Menu

Discussion Forum

New Discussion
Filter

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 The information provided may be out of date.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.

Simple convection model not converging

Posted May 15, 2012, 11:12 a.m. EDT Low-Frequency Electromagnetics, Fluid & Heat, Modeling Tools & Definitions, Parameters, Variables, & Functions, Studies & Solvers Version 4.2a 2 Replies

Ricardo Aguilar
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hello all,

I have a simple beam model in which the beam is made out of silicon and it is sorrounded by air. When I apply a voltage to the beam, it creates temperature gradient along the beam due to the resistivity of the beam being dependent on the temperature at which it is at. To incorporate the convection due to the air, I went ahead and used two physics to mimic this phenomenon: electric currents (to compute the resistance of the beam and the power) and non-isothermal flow (to compute the temperature, including the effect of convection). The model itself is a one quarter model, allowing for symmetry, which I took care of with the simple boundary condition of electric and thermal insulation. The reason behind coupling these is so that I can see the effect of convection, given the fact that the heat transfer coefficient of air will change due to temperature. When I run my model with time dependent I get the following error:

"Failed to find consistent initial values.
Last time step is not converged."

When I run it with stationary I get the following error:

"Failed to find a solution.
Maximum number of segregated iterations reached.
Returned solution is not converged."

Any help is greatly appreciated.

Thanks.

Attachments:

2 Replies Last Post May 16, 2012, 11:12 a.m. EDT
John H.
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 16, 2012, 3:49 a.m. EDT
I'm far from being an expert on heat transfer — in fact, I only looked at your model because I would like to learn more about it at some point. That being said, I noticed that when you change the setting for "Absolute pressure" in the "Fluid 1" node from "User defined" to "Pressure nitf/fluid1", the study runs through. This makes sense to me after I read the doc on that subject.

What doesn't make sense to me is your "Temperature 2" node, where you set the temperature to "T". Given that T is the dependent variable in your problem, this definition seems redundant, if not circular, to me. Also, I was expecting to see some kind of volume force somewhere in the model. How else can there be convection?
I'm far from being an expert on heat transfer — in fact, I only looked at your model because I would like to learn more about it at some point. That being said, I noticed that when you change the setting for "Absolute pressure" in the "Fluid 1" node from "User defined" to "Pressure nitf/fluid1", the study runs through. This makes sense to me after I read the doc on that subject. What doesn't make sense to me is your "Temperature 2" node, where you set the temperature to "T". Given that T is the dependent variable in your problem, this definition seems redundant, if not circular, to me. Also, I was expecting to see some kind of volume force somewhere in the model. How else can there be convection?
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 16, 2012, 11:12 a.m. EDT
Hi

indeed when you mix several physics, you must check that all dependent variables have enough BCs to give an unique solution. You have not defined any outles, so there is no BC fixing for "p". Add at least one point in a corner of p=0 for a relative pressure gauge fixing.

Go through all dependent variables first

Next I'm astonishhed you have not added some body force on your "fluid" such as a gravity load

--
Good luck
Ivar
Hi indeed when you mix several physics, you must check that all dependent variables have enough BCs to give an unique solution. You have not defined any outles, so there is no BC fixing for "p". Add at least one point in a corner of p=0 for a relative pressure gauge fixing. Go through all dependent variables first Next I'm astonishhed you have not added some body force on your "fluid" such as a gravity load -- Good luck Ivar

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.

Suggested Content