Ground control research continues to be a major part of the portfolio reviewed by SHMRAB. The majority of the work was carried out through ECSC and RFCS collaborative projects by Rock Mechanics Technology Ltd (RMT), UK Coal Mining Ltd and the University of Nottingham. HSL also contributed with metallurgical studies of rockbolts. Progress made during 2005 is reported below.
This project, ECSC PR132, was completed in June 2005. The main objectives of RMT’s contribution were as follows:
These objectives were successfully met. The short encapsulation pull test was developed under previous Projects to provide a more realistic approach to the laboratory assessment of rock bolt performance than was provided by previously available tests such as the double embedment test. In particular, the laboratory short encapsulation test (LSEP) provides a simulation of the reinforcement mechanisms mobilised in-situ through the use of rock samples subjected to stress regimes found underground. In comparison other tests are highly artificial and therefore limited in value. The LSEP was adopted as the system performance test in the revised version of BS7861:Part1 for rockbolts which is ready to go out for consultation.
Under this Project, an LSEP methodology was successfully further developed to allow the assessment of flexible long tendon system performance. The methodology was used to provide benchmark performance data for flexible reinforcement support systems intended for embedment in encapsulated resin. Design of rock sample, embedment length and loading systems were optimised. The embedment length adopted, 450mm, was considerably longer than that for rock bolts (160mm) as it was felt necessary to subject the tendons to loads approaching their yield strength prior to bond failure. Benchmark data was determined from comprehensive testing of systems currently available, and it is anticipated that the methodology and benchmark data will be incorporated into BS7861:Part2 covering flexible reinforcement systems for coal mines which is currently being revised. Only one commercially available system, other than the benchmark product, was able to comply with the proposed performance benchmarks.
Work was carried out to further adapt the LSEP methodology for assessment of long tendons embedded in cementitious grouts. This encountered problems associated in particular with maintaining the geometry of the tendon sample outside the test encapsulation length and applying the load to the tendon. Nevertheless, sufficient progress was made to establish the methodology, and consistent results from one type of tendon were obtained. The work to establish performance benchmarks was not completed prior to the end of this European Project but was completed before the end of the accompanying HSE Project, three months later. The embedment length adopted for these tendons was 320mm, as the load transfer characteristics were found to be superior to those of the resin encapsulated tendons and so yield loads were approached with shorter embedments.
The LSEP methodology was also used to assess new rock bolt systems such as a permanently pressurised hydraulically inflated rock bolt, and provided valuable data for comparison of UK rock bolt systems with those used in South Africa and the United States. A combination of laboratory methods, used to evaluate samples installed underground, allowed the assessment of innovative bolt installation systems trialled in South African coalmines, and showed the importance of spin time in optimising resin bond performance. Spin time optimisation windows were quantified. Lack of suitably developed consumables prevented full implementation of an optimised system of bolt installation designed to minimise operator judgment. Nevertheless, sufficient work was done to indicate the value of such a system. Assessment of reinforcement systems employing PUR as the embedment media showed that PUR is unlikely to be suitable for this application until bond performance is substantially improved.
A paper describing this research was presented to the 24th International Conference on Ground Control in Mining in the USA in August 2005.
The objectives of the University of Nottingham’s input to the project were to investigate the use of pre-tensioned bolts and cables in coal mining environment and the beneficial affects of the pre-stressing supports on the immediate strata surrounding a coal mine excavation. The mechanical performance of fully-grouted rock-bolts was investigated in relation to the effectiveness of (i) pre-tensioning of the bolt and (ii) the borehole roughness.
Plane-strain finite-difference models comprising pre-tensioned bolts were examined under both static and dynamic loading conditions. By variation of the rock-mass properties, bolt properties, and by using joint and slip elements in the model, a better understanding of the system was gained.
Direct-shear testing was utilized to investigate the effect of rock surface roughness angle on enhancing the strength of the rock/resin interface. The study suggested that under a high stress condition it may not be an effective practice to enhance the borehole roughness of the bolting system.
This project, ECSC PR135, was also completed in June 2005. RMT was Project Co-ordinator and submitted the final report to the European Commission in November 2005. The overall aims were to develop improved and lower cost methods of in-situ stress measurement for the European coal mining environment and to enable analyses of rock stress concentrations in the proximity of roadways and longwalls in order to decrease the number of hazardous situations in coal mines and improve mine planning.
Three methods of in-situ stress measurement, namely; overcoring the CSIRO HI cell, Laboratory Acoustic Emission Analysis (Kaiser effect) and Borehole Ovalisation were investigated in detail. RMT’s part of the Project mainly concentrated on the Kaiser effect though some work was also undertaken on improving overcoring techniques and instrumentation. It was concluded that overcoring the CSIRO HI cell was a reliable method and that errors due to rock anisotropy, whilst potentially significant, could be corrected with a good knowledge of the rock properties. There remains considerable uncertainty over applicability of the Kaiser effect and it should only be used to supplement information from other measurements. However good agreement was achieved with overcoring measurements from several mines and the effect was shown to occur in some building materials. This requires further investigation. Borehole ovalisation shows considerable promise but requires good input data and knowledge of rock properties.
Work on improved stress analysis produced an improved understanding of the relationship between stiffness and confinement for intact and jointed rock (Nottingham University), showed that it is possible to develop an understanding of the natural stress field through modelling the tectonic, structural and erosional history of a coal basin and developed a means of modelling coal mining induced seismic activity. The applicability of stress analysis through modelling was put into context through a study of uncertainty in geotechnical design.
UoN contributed to the project with workto investigate:
In the first study, the phenomenon of the stress-dependency of rock stiffness was investigated experimentally on four types of rocks: sandstone, siltstone, mudstone and granite. The elastic modulus obtained from the test results showed a general increase with the increase of the confining pressure in triaxial testing, particularly in the broken rocks. Based on the experimental data, new empirical equations describing the relationship between the confining pressure and the elastic modulus were obtained for inclusion in numerical modelling. This experimental study has been further progressed in the continuing RFCS project GEOMOD.
The ability to model the stability of main mine roadways was investigated, applying numerical modelling to two case studies. Each of the two modelled coal mines had its own distinct support system and environment. The first site chosen was UK Coal’s Harworth mine which is an underground mine working at a depth of 1000m. The second was at Hay Royds, a small privately owned coal mine, which used a room and pillar support system at an approximate depth of 200m.
The final part of the research concentrated on a statistical study of the uncertainty of geotechnical design in tunnels. Natural rock’s geomechanical properties are inherently variable so excavation design carries with it a degree of uncertainty. This uncertainty is not represented in the results of numerical modelling, unless the operator declares a factor of safety. UoN researchers have combined the Monte Carlo statistical analysis technique with numerical modelling to vary the parameters used when setting up a numerical model of a typical mine roadway to establish those that have a significant effect on the results of the model and those that do not. The form and distribution of the analysis results has been shown to provide useful and quantifiable uncertainty information about the potential roof deformation of the roadway with respect to input parameter’s uncertainty.
RMT is the Co-ordinator for this project, RFC-CR-03011 (GEOMOD). It mainly involves the development of numerical modelling tools for improved mine layout and support design. This includes considerable rock testing, incorporation of the rock testing results into modelling techniques and improved scaling of laboratory test results to in-situ conditions.
RMT’s work has concentrated on developing new three-dimensional modelling techniques to solve practical problems of mine layout and support design. During 2005, RMT has been able to perfect a means of using a simplified three dimensional strain softening model to examine rockbolt and cable bolt support row spacing, which was not possible with the previous 2D modelling. This is now being used routinely in consultancy work to advise on optimum support systems.
RMT also commenced work on the HSE supported modelling of pre-tensioned cable bolts. An exercise was undertaken to compare the behaviour of tensioned and untensioned cable bolts and truss supports in the event that the main rockbolt reinforcement system was to fail. Care was taken to simulate in the model the installation processes for each of these support systems. The models showed little or no advantage in tensioning either cable bolts or trusses and indicated that the cable bolts provided much more roof control under moderate stresses. However the truss systems did provide an eventual safety net to catch the roof when it had completely broken, though this did not prevent high levels of convergence occurring. There is still much more work to be done on modelling tensioned support systems. A current failing of the modelling is that it does not take into account any remobilisation of inter-bedding friction lost through roof sag prior to cable installation. New modelling techniques are being examined to achieve this.
Good results have been achieved in improved modelling of inter-seam interaction effects, particularly for cases where new extraction districts are planned above previously worked panels. This can lead to highly problematic support conditions as was found when the Stanley Main was worked above the Barnsley longwalls at Riccall mine. Improved predictability of conditions is required to optimise such layouts and ensure adequate support. One of the methods used to do this has been the use of displacement discontinuity elements to represent the goaf area.
Another approach examined under the Project has been the statistical analysis of telltale data to identify “typical” rock bolted roof behaviour for a particular set of geotechnical conditions. This has proved very valuable in both identifying the range of outcomes possible and for predicting probable support conditions in new areas of a mine, especially under unusual stress conditions. It is anticipated that this approach will be further developed to examine the behaviour of tensioned cable bolts.
UoN’s contribution aims to gain a comprehensive understanding of the mechanical properties of coal and coal measure rocks is required. In particular, the effect of confining pressure and time on the mechanical behaviour of selected coal measure rocks was reviewed, experimentally investigated, and numerically implemented.
As reported above under PR 135, laboratory tests on fractured coal measure mudstone showed an increasing contribution of strain rate to the variation in stiffness and strength with increasing confinement. An anti-trigonometric function was fitted to these results and implemented in FLAC4.0. The FLAC model of confinement/strain rate against stiffness was constructed and found to fit the laboratory test data well.
Creep testing was carried out on UK coal measure rocks; the result data was fitted to Burger’s model but more work is required to scale from laboratory to the field. The effect of rock damage condition on creep deformation properties was found significant. Therefore, a double creep model is proposed to include pre- and post- failure creep properties. Under this task, a numerical model of a generic prototype of a coal mine is being used to validate the new procedures of double creep properties, i.e. both intact and broken materials will have their own creep properties.
Work on this project continues until mid-2006.
This new RFCS Project, RFC-CR-05002 (MONSUPPORT), commenced midway through the year. RMT’s main activities will be continuation of the comparisons of tensioned and untensioned cable bolt behaviour, feeding back to the Committee revising BS7861:Part 2, and continued development and application of geotechnical instrumentation.
Work commenced on laboratory investigations for development of systems for measuring the relative performance of tensioned and untensioned flexible rock reinforcement consumables. The emphasis was on completing work on untensioned systems. The testing comprised a suite of double embedment shear tests on Mini-Cage cable bolts. This has now been completed and recommendations made regarding the tests and acceptance criteria for untensioned systems to be incorporated into a revised British Standard on Flexible Reinforcement Systems.
During the first six months of the Project RMT examined many physical principles which are, or have the potential for use in, coal mining geotechnical monitoring tools As a result of this analysis RMT have prioritised developing the following measuring devices during this Project:
RMT commenced work on development of an optical integrity monitoring rock bolt based on the principle that a broken optical fibre will no longer transmit light. A special bolt which utilises this principle successfully would indicate its integrity visually without any requirement for special interrogation equipment. A programme of laboratory tests was undertaken investigating different types of fibre, sheathing methods and adhesives. Two sets of prototypes were produced, one based upon the transmission of light from a cap lamp and one based upon using an integral radio isotope light source. Both types of “Optobolt” were installed underground by UK Coal at different mines and performance information from these trials is awaited.
A spring loaded, two way, remote reading extensometer has been developed for use in coal mine roof, ribs and floor. Previous models relied upon gravity for tensioning the anchor wire and were therefore unsuitable for measurements in non-vertical geometries.
New designs of four way and three way remote reading extensometers have been developed. These differ from previous designs in that each bay is measured independently with a full 75mm range of displacement. These also incorporate spring loading, and so can be used in non vertical applications. A prototype groutable extensometer has also been developed and this is being trialled in a brick lined rail tunnel in the UK.
Two ATEX certifications for RMT’s remote reading telltale and extensometer systems were applied for and granted. One covered a battery powered, handheld portable readout. This provides considerable additional versatility to the system. It enables the system to be used without the need for full scale surface monitoring or an underground IS power supply. The second ATEX approval has enabled the 3 and 4 way transponders described above to be covered by the ATEX certification and has removed the need to supply a long length of cable attached to each transponder making them lighter and less vulnerable to damage during transportation.
Sixteen four way, spring loaded, remote reading extensometers were manufactured for measurement of roadway conditions in a trial of rockbolting at a coal mine in the Czech Republic. Six, two anchor, spring loaded, remote reading extensometers, were manufactured and supplied to an Indian coal mine for monitoring rib dilation during de-pillaring operations. Twelve, two anchor, spring loaded, remote reading extensometers were manufactured and installed in the hanging wall of a haulage roadway in a South African Copper mine. In all the above cases the transponders will be read from underground with a handheld portable readout. The performance of all these instruments will be monitored during the Project.
A full remote reading telltale system, comprising 13 dual height telltales, was installed in July 2005 in 301’s maingate at Daw Mill Colliery. This was the first such installation of the system in a UK mine and will allow a thorough evaluation of the system hardware and software, which will be updated as it is improved. The site at Daw Mill was chosen as the monitored section of maingate was to be sealed off after completion of face production in order to prevent spontaneous combustion before it needs to be salvaged to equip the next face. It was therefore important to monitor the condition of the sealed section of rockbolted maingate remotely. The system has now been running at Daw Mill for 6 months with excellent reliability.
Work has commenced on identifying whether it will be possible to develop a replacement for the various models of sonic extensometer, which have been discontinued by their manufacturers due to the obsolescence of a common component. Their loss is a major issue worldwide. Initial indications are positive with the additional possibility of incorporating a data logging facility in the new instrument. Work has also commenced on developing a new resistive strain gauge monitoring instrument for the industry, which will also include a data logging facility.
HSL completed a study of the two widely available rockbolts to ascertain if one would be more susceptible to failure than the other. Rockbolts are designed to prevent roof falls in mines by providing support to the rock strata that make up the roof. They are generally inserted into holes previously drilled into the roof and held in place with resin. In addition to the potential axial load on the bolts due to loose strata, it is possible that the different strata layers can move horizontally in respect to each other. Under these conditions the bolt can experience shear and bending stresses. The actual loading experienced can be very complex and depend on a number of factors, such as rock hardness, resin strength and resin bonding.
Numerical modelling of the full loading conditions would be highly complex, requiring software that could model a large range of different material types and failure modes, and would require very detailed data on the material properties of the resin, rock and the bolts. The results would then be applicable to the conditions modelled, and may not be representative of other operating conditions, with different resin or rock hardness, for example. The HSL work therefore concentrated on the difference between two of the different geometries of rockbolts available, to ascertain if one would be more susceptible to failure than the other. This was achieved by looking at different loading cases; simple axial loading, simple bending and combined axial bending, and comparing the results to those of a plain cylinder of similar diameter. The main findings were:
Following a fall of ground at a coal mine, the recovered halves of six broken rockbolts were submitted to HSL by HSE’s Mines Inspectorate for metallurgical examination. The rockbolts were examined visually and a series of mechanical tests were carried out. The materials and manufacture of the bolts were compared with BS 7861-1:1996 Strata reinforcement support system components used in coal mines – Part 1: Specification for rockbolting, and with the draft revision of this standard, and an assessment was made of the likely mode of failure of the bolts
The investigation has been completed and the following conclusions drawn:
Several KT rockbolts, which appeared to have failed with low levels of plastic deformation, have been recovered by Mines Inspectors from falls of ground in UK coal mines and examined by HSL. Tests on some batches of these rockbolts indicated that the steel was of relatively low toughness and hence may be susceptible to brittle failure at stress concentrations, such as may be provided by the rib profile. It was also thought that where strata movement was concentrated at one horizon in the mine roof, that the enhanced rib profile of the KT rockbolt may prevent any debonding of the resin occurring and hence concentrate the load on a very short length of the rockbolt. In this situation, although the rockbolt may be behaving in a ductile manner on a microstructural scale, the short effective gauge length may result in a failure with minimal global plastic deformation.
HSL carried out a series of tensile tests to evaluate the effect of the rib profile of the KT rockbolt on its in-air mechanical properties, and to investigate whether increased resin bond strength resulted in a low elongation to failure in the KT rockbolt when the initial gauge length was small. The experimental work on this project has been completed and the report is in its final draft stages. The main findings were: