Learning Forum

[Preston Goulson]

Hello. I'm running a nonlinear static structural analysis with pressure and imported thermal loads. It's a tube heat exchanger, where the welds are approximated as bonded edges between a shell mesh for the tube, and a solid mesh for the header. I keep running into a situation where the analysis bisects over and over again, even though each time, the bisection occurs after the force convergence value drops below the criterion. The step eventually converges, but the analysis will fail due to an unconverged solution a little while later. If I go into the output file, the only relevant warning I can find is about large increments of plastic strain. I've upped the max number of substeps to 1000, and even tried adding the command snippet NEQIT, 100. Still get the same bisection behavior. Any idea what I can do to fix this? Thanks in advance.

[Akshay Maniyar]

Hi, 

The issue you're encountering, where the analysis bisects repeatedly even after the force convergence value drops below the criterion, could be due to several factors. One possible reason could be the large increments of plastic strain that you mentioned. Only a 15% change in plastic strain is allowed in a single substep. This could be causing the solution to diverge, leading to repeated bisections. Increasing the maximum number of substeps and using the NEQIT command, as you've done, are good steps to try and mitigate this issue. However, if these steps are not resolving the problem, it might be necessary to review your model and the applied loads. You can check the below video and see if it helps.

Convergence Problems in Metal Plasticity | Ansys Courses

 

Thanks,

Akshay Maniyar