Hello everybody.
My task is to write 2d UMAT for nonlinear viscoelasticity. I am trying to write a 1d UMAT for non-linear viscoelasticity based on the Schapery's model (paper attached below) and then extend it. I get "Excessive distortion error". I tried printing variables in the .dat file. Problems are
I will attach my FORTRAN code for reference. You see I am newbie to Abaqus. I would like to have your suggestions and ideas.
EDIT 1
- It looks like I should not use nlgeom for this analysis. Anyway I was able to successfully run the code with a sample 1D truss (single element). But the result is not as expected. When trying to simulate relaxation under constant strain, stress looks to increase and then go constant after a certain time period which is not correct. Any suggestions would help.
- If somebody could just skim the attached paper, I was not able to find a starting value to initialize a parameter (q_ij^time) for the 1st time step. The result changes with different initialization values and the paper does not provide any info on that. Would be really helpful if someone could help me in this.
Thanks a lot for your time
Shreeraman S
Hello Mr. Beizaee
First of all thanks for the answer. I got the point that you are trying to drive in. The problem is,
Thanks and Regards
Shree
Hi Shreeraman,
1. UMAT is called twice (I vaguely remember that once I checked) during each time step. Once to get the stiffness matrix and solve K.(delU) = (del F) and get (del U). Once you get (del U), abaqus calculate (DSTRAN) and that is supplied to user to calculate stress increment and update the stress. So at the first time step, when you print DSTRAN, once you get it zero and 2nd time you should get some value for it, if loads are applied. Please check it.
2. As Shahriyar suggested, looking at the creep strain levels in Fig.8 and 14 seems that strain are very low. So you can first try the small deformation analysis with NLGEOM=OFF only.
Hi Mr.Faye
Thanks for the response. I am able to run the UMAT now with nlgeom=OFF. Just that the results are not correct. As posted above, I still could not find a starting value for the parameter (q_ij^time) in any related papers and unfortunately the results seem to be sensitive on the initial (assumed) value.
Thanks and Regards
Shree
- Badges
- Science topic
- Similar topics
- Engineering
- Mechanical Engineering