Parameter determination of Chaboche kinematic hardening model using a multi objective Genetic Algorithm

• Published in: Computational materials science 2011, Vol.50 (3), p.1114-1122
• Main Authors: Mahmoudi, A.H., Pezeshki-Najafabadi, S.M., Badnava, H.
• Format: Article
• Language: English
• Subjects: Chaboche model; Condensed matter: structure, mechanical and thermal properties; Exact sciences and technology; Fatigue, brittleness, fracture, and cracks; Genetic Algorithm; Hysteresis loop; Kinematic hardening; Mechanical and acoustical properties of condensed matter; Mechanical properties of solids; Physics; Ratcheting; Uniaxial and biaxial loading
• ISSN: 0927-0256; 1879-0801
• DOI: 10.1016/j.commatsci.2010.11.010

► A Chaboche model that accurately predicts both ratcheting and hysteresis loop. ► A Genetic Algorithm improved ratcheting prediction and hysteresis loop model. ► Two fitness functions appeared to provide better results. Chaboche model is a powerful tool to evaluate the cyclic behavior under different loading conditions using kinematic hardening theory. It can also predict the ratcheting phenomenon. To predict the ratcheting, it is required to determine the material parameters under strain control conditions. Although, these parameters can model the hysteresis loop fairly accurately, their ratcheting prediction does not have the same quality. A set of material parameters that could accurately predict both ratcheting and hysteresis loop is of great importance. The available models, generally for low cycle fatigue, are mostly complex and nonlinear. Therefore, an optimization procedure can be used for parameter determination and consequently improving the prediction of these models. Genetic Algorithm is a numerical approach for optimization of nonlinear problems. Using a multi objective Genetic Algorithm for Chaboche model, a set of parameters was obtained which improved both ratcheting prediction and hysteresis loop model. Two fitness functions were used for this approach. The proposed model was verified using Hassan and Corona’s experimental data conducted on CS 1026 low carbon steel. The model indicated a very good agreement in the case of uniaxial loading with the experimental data. The results of proposed model for biaxial loading histories are similar to the model by Hassan and his co-workers.