Abstract
The application of macroscopic damage models is often made difficult by the fact that the calibration of the model parameters is associated with great experimental effort. Therefore, a methodological, simulative approach is presented to determine the damage model parameters with significantly less material input for the actual, local material state. Virtual experiments were performed on three-dimensional, statistically Representative Volume Elements (3D-RVE) of the microstructure to incorporate microstructure-based influences. For the generation of 3D-RVE, the in-house software DRAGen was used, which generated a geometric model based on input data gathered from EBSD analyses on the ferritic-bainitic reactor pressure vessel steel 22NiMoCr3-7. Mechanical properties were assigned to the constituents of the microstructure model by utilizing a phenomenological crystal plasticity model, which was inversely calibrated on the experimental flow curve. In addition, damage criteria on the microscale following the macroscopic material behavior are needed for cleavage and ductile fracture mechanism. Finally, the macroscopic damage model can be calibrated locally by micromechanical simulations. The virtual experiments were conducted with prescribed deformation boundary conditions, and different stress states were simulated by varying them.
