You are on page 1of 301

University of Wollongong

Research Online
University of Wollongong Thesis Collection

University of Wollongong Thesis Collections

2001

Shear behaviour of fully grouted bolts under


constant normal stiffness condition
Ashitava Dey
University of Wollongong

Recommended Citation
Dey, Ashitava, Shear behaviour of fully grouted bolts under constant normal stiffness condition, Doctor of Philosophy thesis, Faculty
of Engineering, University of Wollongong, 2001. http://ro.uow.edu.au/theses/1837

Research Online is the open access institutional repository for the


University of Wollongong. For further information contact Manager
Repository Services: morgan@uow.edu.au.

SHEAR BEHAVIOUR OF FULLY GROUTED BOLTS


UNDER CONSTANT NORMAL STIFFNESS CONDITION

A thesis submitted in fulfilment of the requirements


for the award of the degree of

DOCTOR OF PHILOSOPHY

from

UNIVERSITY OF W O L L O N G O N G

by

Ashitava Dey
B.E. (Cal), M.Tech (ISM), AMIE (India).

Faculty of Engineering
2001

CERTIFICATION

I, Ashitava Dey, declare that this thesis, submitted in fulfilment of the requirements for
the award of Doctor of Philosophy, in the Faculty of Engineering, University of
Wollongong, is wholly m y o w n work unless otherwise referenced or acknowledged. The
document has not been submitted for qualifications at any other academic institution.

Ashitava Dey
3rd July 2001

ACKNOWLEDGEMENTS

I wish to express m y sincere gratitude to m y thesis supervisors Dr. N . I. Aziz and


Prof. B. Indraratna, Faculty of Engineering, University of Wollongong, for providing
the opportunity and excellent facilities to conduct my research work on this
challenging project. Their generous support (technical, financial, and in some cases
personal), guidance, encouragement, tolerance and friendly attitude have been
invaluable during the entire period of the research work.

Special thanks to Mr. Alan Grant, Senior Technical Officer, Faculty of Engineering
for his help with the equipment fabrication and the laboratory testing. Thanks are
extended to Mr. Ian Laird, Mr. Richard Webb, Mr. Bob Rowlan, Mr. Ian Bridge, Mr.
Ravi Narayan (undergraduate student) and other technical staff of the Faculty of
Engineering for their generous assistance at various stages of research work.

I greatly appreciate the contributions made by Dr. Asadul Haque, Geotechnical


Engineer, Railway Services Australia, through his critical comments and advice with
respect to some complex areas of this study, and the extent of help provided by Mr.
John Sandona, Under Manager, Tower Colliery, BHP, in conducting field
investigations.

Naturally, there were many more individuals who helped in one way or another in
bringing this thesis to a successful conclusion, I thank you all.

iii

A s always, I wish to thank all m y family members, especially m y wife Swati, for her
continuous inspiration and support in all aspects of my life. Without their patience,
understanding and encouragement this thesis would never been completed. Finally, a
little thanks to tiny Arko for his amazing word "nanga", and the fun, joy and
happiness he brought into our family.

Three years in the making! A cast of thousands!

iv

ABSTRACT

Research on rock bolting technology has been extensively pursued, in both civil and
mining engineering, for several decades. The study of bolting technology embraced
initially rigid re-bars and subsequently flexible cable bolts. Generally, the past studies
were focused on bolt strength and their load bearing capacities in relation to physical
size (length and diameter) and material properties. Very little work has been reported
on the influence of bolt surface configuration on the load transfer mechanisms
between the bolt and the host rock. Also, previous research on the influence of
bolting on rock joint reinforcement and stabilisation has been carried out under
Constant Normal Load (CNL) conditions, which does not reflect the actual
deformation behaviour of reinforced joints. This matter has lately been recognised,
and some recent publications have highlighted the shortcomings of the tests
conducted under CNL conditions. Accordingly, this thesis is concerned with various
aspects of bolt research related to the better understanding of the load transfer
mechanisms between the bolt/resin/rock, and the impact of infill on the reinforced
joint shear strength under Constant Normal Stiffness (CNS) condition. Two types of
bolts, commonly used in Australian coal mines were used for both laboratory and
field investigations, and referred to as type I and type II bolts, respectively.

The shear behaviour of bolt/resin interface was studied under CNS condition in the
laboratory by testing cast resin/plaster samples in a specially constructed CNS
apparatus. The samples were tested for various loading cycles at initial normal stress
(rjno) levels ranging from 0.1 to 7.5 MPa, representing the in-situ horizontal stress

condition in the field. Laboratory test results indicated that, bolts with deeper rib
profile and wider rib spacing (e.g. bolt type I) offers higher shear resistance at low
confining pressures less than 6 MPa, whereas, bolts with shallower rib profile with
narrower rib spacing (e.g. bolt type II) offers marginally higher shear resistance at
confining pressures exceeding 6 MPa. The maximum dilation of the bolt/resin
interface occurs at a shear displacement of about 60% of the bolt rib spacing.

The laboratory study on the shear behaviour of the bolt/resin interface of fully
grouted bolts was supported with field investigations in two local coal mines. At
West Cliff and Tower Collieries, in the Southern Coalfield of Sydney Basin, NSW,
Australia, 18 instrumented bolts and 6 extensometer probes were installed at three
different sites. Field investigations in those mines revealed that, the load transfer on
the bolt is influenced by; a) the confining stress condition in the field, b) the strata
deformation, and c) the surface profile of the bolts. The influence of front abutment
pressure was observed by sharply increasing load build up on the bolts, when the
approaching longwall face was 60m and 150m from the test sites, in the travelling
and belt roads, respectively. The field study also showed that, under the influence of
low horizontal stress (both in magnitude and the direction), type I bolt offered
significantly higher shear resistance, whereas under high influence of high horizontal
stress, type II bolt offered marginally greater shear resistance at the bolt/resin
interface, which were in accordance with the findings of laboratory testing.

The shear behaviour of bolted and non-bolted joints containing infill material, up to
7.5 mm in thickness, was studied under various initial normal stress levels between

vi

0.13 M P a and 3.25 M P a , at a constant strain rate of 0.5 m m / m i n and a constant


stiffness of 8.5 kN/mm. Significant reduction in shear strength was observed when
the joint contained a layer of clay infill of 1.5 mm. Bolting contributed to increasing
the strength and stiffness of the joint composite, except at large normal stress levels
and at high infill thickness. The dilation and overall friction angle for bolted and nonbolted joints were also compared along with stress profiles. At high infill thickness
(t/a>l), the shear behaviour under both CNL and CNS conditions was found to be
similar for both bolted and non-bolted joints, while at low infill thickness (t/a<Q.3),
the CNL strength envelope plotted significantly above the CNS em-elope. Thus, for
rough joints with little or no infill, the CNS behaviour is more realistic in practice,
especially for underground mining conditions in bedded or jointed rock.

An analytical model is presented using Fourier transform and the hyperbolic stressstrain simulation, to predict the shear behaviour of both clean and infilled bolted
joints. The values of shear stress, normal stress, and dilation predicted by the
analytical model were tallied favourably with the laboratory results. The UDEC
model presented in the thesis underestimated the shear strength when compared with
laboratory tested values, whereas it overestimated both the normal stress and dilation,
especially at high initial normal stress conditions. Identical shear behaviour was
observed, in both CNL and CNS UDEC model predictions, at low shear
displacements and at high initial normal stress conditions, whereas the shear stress
predictions were different at low normal stress conditions.

vn

CONTENTS

Page
TITLE PAGE i
CERTIFICATION ii
ACKNOWLEDGEMENTS iii
ABSTRACT v
CONTENTS viii
LIST OF FIGURES xv
LIST OF TABLES xxiv
LIST OF SYMBOLS AND ABBREVIATIONS xxv

Chapterl. INTRODUCTION
1.1 General background 1
1.2 Key objectives 5
1.3 Outline of the thesis 5

Chapter 2. ROCK BOLTING PRACTICES


2.1 Introduction 9
2.2 Review of typical rock bolts and accessories 10
2.3 Rock bolting theories 13
2.4 Support design philosophy 16
2.4.1 Estimation of rock load 16
2.4.2 Design of suitable support system 20
2.5 Review of rock bolt research 22

Contents (continued..)

2.6 T h e concept of constant normal stiffness and its applicability in rock

47

bolting
50

2.7 Conclusion

Chapter 3. EXPERIMENTAL STUDY OF THE FAILURE MECHANISM


OF BOLT/RESIN INTERFACE
3.1 Introduction

51

3.2 Bolt surface preparation

52

3.3 Sample casting

54

3.4 C N S testing apparatus

56

3.5 Testing of bolt/resin interface

59

3.6 Shear behaviour of bolt/resin interface

59

3.6.1 Effect of normal stress on stress paths

59

3.6.2 Dilation behaviour

62

3.6.3 Effect of normal stress on peak shear

64

3.6.4 Effect of cyclic loading on peak shear

65

3.6.5 Overall shear behaviour of type I and type II bolts

67

3.6.6 Effect of normal stiffness

69

3.6.7 Effect of bolt surface profile configuration

70

3.7 Empirical model for predicting the shear resistance at bolt/resin

71

interface
3.8 Conclusion

76

Chapter 4. FIELD INVESTIGATION FOR LOAD TRANSFER


MECHANISM OF FULLY GROUTED BOLTS
4.1 Introduction

77

4.2 Site description

78

ix

Contents (continued..)

4.2.1 West Cliff Colliery (Mine 1)


4.2.2 Tower Colliery (Mine 2) 81
4.3 Instrumentation 86
4.3.1 Instrumented bolts 86
4.3.2 Sonic probe extensometer 90
4.3.3 Intrinsically safe strain bridge monitor 92
4.4 Field monitoring and data analysis 93
4.4.1 West Cliff Colliery 93
4.4.2 Tower Colliery 100
4.4.2.1 Load transfer during the panel development and the 101
longwall retreating phases
4.4.2.2 Load transfer in the belt and travelling road 103
4.4.2.3 Behaviour of strata deformation 107
4.4.2.4 Impact of strata deformation of the load transfer in the 112
bolts
4.4.2.5 Comparison of load transfer in type I and type II bolts 115
4.5 Bolt selection guidelines 124
4.6 Conclusions 125

Chapter 5. SHEAR BEHAVIOUR OF ROCK JOINTS


5.1 Introduction 127
5.2 Review of unfilled joints 128
5.3 Review of unfilled reinforced joints 133
5.4 Review of infilled joints 145
5.5 Concluding remarks 153

78

Contents (continued..)

Chapter 6. SHEAR BEHAVIOUR OF UNFILLED BOLTED JOINTS


6.1 Introduction 154
6.2 Laboratory investigation 154
6.2.1 Selection of model material 154
6.2.2 Preparation of bolted joints 155
6.2.3 CNS direct shear testing apparatus 158
6.2.4 Laboratory experiments 159
6.2.5 Processing of test data 160
6.3 Shear behaviour of non-bolted joints 161
6.3.1 Shear stress 161
6.3.2 Dilation 162
6.3.3 Normal stress 163
6.4 Shear behaviour of bolted joints 164
6.4.1 Shear stress 164
6.4.2 Dilation 165
6.4.3 Normal stress 166
6.5 Strength envelope 168
6.6 Conclusions 169

Chapter 7. SHEAR BEHAVIOUR OF INFILL BOLTED JOINT


7.1 Introduction 171
7.2 Laboratory investigation 171
7.2.1 Sample preparation 171
7.2.2 Testing procedure 174
7.3 Shear behaviour of infilled non-bolted joints 176

xi

Contents (continued..)

7.3.1 Shear stress

176

7.3.2 Normal stress

178

7.3.3 Dilation

179

7.3.4 Stress path

180

7.4 Shear behaviour of infilled bolted joints


7.4.1 Shear stress

182
182

7.4.2 Normal stress

184

7.4.3 Dilation

184

7.4.4 Stress path

186

7.5 Comparison of the shear behaviour of infilled bolted and non-bolted

187

joints
7.5.1 Overall shear behaviour

187

7.5.2 Effect of infill on shear strength

190

7.5.3 Profile of shear plane

192

7.5.4 Strength envelopes

194

7.6 Conclusions

196

Chapter 8. MODELLING OF THE SHEAR BEHAVIOUR OF


REINFORCED JOINTS
8.1 Introduction

198

8.2 Analytical model for shear behaviour of unfilled reinforced joints

199

8.2.1 Definition of Fourier series and its applicability in joint modelling 199
8.2.2 Prediction of variation in normal stress

202

8.2.3 Prediction of shear stress

203

8.3 Analytical model for shear behaviour of infilled reinforced joints


8.3.1 Modelling of drop in shear strength

xii

206
206

Contents (continued..)

8.3.1.1 Impact of infill thickness on peak shear strength


8.3.1.2 Relationship between normalised strength drop and 207
t/a ratio
8.3.1.3 Modelling of normalised strength drop 208
8.3.2 Modelling of shear strength of infilled bolted joints 211
8.4 Computer program for calculating Fourier coefficients and the 212
shear strength
8.5 Verification of analytical model 212
8.6 Numerical model of reinforced joints under CNS conditions 216
8.6.1 General background 217
8.6.2 Global reinforcement 217
8.6.3 Material properties 222
8.6.4 The continuous yielding model 225
8.6.5 Conceptual CNS model for reinforced joints 226
8.6.6 UDEC modelling of reinforced joints under CNS condition 227
8.6.7 Validation of the model 229
8.6.8 Comparison of CNL and CNS shear behaviour 234
8.7 Conclusion 236

Chapter 9. CONCLUSIONS AND RECOMMENDATIONS


9.1 Conclusions 237
9.2 Recommendations for future research 242

REFERENCES 245
APPENDICES
Appendix I 256

xiii

206

Contents (continued..)

Appendix IIA 260


Appendix IIB 264
Appendix IIC 265
Appendix IID 266
Appendix fflA 269
Appendix IIIB 272

xiv

LIST OF FIGURES

Figures

Page

Figure 2.1: Suspension theory.

14

Figure 2.2: B e a m building theory.

14

Figure 2.3: Keying theory.

15

Figure 2.4: The ground response curve (after Deere et al., 1969).

19

Figure 2.5: Support management plan (after Strata Control Technology, 1998).

21

Figure 2.6: Failure modes of rock bolt system a) failure of rock mass, b) failure

23

of bolt/grout/resin interface, and c) failure of bolt (after Littlejohn &


Bruce, 1975).
Figure 2.7: Stress distribution along a resin anchor, a) stress situation, and b)

25

theoretical stress distribution (after Farmer, 1975).


Figure 2.8: Load displacement, strain distribution and computed shear stress

26

distribution curves - 500 m m resin anchors in limestone, a) strain


distribution, b) stress distribution (after Farmer, 1975).
Figure 2.9: Results from instrumented bolts a) normal stress in the bolt, and b)

27

shear stress at the bolt/grout interface (after Dunham, 1976).


Figure 2.10: Zones of influence of a grouted bolt surrounding a tunnel

29

(after Adali and Russel, 1980).


Figure 2.11: Influence of grouted bolts on tunnel convergence

34

(after Indraratna and Kaiser, 1990).


Figure 2.12: Schematic diagram reflecting the geometry of a rough bolt

35

(after Yazici and Kaiser, 1992).


Figure 2.13: Schematic diagram relating components of bond strength model

35

(after Yazici and Kaiser, 1992).


Figure 2.14: The modified H o e k Cell (after Hyett et al., 1995).

37

Figure 2.15: M o d e s of failure of grouted bolt Cell (after Hyett et al., 1995).38
Figure 2.16: Detail of short embedment pull test (after Benmokrane et al., 1995). 40
Figure 2.17: Influence of anchor length on the load-displacement behaviour
(after Benmokrane et al., 1995).

41

List of Figures (contd....)

Figure 2.18: Schematic diagram of bolt anchorage a) long anchor, b) short

43

anchor (after Serrano and Olalla, 1999).


Figure 2.19: Failure mechanism of long and short anchored bolts

44

(after Serrano and Olalla, 1999).


Figure 2.20: Stress components in a small section of a bolt

45

(after Li and Stillborg, 1999).


Figure 2.21: Shear stress along a fully grouted rock bolt subjected to an axial

46

load after decoupling occurs (after Li and Stillborg, 1999).


Figure 2.22: Shear stress along a fully grouted rock bolt subjected to an axial

46

load before decoupling occurs (after Li and Stillborg, 1999).


Figure 2.23: Behaviour of bolted joint in the field under Constant Normal

49

Stiffness
Figure 2.24: Dilation behaviour of joint plane a) two smooth planes,

49

b) bolt and resin interface.

Figure 3.1: Hollow bolt segment. 53


Figure 3.2: Flattened bolt surface.

53

Figure 3.3: Bolt surface welded on the top shear box plate.

54

Figure 3.4: Arrangement of casting resin samples.

55

Figure 3.5: Cast resin block.

56

Figure 3.6: Cast resin samples inside the bottom shear box.

56

Figure 3.7: Line diagram of the C N S apparatus (modified after

57

indraratna et al., 1997).


Figure 3.8: A closer view of the C N S apparatus (modified after

58

Indraratna et al., 1999).


Figure 3.9: Shear stress profiles of the type I bolt from selected tests.

61

Figure 3.10: Dilation profiles of the type I bolt from selected tests.

63

Figure 3.11: Dilation profiles of type I bolt for first cycle of loading.

63

Figure 3.12: Dilation profiles of type II bolt forfirstcycle of loading.

64

Figure 3.13: Shear stress profiles of type I bolt forfirstcycle of loading.

65

Figure 3.14: Shear stress profiles of type II bolt forfirstcycle of loading.

65

Figure 3.15: Variation of peak shear stress with normal stress for type I bolt.

66

xvi

List of Figures (contd....)

Figure 3.16: Variation of peak shear stress with normal stress for type II bolt.
Figure 3.17: Comparison of stress profile and dilation of type I and

67
68

type II bolts for first cycle of loading.


Figure 3.18: Variation of peak dilation with initial normal stress.
Figure 3.19: Comparison of predicted values of shear stresses and the values

71
74

obtained from laboratory experiments for type I bolt.


Figure 3.20: Comparison of predicted values of shear stresses and the values

75

obtained from laboratory experiments for type II bolt.

Figure 4.1: Geographical location of West Cliff Colliery (Mine 1), and 78
Tower Colliery (Mine 2).
Figure 4.2: Geological sections showing Bulli seam and the associated strata

79

near the instrumentation site at West Cliff Colliery.


Figure 4.3: General mine layout of West Cliff Colliery.

80

Figure 4.4: General layout of the panel under investigation indicating

80

instrumentation site at West Cliff Colliery.


Figure 4.5: Detail site plan of instrumented bolts at West Cliff Colliery.
Figure 4.6: Borehole section showing the geological formation above

81
82

Bulli seam, Tower Colliery.


Figure 4.7: General mine layout of Tower Colliery indicating the panel

83

under investigation.
Figure 4.8: Detail layout of the panel under investigation at Tower Colliery.

84

Figure 4.9: Detail layout of the instrumentation site at Tower Colliery.

84

Figure 4.10: Instrumentation naming convention at Tower Colliery,

85

a) for bolts, and b) for extensometers.


Figure 4.11: Strain gauge and bolt layout for West Cliff Colliery (Mine 1).

87

Figure 4.12: Strain gauge and bolt layout for Tower Colliery (Mine 2).

88

Figure 4.13: Bolt segment showing channels.

89

Figure 4.14: A section of an instrumented bolt showing the strain gauge

89

and wirings through the silicon gel.

xvii

List of Figures (contd....)

Figure 4.15: Various components of an extensometer, a) readout unit,

91

b) probe, and c) photograph showings the process of taking


readings in the underground.
Figure 4.16: Photograph showing the process of taking reading with S B M

93

in the underground.
Figure 4.17: Load transferred on the type II bolts, West Cliff Colliery.

95

Figure 4.18: Load transferred on the bolts installed at the left side of the C/T,

96

West Cliff Colliery.


Figure 4.19: Shear stress developed at the bolt/resin interface of the bolts

97

installed at the left side of the C/T, West Cliff Colliery.


Figure 4.20: Load transferred on bolts installed at therightside of the C/T,

98

West Cliff Colliery.


Figure 4.21: Shear stress developed at the bolt/resin interface of the bolts

99

installed at therightside of the C/T, West Cliff Colliery.


Figure 4.22: Roof condition in the gate roads at Tower Colliery,

100

a) left side, and b)rightside.


Figure 4.23: Load transferred on the bolt T R A 1 , during the panel

101

development and longwall retreating phases, Tower Colliery.


Figure 4.24: Load transferred on the bolt B R A 1 , during the panel development

102

and longwall retreating phases, Tower Colliery.


Figure 4.25: M a x i m u m load transferred on the bolt BRJ2, for a particular

103

face position, Tower Colliery.


Figure 4.26: Load transferred on the bolts T R J 2 and B R J 2 during the panel

104

development phase, Tower Colliery.


Figure 4.27: Load transferred on the bolts T R J 2 and B R J 2 during the longwall

105

retreating phase, Tower Colliery.


Figure 4.28: M a x i m u m load transferred on the bolts T R J 2 and BRJ2, for

106

a particular face position, Tower Colliery.


Figure 4.29: Strata deformation recorded in the extensometer T R 1 ,

108

Tower Colliery.
Figure 4.30: Strata deformation recorded in the extensometer T R 2 ,
Tower Colliery.

xviii

109

List of Figures (contd....)

Figure 4.31: Strata deformation recorded in the extensometer T R 3 ,

110

Tower Colliery.
Figure 4.32: M a x i m u m deformation recorded in the extensometer T R 1 ,

111

for a particular face position, Tower Colliery.


Figure 4.33: Three-dimensional surface generated from the m a x i m u m strata

112

deformations recorded in the travelling road, Tower Colliery.


Figure 4.34: Relative displacements between the two consecutive extensometer

113

probes in B R 2 , Tower Colliery.


Figure 4.35: Load transferred on the bolt B R A 2 , during the longwall

114

retreating phase, Tower Colliery.


Figure 4.36: M a x i m u m deformation recorded in the extensometer T R 2 ,

114

for a particular face position, Tower Colliery.


Figure 4.37: M a x i m u m load transferred on the bolt T R A 2 , for a particular

115

face position, Tower Colliery.


Figure 4.38: Load transferred on the type I bolts, installed in the travelling road, 116
Tower Colliery.
Figure 4.39: Load transferred on the type I and type II bolts, installed at the

117

left side of the belt road, Tower Colliery.


Figure 4.40: Shear stress developed at the bolt/resin interface of the type I and

118

type II bolts, installed at the left side of the belt road,


Tower Colliery.
Figure 4.41: Load transferred on the type I and type II bolts, installed at the

119

right side of the belt road, Tower Colliery.


Figure 4.42: Shear stress developed at the bolt/resin interface of the type I and

120

type II bolts, installed at therightside of the belt road,


Tower Colliery.
Figure 4.43: Load transferred on the type I and type II bolts, installed at the

121

middle of the travelling road, Tower Colliery.


Figure 4.44: M a x i m u m load transferred on type I and type II bolts for a
particular face position, installed at the middle of the travelling
road, Tower Colliery.

xix

122

List of Figures (contd....)

Figure 4.45: Shear stress developed at the bolt/resin interface of the type I and

122

type II bolts, installed at the middle of the travelling road,


Tower Colliery.
Figure 4.46: Three-dimensional surface generated from the m a x i m u m load

123

transferred on type I and type II bolts, installed in the travelling


road, Tower Colliery.

Figure 5.1: Bilinear strength envelope for saw-tooth asperity (after Patton, 1966). 129
Figure 5.2: Idealised penetration of micro asperities of concrete into rock

132

surface (after Johnston and L a m , 1989).


Figure 5.3: Deformation of bolted joint during shearing

134

(after Indraratna et al., 1999).


Figure 5.4: Shear force versus joint displacement (after Bjurstrom, 1974).

135

Figure 5.5: Components of shear resistance offered by a bolt

13 7

(after Bjurstrom, 1974).


Figure 5.6: Arrangement for bolt shear testing (after Hass, 1981).

137

Figure 5.7: Effect of bolt inclination on the bolt shear resistance

140

(after Aydan and K w a m o t o , 1992).


Figure 5.8: Axial and shear stress on bolt at various normal stresses

141

(after Aydan and K w a m o t o , 1992).


Figure 5.9: Resistance mechanism of a reinforced rock joint (after Ferrero, 1995). 142
Figure 5.10: Force components and deformation of a bolt, a) in elastic zone,

144

and b) in plastic zone (after Pellet and Boulon, 1998).


Figure 5.11: Evolution of shear and axial forces in a bolt, a) in elastic zone,

145

and b) in plastic zone (after Pellet and Boulon, 1998).


Figure 5.12: Shear strength of mica infilled j oint (after G o o d m a n , 1970).

147

Figure 5.13: Shear strength of kaolin infilled joint (after Ladanyi and

148

Archambault, 1977).
Figure 5.14: Variation of shear strength with t/a ratio

149

(after Phien-Wej et al., 1990).


Figure 5.15: Effect oft/a ratio on shear strength of infilled joints
(after Papalingas et al., 1993).

xx

150

List of Figures (contd....)

Figure 5.16: Effect of infill on strength envelope, a) joint with asperity

152

angle 9.5, and b) joint with asperity angle 18.5


(after Indraratna et al., 1999).

Figure 6.1: Arrangement for sample casting. 156


Figure 6.2: A typical joint sample.

157

Figure 6.3: A typical joint sample inside the bottom shear box.

158

Figure 6.4: Shear stress profile for non-bolted joints.

162

Figure 6.5: Dilation profile for non-bolted joints.

163

Figure 6.6: Normal stress profile for non-bolted joints.

164

Figure 6.7: Shear stress profile for bolted joints.

165

Figure 6.8: Dilation profile for bolted joints.

166

Figure 6.9: Normal stress profile for bolted joints.

167

Figure 6.10: Variation of percentage increase in normal stress with initial

167

normal stress.
Figure 6.11: Strength envelope for bolted and non-bolted joints.
Figure 6.12: Variation of difference between peak shear stresses for bolted

168
169

and non-bolted joint with initial normal stress.

Figure 7.1: A typical saw tooth joint with infill. 172


Figure 7.2: Shear strength envelope of clay infill.

173

Figure 7.3: Process of placing infill on top of joint surface.

173

Figure 7.4: Installation of bolt perpendicular to the infilled joint.

174

Figure 7.5: Samples after testing; top: non-bolted, and bottom: bolted.

176

Figure 7.6: Variation of shear stress for infilled non-bolted joint, a) according

177

to initial normal stress, and b) according to infill thickness.


Figure 7.7: Variation of normal stress for infilled non-bolted joint, a) according

179

to initial normal stress, and b) according to infill thickness.


Figure 7.8: Variation of dilation for infilled non-bolted joint, a) according

181

to initial normal stress, and b) according to infill thickness.


Figure 7.9: Stress paths for infilled non-bolted joint at an initial normal stress
of 0.13 M P a .

xxi

182

List of Figures (contd....)

Figure 7.10: Variation of shear stress for infilled bolted joint, a) according to

183

initial normal stress, and b) according to infill thickness.


Figure 7.11: Variation of normal stress for infilled bolted joint, a) according

185

to initial normal stress, and b) according to infill thickness.


Figure 7.12: Variation of dilation for infilled bolted joint, a) according to initial

186

normal stress, and b) according to infill thickness.


Figure 7.13: Stress paths for infilled bolted joint at an initial normal stress

187

of 0.13 M P a .
Figure 7.14: Shear stress variation with shear displacement at 2.5 m m infill,

188

for bolted and non-bolted joints.


Figure 7.15: Shear strength envelopes of bolted and non-bolted joints

189

at 2.5 m m infill.
Figure 7.16: Dilation profiles for bolted and non-bolted joints at 2.5 m m infill.

189

Figure 7.17: Variation of % drop in peak shear strength with infill thickness.

190

Figure 7.18: Variation of shear stress with shear displacement for bolted

191

and non-bolted joints at CT0 = 0.63 M P a .


Figure 7.19: Relative location of shear plane through infilled joints

192

at o"no = 1.25 M P a .
Figure 7.20: Stress paths and strength envelopes for 2.5 m m infill

195

thickness (t/a=0.5) for bolted and non-bolted joints.


Figure 7.21: Stress paths and strength envelopes for 7.5 m m infill

195

thickness (t/a=1.5) for bolted and non-bolted joints.

Figure 8.1: Periodic function f(x) with a period of 2T. 200


Figure 8.2: Dilation profile segmented into m equal parts.

201

Figure 8.3: Orientation of bolt with respect to joint plane.

204

Figure 8.4: Variation of peak shear stress with t/a ratio.

207

Figure 8.5: Variation of Normalised Strength Drop ( N S D ) with t/a ratio.


Figure 8.6: Straight line formulation of N S D ; a) for non-bolted joint, and

208
210

b) for bolted joint.


Figure 8.7: Comparison of shear stress profiles predicted by the model and
from laboratory tests for unfilled joints.

xxii

213

List of Figures (contd....)

Figure 8.8: Comparison of dilation profiles predicted by the model and

214

from laboratory tests for unfilled joints.


Figure 8.9: Comparison of peak shear stress predicted by the model and

215

from laboratory tests for infilled bolted joints.


Figure 8.10: Comparison of normal stress profiles predicted by the model

215

and from laboratory tests for infilled bolted joints (t=1.5 m m ) .


Figure 8.11: Comparison of dilation profiles predicted by the model and

216

from laboratory tests for infilled bolted joints (t=1.5 m m ) .


Figure 8.12: Conceptual model representing the mechanical behaviour of

218

fully grouted bolt.


Figure 8.13: Line diagram showing the axial behaviour of bolt in U D E C .

220

Figure 8.14: A typical bolt element passing through triangular block.

221

Figure 8.15: Line diagram showing the shear behaviour of grout in U D E C .

222

Figure 8.16: Conceptual C N S model of reinforced joints.

226

Figure 8.17: M e s h generation by U D E C .

229

Figure 8.18: Comparison of shear stress profiles predicted by U D E C model

231

and from laboratory tests.


Figure 8.19: Comparison of dilation profiles predicted by U D E C model

232

and from laboratory tests.


Figure 8.20: Comparison of normal stress profiles predicted by U D E C model

233

and from laboratory tests.


Figure 8.21: Comparison of shear stress profiles predicted by U D E C model
under C N L and C N S conditions.

xxiii

235

LIST OF TABLES

Tables

Page

Table 2.1: Description of various types of bolt.

11

Table 2.2: Description of various bolt accessories.

12

Table 2.3: Geomechanics classification guide for excavation and support in

18

rock tunnel (after Bieniawski, 1987).


Table 3.1: Specification of bolt types.

54

Table 3.2: Calculated values of a, b and c.

75

Table 7.1: Dilation or compression of joints for various t/a ratios.

xxiv

193

LIST OF SYMBOLS AND ABBREVIATIONS

Symbols

Asperity height

&n5 b n

Fourier coefficients

Area of joint plane

Ab

Bolt cross sectional area

Young's modulus

hxp

Shear displacement at peak shear

Asperity angle

kb

Bolt stiffness

kn

Bolt-joint composite stiffness

kr

Joint stiffness

ks

Joint shear stiffness

Infill thickness

Period of cycle

a,P

Hyperbolic constants

8V

Joint dilation

8v(h)

Joint dilation at any shear displacement

<|>b

Basic friction angle

^bn(h)

Effective normal stress on the plane of bolt/joint composite at any shear


displacement

0"c

Uniaxial compressive strength

o-n

Normal stress

O"no

Initial normal stress

a n (h)

Normal stress at any horizontal displacement

o"t

Tensile strength

Shear stress

Peak shear stress

XXV

tp infilled

Infilled joint shear strength

t P unfilled

Unfilled joint shear strength

Ax p

Drop in peak shear strength

Inclination of bolt with respect to the joint plane

Abbreviations

BSM

Bond Strength Model

CNL

Constant Normal Load

CNS

Constant Normal Stiffness

JCS

Joint wall Compressive Strength

JRC

Joint Roughness Coefficient

NATM

N e w Austrian Tunnelling Method

NSD

Normalised Strength Drop

OBE

O p t i m u m Beaming Effect

RRU

Reinforced Rock Unit

SBM

Strain Bridge Monitor

UDEC

Universal Distinct Element Code

UOW

University of Wollongong

USBM

United States Bureau of Mines

xxvi

Chapter 1

INTRODUCTION

1.1 GENERAL BACKGROUND

Since early civilisation, man has been involved in the development of stable
infrastructure, and in the process, a number of methods have been developed and
evolved to stabilise the man-built structures. In mining engineering and underground
rock structures, bolting is considered to be one such technique for improving the
structural stability. According to Kareby (1995), rock bolting in various forms has
been used as early as the nineteenth century. Development has progressed from
wooden dowels (nineteenth century), to expansion shell bolts (early twentieth
century), and subsequently to cement and resin grouted rebars (1950) and friction
stabilisers such as swellex and split set bolts (1980).

However, in Australia, rock bolts were not used extensively until the construction of
Snowy Mountain Hydroelectric Scheme in 1948. Since then, stabilisation of rock
mass by systematic bolting has become a de facto standard of stabilising the
superficial workings (e.g. slopes, dams etc.) as well as the underground excavations.
Hundreds of millions of bolts are installed annually in various construction projects,

Chapter 1: Introduction

and a recent survey revealed the worldwide usage in excess of 500,000,000 units per
annum (Windsor, 1997).

In recent years, the range of application of rock bolts has been broadened both in
mining and civil engineering. This is due to advances made in the understanding of
bolt failure mechanisms as well as the improvement attained in strata control
technology. This is clearly demonstrated by the increased use of rock reinforcement
in underground excavations as an alternative to traditional forms of support such as
timber, crossbar, wooden chock etc.

The rock bolt research gained momentum following the introduction of New
Austrian Tunnelling Method (NATM) in the early sixties, where fully grouted bolts
were employed as a major support component in rock tunnels. Bjurstrom (1974) and
Hass (1981) conducted systematic laboratory studies on the behaviour of fully

grouted bolts to predict the shear strength of bolted joints as a function of the fricti
angle, normal stress and bolt inclination. It was concluded that inclined bolts were
stiffer, and that they contributed more effectively towards enhanced shear strength of
discontinuities than the bolts installed perpendicular to the joint plane. Based on
shear testing of model gypsum joints, Dight (1982) found that normal stress acting on

the joint plane had little influence on the bolt shear resistance, and that joint dilata
was related to the bolt inclination. Indraratna and Kaiser (1990) presented a
comprehensive analytical model for predicting the stability of circular tunnels with
grouted bolts, where the convergence was determined as a function of the bolt
spacing, the tunnel diameter and the bolt friction. In a recent paper, Pellet and Egger

Chapter 1: Introduction

(1996) introduced an analytical model for evaluating the role of bolts on the shear

strength of rock joints. The interaction of the axial and shear forces mobilised in the
bolt as well as the large plastic displacements of the bolt during the loading process
were considered in the model. However, none of these studies include the influence
of bolt surface (profile) configuration on the load transfer mechanism of grouted
bolts.

During the last three decades, various researchers have extensively examined in the
laboratory, the shear strength parameters of both natural and artificial infilled rock
joints (e.g. Goodman, 1970; Kanji, 1974; Ladanyi & Archambult, 1977; Lama, 1978;

Barla et al., 1985; Bertacchi et al., 1986; Pereira, 1990; Phien-wej et al., 1990 and d
Toledo & de Freitas, 1993). All these tests were carried out under the conventional
Constant Normal Load (CNL) conditions, and not under Constant Normal Stiffness
(CNS) conditions, as there was insufficient understanding of the relevance of CNS
testing in underground mining conditions.

Johnstone & Lam (1989), Skinas et al. (1990) and Indraratna et al. (1999) have
described the importance of CNS condition to simulate the actual shear behaviour in
the field. For non-planar discontinuities, shearing often results in dilation as one
asperity rides over another. If the surrounding rock mass is unable to deform
sufficiently, then an inevitable increase in the normal stress occurs during shearing.
Therefore, the CNL condition is unrealistic in circumstances where the normal stress
in the field changes considerably during the shearing process, such as in some
underground mining situations. The deformation behaviour of a bolted joint is

Chapter 1: Introduction

governed by the applied normal stress, the stiffness of the joint-bolt composite, and
the frictional properties of the joint interface. The CNS method is more realistic than
the conventional CNL approach in such situations as discussed by Indraratna et al.
(1999).

Leichnitz (1985) was the first to report on the laboratory studies of rock joints under
CNS condition, and he verified that both the shear force and dilation were functions
of the normal force and shear displacement. Since then, similar findings have been
reported by Johnstone & Lam (1989), Van Sint Jan (1990), Ohnishi and Dharmaratne
(1990), Skinas et al. (1990) and Haberfield & Johnstone (1994). More recently,
Indraratna et al. (1999) discussed the shear behaviour of soft joints under CNS
condition for both clean and infilled joints, where an analytical model was developed
to predict the shear behaviour of soft joints using Fourier transforms.

Despite the significant advances made in strata reinforcement technology, the


mechanisms of bolt-rock interaction are still complex and they remain to be a grey
area for research. To the best of writer's knowledge, no research work has been
reported on the influence of the surface geometry of bolts on the load transfer
mechanism, and the effect of infill material on the overall shear strength of bolted
joints under CNS condition. Accordingly, the research study reported in this thesis is
aimed to address various issues in order to provide solutions to the complex problem
of bolt and rock interaction. In particular, the study will focus on the shear strength
bolt/resin/rock interface and the shear behaviour of infilled bolted joints.

Chapter 1: Introduction

1.2 KEY OBJECTIVES

The objectives of the present research work include:

The study of the shear behaviour and load transfer mechanism in


rock/grout/bolt interfaces both in the laboratory and in the field.
The study of the influence of bolt surface geometry on the load transfer
mechanism of fully grouted bolts.
The study of the shear behaviour of both unfilled and infilled bolted joint
mainly under CNS condition.
Development of an analytical model to predict the shear behaviour of both
unfilled and infilled bolted joint, and
Numerical modelling of bolted joint under CNL and CNS conditions using
Universal Distinct Element Code (UDEC).

1.3 OUTLINE OF THE THESIS

The thesis consists of nine chapters and is organised as follows:

This chapter has provided a brief introduction of the development and evolution of
rock bolting technology. Chapter 2 is concerned with the present support design
philosophy for rock engineering. It contains an examination of the current roof

Chapter 1: Introduction

bolting practices in thefieldand the popular rock bolting theories currently employed
in practice. A brief description of available approaches for the determination of rock
loads and the associated design procedures are also discussed.

Chapter 3 describes the experimental studies undertaken on the shear behaviour and
failure mechanism of the bolt/resin interface of fully grouted bolts. It contains the
principles and the detailed design of the large-scale CNS direct shear apparatus,
sample preparation and the experimental procedures adopted for the study. The shear
behaviour of flattened bolt surface of two most popular bolt types is examined under
CNS condition with initial normal stress levels of 0.1 to 7.5 MPa. The influence of
bolt profile configuration on the effective shear strength at the bolt/resin interface is
described together with the failure mechanism of the bolt/resin interface.

Chapter 4 includes the analysis of field investigation carried out in two coal mines in
the Illawarra region of NSW, Australia. As extensive instrumentation procedure of
bolts, site description, field installation procedures, description of field monitoring
instruments and data collection are contained in this chapter. The results from
eighteen instrumented bolts installed at three different mine sites are elucidated. The
chapter also contains the guidelines for selecting a suitable bolt surface profile under
various confining stress conditions.

Chapter 5 is devoted to reviewing the influencing factors governing the shear


behaviour of both unfilled and infilled bolted joints. It contains a review of selected
past studies, the existing models simulating the shear behaviour of joints and limited

Chapter 1: Introduction

literature on infilled joints tested under C N S condition. T o the best of writer's


knowledge, no suitable literature was available to report on the shear behaviour of
infilled bolted joint under CNS condition.

Chapter 6 presents the result and discussion of the shear behaviour of both unfilled
bolted and non-bolted joints under CNS condition. The interpretation of data relates
to stress paths, stress-strain relationship, strength envelope and dilation behaviour of
unfilled joints. The bolt contribution on the shear strength of unfilled joint is also
described.

Chapter 7 presents the results and discussions pertinent to the CNS behaviour of both
infilled bolted and non-bolted joints. The analysis of the test results was conducted
in relation to the shear strength, dilation and normal stress responses. The effect of
infill thickness on the overall shear strength of both bolted and non-bolted joint is
explained. The influence of bolt in relation to the infill thickness on the shear
strength of bolted joint is also described. The concept of Normalised Strength Drop
(NSD) associated with the increase in infill thickness for bolted and non-bolted joints
is introduced.

Chapter 8 introduces an analytical model representing the general shear behaviour of


bolted joints applicable to both unfilled and infilled bolted joints under CNS
condition. Based on the Fourier transforms, the dilation response during shearing
process of bolted joint is accurately modelled, and finally, a model for the prediction
of shear behaviour of unfilled bolted joint is developed. The shear behaviour of

Chapter 1: Introduction

infilled bolted joint is modelled using the unfilled joint model and the N S D . A
comparative study of the current model predictions with laboratory results, as well as
the results from UDEC analysis are presented at the end of the chapter.

Conclusions and recommendations are given in the final Chapter 9. It summarises the
findings of this research study in relation to; (a) the influence of bolt surface
geometry on the overall shear behaviour of bolt/resin interface, (b) the selection of a
suitable bolt profile configuration depending on the prevailing in-situ stress
condition, (c) the shear behaviour of unfilled bolted joints, and (d) the shear
behaviour of infilled bolted joints. Recommendations on the possible future lines of
investigation are also presented in this chapter. A list of references and appendices
follow Chapter 9.

Chapter 2

ROCK BOLTING PRACTICES

2.1 INTRODUCTION

Since its introduction as a support system to mining and civil engineering, rock
bolting has evolved through different phases of technological transformations. In the
early stages, wooden dowels were used to prevent roof falls in coal faces in the
British Coal Mines (Beyl, 1945-46). In recent years, a variety of rock bolting systems
have been developed and, in fact, they have become the most common method of
ground support in most Australian mines and most rock engineering operations.

The early studies on the fundamental of bolting techniques date back to the studies
carried out by USBM in 1948. Since then, substantial research has been carried out,
covering various aspects of bolting technology such as design philosophy, bolt
configuration, bolt load transfer mechanism, installation procedures for effective
ground stabilisation and so on.

This chapter is concerned with the literature review of rock bolting technology. It
provides a critical review of the bolt types, its application and theories related to

Chapter 2: Rock Bolting Practices

bolting and contemporary research work carried out in this line. Particular emphasis
is given to the behaviour of rock bolt under Constant Normal Stiffness (CNS)
condition, as CNS is the primary basis of the research work undertaken in this thesis.

2.2 REVIEW OF TYPICAL ROCK BOLTS AND ACCESSORIES

Although a number of different types of rock bolts and its variants are in use today,
they may be broadly classified into the following four major groups:

Mechanically anchored rock bolts


Frictional rock bolts
Grouted rock bolts
Grouted cable bolts

Table 2.1 shows a brief description of each type of bolt together with the relative
advantages and disadvantages.

In addition to the bolts used in a reinforcement system, a number of other


components are also used to make reinforcement systems more effective, as well as
for obtaining better load transfer characteristics. These include face plate, antifriction washer, w-strap, wire mesh etc. Table 2.2 shows a brief description of each
component.

10

a fa
a

E
E
o
U

2
E a.
o x_ 5P

c -S
S o _;

fl
a> ^_.
o c

co ...

__ _>
-5
cu O
__
co

CU
bO
CS

CVD

a I"3

u
E CO
- o
a.
_> O OO
- 5 To
>J ON
<u ,,

o g.

9J

> 3
CO

o o
o -o

o S 5
" o o
J

U J3

E
es
_.
wo
es

n
_

Eli

c
o

C/3

*'l

>

CO

_
00

T3

es
-5

n)

bt)

cs

CS

-__! O.

C_
0) c -D
C3
co o -41 CO

*I

o ^3
-o o
_> o

1>
1

a>

-J

-a
<i>

WD
es
a
es
>
"O
<

cs eg

CS cti

>> <U

- "5 T3 <u 4_.

C+H

CS

>
o

cs __> ._> __ . 2

cS

_> o

fl u

f3

-O -rt

c_

cS E
_9
o

o
CS
<U
>
DcS
U . J
^_
Q , CD " o
_:-__

S _o o
_=
* _5
E a.ac
'
Q __ o
<u a

cS cs

X)

uo

CS

a
CS

cS
si o
+CO

__ o
fl
^3
S
i/i o

o o

c
u
'E o_S 3
_o g jS TJ
_0 CS
CL,
1/0

'to 3 <3 O
fl cr oo p
<_> G fe

ILI

u .13 C_

__

5 . __

2 -

OH

3
co

- -a

<J

-w
CO

. ^*

& hE** o -
_^

*_bl> o

C
O

O (
_

^
X)

3 '-5

WJ

e
o

_>
o
.E o &"j
"a. i w g
. cs fa u

"S " o

c
o
oo
^^>n
_
u
(U
_.
.>
u
<
U
fl
o
o CS X) CO
fl

:A

._? co

CS

a. +j
u
o
o
5 -3
_ r>
CO
_b S
u _

_J "" N _o

4J X

VJ

. o
Q_) U

C *H

_-> u - ei
O

_) CB

5 o
C_ O

Q.
fl

es it

u
TJ
_2
CS

"-_

cS

2 _ro
"5 "O

<

a.
"o

*o -a

_*J

cS o
_n o

a
cs
o
'fl
cS
_=
o

CS
oo
s_>
>

o
r-' cS

_
>

'3

'
c

<L> co
=_ oo oc
c.
u
0J
p.
00 "* X
_J

-J <_
GO CO

o
__
-a
CU
3
O

_.

c
<_

s
CJ

CJ

3
O

.
0J oo

h _3

_0 g

o
3 0009 03243424 8

T3
CU

__

2 1
o 5

o
X __ __ 3 .-t_

_. CQ DC

CU

,3

-*>

o
x>

CU

Pu
>>

<u
o

s
O

oo

l~c

.
G

ti

io

cr
o

-*J

oo
O

X3
-4

-G

oo

TJ

<U

CO

P,
-G
3

00

ft
>-,
-*

CU

"o
X>

,o
C+-I

CO

oo

u
U)
c.
-3

E
o
u

-^
<u
o

>-,
G
c.

c
<u
(U

3
O

,3

<U

-I1

"3

-t->

3
<u
I-c
o
G

'5

_>
-3
O

1-1

(U

oo

CO

jd

CO

TJ

O
O

C+H

-4'

_>

,-i

G
<u
(U
TJ

'>

<u
G
O
G
o

o
OH

o
c

-O
(U
_>

(U
rO

<u
-G
G
*-*

_fi
TJ
U
00

3
G
<U

o
o

3
G
3

%>

k.

s
cs
u

- /

WD

* '

CS

>
(N

u
E5

o
co
co
<D
U

o
o

o
_>

ft

c
o
CO

Q
ri
c4
<u

o
o
o
_-)

y
3
TJ
<U
s-

u
-*

co

CJ

_2
"Hi
(U

o
3
(X,

oo

__:
<u
o
<+-4

TJ
G
(3
-+^

G
G

'-G
H
3

G
Dt

.!H.

o
O

ss
G

S-i

O
G
O
O

<4=l

TJ
(U
1 \
G
OH

O
-4

<U

O
CO
t-l

(
U
OH

O
^

fr
3

'3T>

G
-G
G

<
>u
CU
!H

G\
0

00

i>

oo
U

o
'3

G
O

p*

s_

CU

__!
(U

G
G
O

TJ

>
O
G

>.
43
(U
o
,3
C+H

<u
G

-4
00

Jh,

-(>

-4

CU

3
*

_o

>J

4=>

CU

WD
es

TJ

<u

3
TJ
ed
O
u
.G

<u
-G
co
3

a,
3

<U

u
1-

>>

>\

;;
3

s
s
00

00

(U
-^
*3Xi

^->
G

TJ
_>
GH

C+H

<u

E
a,
G

c
(Ur
0

3
p00

sO

O
-.
O

<+H

__
3

_:

3
TJ

00

(U
O
(U

O.

Chapter 2: Rock Bolting Practices

2.3 R O C K BOLTING THEORIES

Many researchers have put forward their opinions as to how roof bolts act to support
the immediate roof. The most widely accepted theories include:

Suspension theory
Beam building theory
Keying theory

Whenever an underground opening is made, the immediate strata directly overhead


tend to sag. If not properly and adequately supported, the laminated immediate roof
could separate from the main roof and may collapse. Roof bolts, in such situations,
nail the immediate roof to the self-supporting main roof by the tension applied to the
bolts. Suspension theory assumes that the rock near the excavation is weak and
fractured, however, rocks that are high above in the roof are relatively stronger in
comparison with the immediate roof rocks. The length of the bolt should be long
enough to penetrate into the stable rock above the immediate roof. Bolts should be
strong enough to hold the dead weight of fractured rock beneath the stable strata.
Figure 2.1 shows the action of bolts as the basis of the suspension theory.

In many cases, the strong, self-supporting main roof may not necessarily be located
within the distance that ordinary roof bolts can reach to form firm anchors for

suspension. In such a situation, it is assumed that the bolt acts by laminating thin and
weak individual layers of rock together into a thicker and stronger beam of rock.

13

Chapter 2: Rock Bolting Practices

Such a system will transmit horizontal shear force from one layer to another and
reduce the horizontal displacement between layers. When the theory is applied in
respect to tensioned bolts, the frictional resistance between the thin layers is
increased due to the increased normal force acting perpendicular to the bedding
plane. Non-tensioned fully grouted bolts will also transmit shear forces between the
bedding planes as a result of the shear stiffness of the bolt/grout composite. Figure
2.2 explains the action of bolts according to the beam building theory.

Strong
competent
strata
Weak rock in
immediate roof

Figure 2.T. Suspension theory.

Composite
beam

Figure 2.2: B e a m building theory.

14

Chapter 2: Rock Bolting Practices

W h e n the roof strata are highly fractured and blocky, or the immediate roof contains
one or several set of joints with different orientations, roof bolting significantly
increases frictional forces along fractures, cracks and weak planes. Sliding and/or
separation along weak interface is thus prevented or reduced as shown in Figure 2.3.
The keying effect mainly depends on active bolt tension or under favourable
circumstances, passive tension due to rock mass movement. It has been observed that
bolt tension produces stress in the stratified roof, which are compressive both in the
direction of the bolt axis and orthogonal to the bolt. Superposition of the compressive
area around each bolt forms a continuous compressive zone in which tensile stresses
are offset and the shear strength is improved.

Highly fractured
and blocky
immediate roof

Figure 2.3: Keying theory.

15

Chapter 2: Rock Bolting Practices

2.4 SUPPORT DESIGN PHILOSOPHY

The primary concern regarding the support design is that whether the structure under
consideration is self-supporting, and if not, what kind of strategies must be followed
for the overall stability of the structure during the entire life span of the roof. The
selection of support members is not only closely associated with their mechanical
functions but also with the availability, suitability and economic viability of the
support elements. Currently, the support design philosophy consists of two
fundamental steps:

Estimation of rock load


Design of suitable support system

2.4.1 Estimation of Rock Load

There are several approaches to determine the load on rock and these are classified
into following groups:
Rock mass classification approach
Semi-empirical approach
Structural defect approach
Rock-support interaction approach

In the rock mass classification approach, the immediate roof is classified into
different categories depending on the values assigned to various structural and

16

Chapter 2: Rock Bolting Practices

geological parameters of the roof rock and the method of excavation (Terzaghi, 1946;
Deere et al., 1969; Bieniawaski, 1973; Barton et al., 1973). Over the years, the rock
mass classification approach has gained increasing popularity because of its
simplicity of use. Table 2.3 show the guidelines for geomechanics classification for
excavation and support in rock tunnels as proposed by Bieniawaski, 1987.

According to the semi-empirical approach it is assumed that a ground arch around


the opening occurs and the boundaries of the loosening zone are restricted between
the ground arch and the excavation boundary. The fundamental reasoning for this

assumption originates from the fact that the rock mass is incapable of resisting tensile
stresses and is sufficiently strong while under compression. The original theory was
proposed by Janssen (1895) for soils and was applied to tunnelling by Terzaghi
(1946).

The structural defect approach involves the determination of the dimensions of


potentially unstable blocks or layers of rock in relation to the special orientation of
discontinuities and the geometry of the opening. The total potential volume of rock
prone to fall will vary depending on the number of sets of joint planes and the tensile
strength of the rock or the shear strength of joint planes, whichever is weaker. When
the number of joint sets is two or more, the determination of potentially unstable
blocks sliding or toppling into the excavation, can be determined by block theory
(Wittke, 1964; John, 1970; Londe and Vormeringer, R., 1970, Goodman and Shi,
1984). This theory can effectively evaluate the possibility of sliding or falling of
unstable rocks and resulting loads.

17

CO

_>
*_: CO

tiO
,

VI
-t

'
X)

>.-=
>

C3
<U
-3
O in

VI
4i

CO

2 cu
cr
3u. -

OO
<u

CU
3

to

t
<U
"o _3
>
a.

<u <U
_.
O

3
3

u
3

<u
to
-3 3 _
-5 cu '5b '3a
<u o
to u
OS
_> D.
.2 _.

*- E
abJH

c
o

CU

-O

CO

T3

-a

(U

*_

_.
+-

t
o

o
in

3
__

'3

'5
o_. cr

cr
.
<u

to

Q.
3
CO
D.
O
3

<u
_>
!_
.

CU

o
in

oin

o
o
. "3
_. 3
E
3
3

_>
_.

co
CU

3
P .

'co

cu

IS
c75
_

^ &
o
6

o
10

_=

CO

o
o
CN
i
KTi

o
___(

3
in

3
<U

Jfl
CO

O
_.
o

3
'% __

O
3

<U

i_

3 _o CO ' '%
D. CO +; i n .3
O
~
co 3
SO O "3
o
O (U CO
3 o

_c
CO
o
*_

__
_>>
co
cd

3
3

CU

o
E 3

3 "ed

__

CO

-4

__,

t_ w

'1

<u
o -

"5

_-

_D
O

ed

CO

CO

CU

'

CO

B 9- 'I

g>^
t

in

.a "S__

'13 -

o
J

CO
CU

3 '

CO

_ !

c.

E
E S <u

'_!

?.

bO g

CO

-D

CO

3
3
3

CO

3
O

"o

_5

-a

CU
.

a>
bo
2
_r
in

r-i

CU
0

fl" _>

3
"3
3
3
3

CU

"3
3
3
CO

>>

CS
D.
CO

to

3
O

_J

"5 _3<u>

00

"5
EQ
T3
3
3

>^

CVO

00

*-?
CO

CU

ft
bO

CU
Q.
O

*3

<U

>
St
o

o
3
3

(U
O
3
3

>

-a

"3
<u 3

*"~|

CU

(U

CU

_.
cC
3

__

"3

CO

j - 4>
O
O
C
3
<U 3
-O
>
-rt "3
CU 3

_> __
to
C <U
3 CS
3
Q.
3
3 O

_?-

3 o
_>

Q.
J n. o
O
o
H
H

'

__!
O

Vi

CJ

a
u
*o

o
o

>
-3
_.
O oo
O
SO (_.

o
oo

.M -1,

Ho

3 __
OH __

<u _>
> c_ 3

o
O

__
o
o c_
'._

8^

Eg
5

to __

'1 o

T5

3
O
O

o
a. 3
CO
o. -3
i

"ed
*i

CO

_.

_. c_
o 5
o ^
>
c_ c_

<C
3

'5

3
(U - J '
J_ 3
(U
O.
.
3
o s_

<U

-*i

CO

-a

a.

CO

fl

3
3
O
O

u
o

3
3

t:
> o
-a CO
D.
3

'S5

CO
CO

a. >
"3
>
3 3
CO
3 _Q. i n
_; O(U
(U
i
o
_J in

_E

Vi

_.
<u

__

cS
o
Q. fc E
_3 3 o
O

a
E
ii
-5 <? 3E
'CU
o

o
en _" ~ U
O U

5
3

-*> t;
ii .
o c cu o o
Q.
a. Q.
Q.
D. i
a.
3
3 TJ >
3
CO
CO

_>
3 __
in

>

to

_ ,3

-5 o3
o
3

E
o
S
^O

CU

_s o
_o

__
o

o o
o V
o
Q.

<u __
> G_

>

3
CU

bD
fc- 3
O

_= _2
00 JO

Chapter 2: Rock Bolting Practices

In the rock support interaction approach, the rock load is based on the interaction
between rock mass and support system. The basic idea of this approach was first
suggested by Fenner (1938) and was introduced to the real tunnelling practice by
Rabcewicz (1964) and was modified by Deere et al. (1969). Figure 2.4 shows a
typical ground reaction curve for rock tunnels. The ground reaction curve displays
the load that must be applied to the roof or the walls of the tunnel to prevent further
movement. This approach later gave birth to a new technique of tunnelling known as
the New Austrian Tunnelling Method (NATM). The principle of this approach is well
illustrated by a ground response - support reaction curve. The load supported by the
support members will be the one at which ground response curve intersects the
support reaction curve.

AC - properly designed support:


equilibrium at C
massive rock mass

OD - radial deformation for


stable unlined tunnel

jointed rock mass

AeE - support yields before


stabilizing opening
AF - support too flexible
6H - support too delayed

ground reaction curve


C

ground arch
railing

it-

B D
6
RADIAL DEFORMATION, u

Figure 2.4: T h e ground response curve (after Deere et al., 1969).

19

Chapter 2: Rock Bolting Practices

2.4.2 Design of Suitable Support System

Once the rock load has been calculated using any of the methods discussed above, a
suitable support system may be designed depending on the availability, working
conditions and the economic viability of the support elements. It has been a common
practice to support the longwall gate roads by means of full column resin bolts along
with wire mesh and W-straps. The support density depends on the estimated rock
load and the length and capacity of the bolt. Figure 2.5 shows the flow chart of a
typical support management plan for a mine. It consists of four distinct phases:

Design
Implementation
Monitoring
Response

The design phase includes the selection of a suitable support system depending on
the existing geotechnical data at hand. In the implementation phase all the designed
support elements are put forward into the appropriate locations. Once the support
system is installed, the next step is to monitor the performance of the support system.
The response of the support design team will depend upon the analysis of data
obtained from monitoring. The necessary action required at various stages is
explained in Figure 2.5.

20

Monitor performance of
roadway behaviour and
support system

c
'J. w_

2 .a

<

>>

c
<

_
E-t

<

n
V
a

d
u

in

_3 1- I
c
* E f=
b
_ _
_
CO
>->
cc- s
*
u

_
u
bB
cs

CO

u.

ON

s_
^o
"o
o

o
D-

Implement

^1

C
'C cfl

c
o
U

cd

ta
is
&_

2 _ 3>
_
Q

_J o
_

u
ct!

o
-3

cd
C
cd

V-

s -a
s _
co
W 2 00
S

-_
C
_

O S
_5 c

tS 2
w S

_
_ >CS
_3 J3
_
J_
sa
E _
_ n.
_

05
'_

co

_ E
_

C ca
< Q

3
O

"_ co

(J J3 .
< 3CO ."C
ti
H
CA c. c

CO

c*O

cd

o
OH

CO
in

rsi
u

I-I

cs
a

Iff

-_
CO

3 wa
O

8 __>

3 *
_ cs
fi t_
" tj a

& s -_c _

.5 '3

|_.

M pj_>
DO

c
< B

5
o
_!

_>

s _o
M 60 O

c.-I
2
i c
2 2
-2 3
S3 ;s
_!

O. O w
u co > _
04* E j Q 3
C_ C_ _.

cs
no
"3
co
<u

Sill

TJ
C
CS

r>n
_ c
o E
C
co
2
J3

a _

_ JCaan
EC3 c*.
_ o
.O
O

- _
nn u >
->
4J
CO
_
9) to J2
'_
a.

>
c

CO

-1
ret

CO

CJ C
D.

<N

Chapter 2: Rock Bolting Practices

2.5 REVIEW OF R O C K BOLT RESEARCH

Traditionally, geomechanical problems are collectively examined using a


combination of experimental, analytical and numerical methods. However, the
selection of any appropriate combination of approaches is dependent on the
prevailing properties of the resulting physical phenomenon. Numerous investigations
have been carried out in the past to study the effect of reinforcement by rock bolts.
Some of the selected research work, mainly in the area of load transfer and failure of
rock bolt system, are described in the following paragraphs. The research work in
relation to the shear strength of reinforced joint is reviewed in Chapter 5.

The first systematic study on the failure of rock bolt system was carried out by
Littlejohn and Bruce (1975). Three modes of failure of rock bolt system has been
suggested, which are attributed to:

Failure of rock mass,


Failure of rock bolt, and
Failure of bolt/grout/rock interface.

All three modes of failure are shown in Figure 2.6. The estimated bolt load was
given by:

P = 0.lGc7trL (2-1)

22

Chapter 2: R o c k Bolting Practices

where,
P = m a x i m u m normal force in bolt,
a c = uniaxial compressive strength of surrounding rock,
r = borehole radius, and
L = bond length.

/ > v > > > * I J W J * y.'. *.'_ '*. '_y


s % V. \ \ \ > : i A S S S ; S S S

:0! /
\

\__V

c)

b)

a)

Figure 2.6: Failure modes of rock bolt system a) failure of rock mass, b) failure of
bolt/grout/resin interface, and c) failure of bolt (after Littlejohn &
Bruce, 1975).

Farmer's (1975) mathematical model on stress distribution along the encapsulated


bolt length compared favourably with pullout test results of the instrumented bolt.
Farmer (1975) proposed that, the shear stress in resin annulus is a function of bolt
extension and modulus of rigidity of the grout as follows:

; When

(R-a)

23

R -a <a

(2.2)

Chapter 2: Rock Bolting Practices

" aln(R/a)G*

;When

"a>a

where,
xx = shear stress in the resin annulus,
c^x = extension in the bolt,
a = radius of bolt,
x = distance along the length of bolt starting at free end of grout,
R = radius of borehole, and
Gg = shear modulus of grout.

Shear stress distribution along a typical resin anchor was given by the following
equation and is shown in Figure 2.7:

o.2-'

=
<*_

0Ae

(2.4)

where,
CTO = stress on bolt at x equal to zero.

The pull test showed that there were stress variations along the bolt length.
Maximum stress values were recorded at the free end and the minimum at the far
end. This was attributed to the interaction between the bolt and grout interface. The
24

Chapter 2: Rock Bolting Practices

theoretical stress distributions were compared favourably with the laboratory results
as shown in Figure 2.8.

~~fci

___.
Oo

002

O04

006

006

01

- *0I exp {-0-2-g-)

L=23o

Figure 2.7: Stress distribution along a resin anchor, a) stress situation, and b)
theoretical stress distribution (after Farmer, 1975).

25

Chapter 2: Rock Bolting Practices

Pull out failure


IOO0
-- 800
-- 600
-- 400

3>
a
"3
u
C/3

200

500

400

300

200

100

Distance along rod, x ( m m )

a)
3000

20O0

CM
ii

_.
*->

1000

!
-.
cs
0)

500

400

200

300

100

t/3

Distance along rod, x ( m m )

b)
Figure 2.8: Load displacement, strain distribution and computed shear stress
distribution curves - 500 m m resin anchors in limestone, a) strain
distribution, b) stress distribution (after Farmer, 1975).

Farmer's mathematical model w a s supported by the work of D u n h a m (1976) w h e n


carrying out pullout tests on strain gauged instrumented bolts, particularly at small
pulling force. However, as the load increased, the contact between the bolt and the

26

Chapter 2: Rock Bolting Practices

grout w a s lost resulting into a drop in the value of shear stress along the increasin;
length (see Figure 2.9).

C/3

cd

100

wo
ioo
Embedment ( m m )

200

Embedment (mm)

a)

b)

Figure 2.9: Results from instrumented bolts a) normal stress in the bolt, and b) shear
stress at the bolt/grout interface (after D u n h a m , 1976).

Use ofPlasticity and Limit Equilibrium Theories

Adali and Russel (1980) conducted studies on physical models of circular tunnel with
grouted bolts to calculate the size and number of bolts. Outer stress conditions were

27

Chapter 2: Rock Bolting Practices

used as an input to the model. A potential function V was introduced in the basic
equilibrium equation, and accordingly:

d(Jr

dr

<Jr -<70 ^
+
+ Q = 0

d2<f>

1 d(f>
r dr dr dr

dv

where,
crr = radial stress,
ere = tangential stress,
r = distance from centre to point of interest,
Q = radial body force per unit volume, not due to gravity,
V = potential function, and
<|> = stress function.

The whole space was divided into three regions as shown in Figure 2.10. These
regions were described as the fractured zone or the protective zone, influence zone of
grouted bolts, and the outer zone respectively. A solution was provided for the state
of stress and deformation in each region. Although an elastic model is a rather rough
approximation of the actual situation and tend to predict values that are higher than
the actual values, nevertheless, it gave a general idea on the properties of a tunnel
boundary conditions supported by grouting. It was found that if the rock is not highly

28

Chapter 2: Rock Bolting Practices

fractured, the protective zone becomes larger and the pressure on the tunnel wall
increases, thus requiring greater number of bolts for effective support.

Figure 2.10: Zones of influence of a grouted bolt surrounding a tunnel (after Adali
and Russel, 1980).

Lang and Bischoff (1981) studied the stability of reinforced rock structure and
developed the concept of Reinforced Rock Unit (RRU) which was analogous to the
vissour in the old masonry structure. Using limit analysis procedure and
characteristic properties for the rock and rock reinforcement, functional equations
were developed for the stability of reinforced rock units relative to one another.
Along with the solution for the design and utilisation of rock reinforcement system,
the following equation was formulated to determine the stability index for the
resulting reinforced rock structure:

29

Chapter 2: Rock Bolting Practices

J_

T = a- (1
K//
^

e~KMD/R

-)
/R
yR \-e~^

(2.6)

where,
T = bolt tension,
a = factor depending on the time of bolt installation after excavation (0.5 1.0),
y = unit weight of rock,
R = shear radius = A/P,
A = area of reinforced rock unit supported by a bolt,
P = shear perimeter of reinforced rock unit,
K = ratio of average horizontal to average vertical stress,
[i =tan((p),
(p = angle of internal friction,
c = apparent cohesion of the rock mass,
h = in-situ horizontal stress,
D = height of de-stressed rock above the surface of the opening, and
1 = length of bolts.

It was shown that adequate and safe design of a rock reinforcement system requires
consideration of the following;

rock properties,

30

Chapter 2: Rock Bolting Practices

in situ stress field,


bolt material (steel) properties,
bolt pattern: length and spacing,
bolt tension,
time of installation,
modes of failure, and
mining procedures and operations.

Tao and Chen (1983) proposed a simple mathematical model for predicting stress
distribution along the bolt based on field studies on the stress distribution along
instrumented bolt in yielding rock. The concept of neutral point was subsequently
introduced. In yielding ground a fully grouted bolt is essentially divided into a pickup length and an anchor length. The pick-up length restrains the rock from
deforming, whereas the anchor length is restrained by rock. The shear strain
distribution (xz) is given by:

1 dQ{z)
T= = --^y7U2 dz

(2.7)

where,
d = bolt diameter, and
Q(z) = axial stress distribution along the bolt.

The neutral point of the bolt is given by the limit equilibrium analysis:

31

Chapter 2: Rock Bolting Practices

p=

(28)

i_[i + (_/_)]

where,
p = radial distance to the neutral point,
L = bolt length, and
a = tunnel radius.

Indraratna and Kaiser (1990) studied the failure mechanism of surrounding rock mass
in a tunnel reinforced by fully grouted bolts. The basic equilibrium equation for
tunnel was modelled by introducing equivalent strength parameters in relation to bolt
density. The following model was proposed along with a complete set of solutions
for stress distribution, strain and deformation surrounding the tunnel:

d<Jr

dr

(\ m* )<Jr
V
' =

(Jcr *
(2.9)

where,
crr = radial stress field,
r = distance from tunnel centre to point of interest,
m*=m(l+p),
m = internal friction parameter used in Mohr-Coulomb failure criteria,
(3 = bolt density parameter and is given by limit equilibrium:

32

Chapter 2: Rock Bolting Practices

ndXa
SL.ST

d = bolt diameter,
A, = friction factor for bolt grout interaction,
a = tunnel radius,
S L = longitudinal bolt spacing,
S T = circumferential bolt spacing,
o"cr* = crcr (1 + P), and
CTcr = equivalent post peak compressive strength, upon yielding.

The above analytical model was verified by conducting studies on laboratory


simulated physical models of a tunnel supported with fully grouted bolts. It was
shown that the greater the bolt density, the lesser w a s the tunnel wall convergence
(see Figure 2.11).

Yazici and Kaiser (1992) conducted studies on bond strength of grouted cable bolts
and presented a conceptual model for fully grouted cable bolts, k n o w n as Bond
Strength Model ( B S M ) . The simplified bolt surface w a s assumed to be zigzag in
shape as shown in Figure 2.12, which does not necessarily represent the actual bolt
surface profile. T h e concept of bond strength model is explained in Figure 2.13,
which consists of four quadrants. T h efirstquadrant shows the variation of bond
strength with axial displacement. The second quadrant relates to the pressure at the
bolt-grout interface to the bond strength. T h e third quadrant shows the relation

33

Chapter 2: Rock Bolting Practices

between axial and lateral displacement, and lastly the fourth quadrant depicts the
behaviour of grout during dilation.

in

,E

CM

/
/

I
J

/
/
1 /
j /

c_

Cfl
Cfl

_-

00

"

1 I / /
' // / /
/ // / / /

b
r(O

"

LO

//// /
~ ' /////

xt

<_>

.H

CU
CH

/^

y<

on
Csl

/ /
/

'''n

_
_

'

/'

^
/

/////

\\\ L
-W

fl:

no baits, 0 = G
B: /S* - 0.073

<

s s
'
yS y
/
-S / /
///

/
/

/ ill/ '
\ ill /

i f /
/

B
JT

y ^/ '

If/
/ 1

/ / / / . /

<L>

s_
+->

_-^-"*~^'^

/^"^

I / /
/ / /

CD

I
/

^^^

'

\
"

D^,

<n

C:

if
-

"ijf
jf
jr

^ " '145

0; 0 0.220
E: 0 = 0.291
F: na bolts, elastic.

_
^

i_., i

.__

i, ..

10

Tunnel convergence (mm)

Figure 2.11: Influence of grouted bolts o n tunnel convergence (after Indraratna and
Kaiser, 1990).

34

Chapter 2: Rock Bolting Practices

Plan

v<^H^iiiifcK
Detai at A:

Bolt

>sAJj
'/'

f'

-7'

't
i =tan-

Axial lorce

Figure 2.12: Schematic diagram reflecting the geometry of a rough bolt (after Yazici
and Kaiser, 1992).

UWmaSe lateral
displacement

f\

Figure 2.13: Schematic diagram relating components of bond strength model (after
Yazici and Kaiser, 1992).

35

Chapter 2: Rock Bolting Practices

Using the equality of small angle tangents to the angles in radia, the bond strength
was proposed in terms of friction and dilation angle given by:

x = a.tan </>< lo 1-

f a V

+ </>

(2.10)

\ (7 lim J

where,
x

= bond strength,

a = radial stress at the bolt grout interface,


aiim = limiting radial stress,
p = coefficient of reduction of dilation angle, and
<j) = friction angle between steel and grout.

The ultimate bond strength at the bolt grout interface w a s limited by the grout
strength. It was concluded that shearing off of the rough edges in the resin/rock
annulus eventually prevents further dilation, but the pressure on the bolt remains
constant during the pulling of the bolt. The major shortcoming of the above theory is
that it assumes the bolt surface is zigzag in nature (see Figure 2.12), which is not
realistic.

Hyett et al. (1995) proposed a constitutive law for bond failure of fully grouted cable
bolts based on a series of pull tests performed in a modified Hoek cell as shown in
Figure 2.14. A fully grouted seven-wire strand cable was used and the confining
pressure at the outside cement annulus was maintained constant. The bond strength
was shown to increase with the confining pressure. The radial dilation was attributed

36

Chapter 2: Rock Bolting Practices

to an unscrewing failure mechanism along the majority of the test section. Test
results were used to develop a frictional-dilational model for cable bolt failure. Three
failure mechanisms were proposed; dilational slip, non-dilational unscrewing and
shear failure of the cement flutes as shown in Figure 2.15. The axial forces
corresponding to the three failure mechanisms were by a series of equations as
follows.

15.2 m m seven-wire strand


Type 10 Portland cement annulus
Pressure vessel end cap
Specimen end cap
15 m m P V C tube for de-bonding
A B S pipe to support end of specimen
and overcome end-effects
Neoprene bladder
Cantilever strain gauge arms
High pressure electrical feed through
10. High pressure fitting
11. Pressure transducer

Figure 2.14: The modified H o e k Cell (after Hyett et al., 1995).

37

Chapter 2: Rock Bolting Practices

Rotational slip (unscrewing)


C = Torsional rigidity of the
cable
L f = Free length between the
test section and anchor
point
P, = Radial pressure
ac = U C S of the grout

Figure 2.15: M o d e s of failure of grouted bolt Cell (after Hyett et al., 1995).

For dilational slip, after splitting of the cement annulus,

F a = A l P 1 tan((Ls + i)

(2.11)

For non-dilational unscrewing, and

R =

_!/ t a n * .
(2.12)

sin a
38

Chapter 2: Rock Bolting Practices

For shear failure of the cement flutes,

Fa = Ai (T0 + Pitan^g) (2.13)

where,
Fa = axial load on cable,
Ai = apparent cable grout interface contact area,
Pi = radial pressure at r = rh inner radius of cement annulus,
<|>gS = friction angle between grout and steel,
i = dilation angle of the cable grout interface,
a = pitch angle,
Q = component of axial force related to untwisting of the cable,
T0 = Mohr-Coulomb cohesion intercept for the grout, and
<J>g = internal angle of friction for the grout.

The bond strength was found to depend on the grout quality; the radial stiffness of
the borehole wall and the mining induced distressing.

Benmokrane et al. (1995) conducted laboratory studies on pull test of short


encapsulation length (see Figure 2.16). Six types of cement based grouts and two
types of rock anchors were used. The following empirical equation was derived for
the estimation of anchor pullout resistance for the given embedment length:

39

Chapter 2: Rock Bolting Practices

P =a+b

AL
(2.14)

where,
P = ultimate pull-out load,
AL = anchored length,
<f> = diameter of anchor, and
a, b = constants, depend on grout and anchor type.

ALoad
Steel tendons
Stranded cable * = 15.8 m m )
Threadbar (* = 15.8mm)

Concrete cylinders
(200 m m )
Cement grout
Drilled hole
(38 m m )
A L = anchored length

Model A

Figure 2.16: Detail of short embedment pull test (after Benmokrane et al., 1995).

It was noted that the slip at the unloaded end started at approximately 8 0 % of the
failure load (see Figure 2.17). Based on the pullout results, an analytical model for
the bond-stress-slip relation was proposed for both the threaded bars and the stranded
cables respectively:

40

Chapter 2: Rock Bolting Practices

T = m.S + n

(2.15)

where,
T

= shear bond stress at anchor grout interface,

S = slip between anchor and grout, and


m, n = coefficients which depend on the type of anchor, grout and stages of
shear.

14
(a) Deformed bar

loaded end (L)


unloaded end (U)

r~
10

20

30

Slip (mm)

0>) Stranded cable

loaded end (L)


unloaded end (U)

20

30

Slip (mm)

Figure 2.17: Influence of anchor length on the load-displacement behaviour (after


Benmokrane et al., 1995).

41

Chapter 2: Rock Bolting Practices

Stankus and Peng (1996) proposed a n e w concept for roof support design using
Optimum Beaming Effect (OBE). The OBE was defined as the roof beam that has no
separation above or within the bolted range and having the shortest possible bolt. The
OBE concept was considered to be an effective tool with regard to saving in material
cost, material handling and drilling time in addition to improving the roof control
problems.

Serrano and Olalla (1999) described the tensile resistance of rock anchors based on
Euler's variational method and Hoek and Brown (1980) rock mass failure criterion.
Depending on the slenderness ratio (L/D; L = anchor length, D = anchor diameter),
two types of failure surfaces were obtained e.g. short and long (see Figure2.18).
Figure 2.19 shows the deformation behaviour of short and long anchors respectively.
In case of long anchors, nearly all the failure surface coincided with a cylindrical
surface defined by its diameter and length. In the case of short anchors, convex
surfaces were obtained. For both short and long anchors, a formula was proposed for
the pullout strength. The values of ultimate pullout strength, depending on the rock
type and its RMR indices, were obtained and compared with published data. Because
of its dependency on RMR, the model is empirical in nature, and thus would be site
specific.

Li and Stillborg (1999) proposed an analytical model for predicting the behaviour of
bolts under three different conditions; (a) for bolts subjected to a concentrated pull
load in pullout test, (b) for bolts installed in uniformly deformed rock masses, and (c)
for bolts subjected to the opening of individual rock joints. The stress component in

42

Chapter 2: Rock Bolting Practices

a small section of a bolt is shown in Figure 2.20 and the conceptual model for force
equilibrium equation along the axis of the bolt was expressed by:

A dab
T

b =

(2.16)

7tdb dx

where,
Tb = shear stress at the bolt interface,
A = cross sectional area of the bolt,
db = bolt diameter,
a\, = axial stress on the bolt, and
x = small length of bolt considered.

Short anchor

Long anchor

P___--I__Cfc-____t____-____-_-^_-i_er

'ult

b)

a)

Figure 2.18: Schematic diagram of bolt anchorage a) long anchor, b) short anchor
(after Serrano and Olalla, 1999).

43

Chapter 2: Rock Bolting Practices

(L-L e ):( n -n)-<f-

c;nc-

L-

4-

L=Anchor length
D=Anchor diameter

X.Y= Axes
x,y=Adimensionol axes {X/D; Y/0)

Lc=Critical anchor length

ri = _/D; n c = L c / 0

T,a = Stresses acting on rock anchor interfoce

ure 2.19: Failure mechanism of long and short anchored bolts (after Serrano and
Olalla, 1999).

44

Chapter 2: Rock Bolting Practices

dx

Figure 2.20: Stress components in a small section of a bolt (after Li and Stillborg,
1999).

The model was based on the description of the mechanical coupling at the interface
between the bolt and the grout medium for grouted bolts, or between the bolt and the
rock for frictionally coupled bolts. It was suggested that for rock bolts in pull out
tests, the shear stress of the interface attenuates exponentially with increasing
distance from the point of loading when the deformation is compatible across the
interface. Decoupling was found to start at the loading point when the applied load
was large enough, and then propagated towards the far end of the bolt with a further
increase in the applied load. Accordingly, the shear stress distribution along a fully
grouted rock bolt before and after coupling is shown in Figures 2.21 and 2.22
respectively. The magnitude of the shear stress on the decoupled bolt section was
dependent on the coupling mechanism at the interface. For fully grouted bolts, the
shear stress on the decoupled section was lower than the peak shear strength of the
interface, while for frictionally coupled bolts, it was approximately the same as the

45

Chapter 2: Rock Bolting Practices

peak shear strength. It w a s also revealed that the face plate plays a significant role in
enhancing the reinforcing effect.

OT
V)

2
*^
V)

_.
ra
ca
__

*e

*t

x2

Figure 2.21: Shear stress along a fully grouted rock bolt subjected to an axial load
after decoupling occurs (after Li and Stillborg, 1999).

Figure 2.22: Shear stress along a fully grouted rock bolt subjected to an axial load
before decoupling occurs (after Li and Stillborg, 1999).

Tadolini et al. (2000) studied the performance of high capacity tensioned cables for
supporting a tailgate. A combination of tensioned cable systems consisting of 4.5 m 7
strand cables (18 mm diameter) with a designed capacity of 58 tons was found to
successfully maintain the roof of longwall tailgate. The formation of roof guttering
was largely reduced by the application of tensioned cables across the failed zone. It
46

Chapter 2: Rock Bolting Practices

was also found that the application of tension on the point-anchored cable supports
m o v e d the roof upward during jacking process, but failed to m a k e the roof stiffer.

Witthaus and Ruppel (2000) described the successful implementation of bolting in


very deep and highly stressed multi-seam mining conditions. Different planning
elements such as the results of the observation of physical models, numerical models
(using F L A C and G E D R U ) for calculation of stress-distribution, laboratory and
underground support testing were used. The knowledge of in-situ strata behaviour
including geotechnical classifications was used to successfully implement the support
system. M o r e emphasis w a s given to regular monitoring of support performance and
re-estimation of various parameters using deviations of actual parameters from the
planned ones. In the same context, Wittenberg and Ruppel (2000) emphasised the
importance of a comprehensive quality management

system for successful

application of rock bolts in highly stress conditions.

2.6 THE CONCEPT OF CONSTANT NORMAL STIFFNESS AND ITS


APPLICABILITY IN R O C K BOLTING

In the past, various research work have been carried out on reinforced rock joints in
the laboratory using conventional direct shear apparatus, where the normal stress
acting on the joint interface w a s considered to be constant throughout the shearing
process. If the shearing surface is smooth enough and offering little or negligible
dilation, then the testing under the conventional Constant Normal Load ( C N L )
condition is adequate to represent shear behaviour. For non-planar discontinuities,

47

Chapter 2: Rock Bolting Practices

however, shearing often results in dilation as one asperity rides over another. If the
surrounding rock mass is unable to deform sufficiently, then an inevitable increase in
the normal stress occurs during shearing which is dependent on the applied initial
normal stress, the stiffness of the host medium and the dilation of the shearing
interface. Therefore, the CNL condition is unrealistic in circumstances where the
normal stress in the field changes considerably during the shearing process, for
example, as in the case in many underground mining situations. Thus, shearing of
rough surface no longer takes place under CNL condition; rather it is the stiffness of
the surrounding rock mass that remains relatively constant and, governs the shear
behaviour. Accordingly, this mode of shearing is defined as shearing under Constant
Normal Stiffness (CNS).

Johnstone & Lam (1989), Skinas et al. (1990) and Indraratna et al. (1999) have
described the importance of CNS condition to simulate the actual shear behaviour in
the field. Figure 2.23 shows a schematic diagram of a bolted joint, whose
deformation behaviour is governed by the applied normal stress, the stiffness of the
joint-bolt composite, and the frictional properties of the joint interface. The CNS
method is more realistic than the conventional CNL approach in such situations.
Figure 2.24 describes the shearing process of bolt/resin/rock interface. As indicated
earlier, in case of smooth surface (see Figure 2.24a) no dilation occurs and hence,
both CNL and CNS will produce same result. However, in case of shearing of
bolt/resin/rock interface, the rough bolt surface rides over the resin and/or rock (see
Figure 2.24b), hence, the CNS is more appropriate than the CNL. Indraratna et al.

48

Chapter 2: Rock Bolting Practices

(1999), and Indraratna & Haque (2000) have discussed m a n y other situations where
CNS is more realistic as compared to CNL.

tzzzzzzzzzz/

On f (kn,8v)
kn f (kr,kb)
kr & kb=rock & bolt stiffness
S v = joint dilation

T
joint plane

I bolt

Figure 2.23: Behaviour of bolted joint in thefieldunder Constant Normal Stiffness.

a n = a n o + k.5v/A

n = a no

* _ . _ f e -

Or
a) CNUCNS
Rough surface
Dilation

Smooth surface
No dilation

Figure 2.24: Dilation behaviour of joint plane a) two smooth planes, b) bolt and resin
interface.

49

Chapter 2: Rock Bolting Practices

2.7 C O N C L U S I O N

A brief description of the existing rock bolting practices has been given. Today, fully
grouted rock bolts have become the industry standard for supporting underground
roadways. The successful implementation of roof support requires a detailed plan
indicating geotechnical behaviour of host medium, estimated rock load, support
design, monitoring of support performance and re-establishment of whole cycle, if
necessary.

To date, most of the rock bolt research is concerned mainly with the estimation of
rock load and the load transfer mechanisms via the bolt-ground interaction. Some of
the reported research work aimed at studying the failure mechanism of the bolt/resin
interface or the rock mass itself. Surprisingly, the study on the influence of bolt
surface geometry on the failure mechanism of bolt/grout interface is scarce, and the
limited research work in this line was reported by Yazici and Kaiser (1992) with an
assumption of a zigzag bolt surface, which is uncommon in mines. Accordingly, the
need to effectively investigate the impact of bolt surface geometry on the failure
mechanism of bolt/resin/rock interface is of paramount importance. Thus, this thesis
is aimed at examining the failure mechanism of such bolt/resin interface under
Constant Normal Stiffness condition, which in underground mining situations is
considered to be more realistic as compared to the conventional Constant Normal
Load condition.

50

Chapter 3

EXPERIMENTAL STUDY OF THE FAILURE MECHANISM


OF BOLT/RESIN INTERFACE

3.1 INTRODUCTION

One of the most important parameters in rock/resin/bolt load transfer mechanism is


the surface condition of the bolt, i.e. the bolt surface roughness. The study of load
transfer mechanism would remain incomplete without the relationship between the
bolt surface roughness and the shear behaviour of bolt/resin interface, as the surface
roughness dictates the level of interlocking between the bolt and the resin surfaces. A
realistic outcome of such study w a s considered possible if the tests were conducted
under Constant Normal Stiffness ( C N S ) condition, because the normal stress acting
on the bolt/resin interface changes, however the stiffness remains constant during
shearing process. The existing C N S testing machine as described by Indraratna et al.
(1997) was modified to incorporate the surface profile of the bolt for direct shear
testing of cast resin blocks.

A series of tests were conducted on the flattened bolt surface of two most popular
bolt types currently being used in Australian coal mines, and are referred to in this

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

thesis as type I and type II bolts, respectively. The flattened bolt surface was used

print the image of the bolt surface on cast resin samples, such that the resin and bo
surfaces exactly match with each other during testing. The testing program included

the shearing of bolt/resin interface of two bolt types at different initial normal st
levels, ranging between 0.1 and 7.5 MPa, at a constant shearing rate of 0.5 rnm/min.
Each sample was subjected to several cycles of loading to observe the effect of
repeated loading on the bolt/resin interface.

This chapter contains the detailed design of a large-scale CNS apparatus for direct
shear testing, the preparation of the bolt surface and the resin samples used for
testing. A brief review of the laboratory testing program is discussed together with
the relevant instrumentation and monitoring procedures implemented. Also discussed

is the analysis of test data and the development of an empirical model to predict the
shear strength of bolt/resin interface.

3.2 BOLT SURFACE PREPARATION

A 100 mm length of a bolt was selected for the surface preparation for CNS shear
testing. The specified length of bolt was cut and then drilled through as shown in
Figure 3.1. The hollow bolt segment was then cut along the bolt axis from one side
and preheated to open up into a flat surface as shown in Figure 3.2. The surface

features of the bolt (ribs) were carefully protected while opening up the bolt surfac

The flattened surface of the bolt was then welded on the bottom plate of the top shea

52

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

box of the CNS testing machine as shown in Figure 3.3. Although these flattened bolt
surfaces may not ideally represent the complex behaviour of circular shaped bolt
surface observed in the field, nevertheless, they still provided a simplified basis for
evaluating the impact of the bolt surface geometry on the shear resistance offered by
a bolt. Table 3.1 shows the detailed specification of two types of bolts used for the
study.

Figure 3.1: Hollow bolt segment.

Figure 3.2: Flattened bolt surface.

53

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

Figure 3.3: Bolt surface welded on the top shear box plate.

Table 3.1: Specification of bolt types.

Bolt

Core

Finished

Diameter

Diameter

(mm)

(mm)

Typel

21.7

Type II

21.7

Rib Spacing Rib Height


(mm)

(mm)

24.4

28.5

1.35

23.2

12.5

0.75

3.3 SAMPLE CASTING

The welded bolt surface on the bottom plate of the top shear box was used to print
the image of bolt surface on cast resin samples. The arrangement for casting the
sample is shown in Figure 3.4. For obvious economic reasons, the samples were cast
in two parts. Nearly three-fourth of the mould was cast with high strength casting
plaster and the remaining one-fourth was topped up with a chemical resin commonly

54

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

used for bolt installation in coal mines. A curing time of two weeks was allowed for
all specimens before testing was carried out. The properties of the hardened resin
after two weeks were: uniaxial compressive strength (ac) = 76.5 MPa, tensile
strength (at)= 13.5 MPa, and Young's modulus (E) = 11.7 GPa. The cured plaster
showed a consistent ac of about 20 MPa, rjt of about 6 MPa, and E of 7.3 GPa. Such
model materials were suitable to simulate the behaviour of a number of jointed or
soft rocks, such as coal, friable limestone, clay shale and mudstone, and were based
on the ratios of oVo-t and aJE applied in similitude analysis (Indraratna, 1990).
Figure 3.5 shows a typical cast sample and Figure 3.6 shows such a sample placed in
the shear box of the CNS testing machine. The resin samples prepared in this way
matched exactly with the bolt surface, allowing a close representation of the
bolt/resin interface in practice.

Figure 3.4: Arrangement of casting resin samples.

55

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

Figure 3.5: Cast resin block.

Figure 3.6: Cast resin samples inside the bottom shear box.

3.4 C N S T E S T I N G

APPARATUS

Figure 3.7 is a general view of the C N S testing apparatus used for the study.
Designed and built in house, the equipment was basically a modified version of the

56

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

type used by researchers earlier at Monash University (Johnstone and Lam, 1989).
The modified CNS apparatus incorporated the bolt surface profile in the top shear
box.

C N S shear apparatus
Jack connected
to p u m p
Load cell connected
to digital strain meter

Spring of known
stiffness (kn)

Fixed frame

Load capacity:

Normal = 180 kN
Shear = 120 kN

Shear displacement rate = 0.3 -1.7 mm/min


Normal stiffness = 8.5 kN/mm
Box size: Top 250x75x150 mm
Bottom 250x75x100 m m

Figure 3.7: Line diagram of the C N S apparatus (modified after Indraratna et al.,
1997).

As can be seen in Figure 3.7 the equipment consisted of a set of two large shear
boxes to hold the samples in position during testing. The size of the bottom shear

57

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

is 250x75x100 mm while the top shear box is 250x75x150 mm. A set of four springs
are used to simulate the normal stiffness (kn) of the surrounding rock mass. The top
box can only move in the vertical direction along which the spring stiffness is
constant (8.5 kN/mm). The bottom box is fixed to a rigid base through bearings, and
it can move only in the shear (horizontal) direction. The desired initial normal stress
(ono) is applied by a hydraulic jack, where the applied load is measured by a
calibrated load cell. The shear load is applied via a transverse hydraulic jack, which
is connected to a strain-controlled unit. The applied shear load can be recorded via
strain meters fitted to a load cell. The rate of horizontal displacement can be varied
between 0.35 and 1.70 mm/min using an attached gear mechanism. The dilation and
the shear displacement of the joint are recorded by two LVDT's, one mounted on top
of the top shear box and the other is attached to the side of the bottom shear box.
Figure 3.8 shows a closer view of the CNS apparatus.

Figure 3.8: A closer view of the C N S apparatus (modified after Indraratna et al.,
1997).

58

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

3.5 TESTING OF BOLT/RESIN INTERFACE

A total of 12 samples were tested for two different types of bolt surface at initial
normal stress (ano) levels ranging from 0.1 to 7.5 MPa. Each sample for bolt type I
was subjected to five cycles of loading in order to observe the effect of repeated
loading on the bolt/resin interface. Samples for bolt type II were subjected to only
three cycles of loading, as it was found that the stress profile did not vary
significantly after the third cycle of loading. The stress profile, as described above, is
defined as the variation of shear (or normal) stress with shear displacement for
various cycles of loading. The o-no applied to the samples represented typical
confining pressures, which might be expected in the field, especially in shallow
mining conditions. A constant normal stiffness of 8.5 kN/mm (or 1.2 GPa/m when
applied to a flattened bolt surface of 100 mm length) was applied via an assembly of
four springs mounted on top of the top shear box. The simulated stiffness was found
to be representative of the soft coal measure rocks. An appropriate strain rate of 0.5
mm/min was maintained for all shear tests. A sufficient gap (less than 10 mm) was
allowed between the upper and lower boxes to enable unconstrained shearing of the
bolt/resin interface.

3.6 SHEAR BEHAVIOUR OF BOLT/RESIN INTERFACE


3.6.1 Effect of Normal Stress on Stress Paths

Figure 3.9 shows the shear stress profiles of the bolt/resin interface for selected
normal stress conditions for the type I bolts. The difference between stress profiles

59

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

for various loading cycles was negligible at low values of o-no (Figure 3.9a). This was
gradually increased with increasing value of ano reaching a maximum between 3 and
4.5 MPa (Figure 3.9b). Beyond a 4.5 MPa confining pressure, the difference between
stress profiles for the loading cycles I and II decreased again (Figure 3.9c). A similar
trend was also observed for the type II bolt surface (not shown in figure).

At low ono values, the relative movement between the bolt/resin surfaces caused an
insignificant shearing and slickensiding of the resin surface, thus keeping the surface
roughness nearly unchanged. For each additional cycle of loading, the shear stresses
marginally decreased, especially in the peak shear stress region (see Figure 3.9a).
However, as the value of ono was increased, the shearing of the resin surface was also
increased, and the difference in stress profiles for various cycles of loading became
significant. At moderate normal stress levels (between 3 and 4.5 MPa), the shearing
and slickensiding of the contact surface was not sufficient to smoothen the resin
surface excessively in the first cycle, and further cyclic loading was required to cause
a significant difference in the observed stress profiles. This is indicated by the
continuously decreasing gap between the shear stress profiles in Figure 3.9b. As the
initial normal stress was increased beyond this level, the difference in shear stress
profiles for various cycles of loading was decreased due to excessive shearing and
smoothening of the resin surface, after the first cycle of loading. At this stage, no
further decrease in stress profiles was observed, i.e. after the second cycle of loading
as shown in Figure 3.9c. The difference in stress profile after the third cycle was
negligible at almost all normal stress levels, thus indicating no change in surface
properties after the third cycle of loading.

60

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

0.6 -

ano = 0.1 M P a

0.5 Q.

2
in
in

a
_
_.

4-

0.4 0.3 -

0.2 -

a
IA

(afl
0.1
0.0
4

10

Shear displacement (mm)

a n o = 4.5 M P a

-_r- Cycle - II
-x Cycle - IV
Cycle - V

10

12

14

16

18

Shear displacement (mm)


ono = 7.5 M P a

c)
ra
0.
0)

n
<u
_:
(0

-* Cycle - IV
Cycle-V

10

15

20

Shear displacement (mm)

Figure 3.9: Shear stress profiles of the type I bolt from selected tests.

61

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

3.6.2 Dilation Behaviour

Figure 3.10 shows the dilation profiles from selected tests for the bolt type I. The
dilation of the bolt/resin interface for both bolt types was characterised by the initial
contraction, and then subsequent increase in dilation for all the loading cycles. This
initial contraction may be due to the early degradation of very fine irregularities of
the resin surface. The dilation was reduced with increasing value of ano (e.g., peak
dilations in Figure 3.10b are smaller as compared to those in Figure 3.10a), and
correspondingly, the increase in normal stress from its initial value was also found to
decrease with increasing ano, as per Equation 3.1 described later. This gives an
indication that the behaviour of rock/resin interface at CNS condition may approach
the CNL condition at very high ono values, where no significant increase in normal
stress is expected to be observed from its initial value throughout the shearing
process. As expected, the gap between the dilation profiles for various loading cycles
remained constant at low initial normal stress levels (Figure 3.10a), and progressively
decreased at moderate normal stress levels (Figure 3.10b).

For the first cycle of loading, Figures 3.11 and 3.12 show the variation of dilation
with shear displacement at various normal stresses for type I and type II bolts,
respectively. For various values of an0, the maximum dilation occurred at a shear
displacement of 17 - 18 mm and 7-8 mm, for type I and type II bolts, respectively
(Figures 3.11 and 3.12). The distance between the ribs for both bolt types is shown in
Table 3.1. Therefore, it may be concluded that the maximum dilation occurred at a
shear displacement of about 60% of the bolt rib spacing.

62

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

a)

o-no = 0.1 M P a
- Cycle

Cycle II

_Cycle- III
--*- Cycle -IV
* Cycle -V

10

-0.5

20

30

Shear Displacement (mm)

b)

"no = 4.5 M P a
- Cycle I

0.8
E 0.6
E, 0.4
c
o 0.2
+3
0

-0.2 0
D

Cycle - II
~*-*

- A - Cycle - III
-*-- Cycle - IV
-A Cycle - V

10

15

20

Shear Displacement ( m m )

Figure 3.10: Dilation profiles o f the type I bolt from selected tests.

Dilation Profile

-* 0.1 MPa
-m1.5 MPa

-A3MPa
-*-4.5 MPa
-*-6MPa
-7.5 MPa

25
Shear Displacement ( m m )

Figure 3.11: First loading cycle dilation profiles for the type I bolt at various a n
values.

63

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

Dilation Profile

-0.1 MPa
-1.5 MPa1
-3 MPa
-4.5 MPa)
-6MPa t
t

-7.5 MPa
I

10

12

14

Shear Displacement (mm)

Figure 3.12: First loading cycle dilation profiles for the type II bolt at various or
values.

3.6.3 Effect of N o r m a l Stress on P e a k Shear

At various normal stresses, Figures 3.13 and 3.14 show the variation of shear stress
with shear displacement for the first cycle of loading, for both type I and type II bolts,
respectively. The shear displacement for peak shear stresses increased with
increasing value of ano for both bolt types. This was due to the increased amount of
resin surface shearing with the increasing value of ano. However, there was a gradual
reduction in the difference between the peak shear stress profiles with increasing
value of CTno. The shear displacement required to reach the peak shear strength is a
function of the applied normal stress and the surface properties of the resin, assuming
that the geometry of the bolt surface remains constant for a particular type of bolt as
evident from Figures 3.13 and 3.14.

64

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

Shear Stress Profile

m
o.

-0.1 MPa

10

-1.5 MPa

-3MPa

-4.5 MPa

(0
_.

-6MPa
-7.5 MPa

n
a>

10

15

20

25

Shear Displacement (mm)

CO

Figure 3.13: Shear stress profiles of type I bolt forfirstcycle of loading.

Shear Stress Profile


7,

0.1 W _

5-

-*--1.5K/Pa

<n 4

-A3MPa

i= 3(0

- X 4.5 M _

i5 2

-6MPa

_^~^

0)

(0

1 > *
c
2
4
6
8

*
10

-^-7.5 MPa

r
12

1
14

Shear Displacement (mm)

Figure 3.14: Shear stress profiles of type II bolt for first cycle of loading.

3.6.4 Effect of Cyclic Loading on Peak Shear

For various loading cycles, Figures 3.15 and 3.16 show the variation of peak shear
stress with normal stress applied for type I and type II bolts, respectively. For the type

65

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

I bolt surface, the graphs of cycle I through cycle III s h o w a bi-linear trend, whereas
the graphs representing cycles IV and V show only a linear trend. For the type II bolt
surface, only cycle I shows a bi-linear trend and cycles II and III show a linear trend.
At low initial normal stress, the shearing of resin surface is negligible, and hence, the
rate of increase of peak shear stress with respect to normal stress is high. At higher
normal stress, the degree of shearing of resin surface is greater, and some of the

energy is thus utilised to shear off the resin surfaces. As a result, there is retarded rate
of increase in peak shear stress with respect to the normal stress. As the samples are
loaded repeatedly, the resin surfaces become smoothened reducing the surface
roughness, and as a result, the rate of increase of peak shear stress is likely to remain
constant with respect to the normal stress.

Initial normal stress


N o r m a l stress at peak shear

Cycle I
Cycle I

a.

--A--- Cycle N
* C y c l e II
...%... Cycle III

* C y c l e in
...+... Cycle V

_
_:
_:

1 C y c l e IV
....... CycleV

0 2

Cycle V

in

a.
1

Normal stress (MPa)

Figure 3.15: Variation of peak shear stress with normal stress for type I bolt.

66

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

Initial normal stress


Normal stress at peak shear

CL 5
to

S 4
_
n 3
a

..--.- Cycle 1

-C

.--EJ-. - Cycle 1

w
(0 *

o 1
0.

Cycle III
- Cycle
3

Normal stress (MPa)

Figure 3.16: Variation of peak shear stress with normal stress for type II bolt.

3.6.5 Overall Shear Behaviour of T y p e I and T y p e II Bolts

Figure 3.17 shows the stress profiles of both type I and type II bolts plotted together
for the first cycle of loading. The following observations were noted:
The shear stress profiles around peak were similar for both bolt types.
However, slightly higher stress values were recorded for the bolt type I at low
normal stress levels, whereas slightly higher stress values were observed for the
bolt type II at high normal stress levels, in most cases.
Post peak shear stress values are higher for the bolt type I indicating better
performance in the post-peak region.
Shear displacements at peak shear are higher for the bolt type I indicating the
safe allowance of more roof convergence before instability stage is reached.

Dilation is greater in the case of bolt type I.

67

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

Shear Stress Profile of Type I and Type II Bolts


- 0.1 MPa I
- 1.5MPal
- 3MPal
- 4.5 MPa I

*___*

-- 6MPal
- --- - 7.5 MPa I
*- 0.1 MPa II
- 1 . 5 MPa 1
-e3MPa(
- 4 . 5 MPa I

AA_j j A j A A A A A
4

_. A

10

j
12

-*6MPaH

14

- 7 . 5 MPa II

Shear Displacement (mm)

Normal Stress Profile of Type I and Type II Bolts


. 0.1 MPa I

A
CL

(0
(0
0)
_.
4-1

CO

- 1.5 MPa I

. 3MPal

4.5 MPa I

O^B-O B-B B "

m*******-*^

. Q. ,Q.. a
B" -B- - e -B-

- 6MPal

- -- - 7.5 MPa I
- -0.1MPa
-m1.5 MPa i

- Q

-B3

MPa II

- 4 . 5 MPa

_. 0 2 4 6

o
z

_*___*:

++-*-

*********
4
2

*^l

-jlHIHie-*^!^*^^-*'

4=_fe-_L
8

10

12

- * 6 MPa II

14

- 7 . 5 MPa n

Shear Displacement (mm)

Dilation Profile of Type I and Type II Bolts

A--- 0.1 MPa I


-.. 1.5MPal
-a---3 MPa I
--- 4.5 MPa I
X---6MPal
----7.5MPaI
-_,0.1 MPa II
- a 1 . 5 MPa II
-a3MPa
- 4 4 . 5 MPa II
-*6MPaH
- 7 . 5 MPa ll

Shear Displacement (mm)

Figure 3.17: Comparison of stress profile and dilation of type I and type II bolts for
first cycle of loading.

68

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

3.6.6 Effect of N o r m a l Stiffness

The laboratory experiments were carried out with spring assembly with an effective
stiffness of 8.5 kN/mm. In practice, the stiffness of resin/rock system will be usually
higher than the laboratory simulated stiffness. As the stiffness increases, the effective
normal stress on the bolt/resin interface at any point of time will also increase, as per
the following equation:

Gn =

<Jno +

kn.dv
A

(3-D

where,
rjn = effective normal stress,
ano = initial normal stress,
kn = system stiffness,
5V = dilation, and
A = area of the bolt surface.

In general, higher values of effective normal stresses should be observed for the type
I bolt as compared to the type II bolt as long as the confining pressure remains low.
Higher values of effective normal stress will have a direct positive impact on the
peak shear stress values and, therefore, when installed in the field, the type I bolt
would outperform the type II bolt, particularly at low confining pressure conditions.

69

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

However, in deep mining conditions, where the high stress conditions prevail, the
reverse situation should be observed. This phenomenon is explained in Section 3.6.7.

3.6.7 Effect of Bolt Surface Profde Configuration

In a relative displacement situation, however minute the movement is between resin


and bolt surface, the complete bolt surface is no longer in contact with the whole
resin surface. The contact zone between the resin and the bolt surface at that point is
usually restricted to the bolt rib and its surrounding area only. At low normal stress
condition, the stress concentration on the resin surface around the bolt ribs is not
sufficient to cause resin failure. However, at high normal stress levels, the stress
concentration on the resin surface due to the projected ribs of the bolt surface reaches
a value which is greater than the compressive strength of the resin, causing it to crush
around that zone.

The stress concentration is more evenly distributed on the resin surface for the bolt
type II with rib spacing of 12.5 mm, as compared to the bolt type I with a rib spacing
of 28.5. At low normal stress values, the stress concentration on the resin surface is
not high enough to crush it, thus its integrity remains intact. As a result, the bolt with
deeper ribs will offer higher resistance due to higher dilation (see Figure 3.18).
However, at a high normal stress level where the concentrated stress around ribs is
high enough to crush the resin surface, the bolt that has lower rib spacing provides a
more evenly distributed stress on the resin surface. This in indicated by a relatively
flatter diminishing peak dilation curve for type II bolts as compared to type I bolts

70

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

(Figure 3.18}. In other words, the bolts with lowerribspacing would offer a greater
resistance at high normal stress conditions.

1.4 -,
1.2 -

------ Type Ibolti

I 1.1 -8
re

=5

0.6

__
re

__

04
0.2 -

"

"

13

Initial normal stress (MPa)

Figure 3.18: Variation of peak dilation with initial normal stress.

3.7 EMPIRICAL MODEL FOR PREDICTING THE SHEAR


RESISTANCE AT BOLT/RESIN INTERFACE
Based on the above study a mathematical model was developed to predict the shear
resistance of bolt/resin interface at any normal stress within the range of experimental
stresses. The shear stress at any particular point of time could be expressed as a
function of the shear displacement, the normal stress and the surface properties of the
bolt. Examination of the shapes of shear stress profiles at various normal stresses
suggests the following empirical relationship (Daniel and W o o d , 1980):

71

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

x = a.x\ecx

(3.2)

where,
T = predicted shear stress,
x = shear displacement,
e = natural log base,
a = f (normal stress),
b = f (normal stress), which takes into account the effect of shearing of
asperities at high normal stress values, and
c = f (surface properties) i.e. constant for a particular set of bolt-resin
interface and is equal to -0.092 and -0.29 for both type I and type II bolt,
respectively.

Assuming that the parameters a and b are constant for any particular normal stress,
the peak shear stress can be obtained by differentiating Equation 3.2 with respect to
the shear displacement and equalling it to zero. Therefore, for x to be maximum,

dx
=0
dx

or, x =

(3.3)

B y incorporating the value of x in Equation 3.2,

X max a.

(-2LY

(3.4,

V c.e>

72

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

For the set of data generated from the laboratory experiments the following
relationship was obtained based on best-fit regression analysis.

If N = initial normal stress:

for type I bolt;


-1.9

a = 0.425 +2.79 e

(3.5)

b = 0.405 + 0.08 N for N < 3 MPa

(3.6a)

= 0.580 + 0.0265 N for N > 3 MPa

(3.6b)

for type II bolt;

-1.29

a = 0.45 +3.48 e

(3.7)

b = 0.525 + 0.142 N for N < 4.5 MPa

(3.8a)

= 0.95 + 0.0467 N for N > 4.5 MPa

(3.8b)

The two different functional values of b could be explained from Figures 3.15 and
3.16 where the relationship is bi-linear. The first and steeper segment is continued up
to 3 MPa for type I bolt and 4.5 MPa for type II bolt, after which the second and
flatter segment represents the effect of shearing of asperities at higher normal stress
levels. The above anomaly necessitates the use of different functional values of b for

73

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

different normal stress levels. Table 3.2 shows the calculated values of a, b and c
using the above equations for various normal stress levels.

Figures 3.19 and 3.20 show the predicted shear stresses for any shear displacement at
any particular normal stress, and the corresponding values based on laboratory
observations for both type I and type II bolts. The predicted values are in close
agreement with the experimental values throughout the loading period.

Experimental
Model

10

15

20

Shear displacement (mm)

Figure 3.19: Comparison of predicted values of shear stresses and the values obtained
from laboratory experiments for type I bolt.

74

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

Figure 3.20: Comparison of predicted values of shear stresses and the values obtained
from laboratory experiments for type II bolt.

Table 3.2: Calculated values of a, b and c.


Normal
stress
(MPa)

Type I bolt

Type II bolt

0.1

0.43

0.41

-0.092

0.45

0.54

-0.29

1.5

1.2

0.52

-0.092

1.9

0.74

-0.29

3.0

1.9

0.65

-0.092

2.7

0.95

-0.29

4.5

2.25

0.699

-0.092

3.06

1.16

-0.29

6.0

2.46

0.739

-0.092

3.25

1.23

-0.29

7.5

2.59

0.78

-0.092

3.38

1.30

-0.29

75

Chapter 3: Experimental Study of the Failure Mechanism of Bolt/Resin Interface

3.8 CONCLUSION

It can be inferred from this experimental study that:


The shear behaviour of the bolt surface at various confining pressures directly
affects the load transfer mechanism from the rock to the bolt.
The type I bolt offered higher shear resistance at low confining pressure (below
6.0 MPa), whereas, the type II bolt offered greater shear resistance at high normal
stress conditions exceeding 6.0 MPa. This was attributed to the surface profile
configuration of the bolt i.e., the spacing and the depth of the rib.
The bolt with deeper rib offered higher shear resistance at low normal stress
conditions, while the bolt with closer rib spacing offered higher shear resistance
at high normal stress conditions.
The impact of repeated loading on the effective shear resistance of the bolt/resin
interface was influenced by the magnitude of the applied normal stress, the
number of loading cycles and the surface geometry of the bolt.
The maximum dilation occurred at a shear displacement of nearly 60% of the rib
spacing.
The bolt type I showed better performance than that of bolt type II when
considering the shear behaviour at low normal stress, dilational aspects, and the
post-peak behaviour.
The empirical model could predict the complete shear stress profile at various
normal stress conditions, which was in close agreement with the laboratory
results.

76

Chapter 4

FIELD INVESTIGATIONS

4.1 INTRODUCTION

In an attempt to verify the laboratory findings, with respect to the influence of bolt
surface profiles on load transfer mechanisms under different testing environments, a
program of field investigations were undertaken in two local coal mines (West Cliff
and Tower Collieries) in the Southern Coalfields of Sydney Basin, NSW, Australia.
Both selected sites were at the development headings, serving the newly developed
retreating longwall faces. The examination of the bolt behaviour was undertaken
with respect to the prevailing in-situ stress conditions, both in magnitudes and the
stress directions, affecting the stability of the development headings. Two different
types of fully instrumented bolts were installed, and the methodology of their
installation, positioning, and regular monitoring is the subject of detailed discussion
in this chapter. The chapter also contains the guidelines for selecting a suitable bolt
surface profile under various confining stress conditions.

Chapter 4: Field Investigations

4.2 S I T E D E S C R I P T I O N

4.2.1 West Cliff Colliery (Mine 1)

West Cliff Colliery is located in the Southern Coalfields of Sydney Basin, NSW,
Australia (see Figure 4.1). The mine was selected as the first site for field
investigation. West Cliff Colliery currently extracts coal from the Bulli seam at a
depth of 465m and produces around 1.45 million tonnes per annum from the 200m
wide longwall face and development operations. The average seam thickness is 2.7m,
and a representative geological section close to the experimental site is shown in

Figure 4.2. The roof strata consisted of sandstone/shale/siltstone combination, which


could be considered to be stable.

Figure 4.1: Geographical location of West Cliff Colliery (Mine 1), and Tower
Colliery (Mine 2).

78

Chapter 4: Field Investigations

Shale (32-48 M P a )
Siltstone (56 M P a )

4.5m

Sandstone (65-88 MPa)

Siltstone (56 MPa)


Sandstone (65-88 M P a )

Coal, Bulli Seam (7-15 MPa)


Shale (32-48 M P a )

5.0m

Sandstone (65-88 MPa)

Figure 4.2: Geological sections showing Bulli seam and the associated strata near the
instrumentation site at West Cliff Colliery.

Figure 4.3 shows the general layout of the mine and the location of the longwall
panel. Detailed layout of the panel under investigation (panel 512) is shown in Figure
4.4. Six, 2.1m long instrumented bolts (three of each type I and type II bolts as used
in the laboratory experiments) were installed in the longwall panel cut-through (C/T
7). Figure 4.5 shows the details of the instrumented bolt installation pattern at C/T 7.
In order to avoid excessive and non-uniform loading of the bolts (because of wide
excavation at the cut-through intersections), the bolts were installed near the middle
length of the cut-through. The regular pattern of bolting in the site was 6 bolts in a

79

Chapter 4: Field Investigations

row and the spacing between the rows was lm. At the test site, three bolts were
installed in each row, and in all there were two rows of instrumented bolts, with each
row containing one type of bolts. Also shown in Figure 4.5 are the magnitude and the
direction of the prevailing principal horizontal stress. Following installation of
instrumented bolts in the month of January 1999, field data was subsequently
collected at regular intervals, until the longwall face past the instrumented cutthrough in June 1999.

Figure 4.3: General mine layout of West Cliff Colliery.

Figure 4.4: General layout of the panel under investigation indicating instrumentation
site at West Cliff Colliery.

80

Chapter 4: Field Investigations

Right side

Type I bolts
Type II bolts

C/T

C/T
< fjj

lr
N u m b e r e d instrumented bolts

Ordinary bolts
Direction of Development
,-f^i^ Roof Guttering

Figure 4.5: Detail site plan of instrumented bolts at West Cliff Colliery.

4.2.2

T o w e r Colliery (Mine 2)

Tower Colliery is also located in the Illawarra Coal Measures of N S W , Australia, and
the geographical location for the mine is shown in Figure 4.1. The mine was selected
for the second site of field monitoring because of its proximity to the University of
Wollongong. The average seam thickness of the Bulli seam in the instrumented site
was 3.3m at a depth of 515 m, and the average annual production from the 230 m
wide longwall face and other development operations is around 1.4 million tonnes.
Figure 4.6 shows the geological section of a borehole close to the experimental site,
illustrating the immediate roof stratification above the Bulli seam. The roof rock
consisted of mainly sandstones, mudstones, siltstones and occasional weak bands of
shale, and is classified as moderate to strong roof. Figure 4.7 shows the general

81

Chapter 4: Field Investigations

layout of the mine and the location of longwall panel. Twelve instrumented bolts,
each 2.4 m long were installed in two adjacent main entry roads to longwall panel
(panel 18) as shown in Figure 4.8. The general arrangement for instrumented bolt
installations in both gateroads was the same as at West Cliff Colliery, and is shown
in Figure 4.9.

1.2m

Mudstone (30-40 MPa)

1.8m

Sandstone (70-90 MPa)

0.5m

Shale (52 MPa)

1.5m

Sandstone (70-90 MPa)


Shale (55 MPa)

0.8m

Sandstone (70-90 MPa)

3.3m

Coal, Bulli Seam (7-15 MPa)

1.8m

Siltstone (32-46 MPa)


Coaly partings

0.5m

Mudstone (30-40 MPa)

Figure 4.6: Borehole section showing the geological formation above Bulli seam,
Tower Colliery.

82

Chapter 4: Field Investigations

Panel under investigation

Figure 4.7: General mine layout of Tower Colliery indicating the panel under
investigation.

83

C h a p t e r 4: Field Investigations

Slight
guttering

Severe
guttering

Travelling road

Installation site

________*_:___^._____/_______^i
11

Ifii

\r~

v 'i 11

1(

r-

Direction of
panel retreat
\cir~i

IDDCZDCZlI
OC_____.I___[

D-ZZX

7____7_I

DOC

-ZL

Principal horizontal
stress direction

Major dyke

Figure 4.8: Detail layout of the panel under investigation at Tower Colliery.

Left side

Right side

Direction o f panel retreat

~y^T Instrumented bolts


^_* Ordinary bolts
CD

Extensometers

Direction of development
Roof guttering

Figure 4.9: Detail layout of the instrumentation site at Tower Colliery.

84

Chapter 4: Field Investigations

In addition to the instrumented bolts, three extensometer probes were also installed
between the two rows of instrumented bolts in each gateroad. It is noted that the
instrumented site at West Cliff Colliery did not have extensometric installations, due
to unavailability of equipment at the time of instrumentation. The notations adopted
for the bolts and extensometer probes in Mine 2 is explained in Figure 4.10. Field
monitoring commenced soon after the installation in May 2000, and ended in
October 2000, when the approaching longwall face overran the instrumented site.

1: Bolts installed at the left side


2: Bolts installed at the middle
3: Bolts installed at therightside

TR: Travelling Road


BR: Belt Road

TR J X
J: Type 1 bolts
A: Type II bolts

(a)

1: Extensometers installed at the left side


2: Extensometers installed at the middle
3: Extensometers installed at therightside

TR: Travelling Road


BR: Belt Road

>TRX
(b)
Figure 4.10: Instrumentation naming convention at T o w e r Colliery, a) for bolts, and
b) for extensometers.

85

Chapter 4: Field Investigations

4.3

INSTRUMENTATION

4.3.1 Instrumented Bolts

Bolts selected for instrumentation in both mines were identical except in

which were similar to the normal bolting cycles in respective mines. At We

Colliery (Mine 1), the production bolt was 2.1 m in length, whereas at Tow

Colliery (Mine 2), the length of the bolt was 2.4 m. In both cases, the bo
was 21.8 mm. The first step involved in the bolt instrumentation process
two identical, diametrically opposite channels of 6 mm wide and 3 mm deep
along the bolt axis, leaving outer 100 mm intact as shown in Figures 4.11

The intact bolt configuration without any channel in the first 100 mm ens

stability and integrity of the strain gauges and the corresponding wiring

installation of bolt in the field. This has been the normal practice in b
instrumentation (Signer and Jones, 1990b). Figure 4.13 shows a section of
engraved channels.

Once the channels were cut, they were smoothened with sand paper and wipe

with alcohol solvents. A total of 18 strain gauges were mounted on each bo

each channel). The spacing between the strain gauges mounted on 2.1 m bolt

installed at West Cliff Colliery, was 200 mm (see Figure 4.11). On the oth

the spacing on 2.4m long bolts, installed at Tower Colliery, was kept at 2

intervals (see Figure 4.12). The variation in the strain gauge spacings on

installed in two mines was necessary in order to accommodate the variatio

86

Chapter 4: Field Investigations

lengths used. The slots were thenfilledwith silicon gel to cover up the strain gauges,
and allowed to harden for a week prior to installation in the field. Figure 4.14 shows
a section of an instrumented bolt with strain gauges and wirings visible through the
silicon cover.

-H 6 h-

100

200

'n

4
200

i:

.t
21.8

Cross Section

200

''

Slot: 6 mm x 3.0 mm
Slot Area : < 10 %

3 :

200

"f
i

P,

All Strain Gauges are 5 m m ,


120 Q , Gauge Factor 2.15

200 '

5:

200
V
A

o
o

Reduction in bolt
strength 10 %

cs

6 :
.

'

200
7:

- * -

200
D
1

8:

9:

Strain Gauges

200

,
i ''

300 1

>'

.pj
T^P

'

Jr

Long Section
Figure 4.11: Strain gauge and bolt layout for West Cliff Colliery (Mine 1).

87

Chapter 4: Field Investigations

, 100

-f:

-H 6 K-

jk

1 t

"\

21.8

Cross Section

250 (
>
k

250 (
>
>k

Slot: 6 mm x 3.0 mm
Slot Area : < 10 %

250 (

"+<

>f

All Strain Gauges are 5 m m ,


120 Q , Gauge Factor 2.15

250 '
k

2f50 '
(
>r
>k

c
cr
r

'
6

Reduction in bolt
strength 1 0 %

'

250 '

-+:

250 (

D
>
>i

Strain Gauges

250 (

>

<
i

*1
i

21

30 1

>* ^

pJ

1f

L o n g Section

Figure 4.12: Strain gauge and bolt layout for T o w e r Colliery (Mine 2).

88

Chapter 4: Field Investigations

Figure 4.13: Bolt segment showing channels.

Figure 4.14: A section of an instrumented bolt showing the strain gauge and wirings
through the silicon gel.

89

Chapter 4: Field Investigations

4.3.2

Sonic Probe Extensometer

Bed separation and general ground deformation caused by face movement was
monitored by using GEOKON 701 multi-point sonic probe extensometer. During
monitoring, the extensometer probe (8 m long) was inserted in the guide tube located
along the 38 mm diameter boreholes. Placed at a predetermined interval along the
borehole were magnetic anchor points. A total of twenty reference magnets were
housed in each borehole, and the position of each anchored magnet was picked up by
the inserted extensometer probe.

The principle of the sonic probe relies on the magnetic properties of the probe
material. An electrical pulse in the probe creates a magnetic field, which interacts
with the field of a permanent magnet located in a borehole anchor. The change in the

magnetic field generates, a torsional stress pulse in the probe material, which travels
along the probe at the speed of sound. The arrival time of the stress pulse at the end
of the probe is detected, and this is a reliable measure of magnet position. The
position of each anchored magnet was measured and displayed directly on the
GEOKON 701 readout box which has a least reading of 0.025 mm. Figure 4.15
shows various components of a sonic probe extensometer.

90

Chapter 4: Field Investigations

a)

b)

c)
Figure 4.15: Various components of an extensometer, a) readout unit, b) probe, and
c) photograph showings the process of taking readings in the
underground.

91

Chapter 4: Field Investigations

4.3.3 Intrinsically Safe Strain Bridge Monitor

An intrinsically safe Strain Bridge Monitor (SBM), IS2000, was used for the
underground measurement of strains developed in the instrumented bolts. SBM is set
to operate with the more commonly used 120Q strain gauges. By choice of

appropriate bridge circuit, it was possible to measure the strain in a single gauge, t

gauges or four gauges. The quarter bridge configuration was restricted to 120__ strain
gauges only. As the SBM had a fixed gauge factor setting of 2.00, the actual strain
measurement indicated by the display could be calculated as follows:

2V
E =G

For quarter bridge configuration

(4.1)

E =
G

For half bridge configuration

(4.2)

E = For full bridge configuration (4.3)


2G

where,
E = the mean actual strain measured by an active gauge,
Vd= the change in SBM reading, and
G = the gauge factor of the strain gauge.

Figure 4.16 shows a general view of the SBM, while taking readings in underground.

92

Chapter 4: Field Investigations

Figure 4.16: Photograph showing the process of taking reading with S B M in the
underground.

4.4 FIELD MONITORING AND DATA ANALYSIS

4.4.1

W e s t Cliff Colliery

Site monitoring was conducted at regular intervals from the instrumented site. The
primary horizontal stress around the region was estimated to be in the order of 16
MPa (Strata Control Technology, 1996), and its direction was as shown in Figure 4.5.
The direction and the impact of horizontal stress were manifested clearly by

excessive guttering at the left side of the C/T. Thus, the bolts at the left side of the c
through were subjected to excessive shear loading as compared to the bolts at the

93

Chapter 4: Field Investigations

right side of the cut-through. A s can be seen in Figure 4.17, the bolts at the left side
of the C/T experienced significantly higher load transfer in comparison to the bolts at
the right side. The maximum load on the bolt at the left side was 220 kN, whereas the
maximum load at the right side bolt was only 19.46 kN. Therefore, it may be
concluded that the horizontal stress direction played an important role on the bolt
load transfer mechanism. The redistribution of in-situ stress, because of heading
development, has obviously caused differential impact on the reinforcement
elements, such as bolts. The acute angle formed by the prevailing principal
horizontal stress and the heading development direction caused the left side of the
roadway to be under relatively higher shear loading as compared to the other side.

Figures 4.18 and 4.19 show the load transferred to the bolts and the corresponding
shear stresses at the bolt/resin interface for both bolt types at the left side of the cutthrough, respectively. No load build up comparison was possible for bolts installed at
the middle of the cut-through, as the bolt number 2 of type I (see Figure 4.5)
malfunctioned after a short period of installation. The shear stress at the bolt resin
interface was calculated by using the following equation:

Fl- F2
AT =

-
Ttdl

(4.4)

where,
AT = average shear stress at the bolt-resin interface,
Fi = axial force acting in the bolt at strain gauge position 1, calculated from
strain gauge reading,

94

Chapter 4: Field Investigations

F2 = axial force acting in the bolt at strain gauge position 2, calculated from
strain gauge reading,
d = bolt diameter, and
1 = distance between strain gauge position 1 and strain gauge position 2.

_ 1/13/99

Type II bolts (installed at the left side)

-m 1/13/99
-___ 1/14/99
-*--1/25/99
-*_- 2/3/99
- _ 2/12/99
_ ) _ _ 3/4/99
5/18/99
5/26/99
-* 6/2/99

250

- B 6/4/99
-A6/10/99

Load on bolt(kN)

-X6/17/99

Type II bolts (installed at the middle)


.1/13/99

b)

1/13/99
.1/14/99
.1/25/99
.2/3/99
.2/12/99
.3/4/99
.5/18/99

250

Load on bott(kN)

-1/13/99

T y p e II bolts (installed at the right side)

-*_-1/13/99
___-1/14/99
-*1/25/99

c)

-*_ 2/3/99
- _ 2/12/99
_l

3/4/99

___-5/18/99
___ 5/26/99
-06/2/99

150

100

200

250
- O 6/4/99
-A6/10/99

Load on bolt(kN)

___fi/i7/aa

Figure 4.17: Load transferred o n the type II bolts, West Cliff Colliery.

95

Chapter 4: Field Investigations

T y p e II bolts (installed at the leftside)

-1/13/99
-1/13/99
-A1/14/99
-X1/25/99
-* 2/3/99
-2/12/99
- ( _ 3/4/99
5/18/99

5/26/99

-06/2/99
-B6/4/99
0

50

100

150

200

250

-A 6/10/99
-*6/17/99

Load on bolt(kN)

cmnmn

Type I bolts (installed at the leftside) -#1/13/99


-m-1/13/99
-41/14/99

E
E

-K1/25/99

+*

-*-2/3/99

__

-2/12/99

._?
"55

-13/4/99
5/18/99

__
*-

o
o
0. -50

5/26/99
-6/2/99

50

100

150

200

250

- B 6/4/99
-A6/10/99

Load on bolt(kN)

-*6/17/99

Figure 4.18: L o a d transferred o n the bolts installed at the left side of the C/T, W e s t
Cliff Colliery.

96

Chapter 4: Field Investigations

T y p e II bolts (installed at t h e leftside)

-1/13/99
-m 1/13/99

1800 _

-41/14/99
-*-1/25/99
-*- 2/3/99
--2/12/99
-l_ 3/4/99
5/18/99

5/26/99

-e6/2/99
-B 6/4/99

-20

-10

10

20

-& 6/10/99
-K-6/17/99

Shear stress (MPa)

ennmn

Type I bolts (installed at the leftside)

-1/13/99
-1/13/99

1800

-^1/14/99

E
E,

-*1/25/99
-*_ 2/3/99

4-1

__

--2/12/99

._?
'55
o
o

_f3/4/99
5/18/99
5/26/99

__

-06/2/99
-B6/4/99

-20

-10

10

Shear stress (MPa)

20

-A 6/10/99
-* 6/17/99

Figure 4.19: Shear stress developed at the bolt/resin interface of the bolts installed at
the left side o f the C/T, W e s t Cliff Colliery.

Figure 4.18 s h o w s the load transferred o n the bolts, both type I and type II, installed
at the left side o f the cut-through. A s observed f r o m the figure, the axial load transfer
o n both type II and type I bolts w a s nearly o f the s a m e magnitude (Figures 4.18a and
4.18b), h o w e v e r , the shear stress sustained b y the bolt/resin interface w a s slightly
higher in the type II bolt as c o m p a r e d to the type I bolt (Figures 4.19a and 4.19b).

97

Chapter 4: Field Investigations

Figures 4.20 and 4.21 s h o w the axial load transferred o n the bolts and the
corresponding shear stress on type I and type II bolts, installed at therightside of the
cut-through, respectively. There is a considerable difference between the loads
transferred in type I and type II bolts. Clearly, the axial load transferred on the type I
bolts w a s m u c h higher than that of type II bolts (see Figures 4.20a and 4.20b) due to
the impact of both magnitude and direction of prevailing principal horizontal stress in
the instrumented site. T h e shear stress sustained by the bolt/resin interface w a s also
higher in the type I bolts as compared to the type II bolts (see Figure 4.21a and
4.21b).
_#1/13/99

T y o e II bolts (installed at the right side)

-m 1/13/99
__1/14/99
-X1/25/99

E
E
*
_:

_* 2/3/99
--2/12/99

._?
'3

H 3/4/99

o
o

-_5/18/99

a)

-___ 5/26/99

c_

-6/2/99
-Et- 6/4/99

80

-60

-40

-20

20

40

60
-_ 6/10/99

L o a d o n bolt(kN)

-x6/17/99

T y p e I bolts (installed at the right side)

-1/13/99
-m- 1/13/99
-41/14/99
-X1/25/99
-3K- 2/3/99
-*_ 2/12/99
H 3/4/99
-___ 5/18/99
5/26/99
-6/2/99
_g_ 6/4/99

-80

-60

-40

-20

20

40

60

-A6/10/99
- * ^ 6/17/99

L o a d o n bolt(kN)

_* 6/29/99

Figure 4.20: Load transferred on bolts installed at therightside of the C/T, West
Cliff Colliery.

98

Chapter 4: Field Investigations

T y p e II bolts (installed at the right s i d e )

1/13/99
-1/13/99
-A 1/14/99
*1/25/99
* - 2/3/99
_ 2/12/99 |
43/4/99 :
-

5/18/99| |

5/26/99 |

a)

-06/2/99
-B-6/4/99

-10

-5

10 -<_ 6/10/99

S h e a r stress ( M P a )

-*_ 6/17/99

T y p e I bolts (installed at the right side)

-1/13/99
-1/13/99
-*_ 1/14/99
-*_1/25/99
f

-* 2/3/99
-2/12/99 ||
-13/4/99

_-5/18/99
5/26/99
-06/2/99
-B6/4/99
-tr- 6/10/99!

-^6/17/99 J
-*6/29/99

Figure 4.21: Shear stress developed at the bolt/resin interface of the bolts installed at
therightside of the C/T, West Cliff Colliery.

The examination of the load transferred and subsequent shear stress build up on the
bolts at therightside have revealed a different picture in comparison with the bolts
installed at the left side of the C/T. Clearly, the level of load build up on the bolts
installed at the left side of the C/T w a s influenced by the presence and the direction
of high horizontal stress.

99

Chapter 4: Field Investigations

4.4.2 T o w e r Colliery

Following the installation of instrumented bolts and extensometers, the site was
monitored at regular intervals. The primary horizontal stress around the region was
estimated to be 25 MPa (Strata Control Technology, 1996), and the direction was
nearly parallel to the axis of the gateroads as shown earlier in Figure 4.9. The
regional stress pattern indicated that, the direction of the principal horizontal stress
should be near perpendicular to the axis of the gateroads. However, the presence of a
major dyke at the outbye of the instrumented sites caused reorientation of the
prevailing horizontal stress direction (see Figure 4.8). Severe guttering and
deteriorations in general roof conditions were observed up to the location of the dyke,
and beyond that, there was a relative improvement in ground conditions (including
the instrumentation site) with minor guttering observed along the left side of both the
access roads as shown in Figure 4.22a. No guttering was observed on the other (right)
side of both the gateroads (see Figure 4.22b). Thus, it can be inferred that the bolts
on either side of the gateroads were subjected to differential shear loading, with the
left side being subjected to sightly higher loading than that on the right side.

Figure 4.22: R o o f condition in the gate roads at Tower Colliery, a) left side, and b)
right side.

100

Chapter 4: Field Investigations

4.4.2.1 L o a d

transfer during the panel development and the longwall

retreating phases

Figures 4.23 and 4.24 show the overall load transferred on the type II bolts during
panel development and subsequent longwall retreating phases, installed at the left
side of the travelling road and the belt road (numbered as TRA1 and BRA1 in Figure
4.9), respectively. As expected, the load transferred on the bolts during the panel
development stage was relatively low as compared to the longwall retreating phase in
both the travelling and belt roads. The load developed up due to the front abutment
pressure of retreating longwall face was, on the average, 5 to 8 times higher than that
of panel development phase (see Figures 4.23 and 4.24).

5.0 m

TRA1 (Panel development phase)

~_16 m
20 m
_v- 39.7 m

2500-

If
4

1500 -

JC
D)
tt

1000 - iT

o
o
(_

- * - 62.7 m
-64.7 m

200&N

t110.7 m

176.7 m
198.7 m

500-

270.3 m
279.9 m

&
i

-20

20

40

60

80

100

355.4 m

Load on bolt (kN)


* 355.4 m

TRA1 (Longwall retreat phase)

-*261.9 m
-252.1 m
- - 250.3 m

193.9 m

2500 | 2000 -

158.4 m

_: 1500
1 1000- j w

I 500-

~~"-~-

Ir^"^

u ()

20

40

-*-124.7 m
-_104.4 m

-________

-wt-94.1 m
-^-89.7 m
i

-*- 81.4m

60

Load on bolt (kN)

80

100

-65.2 m
-n38.3m j,

17.9 m

Figure 4.23: Load transferred on the bolt T R A 1 , during the panel development and
longwall retreating phases, Tower Colliery.

101

Chapter 4: Field Investigations

BRA1 (Panel development phase)

2.7 m
19.1 m

2500

_
E

2000/

25 m

25 m
25 m

1500 -

36.3 m
38.3 m

1000 -

43.3 m ,

o
o
(
_

500

138.3 m
153.2 ml

0
-20

20

40

60

80

100

168.5 m j
168.5 m

Load on bolt (kN)

-X- 264.1 m
v- 264.1 m

BRA1 (Longwall retreat phase)

_*264.1 m
_-.264.1 m
-+_ 254.5 m

E
E

___ 244.7 m
-, 242.9 m

J_

-_ 186.5 m

o>

151 m

o
o
(
_

117.3 m
97 m
86.7 m

20

40

60

Load on bolt (kN)

82.3 m

80

100

74 m
.57.8 m

._!

Figure 4.24: Load transferred on the bolt B R A 1 , during the panel development and
longwall retreating phases, Tower Colliery.

The m a x i m u m load transferred to the bolt B R J 2 at a particular face position, during


panel development and longwall retreating phases, is shown in Figure 4.25. The load
transferred during the panel development stage became constant within 40m advance
of development headings, away from the instrumented sites. However, the load build
up due to the front abutment pressure of the approaching longwall face began to
increase significantly, w h e n the distance was 150m from the test sites.

102

Chapter 4: Field Investigations

BRJ2
120

z
_< mn

Panel development
Longwall retreat

**

80

o
E 60
3
E 40
X

20

%-

50

100

150

200

250

300

Face distance from the test site (m)

Figure 4.25: M a x i m u m load transferred on the bolt B R J 2 , for a particular face


position, Tower Colliery.

4.4.2.2 L o a d transfer in the belt and travelling road

Figure 4.26 shows the load transferred on the bolts BRJ2 and TRJ2 (both type I)
respectively, during the panel development stage. The corresponding load transferred
on the same bolts, during the longwall retreating phase, is shown in Figure 4.27. No
significant changes in load transfer were observed, either during the panel
development (Figure 4.26) or during the longwall retreating (Figure 4.27) phases.
However, the load transferred on the belt road, in general, was slightly greater as
compared to the travelling road during both panel development and longwall
retreating phases. The differential response by the instrumented bolts in the gateroads
was expected because of the relative positions of the gateroads with respect to the
mined longwall panel as shown in Figure 4.8.

103

Chapter 4: Field Investigations

TRJ2
-5.0 m

-16m

2500

20 m I

E 2000
E.
_: 1500

-39.7 m
-62.7 m

._?
_: 1000

-64.7 m

o
o

-110.7 mj

500

-176.7 m
198.7 m i

0
5

10

15

-270.3 m

20

279.9 m

Load on bolt (kN)

355.4 m !

BRJ2

- 2.7 m
-_19.1 m

2500

25 m
-* 25 m

1 2000

-^-25m
_f 1500

-36.3 m
-i 38.3 m

._?
"3 1000
o
o
cc

500

43.3 m
138.3 m

-o153.2 m

0
10

15

Load on bolt (kN)

20

25

-D

168.5 m

-A-168.5 m
-x264.1 m

Figure 4.26: Load transferred on the bolts TRJ2 and BRJ2 during the panel
development phase, Tower Colliery.

104

Chapter 4: Field Investigations

-x- 355.4 m
x 261.9 m

TRJ2

-*252.1 m

2500

-+- 250.3 m
193.9 m
158.4 m
-124.7 m
-_ 104.4 m
-_-94.1 m

| 2000
_: 1500
!_ 1000

500

-x-89.7 m

0
20

40

60

80

100

120

Load on bolt (kN)

BRJ2
2500
-g 2000
E
*T 1500
O)

I 1000
s
D; 500

20

40

60

80

Load on bolt (kN)

100

120

-*- 81.4 m
-65.2 m
-i38.3 m
-264.1 m
-254.5 m
-244.7 m
-242.9 m
-186.5 m
-151 m
-117.3m
-97 m
-86.7 m
-82.3 m
-74 m
-57.8 m
-30.9 m
-10.5 m

Figure 4.27: Load transferred on the bolts T R J 2 and B R J 2 during the longwall
retreating phase, Tower Colliery.

Figure 4.28 shows the m a x i m u m load transferred on to B R J 2 and T R J 2 for a


particular face position, during the panel development and the longwall retreating
phases. In both gateroads, the load on the bolts started to build up immediately after
their installation during the panel development stage, and then became constant when
the heading development face was about 50m away from the test sites (see darker
lines in Figures 4.28a and 4.28b). During the longwall retreating phase, however, the
impact of front abutment pressure was observed (by sharply increasing load), when

105

Chapter 4: Field Investigations

the approaching longwall face was around 60m away from the test site in case of
travelling road (Figure 4.28a).

TRJ2

a)

120

100
a

80

n
o
E

60

- Panel development
- Longwall retreat

40

E
20
"5
*-
CO

=_

--_[-

100

200

400

300

Face distance from the test site (m)

b)

BRJ2
120
100
o

- Panel development
* Longwall retreat

80

ra

60 -

40 -

20

"I
s

_-

0
0

50

100

150

200

250

300

Face distance from the test site (m)

Figure 4.28: M a x i m u m load transferred on the bolts TRJ2 and BRJ2, for a particular
face position, Tower Colliery.

106

Chapter 4: Field Investigations

In case of belt road, the same was observed w h e n the approaching longwall face was
about 150m away from the test site (Figure 4.28b). Thus, the impact of longwall face
movement was more prominent in case of belt road as compared to the travelling
road. It is noted that, the travelling road was adjacent to the solid longwall block
(longwall panel 19, yet to be mined), whereas the belt road was driven in a relatively
disturbed zone between the present longwall face and the travelling road, thereby
influencing the load transfer mechanism.

4.4.2.3 Behaviour of strata deformation

Figures 4.29 - 4.31 show the overall roof deformation recorded from the
extensometry readings in the travelling road. As expected, the maximum deformation
was recorded in the middle of the road (Figure 4.30) because of the prevailing near
parallel principal horizontal stress direction, and was aggravated by the deadweight
of the separated sagging roof. The amount of roof deformation recorded at the left
side of the gateroad (Figure 4.29) was relatively small as compared to the middle
section (Figure 4.30), but was greater than that on the right side (Figure 4.31) of the
roadway. Also, the acute angle between the horizontal stress direction and the axis of
the gateroads caused some shearing in the immediate roof at the left side of the
gateroads, while the other side of the gateroads was relatively stable, with negligible
amount of strata deformation (Figure 4.31). No significant strata deformation was
observed at a horizon level of more than 3m from the roof level. Thus, it may be
suggested that, in addition to the regular bolt pattern at Tower Colliery, occasional
use of longer secondary reinforcement units (e.g. cable bolts) may be required for

107

Chapter 4: Field Investigations

effective heading stabilisation. However, it is difficult to suggest similar strata


reinforcement pattern at outbye side of the dyke, because of varying stress conditions.

Tower-MG18-TR1

-16 m
-_-20m
-_ 39.7 m

8-

- x - 176.7 m
- * - 198.7 m
__ 270.3 m

7*
!i

i 279.9 m
355.4 m

I i

261.9 m

i i

193.9 m

5,

-x-158.4 m
*124.7 m

i;

o
_

355.4 m

_104.4 m

49 r

a
X

3?

+94.1 m

89.7 m

-81.4 m
- 65.2 m
-_38.3 m

-*- 17.9 m

T 7

-^--2m
~x- -25.7 m
U

-10

-36.3 m

QMTA

X | W

10

20

30

40

50

60

Displacement (mm)

Figure 4.29: Strata deformation recorded in the extensometer T R 1 , Tower Colliery.

108

Chapter 4: Field Investigations

Tower-MG18-TR2

-16m
-_-20m
_ 39.7 m
-*-176.7 m
-*- 198.7 m
-270.3 m
-+-279.9 m
355.4 m
355.4 m
o

261.9 m

193.9 m
-*- 158.4 m

S
c
x

-*- 124.7 m
.*_ 104.4 m
_+_ 941 m

89.7 m
81.4 m

--65.2 m
-_ 38.3 m
-_17.9 m
m

-M2

-*- -25.7 m
-10

10

20

30

40

50

-36.3 m

Displacement (mm)

Figure 4.30: Strata deformation recorded in the extensometer T R 2 , Tower Colliery.

109

Chapter 4: Field Investigations

Tower-MG18-TR3

--16m
-_20 m
* 39.7 m
-x-176.7 m
-* 198.7 m

6-

--270.3 m
-t279.9 m

355.4 m

355.4 m

261.9 m

s.

193.9 m
- x - 158.4 m

a
a
x

-*124.7 m

3'

-* 104.4 m
- 4 94.1 m
- 89.7 m
-81.4 m
-65.2 m
38.3 m
-17.9 m
-2 m
-25.7 m

-10

10

20

30

40

50

60

-36.3 m

Displacement (mm)

Figure 4.31: Strata deformation recorded in the extensometer T R 3 , Tower Colliery.

Figure 4.32 shows the m a x i m u m deformation (recorded from T R 1 ) with respect to


the face distance from the test site, during the panel development and the longwall
retreating phases. A s can be seen from the figure, a negligible amount of roof
deformation was recorded during the panel development stage. The roof deformation
began to build up, when the approaching longwall face was about 50 m from the
instrumented site. The pattern of deformation was similar to the pattern of load build
up on the bolts as discussed in Section 4.4.2.2.

110

Chapter 4: Field Investigations

TR1

Panel developm
Longwall retreat

JO

200

300

400

Face distance from the test site (m)

Figure 4.32: M a x i m u m deformation recorded in the extensometer T R 1 , for a particular


face position, Tower Colliery.

Figure 4.33 shows the three-dimensional surface profile of the maximum


deformations recorded by the extensometers in the travelling road, for a particular
face position during the longwall retreating phase. As discussed before, the
maximum deformation was found to occur near the middle section of the roadway,
with significant deformation occurring when the approaching longwall face was less
than 60m from the instrumented site.

Ill

Chapter 4: Field Investigations

Figure 4.33: Three-dimensional surface generated from the m a x i m u m strata


deformations recorded in the travelling road, Tower Colliery.

4.4.2.4 Impact of strata deformation on the load transfer in the bolts

Figure 4.34 shows the relative displacement between the magnetic reference anchors
along the borehole, for the extensometer BR2, installed in the middle of the belt road.
The relative movement was initiated mainly at two levels i.e. at 600 mm and at 2300
mm levels above the roof. The load transferred on the bolt BRA2 (installed close to
extensometer BR2) is shown in Figure 4.35. The maximum load transferred on to the
bolt at 600 mm above the roof was in close agreement with the location of the
maximum strata deformation. Thus, there exists a direct relationship between the
load transfer in the bolt and the relative strata deformation at any particular horizon.

112

Chapter 4: Field Investigations

N o comparison between the strata deformation and the load build up on the bolt
could be drawn at 2300 mm level, as the position of the last strain gauge was at 2200
mm above the roof level.

Tower-MG18-BR2

19.1 m
-_-25m
~_25 m

8-

- x 25 m
- * 36.3 m
38.3 m

7;?

t 153.2 m

] [

168.5 m

168.5 m

264.1 m

_-.- 264.1 m
;i

264.1 m

<i

_*-. 254.5 m

4) l

-*244.7 m

c
_:

%
o
o
r_

186.5 m

_?

151 m
3^

117.3 m
97 m

<

_- 86.7 m

2
i

> J

- _ 82.3 m

x74 m

1
i| /*"

-*57.8 m
-30.9 m
-H10.5 m

-5

()

10

15

20

25

30

35

Relative displacement (mm)

Figure 4.34: Relative displacements between the two consecutive extensometer


probes in B R 2 , Tower Colliery.

113

Chapter 4: Field Investigations

-^-264.1 m
-*- 264.1 m
-*-264.1 m
254.5 m
244.7 m |
242.9 m
--186.5 m
-_-151m |
-_-117.3m
-*-97m

BRA2
2500 n

5ooo~^

1500

_=
O)

_o
o

1000 J

0_

500 -i

-50

-H*-86.7m

50

100

150

250 --82.3 m

200

i74 m

Load on bolt (kN)

57.8 m :

Figure 4.35: Load transferred on the bolt B R A 2 , during the longwall retreating phase,
Tower Colliery.

The m a x i m u m strata deformation recorded on T R 2 , and the m a x i m u m load


transferred on the bolt TRA2 due to front abutment pressure, with respect to the
approaching longwall face position are shown in Figures 4.36 and 4.37, respectively.
Similar trends in the maximum strata deformation and the maximum load transferred
on the bolt were observed, both showing significant increase when the approaching
face was about 60m from the test site.

TR2

E
E.
c
o

Panel development
Longwall retreat

CO

a
o

o
-

2
o -100

200

100

300

400

Face distance from the test site (m)

Figure 4.36: M a x i m u m deformation recorded in the extensometer T R 2 , for a


particular face position, Tower Colliery.

114

Chapter 4: Field Investigations

TRA2
120
o

ra

Pi

o
E

Lc

development
rail retreat

100

200

-__-

300

400

Face distance from the test site (m)

Figure 4.37: M a x i m u m load transferred on the bolt T R A 2 , for a particular face


position, Tower Colliery.

4.4.2.5 Comparison of load transfer in type I and type II bolts

Figure 4.38 shows the load transferred on the bolts TRJ1, TRJ2 and TRJ3, installed
at the left, middle and the right side of the travelling road. The maximum load
recorded on the above bolts was 39 kN, 97.6 kN and 33.5 kN, respectively. The bolt
at the left side was subjected to relatively higher load as compared to the bolts at the
right side, which may have been due to the influence of the orientation of principal
horizontal stress, striking the gateroads with an acute angle from the left side (Figure
4.9). When compared with other bolts, the bolt at the middle of the road recorded the
maximum value because of the dominant role of excessive strata deformation in the
middle of the gateroads.

115

Chapter 4: Field Investigations

-*- 355.4 m

TRJ1

x- 261.9 m
-252.1 m

250.3 m
193.9 m

-158.4 m
--124.7 m
-104.4 m
-_-94.1 m
-x-89.7 m
40

60

80

-*- 81.4 m
120 -*- 65.2 m

100

I38.3 m

Load on bolt (kN)

17.9 m

^355.4 m
-*~261.9m
--252.1 m

TRJ2
2500 1

i 250.3 m
193.9 m
158.4 m
-+- 124.7 m
-m 104.4 m
-_-94.1 m

| 2000
S 1500
._?
j_ 1000
o 500
cc
20

40

60

80

100

120

Load on bolt (kN)

-*- 89.7 m
-*-81.4m
--65.2 m
-438.3 m

-*- 355.4 m

TRJ3

-*- 261.9 m
-*- 252.1 m
250.3 m
193.9 m

158.4 m

-*-124.7 m
-*-104.4 m
-Tt-94.1 m
-^-89.7 m
-81.4 m

-20

20

40

60

80

Load on bolt (kN)

100

120. - 65.2 m
-i38.3 m

17.9 m

Figure 4.38: Load transferred on the type I bolts, installed in the travelling road,
T o w e r Colliery.

116

Chapter 4: Field Investigations

Figure 4.39 shows the pattern of load transferred on the type I (BRJ1) and type II
(BRA1) bolts, installed at the left side of the belt road. The load transferred on the
type I bolts was relatively smaller as compared to the type II bolts. The maximum
load transferred on BRJ1 and BRA1 was 41.7 kN and 84.6 kN, respectively. The
corresponding shear stress developed at the bolt/resin interface for both bolts is
shown in Figure 4.40. The comparative values observed from the shear stress profiles
of BRJ1 and BRA1 suggests that, the bolt type II offered better load transfer
characteristics when subjected to higher shear loading, caused by the influence of the
horizontal stress.

-*- 264.1 m
-* 264.1 m

BRJ1

--264.1 m
. 254.5 m

244.7 m

2500-

2000-

E
*T

1
1500

CD

1000-

>_.
o
__

500i

__C_L^'WV^

-20

"""

I V . \.^"_i
a. \ L l T I
i

20

40

60

80

100

Load on bolt (kN)

- 242.9 m
-+- 186.5 m
-_i151 m
-_ 117.3 m
-*-97 m
-*-86.7m
- 82.3 m i
i74 m

57.8 m
-30.9 m
-H~ 264.1 m
-*- 264.1 m

BRA1

-^-264.1 m
2500 -i
E
E.
*T

i 254.5 m
244.7 m

2000"
"
1500 -

1000

-151 m
^_117.3 m

CO

J_
|

500
i

-20

()

242.9 m

- 186.5 m

20

40

60

Load on bolt (kN)

80

100

-x97 m
-*- 86.7 m
-* 82.3 m
i74 m

57.8 m '

Figure 4.39: Load transferred on the type I and type II bolts, installed at the left side
of the belt road, Tower Colliery.

117

Chapter 4: Field Investigations

254.5 m
244.7 m
242.9 m
-186.5 m
-*-151 m
-_ 117.3 m
-x-97m
-*- 86.7 m
- 82.3 m
i74 m
57.8 m
30.9 m
-10.5 m

BRJ1

-2.0

0.0

2.0

4.0

Shear stress (MPa)

--264.1 m

BRA1

* 254.5 m

244.7 m

242.9 m
i

-186.5 m
-*151 m
-*-117.3 m
-x-97m
-*- 86.7 m
*- 82.3 m
-4.0

-2.0

0.0

2.0

Shear stress (MPa)

4.0

-474 m

57.8 m

30.9 m

Figure 4.40: Shear stress developed at the bolt/resin interface of the type I and type II
bolts, installed at the left side of the belt road, Tower Colliery.

The neutral point (zero shear stress position), as defined by Tao and Chen (1983), for
both BRJ1 and BRA1 occurred at around 500 mm above the roof level. Accordingly,
the pickup length (from roof level to the neutral point) and the anchor length (from
neutral point to the end of the bolt) were calculated to be 0.5m and 1.8m,
respectively. Thus, it can be inferred that, the location of the neutral point is
independent of the bolt type.

118

Chapter 4: Field Investigations

Figure 4.41 shows the load transferred in type I (BRJ3) and type II ( B R A 3 ) bolts,
installed at the right side of the belt road. As expected, the load transferred in type I
bolt (38.1 kN) was relatively higher as compared to the type II bolt (18.6 kN).
Because of the lower influence of the horizontal stress on the bolts, the shear stress
developed at the bolt/resin interface in type I bolt was relatively higher than in type II
bolt (see Figure 4.42), thus, reconfirming the superior load transfer characteristics of
the type I bolts under lower levels of prevailing horizontal stress.

BRJ3

o
o
DC

-10

10

20

30

40

50

Load on bolt (kN)

BRA3

.SP
o
o

-10

10

20

30

Load on bolt (kN)

40

50

-K264.1 m
* 264.1 m
--264.1 m
-H-254.5 m
244.7 m
242.9 m
-186.5 m
-_151 m
-_117.3m
-*-97m
-*- 86.7 m
--82.3 m
-H74 m
57.8 m
*-264.1 m
*- 264.1 m
264.1 m
4254.5 m
244.7 m
242.9 m
186.5 m
151 m
117.3 m
x 97 m
* - 86.7 m
82.3 m
+-74m
57.8 m

Figure 4.41: Load transferred on the type I and type II bolts, installed at therightside
of the belt road, T o w e r Colliery.

119

Chapter 4: Field Investigations

BRJ3

--264.1 m
-+ 254.5 m

2500

244.7 m

242.9 m

E
E

-186.5 m

si

-_151 m

._?
"55
_r

-_-117.3m
-*97 m
-*86.7 m

<-

--82.3 m

o
o

-\74 m

-0.5

0.

0.0

0.5

1.0

1.5

57.8 m

-^-30.9 m

Shear stress (MPa)

-10.5 m
-* 264.1 m

BRA3

-+- 254.5 m

244.7 m

-242.9 m
-*- 186.5 m
---151 m

.E?
S
o
o

-_-117.3 m
-* 97 m
-*-86.7 m
- 82.3 m

0_

-i74 m

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Shear stress (MPa)

57.8 m
30.9 m

--10.5 m

Figure 4.42: Shear stress developed at the bolt/resin interface of the type I and type II
bolts, installed at the right side of the belt road, Tower Colliery.

The neutral point for both B R J 3 and B R A 3 w a s found to be at around 820 m m above
the roof level. The neutral point appears to be dependent on the stress condition at the
installed bolt location and the strata deformation behaviour. This was evidenced by
the different neutral point locations (for the same type of bolt) of 500mm on the left
side and 8 2 0 m m on therightside of the gateroad respectively.

120

Chapter 4: Field Investigations

O n the other hand, the pattern and the magnitude of the load transferred on the type I
(TRJ2) and type II (TRA2) bolts, installed at the middle of the gateroad, is shown in
Figure 4.43. The near parallelism of the direction of the principal horizontal stress to
the axis of the gateroads caused insignificant shear loading on the bolts installed at
the middle of the gateroads. The strata deformation due to the dead weight of the
immediate roof, which affected both bolt types equally, caused almost equal levels of
load transfer on both type of bolts. This is clearly demonstrated by the pattern of load
distribution shown in Figure 4.44, where-by the load profiles for both bolt types I and
II are similar during the entire longwall retreating phase. Similarly, the shear stresses
developed at the bolt/resin interface of these two bolts also showed identical
behaviour (see Figure 4.45).
-355.4 m

TRJ2

-261.9 m
-252.1 m

2500

250.3 m

| 2000

-193.9 m

_: 1500
_?
j? 1000
o 500
a.

-158.4 m
-124.7 m
-104.4 m
-94.1 m
-89.7 m
20

40

60

80

100

120

Load on bolt (kN)

-81.4 m
-65.2 m
-38.3 m

TRA2

-355.4 m
-261.9 mi
-252.1 m!
-250.3 m
-193.9 m
-158.4 m

._?
'3
_:
*^
o
o
or -20

-124.7 m
-104.4 mi
-94.1 m !
-89.7 m \
-81.4 m j

20

40

60

Load on bolt (kN)

80

100

120

-65.2 m |
-38.3 m ;
-17.9 m i

Figure 4.43: Load transferred on the type I and type II bolts, installed at the middle of
the travelling road, Tower Colliery.

121

Chapter 4: Field Investigations

120

_? 100
~

80

TRJ2
TRA2

(0

o 60
E 40
3
E
x
m

20
-#100

200

300

400

Face distance from the test site (m)

Figure 4.44: M a x i m u m load transferred on type I and type II bolts for a particular
face position, installed at the middle of the travelling road, Tower

Colliery.

--355.4 m

TRJ2

-_-261.9 m
-_- 252.1 m
-x-250.3 m
-*- 193.9 m

E
E

-158.4 m
-+-124.7 m

._?

==rb=

o
o
(
_

104.4 m |
94.1 m i

50^

_r=-K

, 89.7 m i
-_- 81.4 m

-6.0

-4.0

-2.0

0.0

2.0

4.0

6.0

65.2 m
-x- 38.3 m

Shear stress (MPa)

__ 17 Q m j

-* 355.4 m
-_261.9 m

TRA2

-A-252.1 m

2500
E
E.
_:
_?

-x 250.3 m
-*- 193.9 m
-^158.4 m
-4124.7 m

_*"

JQOO^j

104.4 m

94.1 m

o
o

500 >

89.7 m
81.4 m

-6.0

-4.0

-2.0

0.0

2.0

Shear stress (MPa)

4.0

6.0

65.2 m
-*- 38.3 m
-*- 17.9 m

Figure 4.45: Shear stress developed at the bolt/resin interface of the type I and type II
bolts, installed at the middle of the travelling road, Tower Colliery.

122

Chapter 4: Field Investigations

The three-dimensional surface profiles generated from the m a x i m u m load transferred


with respect to the face distance from the instrumented site, on type I and type II
bolts, installed in the travelling road is shown in Figure 4.46. Both surfaces indicated
that the maximum load build up occurred at the middle of the gateroad. Some
noticeable amount of load build up occurred when the retreating longwall face was at
a distance of less than 60 m from the instrumented site in both bolt types.

Figure 4.46: Three-dimensional surface generated from the m a x i m u m load transferred on


type I and type II bolts, installed in the travelling road, Tower Colliery.

123

Chapter 4: Field Investigations

4.5 BOLT SELECTION GUIDELINES

The laboratory experiments discussed in Chapter 3 suggested that the type I bolts
offered higher shear resistance as compared to the type II bolts at low normal stress
levels (confining stress condition in the field) on the bolt/resin interface. At higher
normal stress levels, the type II bolts offered marginally higher shear resistance as
compared to the type I bolts. Similar phenomenon was also observed from the field
investigations. When the bolt was subjected to high shear loading in the field, the
shear stress sustained by the type II bolt was marginally higher than the type I bolt;
whereas when the bolt was subjected to low shear loading, the shear stress sustained
by the type I bolts was significantly greater than the type II bolts. Thus, the degree of
shear stress sustained by the bolt was influenced by the level of the applied initial
normal stress on the bolt/resin interface or the confining pressure (in the field) and
the surface profile configuration of the bolt. Accordingly, bolts with deeper and
wider rib spacing should be used in section of the roadways subjected to the
influence of low horizontal stress, whereas, bolts with shallower and narrower rib
spacing should be used in areas under the influence of high horizontal stress (as
evidenced by the excessive roof guttering).

124

Chapter 4: Field Investigations

4.6 CONCLUSIONS

The field investigations conducted in two coal mines in the Illawarra region of NSW,
Australia, revealed the following aspects of the load transfer mechanism.

The load transfer on the bolt was influenced by; a) the confining stress condition,
b) the extent of strata deformation, and c) the surface profile roughness of the
bolts.
The load transferred on the bolts, during the longwall retreating phase, was
relatively greater than that of panel development phase.
The face movement did not influence the load transferred on the bolts, when the
development face moved beyond 50m away, or the approaching longwall face
position was 150m away from the test sites.
There was no significant variation of the load transfer magnitude on the bolts in
either gateroads (belt road and travelling road) during the panel development
stage.
The influence of front abutment pressure build up on the gateroads appears at
different face positions. The load build up on the bolts in the belt road occurs
when the longwall face is less than 150m from the test site, whereas, the same
build up on the travelling road starts when the face position is less than 60m.
The maximum strata deformation was found to occur in the middle of the
gateroads.
The maximum load transferred along the bolts was found to occur at the level of
maximum deformation.

125

Chapter 4: Field Investigations

The neutral point w a s found to be independent of the type of bolt used, and its
value changes with varying stress conditions and strata deformation
characteristics.
The field study showed that, under the low influence of horizontal stress (both in
magnitude and the direction), the type I bolt offered higher shear resistance,
whereas under high influence of horizontal stress, the type II bolt offered better
shear resistance at the bolt/resin interface. Such findings were also observed in
the laboratory studies, and they provide a useful guide for future selection of
appropriate bolt types for given stress conditions.

126

Chapter 5

SHEAR BEHAVIOUR OF ROCK JOINTS

5.1 INTRODUCTION

Joints in rock can be open or closed. Closed joints may contain infill, which is either |
loose or cemented. The nature of joint fillings, thus has considerable influence on the i
strength properties of the rock mass. Rocks with weaker infill joints are generally |
weaker than the same type of rocks with well-cemented joints.

Irrespective of the fillings, rock bolting is widely used for strengthening rock joints,
and the application of bolts has been the subject of considerable research during the
past several decades. Much of the studies were concerned with the effective
application of bolts with respect to the bolt orientation with the joint plane. The joint
orientation is of significant concern in tunnel construction and underground coal
mining heading development, particularly in weak rock stratification. This chapter, is
therefore, concerned with the critical review of the role of bolting in jointed rock
mass stabilization.

Chapter 5: Shear Behaviour of Rock Joints

5.2 REVIEW OF UNFILLED JOINTS

In recent years there has been a growing interest on the study of the shear behaviour
of rock joints under Constant Normal Stiffness condition, instead of the traditional
method of testing under Constant Normal Load conditions. The pioneering work of
Patton (1966), based on a series of tests on regular saw-tooth shaped artificial joints,
was carried out under CNL condition, and test results closely fitted to a bi-linear
shear strength envelope as described in Figure 5.1. Accordingly:

p = Q X tanftyi) +i) for asperity sliding (5.1)

p = c + n X tan(Vb) for asperity shearing (5.2)

where,
TP = peak shear stress,
an = normal stress,
([>_ = basic friction angle,
c = cohesion intercept, and
i = initial asperity angle.

According to Patton (1966), the sliding of asperities takes place under low normal
stress, and beyond a certain normal stress value, shearing through asperities may also
result. It has been observed that the peak shear strength predicted by Patton's model,

128

Chapter 5: Shear Behaviour of Rock Joints

especially at low to m e d i u m normal stress conditions, generally overestimates the


actual shear strength.

Normal stress
Figure 5.1: Bilinear strength envelope for saw-tooth asperity (after Patton, 1966).

Barton (1973) also introduced a non-linear strength envelope for non-planar rock
joints, tested under Constant Normal Load (CNL) condition, and was given by:

= tan <kb+JRCxlog]0

(5.3)

where,
<|>b = $ " (dn + S n ),
= friction angle,
d = peak dilation angle which decreases with the increase in normal stress,

129

Chapter 5: Shear Behaviour of Rock Joints

s = angle due to shearing of asperities which increases with the increase of the
normal stress as more surface degradation occurs,
JRC = Joint Roughness Coefficient, and
ac = uniaxial compression strength.

The JRC-JCS model developed by Barton (1973, 1976) and Barton and Bandis
(1982, 1990) has almost dominated the practice of rock joint engineering and is now
considered to be a de facto international standard. However, despite this recognition,
there remains a concern because of the highly empirical basis of the model, which
may lead to more conservative design solutions. Others engaged in similar research
work include Ladanyi and Archambault (1970), Barton (1973, 1976, 1986), Hoek
(1977, 1983, 1990), Hoek and Brown (1980), Bandis et al. (1983), Hencer and
Richards (1989), Kulatilake (1992), Kulatilake et al. (1993), Saeb and Amadei
(1992), and Brady and Brown (1993), where all these researchers conducted their
studies under CNL or zero stiffness condition.

The joint research under Constant Normal Stiffness condition is a relatively new
concept. The importance of CNS condition, with regard to the shear strength of any
joint plane, has been described in Chapter 2. The shear strength of non-planar joint
increases due to application of external normal stiffness, Kn, which allows a joint to
shear under restricted dilation. As a result, the peak shear stress of joints would be
higher under CNS condition in comparison to that of CNL condition. Leichnitz
(1985) was the first to report on the laboratory studies of rock joints under CNS
condition, and he found that both the shear force and the dilation were functions of

130

Chapter 5: Shear Behaviour of Rock Joints

the normal force and shear displacement. Since then, similar findings have been
reported by Benmokerane and Ballivy (1989), Van Sint Jan (1990), Ohnishi and
Dharmaratne (1990), Skinas et al. (1990), Haberfield & Johnstone (1994), and
Haberfield and Seidel (1998).

Johnstone & Lam (1989) developed an analytical method for the shear resistance of
concrete/rock interface under CNS condition. By assuming the penetration of microasperities of concrete into the rock surface, when the contact normal stress exceeds
the uniaxial compressive strength (see Figure 5.2), they formulated the following
equation for the mobilised cohesion, cm:

sl

-1

=COS
m

(5.4)

qu J

71
where,
csi = cohesion of rock for asperity sliding,
aR = actual contact normal stress, and
qu = uniaxial compressive strength.

131

Chapter 5: Shear Behaviour of Rock Joints

| 0 < f f n < qu

Shear
plane

(c)

Figure 5.2: Idealised penetration of micro asperities of concrete into rock surface
(after Johnston and L a m , 1989).

In another paper, Haberfield and Seidel (1998) have extended the C N S technique to
include soft rock/rock and concrete/rock joints. The following theoretical model was
proposed, based on the careful observation of laboratory direct shear testing on joints
containing two-dimensional roughness:

* = A~ Z ajanj tan{$b + ij)

132

(5.5)

Chapter 5: Shear Behaviour of Rock Joints

where,
t = shear strength of joint,
A = total joint contact area,
n = number of asperities,
aj = contact areas of the individual asperities,
anj = normal stresses acting on the individual asperities,
<)>b = friction angle, and
ij = variable asperity angles.

The processes modelled also included asperity sliding, asperity shearing, post-peak
behaviour, asperity deformation, and distribution stresses on the joint interface.
Model predictions were compared with the laboratory results, and although the model
predictions compared favourably with laboratory test results for concrete/rock joints,
nevertheless, the strength values were over-predicted when applied to rock/rock
joints.

5.3 REVIEW OF UNFILLED REINFORCED JOINTS

The shear behaviour of reinforced joints has remained the focus of research in
geotechnical engineering particularly in; a) the contribution of bolts to the shear
strength of reinforced joints, b) the deformation behaviour, and c) the stability of
reinforced joints. Figure 5.3 shows typical deformation behaviour of a reinforced
joint under shear loading.

133

Chapter 5: Shear Behaviour of Rock Joints

Normal stiffness (* J
= Constant

^ > Shear stress

Shear fracture

Tensile stress
[- 0(*n. .)]

Figure 5.3: Deformation of bolted joint during shearing (after Indraratna et al., 1999).

The study of reinforced joints w a s first initiated by Bjurstrom (1974). Bjurstrom


conducted a series of shear tests, both in the laboratory and in the field on granite
specimens reinforced by fully grouted steel bolts, to determine the influence of
various factors affecting the shear strength of reinforced rock joints. He found that
both the rock strength and the bolt orientation with respect to the joint plane
influenced the strength of the reinforced joint (see Figure 5.4). Bjurstrom reported
that, the bolt failure characteristics were dependent upon the angle between the bolt
and the joint planes, and the bolt failure in tension occurred when the angle was less
than 35. He also suggested that, the shear strength of such rock was dependent on
the following three parameters.

134

Chapter 5: Shear Behaviour of Rock Joints

=30. ds25 m m , <Tn=0,5 MPo


=30_ d=25mm, <T;=0MPa
^cr-70*, d=16mm. <j;=3,5MPa

CO

to

20

30

40

sT^

Shear displacement ( m m )

Figure 5.4: Shear force versus joint displacement (after Bjurstrom, 1974).

(i) Shear resistance due to reinforcement effect:

T b = P(Cos9 +nSin0)

(5.6)

where,
P = bolt tension,
0 = angle between the bolt and the joint,
|j, = tan((|)), and
<|> = angle of internal friction.

(ii) Shear resistance due to dowel effect:

T d = 0 . 6 7 d 2 ( a s . G c ) 0.5

(5.7)

where,
d = bolt diameter,

135

Chapter 5: Shear Behaviour of Rock Joints

a s = bolt yield strength, and


ac = uniaxial compressive strength of rock.

(iii) Shear resistance due to friction of joint itself:

Tj = Aj an. tanxfc (5.8)

where,
Aj = joint area,
an = normal stress on joint, and
(j)j = joint friction angle.

The total resistance offered by a bolt was given by the summation of all the above
three components as shown in Figure 5.5.

Hass (1981) conducted laboratory studies on limestone specimens with an artificially


cut joint, reinforced by full scale rock bolts of various types, which intersected the
shear plane at different orientations as shown in Figure 5.6. It was concluded that, the
orientation of rock bolts relative to the shear plane had a pronounced effect on the
shear resistance offered by a bolt. He suggested that the bolts would act more
effectively when they are inclined at an acute angle to the shear surface than the
opposite direction, as they tend to elongate as shearing progress. The shear strength
of a bolted joint was found to be the sum of the bolt contribution and the friction

136

Chapter 5: Shear Behaviour of Rock Joints

resulting from the normal stress on the shear plane. The contribution due to shearing
of asperities was not considered in the model.

Actual

Resistance

V)
in

t
+J

</>
L
VI
3S

D o w e l Effect

ID)

Shear Displacement

Figure 5.5: Components of shear resistance offered by a bolt (after Bjurstrom, 1974).

SHEAR
FORCE

rSTRAW

GAGES

_______________: h

24"

"-_l_.6"._j

h12

NORMAL BOLT,

INCLINED BOLT,

045*

v \

12"

_*_'

INCLINED BOLT,

_*-45*

Figure 5.6: Arrangement for bolt shear testing (after Hass, 1981).

137

Chapter 5: Shear Behaviour of Rock Joints

Based on several laboratory and field results, Spang and Egger (1990) provided a
detail analysis of the behaviour of fully grouted bolts in jointed rocks. The results
were used to explain the mode of action of fully bonded, untensioned rock bolts in
stratified or jointed rock mass. A three dimensional model was implemented using
the Finite Element Method (FEM) to support the laboratory findings. Three
distinguished stages of behaviour (elastic, yielding, and plastic) were observed during
shearing of the grouted joints. A mathematical model was proposed to determine the
shear resistance of bolted rocks:

T0 = Pt [1.55 + 0.011

GC107

Sin2(a + i)] a*-14(0.85 + 0.45 tan(j))

(5.9)
where,
T0 = maximum shear resistance of grouted rock,
Pt = maximum tensile load of the bolt,
ac = uniaxial compressive strength of rock,
a = inclination of bolt with respect to joint plane,
<|) = angle of internal friction, and
i = asperity angle.

The above equation is valid for rocks with the uniaxial compressive strength within
the range of 10 to 70 MPa and for angles between the bolt and the joint surface
within the range of 60 to 90 degrees.

138

Chapter 5: Shear Behaviour of Rock Joints

Aydan and K a w a m o t o (1992) used stability analysis method for slopes and
underground openings under various loading conditions against flexural toppling
failure. A method was suggested to take into account, the reinforcement effect of
fully grouted rock bolts. It was reported that the installation angle of rock bolts, 0,
must be between 0 and +90 in order to make their effective use. Theoretically, bolts
should be most effective when the installed angle 0 is equal to 90 - <|> (<j) = joint
friction angle), as shown in Figure 5.7. The proposed analytical model was validated
by the laboratory tests under controlled environment. The reinforcing effect of a rock
bolt on the shear resistance of a discontinuity plane was expressed in the following
form:

Tb = AbGb.(l + tan(j).Sin20) (5.10)


___

where,
Tb = shear resistance due to bar,
Ab= cross section of bar,
Gb = magnitude of stress in the bar in the direction of shearing,
<|) = friction angle of discontinuity, and
0 = angle between bolt axis and discontinuity.

The output from the analytical model was compared favourably with those from the
numerical model using FEM as shown in Figure 5.8. In case of a bolt placed
perpendicular to the joint plane, Aydan and Kawamoto (1992) concluded that the bolt

139

Chapter 5: Shear Behaviour of Rock Joints

contribution w a s independent of the joint friction angle, which is not the case in
reality, due to stress rotation upon shearing.

0'

1
rfr

\9,
-90'

90

1(0,^)= cos $ lamp sind

-90

-60 / /GO/ /0

OO

60

90

.5
-1.

-90

-60

N 0

\-3(X

30

60

0
-15
-30"
45

'90

o. Compression (-)

o. Tension (+)
b

..5

-1.

Figure 5.7: Effect of bolt inclination on the bolt shear resistance (after Aydan and
K w a m o t o , 1992).

140

Chapter 5: Shear Behaviour of Rock Joints

0
6)

6 kN
7 *
8.5 -

Thpnrptlcal
FEM

'->
10
15
DISTANCE ALONG BOLT cm

20

5
10
)5
DISTANCE ALONG BOLT cm

20

Figure 5.8: Axial and shear stress on bolt at various normal stresses (after Aydan and
K w a m o t o , 1992).

Ferrero (1995) proposed a shear strength model for reinforced rock joints, based on
the numerical and laboratory studies on large size shear blocks. Ferrero suggested
that, the overall strength of the reinforced joint could be attributed to the combination
of both dowel effect and the incremental axial force due to the bar deformation, as
shown in Figure 5.9. The proposed analytical model was expressed by:

F = T r Cosa - Q.Sina - (Tr Sina - Q.Cosa) tan<|)

141

(5.11)

Chapter 5: Shear Behaviour of Rock Joints

where,
F = global reinforced joint resistance,
Tr = tensile load in bolt,
Q = force due to dowel effect,
a = angle between the joint and the dowel axis, and
<j) = joint friction angle.

The above model is mainly applicable to the bolts installed perpendicular to the joint
plane in stratified horizontal bedding planes.

Figure 5.9: Resistance mechanism of a reinforced rock joint (after Ferrero, 1995).

Windsor (1997) proposed a method for the design of reinforcement for excavation in
jointed rocks using the systems approach. The method was based on identifying and
stabilising blocks of rock that may form at the mine excavation boundary, which is
the core of block theory. According to Windsor (1997), the load transfer concept
consisted of three basic mechanisms:

142

Chapter 5: Shear Behaviour of Rock Joints

Rock movement and load transfer from an unstable zone to the reinforcing
element.
Transfer of load from the unstable region to a stable interior region via the
element.
Transfer of the reinforcing element load to the stable rock mass.

The above mechanisms were supported by both the mathematical and numerical
models, but without any experimental or field verification.

Pellet and Boulon (1998) proposed an analytical model for predicting the
contribution of a bolt to the shear strength of rock joints. The model took into
account, the interaction of the axial and shear forces mobilised in the bolt, as well as
its large plastic deformation occurring during the loading process. The shape of the
stressed bolt and the failure envelope for both elastic and plastic deformations are
shown in Figures 5.10 and 5.11, respectively. The proposed failure criteria is in the
form of:

7lDb

Qof =

Nof

(Jec-J 1 ~ 16
VTlDb

Gec

(5.12)

where,
Qof = shear force acting on the bolt at the joint plane at the point of failure of
the bolt,

143

Chapter 5: Shear Behaviour of Rock Joints

Db = bolt diameter,
o e c = failure stress of bolt, and
N_f = axial force acting on the bolt at the joint plane at the point of failure of
the bolt.

^^/^/^/^/m^^^mmr~

a)

^mmmsm

b)
Figure 5.10: Force components and deformation of a bolt, a) in elastic zone, and b) in
plastic zone (after Pellet and Boulon, 1998).

144

Chapter 5: Shear Behaviour of Rock Joints

relationship between axial and shear


' forces in elastic conditions (eq. 5.26)
axial and shear forces
at the yield limit (eq. 5.27)
yield limit (eq. 5.7b)

a)

failure criterion ( eq. 5.14b)


axial and shear forces
at failure (eq. 5.41)
Qoe=Qof

b)
Figure 5.11: Evolution of shear and axial forces in a bolt, a) in elastic zone, and b) in
plastic zone (after Pellet and Boulon, 1998).

The axial force acting o n the bolt is difficult to determine without the relevant
instrumentation, and the intermediate steps necessary to calculate the failure strength
are also complicated. In addition, the model assumes a constant rock reaction during
shearing process, irrespective of the applied normal stress and the bolt material.

5.4 REVIEW OF INFILLED JOINTS

A s described earlier, the presence of joints in the rock mass plays an important role
on the overall shear and deformability behaviour of rocks, in addition to in-situ stress

145

Chapter 5: Shear Behaviour of Rock Joints

and hydro-geological conditions. T h e behaviour of a joint is governed by the fact


whether the joint is open or closed with infill material. Typical infill materials
existing within the joint interfaces may be divided to; a) loose material from the
surface e.g. sand, clay etc., b) deposition by ground water flow, c) loose material
from tectonically crushed rock, and d) products of decomposition and weathering.
The thickness of infill material can vary from a fraction of a micron to several
centimetres, depending on the situation in which the infill was deposited. For smooth
joints, the thickness of the infill material does not play a significant role, and for
rough joints the interaction between the infill/asperity and asperity/asperity could
occur, depending on the geometry of the asperities and the infill thickness. When the
infill thickness (t) is sufficiently large, for instance, more than twice the asperity
height (a), the shear behaviour would be governed by the characteristics of infill
material alone. Clearly, the t/a ratio appears to play an important role in the
behaviour of infilled joints.

During the last three decades, various researchers have extensively evaluated, the
laboratory shear strength parameters of both natural and artificial infilled rock joints
(e.g. Goodman, 1970; Kanji, 1974; Ladanyi & Archambault, 1977; Lama, 1978;
Barla et al., 1985; Bertacchi et al., 1986; Pereira, 1990; Phien-wej et al., 1990, and de
Toledo & de Freitas, 1993). Parameters considered by various researchers affecting
the infill joint behaviour include the joint type, infill thickness and characteristics,
drainage conditions, boundary conditions, infill-rock interaction, and the system
stiffness.

146

Chapter 5: Shear Behaviour of Rock Joints

Goodman's (1970) research work on saw-tooth shaped joints,filledwith crushed


mica, found that, the shear strength of the joint was greater than that of the infill
alone up to a t/a ratio of 1.25 as shown in Figure 5.12. Similar findings were reported
also by Ladanyi and Archambault (1977), from their studies on concrete blocks with
kaolin clay infill (Figure 5.13). They also found that, the value of the shear strength
increased with the increasing asperity angle, and with the decreasing t/a ratios.

-s
c
__

200

T"""7 T

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Relative thickness, (t/a)
Figure 5.12: Shear strength of mica infilled joint (after Goodman, 1970).

Research studies on sand and clay infilled joints of limestone, sandstone and marl
undertaken by Tulinov and Molokov (1971) found that, a thin layer of sand did not
have any significant influence on the frictional behaviour of hard rocks. In contrast,
the tests on soft rocks showed the opposite result. A study by Lama (1978) indicated
that the strength of infill joint approached the strength of infill material when the t/a
ratio exceeded the critical value of unity. In some cases, the joint shear strength
dropped to that of infill, with a t/a ratio of less than unity. Kutter and Ruttenberg

147

Chapter 5: Shear Behaviour of Rock Joints

(1979) study showed that, the shear strength of the clay infilled joint increased
significantly with increasing surface roughness, while for sand filled joint, the
increase was insignificant. Phien-wej et al (1990) study on a series of tests on sawtooth shape gypsum samples revealed that, the strength of the joint becomes equal to
that of the infill material, when the t/a ratio approaches two (Figure 5.14).
15
asperity angle

30
A 45

10

en

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Relative thickness, (t/a)

1.8 2.0

Figure 5.13: Shear strength of kaolin infilled joint (after Ladanyi and Archambault,
1977).

Papalingas et al. (1993) reported a detailed shear tests on plaster-cement modelled


joint filled with each of kaolin, marble dust and pulverised fuel ash. He found that,
the shear strength of joints containing kaolin became constant at a t/a ratio of about
0.6, and for marble dust or fuel ash, the values were between 1.25 and 1.50. There
was a marked reduction in shear strength, with the addition of a thin layer of infill
material. As can be seen from Figure 5.15 that, any further increase in the t/a values
beyond 1.5 it would not result in any significant increase in the peak shear strength.

148

Chapter 5: Shear Behaviour of Rock Joints

This is clearly evident by the asymptotic shape of the curves to the t/a axis. De
Toledo and de Freitas (1993) study on ring shear tests on toothed Penrith sandstone
and Gault clay exhibited two different levels of peak shear stress (the soil peak and
the rock peak) along the joint shear stress path. The reduction in soil peak shear
stress was observed up to a t/a ratio of unity, and beyond that, it tapers off and
becomes parallel to the displacement axis. The rock peak shear stress, on the other
hand, was the same, regardless of the level of consolidation pressure applied on the
infill material.

0.0

0.4

0.8
1.2
Relative thickness (t/a)

1.6

2.0

Figure 5.14: Variation of shear strength with t/a ratio (after Phien-Wej et al., 1990).

149

Chapter 5: Shear Behaviour of Rock Joints

100

cs

- e ~a~~
B-B

Vi

u
es

40

Normal stress = 50 kPa

--a-.

V)

es

-g

0)
JS

__

Normal stress = 100 kPa

~-s-

20

'
0.0

-e-

Q-

I l I I I l I i I i
0.2

0.4

0.6

0.8

1.0

-o

I I I I l
1.2

1.4

1.6

1.8

2.0

t/a ratio
Figure 5.15: Effect oft/a ratio on shear strength of infilled joints (after Papalingas et
al., 1993).

The study by D e Toledo and de Freitas (1993) on C N L or zero stiffness condition


indicated that different results would be obtained if the normal stiffness of the
apparatus were changed. In contrast, testing by Cheng et al. (1996) on Johnston paste
(mixture of mudstone powder, cement and water) infilled concrete/rock interfaces
under CNS condition, showed that, unlike in CNL conditions, the presence of a very
thin smear zone (t<l mm) did not significantly reduce the shear strength of joints.
Almost the same strength values were observed for dry cast joints with and without
infill, and up to infill thickness of 2 mm. This conclusion, however, contradicted the
findings from CNL testing reported by others (Goodman, 1970; Lama, 1978;
Papalingas et al., 1993).

150

Chapter 5: Shear Behaviour of Rock Joints

Indraratna et al. (1999) conducted C N S tests on two different asperity angles of 9.5
and 18.5, with joints infilled with soft clays described as Type I and Type II joints,
respectively. The samples were tested under various initial normal stress values,
ranging between 0.3 MPa and 1.10 MPa, respectively, and at a constant normal
stiffness of 8.5 kN/mm. It was found that, the presence of infill in joints contributed
to significant reduction in both the shear strength and the friction angle (Figure 5.16).
The results also showed that, the effect of asperities on the shear strength was
significant, with up to a t/a ratio of 1.4. The infill alone had a controlling influence on
the shear behaviour beyond the stated critical ratios. The drop in peak shear stress
under CNS condition was modelled by a hyperbolic relationship. A Fourier transform
method was also introduced to predict both the dilation and the shear strength of
infilled joints under CNS condition.

To the best of author's knowledge, no suitable literature was available at the time of
writing this thesis to report on the shear behaviour of infilled bolted joints under CNS
condition.

151

Chapter 5: Shear Behaviour of Rock Joints

2-0-i
Type I joint

1-5-

Clean
O f - 15 m m
A f = 2-5 m m
f = 3-5 m m
D f -- 4-5 m m

_
CL

_
_:
ca
o
0.

Decrease in peak friction


angle due to increase in
infill thickness (1-5-4-5 m m )

Normal stress. M P a
(a)

2-5-i
Clean joint

a.

_
_:

Normal stress: M P a
(b)

Figure 5.16: Effect of infill on strength envelope, a) joint with asperity angle 9.5 ,
and b) joint with asperity angle 18.5 (after Indraratna et al., 1999).

152

Chapter 5: Shear Behaviour of Rock Joints

5.5 CONCLUDING REMARKS

It can be inferred, that the critical review reported in this chapter leads to the
following conclusions:

Most unfilled joint research has been conducted under C N L condition.


All the reinforced joint research (reported in literature) was conducted
under CNL condition.
Only a handful of research on both unfilled and infilled joints has been
reported recently under CNS condition.
No research work has been reported on infilled bolted joints.

Having recognised the relevance and the importance of evaluating the shear
behaviour of bolted joints under Constant Normal Stiffness condition; there is a
definite need for understanding the shear behaviour of both bolted and non-bolted
joints. In this thesis, a critical examination of the shear behaviour of unfilled and
infilled joints containing fully grouted bolts is reported, together with a mathematical
model to predict the shear resistance of both unfilled and infilled bolted joints using
Fourier analysis.

153

Chapter 6

SHEAR BEHAVIOUR OF UNFILLED BOLTED JOINTS

6.1 INTRODUCTION

This chapter deals with the shear behaviour of unfilled bolted joints under Co

Normal Stiffness conditions. The laboratory investigation was carried out usin

specially cast saw-toothed blocks bolted with 3 mm diameter rods. The tests we

conducted in a CNS testing machine as described in Chapter 3. The test results


critically analysed and supplemented with a mathematical model later (Chapter

6.2 LABORATORY INVESTIGATION

6.2.1 Selection of Model Material

The idealised joints were prepared from gypsum plaster (CaS04.H20 hemihydrate,

98%). Gypsum was selected for the model material mainly because it was easy to

cast, universally available and inexpensive, as well as widely used by geotech


researchers. Gypsum plaster can be mould into any shape when mixed with water

appropriate proportions, and the long-term strength is independent of time onc

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

hardened and dried. The initial setting time for the gypsum used in the study was in
the order of 30 minutes when mixed with water in the ratio of 2:1. The basic physicomechanical properties of plaster (after a curing period of two weeks in the oven at a
constant temperature of 40 C) were determined by performing several tests on 50
mm diameter specimen. The cured plaster showed a consistent uniaxial compressive
strength (rjc) of about 20 MPa, tensile strength (o\) of about 6.0 MPa and a Young's
modulus (E) of 7.3 GPa. With the ability to alter strength properties by varying the
mixing ratio, gypsum is a suitable material to simulate the behaviour of a number of
jointed or weak soft rocks, such as coal, friable limestone, clay shale and mudstone,
based on the ratios of af<jt and o*c/E relevant in similitude analysis (Indraratna,
1990).

6.2.2 Preparation of Bolted Joints

Fully mated joints of plaster moulds with saw tooth asperity angle of 18.5 were cast
(Figure 6.1) using a mixing ratio of 2:1 by weight of plaster and water. Although
these regular shaped asperities may not ideally represent the complex, irregular and
wavy shaped joint profiles observed in the field, nevertheless, they provide a
simplified basis for evaluating the overall shear strength of bolted joints. The bottom
block (250x75x100 mm) was cast inside the bottom mould containing the required
surface profile and left for two hours to cure. The top specimen (250x75x150 mm)
was then cast on top of the bottom specimen and subsequently, the whole assembly
was left for an additional two hours to allow the top section to cure. Masking tape
was used between two moulds to obtain a fully mated joint surface. During specimen

155

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

preparation, mild vibration was applied to the mould externally to eliminate any
entrapped air inside the sample. The mould was then stripped and allowed to cure in
an oven for 14 days at a constant temperature of 40 C. All the samples were then
allowed to climatise to the room temperature prior to bolt installation. Figure 6.2
shows a typical joint sample.

Figure 6.1: Arrangement for sample casting.

156

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

30 mm

Figure 6.2: A typical joint sample.

Following sample curing, a 5 m m hole was drilled through the middle of the joint
blocks, and a 3 mm <j) threaded steel bolt was installed in the hole using a FosrocChemfix grout (organic resin) commonly used in jointed and stratified roofs of coal
mines in Australia. For the purpose of simplicity, the bolts were installed
perpendicular to the joint plane. This is a common practice in underground coal
mines, where bolts are usually installed perpendicular to the horizontally bedded
mine roofs. A uniform setting time, up to three hours, was allowed for all specimens
before testing. The properties of the hardened resin after three hours were, ac= 76.5
MPa, o-t = 13.5 MPa and E = 11.7 GPa. The change of properties of the grout beyond
three hours was not significant. The uniaxial compressive strength of the model

157

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

material was found to be in the range of 18 to 22 M P a , which is close to the strength


of soft roof rocks found in several underground mine openings.

6.2.3 CNS Direct Shear Testing Apparatus

The detailed description of the CNS shear apparatus has already been discussed in
Chapter 3 (see Section 3.4). The cured specimens were placed inside the top and
bottom shear boxes and screwed tightly to hold in position. Figure 6.3 shows one of
the samples placed inside the bottom shear box. The bottom shear box was then
placed inside the lower specimen holder of CNS machine. The top shear box was
then inserted within the upper specimen holder. Once the shear box containing the
bolted (or non-bolted) joint sample was in place, the top shear box was attached
tightly to the top plate. For bolted joint, however, both shear boxes were screwed
outside and then placed into the CNS specimen holder together. A 10 mm gap was
kept between the top and bottom section of shear box to enable shearing of the bolted
joint plane.

Figure 6.3: A typical joint sample inside the bottom shear box.

158

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

Prior to the commencement of shearing, a predetermined normal load was applied


through a hydraulic jack, operated either manually or by an electric pump. The digital
strain meter fitted to the normal load cell indicated the applied level of initial normal
load. A LVDT mounted on top of the specimen was used to monitor the vertical
displacement of the upper block during shearing.

The shear load was applied to the bottom shear box via a transverse hydraulic jack,
which was connected to a strain-controlled unit. The applied shear load was recorded
via digital strain meters fitted to a calibrated load cell. The rate of horizontal
displacement for the shear boxes could be varied between 0.35 and 1.70 mm/min
using an attached gear mechanism (helical worn transmission system). A sufficiently
low and constant strain rate of 0.5 kN/mm was applied via the bottom shear box,
causing it to move in a horizontal direction, while the movement of the upper shear
box was restricted to the vertical direction only. The slow rate of shearing ensured
fully drained behaviour of saturated infill (described in the following chapter), and
allowed to compare the contribution of bolts in both unfilled and infilled joints. The
dilation and the shear displacement of the joint were recorded by two LVDT's, one
mounted on top of the top shear box and the other attached to the side of the bottom
shear box.

6.2.4 Laboratory Experiments

The shear behaviour of unfilled bolted joint was studied in the laboratory by
conducting a test program on a series of modelled saw-tooth bolted and non-bolted

159

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

joints under Constant Normal Stiffness condition. The laboratory experiments were
carried out in two steps. Firstly, a series of test were conducted on non-bolted joints
at initial normal stress levels ranging between 0.13 MPa and 3.25 MPa, respectively.
Subsequently, another series of bolted joints were tested at the same initial normal
stress condition. The purpose of these two series of tests was to determine the effect
of bolt on the shear strength of joint, and also to determine the strength envelope of
bolted and non-bolted joints. The applied normal stress was limited to a maximum
of 3.25 MPa due to the capacity of the CNS equipment. A constant normal stiffness
of 8.5 kN/mm was simulated via an assembly of four springs mounted on top of the
top shear box as shown in Figure 3.8. A constant strain rate of 0.5 mm /min was
maintained for all the shear tests conducted. At the end of each test, the specimen
was brought back to its initial position by reversing the shear direction. The shear
boxes were dismantled afterwards, and the mode of failure was observed. The final
joint profile was mapped where warranted.

6.2.5 Processing of Test Data

The test results were processed based on the assumption that the normal and shear
load acted uniformly on the whole joint plane. The values of the shear and normal
stresses were calculated using the following equations:

160

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

_____

%h

<6"

~ WxL

nh = ^l (6.2)

Where,
TJ, = shear stress at a shear displacement of h,
ah = normal stress at a shear displacement of h,
Sh = shear load at a shear displacement of h,
Nh = normal load at a shear displacement of h,
W = width of specimen, and
L = length of the specimen.

6.3 SHEAR BEHAVIOUR OF NON-BOLTED JOINTS

6.3.1 Shear Stress

The shear stress profiles for non-bolted joints are plotted in Figure 6.4. The peak
shear strength for joint varies from 1.70 to 4.12 MPa depending on the level of
applied initial normal stress, rjno. As the value of a0 was increased, the peak of shear
stress also increased. The gap between the stress profiles was significant at low cjno
levels, but gradually became marginal as ano was increased, indicating the shearing of
asperities at high normal stress levels. The shear displacement corresponding to peak

161

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

shear stress decreased with the increasing

CT0

values. This is indicated by the inclined

arrow as shown in Figure 6.4. The post peak behaviour of stress path was not
analysed in the present study.

4
6
8
Shear displacement (mm)

10

12

Figure 6.4: Shear stress profile for non-bolted joints.

6.3.2 Dilation

Figure 6.5 shows the dilation behaviour obtained from the laboratory experiments for
non-bolted joints. The initial negative dilation may be attributed to the sample
compaction, closure of pores and the initial settlement of fine irregularities along the
joint plane. The maximum dilation for non-bolted joints ranged between 3.07 mm
and 1.07 mm, corresponding to an initial normal stress of 0.13 MPa and 3.25 MPa,
respectively. As the normal stress increased, the dilation curves become 'dome'
shaped suggesting the increased shearing of asperities at a shear displacement close
to the asperity length. The dilation curves followed a downward trend after attaining
the peak shear strength values.

162

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

3.5
3.0

0.13 MPa

2.5
? 2.0

J 1.5
'&
2
Q 1.0
0.5
0.0
-0.5
Shear displacement (mm)

Figure 6.5: Dilation profile for non-bolted joints.

6.3.3 Normal Stress

The normal stress profiles for the non-bolted joints are plotted in Figure 6.6. As with
the dilation, the normal stress also showed a small initial drop and then began to
increase steadily up to a shear displacement, which was slightly higher than the value
corresponding to the peak shear stress. This steady increase in normal stress shows
the main difference between the Constant Normal Stiffness testing and the
conventional Constant Normal Load testing, where in the later, the stress profiles
remain unchanged (flat straight lines) from the applied initial normal stress levels.
The maximum values of normal stress recorded were 1.47 MPa and 3.94 MPa
corresponding to the initial normal stress levels of 0.13 MPa and 3.25 MPa,
respectively. The slope of the normal stress profiles decreased with increasing initial
normal stress levels, suggesting more extensive shearing of asperities at elevated
normal stress levels.

163

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

CO
Q.

4.0

3.25 MPa

3.5

2.50 MPa

30

1.87 MPa

E 2.5
w
(0
2.0

1.25 MPa
0.63 MPa

(0
(0

F 1.5

0.13 MPa

_i

z 1.0
0.5
0.0
5

10

15

Shear displacement (mm)

Figure 6.6: Normal stress profile for non-bolted joints.

6.4 SHEAR BEHAVIOUR OF BOLTED JOINTS

6.4.1 Shear Stress

Figure 6.7 shows the shear stress profile for bolted joints. A s expected, the peak
shear stress increased with the increase in the level of applied initial normal stress,
o"no- The peak shear strength for bolted joint varied from 2.04 to 4.28 MPa depending
on the magnitude of applied normal stress. The shear displacement at peak shear
stress decreased with the increasing ano. The higher peaks attained by bolted joints
are clearly attributed to the effect of bolting, as the bolted joint offered stiffer
resistance (i.e. increased modulus) as compared to non-bolted joint. The shear
displacement corresponding to the peak shear stress decreased with the increasing

164

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

rjno, however, the rate of decrease w a s slightly diminished w h e n compared with nonbolted joints.

4.5
3.25 MPa
2.50 MPa

4.0
3.5

^^m^^^^H^

1.87 MPa

OL

1.25 MPa

3.0
3 2.5

0.63 MPa

| 2.0

0.13 MPa

2 1.5
w 1.0
0.5
0.0
0

10

12

Shear displacement (mm)

Figure 6.7: Shear stress profile for bolted joints.

6.4.2 Dilation

The dilation profiles for bolted joint are plotted in Figure 6.8. Similar to non-bolted
joints, the dilation profiles were characterised by initial negative values, which began
to increase subsequently reaching m a x i m u m values between 2.77 m m and 0.60 m m ,
corresponding to an initial normal stress level of 0.13 M P a and 3.25 M P a ,
respectively. The dilation profiles for bolted joints appeared to be relatively flat as
compared to the non-bolted joints, which could be attributed to the additional
resistance offered by the bolt. Also, the peak dilation for bolted joints was smaller as
compared to the non bolted joints for all applied normal stress levels However, the
difference between the dilations of bolted and non-bolted joints decreased with

165

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

increasing normal stress indicating the reduced bolt contribution at high normal stress
levels. T h e dilation profiles of bolted joints were relatively flatter as compared to the
non-bolted joints, indicating an increased degree of shearing of asperities in bolted
joints due to increased joint stiffness.

0.13 M P a
0.63 MPa
sJ*" 1.25 MPa
lVx 1.87

10

MPa

12

Shear displacement (mm)

Figure 6.8: Dilation profile for bolted joints.

6.4.3 N o r m a l Stress

Figure 6.9 shows the normal stress profiles for bolted joint tests. T h e peak normal
stress varied between 1.41 M P a and 3.68 M P a corresponding to the initial normal
stress level of 0.13 M P a and 3.25 M P a , respectively. Figure 6.10 shows the
percentage increase in normal stress with respect to the applied initial normal stress,
c no . It can be inferred that the increase in normal stress drops asymptotically with
increasing a no . T h e gap between the plots for bolted and non-bolted joints diminishes
with increasing a no , implying a reduced bolt contribution at higher normal stress

166

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

levels. In general, the normal stress profiles for bolted joints were lower than those
of non-bolted joints because of the resistance offered by the bolt in reducing or
inhibiting joint dilation.

3.25 M P a
2.50 MPa
*********

2
4
6
8
Shear displacement (mm)

10

12

Figure 6.9: Normal stress profile for bolted joints.

1200
(0
(0

- non-bolted
-bolted

o 1000
(0
"
800

_.

o
0)
c

600

o
o
c

200

(0 400
fl.

0.5

1.5

2.5

3.5

Initial normal stress (MPa)

Figure 6.10: Variation of percentage increase in normal stress with initial normal
stress.

167

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

6.5 STRENGTH ENVELOPE

Figure 6.11 shows the strength envelopes for both bolted and non-bolted joints,
which can be considered bi-linear with increasing normal stress. The gap between
the two envelopes decreased marginally with the increasing an0, confirming the
reduced bolt contribution and increased asperity shearing at higher ano. This is also
evident from Figure 6.12 where the difference between the peak shear stresses for
bolted and non-bolted joints are plotted against the respective o _ The slope of the
failure envelope can be assumed to be bi-linear, implying that the starting point of
asperity degradation initiates at a normal stress of around 1.25 MPa.

4.5
4
CL 3.5

3
c
0)
_.
CO
_.
to
0)

to

2.5
2
1.5
1
0.5
+

Initial normal stress (MPa)


Figure 6.11: Strength envelope for bolted and non-bolted joints.

168

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

0.4
^M-*,

CO

n.
5
__
4

0.35

^^-,-v^iiiii^

^^^
\.

0.3 -

^s.

co
_.

0.25

_.
CO

0.2

^\^^
^^\^^
^"""'A

(0

0)

>v

\.
\

__
(0

_c

0.15 -

^|-

c_.

0.1
0.05 -i

()

J
1

1
1

|
1

Initial normal stress (MPa)

Figure 6.12: Variation of difference between peak shear stresses for bolted and nonbolted joint with initial normal stress.

6.6 C O N C L U S I O N S

A series of tests were conducted using the C N S apparatus for both bolted and nonbolted joints at an initial normal stress (rjno) ranging from 0.13 MPa to 3.25 MPa.
While the peak shear stress increased with the increasing level of

CT0,

the shear

displacement at peak shear stress decreased with the increasing rjno for both bolted
and non-bolted joints. The bolt affected an increased peak shear stress and enhanced
stiffness of the bolt-joint composite. The level of bolt contribution appeared to
decrease with the increasing a

169

Chapter 6: Shear Behaviour of Unfilled Bolted Joints

The strength envelopes of both bolted and non-bolted joints showed an approximate
bi-linear trend. The degradation of asperities was initiated at a normal stress of
around 1.25 MPa. The dilation behaviour indicated the compression of fine
irregularities at early stages and a smoothening effect at increased values of rjn0. The
magnitudes of dilation and normal stress profiles were less for bolted joints in
comparison with the non-bolted joints.

The shear tests described here revealed the importance of bolt contribution in
stabilising rock joints under Constant Normal Stiffness conditions. The joint blocks
tested were fully mated and no infill was introduced between them. In reality, the
joints may contain weak or soft clayey infill between the joint planes, which
considerably reduce the overall shear strength of the joint. The role of infill on the
shear behaviour of reinforced joints still needs to be quantified. Therefore, the
following chapter will focus on the shear behaviour of infilled bolted joints.

170

Chapter 7

SHEAR BEHAVIOUR OF INFILLED BOLTED JOINTS

7.1 INTRODUCTION

This chapter deals with the shear behaviour of infilled bolted joints under Constant
Normal Stiffness conditions. The laboratory investigation was carried out using
specially cast saw-toothed blocks containing 0 to 7.5 mm thick infill material and
bolted with 3 mm diameter rods. The tests were conducted in a CNS testing machine
in a similar manner to the testing of unfilled bolted joint. The results of the test were
critically analysed and further supplemented with a mathematical model to predict a
complete stress profile of the reinforced jointed blocks containing infill material,
which is described in the following chapter.

7.2 LABORATORY INVESTIGATION

7.2.1 Sample Preparation

Figure 7.1 shows a typical saw toothed joint prepared from gypsum, with a layer of
clay infill. The procedure for sample casting was the same as that of unfilled joints
described in Section 6.2.2.

Chapter 7: Shear Behaviour of Infilled Bolted Joints

Figure 7.1: A typical saw tooth joint with infill.

Clayey materials have been widely used in the past for laboratory studies, because
clay infilled joints often contribute to significant rock mass instability
(displacements). The ordinary bricklaying clay was used in the current study as it is
widely available and also its composition can be varied depending on the testing
requirement. Placement moisture content of about 27% was ensured for all the tests
by keeping the clay material inside a sealed container. The corresponding shear
strength envelope of clay is shown in Figure 7.2. While placing the clay on top of the

172

Chapter 7: Shear Behaviour of Infilled Bolted Joints

bottom specimen, a s a w tooth shape metal strip w a s used to obtain the uniform
thickness of the infill material all along the surface of the specimen. Figure 7.3 shows
the process of placing a clay seam on the model joint. The thickness of the clay layer
varied between 1.5 mm and 7.5 mm depending on the purpose of testing.

0.035
0.03
. 0.025
I

0.02

0.015

Strength envelope
Moisture content: 2 7 %

V)

__
co

0.01
0.005
0
0.05

0.1

0.15

0.2

0.25

0.3

Normal stress (MRa)

Figure 7.2: Shear strength envelope of clay infill.

Figure 7.3: Process of placing infill on top of joint surface.

173

Chapter 7: Shear Behaviour of Infilled Bolted Joints

Following clay layer placement, the top block was mounted on the bottom section,
thus sandwiching the clay layer in between. The two blocks were then bolted together
by installing a 3 mm bolt perpendicular to the joint plane using Fosroc-Chemfix resin
as used with unfilled joints. Figure 7.4 shows the jointed specimen along with the
threaded bolt (3 mm diameter) ready to be installed in the specimen. The properties
of the hardened resin and the joint material have been described in Section 6.2.2.

Figure 7.4: Installation of bolt perpendicular to the infilled joint.

7.2.2 Testing Procedure

More than forty model joints were tested, and the laboratory experiments were
carried out in two steps. Firstly, the shear tests were conducted on non-bolted joints
and then similar tests were carried out with bolted joints. This testing arrangement
allowed determining the effect of bolting on the shear strength of both unfilled and

174

Chapter 7: Shear Behaviour of Infilled Bolted Joints

infilled joints. The tests were conducted at various initial normal stress levels ranging
between 0.13 MPa and 3.25 MPa, with thickness of the infill clay increased
incrementally up to a maximum of 7.5 mm in five steps, providing an infill thickness

to asperity height ratio (t/a) of 0, 0.3, 0.5, 1 and 1.5, respectively. In this way, the ro
of infill material on the overall shear strength magnitude of both bolted and nonbolted joints could be quantified. A constant normal stiffness of 8.5 kN/mm was
applied via an assembly of four springs on fop of the top shear box. This simulated
stiffness was found to be representative of the coal measure rocks in the Illawarra
region of New South Wales. An appropriate strain rate of 0.5 mm/min was
maintained for all shear tests, which would also ensure a uniform drained condition
of infilled joints. A sufficient gap (less than 10 mm) was allowed between the upper
and lower boxes to enable unconstrained shearing of the infilled joint. A pair of thin
metal plates was used along the sides of the joint plane to prevent any loss of infill
material during testing.

Following testing, the collected data was processed in the same manner as described
in Section 6.2.5. At the end of each test, the specimen was brought back to its initial
position by reversing the shear direction. The shear boxes were dismantled
afterwards, and the mode of failure was observed. The final joint profile was mapped,
where warranted. The asperity profile for each samples were recorded for further
investigation. Figure 7.5 shows one of the tested samples for both bolted and nonbolted joints, respectively.

175

Chapter 7: Shear Behaviour of Infilled Bolted Joints

Figure 7.5: Samples after testing; top: non-bolted, and bottom: bolted.

7.3 S H E A R B E H A V I O U R O F INFILLED N O N - B O L T E D JOINTS

7.3.1 Shear Stress

The shear stress profiles for infilled non-bolted joints, with 1.5 m m (t/a=0.3) infill

and 0.13 MPa initial normal stress, are plotted in Figure 7.6. The peak shear st
for joints varied between 0.14 M P a (corresponding to 7.5 m m infill and ano=0.13

176

Chapter 7: Shear Behaviour of Infilled Bolted Joints

M P a ) and 3.43 M P a (corresponding to n o infill and a no =3.25 M P a , not shown in


figure), respectively, depending on the level of applied initial normal stress, o-no and
the infill thickness, t. The increase in the peak shear stress was affected by ano. The
gap between the stress profiles was generally high at low ano, and gradually becomes
marginal with the increasing __, implying the shearing of asperities at high normal
stress levels. The variation of shear stress between no infill and an infill of 1.5 mm
thickness was very prominent, which was indicated by the large gap between them as
shown in Figure 7.6b. Thus, a thin layer of infill was sufficient to reduce the shear
strength of joints significantly.
1.5 m m infill thickness

a)
1.87 MPa
1.25 MPa

._-"""""- 0.63 MPa


,**- 0.13 MPa

15

10

Shear displacement (mm)

0.13 M P a initial normal stress

b)

1.5 m m

10

15

Shear displacement (mm)

Figure 7.6: Variation of shear stress for infilled non-bolted joint, a) according to
initial normal stress, and b) according to infill thickness.

177

Chapter 7: Shear Behaviour of Infilled Bolted Joints

7.3.2 N o r m a l Stress

The normal stress profiles for infilled non-bolted joints are plotted in Figure 7.7. The
normal stress profiles show a small initial drop, and then the steady increase up to a
shear displacement of slightly higher than at which the peak shear stress occurs. This
steady increase in normal stress was attributed to the impact of the Constant Normal
Stiffness testing, which is discussed in Section 6.3.3. The maximum values of the
normal stress recorded were 0.08 MPa (corresponding to 7.5 mm infill and <Tno=0.13
MPa) and 3.04 MPa (corresponding to no infill and o-no=3.25 MPa, not shown in
figure), respectively, depending on ano and t. The increase in normal stress values
was relatively high at low infill thickness in comparison with thicker infill. In fact,
there were, at times, some small decrease in normal stress was observed, when the
infill thickness was around 7.5 mm as shown in Figure 7.7b. Also, the slope of the
normal stress profiles decreased with the increasing ano, suggesting extensive
shearing of asperities at higher normal stress levels.

178

Chapter 7: Shear Behaviour of Infilled Bolted Joints

1.5 m m infill thickness

a)
3.0
1.87 MPa
1.25 MPa
0.63 MPa

0.13 MPa

10

12

14

16

Shear displacement (mm)

0.13 MPa initial normal stress

b)

1.6
1.4

.*
_**

10

X
_X*

0.8 H

z
0.2
0.0

.-*""

1.5 m m

. . -

to

1 0.6
o 0.4

0 mm

__'

_***

__"

2.5 mn)
5.0 m m I

<*;*)ee<e3__*xxxxxx><xx^^^#<>^>^x-<-

10

7.5 m m
12

14

16

Shear displacement (mm)

Figure 7.7: Variation of normal stress for infilled non-bolted joint, a) according to
initial normal stress, and b) according to infill thickness.

7.3.3 Dilation

Figure 7.8 shows the dilation behaviour obtained from the laboratory experiments for
the infilled non-bolted joints. The initial negative dilation may be attributed possibly
to the compaction of the infill, the closures of pores and the initial settlement of fine
irregularities along the joint plane. The maximum dilation for non-bolted joints

179

Chapter 7: Shear Behaviour of Infilled Bolted Joints

ranged between 3.07 mm (corresponding to no infill and ano=3.25 MPa, not shown in
figure) and -0.44 mm (corresponding to 7.5 mm infill and rjno=0.13 MPa),
respectively, depending on <.. and t. The gap between the dilation curves was
relatively higher at high infill thickness levels (e.g. between 5 mm and 7.5 mm infill)
in comparison with low infill thickness levels (e.g. between 1.5 mm and 2.5 mm
infill). Thus, the impact on joint dilation is highest when the t/a ratio approaches
unity, and was considered as critical t/a ratio by previous researchers as discussed in
Chapter 5. The dilation curves followed a downward trend after attaining the peak
shear strength values.

7.3.4 Stress Path

In order to obtain the stress paths, the shear stress values were plotted against
changing normal stress values. Figure 7.9 shows the stress path for infilled nonbolted joints at

CTO

= 0.13 MPa. The interesting point to note here is, the reversal of

the direction of stress paths with infill thicknesses between 5 and 7.5 mm (t/a ratio of
1 and 1.5, respectively). At the stated infill thickness levels, the shearing of asperities
was almost absent, and the shear behaviour was mainly governed by the infill alone.
The bandwidth (i.e. the extent of variations of shear stress with respect to the normal
stress at any particular t) of the stress paths tends to decrease with increasing level of
infill thickness, indicating the reduced dilation or even compression of joints at those
t/a ratios. This clearly demonstrates the impact of diminishing contacts between
asperities with the increasing t/a ratios.

180

Chapter 7: Shear Behaviour of Infilled Bolted Joints

1.5 m m infill thickness

a)
..*** 0.13 MPa
-^""" 0.63 MPa
1.25 MPa
1.87 MPa

-0.5

10

12

14

16

Shear displacement (mm)


~0.13 M P a initial normal stress

b)

3.5^ 0 m m

3.0 -

ilation (mm)

2.5
2.0

0.5 -

1.5

_*

_.__ 1.5 m m

*"

^x**^ 2.5 m m

u^e^

1.0-

0.0 <
-0.5
-1.0

..^y-y -xx- ' - - ' w " ' * 5.0 m m

IJJJ^AA*^^

^x--x-xxv,,y
2
4 " ' *' x 6'"^ , 3< _ , *^^M_e^^x4_ ; ^xx 14
7.5 m m

16

Shear displacement (mm)

Figure 7.8: Variation of dilation for infilled non-bolted joint, a) according to initial
normal stress, and b) according to infill thickness.

181

Chapter 7: Shear Behaviour of Infilled Bolted Joints

0.13 M P a initial normal stress

1.8 -,

0 mm

1.6
1.4
Q.
(0
(0

1.2 1.0

e
_. 0.8 _
.
(0
CQ
0.6 -

0.4
0.2 -

0.0
0.5

1.0

1.5

Normal stress (MPa)

Figure 7.9: Stress paths for infilled non-bolted joint at an initial normal stress of 0.13

MPa.

7.4 SHEAR BEHAVIOUR OF INFILLED BOLTED JOINTS

7.4.1 Shear Stress

Figure 7.10 shows the shear stress profile for infilled bolted joints. A s expected, the
peak shear stress increased with the increasing ano. The peak shear strength for bolted
joint varies between 0.20 MPa (corresponding to 7.5 mm infill and tTno=0.13 MPa)
and 3.57 MPa (corresponding to no infill and ano=3.25 MPa, not shown in figure),
respectively, depending on orno and t. In general, the shear displacement at peak shear
stress decreased with increasing a no . The higher peaks for bolted joints could be

182

Chapter 7: Shear Behaviour of Infilled Bolted Joints

attributed to the effect of bolting, as the bolted joints offered stiffer resistance (i.e.
increased modulus) as compared to the non-bolted joint, which appeared to reduce
because of the increasing o-no. The variation of shear stress with the infill thickness
(Figure 7.10b) indicated that, at low infill thickness levels, the fall in shear strength
was significant. However, with increased infill thickness, the rate of fall in shear
strength began to decrease indicating the reduced impact of infill beyond a certain
infill thickness level.

1.5 m m infill thickness

a)
.-XX'X-\y-'-Xyy. xxx .

1.87 MPa
l.25 MPa
0.63 MPa

10

Shear displacement (mm)

0.13 M P a initial normal stress

b)
0 mm

Zr<-x<*-^*^^^<'''^^:*'

-'"

10

50mm

15

20

Shear displacement (mm)

Figure 7.10: Variation of shear stress for infilled bolted joint, a) according to initial
normal stress, and b) according to infill thickness.

183

Chapter 7: Shear Behaviour of Infilled Bolted Joints

7.4.2 N o r m a l Stress

Figure 7.11 shows the normal stress profiles for infilled bolted joints from the
laboratory tests. The peak normal stress varied between 0.09 MPa (corresponding to
7.5 mm infill and o-no=0.13 MPa) and 2.87 MPa (corresponding to no infill and
o-no=3.25 MPa, not shown in figure). Normally, the normal stress was found to
increase, up to a peak level, with increasing shear displacement. However, there was
one exception, in which the normal stress was found to decrease when both CT0 and t
was high. Such behaviour might have beeen caused by excessive compaction of
thicker infills, instead of the usual joint dilation, as observed in other situations.

7.4.3 Dilation

The dilation profiles for infilled bolted joints are plotted in Figure 7.12. Similar to
non-bolted joints, the dilation profiles were marked with the initial negative values,
and then began to increase, reaching the maximum values between 2.97 mm
(corresponding to no infill and an0=3.25 MPa, not shown in figure) and -0.27 mm
(corresponding to 7.5 mm infill and orno=0.13 MPa), respectively, depending on an0
and t. The dilation profiles for the bolted joints appears to be relatively flatter as
compared to the non-bolted joint, which could be attributed to the additional
resistance offered by a bolt. Also, the absolute value of peak dilation for the bolted
joints was less as compared to the non-bolted joints for all levels of ano, except when
the dilation is negative. Thus, the bolt acts to prevent either dilation or compression

184

Chapter 7: Shear Behaviour of Infilled Bolted Joints

due to its axial stiffness. T h e critical t/a ratio appeared to be similar as of non-bolted
joints as shown in Figure 7.12.

1.5 m m infill thickness

a)

3.0
_X-X\\-.

Id
0.

/^

x*x'

2.0
0)

&_fi__rfWkA&i! j 2 5 M P a

...""""" 0.63 MPa

1.5

_. 10

o
z

1.87 M P a

x**

in
in

_.
in

\N,

,-x/"v

2.5

,-.-^
*** 0.13 M P a

0.5
<.+****

*********
+*+++**********

0.0
4

10

12

16

14

Shear displacement (mm)

0.13 M P a initial normal stress

U\

1.6 _ 0 m m

1.4 -

***-

oS 1.2 .f 1 0 -

0.8-

_#
. 1.5 m m
0.6
__**^
___"""
2.5 m m
o 0.4
z
0.2 -_^4**_^UB^J,*BP?^?!,'^"^ <'' -^>^"-"->-^J.-N- ^ ^<-- - -xxx*- - 5.0 m m
l
!
;
75 m m
0.0 ] * " * ^ w * w s w S i ^ ^ ^
CO

XLeSSSZ~~~~~~~
i

C)

10

12

14

16

Shear displacement (mm)

Figure 7.11: Variation of normal stress for infilled bolted joint, a) according to initial
normal stress, and b) according to infill thickness.

185

Chapter 7: Shear Behaviour of Infilled Bolted Joints

1.5 m m infill thickness

10

12

16

Shear displacement (mm)

b)1

0.13 M P a initial normal stress

Omm
1.5 m m

2.5 m m

7.5 m m
Shear displacement (mm)

Figure 7.12: Variation of dilation for infilled bolted joint, a) according to initial
normal stress, and b) according to infill thickness.

7.4.4 Stress Path

The stress paths for the infilled bolted joints, at the initial normal stress of 0.13 M P a ,
are plotted in Figure 7.13. The bandwidth of individual stress paths appeared to be
narrow as compared to the non-bolted joints, which was discussed in Section 7.3.4.
Clearly, the presence of the bolt across the joint plane offered a significant resistance
against shearing and dilation as compared to the non-bolted joints.

186

Chapter 7: Shear Behaviour of Infilled Bolted Joints

0.13 M P a initial normal stress


2.0

0 mm

1.8
1.6

"co"

Shea r stress (M

Q.

1.4
1.2
1.0

7.5 m m

0.8

J*

^r**^

0.6
0.4 ,
0.2 .

^_. 1.5 m m

^d/B5w2.5 mm
5.0 m m

0.0

00

0.5

1.0

1.5

Normal stress (MPa)

Figure 7.13: Stress paths for infilled bolted joint at an initial normal stress of 0.13

MPa.

7.5 COMPARISON OF T H E SHEAR BEHAVIOUR OF INFILLED BOLTED


A N D NON-BOLTED JOINTS

7.5.1 Overall Shear Behaviour

Figure 7.14 shows the shear stress profile of both the bolted and non-bolted joints for
the various levels of initial normal stress (o-o) applications at 2.5 mm infill thickness.
The bolted joints show increased stiffness and greater peak shear stress at all levels of
initial normal stress. At low initial normal stress, the percentage increase in the peak
shear stress is significant (about 30% at <jn<0.63 MPa) in comparison with that at
high normal stress (about 12% at orn=1.87 MPa). This implies that the effect of the

187

Chapter 7: Shear Behaviour of Infilled Bolted Joints

bolt o n increasing the shear resistance is marginally reduced as the normal stress
increased. The increased stiffness of the bolted joint is clearly attributed to the added
stiffness of the grouted bolt.

Figure 7.14: Shear stress variation with shear displacement at 2.5 m m infill, for
bolted and non-bolted joints.

Figure 7.15 illustrates non-linear strength envelopes plotted for both bolted and nonbolted joints. It is clear that at low normal stress levels (say less than 0.63 MPa), the
apparent friction angle, expressed as the ratio of peak shear stress to normal stress,
has increased significantly due to the presence of the bolt. Also, the apparent
cohesion appears to be increased slightly. At higher normal stress (an>1.25 MPa), the
apparent friction angle for the bolted and non-bolted joints is similar (i.e. the slope of
two curves in Figure 7.15), indicating the diminished influence of the bolt at high
normal stress levels. Figure 7.16 shows the corresponding dilation profiles of both
bolted and non-bolted joints for various levels of normal stress. As expected, the
dilation of the bolted joint is less than that of the non-bolted joint, thus verifying the

188

Chapter 7: Shear Behaviour of Infilled Bolted Joints

favourable effect of grouted bolts. T h e slope of the dilation curves for bolted joints
is smaller in comparison with non-bolted joints, suggesting a greater degree of
asperity degradation probably occurring in bolted joints. This aspect is elaborated
later in this chapter.

with bolts

0.5

1.5

Normal stress (MPa)

Figure 7.15: Shear strength envelopes of bolted and non-bolted joints at 2.5 m m
infill.

bolted

non-bolted
no
0.13 MPa

0.13 M P a

0.63 MPa

~ 0.63 M P a

1.25 MPa
1.87 MPa

il 1.25 MPa
| 1.87 MPa
_3

_5

-0.5

Shear displacement (mm)

Figure 7.16: Dilation profiles for bolted and non-bolted joints at 2.5 m m infill.

189

C h a p t e r 7: Shear Behaviour of Infilled Bolted Joints

7.5.2 Effect of Infill on Shear Strength

Figure 7.17 shows the percentage drop in peak shear strength of bolted and nonbolted joints with infill thickness, in relation to the corresponding shear strength of
the unfilled non-bolted joints. A significant drop in peak shear strength was observed
even with small infill thickness of 1.5 mm (t/a=0.3), for both bolted and non-bolted
joints.

t/a ratio
0.2

0.4

0.6

0.8

1.2

1.4

1.6

20
_:
1
c

G n o = 1-87 M P a

non-bolted
bolted

" 40

a.
o

60

80

100

rj =0.13 M P a
w
no

Figure 7.17: Variation of % drop in peak shear strength with infill thickness.

Based on the laboratory observations, the following hold for all t/a ratios:

a) Non-bolted joints show a greater drop in shear strength,


b) The greater the t/a ratio, the smaller will be the bolt contribution, and
c) The greater the initial normal stress, the smaller is the % drop in shear
strength.

190

Chapter 7: Shear Behaviour of Infilled Bolted Joints

Moreover, the bolt contribution seems to decrease with the increasing normal stress
for all t/a ratios (i.e. for an0=1.87 MPa, the gap between the dotted and solid line is
smaller than the corresponding gap at ano=0.13 MPa).

Figure 7.18 illustrates the shear stress profile of bolted and non-bolted joints at an
initial normal stress of 0.63 MPa. The difference between shear stress profiles for no
infill and 1.5 mm thick infill (t/a=0.3) is large for both bolted and non-bolted joints.
The corresponding difference of curves between t=1.5 and t=2.5 is considerably
smaller. This demonstrates that a small infill thickness is sufficient to reduce the
shear strength of joint considerably, irrespective of whether the joint is bolted or not.
It is observed that the difference in peak shear strength between bolted and nonbolted joints tapers off as the infill thickness is increased from 0 to 7.5 mm. On the
basis of the data plotted in Figures 7.17 and 7.18, it is clear that the contribution of
the bolt in terms of increasing the joint shear strength diminishes quickly with the
increasing infill thickness.

infill thickness (t)

bolted

non-bolted

0 mm

infill thickness (t)

IB

0.

0 mm

5
M
(A

-g

1.5 m m

1.5 mm

2.5 m m i^ sfcl ^&^,_.

<xx>v ^^..^'^iiijv^.. 2.5 mm

_:
)7.5 5m mm m::
/t&wM
rta.ut!_5___5_Kix^H.ta5__i-,^^

>;sHxt;:a:s___^u.i0jj_a^_;i;__3___9_ 7.5 mm

15

15

Shear displacement (mm)

Figure 7.18: Variation of shear stress with shear displacement for bolted and nonbolted joints at cno = 0-63 MPa.
191

Chapter 7: Shear Behaviour of Infilled Bolted Joints

7.5.3 Profile of Shear Plane

Figure 7.19 s h o w the profiles of shear planes for selected bolted joints at an initial
normal stress of 1.25 MPa. The shear planes were estimated from the measured
dilation corresponding to the horizontal displacement, and they are shown by dashed
lines marked on the figure. The shear plane passes through both infill and asperity
when the t/a ratio is 0 to 0.5, and slightly below the crown of the asperity when the

t/a ratio is unity (Figure 7.19c). For t/a ratio greater than unity (e.g. t/a=l .5), the sh
plane passes entirely through the infill (Figure 7.19d), in which case the shear
behaviour is governed by the infill alone. Therefore, for infill joints tested in this
study, a t/a ratio of 1 to 1.5 can be considered to be critical.
15

15-

b)
t/a = 0.5

10
shear plane

10

15

10

15

Asperity base length ( m m )

Asperity base length ( m m )


15^

15 T

(5.0)

10

15

Asperity base length ( m m )

10

15

Asperity base length ( m m )

Figure 7.19: Relative location of shear plane through infilled joints at a n o - 1-25
MPa.

192

Chapter 7: Shear Behaviour of Infilled Bolted Joints

The values (1.9, 3.8, 4.8 and 6.7) shown on therighthand side of each diagram in
Figure 7.19 indicate the elevation of the shear plane for bolted joints measured in
mm. The values shown within brackets (2.2, 4.0, 5.0, and 6.8) correspond to the

elevation of the shear plane for non-bolted joints, at the same t/a ratios. For all t/

ratios, the shear plane for bolted joints always passes along a slightly lower elevat

as compared to non-bolted joints, indicating either the reduced dilation or increased


compression attributed to bolting. The maximum dilation or compression of the joint

can be determined by subtracting the infill thickness from the elevation of the shear
plane. Table 7.1 summarises the maximum dilation or compression for each t/a ratio,
where the dilation is marked as positive and compression is marked as negative.

Table 7.1: Dilation or compression of joints for various t/a ratios


Joint type

Infill
thickness

t/a

Elevation of
shear plane

Dilation or
compression

(mm)

(mm)

Comments

(mm)
0

2.2

2.2-0 = 2.2

Bolt decreases
dilation

1.9

1.9-0=1.9

@14%.

Non-bolted

2.5

0.5

4.0

4.0-2.5 = 1.5

Bolt decreases
dilation

Bolted

2.5

0.5

3.8

3.8-2.5 = 1.3

@13%.

Non-bolted

5.0

1.0

5.0

5.0-5.0 = 0

Bolted

5.0

1.0

4.8

4.8-5.0 = -0.2

Slight compression of
bolted joint.

Non-bolted

7.5

1.5

6.8

6.8-7.5 = -0.7

Bolted

7.5

1.5

6.7

6.7-7.5 = -0.8

Non-bolted
Bolted

193

Bolted joint
undergoes
greater
compression.

Chapter 7: Shear Behaviour of Infilled Bolted Joints

From Table 7.1, it is clear that as the infill thickness is increased, the joints undergo
compression rather than dilation. This indicates that the influence of asperities
decreases as t/a ratio increases. When t/a<l, the infilled joints show dilation, whereas
for t/a>l, the same undergo compression. Even at high t/a ratios, the bolt
demonstrates the effect of preventing dilation (increasing compression). Therefore, it
can be inferred that, despite the reduced bolt effectiveness at increased t/a ratios with
respect to shear strength (Figure 7.17), nevertheless the bolt maintains the ability to
reduce dilation and/or increase compression of infilled joints.

7.5.4 Strength Envelopes

Figures 7.20 and 7.21 show the variation of the shear stress with the normal stress
along with the corresponding strength envelopes for bolted and non-bolted joints, for
2.5 mm and 7.5 mm infill thickness respectively. At low infill thickness (2.5 mm or
t/a ratio = 0.5), there exist a clear difference in strength envelopes between bolted and
non-bolted joints and, as expected, the apparent friction angle or shearing resistance
(i.e. shear stress to normal stress ratio) is greater for bolted joints as shown in Figure
7.20. As the infill thickness was increased to 7.5 mm or t/a=1.5 (Figure 7.21), the
difference between strength envelopes of bolted and non-bolted joints appears to
have reduced, verifying the earlier findings that, at high t/a ratios, the shear behaviour
is governed by the infill, with little or no contribution from the bolts. At low infill
thickness, the linearised CNL strength envelope (shown by dashed lines) lies
significantly above the CNS envelope (Figure 7.20) and, as the infill thickness is
increased (t/a>1.0), the CNL and CNS strength envelopes for both bolted and non-

194

Chapter 7: Shear Behaviour of Infilled Bolted Joints

bolted joints tend to become close to each other as shown in Figure 7.21. This
suggests that the concept of Constant Normal Stiffness is more relevant for rough
joints with little or no infill, whereas for clayey material (soils in general), the shear
behaviour can be represented sufficiently by the conventional shear strength envelope
(Constant Normal Load) alone.

C N S envelope

bolted

42o

non-bolted
C N L enve lopex
1.6

C N S envelope

cs
Q.

&0.-\3IJPa

5
_
n
2

<

0.63 MPa

1.25MP-

\- 1.87 MRa

co
1

Normal stress (Mpa)

Figure 7.20: Stress paths and strength envelopes for 2.5 m m infill thickness (t/a=0.5)
for bolted and non-bolted joints.

Figure 7.21: Stress paths and strength envelopes for 7.5 m m infill thickness (t/a=1.5)
for bolted and non-bolted joints.

195

Chapter 7: Shear Behaviour of Infilled Bolted Joints

7.6 CONCLUSIONS

It can be inferred from this study that, there occurs a significant drop in peak shear
strength with relatively thin infill thickness of 1.5 mm (t/a = 0.3), for both bolted and
non-bolted joints. The bolt contributed to the increased peak shear stress and the
stiffness of the bolt-joint composite. The overall shear strength of bolted joint was
increased up to 30% at low normal stress levels, but was less than 12% at high
normal stress levels, thus indicating the reduced bolt effectiveness at high normal
stress levels. The CNS strength envelope of infilled joints was found to be nonlinear, and in bolted joints the apparent friction angle was found to increase
considerably at low normal stress. However, the effect of bolting was reduced with
the increased normal stress. The increase in joint shear strength due to bolting was
also influenced by the infill thickness. The greater the t/a ratio, the smaller the role of
bolting is in influencing the joint shear strength. Nevertheless, the presence of bolt
contributed to minimising the joint dilation. On the basis of the current study, bolts
reduce dilation of joints when t/a < 1, and increase joint compression for t/a > 1.

Shear plane passes through the infill or asperity or both, depending on the infill
thickness to asperity height ratio, t/a. Beyond a critical t/a ratio of 1.0, the effect of
bolting or the influence of asperities becomes marginal, and the joint behaviour is
dictated mainly by the infill characteristics. At low infill thickness, the shear
behaviour of bolted joints under CNS condition is different from its behaviour under
conventional CNL condition. At t/a ratios exceeding unity, the difference between the
CNS and CNL envelopes is insignificant, implying that the conventional shear

196

Chapter 7: Shear Behaviour of Infilled Bolted Joints

strength envelope ( C N L ) can be adequate for representing the behaviour clayey infill.
Nevertheless, for rough joints with little or no infill, the CNS behaviour is more
realistic in practice, especially for underground mining conditions in bedded or
jointed rock.

197

Chapter 8

MODELLING OF THE SHEAR BEHAVIOUR OF


REINFORCED JOINTS

8.1 INTRODUCTION

In Chapter 5, various models for predicting the shear strength of both bolted and nonbolted joints were reviewed. Most of these models, based on experimental, analytical
or numerical techniques have been developed to function under CNL conditions,
typically applicable to shearing across planar joints. These models may not
necessarily provide adequate information when applied to non-planar joints as
discussed in Chapter 2. Accordingly, in this chapter the formulation of an analytical
model based on the Fourier analysis has been described, along with the development
of a numerical model using Universal Distinct Element Code (UDEC) to predict the
shear behaviour of such joints under CNS conditions.

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

8.2 ANALYTICAL MODEL FOR SHEAR BEHAVIOUR OF UNFILLED


REINFORCED JOINTS

8.2.1 Definition of Fourier Series and its Applicability in Joint Modelling

A function f(x) is said to be periodic with a period 2T, if for all x, f(x + 2T) = f(x

where T is a positive constant (see Figure 8.1). For example, sin (x + 27t) is a perio
function with a period of 27t. The Fourier expansion corresponding to f(x) can be
defined as (Spigel, 1974):

/w=v ^K
2

cos

ruix , . nnx^

~v~+^sin r i

rrv

(8.1)

T J

where, the Fourier coefficients a and b n are given by:

1 f

U7DC
X C0S

"n=-lTf( )

1 f

dX

^f-

(8-2)

Y17JX

bn=-lTf(x)smrdx

(8.3)

199

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

2T

Figure 8.1: Periodic function f(x) with a period of 2T.

Fourier series have been used as a good mathematical tool for defining various
continuous functions of any kind. The application of Fourier series in the field of
rock mechanics has lately been introduced by Indraratna et al. (1999). When applied
to rock joints, Fourier series can be used to model the joint dilation profile of both
unfilled and infilled reinforced joint with reasonable accuracy. Accordingly:

Sv{h) = v + 2_K cos^^


=i
2 t_TV
T

+ b sm

n -rjr- .
T J

(8.4)

If the total shear displacement is divided into m equal parts as shown in Figure 8.2,
then the Fourier coefficients a and b n can be rewritten as:

200

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

2 m~l C 0 S

n~ _ E ^

m-\

2k7Tn
(8.5)

. 2/br

Z^sin m

mTZ
A=0

(8.6)

where,
5v(h) = dilation at any shear displacement of h,
an, bn = Fourier coefficients, and
T = period of the function 8v(h).

Total shear displacement is


divided into m equal parts

> h

Figure 8.2: Dilation profile segmented into m equal parts.

201

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

8.2.2 Prediction of Variation in N o r m a l Stress

O n c e the dilation of the bolted joint is calculated for a particular shear displacement,
the effective normal stress on the joint plane could be calculated using the following
equation:

n(h) = CTno+

(8-7)

where,
an(h) = effective normal stress at any shear displacement h,
CTno

initial normal stress,

A = area of joint plane; and

_ . =

(8.8)

+
kj

kb

where,
kn = bolt-joint normal stiffness,
kj=joint normal stiffness, and
kb = bolt normal stiffness.

202

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

8.2.3 Prediction of Shear Stress

As indicated by Johnstone and Lam (1989), the joint asperity angle does not remain
constant over the entire length of shear displacement, and a similar phenomenon can
also be observed in the case of bolted joints. As the bolted joint is sheared, the
asperities start to wear depending on the degree of applied initial normal stress on the
joint plane, shear displacement and the characteristics of bolt/grout interface. Thus
the shear stress of a bolt-joint composite, Th, at any point could be given by:

T_i

Shear stress component


due to shear stiffness of
the bolt

Shear stress component


due to effective normal
stiffness of the bolt and
joint

(8.9)

If the bolt is installed at an angle 9 with respect to the joint plane as shown in Figure
8.3, the shear stress component due to bolt shear stiffness is given by:

r*W =

k.Ji
(8.10)

AJsmO
b

where,
Tbs(h) = shear stress component due to bolt shear stiffness at a shear
displacement of h,
ks = bolt shear stiffness,
A b = cross sectional area of the bolt, and

203

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

0 - angle of inclination of bolt with respect to the joint plane.

ss/s"'"'"

(T = f(kn,8y)
kn = f(kr,kb)
kr & kb=rock & boh stiffness
5 V = joint dilation

T
\

joint plane

i f bolt

_L
t

Figure 8.3: Orientation of bolt with respect to joint plane.

For calculating the shear stress component due to normal stiffness of the bolt-joint
composite, the shear strength model (based on the energy principle) proposed by
Sidel and Haberfield (1995) for non-bolted joint was used as the initial basis of the
current model. In a simple form, their model can be written as:

K}

nK ;

l-tan(^)tan(/J

(8.H)

where,
t(h)

= shear stress ofjoint at a shear displacement of h,

on(h) = effective normal stress on the joint plane at a shear displacement of

<t>b

angle of internal friction of joint material,


angle of asperity at zero shear displacement, and

204

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

in = asperity angle at an shear displacement of h.

By introducing the bolt component into the above joint model, the overall shear
strength of a bolt-joint composite can be represented by:

AJ sinO

bnK

tan(^) + tan(i)
_l-tan(^)tan(/J

(8.12)

where,
.(h)

- Shear stress of bolt-joint composite at a shear displacement of h,

and
o"bn(h) = Effective normal stress on the joint plane of bolt-joint composite at
a shear displacement of h and can be calculated by Equation 8.7.

The peak shear stress, xp could be obtained by differentiating Equation 8.12 and then
equalling it to zero. The simple solution for the above equation could not be
achieved due to the complexity involved in handling the Fourier terms using ordinary
calculus. The mathematical package, MATHEMATICA was subsequently used to
obtain the final solution. Appendix I provides the detailed steps involved in solving
the above equation. In a simplified form, the peak shear strength of a bolted joint, xp,
is given by:

k..h
S
T
A / sin 6

+ < no +

0
a

v2

2nh \
+ a, cos
!

205

tan(<f>b) + tan(i)
\-tan((j)b)tan(i^ )

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

(8-13)
where,
hTp = shear displacement at peak shear, and
ihxp = effective asperity angle at a shear displacement of h^.

8.3 ANALYTICAL MODEL FOR SHEAR BEHAVIOUR OF INFILLED


REINFORCED JOINTS

The development of analytical model for predicting the shear strength of i


bolted joints was based on the model developed for unfilled joint and the
strength drop. The model consisted of two parts: first, the shear strength of bolted
joints without any infill was calculated (using the model as per Equation 8.13), and
then the drop in shear strength due to infill was calculated using a hyperbolic model.
In the final step, the shear strength of infilled bolted joint was calculated by
deducting the strength drop due to infill from the unfilled joint strength.

8.3.1 Modelling of Drop in Shear Strength

8.3.1.1 Impact of infill thickness on peak shear stress

Figure 8.4 shows the peak shear stress values for bolted joints, obtained from
laboratory testing and plotted against the respective t/a ratios. Clearly, there is a sharp
drop in strength values from no fill to infill thickness of 1.5 mm (i.e. t/a=0.3). This

206

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

drop, however, tapers off asymptotically w h e n the infill thickness is increased


beyond 1.5 mm (t/a=0.3). The drop in shear strength is significantly influenced by the
t/a ratio while its value is less than unity. As soon as the value oft/a ratio exceeds
unity, the shear strength curves become almost parallel to the x-axis.

0.2

0.4

0.6

0.8

1.2

1.4

1.6!

t/a ratio

Figure 8.4: Variation of peak shear stress with t/a ratio.

8.3.1.2 Relationship between normalised strength drop and t/a ratio

Figure 8.5 shows the Normalized Strength Drop ( N S D ) , defined as the ratio of drop
in peak shear strength divided by the initial normal stress, against t/a ratio for bolted
joints. In order to quantify the impact of infill on the NSD of bolted joints, the shear
strength of the bolted joint with no infill was taken as the reference point for
calculating NSD. As can be seen from Figure 8.5, the NSD increases sharply with
infill thickness as low as 1.5 mm (t/a = 0.3), indicating a sudden drop in shear
strength. Further drop in NSD becomes insignificant at t/a ratios exceeding unity or

207

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

at the critical t/a value as indicated in Figure 8.5. At low normal stress values, the
NSD is more sensitive to the changing t/a ratio. As the normal stress increases
(say 1.25 MPa), NSD curves become almost parallel to the x-axis indicating reduced
impact of normal stress on the overall drop in shear strength.

t/a ratio

o.o

Figure 8.5: Variation of Normalised Strength Drop ( N S D ) with t/a ratio.

8.3.1.3 Modelling of normalised strength drop

A s indicated by Duncan & Chang (1970), Indraratna et al. (1999) and others, N S D is
usually modelled using the hyperbolic relationship given by Equation 8.14:

tla
P + a(tla)

(8.14)

= /3 + a(tla)

(8.15)

NSD =
or,

tla

NSD

208

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

where,
ct,p = Constants, values being dependent on the effective normal stress on the
joint plane and the surface roughness of joint plane.

Equation 8.15 is a straight line representing the x-axis as t/a ratios and the y-axis as
(t/a)/NSD respectively. Figures 8.6a and 8.6b show the above linear relationship for
the results obtained from laboratory experiments for both bolted and non-bolted
joints. The slope and the intercept of these lines are also tabulated in Figure 8.6. The
shear strength of bolted and non-bolted joints at no infill was taken as the respective
datum levels. The correlation coefficients for the above equation for the laboratory
data set were found to vary between 0.990 and 0.999 for both bolted and non-bolted
joints. Therefore, Equation 8.14 can be assumed to predict NSD very accurately for
the experimental data for both bolted and non-bolted joints. Thus, using Equation
8.14, the strength drop due to infill can be calculated for both type of joint with
reasonable accuracy.

209

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

a)

non-bolted

Va ratio
oo 10 0
20 ,
30
40 H

0.13 MRS

so -;

0.63 rVF_

60 4

A 1.25 MPa

70 \
80 \
90 i
00

. 1.87 MPa

cr(Mpa)

0.13

-.011

-.076

0.63

-.117

-.280

1.25
1.87

-.161
-.156

-.396
-.555

b ) bolted
t/a ratio
0.00

1.6

-0.10
-0.20
-0.30

O -0.40
CO
? -0.50
co
5 . -0.60
-0.70
-0.80

0.13 MP_
0.63 MPa
A 1.25 MPa'
x 1.87 MPa

o-n(Mpa)

0.13

-.012

-.072

0.63

-.106

-.235

1.25

-.178

-.366

1.87

-.189

-.513

-0.90
-1.00

Figure 8.6: Straight line formulation of N S D ; a) for non-bolted joint, and b) for
bolted joint.

210

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

8.3.2 Modelling of Shear Strength of Infilled Bolted Joints

Once the strength drop is calculated using the method described above, the shear
strength of infilled bolted joint can be calculated by:

hi

=(0
i. ,

P /infilled bolted

-AT

V P /unfilled bolted

(8.16)

V P A
where,
TP = peak shear strength of bolted joint, and
ATP = drop in shear strength due to infill.

Now, putting the value of peak shear strength of uniflled bolted joint from Equation
8.13 in the above equation, we get:

k..h
S
-

/'''infilled baited

AJsinO

+ <no +

ar

27ih,M
+ ax cos

f
tan((j)b) + tan(i)
\-tan((/)b)tan(ih )

t/a
-a

no

V a
where,
[T ).

= peak shear strength of infilled bolted joint.

211

(8.17)

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

Equations 8.13 and 8.17 were used in a computer program to calculate the shear
stress values at various shear displacements and at different initial normal stress
conditions for unfilled and infilled bolted joints, respectively.

8.4 COMPUTER PROGRAM FOR CALCULATING FOURIER


COEFFICIENTS AND THE SHEAR STRENGTH

For faster processing, an in-house computer program was written, in Quick Basic, to
calculate the Fourier coefficients and the corresponding shear strength of reinforced
joints. T h e computer code for this program is given in Appendix II. The input data
for the program include various physical parameters relating to joint such as the joint
basic friction angle, shear displacement intervals, stiffness of the system, initial
normal stress etc. T h e outputs from the program include the Fourier coefficients,
peak shear stress, and the values of shear stress, normal stress and dilation at various
shear displacements. Using the output, the complete stress profiles for joints can be
plotted and compared with the laboratory results. A typical sample input and output
from the program is also shown in Appendix II.

8.5 VERIFICATION OF ANALYTICAL MODEL

Using the above mention computer program, the Fourier coefficients were calculated
for various values of infill thickness and at different initial normal stress levels.
Figure 8.7 shows the shear stress profile calculated from the model (as per Equation

212

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

8.12) and those obtained from the laboratory experiments for unfilled joints. A s can
be seen from Figure 8.7, the shear stress profile predicted by the Equation 8.12 was
in close agreement with the laboratory results. The predictions for non-bolted joints
(using the model of Indraratna et al., 1999) were closer as compared to bolted joints
(using the current model presented in this thesis). This is probably due to the factor
that, the shear stiffness of the bolt was assumed to be constant even at the plastic
deformation stage in the model, which simplifies the actual deformation behaviour of
the bolt. Figure 8.8 shows the dilation behaviour obtained from laboratory
experiments and the model predictions (Equation 8.4), both for bolted and non-bolted
joints without any infill. As can be seen, the dilation predictions were very close to
the actual values obtained from the laboratory experiments.

Figure 8.7: Comparison of shear stress profiles predicted by the model and from
laboratory tests for unfilled joints.

213

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

Figure 8.9 shows the peak shear stress of infilled bolted joints at all infill thickness
levels obtained from the laboratory results and model predictions. It can be seen from
the figure that by utilising the strength drop criterion, the predicted shear stress
values were in close agreement with the laboratory results. The normal stress and
dilation predicted from the model for 1.5 mm infill thickness along with respective
laboratory results are shown in Figures 8.10 and 8.11, respectively. As expected, the
model predicted the dilations values with greater accuracy as compared to shear and
normal stress values. Given the recognised constraints or limitations, the model was
able to predict, with reasonable accuracy, the stress and dilation parameters defined
in the above Equations 8.12, 8.13 and 8.17.

Experiment

3.5

Model predictions

3.0 -

0.13 MPa u.
%,
0.63 MPa _\.

bolted

2.5 -

N \
NX

15

10

.. 0.63 MPa

^ ^

1.0 -

Jt> 1.87 MPa

1.5

1
.
8
7 MPa / - ^ S A \

3.25 MPa s^~

non-bolted

0.13 MPa

2.0 -

^N_ \

E,
o

_. _/

/ -^*iz 3.25 MPa

L 0.5 -

^^O^,
1

5
()
-0.5 J
Shear displacement (mm)
5

10

15;

Figure 8.8: Comparison of dilation profiles predicted by the model and from
laboratory tests for unfilled joints.

214

apter 8: Modelling of the Shear Behaviour of Reinforced Joints

4.0

Experiment

0 mm

3.5 -i

Model predictions
3.0 i
2.5
2.0
1.5
1.0
0.5
0.0 4

05

1.5

Initial normal stress (MPa)

Figure 8.9: Comparison of peak shear stress predicted by the model and from
laboratory tests for infilled bolted joints.

Model predictions
Laboratory

1.25 M P a
0.63 MPa

0.13 M P a

-1

11

13

15

Shear displacement (mm)

Figure 8.10: Comparison of normal stress profiles predicted by the model and from
laboratory tests for infilled bolted joints (t=l .5 m m ) .

215

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

2.5
0.13 M P a

Model predictions

20
-0.5
Shear displacement (mm)

Figure 8.11: Comparison of dilation profiles predicted by the model and from
laboratory tests for infilled bolted joints (t=1.5 m m ) .

8.6 NUMERICAL M O D E L OF REINFORCED JOINTS UNDER CNS


CONDITIONS

Shear behaviour of reinforced joints have been studied in the past using numerical
tools by various researchers (Dar and Smelser, 1990; Weishen et al, 1990; Ferrero,
1995; Marence and Swoboda, 1995; Benmokerane et al., 1996; Pellet and Egger,
1996; Kinashi et al., 1997; Windsor, 1997; and Kharchafi et al., 1998). The main
advantage of numerical methods is that the model can be built or modified easily and
quickly without conducting further laboratory experiments. So far, most models were
developed to suit the conventional Constant Normal Load conditions, and therefore,
not suitable to be used in the present study. Accordingly, a new numerical model was
developed to take into consideration for the study of shear tests conducted under

216

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

Constant Normal Stiffness conditions. The new model uses UDEC version 3.1 to
study the shear behaviour of reinforced joints. The following write-up describes the
details of this numerical model.

8.6.1 General Background

Ground reinforcement consists of bolts installed in holes drilled in the rock mass.
Two types of reinforcement models are provided in UDEC: local and global
reinforcement. A local reinforcement model considers only the local effect of
reinforcement, where it passes through existing discontinuities. A global
reinforcement model, on the other hand, considers the presence of the reinforcement
along its entire length throughout the rock mass. The global reinforcement model was
used in the present model formulation.

8.6.2 Global Reinforcement

In assessing the support provided by the rock reinforcement, it is often necessary to


consider not only the local restraint provided by reinforcement, where it crosses
discontinuities, but also the restraint to intact rock that may experience inelastic
deformation in the failed region surrounding an excavation. Such situations arise in
modelling inelastic deformations associated with failed rock and/or reinforcement
system (e.g. bolts) in which the bonding (grout) may fail in shear over a given length
of reinforcement.

217

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

Bolt elements in U D E C allow the modelling of shearing resistance along their length,
as represented by the frictional resistance (bond) between the grout and the bolt
surface or the host medium. The bolt is assumed to be divided into a number of
segments of length, L, defined by nodal points located at each end. The mass of each
segment is lumped at the nodal points. Figure 8.12 shows a conceptual model for
representing the mechanical behaviour of a fully grouted bolt. Shearing resistance of
such a bolt is represented by spring/slider connections between the structural nodes
and the block zones in which nodes are located.

Reinforcement
element (bolt)

Figure 8.12: Conceptual model representing the mechanical behaviour of fully


grouted bolt.

The behaviour ofgrouted bolts

The axial behaviour of conventional reinforcement systems m a y be assumed to be


governed entirely by the reinforcing element itself. Being slender, the reinforcing

218

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

element offers little bending resistance, and is treated as one-dimensional m e m b e r


(constant strain elements) subjected to uniaxial tension or compression. A onedimensional constitutive model would, thus be adequate for describing the axial
behaviour of the reinforcing element. The axial stiffness is described in terms of; (a)
the reinforcement cross-sectional area, A, and (b) the Young's modulus, E. The
incremental axial force, AFl, is thus calculated from the incremental axial
displacement by:

EA
AF

=- A M

(8.18)

____>

where,
Aul = Auitj
= Auiti + AU2t2

-(-r'-r-'Mtf1-"^

uj"1, u\b] etc. are the displacements at the bolt nodes associated with each bolt
element. Subscript 1 corresponds to the x-direction and subscript 2 to the y-direction.
The subscripts [a], [b] refer to the nodes. The direction cosines ti and t2 refer to the
tangential (axial) direction of the bolt segment.

Figure 8.13 shows a line diagram of the axial behaviour of a grouted bolt. A tensile
yield limit and a compressive yield force limit can be assigned to the bolt.

219

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

Accordingly, bolt forces cannot develop that are greater than the tensile or
compressive strength limits.

Tensile force

Yield limit

Bolt Young's
modulus
Extension

Yield limit

Compressive force

Figure 8.13: Line diagram showing the axial behaviour of bolt in U D E C .

Calculation of the relative displacement at the grout/rock interface employs an


interpolation scheme to compute the nodal displacement along the bolt axial
direction. In UDEC, each node on the bolt is assumed to exist within constant strain
triangular zone (called host zone) as shown in Figure 8.14. The interpolation scheme
uses weighting factors, which are based on the distance to each grid points of the host
zone as given by:

Auxp = WxAuxX + W2Aux2 + W3Aux3

220

(8.19)

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

where,
A u x p = incremental x-component of displacement at the nodal point,
Auxi = incremental grid point displacements (i = 1..3), and
Wj= weighting factors (i = 1..3).

Constant strain Finite


Difference triangle
Grid points

Bolt nodal point


Figure 8.14: A typical bolt element passing through triangular block.

Figure 8.15 represents the shear behaviour of the grout annulus. During relative
displacement between the bolt/grout interface and the grout/rock interface, the shear
behaviour is governed by the grout shear stiffness. The maximum shear force that can
be developed, per length of element, Fxmax/L, is limited by the cohesive strength of
the grout.

221

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

Force/length
.

jjmax

L ~
A

Grout shear
/

stiffness
Jill 1 llv.Jj
p.

Relative shear displacement

Tftnax

L
Compressive force

Figure 8.15: Line diagram showing the shear behaviour of grout in U D E C .

8.6.3 Material properties

A number of bolts were tested in the laboratory to determine the area, density,
Young's modulus, and yield force resistance of the bolt to be input into the model.
However, the properties related to the grout were found to be difficult to estimate,
and the grout annulus was thus assumed to behave as an elastic-perfectly plastic
solid. The incremental shear force, AFl, caused by the incremental relative shear
displacement, Aul, between the bolt surface and the borehole surface was related to
the grout stiffness, KDOn_, by the following relationship.

AFl = Kbond All* (8.20)

222

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

The value of Kbond could be determined from pullout tests. Alternatively, the stiffness

can be calculated from a numerical estimate for the elastic shear stress, TG, obtai

from an equation describing the shear stress at the grout/rock interface (St. John
Van Dillen, 1983):

_
T

Aw

" (D/2 + t)\n(l + 2t/D)

(8 21)

where,
G = grout shear modulus,
D = reinforcing diameter, and
T = grout annulus thickness.

Consequently, the grout shear stiffness, Kbond, is thus given by:


2TTG
Kbond =

\n(l + 2tlD)

(822)

In U D E C , the following expression has been found to provide a reasonable estimate


of Kbond (St. John and Van Dillen, 1983), which was has been incorporated in the
current formulation by the expression:

2/rfj

Kh

A
bond

T
r
10111(1 + 2///))

(8-23)
V

223

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

The m a x i m u m shear force per bolt length in the grout is thus the bond cohesive
strength, and its value can be estimated from the results of pullout tests at various
confining pressures, which at times becomes difficult to conduct in the laboratory.
Otherwise, it may be approximated from the peak shear strength relationship as
shown below (St. John and Van Dillen 1983):

tpeak

Ti QB (8.24)

where,
TI = approximately half of the uniaxial compressive strength of the weaker of
the rock and the grout, and
QB = quality of the bond between the grout and rock,
= 1 for perfect bonding.
Neglecting the frictional confinement effects, Sb0nd, may then be obtained from:

Sbond = 7C(D+2t) Tpeak (8.25)

Failure of reinforcing system does not always occur at the grout/rock interface.
Failure may occur at the bolt/grout interface, which is often the case for ground
reinforcement. In such cases, the shear stress should be evaluated at this interface by
replacing expressions (D+2t) with (D) in Equation 8.25.

224

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

8.6.4

T h e Continuous Yielding model

The continuous yielding joint model (Cundall and Hart, 1984) was proposed to
simulate, in a simpler manner, the internal mechanism of progressive damage of
joints under shearing, and was used in the formulation of present model. It is more
realistic than the standard Mohr-Coulomb joint as it considers the non-linear
behaviour observed in physical tests such as joint degradation, normal stiffness,
dependence on normal stress, and decrease in dilation angle with plastic shear
displacement. In this model the constitutive relationship is defined by:

Acr

= knAun (8-26)

where,
k = normal stiffness
= CCnGnn in which ctn and P are model constants.

For, shear loading, the model considers an irreversible non-linear behaviour from the
onset of shearing. The shear stress increment is calculated by:

Ax = ksAus (8.27)
where,
ks = normal stiffness,
= OLPs in which cts and p\ are model constants.

225

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

8.6.5

Conceptual C N S Model for Reinforced Joints

Figure 8.16 shows the conceptual model used in the UDEC analysis for simulating

the joint conditions as tested in the laboratory. The block sizes and the asperity

dimensions were proportional to the actual model dimensions used in the laboratory
The Boundary conditions applied to the model ensured the following laboratory
conditions.

///'////////

x
II

Uy = 0

Spring
k_ = 8.5 kN/mm

////////////

I.
Fully grouted
Bolt (constant
strain bar element)

Top
specimen

Joint plane

77777777777/

777/7777777,

240 m m

Figure 8.16: Conceptual C N S model of reinforced joints.

226

E
E
o
o

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

The bottom shear box moves only in x direction i.e. displacement in y

direction is zero, whereas displacement in x direction is free.


The top shear box moves only in y direction i.e. displacement in x direction is
zero, whereas displacement in y direction is free.
The spring block moves only in y direction. All the movements of the block
were restricted to the lower part of the spring, which was governed by the
upward movement of the top specimen. The upper part of the spring block
was fixed with the rigid load cell assembly, which in turn was attached to the
body of the CNS equipment.

8.6.6 UDEC Modelling of Reinforced Joints under CNS Condition

As with every other UDEC model, the first step was to create the first block. It was
then split into three blocks representing lower joint block, upper joint block and the
spring block. The splitting of lower and upper joint blocks involved the creation of
the joint plane asperities, which was achieved by incorporating a FISH function
called placecrack into the model. Once the blocks were created, they were discretised
into appropriated sizes using the gen quad or gen edge function in UDEC. The
material properties were then assigned to each block using UDEC prop and change
functions. Finally the bolt was introduced into the model using cable function in
UDEC and the boundary conditions and normal stress were applied. The following
material properties were used in the model.

227

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

Joint material properties:


Bulk modulus:

1.4 GPa

Shear modulus:

0.79 GPa

Density:

1400 Kg/m3

Joint properties:
Joint normal stiffness:

0.45 GPa/m (correspond to 8.5 k N / m m over an


area of 250mm x 75mm)

Joint shear stiffness:

0.045 GPa/m

Joint friction angle:

33.5

Bolt/grout properties:
Bolt modulus (E)

98.6 G P a

Grout bond stiffness:

1.12x10'N/m/m

Grout cohesion:

1.75xl0 5 N/m

Figure 8.17 shows the output from U D E C for discretised joint block and the bolt
along with its node points. All the models were run for 300,000 iterative steps and
saved in UDEC sav file format. Appendix III shows the UDEC code used to
formulate the reinforced joint model under CNS condition.

228

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

B UDEC 3.10 Plot Window

HBE
Spring block, K n = 8.5 k N / m m

Figure 8.17: M e s h generation by U D E C .

8.6.7

Validation of the M o d e l

The numerical model, as described above, w a s validated by comparing its output


with the laboratory results. Figure 8.18 indicates the variations of shear stress with
shear displacement from both the numerical model and the laboratory tests for
selected initial normal stress conditions. At low normal stress conditions (Figure
8.18a), both the laboratory results and the UDEC model predicted the shear stress
values in close agreement with each other. As the initial normal stress was increased,
there was an increasing gap between the UDEC predictions and the laboratory

229

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

results, as shown in Figure 8.18c. Such behaviour was attributed to the continuous
yielding model used in this study, which did not allow the shearing of asperities. In
practical situations where the high normal stress condition prevails, the shearing of
asperities is inevitable before reaching the peak shear. Thus the peak shear stress at
high normal stress conditions has two components; (a) the shear component due to
normal stress, and (b) the shear stress change due to degradation of asperities. The
component (b) is missing in UDEC model, and hence, the model predicted lower
shear stress values at high normal stress conditions, when compared with the
laboratory results.

Figure 8.19 shows the comparison of dilation profiles from the laboratory tests and
UDEC predictions. Again, similar trends in behaviour were observed for dilation
profiles, as in the case of shear stresses, i.e. at low normal stress conditions,
predictions were very good, however, at high normal stress conditions, dilations were
over-estimated by UDEC. Normal stress predictions from UDEC model are shown in
Figure 8.20 along with laboratory results. Again, greater deviations were observed at
higher normal stress conditions.

230

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

2.5

= 0.13 M P a

fJno - 0.63 M P a

U D E C model

. 1.5

U D E C model
"

Laboratory
experiment

Laboratory
experiment

0.5
0.5

0 *,
0

0 _
5

10

15

20

Shear displacement (mm)

a)

4.5 -,

3.5

3 1.

3-

_
_

_ 2.5-

A /

UDEC model

2-

10
M

_.

_
_

IB

00

a n o = 1.87 M P a

/V^*~

3.5 -

1-5 -

j
if

15

b)

a no = 1.25 MPa

10

Shear displacement (mm)

\ Laboratory
experiment

o
x:
co

/ \
/[
f

\
\

^
Laboratory
experiment

2.5 2

If

UDEC model

1.5
1
0.5 -

0.5 -

0 ,
5

10

15

10

15

-0.5 J

Shear displacement (mm)

Shear displacement (mm)

c)

d)

Figure 8.18: Comparison of shear stress profiles predicted by U D E C model and from
laboratory tests.

231

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

2.5

ano = 0.13 M P a

a n o = 0.63 M P a

U D E C model
U D E C model

B.

Laboratory
experiment

Laboratory
experiment

15
5

-0.5

10

15

S h e a r d is placement ( m m )

S h e a r d is placement ( m m )

b)

a)
0.8

a n o = 1.25 M P a

CT

no-

1.87 MPa

0.7
0.6
0.5

U D E C mode

a.
0.4
2

U D E C model

0.3 H

_
_

0.2 -

Laboratory
experiment

it)

0.1

(A

Laboratory
experiment

_
n
9
JC

10

-0.1
-0.2

15

ri^J^

10

15

Shear displacement (mm)

Shear displacement (mm)

d)

c)

Figure 8.19: Comparison of dilation profiles predicted by U D E C model and from


laboratory tests.

232

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

2.5

crno = 0.63 M P a

co
Q.

= 1.5

U D E C model

(0

in

o
IB

o
C

CO

Laboratory
experiment

0.5

10

15

Shear displacement (mm)

a)

a n o = 1.25 M P a
2.5

U D E C model

to

Laboratory
experiment

0.5

10

15

Shear displacement (mm)

b)
Figure 8.20: Comparison of normal stress profiles predicted by U D E C model and
from laboratory tests.

233

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

8.6.8 Comparison of CNL and CNS Shear Behaviour

In addition to the CNS UDEC model as discussed above, a UDEC code was also
written to predict the shear behaviour of reinforced joints under the CNL conditions,
for the purpose of comparison. This was achieved by making the joint normal
stiffness to be zero in the CNS model. Figures 8.21a and 8.21b show the shear stress
predictions under both CNL and CNS conditions, at initial normal stress levels of
0.13 MPa and 1.87 MPa, respectively. In either case, the CNL and CNS predictions
were close to each other at low shear displacements. At increased shear displacement,
the difference between the CNL and CNS increases. Because only limited
degradation of asperities occurs at low shear displacements, it can be inferred that, at
small shear displacement, the CNL predictions are acceptable.

The peak shear stress predicted from the CNL model was smaller as compared to the
CNS model, especially at low initial normal stress levels (see Figure 8.21a). At low
CNS shear stresses, the joint shear response is influenced by the increased normal

stress associated with the joint dilation. In the case of CNL, the joint dilation does no
contribute to increase the shear strength, as the normal stress is kept constant.
Therefore, the CNL shear stresses tend to be smaller when compared with the CNS

predictions. At high normal stress conditions, the joint dilation is relatively small due
to the degradation of asperities. Thus, the corresponding joint shear response, is
governed mainly by the frictional properties of the joint. Therefore, both CNL and
CNS models predict similar values at high initial normal stress conditions (see Figure

234

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

8.21b). In view of the above, it m a y be concluded that, at high normal stress levels,
both CNL and CNS model can be applied with equal confidence or reliability.

10

20

15

Shear displacement (mm)

a)

3.5

rjno=1.87MPa

to
M
<U
_.
4-1
If)

ro
a
J_

to
2

10

Shear displacement (mm)

b)
Figure 8.21: Comparison of shear stress profiles predicted by U D E C model under
CNL and CNS conditions.

235

Chapter 8: Modelling of the Shear Behaviour of Reinforced Joints

8.7 CONCLUSION

Based on the Fourier analysis and the hyperbolic predictions, an analytical model was
formulated to predict the stress and dilation profiles for both unfilled and infilled
bolted joints under Constant Normal Stiffness condition. A BASIC program was
written to use the analytical model to predict the Fourier coefficients, stress
components, dilation and the peak shear strength. A numerical model of unfilled
joints was also presented using UDEC version 3.1 to predict the shear behaviour
under CNS conditions. A summary of conclusions drawn in this chapter includes:

The predicted values of shear stress, normal stress, and dilation of unfilled
bolted joint by the model were in close agreement with the laboratory results.
Hyperbolic equation predicted the drop in shear strength for infilled bolted
joints to a very high accuracy with a regression coefficient exceeding 0.99.
The analytical model of infilled bolted joint using the unfilled model and the
hyperbolic equation predicted the shear behaviour of such joints close to the
laboratory observations.
The UDEC model underestimated the shear strength when compared with
laboratory testing, whereas it overestimated both normal stress and dilation
especially at high initial normal stress conditions.
The CNL and CNS predictions were dissimilar at low to medium normal
stress conditions.
Both CNL and CNS models predicted similar values at small shear
displacements and as well as at high initial normal stress conditions.

236

Chapter 9

CONCLUSIONS AND RECOMMENDATIONS

9.1 CONCLUSIONS

Several aspects of the shear behaviour of fully grouted bolts under Constant Normal
Stiffness condition were studied, both in the laboratory and in the field. The
laboratory testing programme included; (i) the study of the shear behaviour and
failure mechanism of the bolt/resin interface of two most popular bolt types, currently
in use in Australian coal mines, (ii) the study of the shear behaviour of model joints
made from gypsum plaster, with and without bolting, and (iii) the study of the shear
behaviour of bolted and non-bolted joints containing clay infill. The laboratory study
on the shear behaviour of the bolt/resin interface of fully grouted bolts was supported
with field investigations in two local coal mines, namely West Cliff and Tower
Collieries, in the Southern Coalfields of Sydney Basin, NSW, Australia. An
analytical model was presented using Fourier transforms, to predict the shear
behaviour of both unfilled and infilled bolted joints. The following paragraphs
describe the main conclusions drawn from this study.

Chapter 9: Conclusions and Recommendations

Rock Bolt and Joint Research

To date, most of the rock bolt research is concerned mainly with the estimation of
rock load and the load transfer mechanisms via the bolt-ground interaction. While
some of the reported research work is aimed at studying the failure mechanism of the
bolt/resin interface or the rock mass itself, only a limited research work is reported on
the influence of bolt surface geometry on the failure mechanism of bolt/grout
interface. The past research on rock joints and their stability have been reported
widely, and they were mainly conducted under the Constant Normal Load condition.
Only lately, there has been limited reporting on the shear behaviour of both unfilled
and infilled joints under Constant Normal Stiffness condition, yet no research work
has been reported on the shear behaviour of bolted joints with infill.

Shear Behaviour of Bolt/Resin Interfaces

The shear behaviour of the bolt surface at various confining pressures directly
affects the load transfer mechanism from the rock to the bolt.
Bolts with deeper rib profiles and wider rib spacing (e.g. bolt type I) offered
higher shear resistance at low confining pressures less than 6 MPa, whereas, bolts
with shallower rib profile with narrower rib spacing (e.g. bolt type II) offered
marginally higher shear resistance at confining pressures beyond 6 MPa. Such
findings were supported by both laboratory and field investigations.
The maximum dilation of the bolt/resin interface occurred at a shear displacement
of about 60% of the bolt rib spacing.

238

Chapter 9: Conclusions and Recommendations

In general, the shear behaviour at low normal stress and the post-peak behaviour

was relatively superior for type I bolts than type I bolts.

Field Investigations and Load Transfer Mechanism of Fully Grouted Bolts

Field investigations at both West Cliff and Tower Collieries revealed the following:

The load transfer on the bolt was influenced by; a) the confining ground stress
conditions, b) the strata deformation, and c) the surface profile of the bolts. The
load transfer on type I bolts (50 kN at West Cliff Colliery and 39 kN at Tower
Colliery) was relatively higher as compared to type II bolts (19 kN at West Cliff
Colliery and 21 kN at Tower Colliery) when subjected to low shear loading. On
the other hand, under high shear loading conditions, load transfer on type II bolts
was marginally higher as compared to type I bolts.
There was no significant level of load build up on the bolts, when the
development face advanced beyond 50 m away from the test site. However, the
influence of the approaching of longwall face was detected when the face was
around 150 m away.
There was no significant variation of the load build up on the bolts in the two
gateroads, when the face distance from the test site was more than 150m.
The influence of front abutment pressure build up on the gateroads appears at
different face positions. The load build up on the bolts in the belt road occurs
when the longwall face is less than 150m from the test site, whereas, the same
build up on the travelling road starts when the face position is les than 60m.

239

Chapter 9: Conclusions and Recommendations

The m a x i m u m load transferred along the bolts was found to occur at the level of

maximum deformation.
The field study showed that, under the low influence of horizontal stress (both in
magnitude and the direction), type I bolt offered significantly higher shear
resistance, whereas under high influence of horizontal stress, type II bolt offered
marginally greater shear resistance at the bolt/resin interface. Such findings were
also observed in the laboratory, and they provide a useful guide for selecting
appropriate bolt types for given ground stress conditions.

Shear Behaviour of Unfilled Bolted Joints

The peak shear stress was increased with the increasing level of ano, and the shear
displacement at peak shear stress decreased with the increasing ano for both bolted
and non-bolted joints. The bolt affected an increased peak shear stress and an
enhanced stiffness of the bolt-joint composite, however, the level of bolt contribution
appeared to decrease with the increasing o-no. The bolt increased the joint shear
strength by more than 20% at an0 = 0.13 MPa, whereas the corresponding increase at
CTno = 3.25 was merely 6%. The strength envelopes of both bolted and non-bolted
joints showed an approximate bi-linear trend. The magnitudes of dilation and normal
stress profiles were less for bolted joints in comparison with the non-bolted joints.

240

Chapter 9: Conclusions and Recommendations

Shear Behaviour of Infilled Bolted Joints

The peak shear strength of both bolted and non-bolted infilled joints decreases
significantly with a thin infill layer of 1.5 mm (t/a = 0.3). The overall shear strength
of the bolted joint was increased by up to 30% at low normal stress levels, but not
greater than 12% at high normal stress levels, indicating the reduced effectiveness of
the bolt at high normal stress levels. The CNS strength envelope of infilled joints was
found to be non-linear, and in bolted joints, the apparent friction angle was found to
increase significantly at low normal stress. The greater the t/a ratio, the less important
the role of bolting is in affecting the joint shear strength. Bolts reduce dilation of
joints when t/a < 1, and increase joint compression when t/a > 1.

The shear plane passes through the infill or asperity or both, depending on the infill
thickness to asperity height ratio, t/a. Beyond a critical t/a ratio of 1.0, the effect of
bolting and the influence of asperities become marginal, and the joint behaviour is
dictated mainly by the infill characteristics. At low infill thickness, the shear
behaviour of bolted joints under CNS condition is distinctly different from the
behaviour under conventional CNL. At t/a ratios exceeding unity, the difference
between the CNS and CNL envelopes is marginal, implying that the conventional
shear strength envelope (CNL) can adequately represent the behaviour of clayey
infill. Nevertheless, for rough joints with little or no infill, the CNS behaviour is
more realistic in practice, especially for underground mining conditions in bedded or
jointed rock.

241

Chapter 9: Conclusions and Recommendations

Modelling of the Shear Behaviour of Reinforced Joints

The values of shear stress, normal stress, and dilation of both unfilled and
infilled bolted joints predicted by the analytical model were in close
agreement with the laboratory results, for o range of 0.13 MPa to 3.25 MPa.
The drop in shear strength of infilled joints, as predicted by the hyperbolic
equation, was accurate with a regression coefficient exceeding 0.99.
The UDEC model underestimated the shear strength when compared with
laboratory tested values, whereas it overestimated both the normal stress and
dilation, at high initial normal stress conditions (i.e. ano > 1.25 MPa).
For CTno > 1-25 MPa, both CNL and CNS models predicted identical shear
stress values at small shear movements, however, at o-no < 1.25 MPa, the
shear stress for both CNL and CNS were different and influenced by the value
ofrjno.

9.2 RECOMMENDATIONS FOR FURTURE RESEARCH

The study of the shear behaviour of grouted bolts under CNS should be extended to
address the following, which have not been fully addressed within the scope of this
study.

242

Chapter 9: Conclusions and Recommendations

Bolt surface studies:

Bolt surface studies reported in this thesis should be advanced to include


other rigid bolt configurations as well as cable bolts.

The bolt surface study should be extended by using numerical tools such as
UDEC, FLAC, ABACUS etc., which will lead to the design of new bolt
surface profiles to suit various in-situ stress conditions.

The field investigations should be extended to include the bolt performance


at a wide range of in-situ stress conditions. With the use of more extensive
instrumentation, such as the measurement of in-situ stress in the
instrumentation site and the use of horizontal strata movement measurement
techniques, a more comprehensive design scheme for grouted bolts may be
accomplished.

Infilled bolted joint studies:

The laboratory studies of infill bolted joint introduced in the thesis should be
extended to; (a) bolts installed at various angles to the joint plane, (b)
various types of infill material, (c) joints profiles with closer representation
of natural joints, and (d) varying stiffness conditions, and (e) the influence
of pore pressures on infilled joint shear strength.

243

Chapter 9: Conclusions and Recommendations

Further studies m a y be conducted to compare the applicability of C N L and


CNS techniques for various surrounding environmental conditions, such as,
low, medium or high normal stress (i.e. an0 < 0.25 MPa, ano 1.0 MPa and
ano > 1.25 MPa), thin or thick infill layer (i.e. t/a ratio of as low as 0.1 and
as high as 5), and small or large shear displacements (i.e. h < 1 mm and h >
1 mm), as well as the influence of pore pressures.

The Set theory may be applied by using Venn diagrams to draw a boundary
between the CNL and CNS.

Further field investigations should be conducted to involve the in-situ study


of infilled bolted joint behaviour by using in-bed shear movement
measurement techniques, and validate it with analytical models based on the
field data.

244

REFERENCES

Adali, S. and Russel, R. (1980). Preliminary assessment of effects of grouted bolts in mine
tunnels. Trans. Instn. Min. Metall. (Sec A : Min. Industry), Oct., pp. A190-A197.

Aydan, O. (1989). The stabilisation of rock engineering structures by rock bolt. Ph.D. thesi
Nagoya University, Japan. 204p.

Aydan, O. and Kawamoto, T. (1992). The stability of slopes and underground openings
against flexural toppling and their stabilisation. Rock Mech. & Rock Engg., Vol. 25
No. 3, pp. 143-165.

Bandis, S. C, Lumsden, A. C. and Barton, N. R. (1983). Fundamentals of rock joint


deformation. Int. J. Rock Mech. Min. Sci. and Geomech. Abstr., Vol. 20, No. 6, pp.
249-268.

Barla, G., Forlati, F. and Zaninetti, A. (1985). Shear behaviour of filled discontinuities.
Int. Symp. on Fundamentals of Rock Joints, Bjorkliden, pp. 163-172.

Barton, N. (1973). Review of a new shear strength criterion for rock joints. Engineering
Geology, Vol. 7, pp. 287-332.

Barton, N. (1976). Rock mechanics review, the shear strength of rock and rock joints. Int. J
Rock Mech. Min. Sci. & Geomech. Abstr, Vol. 13, pp. 255-279.

Barton, N. (1986). Deformation phenomena in jointed rock. Geotechnique, Vol. 36, No. 2,
pp. 147-167.

Barton, N. and Bandis, S. (1982). Effects of block size on the shear strength of jointed roc
23rd U.S. Symp. on Rock Mech., Berkeley, C A , pp. 739-760.

References

Barton, N. and Bandis, S. (1990). Review of predictive capabilities of JRC-JCS model in


engineering practice. Proc. Int. Symp. on Rock Joints, Leon, Norway, Balkema, pp.
603-610.

Barton, N., Lien, R. and Lunde, J. (1974). Engineering classification of rock masses for the
design of tunnel support. Rock Mechanics, Vol. 6, pp. 191-235.

Benmokrane, B. and Ballivy, G. (1989). Laboratory study of shear behaviour of rock joints
unde constant normal stiffness conditions. Rock Mechanics as a Guide for Efficient
Utilization ofNatural Resources, Khair (ed), Balkema, Rotterdam, pp.899-906.

Benmokrane, B., Chennouf, A. and Mitri, H. S. (1995). Laboratory evaluation of cementbased grout and grouted anchors. Int. J. Rock Mech. Min. Sci. & Geomech. Abstr.,
Vol. 32, N o . 7, pp. 633-642.

Benmokrane, B., Xu, H. and Bellavance, E. (1996). Bond strength of cement grouted glass
fibre reinforced plastic (GFRP) anchor bolts. Int. J. Rock Mech. Min. Sci. & Geomech.
Abstr., Vol. 33, N o . 5, pp. 455-465.

Bertachhi, P., Zaninetti, A., Barla, G. and Foralti, F. (1986). Laboratory test on the shear
behaviour of filled discontinuities. Proceedings of the International Symposium

on

Engineering in Complex Rock Foundation, Beijing, Kie (ed), Oxford, Pergamon, pp 262270.

Beyl, Z. S. (1945-1946). Rock pressure and roof support. Colliery Engineering, Sept. 1945 Oct. 1946., pp 36-42.

Bieniawski, Z. T. (1973). Engineering classification of jointed rock masses. Trans. S. Afr. I


Civ. Eng., Vol. 15, pp. 335-344.

Bieniawski, Z. T. (1987). Strata Control in Mineral Engineering. Balkema Publication,


Rotterdam, 212p.

246

References

Bjurstrom, S. (1974). Shear strength of hard rock joints reinforced by grouted untensioned
bolts. Proceedings 3rd Int. Conf ISRM Congress, Denver,
pp. pp. 1194-1199.

Brady, B. H. and Brown, E. T. (1993). Rock mechanics for underground mining. Chapman
and Hall, 571 p.

Cheng, F., Haberfield, C. M . and Sidel, J. P. (1996). Laboratory study of bonding wall
smear
in rock socketed piles. Proc. 7th ANZ Conf On Geomechanics, Adelaide, pp. 69-74,

Cundall, P. A. and Hart, R. D. (1984). Analysis of block test no. I inelastic rock mass
behaviour: phase 2 - A characterisation

of joint behaviour (final report). Itasca

Consulting Group, Rockwell, Hanford Operations, Subcontract SA-957, 49p.

Daniel C. & F. F. Wood. (1980). Fitting Equations to Data. John Willey & Sons., New York,
312p.

Dar, S. M. and Smelser, T. W. (1990). Laboratory and numerical modelling of coal mine roof
reinforcement systems. Mechanics of Jointed and Faulted Rock, Rossmanith (ed),
Balkema, pp. 455-461.

Deere, D. U., Peck, R. B., Monsees, J. E. and Schmidt, B. (1969). Design of tunnel liners an
support systems. U S Dept. of Trans., Washington D. C , PB-183, 799p.

De Toledo, P. E. C. and de Freitas. (1993). Laboratory testing and parameters controlling th


shear strength offilledjoints. Geotechnique, Vol. 43, No. 1, pp. 1-19.

Dight, P. M. (1982). Improvement to the stability of rock walls in open pit mines. Ph.D.
thesis, Monash University, Victoria, Australia, 282p.

Duncan, J. M. and Chang, C. Y. (1970). Non-linear analysis of stress and strain in soils. J.
Soil Mech. Found. Div. ASCE, Vol. 96, No. 5, pp. 1629-1653.

247

References

Dunham, P. K. (1976). Anchorage tests on strain gauged resin bonded bolts. Tunnels and
Tunnelling, pp. 73-76.

Farmer, I. W. (1975). Stress distribution along a resin grouted anchor. Int. J. Rock Mech. Mi
Sci. & Geomech. Abstr., Vol. 12, pp. 347-351.

Fenner, R. (1938). Research on the notion of ground stress. Gluckauf Vol. 74, pp. 681-695.

Ferrero, A. M. (1995). The shear strength of reinforced rock joints. Int. J. Rock Mech. Min.
Sci. & Geomech. Abstr., Vol. 32, No. 6, pp. 595-605.

Goodman, R.E. (1970). The deformability of joints in determination of the insitu modulus of
deformation of rocks. Special Technical Publication (ASTM), No. 477, pp. 174-196.

Goodman, R. E. and Shi, G. H. (1984). Block theory and its application to rock engineering,
Prentice-Hall Inc., Englewood Cliffs, NJ, 232p.

Haas, C. J. (1981). Analysis of rock bolting to prevent shear movement in fractured ground.
Mining Engineering, Vol. 33, pp. 698-704.

Haberfield, C. M. and Johnstone, I. W. (1994). A mechanistically based model for rough rock
joints. Int. J. Rock Mech. Min. & Geomech. Abstr., Vol. 31, No. 4, pp. 279-292.

Haberfield, C. M. & Seidel, J. P. (1998). Some recent advances in the modelling of soft rock
joints in direct shear. Proc. Int. Conf. Geomechanics and Ground Control in Mining and
Underground Construction, Wollongong, Aziz and Indraratna (eds), Vol. 1, pp. 71-84.

Haque, A. (1998). Shear behaviour of soft rock joints under constant normal stiffness. Ph.D.
thesis, University of Wollongong, 255p.

Hencer, S. R. and Richards, L. R. (1989). Laboratory direct shear testing of rock


discontinuities. Ground Engineering, pp. 3-24.

248

References

Hoek, E. (1977). Rock mechanics laboratory testing in the context of a consulting engineering
organization. Int. J Rock Mech. Min. Sci. & Geomech. Abstr., Vol. 14, pp. 93-101.

Hoek, E. (1983). Strength of jointed rock masses. Geotechnique, Vol. 33, No. 3, pp. 187-223

Hoek, E. (1990). Estimating Mohr-Coulomb friction and cohesion values from the HoekBrown failure criterion. Int. J Rock Mech. Min. Sci. & Geomech. Abstr., Vol. 27, pp.
227-229.

Hoek, E. and Brown, E. T., (1980). Underground excavation in rock. The institution of
Mining and Metallurgy, U.K., 682p.

Hoek, E. and Brown, E. T. (1997). Practical estimates of rock mass strength. Int. J Rock
Mech. Min. Sci., Vol. 34, No. 8, pp. 1165-1186.

Hyett, A. J., Bawden, W. F., Macsporran, G. R. and Moosavi, M. (1995). Constitutive law for
bond failure of fully-grouted cable bolts using a modified Hoek Cell. Int. J. Rock
Mech. Min. Sci. & Geomech. Abstr., Vol. 32, No. 1, pp. 11-36.

Indraratna, B. (1990). Development and applications of synthetic material to simulate soft


sedimentary rocks. Geotechnique, Vol. 40, No.2, pp. 189-200.

Indraratna, B. and Kaiser, P. K. (1990). Analytical model for the design of grouted rock bo
Int. Jnl. for Numerical and Analytical Methods in Geomechanics, Vol. 14, pp. 227-251.

Indraratna, B., Haque, A. and Aziz, N. (1997). Shear strength characteristics of soft rock
joints based on constant normal stiffness testing. Proc. 16' Int. Conf on Ground
Control in Mining, Peng (ed), Morgantown, U S A , pp. 267-273.

Indraratna, B. Haque, A. and Aziz, N. (1999). Shear behaviour of idealized infilled joints
under constant normal stiffness. Geotechnique, Vol. XLIX, pp. 331-355.

249

References

Indraratna, B. and Haque, A. (2000). Shear behaviour of rock joints. Balkema, Rotterdam,
164p.

Janssen, H. A. (1895). Experiments on grain pressures in soils. Zeitschrift, Verein Deutscher


Ingenieure, Vol. 74, pp. 681-695.

John, K. W. (1970). Engineering analysis of three dimensional stability problems utilising th


reference hemisphere. Proc. of the 24th US Symp. on Rock Mech., pp. 13-25.

Johnstone, I. W. and Lam, T. S. K. (1989). Shear behaviour of regular triangular


concrete/rock joints - analysis. Journal of Geotechnical Engineering, Vol. 115, pp. 711727.

Kanji, M.A. (1974). Unconventional laboratory tests for the determination of the shear
strength of soil-rock contacts. Proc. 3rd Congr. Int. Soc. Rock Mech., Denver, Vol. 2, pp.
241-247.

Kareby, K. (1995). Extending the traditional role of rock bolt anchors in theory and practice
Wideman (ed), Balkema, pp 335-343.

Kharchafi, M., Grasselli, G. and Egger, P. (1998). 3D behaviour of bolted rock joints:
experimental and numerical study. Mechanics

of Jointed and Faulted Rock,

Rosmanith (ed), Balkema, pp. 299-304.

Kinashi, H., Amano, S., Tsuchihara, H., Yoshioka, H. and Michihiro, K. (1997). Numerical
study on reinforcement effect of cable bolt in discontinuities. Rock Stress, Sugawara
& Obara (eds), Balkema, pp. 553-558.

Kulatilake, P. H. S. W. (1992). Joint network modelling and some scale effects in rock
masses. J. ofGeomechanics., Vol. 91, pp. 139-151.

250

References

Kulatilake, P. H. S. W., Wang, S. and Stephansson, O. (1993). Effect of finite size joints on
the deformability of jointed rock in three dimensions. Int. J Rock Mech. Sci. &
Geomech. Abstr., Vol. 30, No. 5, pp. 479-501.

Kutter, H. K. and Ruttenberg, A. (1979). The residual shear strength of filled joints in ro
Proc. 4th Int. Congr. Rock Mech., Montreux, Vol. 1, pp. 221-227.

Ladanyi, H. K. and Archambault, G. (1970). Simulation of shear behaviour of a jointed rock


mass. Proc. IIth Symp.on Rock Mech, Urbana, Illinois, pp. 105-125.

Ladanyi, H. K. and Archambault, G. (1977). Shear strength and deformability of filled


indented joints. Proc. 1st Int. Symp. Geotech. Structural Complex Formations, Capri, pp.
317-326.

Lama, R. D. (1978). Influence of clay fillings on shear behaviour of joints. Proc. 3rd Congr
Int. Assoc. Eng. Geol., Madrid, pp. 27-34.

Lang, T. A. and Bischoff, J. A. (1981). Research study of coal mine rock reinforcement.
U S B M Open File Report, No. 72-82, 182p.

Leichnitz, W. (1985). Mechanical properties of rock joints. Int. J. Rock Mech. Min. Sci. &
Geomech. Abstr, Vol. 22, pp. 313-321.

Li, C. and Stillborg, B. (1999). Analytical models for rock bolts. Int. J Rock Mech. Min. S
& Geomech. Abstr., Vol. 36, pp. 1013-1029.

Littlejohn, G. S. and Bruce, D. A. (1975). Rock anchors - state of the art, part I: design.
Ground Engineering, pp. 25-32.

Londe, P. G. and Vormeringer, R. (1970). Stability of slopes - graphical methods. J Soil.


Mech. Foundns. Div., ASCE, Vol. 96, pp. 1411-1434.

251

References

Marence, M . and Swoboda, G. (1995). Numerical model for rock bolts with consideration of
rock joint movements. Rock Mechanics and Rock Engineering, Vol. 28, No. 3, pp.
145-165.

Mark, C, Mucho, T. P. and Dolinar, D. (1998). Horizontal stress and longwall headgate
ground control. Mining Engineering, January, pp. 61-68.

Ohnishi, Y. and Dharmaratne, P. G. R. (1990). Shear behaviour of physical models of rock


joints under constant normal stiffness conditions. Proceedings Int. Symp. Rock Joints,
Leon, Barton & Stephanson (eds), Balkema, Rotterdam, pp. 323-327.

Papaliangas, T., Hencher, S. R., Lumsden, A. C, and Manolopoulou, S. (1993). The effect of
frictionalfillthickness on the shear strength of rock discontinuities. Int. J. Rock Mech.
Min. Sci. & Geomech. Abstr., Vol. 30, No. 2, pp. 81-91.

Patton, F. D. (1966). Multiple modes of shear failure in rocks. Proc. 1st Cong. ISRM., Lisbon
pp. 509-513.

Pellet, F. (1994). Strength and deformability of jointed rock masses reinforced by rock bolts
English translation of the doctoral thesis 1169, Swiss Federal Institute of Technology,
Lausanne, Switzerland, 206p.

Pellet, F. and Boulon, M. (1998). A constitutive model for the mechanical behavior of bolted
rock joint. Mechanics of Jointed and Faulted Rock, Rossmanith (ed), Balkema, pp.
179-183.

Pellet, F. and Egger, P. (1996). Analytical model for the mechanical behaviour of bolted rock
joints subjected to shearing. Rock Mechanics and Rock Engineering, Vol. 29, pp. 73-97.

Pereira, J.P. (1990). Mechanics of filled discontinuities. Mechanics of Jointed and Faulted
Rock, Rossmanith (ed), Balkema, Rotterdam, pp. 375-380.

252

References

Phien-wej, N., Shrestha, U. B. and Rantucci, G. (1990). Effect of infill thickness on shear
behaviour of rock joints. Rock Joints, Barton & Stephansson (eds), Balkema, Rotterdam,
pp. 289-294.

Rabcewicz, L. (1964). New Austrian tunnelling method. Water Power, Nov. pp 453-457.

Saeb, S. and Amadei, B. (1992). Modelling rock joints under shear and normal loading. Int. J
Rock Mech. Min. Sci. & Geomech. Abstr., Vol. 29, No. 3, pp. 267-278.

Serrano, A. and Olalla, C. (1999). Tensile resistance of rock anchors. Int. J Rock Mech. Min
Sci. & Geomech. Abstr., Vol. 36, pp.449-474.

Sidel, J. P. and Haberfield, C. M. (1995). The application of energy principles to the


determination of the sliding resistance of rock joints. Rock Mech. & Rock Eng., Vol.
28, No. 4. pp. 211-226.

Signer, S. P. and Jones, S. D.( 1990a). A case study of grouted roof bolt loading in a two-e
gate road. P' Conference on Ground Control in Mining, West Virginia University,
Morgantown, June 4-6, pp. 35-41.

Signer, S. P. and Jones, S. D. (1990b). Case study on the use of grouted roof bolt loading i
two-entry gate road. 31st US Symp. on Rock Mechanics, pp. 145-152.

Skinas, C. A., Bandis, S. C. and Demiris, C. A. (1990). Experimental investigations and


modelling of rock joint behaviour under constant normal stiffness. Proceedings Int. Symp.
Rock Joints, Leon, Barton & Stephanson (eds), Balkema, Rotterdam, pp. 301-308.

Spang, K. and Egger, P. (1990). Action of fully-grouted bolts in jointed rock and factors o
influence. Rock Mechanics and Rock Engineering, Vol. 23, pp. 201-229.

Spigel, M. R. (1974). Theory and problems of Fourier analysis with application to boundary
value problems. Schaum's outline series, McGraw-Hill Inc., 191p.

253

References

St. John, C. M. and Van Dillen, D. E. (1983). Rock bolts: a new numerical representation and
its application in tunnel design. Proc. 24th U.S. Symp. on Rock Mech, Texas, June, pp.
13-26.

Stankus, J. C. and Peng, S. S. (1996). A new concept for roof support. Coal Age, September,
pp. 67-72.

Strata Control Technology. (1996). Influence of Sandstone Roof on Longwall Mining at


Tower Colliery. Report No. T O W 1 1 3 0 , 4p.

Strata Control Technology. (1998). An introduction to roof bolting. SCT, Wollongong, 44p.

Tadolini, S. C, Kraus, P. and Worbois, A. (2000). High capacity tensioned cable bolts. 19th
International Conference on Ground Control in Mining, West Virginia University,
Morgantown, August 8-10, pp. 59-66.

Tao, Z. and Chen, J. X. (1983). Behaviour of rock bolts as tunnel support. Rock Bolting Theory and Application in Mining and Underground Construction, Stephansson (ed),
Balkema, pp. 87-92.

Terzaghi, K. (1946). Rock tunnelling with steel supports. Commercial Shearing and
Stamping Co., Youngston, Ohio, pp. 119-140.

Tulinov, R. and Molokov, L. (1971). Role of joint filling material in shear strength of roc
ISRMSymp., Nancy, Paper 11-24.

Van Sint Jan, M. L. (1990). Shear tests of model rock joints under stiff normal loading.
Proceedings Int. Symp. Rock Joints, Leon, Barton &

Stephanson (eds), Balkema,

Rotterdam, pp. 323-327.

Weishen, Z., Zhuoyuan, L., Ping, W. and Caizhao, Z. (1990). Research on strength behavior
of jointed rock mass by numerical and physical simulation. Mechanics of Jointed and
Faulted Rock, Rossmanith (ed), Balkema, pp. 389-397.

254

References

Windsor, C. R. (1997). Rock reinforcement systems. Int. J Rock Mech Min. Sci. & Geomech.
Abstr., Vol. 34, pp. 919-951.

Wittenberg, D. and Ruppel, U. (2000). Quality management for grouted rock bolts. 19th
International Conference on Ground Control in Mining, West Virginia University,
Morgantown, August 8-10, pp. 249-254.

Witthaus, H. and Ruppel, U. (2000). Rock bolting for highly stressed roadways. 19th
International Conference on Ground Control in Mining, West Virginia University,
Morgantown, August 8-10, pp. 234-240.

Wittke, W. (1964). A numerical method of calculating the stability of slopes in rocks with
systems of plane joints. Felsmechanik und Ingenieur-geologie, Suppl. 1, pp 103-129.

Yazici, S. and Kaiser, P. K. (1992). Bond strength of grouted cable bolts. Int. J. Rock Mec
Min. Sci. & Geomech. Abstr., Vol. 29, No. 3, pp. 279-292.

255

Appendix I

SOLUTION OF EQUATION 8.12 FOR SHEAR STRENGTH OF


R E I N F O R C E D JOINTS

The original equation is given by,

T(h) =

k^h +
A J sin 6

+ tan(i)
M-tan(^)
-tan(^)tan(/ )

(Al)

Putting the values of a b n from Equation 8.7 w e obtain,

T(h) =

k.Ji

AJsinO

<7no +

KAWi

tan(^) + tan(Q
J l-tan(^)tan(/,)

(A2)

N o w introducing the Fourier terms for dilation as per Equation 8.4 and assuming
that the dilation function is an even function, we obtain,

f
ks.h
T(h) =
<J+
no
Ab/s'm0 +

2n7ix\W
cos

tan(^) + tan(/)
_l-tan(^)tan(/A)

V ^

=i

)))

(A3)

Appendix I: Solution of Equation 8.12

knowing that,

tan(ih) =

d(Sv)
dh

~Tl;{a"cos~T-

(A4)

n=\

and substituting the value of tan(ih) in Equation A3, then,

r(h) =

+
AIsm6
V

k..fa0
no

2n7DcXs\

-,[

A V2

n=\

J)

tan(^,) + tan(/)
(

l-tan(^)tan
V

2n

oo

a C0S

~Tia{ "

(A5)

2n7tx\

J)

dr(h)
For peak shear stress, approaches to zero, and differentiating Equation A 5
dh
with respect to h, w e obtain,

257

Appendix I: Solution of Equation 8.12

dx
= C -C
~dh _q
^

C 3 sin(./z) + C3C4 +

<J
no

cos(th)

Cj +
A

2ax

{l + C 4 sin///.)}2

(A6)

Where,

k.h

c,= A.lsinO
C2 =tan(^)+tan(i)
t=

2n

k
A

CA=axttm(<f>b)

dz
for =0, in Equation A6, then,
dh

C 4 2 sin2 (./*) + (2C,C4 - C 2 C 3 )sin(rt) - C 2 C 5 cos(./z) +


(A7)

(C1-C2C3C4) = 0

Where,

258

Appendix I: Solution of Equation 8.12

C5=anoC4t+aCiC*
2a

The above equation can be rewritten in the following form,

(c;

y+(2c/c6y+(c62+2c7c42+c2c5y+
(2C6C7> + (c72-C2C5)=0

<A8)

Where,

Q=(2C,C4-C2C3)
C7 \Cj C2C.3C4J

y = sin(th)

The Equation A 8 is a fourth order linear equation of y, which was solved using
MATHEMATICA.

259

Appendix: IIA

Program code for calculating Fourier constants and stress parameters

'PROGRAM CODE FOR REINFORCED SHEAR STRENGTH MODEL


CLS
CONST PI = 3.141592
INPUT "Input complete name of data file"; infile$
DIM maxcount AS INTEGER 'maxcount= maximum number of data points
INPUT "provide maximum number of data points"; maxcount
DIM a(0 TO maxcount) AS SINGLE, b(l TO maxcount) AS SINGLE
OPEN infile$ FOR INPUT AS #1
LINE INPUT #1, titlel$
LINE INPUT #1, title2$
INPUT #1, SDStart!, SDEnd!, SDInterval!, SW, SL, Stiffness, INS,
FricAngl, Asp_angle
Div_count% = (SDEnd! - SDStart) / SDInterval!
DIM Dilation(0 TO Div_count%) AS SINGLE
DIM EstDil(0 TO Div_count%) AS SINGLE
DIM NS(0 TO Div_count%) AS SINGLE
DIM ShearStss(0 TO Div_count%) AS SINGLE
DIM Slope(0 TO Div_count%) AS SINGLE
FOR x = 0 TO Div_count%
INPUT #1, Dilation(x)
NEXT x
INPUT #1, Infill_type%
IF Infill_type% = 0 THEN GOTO EndRead
INPUT #1, TAStart, TAEnd, NumStep%, Alfa, Beta, Rf
DIM PeakShearDrop(0 TO NumStep%) AS SINGLE
DIM InfillPeak(0 TO NumStep%) AS SINGLE
EndRead:
CLOSE #1
'Calculation of Fourier coefficients
FourierCoeff:
REDIM a(0 TO maxcount) AS SINGLE, b(l TO maxcount) AS
FOR n = 0 TO maxcount
FOR k = 0 TO Div count. - 1

Appendix H A : Program Code for Fourier Coefficients and Shear Strength

u * ^ ~ 3 ( n ) + Dila tion(k) * COS (2 * Pi * k * n / Div count%)


NEXT k

a(n) = a(n) * 2 / Div count!


NEXT n
FOR n = 1 TO maxcount
FOR k = 0 TO Div_count% - 1
b(n) = b(n) + Dilation(k) * SIN(2 * Pi * k * n / Div count.)
NEXT k
b(n) = b(n) * 2 / Div_count%
NEXT n
'Calculation of dilation
Dilation:
EstDil(O) = 0
NS(0) = INS
T = SDEnd! - SDStart!: k = 1
FOR x = (SDStart! + SDInterval!) TO SDEnd! STEP SDInterval!
y = 0
FOR n = 1 TO maxcount
y = y + a(n) * COS(2 * Pi * n * x / T) + b(n) * SIN(2 * Pi *
n * x / T)
NEXT n
y = a(0) / 2 + y
EstDil(k) = y
NS(k) = INS + Stiffness * EstDil(k) * 1000 / (SW * SL)
k = k + 1
NEXT

'Calculation of shear stress


ShearStress:
ShearStss(O) = 0: PeakShear = 0: NormAtPeak = INS
FOR k = 1 TO Div_count%
Slope(k) = ATN((EstDil(k) - EstDil(k - 1) ) / SDInterval!)
ShearStss(k) = 1000 * (1.72 + .65 * LOG(k)) / (SL * SW) + NS(k)
* (TAN(FricAngl * Pi /
180) + TAN(Asp_angle * Pi / 180)) / (1 - TAN(Pi *
FricAngl / 180) *
TAN(Slope(k)))
IF PeakShear < ShearStss(k) THEN
PeakShear = ShearStss(k)
NormAtPeak = NS(k)
DisAtPeak = k * SDInterval! + SDStart!
ELSE
END IF
NEXT k

261

Appendix H A : Program Code for Fourier Coefficients and Shear Strength

'Calculation of impact of infill


Infill:
IF Infill_type% = 0 THEN GOTO Printing
TAIntvl = (TAEnd - TAStart) / NumStep%
FOR k = 0 TO NumStep.
PeakShearDrop(k) = INS * (TAStart + TAIntvl * k) / (Alfa *
(TAStart + TAIntvl * k) +
Beta)
InfillPeak(k) = PeakShear - PeakShearDrop(k)
MaxDrop = PeakShear - (INS * Rf / Alfa)
IF InfillPeak(k) < MaxDrop THEN InfillPeak(k) = MaxDrop
NEXT k
INPUT "Input the value of t/a ratio for calculating the infill Peak
shear stress "; TA
PeakShearDrop = INS * TA / (Alfa * TA + Beta)
InfillPeak = PeakShear - PeakShearDrop

'Printing of final output to a date file


printoutput:
INPUT "Type result file name"; Outfile$
OPEN Outfile$ FOR OUTPUT AS #2
PRINT
PRINT
PRINT
PRINT
PRINT
PRINT

#2,
#2,
#2,
#2,
#2,
#2,

"Fourier Coefficients for:"


titlel$: PRINT #2,
titlel$: PRINT #2,
"
a
b"
USING "

FOR n = 1 TO
PRINT
PRINT
PRINT
NEXT n

(0)

maxcount
#2, USING " (##) "; n;
#2, USING " ###.#####"; a(n);
#2, USING " ###.#####"; b(n)

PRINT #2,
PRINT #2, " Disp(mm)
FOR k = 0
PRINT
PRINT
PRINT
PRINT
PRINT
NEXT k

###.#####"; a(0)

Dil(mm)

NS(MPa)

Slope(o)

SS(MPa)"

TO Div_count%
#2, USING ' ####.###; SDStart! + k * SDInterval!;
#2, USING ' ####.#####"; EstDil(k);
#2, USING ' ####.#####"; NS(k);
####.##"; Slope(k) * 180 / Pi;
#2, USING "
####.#####"; ShearStss(k)
#2, USING '

PRINT #2, USING "Peak shear stress ####.#### MPa "; PeakShear

262

Appendix H A : Program Code for Fourier Coefficients and Shear Strength

PRINT #2, USING "Normal stress at peak shear ####.#### M Pa ";


NormAtPeak
PRINT #2, USING "Shear displacement at peak shear ###.## mm ";
DisAtPeak
IF Infill_type% = 0 THEN GOTO EndPrint
PRINT #2,
PRINT #2, " T/A
PeakShearDrop
InfillPeak "
FOR k = 0 TO
PRINT
PRINT
PRINT
NEXT k

NumStep%
#2, USING "###.## "; (TAStart + TAIntvl * k ) ;
#2, USING "
####.##### ; PeakShearDrop(k);
#2, USING "
####.##### ; InfillPeak(k)

PRINT #2, USING " PeakShearDrop = ####.#### at T/A = ##.## ";


PeakShearDrop; TA
PRINT #2, USING " InfillPeak = ####.#### at T/A = ##.## "; InfillPeak;
TA
CLOSE #2
PRINT "The output from the program is written to"; Outfile$; "file"
EndPrint:
END

263

Appendix: IIB

Program code for converting data into input format

'PROGRAM CODE FOR CONVERTING DATA FROM EXCEL TO BASIC PROGRAM INPUT
FORMAT
CLS
INPUT "Input complete name of data file"; infile$
OPEN infile$ FOR INPUT AS #1
INPUT #1, nlim%
DIM Dilation(0 TO nlim%) AS SINGLE
DilRead:
FOR x = 0 TO nlim% - 1
INPUT #1, Dilation(x)
NEXT x
EndRead: CLOSE #1

printtofile:
INPUT "Type name of the file for output to be written"; outfile$
OPEN outfile$ FOR OUTPUT AS #2
FOR k = 0 TO nlim% - 1
PRINT #2, USING "###.##"; Dilation(k);
NEXT k
j = nlim% - 1
PRINT #2,
FOR k = 0 TO nlim%
- 1
PRINT #2, USING "###.##"; Dilation(j);
j = j - 1
NEXT k
FilePrintEnd:
PRINT "Please use the data in "; outfile$; "for input to the Fouri
analysis
program"
CLOSE #2
END

Appendix: IIC

Sample input file

Data file for Fourier Coefficient calculation


No of data points = 24/0.25 + 1
0 24 .25
75 250
8.5 0 .63 37 .5 18. 5
0.00 0.03 0.07 0.10 0.15 0.21 0.26 0 32
0.67 0.73 0.81 0.88 0.96 1.05 1.11 1 18
1.46 1.53 1.59 1.67 1.74 1.81 1.88 1 91
2.16 2.21 2.26 2.30 2.35 2.39 2.43 2 46
2.61
2.58 2.56 2.53 2.49 2.46 2.43 2.39 2 35
2.11 2.06 2.01 1.96 1.91 1.88 1.81 1 74
1.39 1.35 1.30 1.23 1.18 1.11 1.05 0 96
0.58 0.53 0.46 0.39 0.32 0.26 0.21 0 15
0

0.39
1.23
1.96
2.49

0.46
1.30
2.01
2.53

0.53
1.35
2.06
2.56

0.58
1.39
2.11
2.58

2.30
1.67
0.88
0.10

2.26
1.59
0.81
0.07

2.21
1.53
0.73
0.03

2.16
1.46
0.67
0.00

Appendix: IID

Sample output file

Fourier Coefficients for:


No of data points = 24/0.25 + 1
a b
( 0) 2.81313
( 1)
-1.15640
( 2)
-0.08945
( 3)
-0.08796
.sp (mm)
0.000
0.250
0.500
0.750
1.000
1.250
1.500
1.750
2.000
2.250
2.500
2.750
3.000
3.250
3.500
3.750
4.000
4.250
4.500
4.750
5.000
5.250
5.500
5.750
6.000
6.250
6.500
6.750
7.000
7.250
7.500
7.750
8.000
8.250
8.500
8.750
9.000

Dil(mm)
0.00000
0.07768
0.09239
0.11661
0.14990
0.19170
0.24128
0.29781
0.36037
0.42798
0.49964
0.57433
0.65107
0.72891
0.80701
0.88461
0.96105
1.03583
1.10855
1.17896
1.24693
1.31247
1.37569
1.43678
1.49602
1.55372
1.61025
1.66594
1.72113
1.77610
1.83108
1.88621
1.94153
1.99698
2.05242
2.10757
2.16206

0.00000
0.00000
0.00000
NS(MPa)
0.63000
0.66522
0.67188
0.68286
0.69796
0.71690
0.73938
0.76501
0.79337
0.82402
0.85650
0.89036
0.92515
0.96044
0.99585
1.03102
1.06568
1.09958
1.13254
1.16446
1.19528
1.22499
1.25365
1.28134
1.30819
1.33435
1.35998
1.38523
1.41025
1.43517
1.46009
1.48508
1.51016
1.53530
1.56043
1.58543
1.61013

Slope(o)
0.00
17.26
3.37
5.53
7.59
9.49
11.22
12.74
14.05
15.13
15.99
16.63
17.06
17.30
17.35
17.24
17.00
16.65
16.22
15.73
15.21
14.69
14.19
13.73
13.33
13.00
12.74
12.56
12.45
12.40
12.40
12.43
12.48
12.51
12.50
12.44
12.30

SS(MPa)
0.00000
1.05424
0.89112
0.94270
0.99643
1.05376
1.11482
1.17913
1.24581
1.31370
1.38147
1.44778
1.51139
1.57124
1.62655
1.67689
1.72214
1.76252
1.79848
1.83069
1.85993
1.88704
1.91287
1.93821
1.96377
1.99016
2.01784
2.04712
2.07816
2.11091
2.14516
2.18048
2.21628
2.25176
2.28601
2.31797
2.34657

Appendix IID: Sample Output File

9.250
9.500
9.750
10.000
10.250
10.500
10.750
11.000
11.250
11.500
11.750
12.000
12.250
12.500
12.750
13.000
13.250
13.500
13.750
14.000
14.250
14.500
14.750
15.000
15.250
15.500
15.750
16.000
16.250
16.500
16.750
17.000
17.250
17.500
17.750
18.000
18.250
18.500
18.750
19.000
19.250
19.500
19.750
20.000
20.250
20.500
20.750
21.000
21.250
21.500
21.750
22.000
22.250
22.500
22.750
23.000

2.21544
2.26718
2.31668
2.36331
2.40641
2.44534
2.47949
2.50829
2.53123
2.54793
2.55807
2.56147
2.55807
2.54793
2.53123
2.50829
2.47949
2.44534
2.40641
2.36331
2.31668
2.26718
2.21544
2.16206
2.10757
2.05242
1.99698
1.94153
1.88621
1.83108
1.77610
1.72113
1.66594
1.61025
1.55372
1.49602
1.43678
1.37569
1.31247
1.24693
1.17896
1.10855
1.03583
0.96105
0.88461
0.80701
0.72891
0.65107
0.57433
0.49964
0.42798
0.36037
0.29781
0.24128
0.19170
0.14990

1.63433
1.65779
1.68023
1.70136
1.72091
1.73856
1.75404
1.76709
1.77749
1.78506
1.78966
1.79120
1.78966
1.78506
1.77749
1.76709
1.75404
1.73856
1.72091
1.70136
1.68023
1.65779
1.63433
1.61013
1.58543
1.56043
1.53530
1.51016
1.48508
1.46009
1.43517
1.41025
1.38523
1.35998
1.33435
1.30819
1.28134
1.25365
1.22499
1.19528
1.16446
1.13254
1.09958
1.06568
1.03102
0.99585
0.96044
0.92515
0.89036
0.85650
0.82402
0.79337
0.76501
0.73938
0.71690
0.69796

12.05
11.69
11.20
10.56
9.78
8.85
7.78
6.57
5.24
3.82
2.32
0.78
-0.78
-2.32
-3.82
-5.24
-6.57
-7.78
-8.85
-9.78
-10.56
-11.20
-11.69
-12.05
-12.30
-12.44
-12.50
-12.51
-12.48
-12.43
-12.40
-12.40
-12.45
-12.56
-12.74
-13.00
-13.33
-13.73
-14.19
-14.69
-15.21
-15.73
-16.22
-16.65
-17.00
-17.24
-17.35
-17.30
-17.06
-16.63
-15.99
-15.13
-14.05
-12.74
-11.22
-9.49

267

2.37072
2.38943
2.40190
2.40749
2.40590
2.39706
2.38123
2.35889
2.33076
2.29769
2.26062
2.22052
2.17834
2.13498
2.09123
2.04778
2.00523
1.96403
1.92454
1.88700
1.85158
1.81832
1.78721
1.75813
1.73093
1.70536
1.68114
1.65794
1.63537
1.61306
1.59061
1.56764
1.54381
1.51883
1.49248
1.46462
1.43519
1.40424
1.37190
1.33838
1.30396
1.26897
1.23377
1.19873
1.16427
1.13077
1.09861
1.06817
1.03983
1.01394
0.99085
0.97093
0.95452
0.94200
0.93375
0.93014

Appendix IID: Sample Output File

23.250
23.500
23.750
24.000

0.11661
0.09239
0.07768
0.07275

0.68286
0.67188
0.66522
0.66298

-7.59
-5.53
-3.37
-1.13

Peak shear stress 2.4075 MPa


Normal stress at peak shear
1.7014 MPa
Shear displacement at peak shear 10.00 mm

0.93155
0.93837
0.95096
0.96962

Appendix: IIIA

U D E C code for calculating shear strength of reinforced joints under


C N S condition

File "C:\Itasca\UDEC\boltedjn.dat"
UDEC code for determining the shear strength of reinforced joints
new

def placecrack ;define fish fn to create joint profile


r

crack_ht = 0.005
crack_ln = 0.015
stx = -0.06
sty = 0
fnx2 = -0.045
fny2 = -0.005
loop while stx < 0.3
command
crack (stx,sty) (fnx2,fny2)
endcommand
stx = fnx2
sty = fny2
fnx2 = fnx2 + crack_ln
fny2 = fny2 + crack_ht
command
crack (stx,sty) (fnx2,fny2)
endcommand
stx = fnx2
sty = fny2
fnx2 = fnx2 + crack_ln
fny2 = fny2 - crack_ht
endloop
end
**********************************************************

Block creation
round 0.001
bl (-0.06,-0.1) (-0.06,0.30) (0.30,0.30) (0.30,-0.1)

Appendix IHA: U D E C Code for Calculating the Shear Strength of Reinforced Joints

placecrack

;call fish function placecrack

crack (0,0.30) (0,0)


crack (0.24,0.30) (0.24,0)
r

crack (0,0.15) (0.24,0.15)


t

del range (-0.06,0) (0,0.30)


del range (0.25,0.30) (0,0.30)

;left side
;right side

**********************************************************
Mesh generation
gen edge 0.03 range bl 4162 ;upper block
;lower block
gen edge 0.03 range bl 689
;spring block
gen quad 0.1 0.1 range bl 4502
**********************************************************
Assign material and cable properties
prop mat = 1 d = 1.4e-3 k = 1400 g = 792 ; all
change mat=2 range bl 4502
; spring
prop mat=2 d=5e-3 k=23 g=34.5
**********************************************************
Define joint model(C-Y model)
change jcons=3
set jcondf=3
set ovtol=0.01
set add_dil on
set del off
**************************** ******************************

Define joint properties


prop jmat=l jkn=453 jks=45.3 jen=0.0 jes=0.0 jfric=33.5
prop jmat=l jif=66.0 jr=5e3
change jmat=2 range 0,0.24 0.14,0.16
prop jmat=2 jkn=4 53 jks=10 jen=0.0 jes=0.0 jfric=0.0
prop jmat=2 jif=0.0 jr=0.0
*************

*********************************************

Place cable
cable (0.125,-0.09) (0.125,0.09) 19 3 181e-6 5
properties of cable
prop mat=3 cb_dens 7500 cb_ymod=98.6e9
prop mat=3 cb_yield=232e3 cb_ycomp=lel0

270

Appendix IHA: U D E C Code for Calculating the Shear Strength of Reinforced Joints

;properties of grout
prop mat=5 cb_kbond=l.12e7 cb_sbond=l.75e5
;***************************************** + i^^ + viJt + ii^^

;apply boundary conditions


bound xvel=0 range -0.01, 0.01 0.01,0.30
bound xvel=0 range 0.23,0.25 0.01,0.30
bound yvel=0 range -0.06,0.3 -0.11,-0.09
;**********************************************************

;apply normal load


bound stress (0,0,-1.25) range 0,0.24 0.29,0.31
.************************************************** ********

hist unbal
step 3000
bound yvel=0 range 0,0.24 0.29,0.31
,
; call fish program for cal. stress values
call jn_str.fis

271

Appendix: M B
Fish program code for calculating stress values

File "C:\Itasca\UDEC\jn_str.fis"
Fish function to calculate the average stress/disp.
Variable initialization
def av_str
whilestepping
sstav=0.0
nstav=0.0
njdisp=0.0
sjdisp=0.0
ncon=0
j1=0.100

; joint length

**********************************************************
Calculation of average shear strength and disp
ic=contact_head
loop while ic # 0
if c_y(ic) > -0.0025 then
if c_y(ic) < 0.001 then
if c_y(ic) > -0.00295 then
neon = neon + 1
sstav = sstav + c_sforce(ic)
nstav = nstav + c_nforce(ic)
njdisp = njdisp + c_ndis(ic)
sjdisp = sjdisp + c_sdis(ic)
endif
endif
ic = c_next(ic)
endloop
if neon # 0
sstav = sstav/jl
nstav = nstav/jl
njdisp = njdisp/ncon
sjdisp = sjdisp/ncon
endif
end
**********************************************************
Recording of calculated data
reset hist jdisp
hist unbal nc 1

Appendix IHB: UDEC FISH Function used in Model Formulation

hist
hist
hist
hist

n=1000
n=1000
n=1000
n=1000

sstav
ydisp
xdisp
ydisp

nstav njdisp sjdisp


0.050,0.011
0.050,0.005
0.050,0

;top-middle bolt
;left-middle bolt
;middle-b/r intrface

**********************************************************

Apply shear load by imposing shear velocity on bottom block


bou xvel = -0.03 range (-0.001,0.101) (0,1) ;top block
step 300000
save JS125.sav

273