Crack interaction modelling

The growth characteristics of short fatigue cracks under axial loading were investigated using specimens of the ferritic-martensitic steel F82H-mod. I nterest focused on crack propagation due to coalescence, which proved to be the dominant mechanism of crack growth. Crack propagation due to coalescen ce under a certain loading state is strongly influenced by the microstructu re of the material on the one hand and by the interaction of cracks on the other. This study deals with an elasto-plastic fracture mechanics analysis of two interacting cracks neglecting the microstructural influence. Finite element calculations based on a Ramberg-Osgood model for the material prope rties were performed to quantify the interaction of two cracks in terms of an interaction function Y depending on the material and crack configuration . Finally, a neural network was trained to determine the interaction functi on for two cracks within the range of interest.