Introduction
Rockbolts are increasingly relied on as a key component in the primary support mechanism of many underground mines. In the Australian coal mining industry, for example, over 5 million rockbolts are installed each year at a cost of over $A35 million. Previous research by UNSW, Strata Control Technology Pty Ltd (SCT) and Powercoal Ltd has found that over 30% of rockbolts ‘are not providing optimum performance in coal mining environments’ (Galvin et al 2001).
A research initiative has been launched combining the skills and experience of industry and research expertise in the university to develop an understanding of fully encapsulated rockbolts. The broad objective being to improve the performance of rockbolt systems and hence improve the overall safety in mines. This initiative resulted in the establishment of a test facility at UNSW that operates within a controlled laboratory environment.
As part of this initiative, the research aimed to quantify the sensitivity to changes in various rockbolt parameters on anchorage performance. This paper outlines the results found to date with regard to the thickness of the resin annulus on anchorage performance of fully encapsulated rockbolts.
Test facility
Design objectives
The desirable attributes of a rockbolt test facility were seen as:
· the facility should be capable of examining a wide range of parameters associated with the installation of rockbolts and of replicating a wide range of conditions;
· tests should be carried out under controlled conditions to better ensure the repeatability of results;
· the facility should be available for use by industry (both suppliers of rockbolt systems and industry end-users) for such purposes as independently assessing the performance of new products or changes in the method of installation.
The design of the new test facility incorporates a hydraulic ram similar to that used in most rockbolt pull-out tests. The ram can apply various load conditions to a rockbolt. A bi-axial cell is used to hold the test specimen containing a fully encapsulated rockbolt. The test specimen may either be a sample of rock replicating the conditions in a particular mine or, a man-made material. The advantage of the latter is it mitigates many of the problems that can arise due to the variability in material properties between rock samples.
Facility features
The test facility at the UNSW Mining Research Centre uses a modified workshop lathe as the test platform. The main components of the facility include:
· a bi-axial cell with an internal diameter of 145 mm, length of 200 mm and rated maximum confinement pressure of 30 MPa mounted to the bed of the lathe;
· servo-control hydraulic system used for precise control of the loading rate of a 300 kN capacity hollow core ram during a pull-out test;
· computerised system to control the applied load through the hydraulic system and monitor the actual load and displacement of the rockbolt.
Test sample preparation
Test samples
A cementitous grout (Celtite MG75S) was selected in place of cored rock samples in the test program. The grout had a strength of approximately 75 MPa.
In order to ensure uniform material properties, a single batch of over 100 test samples was prepared and cast in plastic moulds. Each core had a diameter of 145 mm and length of 200 mm. The samples were cured for at least 28 days before testing.
A hole was drilled in each sample to a depth of 175 mm using a drill rod mounted on the lathe. A chisel bit was used for the 26 mm hole and finger bits were used for the larger holes. A constant rotation speed and feed rate was used with water flushing to ensure uniform roughness of the borehole.
Rockbolt anchorage
A Celtite 24 mm extra high strength CX rockbolt was used in the test program with a basic profile design as shown in Figure 1. The rockbolt has an inner core diameter of 21.7 mm, a diameter across the ribs of 22.8 mm and rib spacing of 10 mm. The rockbolt has an ultimate tensile strength of 344 kN.
Figure 1. Profile of the rockbolt used in the test program.
Resin cartridges were initially used for anchoring, however, problems with poor mixing between the resin and catalyst and with the plastic packaging led to inconsistent results. The poor mixing was exacerbated by the short length of encapsulation.
A mix-and-pour resin was subsequently used in the test program. After mixing, the resin was injected into the hole into which the spinning rockbolt was rammed. The rockbolt was supported in the chuck while the resin was allowed to set for 10 min. The resin was then left to cure for a further 48 h with the rockbolt and sample standing vertically.
It was observed that with the change from a cartridge resin to a mix-and-pour resin there was a near two-fold increase in the maximum pullout load to approximately 200-250 kN.
Experimental program
Procedure
In summary, the test procedure involved a load being applied between the rockbolt and end surface of the test sample. This tensile load is intended to simulate the induced load on a rockbolt when separation occurs between partings in rock strata.
Figure 2. Test set-up showing the arrangement of the bi-axial cell, hydraulic ram, pressure transducer and LVDT.
During each test, the outer surface of the test sample was subjected to a confinement of 10 MPa within the bi-axial cell. Before a pullout test began, a valve was closed to stop the flow of hydraulic fluid to the cell. The level of confinement simulates in situ field conditions but it was also the minimum level necessary to support the sample in the cell during drilling and pullout test.
A pressure transducer monitored any pressure change in the bi-axial cell during each test.
The tensile load was applied by a hollow core hydraulic ram as shown in Figure 2. A pressure transducer measured the load on the rockbolt. An LVDT measured the displacement of the rockbolt as it was drawn out from the hole by the ram. The data acquisition system recorded the load of the ram and in the bi-axial cell and as well as the displacement at a rate of 20 readings/s.
Test parameters
A combination of the load on the rockbolt and its displacement was used to assess the performance of the rockbolt anchorage system. Using this data, a load/displacement curve was drawn after each test. Based on this curve, the following could be determined.
· Maximum pullout load (or MPL), that is the peak resistance sustained by the anchorage system.
· Stiffness of the system within the elastic region.
· Displacement required to achieve the MPL.
· Stiffness in the post-failure region.
· Residual stiffness of the system.
· Change point from post-failure to residual resistance.
· Load resistance at a nominal displacement of 50 mm.
Results
The effect on anchorage performance of increasing hole diameter while maintaining rockbolt diameter was investigated and has been reported (Hagan and Weckert, 2002). Hole diameters of 26, 28 30 and 32 mm with corresponding thickness of the chemical resin annulus of 2, 3, 4 and 5 mm were examined. The test at each resin annulus was replicated up to six times.
Observations
Reasonable repeatability was observed for each level of resin annulus as illustrated in Figure 3. This figure shows the load/displacement curve for the 3 mm annulus test.
Figure 3. Load/displacement curve for an anchorage system with a 3 mm annulus.
A different behaviour was observed when the hole was opened out to 32 mm when the resin annulus reached 5 mm as shown in Figure 4.
Figure 4 illustrates that the results were again reasonably consistent. However, while initially the stiffness of the anchorage system was similar to that observed at smaller resin annulus, above a load of about 40 kN and in one instance 110 kN, a change occurred that resulted in a much greater displacement before maximum load was achieved. This displacement was of a similar magnitude to the rockbolt rib spacing.
Figure 4. Load/displacement curve for an anchorage system with a 5 mm annulus.
The results from the test program are summarised in Table 1.
Table 1
Summary of results
|
|
|
Annulus thickness (mm) |
|||
|
|
units |
2 |
3 |
4 |
5 |
|
Limit of elastic behaviour - load |
kN |
180 |
180 |
190 |
60 |
|
Limit of elastic behaviour - displacement |
mm |
1.50 |
1.80 |
1.60 |
1.00 |
|
Stiffness within elastic region |
kN/mm |
99.4 |
85.0 |
100 |
40.0 |
|
Maximum Pullout Load (MPL) |
kN |
225 |
245 |
240 |
185 |
|
Displacement at MPL |
mm |
4.80 |
5.10 |
4.80 |
13.5 |
|
Stiffness within post-failure region |
kN/mm |
-16.9 |
-23.0 |
-26.6 |
-16.3 |
|
Residual load at post-failure/residual inflection point |
kN |
125 |
120 |
125 |
90 |
|
Displacement at post-failure/residual inflection point |
mm |
11.5 |
12.5 |
12.0 |
23.8 |
|
Residual stiffness |
kN/mm |
-1.49 |
-1.42 |
-0.94 |
-1.27 |
|
Residual load at 50 mm displacement |
kN |
60 |
70 |
90 |
45 |
There was little measurable change observed in the pressure of the bi-axial cell during each test. The experimental noise tended to mask any changes that might have otherwise occurred. It might be expected that some change would occur during a pullout test as the resin dilates and contracts with the movement of the rockbolt ribs. Unfortunately the current monitoring arrangement tended to even out any transient changes in stress that might occur along the length of the test sample. Alternate arrangements to monitor any induced stress changes are being considered in future experiments.
Analysis
Little difference was observed in the curves for resin annulus thicknesses of 2, 3 and 4 mm as indicated in the summary graph in Figure 5. The performance of the anchorage systems in these instances exhibited a relatively high as well as consistent level of stiffness up to the point of maximum pullout load (MPL); the latter being the maximum load bearing capacity of the anchorage system.
Figure 5. Superimposed curves showing the average results for each annulus thickness.
This initial elastic behaviour reflected the material properties of the rockbolt component in the anchor system as well as the cohesiveness between the rockbolt, resin and rock. As the MPL is less than the UTS of the rockbolt, the MPL is likely to indicate failure of either the resin/rock or resin/rockbolt interface or both.
Beyond the MPL, the resistance to the externally applied load fell away with further displacement of the rockbolt until a residual resistance level was reached for the anchorage system. It is interesting to note that this residual resistance still represented a reasonably high value equivalent to about 70% of the MPL.
Consequently even after failure of the resin interface, a fully encapsulated rockbolt can still provide an appreciable level of resistance against separation of rock strata.
It should be cautioned, however, that the level of this residual resistance might be dependent on the nature of material properties of the surrounding rock mass and further testing would be required to confirm this.
In each case, except for the 5 mm annulus, the limit of elastic behaviour and the MPL were fairly consistent at approximately 180 kN and 240 kN respectively. The lowest MPL and post-failure stiffness were both associated with the smallest annulus which may indicate the need for a minimum amount of resin to ensure good bonding and load transfer between a rockbolt and rock.
The results indicated that a change in anchorage behaviour occurred at 5 mm. At this annulus it is possible that the material properties of the resin come into play and the resin no longer was solely a medium to facilitate the transfer of load between the rockbolt and rock.
The impact of too large a resin annulus was a reduction in the MPL. In this test program it was found that a change from 4 to 5 mm led to a near 25% reduction in the load bearing capacity of the anchorage system. Extrapolating this observed behaviour to even larger hole diameters and hence a greater resin annulus, it is possible that at best the MPL for the anchorage system would be maintained at this lower level but it is more likely that it would reduce even further. Follow-up tests will be required to confirm this trend.
Significantly at this large resin thickness there was a corresponding three-fold increase in the amount of displacement needed before the peak load was achieved.
In practical terms this would indicate that for large resin annulus, a higher degree of relaxation in the bedding, i.e. a greater amount of separation between strata, must occur before the same load is achieved compared with a smaller annulus system. Alternatively, the ability of the anchorage system to resist separation reduces with large resin annulus. Hence the anchorage system with an unduly large resin annulus is less likely to act as an effective rock support mechanism.
Conclusion
The test program indicated that there was an optimum range of resin annulus thickness within which there was little change in the performance of a fully encapsulated rockbolt anchorage system.
Either side of this optimum range there was a reduction in the MPL as well as other properties of the anchorage system. For example, it was found that for the case of a 21.7 mm rockbolt used in the test program when resin annulus reached 5 mm in a 32 mm diameter hole, there was a reduction of nearly 25% in MPL from that achieved within the optimum annulus range. This can significantly degrade the capability of the rockbolt to bind together rock strata. It is yet to be demonstrated whether the optimum range of resin annulus and hence allowable tolerance of the hole diameter varies with the diameter of a rockbolt.
The test program also indicated that a fully encapsulated rockbolt anchorage system can still provide a reasonable level of resistance to the separation or relative displacement between strata even when the maximum load bearing capacity of the anchorage system has been exceeded.
These findings are in general agreement with recommendations by suppliers of rockbolt systems. The findings impress the importance of matching the correct hole size for a given rockbolt diameter.
Acknowledgements
The author acknowledges the support of the Australian Coal Association Research Program (ACARP) for funding the research project. The project has also been supported by Celtite Pty Ltd which provided advice and supply of test materials and, by Strata Control Technology Pty Ltd. The author wishes to thank the contributions made by Steven Weckert and Daniel Peel to the project.
References
Galvin, J M, Offner, J C, Whitaker, A, Fabjanczyk, M and Watson, J O, 2001. Establishing anchorage and failure mechanisms of fully encapsulated roof support systems – End of grant summary report. ACARP Project C7018.
Hagan, P C and Weckert, S, 2002. Anchorage and failure mechanisms of fully encapsulated rockbolts (Stage 2) – interim progress report. ACARP Project C10022, April.
Original version of manuscript published in Proceedings of 10th International Conference on Rock Mechanics, September, 2003, (South African Institute of Mining and Metallurgy: Johannesburg).