- Email: [email protected]
Field evaluation of cable bolt supports
Int. J. Rock Mech. Min. Sci. & Geomech. Abstr.
Printed in Great Britain
Vol.30, No.7, pp. 1431-1434, 1993
0148-9062/93 $6.00 + 0.00 Pergamon Press Ltd
Field Evaluation of Cable Bolt Supports J.M. G O R I S * L.A. M A R T I N * T.M. B R A D Y *
INTRODUCTION
~'.'.-~
The U.S. Bureau of Mines, in cooperation with the Homestake Mining Co., Lead, SD, and the University of Utah, Salt Lake City, UT, conducted a field demonstration to evaluate the behavior of cable bolt supports in three separate areas of the Homestake Mine. This paper presents the results of the demonstration at one of these sites, located between the 6500 and 6650 levels of the 45-48N stope, where both conventional and birdcage cable bolts had been installed. The primary objectives of the project were to (1) monitor the behavior of the rock supports with extensometers and cable bolt strain gauges and (2) model the rock masses using the Fast Lagrangian Analysis Continuum (FLAC) numerical code. Field data collected from the extensometers and cable bolt strain gauges were used to verify the models. In addition, conventional and birdcage cable bolt supports were compared to evaluate their relative effectiveness. Work was conducted as part of the Bureau's program to improve mineral resource recovery.
~
5"..
,~
~
.~'~Oi
Instrumentation
~.~
~"~IX'9~ ~ ~.:-':".~,. |'.A ~ ~
~/l~dding
/
dip
Fracture dip
i-:2.
Fig. 1. Cross section of geology surrounding 45-48N stope.
GROUND C O N T R O L
The Homestake Mine is located in the Black Hills of South Dakota in the town of Lead. The three major formations in the Lead area, the Poorman, the Ellison, and the Homestake, are all Precambrian in age. The 45-48N stope is located in the Homestake Formation and lies between the 6500 and 6650 levels. The ore body is approximately 61 m long by 13.7 m wide. Between the 6500 and 6650 levels, it dips from 35 ° to 60 ° east with a near-vertical plunge to the south. The contact between the Homestake Formation and the overlying Ellison creates planes of weakness along localized graphite layers, and these can cause ground control problems during mining. Figure 1 shows a typical cross section of the 45-48N stope area. In addition to geologic conditions, the geometry of the stope and the sequence of mining have a great effect on the stability of the walls and back. Stress concentrations from earlier mining can be sufficient to cause slabbing of the back. One of the primary reasons that ground control engineers at the I-Iomestake Mine initiated a cable bolt support program in 1977 was to control such slabbing [1; 2].
Ground control for the 45-48N stope consisted of double 18.3-m-long conventional and birdcage cable bolt supports as well as 1.5-m-long Split Sets.** Stations for the cable support patterns began at the south end of the stope at station 0 and continued north to station 16 (Fig. 2). Conventional cables bolts were installed on 3-m centers between stations 0 and 8. Birdcage cables were placed on 3.4-m centers between stations 9 and 16. A typical cross section of the stope showing the fan-shaped configuration of the cable supports is shown in Fig. 3. The conventional cable bolts used were made from high-strength steel with an ultimate strength of approximately 258 kN and a modulus of elasticity of approximately 2.03 x 105 MPa. The bolts were 15.9 mm in diameter and consisted of seven wires. The birdcage cables were developed as a means of improving the pullout resistance of cable supports. They are made from conventional cables by separating the seven wires of a conventional cable and then recombining them to form an open cable having a series of nodes and antinodes spaced about 178 mm apart along the cable. Researchers at the Bureau's Spokane Research Center conducted laboratory pull tests on both single and double birdcage cables to determine their load-displacement behavior when compared to single conventional cables [3-4]. Basically, these tests showed that both single and double birdcage cables have higher load-carrying capacities than conventional cables.
*Spokane Research Center, U.S. Bureau of Mines, Spokane, WA 99207, U.S.A.
**Reference to specific manufacturers, products, or trade names does not imply endorsement by the U.S. Bureau of Mines.
45-48N STOPE
1431
1432
ROCK MECHANICS
Scale, ft 0 1020
Hq
LEGEND Stope on 6570 level - . , ,.-, Drifts on 6500 level 1o Extensometer 1• Instrumented cable bolt STATION LINE 161514131211109 87654
3 21
0
036
Cross-cut - i ' ~ ' ~ _ " " ~ - " = ' / " ~ " ~
' "
-'~e
" ~ ' ~ % .
Bulkh~ea~crologger
%
Fig. 2. Plan view of cable bolt supports from stope on 6580 level,
Ddftr"l
6500 level
Instrumented cable bolt I
/ / 4.6 rn
T
- - 6583
L_~
tope
level
IN T H E 1 9 9 0 s
groutable rebar anchors, depending on the depth of the hole, and were read electronically with 102-mm rangelinear potentiometers placed at the collar of the hole in the drift on the 6500 level. The cable support strain gauges were installed to determine the amount of load being placed on selected cables. Each gauge is 0.69 m long and uses a thin (0.25 mm in diameter), nickel-chromium wire that is wound into spiral grooves between the outer strands of the cable. Readouts from the gauges are in millivolts, and a calibration curve is provided so that these readings can be converted to load. Electrical signals from both the extensometers and cable support strain gauges are read by a microcomputer datalogger (MICROLOGGER). The locations of the MICROLOGGER and other instruments are shown in Fig. 2. RESULTS During the 24-month period the instruments were monitored, three lifts were mined in the 45-48N stope, and major blasting occurred in adjacent cut-and-fill stopes as well as in vertical crater retreat stopes. It is unclear how much these other mining activities around the stope influenced readings from the instruments because records of the blasts were not kept by mine personnel. Figure 4 shows results from extensometer 2 located at station 10.5 (Fig. 2), This is the same station where rock cores were obtained and where the numerical analysis of the 45-48N stope was conducted. Leveling of the curves indicates that the anchors were lost and the potentierectors became stationary. In studying these curves, the sequence in which the anchors were lost relative to time appears reasonable; that is, the deeper anchors (6, 5, 4, etc.) were lost first as mining progressed from the 6570 toward the 6500 level.
.4 m----~ KEY ..... Anchor I . . . . . . . . . . - A n c h o r 2 . . . . . - --............... Anchor 3 . . . . .
Fig. 3. Cross section of 45-48N stope. Table 1. Instruments placed in 45-48N stope. Instrument Extensorn~r: 1 ....... 2 ....... 3 (uphole) . . 4 ....... Cable: 1 ....... 2 ....... 3 ....... 4 .......
Station
Type of support in area
E 80
E
g 6o12.5 10.5 8.0 1.5
Birdcage cables Birdcage cables Not applicable Conventional cables
3.0 12.0 14.0 2.5
Conventional cables Birdeage cables Birdcage cables Conventional cables
INSTRUMENTS Instruments placed around the 45-48N stope consisted of extensometers grouted into holes between cable support stations and strain gauges attached to cable supports. Table 1 shows the type of support used in the vie'mity of each instrument. The exteusometers had either five or six
,f
I lift
~
m40 O. if)
I
Fifth liftl ~lme~ll
~
. !
.m.Io.II. II. |
Anchor4 Anchor5 Anchor 6 !
I
m!I n l J an| = auan. l a . auunm m m
._----=..............
20
lift n
0
100
i
I
I
200 300 400 500 600 TIME, days sirme 09/30188
700
Fig. 4. T i m e - d i s p l a c e m e n t c u r v e s for e x t e n s o m e t e r s 2.
It is important to note that in Fig. 4, all the displacement values are positive, which indicates that the rock at each anchor location was moving down and away from the drift on the 6500 level and toward the stope. Displacements of anchors 1 through 5 for extensometer 2 during the first 200 days were most likely caused by floor movement in the drift on the 6500 level.
ROCK MECHANICS IN THE 1990s
24
Third
0 -24
I
Gauge 2
lift . . . . . . . . Gauge 3
Ei2
c~ -48 1
,,.,~< 96 0
.. ... ..... ... . =u o, Gauge5
lift I
I
-24 150
t
I
t
100 200 300 400 500 600
I 240~
I
d
I
'I
I
7OO
I
I I I 250 350 450 550 TIME, dayssince09/30/88
650
Fig. 5. Time-load curves for instrumented cable bolts at stations 12 (A) and 2.5 (B).
Data shown in Fig. 5A were received from the instrumented cable bolt at station 12 (Fig. 2), which is close to extensometer 2 at station 10.5. The data are scattered because the electrical connections periodically shorted out and made contact again, so little information was revealed about the behavior of the rock. This is unfortunate because reliable data would have helped determine the effectiveness of the cable supports. Strain gauges on these cables were the first to be installed by the crew under field conditions, and it is felt that the installation technique may not have been adequate to protect the electrical connections from water. The strain gauges on the instrumented cable at station 14 (Fig. 2) responded very well for 3 months (Fig. 5B), but then were lost due to mining. The data collected at this site did not indicate whether or not the birdeage cables were more effective than conventional cables. Comments from the cable bolting crew indicated that the birdcage cables were more difficult to install without a special cable pusher; however, miners working in the 45-48N stope believed the birdcage cables supported the reek better than did conventional cables. This conclusion was based on over 40 years of mining experience and visual observations of reek in the stope. Results from pull tests conducted by the Bureau show that double birdcage cables can carry almost twice the load per foot of embedment as can double conventional cables [4]. NUMERICAL MODEL A numerical analysis of the 45-48N stope between the 6500 and 6650 levels was conducted using the computer code FLAC. This is a two-dimensional, finite-difference code that uses the explicit finite-difference method to solve basic equations of motion (force = mass × acceleration) and to calculate the out-of-balance forces acting at each node. To model the 45-48N stope region, a computer mesh was developed using the geology and stope geometry at
1433
station 10.5 (Fig. 2). The model assumes simple mine geometry and does not include any previously mined stopes or drifts in the vicinity of the 45-48N stope other than those shown in Fig. 1. The FLAC code allows for cut-and-fill excavation sequences and installation of rock supports, which are extremely important in modeling mining operations. FLAC also allows for the input of different isotropic material properties for each element. In modeling the 4548N stope region, three different isotropic rock materials were used, as well as a sand material that represented the backfill. Modeling began by allowing the model to first reach a state of equilibrium before excavation, that is, a state of zero displacement of the rock at any point in the mesh. Next, an initial lift was excavated by changing the material properties of an area 4.6 m high by about 9.4 m wide (this represented the mined-out area of the stope) from rock to a void. The program was again run and all forces were allowed to reach a state of equilibrium. Eighteenmeter-long cable bolts were then "installed," the area previously mined out was "backfilled" by changing the material properties of the void to simulate backfill, and the next lift was "excavated." This process continued until the fourth lift had been mined; at this point, the stope floor was at the 6590 level. Then a new series of cable bolts was installed. Mining continued for another three lifts (the final lift was 7). The stope was then at the 6545 level just 13.7 m below the drift on the 6500 level. This modeling sequence of simulated cutting and filling was conducted a number of times to help fine-tune the model, that is, to determine which combination of rock properties resulted in the closest correlation with the measurements obtained from the field instruments. Results from the computer runs are given as displacement (in millimeters) of the rock at each node in the mesh as well as for four specific points corresponding to anchefs 2, 4, 5, and 6 for the extensometer at station 10.5 (Fig. 2). These anchors were located 3.4, 11.6, 15.5, and 19.5 m, respectively, from the collar of the hole on the 6500 level. It should be noted that the FLAC program indicates positive displacement as up toward the drift on the 6500 level, which is the opposite direction shown by field results. Results from the field data and the numerical model are compared in Table 2. Field data were corrected for displacements caused by extensometer head movement (indicated by anchor 1) by subtracting the displacement reading for anchor 1 from the initial data for anchors 2 though 6 (Fig. 4). The field data and the calculations from the numerical model correlated fairly well. However, as mentioned before, the simulation was conducted a number of times to help fine-tune the model and determine which combination of rock properties resulted in the closest correlation with the results obtained from the field instruments. The results in Table 2 represent the best correlation. Figure 6 shows a plot of the data from Table 2.
1434
ROCK MECHANICS IN THE 1990s Table 2. Anchor displacement for extensometer and numerical model, millimeters. Anchor
Lift number 5 6 7
2 ......
Initial field data 12.7 15.2 35.6
Correction for head movement -12.7 -15.2 -17.8
Corrected field data 0.0 .0 17.8
Model data -5.1 .0 .0
4 ......
5 6 7
30.5 38.1 55.9
-12.7 -15.2 -17.8
17.8 22.9 38.1
12.7 22.9 22.9
5 ......
5 6 7
25.4 30.4 33.0
-12.7 -t5.2 -17.8
12.7 15.2 15.2
7.6 22.9 22.9
6 ......
5 6
55.9 (1)
-12.7 (1)
43.2
17.8
7
(')
(I)
~Anchor lost through blasting. 50
i
q 20 LU ~10
=
=
o
/[ i
i
"
-10 0 10 20 30 NUMERICAL MODEL DATA, mm Fig. 6. Field versus numerical model displacement data for extensometer at station 10.5. The correlation coefficient for the data is 0.404554, and although this value is low, it does show fairly good correlation between field and numerical model data. Successful use o f F L A C depends on knowing rock mass properties. F r o m this modeling effort, it appears that F L A C could be used to predict rock mass behavior. at least in a general sense, because o f its ability to model cut-and-fill sequences. CONCLUSIONS Data collected from instruments placed in the rock above the 45-48N stope did not indicate whether or not birdcage cables were more effective than conventional cables at supporting the roof. The extensometers worked very well; however, data received from several o f the cable strain gauges installed early in the project were very scattered and revealed little about the behavior o f the
rock. It is felt that the procedures used to install these gauges may not have been adequate to protect electrical connections. Strain gauges installed later in the same rock mass responded more consistently (Fig. 5B). The cable installation crew working i n t h e 4 5 - 4 8 N stope commented that birdeage cables were m o r e difficult to install without a special cable pusher; however, miners working in the stope said they believed the birdeage cables were supporting the rock better than conventional cables. N u m e d e a l modeling o f the 45-48N stope using the F L A C code provided displacement values comparable to field extensometer measurements. At a minimum, the use o f this code could assist ground control engineers in conducting a general analysis o f the behavior o f rock masses in a cut-and-fill mining operation. Based on data collected, it was concluded that extensometers and cable bolt strain gauges a r e practical and effective instruments for collecting essential field data.
REFERENCES 1. Schmuck C. H. Cable bolting at the Homestake Mine. M/n. Eng. (LiRlcton, CO), v. 31, No. 12, Dec., t677-i681 (1979). 2. Goris J. M., Duan F. and Pfarr J. Evaluation of cable supports at the Homestake Mine. CIM Bull., v. 84, No. 947, Mar., 146-150 (1991). 3. Gods J. M. Laboratory evaluation of cable bolt supports (in two parts) 1. Evaluation of supports using conventional cables. U.S. Bur. of Mines RI 9308, 23 pp. (1990). 4. Goris J. M. Laboratory evaluation of cable bolt supports (in two parts) 2. Evaluation of supports using conventional cables with steel buttons, birdcage cables, and epoxy-coated cables. U.S. Bur. of Mines RI 9342, 14 pp. (1990).
Printed in Great Britain
Vol.30, No.7, pp. 1431-1434, 1993
0148-9062/93 $6.00 + 0.00 Pergamon Press Ltd
Field Evaluation of Cable Bolt Supports J.M. G O R I S * L.A. M A R T I N * T.M. B R A D Y *
INTRODUCTION
~'.'.-~
The U.S. Bureau of Mines, in cooperation with the Homestake Mining Co., Lead, SD, and the University of Utah, Salt Lake City, UT, conducted a field demonstration to evaluate the behavior of cable bolt supports in three separate areas of the Homestake Mine. This paper presents the results of the demonstration at one of these sites, located between the 6500 and 6650 levels of the 45-48N stope, where both conventional and birdcage cable bolts had been installed. The primary objectives of the project were to (1) monitor the behavior of the rock supports with extensometers and cable bolt strain gauges and (2) model the rock masses using the Fast Lagrangian Analysis Continuum (FLAC) numerical code. Field data collected from the extensometers and cable bolt strain gauges were used to verify the models. In addition, conventional and birdcage cable bolt supports were compared to evaluate their relative effectiveness. Work was conducted as part of the Bureau's program to improve mineral resource recovery.
~
5"..
,~
~
.~'~Oi
Instrumentation
~.~
~"~IX'9~ ~ ~.:-':".~,. |'.A ~ ~
~/l~dding
/
dip
Fracture dip
i-:2.
Fig. 1. Cross section of geology surrounding 45-48N stope.
GROUND C O N T R O L
The Homestake Mine is located in the Black Hills of South Dakota in the town of Lead. The three major formations in the Lead area, the Poorman, the Ellison, and the Homestake, are all Precambrian in age. The 45-48N stope is located in the Homestake Formation and lies between the 6500 and 6650 levels. The ore body is approximately 61 m long by 13.7 m wide. Between the 6500 and 6650 levels, it dips from 35 ° to 60 ° east with a near-vertical plunge to the south. The contact between the Homestake Formation and the overlying Ellison creates planes of weakness along localized graphite layers, and these can cause ground control problems during mining. Figure 1 shows a typical cross section of the 45-48N stope area. In addition to geologic conditions, the geometry of the stope and the sequence of mining have a great effect on the stability of the walls and back. Stress concentrations from earlier mining can be sufficient to cause slabbing of the back. One of the primary reasons that ground control engineers at the I-Iomestake Mine initiated a cable bolt support program in 1977 was to control such slabbing [1; 2].
Ground control for the 45-48N stope consisted of double 18.3-m-long conventional and birdcage cable bolt supports as well as 1.5-m-long Split Sets.** Stations for the cable support patterns began at the south end of the stope at station 0 and continued north to station 16 (Fig. 2). Conventional cables bolts were installed on 3-m centers between stations 0 and 8. Birdcage cables were placed on 3.4-m centers between stations 9 and 16. A typical cross section of the stope showing the fan-shaped configuration of the cable supports is shown in Fig. 3. The conventional cable bolts used were made from high-strength steel with an ultimate strength of approximately 258 kN and a modulus of elasticity of approximately 2.03 x 105 MPa. The bolts were 15.9 mm in diameter and consisted of seven wires. The birdcage cables were developed as a means of improving the pullout resistance of cable supports. They are made from conventional cables by separating the seven wires of a conventional cable and then recombining them to form an open cable having a series of nodes and antinodes spaced about 178 mm apart along the cable. Researchers at the Bureau's Spokane Research Center conducted laboratory pull tests on both single and double birdcage cables to determine their load-displacement behavior when compared to single conventional cables [3-4]. Basically, these tests showed that both single and double birdcage cables have higher load-carrying capacities than conventional cables.
*Spokane Research Center, U.S. Bureau of Mines, Spokane, WA 99207, U.S.A.
**Reference to specific manufacturers, products, or trade names does not imply endorsement by the U.S. Bureau of Mines.
45-48N STOPE
1431
1432
ROCK MECHANICS
Scale, ft 0 1020
Hq
LEGEND Stope on 6570 level - . , ,.-, Drifts on 6500 level 1o Extensometer 1• Instrumented cable bolt STATION LINE 161514131211109 87654
3 21
0
036
Cross-cut - i ' ~ ' ~ _ " " ~ - " = ' / " ~ " ~
' "
-'~e
" ~ ' ~ % .
Bulkh~ea~crologger
%
Fig. 2. Plan view of cable bolt supports from stope on 6580 level,
Ddftr"l
6500 level
Instrumented cable bolt I
/ / 4.6 rn
T
- - 6583
L_~
tope
level
IN T H E 1 9 9 0 s
groutable rebar anchors, depending on the depth of the hole, and were read electronically with 102-mm rangelinear potentiometers placed at the collar of the hole in the drift on the 6500 level. The cable support strain gauges were installed to determine the amount of load being placed on selected cables. Each gauge is 0.69 m long and uses a thin (0.25 mm in diameter), nickel-chromium wire that is wound into spiral grooves between the outer strands of the cable. Readouts from the gauges are in millivolts, and a calibration curve is provided so that these readings can be converted to load. Electrical signals from both the extensometers and cable support strain gauges are read by a microcomputer datalogger (MICROLOGGER). The locations of the MICROLOGGER and other instruments are shown in Fig. 2. RESULTS During the 24-month period the instruments were monitored, three lifts were mined in the 45-48N stope, and major blasting occurred in adjacent cut-and-fill stopes as well as in vertical crater retreat stopes. It is unclear how much these other mining activities around the stope influenced readings from the instruments because records of the blasts were not kept by mine personnel. Figure 4 shows results from extensometer 2 located at station 10.5 (Fig. 2), This is the same station where rock cores were obtained and where the numerical analysis of the 45-48N stope was conducted. Leveling of the curves indicates that the anchors were lost and the potentierectors became stationary. In studying these curves, the sequence in which the anchors were lost relative to time appears reasonable; that is, the deeper anchors (6, 5, 4, etc.) were lost first as mining progressed from the 6570 toward the 6500 level.
.4 m----~ KEY ..... Anchor I . . . . . . . . . . - A n c h o r 2 . . . . . - --............... Anchor 3 . . . . .
Fig. 3. Cross section of 45-48N stope. Table 1. Instruments placed in 45-48N stope. Instrument Extensorn~r: 1 ....... 2 ....... 3 (uphole) . . 4 ....... Cable: 1 ....... 2 ....... 3 ....... 4 .......
Station
Type of support in area
E 80
E
g 6o12.5 10.5 8.0 1.5
Birdcage cables Birdcage cables Not applicable Conventional cables
3.0 12.0 14.0 2.5
Conventional cables Birdeage cables Birdcage cables Conventional cables
INSTRUMENTS Instruments placed around the 45-48N stope consisted of extensometers grouted into holes between cable support stations and strain gauges attached to cable supports. Table 1 shows the type of support used in the vie'mity of each instrument. The exteusometers had either five or six
,f
I lift
~
m40 O. if)
I
Fifth liftl ~lme~ll
~
. !
.m.Io.II. II. |
Anchor4 Anchor5 Anchor 6 !
I
m!I n l J an| = auan. l a . auunm m m
._----=..............
20
lift n
0
100
i
I
I
200 300 400 500 600 TIME, days sirme 09/30188
700
Fig. 4. T i m e - d i s p l a c e m e n t c u r v e s for e x t e n s o m e t e r s 2.
It is important to note that in Fig. 4, all the displacement values are positive, which indicates that the rock at each anchor location was moving down and away from the drift on the 6500 level and toward the stope. Displacements of anchors 1 through 5 for extensometer 2 during the first 200 days were most likely caused by floor movement in the drift on the 6500 level.
ROCK MECHANICS IN THE 1990s
24
Third
0 -24
I
Gauge 2
lift . . . . . . . . Gauge 3
Ei2
c~ -48 1
,,.,~< 96 0
.. ... ..... ... . =u o, Gauge5
lift I
I
-24 150
t
I
t
100 200 300 400 500 600
I 240~
I
d
I
'I
I
7OO
I
I I I 250 350 450 550 TIME, dayssince09/30/88
650
Fig. 5. Time-load curves for instrumented cable bolts at stations 12 (A) and 2.5 (B).
Data shown in Fig. 5A were received from the instrumented cable bolt at station 12 (Fig. 2), which is close to extensometer 2 at station 10.5. The data are scattered because the electrical connections periodically shorted out and made contact again, so little information was revealed about the behavior of the rock. This is unfortunate because reliable data would have helped determine the effectiveness of the cable supports. Strain gauges on these cables were the first to be installed by the crew under field conditions, and it is felt that the installation technique may not have been adequate to protect the electrical connections from water. The strain gauges on the instrumented cable at station 14 (Fig. 2) responded very well for 3 months (Fig. 5B), but then were lost due to mining. The data collected at this site did not indicate whether or not the birdeage cables were more effective than conventional cables. Comments from the cable bolting crew indicated that the birdcage cables were more difficult to install without a special cable pusher; however, miners working in the 45-48N stope believed the birdcage cables supported the reek better than did conventional cables. This conclusion was based on over 40 years of mining experience and visual observations of reek in the stope. Results from pull tests conducted by the Bureau show that double birdcage cables can carry almost twice the load per foot of embedment as can double conventional cables [4]. NUMERICAL MODEL A numerical analysis of the 45-48N stope between the 6500 and 6650 levels was conducted using the computer code FLAC. This is a two-dimensional, finite-difference code that uses the explicit finite-difference method to solve basic equations of motion (force = mass × acceleration) and to calculate the out-of-balance forces acting at each node. To model the 45-48N stope region, a computer mesh was developed using the geology and stope geometry at
1433
station 10.5 (Fig. 2). The model assumes simple mine geometry and does not include any previously mined stopes or drifts in the vicinity of the 45-48N stope other than those shown in Fig. 1. The FLAC code allows for cut-and-fill excavation sequences and installation of rock supports, which are extremely important in modeling mining operations. FLAC also allows for the input of different isotropic material properties for each element. In modeling the 4548N stope region, three different isotropic rock materials were used, as well as a sand material that represented the backfill. Modeling began by allowing the model to first reach a state of equilibrium before excavation, that is, a state of zero displacement of the rock at any point in the mesh. Next, an initial lift was excavated by changing the material properties of an area 4.6 m high by about 9.4 m wide (this represented the mined-out area of the stope) from rock to a void. The program was again run and all forces were allowed to reach a state of equilibrium. Eighteenmeter-long cable bolts were then "installed," the area previously mined out was "backfilled" by changing the material properties of the void to simulate backfill, and the next lift was "excavated." This process continued until the fourth lift had been mined; at this point, the stope floor was at the 6590 level. Then a new series of cable bolts was installed. Mining continued for another three lifts (the final lift was 7). The stope was then at the 6545 level just 13.7 m below the drift on the 6500 level. This modeling sequence of simulated cutting and filling was conducted a number of times to help fine-tune the model, that is, to determine which combination of rock properties resulted in the closest correlation with the measurements obtained from the field instruments. Results from the computer runs are given as displacement (in millimeters) of the rock at each node in the mesh as well as for four specific points corresponding to anchefs 2, 4, 5, and 6 for the extensometer at station 10.5 (Fig. 2). These anchors were located 3.4, 11.6, 15.5, and 19.5 m, respectively, from the collar of the hole on the 6500 level. It should be noted that the FLAC program indicates positive displacement as up toward the drift on the 6500 level, which is the opposite direction shown by field results. Results from the field data and the numerical model are compared in Table 2. Field data were corrected for displacements caused by extensometer head movement (indicated by anchor 1) by subtracting the displacement reading for anchor 1 from the initial data for anchors 2 though 6 (Fig. 4). The field data and the calculations from the numerical model correlated fairly well. However, as mentioned before, the simulation was conducted a number of times to help fine-tune the model and determine which combination of rock properties resulted in the closest correlation with the results obtained from the field instruments. The results in Table 2 represent the best correlation. Figure 6 shows a plot of the data from Table 2.
1434
ROCK MECHANICS IN THE 1990s Table 2. Anchor displacement for extensometer and numerical model, millimeters. Anchor
Lift number 5 6 7
2 ......
Initial field data 12.7 15.2 35.6
Correction for head movement -12.7 -15.2 -17.8
Corrected field data 0.0 .0 17.8
Model data -5.1 .0 .0
4 ......
5 6 7
30.5 38.1 55.9
-12.7 -15.2 -17.8
17.8 22.9 38.1
12.7 22.9 22.9
5 ......
5 6 7
25.4 30.4 33.0
-12.7 -t5.2 -17.8
12.7 15.2 15.2
7.6 22.9 22.9
6 ......
5 6
55.9 (1)
-12.7 (1)
43.2
17.8
7
(')
(I)
~Anchor lost through blasting. 50
i
q 20 LU ~10
=
=
o
/[ i
i
"
-10 0 10 20 30 NUMERICAL MODEL DATA, mm Fig. 6. Field versus numerical model displacement data for extensometer at station 10.5. The correlation coefficient for the data is 0.404554, and although this value is low, it does show fairly good correlation between field and numerical model data. Successful use o f F L A C depends on knowing rock mass properties. F r o m this modeling effort, it appears that F L A C could be used to predict rock mass behavior. at least in a general sense, because o f its ability to model cut-and-fill sequences. CONCLUSIONS Data collected from instruments placed in the rock above the 45-48N stope did not indicate whether or not birdcage cables were more effective than conventional cables at supporting the roof. The extensometers worked very well; however, data received from several o f the cable strain gauges installed early in the project were very scattered and revealed little about the behavior o f the
rock. It is felt that the procedures used to install these gauges may not have been adequate to protect electrical connections. Strain gauges installed later in the same rock mass responded more consistently (Fig. 5B). The cable installation crew working i n t h e 4 5 - 4 8 N stope commented that birdeage cables were m o r e difficult to install without a special cable pusher; however, miners working in the stope said they believed the birdeage cables were supporting the rock better than conventional cables. N u m e d e a l modeling o f the 45-48N stope using the F L A C code provided displacement values comparable to field extensometer measurements. At a minimum, the use o f this code could assist ground control engineers in conducting a general analysis o f the behavior o f rock masses in a cut-and-fill mining operation. Based on data collected, it was concluded that extensometers and cable bolt strain gauges a r e practical and effective instruments for collecting essential field data.
REFERENCES 1. Schmuck C. H. Cable bolting at the Homestake Mine. M/n. Eng. (LiRlcton, CO), v. 31, No. 12, Dec., t677-i681 (1979). 2. Goris J. M., Duan F. and Pfarr J. Evaluation of cable supports at the Homestake Mine. CIM Bull., v. 84, No. 947, Mar., 146-150 (1991). 3. Gods J. M. Laboratory evaluation of cable bolt supports (in two parts) 1. Evaluation of supports using conventional cables. U.S. Bur. of Mines RI 9308, 23 pp. (1990). 4. Goris J. M. Laboratory evaluation of cable bolt supports (in two parts) 2. Evaluation of supports using conventional cables with steel buttons, birdcage cables, and epoxy-coated cables. U.S. Bur. of Mines RI 9342, 14 pp. (1990).