A systematic approach to correlate experimental stress and damage measures for hard rock

Abstract

The development of a constitutive model for hard rock goes hand in hand with conventional experiments, which provide information about the intrinsic failure properties of the material. To date most models are calibrated and validated using the macro stress-strain responses from uniaxial and triaxial tests, with only a few studies including the effects of damage accumulation. Therefore, it is essential to address these mechanisms in a constitutive model, not only in the formulation but also in the experimental data sets used for validation. In this study, a systematic approach for coupling experimental measures of stress and damage is proposed. With acoustic emission energy recorded during testing, a damage variable was derived which can be used for calibration of a constitutive model in the damage-plasticity framework. Then by cross-referencing the damage and stress values for specific time increments, a coupled relationship to link the responses was developed. The evolving failure surface was further studied to determine the relationship between the development of stress and damage throughout triaxial rock tests. After peak stress, the stress-damage relationship follows a linear relationship until residual failure where frictional sliding dominates. This important finding reveals, for a large range of pressures, that stress-damage evolution in triaxial space is variable with confinement. This could lead to plastic strain and stress dependent damage evolution laws for more robust constitutive modeling of hard rock.