Large-scale characterisation of cemented rock fill performance for exposure stability analysis

Abstract

A novel large-scale geomechanical laboratory testing program is developed to provide full characterization of the strength and deformational response of cemented rockfill (CRF) material for use in exposure stability analyses. Uniaxial compressive strength (UCS), direct shear, and three-point bending tensile tests are performed on large-scale samples that require the use of novel processes and laboratory equipment. For the first time, the shear properties of large-scale CRF samples have been investigated using direct shear tests under constant normal stiffness (CNS) boundary conditions. Large-scale three-point bending tests are also conducted to obtain true tensile strengths without having to infer them from UCS test results. For up to 28 days curing time, the effects of particle size distribution and binder quantity are experimentally examined for each compression, tension, and shear failure mode through a study of the full stress–strain response. In order to obtain accurate laboratory results, findings are compared against published large-scale tests from various mine sites. The results show that the large-scale samples completed herein provide a reliable UCS estimate with standard deviations of less than one. It is also found that the direct shear responses are substantially larger for simulated CRF: Sidewall contacts than CRF: CRF contacts, and tensile strength responses are higher than previously estimated at 10% of the UCS strength. The effect of particle size distribution on the geomechanical response of the CRF material is highlighted through the large-scale sample testing. The findings of this research study provide increased technical understanding for the development of CRF structures in underground mines that are cost-effective, safe, and durable. Additionally, this research study establishes a practical testing methodology that overcomes the challenges that occur in collecting quantitative geomechanical data from large-scale CRF samples for design purposes.