Figure 2-50: Results of pullout test with 1000 mm (left) and 2000 mm embedment lengths
(Satola & Hakala, 2001) .......................................................................................................... 82
Figure 2-51: Effect of embedment length of pullout load (Martin, 2012) ............................... 83
Figure 2-52: Critical length for the effect of embedment length (Thompson & Villaescusa,
2014) ........................................................................................................................................ 83
Figure 2-53: Effect of embedment length on the maximum applied pullout load (Ma et al.,
2016) ........................................................................................................................................ 84
Figure 2-54: Effect of confinement pressure on the pullout results (J. Chen, 2016) ............... 85
Figure 2-55: Effect of grout UCS on the bond strength (Kilic, Yasar, & Atis, 2002) ............. 85
Figure 2-56: Correlation between pullout force and compressive strength (Meloni et al., 2013)
.................................................................................................................................................. 86
Figure 2-57: Effect of w/c and confining pressure on pullout test results (Hagan & Li, 2017)
.................................................................................................................................................. 87
Figure 2-58: Effect of grout type on pullout test results (Farah & Aref, 1986) ....................... 88
Figure 2-59: Effect of different additives of grout on pullout behaviour (Benmokrane et al.,
1995) ........................................................................................................................................ 88
Figure 2-60: Comparison of resin (left) and cement grout (right) in pullout tests (Pullan &
Hagan, 2018) ............................................................................................................................ 89
Figure 2-61: Effect of loading rate on the pullout test results (Hagan & Li, 2017)................. 90
Figure 2-62: Absorbed energy and applied load in dynamic pullout test of rock bolt (Thompson
et al., 2015) .............................................................................................................................. 91
Figure 2-63: Absorbed energy and applied load in dynamic pullout test of cable bolt
(Villaescusa et al., 2007) .......................................................................................................... 92
Figure 2-64: Typical load and displacement graph of the dynamic test (Crompton et al., 2018)
.................................................................................................................................................. 92
Figure 2-65: Proposed angle double shear testing setup design (G. Yang, 2019) ................... 93
Figure 2-66: Numerical model of dynamic shear testing of rock bolts (Khaleghparast, 2021)
.................................................................................................................................................. 94
Figure 2-67: Numerical sketch model of combined pull and shear test of rock bolt (Saadat,
2019) ........................................................................................................................................ 95
Figure 2-68: Effect of pullout load on normal load on the rock joint (Saadat, 2019) ............. 96
Figure 2-69: Equivalent shear test in each stage of the pullout load ....................................... 96
Figure 2-70: Distribution of compression and tension load on the rock joint and tendon during
a combined pull-shear test (Saadat, 2019) ............................................................................... 97
Figure 2-71: Single shear impact testing setup for rock bolts, (a) diagram, (b)view, and (c)
loading diagram (Pytlik, 2020) ................................................................................................ 97
Figure 2-72: Multi-purpose tendon testing setup (Kang et al., 2020) ...................................... 98
Figure 2-73: Shear behaviour of rock bolts in angled shear test combined with pullout test
(Kang et al., 2020) ................................................................................................................... 99
Figure 2-74: Dynamic double shear models generated with FEM software (ABAQUS)
(Tahmasebinia et al., 2018a, 2021) .......................................................................................... 99
Figure 2-75: Comparison of double shear experiment, analytical and numerical models
(Tahmasebinia et al., 2021) .................................................................................................... 100
Figure 2-76: Combined loading setup (Pinazzi et al., 2021) ................................................. 100
Figure 3-1: Mechanism of the perpendicular double shear test ............................................. 107
Figure 3-2: Cylindrical double shear test apparatus (MK-IV) ............................................... 108
Figure 3-3: Prepared mould for casting concrete samples ..................................................... 112
Figure 3-4: Double shear test assembly procedure ................................................................ 112
Figure 3-5: Double shear test assembly procedure ................................................................ 113
Figure 3-6: Static (left) and dynamic (right) double shear test setups ready for the test ....... 114
Figure 3-7: Test sample after the test ..................................................................................... 115
Figure 3-8 Mechanism of the angled double shear test ......................................................... 116
Figure 3-9: Design of 30 and 45-degree angle double shear tests ......................................... 116
Figure 3-10: Concrete samples before and after casting ........................................................ 117
Figure 3-11: Angle shear test preparation process (Part A) ................................................... 117
Figure 3-12: Angle shear test preparation process (Part B) ................................................... 118
Figure 3-13: Angle shear test preparation process (Part C) ................................................... 118
Figure 3-14: Applying and measuring pretension load.......................................................... 119
Figure 3-15: Testing machine and instrumentation facilities ................................................ 119
Figure 3-16: Grout samples poured in 50 mm cubic moulds................................................. 121
Figure 3-17: Comparison of UCS in different types of grout ................................................ 121
Figure 3-18: Left: 50 mm cube samples cast for compression tests for UCS........................ 122
Figure 3-19: UCS grout strength of grout Type 4 with respect to the curing age (w/c= 20 %)
................................................................................................................................................ 122
Figure 3-20: UCS strength of grout type 1 with respect to the curing age (w/c= 35%) ........ 123
Figure 3-21: Failure mode of grout samples after UCS test .................................................. 125
Figure 3-22: Satisfactory failure of a perfectly hardened concrete sample (BS 12390-3:2002,
2002) ...................................................................................................................................... 125
Figure 3-23: Effect of w/c ratio on the strength of grout Type 4 ........................................... 126
Figure 3-24: Comparison of Old and New bags in 24-hour test ............................................ 127
Figure 3-25: Effect of curing condition on the strength of the grout ..................................... 128
Figure 3-26: Lateral and vertical strain measurement of grout samples ................................ 128
Figure 3-27: Stress-Strain graph of a grout sample ............................................................... 129
Figure 3-28: Cubic moulds prepared for resin casting........................................................... 130
Figure 3-29: Cubic and cylindrical resin samples.................................................................. 130
Figure 3-30: Cubic sample of resin under compression test .................................................. 131
Figure 3-31: UCS results of resin samples ............................................................................ 131
Figure 3-32: Preparation of concrete samples for UCS test .................................................. 133
Figure 3-33 Tensile test result of 15.2 mm cable bolt (Thompson & Villaescusa, 2014) ..... 135
Figure 3-34: Tensile test results of wires of 15.2 mm cable bolt ........................................... 136
Figure 3-35: Tensile test results of wires of 63 t Sumo cable ................................................ 136
Figure 4-1: Test 1: Static perpendicular double shear test using weak grout and Plain 15.2 mm
cable bolt ................................................................................................................................ 143
Figure 4-2: Test 3: Perpendicular double shear test using medium grout and plain 15.2 mm
cable bolt ................................................................................................................................ 143
Figure 4-3: 15.2 mm cable bolt after static shear test ............................................................ 144
Figure 4-4: Trace of the bent cable bolt on the soft (left) and hard (right) grout and concrete
sample after the static shear test............................................................................................. 144
Figure 4-5: Failure modes of wire in perpendicular static double shear tests ....................... 145
Figure 4-6: Static 30-degree inclined double shear test using (Test 6).................................. 146
Figure 4-7: Static 45-degree inclined double shear test using (Test 8).................................. 146
Figure 4-8: The shear load of different angled double shear tests ......................................... 147
Figure 4-9: Axial load of different angled double shear tests ................................................ 148
Figure 4-10: Failure modes of wire in inclined static double shear tests............................... 149
Figure 4-11: Shear surface being filled and cemented by leaked grout ................................. 149
Figure 4-12: Damage to the concrete surface due to high frictional forces ........................... 150
Figure 4-13: Dynamic double shear (Test 16) ....................................................................... 152
Figure 4-14: Dynamic double shear (Test 17) ....................................................................... 153
Figure 4-15: Dynamic shear load in the double shear test of cable bolts using different impact
energies .................................................................................................................................. 153
Figure 4-16: Axial load in the dynamic double shear test of cable bolts using different impact
energies .................................................................................................................................. 154
Figure 4-17: Propagation of radial cracks in dynamic shear tests ......................................... 154
Figure 4-18: Failure modes cable bolt after dynamic double shear tests ............................... 156
Figure 4-19: Failure modes of wire in dynamic double shear tests ....................................... 157
Figure 4-20: Dynamic shear test results of high-strength cable bolt (test 23) ....................... 158
Figure 4-21: Massive radial crack in the middle concrete block ........................................... 158
Figure 4-22: No failure in the cable bolt after the dynamic shear test .................................. 158
Figure 5-1: Possible failure modes of anchorage under axial loading condition ................... 162
Figure 5-2: Simplified schematic drawing of pullout mechanism ......................................... 165
Figure 5-3: Test sample before and after the test ................................................................... 166
Figure 5-4: Introduction of pullout test setup components (top), 3D perspective view (bottom
left), and vertical cross-section (bottom right) of the pullout test rig .................................... 167
Figure 5-5: Strain of the SDPA under 1500 kN static load ................................................... 168
Figure 5-6: Casting concrete samples in cardboard molds (left), riffled surface of the hole
(right) ..................................................................................................................................... 170
Figure 5-7: Externally confined sample by a split steel cylinder........................................... 170
Figure 5-8: Cable bolts being measured and marked............................................................. 171
Figure 5-9: Grouting the samples and hollow tube of cable bolts ......................................... 171
Figure 5-10: Grouting the anti-rotation tube .......................................................................... 172
Figure 5-11: Rubber sheet covering the sample (left), and confinement cage model (right) 172
Figure 5-12: Installing load-transferring shafts ..................................................................... 173
Figure 5-13: Assembling the seat and the fixed frame .......................................................... 174
Figure 5-14: (a) Setting up the anti-rotation system, (b) mounting load cell, and (c) installing
B&W ...................................................................................................................................... 174
Figure 5-15: Applying pretension load .................................................................................. 175
Figure 5-16: Static pullout testing setup ................................................................................ 176
Figure 5-17: Strain gauges installed on specimens ................................................................ 177
Figure 5-18: LVDTs mounted to record displacement .......................................................... 177
Figure 5-19: Acuity laser pointing at the bottom of the sample ............................................ 178
Figure 5-20: High-speed camera setup .................................................................................. 179
Figure 5-21: Bird-eye view (adopted from (Kaewunruen, 2007)) (left) and inside view of the
dynamic drop hammer (right) ................................................................................................ 180
Figure 5-22: Comparison of energy-based dynamic facilities for pullout test ...................... 181
Figure 5-23: Radial cracks in confined samples using split steel cylinders .......................... 182
Figure 5-24: Numerical simulation of the effect of confinement .......................................... 183
Figure 5-25: Preparation and casting of concrete in cylindrical steel pipes .......................... 184
Figure 5-26: Sticking strain gauges on the internal steel confinements ................................ 184
Figure 5-27: Location of strain gauges installed on the internal confinement ...................... 185
Figure 5-28: No crack in the confined samples after the pull tests ........................................ 185
Figure 5-29: Development of an anti-rotation system ........................................................... 187
Figure 5-30: Pullout test result of 15.2 mm plain cable in medium strength grout carried out by
SDPA ..................................................................................................................................... 188
Figure 5-31: Pullout test result of plain 15.2 mm cable bolt carried out by Steel Split Set (SSS)
................................................................................................................................................ 189
Figure 5-32: Comparison of unit pullout test results gained from SDPA and SSS ............... 189
Figure 6-1: Short encapsulation- Plain- weak Strength grout ................................................ 196
Figure 6-2: Short encapsulation, unbulbed, medium Strength grout ..................................... 196
Figure 6-3: Long encapsulation, unbulbed, medium Strength grout ..................................... 197
Figure 6-4: Long encapsulation, unbulbed, hard grout .......................................................... 197
Figure 6-5: Effect of grout type on unit pullout load ............................................................. 198
Figure 6-6: Post-test inspection of concrete samples ............................................................. 198
Figure 6-7: Cracks in the grout and PVC tube due to radial stress concentration ................. 199
Figure 6-8: Failure surface of different grouts after the pullout test...................................... 199
Figure 6-9: Cable bolt tensile failure at the neck of B&W .................................................... 200
Figure 6-10: Polished surface of the cable bolt due to the high friction Cable bolt elongation
................................................................................................................................................ 200
Figure 6-11: Measuring the difference of pull in and pullout length ..................................... 201
Figure 6-12: Bulbed cable being elongated after the pullout test .......................................... 201
Figure 6-13: Plain 63 t Sumo- Grout- Static pullout test ....................................................... 203
Figure 6-14: Plain 63 t Sumo- Resin- Static pullout test ....................................................... 203
Figure 6-15: Bulbed 63 t Sumo- Grout- Static pullout test.................................................... 204
Figure 6-16: Bulbed 63 t Sumo- Resin- Static pullout test .................................................... 204
Figure 6-17: Minimal radial cracks on concrete sample ........................................................ 205
Figure 6-18: Radial strain of concrete sample in static pullout test of plain Sumo cable bolt
................................................................................................................................................ 206
Figure 6-19: Radial strain of concrete sample in static pullout test of bulbed Sumo cable bolt
................................................................................................................................................ 206
Figure 6-20: Cone-shape stress zone on the top side of the concrete sample B&W response to
the load ................................................................................................................................... 207
Figure 6-21: Status of the B&W before (left) and after (right) static pullout test ................. 208
Figure 6-22: Response of the B&W to the static pullout load in grouted plain cable bolt .... 208
Figure 6-23: Response of the B&W to the static pullout load in resined plain cable bolt..... 208
Figure 6-24: Response of the B&W to the static pullout load in grouted bulbed cable bolt. 209
Figure 6-25: Status of the grout surface after static pullout test of plain (left) and bulbed (right)
cable bolt. ............................................................................................................................... 210
Figure 6-26: Status of the resin surface after static pullout test of bulbed cable bolt ............ 210
Figure 6-27: Evidence of unwinding of the plain cable bolt during the static pullout test .... 211
Figure 6-28: Unwinding and cable failure of the bulbed cable bolt in static pullout test ...... 212
Figure 6-29: Dynamic pullout test of cement grouted plain cable bolt ................................. 213
Figure 6-30: Dynamic pullout test of resined unbulbed cable bolt ........................................ 214
Figure 6-31: Dynamic pullout test of grouted bulbed cable bolt ........................................... 214
Figure 6-32: Dynamic pullout test of resined bulbed cable bolt ............................................ 215
Figure 6-33: negligible radial cracks on the surface of the concrete samples ....................... 215
Figure 6-34: Radial strain of concrete sample in dynamic pullout test of unbulbed Sumo cable
bolt ......................................................................................................................................... 216
Figure 6-35: Radial strain of concrete sample in dynamic pullout test of bulbed Sumo cable
bolt ......................................................................................................................................... 216
Figure 6-36: Deformation of the B&W before and after the dynamic pullout test................ 217
Figure 6-37: Comparison of tests with fixed B&W versus tests with free B&W in grouted plain
cable bolt ................................................................................................................................ 218
Figure 6-38: Comparison of tests with fixed B&W versus tests with free B&W in resined
bulbed cable bolt .................................................................................................................... 218
Figure 6-39: Comparison of tests with fixed B&W versus tests with free B&W in grouted
bulbed cable bolt .................................................................................................................... 219
Figure 6-40: Status of the grout surface after dynamic pullout test of plain (left) and bulbed
(right) cable bolt ..................................................................................................................... 220
Figure 6-41: Evidence of unwinding and cable failure of the bulbed cable bolt in dynamic
pullout test.............................................................................................................................. 220
Figure 7-1: Generated geometry of the MW9 cable bolt by SpaceClaim.............................. 228
Figure 7-2: Geometry of the grouted cable bolt..................................................................... 228
Figure 7-3: Perspective, Longitudinal, and cross-sectional view of the meshed pullout test
model...................................................................................................................................... 229
Figure 7-4: Applied boundary conditions on the pullout test model ..................................... 230
Figure 7-5: Friction coefficient is a function of relative velocity and pressure (LS-DYNA,
2018) ...................................................................................................................................... 232
Figure 7-6: General shape of concrete model yield surface for CSCM model in two dimensions
................................................................................................................................................ 237
Figure 7-7: Geometry and meshing of the grout cube ........................................................... 239
Figure 7-8: Comparison of failure in the numerical model and experimental compression test
................................................................................................................................................ 240
Figure 7-9: Comparison of numerical and experimental compression test results on cement
grout ....................................................................................................................................... 240
Figure 7-10: Probes for measurement of the axial load ......................................................... 242
Figure 7-11: Comparison of numerical and experimental results for grout and 63 t Sumo .. 243
Figure 7-12: Ratio of kinetic energy/total internal energy of the simulations ....................... 243
Figure 7-13: Iso-surface image of plastic zones in grout ....................................................... 243
Figure 7-14: Comparison of grout failure in the numerical and experimental tests .............. 244
Figure 7-15: Stress distribution in the cable bolt during the test .......................................... 245
Figure 7-16: Rotational displacement of the bottom end of the cable ................................... 245
Figure 7-17: Simplified model of cable bolt .......................................................................... 247
Figure 7-18: Comparison of a simplified numerical model with experiments and the detailed
model...................................................................................................................................... 247
Figure 7-19: Comparison of the effective plastic strain of the simplified model and the detailed
model...................................................................................................................................... 248
Figure 8-1: Comparison of failure modes of strand wires in different angles of shear test ... 252
Figure 8-2: Numbering of the wires of 15.2 mm cable bolt strand........................................ 252
Figure 8-3: Chance of shear failure in a double shear test for different wires of a 15.2 mm cable
bolt ......................................................................................................................................... 252
Figure 8-4: Comparison of plain and bulbed cable bolts in perpendicular double shear tests
................................................................................................................................................ 253
Figure 8-5: Shear load in the double shear test perpendicular to bolt axis test of cable bolts
using different cementitious grout strengths .......................................................................... 254
Figure 8-6: Axial load in the perpendicular double shear test of cable bolts using different
cementitious grout strengths .................................................................................................. 254
Figure 8-7: Comparison of shear load in different static and dynamic double shear tests .... 255
Figure 8-8: Comparison of energy in different static and dynamic double shear tests .......... 256
Figure 8-9: Comparison of axial load in different static and dynamic double shear tests ..... 256
Figure 8-10: Results of the shear test at different angles ....................................................... 257
Figure 8-11: Hinge angle of cable bolts in perpendicular double shear tests ........................ 258
Figure 8-12: Results of static pullout tests on 63 t Sumo cable bolt...................................... 258
Figure 8-13: Comparison of normalized energy in static and dynamic pullout tests ............ 259
Figure 8-14: Grout damage intensity after static and dynamic tests with plain cable ........... 261
Figure 8-15: Almost 40-degree rotation in the direction of unwinding of the cable bolt during
the static test ........................................................................................................................... 262
Figure 8-16: Rotational displacement of the bottom end of the cable ................................... 263
Figure 8-17: Failure of the bulbed cable bolt in static and dynamic pullout tests ................. 263
Figure 8-18: Pullout test result of 15.2 mm plain cable in medium strength grout carried out by
SDPA ..................................................................................................................................... 264
Figure 8-19: Theory of stick-slip behaviour .......................................................................... 265
Figure 8-20: Components of a cable bolt support .................................................................. 266
Figure 8-21: Mechanism of cable bolt and B&W (Thompson & Villaescusa, 2014) ........... 266
Figure 8-22: B&W deformation in the static pullout test ...................................................... 267
Figure 8-23: B&W deformation in static and dynamic pullout tests ..................................... 268
Figure 8-24: Status of B&W before and after static test of 15.2 mm bulbed cable ............... 268
Figure 8-25: Possible failure mechanism of cable bolts in pullout test ................................. 269
Figure 8-26: High stress and opened cracks around the bulbed area ..................................... 271
Figure 8-27: Relation between elongation and the absorbed energy ..................................... 272
Figure 8-28: Comparison of cable lengths before and after tests .......................................... 273
Figure 8-29: Normalized pullout load.................................................................................... 273
Figure 8-30: Effect of grout strength on the pullout load ...................................................... 274
Figure 8-31: Effect of grout strength on the effective plastic strain ...................................... 274
Figure 8-32: Effect of coefficient of friction on pullout load ................................................ 275
Figure 8-33: Effect of friction coefficient on the shear stress and plastic zone distribution . 276
Figure 8-34: Effect of shear resistance on the axial pullout load .......................................... 276
Figure 8-35: Effect of shear resistance on the effective plastic strain ................................... 277
Figure 8-36: Static pullout tests of plain Sumo cable bolt ..................................................... 278
Figure 8-37: Axial load on cable bolt with 300 mm embedment length ............................... 279
Figure 8-38: Axial load on cable bolt with 450 mm embedment length ............................... 279
Figure 8-39: Axial load on cable bolt with 600 mm embedment length ............................... 279
Figure 8-40: Comparison of axial load in different embedment lengths ............................... 280
Figure 8-41: Comparison of unit axial load in shear tests and pullout tests for 15.2 mm cable
bolt ......................................................................................................................................... 281
Figure 8-42: Comparison of unit axial load in shear tests and pullout tests for 63 t Sumo cable
bolt ......................................................................................................................................... 281
Figure 8-43: Debonding and pullout of the cable bolt in a single shear test, reported by (Aziz,
Rink, et al., 2017)................................................................................................................... 282
Figure A-1: Test 2: Double shear using slow-set grout- Bulb cable ..................................... 319
Figure A-2: Test 4: Double shear using quick-set grout ........................................................ 319
Figure A-3: Test 5: Double shear using quick-set grout ........................................................ 319
Figure A-4: Perpendicular double shear using grout- 63 t cable bolt .................................... 320
Figure A-5: Perpendicular double shear using grout- 63 t cable bolt .................................... 320
Figure A-6: Perpendicular double shear using strong grout- 63 t cable bolt ......................... 321
Figure A-7: Perpendicular double shear using strong grout- 63 t cable bolt ......................... 321
Figure A-8: Static 45 degree inclined double shear test using (Test 7) ................................. 322
Figure A-9: Static 45 degree inclined double shear test using (Test 9) ................................. 322
Figure A-10: Dynamic double shear (Test 16) ...................................................................... 323
Figure A-11: Dynamic double shear (Test 17) ...................................................................... 323
Figure A-12: Dynamic double shear (Test 18) ...................................................................... 324
Figure A-13: Dynamic double shear (Test 19) ...................................................................... 324
Figure A-14: Dynamic double shear (Test 22) ...................................................................... 325
Figure A-15: Cross-section view of the SDPA ...................................................................... 326
Figure A-16: Perspective and cross-section view of the main seat ...................................... 327
Figure A-17: Top view and cross-section of the main seat ................................................... 328
Figure A-18: Top view and cross-section of central part of the main seat ............................ 329
Figure A-19: Bottom view of loading platform ..................................................................... 330
Figure A-20: Reinforcement design of the loading platform................................................. 331
Figure A-21: Channel beam used in the main seat and fixed frame ...................................... 332
Figure A-22: I-beam used in the fixed frame......................................................................... 333
Figure A-23: Angles used in the fixed frame and the main seat ............................................ 334
Figure A-24: Bearing holder .................................................................................................. 335
Figure A-25: Shaft holder ...................................................................................................... 336
Figure A-26: Anti-rotation tube ............................................................................................. 337
Figure A-27: Anti-rotation plate ............................................................................................ 338
Figure A-28: confinement cage rods...................................................................................... 339
Figure A-29: Von-Mises failure criteria evaluation for moving parts of SDPA ................... 339
Figure A-30: Pullout test setup being lifted by roof crane and placed under loading machine
................................................................................................................................................ 340
Figure A-31: Test 1- Short encapsulation- 15.2 mm Plain- weak grout ................................ 342
Figure A-32: Test 2- Short encapsulation- 15.2 mm Plain- weak grout ............................... 342
Figure A-33: Test 3- Short encapsulation- 15.2 mm Plain- weak grout ................................ 343
Figure A-34: Test 4- Long encapsulation- 15.2 mm Plain- weak grout ................................ 343
Figure A-35: Test 5- Long encapsulation- 15.2 mm Bulbed- weak grout ............................. 344
Figure A-36: Test 6- Short encapsulation- 15.2 mm Plain- medium grout ........................... 344
Figure A-37: Test 7- Short encapsulation- 15.2 mm Plain- medium grout ........................... 345
Figure A-38: Test 8- Short encapsulation- 15.2 mm Plain- medium grout ........................... 345
Figure A-39: Test 9- Long encapsulation- 15.2 mm Plain- medium grout ........................... 346
Figure A-40: Test 10- Long encapsulation- 15.2 mm Plain- medium grout ......................... 346
Figure A-41: Test 11- Long encapsulation- 15.2 mm Plain- medium grout ......................... 347
Figure A-42: Test 12- Long encapsulation- 15.2 mm Bulbed- medium grout ...................... 347
Figure A-43: Test 13- Short encapsulation- 15.2 mm Plain- strong grout ............................ 348
Figure A-44: Test 14- Long encapsulation- 15.2 mm Plain- strong grout............................. 348
Figure A-45: Test 15- Long encapsulation- 15.2 mm bulbed- strong grout .......................... 349
Figure A-46: Test 16- Long encapsulation- 15.2 mm bulbed- strong grout .......................... 349
Figure A-47: Test 17- Long encapsulation, Plain 63 t Sumo- cement grout ......................... 350
Figure A-48: Test 18- Long encapsulation, Plain 63 t Sumo- cement grout ......................... 351
Figure A-49: Test 19- Long encapsulation, Plain 63 t Sumo- cement grout ......................... 351
Figure A-50: Test 20- Long encapsulation, bulbed 63 t Sumo- cement grout ...................... 352
Figure A-51: Test 21- Long encapsulation, bulbed 63 t Sumo- cement grout ...................... 352
Figure A-52: Test 22- Long encapsulation, Plain 63 t Sumo- resin ...................................... 353
Figure A-53: Test 23- Long encapsulation, Plain 63 t Sumo- resin ...................................... 353
Figure A-54: Test 24- Long encapsulation, bulbed 63 t Sumo- resin .................................... 354
Figure A-55: Test 23- Long encapsulation, bulbed 63 t Sumo- resin .................................... 354
Figure A-56: Test 26- Long encapsulation, Plain 63 t Sumo- cement grout ......................... 355
Figure A-57: Test 27- Long encapsulation, Plain 63 t Sumo- cement grout ......................... 356
Figure A-58: Test 28- Long encapsulation, Plain 63 t Sumo- cement grout ......................... 356
Figure A-59: Test 29- Long encapsulation, Plain 63 t Sumo- resin ...................................... 357
Figure A-60: Test 30- Long encapsulation, bulbed 63 t Sumo- cement grout ...................... 357
Figure A-61: Test 31- Long encapsulation, bulbed 63 t Sumo- cement grout ...................... 358
Figure A-62: Test 32- Long encapsulation, bulbed 63 t Sumo- cement grout ...................... 358
Figure A-63: Test 33- Long encapsulation, bulbed 63 t Sumo- cement grout ...................... 359
Figure A-64: Test 34- Long encapsulation, bulbed 63 t Sumo- cement grout ...................... 359
Figure A-65: Test 35- Long encapsulation, bulbed 63 t Sumo- resin .................................... 360
Figure A-66: Test 36- Long encapsulation, bulbed 63 t Sumo- resin .................................... 360
Figure A-67: CSCM material input parameters for grout ...................................................... 361
Figure A-68: CSCM material input parameters for resin ...................................................... 361
Figure A-69: Input parameters of tiebreak contact for grout and cable bolt ......................... 361
Figure A-70: Input parameters of tiebreak contact for grout and cable bolt ......................... 362
Figure A-71: Shear stress in UCS test of grout...................................................................... 362
Figure A-72: failure mechanism of tensile test of the cable bolt ........................................... 362
Figure A-73: Simulation of tensile test of 15.2 mm using ANSYS....................................... 363
Figure A-74: Lateral stress on the grout ................................................................................ 363
Figure A-75: 2nd principal deviatoric stress ........................................................................... 363
Figure A-76: Elongation of the cable as a function of relative displacement ........................ 364
Figure A-77: Axial and lateral stress in low-friction pullout test .......................................... 364
Figure A-78: Axial stress in high-friction pullout test ........................................................... 364
Figure A-79: Axial stress distribution in the simplified pullout test model .......................... 365
Figure A-80: Propagation of plastic zone in different test intervals ...................................... 365
Chapter 1 | Introduction

Introduction

CHAPTER 1
Introduction

Rationale
Australia is the leading global producer of gold, iron ore, lead, zinc, and nickel, and it has the
world's largest uranium deposits and fourth-largest coal resources
(https://www.amsj.com.au/australian-mining/). Mining accounts for about 8% of the Gross
Domestic Product (GDP), and the mining sector contributes about 15% to the overall
Australian economy. There are around 350 operating mines in Australia, with one in three
operating underground (https://www.amsj.com.au/australian-mining/). Although mining is one
of the most profitable industries in Australia, it ranks fifth in worker fatalities by all industry
groups across the country. During 2015-2019, the Construction and Mining Labourers Group
accounted for 34 fatalities, representing 36 % of all industry fatalities. Most of the mining
fatalities occurred in the primary mining jurisdictions of Queensland, Western Australia and
New South Wales (AMSJ, 2020). Western Australia reported 52 mining fatalities over twelve
years from 2000 to 2012, while 27 were related to an unsafe workplace situation. Five out of
52 cases (almost 10%) of fatalities in Western Australia directly resulted from mine instability
(Government of Western Australia, 2012). This concern intensified with the need for more
underground minerals, thus encouraging the mining industry to extract deeper ores. Currently
there are several deep underground mines worldwide, the notable ones include Mount Isa Mine
of Queensland with the depth of mining extending to below 1.9 km and the Mponeng Gold
mine of South Africa, with the depth of operation of below 3.84 km as reported 2018. The
deeper the mine progresses, the higher stress concentration around the excavations would be
expected. From the ground stability point of view, the challenges of ground control can be more

1
Chapter 1 | Introduction

severe and life-threatening; hence the need for effective ground reinforcement methods is
paramount to maintain a stable opening at such depths.

Rock bolts (or dowels) and cable bolts (also referred to as tendons) are among the most
common supporting systems of both mining and civil engineering excavations. Nowadays, they
are installed during the initial stages of mining (primary support) or when the drift openings
intersect with other drifts and when nearby mining activity takes place causing multiple
openings (secondary support). Due to the plate tectonics and rock mass qualities, tendons might
be subjected to static and dynamic axial and shear loading. To this, studying the behaviour of
tendons under different loading conditions leads to the efficient and reliable design of
supporting systems.

Background of the Study


In underground mining and generally in tunneling, ground anchors are used as primary and
secondary supports installed in both active and passive forms (Indraratna & Kaiser, 1990). The
use of rock bolts for ground support goes back to 1872 in a slate quarry located in North Wales,
UK, and in 1918 in the mines of Upper Silesia (now Poland) (Lang, 1961). By the 1920s and
1930s, rock bolts were used in several underground mines in the United States where, in 1952,
the annual consumption had reached 25 million (Lu, 1990). In Australia, the mining industry
started to use rock bolting technology in the Snowy Mountains hydroelectric scheme between
1949 and 1969 (Bolstad et al., 1984).

The application of cable bolts as ground supports was initially recorded in the Willory mine
(Canada) and the Free State Geduld Mines Ltd in South Africa in the 1960s (C. Windsor, 1992).
By 2016, 1.4 million cable bolts were used in the United States (S. Tadolini et al., 2016). Since
then, a great variety of cable bolts and rock bolts have been designed and manufactured in order
to improve the reliability of the ground support system. Among all, the 15.2 mm and 18.0 mm
cable bolts were reported to be the top selling types (S. Tadolini et al., 2016).

The success of the tendons technology usage as a popular supporting system is due to the
following advantages (Guglielmetti et al., 2007; Hutchinson & Diederichs, 2002; Mogi, 2006):

• Not constrained by underground excavation cross-section


• Swift installation after excavation
• Saving in the operational area (optimum excavation profile)
• Providing active support and safer strata control by pretension force, and

2
Chapter 1 | Introduction

• Economical in comparison with other ground support systems

Besides the above-stated advantages, the drawback of tendons is that the loading mechanism
of each tendon might be different based on their angle of installation with respect to joints and
weak planes; consequently, the design of the support system using tendons faces complexities.
For instance, Figure 1-1 presents how ground tendons could undergo various loading
conditions based on the ground condition. In general, the failure modes of ground tendons can
be categorized into three classes, shear, axial, or a combination of both (Hutchinson &
Diederichs, 2002). In fact, rarely is it seen to have a pure form of shear or axial failures in the
field, as complications in ground deformation often present varied challenges for underground
work for effective support system application.

Figure 1-1: Different loading modes on ground tendons (Barley & Windsor, 2000)

Loading conditions can vary from slow-rate deformations (quasi-static) to instant movements
due to sudden load applications (dynamic). Static loading is caused by gravitational movements
of the rock blocks or bedding layers over time as mining progresses. At the same time, dynamic
loads are consequences of local coal or rockburst as well as global earthquakes or geo-seismic
activities. Different approaches have been used by researchers over the years to interpret the
failure mechanism of tendons. Field experiments (Bawden et al., 1992; L. B. Martin, 2012;
Stillborg, 1984), analytical developments (Blanco Martín, Tijani, et al., 2011; Cao et al., 2014;
Farmer, 1975; Ghadimi et al., 2015; He et al., 2015; X. Li et al., 2015), laboratory experiments
(Aoki et al., 2002; Aziz et al., 2015; Benmokrane et al., 1995; Moosavi et al., 2005; Stillborg,
1984) and numerical solutions (Aziz, 2007; Cao et al., 2013; J. Chen, 2016; Nemcik et al.,
2014; Vallejos et al., 2020) are some of the common approaches. Current approaches can be
summarized as shown in Figure 1-2. As can be seen, the shear and axial behaviour of tendons
has been studied separately using different mechanisms. Single shear, double shear, and angled

3
Chapter 1 | Introduction

shear tests are the common methods of shear behaviour studies. Pullout tests and tensile tests
of the tendons would measure the resistance of the ground tendon against the axial loading.
Numerical models have been developed using different solving methods such as Finite Element
Method (FEM) (Aziz et al., 2005; Ferrero, 1995; Grasselli, 2005; Mirzaghorbanali et al., 2017)
and Discrete Element Method (DEM) (Shang et al., 2018; S. Tadolini et al., 2016) to predict
the behaviour of bolts under axial and shear loading conditions; however, the focus has been
on rock bolts and cable bolts have hardly been the topic of these studies.

Ground
anchors study

Static Dynamic

Combination Combination
Shear Axial Shear Axial
of both of both

Numerical/ Numerical/ Numerical/ Numerical/ Numerical/ Numerical/


Experimental Experimental Experimental Experimental Experimental Experimental
Analytical Analytical Analytical Analytical Analytical Analytical

FEM
Analytical FEM Impact
FEM Single shear (Only Rock Pullout Double shear Double shear NONE NONE NONE
calculations (Rock bolts) pullout
bolts)

Double shear Tensile

Angled shear

Figure 1-2: Classification of studies around ground tendons

Motivations of the Study


There have long been life-threatening hazards associated with coal bursts during underground
coal mining in Poland, Russia, China, and the United States (Dou et al., 2009; Mark, 2018). The
first incidence of coal burst was reported in the UK in 1783 (Whyatt, J, W. Blake, 2002).
Despite decades of extensive research conducted by international mining and geotechnical
scholars, coal burst frequency and severity is still increasing in these above mentioned
countries. Because of the shallow mining depth, simple geological conditions, advanced
mining technology, and reasonable geotechnical design, coal bursts have not been regarded as
a safety hazard in Australian coal mines, as there has been no case of coal bursts since 2014
(X. Yang et al., 2019). However, an increase in the number of recorded coal bursts and rock
bursts events internationally, has attracted the local researchers and industry partners to the
topic so that there are more than 15 research topics in progress with an estimated value of 4
million AUD sponsored by the Australian Coal Association Research Program (ACARP)
(ACARP, 2022).

Clearly, the interest in this topic has increased in recent years; nevertheless, there remains a
lack of a close link between rock burst management and tendon support usage for effective

4
Chapter 1 | Introduction

outcomes. This sets the motivation for the increasing interest in researching this field,
particularly on the dynamic aspect of ground reinforcement management.

Key Objectives
The main objective of this thesis is to study the performance of the cable bolts subjected to
various loading conditions with specific attention to dynamic events. The key objectives
include:

• Development of laboratory testing methods for dynamic loading of ground tendons


under shear loading conditions.
• Design and development of a new testing facility to provide the opportunity for
implementation of pullout tests under both static and dynamic conditions.
• Comparative study of the performance of tendons under different installation angles.
• Comparative experimental study of static and dynamic loading of cable bolts under
both shear and axial loading.
• Understanding the pullout mechanism of cable bolts with the assistance of numerical
modeling using the Finite Element Method.
• Numerical modeling of the pullout behaviour of cable bolt using the Finite Element
Method.

Methodology
Development of tendon technology testing facilities, laboratory testing, and numerical
modeling are the main objectives of this study. Three different testing apparatuses, including
Static and Dynamic Cylindrical Double Shear testing rig (known as MK-IV), Static Angled
Double Shear testing rig, and Static and Dynamic Pullout testing Apparatus, have been
employed to experiment with different loading modes of tendons. The latest apparatus was
designed, built, and fabricated as part of this research. Numerical tools such as 3D Finite
Element Element modeling software (ANSYS) and LS-DYNA were used to simulate the cable
bolts' behavior under axial and shear loading conditions. The interaction between bolt and grout
was carefully observed in the numerical models.

To accomplish the numerical models, design drafts were drawn using AutoCAD in 2D and
turned into 3D models using SolidWorks. The final design was later exported as an Initial
Graphic Exchange Specification (IGES) file for static structural and explicit dynamic analysis

5
Chapter 1 | Introduction

in ANSYS Workbench. Also, the numerical models of the pullout tests were generated using
LS-DYNA in order to provide both implicit and explicit analysis on the same platform.

Thesis Structure
The thesis consists of nine chapters, followed by a bibliography and appendices. The thesis has
been structured as follows:

Chapter 1 provides an overview of the topic, looks at the motivations of this research,
illustrates the key objectives, and forms the thesis outline.

Chapter 2 reviews the literature on rock bolting and cable bolting with a greater focus on axial
loading. The chapter commences with the fundamental rationale of the topic, followed by an
introduction to ground anchor components. Then, attention has been paid to the history of
developments in laboratory facilities for static and dynamic axial testing (pullout). The chapter
continues with assessment of laboratory, analytical and numerical achievments and then
summarizes the effect of different parameters on pullout strength of the ground anchors.
Finally, recent developments in shear testing of tendon has been reviewed. The chapter aims
to document the research roadmap from the past to the present.

Chapter 3 explains the sample preparation procedure for three different sets of tests, including
static perpendicular double shear tests, static angled double shear tests and dynamic
perpendicular double shear tests. The perpendicular double shear setup (also known as MK-
IV) - the most recent modification of double shear boxed developed at the University of
Wollongong- is used for both static and dynamic shear tests. A newly built angle double shear
setup with the ability to test the two different installation angles of 30 and 45 degrees is also
introduced in this chapter. The chapter finishes with an introduction to the required materials
for laboratory studies. These include cable bolts, cement-base grout and resin, and concrete.
The mechanical properties of each material have been examined and tabulated in this chapter.

Chapter 4 presents the results of laboratory shear tests. The chapter starts with the
experimental plan and summarizes the test numbers and specifications. Then, the results of
static perpendicular double shear, static angle double shear, and dynamic perpendicular double
shear are demonstrated, respectively. Detailed test results can be found in the Appendices.

Chapter 5 explains the design and manufacturing process and introduces the new Static and
Dynamic Pullout test Apparatus (SDPA). The chapter starts with an explanation of the

6
Chapter 1 | Introduction

mechanism of testing with SDPA. Then the design concept and reasoning are introduced,
manufacturing of the various components and sample preparation are detailed. Different
modifications to the setup are listed in order to clarify the development process. Eventually,
the performance of the SDPA is verified by the implementation of trial tests and comparison
of the results with previous studies. Detailed design drawings can be found in the attached
Appendices.

Chapter 6 presents the results of an experimental study on the static and dynamic pullout
testing of cable bolts. Two types of cable bolts, including 15.2 mm seven-wire and nine-wire
Sumo (also called 63 t cable bolts) and two different bonding agents (cementitious grout and
mastic resin), have been selected for testing in both static and dynamic modes. The chapter
begins with a briefing of the experimental plan and continues with static pullout test results. It
is then followed by dynamic pullout test results. Detailed test results and graphs can be found
in the attached Appendices.

Chapter 7 expands on the use of numerical tools for further investigations of pullout
behaviour, specifically for cable bolts. A detailed numerical model was generated to investigate
the interaction between the cable bolt and grout. This is then followed by a simplified model
which has been inspired by the parameters obtained from the detailed model. The simplified
model is proposed to decrease the runtime and increase efficiency of the numerical model. Both
models are verified with laboratory experiments. Sensitivity analysis and parametric study
based on the developed model are later discussed in chapter eight. Further details of the
numerical models, input parameters of models, and some results have been attached in the
appendices.

Chapter 8 widens the understanding of shear and axial behaviour based on the laboratory and
numerical analysis. The chapter discusses the shear test results to demonstrate the failure
mechanism of cable bolts under static and dynamic shear loadings. It is followed by a
parametric study on some parameters affecting the shear test results. Similarly, the results of
pullout tests are discussed, and the failure mechanism is explained. A comprehensive
parametric study covers different aspects of the axial loading of tendons. Finally, shear and
axial test results are combined to predict the behaviour of tendons under combined loading
conditions.

7
Chapter 1 | Introduction

Chapter 9 draws the previous chapters together and highlights the key findings of the study
on static and dynamic shear and axial loading of ground tendons. It summarizes the conclusions
and provides recommendations for further studies.

Scope
This thesis mitigates the issues related to ground tendons in laboratory scale. For this,
experimental approaches have mainly been considered. An effort has been made to optimize
the number of tests in order to minimise the environmental effects of tests and lessen waste
production. It has been guaranteed that a sufficient number of tests have been undertaken to
verify the validity of the results. Consistency of the results and performance of the employed
testing facilities were already proved by previous researchers such as Rasekh (2017), Yang
(2019), and Khaleghparast (2021). Hence, test results were firstly verified based on previous
studies carried out using the same testing facilities. Furthermore, tests were repeated at least
two to three times until consistent results were obtained. In completion of experiments,
numerical simulations were utilized to expand the knowledge. Numerical models were verified
based on the experiments and through trial and error process. Field work was out of the scope
of this research.

A research team on rock bolts and cable bolts at the Civil, Mining, and Environment (CME)
faculty of the University of Wollongong (UOW) has devoted over 20 years to different aspects
of the study of ground anchors. Years of laboratory and analytical research have been carried
out by researchers to develop the current facilities and produce current test results. This
research has followed the same path as previous studies, and every effort has been made to
expand the previous findings and take a step forward in this field of research. The novelty of
this thesis is design and manufacturing of the new static and dynamic pullout testing setup
which is introduced in Chapter 5. The author would credit the novelty of double shear and
angle shear testing to former colleagues, including Prof. Naj Aziz (2019), Dr. Ali
Mirzaghorbanali (2013), Dr. Haleh Rasekh (2017), Dr. Guanyu Yang (2019), and Dr. Saman
Khaleghparast (2021). Thus, it is necessary to refer to their encyclopedic research for further
detail about the literature review of double shear studies, and coverage of this section is out of
the scope of this research.

It is also noteworthy that the new pullout test apparatus has been designed to comply with the
current technologies and facilities at UOW. Hence, the improvement of the current laboratory
testing facilities of the university also does not fit into the scope of this research.

8
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Literature Review of Axial and Shear Loading of


Tendons

CHAPTER 2
Literature Review
of
Axial and Shear Loading of Tendons

Introduction
With an increasing trend of mining to go deeper into the earth, more geotechnical unknowns
are added to the equilibrium equations. In addition, more effort is required to predict and
overcome unforeseen challenges of deep hard rock or coal mining and tunneling. Lang (1961)
discussed the theory of rock reinforcement and the application of rock bolts. They explained
that instead of supporting underground spaces, effort must be made to help the rock support
itself by maintaining its integrity. Later, studies on the possible failure modes of the tendons
have become a vital topic of research in the last three decades. Studies by (Hibino & Motojima,
1981; Indraratna & Kaiser, 1990; Spang & Egger, 1990; Stillborg, 1984) in the late 80s and
early 90s on the design and installation of both rock bolts and cable bolts, the effect of coating
on tendons (Dorsten et al., 1984; Goris & Conway, 1988), the effect of tendons profile and
geometry (Fuller & Cox, 1975; Hyett et al., 1995; Jalalifar et al., 2006; Kaiser et al., 1992;
Renwick, 1992; Stillborg, 1984; C. Windsor, 1992) are only some of the early age studies in
this field.

Based on a rough estimation, it is estimated that more than 500 million anchors, including rock
bolts and cable bolts, have been used worldwide as a supporting system thus far (Aziz, 2014).
It means thousands of tonnes of metal have been buried back into the earth to provide stability
for the excavations. Occasionally, hundreds of meters of tendons in the ground fail to control
ground movements allowing unexpected catastrophes and fatalities. Figure 2-1 shows one of

9
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

the recent rockburst and mine collapses in one of the Australian mines. As can be seen in the
photo, despite the considerable number of anchors, the stored energy of the ground has
dominantly failed the supporting system.

Figure 2-1 Ruckburst in Australian mine

Rockburst and coal bursts, common phenomena of deep excavations, release a significant
amount of energy in a short period (Mazaira & Konicek, 2015; Y. Wu et al., 2019). According
to the seismic data gathered from Polish hard coal mines, seismic energy levels vary from
hundreds of KiloJouls for low seismicity to more than 1 MJ for the very high seismically active
areas (Mutke et al., 2015). Although it is not feasible to prevent energy release due to dynamic
loading, energy can be dissipated by energy-absorbing sources, leading to a less catastrophic
incident. The axial load-bearing capacity of the tendon structure is the parameter that indicates
the strength of the roof support against applied loads. This is usually measured by pullout test
either in situ or in the laboratory.

Research on the pullout behaviour of tendons (specifically cable bolts) is undertaken under
both static and dynamic conditions; however, simulation of the dynamic load in the in situ state
is strenuous or might be impractical. During the last three decades, laboratory pullout tests have
been carried out using different techniques and technologies, such as; a simple single
embedment technique (Aziz et al., 2006; Aziz & Webb, 2003; Benmokrane et al., 1995) or by
double-embodiment tests (Bigby & Reynolds, 2005; Thomas, 2012). These pullout tests are

10
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

carried out by encapsulating tendons in a steel tube, in a concrete block, or an artificial rock
(Ito et al., 2001; L. B. Martin, 2012), and in situ tests (Aziz, Mirzaghorbanali, et al., 2016;
Compton & Oyler, 2005; Stillborg, 1984). Also, the examination of tendons under dynamic
loading conditions on a laboratory scale has been the topic of study in the last three decades all
around the world, such as in Australia, Canada, China, South Africa and the USA (Player et
al., 2004; Player & Cordova, 2009; St-pierre, 2007; Tannant et al., 1995). Some of the most
important findings from these mentioned studies can be noted as:

• Borehole properties: As the first and maybe most crucial parameter of the borehole, the
diameter can affect the required pullout force in different ways. Larger boreholes result
in a lower pullout force for the plain cable bolts, whereas larger boreholes in bulbed
cables can positively affect the pullout force (Thomas, 2012). The borehole surface
must not be smooth; otherwise, the chance of debonding on the contact face of the grout
and borehole would be the more dominant mode of failure (Cao, 2012)
• Cable type and geometry: In general, plain, indented, and bulbed cables are the most
common type of cable bolts. Bulbed cables have consistently shown better performance
than the other two, while plain and indented cables had very similar peak loads (Craig
& Holden, 2014; Craig & Murnane, 2013; Thomas, 2012).
• Confinement pressure: radial confinement around the tendon crucially affects the peak
pullout load(Coates & Yu, 1970). Once the cable bolt is pulled, the radial force to the
surrounding rock increases significantly. In experiments with insufficient lateral
confinement, the cable bolt pulls out due to radial cracks on the sample (Hagan & Li,
2017; Hyett, Bawden, & Coulson, 1992; Ito et al., 2001; L. B. Martin, 2012).
• Grout strength: aside from the rheological properties of each grout material, the
water/cement ratio is a vital factor in defining the strength of the grout. Experiments
unanimously state that the stronger the grout, the higher the pullout load expected (J.
Chen, 2016; Hyett et al., 1995; Kilic, Yasar, & Celik, 2002; MacSporran, 1993). Also,
the longer curing time results in higher strength; however, changes in grout strength
after 21 days occur very slowly (Hassani et al., 1992; X. Li, 2019; Mirza et al., 2016).
It is critical for geotechnical practitioners to have access to sufficient information on the
performance of ground support systems to facilitate decision-making and result in efficient
design for mines and tunnels. Whilst it is accepted that laboratory-based testing is not fully
representative of rock bolts and cable bolts' performance, repetition of the tests and obtaining
consistent test results is still a practical and acceptable solution for simulation of the actual

11
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

field conditions in a safe and under-control situation. Large-scale field experiments can result
in more realistic results; however, they are normally more expensive, more time-consuming
and out of the scope of academic research.

The experimental study of tendons under axial loading is not limited only to mining and
tunneling engineering. The bond quality between rebar and concrete in structural engineering
is also a similar interesting topic in civil and structural engineering (Solomos & Berra, 2010;
Tassios, 1980; Yeih et al., 1997). The study of the pullout behaviour of tendons could be
classified into two main groups of static and dynamic tests (Figure 2-2). Due to their different
applications and different range of strength, rock bolts and cable bolts are separately classified.
Rock bolts have been in use in different forms such as wooden dowels, expansion shell bolts,
FRP1, GFRP2, Basalt, or steel rebars (Blanco Martín, Hadj-Hassen, et al., 2011; X. Li, 2016;
W. Wang et al., 2018). The introduction of energy-absorbing rock bolts opened the window of
studies to investigate the dynamic behaviour of rock bolts. (Ansell, 2005; C. C. Li et al., 2014,
2021; Sharifzadeh et al., 2020; Tannant et al., 1995). Similar to rock bolts, cable bolts are
mainly made of steel; however, Hoehn et al. (2021) introduced a new type of cable bolt made
of carbon fibres. The number of studies on the behaviour of cable bolts, especially in dynamic
loading, is not comparable with rock bolts, and the need for further studies is still necessary.

Pullout test

Static Dynamic

Cable bolts Rock bolts Cable bolts Rock bolts

Figure 2-2: General classification of pullout test experiments

In 1992, Yan categorized the anchorage bond tests carried out in civil engineering based on the
type of specimens used (C. Yan, 1992). Figure 2-3 shows different types of specimens used in
debonding tests:

1. Short embedment length (Type A) or long embedment (Type B~E)

1
Fiber Reinforced Polymer
2
Glass Fiber Reinforced Polymer

12
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

2. Concrete stress in the axial direction (Types A~D), compression (Types A, B, and E),
or tension (Types C and D)
3. Centric loading (Types A~D) or eccentric (Type E)
4. Reinforcement loaded on one side (Types A~C and E) or two sides (Type D)

A B

C D

E
Figure 2-3: Different types of specimens used in debonding test (after (C. Yan, 1992)

In short-encapsulated bolts, the pullout test usually results in cone failure of the host medium.
As the embedment length increases, the failure surface may change to the bond between the
encapsulation material and the host medium, the interface of rebar and adhesive, or a
combination of all three. Longer embedment lengths may lead to anchor failure. In mining and
tunneling, cable bolts are usually four to eight meters long; thus, axial failure mode would be
among any modes of B to E in Figure 2-4.

Figure 2-4: Bonded Anchor Failure Modes (adapted from (Droesch, 2015))

13
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Pull testing in the mining industry follows the same concept as in civil engineering; however,
there are slight differences. In both mining and civil engineering, the pullout test aims to
investigate the bond between three main components, tendon, encapsulation material (such as
grout and resin), and the host medium (Rock or soil, and concrete). In civil engineering
applications, bolt are used with concrete or other homogenized cementitious materials. Also,
due to the short embedment of bolts in civil engineering, cone failure more common scenario
of failure of anchors. Unlike homogenized concrete, the ground condition and quality are
unpredictably changing, so the bond quality between all three components (tendon, grout, and
ground) may vary in each section of the anchor. Frequency of discontinuities, groundwater
existence, apertures, heterogeneity of intact rock, and rock properties such as roughness and
strength are some of the effective parameters in the quality of bonds. These make the failure
modes of tendons installed in heterogeneous ground hardly predictable. Jeremic and Delaire
1983 and Stillborg (1984) categorized the possible failure of the cable bolt system and provided
the following possibilities of failure (Figure 2-5):

A. Cable rupture,
B. Cable/grout interface failure,
C. Grouting material failure,
D. Grout/rock interface failure, and
E. Rock mass surrounding the
borehole.

Figure 2-5 Possible failure modes of the cable bolting system (Jeremic and Delaire, 1983)

While traditionally, the most critical part of a bolted system was regarded as the bolt itself,
Tincelin (1991) believed the most important part of a tendon-reinforced system is the bond
between the tendon and the surrounding rock. Nothing else matters if this bond is insufficient,
as load transfer cannot occur. This bond initially fixes the cable in place and provides normal
load transfer and, later, when it fails, provides frictional resistance against axial movement.
According to this theory, several analytical (Hyett et al., 1995; Ren et al., 2010; Z. Wu et al.,
2010) and numerical models (Delhomme & Debicki, 2010; Ma et al., 2016) have been
developed to analyze the failure of the cable and grout interface.

Studies on the pullout failure of reinforced concrete have brought up a debate among
researchers about possible failure modes. Ottosen (1981) concluded that pullout load directly
depended on the compressive strength of concrete. Bickley's in-field experiments were carried

14
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

out based on ASTM C900-87. Results revealed a linear correlation between the concrete
strength and the ultimate pullout force; however, there was no relationship between them
(Bickley, 1993). Ballarini et al. (1986) utilized linear-elastic fracture mechanics in a two-
dimensional model and this determined that ultimate strength depends on fracture toughness.
Studies by Stone and Carino (1984) revealed that although there is an excellent correlation
between ultimate pullout load and compressive strength, these two are not directly related to
each other. In fact, according to their experience, failure occurs when sufficient aggregate
particles have been pulled out of the mortar matrix. Also, both compressive strength and
ultimate pullout load are influenced by this factor (Stone & Carino, 1984). Non-linear fracture
mechanics and the discrete cracking in finite element study by Hellier (1987) did not show any
relation between pullout failure and compressive failure of concrete.

All of the above debates resulted in more research on the failure mechanism of different
tendons, including rock bolts and cable bolts, and assessment of affecting parameters on the
pullout behaviour of tendons. This chapter reviews the literature studies on characteristics of
rock bolts and cable bolts under axial loading. Even though the main focus of this thesis is on
cable bolts; however, if there was no relevant study for the cable bolts, similar studies on the
rock bolt were substituted.

Ground Anchor Components


Ground anchorage systems consist of two main components, the bolt, and the anchorage
system. Bolts can be divided into two main groups of rock bolts (Dowel) and cable bolts.
Tendons can be anchored to the surrounding rock in three mechanisms: mechanical anchors,
frictional anchors, or grout encapsulation (Cybulski & Mazzoni, 1989; Hoek & Wood, 1988;
Rajapakse, 2008). A variety of tendons and encapsulation materials have been developed for
different loading conditions. This section briefly reviews the components involved in a ground
anchor system.

2.2.1 Rock Bolts and Cable Bolts


Rock bolts and cable bolts (also called tendons) are referred to as the high-tensile strength
elements that support the immediate roof of tunnels or loose blocks of rocks.

15
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-6: Ground support system with rock bolts(S. Yan et al., 2019)

As ground support, tendons are to prevent discrete rock block displacement (Hudson, 2003). A
wide range of rock bolts and cable bolts have been produced for different purposes. Windsor
(1992) listed different types of cables based on modifications applied to them. Khaleghparast
(2021) provided a list of cable bolts produced for the mining industry in Australia by three
main suppliers, Jennmar, Megabolt, and Minova. Mechanical properties of the cables were
mentioned in his shortlist. According to the literature, the main components of every rock bolt
or cable bolt are: a rock bolt or a strand, a bearing plate, and Barrel and Wedge (B&W) (Figure
2-7). Nuts might be replaced with B&W in rock bolts.

Cable bolt Rock bolt


Figure 2-7: Main components of cable bolts

Modifications to the cable profile were mainly made to increase the efficiency of cable bolts
under different loading conditions. Some of the modifications resulted in the introduction of
indentation, insertion of buttons or steel disks to create swaged cables (Schmuck, 1979), and
the use of various bulging cavities along the cables with different terms such as bulb (Garford
Pty Ltd, 1990), ferruled (Renwick, 1992), nutcage (Hyett et al., 1993), or birdcage (Hutchins
et al., 1990), and sheathing or coating of the cable by epoxy (Dorsten et al., 1984). Barley
(2000) described nutcage cable as a smaller version of birdcage cables that can be installed in
pairs in boreholes. Minicage cables are made by compressing cables at certain points, so the
stands create a bulging shape. Birdcage cables are made by unwinding strands at certain points.

16
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

As a comparison, nutcage cables are made using a different technique in which the nuts or
spacers are inserted at certain locations along the cable around the central wire so that when
the strands are wound together, an empty enlarged shape is created. Some of the common
modifications on the strand profile can be summarized in Figure 2-8.

Single strand Plain Indented Drawn

Coated Sheathed Coated Encapsulated

Birdcaged Antinode Node

Bulbed Antinode Node

Ferruled (nutcage) Antinode Node

Figure 2-8: Common modifications on cable bolt profile, adopted from (Barley & Windsor,
2000; C. Windsor, 1992)

It should be noted that modifications of tendons are not limited to the above mentioned options,
and many different cable bolts and rock bolts with different anchor systems have been
registered as a patent. These designs were introduced for specific applications and were not as
common as other types. Combination cable bolts (Eaton et al., 1998), tensionable cable bolts
(Stankus, 2001), and mechanically anchored cable bolts (Hedrick, 2005) are some of these
(Figure 2-9).

The earliest rock bolts were made of wooden or bamboo-grouted dowels (Mah, 1994).
Nowadays, tendons are usually made of steel, galvanized steel, and fibreglass. Hoehn (2021)

17
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

designed a carbon fibre cable bolt for ground support. The structure of a cable bolt is mainly
made of several wires (at least seven) woven around a king-wire or a hollow tube. The surface
of the wires can be smooth or indented to increase the frictional surface (S. Tadolini et al.,
2016). Figure 2-10 from Satolla (Satola & Hakala, 2001) compares different types of bolts with
regard to their physical features.

a)

b)

c)

Figure 2-9: Some of the registered patents for modification of cable bolts (Eaton et al., 1998;
Hedrick, 2005; Stankus, 2001)

Figure 2-10: Different bolt types in terms of physical features, left to right: Rebar, Bulbed
cable, plain cable, hot-dip galvanized steel cable, epoxy coated steel cable (Satola & Hakala,
2001)
18
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

2.2.2 Grout
Encapsulation material or grout is mainly divided into cementitious grouts and resin. The
strength of the cementitious grout is dependent on the percentage of water (Aziz, Craig, et al.,
2016; Aziz, Majoor, et al., 2017; Mirzaghorbanali, 2019), additives and composition
(Mirzaghorbanali, 2019; Quanji, 2010) and curing time (Hassani et al., 1992; Mirza et al.,
2016).

Goris (1990) studied the effect of the W/C ratio and reported the higher the Uniaxial
Compressive Strength (UCS) of the grout, the higher the peak load. Comprehensive studies on
the effect of the W/C ratio on cement-based grout strength have been studied by researchers
such as (Benmokrane et al., 1995; J. Chen, 2016; Kilic, Yasar, & Celik, 2002; Mirzaghorbanali,
2019). Studies showed that there is an optimum value for the W/C ratio, and excess water
results in a decrease in the grout strength. According to Kilic et al. (2002), this is due to the
fact that extra water which was not used by the cement for hydration tends to evaporate during
the curing process and creates inhomogeneous internal structures as a result of capillary
porosity. Chen (2016) conducted a series of tests on cement-based grouts to study their
mechanical properties. Based on the results of uniaxial compressive strength tests, it was
concluded that all the strengths and young moduli were reduced by increasing the W/C ratio.
This was also concluded by Hassani (1992), Chen and Mitri (2005) and Mirzaghorbanali
(2016).

Figure 2-11: Effect of w/c ratio on UCS strength (J. Chen, 2016)

Based on the results of the experiments carried out by Hyett (1995), a lower W/C of 0.3
compared to 0.5 could have more than 50% improvement in the load capacity of the system
but causes higher radial dilation.

19
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

The curing time of cementitious grouts can be regarded to have the same effect as the strength
of the grout because grouts will increase in strength the longer they are cured. However, it is a
known fact that cement grout does not have a linear curing curve, and in the first seven days,
the strength increases much faster relative to the growth from one week to four weeks. As
reported by Mirzaghorbanali (2016), for a thixotropic grout (Stratabinder), 28 days of curing
time only contributes a small amount to the load capacity of the cable bolts in pullout tests.

Resin can be divided into the two main types; being of oil-based and water-based. Rock anchor
resin grout is comprised of two parts, including a filled polyester resin and catalysed filler
(Minova, 2020). The strength of the various resins can be related to the chemical composition
and fillers (Aziz et al., 2013, 2014). The main difference between resin and cementitious grout
is the curing time. While it takes a few days for cement-based grout to reach its maximum
compressive strength, resin sets in a few minutes and reaches high strength values in a few
hours (Pritchard & McClellan, 2011; Pullan & Hagan, 2018).

Development of Laboratory Pullout Testing Technologies over


Time
This section particularly aims to review various laboratory testing facilities being developed
over time. Facilities have mainly been divided into two main groups of static and dynamic
devices and have been sorted out chronologically.

2.3.1 Previous Static Pullout Testing Apparatus


The concept of pullout testing exists in civil and mechanical engineering as well as mining and
tunneling engineering; however, the application is different. Nonetheless, the purpose of all
pullout testing facilities is to evaluate the quality of the bond between tendons and the
surrounding media. In civil engineering, the pullout test can be destructive or non-destructive
and has several purposes:

• To evaluate the strength of concrete (not the tendon) during the pullout practice,
focusing on the damaged zone of the concrete sample (failure cone).
• To investigate section failure of the reinforcing bar. The steel bar is pulled out along
structural adhesive (resins or glue) in this mode.
• To assess the combination of both above in which the upper part of the concrete cone
is damaged, and the lower part is pulled out along the interface of structural glue and
concrete interface.

20
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Table 2-1 summarized the development of different static pullout tests sorted chronologically.
The main feature of each setup and the main issue accompanied by each one have been
mentioned as comments.

21
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Table 2-1: Summary of apparatus and test methodologies used in the literature to study the
axial performance of bolts

ASTM C234-71
(Oland & Callahan, 1978)
Features: 152 mm cube encapsulation. Max
diameter of rebar: 19.05 mm

Comments: setup failed to produce reasonable


results as all concrete cubes broke laterally. Lack
of existence of a confinement system was the main
issue.

Single embedment pullout test


(C. Yan, 1992)

Features: Used for steel rebars.

Comments: setup was not strong enough to test


high-strength bolts

ASTM C900: Standard pullout test of


hardened concrete
(ASTM C900, 2013)

Features: Initially developed for testing concrete


strength. Bar was cast in place.

Comments: setup was inspiring for further in-field


studies (Ottosen, 1981; Stone et al., 1986).
Encapsulation length was short and not
representative of the actual rebar pullout test.

22
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts
Split-pull Setup
(Fuller & Cox, 1975)

Features: Double embedment. Designed for cable


bolts. Cementitious grout was used. The embedment
length was variable from 100 to 700 mm. Constant
stiffness was applied from the steel pipe.

Comments: Due to the use of thick steel pipes as


confinement, results were typically greater than in
situ tests.

Long embedment pull test


(Stillborg, 1984)

Features: Rigid steel confinement was replaced with concrete to simulate the rock mass.
The embedment length could be up to 25 times the cable diameter.

Comments: Although the confinement was working more realistically, the free end of the
cable was not restrained against free rotation, which could significantly affect the results.

23
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts

Double embedment setup


(Hutchins et al., 1990)

Features: two equi-length pipe tubes


anchored with the bonding agent.

Comments: Equal embedment length


gave an equal chance of failure at
either of the two embedded sides and
made the laboratory measurements
difficult. Steel pipes overestimate the
strength of the host medium

Modification of Hutchins pull test


(Hyett, Bawden, & Reichert, 1992)

Features: stress concentration at the threaded


pulling head was addressed by resting the
support test section. Various confining
materials such as steel, aluminum, PVC, and
plastic heat shrink were used. The embedment
length of the two sides was chosen differently
so that pullout could occur on the shorter side.

Comments: pulling head was covering the grout surface at the pullout zone, which was
influencing the pullout initial stiffness. Chosen materials for confinement have a plasticity
limit, which cannot resemble hard rock conditions. The minimum speed of the test was 0.3
m/s (18 mm/min) which was not representing the static test.

24
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts

Modification of Hyett pull test


(MacSporran, 1993)
Features: constant radial pressure
was added by a Modified Hoek
Cell (MHC)

Comments: the new design led to


a 25 mm unbounded cable length
between two embedded parts. This
could result in tensile failure of the
tendon prior to pullout. The
embedment length of the sample
was restrained to 250 mm due to
the size of the Hoek cell.

Short and long encapsulation pull test


(Benmokrane et al., 1995)

Features: similar to the design of Stillborg (1984), short and long embedments were 4 𝞍
and 20 𝞍 respectively. (𝞍=15.8mm).

Comments: The loading frame was limited to 270 kN with a rate of 5 kN/s, which is only
suitable for low-strength cable bolts. A loading rate of 5 kN/s causes the test to be completed
in less than one minute, which is relatively quick for a static pull test.

25
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts
Pullout test setup
(Ito et al., 2001)
Features: Both rock bolts and cable bolts could be
tested. A hollow-ram jack with a loading rate of 0.05
kN/min was used. Artificial rock
(500*500*1000mm3) was used to simulate the real
field condition.

Comments: the UCS and elastic modulus of the


rock material was 82.3 MPa and 35.1 GPa,
respectively, which represent the hard rock
condition only

Short encapsulation pull setup


(Aoki et al., 2002)
Features: used for testing of bulbed cables.
Encapsulation length was 350 mm. Used steel tube
as confinement.

Comments: tests with bulbed cables ended up with


the failure of steel confinement, and the study was
not successful.

Long embedment pull test


(Aoki et al., 2003)

Features: Used for testing the bulbed


cables. Loading rate was 10 kN/min

Comments: loading rate was not


representing static pull tests.

26
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts

Biaxial loading pull test


(Clifford et al., 2001)

Features: biaxial cell was used as


confinement that could apply various
external pressure up to 10 MPa during
the test. The cylindrical sandstone
sample was 142 mm wide and 320 mm
long.

Comments: Results were overestimated


and higher than field and laboratory tests
due to the rigidity of the confinement.

ASTM F432-19
(ASTM.F432, 2004)
Features: determines the effectiveness
of chemical grouting by defining the
speed index (earliest time when
anchorage load reaches to 4000lb)

Comments: steel confinement


overestimates the rock loading capacity.
Only used for cable bolts

27
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts
Bed separation pull test
(Weckert, 2003)

Features: load transferred through the biaxial cell to a rear plate bolted to the bottom of the
samples to simulate bed separation.

Comments: there was no confinement or load at the top surface of the samples.

28
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts
Constant pressure confinement pull test
(Moosavi et al., 2005)

Features: a new Hoek Cell was instrumented internally by cantilever strain gauge arms to
measure the dilation during the tests while the pressure was kept constant. The diameter of
the samples were 61 mm, and the embedment length was 100-150 mm.

Comments: encapsulation length was too short, and the diameter was limited to 61 mm.
The setup was used only for rock bolts.

Double embedment pullout test setup


(Satola & Hakala, 2001)

Features: designed for both rock bolt and cable bolt with long embedment length (2000
mm).

Comments: maximum loading capacity was 350 kN. The maximum pullout was limited to
300 mm. Steel pipe was used as confinement.

29
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts

British Standard 7861-1


(BS 7861-2, 2009)
Features: The first standard test procedure
included cable bolts. Consisted of two 125
mm anchored sections with no gap in between.
Tubes were internally rifled (threaded) to
induce failure at the cable/grout interface. The
setup would fit inside a tensile testing
machine. Rotation of cable bolt was stopped
using a pin between the two sections.

Comments: similar to the setup of (Hutchins


et al., 1990), the use of a metal pipe proved
inconsistency between the laboratory and field
data.

Update of British Standard 7861-1


(Clifford et al., 2001)(BS 7861-2, 2009)
Features: Used rock samples for realistic
results. The loading rate of 1mm/min is set to
be representative of static tests. A biaxial cell
with 10 MPa lateral pressure was used.

Comments: the standard did not consider


other loading conditions, such as dynamic
loading. The biaxial 10 MPa pressure was
deemed unrealistic especially considering the
long service life of cable bolts. Rotation was
not controlled.

30
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts
Modification of (Clifford et al., 2001) pull
setup
(Thomas, 2012)
Features: Included anti-rotation tube and
B&W. Biaxial load cell were removed, and
sandstone (142 mm diameter) was used as
confining material. The Embedment length
was 320 mm.

Comments: was not strong enough for long-


embedment tests. lateral confinement was not
high enough and radial crack obsevered.

Improved confinement pull test


(L. B. Martin, 2012)

Features: pressure vessel and bladder were


used to apply confinement pressure and
constant stiffness up to 25 MPa. Rock
sample was used for encapsulation. Both
resin and grout were tested.
Comments: Although the setup was
designed flawlessly, the testing embedment
length hardly exceeded 250 mm in the tests.
In the tests using resin as an encapsulation
material, the embedment length was 130
mm.

31
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts
Unconfined single embedment setup
(Holden & Hagan, 2014)
Features: No confinement for concrete was
used. Embedment length 280 mm.

Comments: concrete samples failed


radially, and the test results could not be
reliable. It was found that the confining plate
on top of the concrete sample must not cover
the grout column. Debonding occurred
between grout and concrete. The top plate
changed the results by adding extra
resistance against the pullout load.

Improvement of
(Holden & Hagan, 2014)
(J. Chen, 2016)

Features: an external split


pipe confinement was used to
guarantee constant stiffness
by tightening ten bolts (40 N-
m of torque on each).

Comments: the hairline gap between split pipes was big enough to cause radial crack
occurring aligned with the gap. The radial crack could be seen in most samples, especially
in tests with bulbed cable bolts.

32
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts
Improvement of
British Standard 7861-1
(Aziz, Mirzaghorbanali, et al., 2016)
Features: setup consists of two
internally threaded embedment
sections, one longer (230 mm) than the
other side, to impose failure on the
shorter section (170 mm). Anti-
rotation system used for cable bolts.

Comments: steel confinement


overestimated the stiffness of the real
field host medium.

Improvement of (J. Chen, 2016)


(Hagan & Li, 2017)

Features: an un-grouted section was added


to the end of the borehole, so the engaged
length stays constant. The encapsulation
length was increased to 360 mm.

Comments: radial cracks with the bulbed


cable bolts were affecting the test results.

33
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-1 Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts
Improvement of
(Hagan & Li, 2017)
(Rastegarmanesh, 2022)
Features: Similar to (Hagan & Li,
2017). A more robust loading jack
was used to test stronger cable
bolts.

Comments: Radial cracks along


with the hairline gap of the
confinement steels were inevitable.

Multi-purpose tendon testing


setup
(Kang et al., 2020)
Features: can test different loading
axial rates simultaneously with
shear loading.

Comments: setup has been


designed only for rock bolts.

2.3.2 Previous Dynamic Pullout Testing Apparatus


The number of laboratory studies is substantially smaller for monotonic and strongly dynamic
loading, such as impact loading or blasting. Designs of dynamic pullout testing of tendons in
the laboratory were based on gravitational force and free fall induced energy. Hadjigeorgiou
and Potvin conducted a comprehensive study on some of the dynamic pull testing facilities
(Hadjigeorgiou & Potvin, 2007). Most of the developed dynamic testing facilities work based
on the free fall of the drop hammer on the testing sample. Yi and Kaiser (1994b), Kaiser et al.
(1996), Player (2000), and Ortlepp and Swart (2002) stated that the main drawback of a drop
test is that energy dissipation and unwarranted shaking in the loading frame cannot be
measured. In addition, a great inertia force is transferred in the moment of impact, adding
complexity to the analysis of dynamic tests. Banthia (1989) explained that the inertia force

34
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

could be neglected only when the acceleration of the sample remains considerably small. Table
2-2 updates the list of dynamic pullout testing facilities developed for laboratory or in-field
experiments to the best of the knowledge of the author. There still might be some more setups
missing from the Table 2-2. Also, this is a shortlist of the feature, and readers are always
encouraged to read the main reference for more details. Table 2-2 has been sorted
chronologically.

35
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Table 2-2: Summary of apparatus and test methodologies used in the literature to study the
axial performance of bolts under dynamic loads

Split Hopkinson bar testing equipment


(Vos & Reinhardt, 1982)

Features: initially used for impact tensile


strength of concrete. Then it was modified for
the dynamic debond test.
Comments: only embedment lengths shorter
than 200 mm could be tested. Only rebars were
tested.

University of British Columbia Impact test


(C. Yan, 1992)
Features: Based on the transformation of
energy by the free fall of the hammer. It was
aimed at the testing of plain, polypropylene
fibre-reinforced, and steel-reinforced concrete.
A 345 kg hammer hits the sample from a 2.4 m
height. This weight later increased to 505 kg.
Both push-in and pullout of the rebar could be
done by removing and adding the anvil support
Comments: the static tests of these studies
could not be done with the same facility. The
embedment length was no more than 63.5 mm.

36
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-2: Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts under dynamic loads

CanMet Dynamic test facility


(Plouffe et al., 2008)
Features: Based on the transformation of
energy by the free fall of the hammer. The
Hammer weight was 3000 kg, and the
maximum height could be 2.1 m. The
maximum applied energy was 62 kJ.
Comments: the static tests of these studies
could not be done with the same facility.
Steel pipe and grout were used as
confinement which can overestimate the
results.

SRK setup 1
(Ortlepp & Stacey, 1997)

Features: introduced for dynamic testing of wire mesh and shotcrete. Could apply 70 kJ
impact energy with 8.8 m/s velocities.
Comments: the current design was not targeting the performance of tendons.

37
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-2: Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts under dynamic loads

SRK setup 2
(Stacey & Ortlepp, 1999)
Features: SRK setup 1 was
redesigned for rock bolt
dynamic axial testing. 80 kJ
energy could be transferred.
Comments: measurement of
the absorbed energy was not
easy due to the complexity.
There was no method to
measure load. Displacement
and failure were assessed by
photos before and after the test.
Only used for rock bolts.
SIMRAC project 616 (Terra-tek)
(Güler et al., 2001)

Features: It was designed for the dynamic loading of the rock mass. 30 kJ of energy (1000
kg hammer free falls from 3 m height) could be transferred.
Comments: It was not designed for tendons testing. Measurement of load and the absorbed
energy was not practical.

38
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-2: Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts under dynamic loads

Louisiana State University


(Weathersby, 2003)

Features: was designed for both static and dynamic pullout tests of rebars. A 200-Kip static
and dynamic Loader was utilized for pulling a 500 mm long rebar out of concrete samples.
Almost 900 kN (200,000 lb) of load could be applied in less than 2 ms.
Comments: The majority of 33 tests carried out by this device resulted in radial failure of
the concrete sample or rupture of the steel before being pulled out.
Noranda Technology Centre
(NTC) impact test rig
(Gaudreau et al., 2004)
Features: was designed to test
modified cone bolts. The drop mass
was 1000 kg, and the maximum drop
height was 2 m leading to a
maximum 20 kJ of energy.
Comments: tests were mainly
unsuccessful due to the failure of the
rock bolt on threads or nuts prior to
being pulled out.

39
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-2: Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts under dynamic loads

SRK/Duraset wedge-block loading device


(Ortlepp & Erasmus, 2005)

Features: used 1.2 m long cable bolt for testing. Could apply 390 kJ energy with a 10000
kg weight free falling from the maximum height of 4 m. Maximum velocity was 8.9 m/s.
Comments: Using steel pipes as confinement overestimated the results. Limited results are
available.
Royal Institute of Technology in
Stockholm
(Ansell, 2005)
Features: Based on momentum
transformation by the free falling of
the sample. Free falls were halted
by an impact between two large
steel beams spanning the shaft used
and a 1.5 m long beam attached to
the nut of the rock bolt. The
maximum drop height was 5 m.
Comments: Only rock bolts were
tested. In velocities higher than ten
m/s rock bolt, tensile failure has
been recorded.

40
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-2: Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts under dynamic loads

Minerals & Research Institute of


Western Australia (MRIWA)
(Player et al., 2004)
Features: Based on momentum
transformation by the free falling of the
sample hitting the buffers. The weight of
the free-falling part was about 4500 kg,
The maximum velocity was 10 m/s, and
the equivalent applied energy was up to
225 kJ.
Comments: velocities more than 7 m/s
caused damage in the buffers. Results were
totally affected by the stiffness of the
buffers.
Steel tubes and grout were used for
encapsulation which overestimated the
results. The setup was not designed for the
static test to compare the results.
Hopkinson bar techniques impact test
(Solomos & Berra, 2010

Features: monotonic dynamic loads with minimal rise time and a plateau of sufficiently high
amplitude were applied. The embedment length was 100 and 200 mm.
Comments: dominant failure modes in their test were concrete failures. Only tested rebars
and static tests could not be undertaken.

41
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Continuation of Table 2-2: Summary of apparatus and test methodologies used in the
literature to study the axial performance of bolts under dynamic loads

Dynamic Impact Tester (DIT)


(Crompton et al., 2018)

Feature: similar to one designed by (Kaiser et al., 1992). Two different modes of single and
double embedment with a maximum length of 3.5 meters were implemented. The drop mass
could be in the range of 551 to 3171 kg, and the maximum drop height was 2.1 m. The
maximum impulse energy of 65 kJ with the equivalent impact velocity of 6.42 m/s
Comments: effect of rigid confinement leads to an overestimation of the tendons' capacity.
Multi-purpose tendon testing setup
(Kang et al., 2020)

Features: uses a pendulum to apply the impact load.


Comments: setup has been designed only for rock bolts. The impact load is limited to 11 kJ,
which is insufficient for cable bolt testing.

42
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Pullout Mechanism
Many different experimental, analytical, and numerical tools have been used to study and
interpret the failure mechanism of tendons during the pullout test. According to Mylrea (1948),
the bonding mechanism for smooth round bars is as follows. Initially, the load is carried by
adhesion. With increasing load, more of the bar begins to slip, resulting in a reduced adhesive
bond on a smaller portion of the yet-bonded length. As the load increases, the location of the
maximum bond stress in the pullout specimen moves toward the unloaded end of the bolt while
maintaining a constant value. As the whole bar slips, the bond stress is nearly uniform along
the length of the bolt. Upon further slip, the bond stress intensity gradually diminishes.

Potvin (1989) stated that the interface between the cable bolt and grout is the most plausible
failure mode during pullout loading. Potvin deduced that the cable/grout interface has less
surface area, which has a direct correlation to fewer shear bonds. Hence, the properties of the
surface of the cable and the grout properties are determinative in the pullout mechanism.

Martin et al. (2000) conducted a series of pullout tests on a seven-wire cable bolt. The king
wire was instrumented using 10 strain gauges at various lengths of up to 1.22 m, and the cable
was embedded in concrete blocks using resin-based grout. The findings show higher strain
recorded near the collar and decrease in strain values recorded by further gauges. Also, the
interpreted load based on the strain gauge outputs was relatively close to the measured load.
The distance between the curves gets slightly bigger for further strain gauges, indicating a
nonlinear decay in load values. Furthermore, the last strain gauge did not record any value,
which means the bond was intact.

Studies by Hyett et al. (1995) and Aoki et al. (2002) on the pullout test of 15.2 mm cable bolts
(also called seven-wire strand (Dorsten et al., 1984)) showed the pullout mechanism of the
cable includes linear increases in load in response to the debonding force followed by a peak
load and then, a post-peak behaviour. While in Hyett et al. (1995), post-peak behaviour showed
a progressive increase of load and a sudden drop and oscillations (Figure 2-13). Aoki et al.
(2002) experienced an increase in load up to a certain amount and then followed by a steady
residual condition (Figure 2-13). On many occasions, researchers ended up with the tensile
failure of a bulbed 15.2 mm cable before recording any pullout displacement (Aoki et al., 2002;
Thenevin et al., 2017).

43
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-12: Calculated load and strain gauges on the king-wire of the 15.2 mm cable bolt (L.
A. Martin et al., 2000)

Thenevin et al. (Thenevin et al., 2017) experimented with pullout tests on different tendons,
including FRP3 and HA254 rock bolts as well as plain and bulbed cable bolts. In their study,
the pullout failure mechanism of rock bolts was analysed in terms of the following parameters:
Yield bond strength, peak or maximum load and residual load (Figure 2-14). Yield bond
strength is inferred as the load in which the slope of the load-displacement curve falls below a
stiffness of 20 kN/mm. Also residual load was defined as the load measured at an axial
displacement of 50 mm.

However, the pullout mechanism of cable bolts was not similar to rock bolts. It was concluded
that the failure mechanism before the peak load in both types of tendons is similar (interface
adhesion, friction, and mechanical interlock), however, the post-peak behaviour was followed
by a small load drop (due to cable geometry) and further followed by a progressive increase or
reduction of load with small oscillations (Figure 2-15). The residual load in the cable bolt might
be larger than the peak load as a result of different factors such as embedment length,
confinement pressure and cable geometry (Thenevin et al., 2017).

3
Fibre-Reinforced Plastic
4
Threaded steel bar manufactured by Riva Acier

44
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-13: Pullout mechanism of 15.2 mm cable bolts, Left: plain cable bolt and cement
grout (Hyett et al., 1995), Right: P=plain, B=Bulbed cable bolt and cement grout (Aoki et al.,
2002)

Figure 2-14: Typical pullout mechanism of FRP rock bolts (Thenevin et al., 2017)

Figure 2-15: Typical pullout mechanism of Reflex cable bolts with different embedment
lengths (Left) and confinement pressure (Right) (Thenevin et al., 2017)

45
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Li (2019) expected cable/grout failure in their experiments; however, two more failure modes
occurred unexpectedly. Grout/confinement interface failure and cable bolt failure were two
additional failure modes. Different failure mechanisms were compared in Figure 2-16. As can
be seen, for the case of cable/grout failure mode, it is the only mode in which the applied load
had a progressive increase followed by a steady state after adhesion failure.

Figure 2-16: Comparison of different failure modes (X. Li, 2019)

Mah (1994) carried out pullout tests on newly developed fibreglass cable bolts and presented
a typical load-displacement curve for uncoated POLYSTAL5. In this study, the pullout
mechanism was divided into three main stages. Stage one was dominated by a progressive
adhesion failure between the tendon and grout. Stage two was dominated by what appeared to
be grout/cable interfacial shearing. It was believed that a layer of helically wrapped fibre or
perhaps a few layers of unidirectional fibre remained bonded to the grout. Stage three was
dominated by stick-slip failure mode between the cement and cable bolt (Figure 2-17).

5
Glass fiber reinforced unsaturated polyester

46
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-17: Pullout mechanism of uncoated POLYSTAL (Mah, 1994)

Benmokrane et al. (1995) performed a series of pullout tests on 15.8 mm cables (270 KN) and
15.8 mm rock bolts (200 KN). The authors suggested that the mechanical interlock for cable
bolts was not as high as rock bolts. The difference in load values between the two tendons
stems from the rotation of the cable bolts out of the concrete sample. Unlike cable bolts, rock
bolts did not rotate when being pulled out . The negative effect of the rotation on the pullout
load has also been mentioned by Stillborg (1984), Martin (2012), Cao et al. (2013), Hagan
(2017), Rastegarmanesh (Rastegarmanesh et al., 2022) and many more. The low torsional
capacity of cable bolts means they can store rotational energy and use this to complicate the
pullout procedure. This strain energy can be due to both winding and unwinding. Considering
the rotational movement of the cable bolts, Forbes and Vlachopoulos (2016) concluded that the
failure mechanism of the cable bolt can be summaraized into dilational slip between the cable
and the grout (due to the rotation), shear failure of the grout flutes (ridges) and the unscrewing
of the cable. Forbes and Vlachopoulos believed plain cable bolts almost always fail because of
the loss of bonding unless the confinement or the embedment length is high enough to induce
the cable rupture.

The comparison of the cable bolt and rock bolt pullout conducted by Martin 2012 suggested
that the post-peak behaviour of these two is fairly different. In contrast to rock bolts post-peak,
where most of the load is lost, cable bolts have a much higher or even identical load values (to
the peak load) in the post-peak. Martin believed this is due to the relatively smoother surface
profile of the cables compared to rock bolts which does not destroy the bonding agent as much.

47
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Also the oscillations seen in the cable bolts can also be associated with the surface profile of
the cable. This was consistent with Martin’s (2012) explanation for the more noticeable
presence of the oscillation in the FRP rock bolts compared to HA24 bolts, which was attributed
to the smoother indentation angle and the softer fibre-reinforced polymer.

Stillborg (1984) suggested that the residual behaviour of the cable is dominated by the cable
strand moving inside and crossing over the grout ridges (flutes - cable imprints). This justifies
the wavy behaviour in the post-peak section of pullout test graphs. Chen (2016) also interprets
the oscillation seen in the peak zone of the pullout curve as the slip-lock mechanism, in which
the mobilized bulb interacts with the grout ridges. After a certain point, this phenomenon
diminishes, and the frictional load decreases linearly due to the engaged length getting shorter.
The experiment showed that these oscillations only happened in the stronger confining medium
of the bulbed cable.

Figure 2-18: Performance of MW9 cable bolt in pullout test (J. Chen, 2016)

Martin (2012) also noticed an oscillation in the tests using FRP bolts encapsulated in resin.
Visual investigation of the samples showed that this phenomenon happens due to the rib
imprints on the grout not being destroyed fully during the test (Figure 2-19). It was believed
that the behaviour is due to the smoother structure of FRP in comparison to steel bars. Thenevin
et al. (Thenevin et al., 2017) mentioned that the evolution of oscillations depends on the
confinement pressure and slight untwisting structure of cable bolts.

48
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

FRP

Flexible Cable bolt

Figure 2-19: Pullout behaviour of FRP bars and cable bolts encapsulated in resin (L. B.
Martin, 2012)

Researchers defined different parameters to measure the stiffness of the pullout test, as there
has never been a unique testing method. Thompson (2012) used a single scalar parameter
defined as the Load Transfer Index (LTI), which was the maximum pullout load over the
displacement at which it is reached. Consequently, a high LTI can be achieved by high peak
loads at small displacements.

𝑃𝑒𝑎𝑘 𝑙𝑜𝑎𝑑
𝐿𝑇𝐼 =
𝐷𝑖𝑠𝑝𝑙𝑎𝑐𝑒𝑚𝑒𝑛𝑡 𝑎𝑡 𝑝𝑒𝑎𝑘 𝑙𝑜𝑎𝑑

It was also mentioned that LTI is not necessarily determinative of the behaviour, and the load
curve must be studied (Thompson et al., 2012).

Yuan et al. (2004) suggested that “ultimate load” refers to the highest peak load measured
during the pullout test, and “bond strength” refers to the shear strength of the interface of grout
and the reinforcement element. Hence, it is crucial to use uniform terminology to avoid
confusion.

49
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

In the experimental studies of Chen and Mitri (2005), the concept of “bond shear strength” was
defined in the relationship with pullout load (P), embedment length (L) and circumference of
the bolt (c). A similar concept was employed by (L. B. Martin, 2012).

𝑃
𝜏= Equation 2-1
𝑐(𝐿 − 𝑠)

where s is the slippage length. While the equation was offered in the study of seven-wire cable
bolts, the equation simply considers the surface of the cable bolt as a smooth surface without
considering the possible shear resistance generated by the geometry of the surface of cable
bolts.

Salcher and Bertuzzi (2018) conducted 181 field pullout tests on different tendons, including
GRP rock bolts and MW9 cable bolts. Cement grout and resin was selected as an encapsulation
material, and tests were carried out in different lithology, including sandstone, shale, and
interbedded sandstone and shale. In order to compare the test results, Pullout test results were
normalised with respect to the bond length to allow a comparison between the stiffness of
different tests.

Yazici and Kaiser (1992) developed the Bond Strength Model (BSM) theory for cable bolts, as
it was believed that previous models were mostly developed based on rock bolts. According to
the theory, the bond strengths of bolts are primarily frictional and depend on the pressure at the
bolt-grout interface. The pressure increase at this interface is a function of the grout dilation or
radial movement caused by the rough surfaces of twisted cables. Consequently, normal stress
in that interface is of importance. While BSM is regarded as one of the most prolific
contributions to the theoretical modeling of the cable bolt to date, it was only focused on plain
cables and neglected the effect of rotation, stress changes and the relationship between the axial
and lateral displacement at the interface. Also, calculating the dilation is difficult for
calculations. Later in 1993, Diederichs (1993) added the effect of rotation to BSM.

Analytical Studies
Analytical models of the pullout behaviour of tendons have mainly focused on either the
interaction of the tendon and the encapsulation material or the failure of the encapsulation
material from the host medium (rock). As a simplified model for interaction between the steel
bar and concrete, (Hawkes & Evans, 1995) introduced a theoretical model by assuming a linear

50
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

relationship between the bond and axial stress for a smooth steel bar encapsulated in concrete.
Experimental results showed significant agreement with the theoretical model in Equation 2-2:
d2 f  4 Al

P= 1 − e d
 Equation 2-2
4A  
Where d is Bar diameter (m), f is bond stress at the free end (Pa), I is the distance where the
bond stress reaches zero and A is constant.
Efforts of Mylrea (Mylrea, 1948) and Nilson (Nilson, 1972) in determining equations for
finding the bond stress in samples of rebar and concrete showed that the bond stress at any
point is a function of slip at that particular point and that the bond-slip curve cannot be used
for any other point along the bar. However, Nilson proposed that the maximum bond stress for
steel rebar and concrete can be equated as:

𝑢 < (1.43𝑐 + 1.5)𝑓𝑐′ Equation 2-3

Where 𝑓𝑐′ = concrete strength in psi, and c is the distance from the loaded end, in inches. The
proposed equation did not consider other effective parameters, such as rebar diameter and
confining pressure.

During the 1980s and 1990s, several researchers, such as Aydan et al. (1985), Rajaie (1990),
and Hassani (1992) proposed different tri-linear models to delineate the shear load distribution
along the encapsulated length of both rock bolts and cable bolts during the pullout. Aydan et
al. studied the effect of the strength of confinement, and Rajaie (1990) developed the model to
study the load distribution along the encapsulation length.

Pullout experiments of Benmorkrane (1995) on cable bolts with an embedment length of eight
times that of the cable radius showed that the pullout failure curves of tendons could be divided
into the following tri-linear model for the failure of tendons (Figure 2-20).

1. Ascending elastic response corresponding to the elastic shear stress and strain along the
cable,
2. Descending plastic response due to decoupling and high shear slip and
3. Horizontal response due to the residual resistance corresponding to the frictional
behaviour.

51
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-20: Idealized bond-slip model (Benmokrane et al., 1995)

Figure 2-20 can be presented as Equation 2-4 below, where F(W) is the pullout force, Rb is bolt
diameter, and L is the encapsulation length.

1
 W for 0  W  1
 1
  −   − 
 b (W ) =  1 2 W + 2 1 1 2 for 1  W   2 Equation 2-4
 1 −  2 1 −  2
 2 for W   2


F (W )
 b (W ) =
2 Rb L Equation 2-5

Benmokrane assumed the shear load was uniformly distributed along the embedment length.
Similarly, Xiao and Chen (Xiao & Chen, 2008) also defined the same three phases (Figure
2-21) including: elastic phase (I) (Ge is the elastic shear modulus of anchorage), softening phase
(II) (k2 is the elastoplastic shear modulus of anchorage and the soil interface), and residual
phase (III) for shear stress between the anchorage system and the anchor during a pullout test.
Unlike Benmokrane (1995), they believed different lengths of the anchorage experience
different phases. For instance, as the pullout load increases, when L1 is in the elastic phase, L2
and L3 would be in softening and residual phases, respectively (Figure 2-21).

52
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-21: (left) Relationship between shear stress and strain of soils, (Right) Schematic
diagram of shear failure phases (Xiao & Chen, 2008)

Although their model could calculate interface shear strength for both the elastic and softening
stages as well as for the bolt axial stress, it could not determine the ultimate load capacity (Ren
et al., 2010). Wu et al. (2010) defined a tri-linear bond stress slippage model for FRP bars, and
grout similar to what (Benmokrane et al., 1995; Xiao & Chen, 2008) introduced (Figure 2-22).

In Figure 2-22,  is the shear slip (m), k = is the stiffness in the elastic stage (Pa/m), p =

shear strength (Pa), and  f is the frictional (residual) strength. In this model, three phases of
elastic (pre-peak), elastoplastic ultimate load drop and residual are defined based on the term
(Equation 2-6):

 = ms + f Equation 2-6
Where  is the shear stress at the bolt/grout interface, s = shear slip, m and f are coefficients
from the trilinear model (Equation 2-7).
 1
 m = m1 = s , f = 0 0  s  s1
 1

 1 −  2  S − S
 m = m2 = ,f = 2 1 1 2 s1  s  s2 Equation 2-7
 s1 − S2 s1 − S2
m = 0, f =  2 s2  s

53
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-22: Tri-linear law with characteristic parameters (Z. Wu et al., 2010)

Martin (2012) also proposed three theoretical models for each section of the pullout test curve
(Figure 2-23). Experiments by Martin were carried out using a biaxial cell that could preserve
either confinement stiffness or confinement pressure constant during the tests. Due to this
feature, a formulation was offered to calculate the interface radial pressure in the constant
stiffness test. Later, a mode for shear stress for the cable was introduced in which the two

components, one for adhesions and mechanical interlock (  c ) and one for friction (  v ) could

be added together to replicate the pullout behaviour of the rock bolts (Equation 2-8).

Figure 2-23: Slippage model of (L. B. Martin, 2012)

54
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

𝜏𝑏 (𝑊, Δ𝑝𝑏 , 𝑝𝑏0 ) = 𝜏𝑣 (Δ𝑝𝑏 , 𝑝𝑏0 ) + 𝜏𝑐 (𝑊, 𝑝𝑏0 )


𝜏𝑣 (Δ𝑝𝑏 , 𝑝𝑏0 ) = Δ𝑝𝑏 tan 𝜑(𝑝𝑏0 ) Equation 2-8
𝜏𝑝
𝑊 for 0 ≤ 𝑊 ≤ 𝑊𝑝
𝑊𝑝
2
𝜏𝑐 (𝑊, 𝑝𝑏0 ) = 𝑊𝑟 − 𝑊 Equation 2-9
𝜏𝑟 + (𝜏𝑝 − 𝜏𝑟 ) ( ) for 𝑊𝑝 ≤ 𝑊 ≤ 𝑊𝑟
𝑊𝑟 − 𝑊𝑝
{𝜏𝑟 for 𝑊 ≥ 𝑊𝑟
Where W is displacement along the Z axis.
The model reportedly was also valid for constant radial pressure cases; however, it is essential
to note that the tests did not include radial measurements to accurately study the radial
behaviour of the interface. Radial pressure values were calculated using a formula which
was rather less precise than experimental measurements. Furthermore, the theoretical
framework proposed by Martin (2012) was only applicable when the grout and the
confining medium were in an elastic state and the radial crack would not propagate. The
assumption is no longer valid if the sample cracks as the stiffness changes.

Yuan et al. (2004) proposed a bilinear bond-slip model and derived the full-range solutions for
a bonded joint between the fibre-reinforced polymer (FRP) composite and concrete. According
to this model, the bond shear stress increases linearly with the interfacial slip until it reaches
the peak stress 𝜏f at which the value of the slip is denoted by 𝛿1 . Interfacial softening (or micro-
cracking) then starts with the shear stress reducing linearly with the interfacial slip. The shear
stress reduces to zero when the slip exceeds 𝛿f , signifying the shear fracture (or debonding or
macro-cracking) of a local bond element. The absence of residual shear strength after
debonding implies that friction and aggregate interlocking over the debonded length of the joint
is ignored.

Figure 2-24: Bilinear local bond-slip model (Yuan et al., 2004)


55
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Equation 2-10 presents the proposed model in which

𝜏f
𝛿 when 0 ≤ 𝛿 ≤ 𝛿1
𝛿1
𝑓(𝛿) = 𝜏f Equation 2-10
(𝛿 − 𝛿) when 𝛿1 < 𝛿 ≤ 𝛿f
𝛿f − 𝛿1 f
{0 when 𝛿 > 𝛿f

A typical load-displacement curve was described in Figure 2-25. In the proposed model, Yuan
considered five steps, including 1) elastic (OA), 2) elastic-softening (AB), 3) elastic-softening-
debonding (BCD), 4) softening-debonding (DE), and 5) debonding.

Figure 2-25: Typical full-range theoretical load-displacement curve

Ren et al.(Ren et al., 2010) used the bilinear bond-slip model and the five-stage failure theory
of Yuan (2004) to propose a theoretical approach to predict the full load-displacement
behaviour of fully grouted rock bolts. However, the bilinear model was improved to trilinear
(Figure 2-26). Ren et al. stated that the horizontal line represents the non-zero residual shear
strength due to friction (Equation 2-11). In this model, k is the ratio of the residual strength 𝜏r
to the peak stress 𝜏f .

56
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-26: Trilinear bond-slip model (Ren et al., 2010)

Equation 2-11

As a result, deriving analytical solutions become significantly more involved, and the solutions
are more complicated. The basis of the model is the propagation of a softening length and a
shear slip δ from the loading end to the free end along the cable (Figure 2-27). The main
assumptions of the study stated that the bolt remains in an elastic state during the whole test
and axial displacement of the bolt always equals the shear slip of the grout bolt interface.

57
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

(Fig. 5)

Figure 2-27: Evolution of shear stress. (a, b) Elastic stage; (c, d) elastic–softening stage; (e, f)
elastic– softening–debonding stage; (g) softening–debonding stage; (h, i) debonding stage; I,
II and III represent elastic, softening and debonding stress state, respectively (Ren et al.,
2010)

Using the tri-linear model previously developed by We et al. (2010) and the shear stress
evolution theory of Yuan (2004), Chen et al. (2015) proposed an extended model with five
stages including 1) elastic, 2) elastic-softening, 3a) elastic-softening-debonding, 3b) pure
softening, 4) softening-debonding, and 5) debonding. Figure 2-28 illustrates the stages of stress
distribution and failure propagation until debonding occurs. The authors stated that the
maximum pullout load occurs in the elastic-softening-debonding stage (stage 3). The maximum
axial and shear load in this stage is calculated using Equation 2-12 and Equation 2-13.

58
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Stage 2: Elastic-softening

Stage 3a: Elastic-softening-debonding

Stage 3b: Pure softening

Stage 4: Softening-debonding

Stage 5: Debonding

Figure 2-28: Five stages to debonding (J. Chen et al., 2015)

59
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

4𝜏𝑝
𝜎𝑏 (𝑥) = [ tan ℎ (𝑎(𝐿 − 𝑎𝑠 − 𝑎𝑓 )) cos (𝜔(𝐿 − 𝑎𝑠 − 𝑎𝑓 ))
𝐷𝛼
4𝜏𝑝
− sin (𝜔(𝐿 − 𝑎𝑠 − 𝑎𝑓 ))] cos(𝜔𝑥)
𝐷𝜔
Equation 2-12
4𝜏𝑝
+[ tan ℎ (𝑎(𝐿 − 𝑎𝑠 − 𝑎𝑓 )) sin (𝜔(𝐿 − 𝑎𝑠 − 𝑎𝑓 ))
𝐷𝛼
4𝜏𝑝
+ cos (𝜔(𝐿 − 𝑎𝑠 − 𝑎𝑓 ))] sin(𝜔𝑥)
𝐷𝜔

𝜔
𝜏(𝑥) = −𝜏𝑝 [ tan ℎ (𝛼(𝐿 − 𝑎𝑠 − 𝑎𝑓 )) cos (𝜔(𝐿 − 𝑎𝑠 − 𝑎𝑓 ))
𝛼
− sin (𝜔(𝐿 − 𝑎𝑠 − 𝑎𝑓 ))] sin (𝜔𝑥)
𝜔 Equation 2-13
+ 𝜏𝑝 [ tan ℎ (𝛼(𝐿 − 𝑎𝑠 − 𝑎𝑓 )) sin (𝜔(𝐿 − 𝑎𝑠 − 𝑎𝑓 ))
𝛼
+ cos (𝜔(𝐿 − 𝑎𝑠 − 𝑎𝑓 ))] cos (𝜔𝑥)

And,

4𝑘𝐸𝑔 1 𝐷2
𝛼2 = + [ ]
𝐷 + 2𝑡 𝐸𝑏 4𝐸𝑚 𝑏(𝐷 + 𝑏 + 2𝑡)
𝐷 (𝐸𝑔 + 𝐷𝑘(1 + 𝑣𝑔 )ln 𝐷 )

Where  ( x) and  ( x) are axial and shear stress at point x, D = bolt diameter and Eg , Eb ,
and Em are the young’s moduli of the grout, bolt and the confining medium.
Ren (Ren et al., 2010) used the same tri-linear and

Windsor (1997) classified the interface of the cement and resin with rock bolts as being
Continuously Mechanically Coupled (CMC). Li and Stillborg (1999) stated that the shear
strength of an interface comprises three components: adhesion, mechanical interlocking, and
friction. In their laboratory and experimental pullout testing of rock bolts, Li and Stillborg
(1999) proposed a series of theoretical models for different cases. For the laboratory case, the
authors proposed that the decoupling will start at the vicinity of the borehole collar and
propagate into the rest of the bolt during the test. Therefore, four stages were assumed for a
bolt pullout process as follows. In the first stage, the shear load is zero due to decoupling,
followed by the Sr value representing a partially decoupled state. In the last stage, the load is
reduced exponentially to a point where no load is applied after a certain point along the bolt
length (Figure 2-29).

60
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-29: Distribution of shear stress along a fully grouted rock bolt subjected to an axial
load (C. C. Li & Stillborg, 1999)

When the interface is coupled (x2=0), the exponential decay of the load along the cable can
be obtained based on Equation 2-14.

 b ( x) =  bo exp −2 x / d b

2
2Gr Gg Equation 2-14
2 =
Eb Gr ln ( d g / db ) + Gg ln ( d 0 / d g ) 

Where  bo is the axial stress on the bolt loading point, db and d g are bolt and hole diameters,

d 0  10d g , Gr and Gg are the rock and grout shear moduli.


The bond shear strength model for each of the four stages was stated as Equation 2-15. In these
equations, 𝜏𝑏 (x) in shear stress (Pa), 𝑠𝑟 is the residual strength (Pa), 𝜔 is the ratio between
peak and residual strength, 𝑑𝑏 in the diameter of the bolt (m) and 𝛼 is a coefficient.

𝜏𝑏 (x) = 0

𝜏𝑏 (x) = 𝑠𝑟

𝑥 − 𝑥1 Equation 2-15
𝜏𝑏 (x) = 𝜔𝑠𝑝 + (1 − 𝜔)𝑠𝑝
Δ
𝑥−𝑥2
−2𝛼( )
𝜏𝑏 (x) = 𝑠𝑝 𝑒 𝑑𝑏

Using CMC bolts, Martin (2012) introduced a new solution for the behaviour of the fully
grouted bolt in the pullout process. The author simplified the model by extracting the radial

61
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

and axial stress in a plane stress condition out of the equation of the motion, where  is the
Cauchy stress tensor and f is the unit volume force on the grout and rock ( f was considered
zero). Solving the radial and axial stress using the equation resulted in the equations of radial
and tangential stress and the shear stress along the bolt (Equation 2-16).

 rr =
Ri2 pi − R 2j p j

( p − p )R R
i j i
2 2
j 1
R −R 2
j i
2
R −R
2
j i
2
r2

  =
Ri2 pi − R 2j p j
+
( p − p )R R
i j i
2 2
j 1
R −R 2
j i
2
R −R
2
j i
2
r2
Equation 2-16
 Eg   ur  w  Eg w
  + = , r  [ Rb , Rg ]

 =
2 (1 +  g )   Z  r  2 (1 +  g )  r

 Er   ur  w  Er w
 2 (1 + )   Z  r  2 (1 + )  r , r  [ Rg , Rr ]
 +  =
 r r

Based on the results of the pullout tests, the equation is then solved by a reduced variable, and
the model input is adapted from a known function. Accordingly, the trilinear model was
selected by the author in order to simplify the pullout results. The analytical model proposed a
critical length where the maximum elastic interface of the interface occurs.

2.6 Rb
Lc 
f1 Equation 2-17

Martin used the proposed model on the laboratory pullout tests of Benmokrane 1995 and the
field tests of Chen and Ren 2008 and elicited an acceptable match. Three cases of models for
the laboratory tests were tested. While the trilinear model could predict the peak and residual
values, a non-linear model was more accurate. Nevertheless, the original trilinear model was
recommended due to ease of use.

Empirical studies by Forbes and Vlachopoulos (2016), on shear load distribution along the
seven-strand and Garford cable bolts embedded in cementitious grout resulted in:

rb E  xn −  xn−1
x = ( ) Equation 2-18
2 x n − x n−1

Where rb and E are the bolt radius and Young’s modulus,  xn is the strain at the point x .
n

62
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Numerical Studies
This section presents the numerical models developed to simulate the behaviour of rock and
cable bolts under axial loading. If required, the specific properties of the models in the literature
have been mentioned in the numerical modeling chapter as well.

Morsy et al. (2004) used a three-dimensional finite element numerical model to study the load
distribution along the bolts based on pullout test results. In his model, the traditional Mohr-
Coulomb criterion was used for the interaction between the cable and grout. It was seen that
for low values of load, the distribution is exponential, whereas, for the higher loads, the
behaviour becomes more linear.

Meloni et al. (2013) used the FEM model to examine the concrete strength during a non-
destructive pullout test. As plastic deformations were not expected, no frictional contact with
the borehole surface was taken into account along the inserted rod. In fact, the only hypothesis
of the research was that only compressive stress is effectively acting, which is not close enough
to the character of long embedment rock bolts and cable bolts. The authors stated that according
to accurate observations of the drilling and the insert surfaces, only the upper 10 mm of the
anchor head was assumed to be responsible for the mechanical interlocking and, thus, in contact
with the surrounding concrete (Figure 2-30).

FLAC and UDEC, as finite-difference and discrete element methods, have been repeatedly
used in mining and civil numerical analysis due to the variety of predefined tools. The rock
bolt command in both pieces of software relates to a two-dimensional element that can provide
resistance to the axial and lateral movements of the block. While the cable command refers to
a one-dimensional element that can only undergo axial loading and cannot withstand bending.
Both pieces of software considers the existence of the grout through its properties, not as a
separate medium (Itasca, Consulting Group, 2014; ITASCA, 2011).

63
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-30: 2D FEM model for pullout test of short bolts (Meloni et al., 2013)

Ma et al. (Ma, 2014; Nemcik et al., 2014) employed FLAC2D to model rock bolts embedded
in resin. The author stated that if used without modification, rock bolts or cable elements in
FLAC cannot model the interfacial shear bond stress-weakening mechanism between the bolt
and the rock mass. Hence, a non-linear bond-slip relationship was proposed to adjust the
interfacial shear bond stress locally along the bolt as a function of relative shear displacement
(Equation 2-19).

Equation 2-19

Where Fs is the shear force developed in the shear coupling spring, csscoh is the cohesive
strength of the shear coupling spring, 𝜎𝐶′ is mean effective confining stress normal to the rock
bolt element, cssfric is the friction angle of the shear coupling spring and perimeter is the exposed
perimeter of the element. The shear stress was assumed to be the cohesion load over the surface
area it applied to.

As the model was only targeting the shear behaviour. The normal coupling spring was not taken
into account (Figure 2-31) (Ma et al., 2016).

64
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-31: Rock bolt element with only shear coupling spring (Ma et al., 2016)

Chen (2016) utilised FLAC2D to model pullout experiments completed on plain Superstrand
and bulbed MW9 cables. The Rock bolt element in FLAC was chosen as it can model the
strain-softening behaviour seen in the tests (especially MW9 cable). Calibration of the model
was carried out using UCS testing of the confining medium and the grout properties were
extracted using a direct shear test. The result showed a very good correlation to the
experimental data (Figure 2-121). The numerical model also predicted that in the pullout test
of the cable bolt in weak and strong media, a confining pressure of 2 and 9.8 MPa for plain
cables and 3.7 and 10.5 MPa for the bulb cables, was estimated, respectively

Chen (2016) modeled pullout experiments on plain Superstrand6 and bulbed MW97 cables
within a modified LSEPT setup using FLAC2D. Among the FLAC elements, Rock Bolt was
selected because it is capable of modeling strain softening behaviour. A UCS test was
performed to calibrate the model, and direct shear test was used to extract grout properties.
Using the numerical model, it was predicted that the confining pressure of the cable bolt would
vary between 2 and 9.8 MPa for plain cables and 3.7 and 10.5 MPa for bulb cables. The result
showed a very good correlation to the experimental data (Figure 2-32)

6
Outer diameter of 21.8mm, made of 19 smooth steel wires
7
Outer diameter of 28mm, made of 9 smooth steel wires

65
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-32: Comparison between numerical and laboratory results on plain cable installed in
(left) weak confining medium, (right) strong confining medium (J. Chen, 2016)

Two approaches have been used to solve the dynamic loading of tendons: the lumped-mass
model and the dynamic deformation model in a continuum media. Lumped-mass describes the
reinforcement elements as discrete masses connected to each other by springs and dampers.
Research of Tannent et al. (1995), St-Pierre (2007), and Thompson et al. (2015) are some
examples of the application of the lumped-mass model. Dynamic deformation models use
discrete elements to describe the behaviour of the reinforcement as a continuous and
deformable medium. Ansell (2005), Yi and Kiaser (1994b), and Mirarco et al. (2018) employed
this model for analysis of the dynamic loading of rock bolts.

Mirarco et al. (2018) used FLAC3D to simulate the dynamic pullout test performed by the
Canmet-MMSL test rig. They simplified the modelling of the impact load to a simplified
damped oscillator (Figure 2-33). This concept was already being used by St-pierre (2007) as
well.

Figure 2-33: Free-body diagram of the model (Mirarco et al., 2018; St-pierre, 2007)

Two differential equations were presented to describe the motion of the rock bolt (Equation
2-20) and grout (Equation 2-21).

66
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

For the rock bolt:

𝑚𝑥̈ 𝑏 + 𝑐𝑏 (𝑥̇ 𝑏 − 𝑥̇𝑔 ) + 𝑘𝑏 (𝑥𝑏 − 𝑥𝑔 ) − 𝐹𝑓𝑘 + 𝑚𝑔 = 0 Equation 2-20

For the grout:

𝑚𝑔 𝑥̈𝑔 − 𝑐𝑏 (𝑥̇ 𝑏 − 𝑥̇𝑔 ) − 𝑘𝑏 (𝑥𝑏 − 𝑥𝑔 ) − 𝑐𝑔 𝑥̇𝑔 − 𝑘𝑔 𝑥𝑔 + 𝐹𝑓𝑘 = 0 Equation 2-21

Where:

m = mass of hitting hammer


mg = grout mass
g = gravity
𝐹𝑓𝑘 = friction forc between grout and bolt
𝑘𝑔 , 𝑘𝑏 = stiffness of grout and bolt.
𝑐𝑔 , 𝑐𝑏 = viscous damping of grout and bolt
𝑥𝑔 , 𝑥𝑏 = displacement of grout and bolt
𝑥̇𝑔 , 𝑥̇ 𝑏 = velocity of grout and bolt
𝑥̈𝑔 , 𝑥̈ 𝑏 = acceleration of grout and bolt
Although the model was built based on a Canmet rig (Yi & Kaiser, 1994b), results were
compared with a WASM (Player et al., 2004) project, and there was good agreement between
them. However, in both the actual tests and numerical simulations, tensile failure of the rock
bolt was dominant; thus, the model was not representing the bond behaviour.

Pitrakkos et al. (2010) modeled the actual anchor bolts in concrete using ANSYS. In their
model, the component method was used for the bolt/concrete interface. The component method
allows a joint to be characterised by the individual properties of its components. Although the
model perfectly fitted with the pullout peak behaviour, the post-peak behaviour could not be
modeled with this method. Besides, the pullout models were not verified with actual
experiments.

67
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-34: Numerical and experimental model of anchor bolt (Pitrakkos et al., 2010)

Tahmasebinia et al. (2021) adopted the dynamic pullout test results of rock bolts carried out by
Villaescusa (2013) to generate a full-scale FEM using ABAQUS/Explicit. This model was
accompanied by their previous 3D models of dynamic double shear (Tahmasebinia et al.,
2018b, 2018a) to expand the understanding of a tendons behaviour under different loading
conditions. It was concluded that the strain rate could significantly enhance the tensile and
shear capacity when the cable bolt is subjected to dynamic load. For two main reasons, the
pullout test models of Tahmasebinia could not evaluate the cable/grout interface. The full-scale
model could not concentrate on the details. In addition, models were mainly targeting the
relationship between the load and fracture energy of cable bolts. In fact, tensile failure of the
tendon was the only failure mode in their model.

68
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-35: Full-scale FEM model of dynamic pullout test (Tahmasebinia et al., 2021)

Parametric Studies

2.7.1 Tendons Properties (Cable bolt and rock bolt)


Satola and Aromaa (2003) and later, Satola (2007) carried out pullout tests on various tendons
such as galvanised steel rock bolts, epoxy-coated cable bolts, galvanised cable bolts, plain steel
and bulbed steel cable bolts. Results showed that both galvanaised and epoxy coating increased
the load capacity and stiffness of the cable bolt (Figure 2-36). Goris and Conway (1988) had
the same observations.

69
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-36: Effect of tendon properties on pullout load capacity (Satola, 2007)

Goris (1991) compared the effect of the epoxy grout and birdcage with conventional 15.2 mm
cable bolts. The maximum pullout load was recorded in birdcage cables, followed by epoxy-
coated cables. Table 2-3 shows the comparison of the plain, birdcage, and coated cable bolts.
It can be seen that although the average displacement in plain and epoxy-coated tests was close,
the maximum pullout load in the epoxy cable was significantly higher than in the plain cable.

Table 2-3: Effect of birdcage and epoxy coating on pullout load

Plain Birdcage Epoxy-coated


Ave. Max pullout load (kN) 88.1 151.1 123.9
Ave. Max displacement at max load 45.2 6.8 52.83

Dorsten (1984) studied the effect of coating on cable bolts and realised that the corrosion of
steel cables could be avoided by using epoxy to cover the cable surface. Fuller and Cox (1975)
conducted a systematic laboratory study on cable bolts. They proposed that rust can increase
the peak value of the test by improving the friction and mechanical interlocking compared to
smooth steel wires. Stillborg (1984) also suggested that a cable's clean and rough surface can
significantly improve its bonding capacity. In a series of experimental studies, Stillborg (1984),
Hutchins et al.(1990), and Satola (1999) reported that surface contamination by grease or other
material does not have as severe of an effect on the peak load of cable bolts.

Kilic (2002) reported that rock bolts with surface modifications (indentation) could carry up to
5 times more load than smooth bolts. It is well-known in civil and mining engineering that
70
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

these modifications create more surface area and increase both the bond and residual load.
After conducting tests on plain and indented PC strand cables, Tadolini et al. (2012) suggested
that while every indentation to the steel wire would improve the load capacity of the cable bolt,
the deeper the indentation, the higher the load gets and the higher the stiffness becomes. Aziz
et al. (2016) conducted a small-scale pullout on plain and indented Superstrand cables. They
reported that indentation changed behaviour from perfectly plastic to strain softening by
introducing a load drop after a pronounced peak load. More importantly, in all cases, indented
cable vastly improved the peak load of the system while maintaining the same initial stiffness.
This setup, however, was only made for a certain diameter and should be retested for each cable
according to the recommended diameter.

Figure 2-37: Effect of cable surface on pullout load (Aziz et al., 2016)

Similarly, Li (X. Li, 2019) found a significantly higher pullout load with indented cable bolts;
however, surface indentation changed the pullout behaviour. While the plain cable normally
has a perfectly plastic curve, indentation provides a clear and distinguishable load drop after
the first failure, followed by strain-softening behaviour.

71
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-38: Effect of indentation on pullout load capacity (X. Li, 2019)

Hutchins et al. (1990) studied the effect of the birdcages and compared the results with standard
strands. They reported that birdcages significantly increase the load-bearing capacity (Figure
2-39).

Figure 2-39: Comparison between birdcaged and standard cable bolts (Hutchins et al., 1990)

Thomas (2012) studied 14 different cable bolts, and the results showed that bulbed cables
possess more than a 400% higher load-bearing capacity compared to plain cables. In fact, large
boreholes were reported to decrease the capacity of plain cables, whereas bulbed cables had an
increase in the peak load. It was maintained that this was due to the high strength of the grout
(80 MPa) compared to the confining medium (simulated rock with 25 MPa compression
strength).

72
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Craig and Murnane (2013) conducted in situ pullout tests in weak coal roofs in Australia. Three
types of indented cable bolts (Superstrand, TG, and SUMO) and one plain cable (SUMO) were
encapsulated into 400 mm long mudstone. It was seen that the SUMO bulbed cable had a more
outstanding performance compared to the indented but unbulbed cables. Moreover, Craig and
Holden (2014) compared a series of in situ pullout tests on four cable types (Indented SUMO,
TG, Superstrand, and plain SUMO). The result showed that plain and indented SUMO cables
had quite a similar peak load to the plain cable having slightly more displacement at the peak.
The authors also reported a big gap between the laboratory and the in situ test outcomes,
especially in the case of the intended, unbulbed, and smaller cable (Superstrand). On the other
hand, this study proves bulbed cables are less sensitive to the type and the condition of the test.

Hagan (2017) studied the performance of five cables of MW9S8, Twin Strand9, nine wire
Sumo, TG10, and plain Superstrand11. All the cables showed vastly different behaviour in the
pullout test (Figure 2-40). Although all the modified cables showed similar initial stiffness
behaviour, the peak values and the post-peak behaviours were different.

Figure 2-40: Effect of cable properties on pullout test results (Hagan & Li, 2017)

In another experiment, two or more cable bolts were inserted in one borehole (Stillborg, 1984).
Pullout experiments of Stillborg (1984) and Goris (1990) on both single and twin 15.2 mm
strand cables revealed that the capacity of the twin strand cable is more than twice that of the
single strand cable. The system acts like two single strands combined, whereas after this point,

8
Outer diameter of 28mm with nine spiral wires around a hollow tube
9
A bundle of two 15.2 mm cable bolt
10
Outer diameter of 28mm with nine smooth wires around a hollow tube (Product of Jennmar)
11
Outer diameter of 21.8mm with 19 smooth steel wires (Product of Jennmar)

73
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

it seems the system acts as one new system with a larger stiffness instead of the sum of two
cables acting independently.

Stillborg (1984) conducted a series of rock bolt pullout tests in reinforced concrete blocks to
study the behaviour of different designs and bonding agents on load-bearing capacity. It was
observed that the bolt diameter has a significant effect on the results by increasing the load
peak load value. Moreover, the 20 mm steel rebar encapsulated in grout and resin showed
almost similar behaviour. In contrast, the 22 mm resin-encapsulated fibre glass bolt when
compared to the rest of the steel rock bolts in this study (Split set, Expansion shell, Swellex,
and steel rebar) showed higher load with significantly less stiffness behaviour.

During a pullout test, Forbes and Vlachopoulos (2016) measured the strain along a two-meter-
long seven-wire cable bolt embedded in cementitious grout. The effect of the bulb on the cable
bolt was assessed, and the results of plain and bulbed cables were compared. Results showed
that the existence of the bulb significantly reduced the strain of the cable during the pullout
event (Figure 2-41). As a result, modified cable bolts with various bulbs were created to
increase the radial pressure on the cable and to increase the stiffness of the cable bolt system.
The author believed that non-simultaneous debonding occurs along the cable length, especially
for the longer length, with a debonding front moving deeper into the borehole, as seen in strain
and shear bond graphs. After a certain load, a section of the cable is debonded, and a new and
longer section picks up the load. This causes a drop in the load and audible sounds due to the
radial fracturing of the grout.

74
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-41: Comparison of strain in bulbed and plain cables during the pullout test (Forbes
& Vlachopoulos, 2016)

2.7.2 Confinement Medium


Theoretical models developed by Coated and Yu (1970) and experiments of Hyett (1992)
discovered that the harder the confining material, the higher the bond strength. However,
simulation of confinement so that they could represent the real-field condition was the
challenge. For this, various confinement methods have been utilized for pullout testing. Steel

75
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

pipes with a diameter close to the borehole size, concrete samples, artificial rock and rock cores
were some of the options used in the previous studies. However, it was found that tendons
confined in steel tubes usually performed well, which resulted in an overestimation of the test
results (Benmokrane et al., 1995). Hagan (2015; 2014) observed the effect of strength of
confinement from 11 to 64 MPa and stated that confinement strength has reportedly caused the
most adverse effect on all cable types. They concluded that cable bolting should be studied as
a whole system consisting of the cable, grout and rock; the output can be fairly different from
time to time. In previous research, concrete and artificial rock or actual core samples were
recommended (Goris et al., 1996; Ito et al., 2001). On the negative side, lateral and radial crack
propagation during the test is one of the biggest challenges of using concrete (D. Li et al., 2020;
G. Yang et al., 2018). Although using half-cylindrical steel confinements in some studies was
successful (Aziz et al., 2022; Mckenzie & King, 2015), severe radial cracks were observed
when using high-strength cable bolts (Hagan & Li, 2017; Rastegarmanesh, 2022). Researchers
such as MacSporran (1993), Moosvai (2005) and Martin (2012) suggested using a modified
Hock Cell to provide the required lateral pressure. This limits the size of the samples to the size
of cells.

Rajaie (1990) reported that the diameter of the cylindrical rock (concrete) sample used in the
pullout test have a significant effect on the ultimate load of the system. Through experiments
at various diameters, it was suggested that this effect was minimised for samples larger than
250 mm in diameter. While the tests were limited to only plain cables, Holden and Hagan
(2014) studied the size effect of the pullout load of bulbed cable bolts. Their setup had four
sizes of 150, 215, 300, and 40 mm external diameter samples. The bulb centre was placed at a
fixed distance of 140 mm from the bottom. It was seen that higher diameters of the sample
resulted in higher peak load (from 110 kN to 140 kN).

76
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-42: Effect of confinement size on pullout test results (Rajaie, 1990)

In a similar study, Chen (2016) studied the size effect of the sample on the pullout load in
confined and unconfined conditions. A bulbed SUMO cable was encapsulated in artificial rock
(cement-based) with external diameters of 150 to 508 mm. In the confined scenario, the
samples were confined inside holder tubes to create a constant stiffness condition. In both
cases, there seems to be a bilinear relationship between the sample diameter and the load.

Figure 2-43: Effect of sample size on pullout load (J. Chen, 2016)

Coated and Yu 1970 presented a theoretical model to evaluate the bond stress of the cylindrical
tendons in a pullout test and analysed the effect of the bolt and rock elasticity on the bond
stress. It was discovered the harder the confining material, the higher the bond strength.

Macsporran (1993) reported the effect of radial pressure and the water-cement ratio (w/c) of
the grout on a plain cable. It was found that lower w/c will result in both higher strength of the

77
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

grout and pullout load, while for radial pressure, higher pressures would significantly increase
the peak load of the system (Figure 2-44).

Figure 2-44: Effect of confinement pressure on pullout load tests using two different w/c
ratios for grout (left: w/c: 0.3, right: w/c: 0.4) (MacSporran, 1993)

Moosavi et al. (2005) conducted pullout tests on 20 mm Dywidag and 22 and 28 mm samples.
The confining medium was made of 30 and 42 MPa cement grout. Results showed lower radial
dilation when the confining pressure increases. After a point, radial compression is seen
pronouncedly, especially for the weaker grout. In general, a nonlinear relationship was
observed between the confining pressure and bond capacity.

Figure 2-45: Effect of confining pressure on the bond capacity (Moosavi et al., 2005)

Also, in their study, Moosavi et al. (2005) studied the path dependency of the pullout test on
rock bolts using a Modified Hoek Cell (MHC). In the test, after a certain pullout load, the
confining pressure was increased, and this was repeated until the end of the test. When
compared to the full pullout tests at each of the fixed confining pressures, it was seen that at
each confining pressure, the graphs followed the path of the corresponding confining pressure.
As a result, the authors claim no path dependency for the testing setup and maintain that the

78
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

bond capacity would be the same regardless. However, the confining pressure load here seemed
to ascend to the end of the test.

Figure 2-46: Effect of confinement pressure on pullout test (Moosavi et al., 2005)

Experimental studies of Thomas (2012) concluded that laboratory confinement pressure would
be in the range of 5-10 MPa, which is not essentially a true reflection of the ground condition.
They proposed two main reasons for this statement, including that there is no reliable
information about the magnitude of the confinement pressure in the real field, and confinement
is a dynamic variable and varies significantly both along the length of the cable and during the
life of the cable. It was also found that the applied confining pressure has a very significant
impact on the strength and stiffness of the cable. As such, it was a concern that almost any
variation in the applied pressure could have a misleading or masking impact on the measured
results. It is notable in Figure 2-47 that the strength and stiffness of the cable being tested in
the biaxial cell were both increased compared to the tests conducted in a passive steel cylinder
(Thomas, 2012).

79
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-47: Effect of active and passive confinement on the test results (Thomas, 2012)

Aside from lateral and radial confinement, it was found that axial confinement also affects the
results. Khan (1994) noticed the effect of surface confinement during the pullout test of cable
bolts. The study suggested that the confinement of the borehole collar can have a major effect
on the pullout load result by comparing three different sized bearing plates. It was suggested
that covering the grout annulus is not advisable as it forces the failure along the cable/grout
interface and, more importantly, changes the stress distribution in the grout column.

Hagan (2004) introduced a pullout testing setup with two loading modes. The main difference
between the two setups was the existence of axial confinement. The setup without the top
confinement replicated bed separation, while the one with top confinement represented the
traditional pullout assembly. The study concluded that the existence of the top confinement
could increase the pullout load, albeit not noticeably. Similar results were concluded by
Weckert (2003) for the bed separation scenario and Chen (2016) for bulbed cable bolts.

80
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-48: Effect of top confinement on pullout test results (Hagan, 2004)

2.7.3 Embedment Length


Fuller and Cox (1975) stated that in their experiments with embedment lengths of 100 to 700
mm, a linear relationship between the peak value and the embedment length exists. Goris
(1990) found similar results for the embedment length between 203 to 812 mm.

A Laboratory study by Benmokrane (1995) on both threadbar and cable bolts showed that a
longer embedment length increases the required pullout load but does not change the failure
mode (Figure 2-49). Hence, embedment length does not change the constitutive model.

Figure 2-49: Effect of embedment length on the pullout behaviour of cable bolt (left), and
rock bolt (right) (Benmokrane et al., 1995)

Martin et al. (2011) proposed that the bond-slip relationship represents an inherent
characteristic of the bolt–grout joint interface and that it is independent of the boundary
conditions, which are applicable for rock bolts with and without free-end slip. Hence, the same
bond–slip relationship can be input into the rock bolt models with any encapsulation length.

81
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Sotala (Satola & Hakala, 2001) assessed the effect of embedment length using a long
embedment double shear testing machine. In their study, five different tendons, including rock
bolts and cable bolts, were tested with two different embedment lengths of 1000 mm and 2000
mm. Results showed that different tendons responded differently to the increase in embedment
length. While the pullout load increased significantly in plain steel cable bolts, the loading
capacity of epoxy coated and galvanised steel cable bolts was not noticeable (Figure 2-50).

Figure 2-50: Results of pullout test with 1000 mm (left) and 2000 mm embedment lengths
(Satola & Hakala, 2001)

The role of embedment length has been investigated using various sensitivity analyses
proposed by Martin (2012). As a result, embedding length greatly influences the results, and in
short lengths, the shear distribution in uniform and bond-slip models is more accurate,
suggesting that tests with shorter embedding lengths will be more useful in cases where
interface behaviour needs to be studied. It was also stated that in situ pullout tests do not allow
the determination of an accurate bond-slip model for the rock bolt-grout interface because the
embedment lengths tested are too long. Having said that, shorter lengths typically have higher
scatter, which means multiple data points (tests) should be available, according to Hyett (1992).
Furthermore, Martin argued that tendon-grout interface constitutive law is intrinsic, which does
not depend on embedment length or bond-slip model. Finally, the effect of embedment length
on axial load has been plotted in Figure 2-51 (L. B. Martin, 2012).

82
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-51: Effect of embedment length of pullout load (Martin, 2012)

Thompson and Villaescusa (2014) evaluated the critical embedment length of a plain cable bolt
through 5 pullout tests. The critical length was introduced as the length where the cable ruptures
instead of sliding out of the borehole (grout/cable interface failure). It was seen the relationship
between the peak and embedment length was exponential. The critical length was predicted by
intersecting a linear line passing through the rupture loads at 3000 and 3500 mm and the
exponential sliding graph for 500, 1000, and 1500 mm lengths.

Figure 2-52: Critical length for the effect of embedment length (Thompson & Villaescusa,
2014)

Ma et al. (2016) investigate the effect of encapsulation length using numerical models. The
encapsulated lengths and rock bolt element numbers used in these models were, respectively,
2.1-meter-long bolt with 18 rock bolt elements, 4.2-meter-long bolt with 36 rock bolt elements,

83
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

7-m-long bolt with 60 rock bolt elements, 9.8-m-long bolt with 84 rock bolt elements and 13.3-
m long bolt with 114 rock bolt elements. The bolt load-carrying capacity increases with an
increase in grouted rock bolt length (Figure 2-53). In addition, increasing the encapsulated rock
bolt length will decrease the slip of the unloaded end, and when the contributing length of the
rock bolt is long enough, there is no displacement at the unloaded end during the test.

Figure 2-53: Effect of embedment length on the maximum applied pullout load (Ma et al.,
2016)

Chen (2016) applied the confinement pressure by applying torque on the bolts and nuts holding
the cylindrical steel confinement together. Results revealed that applying extra torque to the
bolts can affect the pullout test results (Figure 2-54); however, it was impossible to figure out
the equivalent confinement pressure. Moreover, radial cracks of the sample were undeniable,
even in high-torque samples.

84
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-54: Effect of confinement pressure on the pullout results (J. Chen, 2016)

2.7.4 Grout Properties (cement or resin grouts)


Kilic et al. (2002) implemented several pullout tests using different water/cement ratios of
grout and recorded the bond strength of rock bolts. The Uniaxial Compressive strength (UCS)
of the selected grout did not exceed 45 MPa. Results showed a logarithmic relation between
the UCS and the bond strength (Figure 2-55). It was concluded that extra water, which was not
used by the cement for hydration, tends to evaporate during the curing process and creates
inhomogeneous internal structures as a result of capillary porosity.

Figure 2-55: Effect of grout UCS on the bond strength (Kilic, Yasar, & Atis, 2002)

Meloni et al. (2013) conducted an analytical and numerical investigation and compared the
results with experiments of pullout test for short bolts according to European Standard UNI EN
12504-3. Non-destructive pullout tests were used to estimate the compressive strength of the

85
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

concrete, and there was good agreement between the predicted result and actual test results. A
correlation determined a quadratic polynomial curve was the best fit for the results.

Figure 2-56: Correlation between pullout force and compressive strength (Meloni et al.,
2013)

Hagan (2017) tested two grout strengths of 80 and 62 MPa by changing the W/C of a
cementitious grout (Stratabinder HS). The pullout results showed a strong correlation between
the strength of the grout and the peak load value, while the initial stiffness was only affected
negligibly. More importantly, by mixing the confining medium-strength simultaneously with
the grout strength, the residual resistance of the weaker rock was more adversely affected by
the weaker grout (Figure 2-57).

Studies on the effect of curing time by Hassani et al. (1992) and Aziz et al. (2016) showed that
a longer curing time significantly improves the stiffness and the ultimate load of the system. In
fact, the curing time of cementitious grouts can be regarded to have the same effect as the
strength of the grout because grouts will increase in strength the longer they are cured.
Mirzaghorbanali (2016) concluded that cementitious grout does not have a linear curing curve,
and in the first 7 days, the strength grows much faster relative to the growth from one week to
four weeks (Figure 2-37).

86
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-57: Effect of w/c and confining pressure on pullout test results (Hagan & Li, 2017)

Li (2019) conducted pullout experiments on plain and indented cables with regards to the
effects of grout age. It was seen that in all cases, longer curing time (Stratabinder cementitious
grout) provided higher peaks and residual loads while maintaining similar stiffness.

Farah and Aref (1986) performed pullout tests on seven strand plain cable grouted using a
normal cement-based grout and an aggregate mixed cement-based grout similar to concrete.
The study concluded that the aggregate mixed grout increased the cable bolts ultimate bond
value and provided a higher ductility.

Benmokrane (1995) plotted the pullout load of rebars encapsulated in grout with different
additives (Table 2-4). Results revealed that even though some of the grouts had higher
compressive strength, it did not necessarily cause a higher pullout load. It was found that grout
with a swelling agent creates swelling pressure which effectively enhances the shear bond
resistance. The results of Stillborg (1984) confirmed the same experience.

87
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-58: Effect of grout type on pullout test results (Farah & Aref, 1986)

Table 2-4: Cementitious grout with different additives for pullout test, after (Benmokrane et
al., 1995)
Grout type Main feature UCS (MPa)
CG1 Plain grout (type 10 portland) 52.6
CG3 CG1 with swelling agent 40.3
CG4 CG1 with silica fume and superplasticizer 59.6
CG6 CG1 with sand 51.8

Figure 2-59: Effect of different additives of grout on pullout behaviour (Benmokrane et al.,
1995)

Hassani et al. (1992) studied the effect of some silica powder and concluded that silica powder
addition to cement could increase the bond capacity and improve the post-peak behaviour of
the system. Benmokrane et al. (1995) studied the effect of aluminum powder, silica powder,
and superplasticiser on the pullout behaviour of cable bolts. While the aluminum powder
increased the peak load in the test, the silica powder mixed with superplasticiser seemed to
have a minor effect on the load.
88
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Matin (2012) reported on 23 pull test results of HA25 rock bolts encapsulated using resin
(Lockset SF) and cementitious grout (W/C =0.5) in the ANDRA underground Research
Laboratory of France. The results showed that the cementitious grouts compared to resin
provided better consistency and less scatter of the data.

Pullan (2018) compared the pullout behaviour of cable bolts using resin and cement grout with
different confinement stiffnesses. In this study, when the confinement was weak, failure with
the resin grout tended to occur at the grout and rock interface. In contrast, the cement grout
failed at the cable and grout interface. When strong confinement was used, both resin and
cement grout achieved close peak pullout loads.

Figure 2-60: Comparison of resin (left) and cement grout (right) in pullout tests (Pullan &
Hagan, 2018)

2.7.5 Loading Condition


The loading rate on the tendons can be varied from static to dynamic. Long-term ground
settlements are classified as static loading conditions. In this case, no impact load is applied on
the tendon. In contrast, rockburst, coal burst, blasting, and ground seismic activities can release
a vast amount of impact energy in less than a second (M. Cai, 2013; Kaiser & Cai, 2012; Mark,
2018; Ortlepp & Stacey, 1994). In other words, the static motion refers to inertia with zero
acceleration, and constant and relatively slow displacement rate (Pariseau, 2006).

Hagan and Li (2017) studied the effect of displacement rate on the pullout test results. Four
displacement rates between 1.8 to 16.2 mm/min were examined, and it was shown that the
displacement rate appears not to affect the initial stiffness, peak value, and post-peak behaviour
in any meaningful way. In other words, as long as the rate of displacement does not cause any
dynamic load (or acceleration is zero), the load is deemed static and elicits similar reactions
from the tendon.

89
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-61: Effect of loading rate on the pullout test results (Hagan & Li, 2017)

Hassani et al. (1992) studied used shotcrete grout instead of conventional grout in the pullout
test with the loading rate ranging between 0.1 mm/min to 600 mm/min and reported that the
loading rate does not affect the pullout results. This proved that the main difference between
static and dynamic loading is the initial acceleration and effect of impact loading.

It was stated that the ground velocity in a rockburst incident might reach 50 m/s (Ortlepp &
Stacey, 1997). Released energy in the longwall face during a low-hazardous rockburst event
can be 103 J, while in a high-hazardous case, it surges to 5*106 J (Mutke et al., 2015). The
amount of released energy depends on the mass and velocity of the ejected rock mass (Kaiser
et al., 1996; Ortlepp & Stacey, 2000). Hence, different researches including Wagner (1984),
and Ortlepp and Stacey (1994) suggested that in highly dynamically active ground, tendons
should be designed, based on energy rather than loading capacity. To accommodate this, many
different testing rigs have been developed to enable an understanding of tendons behaviour
under dynamic loading.

Stillborg (1984) conducted a series of in situ pullout tests on fully grouted cable bolts and
reported that pullout load is influenced by blast-induced dynamic loading as a function of
particle velocity, especially for values more than 500 mm/s. Similarly, Farah and Aref (1986)
1986 performed dynamic pullout tests on seven-strand plain cable bolts and reported that
concrete with aggregate provided a higher ultimate bond load than static tests.

Player et al. (2012; 2009) examined the behaviour of several reinforcement systems in dynamic
mode. Conventional and dynamic rock bolts, as well as 15.2 mm cable bolts, were among their
selected reinforcements. It was found that the correct selection of the nut and washer plays an
integral role in maximizing the load transfer through the partial thread. It was identified that

90
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

the breakage for static and dynamic axial loading was the same as long as the 2 m critical
embedment length for the 15.2 mm plain strand cable was considered; otherwise, slip occurs.

Thompson (2015) tested rock bolts in dynamic mode and noticed that the energy absorbed by
the reinforcement system is over almost 32kJ and the force on the rebar exceeded 130 kN.
However, the absorbed energy by the reinforcement was almost 50% of the input energy which
means a high level of energy dissipation (Figure 2-62).

As part of the effort of the same research team in Western Australia (Player et al., 2004;
Thompson et al., 2015), Villaescusa (2007) reported on dynamic test results of seven-strand
cable bolts. In their study, only two of the tests resulted in the pullout of the cable bolts, and
the rest ended with the tensile failure of the cable bolt before being pulled out. Results of the
dynamic test of the cable bolt showed that the energy absorption of the cable bolt was almost
80% of the input load (Figure 2-63).

Figure 2-62: Absorbed energy and applied load in dynamic pullout test of rock bolt
(Thompson et al., 2015)

Comparison of the results of rock bolts and cable bolts in a multi-year study by Player,
Thompson, and Villaescusa (Player, 2012; Player et al., 2009, 2013; Player & Villaescusa,
2008; Thompson et al., 2015) revealed that cable bolts absorb significantly higher energy
during the pullout tests.

91
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-63: Absorbed energy and applied load in dynamic pullout test of cable bolt
(Villaescusa et al., 2007)

Crompton (2018) tested the rock bolts being impacted by an 1181 kg hammer with a speed of
5.4 m/s. The typical load and displacement graph of the tests showed that dynamic pullout tests
experienced a peak displacement while the final displacement would be lower. In other words,
there is an elastic response from the elongated tendon which results in a reduction of the final
pullout displacement. Hence, drawing of load-displacement graph does not make sense and
both load and displacement were plotted versus time.

Figure 2-64: Typical load and displacement graph of the dynamic test (Crompton et al., 2018)

Review of the Recent Improvements in Shear Testing of Tendons


As part of the research team of tendon studies at the University of Wollongong, the main work
associated with shear experiments in the last eight years has been thoroughly covered by other
colleagues, including Yang (2019) and Khaleghparast (2021). In this section, the outcomes of
the last two research projects are reviewed. In addition, recently released studies are added, and
finally, dynamic shear testing of cable bolts is reviewed.

92
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Yang (2019) extended cable bolting studies by performing static cylindrical double shear tests
on high-strength cable bolts. A numerical model using the Distinct Element Method (DEM)
was generated in Particle Flow Code (PFC) 2D software in order to simulate crack propagation
around the cable bolts when shear load was applied. Effect of pretension load, confinement
stiffness were assessed on different cable bolts including MW9 and MW10 (both plain and
indented). It was concluded that the larger the pre-tension force, the smaller the shear
displacement at the cable failure. Also, not only did indentation on the surface of the cable bolt
enhance the performance in shear, but also it causes immature failure of the wires. Four
possible factors have contributed to the abnormal failure of indented wires in the cable strand
include: a) Loss of wire strength through the indentation process, which could be as much as
10%; b) Localised stress on each indented wire due to the indentation process; c) Excessive
bending of wires at the hinge points, particularly when the wires are positioned lateral to the
direction of shear; and d) Indentation interlocking and friction effect of wires. In addition,
assessment of the effect of stiffness of confinement indicated that concrete samples with
internal reinforcement recorded lower shear load on the cable at failure. It was inferred that
greater stiffness of the host medium reduces the displacement.

The internal confinement of the concrete block contributes to increased concrete strength and
stiffness, which minimises early concrete deformation around the tendon close to sheared joint
faces. Meanwhile, the reinforced concrete could also reduce deformation depth at the hinge
points by 50%, resulting in reduced vertical cable displacement. Yang (2019) also proposed
the sketch design of the angled double shear testing for cable bolts (Figure 2-65).

Figure 2-65: Proposed angle double shear testing setup design (G. Yang, 2019)

93
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Khaleghparast (2021) concentrated on static and dynamic testing of solid rock bolts. The
development of dynamic testing machines and technologies has been fully covered by the
author in chapter two of his thesis. The mechanism of failure of the tendon under static and
dynamic shear loading has also been covered in chapters two and three. A detailed numerical
model was developed by the FEM method using LS-DYNA (Figure 2-66).

It was found that in dynamic tests, the strength of the concrete materials can significantly affect
the dynamic test results, as compressive strength is influenced by the rate of loading. The higher
rates of loading can result in more resistance of the concrete material against the applied load.
Energy dissipates easier in softer materials compared to a brittle material. On the other hand,
increasing the concrete strength resulted in lower shear displacement at the point of failure.
The results were in accordance with what Yang (2019) concluded for static testing of cable
bolts. Ultimately, no constant relationship was found between the concrete strength and load-
bearing capacity of the rock bolt subjected to the dynamic shear load.

Figure 2-66: Numerical model of dynamic shear testing of rock bolts (Khaleghparast, 2021)

The dynamic shear load of rock bolts is equivalent to its static double shear load when the
effect of friction is removed. The assumption is that when the specimen is subjected to dynamic
double shear load, the effect of friction becomes negligible because the friction is a time-

94
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

dependent factor, and the impact load occurs in a very short period of time, therefore the time
effect can simply be considered ineffective. The ratio of dynamic shear strength to static shear
strength was measured as being approximately 0.7. The effect of pretension was found to be
ineffective in regard to dynamic double shear testing. While the trend in literature is that the
peak shear load decreases by increasing the pretension load of rock bolts as well as cable bolts
when they are tested on static double shear testing apparatus with a set rating load of 1 mm/min
(Khaleghparast, 2021).

Saadat (2019) used PFC2D and Distinct Element Method (DEM) to simulate the behaviour of
rock bolts under combined shear and pull loading conditions with the infilled joints (Figure
2-67). Saadat summarized the numerical models developed for shear testing in Chapter two of
his thesis.

Figure 2-67: Numerical sketch model of combined pull and shear test of rock bolt (Saadat,
2019)

The idea of the model was to apply pre-tension stress in the form of pullout load on the rock
bolt. Then the direct single shear test of the bolted rock joint can be carried out. Four different
stages including linear elastic, pre-peak hardening, post-peak softening, and residual stage
characterised the failure mechanism of fully grouted rock bolts during the pullout experiment.
The equivalent normal load on the rock joint in each step has been demonstrated in Figure 2-68.
As shown below, the occurrence of pullout to the pre-peak hardening stage results in a rapid
increase of normal load on the rock joint. The results from the shear test model in each stage
of the pullout loading indicates that the highest shear load was recorded when the pullout load
was in the pre-peak stage (points O and P) (Figure 2-69).

95
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-68: Effect of pullout load on normal load on the rock joint (Saadat, 2019)

Figure 2-69: Equivalent shear test in each stage of the pullout load

The compression and tension load distribution on the concrete and the tendon during the pullout
loading indicates that even in the post-peak softening stage of the pullout, the compression load
on the rock joint is still higher than before the tension (Figure 2-70).

96
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-70: Distribution of compression and tension load on the rock joint and tendon during
a combined pull-shear test (Saadat, 2019)

Pytlik (2020) conducted a series of static and dynamic single-shear testing on high-strength
rock bolts. The testing setup acts like the guillotine box suggested by British Standard (BS
7861-2, 2009), which has previously been criticized by Aziz et al. (2016) for underestimating
the shear strength and not being able to replicate real field conditions.

(a) (b) (c)


Figure 2-71: Single shear impact testing setup for rock bolts, (a) diagram, (b)view, and (c)
loading diagram (Pytlik, 2020)

97
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Kang (2020) introduced a testing facility for different loading conditions of tendons including
tension, torsion, bending, perpendicular shear testing, angle shear testing (up to 30 degrees),
and axial impact.

Figure 2-72: Multi-purpose tendon testing setup (Kang et al., 2020)

Four 22-mm rock bolts were tested in 20-degree static angled shear. Although pretension load
was applied on all of the bolts in the form of the torque on the nuts, two of the bolts experienced
being pulled out up to 170 kN axial load. It was concluded that for the tests without applied
pullout load, there was no significant difference between the maximum shear load of
perpendicular shear tests and 20-degree declined samples. However, when the bolts were
pulled out by 170 kN of load, the maximum shear force experienced a considerable decrease.

Between 2018 to 2021, Tahmasebina et al. (2018a, 2018b, 2021) generated a series of dynamic
double shear test models for rock bolts using the Finite Element Method. Although it was
believed that the previous model adequately modeled the dynamic double shear, the second
model in 2021 was developed to provide more realistic outcomes. The new model could
simulate the exact contact surface between the hammer and the sample. The obvious drawback
of both models is the high ductility of the concrete sample, which will absorb a considerable
amount of energy. This can be seen in the significantly high deformation of concrete and
minimal displacement of the tendon. Also, severe penetration of elements can be observed in
the concrete medium. Moreover, the comparison of static double shear experiments and
analytical and numerical models showed no sensible agreement.

98
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-73: Shear behaviour of rock bolts in angled shear test combined with pullout test
(Kang et al., 2020)

Figure 2-74: Dynamic double shear models generated with FEM software (ABAQUS)
(Tahmasebinia et al., 2018a, 2021)

99
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Figure 2-75: Comparison of double shear experiment, analytical and numerical models
(Tahmasebinia et al., 2021)

Pinazzi et al. (2021) introduced a setup to examine rock bolts under combined loading
condition when both axial and shear loads could be applied simultaneously (Figure 2-76). The
current setup was used to test Australian rock bolt and indicated that under combined load
condition with significant shear load, fifty percent of the tensile load capacity should be
assumed in the design. No test was done on the cable bolts.

Figure 2-76: Combined loading setup (Pinazzi et al., 2021)

Singh et al. (2020; 2021) proposed a new analytical model shear response of fully grouted rock
bolt considering the effect of plastic strain hardening of steel on post-elastic stage. Results
showed that models that do not consider the effect of strain hardening overestimate the ultimate
shear displacement by up to 35%. This was also noted that the effect of strain hardening on
shear behaviour depends on the properties of the host medium. In other words, the outcomes
of the analytical model varies based on the mechanical properties of host medium.

100
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

Singh et al. (2022) utilized a new optical fiber instrument to study the behaviour of rock bolts
under perpendicular and angled shear loading. The research noted that fiber optic can provide
continuous measurement of strain along the rock bolt so that further analysis of the behaviour
can be done more accurate. Laboratory experiments on perpendicular angled double shear box
and numerical models on angled shear test using FLAC3D were used to verify the analytical
analysis and developed equations. Even though the research aimed to increase accuracy to the
behaviour of rock bolts under shear loading, no attempt were made to expand the understanding
on cable bolts. In addition, only perpendicular and 80° angled tests were implemented,
therefore, the higher installation angles were not examined.

Summary
Aligned with the systematic research carried out on the behaviour of tendons under different
loading condition, this chapter reviews the studies carried out in the field of axial loading of
tendons. Comprehensive literature review of tendons under static and dynamic loading
conditions were carried out in the PhD theses of Yang (2019) Khaleghparast (2021). After
introducing the ground anchor systems, laboratory facilities for pullout testing of tendons were
described with their specific features and drawbacks. Testing facilities were tabulated in two
different tables separating static and dynamic testing facilities. Different variables, such as
different types of tendon, grout, confinement method and testing mechanism were targeted for
the laboratory studies. It was recognized that dynamic pullout testing of tendons was usually
limited to rock bolts or low-strength cable bolts (15.2 mm cable bolts). Dynamic pullout tests
of 15.2 mm cable bolts did not produce desirable results, and tensile failure of the cable
occurred prior to pullout. Also, setups did not consider the behaviour of the barrel and wedge
as part of the test. Although many different setups were introduced, the main focus of the
majority of them was on rock bolts rather than cable bolts.

After introducing the pullout testing rigs, the pullout mechanism of cable bolts was analysed
according to laboratory experiments, analytical models and equation derivation along with
computer numerical analysis. Laboratory experiments proposed a three-step failure for cable
bolts, including adhesion failure, cable and grout interface failure, and residual failure. The last
stage could also occur in terms of stick-slip failure. Analytical models also introduced trilinear
failure modes for the tendon/grout interface under the pullout test. The models proposed the
first stage as an ascending elastic response corresponding to the elastic shear stress and strain
along the cable. The second stage is a descending plastic response due to decoupling and high

101
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

shear slip. The last stage is a horizontal response due to the residual resistance corresponding
to the frictional behaviour.

Once the pullout mechanism was fully understood, the effect of different parameters on pullout
behaviour was reviewed. Tendons' physical properties, confinement medium features,
embedment length, grout properties, and loading condition were among the affecting
parameters. The effect of bulbs, birdcages, and indentation on the cable bolt was compared
with plain cable bolts. It was seen that the modifications always did not result in improvement
in the pullout behaviour. Even though features such as spiral wire enhanced the maximum
pullout load, the residual behaviour was thoroughly worsened compared to plain cable bolts.
Bulb and birdcage amended the loading capacity of cable bolts considerably.

The effect of confinement pressure on pullout tests indicated that using steel as confinement
material overestimates the pullout load of the cable due to the incomparably high rigidity of
steel compared to rock. Using concrete, artificial rock or real rock core were suggested. The
minimum suggested size for the concrete confinement was to be 300 mm with a 45 mm
borehole in the centre. Consistent confinement pressure of resilience of confinement under
radial and lateral load can significantly impacted the pullout test results.

The effect of different encapsulation materials, including cement-based grout and resin, in the
pullout test showed that maximum pullout load values were in directly related to the strength
of the encapsulation material. However, the post-peak behaviour was dependent on the type of
encapsulation material. Cementitious grouts experienced higher post-peak values than resin.
Also, stick-slip behaviour was usually observed in the samples encapsulated with cement grout.

Laboratory tests undertook the examination of different loading rates in the pullout test. Rock
bolts and low-strength cable bolts were mainly examined under dynamic testing, and high-
strength cable bolts were seldom seen among the successful tests. Dynamic test results were
presented in terms of absorbed energy and displacement. Results stipulated that the absorbed
energy of cable bolts was remarkably higher than rock bolts.

For those studies by Yang (2019) and Khaleghparast (2021) recently completed and carried out
on shear loading of tendons in both static and dynamic modes, recent studies were reviewed,
and developments in laboratory testing methods and numerical analysis were evaluated. Newly
developed testing setups have a propensity for simultaneous axial and shear testing.

To sum up, following research gaps are felt:

102
Chapter 2 | Literature Review of Axial and Shear Loading of Tendons

• Study of the behaviour cable bolts under shear loading is limited to perpendicular static
loading. There is no record of dynamic shear test of cable bolts in any angles by the
time of prepration of this document. Even though, dynamic shear test of rock bolts
installed perpendicular to shear surface has been undertaken, the gap is felt for cable
bolts.
• Angle shear test of rock bolts in static loading mode has already been carried out;
however, angle shear test of cable bolts still requires more investigations.
• Pullout test of cable bolts has been undertaken in both static and dynamic modes,
however, since the testing mechanisms were different, lack of a meaningful comparison
of the results of static and dynamic tests is felt. Majority of the current pullout testing
machine are only capable of running either static or dynamic test. Therefore, design and
development of a testing facility that can implement both static and dynamic tests would
be a great opportunity to understand the effect of loading rate on the performance of
cable bolts.
• Numerical and analytical models developed for cable bolts simplify the geometry of
the cable bolt to a solid smooth rebar. Consequently, current models could not predict
the effect of geometry of the cable bolt in their axial loading capacity and their
performance in pullout tests.

103
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Laboratory Study on Static and Dynamic Shear Loading


of cable bolts

CHAPTER 3
Laboratory Study on
Static and Dynamic Shear Loading of Cable Bolts

Introduction
For decades, there has been a significant reliance on tendons (rock bolts and cable bolts)
technology for ground reinforcement in mining and civil engineering. Various studies have
been reported on the axial strength characterization of tendons; nevertheless, a few of these
studies, specifically experimental, have reported on shear behaviour. Studies by Dulacska
(1972), Hyett et al. (1992), Ferrero (1995), and Goris et al. (1996) are some of the pioneering
experimental studies of tendons subjected to shear load. Hence, the shear failure of tendons in
underground applications with respect to seismic events, as well as an understanding of the
performance of tendons under dynamic loading conditions, still requires more attention from
researchers in the field. Compared to static loading analysis, equations and analysis of dynamic
loads originated from seismic sources and their effect on supporting system are more
complicated (N. Jiang & Zhou, 2012); however, it can still be practical by assuming some
simplifications. In static loading, the combination of factors, including gravity, stress
distribution, time and friction, act simultaneously to control or disturb the stability condition
of the excavated area. This is while in dynamic loadings such as in blasting, rockburst or any
other seismic sources, the components of time and friction are negligible. In practice, tendons
might be subjected to axial or shear and most possibly/or probably a combination of both
simultaneously.

Since the beginning of the twenty-first century, the Australian mining industry has invested in
developing testing equipment and facilities to better understand the shear behaviour of tendons
104
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

with a greater tendency to cable bolts. Shear testing facilities developed at University of
Wollongong (UOW) (Aziz et al., 2003, 2019; Rasekh et al., 2017) and axial dynamic testing
facility of Western Australia School of M 3ine (WASM) (Villaescusa et al., 2005) are among
them. As one of the active universities in this field of research in Australia, University of
Wollongong has introduced several laboratory technologies to examine the shear behaviour of
both rock bolts and cable bolts in the last two decades. Five types of testing machines have
been developed at UOW in order to examine the variety of tendons in different laboratory
scales. In 2003, a rectangular double shear testing rig for rock bolts (MK-I) was introduced by
Aziz et al. (2003). In all of the designs, tendons were installed and encapsulated inside concrete
samples and cementitious grout or mastic resin to replicate the field's similar condition. Steel
confinements were utilized to confine concrete samples and prevent possible radial cracking of
the concrete. Later in 2010, a larger scale rectangular double shear testing rig was intruded
(MK-II). The new rig was able to test cable bolts as well as rock bolts (Craig & Aziz, 2010);
however, the frictional forces between the concrete blocks added extra resistance force to the
shear resistance of the tendon. This made the interpretation of the results much more
complicated. Meanwhile, in 2015, another ACARP project on shear studies, aided in a
collaboration between Megabolt Australia and UOW and the single shear testing of cable bolts
was launched. Although the mentioned setup could test high-strength cable bolts, test samples
were massive, and test implementation was time-consuming. More importantly, the axial load
on the tendon led to the axial pullout of the tendons from the encapsulation material. Hence,
results of the shear test were negatively affected by axial deformation of the cable bolt.

Later in 2017, the previously developed MK-II was modified and introduced as MK-III
(Mirzaghorbanali et al., 2017) to prevent friction between concrete blocks. Steel braces around
the concrete blocks were installed to hold the side blocks in their place and neutralize the effect
of the moment. Even though unnecessary friction between concrete blocks was removed and
the results were only representative of shear loading capacity of cable bolt, the confinement
system was not performing effectively, and both lateral and radial cracking of the samples was
inevitable.

Being aware of the deficiencies of MK-III, in 2019, the cross-section of the concrete samples
was altered to a circle instead of a square (Aziz et al., 2019). Thus, the rectangular samples
were replaced by cylindrical samples, which resulted in more effective and homogenous
confinement in a circular cross-section. This setup (also called MK-IV) has been the latest

105
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

development of horizontal double shear testing rigs at UOW and most likely in the world at the
time of this study.

Following previous efforts carried out at UOW, in 2020, the MK-I setup was selected to be
subjected to dynamic loads in order to examine the behaviour of rock bolts under dynamic
loading conditions (Khaleghparast et al., 2020). Among the available testing setups, MK-I was
smallest; thus the minimum amount of dynamic energy was required for the tests. Based on the
properties of the available impact testing machine and existing limits of energy transformation,
MK-I was the most appropriate setup. Even though the examinations perfectly compared the
behaviour of rock bolts under both static and dynamic loads; MK-I was not strong enough to
test the stronger tendons such as cable bolts.

As part of the current study, in 2021, the design of the angle shear setup developed by Grasselli
(2005) was modified and resulted in a new angle shear testing rig to better simulate the shear
behaviour of the cable bolts (Aziz et al., 2021). Design of Grasselli was strong enough to
undertake shear test of rock bolts, whereas design of Aziz was externally reinforced so that
double shear test of cable bolts was practical. The new setup was capable of implementing
angled shear tests at two different angles of 30 and 45 degrees. Accompanied by previous
horizontal (90 degrees) double shear test results, angled shear test results could potentially
develop a comprehensive understanding around the effect of installation angle on the behaviour
of tendons. Test results will be discussed in chapter four.

Furthermore, in alignment with previous studies carried out on the shear behaviour of tendons
at the University of Wollongong, the shear behaviour of cable bolts under dynamic shear
loading conditions was aimed to be achieved as part of the milestones of this thesis. For this,
the latest introduced double shear testing rig (MK-IV) was reinforced to be subjected to
dynamic and static loads. Comparison of the static and dynamic shear tests could amend the
understanding of the behaviour of the tendons under different loading conditions.

This chapter summarizes the process of sample preparation for static horizontal and angled
shear as well as dynamic horizontal shear tests.

Mechanisms and Preparation of Perpendicular Double Shear Test


The mechanism of double shear testing of cable bolts is based on the movement of the middle
block between two fixed side blocks while the cable bolt connects all three blocks together
(Haile et al., 1998; X. Li et al., 2017). Aziz et al. (2003) developed the first double shear box.
106
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

In this test, the side blocks were fully constrained against movement. In order to replicate the
real field situation, it is vital to apply a pretension load to the cable bolt. In addition, in real
field conditions, the length of the encapsulated bolt may exceed eight meters, while in the
laboratory scale test, the whole cable length may vary from one to two meters depending on
the type of the shear box being used. Hence, it is essential to ensure that the excessive load on
the side blocks does not cause the concrete to crack or pullout of the cable bolt. To prevent the
potential pullout, both ends of the cable bolt should be gripped by Barrel and Wedge. This can
represent a very long encapsulation length where no pulling out occurs of the cable bolt.
Finally, rock discontinuities are usually in-filled so that there is friction between blocks of rock
during the shear process. In these sets of tests, the effect of friction between block is neglected
as a simplification assumption. Figure 3-1 shows the mechanism of the perpendicular double
shear test setup. Cylindrical concrete blocks with a e diameter of 300 mm are used for
perpendicular double shear. Green arrows show the direction of the applied load, and red
supports represent the boundary conditions. The yellow arrows demonstrate the pretension load
and the confining load applied on the B&W. The load on the middle block can be applied both
statically and dynamically.

Figure 3-1: Mechanism of the perpendicular double shear test

In order to simplify the experiments, the followings are assumed:

• There is no friction between shearing surfaces, and both shear surfaces are smooth
• Concrete blocks are rigid enough to prevent lateral, or radial cracks propagating
• Cable bolts are long enough to resist being pulled out
• Energy absorption of the test setup is equal in all tests

Figure 3-2 shows the general view of a new cylindrically shaped Double Shear Testing Rig
(MKIV). The 300 mm diameter steel circular clamps permit the application of external
107
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

confinement to the cylindrical concrete medium. The outer Double Shear cylinder sides rest on
support cradles to provide stable positioning during the shearing stage, allowing the longer
central part of the box to shear vertically down with minimal lateral movement. With the use
of four steel trusses and a 30 mm thick reinforced side plate, this arrangement permits a
minimum of contact occurring between the concrete blocks' side faces during shearing, thus
preventing part of the shearing force from being spent on overcoming the medium joint sides
rubbing friction. However, for higher axial cable pretension loads, the two 30 mm thick steel
plates are further reinforced laterally with; (a) welding a 30 mm thick and 100 mm bar across
the top half and (b) a 20 mm thick and 200 mm square plates inserted on the cable to minimise
their inward bending. The whole assembly is mounted on the outer base cradle half-cylinders,
fastened, and secured to the carrier base frame using eight 20 mm diameter threaded bars. The
concrete steel clamp is fastened to the lower half of the full clamp circle with three bolts pre
the side flanges. The cable bolt is inserted into the full length of the shearing block, fitted with
a load cell on each end, held tight with a barrel and wedge, and then pretensioned to the desired
axial load.

1. Cable bolt
2. Barrel and Wedge
3. Load cell
4. Lateral confining plates
5. Lateral trusses
6. Carrier base frame Base body
7. Steel bolts and nuts
8. Side blocks seats
9. Concrete radial Confinement

Figure 3-2: Cylindrical double shear test apparatus (MK-IV)

During the static test, the applied load (f) causes vertical displacement (l) of the middle block
until the failure of the cable bolt occurs on one or both sheared surfaces. As f is distributed on
both shear surfaces, the load on each surface would be F = f/2. Also, the absorbed energy of
the cable bolt before failure would be the integration of the load in displacement intervals
(∫ 𝑓𝑑𝑙). According to the experimental studies of Hass (1976) on rock bolts, the shear strength
108
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

of a joint is the sum of the bolts contribution, the friction of the shear surface, and the shear
resistance of the asperities (Haas, 1976). Analytical and numerical studies of Spang and Egger
(1990) resulted in a general equation to calculate the shear resistance of the bolts installed at
different angles to the shear surface (Equation 3-1)

𝑇0 = 𝑃𝑡 [1.55 + 0.011𝜎𝑐1.07 𝑆𝑖𝑛2 (𝛼 + 𝑖)]𝜎 −0.14 (0.85 + 0.45 tan 𝜑) Equation 3-1

Where:

𝑇0 = maximum shear resistance of grouted rock


𝑃𝑡 = maximum tensile load of the bolt
𝜎𝑐 = uniaxial compressive strength of the rock
𝛼 = inclination of bolt with respect to joint plane
𝜑 = internal friction angle, and
𝑖 = asperity angle
Considering that bolts in horizontal double shear are installed perpendicular to the shear
surface, Equation 3-2 can be rewritten as Equation 3-1.

𝑇0 = 0.85𝑃𝑡 [1.55 + 0.011𝜎𝑐1.07 ]𝜎 −0.14 Equation 3-2

Although Equation 3-1 was initially derived for rock bolts, it was used in the preliminary steps
of the design in this study to roughly estimate the required load for the shear failure of the cable
bolts.

In the next stage, after the implementation of the static tests, the load-displacement graph
obtained from static tests can be utilized in order to calculate the absorbed energy by the bolt
using the following Equation:

Equation 3-3
𝐸𝑠𝑡𝑎 = ∫ 𝑓 𝑑𝑙

Where 𝐸𝑠𝑡𝑎 is the static energy, f is the applied load and dl is displacement. Assuming the
required energy for cable failure in static and dynamic loading conditions are equal (𝐸𝑠𝑡𝑎 =
𝐸𝑑𝑦𝑛 ) then the same energy would be applied in the dynamic test. The kinetic energy required
for dynamic testing is provided by the free fall of a drop hammer on the shear test setup. A
specific energy can be applied by the selection of specific heights of free fall. Therefore,
potential energy of energy turns into kinetic energy and other dissipating energies such as heat,

109
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

sound and vibrations. Once the hammer hits the sample, momentum as well as work and energy
laws can be employed to calculate the work done by the impact. In the case of the collision of
the drop hammer and the middle block, the collision can be considered as partially inelastic,
however, in order to simplify the math, the collision is considered perfectly inelastic. It other
words, it is assumed that the hammer and the block stick together during the motion and move
with each other after the first impact.

Then, the momentum of the drop hammer and moving middle block can be represented as:

𝑚ℎ 𝑣ℎ + 𝑚𝑏 𝑣𝑏 = (𝑚ℎ + 𝑚𝑏 )𝑣𝑓 Equation 3-4

Where:

𝑚ℎ = mass of the hammer


𝑣ℎ = velocity of the hammer before impact
𝑚𝑏 = mass of the middle block
𝑣𝑏 = velocity of the middle block before impact
𝑣𝑓 = theoretical final velocity of the total mass after impact
In the Equation 3-4, 𝑣𝑏 = 0, because the middle block is at rest prior to the hammer impact.
Also, for the hammer free falling from the height (h), the final velocity of the hammer before
impacting the middle block is:
𝑣ℎ = √2𝑔ℎ Equation 3-5

Then:
𝑚ℎ Equation 3-6
𝑣𝑓 = √2𝑔ℎ
(𝑚ℎ + 𝑚𝑏 )

And finally, according to the energy conservation law:


1 1 Equation 3-7
(𝑚ℎ + 𝑚𝑏 )𝑣𝑓 2 = (2𝑘)𝑥 2
2 2

Or:
(𝑚ℎ + 𝑚𝑏 )𝑣𝑓 2 Equation 3-8
𝑘=
2𝑥 2
Where:
𝑘 = stiffness of the bolt against shear force
𝑥 = final displacement of the middle block
110
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Based on Equation 3-8, the velocity and displacement of the sample can be recorded during the
experiment and the stiffness of the bolt against the shear load can be calculated.

3.2.1 Sample Preparation for Perpendicular Double Shear Test


Aziz et al. (2019) reported on the procedure of concrete sample preparation for the cylindrical
double shear test; hence this section provides a brief overview of the concrete sample
preparation.

Sample preparation includes three main steps: casting and curing concrete samples, preparation
of the setup for testing, and preparation of the sample on the test day. Based on the design
requirements of a cylindrical double shear box, each test requires three cylinders of concrete
with a diameter of 300 mm. Compressive strength of the blocks was 40 MPa to resemble the
field condition for soft to medium rock (Hudson, 2003) (test results can be found in section
3.4.3). The height of the side blocks is 300 mm, and the height of the middle block is 450 mm.
5 mm cardboard cylindrical moulds were supported by a plywood structure for casting the
samples. Each block has a central cylindrical hole with a diameter ranging between 35-45 mm,
representing the rock borehole. The boreholes were preserved before casting concrete with the
help of steel bars being wrapped with 6 mm clear PVC tubes, the same as that reported by Aziz
et al. (Aziz et al., 2018). Bars were held in place by wooden frames (Figure 3-3).

The day after casting, the concrete blocks were removed from the molds and cured as suggested
(AS-1012.9, 1986). After 28 days, concrete blocks were taken for the second stage of
preparation. In the next stage, the testing setup is assembled. To assemble the test sample, three
half-cylinder 12 mm steel cradles were perfectly aligned and adjusted on their seats by the help
of guiding rods (Figure 3-4-A). While the side seats were fixed, the middle seat could be
removed in the next stages in order to allow free vertical movement of the middle block. Two
of 300 mm cylindrical concrete samples rest on the side cradles, and one 450 mm concrete
cylinder sits on the middle cradle (Figure 3-4-B). It is important to ensure the gap between
cradles remains clear in order to prevent the frictional contact of the concrete blocks. To allow
this, 6 mm single-sided adhesive foam tapes were placed between the blocks (Figure 3-4-C).
This also prevents grout leakage from the boreholes and fills the gap between the concrete
blocks.

111
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Figure 3-3: Prepared mould for casting concrete samples

A B

C
Figure 3-4: Double shear test assembly procedure
112
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Once the concrete was positioned, steel C-section channel braces were screwed together to
clamp the blocks. Braces react against the existing moment, which could result in frictional
force between blocks during the test (Aziz et al., 2019) (Figure 3-5-D). In the next stage, a
cable bolt was inserted in the pre-existing borehole of the concrete cylinders and fixed from
both ends by a pair of proper barrels and wedges. Two sets of load cells were inserted axially
under the B&W to measure the axial load on the cables (Figure 3-5-E). In order to simulate
real field conditions, the cable bolt was pretensioned adequately based on the type of cable
(Figure 3-5-F-1 and F-2). Once pretensioned, the desirable grout mixture is cast inside the
boreholes through precast access holes on the concrete blocks. A Blue Heeler tensioner was
used of pretensioning of the cable ends.

F-1 F-2
Figure 3-5: Double shear test assembly procedure

Once grouted, samples were left for at least 14 days to give the grout enough time to reach the
workable strength (roughly 50 MPa). Cubic samples of the grout were cast to measure the
compressive strength of the grout prior to testing to ensure the required grout strength has been
reached (results can be found in section 3.4.1). In the third and last preparation stage, samples
were prepared for testing. Samples were mounted between the loading plates of the loading
113
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

machines with the side blocks resting on rigid supports. The seat under the middle block was
removed, and the middle block was pushed downward vertically by a compression loading
machine. Loading can be carried out either statically or dynamically. For the static shear test,
the load was applied by a 5000 kN compression testing machine (Figure 3-6). Displacement
and load data were measured by the load cells and LVDTs and recorded by data loggers
connected to a computer. For the dynamic shear tests, the test sample was placed under the
drop hammer (Figure 3-6). As the hammer falls, a laser gate triggers the load cell, which allows
data to be recorded by a data acquisition system. The collected data were transferred to a
computer where the results could be analyzed. A high-speed camera, Fastec Trouble-shooter,
was utilized to capture the high-energy impact between the drop hammer and the MK-IV shear
box with a high degree of accuracy. This allowed an accurate analysis of the displacement of
the central block during the shearing load over a period of time. Shearing displacement of the
central block was also monitored through utilizing a remote displacement reader placed
underneath the central block.

Figure 3-6: Static (left) and dynamic (right) double shear test setups ready for the test

Once the test was finished, the sample was taken out of the compression test machine area, and
the top confinements were unscrewed and taken apart. Figure 3-7 shows a test sample after the
test. As can be seen, the middle block was pushed down until the cable bolt failed on one or
114
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

both shear surfaces. If there are minor radial cracks in the concrete, the results might be
questionable and need to be analyzed carefully. If cracks are fully open, test results are not
accepted.

Figure 3-7: Test sample after the test

Mechanism and Preparation of Angled Shear Test


For the angled double shear test, due to the geometry of the test, the concrete block must be
rectangular, and the embedment length of the cable would be longer than in perpendicular tests.
Also, due to the sharp angle of the installation (30 or 45 degrees), one cable bolt cannot pass
through all three blocks. Thus, it is necessary to use two cable bolts connecting the middle
block to each of the side blocks. Thus each cable bolt passes through one shear surface, and
eventually, they resist the shear load on two different shearing faces. Side blocks must be fully
constrained against rotational and transitional displacements (Figure 3-8). Similar to the
perpendicular double shear test, the load is applied on the middle block, and cable bolts are
gripped from both ends in order to prevent the pullout of the cable during the shear test.
Pretension load can be applied on the cable bolts providing that:

• The pretension load is equal on both cable bolts


• The middle block is fully restrained against uplifting during pretensioning.

Figure 3-9 exhibits how two cable bolts cross the shear surface in two different designs for 30°
and 45° angled shear tests. As can be seen, for the 45° angled samples, cable bolts extrude from
the bottom of the middle block, while for the 30° samples, cable bolts are in the sides of the
middle blocks.

115
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Figure 3-8 Mechanism of the angled double shear test

Figure 3-9: Design of 30 and 45-degree angle double shear tests

3.3.1 Sample Preparation for Angled Double Shear Test


According to the design, concrete samples were poured and cured carefully. Boreholes with
the proper installation angle were preserved in the samples (Figure 3-10).

Due to the practical limitations of the setup, the installation of the cable in the angled double
shear is different from the perpendicular setup. In the angled double shear, a cable bolt was
inserted into the middle block borehole, and the bottom barrel and wedge were installed. In
order to prevent leakage of the grout, all the hole openings were sealed with 6 mm compressible
foam (Figure 3-11). Also, in order to maintain the frictionless gap between the blocks, flat
Teflon sheets were attached to the surface of the blocks. For the 30° angled samples, the bottom

116
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

end of the cable bolt was sealed by 3D printed caps in order to prevent leakage of the grout
onto the shear surfaces (Figure 3-11).

Figure 3-10: Concrete samples before and after casting

(A) (B) (C)


Figure 3-11: Angle shear test preparation process (Part A)

Then, the side blocks were lifted by the roof crane and fitted to the sides of the middle block
so that the cable could pass through the boreholes of the side blocks (Figure 3-12).

In the next step, lateral braces were bolted around the samples in order to constrain
displacement of the side blocks (Figure 3-13-A). Once the blocks were fixed against side
displacement, a reinforced C-Section was bolted to the top of all three blocks in order to avert
the uplifting of the blocks during the pretensioning process (Figure 3-13-B).

117
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

(A) (B)
Figure 3-12: Angle shear test preparation process (Part B)

(A) (B)
Figure 3-13: Angle shear test preparation process (Part C)

Then, the pretension load was gradually and equally applied to each cable bolt to ensure both
sides had an equal or near equal pretension load. If possible, two sets of pretensioners were
hired to simultaneously test both sides. The pretension load was measured by axial load cells
installed beneath the top barrel and wedge (Figure 3-14). When the pretension load reached the
desired load, grout material was injected into the boreholes until they were full.
Finally, the sample was hoisted to the testing machine by a roof crane. The side blocks were at
rest on the rigid beams, and the downward load was applied by the slow-rate loading machine.
Figure 3-15 demonstrates the facilities and logger utilized for measuring the load and
deformation during the test.

118
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Figure 3-14: Applying and measuring pretension load

Figure 3-15: Testing machine and instrumentation facilities

119
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Selection of Testing Materials


The main materials for the tests in this thesis were cable bolts, cementitious grout, and concrete.
In the following section, the mechanical properties of the selected materials have been obtained
by laboratory experiments or by referring to relevant references.

3.4.1 Cementitious Grout


It is a well-known fact that the strength of cement-based mortar increases logarithmically over
the curing time (Neville & Brooks, 2010). Like the majority of cement-based materials, grout
also reaches its peak strength after 28 days, and further strength increase is very slow afterward
(Aziz, Majoor, et al., 2017; Mirzaghorbanali, 2019). However, since the anchors are immediate
roof supports in underground areas, time would be crucial in the geotechnically active medium.
Hence, it is essential to understand what parameters and how they may affect the performance
of the encapsulation material. To do this, four different types of commercialized cementitious
grouts from Australia were selected to be analyzed. The selected grouts were all made of
different types of cement with different mixtures. Grout type 4 was the only one with additives
to accelerate the setting. The commercial names of the products have not been mentioned
intentionally.

1. Grout type 1: Slow-setting grout with recommended water/cement ratio of 30-40%.


2. Grout type 2: Slow-setting grout with recommended water/cement ratio of 34-40%.
3. Grout type 3: Slow-setting grout with recommended water/cement ratio of 35-40%.
4. Grout type 4: quick-setting grout with recommended water/cement ratio of 18-23%.

Grout samples were mixed carefully according to the manufacturer's instructions for mixing.
The average recommended range of water was added to each grout. It means the water ratio
for grout type 1 to 4 was 35%, 37%, 37%, and 20%, respectively. In order to discern the
different mixing conditions of different grouts, cubic samples of 50 mm were poured and tested
as recommended by (ACI Committee 308, 1980). Samples were cured for 28 days according
to cement-based materials curing standards (ACI Committee 308, 1980), and all tests were
carried out using a MaTest loading machine with the loading rate of 0.15 ± 2% MPa/sec as
specified by Standard methods for testing concrete (AS_1012.9, 1986).

120
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Figure 3-16: Grout samples poured in 50 mm cubic moulds

Figure 3-17 shows the results of five uniaxial compression tests carried out on all four types of
grout after 28 days of curing. The result has been sorted in ascending order. As can be seen,
grout types 2 and 3 not only had a lower UCS than for the other two grouts, but also the UCS
results for these samples varied over a wide range and was not constant. In contrast, grout
Types 1 and 4 showed fairly stable and constant behaviour in all the tests and presented
significantly higher compressive strength. It is noteworthy that quick-setting grout (Type 4)
reached the peak of ~80 MPa after 18 days with only 20% of water consumption, while grout
Type 1 met the peak of ~55 MPa with almost twice the amount of water.
100

90

80

70
UCS (MPa)

60
Type 1
50
Type 2
40
Type 3
30 Type 4
20

10

0
1 2 3 4 5
Test Number

Figure 3-17: Comparison of UCS in different types of grout

121
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Due to the greater consistency in the results of the grout Types 1 and 4, these two were selected
for further analysis. The following sections evaluate the effect of curing age, water/cement
ratio, and curing conditions on the mechanical behaviour of cementitious grouts types 1 and 4.

a) Effect of curing age on the strength of quick-setting grout


In this step, regimented grout UCS tests were carried out to compare the strength in the early
ages of a quick-setting grout. More than 45 uniaxial compression tests of the 50 mm grout
cubic samples were tested at different curing ages to illustrate the role of grout age on its
mechanical properties. Hence going forward, cubic samples were subjected to a compression
test at 4 hours, 12 hours, 24 hours, 7 days, 28 days and 32 days after casting, respectively. All
the tests were carried out in the room temperature.

Figure 3-18: Left: 50 mm cube samples cast for compression tests for UCS

summarizes test results of the grout’s uniaxial strength at different curing times from 4 hours
up to 32 days. As clearly shown, fast-setting grout reaches its highest strength in the first 24
hours, followed by a slight increment in the next 32 days. The logarithmic trend of strength is
noticeable as well. Additionally, a wider error bar in the early hours of grouting clarifies
uncertainties in grout strength. As time goes by, lower standard deviations and smaller error
bars reveal the consistency of the test results. In conclusion, more than 85% of the ultimate
strength is gained after the first 24 hours so that the material is classified as quick-setting grout.
Figure 3-19: UCS grout strength of grout Type 4 with respect to the curing age (w/c= 20 %)

122
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Similarly, Figure 3-20 presents the average UCS test results for slow-setting grout over 28
days. In slow-setting grout, almost 85% of the ultimate strength can be achieved after 14-16
days.

70

60

50
UCS (Mpa)

40

30

20

10

0
7 days 16 days 21 days 28 days
Age

Figure 3-20: UCS strength of grout type 1 with respect to the curing age (w/c= 35%)

According to

, substantial changes in grout strength occur in the first 24 hours. Hence, due to the quicker
response to the curing age, grout Type 4 was selected for analysis of the effect of curing age
on the failure mechanism.

123
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Figure 3-21 shows the failure surfaces of grouts at the three different curing ages of 4 hours,
12 hours and 24 hours. It was recognized that shear failure surfaces in samples younger than
24 hours were randomly distributed. In contrast, for samples older than 24 hours, shear surfaces
were perfectly formed on four sides of the samples. This might directly be related to the
uncompleted hydration process of cementitious grout. This also complies with the findings of
(Huang et al., 2019) as well. In their numerical and analytical modeling, Huang et al. (2019)
studied the macro mechanical behaviour of concrete samples. Mesostructured concrete samples
were modeled based on the tomography obtained from computed tomography. According to
their study, when the friction between the rigid loading plate and the concrete specimen is
negligible, as displacement increases, the fracture surfaces remain parallel to the loading
direction during the loading process (Figure 3-21-D and E). The same failure modes were also
observed in the 4-hour and 12-hour aged grout samples (Figure 3-21-A and B). This is while,
in cases where there was high friction between the loading plate and sample surfaces, not only
higher strength of concrete was observed, but also fractures pattern was symmetrically
propagated on four surfaces of the sample. Also, two surfaces contacting the rigid loading
plates remained complete, while cracks were growing into the interior of the sample until the
sample fragmented on free surfaces and fragmented edges split from the main body of the
specimen (Figure 3-21-F) (Huang et al., 2019). This was in agreement with the failure mode
of 24-hour aged grout (Figure 3-21-C).

124
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

A B C

D E F
(a) 4 hours (b) 12 hours (c) 24 hours
Figure 3-21: Failure mode of grout samples after UCS test

Additionally, according to British standard 12390-3:2002 for testing hardened concrete, failure
modes plotted in Figure 3-22 are considered satisfactorily hardened concrete. Any other failure
is considered as either loading malfunctions or insufficient attention to preparation procedure
(BS 12390-3:2002, 2002).

Figure 3-22: Satisfactory failure of a perfectly hardened concrete sample (BS 12390-3:2002,
2002)

b) Effect of Water/cement ratio


The correct water content in cementitious grout accelerates the hydration process of cement
mortar. Different grout materials have their own recommended range of water/cement (W/C)
ratio. While the grout strength is highly dependent on the presence of a sufficient amount of
water, extra water can significantly affect the strength. A regimented study on the effect of
W/C on the strength of cementitious grout has been done by (Aziz, Majoor, et al., 2017; Mirza
et al., 2016).

125
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

The recommended W/C ratio for grout material Type 4 ranged between 20-22%. The effect of
water percentage on the mechanical properties of the grout was examined by adding different
water ratios ranging between 18-23% to evaluate the effect of extra or a shortage of water.
Figure 3-23 depicts the average strength of grout Type 4 after 24 hours of curing. The strength
of the grout in the suggested range of W/C is between 53-58 MPa. Lower water percentages
resulted in lower strength.
70.0

60.0

50.0
UCS (MPa)

40.0

30.0

20.0

10.0

0.0
18 19 20 21 22 23
w/c ratio (%)

Figure 3-23: Effect of w/c ratio on the strength of grout Type 4

c) Effect of moisture exposure on grout


Quick-setting grout has a swift reaction to water; hence, it was expected to see changes in
behaviour after being exposed to humid weather. In order to examine this, two different bags
of grout were kept at two different moisture levels, one was sealed and preserved in a sealed
bucket with zero humidity, and the other one was fully exposed to the standard laboratory air
conditioning (25℃) with the moisture level between 40-60% for 20 days. After 20 days,
samples of each bag were mixed with different w/c ratios and tested after 24 hours. A
comparison of the fresh grout and the exposed bag revealed that the strength of the grout
samples from the bag exposed to moisture was almost 75% of the samples of grout kept in zero
humidity conditions (Figure 3-24).

126
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

70.0

60.0

50.0
UCS (MPa)
40.0

30.0

20.0

10.0

0.0
18 20 20 22 22 22
Old 46.3 48.4 49.5 41.3 44.9 45.1
Fresh 61.3 50.9 54.0 57.8 59.2 59.7
Water/Cement Ratio (%)

Figure 3-24: Comparison of Old and New bags in 24-hour test

d) Effect of curing condition on the grout strength


Due to the nature of the rock in the field and concrete in laboratory conditions, it is expected
that a portion of the water of the grout to be absorbed by the host medium. This may lead to a
change in the w/c ratio of the grout, especially in the contact area with rock or concrete. This
is why grout samples on a laboratory scale are placed in water tanks or maintained in high-
humidity rooms while curing (BS 12390-3:2002, 2002).

In order to examine the effect of curing conditions on the strength of the grout samples, two
different sets of slow-setting grout with the w/c ratio of 32% were cast at the same time. While
one set was fully saturated in the water tank, the second set was left outside the mold in the
standard laboratory air conditioning for 21 days. Each set of saturated samples was taken out
of the water tank 24 hours prior to testing and left at the standard room temperature to dry. A
total number of 18 samples for each set were divided into three parts to be tested in seven-day
intervals. The average UCS test results for each interval are shown in Figure 3-25. As can be
seen, the curing conditions have not significantly impacted the strength of the samples;
however, samples kept in the water tank were slightly stronger.

127
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

70

60

50
UCS (Mpa)
40

30

20

10

0
7 Days 16 Days 21 Days
Age

Soaked in water Left in room condition

Figure 3-25: Effect of curing condition on the strength of the grout

e) Mechanical properties of the grout


Finally, to utilize the grout's mechanical properties for future analysis, two single-axis polyester
linear strain gauges with a gauge size of 20 mm* 1.2mm (TML, 2020) were glued onto the
grout sample in principal directions (Figure 3-26). The average uniaxial compression test
results of the cubic samples led to the mechanical properties of the grout, as listed in Table 3-1.

Figure 3-26: Lateral and vertical strain measurement of grout samples

128
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

100
90
80

Stress (MPa) 70
60
50
40
30
20
10
0
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008
Strain

Figure 3-27: Stress-Strain graph of a grout sample

Table 3-1: Grout properties used in the numerical model


Density Compressive strength Elastic modulus Poisson ratio Friction angle
(kg/m3) (MPa) (GPa) degree
2300 80 20.6 0.22 ~22

3.4.2 Mastic Resin


In some parts of this thesis, specifically in pullout tests (Chapter 5 and 6), the performance of
mastic resins are compared with cementitious grout. Lokset Anchor Pack, a two-part flowable
grout, was selected. It consists of Polyester Pb1 Mix & Pour and Lokset catalyst components
(Minova, 2020) being selected. In order to test the samples, cubic and cylindrical samples were
selected according to the testing procedure proposed by Aziz ((Aziz et al., 2013)). 40 mm cubic
moulds made of Polycarbonate sheet (Figure 3-28) and cylindrical PVC tube with a diameter
of 32 mm and height of higher 60 mm (H:D >= 2 as recommended by Ma (2014)) were prepared
for casting resin samples.

129
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

40*40 mm2

Figure 3-28: Cubic moulds prepared for resin casting

Although the mixture ratio of resin to catalyst was reported by Aziz et al. (2013) to be 2% for
Mix & Pour resin, the resin did not set properly. Hence, the ratio of 5% was selected as it was
suggested by the manufacturer for sausage resin. The sample was taken out of the mould after
20 minutes (Figure 3-29) and tested after one day. According to the manufacturer sheet, resin
samples were expected to reach their optimum strength after six hours (Minova, 2020).

Figure 3-29: Cubic and cylindrical resin samples

Finally, samples were tested in accordance with British Standard for testing resin and cement
compositions for use in construction (BS 6319-1, 1984).
Figure 3-31 presents the result of the compression tests for resin samples in ascending order.
As can be seen, the uniaxial compressive strength of resin was in the range 45-60 MPa. It can
be said that the average strength is between 50-55 MPa which is close to the UCS of grout
Type 1.

130
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Figure 3-30: Cubic sample of resin under compression test

70

60

50
UCS (MPa)

40

30

20

10

0
1 2 3 4 5 6 7 8
Test number

Figure 3-31: UCS results of resin samples

3.4.3 Concrete Blocks


The strength of the concrete samples was chosen to resemble the rock host medium properties.
The uniaxial compressive strength of Sandstone and Shale as the main host medium of coal
layers and tunnels in Australia, depending on their classifications, varies between 12-50 MPa
for Sandstone and 7-40 MPa for shale (Table 3-2) (Bertuzzi & Pells, 2002). Based on
Australian Standards related to geotechnical site investigations, rocks with the UCS ranging
between 20-60 MPa are classified as high-strength material (AS_1726, 2017). Since the
concrete sample needs to replicate the real rock behaviour, efforts have been made to maintain
the strength of the concrete samples in the range of 40-60 MPa. As concrete was provided by
local suppliers in different time intervals, keeping the consistency of strength was not easy;

131
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

however, the strength was tested for each set of casting separately. The maximum aggregate
size of the concrete was 10 mm, and in order to preserve consistency, all the test samples of
the same kind were cast from one batch of concrete supplied by a local supplier.

Table 3-2: Mechanical properties of Sandstone and Shale in Australia, after (Bertuzzi &
Pells, 2002)
Substance
Strength Elasticity
Class
UCS 𝜎𝑡 E
m𝝴
(MPa) (MPa) (GPa)
Sandstone I/II 12-50 2-6 2-9 8-14
Sandstone III 7-25 0.5-3 4-10 6-10
Sandstone IV/V 1-7 0.1-0.5
Sandstone Residual soil <1 0
Shale I/II 7-40 1-4 2-10 7-15
Shale III 2-15 0.1-2 2-10 5-10
Shale IV/V 1-2 <0.2
Shale Residual soil <1 0

To examine the mechanical properties of the ordered concrete, cylindrical samples of two
different sizes, 100mm*200mm and 150mm*300mm were prepared for testing (AS_1012.14,
2018). Samples were cured in the water tank at a standard temperature for at least 28 days,
ensuring 99% of the ultimate strength. Since all the tests could not be done right after 28 days,
the properties of the concrete samples were examined precisely on the same date as the actual
shear test.

In order to prepare the sample for the uniaxial compression test, samples were taken out of the
tank and dried for one day prior to testing. The rough end surface of the cylinder was capped
by plaster (AS_1012.9, 1986). When plaster sets, a compression meter equipped with one axial
LVDT was mounted on the surface of the concrete via two yokes and five firmly tightened pins
as instructed by (AS_1012.17, 1997) (Figure 3-32). It is worth mentioning that since only one
LVDT is measuring the axial strain, the mechanical averaging technique has been utilized. This
means diametrically opposite attachment points mounted yokes, and a pivot rod is placed on
diametrically opposite sides of LVDT. The deformation of the sample is calculated from
(AS_1012.17, 1997):

𝑔𝑒𝑟
𝑑= Equation 3-9
𝑒𝑟 + 𝑒𝑔

Where:
d = Total deformation of the specimen through the effective gauge length (mm)
132
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

g = LVDT reading (mm)


𝑒𝑟 = Eccentricity of the rod (mm)
𝑒𝑔 = Eccentricity of the gauge (mm)
Then, the sample was placed under an axial compression test applied by a 50 MPa MaTest
loading machine.

Figure 3-32: Preparation of concrete samples for UCS test

Once the samples were ready for testing, the load was applied at the rate of 0.33 MPa/sec (20
MPa/min ± 2 MPa) (AS_1012.9, 1986). The results of the compression tests on the concrete
samples are shown in Table 3-3. The average uniaxial compression strength of samples was 55
MPa, and the average elastic modulus was 19 GPa. Also, the average strain on peak
compression load was recorded at 0.0044.

133
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Table 3-3: Results of UCS tests of concrete samples


60

50

Stress (MPa)
40
y = 19643x - 11.107
30

20

10

0
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007
Strain

UCS Statistics
Mean 55.84
Standard Error 0.46
Median 56.22
Standard Deviation 1.13
Sample Variance 1.28
Kurtosis -1.21
Skewness -0.74
Range 2.87
Minimum 54.17
Maximum 57.04
Count 6
Confidence Level (95.0%) 1.19

3.4.4 Cable Bolts

According to the experimental plans for axial and shear tests, two types of cable bolts were
selected and supplied by Jennmar Australia. The property of the cable bolts used for the tests
has been illustrated in Table 3-4. Both types of cables were supplied in two different styles of
bulbed and plain, so that the effect of bulbs could be studied. The length of the cable bolts were
selected based on the required design for shear and axial tests. Based on the design of the testing
machines, the length of the cables for double shear and axial tests were 2000 mm and 900 mm,
respectively.

134
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

Table 3-4: Properties of cable bolts used in this study


63 t Sumo 15.2mm

Property

Number of wires 9 7
Strand diameter (mm) 28 15.2
Bulb diameter (mm) 35 25
Steel area (mm2) 339 280
Mass (kg/m) 2.8 2.3
Typical yield strength (kN) 568 245
Ultimate tensile strength (kN) 635 265
Breaking load with barrel & wedge (kN) 540 -
Elongation (%) 5-7 5-7

Figure 3-33 depicts the tensile strength of 15.2 mm cable bolts reported by (Thompson &
Villaescusa, 2014). The dashed lines in the graph demonstrate the estimate elastic modulus of the
cable bolt according to the experiments. The behaviour of the bolts in the tensile test will be used for
numerical modeling purposes in the following chapters.

Figure 3-33 Tensile test result of 15.2 mm cable bolt (Thompson & Villaescusa, 2014)

The required facilities for tensile testing of the whole strands were not available as the tensile
strength of the cable bolts was beyond the power of current facilities at university. Thus, wires
of strands were tested separately. Figure 3-34 and Figure 3-35 show the results of tensile tests

135
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

of 15.2 mm and 63 t Sumo cable bolt wires. The diameter of wires in 15.2 mm and 63 t cable
bolts were 5 mm and 7 mm, respectively. The ultimate tensile load for 5 mm wires was roughly
40 kN, and the same parameter for 7 mm wire was almost 70 kN. Drops in the load in Figure
3-34 was due to testing machine malfunction, however, it did not affect the ultimate loading
capacity of the wire.
45
40
35
30
Load (kN)

25
20
15
10
5
0
0 1 2 3 4 5 6 7 8 9
Strain (%)

Figure 3-34: Tensile test results of wires of 15.2 mm cable bolt

80

70

60

50
Load (kN)

40

30

20

10

0
0 1 2 3 4 5 6 7
Strain (%)

Figure 3-35: Tensile test results of wires of 63 t Sumo cable

Summary
This chapter introduced two main testing methods for double shear testing of cable bolts
installed at angles and perpendicular to shearing planes. The mechanism of performance as

136
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

well as the preparation stages of the cylindrical double shear box (so-called MK-IV), was
illustrated. In the same manner, the angled double shear test setup was introduced. Designing
of the setup was initially inspired by the design of Grasselli (2005) for shear testing of rock
bolts. Two setups were designed to test cable bolts in 30 and 45 degrees of inclination with
respect to shear planes. Then, the preparation steps were fully explained.

The main materials used for this study include cementitious grout, mastic resin, concrete
blocks, and cable bolts. These materials were selected from the commonly used materials in
the Australian mining industry. An effort has been made to use the most recently produced
materials to keep the research up to date. Also, the same materials will be used for the axial
test in order to produce comparable results.

Cementitious grout, as one of the vital parameters influencing the behaviour of tendons,
especially under axial loading conditions, must be selected carefully. Hence, four different
types of commercialized cementitious grouts were selected for evaluation. All four types were
cement-based grout with different mixtures. Grout Type 4 was the only one with additives to
reduce the optimum curing time to 24 hours instead of 28 days. A comparison of compressive
strength revealed that grout Types 1 and 4 were significantly stronger. Due to the high
sensitivity to the curing time, grout Type 4 was selected for further analysis of the effect of
different casting and curing conditions on the compressive strength of grout. Parameters such
as curing age, water/cement ratio, moisture exposure, and curing condition were selected
parameters for sensitivity analysis. Also, the mechanical properties of grout Type 1 were
determined via laboratory tests. Finally, due to the higher flowability and compatibility of grout
Type 1 with the commonly used grout for the Australian mining industry, the above mentioned
grout was selected to be used in axial and shear tests.

In addition, two different cable bolts introduced as low-strength and high-strength products for
grout control were chosen for this study. A 15.2 mm cable bolt (Low-strength) with a maximum
tensile load of 250 kN was selected for static and dynamic perpendicular tests and static angled
shear tests. Due to the existing restrictions with the current loading machines, it was not
possible to test high-strength cable bolts in the angled shear test. Hence, it was decided to use
15.2 mm cable bolts. 63 t Sumo cable bolt was only selected for static perpendicular double
shear test in order to study the behaviour of wires in detail.

137
Chapter 3 | Laboratory Study on Static and Dynamic Shear Loading of cable bolts

A combination of shear testing rigs, static and dynamic loading machines, and selected
materials provide the opportunity for a regimented experimental plan which can
comprehensively analyze the behaviour of cable bolts under different shear loading conditions.
The results of undertaken shear experiments will be presented in the next chapter.

The novelty of this research can be briefed as follows:

• A setup for examining angle shear testing of cable bolts was employed to test the
tendons in 30 and 45 degrees of installation in addition to perpendicular shear tests.
Thus, effect of angle of installation on the behaviour of cable bolts can be examined.
• Dynamic shear testing of cable bolts has not been carried out. A combination of
currently available setup has been employed to undertake dynamic double shear testing
of cable bolts and provide the opportunity of comparing the behaviour of tendons under
different loading conditions.

138
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

Results and Analysis of Static and Dynamic Double Shear


Tests

CHAPTER 4
Results and Analysis of
Static and Dynamic Double Shear Tests

Introduction
The testing rigs and preparation stage were fully explained in the previous chapter. This chapter
mainly focuses on results and observations of laboratory studies on both static and dynamic
double shear tests. Dynamic double shear tests on cable bolts were running for the very first
time at the University of Wollongong and at the date of this report, no other similar tests on the
cable bolts have been reported.

Three different types of shear test were designed for this research. These tests were aimed to
simulate real field situations of tendons under different installation angles and different loading
conditions. It was expected to achieve the following objectives at the end of this chapter:

• Understand the behaviour of cable bolts encapsulated with grout under static and
dynamic shear loading
• Investigate the effect of the angle of installation on shear loading capacity of tendons
• Provide analysis of the failure modes of wire in tendons under different loading
conditions
Two different cable bolts with significantly different tensile strengths (250 kN and 600 kN)
were initially selected for the shear tests. There were two main reasons for this:

1. Comparing the response of low-strength and high-strength cable bolts under different
loading conditions
139
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

2. Implementing the tests with low-strength cable in cases where high-strength cable
cannot be tested due to limitations.
However, due to the capacity limitations of static loading machine for the angle shear tests
(ultimate capacity= 600 kN), it was not possible to run angled static tests. The reader is referred
to (G. Yang, 2019) for a comprehensive study on perpendicular static double shear test results.

Experimental Plan
In order to find the answers of the mentioned objectives, 23 tests were selected to be done as
per Table 4-1. The number of tests were optimized in order to produce the least amount of
constructional waste. 15.2 mm plain cable bolt was selected for the majority of the tests. Only
two of the 15.2 bulbed cable bolts were tested and results were similar to plain cable bolt,
therefor, no more test was carried out using bulbed cable bolts. Furthermore, 63 t Sumo cable
bolt was tested under static and dynamic loading conditions to investigate the behaviour of
high-strength cable bolts. Only one dynamic shear test could be undertaken using high-strength
cable bolt and it was unsuccessful due to insufficient rigidity of double shear testing rig. a trial
version of test using 63 t Sumo cables was implemented. Three different curing times of
cementitious grout was also chosen in order to study the effect of encapsulation strength on the
shear behaviour.

140
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

Table 4-1: Experimental test plan for static and dynamic shear tests
Test Test Installation Cable type Cable Grout
No. mode angle Profile UCS
Degree MPa
1 Static 90 15.2 mm Plain 10-15
2 Static 90 15.2 mm Bulbed 10-15
3 Static 90 15.2 mm Plain 45-50
4 Static 90 15.2 mm Plain 60-75
5 Static 90 15.2 mm Bulbed 60-75
6 Static 30 15.2 mm Plain 45-50
7 Static 45 15.2 mm Plain 45-50
8 Static 45 15.2 mm Plain 45-50
9 Static 45 15.2 mm Plain 45-50
10 Static 45 15.2 mm Plain 45-50
11 Static 90 63 t Sumo Plain 45-50
12 Static 90 63 t Sumo Plain 45-50
13 Static 90 63 t Sumo Plain 45-50
14 Static 90 63 t Sumo Plain 45-50
15 Static 90 63 t Sumo Plain 45-50
16 Dynamic 90 15.2 mm Plain 45-50
17 Dynamic 90 15.2 mm Plain 45-50
18 Dynamic 90 15.2 mm Plain 45-50
19 Dynamic 90 15.2 mm Plain 45-50
20 Dynamic 90 15.2 mm Plain 45-50
21 Dynamic 90 15.2 mm Plain 45-50
22 Dynamic 90 15.2 mm Plain 45-50
23 Dynamic 90 63 t Sumo Plain 45-50
Perpendicular Static Shear Test Results
In the first part of the study, cable bolts were subjected to a quasi-static shear load. In these
tests, the effect of grout strength was also investigated. For this, five tests were implemented
using grout with different curing times and strengths ranging between 15-60 MPa. Table 4-2
and Table 4-3 show the specifications and results of perpendicular double shear tests carried
out on 15.2 mm cable bolts and 63 t Sumo cable bolts, respectively. For 15.2 mm cable bolt,
the pretension load in most of the tests was in the range of 20-45 kN. Although the pretension
load for this type of cable bolt in the field situation is normally higher than 50 kN, applying
this load under laboratory conditions was technically impractical. For 63 t cable bolts,
persistence of the pretension load was totally depended on the quality of B&W. Although 150
kN of pretension load was applied on all the samples, the applied load did not change in only
two of them and the in the rest, the pretension load dropped to 20 kN. Nonetheless, 20 kN was
high enough to hold the samples tightly together. It is also worth reminding that the strength of
concrete samples is in the range of 40-55 MPa, as it has not been mentioned in the table. As

141
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

shown in Table 4-2, the maximum recorded shear load in the tests was 332 kN. Remembering
the fact that there were two shear surfaces, the shear load on each surface is half of the
maximum shear load. Thus, the shear load of the cable bolt would be 166 kN. This is equal to
roughly 66% of the maximum axial loading capacity of the cable bolt (250 kN). As shown, the
resistance of the tendon against shear load has been increased with the improvement of the
grout strength. In Table 4-2, energy represents work done to displace the cable to its maximum
displacement. Also, a maximum axial load is the maximum axial load (usually at the failure
point) recorded by the load cells. Similarly, Table 4-3 summarizes the results of static
perpendicular double shear tests for 63 t Sumo cable bolt. The highest recorded load on each
shear surface was 483 kN which is 76% of the maximum axial loading capacity of the cable (~
630 kN). As can be seen, the increase in pretension load resulted in reduction of shear load
capacity in tests 14 and 15.

Table 4-2: Static perpendicular shear test results of 15.2 mm cable bolt

No. Grout Cable Pretension Max shear Max axial Max Energy
UCS type load load load displacement
MPa kN kN kN mm kJ
1 15 Plain 21.0 246 182 49 5.7
2 15 Bulbed 20 266 195 61 7.5
3 45 Plain 43.6 332 156 53 10.0
4 60 Plain 23.0 314 227 49 8.3
5 60 Bulbed 21.0 318 230 48 8.7

Table 4-3: Static perpendicular shear test results of 63 t Sumo cable bolt

No. Grout Cable Pretension Max shear Max axial Max Energy
UCS type load load load displacement
MPa kN kN kN mm kJ
11 45 Plain 20.0 967 260 59 42.0
12 60 Plain 20.0 903 430 85 49.2
13 45 Plain 20.0 906 285 61 43.7
14 60 Plain 150.0 799 420 70 38.5
15 45 Plain 150.0 753 285 42 23.1

Figure 4-1 to Figure 4-2 show the test results of static shear tests. In each graph, the continuous
line shows the applied shear load on the middle block, and the dashed line demonstrates the
recorded axial load. When the displacement is zero, the value of axial load shows the pretension
load applied on the cable bolt. The difference between shear and axial load clarifies the load
distribution in each direction (axial or shear). The sudden drops in the load in the graphs
represents the failure of individual wires during the test. Also, Figure 4-2 is an example of the

142
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

sudden failure of all strands simultaneously. It is referred to Appendices of chapter 4 for the
detailed test results and graphs.

300

250

200
Load (kN)

150

100

50

0
0 10 20 30 40 50 60
Displacement (mm)

Axial Shear

Figure 4-1: Test 1: Static perpendicular double shear test using weak grout and Plain 15.2
mm cable bolt

350

300

250
Load (kN)

200

150

100

50

0
0 10 20 30 40 50 60
Displacement (mm)

Shear Axial

Figure 4-2: Test 3: Perpendicular double shear test using medium grout and plain 15.2 mm
cable bolt

143
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

4.3.1 Tests Observations


The concrete sample was further cut in half by a chain saw or fully broken by sledge hammer
for further analysis. Figure 4-3 demonstrates examples of cable bolts after double shear testing.
As can be seen, the cable bolt was symmetrically bent on two shear surfaces until it failed on
one side. The deformation of the cable bolt from its axis of origin was in the range of 48-61
mm.

Figure 4-3: 15.2 mm cable bolt after static shear test

Looking at the concrete lateral cut (Figure 4-4), it was demonstrated that as the cable bolt
hinges, the zone under the hinge point experiences a high compressive stress concentration.
This was also reported by (Ferrero, 1995; Ghadimi et al., 2016; X. Li et al., 2015; Liu & Li,
2017). While the grout at the hinge point was fully crushed in the samples using grout in its
early stages of curing, no trace of crush or chipping out of grout was found in the hinge zone
of the fully set grout. Hence, the strength of the grout as well as the concrete, may significantly
affect the stress distribution pattern of the cable bolt at the hinge point.

Figure 4-4: Trace of the bent cable bolt on the soft (left) and hard (right) grout and concrete
sample after the static shear test

Finally, post-test investigations of the samples revealed that the failure mechanism of the
snapped wires could vary from dominant tensile to dominant shear failure (Figure 4-5);

144
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

however, in most of the double shear tests, a combination of both modes of failure is more
common.
Test 1 Test 3 Test 4 Test 5

Figure 4-5: Failure modes of wire in perpendicular static double shear tests
Angled Static Shear Test Results
Results of the angled shear tests have been summarized in Table 4-4. Five tests were carried
out at two different inclination angles (30 and 45 degrees). It is worth mentioning that the
pretension load was initially applied to all samples; however, the pretension load was lifted
during the transportation process. In addition, tests 9 and 10 were undertaken using a load-
control loading machine, and the rest were undertaken using a displacement-control loading
machine. As can be seen, the average shear load of the inclined cable bolt was roughly 500 kN
(250 kN on each surface). This is while for the zero pretension load cable bolts, the maximum
shear load was increased to 285 kN on each shear plane. In addition, the maximum axial load
recorded in angled tests was in the range of 155 to 206 kN.

Table 4-4: Static angled shear test results

No. Installation Cable Pretension Max shear Max axial Max Energy
angle type load load load displacement
degree kN kN kN mm kJ
6 30 Plain 56 508 206 27 9.2
7 45 Plain 30 501 150 27 8.3
8 45 Plain 48 484 183 23 7.4
9 45 Plain 0 572 NM* 16 4.9
10 45 Plain 0 581 155 16 NM*
*NM = Not Measured due to instrument malfunction

Figure 4-6 demonstrates the load-displacement graph of angled shear tests. The continuous
line shows the applied shear load, while the dashed line represents the axial load recorded by
145
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

the load cells. In four out of the five tests carried out at an angle, all of the wires of the cable
bolts snapped simultaneously. In one of the 45-degree tests, multistep failure of the wires can
be seen (Figure 4-7).

600

500

400
Load (kN)

300

200

100

0
0 5 10 15 20 25 30
Displacement (mm)

Shear load Axial Load

Figure 4-6: Static 30-degree inclined double shear test using (Test 6)

600

500

400
Load (kN)

300

200

100

0
0 5 10 15 20 25 30
Displacement (mm)
Shear load Axial load

Figure 4-7: Static 45-degree inclined double shear test using (Test 8)

Figure 4-8 compares the shear loads recorded in static shear tests with cable bolts installed at
30° and 45° to the shear surface. Test results revealed that the higher the bolt angle of
inclination to the sheared joint plane, the greater the ultimate failure load of the cable bolt. In
other words, the ultimate breaking load of the cable bolt increases with the increase in the angle

146
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

of inclination. This is due to the fact that the applied shearing force tends to pull the cable wires
more axially rather than in shear with an increase in the angle of inclination. In addition, the
recorded displacement in the inclined samples is significantly lower than those installed
perpendicular to the shear surface. This means that the allowed deformation before failure in
the bolt being installed in the angle is smaller than for perpendicular samples.
Figure 4-9 depicts the recorded axial load on the cable bolts in different angled shear tests. The
axial load in the angled shear tests was in the range of 150 to 200 kN. The initial load on the
load axis shows the pretension load in each test. As discussed earlier, the pretension load on
the sample of test 10 was lost during transportation. Thus, the initial axial load in test 10 started
from zero. Comparing the pretension load and ultimate axial load, it can be concluded that
samples with higher pretension load experienced relatively higher ultimate axial load before
failure. The maximum axial load was related to 30° inclined samples, as the applied pretension
load was the highest in this test.

900

800

700
581
600
572
485 501 508
Load (kN)

500

400

300

200

100

0
0 5 10 15 20 25 30 35
Displacement (mm)

30 degree 45 degree- 1 45 degree- 2 45 degree- 3 45 degree- 4

Figure 4-8: The shear load of different angled double shear tests

147
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

250

200
Load (kN)

150

100

50

0
2 7 12 17 22 27 32 37 42
Displacement (mm)

Test 6 Test 6 Test 7 Test 8 Test 10 Test 8

Figure 4-9: Axial load of different angled double shear tests

4.4.1 Tests Observations


While the concrete blocks were internally reinforced and it was labor intensive to disintegrate
the concrete sample and take the cable bolt out; the failed surface of the cable at the shearing
planes could be visually studied. Observation of the failed strands at the shear surface in the
different tests showed that failure of the strands could occur in different modes of shear failure,
tensile failure, or a combination of both. Figure 4-10 depicts the failure mode of strands in
different inclined shear tests. The failure mode of strands in different tests was not following a
specific pattern; however, tensile failure is mainly dominant in the tests. Moreover, as the
results revealed, strands on the sides failed under tensile stress, while in two of the tests, the
upper, lower and middle strands failed in shear.

In addition, in tests 9 and 10 at 45°, the shear failure load was 572 and 581, respectively (286
and 291 per shear surface). This level of force was even greater than the tensile failure load of
the cable bolt (250 kN per cable bolt). Further investigation of the test samples revealed that
the gaps between concrete blocks were majorly filled by grout during the grouting process
(Figure 4-11). Hence, the block was cemented together, and the frictional load between the
blocks affected the results; however, test results were affected only until the bond was removed.

148
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

Test 6: 30° Test 7: 45° Test 8: 45° Test 10: 45°

Figure 4-10: Failure modes of wire in inclined static double shear tests

Figure 4-11: Shear surface being filled and cemented by leaked grout

Figure 4-12 also shows the considerable damage on the concrete surface due to high frictional
interactions between the concrete blocks. The excessive increase in forces is likely to be
attributed to; (a) the inefficiency of the two 100 mm wide single sheet Teflon strips instead of
double sheets, sandwiched between joint faces; (b) that the lateral truss reaction force was not
in line with the applied encapsulated cables force axis against the vertical shearing of the centre
block, and (c) the additional force needed to overcome the butted joints face friction. However,
this additional lateral frictional force between the concrete sides may not fully reach the
estimated 30% of the applied shear force, based on the Mohr–Coulomb Fourier series
mathematical model as reported by (Aziz, Mirza, et al., 2016) and by dynamic testing reported
by (Khaleghparast et al., 2020), but will be of a significant amount proportional to the size of
Teflon strips. Hence, it is essential to enhance the methods for removing the frictional to better
understand the role of only cable bolt in the shear test.

149
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

Figure 4-12: Damage to the concrete surface due to high frictional forces

Dynamic Shear Test Results


In the next stage, the effect of the loading rate was studied by implementing dynamic shear
tests. Table 4-5 summarizes the results of dynamic double shear on perpendicularly installed
cable bolts. In these tests, 15.2 mm cable bolts were subjected to different impact loads. There
were seven tests with different drop heights ranging between 2.0 m to 3.5 m implemented. In
order to ensure the maximum shear load on the shear planes, axial movements of the cable
were restricted by B&W.

Furthermore, each test was repeated with a reinforced B&W to study the flexible deformation
of B&W and its role in the load bearing of the cable bolts during the shear test. To achieve this,
the wedges were thoroughly welded to the cable bolt to restrain the axial movement of the
cable bolt. In Table 4-5, rows with green shading denote the tests in which no failure in the
cable bolt occurred. Results revealed that the 15.2 mm cable bolt failed when the absorbed
energy was higher than 12 kJ. In the samples with restrained B&W, absorbed energy, as well
as maximum displacements, were slightly higher than in samples with free ends. Looking at
the average applied load in dynamic tests, the applied load for dynamic tests was in the range
of 365 to 477 kN. Failure of the cable bolts generally occurred in the samples with the applied
load higher than 400 kN; however, the results of test 14 proved that the applied load could not
be the only determinative factor for failure in the dynamic tests.

150
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

Table 4-5: Dynamic perpendicular shear test results

Test Test Drop B&W Applied Absorbed Max Applied Max


No. Code Height statusenergy energy displacement shear Axial
load load
m kJ kJ mm kN kN
*
16 D1 2 free 11.8 6.2 53 394 102
17* D5 2 welded 11.8 6.7 55 398 91
18 D7 2.5 free 14.7 12.4 70 365 211
19 D8 2.5 Welded 14.7 13.2 123 403 170
20* D2 3 free 17.6 12.0 68 424 111
21 D3 3 welded 17.6 12.3 75 477 286
22 D4 3.5 welded 20.6 15.6 116 533 353
*Highlighted rows = Did not experience cable failure

Figure 4-13 and Figure 4-14 demonstrate the shear and axial loads vs. time in milliseconds for
dynamic tests. Dynamic loading was undertaken in less than 40 milliseconds. Due to the nature
of impact loading and the elastic response of the cable bolts, the middle block bounced back
up slightly. Hence, it can be seen that the displacement of the test sample reached the maximum
peak and then decreased until the sample stopped moving. Consequently, the graph of load vs.
displacement for dynamic tests does not fully represent the results of the tests. It has normally
been preferred to plot recorded data vs. time for dynamic tests (Khaleghparast et al., 2020).

The recorded shear load depicts a high peak in the first five milliseconds, which is due to the
inertia of the impacting objects. This stage does not induce any deformation or increase axial
load. The graph is later followed by load transformation from the hammer tup to the middle
block of the double shear rig. There might be several peaks during the loading stage. If there
are several peaks, their average would be considered the applied shear load (Kaewunruen,
2007). If there is only one peak after inertia, the only peak is considered the peak shear load of
the dynamic shear test (Paterson, 1978).

As both ends of the cables were equipped with load cells, the axial load on the cable during the
impact can be recorded. The dashed line in Figure 4-13 shows the changes in axial load during
the test. As a pretension load of 50 kN was applied on the cable bolts, the initial axial load
started from 50 kN. If the cable bolt did not fail (as for test 16), the axial load and displacement
met their peaks simultaneously. If the cable bolt failed during the test (such as for test 17), the
axial load dropped to zero, and the cable bolt lost its resistance against the shear force. Graphs
of the rest of the dynamic shear tests have been plotted in the appendices.
151
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

1200 80
70
1000
60

Displacement (mm)
800
50
Load (kN)

600 40
30
400
20
200
10
0 0
0 5 10 15 20 25 30 35 40
Time (ms)

Shear load Axial load Displacement

Figure 4-13: Dynamic double shear (Test 16)

Figure 4-15 compares the shear loads recorded for different dynamic tests. As can be seen, the
average peak load in all of the tests is around 400 kN. Compared to the static shear tests, the
average load in dynamic tests is slightly lower (approximately 15%). Albeit, some of the
samples in dynamic tests did not experience cable failure at all. Therefore, the parameter of
load by itself cannot represent the cable bolt’s behaviour in dynamic tests and other parameters
such as absorbed energy must be looked at.

Even though very high peak load and inertia were recorded in all the tests, they did not
necessarily experienced cable failure. This is further proof of the fact that failure of the cable
bolts in dynamic tests does not depend on the inertia stage.

The axial load recorded in the dynamic shear tests has been plotted in Figure 4-16. If the ends
of the cable were not restrained by B&W, there was a high possibility of pullout occurrence in
the cable bolt. This has also been experienced in single-shear tests by (Aziz, Rink, et al., 2017).
So, the data gained from axial load cells in this study can be considered as the applied axial
load on the cable bolt and used as the input parameters of axial loading in the study of axial
loading. According to Figure 4-16, the maximum axial load in different dynamic tests can vary
in the rough range of 100 to 200 kN. This is while the average ultimate shear load in the
dynamic test was 400 kN. This means the axial load was 25 to 50% of the shear load in the
dynamic shear tests. Comparison of the axial load in static and dynamic tests show that

152
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

contribution of axial bearing capacity of the cable bolts in dynamic tests is lower than static
tests.
900 140
800 120
700 100

Displacement (mm)
600
80
Load (kN)

500
60
400
40
300
200 20

100 0
0 -20
0 100 200 300 400 500 600
Time (mS)

Shear load Axial load Displacement

Figure 4-14: Dynamic double shear (Test 17)

1600
1400
1200
1000
Load (kN)

800
600
Average applied load
400
200
0
0 10 20 30 40 50 60
Time (ms)

Test 16 Test 17 Test 18 Test 19

Figure 4-15: Dynamic shear load in the double shear test of cable bolts using different impact
energies

153
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

250

200
Load (kN)

150

100

50

0
0 10 20 30 40 50 60
Time (ms)

Test 16 Test 17 Test 18 Test 19 Test 20

Figure 4-16: Axial load in the dynamic double shear test of cable bolts using different impact
energies

4.5.1 Tests Observations


Once the test was finished, samples were taken out of the double shear box for further studies.
Observation of the concrete samples showed some hairline cracks in the concrete samples. This
is while minor radial cracks could be seen in the samples with internal reinforcement as well
as in static tests (Figure 4-17).

Figure 4-17: Propagation of radial cracks in dynamic shear tests

In the next step, the failure mode of the strands in each test was observed. Figure 4-18 shows
the failure modes of 15.2 mm cable bolts in different dynamic shear tests. The failure modes
of the strand could be tensile with cup and cone trace, shear face with a sharp edge, or a

154
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

combination of both. In order to evaluate the effect of the applied energy on each test, the
deformation of the cable from its original axis has been measured. As shown in the photos of
the cables after the test, cables subjected to higher impact energy deformed more significantly
as expected. Moreover, the comparison of the samples with welded B&W with those with free
ends shows that restraining the B&W caused more severe damage as well as larger deformation
on the cable bolts. Also, it can be seen that the cables with free ends did not experience any
failure, although the same amount of energy was applied.

Figure 4-19 illustrates the failure modes of strands in dynamic shear tests. The failure mode of
the strands does not follow any significant pattern and both tensile and shear modes of failure
can be evenly seen in the tests. Even though dynamic tests were carried out in milliseconds of
time, and it is expected to see more shear failure in the double shear tests, tensile failure is as
frequent as shear failure.

155
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

Test 16: No failure Deformation: 20 mm

cm
Test 17: No failure Deformation: 23 mm

Test 18: 100% Shear Deformation: 58 mm

Test 19: 10% shear, 90% tensile Deformation: 57 mm

Test 20: No Failure Deformation: 31 mm


cm

Test 21: 20% shear 80% tensile Deformation: 46 mm


cm

Test 22: 50% shear, 50% Tensile Deformation: 62 mm


cm

Figure 4-18: Failure modes cable bolt after dynamic double shear tests

156
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

Test 14 Test 15 Test 17 Test 18

Figure 4-19: Failure modes of wire in dynamic double shear tests

4.5.2 Trial Dynamic Shear Test on High-Strength Cable Bolts


As part of the study, it was decided to implement a trial dynamic shear test on high strength 63
t Plain Sumo bolts. The setup was prepared using the same procedure as for previous tests, and
a pretension load of 55 kN was applied to the cable bolt. The highest possible height of the rig
(4.2 m) was selected, which could apply 24.7 kJ of energy in each impact. Figure 4-20
demonstrates the shear and axial load as well as displacement during the dynamic test. The
average shear load of 600 kN and the maximum axial load of 175 kN were recorded as a result
of the impact test. The ultimate displacement of the sample after the test was about 23 mm, and
a great bounce back of the sample can be seen in the displacement graph. The significant
difference between the maximum displacement and the ultimate displacement of the test
reveals the robust elastic reaction of the cable bolt to the dynamic load. This also can be inferred
from shear load graph, where the recorded shear load reaches zero. Due to the high elastic
response to the applied load, the hammer bounces back and load cells detach from the sample
until they hit the sample again. During this time, the shear load could not be recorded.
Once the confinements were taken away from the sample, a massive lateral failure of the middle
concrete block was observed. This failure was aligned with the gap between two cylindrical
steel confinements (Figure 4-21). Using a similar confinement system for practicing pullout
test in the laboratory, Hagan and Li also concluded that even a hairline gap between
confinements could be considered to be the weak point of the confinement system and cracks
most likely propagates in the direction of the gap (Hagan & Li, 2017). Due to the substantial
failure of the concrete, the results of this section could not be used for further analysis. Also,
no failure could be observed when the cable bolt was taken out of the concrete block (Figure
4-22).

157
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

2000 50
1800
1600 40

Displacement (mm)
1400 30
Load (kN)

1200
1000 20
800
600 10
400 0
200
0 -10
0 5 10 15 20 25
Time (ms)

Shear load Axial load Displacement

Figure 4-20: Dynamic shear test results of high-strength cable bolt (test 23)

Figure 4-21: Massive radial crack in the middle concrete block

Figure 4-22: No failure in the cable bolt after the dynamic shear test

158
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

Summary
Three different sets of experiments were carried out to analyze the behaviour of the cable bolts
under different shear loading conditions.

1. Static double shear testing with the bolts installed perpendicular to the shear surface.
2. Static double shear testing with the bolts installed 60 and 45 degrees to the shear surface
(Installation angle 30 and 45, respectively).
3. Dynamic double shear testing with the bolts installed perpendicular to the shear failure.
15.2 mm and 63 t Sumo cable bolts were selected for this set of tests due to its popularity and
ease of testing. Double shear tests were aimed to address targets such as:
• The behaviour of cable bolts encapsulated with grout under static and dynamic shear
loading
• Effect of the angle of installation on the shear loading capacity of tendons
A total number of 22 tests were carried out in different modes, including six static perpendicular
shear tests, five static angled shear tests and seven dynamic perpendicular shear tests. One
vertical load cell, two axial load cells, and two displacement lasers were utilized in order to
record the loads and displacement of the samples during the tests. Once the test was carried
out, samples were taken apart for visual investigations of the cable bolt as well as the concrete
blocks. Failure modes of the cable bolt, radial and lateral cracks of the concrete blocks,
displacement of the middle block after the test, hinge radius of the cable and other deformations
were some of the important points of observations. The following results could be inferred
from the experiments:
• Grout and concrete strength slightly affected the shear load capacity of the cable bolt.
Samples with stronger grout experienced 10% higher shear strength,
• The ultimate shear load of the cable bolt was roughly 50% higher in the angled
installation compared with those installed perpendicular to the shear surface; however,
the allowed displacement before failure in the angled shear sample was almost 40-50%
smaller than perpendicular shear tests.
• The ultimate shear load as well as the absorbed energy of the cable bolt in the dynamic
shear test, was roughly 50% higher than for the perpendicular static tests. Thus, the
weakest performing cable bolt subjected to shear testing relates to the samples installed
perpendicular to the shear surface.

159
Chapter 4 | Results and Analysis of Static and Dynamic Double Shear Tests

• Statistical studies of the chance for shear failure in different strands revealed that the
lowest strand in the shear surface and the king-wire have the highest chance of shear
failure, while the side strands usually fail in tension.
• Contribution of axial and shear loads during both static and dynamic tests were
observed. It was noted that in static perpendicular shear tests, almost 50-70% of the
applied load is acting axially. In other words, there is a great chance that the cable being
pulled out as a result of shear load. In dynamic tests, 25-50% of the applied was acting
axially which was comparatively lower than static tests. It can be inferred that due to
the short time of dynamic tests, load does not travel axially through the cable.
• The output data of the shear tests and recorded axial load can be taken to the next step
where actual pullout tests can verity the response of the cable bolt to the static and
dynamic shear load.

160
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Construction of New Static and Dynamic Pullout Test


Apparatus (SDPA)

CHAPTER 5
Construction
of
New Static and Dynamic Pullout Test Apparatus (SDPA)

Introduction
Axial behaviour of rock bolts under dynamic loading condition has already been examined;
however, similar study on cable bolts can hardly be seen in the literature. Even though a few
testing facilities have been developed in different research centers (such as Noranda
Technology Centre (2004), Royal Institute of Technology (2005), CANMET (2008), etc.)
some general drawbacks were noticed:

• Tests were primarily made on rock bolts rather than cable bolts and no further results
on cable bolt were reported.
• Only 15.2 mm cable bolt was tested in dynamic pullout tests and the mentioned cable
experienced tensile failure before being pulled out. Due to the design restrictions,
application of stronger cable bolt has not been reported.
• Steel tubes were used as confinement material which could overestimate the results
(Crompton et al., 2018; Gaudreau et al., 2004; Player et al., 2004).
• Setups were sometimes very complicated so that damages in the setup could affect the
quality of the tests (Player et al., 2004).

161
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

• Even though the developed setups were able to examine the tendons in dynamic pullout
test, comparison of the results with static or cyclic pullout tests was not possible due to
the differences in testing mechanisms.

Hence, the need for a comprehensive design to study all loading conditions was felt necessary.
Therefore, this chapter introduces the new Static and Dynamic Pullout test Apparatus (SDPA),
which has been designed and built for this research study. Different software packages such as
Solid 3D, AutoCAD, and ANSYS Workbench (Static Structural module) have been used in
order to design and analyze the SDPA. Then, the process of apparatus assembly and sample
preparation have been explained in detail. In the next section, laboratory equipment, loading
machines, and measurement tools which have been used for static and dynamic pullout testing
of tendons are introduced. After implementing the preliminary trial tests, lessons learned are
discussed, and the required modifications and improvements in the performance of the
apparatus are applied. Then, the results of secondary trial tests of the SDPA are compared and
verified with the results of the currently available pullout testing facility at UOW called Steel
Split Sets (SSS). Finally, the selection of the testing materials, including concrete, cementitious
grout, mastic resin, and cable bolt, are discussed. The physical and mechanical properties of
each material are illustrated through laboratory experiments.

Description of Static and Dynamic Pullout Test Apparatus


In general, assuming that the bond between the grout and the surrounding ground is strong
enough, the failure of the cable bolt under axial loading would possibly occur due to (Figure
5-1):

• Debonding of the cable end in the borehole (Mode 1)


• Tensile failure of the cable at the joint section (Mode 2)
• Failure of either the cable or B&W at the outer end of the anchorage system (Mode 3).

3 2 1

Figure 5-1: Possible failure modes of anchorage under axial loading condition
162
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Most of the previous designs of pullout test machines consider modes 1 and 2 in the
experiments (Hutchins et al., 1990; Hyett, Bawden, & Reichert, 1992; MacSporran, 1993),
while only a few have considered the third mode of failure (Clifford et al., 2001; Ito et al.,
2001; Weckert, 2003).

Laboratories at the University of Wollongong have been equipped with different loading
machines for quasi-static (Hydraulic servo control rams for load control and displacement
control) in different sizes, spans, and loading capacities to test a great range of specimens. In
addition, a dynamic impact loading facility is mainly used to simulate dynamic events such as
blasting.

A rig was required to be designed to fit in both static and dynamic loading machines to
implement the pullout test of tendons. Figure 5-2 depicts the schematic drawing of the idea of
the pullout test apparatus. As shown in the Figure 5-2, the idea is to use the static load or kinetic
energy to pull the embedded cable bolt out of the concrete sample. The setup is mainly made
of two sections: fixed and moving elements. According to the design, instead of pulling the
cable out of a fixed confined concrete sample, the cable bolt is fixed in its position while the
concrete sample is pushed downward. In order to hold the cable bolt, a proper set of B&W and
a steel plate is used to grip the cable firmly and transfer the load on the cable bolt to a fixed,
rigid body. The concrete sample is fully confined in a robust confinement cage and moves
downward via the applied load on the moving parts. Then the static or dynamic loads are
applied on the top plate and transferred to the concrete cage via four steel shafts.

The advantages of this design are:

1. Ability to test the cable bolts in both static and dynamic loading conditions
2. Ability to evaluate all three possible failure modes during an axial loading process.
3. Simulating the real-field situation by embedding the cable bolt in the concrete sample
with similar strength to the rock in the field.
4. Confining the concrete in all directions in order to resemble the real field confinement
situation.
Figure 5-3 schematically clarifies the setup's performance before and after applying the
load. As can be seen, once the load is applied, the concrete sample is pushed down as part
of moving elements, including loading plates, shafts, and the concrete cage. The cable bolt
is fixed in its initial position by a B&W leaning on a rigid frame. Axial elongation or tensile

163
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

failure of cable bolt, grout/concrete interface failure, and cable/grout interface are the
permitted failure modes. The fixed and rigid frame is mainly made of columns and a heavily
reinforced seat. The seat is where the bearings are placed, and the B&W rest. The bearing
guides in each shaft hole insures that they are constrained to their vertical movement. In
addition, the seat must be rigid enough to prevent any elastic deformation (bending) during
the test. Angles, channel beams, and other plates ensure the strength of the top base plate
against bending as the main load on the tendon is transferred to it. Detailed drawings of the
selected materials have been attached to the appendices related to this chapter.
Figure 5-4 shows the main components of the new pullout test rig. The body parts
mentioned in the following description are written in italics for ease of understanding. As
illustrated, the static or dynamic load can be applied downward on the loading platform
(1). As the main load is directly applied to the platform, channels, and plates are required
to reinforce the platform. The applied load is transferred to the confinement cage via four
load-transferring shafts (10) and shaft holders (5). The cylindrical shafts are made of solid
stainless steel with a 60 mm diameter. Four bearings (11) provide a smooth and frictionless
track for shafts in order to ensure vertical movement of the shafts. The confinement cage
comprises concrete sample (7), top and bottom confinement plates (4 & 8) and a radial
confinement cylinder (6). The vertical confining load is adjustable by tightening the bolts
up to 60 N.m. Embedment length of the samples can vary between 300-450 mm according
to the dimension of the setup. Previous laboratory pullout setups such as Hagan (2017)
recommended the embedment length to be at least equal or longer than the lay length of the
cable bolt (250 mm for 15.2 mm cable bolt and 280 mm for 63 t Sumo cable bolt). The
cable bolt (2) is held in place by the B&W (2) and steel plate (12). A hollow load cell is
located directly beneath the B&W plate to record the axial load. An anti-rotation tube and
plate (15&16) constrain the rotation of the cable bolt during the test (details on the
development of an anti-rotation system can be found in section 5.5.2.
The dimension of the setup has been selected so that it can fit in currently available loading
machines at UOW. Hence, the maximum height of the setup could not be higher than 1500
mm. according to this, the embedment length of the cable was designed to be no longer
than 450 mm. Figure 5-4 shows the 3D drawing of the designed rig and its vertical cross-
section view. The final dimensions of the assembled rig are 800 mm*830 mm*1400 mm.
The reader is referred to the appendix of chapter five for the detailed drawings.

164
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Fixed parts Moving parts

Figure 5-2: Simplified schematic drawing of pullout mechanism

165
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Before test After test


Figure 5-3: Test sample before and after the test

In order to ensure the rigidity of the design for static and dynamic loading conditions, the stress
and strain of the rig elements under 1500 kN static load were assessed by numerical models.
SOLIDWORKS finite element software was used to generate the models. Although the highest
expected load in the static experiments would not exceed 700 kN, the applied dynamic load
may reach up to 1200 kN, based on the dynamic to static load conversion values (Gilat et al.,
2017; Leeb, 1979). Figure 5-5 exhibits the strain of the moving parts of the SDPA under 1500
kN static load. As shown, the design is sufficiently rigid to carry the applied load without
experiencing plastic deformation. The Von-misses failure criteria were selected to evaluate the
robustness of the design under the maximum applied load (Saanouni, 2003). The highest
deformations were recorded in the centre of the loading platform and shaft holders; however,
the maximum recorded deformation was 1.61 mm, which was in the elastic range of the steel.
The stress concentration zone graph has been attached in the appendices.

166
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

1 Loading platform
2 Tendon or rock bolt
3 Barrel and Wedge (B&W)
4 Top confinement plate
5 Shaft holders
6 Radial confinement cylinder
7 Concrete sample
8 Bottom confinement plate
9 Reinforcement plates
10 Load transferring shafts
11 Bearings
12 B&W plate
13 Lifting wings
14 Hollow load cell
15 Anti-rotation plate
16 Anti-rotation confining tube
17 Fixed rigid columns
18 Reinforced ground

Figure 5-4: Introduction of pullout test setup components (top), 3D perspective view (bottom
left), and vertical cross-section (bottom right) of the pullout test rig

167
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-5: Strain of the SDPA under 1500 kN static load

Three failure modes have been numbered on the cross-sectional view (Figure 5-1). The rig is
capable of monitoring the performance of both the tendon as well as the B&W simultaneously.
In other words, all three failure modes can be assessed by the SPDA.

According to the capacity of loading machines, the embedment length of the tendon sample, ,
can vary by up to 500 mm. In this study, two different heights of 300 mm and 450 mm were
selected in order to produce comparable results to previous studies (Hagan & Li, 2017;
MacSporran, 1993; Stillborg, 1984). Due to the symmetrical load distribution and stress
concentration around the circular geometries, the cylindrical-shaped samples were selected for
testing. The diameter of the tested sample was selected to be larger than 300 mm as it was
recommended as the minimum required diameter for an undisturbed loading zone (Holden &
Hagan, 2014). The borehole size was 45 mm in all the tests to eliminate the effect of borehole
diameter on the results (J. Chen et al., 2017).

The rig can be used to investigate the following:

• Embedment length (up to 450mm)


• Host medium strength
168
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

• Encapsulation material
• Loading modes (Static, quasi-static, cyclic, dynamic)
• Rock bolts and cable bolts with different profiles
• Barrels and wedges and plates
• Borehole diameter

Rig Assembly and Experimental Procedures


In the first step, two different concrete samples (300 mm and 450 mm in height and 310 mm
in diameter) were prepared. The size of the samples was selected based on the minimum
suggested diameter by Chen et al. (J. Chen et al., 2017). Two sets of plywood frames were built
to firmly hold the concrete molds during concrete pouring and to ensure the position of the
precast central hole in the cast concrete cylinders.

Cardboard cylindrical tubes were cut to the desired height and placed firmly on the frame
(Figure 5-6). The 42-45 mm hole in the centre of the concrete block was secured by a 30 mm
cylindrical steel bar wrapped with an 8 mm clear PVC tube. The rod remained in position while
casting the concrete, and 24 hours after casting, it was hammered out of the cast concrete. When
the PVC hose was taken out of the concrete, a spiral groove remained on the surface of the
hole, which is representative of the borehole rifling as it happens after drilling in reality. In
laboratory, the intensity of rifling was increased to provide adequate frictional force between
grout and concrete so that no pullout occurs in the concrete/grout interface (Figure 5-6). All
freshly cast concrete samples were covered with moisturized burlap fabric for 28 days to reach
the optimum compression strength as per the American Concrete Institute standard for curing
concrete (308-92) (ACI Committee 308, 1980).

Concrete samples were then taken out of molds and placed on the centre of 23 mm cylindrical
steel plates and then held by clamped laterally using the radial confinement plates placed
around the concrete and tightened by six bolts and nuts up to 50 N.m (Figure 5-7). The concrete
cylinders were then confined in a rigid manner using two half-circular steel plates, clamped
together with bolts and nuts.

169
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-6: Casting concrete samples in cardboard molds (left), riffled surface of the hole
(right)

Figure 5-7: Externally confined sample by a split steel cylinder

Cable bolts were then cut into 900 mm long pieces and supplied welded on both ends by the
tendon manufacturer (Figure 5-8). Since there are different types of cable bolts in this study,
they were colour-coded and marked before use. 140 mm of the cable bolt was extruded from
the bottom end of the concrete sample. The purpose of this was to monitor the behaviour of the
cable in the bottom end and keep the pullout length constant during the test. In other words, as
the top end is pulled out, the bottom end is pulled into the concrete, so the embedded length
always remains constant.

170
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-8: Cable bolts being measured and marked

The internal hollow tube of cable bolts was grouted using cement grout prior to the next steps.
If the internal tube is not filled properly, the cable would be squeezed during the test
(Rastegarmanesh, 2022). Then, the cable was placed in its proper position in the central hole
of the concrete. The bottom end of the hole was sealed to prevent grout leakage. Then the hole
was filled with grout (cementitious or resin grout) from the top (Figure 5-9). During grouting,
the cable and concrete sample were vibrated to remove the air pockets and pack the aggregate
particles together, leading to an increase in the density and strength of the grout.

Figure 5-9: Grouting the samples and hollow tube of cable bolts

Once the grout was set, the anti-rotation tube was placed or mounted on top of the concrete
samples and filled with the grout. The anti-rotation tube is 160 mm long and 4 mm thick with
flattened sides on the opposite sides (Figure 5-10).

171
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-10: Grouting the anti-rotation tube

The top confining steel plate was placed on top of the concrete sample and connected to the
bottom confining plate by four 25mm internally-threaded steel rods (shown in Green). Eight
bolts were inserted on both ends of the rods and firmly tightened, these being vertical
confinements. The combination of vertical confinement together with the radial confinement
provided a fully-confined environment for the concrete, which should simulate the rock's real
field condition, assuming that the confinement is homogeneous. This confinement arrangement
is known in this study as the confinement cage. In order to increase the friction between
concrete samples and vertical confinement plates, a 3 mm thick rubber sheet was laid on the
concrete sample (Figure 5-11). Once the confinements are tightened and the rubber is squeezed,
constraining the concrete sample and preventing it from possible rotational movements.

Figure 5-11: Rubber sheet covering the sample (left), and confinement cage model (right)

172
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Next, four stainless steel shafts, 60 mm in diameter, were connected to the confinement cage
by steel shaft holders using sixteen M16 bolts (Figure 5-12) for push loading of the concrete
blocks in the confinement cage. Shafts and shaft holders mainly transfer the load from the
loading platform to the lower confinement cage and the concrete sample.

Figure 5-12: Installing load-transferring shafts

Figure 5-13 shows the built seat, bearings, and shafts during the adjusting process. This must
be done to ensure the shafts' smooth, perfectly vertical, and frictionless movement among the
bearings. Next, the seat is placed on the base columns and tightened by 20 bolts to ensure the
stiffness and rigidity of the rig under dynamic loading. The columns were made of four vertical
I-beams and angles connecting the base to the foundation to constrain any possible rig
movement during static and dynamic loading.

Before mounting the loading platform, the cable bolt must be fixed by B&W, and a pretension
load (if required) must be applied. Previous studies showed that cable bolts tend to twist and
unwind during laboratory pullout test (Craig & Holden, 2014; Hagan et al., 2015; D. Singh et
al., 2017). To this, an anti-rotation system is installed and tightened firmly (Figure 5-14). The
anti-rotation restrains the movement of the anti-rotation tube to only vertical displacement.
Since the tube has been grouted to the cable, it prevents the cable bolt from twisting during
loading. The anti-rotation restrains by the sidewalls of the seat through four bolts. It is then
followed by a hollow load cell and B&W on top of them (Figure 5-14). During the loading
process, the load cell records the axial load on the B&W.

173
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-13: Assembling the seat and the fixed frame

In cases where the performance of the B&W is monitored, the tension load is applied on the
tendon and B&W in order to simulate real loading conditions. A hydraulic Blue Heeler jack is
employed to apply up to 100 kN (depending on the cable type) load on the cable prior to testing
(Figure 5-15). It should be noted that since the cable has already been grouted, the tension load
should not exceed the debonding capacity of the tendon.

Anti-rotation plate Hollow load cell B&W and plate

(a) (b) (c)


Figure 5-14: (a) Setting up the anti-rotation system, (b) mounting load cell, and (c) installing
B&W

174
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-15: Applying pretension load

Next, the loading platform was placed and connected to the shafts via four shaft holders. The
loading head was internally reinforced to carry 1500 kN of load in static mode with minimised
bending or deformation. Forklift wings were welded to the loading platform for ease of lifting.
Once the sample was fully assembled, it was hoisted by a roof crane or forklift and placed
under the loading machine (see appendices of this chapter). Depending on the test type, the
loading frame can be used for static and dynamic tests.

Laboratory Testing Machines, Facilities, and Measurement Tools


In order to implement and monitor various tests, two different facilities were employed for
static and dynamic pullout test. Specific types of instrumentations were also employed for
static and dynamic (high-speed) tests. Figure 1.19 shows the details of the static loading
facilities at the University of Wollongong, Figure 5-16. The maximum loading capacity of
the machine is 600 kN (60 t). Either displacement rate or load rate can control the loading
rate. The minimum and maximum displacement rate of the machine is 0.01 mm/min and
30 mm/min, respectively. The load was recorded by an Interface Model 1052EKK-900KN
Load cell (Static error of 0.09%) (Interface force Measurements Ltd., 2022). Potentiometers
175
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

measured the machine's displacement with an error range of ±0.1 mm. Both the data
logging and controlling systems read voltage signals directly and are used to control the
test. An output of between a 0-10V signals was translated to displacement and load. Various
data loggers, including three load cells and two laser scanners, were mounted on the
machine. These were:
• Load cell No.1: Reading axial static load applied by the loading machine.
• Load cell No.2: Reading the axial load right under the B&W.
• Linear Variable Differential Transformer (LVDT) No.1 and 2: Reading the average
displacement of the Wedge relative to the barrel.
• LVDT3: Reading the displacement of the loading top relative to the loading machine
coordinate system.
• Laser No.1: Reading the vertical displacement of the concrete cage.
• Laser No.2: Reading the vertical elongation of the cable.

Figure 5-16: Static pullout testing setup

176
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

5.4.1 Displacement and Strain Measurements


Different measurement tools have controlled displacement using different scales from a
micrometer to 100 mm, including strain gauges (Figure 5-17), LVDTs (Figure 5-18), and high-
accuracy non-contact lasers (Figure 5-19). Polyester foil strain gauges in two different gauge
lengths of 10 and 20 mm (PFL-20-11-1LJC-F) have been used in order to measure the lateral
strain of steel confinement. The gauge resistance of this series is 120±0.5 Ω (TML, 2020). The
maximum recorded deformation by this measurement tool hardly exceeds 0.01 mm. Linear
transducers with a maximum stroke of 50 mm and ±0.5% accuracy have been employed to
record deformations less than 20 mm during the tests.

Figure 5-17: Strain gauges installed on specimens

Figure 5-18: LVDTs mounted to record displacement


177
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Finally, non-contact ACUITY AR550-250 laser displacement transducers measure the


displacement of the sample to the maximum range of 200 mm in both static and dynamic tests
(Figure 5-19). The AR550 high-speed laser sensor is Acuity's fastest triangulation device for
dimensional and distance measurement. The high-speed laser has a linearity of ± 0.15% and
boasts a resolution down to within 0.01% throughout the measurement range. The sampling
rate of the AR550 can be specified, and the high-speed sensor can transmit sampling rates up
to 70 kHz via Ethernet output (Acuity, 2022).

Figure 5-19: Acuity laser pointing at the bottom of the sample

5.4.2 Accelerometer
The piezoelectric accelerometers used in the experiments were Dytran Series 3200B Quarts
shock accelerometers designed to measure the mechanical shock events with an amplitude of
up to 10,000g (where g is the Earth's gravitational acceleration or about 9.81 m/s2). This series
of accelerometers are the Low Impedance Voltage Mode (LIVM) instrumentation, where the
output signals are in units of mV/g. Their mounted resonant frequency is 133 kHz, and the bias
voltage is 8.3 volts of direct current (VDC) (Elkome, 2018). The inner body is electrically
isolated from the mounting surface to eliminate the ground loop noise interfering with the
measurements.

5.4.3 High-Speed Camera


A Memrecam HX-7s high-speed camera with a maximum resolution of 2,560 x 1,920 was used
to record a maximum frame rate of 200,000 frames per second (f/s) (Figure 5-20). It was set at
2000 f/s with an exposure time of 1/2,000 s to monitor displacement as well as the mechanism
of failure (Technology, 2022).

178
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-20: High-speed camera setup

5.4.4 Dynamic Drop Hammer


Dynamic testing of the cable bolt has been executed based on the energy transformation with
the help of a drop hammer. The hammer tup has a weight of 590 kg, which can free fall from
the maximum height of 4.5 m (Figure 5-21). The hammer is held at the desired height by
hydraulic brakes, and its movement is guided via low-friction rollers and guiding columns. The
tup of the hammer is equipped with a 1200 kN dynamic load cell (Interface Model 1200) and
a piezoelectric accelerometer (Dytran Series 3200B quartz shock) (Interface force
Measurements Ltd., 2022). Once the hydraulic brakes are released, the hammer falls freely to
impact the target object. The maximum velocity of the tup hammer at the moment before
impacting would be 9 m/s. A trigger laser activates data loggers just a moment before impact
occurs. The data logger collects 1000 data signals per second. Further information on the drop
hammer is reported by Kaewunruen, (200) and Remennikov & Kaewunruen, (2008). The
maximum potential energy of the free fall is roughly 26 kJ. According to the work and energy
laws, energy is transferred to the impacted object in the form of kinetic, sound, and heat.

179
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-21: Bird-eye view (adopted from (Kaewunruen, 2007)) (left) and inside view of the
dynamic drop hammer (right)

Since the energy mechanism is based on the kinetic energy of the hammer, the possible range
1
of energy produced by the machine can be calculated as 𝑘 = 2 𝑚𝑣 2 (Kappos, 2017). Figure

5-22 shows the possible range of kinetic energy transferred with some of the last recent energy-
based testing rigs introduced in the literature review, (chapter 2). The maximum possible
energy in energy-based rigs would be transferred from the free fall of the drop hammer from
the highest possible height. According to the information on the mass of the hammer, the
maximum drop height and maximum velocities have been mentioned in chapter 2.

Figure 5-22 compares some of the energy-based dynamic pullout facilities developed in the
world. According to the weight of the hammer and maximum drop height, the maximum kinetic
energy in each pullout test can be calculated and plotted in the Figure 5-22. With an
approximate weight of 600 kg and a maximum height of 4.5 m, the impact hammer at UOW
can produce roughly 31 kJ of energy per impact. Referring to the previous successful dynamic
tests such as Yan (1992) and Ansell (2005), it can be said the produced energy by the current
facility atUOW is sufficiently large for the pullout tests.

180
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-22: Comparison of energy-based dynamic facilities for pullout test

Apparatus Modifications
During the first sets of static and dynamic trial tests, the performance of the designed rig was
monitored and compared with previous results found by Hagan et al. (2017), Aoki et al.(2002),
Li et al. (2019), and Pitrakkos et al. (2010). Some instructive lessons were learned at the end
of trial tests, which led to major and minor modifications in the rig's structure. Some of the
main changes made were as follows:

5.5.1 Concrete Confinement


The first sets of trial tests revealed that although the lateral confinement steel plates are
tightened firmly, the hairline gap between the half-cylinder plates is wide enough to cause
radial cracks in the concrete to be propagated (Figure 5-23). As soon as the concrete samples
are cracked, the resistance against the pullout force is diminished. Hence, it is crucial to ensure
that samples do not crack and maintain their rigidity until the end of the testing.

181
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-23: Radial cracks in confined samples using split steel cylinders

In relation to this, mechanical numerical models were developed in ANSYS to predict concrete
responses to the confinement. Concrete samples were modeled, and internal pressure of 9 MPa
was applied inside the borehole in a linear trend. The compressive and tensile strength of the
concrete has been considered initially at 40 MPa and 6 MPa, respectively. Three different
modes of a) Not confined, b) Externally confined, and c) Internally and externally confined
samples were considered and compared for their reactions (Figure 5-24). As can be seen, the
external confinement remarkably prevented crack propagation. In models with internal
confinement, a higher chance of crack propagation was recorded in the inner tube; however,
no crack could propagate so as to reach the outer tube. It should also be remembered that the
use of high internal pressure of 9 MPa was significantly stronger (by 1.5 times) than the
expected pressure in the actual laboratory experiments.

182
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

(a) Not confined

(b) Externally confined

(c) Internally and externally confined


Figure 5-24: Numerical simulation of the effect of confinement

Consequently, it was decided to cast the concrete directly inside the cylindrical steel pipe. The
6 mm thick steel tube was 325 mm in outer diameter and was the nearest available pipe in the
local market for use as the external confinement. Figure 5-25 shows the steel pipes mounted
on the wooden frame prior to and after concrete casting.

183
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Figure 5-25: Preparation and casting of concrete in cylindrical steel pipes

Some concrete cylinders were reinforced internally with 6 mm thick steel pipes with an outer
diameter of 165 mm. The purpose of using internal reinforcement was:

1) Reinforcing samples for stronger cable bolts with higher lateral pressure (Such as
bulbed cable bolts)
2) Installing strain gauges and burying them into the concrete in order to monitor the
lateral strain of the top and bottom of the concrete sample during the pullout test (Figure
5-26). The exact locations of the installed strain gauges have been shown in Figure 5
27.

Figure 5-26: Sticking strain gauges on the internal steel confinements

The comparison of the test results indicated a notable increase in the pullout test quality with
both internal and external confinements, majorly preventing radial cracks in concrete. Detailed
information about the test results will be discussed in the following chapter. Figure 5-28 shares

184
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

some of the pullout test results, which shows that there were no noticeable cracks in the samples
cast in steel tubes.

Figure 5-27: Location of strain gauges installed on the internal confinement

Figure 5-28: No crack in the confined samples after the pull tests

5.5.2 Anti-Rotation System


One of the most crucial issues in the pullout tests is the rotation of the cable bolt during the
loading process. This issue has been previously reported and addressed by either enlarging the
embedment length of the cable or using the anti-rotation sleeves, restraining the rotational
movements of the sleeve according to the type of experiment undertaken (Hagan & Li, 2017;
Thenevin et al., 2017). In this study, the rotation was addressed in two main parts:

185
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

1. a 160 mm long pipe being grouted to the cable and restrained from rotation
2. anti-rotation plates and bolts to hold the pipe firmly in position.
In this method, the grouted anti-rotation tube is held tight by two steel plates. The plates were
rectangles with a half-circle cut on one side of each plate. Once the plates are placed around
the tube, they are tightened to the side body of the rig by four bolts. Consequently, plates act
as a clamp around the tube to prevent rotations. In addition, two extra angles lean back to the
seat to ensure no rotation occurs on the whole anti-rotation system. This design was modified
later due to disadvantages. Figure 5-29 shows three different versions of the anti-rotation
systems designed to prevent the rotation of the cable during the loading process. Drawbacks
and modifications applied to each system have been described as well.

186
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Modification 1 Modification 2 Modification 3

Angle
s
Bolts and angles Modified bolts

Pros and Cons


• Length of the flattened • Flattened surfaces extended • Easier assembly by
surface was short and led to remove the possible removing the angles.
to damage to the tube due damage to the tube. • Fixing the issue with the
to excessive load. • Anti-rotation plates were connection of the bolts
• Anti-rotation plates were designed to be smaller and and the plates.
occupying a lot of space rotational load was to be • Less friction between the
and were hindering the transferred to the seat tube and the plates.
assembling process. through four bolts
• There was a huge destructive
momentum at the point of
bolts being connected to the
plates.
Figure 5-29: Development of an anti-rotation system

SDPA Performance Verification


Figure 8-18 plots a typical axial pullout test result carried out by the SDPA on a 15.2 mm cable
bolt. The embedment length of the cable bolt was 300 mm, and the maximum pullout load was
roughly 120 kN. Generally, each test in SDPA includes two main stages: free deformation and
187
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

debonding. In the first stage, the cable bolt elongates in its elastic stage, and B&W adjusts to
the load with increasing load. In this stage, the cable has not pulled out yet, and the applied
load was spent on elastic elongations. As the test proceeded, the cable bolt started to be
debonded with fluctuations in the load values (stage 2).
140
Stage1:
120 Allowed deformation

100
Load (kN)

80

60
Stage2:
40
Debonding and pullout in the
20 plastic phase
~17 mm ~26 mm
0
0 10 20 30 40 50
Displacement (mm)

Figure 5-30: Pullout test result of 15.2 mm plain cable in medium strength grout carried out
by SDPA

A similar test was later repeated using Steel Split Set (SSS) (Aziz, Mirzaghorbanali, et al.,
2016) apparatus to verify the performance of the SDPA (Figure 5-31). A 15.2 mm cable bolt
was pulled out by SSS. The encapsulation length of the cable was 185 mm. the maximum
pullout load was divided into the length of embedment to acquire a comparative load index.
The pullout loads index for the tests with SDPA and SSS were 0.4 kN/mm and 0.41 kN/mm,
respectively, which were in agreement. In addition, the intermittent drops in the applied pullout
load were also observed in testing with SSS apparatus.

Table 5-1 summarized the result of verification tests carried out on 15.2 mm cable bolts.
Cementitious grout with different w/c ratios was used with both testing setups. Maximum and
unit pullout loads were recorded and compared as listed in the table.

Figure 5-32 show the bar charts of pullout test as reported in Table 1-3. It is clear that the test
results with SDPA appeared to be more consistent in comparison to tests carried out using the
SSS rig, Nevertheless the average value from both tests are similar. This demonstrated the
consistency of SDPA results.

188
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

90
80
70
60
Load (kN)
50
40
30
20
10
0
0 2 4 6 8 10 12 14
Displacement (mm)

Figure 5-31: Pullout test result of plain 15.2 mm cable bolt carried out by Steel Split Set
(SSS)

Table 5-1: Pullout test result carried out by SPDA and SSS

Max Unit
Test Grout Cable Embedment
pullout pullout
Apparatus UCS bolt length
load load
MPa mm kN kN/mm
SPDA 61 15.2 mm 300 132 0.44
SPDA 63 15.2 mm 300 121 0.40
SPDA 63 15.2 mm 300 131 0.44
SPDA 85 15.2 mm 300 135 0.45
SSS 63 15.2 mm 185 81 0.44
SSS 60 15.2 mm 185 61.0 0.33
SSS 63 15.2 mm 185 87 0.47
SSS 85 15.2 mm 185 115.1 0.62

0.70
Unit Pullout Load (kN/mm)

0.60

0.50

0.40

0.30

0.20

0.10

0.00
SDPA SDPA SDPA SDPA SSS SSS SSS SSS
Test Setup

Figure 5-32: Comparison of unit pullout test results gained from SDPA and SSS
189
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

Summary
The chapter introduced a new testing technology that could implement static and dynamic
pullout tests. Static and Dynamic Pullout test Apparatus (SDPA) has been designed, fabricated,
and developed at the University of Wollongong in order to complete cable bolt and rock bolt
testing laboratory facilities. With the help of static and dynamic loading machines available at
the University, ground tendons can be subjected to different loading conditions. In the static
mode, SDPA is a double embedment setup with a 300-450 mm embedment length for the
testing side. Externally and internally confined concrete samples are used to simulate the real
field condition of pullout testing. In addition, the performance of the B&W and the effect of
the pretension load can be studied using the same setup. In the dynamic mode, SDPA is an
energy-based setup. The drop hammer facility of UOW is able to apply an approximate energy
of 26 kJ by free falling a ~600 kg weight from the maximum height of 4.5 m. It is expected
that SDPA can study the following parameters:

• Embedment length
• Host medium strength
• Encapsulation material
• Loading modes (Static, quasi-static, cyclic, dynamic)
• Rock bolts and cable bolts with different profiles
• Barrels and wedges and plates
• Borehole diameter
The SDPA setup preparation process has been fully explained in this chapter, and various stages of
development have been discussed. The apparatus has been gradually modified in order to improve the
test results compared with the previous studies. Concrete confinement methods and an anti-rotation
system were some of these modifications. Subsequent modifications and consequences of changes are
discussed.

The performance of the setup has been verified with another pullout test rig. Steel Split Set (SSS) was
used for short encapsulation and double embedment pullout testing. A 15.2 mm cable bolt was selected
for primary verification tests. The cable was grouted in both setups using the exact same grout material,
and unit pullout loads were compared. A comparison of the results revealed that the performance of the
SDPA tallied well with other recognized methods.

Also, some of the drawbacks of SDPA can be counted as:

190
Chapter 5 | Construction of New Static and Dynamic Pullout Test Apparatus (SDPA)

1. The embedment length of the cable is highly dependent on the loading capacity of the loading
machines.
2. The loading capacity of the rig has been limited to 1500 kN in static mode.
3. The confinement load cannot be controlled. Isotropic and homogenous confinement has been
assumed.

191
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Results and Analysis of Axial Static and Dynamic Pullout


tests

CHAPTER 6
Results and Analysis
of
Axial Static and Dynamic Pullout Tests

Introduction
The new Static and Dynamic Pullout test Apparatus (SDPA) testing mechanism was explained
in detail in chapter 5. Also, pullout test sample preparation steps and practical considerations
were discussed in detail. This chapter focuses on the testing and analysis of different pullout
tests carried out on different cable bolts and encapsulation grouts. Tests were carried out in two
main modes of static and dynamic loading conditions. Two types of cable bolts were used in
the study, and they are referred to as high-strength and low-strength. This allows the
comparison of large-profile cable bolts with small profiles; however, none of the low-strength
cable bolts were subjected to dynamic loads. Instead, the effect of grout age and W/C ratios
was separately studied using low-strength cable bolts and reported upon.

Experimental Program

In this research study, two sets of static and dynamic pullout tests were undertaken. In total, 36
tests were carried out using SPDA and SSS (6 trials + 30 actual tests), as listed in Table 6-1.
Different embedment lengths of 185 mm and 300 mm were chosen for the tests. Two types of
nine-wire, 63 t Sumo cable bolt (28 mm diameter and UTS = 630 kN) and 15.2 mm diameter
cable bolts (UTS: 250 kN) were chosen to represent high-strength and low-strength cable bolts,
respectively. All selected cable bolts were steel, smooth surface, and both plain and bulbed

192
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

types. Cementitious grout with three different W/C ratios were used as anchorage. As results,
three different strengths consisting of weak, medium and strong were obtained. Weak, medium
and strong is representative of UCS in the range of 10-15 MPa, 50-65, and 85-90 MPa,
respectively. Also, one type of mastic resin with 2% catalyst was used. The UCS of the resin
was in a similar range to medium cementitious grout (50-65 MPa). Cylindrical concrete
samples with a 320 mm diameter, 300 mm height, and 35-40 MPa compressive strength were
used for all the tests. Mechanical properties of the selected material were explained and
discussed in detail within Chapter 3. It should be noted that the experimental plan has deeply
considered environmental concerns regarding waste production. Hence, the number of tests has
been optimized to obtain the most meaningful data from the optimum number of tests.

193
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Table 6-1: Experimental test plan for static and dynamic pullout tests
Test No. Test mode Cable type Cable profile Encapsulation material*
1 Static 15.2 mm Plain Weak grout
2 Static 15.2 mm Plain Weak grout
3 Static 15.2 mm Plain Weak grout
4 Static 15.2 mm Plain Weak grout
5 Static 15.2 mm Bulbed Weak grout
6 Static 15.2 mm Plain Medium grout
7 Static 15.2 mm Plain Medium grout
8 Static 15.2 mm Plain Medium grout
9 Static 15.2 mm Plain Medium grout
10 Static 15.2 mm Plain Medium grout
11 Static 15.2 mm Plain Medium grout
12 Static 15.2 mm Bulbed Medium grout
13 Static 15.2 mm Plain Strong grout
14 Static 15.2 mm Plain Strong grout
15 Static 15.2 mm Bulbed Strong grout
16 Static 15.2 mm Bulbed Strong grout
17 Static 63 t Plain Medium grout
18 Static 63 t Plain Medium grout
19 Static 63 t Plain Medium grout
20 Static 63 t Bulbed Medium grout
21 Static 63 t Bulbed Medium grout
22 Static 63 t Plain Resin
23 Static 63 t Plain Resin
24 Static 63 t Bulbed Resin
25 Static 63 t Bulbed Resin
26 Dynamic 63 t Plain Medium grout
27 Dynamic 63 t Plain Medium grout
28 Dynamic 63 t Plain Medium grout
29 Dynamic 63 t Bulbed Medium grout
30 Dynamic 63 t Bulbed Medium grout
31 Dynamic 63 t Bulbed Medium grout
32 Dynamic 63 t Bulbed Medium grout
33 Dynamic 63 t Bulbed Medium grout
34 Dynamic 63 t Plain Resin
35 Dynamic 63 t Bulbed Resin
36 Dynamic 63 t Bulbed Resin
Weak grout: w/c ratio =0.43 ,UCS= 10-15 MPa,
*Grout types: Medium grout: w/c ratio =0.4 ,UCS= 50-65 MPa
Strong grout: w/c ratio =0.35 ,UCS= 85-90 MPa

194
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Static Test Results of 15.2 mm Cable Bolts


The results of static pullout tests carried out on 15.2 mm cable bolts have been discussed in
this section. These series of tests were aimed to evaluate the effect of grout strength on the
axial loading behaviour of 15.2 mm cable bolts. Hence, the tests were undertaken using three
different grout strengths containing 10-15 MPa (Weak), 50-65 MPa (Medium), and 85-90 MPa
(Strong). Table 6-2 summarizes the results of the static pullout tests.

For the samples using weak grout, especially when it was used with the Unbulbed (Plain) cable
bolt, the maximum pullout load could hardly exceed 25 kN, which is about one-tenth of the
tensile strength of the cable bolt. The cable bolt with a bulb proved to be stronger than the plain
cable when used with the weak grout (Pullout load 127.4 kN). In medium and strong grout
samples, the average pullout load for the plain bolts was 120 kN. All of the bulbed 15.2 mm
cable bolts failed in tension before being pulled out when they were used with medium and
stronger grout.
Table 6-2: Static pullout test results of 15.2 mm cable bolt

No. Code Cable Embedme Grout Max Max Unit


type nt length UCS Pullout disp Pullout
load load
mm MPa kN mm kN/mm
1 D2 Plain 185 10-15 19.4 36 0.1
2 D4 Plain 185 10-15 21.47 30.5 0.1
3 D5 Plain 185 10-15 24.25 25.7 0.1
4 D6 Plain 300 10-15 27.22 33.3 0.1
5 D5 Bulbed 300 10-15 127.4 73 0.4
6 D3 Plain 185 50-65 61.8 85.3 0.3
7 D8 Plain 185 50-65 87.3 10.4 0.5
8 D9 Plain 185 50-65 76.7 11.7 0.4
9 D8 Plain 300 50-65 131.8 14.7 0.4
10 D9 Plain 300 50-65 121.9 43.3 0.4
11 D10 Plain 300 50-65 132 69.9 0.4
12 D1 Bulbed 300 50-65 Failed 26 Failed
13 D7 Plain 185 85-90 112.9 20 0.6
14 D7 Plain 300 85-90 137.2 11.1 0.5
15 D15 Bulbed 300 85-90 Failed 57.1 Failed
16 D16 Bulbed 300 85-90 Failed 45.2 Failed

Figure 6-1 to Figure 6-4 depict the load-displacement plots of 15.2 mm cable pullout tests. As
can be seen, for the plain cable bolts encapsulated in weak grout, the maximum pullout load
occurs in a linear increase, and it was followed by a decrease in load after meeting the peak
195
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

load (Figure 6-1). The load-displacement behaviour for the plain cables encapsulated in
medium and hard grout was different. In both short and long-encapsulated samples, sudden
load drop and fluctuations in the load graph can be observed. Similar behaviour was observed
and reported by (Hagan and Li (2017); Hyett et al. (1995); Stillborg (1993). Hagan supposed
that fluctuations in the graph are due to the arrangement of the hydraulic system rather than
behaviour of the cable. He suggested that these fluctuations may have been lessened if a
hydraulic damper or accumulator had been integrated into a servo-control system (Hagan &
Li, 2017). This hypothesis was addressed in this study by using a servo-control loading
machine; however, fluctuations still existed in the results.

30

25

20
Load (kN)

15

10

0
0 5 10 15 20 25 30 35 40
Displacement (mm)

Figure 6-1: Short encapsulation- Plain- weak Strength grout

100
90
80
70
Load( kN)

60
50
40
30
20
10
0
0 2 4 6 8 10 12 14
Displacement (mm)

Figure 6-2: Short encapsulation, unbulbed, medium Strength grout

196
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

160

140

120

100
Load (kN)

80

60

40

20

0
0 10 20 30 40 50 60 70 80

Displacement (mm)

Figure 6-3: Long encapsulation, unbulbed, medium Strength grout

160

140

120

100
Load (kN)

80

60

40

20

0
0 2 4 6 8 10 12 14
Displacement (mm)

Figure 6-4: Long encapsulation, unbulbed, hard grout

Figure 6-5 compares the unit pullout load of different grouts. Unit pullout load is the maximum
pullout load over the length of the embedment and is expressed as applied load /length
(kN/mm). As can be concluded, there is no significant increase in the peak pullout load when
the grout is strengthened from medium to strong grout. Both medium and strong grouts resulted
in a pullout load four times greater than the weak grout.

197
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

0.70

Unit Pullout Load (kN/mm)


0.60

0.50

0.40

0.30

0.20

0.10

0.00

Grout type

Figure 6-5: Effect of grout type on unit pullout load

6.3.1 Tests Observations


Once the tests were completed, the interaction between the cable bolt and grout was
investigated. The concrete samples were also cut in half with the help of a chain saw (Figure
6-6). The following observations were made:

Figure 6-6: Post-test inspection of concrete samples

a) Radial cracks and stress concentration


Inspecting the concrete samples after the test showed that no radial cracking in the concrete
sample occurred. This means the lateral stress induced by axial loading did not exceed the
tensile strength of the concrete samples. However, the failure of the PVC confinement pipe as

198
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

well as radial cracks of the grout inside the anti-rotation tubes, showed the greater tendency of
the cable to unwind during the axial loading (Figure 6-7).

Figure 6-7: Cracks in the grout and PVC tube due to radial stress concentration

b) Encapsulation failure
The contact surface of the plain cable bolt and grout with different strengths was carefully
observed after the test. The grout with weaker strength had more failed ridges, and the ridges
were still attached to the cable bolt. This showed evidence of the shear failure of grout during
the test. Sharper and unbroken ridges could be easily seen in the samples with medium and
hard grout strength (Figure 6-8).

Weak grout Medium grout Hard grout


Figure 6-8: Failure surface of different grouts after the pullout test
199
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

c) Tensile failure of the cable bolt


In the bulbed cable bolts encapsulated with medium and hard strength grout, the axial load
generally resulted in tensile failure of the cable bolts rather than pulled out of the cable bolt.
The failure commonly occurred at the neck of the B&W (Figure 6-9). This type of failure is
due to the fact that pullout load of the bulbed cables was higher than their tensile failure.

Figure 6-9: Cable bolt tensile failure at the neck of B&W

Observing the cable bolt after the test showed traces of high frictional contact, especially on
the upper side of the bulb. This demonstrates the fact that the bulb resists being pulled out and
causes the tensile failure of the tendon. Figure 6-10 compares a cable bolt before and after the
test. The shiny surfaces on the bolt represent the high frictional interaction of the bolt and grout
surface. It was also noted that the bulb was filled with grout and the geometry of the bulb was
preserved, but the grout was fully crushed.

Original bolt

Bolt after pullout test

Figure 6-10: Polished surface of the cable bolt due to the high friction Cable bolt elongation

Measurements of the displacement of the bulbed cable bolt after the test showed that the pulled
out length of the cable was ~56 mm, while it was pulled by only ~30 mm (Figure 6-11). This

200
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

was the first evidence of the elongation of the cable bolt. Once the cable length was compared
with an original cable before testing, it was realised that the cable had stretched during axial
loading (Figure 6-12). In total, the elongation length of the cable was higher than the
displacement induced by the pullout load. It could be inferred that the bulb on the cable acted
as a mechanical anchor leading to the elongation of the cable in the distance between the bulb
and the B&W.

Figure 6-11: Measuring the difference of pull in and pullout length

450 mm

Original

Test 5

Test 12

Figure 6-12: Bulbed cable being elongated after the pullout test

Static Test Results of 63 t Sumo Cable Bolts


Table 6-3 summarizes the results of the experiments of 63 t cable bolts encapsulated with
cementitious grout and resin under static pull testing. The ultimate peak load represents the
highest recorded load in the pullout test, and the unit pullout load is the maximum pullout load

201
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

over the length of the embedment and is expressed as applied load /length (kN/mm). Each test
has been repeated at least twice to confirm the consistency of the results.

The overall observation from the tests revealed that bulbed Sumo cables behaved in a stronger
manner against the pullout load in comparison to plain Sumo cables, regardless of the
encapsulation grout type used. It can be postulated that the bulb increases the pullout load by
increasing the lateral resistance against the confinement. Although concrete confinement did
not experience failure in these tests, radial cracks in Hagan (2017) tests with bulb cables were
the evidence of high lateral load caused by the bulb.

Comparison of the unit pullout loads of the testing with cementitious grout and resin revealed
that the unit pullout loads in the testing with plain cable and resin was lower than 0.85 kN/mm.
In the other cases, the unit pullout load was greater than 1.0 kN/mm, and the greatest was
recorded for the bulb cables with both cement grout and resin. However, more tests are
recommended to increase the level of confidence in these statements.

Table 6-3: Static pullout test result of 63 t Sumo cable bolt encapsulated in cement grout and
resin
Test Code Cable Grout Embedment Ultimate Absorbed Max Unit
No. type type length peak load energy displacement pullout
load
mm kN kJ mm kN/mm
17 D1 Plain Cement- 300 317.9 14.5 71 1.06
18 D15 Plain based 300 330 28.7 114 1.10
19 D13 Plain 300 357.6 38.5 148 1.19
20 D2 Bulbed 300 518.7 15.4 67 1.73
21 D3 Bulbed 300 554.9 14.5 60 1.85
22 D7 Plain Resin 300 256.0 13.3 89 0.85
23 D8 Plain 300 244.8 12.3 97 0.82
24 D9 Bulbed 300 544.1 50.1 120 1.81
25 D17 Bulbed 300 484.5 43.4 118 1.62

Further evaluations of the load-displacement graphs of each test revealed different pullout
patterns for 63 t Sumo cable bolts. As shown in Figure 6-13, the pullout load of plain cable
bolts encapsulated with grout experienced an elastic increase followed by a slight drop when
the displacement was in the range of 25-45 mm. It is suggested that this be called a “Pseudo
yield point” where the deformation of the samples is not reversible. At this stage, the
deformations relate to elastic elongation of the cable bolt and permanent deformation of B&W.
By this stage. Only adhesive bond between the cable and grout has been broken and cable bolt
has not been pulled out yet. This Pseudo yield point is then followed by the gradual increase in

202
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

the applied load to reach the maximum pullout load and then decreased until the point where
the test is terminated manually.
400

350

300

250
Load (kN)

200
Test 2
150 Test 3

100

50

0
0 20 40 60 80 100 120 140 160
Displacement (mm)

Figure 6-13: Plain 63 t Sumo- Grout- Static pullout test

Figure 6-14 shows the load-displacement graph of the plain 63 t Sumo cable bolt encapsulated
by mastic resin. It was observed that, unlike grouted samples, the pullout load does not have
an increasing trend after the Pseudo yield point. Hence, it can be suggested that the axial
resistance of resin drops when the cable bolt initially debonded from the resin surface because
of the characteristics of crushing in flakes as against the brittle cement grout chip formation.
300

250

200
Load (kN)

150 Test 4
Test 5
100

50

0
0 20 40 60 80
Displacement (mm)

Figure 6-14: Plain 63 t Sumo- Resin- Static pullout test

The behaviour of the bulbed 63 t Sumo embedded in cementitious grout and mastic resin under
static axial loading is depicted in Figure 6-15 and Figure 6-16, respectively. Similar behaviour
203
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

was previously seen in plain cable encapsulated in grout as well as reported by Chen (J. Chen
et al., 2017). It means the Pseudo yield point was followed by a surge in load to reach the
ultimate pullout load and then decreased until the test was stopped manually (Figure 6-15 and
Figure 6-16). Regardless of the encapsulation material, the pullout behaviour of the bulbed
cable bolt followed a similar pattern; however, there were some fluctuations in the graph when
grout was used as the encapsulation material (Figure 6-15).
600

500

400
Load (kN)

300 Test 6
Test 7
200

100

0
0 10 20 30 40 50 60 70 80
Displacement (mm)

Figure 6-15: Bulbed 63 t Sumo- Grout- Static pullout test

600

500

400
Load (kN)

300 Test 8

200 Test 9

100

0
0 20 40 60 80 100 120 140
Displacement (mm)

Figure 6-16: Bulbed 63 t Sumo- Resin- Static pullout test

204
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Tests Observations

Once the tests finished, samples were visually assessed for noticeable physical changes during
and after the test. In addition, samples were randomly cut in half for further investigation of
the interaction between cable bolt and encapsulation material. The main observations after the
static tests are as follows:

a) Radial cracks and stress concentration


In most samples, there was no evidence of radial cracking, which could prove the importance
of uniform confinement. Only a few randomly distributed hairline cracks were observed on
some samples (Figure 6-17). The pattern of cracking was not repeated in the second repeat of
the similar test, which proves that the existence of the cracks was coincidental.

Figure 6-17: Minimal radial cracks on concrete sample

Figure 6-18 and Figure 6-19 show the recorded radial strain in the concrete block. The location
of strain gauges is schematically shown in the Figure 5-27. Top and bottom strain gauges were
mounted 50 mm away from the top and bottom of the internally confined steel tube. Figure
6-18 and Figure 6-19 demonstrate the radial strain results for plain and bulbed Sumo cable
bolts embedded with cementitious grout, respectively. As can be seen, the radial strain on the
top part of the sample is larger than the bottom side, meaning that the higher load concentration
occurred on the upper side of the block. The strain rate in the plain cable bolt was slightly lower

205
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

than the bulbed cable bolt, and the strain values in both cases were smaller than the ultimate
tensile strain of concrete. In other words, no tensile crack was expected in the concrete samples.

350 0.0006

300 0.0005

250
0.0004
Load (kN)

200

Strain
0.0003
150
0.0002
100

50 0.0001

0 0.0000
0 20 40 60 80 100 120 140
Displacement (mm)
Pullout load Bottom Strain Top strain

Figure 6-18: Radial strain of concrete sample in static pullout test of plain Sumo cable bolt

600 0.0008
0.0007
500
0.0006
400
0.0005
Load (kN)

Strain
300 0.0004
0.0003
200
0.0002
100
0.0001
0 0
0 10 20 30 40 50 60 70
Displacement (mm)
Pullout load Bottom strain Top strain

Figure 6-19: Radial strain of concrete sample in static pullout test of bulbed Sumo cable bolt

Figure 6-20 depicts the sample of bulb cable encapsulated with grout. Once the sample was cut
in half, the cone-shaped failure zone right on the top side of the bulb was observed. Although
the cone was not ruptured during the test, the micro-cracks were propagated, and the sample
fractured once the confinement load was lifted. It can be concluded from the form of the cone
failure that the area on the top side of the bulb had experienced a higher load concentration.
This statement was previously proven by the data collected from strain gauges as well.
206
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Figure 6-20: Cone-shape stress zone on the top side of the concrete sample B&W response to
the load

The increment of the axial load increases the stress concentration in the B&W. This leads to
relative displacement in the B&W. In other words, deformation of the B&W can lift the stress
concentration from the cable bolt and allow higher level energy absorption when compared to
the rock bolts. In different testing conditions, the deformation of the B&W might alter
depending on the level of generated load. Figure 6-21 exhibits the status of B&W before and
after a static pullout test. As shown, the wedge might be pulled through the barrel up to 10 mm
during the loading process. Figure 6-22 to Figure 6-24 shows the response of B&W to the axial
load in three different testing conditions. The left vertical axis shows the pullout load, and the
right axis shows the simultaneous deformation of the B&W. As can be seen, the major
deformation of the B&W occurs before the pullout reaches the Pseudo yield point; however,
deformation continued if the resistance of the cable bolt against the pullout force had an
increasing trend, as in Figure 6-22 to Figure 6-24. The failure at the B&W; was not observed
in any of the tests using the 63 t SUMO cable bolts.

207
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Figure 6-21: Status of the B&W before (left) and after (right) static pullout test

400 9
8

B&W deformation (mm)


350
300 7
6
Load (kN)

250
5
200
4
150
3
100 2
50 1
0 0
0 20 40 60 80 100 120 140 160
Displacement (mm)

Pullout load B&W deformation

Figure 6-22: Response of the B&W to the static pullout load in grouted plain cable bolt

300 7

6
B&W deformation (mm)

250
5
200
Load (kN)

4
150
3
100
2
50 1

0 0
0 10 20 30 40 50 60 70 80
Displacement (mm)

Pullout load B&W deformation

Figure 6-23: Response of the B&W to the static pullout load in resined plain cable bolt
208
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

600 14

B&W deformation (mm)


500 12

10
400
Load (kN)

8
300
6
200
4
100 2

0 0
0 20 40 60 80 100 120 140
Displacement (mm)

Pullout load B&W deformation

Figure 6-24: Response of the B&W to the static pullout load in grouted bulbed cable bolt.

Table 6-4 summarizes the recorded deformation of the B&W with different tests. As can be
seen, the most frequent deformation occurred in the bulbed cable bolts. This proves the greater
resistance of these types of bolts against the pullout load.
Table 6-4: Maximum deformation of the B&W in different loading conditions.

Test Code Cable Encapsulatio Ultimate peak B&W Maximum


No. type n material load deformation
kN mm
1 D1 Plain Cementitious 317.9 9.6
2 D15 Plain Cementitious 330.0 8.4
3 D13 Plain Cementitious 357.6 7.71
4 D7 Plain Resin 256.0 5.2
5 D8 Plain Resin 244.8 5.7
6 D2 Bulbed Cementitious 518.7 11.8
7 D3 Bulbed Cementitious 554.9 13.8
8 D9 Bulbed Resin 544.1 11.7
9 D17 Bulbed Resin 484.5 15.7
b) Encapsulation failure
Observation of encapsulation material after the experiment revealed different facts about the
failure of the encapsulation material. Figure 6-25 shows that grout ridges are more crushed on
the top side of the sample compared to the bottom side. Due to the nature of static tests, there
is enough time in the static pullout test process to allow the grout ridges to crush and turn into
a smooth powder. Hence, there are fewer sharp edges or dented surfaces on the grout surface.
Also, the black burnt surface of the grout is evidence of high frictional interaction on the surface
of the grout. This can also be interpreted from the polished surfaces of the cable, which clearly
indicate the existence of high friction between powdered grout and cable bolt wire surfaces.

209
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Observations of the resin samples showed that even though the pullout failure occurred in the
resin/cable interface, the adhesive bond between resin and concrete was weak. This was
revealed when the samples split and resin completely detached from the concrete surface.
Compared to cementitious grout, the bond between cementitious grout and concrete was
relatively stronger than resin and concrete. Furthermore, it was observed that the exposed resin
column was powdery. It was also observed that the failure in the resin/cable surface included
more crushed ridges rather than frictionally burnt surfaces (Figure 6-26).

Figure 6-25: Status of the grout surface after static pullout test of plain (left) and bulbed
(right) cable bolt.

Figure 6-26: Status of the resin surface after static pullout test of bulbed cable bolt

210
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

c) Unwinding evidence
Figure 6-27 shows the free surface of the bottom ends of the plain cable before starting the test
and 30 minutes have passed during the test. One of the wires of the cable bolt was painted white
as the movement indicator. The position of the camera was placed strategically to observe any
cable movement. Observations revealed the tendency of the cable bolt to rotate in an unwinding
direction at the free end.

Before starting the static test After 30 minutes of running the static test
Figure 6-27: Evidence of unwinding of the plain cable bolt during the static pullout test

Similar unwinding behaviour can be seen in the bulbed cables; however, there were some slight
differences in their structure. The hollow tube of the cable bolt in the plain cable was perfectly
covered by wires and welded to the wires at both ends. Hence, as the wires unwind, the hollow
tube twists accordingly. In the bulbed cable, there is a gap between wires and the hollow tube
in the bulbed zone, which is filled by grout encapsulation material and does not allow free
twisting of the hollow tube; consequently, the hollow middle tube remains bonded to the grout,
while the wires are unwound during the pullout process. Eventually, the weld between the wires
and tube breaks and is disconnected, leaving the hollow tube in place and not being pulled out
with the rest of the wires (Figure 6-28). This phenomenon is subject of further studies.

211
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Figure 6-28: Unwinding and cable failure of the bulbed cable bolt in static pullout test

Dynamic Pullout Test Results


Table 6-5 summarizes the result of dynamic tests carried out on 63 t Sumo cable bolts using
both grout and resin materials. Recordings of energy absorbed by the drop hammer show the
reaction of the cable bolt to the applied impact load. In all tests, the constant input energy of
14.7 kJ was applied by dropping a 600 kg hammer from a height of 2.5 m (this was the
maximum possible energy based on the properties of the drop hammer). The absorbed energy
and the displacement can be used in order to calculate the required energy for 1 mm pullout of
the sample (Unit energy). According to test results, the highest resistance against the impact
load belongs to bulbed cables regardless of the encapsulation material. The lowest unit energy
occurred with a combination of plain cable bolts and resin. This is noteworthy that this test
only repeated once as the poor performance of resin in pullout tests was in agreement with the
past tests reported by Pullan et al. (2018). The same amount of energy caused larger pullouts
in plain cable bolts in comparison with bulbed cable bolts. The average pullout load was
obtained from the average of summits after 20 mm of displacement.

Figure 6-29 to Figure 6-30 depicts the recorded load on the hammer during the dynamic testing
or dynamic tests. In all the following graphs, the load suddenly surges due to the inertia of the
hammer, then the hammer moves with the same speed as the sample. Hence, the load on the
hammer is almost zero but the sample and the hammer move downward simultaneously. Once
the resistance force of the sample against being pulled out becomes bigger than the downward
force of the hammer, the resistance load is recorded by the hammer. As the graph fluctuates,
the average load of resistance is measured. It is notable that inertia by itself does not cause any

212
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

movement in the sample yet. As inertia is the immediate response of the hammer to the impact
load, it does not also represent the resistance of the sample against being pulled out. However,
in the following figure, distinguishing the inertia and energy transformation stage are following
closely.
Table 6-5: Dynamic pullout test results of 63 t Sumo cable bolt
Test Code Cable type Encapsulation Absorbed Max Unit Average
No. material energy displacement energy pullout load
kJ mm kJ/mm kN
26 4 Plain Grout 8.9 69 0.13 309
27 14-1 Plain Grout 9.6 69 0.14 297
28 14-2 Plain Grout 7.1 56 0.13 294
29 10 Plain Resin 10.5 151 0.07 184
30 5 Bulbed Grout 9.9 48 0.21 465
31 6-1 Bulbed Grout 9.9 47 0.21 555
32 6-2 Bulbed Grout 10.2 47 0.22 668
33 16-1 Bulbed Grout 10.1 42 0.24 562
34 16-2 Bulbed Grout 9.6 33 0.29 614
35 12-1 Bulbed Resin 10.8 47 0.23 306
36 12-2 Bulbed Resin 10.9 47 0.23 173
1200

1000

800
Load (kN)

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
Displacement (mm)

Test 1 Test 2 Test 3

Figure 6-29: Dynamic pullout test of cement grouted plain cable bolt

In the test of plain cable encapsulated with resin, the cable bolt was fully pulled out right after
the impact. The combination of plain Sumo cable and resin showed the weakest performance
against the dynamic load among all the dynamic tests (Figure 6-30). This was expected as the
cementitious grout strength was higher than the chemical resin used.

213
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

1000
900
800
700
Load (kN)
600
500
400
300
200
100
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Displacement (mm)
Test 4

Figure 6-30: Dynamic pullout test of resined unbulbed cable bolt

In the bulbed cable bolt encapsulated with cement grout, the maximum recorded displacement
was less than 50 mm, and the average resistance load from four tests was in the range of 465-
614 kN (Figure 6-31). For the bulbed cables encapsulated with chemical resin, the maximum
displacement was similar to the tests using cement grout; however, the resistance load was in
the range of 173-306 kN (Figure 6-32). As Can be seen in load-displacement graphs of dynamic
tests (Figure 6-31, Figure 6-32 and Figures in Appendices of Chapter 6) the last part of the
graph represents the elastic response of the cable bolt to the dynamic load; hence, ultimate
displacement is lower than the maximum displacement.
1600
1400
1200
1000
Load (kN)

800
600
400
200
0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Test 9 Test 8 Test 5 Test 6

Figure 6-31: Dynamic pullout test of grouted bulbed cable bolt


214
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

1200

1000

800
Load (kN)
600

400

200

0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Test 10 Test 11

Figure 6-32: Dynamic pullout test of resined bulbed cable bolt


6.6.1 Tests Observations

Samples were examined during and after the test for physical changes. In addition, some of the
samples were randomly cut in half for further investigation of the interaction between the cable
bolt and encapsulation material. The main observations of the dynamic tests can be listed as
follows:
d) Radial cracks and stress concentration
Only some negligible radial cracks were observed in the samples of bulbed cable bolts
encapsulated by grout (Figure 6-33). The radial crack intensity decreased significantly in the
samples reinforced with internal confinement. No cracks were observed in any of the samples
tested with plain cables, therefore, the quality of confinement system was approved.

Figure 6-33: negligible radial cracks on the surface of the concrete samples
215
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Also, the recorded radial strain of the sample 50 mm away from the top and bottom of the
concrete sample and roughly 70 mm away from the centre of the sample showed insignificant
strain, which does not induce cracks in the concrete sample. The maximum recorded strain on
the top and bottom parts of the sample for plain cable was 0.0004 and 0.0002, respectively
(Figure 6-34). Compared with plain cable bolts, recorded strain in the bulbed cable was
doubled, reaching 0.00085 on the top side; however, it still does not cause cracks in the
concrete. Therefore, the results were not affected by the weakness of the confinement medium
(Figure 6-35).
1200 0.0004
0.00035
1000
0.0003
800
0.00025
Load (kN)

Strain
600 0.0002
0.00015
400
0.0001
200
0.00005
0 0
0 5 10 15 20 25 30 35
Time (ms)

Main load Strain bottom Strain top

Figure 6-34: Radial strain of concrete sample in dynamic pullout test of unbulbed Sumo cable
bolt
1600 0.001
1400 0.0009
0.0008
1200
0.0007
1000
Load (kN)

0.0006
Strain

800 0.0005
600 0.0004
0.0003
400
0.0002
200 0.0001
0 0
0 5 10 15 20 25 30
Time (ms)

Main load Strain bottom Strain top

Figure 6-35: Radial strain of concrete sample in dynamic pullout test of bulbed Sumo cable
bolt
216
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

e) B&W response to the load


As the dynamic tests were carried out in a fraction of second, LVDTs could not record the real-
time deformation of B&W during the impact loading; however, the initial and final
displacements were measured. Also, a high-speed camera was pointed at the B&W to record
the response during the test. It was found that the main displacements of the B&W occur in the
early stage of impact loading. Figure 6-36 demonstrates the status of B&W before and after the
dynamic test. No failure was observed in B&W in any of the dynamic tests of 63 t SUMO cable
bolts; however, some inward movement of the wedge in the Barrel was expected, as also shown
in Figure 6-36.

Before After
Figure 6-36: Deformation of the B&W before and after the dynamic pullout test

Evaluation of Load-Displacement graphs showed that the absorbed energy of the B&W in
dynamic tests is not significant; however, in cases where the movement of the B&W was
restricted, the final displacement of the sample was roughly 10 mm smaller (Figure 6-37). This
response was similar in both plain and bulbed cable bolts (Figure 6-38 and Figure 6-39).
Obviously, smaller displacement results in, though negligible, less energy absorption.

217
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

1200

1000

800
Load (kN)

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
Displacement (mm)

Fixed B&W Free B&W

Figure 6-37: Comparison of tests with fixed B&W versus tests with free B&W in grouted
plain cable bolt

1200

1000

800
Load (kN)

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Fixed B&W Free B&W

Figure 6-38: Comparison of tests with fixed B&W versus tests with free B&W in resined
bulbed cable bolt

218
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

1200

1000

800
Load (kN)

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Fixed B&W Free B&W

Figure 6-39: Comparison of tests with fixed B&W versus tests with free B&W in grouted
bulbed cable bolt

f) Encapsulation failure
Observations of the encapsulation material after the test showed that in dynamic testing of the
plain cable bolt, grout ridges were mostly sharp and untouched. In the case of bulbed cables,
the ridges were chipped out and dented rather than being fully crushed or ground. The evidence
of a swift load transformation of an impact loading is clearly visible in Figure 6-40.

g) Unwinding evidence
In a similar practice to static testing, a high-speed camera was set to record the reaction of the
bottom of the cable to the dynamic load. Evidence revealed that the cable end tends to rotate
in an unwinding direction during the dynamic and static tests.

With the same mechanism of static tests, it was observed that the hollow tube of the bulbed
cables was broken and detached from the rest of the cable at the point of the bulb. This has
been explained previously in static test observations (Figure 6-41).

219
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Figure 6-40: Status of the grout surface after dynamic pullout test of plain (left) and bulbed
(right) cable bolt

Figure 6-41: Evidence of unwinding and cable failure of the bulbed cable bolt in dynamic
pullout test

220
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

Summary
This chapter discussed the results of the pullout test in both static and dynamic modes. Static
and Dynamic Pullout test Apparatus (SPDA) was employed for implementing static and
dynamic tests with a 300 mm encapsulation length. Also, a Split Set setup was used for 185
mm encapsulated samples. Two different cable bolts of 15.2 mm and 63 t Sumo cable bolts
were tested; however, the 15.2 mm cable bolt was only tested statically. One-half of the cable
bolts had bulbs in a predesigned pattern and location. All of the cable bolts used in this study
were supplied by Jennmar Australia. Two types of encapsulation materials, cementitious grout,
and mastic resin were provided by Minova Australia for this research. All experiments were
aimed at finding the answer to the following issues:

• The behaviour of the encapsulated cable bolts (using both types of grout) under static
and dynamic pullout loading
• Effect of grout strength on pullout load capacity
• Comparison of cementitious grout and chemical resin as encapsulation materials in
pullout test

To achieve these goals, 36 tests were designed to compare 15.2 mm and 63 t Sumo cable bolts
under static loading conditions. Three different grout strengths and mastic resin were used in
different tests to study the behaviour of the tendons with different encapsulation materials. The
graphs of load-displacement of each test was plotted separately, and the absorbed energy of the
test was calculated. The SDPA setup was used as the main setup for the tests; however, the
performance of the mentioned setup was verified by the Split Set setup through a few static
tests. Test samples were carefully inspected after each test, and all the physical changes in
samples, including radial cracks, elongation of the cable, deformation of the B&W and rotation
of the cable bolt, were recorded.

In addition, nine dynamic tests were undertaken using a 63 t Sumo cable bolt encapsulated with
both grout and resin. The experiments were focused on the difference between grout and resin
as the encapsulation material. In addition, bulbed and plain cable bolts were compared under
dynamic loading conditions. The load-displacement graph of each test was plotted, and the
absorbed energy was calculated. The displacement of the sample due to the impact load was
monitored during the test, and the deformation of the B&W was measured manually after each

221
Chapter 6 | Results and Analysis of Axial Static and Dynamic Pullout tests

test. Also, samples were carefully examined regarding physical changes such as radial and
lateral cracks, elongation and rotation of the cable bolt and deformation of the B&W.

The findings from this experimental study demonstrated that the new pullout rig was a useful
and versatile rig for understanding and repeatedly evaluating the behaviour of both types of
cables under both static and dynamic conditions and under different loading conditions. The
reader is referred to the Appendices for detailed test results of this chapter.

222
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

Numerical Modeling for Pullout Loading of Cable Bolts

CHAPTER 7
Numerical Modeling
for
Pullout Loading of Cable Bolts

Introduction
Over the past few decades, the rapid development of efficient mathematical algorithms and the
increasing availability of powerful computers have made it possible to develop realistic
constitutive models. With the development of highly sophisticated constitutive models, the
Finite Element Method (FEM) has become an invaluable tool for predicting ultimate loads and
the associated failure mechanisms. Consequently, more capable and realistic models can be
developed to analyze and design various types of structures.

Studies by (Ansell, 2005; J. Chen et al., 2018; Delhomme & Debicki, 2010; Nemcik et al.,
2014; Nie et al., 2014; P. Singh, Spearing, Jessu, et al., 2020; P. Singh et al., 2022; P. Singh &
Spearing, 2021; St-pierre, 2007; Thompson et al., 2015) are some of the attempts of researchers
in modeling the behaviour the tendons under different loading conditions. While the mentioned
studies modeled the response of the reinforcement elements under static and dynamic loading
conditions, the encapsulation material (grout) role was not fully incorporated into the models.
Vallejos et al. (2020) emphasized the importance of the interaction between the tendon and
grout and considered frictional forces to represent the contact between them.

There are two different approaches that have been widely used for pullout test simulation: the
lumped-mass model and the dynamic deformation model. The lumped-mass models describe
the reinforcement elements as discrete masses array in series connected by springs and dampers

223
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

and have been used in models proposed since the beginning of dynamic study programs for
reinforcement elements (Vallejos et al., 2020).

Tannant et al. (1995) for modeling of rockburst for the CanMet-MMSL test facility, Player et
al. (2004) in the project of WASM dynamic testing facility, and St-Pierre for modeling of the
dynamic behaviour of cone bolt using CanMet-MMSL facility (St-Pierre, 2007) utilized a
lumped-mass approach to model the dynamic response of tendons. Lumped-mass models are
more global and aim to replicate the combined response of the tendon and the grout in the
dynamic testing facilities rather than to describe the behaviour of the internal waves in the
elements. Thus, they produce solutions involving the whole system rather than single elements.
In these models, reinforcement elements are represented by secondary elements, such as
segments connected by nodes. Models developed by (Ansell, 2005; Yi & Kaiser, 1994a) used
a dynamic deformation approach.

The dynamic deformation models are more detail-oriented and focus on resolving problems
related to specific components of a system. These models use discrete elements to describe the
behaviour of the reinforcement element as a continuous and deformable medium. In this
approach, stress and deformation waves characterize the material behaviour in the elastic and
plastic ranges through the propagation of the discretized elements in the medium (Vallejos et
al., 2020).

In this chapter, the pullout tests of cable bolts encapsulated in cementitious grout has been
modeled based on the dynamic deformation approach. The model has been generated in LS-
DYNA using the original subroutine codes of the software. LS-DYNA, as a non-linear Finite
Element program, is capable of running explicit analysis (ANSYS Inc., 1999; LS-DYNA,
2018). The main objective of this chapter is to understand the effect of different parameters on
the performance of cable bolts under axial loading conditions. This chapter aims to fill the gap
of experiments that could not be implemented, or implementation of them requires an irrational
amount of time and energy and produces unessential environmental waste.

Two different geometry models with different levels of detail have been proposed. The first
model has concentrated on the interaction between cable bolts and grout and studies the effect
of cable geometry on axial loading behaviour. High-Performance Computers (HPC) with high-
speed processors have been employed to model complicated geometries.

224
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

The second model, which is a simplified geometry, has been proposed based on the parameters
gained from the first detailed model. The aim of developing a simplified model is to decrease
the runtime and facilitate the modeling for future researchers to implement the models using
personal computers. Furthermore, simplified model provides the opportunity for generation of
different geometries with various lengths of embedment.

Proposing of the Detailed Pullout Test Model


Several steps need to be carried out for successful modeling. The modeling process commences
with the generation of the geometry followed by the next steps such as:

• Mesh generation
• Defining boundary locations and conditions, internal and external forces
• Defining the contacts and interactions
• Defining the material properties and allocating materials to the geometry
• Defining the outputs and plotting the results

The steps of the generation of numerical models are explained in this section.

7.2.1 Solver Selection


The selection of the proper solver reduces the computational time significantly. Based on the
problem type, the solver can be chosen to be explicit, implicit, or quasi-static. The main
difference between these three can be simplified to inertia dominance. In the problems in which
inertia in dominant and high velocity and acceleration exist, the explicit solver can perform
more efficiently. Quasi-static solvers are helpful for conditions where constant velocity exists
and there is an insignificant amount of kinetic energy in the system. In other words, inertia is
not dominant. In this case, an explicit solver can be manipulated to perform as a quasi-static
solver. An analysis can be considered quasi-static if (ANSYS Inc., 1999):

1
𝜔𝑒𝑥𝑐𝑖𝑡𝑎𝑡𝑖𝑜𝑛 < 𝜔𝑛𝑚𝑖𝑛 Equation 7-1
3

Where 𝜔𝑒𝑥𝑐𝑖𝑡𝑎𝑡𝑖𝑜𝑛 is excitation frequency and 𝜔𝑛𝑚𝑖𝑛 is minimum natural frequency of the model
during the test. In the case where there is no acceleration, or the velocity is constant and
relatively low, the problem can be solved by static analysis. In quasi-static analysis, each
increment must meet the equilibrium condition. Thus, it would be impractical and time-
consuming to model large events over a long time with their natural period. Hence, in order to

225
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

analyze the problem with an explicit solver without losing accuracy and in an efficient manner,
two algorithms, including 1) increasing loading rate and 2) mass scaling, must be considered
in the analysis.

Table 7-1 classifies implicit and explicit solutions based on the impact velocity and strain rate.

Table 7-1: Implicit and explicit solver boundary

Impact velocity Strain rate


Solution Effect
(m/s) (/s)
-5
Implicit <10 Static/Creep
<50 10-5-10-1 Elastic
50-1000 10-1-101 Elastic-Plastic
1000-3000 105-106 Primarily Plastic12
3000-12000 106-108 Hydrodynamic13
Explicit >12000 >108 Vaporization of colliding solids

An explicit dynamics analysis is used to determine the dynamic response of a structure in the
presence of stress waves, impacts or time-dependent loads that rapidly change over time.
Momentum exchange between moving bodies and inertial effects are usually essential aspects
of the analysis conducted. Mechanical behaviours that are highly non-linear can also be
modelled using this type of analysis. Nonlinearities may arise from materials (for instance,
hyperelasticity, plastic flows, failures), from contact (for instance, high-speed collisions and
impacts), or geometric deformation (for instance, buckling and collapse). This type of analysis
is suitable for analyzing events that occur on time scales less than one second (usually one
millisecond) (ANSYS Inc., 1999).

ANSYS is one of the most powerful finite element solvers capable of solving complex
problems on both explicit and implicit time integration platforms. The explicit dynamic solver
of ANSYS runs the problems with different time scales from linear dynamic and static to quasi-
static and transient dynamic problems. Generally, explicit dynamic toolboxes are used to
formulate equations of motion and discretize them over time. In other words, strain, stress,
displacements, and reaction forces are derived from time-dependent unknown quantities. As

12
Pressure level ≥ material strength
13
Pressure level >> material strength
226
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

the inertia in static pullout tests of cable bolts is negligible, the procedure can be simulated in
a quasi-static environment using an explicit dynamic solver (ANSYS Inc., 1999).

Equation of motion can be written as:

𝑀. 𝑢̈ (𝑡) + 𝐶. 𝑢̇ (𝑡) + 𝐾. 𝑢(𝑡) = 𝑝(𝑡) Equation 7-2

Where:
M = Mass of moving body
C = damping
K = stiffness
P is force relative to time

In static and quasi-static mode, the first two components can be omitted, and the equation turns
to Hook's law. An Explicit Dynamics analysis typically includes many different types of
nonlinearities, including large deformations, large strains, plasticity, hyperelasticity, material
failure and more.

7.2.2 Geometry Generation


The geometry of the model can be decided to be 2D or 3D according to the type of the problem.
Also, in symmetrical models, it is always preferred to minimise the size of the model by using
all the symmetric axes and generating only a part of the model. Then, the model must have the
exact same mechanical properties to represent the behaviour of the removed symmetrical
sections. Also, in large-scale models with indisputable details, the model must be split into
different sections and analyzed separately. Then, the outputs of each section would be used for
the next models.

Due to the out-of-plane rotation of wires in the cable bolts, it is not possible to consider
symmetry in the models; even though many researchers prefer to simplify the structure of the
rock bolt and cable bolts to a smooth cylinder due to difficulties of the modeling (Jalalifar &
Aziz, 2010; Tahmasebinia et al., 2018b). Nowadays, modeling technologies have been
developed considerably, and there are several different tools to generate very complex
geometries. Solid 3D and SpaceClaim were the selected software for modeling two different
cable bolts of 15.2 mm and MW9. Geometry was saved in the vendor-neutral format of IGES14.

14
Initial Graphics Exchange Specification
227
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

The cable bolt was modeled with full details, including all wires and hollow tube for MW9
Figure 7-1.

Figure 7-1: Generated geometry of the MW9 cable bolt by SpaceClaim

In the next stage, grout material was modeled around the cable bolt (Figure 7-2). Since the
purpose of the study is to understand the interaction between the cable bolt and the grout, it
was decided to replace the surrounding concrete with its reaction force; hence, concrete
material was not modeled. It results in less mesh and element and, consequently, a faster
solution. For the sake of easier meshing, grout geometry was made of three layers with 30 mm
and 45 mm diameters for the innermost and outermost layers, respectively. The geometry
model was exported to LS-Prepost for further stages.

Figure 7-2: Geometry of the grouted cable bolt

7.2.3 Mesh Generation


Meshing has been done by automatic surface meshing for the IGES file. Mesh sizes were
selected through the mesh optimization process to minimise the run time while illustrating
adequate details, especially at the contact surface between the cable bolt and the grout. Thus,
the inner surface of the grout material, which is in touch with the cable bolt, has finer mesh,
and mesh size increases as it gets closer to the outer surface. Since cable bolt was selected to
be a rigid body (details in section 7.2.7), all the meshes allocated to it are not involved in the
calculation process. Figure 7-3 exhibits a perspective view, longitudinal and cross-sectional
228
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

views of the model meshing. This detailed modelling may provide the opportunity to better
understand the factors affecting the pullout testing of cable bolts.

Figure 7-3: Perspective, Longitudinal, and cross-sectional view of the meshed pullout test
model

7.2.4 Boundary Conditions


The model represents the static pullout test conditions. Thus, transitional and rotational
constraints in all X, Y, and Z directions were applied on the outer surface of the grout nodes
(*SPC_SET). Pullout force on the cable was applied as a linear displacement on the set of
nodes at the end of the cable bolt (*Prescribed_Motion_Set). It is also essential to ensure that
the pulling end of the cable does not rotate. This represents the role of B&W in the actual
experiments. Hence, all of the transitional and rotational degrees of freedom were constrained
for the nodes at the end of the cable. Also, the hollow tube inside the cable bolt does not carry

229
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

any load and only moves with the rest of the wire and providing internal rigidity for the cable
bolt. Thus, a separate linear motion was defined for the hollow tube via
*Prescribed_Motion_Rigid keyword. Figure 7-4 schematically shows the boundary conditions.
As a rule for the rest of the figures in this chapter, the moving end of the cable is on the right
side, and the cable moves to the right to be pulled out.

Figure 7-4: Applied boundary conditions on the pullout test model

7.2.5 Contacts and Interactions


As the cable bolt and grout material are in contact, the type of contact should be illustrated.
There are many different algorithms to define the contacts in LS-DYNA. A proper contact card
must be selected based on the type of contacting objects and properties of the contact. In the
pullout test model, three types of contacts have been utilized. The wires of the cable bolt and
central hollow tube were connected by *Automatic_Surface_To_Surface contact. The main
feature of this contact is including static and dynamic friction coefficients in calculations. As
wires are tightly twisted and do not deform during the pullout test, high friction coefficients of
0.9 were selected for both static and dynamic friction modes.

As mentioned before, the grout material in the numerical models has been made of three
separate layers in order to ease the meshing process; however, it is essential to ensure that the
layers are fully bonded and there is no weakness plane in the grout. In order to fix this issue,
the *Tied_Surface_To_Surface contact card was selected. In this contact, each surface of the
master segment is tied to the nearest element surface of the slave segment.

Finally, the most crucial contact card would be the contact between the surface of the cable
bolt and the grout material. In some of the studies in the 1900s, such as In the experiments,
while setting, cementitious grout material creates a chemical bond with the steel cable bolt
(Weathersby, 2003). During the pullout test, once the chemical bond fails, the cable bolt and
grout remain in frictional contact. Depending on the profile of the cable bolt, the material type

230
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

of the cable bolt and the grout, and the roughness of the surfaces, friction coefficients may alter
(Benmokrane et al., 1995; Nemcik et al., 2014; Weathersby, 2003). Previous studies on the
rock bolt have revealed the fact that the chemical bond between the grout and cable bolt can
hardly play 10-20% of shear resistance in dynamic tests (Weathersby, 2003). It can be
concluded that chemical resistance is even less in the static tests. Hence, the effect of chemical
bonds was neglected in the quasi-static models.

Researchers have used different contact cards to model the contact between reinforcements and
concrete. Thilakaranthna (2010) used *AUTOMATIC_SURFACE_TO_SURFACE contact to
model the rebar of columns in the simulation of an impact incident. It was believed that the
mentioned contact avoids unrealistic localized stress and strain occurrence at the contact
interface (Thilakarathna et al., 2010). In the simulation of the stochastic impact of falling rock
on the reinforced concrete slab, Yan et al. (2018) used
*AUTOMATIC_NODE_TO_SURFACE contact for the rebar and concrete. Khaleghparast
(2021) used the *AUTOMATIC_SURFACE_TO_SURFACE contact card to simulate the
interaction between the rock bolt and concrete in the modeling of the dynamic double shear
test (Khaleghparast, 2021). In none of the mentioned studies the bond between reinforcement
and concrete was not the major topic of study. Also, the required bond between the
reinforcement and concrete in these studies was only to create a uniform reinforced structure,
and the models were not directly subjected to debonding load. Xu et al. (2011) modeled the
bolt pullout test of an anchor bolt embedded in concrete. In their study, it was stated that the
determination of the contact coefficients between the rebar and grout is difficult if there are no
experimental results (Xu et al., 2011). Delhomme (2010) modeled the anchor bolts of pedestals
under pullout and relaxation tests using ABAQUS. Based on his study, the interaction between
bolt and concrete can be assumed frictionless.(Delhomme & Debicki, 2010)

In this study, two contact cards were assessed to describe the interaction between the bolt and
grout. Since the mesh size and mesh type of the cable bolt and the grout were not the same size
and type, creating contacts via *AUTOMATIC_NODE_TO_SURFACE is an option.
Connecting the nodes of the slave segment (Grout) to the surfaces of the master segment (Cable
bolt) significantly increases the chance of perfect contact between the two parts. Static and
dynamic friction coefficients are the most vital parameters of the mentioned contact card. As
the second and more suitable option, *CONTACT_TIEBREAK_SURFACE_TO_SURFACE
could be selected. In this setting, the two contacting geometries can carry shear stresses up to

231
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

a certain magnitude across their interface before sliding relative to each other. This state is
known as "sticking." The model defines an equivalent shear stress at which sliding on the
geometry begins as a fraction of the contact pressure. Once the shear stress is exceeded, the
two geometries will slide relative to each other. In the next step, when the bond between the
surfaces fails, static and kinetic frictional resistances act on the surface. Hence, the contact card
is able to present debonding behaviour as well as frictional resistance. Then, the shear failure
criteria of the contact can define debonding force (LS-DYNA, 2018):

2 2
|𝜎𝑛 | |𝜎𝑠 |
( ) +( ) ≥1 Equation 7-3
𝑁𝐹𝐿𝑆 𝑆𝐹𝐿𝑆

NFLS and SFLS refer to tensile and shear failure stress, and 𝜎𝑛 and 𝜎𝑠 refer to interface normal
and shear stresses, respectively. After debonding is completed, the surface interaction is
governed by standard frictional contact constraints for normal and tangential directions. The
friction coefficient is assumed to be dependent on the relative velocity (𝑣𝑟𝑒𝑙 ) of the surfaces in
contact. There is an exponential relation between the friction coefficient and relative velocity
(Figure 7-5).

Figure 7-5: Friction coefficient is a function of relative velocity and pressure (LS-DYNA,
2018)

Therefore, the friction coefficient can be calculated as (LS-DYNA, 2018):

𝜇𝑐 = 𝜇𝑘 + (𝜇𝑠 − 𝜇𝑘 )𝑒 −𝐷𝐶|𝑣𝑟𝑒𝑙 | Equation 7-4

Where 𝜇𝑠 and 𝜇𝑘 are static and dynamic friction coefficients, and 𝐷𝐶 is the decay coefficient.
In this study, the quasi-static test had a maximum speed of 1 mm/min, which is relatively slow.
Hence, it is assumed that the friction coefficient during the test would be equal to the static

232
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

friction coefficient. Chosen contact properties for the model were found based on trial and error
and are listed in Table 7-2 (See appendix for full input parameters):

Table 7-2: Properties of contact between cable bolt and encapsulation material

Contact specifications Grout Resin


Static coefficient of friction (FS) 0.43 0.15
Dynamic coefficient of friction (FD) 0.40 0.11
Exponential decay coefficient (DC) 0.0 0.2
Tensile failure stress (NFLS) 5.0 5.0
Shear failure stress (SFLS) 50.0 70.0
7.2.6 Mass Scaling
In quasi-static analysis, each increment must meet the equilibrium condition. Thus, it would be
impractical and time-consuming to model large events over a long time with their natural
period. Hence, in order to analyze the problem with an explicit solver without losing accuracy
and in an efficient manner, mass scaling can be applied. In mass scaling, it is tried to increase
the step time in such a way that events can be done in fewer increments without affecting the
accuracy (ANSYS Inc., 1999). It should be noted that when mass scaling is applied, the velocity
values are not real; however, the displacement is actual (Wyllie & Mah, 2017). In this study,
there are some elements that require to be artificially increased in mass in order to decrease
their time step. Selective mass scaling was done by using a negative value for the DT2MS card
in the *CONTROL-TIMESTEP keyword.

7.2.7 Material Properties


Among the explicit analysis toolboxes, LS-DYNA is one of the most versatile pieces of
software due to it embracing a variety of material cards. There is a wide range of material cards
with different inputs to simulate the behaviour of different materials such as steel, concrete,
and resin. Selection of the proper material requires a comprehensive understanding of the
features of each material card. In this section, the selection of required materials for this study
has been explained.

a) Encapsulation material (grout/resin)


Similar to concrete, cement-based grout is a quasi-brittle material with a strain-softening
response after the peak load (Markovich et al., 2011; Takazawa et al., 2018). Compressive
strength (𝜎𝑐 ) of the grout is significantly higher than its tensile strength (𝜎𝑡 ) and its failure
mode is in direct relation with the internal friction angle (φ) and cohesion (c) (Aziz, Majoor, et
al., 2017; Mirza et al., 2016). Moreover, the behaviour of the cement-based materials alters
233
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

under different loading rates, especially under impact and dynamic loading. It is known that
the strength of concrete tends to be considerably higher under high strain rates (Markovich et
al., 2009).

The list of available material cards for modeling the concrete structures are (LS-DYNA, 2018):
• *MAT_005: *MAT_SOIL_AND_FOAM
• *MAT_025: *MAT_GEOLOGICAL_CAP_MODEL
• *MAT_072: *MAT_CONCRETE_DAMAGE
• *MAT_072R3: *MAT_CONCRETE_DAMAGE_REL3
• *MAT_078: *MAT_SOIL_CONCRETE
• *MAT_084: *MAT_WINFRITH_CONCRETE
• *MAT_111: *MAT_JOHNSON_HOLMQUIST_CONCRETE
• *MAT_145: *MAT_SCHWER_MURRAY_CAP_MODEL
• *MAT_159: *MAT_CONTINUOUS_SURFACE_CAP_MODEL(CSCM)
• *MAT_172: *MAT_CONCRETE_EC2
• *MAT_272: *MAT_RHT
• *MAT_273: *MAT_CONCRETE_DAMAGE_PLASTIC_MODEL

There are still many more models for the simulation of reinforced concrete that are out of the
scope of this study. Among the available material cards, different features result in different
applications. Table 7-3 compares some of the main features of the currently available card
options. The cards which cannot simulate the damage would not be suitable for the purpose of
this study.
Table 7-3: Features of the material cards for concrete (Adopted from (LS-DYNA, 2018))

SRATE FAIL16 EOS17 THERMAL18 ANISO19 DAM20 TENS21


15

*MAT_072R3 Yes Yes Yes Yes Yes


*MAT_078 Yes Yes Yes
*MAT_084 Yes Yes
*MAT_111 Yes Yes Yes Yes
*MAT_159 Yes Yes Yes Yes
*MAT_172 Yes Yes Yes
*MAT_272 Yes Yes Yes Yes
*MAT_273 Yes Yes Yes Yes

15
Strain-rate effect
16
Failure criteria
17
Equation-of-state
18
Thermal effects
19
Anisotropic/Orthotropic
20
Damage effect
21
Tension handled differently than compression in some manner
234
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

Several research has been carried out to compare different material cards to make the material
selection easier (Fan & Yuan, 2014; Fanaradelli & Rousakis, 2020; Heckötter & Sievers, 2017;
Hu et al., 2016; Salamon & Harris, 2014; Tan et al., 2014; Tu & Lu, 2009; S. Wang et al.,
2017). In their studies, Tu and Lu comprehensively assessed several widely used concrete
materials, including *MAT_00522, *MAT_01723, *MAT_02524, *MAT_07225, *MAT_07826,
and *MAT_09627 material cards (Tu & Lu, 2009). It was found that MAT_072 presented a
satisfactory response to the physical explosion test. In their comparison of different material
models such as *MAT_0058, *MAT_072R311, *MAT_15928, MAT_08429 , and *MAT_27230,
Tan simulated a uniaxial compression test of concrete cubes and cylinders. They noticed that
damage could be accumulated unrealistically fast as soon as slight movement of the material
is detected. However, this is acceptable for the scenarios where the model undergoes substantial
deformation and catastrophic failure (Tan et al., 2014). Fan and Yuan asserted that some
concrete material cards, such as *MAT_16, *MAT_25, and *MAT_111 ignore some essential
features. For instance, the third stress invariant dependence has not been included in the
mentioned models. They declared that *MAT_072R3, *MAT84, and *MAT_159
comprehensively include theories required for low-velocity impact tests (Fan & Yuan, 2014).
For this study, three material cards of *MAT_072R3, *MAT_272, and MAT_159 were
selected to be evaluated.

*MAT_072R34 is one of the comprehensive models which has been introduced and developed
in the 1990s for DYNA3D and later added to LS-Dyna. It was developed and modified (Malvar
et al., 1997, 1998; Malvar & Simons, 1996; Schwer & Malvar, 2005) mainly to analyze
concrete structures subjected to blast loading. The concrete model was initially based on the
pseudo-tensor model implemented in LS-DYNA as *MAT_016 and is now available in LS-
DYNA as *MAT_072. Through the modification of the original model, it was ensured that the
material response corresponded with that observed during the standard uniaxial, biaxial, and
triaxial tests carried out for both tension and compression loads. The material model requires

22
*SOIL/CRUSHABLE_FOAM
23
*ISOTROPIC_ELASTIC-PLASTIC
24
*KINEMATIC_HARDENING_CAP
25
*CONCRETE_DAMAGE_RELEASE3
26
*SOIL/CONCRETE
27
*BRITTLE_DAMAGE
28
*MAT_CSCM
29
*MAT_WINFRITH_CONCRETE
30
*MAT_RHT
235
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

49 input parameters which make it challenging to work with. Markovich et al. stated that the
main difficulty with the use of a realistic but complex constitutive models is the requirement
for a large number of input parameters. Obtaining the correct values for these parameters
requires uniaxial, biaxial and triaxial tests on concrete samples (in tension and compression)
that address concrete properties (Markovich et al., 2009). There is a capability to automatically
generate the parameters by entering only some basic input parameters, such as compressive
strength (Schwer & Malvar, 2005). This also was updated by some new modifications by
(Magallanes et al., 2010); however, Markovich recommended that the material card still needs
to be enhanced for better automatic parameter generation (Markovich et al., 2011).

The RHT model (*MAT_272) was introduced and developed by adding some new features to
the JH31 concrete model (Riedel, 2000; Riedel et al., 1999). The RHT constitutive model is an
advanced plasticity model for brittle materials. RHT is able to model plasticity and shear
damages caused by static as well as dynamic loading. In this model, deviatoric stress in the
material is limited by a generalized failure surface. Some of the main features of the material
are (Borrvall & Riedel, 2011):
• Porous compaction treatment
• Ultimate strength specified independently in compression, tension, and shear
• Elastic yield in the percentage of ultimate strength
• Strain rate and meridian dependence
• Damage and failure

Ji et al. stated that although the RHT model is one of the most worthy of consideration for
cement-based material, the determination of the parameters is mandatory and difficult (Ji et al.,
2005). Tu and Lu (2009) calculated the damage parameters of RHT and found that the given
parameters in AutoDyn cause unexpected strain softening (Tu & Lu, 2009). Studies (Borrvall
& Riedel, 2011) showed that the tensile properties of the concrete model in this material card
need to be improved. In their study, it was stated that an appropriate crack softening law and a
more general strain rate treatment are required for this model (Borrvall & Riedel, 2011). Wang
et al. studied the parameters of the RHT model to simulate fibre-reinforced concrete. They
concluded that the main parameters of the model could be determined according to dynamic

31
Johnson-Holmquist
236
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

and static mechanical experiments. Also, the rest of the parameters need to be checked and
adjusted by numerical simulation of relevant experiments (S. Wang et al., 2017)

*MAT159 was mainly developed to simulate the deformation and failure of concrete in
roadside safety structures impacted by vehicles. The main features of the model are (Murray et
al., 2007):
• Isotropic constitutive equations
• Three stress invariant yield surfaces with translation for pre-peak hardening
• A hardening cap that expands and contracts.
• Damage-based softening with erosion and modulus reduction.
• Rate effects for increasing strength in high-strain rate applications.

Bohara et al. (2019) compared three widely used concrete models, including *MAT_072R3,
*MAT84, and *MAT_159, under cyclic loading in cases where only limited properties are
available. It was found that both *MAT_072R3 and *MAT_159 can generate the required
parameters based on basic information such as unconfined compressive strength; However,
*MAT_159 responded more accurately in relation to softening behaviour and resulted in more
reliable force-displacement behaviour and damage pattern (Bohara et al., 2019).

b) Failure criteria for grout


*MAT_159 is a cap model with a smooth intersection between the shear yield surface and
hardening cap (Figure 7-6) (LS-DYNA, 2018).

Figure 7-6: General shape of concrete model yield surface for CSCM model in two
dimensions

237
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

In this model, the yield surface is formulated in terms of three invariants:

𝐽1 = 3𝑃

1
𝐽2′ = 𝑆 𝑆
2 𝑖𝑗 𝑖𝑗 Equation 7-5

1
𝐽3′ = 𝑆 𝑆 𝑆
3 𝑖𝑗 𝑗𝑘 𝑘𝑖
Where:
𝐽1 = The first invariant of the stress tensor
𝐽2′ = The second invariant of the deviatoric stress tensor
𝐽3′ = The third invariant of the deviatoric stress tensor

And the strength of concrete is modeled by the shear surface in the tensile and low confining
pressure regimes:

𝐹𝑓 (𝐽1 ) = 𝛼 − 𝜆 𝑒𝑥𝑝−𝛽𝐽1 + 𝜃𝐽1 Equation 7-6

Where𝛼, 𝛽, 𝜆, and 𝜃 are selected by fitting the model surface to strength measurements from
triaxial compression tests conducted on plain concrete cylinders.

Also, the damage parameter (𝑑) is used to calculate the softening function. Damage parameters
may vary from zero to one via the following equations:

0.999 1+𝐷
Brittle damage: 𝑑(𝜏𝑡 ) = [ − 1] Equation 7-7
𝐷 1 + 𝐷𝑒 −𝐴(𝜏𝑡−𝜏0𝑡 )

𝑑𝑚𝑎𝑥 1+𝐵
Ductile damage: 𝑑(𝜏𝑐 ) = [ − 1] Equation 7-8
𝐷 1 + 𝐵𝑒 −𝐴(𝜏𝑐−𝜏0𝑐)

Where A, B or C, and D set the shape of the softening curve plotted as stress-displacement or
stress-strain. Also 𝑑𝑚𝑎𝑥 is the maximum damage level than can be attained.

c) Verification of material for grout


In order to verify the selected material model, the UCS test of the cement grout cubes was
simulated. 50 mm cubic samples were generated to model grout compression tests (Figure 7-7).
Then, samples were subjected to vertical load with a defined load rate roughly equal to 0.15
MPa/sec. Table 7-4 shows the input parameters of the material cards in LS-DYNA (See
appendix for detailed input parameters). Selected units for this study are based on the consistent
238
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

system unit required for LS-DYNA. It means mass is in kilogram (kg), length is in millimeters
(mm), time is in second (s), force is in Newton (N), stress is in Mega Pascal (MPa) and Energy
is Joule (J). All the other parameters are selected accordingly. Also, in Table 7-4, Phrases in
the parenthesis are the name of parameters in LS-DYNA.

Figure 7-7: Geometry and meshing of the grout cube

Table 7-4: Input parameters of material keyword for grout and resin

Material keyword inputs Cement Grout


Material keyword *MAT_159
Mass density (RO) 2.252e-9
Unconfined compression strength (FPC) 55
Maximum aggregate size (DAGG) 8
Maximum strain increment (INCRE) 0
Rate effects options 1
Element erosion (ERODE) 1.05
Element recovery (RECOV) 0
Cap retraction option (ITRETRC) 0
Pre-existing damage (PRED) 0
Plotting options (NPLOT) 1

Figure 7-8 compares the failure mode of the grout after the numerical compression test and the
actual test results. The hourglass shape of failure in the grout was also previously seen in the
actual experiments. Also, Figure 7-9 plots the results of numerical and experimental uniaxial
compression tests. It should be mentioned that the strain rate of cement grout was verified by
using the experimental studies of Kohees (2019). As shown, the results of numerical and
experimental tests were in significant agreement; however, the post-failure calculations of the
numerical results could not perfectly model the post-peak behaviour. This cannot affect the

239
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

pullout results as in the actual tests grout parts detach from the main grout body as soon as they
fail.

Figure 7-8: Comparison of failure in the numerical model and experimental compression test

160
140
120
100
Load (kN)

80
60
40
20
0
0 0.005 0.01 0.015 0.02 0.025
Strain

Experimental Numerical

Figure 7-9: Comparison of numerical and experimental compression test results on cement
grout

d) Cable bolt
Many different material cards can be utilized to simulate steel as reinforcement buried in
concrete. In the field of dynamic loading of concrete structures, Thilakarathna (2010) modeled
the axial impact loading of reinforced concrete columns. In their study, longitudinal steel bars
and hoops were modeled as elastic, perfectly plastic material by using *MAT_00332
(Thilakarathna et al., 2010). In the study of impact load on reinforced concrete beams, Jiang et
al. (2012) recommended materials that use the Cowper-Symonds or Johnson-Cook plasticity

32
*MAT_PLASTIC_KINEMATIC
240
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

model to simulate the reinforcements (H. Jiang et al., 2012). *MAT_0031, *MAT_01533, and
*MAT_02434 are some of the suggested materials for elastic-plastic modeling of the
reinforcements. Jiang found that the models using *MAT_003 were in excellent agreement
with the experiments (H. Jiang et al., 2012). Chung et al. (2014) also utilized *MAT_024 to
simulate the steel reinforcement of bridge columns under impact loading conditions (Chung et
al., 2014).

Among all the pullout experiments carried out in this study, only one out of more than 40 tests
resulted in tensile failure of the cable bolt. Hence, this study mainly concentrates on the failure
of the bonding material rather than the plastic failure of the cable bolt. Also, as the hollow tube
inside the cable was filled with grout, it was assumed that the hollow tube was fully rigid. Two
material models, including *MAT_020 for modeling the rigid hollow tube and *MAT_001 for
modeling the wires as isotropic hypoelastic materials, were selected. For bodies defined to be
rigid, the only input parameter would be density. Then, the mass, inertia, and reaction forces
on the rigid body are calculated on a single node in the centre of the mass of the rigid body.
Material properties of elastic wires have also been listed in Table 7-5. No failure or plastic
deformation is expected in this material, and the material is significantly stiffer than the
encapsulation material. Therefore, the runtime would be significantly reduced.
Table 7-5: Input parameters of elastic material for steel

Density (tonne/mm3) Elastic modulus (MPa) Poison ratio


7.8e-9 2.0e5 0.29
7.2.8 Outputs
There are several outputs for the model which need to be considered. The main parameter to
be compared with experiments is the axial load on the cable. This was done by measuring the
load on each 50 mm cable bolt (Figure 7-10). This will present the load transformation
mechanism in the cable bolt. Measurements of the axial load were carried out by
*DATABASE_CROSS_SECTION_PLANE and reported in the format of an ASCII file.

It is also essential to ensure that simulations remain in the range of quasi-static. For this, the
total kinetic energy of the models needs to be measured. Obviously, the recorded energy does
not represent the actually recorded energy of the experiments due to the simplifications.

33
*MAT_JOHNSON_COOK
34
*MAT_PIECEWISE_LINEAR_PLASTICITY
241
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

Nonetheless, the kinetic energy of the model should not exceed 5-10% of the total internal
energy to be considered a quasi-static solution (LS-DYNA, 2018). This output was recorded
as *glsat and *matsum ASCII files.

L250

L150
L400

L350

L300

L200

L100

L50

L0
Figure 7-10: Probes for measurement of the axial load

Furthermore, stress and strain distribution on the grout surface as well as plastic failure zones
were studied through Prepost software. Axial stress, lateral stress, effective plastic stress, and
elastic elongation of the cable bolt were measured via colour contour plots.

Numerical Results
This section presents the results of the simulated cable bolt under static pullout loading
conditions. For verification of the models, results of the 63 t Sumo plain cable pullout test were
used (Figure 7-11). As can be seen, the debonding load and maximum pullout load were almost
accurately estimated by the developed model; however, it should be recalled that debonding in
the numerical model occurred quicker (less displacement before failure) due to the rigidity and
simplicity of the numerical model. Furthermore, the numerical model has not considered the
deformation of the B&W, components of the pullout test facility and other unforeseen
deformations.
Also, Figure 7-12 presents the ratio of total kinetic energy to total internal energy. As shown,
the kinetic energy is significantly lower than the total energy, which classifies the test as quasi-
static.
Figure 7-13 portrays a lateral section of the grout material. This Figure 7-13 concentrates on
recorded plastic grout failure by the Iso-surface technique in LS-DYNA. In iso-surface mode,
the colour contour of stress intensity on the surface is plotted, and the rest of the areas are
hidden. As can be seen, the edges of the grout experienced severe damage while the surface
remained untouched.

242
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

400

350

300

250
Load (kN)

200

150

100

50

0
0 20 40 60 80 100 120 140 160
Displacement (mm)

Numerical Experiment 1 Experiment 2

Figure 7-11: Comparison of numerical and experimental results for grout and 63 t Sumo

3.0E-03
Kinetic energy/Internal energy

2.5E-03

2.0E-03

1.5E-03

1.0E-03

5.0E-04

0.0E+00
0 5000 10000 15000 20000
Test interval

Figure 7-12: Ratio of kinetic energy/total internal energy of the simulations

Figure 7-13: Iso-surface image of plastic zones in grout

243
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

Figure 7-14 compares the failure mode of the grout in the numerical and experimental tests.
The image of the experiment has been produced by 3D scanning and image digitizing
techniques. Both experiment and numerical simulation reveal the fact that stress concentration
has been on the grout edges.

Figure 7-14: Comparison of grout failure in the numerical and experimental tests

The distribution of the axial stress in the cable bolt illustrates that the pulling side of the cable
was experiencing a higher tensile load rather than the bottom end. Figure 7-15 Plots the stress
concentration on different lengths of the cable bolt. In the graph, positive stress represents
tensile stress. Stress level decreases in a roughly linear manner from the top to the bottom end
of the cable bolt.Also, as observed in the experiments, the free bottom end of the cable tends
to rotate and move along the grout surface rather than shear at the grout edges. This was also
observed in the numerical simulations. As shown in Figure 7-16, the marked wire on the cable
bolt was displaced rotationally as the test progressed.

244
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

Figure 7-15: Stress distribution in the cable bolt during the test

Figure 7-16: Rotational displacement of the bottom end of the cable

Proposing a Simplified Pullout Test Model


Since the application of numerical models should be time-efficient and cost-effective, effort
should be made to generate more efficient and accessible models. The actual runtime of the
currently generated model was 1 hour and 10 minutes with system configuration of eigh CPU
cores, 128 GB RAM, and allocating 109 words of memory in a single precision analysis. With
ordinary computers of the day, the maximum number of available CPUs is limited to six and
8-32 GB RAM. Hence, the current model is not easily implemented using these ordinary
computers. Thus, it was decided to propose a new simplified model using the obtained
parameters from the previous modelling stage. There are several advantages and disadvantages
of simplifying the model, which are listed below:
Advantages:
245
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

1. Faster runtime. No need for high-performance computers.


2. Easy and uniform meshing
3. No erosion in materials is required.
4. Parametric studies can be easily done.
5. Sensitivity analysis can be carried out at a reliable level.
Disadvantages:
1. The physical interaction of grout edges and the cable bolt is neglected.
2. The effect of cable geometry is neglected
3. Possible deformations and rotations of the cable bolt cannot be studied.
4. Lateral and radial forces are underestimated.
5. Grout damages and plastic failures are ignored.
For this, the geometry of the model was simplified to a solid cylinder encapsulated in a column
of cementitious grout (Figure 7-17). Due to the existence of the symmetrical surfaces along the
axis line of the cable, only a quarter of the model has been generated. The same materials have
been used for the proposed model; however, the contact card has been reviewed and
recalibrated according to the mesh size and number of nodes.
Input parameters of the contact between the cable bolt and the grout35 have been recalibrated
and verified with the previous model as well as the experiments (Table 7-6).

Figure 7-18 compares the results of the simplified model with the detailed model and the
experiments. As can be seen, there is an acceptable agreement with the peak pseudo-yield point
and the maximum pullout load of the numerical models and the experiments. There are some
significant fluctuations in the simplified model due to coarse mesh size. However, it is not
affecting the solution and results.
Also, a comparison of the effective plastic strain of the detailed and simplified model shows
that the plastic failure of the grout material has been closely estimated by the simplified model
(Figure 7-19), albeit not as accurate as the detailed model and with a small amount of
overestimation.
The calibration of the model was carried out by trial and error and comparison of different
models. Once the model was calibrated, three different embedment lengths of 300, 450, and
600 mm were modeled. These models were used for the sensitivity analysis.

35
*CONTACT_TIEBREAK_SURFACE_TO_SURFACE
246
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

Figure 7-17: Simplified model of cable bolt

Table 7-6: Input parameters of contact keyword for grout


Static coefficient of friction (FS) 0.37
Dynamic coefficient of friction (FD) 0.37
Exponential decay coefficient (DC) 0.0
Tensile failure stress (NFLS) 5.0
Shear failure stress (SFLS) 28.0
450
400
350
300
Load (kN)

250
200
150
100
50
0
0 20 40 60 80 100 120 140
Displacement (mm)

Numerical model Simplified model

Figure 7-18: Comparison of a simplified numerical model with experiments and the detailed
model

247
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

Figure 7-19: Comparison of the effective plastic strain of the simplified model and the
detailed model

More details of the numerical simulation and results can be found in the Appendix A.4.
Sensitivity analysis and parametric study carried out using numerical models and detailed
results presented in chapter 8.

248
Chapter 7 | Numerical Modeling for Pullout Loading of Cable Bolts

Summary
This chapter introduces a new numerical simulation based on the dynamic deformation
approach. In this model, an attempt was made to investigate the physical interaction between
the cable bolt and the grout during the axial loading. For this, a finite element method model
was generated using LS-DYNA, and explicit solvers were used to solve the problem.

Two different geometries, including a detailed cable bolt model and a simplified model, were
proposed for different studies. A detailed model was proposed to study the effect of the
geometry of the cable on axial behaviour. The model was verified using actual experimental
results. Later, the simplified model was proposed to increase the runtime and facilitate the
modeling for future studies based on the empirical parameters acquired from the detailed
proposed model. Once both models were calibrated and verified, a sensitivity analysis of
different parameters was carried out. The effect of grout strength, friction coefficients, shear
resistance and embedment length were the selected parameters of this study.

Results revealed that the behaviour of the cable bolt during the pullout test included two stages:

1. The load rapidly increased to the point where the shear resistance between the cable
bolt and the surface of the grout was surrendered. It is usually followed by a sudden
drop in the load (as in Figure 7-11).
2. Frictional interaction between the cable bolt and the grout causes resistance against the
pullout load. In highly-frictional interactions, the load may increase even higher than
the peak load in the first phase. In low-frictional contacts, a constant resistance was
produced, which was significantly lower than the peak in the first phase.

249
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Discussion on Axial and Shear Loading of Cable Bolts

CHAPTER 8
Discussion
on
Axial and Shear Loading of Cable Bolts

Introduction
In the previous chapters, cable bolts were subjected to different loading conditions, including
double shear loading perpendicular to the bolt axis installed across the sheared joint surface,
the angled shear load at 30 and 45 degrees of inclination to the shear surface and finally, the
axial loading for a pullout situation of the cable bolt. Two different loading rates at quasi-static
and dynamic tests were applied perpendicular to the axially laid cable in the double shear test
and pullout tests to investigate the effect of the rate of loading on the performance of tendons.
Finally, numerical models were used to study the influence of various parameters, such as grout
used for cable encapsulation and the host medium.

This chapter is divided into two main sections: Firstly, the shear test failure mechanism and the
parametric study, and secondly the failure mode and the sensitivity analysis of the pullout test.

A parametric study has been carried out to evaluate the effect of different parameters on the
behaviour of tendons under shear and axial loading conditions. The main purpose of the
parametric study and sensitivity analysis was to evaluate the effect of various parameters on
the behaviour of tendons; however, implementing several experiments was not always efficient
for this purpose. Hence, in some cases, numerical solutions were alternatively employed to
reduce the number of experiments.

250
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Results of shear and axial tests in previous chapters revealed that in shear tests, cable bolts
were subjected to failure. Whereas in axial tests, the bond between the cable bolt and
encapsulation material failed prior to the cable failure. Hence, the failure mechanism of steel
must be considered for shear testing, and the failure mechanism of cementitious material was
studied for pullout tests.

Shear Tests Analysis


Several shear tests carried out at different angles, and with different loading rates revealed the
failure mechanism of cable bolts under the following conditions.

8.2.1 Failure Mechanism of Wires in a Shear Test


Figure 8-1 shows the tensile mode of wires in different static shear tests at different angles.
Different colours show different modes of failure. Failure of wires (if it occurred) may be in
three modes: tensile, pure shear, pure tensile, or a combination of both. Also, wires were drawn
in the same order as they appeared on the shear surface. A simplified numerical model of rock
bolts under shear loading explained by Jalalifar (2006) showed similar load distribution in rock
bolts so that the tensile and the compression loads were expected to occur on the top and the
bottom of the hinge, respectively.

Perpendicular shear test experiments indicated that in most cases, the top strand wire failed in
tension, and the bottom strand wire experienced shear failure. No specific pattern of failure
was observed on the sides and, in the king Wire (Figure 8-1). Yang et al. (2019) suggested that
each wire could be assumed as a single rock bolt and assessed individually. Their tests on
different high-strength cable bolts could not show any specific pattern of failure.

For cable bolts installed at an angle, there was no pattern or trend in the failure mode of the
wires; however, it was observed that cable bolts installed at angles, have encountered tensile
failure more than shear failure (Figure 8-1).

The failure mode of wires in all shear tests carried out on a 15.2 mm (about 0.6 in) cable bolt
was studied statistically. In 18 shear tests, carried out in both static and dynamic modes, wires
were numbered as shown in Figure 8-2. Strand wire number one is the top wire and number 4
is the lowest wire in the shear plane. Wires 2, 3, 5 and 6 are side wires, and number 7 is the
king-wire in the middle of the strand.
Statistical analysis of the failures in strand wires showed the highest chance of shear failure
related to wires number 4, 7, and 1, respectively. Figure 8-3 illustrates the probability of shear
251
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

failure according to the shear test results. The Green and red colour scale demonstrates the
probability of wires being failed in shear
Perpendicular shear

Angled shear
Test 6: 30° Test 7: 45° Test 8: 45° Test 10: 45°

Figure 8-1: Comparison of failure modes of strand wires in different angles of shear test

Figure 8-2: Numbering of the wires of 15.2 mm cable bolt strand

Figure 8-3: Chance of shear failure in a double shear test for different wires of a 15.2 mm
cable bolt

252
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

8.2.2 Effect of Cable Geometry on the Shear Behaviour


Figure 8-4 compares the test results of the perpendicular double shear test using plain and
bulbed 15.2 mm cable bolts. The maximum axial and shear loads recorded in each test have
been shown. Results clearly indicate that bulbs do not significantly change the shear load
capacity of the cable bolt because both ends of the cable bolt were retained by B&Ws, thus
preventing the axial movement of the cable. Hence, it can be said that testing using the Mk-IV
DS rig was not a suitable tool for this purpose, and the results of pullout tests should be
considered with respect to the objective of the test.

Weak grout Strong grout


(20-30MPa) (60-80 MPa)
350 314 318
300 266
246
250 227 230
Load (kN)

195
200 182

150

100

50

0
Plain Bulbed Plain Bulbed
Cable type

Axial load Shear load

Figure 8-4: Comparison of plain and bulbed cable bolts in perpendicular double shear tests

8.2.3 Effect of Grout Strength on Shear Load Capacity


Figure 8-5 and Figure 8-6 compare the results of shear and axial loads recorded from shear
tests on the load applied perpendicular to the bolt axis (zero-degree orientation) with different
grout strengths. It can be stated that the maximum shear load in the weaker grout samples is
roughly 75% of those tested with stronger grout. Also, it can be said that bulbs on the cable
bolt enhanced the shear resistance though the difference was not significant. Furthermore, both
ends of the tested cable bolt were firmly fixed by B&W and did not allow debonding; thus,
bulbs cannot react against the axial load. In order to study the effect of bulbs on the cable, pure
axial loading was required.

The recorded axial load on cable bolts followed the same trend as shear loads; however, the
maximum axial load in all cases, regardless of the grout strength and cable type, was in the

253
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

range of 75-80% of the maximum shear load. It is noted that the result of test 3 has been
removed from Figure 8-6 intentionally due to functional error of measurements during the test.
350
300
250
200
Load (KN)

150
100
50
0
0 10 20 30 40 50 60 70
Displacement (mm)

Test 1 Test 2 Test 3 Test 4 Test 5

Figure 8-5: Shear load in the double shear test perpendicular to bolt axis test of cable bolts
using different cementitious grout strengths
250

200

150
Load (KN)

100

50

0
0 10 20 30 40 50 60 70
Displacement (mm)
Test 1 Test 2 Test 4 Test 5

Figure 8-6: Axial load in the perpendicular double shear test of cable bolts using different
cementitious grout strengths

8.2.4 Effect of Static and Dynamic Loading Rate on Shear Load Capacity
Figure 8-7 shows the applied shear load in different tests, including perpendicular static shear
loading, angled static loading and perpendicular dynamic tests in ascending order. The number
on each bar indicates the test numbers. Comparing the shear loads, it can be inferred that the
load at failure for perpendicular static shear was approximately 50% of angled shear tests
because of two joint surfaces being sheared. A comparison of static and dynamic perpendicular
shear tests showed that dynamic samples experience 30 to 60% higher loads at failure.

254
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

700

9 10
600
18
7 6
8 17
Shear load (kN) 500
16
15
400 14
4 5 3
300 1 2

200

100

0
Perpendicular static Angled Static Perpendicular dynamic

Figure 8-7: Comparison of shear load in different static and dynamic double shear tests

Also, Figure 8-8 compares the energy absorption of the 15.2 mm cable bolt during the double
shear test in both static and dynamic modes. The results have been sorted in ascending order.
As can be seen in Figure 8-8, the energy absorption of cable bolts in perpendicular and angled
shear tests were in the same range. This is while the absorbed energy in the dynamic tests was
roughly 50% higher than in the static tests. The comparison of load and energy bar charts
(Figure 8-7 and Figure 8-8) revealed that high load levels do not necessarily result in high
energy absorption capability. For instance, although the shear load in the angled tests was
higher than the perpendicular tests, the recorded displacements in those tests were smaller than
the perpendicular tests. In fact, as the energy stems from the applied load and displacement, it
can be said that the energy absorption capacity of the cable bolt does not change at different
installation angles. The energy absorption increased by up to 40% in dynamic tests due to the
inertial, damping, and elastic forces (Clough & Penzien, 1995).

Figure 8-9 compares the axial load in different shear tests arranged in ascending order. As can
be seen, the axial load in both static and dynamic perpendicular shear tests are close to each
other, in the range of 150 to 250 kN. The axial load for the perpendicular static test is a function
of grout strength. In shear tests with stronger grout (test 4), a higher axial load has been
recorded. In angled shear tests, since the pretension load has not been equal for all the samples,
it is not easy to conclude a logical trend of changes in the axial load. In dynamic tests with an
equal pretension load for all the samples, it can be seen that samples in dynamic shear tests
experienced higher level axial load in comparison with static tests. The highest recorded axial
load in these tests relates to the dynamic tests numbered 17 and 18 with over 250 kN of the
255
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

applied load. This means that the chance of tensile failure of the wires in the test can be as high
as the shear failure.
18
18
16
14 15
16 17 14
12
Energy (kJ)

5 10
10 6
4 7
2 8
8
9
6 1
4
2
0
1 2 3 4 5

Perpendicular static Angled static Perpendicular dynamic

Figure 8-8: Comparison of energy in different static and dynamic double shear tests

400
18
350

300 17
Axial load (kN)

250 4 5
14 6
2
200 1 8
15
3 7 10
150
16
100

50

0
1 2 3 4 5

Perpendicular static Angled static Perpendicular dynamic

Figure 8-9: Comparison of axial load in different static and dynamic double shear tests

8.2.5 Effect of Angle of Installation on Shear Load Capacity


The comparison of static shear tests at different angles illustrates the fact that the less
perpendicular to the shear surface, the more load capacity of the tendon in shear displacement.
This can be seen and concluded from Figure 8-10 as well. The tensile strength of the cable
bolts is logically higher than the shear strength; thus, cable bolts installed at an angle will
experience higher tensile stress during the shear test. Consequently, higher resistance versus
the applied load was observed.

256
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

900

800
Perpendicular- 1
700
Perpendicular- 2
600 Perpendicular- 3
Load (kN)

500 Perpendicular- 4
Perpendicular- 5
400
30 degree
300 45 degree- 1
200 45 degree- 2
45 degree- 3
100 45 degree- 4
0
0 10 20 30 40 50 60 70
Displacement (mm)

Figure 8-10: Results of the shear test at different angles

Evaluation of the shear tests revealed that the maximum shear load capacity of the tendon in
perpendicular shear tests was roughly 150 kN which is 60% of its maximum tensile strength.
This is while both 30-degree and 45-degree angled shear tests recorded a shear load capacity
of almost 250 kN. This value equals the maximum tensile strength of a 15.2 mm cable bolt. It
means, in angled shear tests, the wires were positioned almost parallel to the loading direction,
which applies a pure axial load to the cable bolt due to the flexibility of the cable bolt. Evidence
of the failure of wires in two tests clearly proved this statement; however, the other two angled
shear tests showed some shear failure of the wires. Measuring hinge angles in perpendicular
double shear tests showed that cable bolts experienced a hinge angle of 36 to 40 degrees when
they failed (Figure 8-11). Considering that the hinge occurs in cable bolts before failure, it is
expected that any cable installed at an angle will experience a high tensile load. For instance,
the cable bolt installed at 45 degrees would be in an 80–85-degree position, almost parallel to
the loading axis. The hinge angle, however, depends on the strength of the grout and the
concrete.

257
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

36°

37°

40°

Figure 8-11: Hinge angle of cable bolts in perpendicular double shear tests

Pullout Tests Analysis


Figure 8-12 compares the results of different static pullout tests using different combinations
of plain cable and grout (P/G), bulbed cable and grout (B/G), plain cable and resin (P/R) and
bulbed cable and resin (B/R). 63 t Sumo cable bolt was used for this section of the study. The
pseudo yield and the maximum pullout load in these tests were compared, and it was found
that in all tests except the P/R (Tests 6 and 7), the maximum pullout load was up to 40% higher
than the pseudo yield. Although the pseudo yield loads of plain cable in both grout (Tests 1, 2,
3) and resin (Tests 6 and 7) were roughly equal, the maximum pullout force in the resin samples
was significantly lower. This represents a totally different behaviour of resin in the second
phase of the pullout test. The difference in behaviour lies in the failure mechanism, which needs
to be assessed in detail in the next section.
600

500

400
Load (kN)

300

200

100

0
1- P/G 2- P/G 3- P/G 4- B/G 5- B/G 6- P/R 7- P/R 8- B/R 9- B/R

Pseudo yield load Maximum pullout load

Figure 8-12: Results of static pullout tests on 63 t Sumo cable bolt

258
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Figure 8-13 compares the normalized pullout energy of 16 tests (eight statics and eight
dynamics) being sorted out in ascending order. The normalized energy is the required energy
to pullout the cable bolt for 1 mm and was calculated from the ratio of total absorbed energy
and total recorded displacement at the end of each test. This comparison study revealed that
the required energy to pullout the cable dynamically is in the range of 50-60% of the static tests
for plain cable and 75-80% for the bulbed cables. Also, evaluation of the results of both static
and dynamic tests (Figure 8-12 and Figure 8-13) shows that the weakest performance was
related to P/R samples in both static and dynamic tests. In contrast, the B/R combination
behaved the strongest in response to static and dynamic pullout load.
0.45
Normalized pullout energy (kJ/mm)

0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
P/R P/G P/G P/G B/G B/G B/G B/R

Dynamic Static

Figure 8-13: Comparison of normalized energy in static and dynamic pullout tests

8.3.1 Failure Mechanism of Encapsulation Material


In order to understand the failure mechanism of pullout tests, several shreds of evidence were
utilized to obtain a conclusion.

a) Post-test visual observations


After the test, concrete host medium samples were split axially to observe the damage intensity
of the grout and concrete. Figure 8-14 compares the failure modes of grout in both static and
dynamic loading conditions using plain and bulbed cable bolts. In general, the following
observations can be recorded:

1. In static tests using plain cables (Figure 8-14-a and d):


a. Broken ridges of grout on the top side of the sample were more severe than on
the bottom side. Ridges on the top 50 mm of the grout column entirely
259
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

disappeared, and the grout surface was smooth. This means that in the static test,
the cable was gradually elongated, which led to an axial stretch of the cable and,
consequently, dents on the grout edges. Obviously, the chipped edges of the
grout are more frequent on the top side, which faces more tensile load than the
bottom end.
b. Getting closer to the bottom end, the ridges were sharper, with slightly broken
edges.
c. The surface of the grout on the grooves was turned black, the effect of high
frictional movements of the cable on the surface of the grout. The black surface
was evenly distributed alongside the grout column.
d. The surface of the resin was eroded to powder, and small chips and dents of the
resin were observed.
2. In dynamic tests using bulbed cables (Figure 8-14-b)
a. Sharp ridges and less chipped edges alongside the grout column could be seen.
There was no significant difference between the failure modes of the grout at
the top and bottom ends, which shows the effect of a fast-loading rate.
b. More frequent burned surface in comparison with the static test, which is
evidence of more frictional interaction between the grout and the cable bolt.
3. In both static and dynamic tests using bulbed cables (Figure 8-14-c and e):
a. Failure modes were totally different on the top half as compared with the bottom
half being separated by the bulb.
b. The top half experienced highly crushed ridges with semi-smooth surfaces and
high frictionally burnt grooves.
c. Widely broken ridges with rough surface and dented edges without any trace of
the high frictional interaction in the area of the bulb and at the bottom half can
be clearly seen. This implies that bulbs create a robust physical anchorage with
the great.
d. Severe shear failure of resin ridges was observed; however, the broken ridges
were perfectly stuck to the surface of the cable and filled the gap between wires.

260
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

(a) Static test- Plain Sumo (b) Dynamic test- Plain Sumo (c) Static test- Bulbed Sumo

(d) Static test- Plain Sumo (e) Static test- Bulbed Sumo
Figure 8-14: Grout damage intensity after static and dynamic tests with plain cable

261
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

b) Rotation of the cable bolt


Visual monitoring of the extruded end of cable bolts in both static and dynamic tests revealed
that even though cable bolts are fully constrained against rotation by anti-rotation systems, the
free bottom end of the cable tends to unwind and slide through the grooves rather than breaking
the ridges in shear.

Figure 8-15 shows the free bottom end of a plain cable before the start of pull testing and after
almost 30 minutes of static testing (30 mm of displacement). One of the wires of the cable bolt
was painted white as the movement indicator. A camera placed strategically to observe any
cable movement showed the tendency of the cable bolt to unwind at the free end. This
behaviour is contrary to what has been recorded in the pullout test of rock bolts (Aziz et al.,
2006). Difference between the geometery of a cable bolt and a rock bolt is the main reason for
different behaviour in the pullout tests.

Before starting the static test After 30 minutes of running the static test
Figure 8-15: Almost 40-degree rotation in the direction of unwinding of the cable bolt during
the static test

Numerical models also revealed that the free bottom end of the cable tends to rotate and move
along the grout surface rather than shear the grout edges. As shown in Figure 8-16 from
numerical simulations, the marked wire on the cable bolt has been displaced rotationally as the
test progressed.

Similar unwinding behaviour can be seen in the dynamic tests, which showed that the rate of
loading did not alter the debonding mechanism of the plain cable bolt. Also, the same
unwinding behaviour can be seen in the bulbed cables; however, there were some slight
differences in their structure. The hollow tube of the cable bolt in the plain cable was perfectly
covered by wires and welded to the wires at both ends. As wires unwind, the hollow tube twists

262
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

accordingly, which is attributed to the effect of the bulb structure. The bulbed cable has a gap
between wires and the hollow tube in the bulb area, which is filled by encapsulation material.
Therefore, the hollow tube is confined by the grout. When the wires are untwisted and pulled
out during the test, radial pressure in the bulb area increases which causes an increase in the
radial pressure in the middle tube. As wires find their way to slide out of the grout, the middle
tube remains confined and fixed. Eventually, the weld between the wires and tube breaks,
leaving the hollow tube in place and not being pulled out with the other wires (Figure 8-17).
This snapped point is likely to occur at a distance of 150 mm from the end of the cable. It is
believed that this behaviour could have been different from the bulbs further away from the
end of the cable.

Figure 8-16: Rotational displacement of the bottom end of the cable

Figure 8-17: Failure of the bulbed cable bolt in static and dynamic pullout tests

263
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

c) Stick-slip behaviour during pullout tests


In most samples of the plain cable bolt encapsulated with grout and some of the bulbed cables
with the same encapsulation material, fluctuations/oscillations were observed in the load-
displacement graph. The oscillation in the load values occurred after the pseudo-yield step and
during the pullout test until reaching the maximum pullout load. An example of fluctuating
behaviour is shown in Figure 8-18. Similar behaviour was observed and reported by (Hagan
and Li (2017), Hyett et al. (1995), Stillborg (1993) without elaborations; Hagan (2017),
however, assumed that the behaviour related to the arrangement of the hydraulic system.
140

120

100
Load (kN)

80

60

40

20

0
0 10 20 30 40 50
Displacement (mm)

Figure 8-18: Pullout test result of 15.2 mm plain cable in medium strength grout carried out
by SDPA

Jaeger (1971) reported similar behaviour in a direct shear test of a granite sample and called it
a "stick-slip" relaxation oscillation (Figure 8-19). They analogize the stick-slip behaviour with
the motion of mass (M) (weight W) against a surface through a spring of stiffness (k) whose
other end is moved at a constant speed (V). The dynamic coefficient of friction (µd) of the
surface is constant and less than that of static friction (µs). When M is at rest, it slips when the
extension of the spring is x1, and the applied force is F1. If the velocity (V) is very slow
compared to the velocity of the slip, the stored energy in the spring is partially lost, and the
force appears to oscillate (Jaeger, 1971). The oscillation in load can be represented as Equation
8-1.

264
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

𝛥𝐹 = 2 (µ𝑠 − µ𝑑 )𝑊 Equation 8-1

Figure 8-19: Theory of stick-slip behaviour

Since the cable bolt is in its elastic elongation stage during the pullout test and considering the
fact of the rotational movement of the cable during the test, it can be concluded that the cable
bolt is frictionally moving on the surface of the grout; hence, stick-slip behaviour is a possible
scenario.

d) Performance of the B&W in load bearing


In its simplest form, the cable bolt setup includes wires, a bearing plate, a barrel, and a two- or
three-part wedge. The structure of cable bolts could be more complicated by adding bulbs (or
nut cages) or indentations on the cable design and confinement plates (flat or domed) (Figure
8-20). Case studies (Thompson & Villaescusa, 2014) concluded that the strength of the tendon
is not the sole indicator of the maximum load capacity of the reinforcement system. Parameters
such as plates, straps, mesh and shotcrete -as external fixtures- are simultaneously playing an
important role in restraining surface support (Thompson & Villaescusa, 2014).

Barrels and wedges are frictionally in touch with the cable bolts (Figure 8-21). The wedge
segments grip the cable by means of tooth-shaped threads forced into the cable wires' surface.
The clamping force (W) highly depends on the friction between the wedge and the barrel. For
normal friction conditions, the clamping force (W) is more than four times the cable pulling
force (V), and it adjusts with the pulling force (Thompson & Villaescusa, 2014).

265
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Figure 8-20: Components of a cable bolt support

Figure 8-21: Mechanism of cable bolt and B&W (Thompson & Villaescusa, 2014)

Static tests revealed that the wedge's frictional movements initiate before the cable is pulled
out, although the recorded deformations were relatively small. A comprehensive examination
of the performance of B&Ws by Thompson (2014) revealed that the slippage of the strand
within the B&W occurs at forces less than 20% of the strand's ultimate capacity. This also
becomes worse when anchors are corroded by moisture in the field (Thompson & Villaescusa,
2014). The proper B&W for a 63 t Sumo cable has a loading capacity of 540 kN (Jennmar,
2022).

Table 8-1 and Figure 8-22 compare the recorded load and deformation of the B&W used in
static and dynamic tests with two different plain and 63 t bulbed Sumo cable bolts. Based on
the recorded deformation of the B&W during static tests, Figure 8-22 demonstrates the amount
of the wedge being pulled in through the barrel with different tensile loads. Since the constant
recording of the deformations was not possible in dynamic tests, the deformation of the B&W

266
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

was measured before and after the test. According to the B&W deformation results of static
tests (lines in Figure 8-22), a second-degree regression line (solid black line) was fitted to
forecast the deformations of larger loads such as dynamic loads (Red circle dots in Figure
8-22). Also, the final deformation of some of the static tests was plotted to assess the accuracy
of the estimating equation (Blue triangle dots in Figure 8-22). There was an acceptable level
of correlation between the experimental results and estimating equation. Estimations and
experimental results of static and dynamic tests revealed that the performance of the B&W is
independent of the loading ratio; however, the B&W of dynamic tests experienced more
significant deformations.
Figure 8-23 shows the status of B&W before and after static and dynamic tests. B&Ws installed
on bulbed cables experienced slightly more deformation, which shows a greater chance of
failure or slippage of the cable; even though, no failure occurred in any of the tests with 63 t
Sumo cable, at the B&W.
Table 8-1: recorded load and deformation of B&W in static and dynamic tests

Test No. P0 P10 T1 T2 T3 T4 T5 T6


Before test Pretensioned Static tests Dynamic tests
1t 10 t Plain Bulbed Bulbed Plain Bulbed Bulbed
deformation (mm) 0 5 7 9 10 8 10 11
Peak load (kN) ~10 ~100 305 500 550 270 553 650
800
700
600
Axial load (kN)

500
400
300
200
y = 8.7742x2 - 31.44x
100
0
0 2 4 6 8 10 12
Wedge displacement (mm)

Static test 1 Static test 2


Static test 3 Static tests for verification
Dynamic tests for verification

Figure 8-22: B&W deformation in the static pullout test

267
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Before test

P0 P10
(1 t pretension) (10 t pretension)

After test

T1: P/G T2: B/G T3: B/G T4: P/G T5: B/G T6:B/G
Static Dynamic
Figure 8-23: B&W deformation in static and dynamic pullout tests

Figure 8-24 compares the status of B&W before and after the static pullout test of the 15.2 mm
cable bolt. In 15.2 mm bulbed cable bolts, the tensile failure of the cable bolt occurred right
below the B&W. Also, sparks were observed during the test due to the high frictional
movement of the wedge and barrel. In none of the cases the slippage of the cable was observed
inside the wedge; however, it did not cause failure.
Before test After test

15.2 mm plain cable 15.2 mm bulbed cable


Applied pretension load: 5 t
Max load: 131 kN Max load: 255 kN
Figure 8-24: Status of B&W before and after static test of 15.2 mm bulbed cable

268
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

e) Proposed pullout failure mechanism


To sum up, according to the experiments as well as the literature review, it can be concluded
that the failure mechanism of cable bolts encapsulated in cement grout or resin can be one of
the possible options in Figure 8-25. The reader is reminded that different tendons might
demonstrate different behaviour.

P3
III
P1 IV
Pullout Load

P2 II

δ1 δ2
Displacement

Figure 8-25: Possible failure mechanism of cable bolts in pullout test

• Pullout load increases in an elastic manner up to P1, where the load overcomes the bond
between the tendon and grout. Elastic elongation of the cable bolt or deformation of the
B&W can also occur in this stage (δ1). Shear resistance in this stage is related to the
adhesion between the cable surface and grout (Mylrea, 1948).
• It is then followed by a softening stage in which the load drops to P2. This stage might
be negligible and not easily discernible, especially if the strength of confinement is not
high.
• Due to several reasons consisting of, but not limited to; the confinement strength,
consistency of the embedment length, and geometry of the cable bolt any of the
scenarios I to IV may occur.
• Case (I): In any of the situations when the confinement pressure is decreasing,
embedment length is reducing or grout material has experienced severe shear failure,
the pullout load dwindles to zero or
• Case (II): post-peak residual behaviour occurs when the cable and grout's frictional and
shear resistance does not exceed the adhesion resistance. The low friction coefficient
269
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

or low shear resistance of the grout could be the main reason. It was mainly observed
with resin and in plain cable bolt application where the friction coefficient was low or
• Case (III): Hardening occurs when the frictional or shear resistance of the cable bolt
and grout is higher than the adhesion. Also, the progress of the pullout load leads to
greater resistance due to the increase of confinement pressure or frictional interaction.
This case is mainly observed with the bulbed cable bolts and high-strength cement
grout. The maximum recorded pullout load in (P3) stage would be higher than the
adhesion debonding load (P1). P3 is considered the maximum pullout load. or
• Case (IV): Stick-slip occurs when the friction coefficient and confinement pressure are
simultaneously high. It results in oscillations in load values; however, the maximum
pullout load would not be affected.

8.3.2 Parametric Study


In this section, the effect of several parameters, including cable geometry, grout properties,
friction coefficient, and embedment length, are assessed with the help of numerical and
experimental results. It should be noted that 63 t Sumo cable bolt and cementitious grout have
been used for the parametric study.

a) Effect of cable geometry on the pullout behaviour


Typically cable bolts have their natural elongation rate before failure ranging between 5-7%
(Jennmar, 2022). According to the analytical and experimental studies carried out on load
intensity alongside the rebar in an axial loading condition, such as those of (Blanco Martín,
Tijani, et al., 2011; Y. Cai et al., 2004; Kilic, Yasar, & Celik, 2002; Nemcik et al., 2014; Ren
et al., 2010), the axial force on the cable can vary widely from the top side to the bottom end.
The debonding mechanism is usually introduced as a progressive failure at different stages,
starting with the initial elastic stage followed by hardening and then softening stages (Ren et
al., 2010). However, in analytical studies, it is usually assumed that the load on the tendon does
not exceed its elastic yield point, and the only possible failure is the shear failure of the
encapsulation material. According to the literature review, there is a linear relationship between
the bolt length and the required pullout load (Kilic, Yasar, & Celik, 2002; Ma et al., 2016;
Thompson & Villaescusa, 2014). In other words, when the encapsulation length of the cable
bolt is very long, the tensile failure of the tendon would be the dominant mode. Accordingly,
the effect of cable geometry on the load distribution in the cable bolt may require further
scrutiny.

270
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Results from both static and dynamic pullout testing showed that the bulbed cables experienced
less debonding, higher elongation, higher axial load and higher energy levels. It can be said
that the geometry of the bulb on the cable bolt acts as a physical anchor which holds the cable
in its position. Bulbs contribute to an increase in the pullout load and debonding capacity of
the cable bolts by up to 60%. The opened conical cracks on the concrete cylinder in high-
strength 63 t Sumo cable bolt samples were clear evidence of high radial stress around the bulb
(Figure 8-26).

Top
Top

Figure 8-26: High stress and opened cracks around the bulbed area

In addition, the recorded elongations of 63 t Sumo cable bolt in pullout tests have been
summarized in Table 8-2. Higher elongation of the bulbed cables can strengthen the hypothesis
of bulbs acting as anchors. Also, there is no significant difference between the elongation rate
of static and dynamic pullout tests, which shows that the loading rate does not affect the
elongation. The recorded relative displacement of the top and bottom of the cable bolt could
roughly measure the elongation of the cable bolt. Recorded elongation displayed that the cable
elongation was in the elastic range of 5-7%; hence, no failure in cables was observed.

Figure 8-27 shows the relation between elongation and the estimated absorbed energy of cable
bolts for different tests. According to the results, the permitted elongation before failure of the
cable bolts stores a portion of energy up to 5 kJ, increasing their load-bearing capacity,
especially in high-energy loading conditions such as dynamic loading.
271
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Table 8-2: Cable bolt elongation measurements in static and dynamic pullout tests

Max Elongation Elongation


Test type
mm %
P/G 11 2.2
B/G 16 3.3

Static
B/G 18 3.7
P/G 16 3.3
P/R 2 2.5
P/G 6 1.2
P/G 10 2.1
Dynamic

P/G 8 1.6
B/G 14 2.9
B/G 17 3.5
B/G 17 3.5
P/R 0 0
6

5
Absorbed energy (kJ)

0
0 5 10 15 20
Elongation (mm)

Figure 8-27: Relation between elongation and the absorbed energy

Similarly, in low-strength 15.2 mm cable bolts, the recorded displacement of the bulbed cables
was smaller, and the pullout load was higher than in similar tests carried out using the plain
cable bolt. Figure 8-28 shows the elongation of the 15.2 mm bulbed cable bolt before and after
the pullout test. As can be seen, more than 30 mm of elongation in the cable has occurred
because of the bulb. Moreover, the tensile failure of the cable bolt was observed in all tested
bulb cables.

272
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

460 mm

Before test

Strong grout

Medium grout

Figure 8-28: Comparison of cable lengths before and after tests

b) Effect of grout strength type on the pullout behaviour


As mentioned in previous chapters, the strength of the grout is a function of many different
parameters, such as w/c ratio, age, preparation method and curing conditions. Therefore, the
evaluation of the effect of grout strength on the pullout behaviour is essential. Figure 8-29
classifies the pullout test results carried out using three different grout strengths. As can be
seen, the required pullout force in the weak grout is considerably lower than the medium with
strong grouts, even though the difference between the last two is not significant.

Figure 8-29: Normalized pullout load

Figure 8-30 compares the effect of grout strength on the pullout load response of the cable bolt.
Three different UCS strengths of 20, 40, and 50 MPa were modeled for cementitious grout.

273
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Results showed that the grout strength could significantly increase the Pseudo yield point,
however, the post-peak response was not dependent on the grout strength.
450
400
350
300
Load (kN)

250
200
150
100
50
0
0 20 40 60 80 100 120
Displacement (mm)

20 MPa 40 MPa 60 MPa

Figure 8-30: Effect of grout strength on the pullout load

Also, Figure 8-31 compares the plastic zone of the grout in each test. It can be clearly seen that
the stronger the grout, the smaller the plastic zone on the grout surface. Also, the stress
concentration is mainly on the grout edges.
20 MPa

40 MPa

50 MPa

70 MPa

Figure 8-31: Effect of grout strength on the effective plastic strain

274
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

c) Effect of static and dynamic friction coefficient on the pullout behaviour


Confinement pressure and coefficient of friction can affect the response of the cable bolt to the
axial load. The effect of different friction coefficients has been plotted in Figure 8-32. Note
that in this part of the study, the static and dynamic friction coefficients have been assumed to
be equal. As can be seen, the maximum pullout load of the cable bolt can be affected
significantly by the friction coefficient, even though the pseudo-yield point is equal in all
samples. In other words, the frictional interaction between the cable bolt and the grout began
when their bond was fully broken.

700

600

500
Load (kN)

400

300

200

100

0
0 10 20 30 40 50 60 70 80 90 100
Displacement (mm)

0.2 0.3 0.4 0.5 0.6

Figure 8-32: Effect of coefficient of friction on pullout load

Figure 8-33 presents the effect of the friction coefficient on the shear stress distribution as well
as propagation of the plastic zone. The shear stress has been shown by colour contours, and
plastic zones have been demonstrated in iso-surface mode. In iso-surface mode, the colour
contour of the stress intensity on the surface is plotted, and the rest of the areas are hidden. As
can be seen, when the friction was low (Fs, Fd= 0.2), there was less stress concentration on the
grout edges. In other words, the cable bolt tended to move along the grooves on the surface of
the grout rather than breaking the grout edges in shear.

In contrast, when the friction coefficient was high (Fs, Fd= 0.6), the stress was concentrated on
the grout edge as it was easier for the grout to be pulled out. This also can be seen in the Iso-
surface plastic zone graphs (Figure 8-33). The intensity of the plastic zones on the edges of the
grout with a higher friction coefficient can be clearly seen in the iso-surface view in Figure
8-33.

275
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Friction coefficient = 0.2

Friction coefficient = 0.6

Figure 8-33: Effect of friction coefficient on the shear stress and plastic zone distribution

d) Effect of shear resistance of contact between cable bolt and grout


The shear resistance between the cable bolt and the grout is one of the controlling parameters
of the contact properties. Figure 8-34 exhibits the effect of shear resistance on the behaviour
of cable bolts being subjected to pullout load. As the graph reveals, the Pseudo yield point is
directly affected by the shear resistance, and there is only a slight improvement in the peak
pullout load. It is proposed that the shear resistance is generated as a result of the chemical
bond between the cable bolt and encapsulation material.
450
400
350
300
Load (kN)

250
200
150
100
50
0
0 20 40 60 80 100
Displacement (mm)

20 MPa 40 MPa 50 MPa 80 MPa

Figure 8-34: Effect of shear resistance on the axial pullout load

276
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Figure 8-35 compares the effect of shear resistance on the effective plastic failure zone in the
grout. When the grout shear resistance was low, more severe damage occurred in the sample.
In contrast, in higher shear resistance, plastic zones were limited only to the weakest points of
the encapsulation material (edges), as indicated in the Figure.

20 MPa

40 MPa

80 MPa

Figure 8-35: Effect of shear resistance on the effective plastic strain

The difference in failure could also be seen in two different experiments of the same kind with
totally different levels of damage (Figure 8-36). While both tests used the same cable bolt
(Plain 63 t Sumo) in static mode, the first test (Figure 8-36-a) experienced severe destruction
in the structure of the grout, while in the second test (Figure 8-36-b), there are sharp grout
edges with traces of high frictional interactions.

277
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

a) severe damage to the grout surface

b) sharp edges and frictional interaction


Figure 8-36: Static pullout tests of plain Sumo cable bolt

e) Effect of embedment length on the pullout behaviour


As testing of different embedment lengths in the laboratory requires manufacturing of new test
equipment such as larger confinement cages and stronger loading machines, therefore, the
effect of length of encapsulation on the response of cable bolt was studied by numerical models.
Figure 8-37 to Figure 8-39 show the recorded axial load on 63 t cable bolt with three different
embedment lengths, including 300 mm, 450 mm, and 600 mm based on numerical simulations.
The axial load values were recorded at 50 or 100-mm intervals on the cable length. For instance,
L0 represents the top of the pullout test sample where the cable extrudes from the top end of
the concrete sample. L300 is a point 300 mm toward the bottom of the concrete sample from
the top of it. As shown, the pseudo-yield point of all tests followed the same pattern. As the
load reaches the pseudo-yield point, the adhesive contact between the cable bolt and the grout
is removed and causes drops in the load values. Oscillations continue until shear resistance and
frictional forces act against the movement of the cable bolt. In this step, the pullout load
significantly increases. The top part of the cable botl (L0) experienced the highest axial load
while the most bottom part experienced the lowest load values.
Figure 8-40 compares the maximum pullout load of each test with different encapsulation
lengths. The graph clearly illustrates that the peak pullout load of longer embedment lengths

278
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

can be higher than 630 kN, which is the ultimate tensile strength of the current cable bolts of
the study. Hence, implementation of the mentioned tests is not practical at a laboratory scale.

Figure 8-37: Axial load on cable bolt with 300 mm embedment length

Figure 8-38: Axial load on cable bolt with 450 mm embedment length

Figure 8-39: Axial load on cable bolt with 600 mm embedment length
279
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Figure 8-40: Comparison of axial load in different embedment lengths

Combination of Shear and Pullout Loading


It is essential to understand the behaviour of the cable bolt in real-field conditions. In the shear
test, carried out in the laboratory, both ends of the cable bolt were fully restricted from the axial
movement, while in the actual field condition, this condition does not exist. It means the axial
pullout load of the cable is possible while the cable is under shear loading. According to the
recording of axial loads in shear tests, the unit axial load per millimeter of embedment length
can be calculated. These values can also be compared with the average unit pullout load gained
from different pullout tests. Figure 8-41 and Figure 8-42 compare the recorded axial load in
shear tests with the required load to pullout according to the pullout test results. As can be seen,
that almost all the shear test samples could experience pullout during the shear test if there was
no constraining load applied by the B&W.

280
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

0.9
0.8 Bulbe Bulbe
Unit load (kN/mm) 0.7 Plain
Plain
0.6
Plain
0.5
0.4
0.3
0.2
0.1
0.0
1 2 3 4 5
Test number

Applied axial load in shear test Required load to pullout

Figure 8-41: Comparison of unit axial load in shear tests and pullout tests for 15.2 mm cable
bolt
1.6
Bulbed Bulbed
1.4

1.2
Unit load (kN/mm)

1.0 Plain Plain


Plain
0.8

0.6

0.4

0.2

0.0
1 2 3 4 5
Test number

Applied axial load in shear test Required load to pullout

Figure 8-42: Comparison of unit axial load in shear tests and pullout tests for 63 t Sumo cable
bolt

A similar experience was observed in the single-shear tests carried out by (Aziz et al., 2018;
Aziz, Rink, et al., 2017). Figure 8-43 shows the failure of an MW10 cable bolt on the shear
surface of a single shear test reported by (Aziz, Rink, et al., 2017). Debonding and pullout of
the cable bolt in their tests were reported even though the embedment length of their sample
was 1800 mm (6 times longer than MK-IV).

281
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Figure 8-43: Debonding and pullout of the cable bolt in a single shear test, reported by (Aziz,
Rink, et al., 2017)

To sum up, although the existence of the B&W on both ends of the double shear test allows
the study of the shear loading of cable bolts, the results might be conservative.

Summary
In the previous chapters, cable bolts were subjected to three different loading angles, including
pure shear tests (perpendicular shear test), angled shear tests and pure axial tests (pullout test).
In this chapter, the failure mechanisms of the tendon for each loading condition were discussed.
Pure shear and pure axial tests were repeated with both static and dynamic loading rates. The
mechanism of failure for each loading condition was evaluated in this chapter. Several
parameters were observed and compared:
• Applied load
• Absorbed energy
• Failure modes of cable bolt components (wires, B&W)
• Failure modes of medium (grout and concrete)

Finally, the effect of different parameters on the pullout test behaviour was assessed using
experimental and numerical analysis methods.

The analysis of shear and axial tests revealed that, while in shear tests, cable bolts were
subjected to failure, in axial tests, the bond between the cable bolt and encapsulation material
failed prior to the cable failure. Hence, the failure mechanism of steel cable bolt must be

282
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

considered as the dominant failure in double shear tests. Also, failure mechanism of
encapsulation materials (cementitious grout or resin) should be considered for pullout tests.

In perpendicular shear tests (zero angle orientation/ applied load perpendicular to the cable
bolt), wires experienced both shear and tensile failures. The tensile failure was detected by the
cup and cone shape of the failed wire, and the shear failure appeared as a sharply cut surface.

Assessment of the effect of different parameters such as the cable geometry, the grout strength,
the loading rate, and the installation angle on the shear behaviour of cable bolts revealed that
the geometry of the cable bolt (plain or bulbed) and the grout strength have a minor effect on
the shear resistance of the tendon.

Although the shear load capacity of angled shear tests was higher than for the perpendicular to
bolt axis samples, the absorbed energy before failure in all static shear tests was the same. In
other words, the product of the integration of the load and the displacement should be constant
in all situations.

The shear load capacity and the absorbed energy in dynamic shear tests were roughly 40%
higher than in static tests. Extra dampening and inertia energy was stored in the steel cable
bolts.

The pullout procedure of cable bolts mainly includes two steps. Step one is when the applied
load overcomes the adhesive bond between the encapsulation material and the cable bolt. The
applied load reaches the so-called "pseudo-yield point" where the debonding occurs. Step two
is when the cable bolt initiates the movement inside the encapsulation material. Based on the
encapsulation material, confinement type, and cable geometry, another peak load in the pullout
test might be experienced in this stage, called the maximum pullout load.

The failure mode of encapsulation material in the pullout tests was divided into two modes:
shear failure and frictional resistance. Obviously, the frictional resistance of the grout was
much lower than the shear resistance; however, it is not easy to measure the share of each one
in these tests. In static tests, where the cable bolt had enough time to elongate its elastic range,
there was a greater chance of shear failure of the grout. In contrast, in dynamic tests, due to the
factor of time, the cable bolt tended to slide through the grooves of the grout in a fully frictional
interaction. Hence, less energy is required for the cable bolt to be pulled out. In general, the

283
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

maximum pullout load and the absorbed energy of dynamic tests were in the range of 60-75%
of the static tests.

A failure mechanism was proposed for cable bolts and grout failure (Figure 8-25). According
to the results of experiments, the pullout mechanism includes the following stages:

Stage 1: Pullout load increases in an elastic manner up to P1, where the load overcomes the
bond between the tendon and grout. Elastic elongation of the cable bolt or deformation of the
B&W can also occur in this stage (δ1). Shear resistance in this stage is related to the adhesion
between the cable surface and grout (Mylrea, 1948).

Stage 2: Stage 1 is then followed by a softening stage in which the load drops to P2. This stage
might be negligible and not easily discernible, especially if the confinement strength is not
high.

Stage 3: Due to several reasons consisting of, but not limited to confinement strength, the
consistency of the embedment length and geometry of the cable bolt, any of the following
scenarios I to IV may occur.

Case (I): The pullout load dwindles to zero when the confinement pressure is
decreasing, the embedment length is reducing or the grout material has experienced
severe shear failure,

Case (II): Residual stage occurs when the cable and grout's frictional and shear
resistance does not exceed the adhesion resistance. This may be attributed to low
friction coefficient or low shear resistance of the grout. This was observed mainly with
a low friction coefficient with the application of resin and plain cable bolts.

Case (III): Hardening stage occurs when the frictional or shear resistance of the cable
bolt and grout is higher than the adhesion. Also, the progress of the pullout load leads
to greater resistance due to the increase in confinement pressure or frictional interaction.
This was mainly observed with the bulbed cable bolts and high-strength cement grout.
The maximum recorded pullout load in this stage (P3) was higher than the adhesion
debonding load (P1). P3 is considered the maximum pullout load.

284
Chapter 8 | Discussion on Axial and Shear Loading of Cable Bolts

Case (IV): Stick-slip occurs when the friction coefficient and confinement pressure are
simultaneously high. It results in oscillations in load values, however, the maximum
pullout load would not be affected.

The effect of different parameters such as cable geometry, grout strength, the static and
dynamic friction coefficient of the surfaces and embedment length on the behaviour of cable
bolts under axial loading was examined with the help of the numerical model and the laboratory
experiments. While the grout strength and shear resistance significantly affect the Pseudo yield
point and do not affect the maximum pullout load, increments of friction coefficients and
embedment length substantially increase the maximum pullout load and do not change the
debonding load.

285
Chapter 9 | Conclusions and Recommendations

Conclusions and Recommendations

CHAPTER 9
Conclusions and Recommendations

Introduction
More than 22 tests in shear and more than 36 pullout tests were carried out to evaluate the
performance of the cable bolts under different loading conditions. Different loading angles,
including perpendicular, 30-degree, and 45-degree to the shear surface as well as direct axial
loading (pullout test), have been undertaken. Static and dynamic loading rates were applied on
two different cable bolts in perpendicular double shear and axial pullout tests.

Two different testing setups were employed for double shear testing of cable bolts in
perpendicular and angled directions. Although the rigs were already built, the perpendicular
double shear rig was reinforced for dynamic testing. No dynamic testing of angled double shear
was practical due to the limitations in laboratory facilities. A new pullout test setup was
designed and built to implement pullout tests in static and dynamic modes (also called SDPA).
The performance of the new rig was verified by tests carried out on steel split cells.

The numerical model was generated using the Finite Element Method software LS-Dyna. A
detailed model of the nine-wire 63 t Sumo cable bolt encapsulated in cement grout was
generated. The purpose of the detailed model was to better understand the interaction between
the grout and cable bolt. A simplified model was later developed based on the results of the
detailed model in order to minimise the runtime as well as increase the efficiency.

The effect of some of the prominent parameters, such as grout type and age, and cable bolt
properties, was examined through laboratory experiments. Numerical models were also
generated to assist with further analysis of affecting parameters as well as minimising the
number of experiments. The effect of embedment length, mechanical properties of grout and
286
Chapter 9 | Conclusions and Recommendations

frictional interaction of cable bolts and grout in the axial test have been assessed by numerical
models.

This chapter summarizes the conclusions drawn from all previous chapters.

Behaviour of the tendons in shear tests


Results of 22 different shear tests were used to understand the behaviour of cable bolts under
shear loading conditions. The following conclusions could be inferred from the experiments:

• It should be stated that some of the findings of this thesis has not been addressed entirely
in the past. There has been a lack of knowledge about dynamic shear testing of cable bolts,
particularly in double shear test.
• Static shear tests revealed that the strength of grout and concrete could slightly affect the
shear loading capacity of the cable bolt. Samples with stronger grout experienced a 10%
increase in shear strength. This result was in agreement with the findings of Khaleghparast
in his similar experimental study on rock bolts (Khaleghparast, 2021).
• The ultimate shear load of the cable bolt was roughly 40% higher in the angled cable bolts
compared with those installed perpendicularly to the loading direction; however, the
allowed displacement before failure in the angled shear sample was significantly smaller.
In other words, the energy absorption of the cable in static loading is constant.
• The ultimate shear load and absorbed energy of the cable bolt in the dynamic shear test
were around 50% higher than in the perpendicular static tests. This can be explained by the
high damping effect of steel materials when encountering high dynamic loads.
• The statistical study of the chances of failure in shear for different strand wires revealed
that almost 60% of the wires fail in tension rather than in shear, even in pure shear tests.
This is because of the ability of the cable bolt to be readily bent. Once the cable bolt bends,
the load would be applied on wires in tension rather than in shear. In general, there was not
a common pattern of failure in shear tests. As a rule of thumb and according to the current
tests in this study, it can be conservatively proposed that the ultimate shear loading capacity
of the cable bolt, when installed perpendicular to the sheared surface, is 60% of its UTS.
The application of this estimation in engineering designs requires more verification.

Behaviour of the tendon in pullout (axial) tests


Results of more than 36 tests and seven trial tests were carried out with the newly designed
Static and Dynamic Pullout Apparatus (SPDA) and showed acceptable consistency. Also, the
287
Chapter 9 | Conclusions and Recommendations

performance of the apparatus was compared with a steel split set pullout test rig. The findings
from this experimental study demonstrated that the new rig was a valuable and versatile tool
for understanding and evaluating the behaviour of both types of cable bolts under both static
and dynamic loading and under different loading conditions. The following conclusions were
inferred from the experiments:

• The main assumption of the study was that the performance of the cable bolt under axial
loading was limited to the interaction between the cable bolt and encapsulation material;
however, this assumption was not necessarily valid for the low-strength cable bolt. In
bulbed 15.2 mm cable bolts, the strength of the cable bolt was lower than the required
pullout load; therefore, the tensile failure of the cable bolt was observed. This assumption
appropriately corresponded to tests with other types of cables.
• Results revealed that the axial pullout failure occurred in two steps; the first step was when
the axial load overcame the bond between the cable bolt and encapsulation material. The
second step was when the cable was pulled out. Accordingly, there were two different peak
loads for each step. The first peak was named Pseudo yield point and the second possible
peak load was called the “Maximum pullout load". While the Pseudo yield point was
dependent on the strength of the bond between two surfaces, the “maximum pullout load”
was controlled by the physical interaction between the cable bolt and grout.
• Experiments carried out with 63 t cable bolts encapsulated in both grout and resin showed
that both materials presented somewhat similar behaviour in the first stage, with close
Pseudo yield points; however, the maximum pullout load in the second stage significantly
dropped in the resin samples. This is while the maximum pullout load in the grout sample
was higher than their Pseudo yield point.
• The behaviour of the bulbed cables was affected by the existence of bulbs. In all of the
samples, the bulb reacted as a mechanical anchor and produced greater radial reaction force,
causing radial failure of samples or tensile failure of the cable bolt. Experiments showed
that the behaviour of bulbed cables was independent of the type of encapsulation material
used.
• Due to the spiral structure of wires in cable bolts, the surface of the encapsulation material
was made of ridges and grooves. Ridges are the sharp edges created due to the gap between
wires, and grooves are the trace on the surface of wires. Once the cable bolt debonds, the
second step causes the movement of the cable along the encapsulation material. The motion

288
Chapter 9 | Conclusions and Recommendations

of the cable can occur either by the failure of the ridges or by moving along the grooves.
The first one causes the shear failure of the encapsulation material, and the latter causes the
twisting of the cable bolt around its axial axis.
• In static tests, when the motion of the cable occurs very slowly, both shear failure of ridges
and untwisting of the cable bolt are possible. If the applied load is lower than the static
frictional resistance of the cable and encapsulation material, the shear failure of the ridges
is dominant. Once the load exceeds the static frictional resistance, the cable bolt moves
along the grooves, causing a black-burn scared surface on the groove surface, which
represents high frictional interaction. Due to the elastic characteristic of the cable bolts, the
stored energy is released, and the cable bolt stops moving until the stored energy exceeds
the static frictional resistance. These leads to fluctuations in the load graph, which is
commonly called stick-slip behaviour.
• In the dynamic pullout test, due to the factor of time, overcoming the dynamic frictional
resistance between the cable bolt and grooves is the simplest way for the cable bolt to be
pulled out. Untwisting of the cable bolt and its movement along the grooves is more feasible
than the shear failure of the ridges. The severely burnt surface of the grooves and scarcely
broken ridges were evidence of this condition.
• Unlike shear tests, the pullout resistance of the cable bolt is limited to the mechanical
properties of the encapsulation material, not the steel tendon. Brittle materials, such as
cementitious grout, dissipate energy by the propagation of cracks (Miller et al., 1999).
Besides, the dynamic friction coefficient of materials is usually smaller than static
coefficients of friction. Therefore, it was observed that less energy was required to pull out
the cables in dynamic tests. It should also be noted that higher load levels have been
recorded in dynamic tests due to the high-impact load. Experiments on 63 t plain cable
bolts revealed that the required energy in the dynamic pullout test is almost 50% lower than
in static tests.

Numerical analysis
The following were inferred from the Numerical study:

• Among the currently commercialized numerical modeling software, the ANSYS package
was selected due to the undeniable power of modeling in different applications. Modeling
detailed geometries is not easy and practical in some other equivalent software. This is
while the generation of very complicated geometries can be easily undertaken in the
289
Chapter 9 | Conclusions and Recommendations

provided platforms by ANSYS. As this study aimed to understand the interaction between
the cable bolt and encapsulation material, a detailed geometry was generated.
• Among all the available solvers in ANSYS, the explicit dynamic solver was the most
appropriate due to its powerful tools for modeling failure and element erosion. However,
the current explicit dynamic modulus of ANSYS was not recommended by the experts and
technical supporters of the software. This was due to the deficiencies of the newly
developed modulus of explicit dynamic. It was unanimously recommended by the experts
in the field to use LS-DYNA explicit dynamic solver.
• Among the available options for modeling the contact between the cable bolt and grout,
*CONTACT_TIEBREAK_SURFACE_TO_SURFACE appeared to present the best
performance. The mentioned contact card is capable of modeling the chemical bond
between the contacts and the physical interaction between the surfaces while both surfaces
are in touch. This has been one of the closest simulations of the pullout tests to the date of
this thesis.
• Once the failure mechanism of the pullout test was fully understood, a simplified model
was developed to decrease the runtime and increase the simplicity. The proposed simplified
model used a simple cylinder instead of a detailed cable bolt with calibrated friction
coefficient values to simulate the same pullout resistance. A comparison of the results
showed that the simplified model's results are in an acceptable alignment with the detailed
model.
• Sensitivity analysis of the parameters such as compression strength, shear strength of grout,
friction coefficients of surfaces and embedment length was carried out using numerical
analysis. The analysis revealed that mechanical parameters of grout, such as compression
and shear strength, can significantly affect the pseudo yield load, while the friction
coefficients of the surfaces impact the maximum pullout load. It can be clearly stated that
the failure modes in axial testing were switching between shear failure of ridges or
frictional motion on the surface of the grooves. Wherever the grout mechanical properties
overcome the frictional resistance of the contact surface, less damage on the ridges and
more frictional motion was observed. In contrast, wherever the friction coefficients were
dominantly higher than the strength of the grout, shear failure of the ridges was recorded.
These outcomes were in good agreement with the experimental results.

290
Chapter 9 | Conclusions and Recommendations

Recommendations
• Comparing solid rock bolts and cable bolts with the same diameter or strength can be
an excellent comparison to optimize the application of tendons. The efficiency of each
type under different loading scenarios can be examined and compared.
• The double shear testing rig was not rigid enough to implement the dynamic shear test
for high-strength cable bolts. It is recommended to reinforce the current design for
future studies. Replacing split cylindrical confinement with samples directly cast in
uniform steel tubes is recommended.
• Measurement of the energy in the current SDPA setup was not accurate because the
load was measured by the load cell installed on the tup hammer. The accurate load must
be calculated from the axial load cell installed on the cable bolt. Due to the limitation
of calibrations and existing inaccuracies in the readings of the mentioned load cell, it is
recommended to repeat a similar experiment with a more reliable load cell and verify
the energy values with the current findings.
• Different cable bolts must be pull-tested to widen the understanding of axial behaviour.
• The embedment length of axial tests was limited to 450 mm. With some simple
modifications, SPDA is able to test longer embedment lengths. This should be done in
the future to expand the knowledge around load distribution of the longer cable bolts.
However, caution must be exercised so that the final pullout load does not exceed the
tensile strength of the tested cable, irrespective of the encapsulated wire length.
• The combination of hook cells and SPDA will allow axial testing of tendons under
different confinement pressures.
• Numerical models for dynamic pullout tests can be undertaken. Proposed friction
coefficients for the simplified model should also be calibrated in the dynamic mode.

291
References
ACARP. (2022). ACARP summary of current projects for the months November, December
2021 and January 2022 (Issue February).
https://www.acarp.com.au/Media/ACARPCurrentProjectsReport.pdf
ACI Committee 308. (1980). Standard Practice for Curing Concrete (ACI 308-92). American
Concrete Institute.
Acuity. (2022). AR550 High Speed Laser Sensor Principles of Operation AR550 Laser Position
Sensors Mechanical Dimensions units in mm. https://www.acuitylaser.com/product/laser-
sensors/short-range-sensors/ar550-high-speed-laser-sensor/
AMSJ. (2020). Mining fatality rate falls in 2020. Australian Mine Safety Journal.
https://www.amsj.com.au/mining-fatality-rate-falls-in-2020/
Ansell, A. (2005). Laboratory testing of a new type of energy absorbing rock bolt. Tunnelling
and Underground Space Technology, 20(4), 291–300.
https://doi.org/10.1016/j.tust.2004.12.001
ANSYS Inc. (1999). ANSYS Theory Reference - Release 5.6. In Theory Reference (Issue
November, pp. 1–1286).
http://research.me.udel.edu/~lwang/teaching/MEx81/ansys56manual.pdf
Aoki, T. A., Maeno, Y., Shibata, K., Obara, Y., & Shibat, K. (2002). Laboratory Pull-Out Tests
of Cablebolts. In ISRM International Symposium - EUROCK 2002 (pp. 599–606).
https://www.sciencedirect.com/science/article/pii/S167477551730001X
Aoki, T. A., Shibata, K., Nakahara, F., Maeno, Y., Kawano, R., & Obara, Y. (2003). Pull tests
of long-embedded cablebolts. 10th congress of the ISRM: Technology roadmap for rock
mechanics.
AS_1012.14. (2018). Australian Standards: Methods of testing concrete, Method for securing
and testing cores from hardened concrete for compressive strength and mass per unit
volume. Standards Australia.
AS_1012.17. (1997). Australian Standards: Methods of Testing Concrete Determination of
The Static Chord Modulus of Elasticity and Poiss. Standards Australia.
AS_1012.9. (1986). Australian Standards: Methods of testing concrete - Method for making
and curing concrete compression, indirect tensile and flexure test specimens, in the
laboratory or in the field. Standards Australia.
AS_1726. (2017). Australian Standards: Geotechnical site investigations (p. 72). Standards
Australia.
292
ASTM.F432. (2004). Standard Specification for Roof and Rock Bolts and Accessories (Vol.
08, Issue Reapproved). ASTM.
ASTM C900. (2013). American Standard for Testing Materials: Standard test method for
pullout strenght of hardened concrete. American Society for Testing and Materials West
Conshohocken, USA.
Aydan, O., Ichikawa, Y., & Kawamoto, T. (1985). Load bearing capacity and stress distribution
along rockbolts with inelastic behaviour of interfaces. 5th International Conference of
Numerical Methods in Geomechanics, 1281–1292.
Aziz, N. (2007). Experimental and numerical study of double shearing of bolt under
confinement. 26th International Conference on Ground Control in Mining, 242–249.
http://ro.uow.edu.au/engpapers/3927
Aziz, N. (2014). Status of Ground Support in Australian Coal Mines with Reference to the
Latest Research on Rock Bolting Technology. International Symposium on Earth Science
and Technology. P1-8.
Aziz, N., Anzanpour, S., Khaleghparat, S., Rastegarmanesh, A., Mirzaghorbanali, A.,
Remmenikov, A., Nemcik, J., Oh, J., & Si, G. (2021). Angled shear testing of 15.2 mm
seven wire cable bolt. Resource Operators Conference, February 2019, 18–20.
Aziz, N., Craig, P., Mirzaghorbanali, A., & Nemcik, J. (2016). Factors Influencing the Quality
of Encapsulation in Rock Bolting. Rock Mechanics and Rock Engineering, 49(8), 3189–
3203. https://doi.org/10.1007/s00603-016-0973-5
Aziz, N., Hillyer, J., Joyce, D., Ma, S., & Nemick, J. (2013). New approach to resin sample
preparation for strength testing. Coal Operators’ Conference, February, 152–155.
Aziz, N., Hossein, J., & Hadi, M. S. (2005). The Effect of Resin Thickness on Bolt-Grout-
Concrete Interaction in Shear. Coal Operators’ Conference, 3–9.
https://ro.uow.edu.au/cgi/viewcontent.cgi?article=1062&context=coal
Aziz, N., Jalalifar, H., & Concalves, J. (2006). Bolt surface configurations and load transfer
mechanism. Coal Operators Conference, 236–245.
Aziz, N., Majoor, D., & Mirzaghorbanali, A. (2017). Strength Properties of Grout for Strata
Reinforcement. ISRM European Rock Mechanics Symposium EUROCK 2017, 191, 1178–
1184. https://doi.org/10.1016/j.proeng.2017.05.293
Aziz, N., Mirza, A., Nemcik, J., Rasekh, H., & Li, X. (2016). A Follow up to Study the
Behaviour of Cable Bolts in Shear : Experimental Study and Mathematical Modelling.
Coal Operators’ Conference, February, 24–31.

293
Aziz, N., Mirzaghorbanali, A., Anzanpour, S., Rastegarmanesh, A., Khaleghparat, S., Nemcik,
J., Oh, J., & Si, G. (2022). Angle Shear Testing of 15.2 mm Seven Wire Cable Bolt. Rock
Mechanics and Rock Engineering, 55(7), 3919–3937. https://doi.org/10.1007/s00603-
022-02847-2
Aziz, N., Mirzaghorbanali, A., Nemcik, J., Heemann, K., & Mayer, S. (2015). Shear strength
properties of plain and spirally profiled cable bolts. Canadian Geotechnical Journal,
52(10), 1490–1495. https://doi.org/10.1139/cgj-2014-0504
Aziz, N., Mirzaghorbanali, A., & Nemick, J. (2016). Load Transfer Characteristics of Plain
and Spiral Cable Bolts Tested in New Non Rotating Pull Testing Apparatus. Coal
Operators’ Conference, February, 32–39.
Aziz, N., Nemick, J., Mirzaghorbanali, A., Foldi, S., & Joyce, D. (2014). Suggested methods
for the preparation and testing of various properties of resins and grouts. Coal Operators
Conference, February 2019, 18–20.
Aziz, N., Pratt, D., Williams, R., Aziz, N., Pratt, D., & Williams, R. (2003). Double Shear
Testing of Bolts. Coal Operators’ Conference, February, 12–14.
Aziz, N., Rasekh, H., Mirzaghorbanali, A., Yang, G., Khaleghparast, S., & Nemcik, J. (2018).
An Experimental Study on the Shear Performance of Fully Encapsulated Cable Bolts in
Single Shear Test. Rock Mechanics and Rock Engineering, 51(7), 2207–2221.
https://doi.org/10.1007/s00603-018-1450-0
Aziz, N., Rink, O., Rasekh, H., Hawkins, E., & Mirzaghorbanali, A. (2017). Single shear
testing of various cable bolts used in Australian mines. Coal Operators’ Conference,
February, 222–230. http://ro.uow.edu.au/coal/659
Aziz, N., & Webb, B. (2003). Study of Load Transfer Capacity of Bolts Using Short
Encapsulation Push Test. Coal Operators Conference, 72–80.
https://ro.uow.edu.au/coal/162/
Aziz, N., Yang, G., Khaleghparast, S., & Marshall, T. (2019). Development of double shear
testing of tendons. Coal Operators’ Conference, February, 150–161.
https://ro.uow.edu.au/coal/732/
Ballarini, R., Shah, S. P., & Keer, L. M. (1986). Failure characteristics of short anchor bolts
embedded in a brittle material. Proceedings of the Royal Society of London. A.
Mathematical and Physical Sciences, 404(1826), 35–54.
Banthia, N., Mindess, S., Bentur, A., & Pigeon, M. (1989). Impact testing of concrete using a
drop-weight impact machine. Experimental Mechanics, 29(1), 63–69.

294
https://doi.org/https://doi.org/10.1007/BF02327783
Barley, A. D., & Windsor, C. R. (2000). Recent advances in ground anchor and ground
reinforcement technology with reference to the development of the art. ISRM
International Symposium, 26.
Bawden, W. F., Hyett, A. J., & Lausch, P. (1992). An experimental procedure for the in situ
testing of cable bolts. International Journal of Rock Mechanics and Mining Sciences &
Geomechanics Abstracts, 29(5), 525–533. https://doi.org/https://doi.org/10.1016/0148-
9062(92)92635-P
Benmokrane, B., Chennouf, A., & Mitri, H. S. (1995). Laboratory evaluation of cement-based
grouts and grouted rock anchors. International Journal of Rock Mechanics and Mining
Sciences And, 32(7), 633–642. https://doi.org/10.1016/0148-9062(95)00021-8
Bertuzzi, R., & Pells, P. J. N. (2002). Geotechnical parameters of Sydney Sandstone and shale.
Australian Geomechanics Journal, 37(5), 41–54.
Bickley, J. A. (1993). Field manual for maturity and pullout testing on highway structures
(Issue SHRP-C-376). Strategic Highway Research Program. ISBN: 030905754X .
Bigby, D., & Reynolds, C. (2005). Development of the laboratory short encapsulation pull test
for a revised British standard on rock reinforcement components used in coal mining.
Proceedings of the 24th International Conference on Ground Control in Mining, 2–4.
Blanco Martín, L., Hadj-Hassen, F., Tijani, M., & Noiret, A. (2011). A new experimental and
analytical study of fully grouted rockbolts. 45th US Rock Mechanics Geomechanics
Symposium, February, 11.
Blanco Martín, L., Tijani, M., & Hadj-Hassen, F. (2011). A new analytical solution to the
mechanical behaviour of fully grouted rockbolts subjected to pull-out tests. Construction
and Building Materials, 25(2), 749–755.
https://doi.org/10.1016/j.conbuildmat.2010.07.011
Bohara, R. P., Tanapornraweekit, G., & Tangtermsirikul, S. (2019). Investigation of concrete
material models for analysis of seismic behavior of reinforced concrete under reversed
cyclic load. Songklanakarin Journal of Science and Technology, 41(4), 952–959.
https://doi.org/https://doi.org/10.14456/sjst-psu.2019.120
Bolstad, D. D., Hill, J. R. M., & Karhnak, J. M. (1984). Rock bolting: Theory and application
in mining and underground construction. In Rock Bolting (1st Editio, pp. 313–320).
Routledge.
Borrvall, T., & Riedel, W. (2011). The RHT concrete model in LS-DYNA. 8th European LS-

295
DYNA Users Conference, January. https://www.researchgate.net/publication/263467071
BS 12390-3:2002. (2002). British Standards: Testing hardened concrete- Part 3: Compressive
strength of test specimens. British Standards Institution.
BS 6319-1. (1984). British Standards: Testing of Resin and Polymer/cement Composition for
Use in Construction. British Standards Institution.
BS 7861-2. (2009). British Standards: Strata reinforcement support systems components used
in coal mines Specification for flexible systems for roof reinforcement. British Standards
Institution.
Cai, M. (2013). Principles of rock support in burst-prone ground. Tunnelling and Underground
Space Technology, 36, 46–56. https://doi.org/10.1016/j.tust.2013.02.003
Cai, Y., Esaki, T., & Jiang, Y. (2004). An analytical model to predict axial load in grouted rock
bolt for soft rock tunnelling. Tunnelling and Underground Space Technology, 19(6), 607–
618. https://doi.org/10.1016/j.tust.2004.02.129
Cao, C. (2012). Bolt profile configuration and load transfer capacity optimisation. PhD Thesis.
University of Wollongong.
Cao, C., Nemcik, J., Aziz, N., & Ren, T. (2013). Analytical study of steel bolt profile and its
influence on bolt load transfer. International Journal of Rock Mechanics and Mining
Sciences, 60, 188–195. https://doi.org/10.1016/j.ijrmms.2012.12.013
Cao, C., Ren, T., Cook, C., & Cao, Y. (2014). Analytical approach in optimising selection of
rebar bolts in preventing rock bolting failure. International Journal of Rock Mechanics
and Mining Sciences, 72, 16–25. https://doi.org/10.1016/j.ijrmms.2014.04.026
Chen, J. (2016). Load transfer mechanism of fully grouted cable bolts (Issue November). PhD
Thesis, University of New South Wales.
Chen, J., Hagan, P. C., & Saydam, S. (2017). Sample diameter effect on bonding capacity of
fully grouted cable bolts. Tunnelling and Underground Space Technology, 68(May), 238–
243. https://doi.org/10.1016/j.tust.2017.06.004
Chen, J., Saydam, S., & Hagan, P. C. (2015). An analytical model of the load transfer behavior
of fully grouted cable bolts. Construction and Building Materials, 101, 1006–1015.
https://doi.org/10.1016/j.conbuildmat.2015.10.099
Chen, J., Saydam, S., & Hagan, P. C. (2018). Numerical simulation of the pull-out behaviour
of fully grouted cable bolts. Construction and Building Materials, 191, 1148–1158.
https://doi.org/10.1016/j.conbuildmat.2018.10.083
Chen, J. Y., & Mitri, H. S. (2005). Shear bond characteristics in grouted cable bolts. 24th

296
International Conference on Ground Control in Mining, Lakeview Scanticon Resort and
Conference Center. P 342-348.
Chung, C. H., Lee, J., & Gil, J. H. (2014). Structural performance evaluation of a precast
prefabricated bridge column under vehicle impact loading. In Structure and Infrastructure
Engineering (Vol. 10, Issue 6, pp. 777–791). Taylor & Francis.
https://doi.org/10.1080/15732479.2013.767841
Clifford, B., Kent, L., Altounyan, P., & Bigby, D. (2001). Systems used in coal mining
development in long tendon reinforcement. 20th International Conference on Ground
Control in Mining, 7–9.
Clough, R. W., & Penzien, J. (1995). Dynamic of structures (3rd Edition). Computers &
Structures, Inc.
Coates, D. F., & Yu, Y. S. (1970). Three-dimensional stress distributions around a cylindrical
hole and anchor. In International Society of Rock Mechanics. International Society of
Rock Mechanics.
Compton, C., & Oyler, D. (2005). Investigation of Fully Grouted Roof Bolts Installed Under
In Situ Conditions. 24th International Conference on Ground Control in Mining, 302–
312.
Craig, P., & Aziz, N. (2010). Shear testing of 28 mm hollow strand “TG” cable bolt.
Proceedings - 29th International Conference on Ground Control in Mining, ICGCM,
169–174.
Craig, P., & Holden, M. (2014). In situ bond strength testing of Australian cable bolts. Coal
Operators’ Conference, February, 147–155.
Craig, P., & Murnane, B. (2013). In-situ pull testing of cable bolts encapsulated with injection
polyurethane. Coal Operators’ Conference, February, 131–136.
https://ro.uow.edu.au/coal/446/
Crompton, B., Berghorst, A., Knox, G., & Mining, N. C. (2018). A new dynamic test facility
for support tendons. Australian Centre for Geomechanics Newsletter, 47, 3–5.
https://www.ncm.co.za/downloads/papers_presentations/ACG_Newsletter_August2018.
pdf
Cybulski, J. A., & Mazzoni, R. A. (1989). Roof support systems continue to evolve. 12th
Annual Institute on Coal Mining Healthy, Safety and Research, 147, 158.
Delhomme, F., & Debicki, G. (2010). Numerical modelling of anchor bolts under pullout and
relaxation tests. Construction and Building Materials, 24(7), 1232–1238.

297
https://doi.org/10.1016/j.conbuildmat.2009.12.015
Diederichs, M. S., Pieterse, E., Nose, J., & Kaiser, P. K. (1993). A model for evaluating cable
bolt bond strength: an update. ISRM International Symposium-EUROCK 1993.
Dorsten, V., Hunt, F. F., & Preston, H. K. (1984). Epoxy coated seven-wire strand for
prestressed concrete. Prestressed Concrete Institute, 29(4), 120–129.
https://doi.org/10.15554/pcij.07011984.120.129
Dou, L. ming, Lu, C. ping, Mu, Z. long, & Gao, M. shi. (2009). Prevention and forecasting of
rock burst hazards in coal mines. Mining Science and Technology, 19(5), 585–591.
https://doi.org/10.1016/S1674-5264(09)60109-5
Droesch, D. (2015). Bonded Anchors in Concrete Under Sustained Loading (Issue July).
Masters Thesis. University of Massachusetts Amherst.
Dulacska, H. (1972). Dowel action of reinforcement crossing cracks in concrete. Journal
Proceedings, 69(12), 754–757.
Eaton, J. R., Calandra Jr, F., Oldsen, J. G., Jan, P., Eaton, J. R., Calandra Jr, F., & Oldsen, J.
G. (1998). United State Patents: US 5785463 Combination cable bolt system (Vol. 12,
Issue 19, pp. 2–4). Google Patents. https://patents.google.com/patent/US5785463
Elkome. (2018). Dytran Instruments 3200B Shock Accelerometer.
https://www.dytran.com/Model-3200Bshock-Accelerometer/
Fan, W., & Yuan, W. C. (2014). Numerical simulation and analytical modeling of pile-
supported structures subjected to ship collisions including soil-structure interaction.
Ocean Engineering, 91, 11–27. https://doi.org/10.1016/j.oceaneng.2014.08.011
Fanaradelli, T. D., & Rousakis, T. C. (2020). 3D finite element pseudodynamic analysis of
deficient rc rectangular columns confined with fiber reinforced polymers under axial
compression. Polymers, 12(11), 1–26. https://doi.org/10.3390/polym12112546
Farah, A., & Aref, K. (1986). Design Considerations of Fully Grouted Cable Bolts. 88th Annual
General Meeting of the CIM, Montreal, Canada, Canadian Instiute of Mining, Metallurgy
and Petroleum, 1–18.
Farmer, I. W. (1975). Stress distribution along a resin grouted rock anchor. International
Journal of Rock Mechanics and Mining Sciences And, 12(11), 347–351.
https://doi.org/10.1016/0148-9062(75)90168-0
Ferrero, A. M. (1995). The shear strength of reinforced rock joints. International Journal of
Rock Mechanics and Mining Sciences And, 32(6), 595–605. https://doi.org/10.1016/0148-
9062(95)00002-X

298
Forbes, B., & Vlachopoulos, N. (2016). A New Look At Cable Bolt and D-Bolt Support Under
Axial Load. 69th Canadian Geotechnical Conference, October 2016.
https://www.researchgate.net/publication/303988820%0AA
Fuller, P. G., & Cox, R. H. T. (1975). Mechanics of load transfer from steel tendons to cement
based grout. 5th Australasian Conference on the Mechanics of Structures and Materials,
189–203.
Garford Pty Ltd. (1990). Garford Bulb anchor, an improved, economical method for rock
stabilisation (pp. 1–4). Patent.
Gaudreau, D., Aubertin, M., & Simon, R. (2004). Performance assessment of tendon support
systems submitted to dynamic loading. Ground Support in Mining and Underground
Construction, June, 561–586. https://doi.org/10.1201/9780203023921-39
Ghadimi, M., Shahriar, K., & Jalalifar, H. (2015). A new analytical solution for the
displacement of fully grouted rock bolt in rock joints and experimental and numerical
verifications. Tunnelling and Underground Space Technology, 50, 143–151.
https://doi.org/10.1016/j.tust.2015.07.014
Ghadimi, M., Shariar, K., & Jalalifar, H. (2016). A new analytical solution for calculation the
displacement and shear stress of fully grouted rock bolts and numerical verifications.
International Journal of Mining Science and Technology, 26(6), 1073–1079.
https://doi.org/10.1016/j.ijmst.2016.09.016
Gilat, A., Kuokkala, V. T., Seidt, J. D., & Smith, J. L. (2017). Full-Field Measurement of Strain
and Temperature in Quasi-Static and Dynamic Tensile Tests on Stainless Steel 316L.
Procedia Engineering, 207, 1994–1999. https://doi.org/10.1016/j.proeng.2017.10.1057
Goris, J. M. (1990). Laboratory evaluation of cable supports, Part 1: Evaluation of Supports
Using Conventional Cables. In International Journal of Rock Mechanics and Mining
Sciences & Geomechanics Abstracts (Vol. 28, Issue 6). https://doi.org/10.1016/0148-
9062(91)91464-3
Goris, J. M. (1991). Laboratory evaluation of cable supports, Part 2: Evaluation of Supports
Using Conventional Cables With Steel Buttons, Birdcage Cables, and Epoxy- Coated
Cables. In International Journal of Rock Mechanics and Mining Sciences &
Geomechanics Abstracts (Vol. 28, Issue 6). States Department of the Interior, Bureau of
Mines. https://doi.org/10.1016/0148-9062(91)91464-3
Goris, J. M., & Conway, J. P. (1988). Grouted flexible tendons and scaling investigations.
International Journal of Rock Mechanics and Mining Sciences and Geomechanics

299
Abstracts, 25(6), 294.
Goris, J. M., Martin, L., & Curtin, R. (1996). Shear behavior of cable bolt supports in
horizontal, bedded deposits. International Conference on Ground Control in Mining,
89(1001). https://www.osti.gov/biblio/405595
Government of Western Australia. (2012). Fatal accidents in the Western Australian mining
industry 2000-2012.
http://www.dmp.wa.gov.au/documents/Reports/RP_FatalAccidentsMINING_2000-
12.pdf
Grasselli, G. (2005). 3D behaviour of bolted rock joints: Experimental and numerical study.
International Journal of Rock Mechanics and Mining Sciences, 42(1), 13–24.
https://doi.org/10.1016/j.ijrmms.2004.06.003
Guglielmetti, V., Grasso, P., Mahtab, A., & Xu, S. (2007). Mechanised tunneling in urban area,
design methodology and construction control. In V. Guglielmetti, P. Grasso, A. Mahtab,
& S. Xu (Eds.), Taylor & Francis (1st Editio). Taylor & Francis e-Library.
Güler, G., Kujipers, J. ., Wojno, L., Milev, A., & Haile, A. (2001). Determine the effect of
repeated dynamic loading on the performance of tunnel support systems (Issue March).
Internal Report.
Haas, C. J. (1976). Shear resistance of rock bolts. Trans. Soc. Min. Eng. AIME, 260(1).
Hadjigeorgiou, J., & Potvin, Y. (2007). Overview of dynamic testing of ground support. 4th
International Seminar on Deep and High Stress Mining, 349–371.
Hagan, P. (2004). Variation in load transfer of a fully encapsulated rockbolt. 23rd International
Conference on Ground Control in Mining, 242–249.
Hagan, P., & Chen, J. (2014). The load transfer mechanism of fully grouted cable bolts under
laboratory tests. Coal Operators Conference, February, 18–20.
https://ro.uow.edu.au/coal/507/
Hagan, P., Chen, J., Hebblewhite, B., Saydam, S., & Mitra, R. (2015). Optimising the Selection
of Fully Grouted Cable Bolts in Varying Geotechnical Environments. ACARP Project
C22010, February.
Hagan, P., & Li, D. (2017). Performance of cable bolts under axial loading subjected to varying
geotechnical conditions. In ACARP Project C24018.
Haile, A., Grave, D., Sevume, C., & Le Bron, K. (1998). Strata control in tunnels and an
evaluation of support units and systems currently used with a view to improving the
effectiveness of support, stability and safety of tunnels. Safety in Mines.

300
Hassani, F., Mitri, H., Khan, U., & Rajaie, H. (1992). Experimental and numerical studies of
the cable bolt support systems. In Rock support in mining and underground construction
(pp. 411–417). A.A. BALKEMA.
Hawkes, J. M., & Evans, R. H. (1995). Bond Stresses in Reinforced Concrete Columns and
Beams. In Journal of the Institute of Structural Engineers (Vol. 24, Issue 10). The
Institution of Structural Engineers.
He, L., An, X. M., & Zhao, Z. Y. (2015). Fully Grouted Rock Bolts: An Analytical
Investigation. Rock Mechanics and Rock Engineering, 48(3), 1181–1196.
https://doi.org/10.1007/s00603-014-0610-0
Heckötter, C., & Sievers, J. (2017). Comparison of the RHT Concrete Material Model in. 11th
European LS-DYNA Conference, May.
Hedrick, N. (2005). Australian patent, AU 2006201823 B2, Cable bolt. Patent.
Hibino, S., & Motojima, M. (1981). Effects of rock bolting in jointy rocks. ISRM International
Symposium, 1057–1062. https://www.onepetro.org/conference-paper/ISRM-IS-1981-171
Hoehn, K., Barnes, S., & Greyvensteyn, I. (2021). Carbolt Self Monitored, Yieldable Carbon
Fibre Cable Bolt for Ground Control. Patent.
Hoek, E., & Wood, D. F. (1988). Rock support. Report. Min. Mag., 159(4), 282–287.
Holden, M., & Hagan, P. (2014). The size effect of rock sample used in anchorage performance
testing of cable bolts. Coal Operators’ Conference, February, 128–136.
https://ro.uow.edu.au/coal/506/
Hu, G., Wu, J., & Li, L. (2016). Advanced Concrete Model in Hydrocode to Simulate Concrete
Structures under Blast Loading. Advances in Civil Engineering, 2016, 1–13.
https://doi.org/10.1155/2016/7540151
Huang, Y., Hu, S., Gu, Z., & Sun, Y. (2019). Fracture behavior and energy analysis of 3D
concrete mesostructure under uniaxial compression. Materials, 12(12).
https://doi.org/10.3390/ma12121929
Hudson, J. (2003). Engineering Rock mechanics. Book. University of london.
Hutchins, W. R., Bywater, S., Thompson, A. G., & Windsor, C. R. (1990). A versatile grouted
cable dowel reinforcing system for rock. AusIMM Proceedings, 25–29.
http://hdl.handle.net/102.100.100/256096?index=1
Hutchinson, D. J., & Diederichs, M. S. (2002). Cable Bolting Practices in Underground Mines.
Book. BiTech Publishers Ltd.
Hyett, A. J., Bawden, W. F., & Coulson, A. L. (1992). Physical and mechanical properties of

301
normal Portland cement pertaining to fully grouted cable bolts. International Symposium
on Rock Support, 341–348.
Hyett, A. J., Bawden, W. F., Macsporran, G. R., & Moosavi, M. (1995). A constitutive law for
bond failure of fully-grouted cable bolts using a modified hoek cell. International Journal
of Rock Mechanics and Mining Sciences And, 32(1), 11–36. https://doi.org/10.1016/0148-
9062(94)00018-X
Hyett, A. J., Bawden, W. F., Powers, R., & Rocque, P. (1993). The nutcase cable bolt.
Proceedings Innovative Mine Deisgn for the 21st Century: Proceedings of the
International Congress on Mine Design, Ontario, Canada, 409–419.
Hyett, A. J., Bawden, W. F., & Reichert, R. D. (1992). The effect of rock mass confinement on
the bond strength of fully grouted cable bolts. International Journal of Rock Mechanics
and Mining Sciences And, 29(5), 503–524. https://doi.org/10.1016/0148-9062(92)92634-
O
Indraratna, B., & Kaiser, P. K. (1990). Design for grouted rock bolts based on the convergence
control method. International Journal of Rock Mechanics and Mining Sciences And,
27(4), 269–281. https://doi.org/10.1016/0148-9062(90)90529-B
Interface force Measurments Ltd. (2022). 1000 FATIGUE RATED HIGH CAPACITY LOAD
CELL (Issue Product Bruchure, pp. 1–3). Bruchure. Interface force Measurments Ltd.
Itasca, Consulting Group, I. (2014). UDEC User’s Manual (v6.0.286). Manual. UDEC -
Universal Distinct Element Code.
ITASCA. (2011). Flac 3D Manual (Vol. 20, Issue 1). Manual. https://doi.org/10.1007/s10726-
010-9200-z
Ito, F., Nakahara, F., Kawano, R., Kang, S. S., & Obara, Y. (2001). Visualization of failure in
a pull-out test of cable bolts using X-ray CT. Construction and Building Materials, 15(5–
6), 263–270. https://doi.org/10.1016/S0950-0618(00)00075-1
Jaeger, J. C. (1971). Friction of Rocks and Stability of Rock Slopes. Geotechnique, 21(2), 97–
134. https://doi.org/10.1680/geot.1971.21.2.97
Jalalifar, H., & Aziz, N. (2010). Experimental and 3D numerical simulation of reinforced shear
joints. Rock Mechanics and Rock Engineering, 43(1), 95–103.
https://doi.org/10.1007/s00603-009-0031-7
Jalalifar, H., Aziz, N., & Hadi, M. (2006). The effect of surface profile, rock strength and
pretension load on bending behaviour of fully grouted bolts. Geotechnical and Geological
Engineering, 24(5), 1203–1227. https://doi.org/10.1007/s10706-005-1340-6

302
Jennmar. (2022). Cable Bolts 63T Sumo Cable Bolt. Product Brochure.
https://www.jennmar.com.au/assets/Uploads/Documents/Coal-Catalogue-Feb-
2021/MM-C-JCB28.pdf
Ji, C., Long, Y., & Wan, W. Q. (2005). Numerical simulation of projectile penetrating into
steel fiber reinforced concrete. Jiefangjun Ligong Daxue Xuebao (Journal of PLA
University of Science and Technology (Natural Science Edition), 6(5), 459–463.
Jiang, H., Wang, X., & He, S. (2012). Numerical simulation of impact tests on reinforced
concrete beams. Materials and Design, 39, 111–120.
https://doi.org/10.1016/j.matdes.2012.02.018
Jiang, N., & Zhou, C. (2012). Blasting vibration safety criterion for a tunnel liner structure.
Tunnelling and Underground Space Technology, 32, 52–57.
https://doi.org/10.1016/j.tust.2012.04.016
Kaewunruen, S. (2007). Experimental and numerical studies for evaluating dynamic behaviour
of prestressed concrete sleepers subject to severe impact loading. PhD Thesis. [University
of Wollongong]. http://ro.uow.edu.au/theses/277
Kaiser, P. K., & Cai, M. (2012). Design of rock support system under rockburst condition.
Journal of Rock Mechanics and Geotechnical Engineering, 4(3), 215–227.
https://doi.org/10.3724/sp.j.1235.2012.00215
Kaiser, P. K., McCreath, D. R., & Tannant, D. D. (1996). Canadian rockburst support
handbook. Book. Geomechanics Research Center, Mirarco.
Kaiser, P. K., Yazici, S., & Nosé, J. (1992). Effect of stress change on the bond strength of
fully grouted cables. International Journal of Rock Mechanics and Mining Sciences And,
29(3), 293–306. https://doi.org/10.1016/0148-9062(92)93662-4
Kang, H., Yang, J., Gao, F., & Li, J. (2020). Experimental Study on the Mechanical Behavior
of Rock Bolts Subjected to Complex Static and Dynamic Loads. Rock Mechanics and
Rock Engineering, 53(11), 4993–5004. https://doi.org/10.1007/s00603-020-02205-0
Kappos, A. . (2017). Dynamic Loading and Design of Structures. In Dynamic Loading and
Design of Structures. https://doi.org/10.4324/9780203301951
Khaleghparast, S. (2021). Investigation of the shear behaviour of rock bolts and tendons under
static and dynamic loading conditions (Issue March). PhD Thesis. University of
Wollongong.
Khaleghparast, S., Anzanpour, S., Aziz, N., Remennikov, A., & Mirzaghorbanali, A. (2020).
Static and dynamic testing of tendons. ISRM International Symposium-EUROCK 2020,

303
February 2019, 18–20.
Khan, U. (1994). Laboratory investigation of the steel cables and composite material tendons
for ground support. PhD Thesis, McGill University.
Kilic, A., Yasar, E., & Atis, C. D. (2002). Effect of bar shape on the pull-out capacity of fully-
grouted rockbolts. Tunnelling and Underground Space Technology, 18(1), 1–6.
https://doi.org/10.1016/S0886-7798(02)00077-9
Kilic, A., Yasar, E., & Celik, A. G. (2002). Effect of grout properties on the pull-out load
capacity of fully grouted rock bolt. Tunnelling and Underground Space Technology,
17(4), 355–362. https://doi.org/10.1016/S0886-7798(02)00038-X
Kohees, M., Sanjayan, J., & Rajeev, P. (2019). Stress-strain relationship of cement mortar
under triaxial compression. Construction and Building Materials, 220, 456–463.
https://doi.org/10.1016/j.conbuildmat.2019.05.146
Lang, T. A. (1961). Theory and practice of rock bolting. Book. Transactions of the American
Institute of Mining, Metallurgical and Petroleum Engineers, 220, 333–348.
Leeb, D. (1979). Dynamic hardness testing of metallic materials. NDT International, 12(6),
274–278. https://doi.org/10.1016/0308-9126(79)90087-7
Li, C. C., Hadjigeorgiou, J., Mikula, P., Knox, G., Darlington, B., Royer, R., Pytlik, A., &
Hosp, M. (2021). Performance of identical rockbolts tested on four dynamic testing rigs
employing the direct impact method. Journal of Rock Mechanics and Geotechnical
Engineering, 13(4), 745–754. https://doi.org/10.1016/j.jrmge.2021.01.003
Li, C. C., & Stillborg, B. (1999). Analytical models for rock bolts. International Journal of
Rock Mechanics and Mining Sciences, 36(8), 1013–1029. https://doi.org/10.1016/S1365-
1609(99)00064-7
Li, C. C., Stjern, G., & Myrvang, A. (2014). A review on the performance of conventional and
energy-absorbing rockbolts Journal of Rock Mechanics and Geotechnical Engineering A
review on the performance of conventional and energy-absorbing rockbolts. Journal of
Rock Mechanics and Geotechnical Engineering, July.
https://doi.org/10.1016/j.jrmge.2013.12.008
Li, D., Li, Y., & Zhu, W. (2020). Analytical Modelling of Load – Displacement Performance
of Cable Bolts Incorporating Cracking Propagation. Rock Mechanics and Rock
Engineering, 53(8), 3471–3483. https://doi.org/10.1007/s00603-020-02123-1
Li, D., Masoumi, H., Hagan, P. C., & Saydam, S. (2019). Experimental and analytical study
on the mechanical behaviour of cable bolts subjected to axial loading and constant normal

304
stiffness. International Journal of Rock Mechanics and Mining Sciences, 113(July 2018),
83–91. https://doi.org/10.1016/j.ijrmms.2018.11.011
Li, D., Masoumi, H., Saydam, S., Hagan, P. C., & Asadizadeh, M. (2018). Parametric study of
fully grouted cable bolts subjected to axial loading. Canadian Geotechnical Journal,
12(November), 1–12. https://doi.org/10.1139/cgj-2018-0470
Li, X. (2016). Study of cable bolt shear strength characteristics for ground reinforcement in
mines. , PhD Thesis, University of Wollongong.
Li, X. (2019). Load transfer characteristics of a cable bolt in DEPT. Shock and Vibration.
https://doi.org/10.1155/2019/2106741
Li, X., Aziz, N., Mirzaghorbanali, A., & Nemcik, J. (2017). Comparison of the shear test results
of a cable bolt on three laboratory test apparatuses. Tunnelling and Underground Space
Technology, 61, 82–89. https://doi.org/10.1016/j.tust.2016.10.003
Li, X., Nemcik, J., Mirzaghorbanali, A., Aziz, N., & Rasekh, H. (2015). Analytical model of
shear behaviour of a fully grouted cable bolt subjected to shearing. International Journal
of Rock Mechanics and Mining Sciences, 80, 31–39.
https://doi.org/10.1016/j.ijrmms.2015.09.005
Liu, C. H., & Li, Y. Z. (2017). Analytical Study of the Mechanical Behavior of Fully Grouted
Bolts in Bedding Rock Slopes. Rock Mechanics and Rock Engineering, 50(9), 2413–2423.
https://doi.org/10.1007/s00603-017-1244-9
LS-DYNA. (2018). LS-DYNA Keyword User’s Manual: Vol. II. Manual. Livermore Software
Technology Corporation (LSTC).
Lu, T. (1990). Anchorage behaviour of underground rockbolts. Masters Thesis. University of
Wollongong.
Ma, S. (2014). Load Transfer mechanism of fully encapsulated rockbolts. PhD Thesis.
Univrsity of Wollongong.
Ma, S., Nemcik, J., Aziz, N., & Zhang, Z. (2016). Numerical modeling of fully grouted
rockbolts reaching free-end slip. International Journal of Geomechanics, 16(1), 1–7.
https://doi.org/10.1061/(ASCE)GM.1943-5622.0000484
Ma, S., Zhu, X., Qin, W., & Hu, S. (2018). Determination of the bond–slip relationship of fully
grouted rockbolts. Environmental Earth Sciences, 77(9), 1–12.
https://doi.org/10.1007/s12665-018-7495-2
MacSporran, G. (1993). An empirical investigation into the effects of mine induced stress
change on standard cable bolt capacity. Masters Thesis. Queen’s University at Kingston.

305
https://elibrary.ru/item.asp?id=5736562
Magallanes, J. M., Wu, Y., Malvar, L. J., & Crawford, J. E. (2010). Recent improvements to
release III of the K&C concrete model. 11th International LS-DYNA User Conference, 1,
37–48. https://www.dynalook.com/international-conf-2010/Simulation-1-4.pdf
Mah, G. P. (1994). The development of a fibreglass cable bolt (Issue May). Masters Thesis,
University of British Columbia.
Malvar, L. J., Crawford, J. E., Facilities, N., Service, E., Hueneme, P., & Engineers, S. (1998).
Dynamic Increase Factors. 28th DoD Explosives Safety Seminar (DDESB), 1–17.
Malvar, L. J., Crawford, J. E., Wesevich, J. W., & Simons, D. (1997). A plasticity concrete
material model for DYNA3D. International Journal of Impact Engineering, 19(9–10),
847–873. https://doi.org/10.1016/s0734-743x(97)00023-7
Malvar, L. J., & Simons, D. (1996). Concrete Material Modeling in Explicit Computations.
Workshop on Recent Advances in Computational Structural Dynamics and High
Performance Computing, 1–30.
Mark, C. (2018). Coal bursts that occur during development: A rock mechanics enigma.
International Journal of Mining Science and Technology, 28(1), 35–42.
https://doi.org/10.1016/j.ijmst.2017.11.014
Markovich, N., Kochavi, E., & Ben-Dor, G. (2009). Calibration of a Concrete Damage Material
Model in LS-Dyna for a Wide Range of Concrete Strengths. IWSRIB_2009_Haifa, April
2015. https://doi.org/10.13140/RG.2.1.3503.4723
Markovich, N., Kochavi, E., & Ben-Dor, G. (2011). An improved calibration of the concrete
damage model. Finite Elements in Analysis and Design, 47(11), 1280–1290.
https://doi.org/10.1016/j.finel.2011.05.008
Martin, L. A., Dreesbach, C. A., & Pakalnis, R. (2000). Evaluation Of Ground Support At A
Trona Mine Using Instrumented Cable And Rebar Bolts. New Technology for Coal Mine
Roof Support, 271–280. https://stacks.cdc.gov/view/cdc/8315
Martin, L. B. (2012). Theoretical and experimental study of fully grouted rockbolts and
cablebolts under axial loads. PhD Thesis, Ecole Nationale Supérieure des Mines de Paris.
Mazaira, A., & Konicek, P. (2015). Intense rockburst impacts in deep underground
construction and their prevention. Canadian Geotechnical Journal, 52(10), 1426–1439.
https://doi.org/10.1139/cgj-2014-0359
Mckenzie, R., & King, B. (2015). Megabolt shear testing program. Presentation. Presentation
to NUGS and BBUGS, 25-26 February.

306
Meloni, D., De Nicolo, B., & Valdes, M. (2013). Finite element model of the pull-Out test for
concrete strength evaluation. Civil-Comp Proceedings, 102(March).
https://doi.org/10.4203/ccp.102.167
Miller, O., Freund, L. B., & Needleman, A. (1999). Energy dissipation in dynamic fracture of
brittle materials. Modelling and Simulation in Materials Science and Engineering, 7(4),
573–586. https://doi.org/10.1088/0965-0393/7/4/307
Minova. (2020). Lokset Anchor Pack Polyester resin anchor grout.
https://www.minovaglobal.com/media/2066/tds_-lokset-anchor-pack_feb2019_-
emea_afr-_en.pdf
Mirarco, G., Innovation, M., Compa, C., Compa, S., & Ingenier, T. (2018). Numerical
modelling of dynamic testing for rock reinforcement used in underground excavations.
4th International Symposium on Block and Sublevel Caving, 767–780.
https://doi.org/https://doi.org/10.36487/ACG_rep/1815_60_Marambio
Mirza, A., Aziz, N., Ye, W., & Nemcik, J. (2016). Mechanical properties of grouts at various
curing times. February, 84–90.
http://ro.uow.edu.au/cgi/viewcontent.cgi?article=2258&context=coal
Mirzaghorbanali, A. (2019). Strength properties of grout for strata reinforcement. Coal
Operators’ Conference, February, 196–202. https://ro.uow.edu.au/coal/641/
Mirzaghorbanali, A., & Nemcik, J. A. (2013). Numerical modelling of cyclic shear behaviour
of rock joints under under constant normal stiffness condition. 406–413.
Mirzaghorbanali, A., Rasekh, H., Aziz, N., Yang, G., Khaleghparast, S., Nemcik, J., Hadi, M.,
& Remennikov, A. (2017). Shear strength properties of cable bolts using a new double
shear instrument, experimental study, and numerical simulation. Tunnelling and
Underground Space Technology, 70(October 2016), 240–253.
https://doi.org/10.1016/j.tust.2017.07.018
Mogi, K. (2006). Experimental rock mechanics. Book. Taylor & Francis/BalkemaIn
Experimental Rock Mechanics (Vol. 3). CRC Press.
https://doi.org/10.1201/9780203964446
Moosavi, M., Jafari, A., & Khosravi, A. (2005). Bond of cement grouted reinforcing bars under
constant radial pressure. Cement and Concrete Composites, 27(1), 103–109.
https://doi.org/10.1016/j.cemconcomp.2003.12.002
Morsy, K., Han, J., Khair, W., & Peng, S. (2004). 3D FEM simulation for fully grouted bolts.
23rd International Conference on Ground Control in Mining, 273–277.

307
Murray, Y., Abu-Odeh, A., & Bligh, R. (2007). Evaluation of LS-DYNA Concrete Material
Model 159. In Report No. FHWA-HRT-05-063 (Issue May).
Mutke, G., Dubiński, J., & Lurka, A. (2015). New Criteria to Assess Seismic and Rock Burst
Hazard in Coal Mines. Archives of Mining Sciences, 60(3), 743–760.
https://doi.org/10.1515/amsc-2015-0049
Mylrea, T. D. (1948). Bond and anchorage. Journal Proceedings, 44(3), 521–552.
Nemcik, J., Ma, S., Aziz, N., Ren, T., & Geng, X. (2014). Numerical modelling of failure
propagation in fully grouted rock bolts subjected to tensile load. International Journal of
Rock Mechanics and Mining Sciences, 71, 293–300.
https://doi.org/10.1016/j.ijrmms.2014.07.007
Neville, A. ., & Brooks, J. . (2010). Concrete Technology (2nd Editio). Book. Pearson
Education Limited.
Nie, W., Zhao, Z. Y., Ning, Y. J., & Guo, W. (2014). Numerical studies on rockbolts
mechanism using 2D discontinuous deformation analysis. Tunnelling and Underground
Space Technology Incorporating Trenchless Technology Research, 41, 223–233.
https://doi.org/10.1016/j.tust.2014.01.001
Nilson, A. H. (1972). Internal Measurement of Bond Slip. ACI Journal Proceedings, 69(7),
439–441. https://doi.org/10.14359/7170
Oland, C. B., & Callahan, J. P. (1978). Bond Between Concrete and Steel Reinforcement at
Temperatures to 149 ° C (No. ORNL/TM_6086).
https://inis.iaea.org/search/search.aspx?orig_q=RN:9401294
Ortlepp, W. D., & Erasmus, P. N. (2005). Dynamic testing of a yielding cable anchor. 3rd
Southern African Rock Engineering Symposium.
Ortlepp, W. D., & Stacey, T. R. (1994). Rockburst mechanisms in tunnels and shafts.
Tunnelling and Underground Space Technology Incorporating Trenchless, 9(1), 59–65.
https://doi.org/10.1016/0886-7798(94)90010-8
Ortlepp, W. D., & Stacey, T. R. (1997). Testing of tunnel support dynamic load testing of rock
support containment systems. Report. Project No. GAP221. Safety In Mines Research
Advisory Committee.
Ortlepp, W. D., & Stacey, T. R. (2000). Performance of tunnel support under large deformation
static and dynamic loading. Comptes Rendus de l’Academie de Sciences - Serie IIa:
Sciences de La Terre et Des Planetes, 331(11), 15–21. https://doi.org/10.1016/S0886-
7798(98)00022-4

308
Ortlepp, W. D., & Swart, A. H. (2002). Extended use of the Savuka dynamic test facility to
improve material and analytical technology in deep-level stope support. In Report No.
GAP 818 (Issue November).
Ottosen, N. S. (1981). Nonlinear finite element analysis of pull-out test. Journal of the
Structural Division, 107(4), 591–603.
Pariseau, W. G. (2006). Design Analysis in Rock Mechanics. In Design Analysis in Rock
Mechanics. https://doi.org/10.1201/9780203968253
Paterson, M. S. (1978). Experimental Rock Deformation-The Brittle Field. Report.
Mineralogical Magazine, 43(326), 317.
Pinazzi, P. C., Spearing, A. J. S. (Sam., Jessu, K. V., Singh, P., & Hawker, R. (2021). Combined
Load Failure Criterion for Rock Bolts in Hard Rock Mines. Mining, Metallurgy and
Exploration, 38(1), 427–432. https://doi.org/10.1007/s42461-020-00289-4
Pitrakkos, T., Tizani, W., & Wang, Z. (2010). Pull-out behaviour of anchored blind-bolt: A
component based approach. International Conference on Computing in Civil and Building
Engineering, September. https://www.researchgate.net/publication/272816762%0APull-
out
Player, J. R. (2012). Dynamic Testing of Rock Reinforcement Systems (Issue April). PhD
Thesis. Curtin University of Technology.
Player, J. R., & Cordova, M. (2009). Dynamic Threadbar Test at WASM. In Internal report
for CODELCO-El Teniente Mine. Western Australian School of Mine.
Player, J. R., Thompson, A. G., & Villaescusa, E. (2009). Dynamic Testing of Threadbar used
for Rock Reinforcement. 3rd CANUS Rock Mechanics Symposium, 1–14.
Player, J. R., & Villaescusa, E. (2008). An Examination of Dynamic Test Facilities. Australian
Mining Technology Conference, December.
https://www.researchgate.net/publication/265566204%0AAN
Player, J., Villaescusa, E., & Thompson, A. G. (2008, September). An Examination of
Dynamic Test Facilities'. In Australian Mining Technology Conference. Melbourne:
Australasian Institute of Mining and Metallurgy.
Player, J. R., Villaescusa, E., & Thompson, A. G. (2004). Dynamic testing of rock
reinforcement using the momentum transfer concept. Ground Support in Mining and
Underground Construction, December 2014, 614–642.
https://doi.org/10.1201/9780203023921-42
Player, J. R., Villaescusa, E., & Thompson, A. G. (2013). Dynamic testing of fully encapsulated

309
threaded bar–resin and cement grouted. 7th International Symposium on Ground Support
in Mining and Underground Construction, 247–264.
Plouffe, M., Anderson, T., & Judge, K. (2008). Rock Bolt Testing Under Dynamic Conditions
at CANMET-MMSL. Proc. 6th Int. Symp. on Ground Support in Mining and Civil
Engineering Construction, Cape Town, S. Afr. Inst. Min. Metall. Symposium Series S, 51,
581–595.
Potvin, Y., Milne, D., & Gendron, A. (1989). Cable bolt research project. In NTC internal
report, WASM.
Pritchard, C. J., & McClellan, R. S. (2011). Injectable resin-a highly adaptable ground support
anchoring system. Canadian Institute of Mining Metallurgy and Petroleum Conference
and Exhibition.
Pullan, E., & Hagan, P. C. (2018). Comparison of the Performance of Resin and Cementitious
Grouting Media for Cable Bolts. Coal Operators’ Conference, February, 18–20.
https://ro.uow.edu.au/cgi/viewcontent.cgi?article=2378&context=coal
Pytlik, A. (2020). Comparative Shear Tests of Bolt Rods under Static and Dynamic Loading.
Studia Geotechnica et Mechanica, 42(2), 151–167. https://doi.org/10.2478/sgem-2019-
0038
Quanji, Z. (2010). Thixotropic behavior of cement-based materials : effect of clay and cement
types. Masters Thesis, Iowa State University.
Rajaie, H. (1990). Experimental and numerical investigations of cable bolt support systems
(Issue June) [McGill University]. internal-pdf://117.243.230.217/Experimental and
numerical investigations of
c.pdf%0Ahttp://digitool.library.mcgill.ca/webclient/StreamGate?folder_id=0&dvs=1566
756605269~216
Rajapakse, R. (2008). Rock Bolts, Dowels, and Cable Bolts. Geotechnical Engineering
Calculations and Rules of Thumb, 303–320. https://doi.org/10.1016/b978-075068764-
5.50025-8
Rasekh, H., Aziz, N., Mirza, A., Nemcik, J., Li, X., Yang, G., & Khaleghparast, S. (2017).
Double Shear Testing of Cable Bolts with No Concrete Face Contacts. Procedia
Engineering, 191, 1169–1177. https://doi.org/10.1016/j.proeng.2017.05.292
Rasekh, H., Mirzaghorbanali, A., Nemcik, J., Aziz, N., & Li, X. (2017). A New Equation for
the Shear Strength of Cable Bolts Incorporating the Energy Balance Theory. Geotechnical
and Geological Engineering, 35(4), 1529–1548. https://doi.org/10.1007/s10706-017-

310
0191-2
Rastegarmanesh, A. (2022). Axial Performance of Cable Bolts Under Various Bonding And
Loading Conditions. PhD Thesis, University of Southern Queensland.
Rastegarmanesh, A., Mirzaghorbanali, A., McDougall, K., Aziz, N., Anzanpour, S.,
Nourizadeh, H., & Moosavi, M. (2022). Axial Performance of Cementitious Grouted
Cable Bolts Under Rotation Constraint Scenarios. Rock Mechanics and Rock Engineering,
55(9), 5773–5788. https://doi.org/10.1007/s00603-022-02950-4
Ren, F. F., Yang, Z. J., Chen, J. F., & Chen, W. W. (2010). An analytical analysis of the full-
range behaviour of grouted rockbolts based on a tri-linear bond-slip model. Construction
and Building Materials, 24(3), 361–370.
https://doi.org/10.1016/j.conbuildmat.2009.08.021
Renwick, M. T. (1992). Ultrastrand–an innovative cable bolt. 5th Underground Operators’
Conference, Australasian Institute of Mining and Metallurgy, 151–156.
Riedel, W. (2000). Beton unter dynamischen Lasten: Meso-und makromechanische Modelle
und ihre Parameter. Masters Thesis. Ernst Mach Institute.
Riedel, W., Thoma, K., Hiermaier, S., & Schmolinske, E. (1999). Penetration of reinforced
concrete by BETA-B-500 numerical analysis using a new macroscopic concrete model
for hydrocodes. 9th International Symposium on the Effects of Munitions with Structures,
315.
Saadat, M. (2019). The Performance of Fully Grouted Rock Bolts Subjected to Combined Pull
and Shear Loads Under Constant Normal Stiffness Condition (Issue August). PhD Thesis,
University of Adelaide.
Saanouni, K. (2003). Numerical modeling in damage mechanics. In Engineering Fracture
Mechanics (1st Editio, Vol. 3). Hermes Science Publications and Kogan Page Limited.
Salamon, J., & Harris, D. (2014). Evaluation of Nonlinear Material Models in Concrete Dam
Finite Element Analysis. In Report No.: DSO-2014-08, for U.S. Department of the
Interior, Bureau of Reclamation (Issue September).
Salcher, M., & Bertuzzi, R. (2018). Results of pull tests of rock bolts and cable bolts in Sydney
sandstone and shale. Tunnelling and Underground Space Technology, 74(January), 60–
70. https://doi.org/10.1016/j.tust.2018.01.004
Satola, I. (1999). Testing of the yielding cable bolt achieved by debonding. 9th ISRM Congress,
290.
Satola, I. (2007). The axial load-displacement behavior of steel strands used in rock

311
reinforcement. In Engineering. Helsinki University of Technology.
Satola, I., & Aromaa, J. (2003). Corrosion of rock bolts and the effect of corrosion protection
on the axial behavior of cable bolts. 10th ISRM Congress.
Satola, I., & Hakala, M. (2001). Corrosion protected cable bolts in long-term reinforcement. In
Research project, part I. Report.
Schmuck, C. H. (1979). Cable bolting at the Homestake gold mine. Mining Engineering,
31(12), 1677–1681.
Schwer, L. E., & Malvar, L. J. (2005). Simplified Concrete Modeling With
*Mat_Concrete_Damage_Rel3. Ls-Dyna User Week 2005, 49–60.
Shang, J., Yokota, Y., Zhao, Z., & Dang, W. (2018). DEM simulation of mortar-bolt interface
behaviour subjected to shearing. Construction and Building Materials, 185, 120–137.
https://doi.org/10.1016/j.conbuildmat.2018.07.044
Sharifzadeh, M., Lou, J., & Crompton, B. (2020). Dynamic performance of energy-absorbing
rockbolts based on laboratory test results. Part I: Evolution, deformation mechanisms,
dynamic performance and classification. Tunnelling and Underground Space Technology,
105(June), 103510. https://doi.org/10.1016/j.tust.2020.103510
Singh, D., Hagan, P., & Li, D. (2017). Variation in anchorage performance of a high capacity
modified bulb cable bolt under differing conditions in a weak confining medium.
February, 18–20.
Singh, P., Jang, H., & Spearing, A. J. S. S. (2022). Improving the Numerical Modelling of In-
Situ Rock Bolts Using Axial and Bending Strain Data from Instrumented Bolts.
Geotechnical and Geological Engineering, 40(5), 2631–2655.
https://doi.org/10.1007/s10706-022-02051-7
Singh, P., & Spearing, A. J. S. (2021). An Improved Analytical Model for the Elastic and
Plastic Strain-hardening Shear Behaviour of Fully Grouted Rockbolts. Rock Mechanics
and Rock Engineering, 54(8), 3909–3925. https://doi.org/10.1007/s00603-021-02439-6
Singh, P., Spearing, A. J. S., & Jessu, K. (2020). Analysis of the Combined Load Behaviour of
Rock Bolt Installed Across Discontinuity and Its Modelling Using FLAC3D.
Geotechnical and Geological Engineering, 38(6), 5867–5883.
https://doi.org/10.1007/s10706-020-01398-z
Singh, P., Spearing, A. J. S., Jessu, K. V., & Caroline Pinazzi da Silva Ribeiro, P. (2020).
Establishing the need to model the actual state of stress along rock bolts. International
Journal of Mining Science and Technology, 30(3), 279–286.

312
https://doi.org/10.1016/j.ijmst.2020.03.002
Solomos, G., & Berra, M. (2010). Rebar pullout testing under dynamic Hopkinson bar induced
impulsive loading. Materials and Structures/Materiaux et Constructions, 43(1–2), 247–
260. https://doi.org/10.1617/s11527-009-9485-z
Spang, K., & Egger, P. (1990). Action of fully-grouted bolts in jointed rock and factors of
influence. Rock Mechanics and Rock Engineering, 23(3), 201–229.
https://doi.org/10.1007/BF01022954
St-pierre, L. (2007). Development and Validation of a Dynamic Model for a Cone Bolt
Anchoring System. Masters Thesis, McGill University.
Stacey, T. R., & Ortlepp, W. D. (1999). Retainment support for dynamic events in mines. In
Rock support and reinforcement practice in mining (pp. 329–333). Routledge.
Stankus, J. (2001). United States Patent, US 6270290 B1, Tensionable cable Bolt. Patent.
Stillborg, B. (1984). Experimental investigation of steel cables for rock reinforcement in hard
rock. Masters Thesis. Lulea University. https://doi.org/10.1016/0148-9062(85)92556-2
Stone, W. C., & Carino, N. J. (1984). Comparison of analytical with experimental internal
strain distribution for the pullout test. Journal Proceedings, 81(1), 3–12.
Stone, W. C., Carino, N. J., & Reeve, C. P. (1986). Statistical methods for in-place strength
predictions by the pullout test. Journal Proceedings, 83(5), 745–756.
Tadolini, S. C., Tinsley, J., & McDonnell, J. P. (2012). The next generation of cable bolts for
improved ground support. 31st International Conference on Ground Control in Mining,
1–7.
Tadolini, S., Derycke, S., & Bhagwat, A. (2016). Characterization of tensile and shear loading
on indented PC-strand cable bolts. International Journal of Mining Science and
Technology, 26(1), 89–95. https://doi.org/10.1016/j.ijmst.2015.11.015
Tahmasebinia, F., Zhang, C., Canbulat, I., Vardar, O., & Saydam, S. (2018a). A numerical
study of the behaviour of fully grouted cable bolts under static and dynamic loading. ISRM
International Symposium - 10th Asian Rock Mechanics Symposium, ARMS 2018,
November.
Tahmasebinia, F., Zhang, C., Canbulat, I., Vardar, O., & Saydam, S. (2018b). Numerical and
analytical simulation of the structural behaviour of fully grouted cable bolts under
impulsive loading. International Journal of Mining Science and Technology, 28(5), 807–
811. https://doi.org/10.1016/j.ijmst.2018.08.012
Tahmasebinia, F., Zhang, C., Wei, C., Canbulat, I., Saydam, S., & Sepasgozar, S. (2021). A

313
new concept to design combined support under dynamic loading using numerical
modelling. Tunnelling and Underground Space Technology, 117(August), 104132.
https://doi.org/10.1016/j.tust.2021.104132
Takazawa, H., Hirosaka, K., Miyazaki, K., Tohyama, N., Saigo, S., & Matsumoto, N. (2018).
Numerical simulation of impact loading for reinforced concrete wall. International
Journal of Pressure Vessels and Piping, 167(October), 66–71.
https://doi.org/10.1016/j.ijpvp.2018.10.010
Tan, S. H., Chan, R., Poon, J. K., & Chng, D. (2014). Verification of Concrete Material Models
for MM-ALE Simulations. 13th International LS-DYNA Users Conference, 16, 1–15.
Tannant, D. D., Brummer, R. K., & Yi, X. (1995). Rockbolt behaviour under dynamic loading:
Field tests and modelling. International Journal of Rock Mechanics and Mining Sciences
And, 32(6), 537–550. https://doi.org/10.1016/0148-9062(95)00024-B
Tassios, T. P. (1980). Properties of bond between concrete and steel under load cycles
idealizing seismic actions. Comite´ Euro-International Du Be´ton, 131.
Technology, N. I. (2022). The MEMRECAM HX series : the MEMRECAM HX-7. Nac Image
Technology. https://www.nacinc.com/products/memrecam-high-speed-digital-
cameras/hx-7s/
Thenevin, I., Blanco-martín, L., Hadj-hassen, F., Schleifer, J., Lubosik, Z., & Wrana, A.
(2017). Laboratory pull-out tests on fully grouted rock bolts and cable bolts: Results and
lessons learned. Journal of Rock Mechanics and Geotechnical Engineering, 9(5), 843–
855. https://doi.org/10.1016/j.jrmge.2017.04.005
Thilakarathna, H. M. I., Thambiratnam, D. P., Dhanasekar, M., & Perera, N. (2010). Numerical
simulation of axially loaded concrete columns under transverse impact and vulnerability
assessment. International Journal of Impact Engineering, 37(11), 1100–1112.
https://doi.org/10.1016/j.ijimpeng.2010.06.003
Thomas, R. (2012). The load transfer properties of post-groutable cable bolts used in the
Australian coal industry. 31st International Conference on Ground Control in Mining, 1–
10.
Thompson, A. G., Player, J. R., & Villaescusa, E. (2015). Simulation and analysis of
dynamically loaded reinforcement systems. September 2004.
Thompson, A. G., & Villaescusa, E. (2014). Case studies of rock reinforcement components
and systems testing. Rock Mechanics and Rock Engineering, 47(5), 1589–1602.
https://doi.org/10.1007/s00603-014-0583-z

314
Thompson, A. G., Villaescusa, E., & Windsor, C. R. (2012). Ground support terminology and
classification: An update. Geotechnical and Geological Engineering, 30(3), 553–580.
https://doi.org/10.1007/s10706-012-9495-4
Tincelin, E., & Fine, J. (1991). Mémento du boulonnage. Presses de l’École Des Mines, Paris.
TML. (2020). TML Strain Gauges Catalogue. Tokyo Measuring Instruments Lab.
Tu, Z., & Lu, Y. (2009). Evaluation of typical concrete material models used in hydrocodes for
high dynamic response simulations. International Journal of Impact Engineering, 36(1),
132–146. https://doi.org/10.1016/j.ijimpeng.2007.12.010
Vallejos, J. A., Marambio, E., Burgos, L., & Gonzalez, C. V. (2020). Numerical modelling of
the dynamic response of threadbar under laboratory-scale conditions. Tunnelling and
Underground Space Technology, 100(June 2019), 103263.
https://doi.org/10.1016/j.tust.2019.103263
Villaescusa, E., Thompson, A., & Player, J. (2007). Dynamic Testing of Rock Reinforcement
Systems Dynamic Testing of Rock Reinforcement Systems. Australian Mining
Technology Conference, October.
Villaescusa, E., Thompson, A., & Player, J. (2013). A decade of ground support research at
the WA School of Mines. 233–245. https://doi.org/10.36487/acg_rep/1304_15_villaescusa
Villaescusa, E., Thompson, A., & Player, J. R. (2005). Dynamic Testing of Rock
Reinforcement Systems. Australian Mining Technology Conference, October, 79–95.
Vos, E., & Reinhardt, H. W. (1982). Influence of loading rate On bond behaviour of reinforcing
steel and prestressing strands. Matériaux et Construction, V15, 3-10.
Wagner, H. (1984). Support requirements for rockburst conditions. Rockbursts and Seismicity
in Mines, 6(November), 209–218.
Wang, S., Yan, L., Liang, Q., & Xu, M. (2017). Research on RHT constitutive model
parameters of fiber reinforced concrete based on experiment and numerical simulation.
135(Iccte), 185–192. https://doi.org/10.2991/iccte-17.2017.31
Wang, W., Song, Q., Xu, C., & Gong, H. (2018). Mechanical behaviour of fully grouted GFRP
rock bolts under the joint action of pre-tension load and blast dynamic load. Tunnelling
and Underground Space Technology, 73(December 2016), 82–91.
https://doi.org/10.1016/j.tust.2017.12.007
Weathersby, J. (2003). Investigation of bond slip between concrete and steel reinforcement
under dynamic loading conditions. Development, May, 263.
http://etd.lsu.edu/docs/available/etd-0321103-

315
171804/%5Cnftp://134.147.28.62/Concrete-frc/Weathersby_bondslip-dynamics.pdf
Weckert, S. (2003). Anchorage and encapsulation failure mechanisms of rockbolts. Masters
Thesis, University of New South Wales.
Whyatt, J, W. Blake, T. W. (2002). 60 years of rockbursting in the Coeur d’Alene District of
Northern Idaho, USA: lessons learned and remaining issues. 109th Annual Exhibit and
Meeting, Society for Mining, Metallurgy, and Exploration, 10.
Windsor, C. (1992). Cable bolting for underground and surface excavations. Rock Support in
Mining and Underground Construction. In International symposium on rock support (pp.
349-366).
Windsor, C. R. (1997). Rock reinforcement systems. International Journal of Rock Mechanics
and Mining Sciences, 34(6), 919–951. https://doi.org/10.1016/S1365-1609(97)80004-4
Wu, Y., Gao, F., Chen, J., & He, J. (2019). Experimental Study on the Performance of Rock
Bolts in Coal Burst-Prone Mines. Rock Mechanics and Rock Engineering, 52(10), 3959–
3970. https://doi.org/10.1007/s00603-019-01794-9
Wu, Z., Yang, S., Zheng, J., & Hu, X. (2010). Analytical solution for the pull-out response of
FRP rods embedded in steel tubes filled with cement grout. Materials and
Structures/Materiaux et Constructions, 43(5), 597–609. https://doi.org/10.1617/s11527-
009-9515-x
Wyllie, D. C., & Mah, C. W. (2017). Rock slope engineering: Civil and mining, 4th edition.
Rock Slope Engineering: Fourth Edition, 1–432. https://doi.org/10.1201/9781315274980
Xiao, S., & Chen, C. (2008). Mechanical mechanism analysis of tension type anchor based on
shear displacement method. Journal of Central South University of Technology, 15(1),
106–111. https://doi.org/10.1007/s11771
Xu, C., Heping, C., Bin, L., & Fangfang, Z. (2011). Modeling of anchor bolt pullout in concrete
based on a heterogeneous assumption. Nuclear Engineering and Design, 241(5), 1345–
1351. https://doi.org/10.1016/j.nucengdes.2011.01.025
Yan, C. (1992). Bond Between Reinforcing Bars and Concrete under Impact Loading (Issue
October). PhD Thesis, University of British Columbia.
Yan, S., Song, Y., Bai, J., & Elmo, D. (2019). A Study on the Failure of Resin End-Anchored
Rockbolts Subjected to Tensile Load. Rock Mechanics and Rock Engineering, 52(6),
1917–1930. https://doi.org/10.1007/s00603-018-1663-2
Yang, G. (2019). The study of shear properties of various cable bolts using single and double
shear technics with the modeling simulations. PhD Thesis, University of Wollongong.

316
Yang, G., Aziz, N., Rasekh, H., Khaleghparast, S., Li, X., & Nemcik, J. (2018). Profile of
Sheared Cable Bolts Strand Wires Publication Details. Coal Operators’ Conference, 343–
352. http://ro.uow.edu.au/coal/714
Yang, X., Ren, T., He, X., & Tan, L. (2019). A review of energy sources of coal burst in
Australian coal mines. Coal Operators’ Conference, February, 287–296.
Yazici, S., & Kaiser, P. K. (1992). Bond strength of grouted cable bolts. International Journal
of Rock Mechanics and Mining Sciences And, 29(3), 279–292.
https://doi.org/10.1016/0148-9062(92)93661-3
Yeih, W., Huang, R., Chang, J. J., & Yang, C. C. (1997). A pullout test for determining
interface properties between rebar and concrete. Advanced Cement Based Materials, 2(5),
57–65.
Yi, X., & Kaiser, P. K. (1994a). Elastic stress waves in rockbolts subject to impact loading.
International Journal for Numerical and Analytical Methods in Geomechanics, 18(July
1993), 121–131.
Yi, X., & Kaiser, P. K. (1994b). Impact testing for rockbolt design in rockburst conditions.
International Journal of Rock Mechanics and Mining Sciences And, 31(6), 671–685.
https://doi.org/10.1016/0148-9062(94)90007-8
Yuan, H., Teng, J. G., Seracino, R., Wu, Z. S., & Yao, J. (2004). Full-range behavior of FRP-
to-concrete bonded joints. Engineering Structures, 26(5), 553–565.
https://doi.org/10.1016/j.engstruct.2003.11.006
Zhao, Y., & Yang, M. (2011). Pull-out behavior of an imperfectly bonded anchor system.
International Journal of Rock Mechanics and Mining Sciences, 48(3), 469–475.
https://doi.org/10.1016/j.ijrmms.2010.09.011

317
A: Appendices

A.1 Appendices of Chapter 4


Table A-1: Experimental test plan for static and dynamic shear tests

Test Test Installation Cable type Cable Grout


No. mode angle Profile UCS
Degree MPa
1 Static 90 15.2 mm Plain 10-15
2 Static 90 15.2 mm Bulbed 10-15
3 Static 90 15.2 mm Plain 45-50
4 Static 90 15.2 mm Plain 60-75
5 Static 90 15.2 mm Bulbed 60-75
6 Static 30 15.2 mm Plain 45-50
7 Static 45 15.2 mm Plain 45-50
8 Static 45 15.2 mm Plain 45-50
9 Static 45 15.2 mm Plain 45-50
10 Static 45 15.2 mm Plain 45-50
11 Static 90 63 t Sumo Plain 45-50
12 Static 90 63 t Sumo Plain 45-50
13 Static 90 63 t Sumo Plain 45-50
14 Static 90 63 t Sumo Plain 45-50
15 Static 90 63 t Sumo Plain 45-50
16 Dynamic 90 15.2 mm Plain 45-50
17 Dynamic 90 15.2 mm Plain 45-50
18 Dynamic 90 15.2 mm Plain 45-50
19 Dynamic 90 15.2 mm Plain 45-50
20 Dynamic 90 15.2 mm Plain 45-50
21 Dynamic 90 15.2 mm Plain 45-50
22 Dynamic 90 15.2 mm Plain 45-50
23 Dynamic 90 63 t Sumo Plain 45-50

318
A.1.1 A: Static perpendicular shear test results
300

250

200
Load (kN)

150
Shear
100 Axial

50

0
0 500 1000 1500 2000 2500 3000 3500 4000
Displacement (mm)

Figure A-1: Test 2: Double shear using slow-set grout- Bulb cable
350
300
250
Load (kN)

200
150 Shear
100 Axial
50
0
0 500 1000 1500 2000 2500 3000 3500
Displacement (mm)

Figure A-2: Test 4: Double shear using quick-set grout

350
300
250
Load (kN)

200
150 Shear

100 Axial

50
0
0 10 20 30 40 50 60
Displacement (mm)

Figure A-3: Test 5: Double shear using quick-set grout

319
1200

1000

800
Load (kN)

600

400

200

0
0 10 20 30 40 50 60 70 80 90 100
Displacement (mm)

Shear load Axial load 1 Axial load 2

Figure A-4: Perpendicular double shear using grout- 63 t cable bolt

1000

900

800

700
Load (kN)

600

500

400

300

200

100

0
0 10 20 30 40 50 60 70 80 90 100
Displacement (mm)

axial load 1 axial load 2 shear load

Figure A-5: Perpendicular double shear using grout- 63 t cable bolt

320
900

800

700

600
Load (kN)

500

400

300

200

100

0
0 10 20 30 40 50 60 70 80 90
Displacement (mm)

axial load 1 axial load 2 Shear load

Figure A-6: Perpendicular double shear using strong grout- 63 t cable bolt
800

700

600

500
Load (kN)

400

300

200

100

0
0 10 20 30 40 50 60
Displacement (mm)

axial load 1 axial load 2 shear load

Figure A-7: Perpendicular double shear using strong grout- 63 t cable bolt

321
A.1.2 B: Static angled shear test results
600

500

400
Load (kN)

300

200

100

0
0 5 10 15 20 25 30
Displacement (mm)

Shear load Axial load

Figure A-8: Static 45 degree inclined double shear test using (Test 7)

700

600

500
Load (kN)

400

300

200

100

0
0 2 4 6 8 10 12 14 16 18
Displacement (mm)

Shear Load

Figure A-9: Static 45 degree inclined double shear test using (Test 9)

322
A.1.3 C: Dynamic perpendicular shear test results
600 60

500 50

Displacement (mm)
400 40
Load (kN)

300 30

200 20

100 10

0 0
0 5 10 15 20 25 30
Time (ms)

Shear load Displacement

Figure A-10: Dynamic double shear (Test 16)


1600 80

1400 70

1200 60

Displacement (mm)
1000 50
Load (kN)

800 40

600 30

400 20

200 10

0 0
0 5 10 15 20 25 30 35
Time (ms)

Shear load Displacement

Figure A-11: Dynamic double shear (Test 17)

323
1000 140.0
900
120.0
800
100.0

Displacement (mm)
700
600
Load (kN)

80.0
500
60.0
400
300 40.0
200
20.0
100
0 0.0
0 5 10 15 20 25 30 35
Time (ms)

Shear load Axial load Displacement

Figure A-12: Dynamic double shear (Test 18)


1400 80

1200 70

60
1000

Displacement (mm)
50
Load (kN)

800
40
600
30
400
20

200 10

0 0
0 5 10 15 20 25 30 35 40
Time (ms)

Shear load Axial load Displacement

Figure A-13: Dynamic double shear (Test 19)

324
900 140
800 120
700 100

Displacement (mm)
600
80
Load (kN)

500
60
400
40
300
200 20

100 0

0 -20
0 100 200 300 400 500 600
Time (mS)

Shear load Axial load Displacement

Figure A-14: Dynamic double shear (Test 22)

325
A.2 Appendices of Chapter 5

A.2.1 Detailed drawings of the designed Static and Dynamic Pullout test Apparatus

1
9

10

2
11

12
13
4

14

5
15

6
16
7
17

18

Figure A-15: Cross-section view of the SDPA

326
f) Main seat (Part No. 13)

Figure A-16: Perspective and cross-section view of the main seat


327
Figure A-17: Top view and cross-section of the main seat

328
Figure A-18: Top view and cross-section of central part of the main seat

329
g) Loading platform (Part No. 9)

Figure A-19: Bottom view of loading platform

330
Figure A-20: Reinforcement design of the loading platform

331
h) Other sections

Figure A-21: Channel beam used in the main seat and fixed frame

332
Figure A-22: I-beam used in the fixed frame

333
Figure A-23: Angles used in the fixed frame and the main seat

334
Figure A-24: Bearing holder

335
Figure A-25: Shaft holder

336
Figure A-26: Anti-rotation tube

337
Figure A-27: Anti-rotation plate

338
Figure A-28: confinement cage rods

Figure A-29: Von-Mises failure criteria evaluation for moving parts of SDPA

339
Figure A-30: Pullout test setup being lifted by roof crane and placed under loading machine

340
A.3 Appendices of Chapter 6

A.3.1 A: Static pullout test results of 15.2 mm cable bolt


Table A-2: Static pullout test results of 15.2 mm cable bolt
No. Code Cable Embedme Grout Max Max Unit
type nt length UCS Pullout disp Pullout
load load
mm MPa kN mm kN/mm
1 D2 Plain 185 10-15 19.4 36 0.1
2 D4 Plain 185 10-15 21.47 30.5 0.1
3 D5 Plain 185 10-15 24.25 25.7 0.1
4 D6 Plain 300 10-15 27.22 33.3 0.1
5 D5 Bulbed 300 10-15 127.4 73 0.4
6 D3 Plain 185 50-65 61.8 85.3 0.3
7 D8 Plain 185 50-65 87.3 10.4 0.5
8 D9 Plain 185 50-65 76.7 11.7 0.4
9 D8 Plain 300 50-65 131.8 14.7 0.4
10 D9 Plain 300 50-65 121.9 43.3 0.4
11 D10 Plain 300 50-65 132 69.9 0.4
12 D1 Bulbed 300 50-65 Failed 26 Failed
13 D7 Plain 185 85-90 112.9 20 0.6
14 D7 Plain 300 85-90 137.2 11.1 0.5
15 D15 Bulbed 300 85-90 Failed 57.1 Failed
16 D16 Bulbed 300 85-90 Failed 45.2 Failed

341
25

20

15
Load (kN)

10

0
0 5 10 15 20 25 30 35 40
Displacement (mm)

Figure A-31: Test 1- Short encapsulation- 15.2 mm Plain- weak grout

25

20

15
Load (kN)

10

0
0 5 10 15 20 25 30 35
Displacement (mm)

Figure A-32: Test 2- Short encapsulation- 15.2 mm Plain- weak grout

342
30

25

20
Load (kN)

15

10

0
0 5 10 15 20 25 30
Displacement (mm)

Figure A-33: Test 3- Short encapsulation- 15.2 mm Plain- weak grout


30

25

20
Load (kN)

15

10

0
0 5 10 15 20 25 30 35
Displacement (mm)

Figure A-34: Test 4- Long encapsulation- 15.2 mm Plain- weak grout

343
140

120

100
Load (kN)

80

60

40

20

0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
Displacement (mm)

Figure A-35: Test 5- Long encapsulation- 15.2 mm Bulbed- weak grout


70

60

50
Load (kN)

40

30

20

10

0
0 10 20 30 40 50 60 70 80 90
Displacement (mm)

Figure A-36: Test 6- Short encapsulation- 15.2 mm Plain- medium grout

344
100

90

80

70

60
Load( kN)

50

40

30

20

10

0
0 2 4 6 8 10 12
Displacement (mm)

Figure A-37: Test 7- Short encapsulation- 15.2 mm Plain- medium grout


90

80

70

60
Load (kN)

50

40

30

20

10

0
0 2 4 6 8 10 12 14
Displacement (mm)

Figure A-38: Test 8- Short encapsulation- 15.2 mm Plain- medium grout

345
160

140

120

100
Load (kN)

80

60

40

20

0
0 2 4 6 8 10 12
Displacement (mm)

Figure A-39: Test 9- Long encapsulation- 15.2 mm Plain- medium grout


140

120

100
Load (kN)

80

60

40

20

0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Figure A-40: Test 10- Long encapsulation- 15.2 mm Plain- medium grout

346
160

140

120

100
Load (kN)

80

60

40

20

0
0 10 20 30 40 50 60 70 80
Displacement (mm)

Figure A-41: Test 11- Long encapsulation- 15.2 mm Plain- medium grout
300

250

200
Load (kN)

150

100

50

0
0 5 10 15 20 25 30
Displacement (mm)

Figure A-42: Test 12- Long encapsulation- 15.2 mm Bulbed- medium grout

347
140

120

100
Load (kN)

80

60

40

20

0
0 5 10 15 20
Displacement (mm)

Figure A-43: Test 13- Short encapsulation- 15.2 mm Plain- strong grout
160

140

120

100
Load (kN)

80

60

40

20

0
0 2 4 6 8 10 12
Displacement (mm)

Figure A-44: Test 14- Long encapsulation- 15.2 mm Plain- strong grout

348
300

250

200
Load (kN)

150

100

50

0
0 10 20 30 40 50 60
Displacement (mm)

Figure A-45: Test 15- Long encapsulation- 15.2 mm bulbed- strong grout
250

200

150
Load (kN)

100

50

0
0 5 10 15 20 25 30 35 40 45 50
Displcaement (mm)

Figure A-46: Test 16- Long encapsulation- 15.2 mm bulbed- strong grout

349
A.3.2 B: Static pullout test results of 63 t Sumo cable bolt
Table A-3: Static pullout test result of 63 t Sumo cable bolt encapsulated in cement grout and
resin
Test Code Cable Grout Embedment Ultimate Absorbed Max Unit
No. type type length peak load energy displacement pullout
load
mm kN kJ mm kN/mm
17 1 Plain Cement- 300 317.9 14.5 71 1.06
18 15 Plain based 300 330 28.7 114 1.10
19 13 Plain 300 357.6 38.5 148 1.19
20 2 Bulbed 300 518.7 15.4 67 1.73
21 3 Bulbed 300 554.9 14.5 60 1.85
22 7 Plain Resin 300 256.0 13.3 89 0.85
23 8 Plain 300 244.8 12.3 97 0.82
24 9 Bulbed 300 544.1 50.1 120 1.81
25 17 Bulbed 300 484.5 43.4 118 1.62

350

300

250
Load (kN)

200

150

100

50

0
0 10 20 30 40 50 60 70 80
Displacement (mm)

Figure A-47: Test 17- Long encapsulation, Plain 63 t Sumo- cement grout

350
350

300

250
Load (kN)

200

150

100

50

0
0 20 40 60 80 100 120
Displacement (mm)

Figure A-48: Test 18- Long encapsulation, Plain 63 t Sumo- cement grout
400

350

300

250
Load (kN)

200

150

100

50

0
0 20 40 60 80 100 120 140 160
Displacement (mm)

Figure A-49: Test 19- Long encapsulation, Plain 63 t Sumo- cement grout

351
600

500

400
Load (kN)

300

200

100

0
0 10 20 30 40 50 60 70 80
Displacement (mm)

Figure A-50: Test 20- Long encapsulation, bulbed 63 t Sumo- cement grout
600

500

400
Load (kN)

300

200

100

0
0 10 20 30 40 50 60 70
Displacement (mm)

Figure A-51: Test 21- Long encapsulation, bulbed 63 t Sumo- cement grout

352
300

250

200
Load (kN)

150

100

50

0
0 10 20 30 40 50 60 70 80
Displacement (mm)

Figure A-52: Test 22- Long encapsulation, Plain 63 t Sumo- resin


300

250

200
Load (kN)

150

100

50

0
0 10 20 30 40 50 60 70 80
Displacement (mm)

Figure A-53: Test 23- Long encapsulation, Plain 63 t Sumo- resin

353
600

500

400
Load (kN)

300

200

100

0
0 20 40 60 80 100 120 140
Displacement (mm)

Figure A-54: Test 24- Long encapsulation, bulbed 63 t Sumo- resin


600

500

400
Load (kN)

300

200

100

0
0 20 40 60 80 100 120 140
Displacement (mm)

Figure A-55: Test 23- Long encapsulation, bulbed 63 t Sumo- resin

354
A.3.3 C: Dynamic pullout test results of 63 t Sumo cable bolt
Table A-4: Dynamic pullout test results of 15.2 mm cable bolt
Test Code Cable type Encapsulation Absorbed Max Unit Average
No. material energy displacement energy pullout load
kJ mm kJ/mm
26 4 Plain Grout 8.9 69 0.13 309
27 14-1 Plain Grout 9.6 69 0.14 297
28 14-2 Plain Grout 7.1 56 0.13 294
29 10 Plain Resin 10.5 151 0.07 184
30 5 Bulbed Grout 9.9 48 0.21 465
31 6-1 Bulbed Grout 9.9 47 0.21 555
32 6-2 Bulbed Grout 10.2 47 0.22 668
33 16-1 Bulbed Grout 10.1 42 0.24 562
34 16-2 Bulbed Grout 9.6 33 0.29 614
35 12-1 Bulbed Resin 10.8 47 0.23 306
36 12-2 Bulbed Resin 10.9 47 0.23 173

1000

900

800

700

600
Load (kN)

500

400

300

200

100

0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
Displacement (mm)

Figure A-56: Test 26- Long encapsulation, Plain 63 t Sumo- cement grout

355
1200

1000

800
Load (kN)

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
Displacement (mm)

Figure A-57: Test 27- Long encapsulation, Plain 63 t Sumo- cement grout
1200

1000

800
Load (kN)

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50 55 60
Displacement (mm)

Figure A-58: Test 28- Long encapsulation, Plain 63 t Sumo- cement grout

356
1000

900

800

700

600
Load (kN)

500

400

300

200

100

0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Displacement (mm)

Figure A-59: Test 29- Long encapsulation, Plain 63 t Sumo- resin


1600

1400

1200

1000
Load (kN)

800

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Figure A-60: Test 30- Long encapsulation, bulbed 63 t Sumo- cement grout

357
1600

1400

1200

1000
Load (kN)

800

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Figure A-61: Test 31- Long encapsulation, bulbed 63 t Sumo- cement grout
1400

1200

1000
Load (kN)

800

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Figure A-62: Test 32- Long encapsulation, bulbed 63 t Sumo- cement grout

358
1200

1000

800
Load (kN)

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Figure A-63: Test 33- Long encapsulation, bulbed 63 t Sumo- cement grout
900

800

700

600
Load (kN)

500

400

300

200

100

0
0 5 10 15 20 25 30 35 40
Displacement (mm)

Figure A-64: Test 34- Long encapsulation, bulbed 63 t Sumo- cement grout

359
1000

900

800

700

600
Load (kN)

500

400

300

200

100

0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Figure A-65: Test 35- Long encapsulation, bulbed 63 t Sumo- resin


1200

1000

800
Load (kN)

600

400

200

0
0 5 10 15 20 25 30 35 40 45 50
Displacement (mm)

Figure A-66: Test 36- Long encapsulation, bulbed 63 t Sumo- resin

360
A.4 Appendices of Chapter 7

A.4.1 Input parameter of materials and contacts

Figure A-67: CSCM material input parameters for grout

Figure A-68: CSCM material input parameters for resin

Figure A-69: Input parameters of tiebreak contact for grout and cable bolt

361
Figure A-70: Input parameters of tiebreak contact for grout and cable bolt
A.4.2 Verification of grout material

Figure A-71: Shear stress in UCS test of grout


A.4.3 Simulation of 15.2 mm cable bolt tensile test (using ANSYS)

Figure A-72: failure mechanism of tensile test of the cable bolt

362
Figure A-73: Simulation of tensile test of 15.2 mm using ANSYS
A.4.4 Detailed pullout test model results

Figure A-74: Lateral stress on the grout

Figure A-75: 2nd principal deviatoric stress

363
Figure A-76: Elongation of the cable as a function of relative displacement
A.4.5 Low friction-deformable cable

Figure A-77: Axial and lateral stress in low-friction pullout test


A.4.6 High friction- deformable cable

Figure A-78: Axial stress in high-friction pullout test


364
A.4.7 Proposed simplified model

Figure A-79: Axial stress distribution in the simplified pullout test model

Figure A-80: Propagation of plastic zone in different test intervals


A.5 Appendices of Chapter 8
Code for excel data screening (Removing NaN data while reading Excel file
%Claen excel File

365
[num,txt,raw] = xlsread('results.xlsx','Sheet1') ;
disp("Data trimmer code")
[rows, columns] = size(num);
nanLocations = isnan(num);
ca = cell(rows, columns); % Initialize cell array.
index = 1;
for row = 1 : rows
if ~nanLocations(row, 1)
ca{index, 1} = num(row, 1); % Only fill cell if it's not nan.
ca{index, 2} = num(row, 2);
index = index + 1;
end
end
xlswrite('results_clean.xlsx', ca);

366

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy