Abstract
When the Linear Parallel Bond Model (PBM) in Particle Flow Code is used to simulate intact rocks, a basic problem is how to determine the microscopic parameters that control the macroscopic properties of the modeled rocks. After simplifying the microscopic parameters of the PBM model, this study proposes a new method of inverse performance of the regression equations of the macroscopic parameters by the gray absolute correlation combined with regression analysis, which solves the drawbacks of the current manual trial parameter calibration method to a certain extent. When this method is applied to the calibration of the microscopic parameters of the siltstone, the numerical simulation results produce good agreement between the results derived from the finite element software (ABAQUS) both in terms of macroscopic parameters and damage forms.