Investigating the effect of the statistical distribution function of the GSI on the probabilistic analysis of rock mass geomechanical parameters

Hoek and Brown suggested a method to estimate the strength and deformation modulus parameters of rock masses. The method was then widely used in rock engineering designs. In such designs, the mean values of Hoek and Brown parameters are often used which are not proper values due to the variability of rock mass properties within a great range of values. In such cases, probability analysis of rock mass properties is highly important. The geological strength index is one of the most important parameters in Hoek and Brown equations. Determination of this parameter includes greater uncertainties than determining other parameters. In this paper, based on the results of rock mechanical tests carried out on rock samples of Gol-Gohar iron ore mine, and the required field surveys, the sensitivity of rock mass geomechanical properties on the type of the statistical distribution function of the geological strength index in statistical analysis of these parameters using Monte Carlo simulation method was investigated. The results showed that the sensitivity of Hoek and Brown equations to determine different rock mass geomechanical parameters varies as the type of the statistical distribution function of the geological strength index changes. The sensitivity of geomechanical parameters such as internal friction angle, cohesion, total strength and rock mass modulus on the type of the statistical distribution function of the geological strength index is much less than parameters such as uniaxial compressive strength and tension strength of rock mass. The greatest variations based on changes of the type of the statistical distribution function of the geological strength index are less than 5% for the internal friction angle, cohesion and total strength, less than 10% for the modulus, and less than 25% for the uniaxial compressive strength and tension strength.