Past finite element analysis of pillar mining effects on the Ross Shaft at the Homestake Mine in Lead, SD, indicated stability of the shaft during mining on both sides of a central pillar that was defined within a much larger original shaft pillar. Experience with pillar mining that began in 1988 verified the earlier forecast of stability. Mining of the remaining central portion of the original pillar began in early 1995. A "cave" occurred in 1996 that led to a temporary cessation of pillar mining; pillar mining began anew in 1997 and continued for a short period. A relatively small portion of the central pillar remains. Previous pillar mining on north and south sides of the central pillar and mining between the central pillar and the Ross Shaft as well as top and bottom sill development have isolated the ore remaining. Thus, one intuitively expects little effect of final mining on the Ross Shaft. Results of the present study confirm this expectation and show very little change in shaft stability as a consequence of mining the ore remaining in the central portion of the original shaft pillar. Comparisons of three-dimensional finite element model displacements with borehole extensometer measurements that began in June, 1944, reinforce this conclusion.
BACKGROUND
The Spokane Research Laboratory, Homestake Mining Company and the University of Utah have been engaged in cooperative rock mechanics research for many years with the goal of improving safety, stability and efficiency. Studies include pillar analyses in vertical crater retreat mining, comparisons of conventional and "bird-cage" cable bolting and shaft pillar mechanics. Laboratory testing for rock properties, determinations of stress in situ, finite element simulations of mining sequences and numerous borehole extensometer measurements of rock mass displacements were done in each study.
As in previous shaft pillar analyses (Pariseau and others, 1995, 1996), the main rock types considered are the Poorman, Homestake and Ellison formations. Each is considered anisotropic (orthotropic) and has nine elastic moduli (three Yomag's moduli, three shear moduli, three Poisson's ratios) and nine strengths (three maconfined compressive, three tensile and three shear strengths). Principal material directions are down foliation dip, parallel to strike and perpendicular to foliation. These properties were originally determined in an extensive suite of laboratory tests. Subsequent comparisons between finite element model displacements with mine extensometer measurements led to scale factors of 0.25 and 0.50 for elastic moduli and strengths, respectively. The present study uses elastic moduli and strengths that are also 0.25 and 0.50.
The present study differs from past analyses (c.a. 1990) in two ways. First, the computer code UTAH3 was modified to produce a more accurate anisotropic factor of safety and to execute faster. UTAH3 is an elastic-plastic finite element program that uses associated flow rules; Drucker-Prager and anisotropic quadratic yield criteria are available. Hydro-mechanical coupling and viscoplastic versions exist, but the mine is dry and no time-dependency was considered in this study. Equation solving is by line iteration using a tangent stiffness that is updated after each load increment. Second, the mining sequence was modified to more accurately reflect actual pillar mining. This latter modification was done following discussions with Homestake personnel in May, 2000. Earlier forecasts of shaft pillar stability were based, in part, on assumed pillar mining plans.