ABSTRACT
Conventional manual calibration of steel material constitutive model parameters is a process of trial and error, with the disadvantages of low efficiency and bad accuracy. The present work aims to propose an automatic calibration process by introducing modern optimization algorithm. Firstly, a series of Q345 and Q235 steel material property tests were conducted, to prepare the data needed to conduct the calibration. Secondly, the material property parameters were determined inversely by particle swarm optimization (PSO). It was found that after 20 times of iteration, the calibration accuracy of Q345 steel is enhanced by 18.73% to 40.44%, and corresponding improvement of Q235 steel is 12.61%-76.20%. Thirdly, the cyclic plasticity characteristics of Q345 and Q235 steels were compared, the results show that both isotropic and kinematic hardening levels of Q235 steel are greater than those of the Q345 steel. The Q-value of Q235 steel is increased by about 60.39%, compared with Q345, and the backstress (C1/γ1) of Q235 is 22.14% larger than that of Q345.
