Error Associated with the use of Triaxial Rock Strength Criteria when Assessing the Stability of Underground Openings

"In this paper we investigate the consequences of using conventional triaxial peak strength criteria in the polyaxial stress conditions that occur around the end of a long circular underground opening. A series of 3d boundary element analyses were performed using Examine3D to determine the stresses induced around the tunnel in different in-situ stress conditions. To discuss the applicability of a triaxial strength criterion in failure prediction, strength factors were computed for both the Ottosen polyaxial strength criterion and the triaxial Hoek-Brown criterion. Assuming that the Ottosen criterion is an accurate indicator of strength factor, the discrepancy from this value due to the Hoek-Brown criterion is the error in prediction.In order to simultaneously show the error in prediction and actual strength factor, a quantitative bivariate colour scheme is used to visualize the results. This allows rapid identification of regions with large error, and where either unsafe or overly conservative conditions occur. We show how the inappropriate use of triaxial criteria rather than polyaxial criteria leads to an under- or over-estimation of strength factor, and can result in uneconomic or unsafe design. Extending the analysis to different in situ stress regimes shows how the in situ stress conditions affect the assessment."