Dear all
I have written a simple VUMAT for elastoplastic analysis of simple 3D element similar to Johnson-Cook plasticity model without strain-rate dependency or temperature involvement; Sigma = A + B*eps**n. I run the code for a single element and the results are in good agreement with results of the defined Johnson-Cook model of the Abaqus. But when I increase the numbers of elements; for instance 3 elements, as the analysis reaches the plastic state and after some increments, one of the parameters becomes NaN which is described as an unidentified parameter and the analysis stops with an error related to excessive distortion of the elements.
Has anyone faced this kind of problem? I would appreciate it if somebody helps me through this. Regards
NaN frequently occurs when you divide through a very small number or get another kind of overflow error.
Just write out all variables to a file and try to see where things go wrong.
Dear Dr. Baeker
Thank you for your helpful response. I did write down the variables in my code and found one of the variables (deqps as plastic strain increment) changes to a negative value in one of the elements. Since it does not make sense to have a negative value for plastic strain increment, I modified the definition of the deqps. As so, the problem solved and the results are in good agreement with the results of Johnson-Cook plasticity
Still, I am a bit confused and surprised why there was no such an error when applying the same code to one element?!
Anyway, I do appreciate your answer. Thanks.
"why there was no such an error when applying the same code to one elemen"
Probably due to a difference in loading or in the time increments?
Dear prof. Baeker
Actually, the problem was related to my misunderstanding of FORTRAN. I had defined a variable named "lodepara" which meant to be the Lode angle parameter of the stress state, but since the FORTRAN takes all the variables beginning with i-n as an integer value, the "lodepara" became zero in the loops which caused the mentioned issue.
- Badges
- Science topic