Dear all,
Currently I am developing codes for the implicit gradient-enhanced damaged plasticity model of concrete in the framework of isogeometric analysis approach.
HOWEVER, when I'm testing my code in the case of monotonic loading, I found there exists unexpected local snapback on the level of Gaussian points within elements, which means that, for a specific Gaussian point k which has already reached plasticity at iteration step n, while at the next iteration step n+1, the yield function value that I calculated from the trial state of this Gaussian point might be F<0. This local snap back leads the rest computations to wrong results.
Could anybody give me some suggestions to solve this unadmissible problem? Thank you!
Best regards,
Cheng Ma