Estimation And Optimization Of Parameters In Chaboche Model Based On Genetic Algorithm

Mechanical materials often work in the condition of alternating stress, especially the materials work in high temperature, not only withstand mechanical loads, but also bear the thermal load, its strength calculation work in mechanical design plays a very important position. So we should know how the material works and the change of stress and strain, estimate its surplus of bearing capacity and its remaining life, in order to judge its real work state.Chaboche Unified Visco-plastic Constitutive Model can descript the unelastic deformation in cyclic loading condition, considering the influence of temperature and time. Parameters of Chaboche model include:elastic modulus E, dynamic sclerosis coefficient a1,C1,a2,C2, isotropic hardening coefficient Q,b, viscosity coefficient Z, n and initial yield stress k. When these parameters are known, Chaboche model can be used to estimate the stress-strain response.In this paper, taking316stainless steel for example, propose a way for the estimation of Chaboche model parameters, according to uniaxial cyclic loading test. Firstly, get experimental data from high temperature cyclic loading experiment of selected material. Secondly, take experiment data as reference, comparing to the result from Chaboche model. The differences between them are used as objective function. The Genetic Algorithms is used to search for the Chaboche parameters to minimize the objective function. Then Least-square method is used for further optimization of Chaboche parameters. Finally, based on genetic algorithm and the Least-square algorithm of parameters estimation and optimization process, taking VB as the platform, develop Chaboche model parameters identification system. This system can estimate parameters of Chaboche model, according to cyclic loading experiment data of certain material, which are used for the prediction of stress-strain response.