Identification Problem of Internal Variables Model of Material

The paper deals with the identification of material model based on the internal variable. The model with one internal variable, which was average dislocation density, was considered. Identification was performed using inverse analysis (IA) of uniaxial compression tests. In this work IA was transformed to an optimization task and the goal function was defined as difference (in Euclid's norm) between measured and calculated parameters: loads in plastometric tests (used to identify flow stress) and stresses in stress relaxation tests (used to identify recrystallization kinetics). Exploring a possibility of making the identification more reliable by application the Sensitivity Analysis (SA) was the main objective of the work. The IA was preceded by SA of the model output with respect to the model parameters to select an efficient optimization algorithm and/or eliminate local minima. Selected results of identification for different materials are presented in the paper, as well.