Determination of Ductile Damage Parameters Using Hybrid Particle Swarm Optimization

Damage models are widely used to describe the ductile damage and fracture of metals. This paper proposes a new approach to determine the damage parameters of the Gurson-Tveergard-Needleman model, which uses a load-displacement curve coupled with finite element method. The load-displacement curve was obtained from a tensile test of a smooth tensile specimen and contained information about the damage and fracture behavior of the tested material. The principle of damage parameters identification is to minimize the deviation between experimental and simulated load-displacement curves by a hybrid particle swarm optimization. As a combination of particle swarm optimization and simulated annealing, the hybrid particle swarm optimization is an economical and effective algorithm to identify damage parameters. The identified damage parameters are also verified by testing and simulating deformation shapes of the smooth tensile specimen which is used for parameters determination. Tests and simulations of notched tensile specimens were carried out to discuss the transformability of the identified damage parameters. It is observed that the value of critical void volume fraction of the Gurson-Tveergard-Needleman model decreases with the increase of stress triaxiality.