This paper presents a methodology, procedures and standards for sillmat design based on three inter-related engineering modeling approaches: centrifuge modeling, analytical modeling and numerical modeling. The study aimed at establishing the effect of stope geometry and stope wall roughness on sillmat strength behaviour, when exposed by undercut mining. The results of the modeling techniques suggest that, for stopes with smooth wall rock conditions, sillmat failure is driven by the fill self-weight and has minimum dependency on the fill binder content. For stopes with rough wall rock conditions, however, wall roughness contributesignificantly to the stability of the sillmat during underfill mining. This study indicates that the stability of paste fill sillmats can be designed with confidence using a combination of the three modeling techniques.
Sillmats are structural elements upon which the safety and economics of overhead mining in steeply dipping ore zones of moderate width, during overhand cut-and-fill mining, underfill mining or sill pillar mining, depend. Generally, sillmats are used to support uncemented or low cement content backfill. Conventional sillmats are cast from cemented sand backfill materials, often undedain by a timber mat; these structures have to be selfsupporting when exposed by undercut mining. Improper design would result in failure of the fill mass, and extensive economic losses associated with loss of production and ore dilution, as well as in safety problems. Sillmat, in this study, refers to the entire cemented paste fill column in the stope, and which will eventually be undercut by mining.
In this paper, centrifuge model studies were combined with analytical and numerical modeling analysis to provide views into the behavior and potential failure modes of mine paste backfill sillmats. Centrifuge modeling was the primary tool used not only to dynamically test sillmat performance on a time-dependent basis but also to accommodate the three-dimensional aspects of the problem. Analytical modeling was carded out using limiting equilibrium analysis, based on a method introduced by Mitchell and Roettger (1989). Numerical modeling was carded out using FLAC (Fast Lagrangian Analysis of Continua), a powerful two-dimensional elastic plastic-finite difference code.
The modeling study incorporated the effect of stope geometry and stope wall roughness on sillmat behaviour and stability performance. Models were prepared to simulate two different stope conditions; stopes 3 m wide, 15 m long, and 30 m high, typical of mining Block 4 of the Battle Mountain GoldGolden Giant Mine, and stopes 5-7.5 m wide, 15 m long, and 40 m high, typical of Block 5. In all cases the stope walls were inclined at 75 ø and smooth, medium-rough and rough rock wall conditions were established for simulating typical boundary modes. Wall closure effects associated with such narrow stopes, surcharge loading and blasting effects were not considered in the study. Paste fill was prepared at 80% pulp density using unclassified tailings mixed with Type 10 Normal Portland cement (NPC) and Type C fly ash (FA). All recipes were prepared at 7% binder content and cured at 14, 28 and 56 days.
