Cemented paste backfill is an increasingly popular technique to improve ground stability in underground mines. This technique is used in several mining methods that require strength evaluation for the vertically exposed cemented backfill following the excavation of an adjacent stope on one side. The critical strength is generally evaluated with an analytical solution proposed by Mitchell et al. (1982). Despite its wide acceptance in academia and application in the mining industry, the Mitchell solution has received only a few updates in the literature, including some new developments given by the authors and colleagues. The original Mitchell solution and its variants were mainly validated against the physical model test results obtained by Mitchell et al. (1982). Even though the Mitchell model debatably assumed a zero backfill friction angle, the required strengths predicted by the Mitchell solution corresponded quite well to those obtained by physical model tests. This study reanalysed the Mitchell solution and its physical model. The testing conditions and procedures for measuring the shear strength parameters are investigated. The stability of the cemented backfill upon removal of a confining wall is analysed with FLAC3D. The comparisons between the numerical modellings, experimental results and analytical solutions are presented, and the applicable range of the classical Mitchell solution is discussed. A new analytical solution is proposed to evaluate the minimum required strength of the cemented backfill confined by two sidewalls exposed on one side and subject to pressure from uncemented backfill on the opposite side. The proposed analytical solution is validated by numerical simulations with FLAC3D. The proposed analytical solution is used to determine the theoretical strength requirement of cemented backfill in primary stopes of an iron mine that employs stage stoping with subsequent backfill mining. The floating Factor of Safety (FS) characterising the current backfilling quality control level of this mine was statistically evaluated with a large amount of uniaxial compressive strength (UCS) data after testing vast drilled samples from field stopes. The engineered strength requirement of the cemented backfill in primary stopes had been finalised by combining the analytical results and floating FS of the mine.
The long-term strength of cemented backfill mass with ordinary Portland cement binder generally decreases with sulphide content due to the formation of expansive phases such as gypsum. This paper investigates the potential of using commercial reactive MgO-activated ground granulated blast furnace slag (MgO-GGBS) in cemented backfill from high sulphide content lead-zinc mine tailings to prevent long-term strength loss. The study focuses on the effect of MgO-GGBS content and the reactive MgO dosage on the unconfined compressive strength (UCS) and the shrinkage/expansion rate. The test results showed that the 28-day UCS of cemented backfill achieved the target strength (?1.0 MPa) with 14 wt% MgO-GGBS content, and the reactive MgO dosage affected the long-term UCS and the shrinkage/expansion rate of cemented backfill body. The main hydration products when using MgO-GGBS were hydrated calcium/magnesium silicate (C-S-H/M-S-H) and hydrotalcite-like phases (Ht). Cemented backfill has a porous opening microstructure. Micro-expansion produced by appropriate MgO content can increase microstructure density, which increases short- and longterm UCS of cemented backfill body, while sustained expansion produced by excessive MgO could destroy the MgO-GGBS microstructure, decreasing the UCS of cemented backfill. We conclude that the mechanical and extension properties of cemented backfill body are highly dependent on the reactive MgO content of the MgO-GGBS. The optimum value of responsive MgO content of MgO-GGBS was 2.57.5 wt% to achieve the long-term stability of cemented backfill.