Anchorage and Failure Mechanisms of Fully Encapsulated Rockbolts: Stage 2

The results of this second stage project have contributed to the state of knowledge and further understanding of the mechanics of fully encapsulated or full column rockbolts (FERB). The project included several investigations intended to improve current rockbolting practices in the field and into the mechanism of rockbolt failure.

Background
The project followed the earlier Australian Coal Association Research Program (ACARP) project C7018. Recommendations from the peer review of that project concerning improvements to the test facility and follow-up investigations formulated the scope for the second stage project.

The defined objectives of the Stage 2 project, with funding of $135,000 provided by ACARP were to:

• enhance and upgrade the control and monitoring capabilities of the testing facility established in the earlier project;
• modify and refine the testing procedures to ensure more reliable research outcomes; and
• expand on the earlier research activities covering a greater level of detail and a wider range of principles that affect the performance of rockbolts.

The project was again undertaken as a collaborative venture combining the collective expertise at the University of New South Wales (UNSW) and Strata Control Technologies (SCT) to ensure the maximum benefit would be gained from the project.

Management &Communication
The project was principally undertaken in two phases over a two and one half year period that began in April 2001. A progress report detailing the interim findings from the first phase was completed in April 2002. The project involved several students in the School of Mining Engineering; seven students at Honours level and one student at Masters level.

Results of Investigations
Considerable insight has been gained as a result of this project into understanding the effect of changes in a range of parameters on anchorage performance as well as the load transfer behaviour and failure mechanisms of fully encapsulated rockbolts.

Several investigations were conducted using the upgraded test facility; these included an investigation into the effect of changes in the design as well as the conditions concerned with the installation of fully encapsulated rockbolts. In particular the investigations focussed on the effect of changes in:

• resin annulus;
• deformation geometry of the rockbolt; and
• resin spin time.

A further investigation was undertaken to better understand the nature of load transfer of a fully encapsulated rockbolt. The various findings are as follows.

• In terms of load transfer, load was found to decrease with distance along an encapsulated rockbolt. The load being greatest nearest the rock free surface (that is nearest the point of load application), gradually reducing with distance along the rockbolt into the rock mass. This finding supports the theory that an externally applied load on an encapsulated rockbolt is gradually transferred into the rock mass, creating a localised stress field that aids in clamping together the rock mass. This finding was predicted in the numerical modelling work by Whitaker and reported earlier in the Stage 1final report. There has been some conjecture as to the nature of the load transfer function. To better understand this, two arrangements of load application were investigated. One arrangement modelled the standard rockbolt pull-test. This tends to confine the surface of the rock surrounding the rockbolt. The second arrangement modelled loading resulting from the separation of the partings in a rock mass. In both cases load transfer occurred over the entire albeit short length of encapsulation. Minor differences were observed in the load transfer function between the two loading arrangements. An earlier test indicated a non-linear rate of load reduction with distance using the first arrangement while in the second arrangement the trend appeared to be linear. In a subsequent repeat test using a rockbolt with twice the number of strain-gauges, differences in the load transfer function between the two loading arrangement were less apparent. Significantly, load transfer appeared to be nonlinear in both cases with 50% of the load transfer taking place within approximately 54 mm or 2.5 rockbolt diameters from the free surface. Interestingly cratering of the test sample immediately surrounding the rockbolt was observed in the laboratory tests with the second loading arrangement, a result that was predicted following the numerical modelling analysis undertaken in Stage 1.
• Following on from these results in terms of the mechanism of rockbolt failure, when a high load is applied to a rockbolt the resulting stress induced in the surrounding rock mass may exceed its strength causing localised rock failure. It is reasonable to consider that the load will transfer along the rockbolt until either confinement of the rock mass is sufficient to withstand the load transfer; the rockbolt intersects a higher strength rock mass; or, the end of the rockbolt is reached. In either case this transfer of load extends the length of localised failure with permanent de-coupling of the rockbolt from the rock mass.
• The strength or load bearing capacity of FERB was found to be independent of changes in resin annulus where resin annulus remained less than 4mm. With greater resin thickness there was a reduction in the magnitude of anchorage strength and an increase in the variability. In addition, the stiffness of the anchorage system reduced as the influence of the properties of the resin in the anchorage system became more prominent. A near halving in anchorage strength and system stiffness was observed with an annulus size of 9mm compared to 2mm. This indicates that the extent of acceptable tolerance when drilling for rockbolt installation is around 4mm.
• Underspinning or overspinning of a fast-set resin cartridge had the same result in terms of diminishing the anchorage performance of a rockbolt and are equally undesirable in practise. Both had a pronounced negative effect in terms of increasing the variability in rockbolt performance. More consistent performance was achieved in those tests that were closer to the recommended spin time.
• A near doubling in anchorage strength was observed with a changeover to a mix-and-pour resin from resin cartridges when anchoring a rockbolt. There was also greater variation with resin cartridges even under supposedly the same conditions.

Future Actions
While the insights gained from Stage 2 represent an advance in the state of knowledge of the mechanics of rockbolts as well as some of the factors involved with the practise of installing rockbolts, it has lead to even more fundamental questions being raised. For example:

• how much difference is there in the anchorage performance of the different types of rockbolts currently available on the market?
• what is the critical range in spin times for other resin cartridges on the market?
• what is the effect of glove fingering on anchorage performance with different cartridge designs - length of encapsulation, type of plastic etc?
• do highloads lead to de-coupling of a rockbolt from the rock mass through load transfer?
• what influence does resin annulus have on the effectiveness of load transfer between a rockbolt and rock mass?