Comparing Shear Strength Dispersion Characteristics of the Triaxial and Direct Shear Methods for Undisturbed Dense Sand

Obtaining representative samples for laboratory shear strength testing forms the foundation of a successful geotechnical investigation program. The selection of suitable laboratory testing methods and the understanding of their limitations in the mining of a mineral sands deposit is discussed. The pit slopes will comprise of cohesionless sand and silty sand. The method of recovering undisturbed samples in a deep drilling program is outlined. The use of undisturbed samples increases confidence and reliability in probabilistic stability analysis. The effective friction angle values from 40 consolidated undrained triaxial and 46 direct shear tests are compared. The direct shear test results are 5% lower than the triaxial test results due to subhorizontal bedding. In addition, the coefficient of variation for the triaxial and direct shear friction angle results are 4.3% and 9% respectively, indicating a larger dispersion in the direct shear test results. The triaxial and direct shear datasets result in distinctly different outcomes for pit slope optimisation studies with substantial financial implications. Incorporating triaxial test data in a 35 degrees pit slope stability assessment results in a 36% higher confidence level of achieving or exceeding a factor of safety of 1.3, relative to using the direct shear test data.