I'm modelling the debonding between two interfaces in a 3D model using a cohesive zone. When I was using Implicit, the CSMAXSCRT and CPRESS showed nice, regular behaviour. With this, I mean that the contours are more or less straight over the width of the interface. When I switched to Explicit however, I found that CSMAXSCRT had several regions behind the crack tip that did not reach 1. This implies that the cohesive law has not reached its initiation value yet. Similarly, CPRESS also has regions with unusually high values (higher than they should be, given the crack opening at that location and the cohesive law that I used).
I have tried all sorts of stabilisation approaches (mass scaling, viscosity in the cohesive law, bulk viscosity, using a time increment lower than the stable time increment predicted by Abaqus, ...), but none of these resolve these issues. Do you have any other suggestions what could be causing this strange behaviour
Kind regards,
Comp
I have tried all sorts of stabilisation approaches (mass scaling, viscosity in the cohesive law, bulk viscosity, using a time increment lower than the stable time increment predicted by Abaqus, ...), but none of these resolve these issues. Do you have any other suggestions what could be causing this strange behaviour
Kind regards,
Comp