Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
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.
Unidirectional coupling of a time dependent study to a stationary study
Posted 16 déc. 2013, 02:27 UTC−5 Studies & Solvers Version 4.3a 4 Replies
Please login with a confirmed email address before reporting spam
I am using structural mechanics module and a stationary study to calculate stresses and strains in a simple 2D axi-symmetrical model. The model has a thermal expansion node.
I have another study step, which is a time-dependent one. This step has a heat transfer module and I am calculating the temperature distribution in the domains for the times from 0 to 24 hours.
I am trying to couple the temperatures calculated in the time-dependent study to the thermal expansion node in my stationary analysis. I can do this for only a certain time, using the 'Values of variables not solved for' option under the 'Values of dependent variables' section in the Study settings.
Instead, I want to be able to use a continuation parameter in the stationary analysis and in each sweep, I would like to get the temperatures at a certain time, say starting from 0 to 1h, 2h... (25 steps).
I hope I could state the problem clearly. Is this possible? I'd appreciate it if anyone could guide me.
Thanks,
--
Tolga
Please login with a confirmed email address before reporting spam
I have a totally unrelated problem, but it also requires a time-dependent solver followed by a stationary solver. I haven't been able to figure out how to automatically apply the stationary solver at all the times from the time-dependent solution.
Have you made progress on this for your problem?
Thank you!
All the best,
Vanessa
Please login with a confirmed email address before reporting spam
Unfortunately, I haven't figured out this problem, yet. For now, I kind of combined the problem to a time dependent solver. But this is a temporary solution, and I don't think it is applicable to all cases.
When I have time I will work on this and post a solution here if I can figure it out. I am thinking maybe using extrusion coupling variables could help accessing the solutions of the time dependent solver.
Kind regards,
Tolga
Please login with a confirmed email address before reporting spam
I'm not sure if it's working totally correctly, but I was able to get time-dependent output from the stationary physics by having three solver configurations:
1) A time dependent solver for the first physics only
2) A stationary solver for the second physics only, where under "dependent variables" it is set to "user defined," and then under "values of variables not solved for" it is set to use all solutions from solver 1.
3) A time dependent solver for the second physics only, again set to use all the solver 1 solutions.
Let me know if this works for you!
All the best,
~Vanessa
Please login with a confirmed email address before reporting spam
I guess I've got a similar problem. Here is a simplified version of it:
(1) u1t + D*u1xx = R; R = f1(u2(x)), D = const
(2) S*u2xx = 0; S = f2(u1(x))
f1 and f2 are (generally) non-linear explicit functions.
u1(x)|t=0 is known.
I need to find u1(x,t), assuming (2) is true for every moment t.
(The relaxation time of the physical process associated with (2) is negligible )
So, (2) should be solved at every time step to obtain u2[t], having u1[t] as an input;
after that, (1) should govern the time step transition u1[t] -> u1[t+1].
But how can it be implemented in COMSOL (I'm trying to use Coef. Form PDE)?
Could you, please, give some hints on the architecture of studies, steps and solvers.
Would be very grateful!