Discussion Forum

Hi

That one I havnt tried, for me "t" remains "t" in all cases, (but it might be possible still, I'm not COMSOL).
Isn't there other ways to split your job, ie. with the solver manager ?

Good luck
Ivar

Hi Ivar,

Thanks for your reply.

I have a pseudo coupled problem to solve. I want to solve a PDE (in the form of PDE general form with Transient setting). This PDE has nothing to do with time. Its form is just similar to the time-related PDE available in COMSOL. (It has to be solved for time = 10 seconds) [tau = 10]
One of the coefficients of this PDE has to be updated in each time step. In order to update this coefficient, I have to solve a transient structural problem (for 5 seconds). [t=5]
When I want to solve this problem, COMSOL mixes these two time parameters, as both of them are defined by "t". Hence I get a wrong answer.

Any idea to fix this problem?

Thank you,
Masoud

Hi
Sorry, there you are using a global approache I would really like to master better, but I havnt had time to study this approach to COMSOL enough yet, I have to look after my clients shallenges first, not just to have fun ;) ( I havnt mastered to get their need overlapping this apprach)

But there are certanly other out there, if not support

Pls keep us informed, I find it interesting
Good Luck
Ivar

Hi again

I had a quick readthrough of the doc again, for me I understand that "t" remains a special variable, that you cannot change suchas the fem.dim ones like u,v,w ....
But you might get around with a script, having two different models structures fem and fem1, you solve one then the other and transfer manually the data between the two (obviously each one with the appropriate structueres and set up required) But this is probybly mquite tedious.

Ivar

Thanks a lot Ivar. I am trying to script it.

I appreciate your points.

Masoud