Page 423 - Rock Mechanics For Underground Mining

YIELDING PILLARS

crown pillar. Although such events had occurred previously in the mine, it was recog-
nised that bursts involving the Beta fault, as this event did, had important implications
for extraction of the taller and wider stopes to the north which were transgressed by
the same fault. The analysis reported by Lee et al. was intended to provide a proven
model of rock mass performance in the block, from which the consequences for future
stoping could be evaluated.
The analysis of rock performance in the block was conducted in two stages. Prelim-
inary plane strain studies were conducted with the boundary element code BITEMJ
(Crotty, 1983), which can simulate slip and separation on faults. These analyses
indicated sense and magnitude of slip on the Beta fault consistent with in situ mea-
surements, as indicated in Figure 13.31a. However, because the plane analysis was
incapable of examining the development of slip along the strike of the Beta fault
as stoping progressed, a more comprehensive three-dimensional analysis of stress
and displacement was conducted. The problem geometry, as defined for input to
the linked boundary element–finite element scheme BEFE (Beer, 1984), is shown
in Figure 13.31b. The results of the analyses are illustrated in Figure 13.31c, for
the state of fault slip before and after mining the eastern extension of the G2 stope.
These are contour plots of the magnitudes of dip shear displacement on the Beta
fault, mapped on the plane of the fault. It is observed that a substantial increase
in the zone and magnitude of slip is indicated over the part of the fault transgress-
ing the crown pillar, attending the mining of the eastern extension of the G2 stope.
Such correspondence of observed and calculated rock mass response provided a ba-
sis for prediction of rock performance during mining of stopes further north in the
block.
Several principles are illustrated by the Mount Charlotte study. First, effective
stope-and-pillar design in irregular orebodies or complex structural settings usually
requires computational methods for design analysis. Second, two-dimensional anal-
ysis has an important rˆole in assessing and characterising rock mass performance,
before more complicated and expensive three-dimensional methods are employed.
Third, field observations combined with appropriate analysis provide a basis for con-
fident prediction of host rock performance as stoping proceeds through a selected
extraction sequence. If analysis suggests a particular stope and layout or stoping se-
quence may induce intolerable rock mass response, a verified mine model similar
to that developed for the Charlotte orebody permits easy assessment of alternative
layouts and sequences.

13.9 Yielding pillars

The discussion in section 13.4 indicated that when the magnitudes of the pre-mining
stresses increased relative to the in situ strength of the orebody rock, an excessive
proportionofanorereserveiscommittedtopillarsupport.Thesolutiontothisproblem
varies with the type of deposit. For a metalliferous orebody where reserves are limited
and the post-peak behaviour of pillars is uncertain, ore in pillars which were initially
designed to perform in an intact, elastic mode may be recovered by the extensive use
of backfill. This procedure is described more fully in Chapter 14. Where reserves
are more extensive, such as coal seams or other stratiform deposits, pillars may be
designed to operate in a plastic mode, i.e. at a factor of safety less than unity.
405