Calibration of multivariable plasticity parameters using classical optimization methods for porous rocks

Optimization is a branch of applied mathematics which is used extensively in almost all areas of decision-making. In particular, classical method is a kind of the optimization methods that can be applied to make decisions. The optimization method can only make a better decision if the algorithm starts with an appropriate initial data. The main purpose of this study is two-folds. First, the material parameters of multivariable plasticity models, such as Mohr-Coulomb and DiMaggio-Sandler are calibrated via the gradient based optimization method by considering the appropriate initial data from porous rock database. Moreover, the calibration of models is done by the methods presented in the Nlopt library with considering the same objective functions and the same initial guess for each parameter. Then, a home-made finite element simulator is used to evaluate the correctness of the proposed material parameters identification framework. The results show the capability of the proposed optimization framework to determine properly the material parameters of multivariable plasticity models.