| In the background of the automotive industry’s development towards safety,environmental protection and energy conservation,TRIP steel has been widely used because of its good combination of strength and ductility and its excellent performance in collision energy absorptions.However,the special microstructure of TRIP steel and the martensitic phase transformation occurred during the deformation process brings difficulties in building a constitutive model.At present,the finite element numerical simulation technology is extensively used to predict the material behavior in stamping forming.But the lack of a constitutive model which can meet the requirements for engineering applicatoin,has become a bottleneck that limits the further development and application of TRIP steel.In this paper,taking TRIP780 as the research object,a constitutive model for TRIP steel based on the homogenization method was established by means of the theoretical analysis,numerical simulation and experimental research.Also,the corresponding UMAT subroutine was developed and a method for identifying the constitutive model parameters of each TRIP steel constituent phase was proposed and realized by using the technology of FEM simulation and optimization.The main research content is as follows:Firstly,according to the characteristics of TRIP steel’s microstructure and its phase trasformation rule,a constitutive model of TRIP steel based on the homogenization method of mesoscopic mechanics was proposed.The residual austenite content of TRIP steel under different deformation was measured by using XRD and then a kinetics model of martensite phase transformation was established based on the measurement result.A representative volume element(RVE)of TRIP steel was constructed based on TRIP steel microstructure shown by SEM.On this basis,a two-layer homogeneous constitutive model of TRIP steel was proposed.The austenite phase and martensite phase involved in the phase transformation was homogenizes by using Eshelby-Mori-Tanaka method in this model,while the other elastoplastic phases were homogenized by Gladman stress mixing method.Based on the two-layer homogenization algorithm of TRIP steel,a UMAT subroutine was developed.The implmentaion of the two-layer homogenization algorithm in UMAT and the processing method of material parameters and state variables in UMAT subroutine are described.A double iteration process with the Eshelby-Mori-Tanaka method was applied in this UMAT to partition the strains between martensite and austenite.A method using numerical simulation and optimization techniques to identify the hardening model parameters of each consitutive phase of TRIP steel was proposed and implemented.In this method,the hardening model parameters of each constituent phase were taken as optimization variables,and the minimum error between calculation results of finite element numerical simulation and experimental measurement was set as optimization goal.By combining the optimization algorithm on the Isight platform to find the optimal solution of the hardening model parameters of each consitutive phase,the corresponding parameters can be calibrated.Uniaxial-tension tests and cyclic shear tests were conducted to verify the above constitutive model and the parameter identification method.Based on the experimental results of uniaxial-tension tests,the effect of martensitic phase transformation on TRIP steel reinforcement was discussed.The comparision with the simulation results of three fixed volume fraction of martensite phase showed that the constitutive model can effectively reflect the influence of martensitic phase transformation on the macroscopic stress-strain relationship of TRIP steel.In order to reflect the Bauschinger effect,transient effect and permanent softening behavior of materials under reverse loading,the Chaboche model was introduced on the basis of isotropic hardening model and its constitutive model parameters are calibrated in this paper.Finally,the numerical simulation results showed that the mixed hardening model can reproduce the material behavior under reverse loading effectively. |
