RP001AECL: Microseismic
Monitoring and Excavation Damage Assessment at Atomic Energy of Canada
Limited's Underground Research Laboratory. R. P. Young and S. Talebi,
September 1988
The purpose of this project
was to seismically characterize the damage induced by the excavation process
to Atomic Energy of Canada's Underground Research Laboratory during the
shaft extension phase (300 m. - 420 m. level). A microseismic whole waveform
monitoring system was installed at this site in December 1987. Hydrophones
and 3D accelerometers (up to 5 kHz) placed in 4 inclined boreholes at the
300 m. level, were used to sense microseismic waves radiated due to crack
propagation in the rock mass.
An increased level of activity
was observed for a period of about 2 hours after the blasts. P and S wave
arrival times of signals were used to source locate the activity. The results
show that crack propagation occurs mainly around the bottom and the walls
of the shaft near the newly created faces. In the horizontal plane, the
activity is located mainly within 1 to 2 meters of the walls of the shaft.
This area should correspond to where most of the damage due to the excavation
occurs. An interesting feature is that the events cluster along a preferentially
NE-SW direction, compatible with observations of overbreaks and stress
measurements in the area. Preliminary results relative to source mechanism
characterisation were also presented.
RP002AECL: Manual of Source
Locations of Microseismic Events Induced by Shaft Extension at the Underground
Research Laboratory. R.P. Young and S. Talebi, March 1989
This report is the second
in relation to Contract #AECL/WS/293/49l75 entitled "Microseismic Monitoring
and Excavation Damage Assessment at Atomic Energy of Canada Limited's Underground
Research Laboratory" between Atomic Energy of Canada Ltd. and Queen's University.
The report is a manual of source locations of microseismic events induced
by extension of the shaft between 324 m. and 443 m. of depth.
Further to the submission
of the first report (Young and Talebi, 1988) on the same subject to AECL
in September 1988 and a seminar presentation at URL in November 1988, some
interest was manifested by the personnel at AECL in P and S wave velocities
used for source location of the events. As mentioned in the previous report,
source location algorithms use a minimum of 4 P and S arrival times to
determine the co-ordinates x, y and z of an event and the time when the
event occurred. It is well known that the accuracy of the solution depends
on several factors, one of which is the accuracy of the values used for
P and S wave velocities. The velocities we used in the first report were
6,25 km/s for P wave and 3,6 km/s for S wave, based on the results obtained
by Wong et al (1983) during a tomography survey at URL. Indeed, the paper
by Wong et al (1983) was the only piece of information we had available
at the time of the processing of the data.
A memo was provided by AECL
on velocity measurements carried out at URL, Room 205 (Hayles and Stevens,
1988). The results for three types of raypaths (horizontal, inclined
at 45° and subvertical) could be summarised as follows:
for P wave Horizontal 5535
+/- 65 m/s
45 degrees 5626 +/-
80 m/s
subvertical 5896 +/
-65 m/s
for S wave Horizontal 3443
+/- 58 m/s
45 degrees 3270 +/-
30 m/s
subvertical 3435 +/-
26 m/s
These results suggest the
presence of an anisotropic velocity field at the URL. They are, otherwise,
lower than those of Wong et al (1983). This apparent discrepancy could
be due to several factors. The area of the rock mass and the range of distances
used by Hayles and stevens (1988) are respectively closer to underground
walls and shorter than those used by Wong et al (1983). A certain scale
effect and/or influence of a destressed area affecting the results of Hayles
and Stevens (1988) cannot be excluded.
Taking all these facts into
account, we decided, however, to use the results of Hayles and Stevens
(1988) in the manual. The velocities used for activity due to blast #49
to blast #70 (324 m. to 370 m. of depth) are those obtained for 45 degree
rays. The velocities obtained for subvertical angles were used to source
locate the activity due to blast #71 to blast #105 (370 m. to 443 m. of
depth). 3D view, plan view and vertical section of source location after
each blast have been presented. Due to the large number of events following
blast #96, it was decided to divide the time period after this blast into
5 shorter periods to get a better presentation. Otherwise, following the
same methodology as that used in the first report, 3D view, plan view and
vertical section of source locations are also presented week by week.
The source locations obtained
using the results of Hayles and Stevens (1988) are slightly different from
those obtained in the first report: discrepancies of a few meters have
been observed in some cases. This is perfectly understandab1e since the
velocities used in the two cases are in some cases different by more than
10%. However, the conclusions of our first report concerning the extent
of the damaged area around the shaft and preferential NE-SW orientation
of source location in some cases are still valid. It seems appropriate
to state that the knowledge of the true in situ velocities in the area
around the shaft is required. Only direct measurement of velocities in
that area could resolve the uncertainties associated with the accuracy
of source locations.
A velocity survey around
the shaft has been planned in the new contract and will take place in March
1989. It is planned to use a series of 1 m. deep boreholes drilled at 3
meter intervals around the shaft in different directions. The results of
this survey will provide the most appropriate set of velocity measurements
to be used in the source location of microseismic events since the raypaths
would be close to those of microseismic waves radiated by the events. The
source locations calculated as a result of this survey will form the basis
of a further report.
RP003AECL: Final Report
On Source Location of Microseismic Events Induced by Shaft Excavation at
the Underground Research Laboratory. S. Talebi and R. P. Young, September
1989
The purpose of this research
project was to seismically characterize the damage induced by the excavation
process to Atomic Energy of Canada's Underground Research Laboratory (URL)
during the shaft extension phase (324 m. - 443 m. of depth). A microseismic
whole waveform monitoring system was installed at this site in December
1987. Hydrophones and 3D accelerometers (up to 5 kHz) placed in 4 inclined
boreholes at the 300 m. level, were used to sense microseismic waves radiated
due to crack propagation in the rock mass.
The microseismic data-set
recorded during shaft extension was source located using P and S wave velocities
from the velocity survey along the shaft carried out in March 1989. The
events are located within a distance of 10 m from the bottom of the shaft
in the vertical direction. The mean radial distance of the events from
the centre of the shaft is 4.28 m (the radius of the shaft is 2.3 m). This
parameter shows an increase with depth below 370 m. Statistical methods
were used to quantitatively analyse the preferential NE-SW orientation
of the events. A clear clustering is observed between N20°E and N60°E.
RP004AECL: In Situ Measurements
of P & S Wave Velocities at The Underground Research Laboratory.
S. Talebi and R. P. Young, September 1989
A velocity survey was carried
out along the shaft at URL using 32 horizontal shallow boreholes drilled
between the 300 m level and the 420 m level for shot locations. The P and
S wave velocities and elastic moduli obtained are in the following range:
P wave velocity:
5.618 km/s to 5.877 km/s
S wave velocity:
3.270 km/s to 3.405 km/s
Dynamic Poisson's ratio:
0.225 to 0.249
Dynamic Young's modulus:
71.2 GPa to 77.5 GPa
A mean value of 5.763 km/s
for P wave velocity and 3.376 km/s for S wave velocity is obtained for
the whole data set. A mean value of 0.244 for the dynamic Poisson's ratio
is obtained using a Wadati diagram. In general, the results show lower
velocities for sub-horizontal and low-angle raypaths and higher velocities
for steep-angle raypaths.
RP005AECL: Source Mechanism
Characterisation of Microseismic Events Induced by Shaft Excavation at
the Underground Research Laboratory. S. Talebi, R.P. Young, D.J.
Rawlence and S.J. Gibowicz, July 1990
This report describes an
investigation of source mechanism and source parameters for the microseismic
events induced by shaft sinking at the URL between 324 m and 443 m of depth
during the period January to August 1988. A S/P wave amplitude ratio
scheme was used for focal mechanism determinations with a two-stage grid
search procedure. Satisfactory results were obtained on synthetic data
with coarse and fine mesh sizes of 5° and 1° respectively. Signals
recorded from production blasts during shaft extension and those from blasting
caps of the velocity survey along the shaft (March 1989) were used to calibrate
the sensors. From an initial 340 events meeting the cut-off criteria, the
algorithm converged to a minimum in 288 cases with no mismatch in P-wave
first motions. The results from 95 solutions with misfit values less than
1 were used for further analysis. These events were divided into three
sub-sets with regard to the bottom of the shaft for each production blast:
top, middle and bottom zones. A very clear decoupling was observed between
the results of the top and bottom zones. The agreement between the P, T
and B axes of focal mechanisms and the orientations of the principal stress
components at URL is best for the top zone sub-set. The results of source
parameter determinations for the same 95 events could be summarised as
follows:
P-wave corner frequency:
1100 Hz to 4300 Hz
S-wave corner frequency:
960 Hz to 3600 Hz
Moment magnitude: -3.6 to
-1.9
Seismic moment: 5103 Nm
to 1.5 106 Nm
Seismic energy: 0.0033 J
to 70 J
Source radius: 0.3 m to
0.7 m
Apparent stress: 0.02 MPa
to 2.4 MPa
Stress drop: 0.04 MPa to
3.0 MPa
The ratio of S-wave energy
to P-wave energy ranges from 1.5 to 70; for one third of events this ratio
is smaller than 10. These events have also low apparent stresses and are
probably associated with some component of tensile failure at the source.
However, shear failure is, in most cases, the dominant mode of failure.
The events associated with the development of the 420 m level have, in
general, larger seismic moments than those caused by the shaft extension.
this suggests that the seismic response of the rock mass, particularly
the magnitude of the events, depends on the direction of excavation. Larger
events tend to occur deeper along the shaft and are more likely to result
from a pure shear failure, in agreement with increasing stress levels with
depth.
RP005aAECL: Borehole Breakouts
a Review of the Theory and a Discussion of their Investigation Using Seismic
Methods. D.J. Rawlence and R. P. Young, January 1990
Borehole breakout and sidewall
failure in highly stressed tunnels are of concern in the stability of boreholes
and the design of safe underground openings. These phenomena are the result
of localised stress concentrations causing spalling or shear failure of
the brittle rock. The purpose of this report is to review the current literature
on the theory of breakout formation and discuss how techniques from seismology
can be used for their investigation.
RP006AECL: Design of a
Microseismic System for the Underground Research Laboratory Mine-by Experiment.
S. Talebi and R.P. Young, October 1990
This report is the first
in relation to Contract WS-30J-54474 between Atomic Energy of Canada Limited
(AECL) and Queen’s University. The report describes a microseismic
monitoring system designed for the mine-by experiment at the 420 m level
of the Underground Research Laboratory (URL), Pinawa, Manitoba.
The report first provides
a summary of the main findings of the shaft extension monitoring project.
Then the guidelines for the design of a microseismic monitoring system
is analysed: the objectives of the experiment, the type and number of sensors
to be used and their coupling to the rock mass, the geometry of the array,
the expected microseismic activity, followed by an examination of the sources
of inaccuracy in source location determinations. The proposed microseismic
system is then described: the array of sensors, the acquisition systems
and the schedule of costs. A discussion on the proposed systems and
the expected results concludes the report.
RP007AECL: Seismic and
Radar Characterisation at the 240 M Level Of The Underground Research Laboratory.
S. Talebi and R.P. Young, November 1990
Following an agreement reached
with the geophysics division of the Whiteshell Nuclear Establishment, radar
and tomographic surveys were carried out in March and April 1989 at the
URL, 240 m level in a pillar between room 207, 208 and 209. This report
summarises the results of this experiment, the processing and the interpretation
of the signals.
The recorded signals had
low noise levels and clear onsets of P waves, allowing for a very accurate
determination of arrival times. Some secondary arrivals were observed on
some signals. A back projection image was produced and used as a starting
point for a simultaneous iterative reconstruction technique routine. The
seismic and radar tomograms show generally higher velocities in the middle
of the pillar and lower velocities close to the edges of the pillar. However,
very little velocity contrast in the image could be attributed to the internal
structural properties of the pillar. Evidence of the influence of anisotropy
on the results was observable, particularly in the case of radar tomograms.
Seismic velocities showed about 6% anisotropy with a maximum velocity oriented
at N56°E (for radar, these values are respectively 13.5% and N78°E).
This is in reasonable agreement with the orientation of the vertical joints
and the maximum principal stress at the 240 m level of the URL.
It was also observed, during
this experiment, that the seismic waves induced by detonating blasting
caps in the surveyed area at the 240 m level were detectable at the sensors
of the microseismic system at the 300 m level. Thus these signals were
recorded on the microseismic system at URL and the data set is available
for any future investigation of the properties of the fracture zone #2
through which the rays have passed.
RP008AECL: Ultrasonic
Imaging and Acoustic Emission Monitoring of Laboratory Hydraulic Fracturing
Experiments in Lac du Bonnet Grey Granite from AECL's Underground Research
Laboratory. T. Chow, S.D. Falls, S.R. Carlson, and R.P. Young, November
1990
Ultrasonic tomography and
acoustic emission data were obtained during laboratory hydraulic fracturing
tests on two large cylinders of Lac du Bonnet grey granite. Compressional
velocities were found to be anisotropic, with the in situ vertical direction
being the most rapid direction in both samples. The velocity anisotropy
is related to the rock's pre-existing microcrack fabric. Compressional
velocities rose over the course of the experiment due to radial penetration
of fluid into the rock. A regression analysis showed that the velocity
changes can be explained by variations in crack density, inferred from
initial velocities, and radial distance from the borehole. The O'Connell
and Budiansky (1974) model was then used to calculate saturation levels
consistent with the observed velocity changes. Acoustic emissions reoccurred
in a few distinct zones over several pressurisation cycles. The AE locations
allowed two distinct fracture planes to be sharply delineated in one sample.
AE source mechanisms showed that shear interactions were abundant at the
microcrack level even though the samples failed macroscopically in tension.
RP009AECL: Analysis of
the Microseismicity Induced by the 420 Level Development at the Underground
Research Laboratory. S. Talebi, B. Feignier and R.P. Young, April
1991
The excavation of the incline
and decline galleries and room 405 at the 420m level of the Underground
Research Laboratory was monitored during the period June-September 1990
using the shaft extension microseismic system. A total of 63 blasts were
recorded. During the excavation, V-shaped notches have been observed mainly
in the Northeast sections of the tunnels which are perpendicular to the
orientation of the maximum principal in situ stress component. The objective
of the present study was to investigate any fundamental differences between
the characteristics of the two microseismic data sets originating from
the two orthogonal sections of the tunnels.
In a preliminary analysis,
two basic parameters were used: the number of events recorded within a
certain time period following each blast and the rate of the activity.
Histograms of the number of events after the blasts showed compatible features
to those observed for the shaft excavation data set: the time period immediately
after the blasts revealed to be one of increased activity, followed by
a rather exponential decay of the number of recorded events. No apparent
distinction, however, could be made between the blasts in Northeast and
Northwest galleries based on the number of events within a certain time
period or histograms of activity after the blasts.
Source parameter determination
was performed for two representative data sets recorded following two blasts
in the incline gallery: one in the Northeast section and the other in the
Northwest section. The results revealed interesting features: the intensity
of the events, estimated from their seismic moment and moment magnitude,
is about 2 times larger for the events following the blast in the Northeast
section. A similar ratio exists between the stress drop estimations while
source radii are very similar. The release of seismic energy in the first
case is of a higher level and more sudden than for the second case. These
results are in agreement with in situ observations of damage during the
excavation.
RP010AECL: Calibration
of the Queen’s Mine-by Microseismic System at the Underground Research
Laboratory. A.J. Feustel and R.P. Young, January 1992
Calibration of the Queen's
Mine-by microseismic recording system was carried out at the URL in August
of 1991. The routine QCAL was written to process the calibration data.
This routine calculates the system gain on any input signal for each channel
and also calculates the total system impulse response of each channel.
The total system response is the product of the system gain and the accelerometer
response over the frequency range of approximately 50 to 10000 Hz.
The system gain for each
channel was typically between 4-5 times, and the sensor response was flat
between 20-2000 Hz and rose slowly to +3 dB deviation from 2000 to 10000
Hz. This report presents the results of the QCAL routine in table form
and includes coefficients determined for best fit polynomials to the system
gain and total system impulse response signals.
RP011AECL: In Situ Velocity
Measurements For The Mine-By Experiment. S. Talebi and R. P. Young,
March 1992
This report summarises the
results of the P- and S- wave velocity measurements carried out at the
420 m level of the URL in the area of the mine-by tunnel in a velocity
and an attenuation survey. In the first survey, blasting caps were used
as the source and only P-wave velocities were measured. In the second survey,
squibs were used as the source and both P- and S- wave velocities were
measured. The best-fit line, to shot-sensor distance versus P--wave travel
time, results in a P-wave velocity of 5,907 m/s for the velocity survey.
Similar diagrams give a P-wave velocity of 5,903 m/s and an S-wave velocity
of 3409 m/s for the attenuation survey. A Wadati diagram gives a value
of 0.245 for the dynamic Poisson's ratio. The raypaths in granite and granodiorite
show no significant difference in P-wave velocity.
RP012AECL: Preliminary
AE/MS Results From The Mine-By Experiment. S. Talebi and R.P. Young,
March 1992
Microseismic activity recorded
following the reaming and rock breaking of a 1meter round of the mine-by
tunnel of the URL is described. The event source locations, in both cases,
cluster along the minimum principal stress direction in the plane perpendicular
to the tunnel axis, above and below the tunnel. In the longitudinal section,
some clustering is observed around the tunnel edges but no events seem
to originate from inside the block of rock to be excavated.
RP013AECL: Monitoring
and Source Location of Microseismicity Induced by Excavation of the Mine-By
Tunnel: Preliminary Analysis. D.S. Collins and R.P. Young. October
1992
The Queen 's Microseismic
System (QMS) monitored seismic activity through out the excavation of the
Mine-by tunnel at the URL from September 1991 to August 1992. A detailed
description of the installed QMS system is given. An on-site Queen's engineer
(David Collins) was present through out the experiment, whose main purpose
was to keep the QMS system acquiring data optimally and communicate any
problems or needs between Queen's and URL personnel. A quantitative description
of the various acquisition parameters used during the experiment are discussed.
The sole purpose of one computer
at the URL was to source locate seismic events in real time, using automated
Queen's software. These auto-source locations are presented on CAD views
showing each of the 50 surveyed excavation rounds. For one round, a comparison
of source locations is made from manually picked and auto-picked P-wave
arrivals. A substantial improvement in source location error is found to
occur from manually picking events, however, a comparison of the locations
on CAD views of the tunnel shows similar patterns of seismic activity.
This demonstrates the effectiveness of auto-source locating as a corroborative
tool in providing an initial estimate of where seismic activity is occurring
in real time.
RP014AECL: Uniaxial Compression
Testing of Large Samples of Lac du Bonnet Granite at Low Strain Rates.
Part 1: Studies of Acoustic Emission Rate. S.D. Falls and R.P. Young,
October 1992
This report deals with acoustic
emission (AE) rate and magnitude data from very slow loading rate uniaxial
compression tests of four 19.4 cm diameter cylinders of Lac du Bonnet Granite
from the 240m level of AECL's URL AE count-rate was measured during the
four tests and is compared to both axial and inelastic volumetric strain.
The curves presented of AE count versus axial strain show some common characteristics
for all samples. The AE rate is initially quite low. The rate of AE activity
begins to increase coincident with the onset of dilatancy. Their are generally
some periods of accelerated AE activity well before ultimate failure. These
are assumed to be related to localised spalling incidents. As failure approaches,
the AE rate accelerates very significantly.
There is a relationship between
inelastic volumetric strain and AE count during dilatancy. As dilatancy
increases, the relative abundance of AE increases more than might be expected
from simple dilatancy for most of the samples. This suggests that progressive
localisation of event hypocentres is occurring rather than random microcrack
locations. This progressive clustering of microcracks indicates that weakening
of specific regions is occurring, which could lead to premature failure
of the samples. The sample that attained the highest strength, showed a
more linear relationship between inelastic volumetric strain and AE count
than the other samples. This indicates that this sample may not have experienced
the same degree of weakening as the other samples.
The magnitudes of events
may be characterised by calculating so-called b-values at various stages
during the tests. This has been done for one of the samples. Results indicated
that event magnitudes are initially fairly low during the period of crack
closure. They then decrease significantly during the quiet period of mostly
elastic strain. Magnitudes then increase to higher values during dilatancy.
Because of data uncertainties, it is difficult to determine whether there
is a significant increase in event magnitudes as failure approaches.
RP015AECL: Acoustic Emission
and Ultrasonic Velocity Study of Excavation-Induced Microcrack Damage in
the Mine-By Tunnel at the Underground Research Laboratory. S.R. Carlson
and R.P. Young, November 1992
Acoustic emission monitoring
and ultrasonic velocity measurements were employed to investigate excavation-induced
microcrack damage in a portion of the Mine-by tunnel wall at Atomic
Energy of Canada Ltd's Underground Research Laboratory. Sixteen channels
of 1.0 MHz, whole-waveform data were gathered over a 3.5 week period coinciding
with tunnel extension rounds 17-20. The array was installed along the Northwest
wall of the Mine-by tunnel in four 1.3-m-long NQ3 boreholes aligned parallel
and arranged in a diamond pattern. The array was positioned as close as
possible to the working face of the tunnel at the time of installation
and was immediately adjacent to a radial array of eight miniature Bof-ex
extensometers.
Acoustic emission event locations
and ultrasonic velocity data indicate a definite, but fairly limited amount
of microcrack damage in the portion of the tunnel wall under study. Compressional
and shear wave velocities rose by 200 to 300 m/s over a distance of one
metre into the tunnel wall. Compressional velocity anisotropies of 12%
were recorded, with the maximum velocity oriented parallel and minimum
velocity orthogonal to the tunnel azimuth. Only a small portion of the
recorded acoustic emission activity located within the array; most events
located directly above it. Those events that did locate within the array
were concentrated near the tunnel free surface. Focal mechanism solutions
indicate a shear mechanism for approximately three quarters of the near-array
events and a tensile mechanism for most of the remainder.
RP016AECL: AE/MS Source
Location Calibration and Velocity Results: Two Surveys Performed After
the Mine-By Tunnel Excavation. D.S. Collins and R.P. Young, March
1993
A velocity and attenuation
survey was performed in August 1992 following the excavation of the Mine-by
tunnel at AECL's Underground Research Laboratory. Velocities were calculated
from both surveys using a schmidt hammer and blasting caps as sources,
and the 16 triaxial accelerometers of the Queen's Mine-by Microseismic
System (QMS) as receivers.
The velocity survey used
sources in boreholes along the length of the Mine-by tunnel, drilled (30-70
cm) into the tunnel wall, resulting in:
Apparent P-wave velocity
5720 ? 110 m/s
Apparent S-wave velocity
3360 ? 50 m/s
Dynamic Poisson's Ratio
0.24 ? O.O1
Dynamic Young's Modulus
73.7 ? 2.2 GPa
The apparent P-wave velocity
(with raypaths through the tunnel removed) is low compared with the velocity
for the undisturbed rock mass, found by others to be 5900 m/s, and suggests
that a damage zone exists around the Mine-by tunnel. The attenuation survey,
with sources fired from outside of the receiver array, measured an apparent
P-wave velocity of 5820 + 50 m/s, somewhat lower than would be expected
for the undisturbed rock mass, suggesting some influence from galleries
on the 420 Level including the Mine-by tunnel.
The accuracy of source locating
microseismic events was found by comparing the known locations of boreholes
from the velocity survey with calculated source locations. The schmidt
hammer source was used as it has a similar energy to a microseismic event
induced from the Mine-by tunnel excavation. It is concluded that the accuracy
of manually source locating events in the central 12 meters of the tunnel,
using a single velocity model and removing tunnel raypaths, is 0.3 ? 0.l
meters.
RP017AECL: The Spatial
and Temporal Distribution of AE/Ms Source Locations Following the Mine-By
Tunnel Excavation of Round 17. R P. Young and D. S. Collins, March
1993
The Queen's Microseismic
Mine-by System (QMS) monitored seismic activity induced from the excavation
of the Mine-by tunnel at AECL's Underground Research Laboratory. A detailed
analysis is presented of events collected during round 17 from March 5
- 19, 1992. A total of 359 events have been manually source located and
presented on CAD views of the tunnel.
The frequency of seismicity
shows a general reduction with time from the start to the end of the round,
however, highs and lows in activity do occur. More seismicity occurs in
the upper notch than the lower notch region. Two main factors could be
causing this: gravity and geology. Rock in the upper notch falls to the
ground as it is damaged, however rock in the lower notch region is confined
by gravity and the weight of loose rock on the ground. Secondly, the geology
of round 17 is granodiorite in the lower two thirds and granite in the
upper third suggesting a lithological influence on seismic activity.
Temporal and spatial trends
are seen in the data when the events are analysed over shorter time periods.
Clusters of activity occur in the lower and upper notch regions, between
rounds 13 and 17. Additionally aseismic zones are discernible, and one
particular zone in the lower region of rounds 15-17 can be correlated with
photographs of this area showing a raised portion (lack of lower notch).
This region may mark a contact between rock of varying strength, possibly
with the aseismic zone being an area in the rock mass having a higher percentage
of granodiorite compared with granite.
RP018AECL: Ultrasonic
Imaging of Damage Induced by Cyclic Loading of Lac du Bonnet Granite.
T. Chow, I. Meglis, and R.P. Young, March 1993
Thorough three dimensional
ray path coverage of transmitted ultrasonic waves were used to determine
the development of microcrack damage in a cylinder of Lac du Bonnet grey
granite that was undergoing uniaxial cyclic loading. 19 of the 28 load
cycles reached a stress that corresponded to the onset of unstable crack
growth. Time of flight measurements were made following selected cycles.
The data show a decrease in velocity and a change in orientation of the
anisotropy of elastic wave propagation with continued load cycling. Wave
velocity was found to decrease monotonically throughout the experiment,
with ray paths orthogonal to the axial direction being affected the most.
The acoustic velocity along the axial direction was least affected and
decreased minimally throughout. Tomographic images indicated progressive
development of microcracks, primarily in lobes near the central circumferential
surface of the cylinder. While only the base slowness was imaged within
the lobes, the relatively constant axial velocities suggest these were
primarily sub-vertical axial splitting cracks. In addition, the slowness
reconstructions were designed to be preferentially sensitive to axial microfractures.
The onset of stable crack
growth was estimated by monitoring volumetric strain reversal in real time)
using axial and circumferential strain gages. Acoustic emission counts
were monitored during the load cycles to aid in the determination of the
onset of -stable and unstable crack growth. The Kaiser effect was noticed
on the first few load cycles only; thereafter AE commenced at approximately
the same load each time the sample was cycled.
RP019AECL: AE Monitoring
During Loading of Lac du Bonnet Granite Blocks Containing Central Holes.
I.L. Meglis, R. Chow and R.P. Young, March 1993
Three blocks of Lac du Bonnet
granite from the Cold Spring Quarry were tested under various load conditions
to investigate the effect of stress path on notch formation. Two blocks
had throughgoing central holes, and were subjected to uniaxial and biaxial
loads, respectively. The third block initially had a partially drilled
central hole and was subjected to a biaxial load. Following loading, the
hole was advanced and the sample was reloaded until on the fourth cycle
the hole was throughgoing. AE studies were used to determine the location
and mechanism of microcracking during the load cycles on the incrementally
drilled sample. Results indicate that activity was associated with fracturing
both close to the face of the partially drilled hole and with the boundary
conditions of the applied load. Most AE events had either shear or complex
source mechanism, though many of the events close to the hole were apparently
implosive events. Notch formation occurred in this sample when the tangential
stress in the hole was approximately 345 MPa, significantly higher than
the unconfined compressive strength of the sample. The damage associated
with the advancing hole was apparently quite localised and therefore does
not appear to have significantly affected notch development.
RP020AECL: Thermally Induced
Fracturing of Lac du Bonnet Granite, S.R. Carlson, D.P. Jansen and R. P.
Young, April, 1993
Concurrent ultrasonic tomography
and acoustic emission monitoring were employed to study thermally induced
fracturing in unconfined samples of Lac du Bonnet pink granite. The samples
were thermally cycled to progressively higher peak temperatures with an
electrical resistance cartridge heater placed in a central vertical borehole.
Tomography data were collected at room temperature before and after each
thermal cycle. Acoustic emission waveforms were recorded during both heating
and cooling phases of each thermal cycle. Acoustic emission events were
located and source mechanisms determined. Macroscopic fractures originated
at the relatively cool outer surfaces, grew inward toward the heat source,
and eventually intersected the heater borehole in both samples. Acoustic
emission locations and ultrasonic tomography clearly delineated the macroscopic
fractures. We attribute the development of the macroscopic fractures to
a thermal gradient cracking mechanism and hoop stresses exceeding the tensile
strength of the sample..
RP021AECL: Source Mechanism
Studies at the Underground Research Laboratory. B. Feignier and R.P.
Young, April 1993
This report describes the
status of research on failure mechanisms of microseismic events induced
by the excavation of the “Mine-by” tunnel at AECL’s Underground Research
Laboratory (URL) in Pinawa, Manitoba. The overall objectives of this
study were to enhance understanding of the rockmass response to the excavation
process and use the information contained in the source of the microseismic
events to determine the mechanism of damage development in the tunnel.
The moment tensor inversion
method was applied to events recorded during and after the excavation of
Round 17 of the mine-by tunnel. This choice is explained by the location
of this round, at the centre of the tunnel, i.e. at the centre of the microseismicity
array where location accuracy and focal coverage are optimal. Also,
an extended time period was allocated to microseismic monitoring, giving
more than 300 hours of recording. Therefore, Round 17 was chosen
as an initial case study.
In the report, the first
thirty seven microseismic events associated with the excavation are analysed.
Their hypocenters correlate with the damage observed underground.
This damage develops in the sigma3 direction similarly to borehole breakouts
being observed in the petroleum industry. A detailed analysis of
the source mechanisms was carried out, using moment tensor inversion, to
investigate the growth of the breakout in its early stage. The results
show that as much as 75% of the mechanisms contain a significant negative
isotropic component. These events are interpreted as closure of cracks,
due to changing stress conditions, that opened during earlier excavation
phases. It was emphasised that “classical” seismological techniques
to investigate source properties (such as fault plane solution, Brune or
Madariaga source models) are inadequate since they are based on the assumption
of a shear failure mechanism which may or may not be applicable to excavation-induced
seismicity..
RP022AECL: Attenuation
Analysis at the AECL Underground Research Laboratory Using the Spectral
Ratio Method; Preliminary Results. A.J. Feustel And R.P. Young, April
1993
A controlled experiment was
carried out in August, 1992 at the AECL Underground Research Laboratory
that was designed to investigate seismic P-wave attenuation by use of the
spectral ratio technique. This experiment was unique because it offered
an opportunity to measure in-situ attenuation for a granitic rockmass whose
volume encompassed a 16 sensor triaxial accelerometer microseismic array.
Results from the experiment for 7 ray paths gave average values of Qp ranging
from 188 to 276. The average of 69 individual calculations was 223 with
a standard deviation of 37. Errors for each individual value were between
19% and 30% with the largest portion of the error (approximately 20%) caused
by the uncertainty in the geometrical spreading constraint of the data.
Qp values measured in a horizontal plane seemed to vary as a function of
path azimuth but large uncertainties in the data reduced the confidence
in this observation.
RP023AECL: The Spatial
and Temporal Distribution of Microseismicity Recorded in Round 17 of the
Mine-By Tunnel. D.S. Collins and R P. Young, November 1993
The seismicity recorded in
the round 17 volume of the Mine-by tunnel was manually source located
and analysed. 113 microseismic events were recorded in this round, over
the length of the experiment, with the first events occurring after the
excavation of round l6.
The 72 events locating in
the upper half of the round cluster and correlate well with the surveyed
breakout notch. The frequency of seismicity in this area peaks 39 days
after it starts, when the tunnel has been excavated to approximately 3
meters past round l7. Virtually no seismicity occurs in this area after
this time. The seismicity shows spatial trends with the locations starting
in a broad distribution around the bottom of the upper notch area, and
then moving upwards towards an apex. The 4l events locating in the lower
half of the round occur intermittently over a 5 month period, with the
last 8 events being recorded after the track and loose rock on the tunnel
floor are removed. A much smaller breakout notch occurs in this lower half
compared to the upper half of the round, and the events are seen to locate
broadly over this lower notch area The seismicity locates slightly outwards
from the tunnel perimeter with time.
A correlation between geology
and seismicity is apparent. Seismicity in the upper half of round 17 clusters
in the overbreak notch formed in the grey granite. The lower half of
round 17 is granodiorite, and although a much smaller notch forms, the
broad zone of seismicity recorded suggests a certain amount of damage has
occurred to the rock mass at this location. The seismicity is plotted
on tunnel profiles, taken at convergence array 415-5, and show that
the upper and lower notches develop with time. From these it is suggested
that the notches develop over a longer time period than the seismicity
suggests. It is seen that after the majority of seismicity in the upper
area of round 17 has occurred, the notch profile continues to evolve for
several subsequent months.
RP024AECL: Initial Progress
Report on Acoustic Emission Monitoring of Stage 1 of the Mine-By Heated-Failure
Tests. Stephen D. Falls and R P. Young, February 1994
The Canadian nuclear fuel
waste disposal concept involves the storage of initially thermally hot
materials in the floor of underground openings. The mine-by heated failure
tests are part of the program to evaluate this disposal scheme using a
variety of rock mechanics and seismic/ultrasonic monitoring techniques.
Phase I of this experiment involved the drilling of a 600 mm diameter borehole
at the 420 m level of Atomic Energy of Canada Limited's (AECL's) Underground
Research Laboratory (URL) near Pinawa, Manitoba. The volume of rock surrounding
the borehole was subsequently heated such that the rock at the 600 mm borehole
wall reached a temperature of 85 C.
This progress report deals
with the initial findings of the acoustic emission (AE) monitoring program.
The AE history is being monitored by two methods. Firstly, the AE energy
outputs from 16 ultrasonic transducers surrounding the test site are being
continuously monitored. There are four sensor boreholes surrounding the
large central borehole, each containing four evenly space transducers.
Monitoring commenced November 28, 1993, continuing throughout the period
of drilling of the 600 mm borehole, a quiet period following the drilling,
and into the heating phase of the experiment. Secondly, full waveform
data is being acquired, to be used for AE source location studies.
RP025AECL: Wave Propagation
Effects of an Underground Opening. S.C. Maxwell and R.P. Young, March,
1994
A p-wave velocity survey
was performed after the excavation of the Mine-By tunnel, at the Underground
Research Laboratory. The survey was performed in order to calibrate the
velocity structure for source locating the excavation induced seismicity.
The tunnel excavation was found to significantly alter the velocity structure,
in particular the amplification of an anisotropic fabric in the rock probably
related to microcracks. The tunnel opening was also causing waves to be
diffracted, which was modelled using an algorithm based on a finite-difference
scheme. Anisotropic velocity inversion was performed to analyse the variation
in velocity immediately around the tunnel. Although the resulting image
had resolution limitation related to limited raypath coverage, a distinct
zone of high velocity was observed roughly corresponding to a zone of increased
stresses where a break-out notch formed. Perpendicular to this zone, a
region of decreased velocity was found associated with a region of decreased
stress. The stress changes are believed to have preferentially opened and
closed pre-existing fractures (maybe excavation-induced) in the rock, thereby
decreasing and increasing the p-wave velocity. The induced microseismicity
is found to be generally associated with the high-velocity region adjacent
to the notch, and appears to be located along the boundary between the
fractured and elastic regions. The comparison of the velocity structure
and the induced seismicity highlights the potential of using a holistic
interpretation of seismicity and velocity data to assess the excavation
damaged and disturbed zones.
RP026AECL: Comparison
of the Excavation-Induced Microseismicity From the Granite and Granodiorite
Sections of the Mine-By Tunnel. D. S. Collins and R. P. Young, April
1994
420 seismic events recorded
in granite and granodiorite regions of the Mine-by tunnel are analysed
for spatial and temporal trends in source locations. In particular the
seismicity in the l meter, round 7 (granite) and round 25 (granodiorite)
volumes are compared.
In both 1 meter volumes,
the first seismicity happens after the excavation of the previous round,
round 6 and 24 respectively. The events after the round 6 excavation occur
directly ahead of the face and it is suggested that this initial cracking
causes zones of weakness that influence the location of the round 7 post-ream
crack. In the upper notch region of the round 7 volume, seismicity peaks
when the tunnel excavation is l - 2 meters past round 7, compared with
the round 25 volume in which the amount of seismicity is high directly
after the excavation of round 25 and then again when the tunnel is approximately
11 meters past round 25. In both l meter volumes, the lower notch development
seems to be confined by the loose rock and track on the floor of the tunnel.
In the round 7 volume the majority of the seismicity in the lower notch
region occurs after the excavation of rounds 7 and 8, and in the round
25 volume, the seismicity peaks when the tunnel is about 3 meters past
round 25.
Two main points are found
from comparing the seismicity in the granite and granodiorite regions:
i) the spatial extent of damage surrounding the tunnel is similar, with
seismicity extending to about 1 meter from the tunnel perimeter in the
notch regions, and seismicity occurring close to the perimeter on the sides
of the tunnel, and ii) seismicity in the granodiorite rounds continue for
a longer time (after a greater number of rounds have been excavated), than
seismicity in the granite rounds, suggesting a difference in the seismic
response time between the two rock types.
RP027AECL: Acoustic Emission
and Ultrasonic Velocity Studies of the Mine-By Heated Failure Test - Phase
1. S.D. Falls and R.P. Young, August, 1994
During the first phase of
the Mine-by Heated Failure Tests, acoustic emission monitoring (AE) and
ultrasonic p-wave velocity studies were undertaken to gain a better understanding
of the processes taking place as a large diameter borehole was drilled,
and then the volume of rock around the hole was heated such that the borehole
wall reached a temperature of 85°C. This was done in room 415 at the
420 m level of the Underground Research Laboratory (URL). The AE monitoring
program was used to temporally and spatially delineate microcrack activity
associated with the development of borehole breakout and damage at the
borehole face. It also showed that microcrack activity was taking place
in the notch in the floor of the room and that this was activated by the
heating and cooling process.
P-wave velocity studies showed
the changing conditions within the rock surrounding the borehole. Large
velocity increases at the start of drilling have been attributed to rapid
saturation of the rock mass indicating the presence of a somewhat permeable
network of cracks in the vicinity of the tunnel, probably induced during
tunnel excavation. The crack density appears to be greatest closest to
the tunnel floor decreasing with depth, and the cracks seem to be aligned
parallel to the length of the tunnel.
RP028AECL: Source Parameters
of Excavation Induced Seismicity from the Mine-By Tunnel. D.S. Collins,
C. Baker and R. P. Young, November, 1994
The Mine-by experiment at
the Underground Research Laboratory, Manitoba involved the excavation of
a tunnel in one meter increments under extreme stress conditions while
being monitored by state-of-the-art geotechnical instrumentation. The large
number of induced seismic events recorded by a 48 channel microseismic
monitoring system, have been processed for source locations and source
parameters using a calibrated, automated software package. A total of 93l7
events were located and the parameters: seismic moment, radiated energy,
source radius, static stress drop, and S/P wave energy ratio, were calculated.
Source parameters were computed on the basis that shear failure was considered
a significant component of the failure process.
It is found that the highest
level of seismicity and the events with the largest source parameter values
occur within a 2 meter region of the active excavation face (between 0.5
meters ahead and 1.5 meters behind the face). Seismicity occurs for a long
distance (up to 20 meters) further behind the active excavation face, with
the upper range (maximum value) of the source parameters, e.g. seismic
moment, decreasing with distance from the face. This envelope of maximum
source parameter values with distance from the face relates to the change
in stress concentration due to the influence of the tunnel. The area surrounding
the tunnel is divided into four regions, namely the upper and lower notch
regions, sidewall region, and region ahead of the face. No significant
differences in the ranges of the source parameters can be seen in these
four areas. The large range of S/P wave energy ratios indicate the events
varied in mechanism from pure shear to shear with volumetric (tensile or
compressional) components of failure. Events with high and low S/P energy
ratios seem to occur spatially together around the tunnel. An analysis
of the events surrounding a one meter section of the tunnel, suggest that
the first events to occur in a new zone of competent, undamaged rock are
smaller in size than those occurring later in this same area. This may
be interpreted as the later events occurring in a damaged zone of rock
with slip over a larger area taking place.
Comparing the events from
the Mine-by tunnel excavation with those from the mine shaft extension,
the most notable difference is the latter seismicity is generally higher
in radiated energy. This may be explained by the different tunnelling directions
and associated stress ratios, as well as the different excavation methods.
Events from both excavations, show a large range of S/P wave energy ratios,
suggesting a complex damage process and the need for further source mechanism
analysis using moment tensor inversion.
RP029AECL: Atomic Energy
of Canada Limited Acoustic Emission and Ultrasonic Velocity Studies of
the Mine-By Heated Failure Test - Phase 2. S.D. Falls and R.P. Young,
March 1995
During the second phase of
the Mine-by Heated Failure Tests, acoustic emission monitoring (AE) and
ultrasonic P-wave velocity studies were undertaken to gain a better understanding
of the processes taking place as a volume of Lac du Bonnet granite was
heated to a temperature of about 85°C and then a large diameter borehole
was drilled within this volume. This was done in room 415 at the 420 m
level of the Underground Research Laboratory (URL). The AE monitoring program
was used to temporally and spatially delineate microcrack activity associated
with the development of borehole breakout and damage at the borehole face.
Borehole breakouts, as delineated by AE source locations developed much
more rapidly under heated conditions than they did under unheated conditions
in the previous phase of the experiment. AE also showed that microcrack
activity was taking place in the notch in the floor of the room and that
this was enhanced by the heating and cooling processes.
P-wave velocity studies showed
the changing conditions within the rock surrounding the borehole. The crack
density appears to be greatest closest to the tunnel floor decreasing with
depth, and the cracks are aligned parallel to the tunnel wall. Changes
in P-wave velocity mirror the temperature profile. Thermally induced stresses
are thought to close existing microfractures to some extent, thus increasing
P-wave velocity.
RP030AECL:Three-Dimensional
Seismic Velocity Imaging of the Mine-By Tunnel. S.C. Maxwell and R.P Young,
August 1995
A p-wave velocity survey
was performed after the excavation of the Mine-By tunnel, at the Underground
Research Laboratory. The survey was performed in order to calibrate the
velocity structure for source locating the excavation induced seismicity.
Maxwell and Young, 1994, presented results of a preliminary analysis of
this controlled-source velocity survey. A new 3D velocity image was computed
for the central portion of the tunnel. A zone of high velocity was observed
roughly corresponding to a zone of increased stresses where a break-out
notch formed. Perpendicular to this zone, a region of decreased velocity
was found associated with a region of tensile stress. The stress changes
are believed to have preferentially opened and closed pre-existing fractures
in the rock, thereby decreasing and increasing the p-wave velocity.
The induced microseismicity is found to be generally associated with the
high-velocity region adjacent to the notch, and appears to be located along
the boundary between the fractured and elastic regions. Passive-source
images were also computed from the induced seismicity data using a simultaneous
hypocentral location and velocity inversion technique. An image of the
upper notch showed decreased velocities in the fractured zone in the notch,
with higher velocities in the intact-high stress region further into the
rockmass. A separate image of the velocity structure immediately ahead
of the tunnel face indicated decreased velocities in the tensile stress
region. A comparison was also made between the source parameters of the
induced seismicity and the velocity structure The greatest seismic deformation
was found to occur with the velocity transition zone, corresponding to
the boundary between the fractured rock and the highly stressed. intact
rock The comparison of the velocity structure and the induced seismicity
highlights the potential of using a holistic interpretation of seismicity
and velocity data to assess the excavation damaged and disturbed zones
RP031AECL: Acoustic Emission
and Ultrasonic Velocity Studies of the Mine-By Heated Failure Test - Phase
3. S.D. Falls and R.P. Young, August 1995
Phase 3 of the Mine-by Heated
Failure Tests, involved the excavation of two 600-mm-diameter boreholes
with approximately 400 mm spacing between them. Much of the emphasis of
this phase is to determine whether there was any interaction between the
two holes. The volume of rock around the boreholes was heated such that
the temperature at the borehole wall reached approximately 85°C, starting
25 days after the completion of drilling. Heaters were on between 28 October
and 14 December, 1994. Acoustic emission (AE) activity was monitored between
23 August, 1994 and 20 January, 1995. This period was several weeks before
drilling started to several weeks after the heating ceased. Throughout
this period three-dimensional P-wave velocity surveys were conducted approximately
once every week. The experiment took place in room 415 at the 420 m level
of the Underground Research Laboratory (URL). The AE monitoring program
was used to temporally and spatially delineate microcrack activity within
a volume extending 1 to 2 metres from the large central boreholes.
Drilling the first hole resulted
in relatively low levels of AE energy release. However, as the drilling
of the second hole commenced, high levels of AE energy were released and
anomalous AE activity continued over a prolonged period relative to the
single-hole results from phase l. During the heating period, the AE energy
rate rose to a peak within a few days, which was much faster than in previous
tests. As in other tests the AE rate began to decay after reaching a peak.
Unlike previous tests, the AE rate began to rise to a second peak of activity
starting after about two weeks of heating. The trend during heating was
that much higher AE levels occurred than in phase l.
P-wave velocity showed a
trend of general decay throughout the test, with the average velocity at
the end of the test over 100 ms-1 slower than the initial velocity. As
with the previous phases, there was a substantial rise in velocity at the
beginning of heating. This change can probably be associated with thermally-induced
stress increases closing existing microcracks. Unlike the previous phases
there was a large drop in P-wave velocity during heating, occurring coincidentally
with the second peak in AE activity described above. This velocity decrease,
coupled with the increase in AE rate, suggests that extensive microcrack
damage occurred during the heating process. A general progressive increase
in P-wave velocity in the rock below the holes indicates that stress may
have been shed to the deeper regions as the shallower parts of the study
volume became damaged.
AE source locations show
that the borehole-breakout activity was limited during the drilling of
the first hole, but developed rapidly in both holes as the second hole
was excavated. The AE locations are more strongly clustered on the side
of each hole adjacent to the neighbouring hole than on the outer sides.
There is no evidence of direct interconnecting large cracks between the
two boreholes. However, during the heating period marked by a velocity
decrease and AE energy peak, there does appear to have been scattered activity
throughout the volume, most concentrated in the web region between the
holes.
RP032AECL: Preliminary
Analysis of the Excavation Damage Zone Around the Mine-By Tunnel by Borehole
Logging Using a Micro-Velocity Probe. S.C. Maxwell, R.E. Murdie and
R.P. Young, March 1996
In order to characterize
the extent and degree of damage around underground excavations, a proto-type
micro-velocity probe was constructed. The probe measures the time of travel
of high-frequency (ultrasonic) seismic waves travelling between two transducers,
and enables the variation in velocity to be logged within boreholes. By
logging radial boreholes around the underground excavation, the velocity
logs can be used to assess the excavation damage zone. Preliminary
trials of the probe at AECL's URL indicated that the probe is both simple
and effective. Several holes in the Mine-by tunnel were logged, which indicated
significant damage in the rockmass immediately around the tunnel. Core
from one of the holes was also logged which indicated that the velocity
of the core was 57% of the in situ velocity.
RP033AECL: Acoustic Emission
And Ultrasonic Velocity Studies Of The Mine-By Heated Failure Test - Phase
4. S.D. Falls and R.P. Young, July 1996
During the fourth phase of
the Mine-by Heated Failure Tests, acoustic emission monitoring (AE) and
ultrasonic p-wave velocity studies were undertaken to gain a better understanding
of the processes taking place as a large diameter borehole was drilled,
and then the volume of rock around the hole was heated such that the borehole
wall reached a temperature of 85°C. In this phase the large diameter
borehole was sealed and held under 100 MPa pressure to simulate the effects
of backfilling the borehole before the rock mass was heated. This was done
in room 415 at the 420 m level of the Underground Research Laboratory (URL).
The AE monitoring program
was used to temporally and spatially delineate microcrack activity associated
with the development of borehole breakout and damage at the borehole face.
This activity was greatest during times of transition, drilling heating
and cooling. Compared to previous phases, the activity during the heating
phase of the experiment was less concentrated in the breakout notches in
the large-diameter borehole. This suppression of activity was related to
the pressurisation of the HFT5 borehole. Increased activity was seen in
these breakout notches during depressurisation events. Microcrack activity
was also taking place in the notch in the floor of the room, which was
activated by the heating and cooling process.
P-wave velocity studies gave
information on the initial conditions and temporal changes in the rock
mass surrounding the HFT5 borehole. Results indicated the presence of an
excavation damaged zone, with microcracks predominantly aligned approximately
parallel to the floor of the tunnel. decreasing in crack density with depth
into the rock. There was a slight drop in P-wave velocity due to drilling
the HFT5 borehole, little change due to pressurisation of the hole. and
a large increase due to thermally induced stress during heating.
RP034AECL: Micro-Velocity
Probe Logs of the Permeability Test Boreholes in the Mine-By and Excavation
Stability Experiments. S.C. Maxwell and R.P. Young, January 1997
A modified design for a micro-velocity
logging probe was tested at AECL in September 1996. The probe is used to
measure changes in ultrasonic interval velocity within boreholes, to investigate
the damage zone around underground excavations. The new probe provided
better quality data compared to the original design, mainly due to improved
transducer contact with the borehole walls. A strategy for logging p- and
s-wave velocities was also implemented. The new probe was used to log holes
in which a SEPPI probe has been used to measure in situ permeability, so
that velocity changes could be related to permeability changes. Logs collected
in extensometer holes in the Mine-by tunnel indicated substantial excavation-induced
damage in the surrounding rock mass. By contrast, relatively little damage
was observed around the Excavation Stability tunnel (418-2). In the tensile
side-wall of the tunnel, damage was observed extending to 70 cm. However,
in the other holes in the floor of tunnel, less than 10 cm of damage was
found. Relatively larger velocity variations due to lithology changes were
found in this tunnel, compared to the Mine-by tunnel which appeared more
homogeneous. Velocity anisotropy around the Mine-by tunnel was also investigated
by comparing borehole logs with cross-hole measurements. In the region
at the end of the tunnel and ahead of the excavation face, a damage region
was found to extend to more than 1 m into the rock. The microfractures
appeared to be subvertical, approximately parallel to the tunnel face.
In the side-wall region of the tunnel, the damage zone was found to extend
to about 50 cm. In this case, the microfractures appear to be subhorizontal.
In both instances, the inferred microfracture orientation agrees with the
orientation of near-field, tensile principal stress.
RP035AECL: Assessing Microcrack
Damage Around the Mine-by Tunnel at the Underground Research Laboratory
Using Ultrasonic Velocity Tomography. I.L. Meglis, T. Chow and R.P. Young,
April 1997
The compressive strength
of rock is influenced strongly by the presence of microcracks, which can
provide nucleation sites for fractures which ultimately lead to failure.
Information on the distribution of microcracks around underground excavations
may therefore be useful in assessing the strength of material adjacent
to the openings. The density of microcracks, their initial size, and any
preferred orientation are all factors which influence the tendency of the
rock to fail under stress. Because these factors also affect both the velocity
and amplitude of ultrasonic waves in rock, they can be assessed in situ
both qualitatively and, potentially quantitatively using measurements of
wave velocity and amplitude. The purpose of this work was therefore to
make a detailed ultrasonic survey around one quadrant of the Mine-by tunnel
at 420 m depth at the Underground Research Laboratory in order to characterize
the crack population which formed in response to the stress changes associated
with underground excavation.
The velocity data from this
survey provide an estimate of the average rock properties within a 1 m
shell around this portion of the tunnel. The raw data are then used to
generate a tomographic image of this shell, which provides a more detailed
picture of the spatial variation of velocity properties within the rock.
Relative changes in crack density and orientation around the tunnel are
inferred from these results. In summary, the material in the region from
the sidewall to about midway to the notch region appears to be extensively
microcracked, resulting in as much as 30% drop in velocity from the intact
values. The microcracking is most concentrated close to the tunnel wall,
but persists to at least 1 m depth into the rock. The region midway between
the notch and the sidewall shows the strongest alignment of cracks, inferred
from the magnitude of the velocity anisotropy. Within the notch region,
in apparent contrast with the extensive microseismic activity reported
by Feignier and Young (1993), the material shows only a small amount of
microcrack damage. However, this observation is consistent with the fact
that much of the original damaged material has spalled away during notch
formation. The existing cracks are apparently concentrated within the few
cms of material adjacent to the tunnel.
It appears therefore that
the stress changes associated with excavation and the subsequent stress
conditions induce grain-scale microcracking all around the tunnel, and
not solely in the notch. In the sidewall region, where both the pre- and
post-excavation stresses are low , the distribution of microcracks is relatively
isotropic. Closer to the notch, where the tangential stress is higher,
the microcracks show a higher degree of preferred orientation. Within the
notch region, it is inferred that the very high tangential stresses produced
a high local density of strongly aligned cracks, which ultimately led to
breakout formation.
Analysis of the strength
and velocity measurements made by CANMET on core samples of Lac du Bonnet
granite indicates a correlation between decreasing unconfined compressive
strength (UCS) and ultrasonic velocity, both of which are related to the
microcrack damage in the samples. Extrapolating this correlation to the
measurements in situ suggests that the strength of much of the material
surrounding the tunnel has been reduced by the excavation-induced microcrack
damage. The current stress acting on the material in the sidewall is still
well below the in situ strength inferred from the velocity measurements.
However, the magnitude of induced thermal stresses, stress changes resulting
from continued excavation, and stress waves generated by blasting and induced
seismicity must be considered.
The material in the notch
region shows little existing crack damage, implying little reduction in
short-term strength. However, the effects of continued strength degradation
from stress corrosion, crack propagation or thermal fracturing are also
unknown. Any cracks which form by such processes in the highly stressed
notch region are likely to be strongly aligned parallel to the tunnel wall
and could contribute to further spalling. This spalling would then induce
further stress changes in the region. Therefore the absence of extensive
microcracking, as is detected in the sidewall, does not necessarily imply
long-term stability for the notch region.
Future work is suggested
on calibrating the correlation between strength, velocity, and crack density
in damaged rocks in order to provide a more useful tool for assessing in
situ strength properties.
|
