Citation
Load transfer mechanics of resin-grouted rock bolts

Material Information

Title:
Load transfer mechanics of resin-grouted rock bolts
Creator:
Tadolini, Stephen C
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
104 leaves : illustrations, charts ; 28 cm

Subjects

Subjects / Keywords:
Rock bolts -- Testing ( lcsh )
Rock bolts -- Testing ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 92-93).
Thesis:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, College of Engineering and Applied Science
Statement of Responsibility:
by Stephen C. Tadolini.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
19783094 ( OCLC )
ocm19783094
Classification:
LD1190.E53 1988m .T32 ( lcc )

Full Text
LOAD TRANSFER MECHANICS OF RESIN-GROUTED ROCK BOLTS
by
Stephen C. Tadolini
B.S., University of Colorado, 1979
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Department of Engineering
1988


This thesis for the Master of Science degree by
Stephen C. Tadolini
has been approved for the
Department of
Civil Engineering
by
Date
/fl. !99b


Tadolini, Stephen C. (M.S., Engineering)
Load Transfer Mechanics of Resin-Grouted Rock Bolts
Thesis directed by Professor Tzong-Hsin Wu
The load transfer mechanics of resin-grouted rock bolts was
investigated, with the aid of ultrasonic measurement systems, under
laboratory and field conditions. The laboratory results revealed
that 94 percent of the load-induced deformations occurred in the
first 15 inches of the bolt. This was determined from pull-tests
conducted on twenty-five 4-ft rock bolts with eighty-eight reflector
signals. The field portion of the investigation, conducted in two
underground coal mines, examined instantaneous and time-dependent
transfer mechanisms. The instantaneous response of partially resin-
grouted rock bolts indicated that an initial load loss was to be
experienced due to bed relaxation and system creep. The time-
dependent transfer mechanisms, studied in two underground test
rooms, illustrated the effects of induced stresses and roof
movements due to environmental, geological, and on-going room
excavation. The ultrasonic instruments provided information that
was not obtainable prior to the development of this measurement
technique.
The form and content of this abstract are approved. I recommend
its publication.
Signed
zong-Hsin Wu


TV
CONTENTS
CHAPTER
I. INTRODUCTION............................................. 1
Background............................................ 2
Investigation Objectives.............................. 2
II. ULTRASONIC MEASUREMENT EQUIPMENT......................... 4
Pulsed-Phase-Lock-Loop Systems........................ 6
Pulse-Echo Systems................................... 12
Current Status..................................... 19
III. LABORATORY INVESTIGATIONS............................... 22
Instrument Description............................... 23
Instrument Operation................................. 25
Experimental Procedure............................... 26
Bolt Preparation..................................... 26
Machining.......................................... 26
Drilling........................................... 28
Calibration........................................ 28
Block Preparation.................................... 33
Construction....................................... 33
Drilling........................................... 33
Bolt Installation
33


V
CONTENTS (Continued)
Laboratory Data Acquisition and Reduction.............. 35
Analysis............................................ 39
Laboratory Conclusions................................. 54
IV. FIELD INVESTIGATION......................... 56
Instantaneous Rock Bolt Transfer Performance........... 56
Time-Dependent Transfer Mechanics...................... 62
Instrumentation..................................... 62
Conclusions and Recommendations........................ 90
BIBLIOGRAPHY................................................... 92
APPENDIX
A. ELONGATION MEASUREMENT DATA............................ 94
B. ELONGATION MEASUREMENTS PLOTTED ALONG
THE BOLT AXIS FOR EACH LOADING CONDITION................ 99
C. TORQUE TENSION RATIO TEST............................. 104


VI
TABLES
Table
1. Drill hole pattern utilized for test blocks........... 29
2. Coefficients of correlation for loading conditions
in zone 1 and zone 2................................. 46
3. Calculated intercept values for lines generated
from section 1 and section 2 for each loading
condition............................................. 47
4. Load transfer relationships........................... 53
5. Realized loads and percentage of applied loads....... 54
6. Calculated initial loads, 24 and 48 hours
after installation.................................... 61
7. Initial ultrasonic signal locations and
subsequent lengths.................................... 67
8. Pressure pad load data................................ 77
9. Ultrasonic bolt load data............................. 78


VII
FIGURES
Figure
1. Pulsed-phase-lock-loop system configuration.............. 7
2. Pulsed-phase-lock-loop system signal.................... 11
3. Pulsed-echo system schematic............................ 13
4. Ultrasonic reflection signal from a
bolt profile............................................. 15
5. Ultrasonic triggered signal from a maximum peak........ 18
6. Bolt Mike S-l ultrasonic instrument..................... 20
7. Raymond PDX-934 Bolt Gage............................... 24
8. Bolt heads and ultrasonic transducer.................... 27
9. High-speed drill and jog assembly....................... 32
10. Bolt hole pattern in test blocks........................ 34
11. Pull collar and resin cartridge prior to placement.... 36
12. Rotary drill and forklift placing bolts................. 37
13. Ultrasonic instrument and load indicator................ 38
14. A bilinear graph illustrating the general data trend.. 40
15. Linear regression analysis of 1,000 lb
and 2,000 lb loading increments......................... 41
16. Linear regression analysis of 3,000 lb
and 4,000 lb loading increments......................... 42
17. Linear regression analysis of 5,000 lb
and 6,000 lb loading increments......................... 43
18. Linear regression analysis of 7,000 lb
and 8,000 lb loading increments......................... 44


VI 1 1
FIGURES (Continued)
Figure
19. Combined slopes of the calculated linear
regression lines......................................... 45
20. Linear regression analysis of 1,000 lb
and 2,000 lb loading increments......................... 48
21. Linear regression analysis of 3,000 lb
and 4,000 lb loading increments......................... 49
22. Linear regression analysis of 5,000 lb
and 6,000 lb loading increments......................... 50
23. Linear regression analysis of 7,000 lb
and 8,000 lb loading increments......................... 51
24. Combined slopes of calculated linear
regression lines......................................... 52
25. Existing supports and experimental roof
bolt locations........................................... 58
26. Compression pad diagram.................................. 64
27. Vertical-displacement gage diagram....................... 65
28. Test site configuration and
instrumentation locations................................ 68
29. Initial instrumentation load isopleth.................... 70
30. Test site 30-day load contours........................... 72
31. Test site 30-day change in load contours................. 73
32. Differential sag station plot............................ 74
33. Test site 70-day load contours........................... 75
34. Test site 70-day change in load contours................. 76
35. Test site 140-day load contours.......................... 80


IX
FIGURES (Continued)
Figure
36. Test site 140-day change in load contours................ 81
37. Test site 238-day load contours.......................... 83
38. Test site 238-day change in load contours................ 84
39. Test room no. 1 pressure pad loads....................... 86
40. Test room no. 1 ultrasonic bolt loads
at the 2-inch and 6-inch levels......................... 87
41. Test room no. 2 pressure pad loads........................ 88
42. Test room no. 2 ultrasonic bolt loads
at the 2-inch and 6-inch levels......................... 89


CHAPTER I
INTRODUCTION
Millions of rock bolts, tensioned and untensioned, are used
each year for structural support. The application ranges from
underground mining environments to geotechnical applications of tie-
back walls and canal lock support. An increasing percentage of
these bolting systems utilize resin-anchor and resin-mechanical
anchor combination bolts as well as mechanical anchor bolts. The
method by which expansion anchor bolting systems carry roof loads is
well understood and documented. However, the method by which resin-
grouted bolting systems carry roof loads or support is much less
understood, mainly because testing procedures used to evaluate point
anchor systems have not been effective on resin- grouted bolts. The
support capability of a given anchorage system installation depends
on such factors as rock characterization, bolt characteristics, and
bolt tension. Measurement of installation and post-installation
bolt tension provides a great deal of information about the
effectiveness of the bolts and the behavior of the supported
structure.


2
Background
Current practice in civil and mining environments is to use
load cells, strain gages, pressure pads, or vibrating-wire
instrumented bolts. An obvious disadvantage of these is that the
bolt so instrumented receives special handling during the
installation process. This may make it nonrepresentative of
typically installed bolts. The cost per bolt with these types of
methods is also high.
Current practice in production is to use torque wrench
readings on a small statistical sample of the bolts, as required by
Federal regulations in the mining industry. This method has the
disadvantage that torque measurements are only about 30% accurate
as related to bolt load. The method may also disturb the anchorage
that has been obtained in the immediate host rock, while the small
sample gives only a vague indication of overall structural
stability. Resin-grouted or concrete grouted bolts or supports
cannot be tested with a torque wrench as the measurement would only
reveal the torsional resistance of the steel.
Investigation Objectives
In an effort to more fully understand the load transfer
mechanics associated with resin-grouted bolting systems so that
optimum column lengths for maximum load dissipation characteristics
can be determined, an investigation was designed and conducted to
evaluate the behavior of full and partial column resin grouted bolts
subjected to a range of end loading conditions, both in the


3
laboratory and under field conditions. Several research projects
have investigated the influence of installation procedures on the
effectiveness or resin-grouted bolts Q-4)11. The results of these
laboratory and field investigations illustrated the necessity for
investigating the nonlinear behavior of the stress distribution in
grouted bolts (5,6).
Underlined numbers in parentheses refer to items in the list
of references at the end of this report.


4
CHAPTER II
ULTRASONIC MEASUREMENT EQUIPMENT
To overcome the problems associated with the determination
of bolt stability and integrity, the use of ultrasonic measurements
was investigated and adapted for use with tension and untensioned
support used in the geotechnical and mining industries. These
ultrasonic measurements provide an easy-to-use, high-accuracy, low-
cost-per-bolt system that is capable of determining the amount of
load being subjected on an installed support system. The principle
involved is simple. The round trip travel time of a pulse or group
of waves through the bolt is measured. This travel time is affected
by bolt stress and strain and changes linearly with respect to bolt
loads. The waves are introduced into the bolt by a piezoelectric
transducer, which also detects the return waves that have reflected
back from the bolt end or other reflector.
With the use of the ultrasonic system, no modifications were
necessary to accommodate strain gages or vibrating wires that can
create zones of induced stress concentrations; these stress
concentrations can produce erroneous test results, particularly at
low stress levels. By measuring the axial deformation of a bolt
throughout its length, calculations can be made to determine the
amount of load induced at these locations by the applied end loading


5
conditions. This information can provide a clearer understanding as
to the overall loading characteristics of a full and partial-column
resin-grouted bolt.
The results of this investigation will provide a more
comprehensive understanding of the load transfer mechanics
associated with grouted roof bolting systems and provide a basis on
which to choose adequate column lengths for maximum load dissipation
characteristics. These parameters become extremely important as new
combination systems, utilizing partial grout columns, are being
introduced in the mining and geotechnical industries.
Making the required time measurements, however, is very
complex because the changes involved are very small. To appreciate
the electrical and mechanical complexity of the instrumentation
problem, it is important to recognize that a typical 5/8-in-diam, 4-
ft-long mine bolt stretches only 0.00064 in per 100 lbf of axial
load. Considering that the signal makes a round trip (8 ft) through
the bolt and that a typical signal velocity is 231,000 in/sec,
causing the instrumentation to change at least one unit for a 100-
Ibf change, means time must be stably resolved to about 5.5 X 10^
sec. A typical probing frequency is 2.25 MHz. Any error source
need only cause an equivalent change of the order of 0.02
microseconds to become a major difficulty. Signal attenuation is
also severe, reaching about 80 dB for the described bolt. The near-
field focal length of typical transducers at frequencies used is
only a few inches. Beyond this, the wavefront spreads out in a
conical fashion. Since bolt signal paths are relatively long, the


6
signal path is quite complex because the waves reflect off the sides
of the bolt, become mode-converted from P- to S-type waves and back,
and constructively and destructively interfere (7).
The research in the ultrasonic field was undertaken by
examining two fundamental approaches to the measurement. Pulsed-
phase-lock-loop (P^L^) systems were studied first. These systems
measure time directly by shifting the probing frequency to maintain
constant output to reflection wavephase differences as the bolt
loads. The frequency change, which is easy to measure, then becomes
an indirect measure of time change. Pulse-echo systems were
examined, evaluated, and selected for use in this investigation.
These systems have the ability to measure travel time directly by
incorporating special electronic techniques to provide the required
time resolution (8).
Pulsed-Phase-Lock-Loop Systems
The pulsed-phase-lock-loop system configuration, shown in
figure 1, uses a toneburst of, typically, six cycles to excite the
transducer. The transducer then detects the reflected toneburst.
The electronic continuation of the driving gated toneburst is then
multiplied with the received toneburst, as described mathematically
in (1):
E0 = M! sin (wjt) X M2 sin (i^t+e) (1)
where E0 is the multiplier output voltage, M is amplitude, 0 is the
phase difference between the reflection and the continuation of the


7
igure 1. Pulsed-phase-lock-loop system configuration.


8
driving toneburst, and and W2 are the angular velocities of the
driving and reflected signals. Using the trigonometric formula (2)
sin A sin B = cos (A-B) Cos (A+B), (1) reduces to
E0 = 1/2 M^M2 [cos (w^t w2t 0) cos (wjt + + 0)] (2)
Equation (2) reduces further, since in this case = u>2> to
E0 = 1/2 [cos (-0) cos {2w1 + 0)] (3)
The second term in (3) is of relatively high frequency compared to
the first term. The function of the low pass filter is to
effectively remove the second term. The filtered voltage is then
used to drive the voltage-controlled oscillator (VCO), which then
changes the driving frequency in direct proportion to its input
voltage, completing the loop.
The loop locks with the two signals in quadrature. The
locking tendency of the loop can be discovered by considering
perturbations from 0 = 90. Assume the phase difference increases
in magnitude from 90. Then cos (-0), which is in the third
quadrant, will reduce from 0 toward -1. The VCO input voltage will
then decrease. This reduces the driving frequency, which reduces 0,
since the origin of the output toneburst is fixed in time, and the
reflection arrival time does not vary with frequency. This
continues until 0 returns to 90. Assume the phase difference
decreases from 90. Then cos (-0), which is now in the fourth
quadrant, will increase from 0 toward 1. The VCO input voltage will
then increase. This raises the driving frequency, which increases


9
0. Again this continues until 0 = 90. Thus the loop locks in
quadrature. The total travel time of the toneburst can be described
in terms of a total phase angle 4> = N X 360 + 0, where 0 is the
fractional phase difference described above, and N is the whole
number of cycles between the toneburst initiation and the start of
the reflection:
where F is the frequency, t is the time, and subscript i is an
initial value.
Consider now that the bolt will be stretched while the loop
is locked, so we get a final <|> as follows:
As discussed previously, the action of the phase-lock loop is to
keep the phase constant, so
i = 2-TTFi t-j
(4)
f = 2nFftf
(5)
subtracting
f i = 2it(Fftf Fjt,)
(6)
(7)
But Ff F^ + AF and tf t^ + At, so from (7)
F^ = F^ + AF t i + At F i + AFAt
(8)


10
Subtracting F^, and dividing both sides by At, this reduces to
AFt-j
0 = -----+ F< + AF (9)
At 1 v
Again, substituting Ff = F^ + AF and simplifying further
At AF
ti" = ^f
(10)
This is the basic relationship between time and the
frequency that is the measured quantity of the P2I_2 instrument (9).
It is necessary to keep track of both the current frequency and the
change in frequency to calculate a number linearly related to bolt
stretch. The important frequency denominator subscript has been
ignored in previously published literature (7). A typical pulsed-
phase-lock-loop system signal is shown in figure 2.
Many laboratory experiments and one in-mine experiment were
carried out during this study with a P2L2 instrument. Using typical
roof bolts, the stability of the instrument in mine or mine-like
conditions ranges to hundreds of hertz, equivalent to a load on the
order of several thousands of pounds. A careful analysis showed
that the principal source of difficulty was that the transducer was
oscillating at its installed resonance rather than at the frequency
of the driving toneburst. This meant that anything that would alter
the installed transducer resonance would distort the desired
reading. Because of this, very small amounts of temperature change,
position change, or couplant change caused stability problems.
Another problem was that altering the instrument to allow it to


I v/div
=----------------n------------
_J____I____I___I____I___I I______I I
1.5/i.sec/d iv
Figure 2. Pulsed-phase-lock-loop system signal.


12
be portable and gassy-mine permissible would present significant
electronic problems.
For the above reasons, further research on the P2L2
technology for this application was given a very low emphasis. A
parallel research effort has shown that the mine bolt load
measurement can more easily be made using pulse-echo technology.
Pulse-Echo Systems
Five commercial strain-measuring pulse-echo systems have
been examined. Three other systems measuring reflected energy were
also examined. None were found suitable for mine-roof bolt
measurements. The most promising unit was selected and modified for
the mine-bolt application. A pulse-echo system schematic is shown
in figure 3. The pulse-echo systems measure the travel time
directly. A single pulse of about 300 V is used to drive the
transducer and start the timing circuitry. The round-trip travel
time is obtained when the reflected signal is detected by the
transducer. From the earlier discussion of time-resolution
requirements, it is evident that the trigger point on the reflected
wave must be very stable. If a straightforward voltage level
trigger is used, it will shift in time as signal amplitude varies.
The only points on a sinusoidal wave that remain stable in time as
amplitude varies are the zero crossings and the peaks. The
circuitry in the trigger-box typically combines a voltage level
criterion with a zero-crossing after the voltage threshold has been
exceeded. Automatic Gain Control (AGC) logic is used to ignore


13
Figure 3. Pulsed-echo system schematic.


14
unwanted reflections and noise. The AGC mainly suppresses the
signal gain for a selected period, and then very suddenly increases
gain for a period corresponding to the approximate arrival time of
the desired reflection. The pulsing process repeats typically about
400 times per second. The timing circuitry averages several hundred
readings to improve accuracy and then updates the display. The
correction logic corrects the time value for stress-caused velocity
change and also for non-stress-inducing, temperature-caused length
change. It also can compute equivalent length change from the
corrected time values and the initial length. A modified commercial
unit has been built to improve gain, gain control, portability, and
provide the potential for permissibility. Ultrasonic reflection
signals from an entire bolt profile are shown in figure 4.
Research has found pulse-echo instruments capable of
resolving 0.0001 in of stretch. An important consideration in any
type of application is that readings be repeatable as the transducer
is removed, then replaced at a later time to take a new reading.
The research program, to date, has shown that repetition accuracy of
0.0001 in can be achieved. This is accomplished by using magnetic
holddown on the transducer to provide a consistent force, ensuring
that the head flat is perpendicular to the bolt axis, and by using a
low-viscosity couplant material.
Nonresearch application requires that the calibration of a
few bolts be applied to a large number of similar bolts. Bolt-to-
bolt manufactured length variations of 0.1 in are common. This
would, of course, ruin the 0.0001-in resolution. To deal with this


Iv/div
15
1 I I I t_____I_I__I__I__1
70/isec/div
Figure 4. Ultrasonic reflection signal from a bolt profile.


16
problem, a uniform distance reflector is established by drilling a
0.040-in-diam hole perpendicular to the bolt axis at, typically, 2
to 15 in from the head. In testing 30 bolts to failure, the hole
has been found not to have any measurable effect on bolt strength.
This probably is due to the smooth flow of stress around the hole.
Because of small errors in hole position and small variations in
bolt structure and composition, uniformity is currently limited to
about 0.0015 in as measured by the instrument. Since the head-
to-reflector hole distance varies up to 15 in, this is equivalent in
terms of bolt load to about 700 lbf. Using individually
calibrated bolts, as was done for research purposes in this
investigation, allows resolution of 50 lbf. If the bolt is bent
more than about 5, accuracy deteriorates significantly owing to
wave diffraction from the curvature and to the outer fibers of the
bolt being stretched past yield. Resolution deteriorates to about
1300 lbf where signal levels are still high enough to provide useful
measurement. This problem can be attacked by using an attachment to
the drill steel to grind a flat spot in the mine roof or slope using
two reflector holes. In the latter approach, times from both holes
are subtracted from each other to remove the near head effects.
Corrosion on the bolt head can also reduce accuracy. To combat
corrosion, a couplant with anticorrosive properties is used, and the
head is cleaned before the measurement is taken. Based on field and
in-mine experience to date, corrosion is not expected to be a
significant problem.


17
To provide a measurement of bolt strain, the instrument must
be programmed to know the initial velocity, the initial length, and
the change in velocity with load. The velocity is set by applying
the instrument to a bolt stub of known length and then adjusting its
velocity setting until the instrument reading equals the known
length. Next, the initial head-to-reflector distance in an unloaded
bolt is measured using the ultrasonic instrument with the velocity
setting so calibrated. Finally, the velocity change with load is
calibrated by setting a stress factor in the instrument. Velocity
changes linearly with stress. To set the stress factor, the above
bolt is loaded to a measured level. Its exact length change is
measured using linearly variable displacement transducers (LVDT's)
accurate to 0.00001 in. The stress factor is adjusted until the
length change shown by the instrument equals the length change
measured by the LVDT's. A typical value of the stress factor is
0.27. The instrument now will measure the true strain of the
bolting system as it is loaded. A triggered signal from a maximum
peak is shown in figure 5.
The first instrument that was modified for these types of
measurements was the Raymond Bolt Gage. This instrument has the
capability of locating the reflective signal from a single location.
However, the temperature has to be updated manually prior to each
measurement. Additionally, the material velocities and the
calibration factors must be input into the instrument prior to each
set of measurements. The output of this instrument is elongation or
distance. Either reading can be converted to stress, strain, or
load with the proper data manipulation.


O.lv/div
18
1
J____I___I____I___ I I 1 l 1
1.5/z sec/d iv
Figure 5. Ultrasonic triggered signal from a maximum peak.


19
The latest advance in this technology is the Bolt Mike S-l.
This instrument directly converts reflectivity time into elongation,
load, stress, and torque. The Bolt Mike S-l updates the temperature
factor 10 times per second and requires no updating while acquiring
measurements. The instrument is capable of monitoring three
sections of a bolt simultaneously. The Bolt Mike S-l, shown in
figure 6, has programming cards inserted into the instrument to make
the required area calculations, signal velocities, and any stress or
load factor that the operator wants to apply to a set of given data.
The instrument has an RS-232 port that allows rapid data
transmi ssion.
Current Status
As stated earlier, the pulse-echo technique is capable of
resolving 50 lbf on an installed bolt. Experience has shown that
once the appropriate gain range has been set by observing the signal
on an oscilloscope, nonexpert personnel can make accurate readings.
In a situation where the calibration of one bolt is applied to a
whole group of bolts, the absolute accuracy is reduced to about
700 lbf, though the instrument will sense 50-1bf changes. This
reduction in accuracy comes from bolt variations in the apparent
head-to-reflector hole time measurement. About half of the effect
is composition, dimension, and forming stress. These variations are
primarily confined to the head and shank areas. With
threaded-tensioned bolts these variations are minimized. The multi-
reflector instrument should remove the head- and shank-area effects.


20


21
The measurement technique passed through a transition stage
from instrumentation improvement to research data acquisition. The
field work described in this paper continues to provide the
information necessary to initiate improvements. The information
acquired to date indicates that reliable measurements can be made on
expansion anchor and all types of resin-grouted bolting systems.
The field data are often difficult to interpret due to the number of
factors and influences that need to be considered in underground
openings as well as geotechnical applications.


CHAPTER III
LABORATORY INVESTIGATIONS
In an effort to understand the load transfer mechanics
associated with resin-grouted bolting systems so that optimum column
lengths for maximum load dissipation characteristics can be
determined, an investigation designed to evaluate the behavior of
full-column, resin-grouted bolts subjected to a range of end-loading
conditions was conducted. As discussed earlier, previous studies
investigated the influence of installation procedures on the
effectiveness of resin-grouted roof bolts. The results of these
laboratory and field investigations illustrated the necessity for
investigating the nonlinear behavior of the stress distribution in
grouted bolts. Therefore, this portion of the investigation
consisted of installing 25 full-column, resin-anchor roof bolts into
two concrete blocks. The resulting loads induced on the bolts were
then determined as they were subjected to a series of tensile loads
applied at the bolt heads. This method of loading has been shown to
be typical of how resin-grouted bolting systems are loaded. The
ultrasonic measurement system, used in this investigation, was
capable of determining the axial deformation of the bolt at any
location. With this system, no modifications were necessary to
accommodate strain gages or vibrating wires that can create zones of
induced stress concentrations; these stress concentrations can


23
produce erroneous test results, particularly at low stress levels.
By measuring the axial deformation of a bolt throughout its length,
calculations can be made to determine the amount of load induced at
these locations by the applied end-loading conditions. This
information can provide a clearer understanding as to the overall
loading characteristics of a full-column, resin-grouted bolt.
The results of this laboratory investigation will render a
more comprehensive understanding of the load transfer mechanisms
associated with fully grouted roof bolting systems and provide a
basis on which to choose adequate column lengths for maximum load
dissipation characteristics. These parameters become extremely
important as new combination systems, utilizing partial grout
columns, are introduced into the mining and construction industries.
Instrumentation Description
The success of the laboratory investigation depended largely
on the use of ultrasonic instrumentation to measure deformation of
the test bolts with a high degree of precision and accuracy. The
repeatability of the instrument had been previously determined, and
an instrument resolution of 0.0001 in had been achieved. Since the
bolts being analyzed required much less physical alteration than
traditional strain-gaged bolts, the decision to use ultrasonic
technology seemed to be a logical one. The ultrasonic instrument
used in the laboratory investigation was the Raymond PDX-934 Bolt
Gage (fig. 7). The method used by this instrument to determine
axial bolt deformation is quite straightforward (10). Basically,


24


25
the instrument transmits an ultrasonic sound wave through a bolt via
a signal transmitter/receiver transducer magnetically attached to
the bolt head. This sound wave is reflected off the far end of the
bolt and returned to the signal transducer. The instrument
determines the amount of time it took for this sound wave to travel
to the end of the bolt and back. By comparing the round-trip travel
time in the same bolt under stress, the change in length can be
determined. This determination is made by noting that the velocity
of sound in a stressed bolt is lower than in an unstressed bolt;
this difference in velocity is proportional to the change in length
of the bolt. The microprocessor in the instrument converts these
time measurements to a change-in-length reading viewed on the front
display panel. It should be noted that the instrument calculates
the length values by Hooke's Law, which assumes that the material
being tested remains linearly elastic. In short, once a
relationship between round-trip travel time increase and elongation
has been determined for a particular lot of bolts, the instrument
can be programmed with this calibration factor and then is ready to
display change in length of any bolt in the lot for any applied
loading condition.
Instrument Operation
The following instrument procedure was used for testing the
bolts in the laboratory investigation. First, the calibration
factor described in the previous section was determined and inputted
into the instrument. Next, a light coating of molybdenum grease was


26
applied to the head of the bolt being tested; this grease provided a
good conductive contact for the transducer/bolt interface. After
adjusting the instrument gain values to properly reflect the signal
off the desired location in the bolt, the bolt was ready for
analysis. Once the calibration factor was determined, the
instrument was ready for analysis of all bolts in both test blocks.
Experimental Procedure
Two separate tasks were necessary in the laboratory phase of
this study. First, two concrete blocks were constructed to
accommodate the bolts being tested. The second task involved
preparing the bolts for analysis. The investigation required that
the overall elongation of each bolt be determined for every loading
condition, and the elongation of intermediate points within each
bolt needed to be observed. Accordingly, small holes were drilled
into each bolt at various locations along its length to provide
additional reflector locations for the ultrasonic signal. The
instrument was capable of delineating between reflector locations by
making adjustments to the strength of the propagating sound wave.
Bolt Preparation
Machining
The bolts modified for testing were grade 40, no. 6 rebar,
resin-type roof bolts. Since the bolt manufacturer stamps the bolt
specifications on each bolt head, each head needed to be lathed
smooth in order to accommodate the ultrasonic signal transducer


27


28
(fig. 8). The foot of each bolt was also lathed smooth
in order to minimize test signal interference and to provide an
additional reflecting signal for the instrumentation.
Dri11inq
Each bolt had three holes drilled orthogonal to the bolt
axis at various locations along its length. Table 1 shows the
spacing pattern used for each bolt. Holes of diameter 0.04 inch
were drilled with a high-speed drill and jig assembly (fig. 9). The
spacing assured that every section of roof bolt length was
adequately represented.
Calibration
The bolts were calibrated with the ultrasonic instrument to
obtain baseline length data prior to installation in the two test
blocks. Since all of the bolts came from the same production lot, a
standardized bolt was constructed to determine the necessary signal
velocity properties required for input into the microprocesor of the
ultrasonic instrument. These values obtained from the standardized
bolt were used for all bolts installed in the two test blocks.
The construction of the standardized bolt consisted of
cutting the bolt to a prescribed length, determined by micrometer
measurements with an accuracy exceeding that of the ultrasonic
instrument. The velocity of the material could then be determined,
and a calibration factor formulated. The average cross-sectional
area of the bolt, also required for elongation calculations, was
also determined from the standardized bolt.


29
Table 1. Locations of Reflector Holes in Test Holes


Table 1. Continued


Table 1. Continued


32
Figure 9
High-speed drill and job assembly


33
Block Preparation
Construction
The two test blocks were constructed of monolithically
poured, 3,500-psi-strength concrete. The dimensions block were 3 ft
by 3 ft by 5 ft. To ensure sufficient concrete strength, the
concrete was allowed to cure for approximately 2 weeks before bolt-
hole drilling began.
Dri11inq
The bolt hole pattern used for both test blocks is shown in
figure 10. A minimum hole separation of 6 in was incorporated into
the hole pattern due to the size of the bolt-pulling apparatus. A
minimum distance of approximately 8 in was used for separation
between the outer bolt holes and the edge of the blocks, because
previous studies showed that if this separation is too small,
cracking of the blocks along their edges can occur as the bolts are
being pulled.
Bolt Installation
The resin used for bolt installation was a two part, single
tube, 60 sec spin time, 48 inch equivalent epoxy-type grout. No
bearing plate was used on the bolts, because the pull-test apparatus
could simply push on the flat concrete surface between the bolts.
The bolt installation equipment included a hand-held, air-
powered rotary bolting drill attached to a 10-ton forklift. The
bolts were installed horizontally so that the forklift could be used
to drive the bolts into the holes.


36 .
36.0
NOTE: All dimensions are In inches.
T
o
T
4
OJ
in
1~
04
in
f
OJ
in
f
l>4
in
4
_L
o
CD
TOP VIEW
Figure 10. Bolt hole pattern in test blocks.
CO
-P*
' 09


35
The procedure used to install the bolts into the test blocks
was simple. First, a bolt and its corresponding bolt hole were
prepared for installation. This included sliding a pull collar on
the bolt, and inserting a resin cartridge into the hole (fig. 11).
Second, the forklift with the rotary drill attached was positioned
in front of the hole (fig. 12). Third, the bolt was forced into the
hole by driving the forklift forward, and at the same time, the bolt
was spun with the rotary drill to begin mixing the resin. After the
bolt was emplaced and was spun for approximately 20 sec, the bolt
continued to be held in place for approximately 30 sec until the
resin was completely cured. The forklift was then retreated. This
procedure was repeated for all bolts in both blocks. The bolts were
allowed to cure for 1 week prior to testing to ensure maximum
anchorage capacity.
Laboratory Data Acquisition and Reduction
The data obtained from the test apparatus were in the form
of elongation measurements obtained from the ultrasonic instrument
and load readings taken from the dial gage attached to the hydraulic
bolt-pulling apparatus. Direct readings were taken from the
ultrasonic instrument and the load indicator (fig. 13). Each bolt
reflector location was individually monitored while load was
applied. The load was applied to each bolt from 1,000 lb to 8,000
lb in 1,000-lb increments; the applied load was monitored by using
the dial gauge attached to the hydraulic ram assembly (fig. 13). At
each loading condition, an elongation reading of a particular


36


37


38


39
reflector location was taken with the ultrasonic instrument. This
provided the change in length of a reflector location with respect
to the bolt head induced by a given applied load. The data obtained
in the lab were transferred to a computer file for reduction. A
spreadsheet program prepared the data for graphic display; a series
of load-vs-deformation graphs were developed using this program.
Analysis
The elongation data (appendix A), which indicated the change
in length of the distance between reflector locations and the bolt
heads for each applied loading condition, were assembled into a
series of graphs for analysis. Appendix B contains the graphs
generated for each loading condition.
Initial examination of the graphs indicated that most of the
deformation of the bolts occurred within the first 20 in of the bolt
heads. To analyze the data in detail, two assumptions were made.
First, the general trend of each graph was assumed to be bilinear,
as shown in figure 14. Second, the intersection of these lines
represented the "break point" or the point representing the location
where the data tended to drop off the initial linear trend.
The first line, named zone 1 for analysis and representing
the data from the 0-in to 15-in range, was calculated for each
loading condition by a linear regression analysis. The analysis did
not extend beyond the 15-in range so that the data representing the
zone where the break point existed were not included in the
analysis. The results from the linear regression analysis are shown
in figures 15 through 18. The results of this analysis, shown in


ELONGATION, In
40
CO
I
o
X
Figure 14. A bilinear graph illustrating the general data trend


ELONGATION, in ELONGATION, in
41
0.0024
0.0022
0.002
0.0018
0.0016
0.0014
0.0012
0.001
0.0008
0.0006
0.0004
0.0002
0
1 3 5 7 9 11 13 15 17
DISTANCE FROM BOLT HEAD, in
Figure 15. Linear regression analysis of 1,000 lb (bottom) and
2,000 lb (top) loading increments.


ELONGATION, in ELONGATION, in
42
DISTANCE FROM BOLT HEAD, in
Figure 16. Linear regression analysis of 3,000 lb (bottom) and
4,000 lb (top) loading increments.


ELONGATION, in ELONGATION, in
43
Figure 17. Linear regression analysis of 5,000 lb (bottom) and
6,000 lb (top) loading increments.


ELONGATION, in ELONGATION, in
44
Figure 18. Linear regression analysis of 7,000 lb (bottom) and
8,000 lb loading increments.


ELONGATION, in
45
Figure 19. Combines slopes of the calculated linear regression
1ines.


46
table 2, for each loading condition revealed excellent straight line
fits of the data. Also, as shown in figure 19, the slopes of the
calculated lines tended to increase, indicating that elastic
conditions were not exceeded in the pull tests.
Table 2. Coefficients of correlation for each loading
condition in zone 1 and zone 2.
Loading condition (lb) Zone 1 Zone 2
1,000 0.55 0.01
2,000 0.80 0.01
3,000 0.90 0.03
4,000 0.93 0.03
5,000 0.95 0.008
6,000 0.95 0.006
7,000 0.96 0.006
8,000 0.96 0.01
The second area, named zone 2 for analysis and representing
the data from the 25-in to 48-in range, was also calculated for each
loading condition by a linear regression analysis. The values for
the coefficient of correlation are also shown in table 2. Due to a
large scatter of the data in this range and to a significantly less
number of observations, the coefficients of correlation (R^) were
much less than those calculated for zone 1 of the graphs. The
scatter was most likely a result of minor variations in individual
bolt properties and installation procedures that resulted in each
bolt behaving slightly differently. These differences in behavior
resulted in each bolt possessing a slight difference in break point
location, and therefore scattering the data in zone 2. However, a
straight line fit seemed to be the most logical choice for these


47
data because the first lines of the graphs exhibited such an
excellent fit to a straight line, and the overall trend appeared to
be two intersecting straight lines. The data did not extend back
from the 15-in mark so that the data representing the zone where the
break point existed were not included in the analysis. The results
of the linear regression analysis are shown in figures 20 through
23. Note that contrary to the lines calculated for the first
portion of the graphs, the slopes of the lines in the 25-in to 48-in
range tended to remain constant (fig. 24), indicating that the axial
strain did not vary significantly for the different loading
conditions.
The break points were determined by calculating the
intersection of the calculated zone 1 and zone 2 lines of the
graph for each loading condition. Table 3 shows the values of the
two calculated lines for each loading condition. The break point
locations did not vary significantly as the applied loading was
increased, indicating that the mechanism by which the bolts
transferred the applied loads into the surrounding material was not
a function of the applied load.
Table 3. Intersection of zone 1 lines and zone 2 lines
Loading Condition (lb) Intersection point (in from bolt head)
1000 12.9
2000 15.2
3000 15.4
4000 15.6
5000 16.1
6000 16.0
7000 15.7
8000 15.3


ELONGATION, in ELONGATION, in
48
DISTANCE FROM BOLT HEAD, in
DISTANCE FROM BOLT HEAD, in
Figure 20. Linear regression analysis of 1,000 lb (bottom) and
2,000 lb (top) loading increments.


ELONGATION, in ELONGATION, in
49
Figure 21. Linear regression analysis of 3,000 lb (bottom) and
4,000 lb (top) loading increments.
ro


ELONGATION, in ELONGATION, in
50
DISTANCE FROM BOLT HEAD, in
Figure 22. Linear regression analysis of 5,000 lb (bottom) and
6,000 lb (top) loading increments.


ELONGATION, in ELONGATION, in
0.0095
0.009
0.0085
0.008
0.0075
0.007
0.0065
0.006
0.0055
1 1 1 1 B
_D

1 1 1 Ln
24 28 32 36 40 44 48 52
DISTANCE FROM BOLT HEAD, in
Figure 23. Linear regression analysis of 7,000 lb (bottom) and
8,000 lb (top) loading increments.


ELONGATION, in
52
0.009
0.008
0.007
0.006
0.005
0.004
0.003
24 28 32 36 40 44 48 52
DISTANCE FROM BOLT HEAD, in
8000 lb 1 ' 1 '
7000 lb
~ 6000 lb
5000 lb
~ 4000 lb
3000 lb
Figure 24. Combined slopes of calculated linear regression lines.


53
Based on these analyses, it is apparent that the test bolts
all exhibited similar deformations when applied to a range of end
loadings. This load transfer was approximately linear from 0-in to
15-in, indicating that the bolts were experiencing a constant strain
(and stress) in this range (table 4). As the applied loading was
increased, the stress in the bolts from 0-in to 15-in increased, but
still exhibited a constant stress throughout this range. The stress
in the 25-in to 48-in range of the bolts was also distributed
uniformly throughout this range, but as the applied loads were
increased, the apparent stress in this range did not appreciably
increase (table 4). This indicates that the magnitude of stress
produced in the 25-in to 48-in range of the bolts was much less
dependent on the applied load than was the 0-in to 15-in range of
the bolts.
Table 4. Axial bolt strains
Applied load (lb) Axial Strain (0 15 in) Axial Strain (25 to 48 in)
1000 6.125 X 10"5 3.263 X 10-6
2000 1.328 X 10'4 5.573 X 106
3000 2.072 X 10"4 1.052 X 10"5
4000 2.820 X 10"4 1.299 X 10"5
5000 3.504 X 10"4 7.961 X lO6
6000 4.191 X 10"4 7.889 X 10"6
7000 4.871 X 10"4 8.704 X 10"6
8000 5.595 X 104 1.347 X 105
The amount of load realized by the bolts at each loading
increment was calculated to determine the amount of actual
dissipation. This value would represent the amount of the applied


54
load that was transferred to the surrounding material at the point
of intersection of the zone 1 and zone 2 calculated lines. These
values were determined by utilizing Hooke's law and a modulus of
elasticity of 30 x 106 psi. Table 5 shows the values calculated
with the percent of load that has been transferred into the
surrounding material at each loading increment.
Table 5. Realized loads and percentage of applied loads.
Applied load (lb) Axial Strain (0-19 in) Zone 1 and 2 intersection pt Calculated load (lb) Pet of di ssipation
1,000 6.125xl0-5 12.9 63 93.7
2,000 1.328xl0-4 15.2 115 94.2
3,000 2.072xl0-4 15.4 178 94.1
4,000 2.820xl0-4 15.6 239 94.0
5,000 3.504X10"4 16.1 287 94.3
6,000 4.191xl0-4 16.0 346 94.2
7,000 4.871X10"4 15.7 410 94.2
8,000 5.595xl0-4 15.3 483 94.0
Laboratory Conclusions
A laboratory investigation designed to evaluate the
characteristics of fully grouted roof bolts subjected to various end
loadings was performed on twenty-five 48-in long grade 40 rebar
bolts using ultrasonic measurement technology. The results of this
investigation revealed that approximately 94 percent of the load-
induced deformations occurred in the first 15 in of the bolts. The
remaining portion of the bolts underwent only minor deformations.
This conclusion was based on a statistical analysis performed on 88
data locations located at various distances from the bolt heads.
The data show that the grout column allows the first 24 in of the


55
bolts to behave as if they were restrained only at the head and at
the 24-in level, with the grout acting as a deformation damper. The
24-in to 48-in portion of the bolts continued to deform, but at a
much lower magnitude.
The use of ultrasonic instrumentation allowed for the
nondestructive determination of elongations of the bolts; this
method did not require extensive bolt modification as is necessary
with traditional strain-gage installations. Ultrasonic
instrumentation, along with the standard pull test equipment, can
allow for nondestructive testing of any solid type of bolt in any
medium.


CHAPTER IV
FIELD INVESTIGATION
The field investigation was divided into two phases. The
first dealt with the instantaneous performance of resin-assisted
bolting. The second was concerned with the long-term monitoring of
an entire support system.
Instantaneous Rock Bolt Transfer Performance
The first experiment analyzed the instantaneous response of
a point-lock resin-grouted bolting system. The experiment was
conducted in the Galatia Coal Mine, located near Galatia, Illinois.
Eight point-lock resin-grouted bolting systems consisted of a
threaded rebar bolt inserted into a traditional 4-in expansion
anchor shell. The 6-ft-long bolts were anchored with a 24-in
equivalent length of a two-piece resin. A test site area was
selected that was two cross cuts outby from the working faces of the
coal mine section. The area had been previously supported with
similar 6-ft bolts spaced on 4-ft centers. Additionally, two
trusses had been installed, as basic mining practice, across the
center of the intersection. Eight previously prepared bolts were
installed among the existing support systems. The bolts were
randomly selected from 12 bolts that had been prepared and


57
calibrated in the laboratory. Reflector locations were drilled at
approximately the 4- and 10-in levels. The reflector holes, 0.04-
inch in diameter, were installed with the high-speed drill. Figure
25 shows the location of the existing supports and the experimental
roof bolts.
The bolts were installed with a roof bolt truss machine to
the pre-determined torque of 225 ft1b. Thirty minutes after
installation, the first set of measurements were taken using the
ultrasonic instrumentation. The amount of pre-load that was applied
to the bolts ranged between 6,042 and 11,025 lb. To convert this to
the torque values of 158 and 277 ft1b simply divide the load values
by 39.8. This is an approximate conversion, which was not deter-
mined for this specific bolt type, but for this type of investiga-
tion where relative values are of concern, it can be safely applied
(11)* The average load on the eight bolts was 8,771 lb or approxi-
mately 220 ft1b of torque had been applied to each bolt. The
difference between the applied torque and the measured torque can be
isolated in the amount of force lost due to the friction generated
between the bolt head, bearing plate, and unlubricated washer. The
torque tension ratio tests are described in Appendix C. Twenty-four
hours after installation, the bolt loads were again monitored with
the ultrasonic instrumentation. The minimum load was 3,923 lb (98
ft*lb) and the maximum bolt load was 11,036 lb (277 ft*lb). The
average load on the installed bolts was 5,875 lb or 148 ft1b. It
should be noted that at this time mining was still on-going in this


58

A 8
A2 A4
JOINT
TRUSS
sTRUSS


.1 .4 5

/
T/
*folNT
o
o
Figure 25. Existing supports and experimental roof bolt locations.


59
section approximately 150 ft from the test area. The immediate roof
of the test area appeared competent, with no visible cracks or
partings.
The subsequent loss of installation loads, approximately 29
pet, can be attributed to the relaxation of the threaded joint
surfaces along small high spots of contact surfaces. Another
parameter that may influence the way bolts tend to "bleed-off" can
best be described as embedment relaxation. When the first bolts are
installed, any separations or joints in the immediate roof area are
compressed. By further compressing this joint over a number of
bolts the first bolts tend to relax, which can cause unloading.
Bolt relaxation tends to take a number of hours as can be seen quite
clearly in this data set. The elastic interaction between the bolt
and strata depends on a number of variables, such as joint
stiffness, size of the joint or separation, and the distance between
the bolts. The possibility also exists that the resin and rock
interface slipped a small amount due to creep, which may have been
either mechanical or thermal. These reasons are only speculative
and require extensive amounts of research beyond the scope of this
investigation (12).
The bolts were monitored approximately 24 hours later or 48
hours after the initial installation. The mining unit had moved to
another area of the section, leaving the face approximately 200 ft
inby from the test area. Some of the bolts increased in load
approximately back to the initial loading conditions, while others
loaded substantially. The maximum and minimum values for the eight


60
bolts was 4,442 lb (111 ft*lb) and 12,327 lb (310 ft*lb),
respectively. The average load for the test bolts installed in the
area was 8,619 lb (216 ft1b). The immediate area showed signs of
loading in the form of cracks and joints, with small sloughages of
material falling out of the immediate roof. The pillars also
appeared to have taken load during this short time period, evidenced
by breakages at the corners and pillar line. This could have been
the result of abutment stresses falling back to the test area,
thereby redistributing the overburden and mining induced loads.
Additionally, a "day-lighted" joint formed diagonally across the
test area. Loading trends and patterns could not be correlated to
this joint, but it may have influenced the deformations, measured by
the ultrasonic equipment, by creating bending and shear motions in
the bolting systems.
The detailed map showing the locations of the existing
bolts, experimental bolts, and trusses also shows the location of
the joint that appeared after the second day (fig. 25). The actual
measurement data are provided in table 6. It should be noted that
the point-lock bolts that were tested had different cross-sectional
areas at distinct location along the length of the bolt. The
effective areas, determined in the laboratory, was approximately
0.447 in^ at the 4-in level and was 0.406 in^ at the 10-in level.
This may affect the way that the bolts initially load or determine
the zone at which yield or failure may occur.
The resin-assisted rock bolts showed instantaneous signs of
load loss when installed with an average load of 8,771 lb or


Table 6. Calculated initial loads, 24 hours and 72 hours after installation
Initial 24 hours 48 hours
Bolt Hole Load Torque Load Torque Load Torque
No. Location (lb) (ft lb) (lb) (ft lb) (lb) (ft lb)
n 4 Inch 10201 256 11036 277 11221 281
10 Inch 9221 232 7580 190 7711 194
n 4 Inch 9640 242 4422 111 4422 111
10 Inch 7709 194 5838 147 6386 160
#4 4 Inch 9945 249 3923 98 7572 190
10 Inch N.D. N.D. N.D. N.D. N.D. N.D.
#5 4 Inch 6042 152 5771 145 6222 152
10 Inch 5996 151 5601 141 12625 317
#6 4 Inch 11025 277 5729 144 12327 310
10 Inch 10035 252 16508 414 9840 247
#7 4 Inch 6276 158 4707 118 N.D. N.D.
10 Inch 11609 291 10980 276 11703 294
#8 4 Inch 6821 172 5377 135 8066 202
10 Inch 8727 219 6376 160 10258 258
#9 4 Inch 10245 247 6129 153 10611 266
10 Inch 12757 320 11248 282 12545 315
N.D. No Data


62
approximately 220 ft1b or torque. The dissipation of load during
the initial 24-hour period was attributed to resin-rock creep, bed
relaxation, and the redistribution of the localized stress field.
The area was subjected to loading during the next 24-hour period,
which prevented the monitoring of this trend.
Time Dependent Transfer Mechanics
A test site was installed to verify the long-term effects of
fully grouted bolts installed in an underground coal mine. The
support was installed during the regular mining cycle so that any
changes in load could be attributed to movements or shift in the
immediate roof area. The test areas that were monitored included a
main entry and an adjoining room. This ensured that the effects of
orientation could be accounted for when the data analysis was
undertaken. Additionally, it was felt that the ultrasonic
instrumentation should be tested during an extended period of time
to assure that repetitive and accurate readings could be obtained.
This also provided an opportunity to examine the transfer mechanics
of resin-grouted bolts over an extended period of time (13). To
verify and compare the roof bolt loads, additional instrumentation
was installed in the test area.
Instrumentation
Three types of instrumentation were used in this
investigation to measure the bolt loads and roof movements.
Compression pads and strain bolts were used directly to measure
loads applied to the bolts after installation. Vertical


63
displacement gages were installed to measure differential roof
displacements during test site monitoring and support operation.
Additionally, observation holes were drilled in each test area to
monitor with a fiber-optic strata-scope the locations and widths of
roof separations.
Each compression pad (fig. 26) consists of a rubber membrane
placed between two steel plates. The compression pads have a
working load limit of 32,000 lb with a calculated accuracy of 200
lb. Readings of the compression pads are monitored with a special
calibrated ring that measures the change in the circumference of the
rubber membrane as it loads and unloads. Laboratory tests on the
ring indicate that when loads exceed 30,000 lb, the accuracy drops
to 1,000 lb. The nature of the pad is to act as a spring between
the bolt head and the roof or wall. In laboratory investigations,
the pad, after being subjected to high loads, failed to rebound to
its specified unloaded circumference. The rubber tends to
permanently deform after continued loading. In this investigation,
positive and negative loads were generated on the pads. However,
the loads were low by comparison, and no permanent deformations were
felt to have occurred.
The vertical-displacement gage (fig 27) consists of four
spring clips used to anchor high-strength, stainless steel
prestretched wire at selected depths in a 1-3/8-in-diam hole drilled
in the roof. The uppermost spring clip is placed in a stable layer
to be used as a base reference for measured displacements. For this
investigation, a hole depth of 7 ft 2 in was used. The remaining


64
8.00"diamr
|l.50"diam|*
Steel-
lo.85"l
diam.
7.00"diam.
Figure 26. Compression pad diagram.


65
Figure 27. Vertical-displacement gage diagram.


66
three clips are placed 5 ft, 2 ft, and 1 ft away from the immediate
roof. The wires from the four spring clips run through a 10-in-long
tube anchored in the collar of the drill hole. The wires go through
numbered holes in the copper cap on the end of the tube and have
small brass fittings that are used as reference points. A loop is
made at the end of the wires so that a 3-lb weight can be attached
to maintain a constant tension on the wire while readings are being
taken. Readings are made with a dial indicator placed between the
cap and the reference point on each wire.
Bolts that were being used to support the roof area were
obtained from the mine and modified to accommodate ultrasonic
measurement readings. The locations for the reflector signals were
located at approximately the 2-in and 6-in levels. The material
velocity for the bolts was predetermined in the laboratory to ensure
expected accuracies of 150 lb. The type of bolt being used in the
investigation was a type 40, 3/4-inch-diam, 6-ft-long tendon
installed in a 1-in-diameter hole. The initial reflector signals
and subsequent readings are shown in table 7.
The test site configuration and instrumentation locations
are shown in figure 28. The height of the openings varied from 6.5
to 8 ft. The approximate overburden in the test area was 350 ft.
The selected mining configuration was the development of 20-ft-wide
openings on 80-ft-centers. This left a final pillar dimension of 60
by 60 ft. The instrumentation was read and evaluated five times
over a 238-day period.


67
Table 7. Initial Ultrasonic Signal Locations and Subsequent
Lengths.
Bolt # Initial Distance 30 days 70 day 140 days 238 days
No. 1 2.0470 5.9958 2.0480 5.9969 2.0457 5.9960 2.0476 5.9963 2.0473 5.9971
No. 7 1.9938 6.0048 1.9950 6.0060 1.9943 6.0050 1.9941 6.0055 1.9942 6.0061
No. 8 2.0318 6.0136 2.0331 6.0150 2.0320 6.0137 2.0322 6.0145 2.0323 6.0156
No. 10 1.9928 6.0368 1.9941 6.0380 1.9936 6.0371 1.9935 6.0373 1.9933 6.0375
No. 9 1.9806 5.9905 1.9811 5.9912 1.9808 5.9907 1.9813 5.9913 1.9813 5.9909
No. 2 2.0758 6.0361 2.0765 6.0367 2.0761 6.0364 2.0766 6.0368 2.0764 6.0366
No. 12 1.9740 6.0066 1.9750 6.0079 1.9744 6.0075 1.9749 6.0077 1.9747 6.0072
No. 4 1.8498 6.0459 1.8509 6.0474 1.8505 6.0471 1.8508 6.0473 1.8505 6.0468


68
1

I
N



Hi
1
6+40
13 15 16
14
lo 8
o 1 7

O

9 12
lo 11
8 9 4
7 2 3
O A O
6 12 2
5 4 1
Pressure pads
Stress bolts
Observation holes
Sag station
Test room 2
Test
room 1
1 Right
Figure 28. Test site configuration and instrumentation locations.


69
The initial loads applied to the bolts varied slightly from
bolt to bolt in both of the test areas. The load is generated by
the upward thrust of the bolter after adequately mixing the resin
but prior to hardening. A load isopleth, shown in figure 29, was
generated to show the loading trends immediately after installation.
In test room 1, the minimum load was 1,025 lbs and the maximum load
was 4,830 lb. The average for the test bolts in room 1 was 2,655
lb. This was slightly lower than the loads generated in test room
no. 2, where the minimum was 2,289 lb and the maximum was 5,097 lb.
The average loads generated on the bearing plates in test room
number 2 was 3,676 lb.
Thirty days after the test area installation, the
instrumentation was again read and evaluated. The loads generated
by the initial force of the roof bolting machine had dissipated when
the roof began dilating. This type of dilation results from
abutment stresses dissipating and transferring forward as the
working faces are advanced during the mining cycles. The average
load in test room 1 had dropped to 1,353 lb, and the average load in
test room 2 dropped to 2,389 lb. To illustrate these changes, load
contours were drawn that illustrate the 30-day loads and the 30-day
changes. These changes are obtained by subtracting the initial
loads from the measured loads, thus negative values are shown. The
trend for the roof dilation can clearly be seen as occurring in the
middle of test room 1 and varying slightly in test room no. 2. The
measurements obtained with the ultrasonic instrumentation also
showed a reduced amount of load during this period. However, the


70
Test
room 1
1 Right
Figure 29. Initial instrumentation load isopleth.


71
initial loads were higher, and the subsequent loads remained larger
than those recorded with the pressure pads. The loads generated at
the 2-in level were approximately 50 pet higher than those
determined at the 6-in level. This would be expected when examined
with the analytical equations developed during the laboratory
investigation. Figures 30 and 31 show the 30-day load and change in
load contours that were developed from the data. The displacements
recorded with the differential sag station also indicated that a
larger amount of vertical displacement in the upward direction had
taken place at the lowest location, 1 ft from the opening roof. The
immediate roof was composed of a laminated shale, and shrinkage due
to moisture lost probably occurred. This would explain additional
load losses measured by the test site instrumentation. The
differential sag station plot is shown in figure 32. Station 1
through 4 are located at 7 ft 10 in, 5 ft, 2 ft, and 1 ft,
respectively.
The instrumentation was read and evaluated 70 days after
installation. The average loads in test room no. 1 dropped to 354
lb. Five of the bolts had no measured loads. The pads in test room
no. 2 averaged 3,277 lb. This represents an increase in the average
load of 888 lb. The 70-day load and change in load contours are
shown in figures 33 and 34. The same general trend continued in
test room no. 1, with unloading occurring due to shale shrinkage and
abutment stress relaxations. The increase in the average load in
test room no. 2 can be attributed to an increase in load in seven of
the bolts and a small decrease in only one of the bolts. The


72
Test
room 1
1 Right
Figure 30. Test site 30-day load contours


73

6M0
H
o o o o
o o o
0 o o
CM CM CM
1 I I
Test roon 2
KEY
Pressure pads
Stress bolts
Observation holes
Sag station
Unloading pounds
Loading pounds
Test
room 1
1 Right
Figure 31. Test site 30-day change in load contours.


DISPLACEMENT. In
74
Figure 32. Differential sag station plot.


75
Test
room 1
1 Right
Figure 33. Test site 70-day load contours.


76
Test room 2
Test
room 1
1 Right
Figure 34. Test site 70-day change in load contours.


77
pillars in this area were experiencing small amounts of rib
sloughage which increased the effective room span dimension, thus
increasing the loads in the test area. The ultrasonically measured
bolts agreed well with the pressure pad data. The average of the
load recorded at the 2-in level in test room no. 1 was 2,739 lb, and
in test room no. 2 the average measured load was 3,297 lb. The same
general trends in the amount of load transfer was measured in the
majority of the bolts. Tables 8 and 9 show the pressure pad load
data information and the ultrasonic bolt load data obtained during
the entire investigation.
The instrumentation was monitored and evaluated 140 days
after the initial installation or 70 days since the last set of
measurements. The loads fluctuated throughout both test areas. The
loads that increased were related to the rib sloughages and slabbing
Table 8. Pressure Pad Load Data (lb).
TEST ROOM 1
Days No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8
Init 4027 2849 1785 2545 4830 2849 1329 1025
30 2089 2849 569 1177 3001 721 417 0
70 1481 0 0 0 0 265 0 1085
140 1177 1085 1353 1754 2422 0 0 1058
238 265 952 1 1754 2022 721 0 0 951
TEST ROOM 2
Days No. 9 No.10 No.11 No.12 No.13 No.14 No.15 No.16
Init 3894 5097 2957 2289 3359 3492 3359 4963
30 1937 5365 1785 1177 873 2241 3492 2241
70 3091 6033 3492 2422 2289 1633 3760 3492
140 3492 6033 2545 2422 2422 1633 4027 3626
238 1329 6301 2241 2422 873 1481 3001 1481


78
Table 9. Ultrasonic Bolt Load Data (lb).
TEST ROOM 1
Days No. 9 No. 2 No. 12 No. 4
Approximately 2-in level
30 3346 4469 6714 7881
70 1338 1915 2686 5015
140 4684 5108 6042 7165
238 4684 3831 4700 5015
Approximately 6-in level
30 1549 1317 2868 3288
70 442 659 1986 2631
140 1770 1537 2427 3069
238 885 1098 1324 1973
TEST ROOM 2
Days No. 1 No. 7 No. 8 No. 10
Approximately 2-in level
30 6474 7977 8480 8646
70 3237 3324 1305 5320
140 3885 1994 2609 4655
238 1942 2659 3261 3325
Approximately 6-in level
30 2431 2649 3085 2634
70 442 441 220 659
140 1105 1545 1983 1098
238 2874 2869 4408 1537


79
that the two areas were undergoing. This was attributed to a larger
roof area being supported by the bolts located near the edge of the
pillars. The loads determined with the ultrasonic instrumentation
increased slightly in both areas. The average loads, determined at
the 2-in level from the ultrasonic bolts in test room no. 1 was
5,750 lb. The average load, calculated from the same level in test
room no. 2, was 3,286 lb. The average load, determined from the
pressure pad instrumentation in test room no. 1, was 1,106 lb, while
the average load in room no. 2 was 3,275 lb. The substantial
difference or conflicting information acquired in test room no. 1
exemplifies the complexities of making underground load
determinations on rock bolts. The transfer mechanics from bolt to
bolt vary considerably when the physical properties of the material,
or the initial bolt loads, due to thrust, vary even slightly.
Additionally, the loads measured by the pressure pads are loads
generated directly on the bearing plate of the bolt. This can be
related to the loads that occur along the length of the bolt, but
may vary significantly if shear is introduced in less than ideal
conditions. Figures 35 and 36 show the 140-day loads and the
changes in load. Referencing back to the differential sag station
data (fig. 32), it should be noted that at this time the roof
appears to have settled back to the normal roof position.
The final time the test site instrumentation was monitored
and evaluated occurred 238 days after the initial installation.
During this time the trends in loading decreased slightly or
remained constant. At this time it was felt that the load transfer
mechanics in this area had been completed and subsequent unloading


80
Test
room 1
1 Right
Figure 35. Test site 140-day load contours.


81
I
1
KEY
Pressure pads
Stress bolts
Observation holes
Sag station
Unloading pounds
Loading pounds
Test room 2
Test
room 1
1 Right
Figure 36. Test site 140-day change in load contours.


82
and loading would be mainly due to roof dilations with seasonal
changes or pillar spalling. The average load determined from the
pressure pad data in test room no. 1 was 833 lb. Two of the
instruments recorded no load on the bearing plates. The ultrasonic
measurements obtained from the 2-in and 6-in levels averaged 4,558
and 1,320 lb, respectively. The load transfer dissipation in this
short span (4-in) showed a reduction in load of approximately 71
pet. The average load determined from the pressure pad data in test
room no. 2 was 2,391 lb. The ultrasonic loads determined at the 2-
in and 6-in levels averaged 2,797 and 2,922 lb, respectively. This
difference is evidenced in the fact that ultrasonic bolts no. 7 and
no. 8 had higher loads at the 6-in level than at the 2-in level.
The resin grout may have become unbonded, or shear may have been
induced into the bolt tendon as the area shifted under varying
degrees of stress and loading conditions. Figures 37 and 38 show
the 238-day loads and changes in load. It is evident that the loads
dropped on the test instrumentation during the 238 days that the
rooms were monitored. This was primarily due to the dilation that
the roof was being subjected to during this period of time. This
type of phenomenon makes data interpretation difficult and
speculative (14). Field conditions dictate that trends or patterns
be analyzed to make practical evaluations and recommendations. The
transfer mechanics of these full-column, resin-grouted rock bolts,
evidenced by the pressure pads and ultrasonic bolts, appear to
follow the same trend that was definitely shown in the laboratory.
Load dissipation transfers rapidly back into the rock mass in a very
short interval. The field test was performed in conjunction with


83
Test
room 1
1 Right
Figure 37. Test site 238-day load contours.


84
.v
6+40
-2000
-1000
+1000
0\

-2000
-3000
3000
2000
1000
Pressure pads
Stress bolts
Observation holes
Sag station
Unloading pounds
Loading pounds
Test
room 1
1 Right
Test room 2
Figure 38. Test site 238-day change in load contours.


85
the laboratory study. If the results from the laboratory study had
been concluded, an additional signal location would have been
provided above the 16-in level to determine if all the load had
dissipated into the rock mass. The isopleths provide a visual
summary of the loading and unloading conditions of the underground
test areas throughout the field investigation. To compare the load
fluctuations of the test instrumentation over the same period of
time, graphs are included that show the loads for the pressure pads
and the ultrasonic bolts. These figures, 39 through 42, start at
the bottom of test area in room no. 1 and progress inby and then
left into room no. 2. The general trends of load compare favorably
for both types of instrumentation in each respective zone. The
level of accuracy for each type of instrument affects the final load
levels.
The long-term field investigation exemplifies the problems
that are associated with obtaining and interpreting geotechnical
data. The information or data obtained is correct, but
interpretation is difficult. However, the transfer mechanics of the
resin-grouted bolts appear to coincide with the information obtained
in the laboratory and the instantaneous experiments conducted in the
first field site. The load dissipation or transfer mechanics
appears to have occurred in the first portion of the rock bolt
tendon. This information becomes important as shorter grout lengths
are being utilized to reduce support cost and increase production in
mining and geotechnical applications.


LOAD, lbs x 10 LOAD, lbs
86
0 20 40 60 80 100 120 140 160 180 200 220 240
TIME, days
Figure 39. Test room no. 1 pressure pad loads.


LOAD, lbs x 10 LOAD, lbs
87
TIME, days
Figure 40. Test room no. 1 ultrasonic bolt loads at the 2-in
(bottom) and 6-in (top) levels.


LOAD, lbs x 10 LOAD, lbs
88
TIME, days
Figure 41. Test room no. 2 pressure pad loads.


89
TIME, days
TIME, days
Figure 42. Test room no. 2 ultrasonic bolt loads at the 2-in
(bottom) and 6-in (top) levels.


90
Conclusions and Recommendations
The objective of this study was to analyze the transfer
mechanics in resin-grouted rock bolts. The investigation was
conducted in the laboratory and the field with the aid of ultrasonic
measurement systems.
The laboratory investigation designed to evaluate the
characteristics of fully grouted bolts and the appropriate transfer
mechanics was performed by subjecting rock bolt tendons to various
end loadings. The experiment was performed on twenty-five 48-in-
long, grade 40 rebar bolts using ultrasonic measurement technology.
The result of this investigation revealed that approximately 94 pet
of the load induced deformations occurred within the first 15 in of
the bolts. The remaining portion of the bolts underwent only minor
deformations. These conclusions were based on statistical analysis
performed on 88 data locations located at various distances from the
bolt heads.
The field evaluation of bolt load transfer mechanics was
investigated in two phases. Phase I examined the instantaneous
transfer mechanics of a resin assisted 6-ft-long bolts anchored with
a 24-in equivalent length of two piece resin at the Galatia Mine
located in Galatia, Illinois. The results indicated that a
reduction in active force or torque took place 24 hours after
installation. This reduction in torque or load averaged
approximately 29 pet. This may have been the result of the bolting
system or embedment relaxation. However, the 24-in length of resin
grout may have been inadequate to resist the initial state of stress


91
to which the bolt was subjected. Movements in the underground test
area prevented verification of this hypothesis. The length of the
required column in resin-assisted systems is an important parameter
as this method becomes more popular because of the economic
benefits. Site-specific evaluations of this type should be
performed to ensure adequate anchorage.
Phase II of the field evaluation included the installation
of two small-scale test sites in an underground coal mine to examine
the time-dependent transfer mechanics in full-column, resin-grouted
rock bolts. The trends and patterns indicated by the
instrumentation showed that fluctuations in load can be expected to
occur in underground excavations as a function of stress relief and
cyclic humidity. The loads, which were low when compared to the
ultimate strength of the bolting systems, corresponded to the roof
movements and pillar/rib sloughages occurring in the immediate area.
The load dissipation characteristics, measured by the ultrasonic
bolts, occurred rapidly along the length of the rock bolt. The
differences in loads between the 2-in and 6-in levels showed this
rapid reduction in applied loads.
The ultrasonic instruments proved to be a valuable
measurement method for investigating the transfer mechanics and load
conditions for rock bolts. The data, obtained throughout the
investigation, provided information that was not obtainable prior to
this measurement technique. Further research efforts should
concentrate on the interpretation of measured loads and stress
conditions under field conditions.