convergence problem
convergence problem
(OP)
Hi, I have a big problem with abaqus cae. I have to verify stresses in a beam under different load (from 1000N to 9000N). Everithing is fine until I reach 7000N and 8000N. In those cases abaqus doesn't find a solution to the problem (if I put 9000N everything's fine again). What is the problem? Maybe a mesh problem?





RE: convergence problem
Rob Stupplebeen
https://sites.google.com/site/robertkstupplebeen/
RE: convergence problem
A) Are there any error messages in the .log/.dat/.msg files? If so, what do they say?
B) If an .sta file was generated, what does the last line say?
C) Open the ODB, look at the displacements. Are there any distortions?
http://www.eng-tips.com/faqs.cfm?fid=376
http://www.eng-tips.com/faqs.cfm?fid=1083
RE: convergence problem
error message: Too many attempts made for this increment
.sta file: THE ANALYSIS HAS NOT BEEN COMPLETED
I tried what "rstupplebeen" suggested with no results. I can not see any displacement because abaqus can not start the step1's analisys.
I wonder why I have problem with 7000N and 8000N and everithing goes fine with 9000N.
RE: convergence problem
Are you using a dynamic implicit? Probably not.
So what rstupplebeen suggested was probably correct.
Try limiting the maximum allowable time increment (the 4th number at *static) as well.
So use something like 1e-3,1,1e-7,1e-2. (<--- this is VERY dependent on your problem, it's just an example).
RE: convergence problem
The funniest thing is that if I put 8200N instead of 8000N (or 6800N instead of 7000N) it works!!!
p.s. Not so funny
RE: convergence problem
http://www.eng-tips.com/faqs.cfm?fid=376
http://www.eng-tips.com/faqs.cfm?fid=1083
RE: convergence problem
Rob Stupplebeen
https://sites.google.com/site/robertkstupplebeen/