INTRODUCTION
Measurements in the deep tabular excavations of the South African gold mines indicate significant time-dependent deformation. Although the host rock consists of hard brittle material, creep-like movements of the order of 0.6 mm/h have been observed in certain areas. This paper describes measurements of time-dependent deformation in three different geotechnical environments. The time-dependent behaviour is the result of the theology of the mining-induced fracture zone surrounding the excavations at these depths, while pre-existing discontinuities such as bedding planes also play a prominent role. It appears that detailed monitoring of stope closure profiles can provide a useful signature of the geotechnical response of the rock mass to the mining process. The closure profiles are simulated using a displacement discontinuity framework that allows for the time-dependent failure processes. Preliminary studies indicated that the initial jump in stope closure after blasting is correlated with the stope face stress.
The design of deep level mine layout configurations and the implementation of effective strata control strategies is often reliant on direct experience of mining conditions in different geotechnical environments. If a basic understanding of the underlying failure processes and mechanisms is available, it is possible that these strategies can be refined to reduce the risk of rockburst and rockfall incidents. Considerable research effort has beon directed over many years towards accomplishing this understanding by detailed observations of the fracture zone surrounding deep level stopes (Leeman 1960, Adams & Jager 1981, Bmmmer & Rorke 1984, Legge 1984) and by the development of theoretical models of stress redistribution in the rock mass (Salamon 1974, Cundall et al. 1994). More recently, researchas shown that deep stopes in hard rock undergo significant time-dependent deformation (Malan 1998). The improved understanding of the time-dependent rock response provides a firm foundation for the assessment of different mining rates or multiple shift operations (Napier & Malan 1997). In this paper, continuous closure data collected in various geotechnical areas of the South African gold mining industry are presented to illustrate the characteristic response of the rock mass in each area. The time-dependent failure processes are modelled by means of a random discontinuity assembly in a viscoplastic displacement discontinuty framework. The simulated behaviour of a small span stope is described to illustrate the correspondence between the closure in the model and the field data.
