Closed a3490489 closed 1 year ago
Hi, The error is related to incorrect assignment of any of these variables (in file userinputs.f) related to array size assignment - maxnloop, maxnslip, maxnmaterial check your user inputs
Another possible issue could be that, these elements behave stiffer than the others which might lead to high stresses. You may try using smaller time steps initially and a conservative time step increase ("paste_front=1.1" in userinputs.f)
Make sure to not to use the GND models.
I will double check this issue and let you know once I'm back in the office.
Dear Vikramroybarc and Eralp,
Thank you very much for the kind reply. I tried to check the userinputs.f and used a smaller initial time steps (0.0001s) with "paste_front"=1.25, 1.1, and 0. But it still shows a similar error. I checked the GND model is not activated, and the material property of ID 2 (Copper?) in the user materials.f was used. The default values of maxnloop(3), maxnslip(12), maxnmaterial(10) in userinput.f were used. The C3D8 model with all the same settings works well, although C3D4 are not. I still don't know where the problem is. Please tell me if there is any possible reason; thanks a lot.
Sincerely, Ta-Te
Hi, They all seem correct. In the "usernputs.f" the element type and total number of elements in the mesh need to be defined. have you checked that?
Thanks, Eralp
Dear Eralp,
Thank you for your reply. Yes, I changed the element number(3862) and element type(C3D4) following the youtube videos. By the way, I also tried to set the element type as C3D8, although the element in the model is actually C3D4, and it works. I am not sure why.
Sincerely, Ta-Te
Ok, thanks.
Could you convert the element type to C3D10 (quadratic tets) in userinputs.f and use that as the element type. Make sure to check you real element type, they might be set to quadratic during Neper generator.
C3D4 is available within the mesh library. C3D8 has 8 integration points so it can still work since the other 7 is not accessed unless GNDs are turned on.
Let me know if this helps, Eralp
Dear Eralp,
Thank you for your reply. I tried to change the element type to C3D10 in userinputs.f, and the element type in the Abaqus model remains C3D4. It works, and the stress distribution is the same when the element type is set as C3D8 in userinputs.f. I also tried to change the real element type to C3D10 in Abaqus, but the stress distribution became weird. I don't know if it is correct. Please see the attached figs.
Sincerely, Ta-Te
Thanks for your feedback. I will check C3D4 element in more convenient time. Apparently, having one integration point causing an allocatable array issue.
For C3D10, make sure your BCs are defined for the mid nodes on the surfaces as well.
Dear Eralp,
Thank you very much for the kind help. For C3D10, yes, it is the BCs problem.
Sincerely, Ta-Te
Dear Eralp
Thank you very much for sharing the CrystalPlasticity codes; it is beneficial. When I followed your Video Tutorials, everything went well except the C3D4 model. It went well when I simulated the C3D8 models from DREAM3D and NEPER. When running the C3D4 model, I did change the element number and element type in the userinputs.f file. The msg file shows the following error. Would you happen to know where the problem should be? All the comments and suggestions are a tremendous help to me.
Sincerely, Ta-Te