Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Two solver steps in time depend. study
Posted 2017年8月22日 GMT-4 13:36 Version 5.3 2 Replies
Please login with a confirmed email address before reporting spam
I am running a time dependent study. When I try and run for long times, i.e. greater than 5 s, it does not converge. However, if I run a solver step until ~0.1 s and then feed that result in as the initial condition in a second step in the solver, I can run it to longer times, i.e. 100 s. The solutions seem reasonable.
My question is: Is this an appropriate method to getting a simulation to converge?
Thank you
My question is: Is this an appropriate method to getting a simulation to converge?
Thank you
2 Replies Last Post 2017年8月22日 GMT-4 14:09
Please login with a confirmed email address before reporting spam
Posted:
7 years ago
Hello Lawrence,
Please see my colleague's contribution to the following thread for an explanation why your model behaves differently depending on the end time.
www.comsol.com/community/forums/general/thread/108101
As that post suggests, you should be able to get convergence in a single solver step by adjusting the settings as he describes.
Best,
Jeff
Please see my colleague's contribution to the following thread for an explanation why your model behaves differently depending on the end time.
www.comsol.com/community/forums/general/thread/108101
As that post suggests, you should be able to get convergence in a single solver step by adjusting the settings as he describes.
Best,
Jeff
Please login with a confirmed email address before reporting spam
Posted:
7 years ago
Dear Jeff,
Thank you for directing me to this post.
Decreasing the "fraction of initial step" worked perfectly. I can now use a single solver step.
Best,
Lawrence
Thank you for directing me to this post.
Decreasing the "fraction of initial step" worked perfectly. I can now use a single solver step.
Best,
Lawrence