Simulation of crack propagation using a GTN ductile damage model based on the virtual crack closure technique

In this study, tensile tests are performed on notched flat specimens to calibrate the Gurson-Tvergaard-Needleman (GTN) model for a selected VVER440 material and the optimization of damage parameters using artificial neural networks (ANN) is presented. The specified parameters are applied to standard Compact Tension (CT) specimens, whose fracture behavior is used to estimate the fracture toughness (geometry independent) material properties with finite element simulation software. In order to compare fracture toughness simulations conducted by means of the classical fracture mechanics approach based on J-Integral, detailed local approach analyses are performed on CT specimens. In the determination of fracture toughness, a new approach is used, the crack propagation is based on the virtual crack closure technique (VCCT), as this allows the calculation of the J-integral in the software during the crack propagation. The results indicate that the new approach gives similar fracture toughness compared to the conventional method and measured values, but with significantly higher computational efficiency.