Simplex reflection analysis of laboratory strength data

Abstract

Linear regression and simplex reflection techniques for fitting laboratory strength data for rocks and soils to linear and nonlinear failure envelopes are presented. The simplex reflection technique is applied to the Hoek-Brown nonlinear failure criterion which is widely used for estimating the strength of jointed rock masses. A comparison between the simplex reflection technique and linear regression analysis, for assessing the parameters associated with the Hoek-Brown criterion, is discussed. The strengths and limitations of the two techniques are compared, and it is established that the simplex technique is superior for fitting the data to any nonlinear criterion. The implementation of the simplex reflection method is presented together with an algorithm in the form of a pseudocode for use on computers. Key words : linear regression, simplex reflection, failure criterion, stress, residuals.