Abstract
Understanding the as placed stress distribution in mine backfills has been a long-standing challenge with significance for rational backfill engineering analysis and design. Specifically, analysis to define backfill strengths required to ensure stability upon exposure (i.e., mining beside or beneath backfill) necessitates assumptions for initial stress distributions within the backfill.
Most theoretical studies of the stress conditions have focussed on “arching” models first proposed for understanding stresses in uncemented granular materials in silos or backfilled trenches. In the 1980’s a series of detailed field and laboratory investigations at Bolliden’s Näsliden mine confirmed that the theories were reasonable for estimating stresses in uncemented hydraulically placed backfill in stopes up to 17 m wide and 50 m high. Since then, substantially more detailed investigations in Cemented Paste Backfill (CPB) have been carried out, but these field results have not been systematically compared with extended arching theories that account for backfill cohesion arising from binder hydration.
The work considered in the proposed paper begins by summarizing CPB field stress measurements from Williams, Kidd, Cayeli, Casa Berardi, Kenowna Bell, and Red Lake mines. It is found useful to present the results in normalized form, using normalized depth d/L (where d is depth below backfill surface and L is the stope’s horizontal hangingwall-to-footwall length) and normalized stress σv/γL. These results are then compared with elastic stress analyses.
The motivation for the elastic analysis is the fact that field sampling and bulk property analyses from Williams, Kidd, Cayeli, and Neves Corvo mines (and, to a lesser extent, Casa Berardi) suggest that the backfill hydrates sufficiently before effective stresses are applied to it such that the subsequent backfill response is elastic. (This phenomenon is explored in detail for Williams mine and is the subject of a companion paper submitted to the conference.) It is found that the normalized elastic response adequately describes the average of the field measurements and captures most field measurements within ±40% of predicted values. Furthermore, application of arching theory provides only a lower-bound estimate of stresses, at best.
Given that an elastic backfill response adequately represents measured field stresses, it is not necessary to consider further effects of failure along the stope sidewalls (in fact, the results suggest such hypothesized failure does not occur). This has important implications for certain design scenarios, notably undercut analysis for tall stopes, as will be considered in this paper.