The primary objective of the research reported in this thesis is to assess the suitability of applying the Distinct Element Method (OEM) numerical modelling technique to analysing underground coal ·mining problems. The DEM was first proposed by Cundall (1971). It uses a discontinuum approach which treats the rock mass as an assemblage of separate or distinct blocks that interact with adjacent blocks at joints. A description of the basis of the DEM is given in Chapter 2 together with the capabilities and limitations of the computer program UDEC (Itasca 1988) which was used to conduct the analyses reported in this thesis. Past applications of the OEM include analysing hard rock surface and underground mining where the deformability and weakness of the rock mass is associated with the joints rather than the strong stiff rock blocks. The OEM has not been applied to any significant extent in analysing underground coal mining operations . which occur in stratified, weak, highly deformable rock masses. This research was designed to evaluate the ability of the OEM to simulate with sufficient accuracy, the response of the rock mass and hence excavation stability and ground support performance to underground ·coal mining. As such the research is based on the application of an available working computer program to new situations rather than the development of a new OEM code.
The performance of program UDEC and hence the DEM is assessed by comparing model predictions (see Chapters 5 to 7) with in situ measurements and observations from two longwall mines. These two case histories, which are detailed in Chapter 4, were selected due to the significantly different in situ conditions that prevail at each mine site. Comparing model performance for the two case histories enabled the ability of the DEM to predict rock mass response for a wide variety of conditions to be evaluated.
Both the overall response of the rock mass to longwall mining (see Chapters 5 and 6) and the detailed behaviour of the rock mass and stability of the small gate roads adjacent to the longwall (see Chapter 7) were analysed. The overall analyses studied large scale stress redistribution and ground movements in the longwall abutments, the interaction between adjacent excavations, and the influence of stress direction on excavation stability. The detailed analyses considered small scale rock fracturing which lead to gate road roof collapse, and the interaction between the rock mass and ground support.
Results from the comparative analyses were used to construct calibrated models to predict rock mass response to different mining scenarios (see Chapter 8). A generalised set of model design guidelines (see Chapter 9) were developed to provide a starting point for conducting future DEM numerical analyses of coal mine excavation stability. Conclusions drawn from the analyses and recommendations on mine design, improvements to the DEM and future research topics are summarised in Chapter 10.
Comparison of numerical analysis results with in situ measurements and observations demonstrated the DEM is suitable for predicting rock mass response to underground coal mining. Realistic model response was obtained for both case history studies when laboratory test values were used in the analyses. This indicates that only a minimal number of calibration assessments will be required prior to conducting predictive mine design analyses.
For reliable predictions to be obtained good quality information concerning the distribution of joints in the rock mass and its behavioural mechanisms are required. In coal mining, the· dominant mechanisms affecting stability are shear movement and separation along the pervasive bedding planes, bending and rotation of the strata, and intact rock fracture. This last mechanism tends to dominate excavation stability and hence ground support performance due to the weak, brittle nature of coal measure rocks.
As the computer program used for the research does not permit fracturing of intact fully deformable blocks once the analysis commences, special high peak strength joints were placed at potential fracture sites. Although this approach is successful, it requires prior knowledge on the location of rock fracturing. To improve the predictive capability of the OEM, the computer program must incorporate the ability to automatically fracture intact blocks during analysis.
The main disadvantage of the OEM is its requirement for extensive computer processing time· which limits its outline application to mine design. To improve computational efficiency, further research is warranted on the following aspects:
- better time step determination
- improved vibration damping
- more efficient house-keeping routines.
When the above improvements are combined with the ability to automatically fracture intact blocks, the Distinct Element Method will become a viable mine design tool for predicting rock mass behaviour, excavation stability and ground support requirements in underground coal mines.