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.

Solver error when mesh increase...

Please login with a confirmed email address before reporting spam

I am using Comsol 4.0a.

I got the tutorial from Comsol for one-way FSI model with joule heating example.

When I follow the tutorial the result is fine and there is no error. But I change the mesh number from fine to finer I got the following error and warning;

Undefined value found.
- Detail: Undefined value found in the equation residual vector.
There are 80 degrees of freedom giving NaN/Inf in the vector for the variable mod1.T
at coordinates: (0,5.5633e-005,9.48117e-005),....

Inverted mesh element near coordinates (0.000115527, 9.18074e-005, 9.88559e-005)

The mesh size is 19918 and minimum element quality 0.4439 with tutorial.

The finer mesh size is 57516 and minimum element quality is 0.3468.

Any comments about this problem is welcome. Thanks in advance

5 Replies Last Post Aug 6, 2010, 4:49 p.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 Aug 5, 2010, 4:24 a.m. EDT
Hi

I would be suspicous particuluarly with respect to the message:

There are 80 degrees of freedom giving NaN/Inf in the vector for the variable mod1.T
at coordinates: (0,5.5633e-005,9.48117e-005),....

for me this looks like there is something missing in the boundary conditions or a T value not defined (could be used in thematerial section if you are using library material depending on T and T is not defined

--
Good luck
Ivar
Hi I would be suspicous particuluarly with respect to the message: There are 80 degrees of freedom giving NaN/Inf in the vector for the variable mod1.T at coordinates: (0,5.5633e-005,9.48117e-005),.... for me this looks like there is something missing in the boundary conditions or a T value not defined (could be used in thematerial section if you are using library material depending on T and T is not defined -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 5, 2010, 8:22 a.m. EDT
I changed the solver setting from segregated to Fully coupled. And there is no more error message.

But I still got the following warning message:

Inverted mesh element near coordinates (0.000116793, 9.90577e-005, 9.88431e-005);

Do I need to concern this message?
I changed the solver setting from segregated to Fully coupled. And there is no more error message. But I still got the following warning message: Inverted mesh element near coordinates (0.000116793, 9.90577e-005, 9.88431e-005); Do I need to concern this message?

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 Aug 5, 2010, 10:52 a.m. EDT
Hi

inverted mesh is a warning that locally the mesh quality is bad, if there is only one/a few and the coordinates provided are not in the middle of the most important feature of your model it's not that important.

nevertheless a "good" model should not have any inverted mesh elements, soits also possible to locate the position an try to understand what is happening, perhaps there are some unecesary details to be removed (defeaturing) or it could be supressed by a local edge pre-meshing with a higher density

As often no clear yes/no

--
Good luck
Ivar
Hi inverted mesh is a warning that locally the mesh quality is bad, if there is only one/a few and the coordinates provided are not in the middle of the most important feature of your model it's not that important. nevertheless a "good" model should not have any inverted mesh elements, soits also possible to locate the position an try to understand what is happening, perhaps there are some unecesary details to be removed (defeaturing) or it could be supressed by a local edge pre-meshing with a higher density As often no clear yes/no -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 6, 2010, 12:22 p.m. EDT
Are you solving coupled problem?
Are you solving coupled problem?

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Aug 6, 2010, 4:49 p.m. EDT
Hello,

The intention of this model is to only demonstrate the use of the Fluid-Structure iteraction feature along with heat transfer physics in COMSOL 4.0a. One can see that the Young's modulus of the flexible obstacle is small and provides a large deformation when influenced by the fluid. This material is not real.
Real FSI situations need to be treated with care but this tutorial will show how to approach such simulations.

best regards,

Mauricio
Hello, The intention of this model is to only demonstrate the use of the Fluid-Structure iteraction feature along with heat transfer physics in COMSOL 4.0a. One can see that the Young's modulus of the flexible obstacle is small and provides a large deformation when influenced by the fluid. This material is not real. Real FSI situations need to be treated with care but this tutorial will show how to approach such simulations. best regards, Mauricio

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.