Post-Peak Multi-Stage Triaxial Testing of Miniature Rock Mass Samples

Triaxial tests were performed on rock core samples with 14 confinement stages per test to investigate the post-peak strength and nonlinearity of the failure envelope for fractured rock masses. The confinement stages were tested progressively on each sample, starting at 0.69 MPa, incrementally ascending to 27.6 MPa, and then descending back down to 0.69 MPa. Three types of samples were tested: massive (no defects), continuous pre-existing defect, and jointed. The jointed samples were prepared to imitate a small-scale rock mass by confining brecciated sandstone core within heat shrink film and then mechanically breaking healed veins between the clasts with Brazilian disk load platens prior to testing. The measured post-peak strength of the samples was variable and dependent on the direction and quantity of pre-existing discontinuities within the sample. A generic failure criterion was fit to the test results to determine strength parameters for evaluation and comparison. Two empirical rock mass strength estimation methods were fit to the test results to assess if empirical methods are able to model the measured failure envelopes. 1. INTRODUCTION Most engineering evaluations rely on empirical methods to estimate rock mass strength properties. Empirical methods generally utilize input parameters that can be readily measured in the laboratory or core shed such as the intact rock strength, fracture shear strength, or the degree of rock mass fracturing. Very few published studies have been performed on fractured rock samples to support or validate existing empirical estimation criteria (González-Fernández et al. 2023). This study presents a novel technique to prepare fissured, but fully interlocked, natural rock samples with mechanically induced fractures on pre-existing weakness planes. These samples are referred to as "miniature rock mass" samples due to their resemblance to a small-scale rock mass. Two of these samples were tested under multi-stage triaxial compression to measure post-peak strength behavior. In addition, two baseline samples were also tested to bracket the potential strength envelopes of a rock mass: one massive sample (no defects) and one sample with a continuous pre-existing defect. The resultant strength envelopes for each sample were then evaluated and fit using two different empirical rock mass strength estimation methods.