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 The information provided may be out of date.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.

Undefined value found for fluid transport with heat transfer

Please login with a confirmed email address before reporting spam

Hi,

I am solving a coupled problem with a energy equation using heat transfer in solids, a fluid mass transport equation with the transport in diluted species, and a vapor mass transport equation using the user defined PDE. The other equation is a Darcy's law to find the total pressure.

I am getting the following error:

Repeated error test failures. May have reached a singularity.
Time : 9.453450314751982e-005
Undefined value found.
Undefined value found in the stiffness matrix.
There are 6 equations giving NaN/Inf in the matrix rows for the variable mod1.c v
at coordinates: (0.0173581,0.00710536), (0.0175,0.00701763), (0.017429,0.00706149), (0.0175,0.00710536), (0.017429,0.00714922), ...
and similarly for the degrees of freedom, NaN/Inf in the matrix columns
Last time step is not converged.

I have checked the equations, BCs many times and still have no idea what I could do about it. Could you please give me any suggestions?


Thanks a million~~



5 Replies Last Post Sep 6, 2012, 10:50 a.m. EDT
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 30, 2011, 1:59 a.m. EDT
Hi

have you check that you initial conditions are OK ? hit the "Study ... Dependent variables", right-click "Compute to Selected" anc check the coherence of all initial values

What puzzles me is that your mod1_P_g is flagged Solved for this field "UNCHECKED", and it's greyed, so COMSOL is not accepting to solve for all physics, this could lead to NaNs, but why ?

Have you tried to use a segregated solver, rather than the direct one. Your model is too complex for me to "take it in 5 min" my ususal time limit here on the Forum ;) (My boss have given me many studies for today too )

This was as seen from v4.2 (I have uninstalled the older ones)

--
Good luck
Ivar
Hi have you check that you initial conditions are OK ? hit the "Study ... Dependent variables", right-click "Compute to Selected" anc check the coherence of all initial values What puzzles me is that your mod1_P_g is flagged Solved for this field "UNCHECKED", and it's greyed, so COMSOL is not accepting to solve for all physics, this could lead to NaNs, but why ? Have you tried to use a segregated solver, rather than the direct one. Your model is too complex for me to "take it in 5 min" my ususal time limit here on the Forum ;) (My boss have given me many studies for today too ) This was as seen from v4.2 (I have uninstalled the older ones) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 30, 2011, 10:57 a.m. EDT
Hi Ivar,

Thank you so much for your time and suggestions.

I took a look at the initial values and the mod1.P_g has zeros everywhere initially, which does not make sense. I have specified it to be the atmospheric pressure. Do you know why and how I could modify this?

"What puzzles me is that your mod1_P_g is flagged Solved for this field "UNCHECKED", and it's greyed, so COMSOL is not accepting to solve for all physics, this could lead to NaNs, but why ?"
--I don't quite get this. Could you please explain a little bit? Thank you so much!

I will check out segregated solvers.

Thanks a lot again. Really appreciate it.


Haolin
Hi Ivar, Thank you so much for your time and suggestions. I took a look at the initial values and the mod1.P_g has zeros everywhere initially, which does not make sense. I have specified it to be the atmospheric pressure. Do you know why and how I could modify this? "What puzzles me is that your mod1_P_g is flagged Solved for this field "UNCHECKED", and it's greyed, so COMSOL is not accepting to solve for all physics, this could lead to NaNs, but why ?" --I don't quite get this. Could you please explain a little bit? Thank you so much! I will check out segregated solvers. Thanks a lot again. Really appreciate it. Haolin

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 30, 2011, 2:38 p.m. EDT
Hi Ivar,

And when I used the segregated solver, another error was given:

Recoverable error in residual, but solver could not recover.
Time : 0.0005509748135944119
Last time step is not converged.

I searched this forum and online but had no luck what this means.....


Your help is greatly appreciated.


Best,
Haolin
Hi Ivar, And when I used the segregated solver, another error was given: Recoverable error in residual, but solver could not recover. Time : 0.0005509748135944119 Last time step is not converged. I searched this forum and online but had no luck what this means..... Your help is greatly appreciated. Best, Haolin

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 31, 2011, 1:14 a.m. EDT
Hi

for me it remains linked to an "ill conditioned matrix" = BCs that are not sufficiently independent or missing or initial values too far off from expected results.

Perhaps you can start with the model somewhat simplified, run a first time, use the results as initial conditions, or as a linear interpolation of the initial conditions. With pressure ayou have also the relative "gauge" pressure and the absolute pressure" thta might play you a game. Finally many of the material properties are T and p dependent, which makes the simulations non-linear, even if they are often just as "correct" if you used constant values for the material properties

--
Good luck
Ivar
Hi for me it remains linked to an "ill conditioned matrix" = BCs that are not sufficiently independent or missing or initial values too far off from expected results. Perhaps you can start with the model somewhat simplified, run a first time, use the results as initial conditions, or as a linear interpolation of the initial conditions. With pressure ayou have also the relative "gauge" pressure and the absolute pressure" thta might play you a game. Finally many of the material properties are T and p dependent, which makes the simulations non-linear, even if they are often just as "correct" if you used constant values for the material properties -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Sep 6, 2012, 10:50 a.m. EDT
Hi,

Was this issue solved? I have downloaded the model you put here to try and learn from it. If you have one, can you contact me and send me the model? I'm having trouble with coupling physics in *my* model.

My E-mail is Shovalwarman@gmail.com

thank you very much,
Shoval
Hi, Was this issue solved? I have downloaded the model you put here to try and learn from it. If you have one, can you contact me and send me the model? I'm having trouble with coupling physics in *my* model. My E-mail is Shovalwarman@gmail.com thank you very much, Shoval

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.