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 The information provided may be out of date.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.
Mutlitphysics node needed? For heat transfer through solid and through fluid in one model
Posted 21.01.2014, 07:22 GMT-5 General, Heat Transfer & Phase Change, Computational Fluid Dynamics (CFD), Structural Mechanics Version 4.4 7 Replies
Please login with a confirmed email address before reporting spam
I simulate heat transfer through a water volume (cylinder shape) totally included within a copper cylinder.
I assign the materials "copper" to the outer cylinder and "water" to the inner cylinder.
I assign the physics:
- "heat transfer in solids" to the copper area and
- "heat transfer in fluids" to the water area.
At this point I get a situation which as I interprete is a hint for some error in reasoning on my side:
Both "heat transfer in solids" and "heat transfer in fluids" offer "thermal insulation" for all their individual cylinder surfaces; the inner fluid cylinder for its 6 surface segments and the outer hollow solid copper cylinder for 12 surfacce segments, 6 outer surfaces and also the 6 inner surfaces which are in contact with the water.
I would like to have a continuity condition at this inner surface.
I further assign a continuous heat source to the outer surface of the solid hollow outer copper cylinder.
If I try (after meshing and including a time dependend study) to run the model it calculates twice the temperatures (two data sets, ht and ht2) within the inner cylinder.
I assume I would have the need to apply a Temperature Coupling node under Multiphysics for I use two different physics nodes for two areas, but I don't find how I could manage to emplace a mulitphysics node - so probably different way is intended.
I assign the materials "copper" to the outer cylinder and "water" to the inner cylinder.
I assign the physics:
- "heat transfer in solids" to the copper area and
- "heat transfer in fluids" to the water area.
At this point I get a situation which as I interprete is a hint for some error in reasoning on my side:
Both "heat transfer in solids" and "heat transfer in fluids" offer "thermal insulation" for all their individual cylinder surfaces; the inner fluid cylinder for its 6 surface segments and the outer hollow solid copper cylinder for 12 surfacce segments, 6 outer surfaces and also the 6 inner surfaces which are in contact with the water.
I would like to have a continuity condition at this inner surface.
I further assign a continuous heat source to the outer surface of the solid hollow outer copper cylinder.
If I try (after meshing and including a time dependend study) to run the model it calculates twice the temperatures (two data sets, ht and ht2) within the inner cylinder.
I assume I would have the need to apply a Temperature Coupling node under Multiphysics for I use two different physics nodes for two areas, but I don't find how I could manage to emplace a mulitphysics node - so probably different way is intended.
Attachments:
7 Replies Last Post 14.04.2016, 08:29 GMT-4
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi,
it is possible to thermally connect both physics modes. However, I think in this case the easiest way would be to use only one mode (e.g. Heat Transfer in Solids) for all domains (copper and water) and to add a Heat Transfer in Fluids Domain within the mode for the water domain. The advantage is that the temperature continuity is automatically realised and there is only one temperature variable.
If there are two temperature modes (e.g. ht1 and ht1), there are by default also two temperature variables T and T2. This might also be the reason for your result (twice the temperature (T) in the inner cylinder).
--
Best Regards
Phillip
it is possible to thermally connect both physics modes. However, I think in this case the easiest way would be to use only one mode (e.g. Heat Transfer in Solids) for all domains (copper and water) and to add a Heat Transfer in Fluids Domain within the mode for the water domain. The advantage is that the temperature continuity is automatically realised and there is only one temperature variable.
If there are two temperature modes (e.g. ht1 and ht1), there are by default also two temperature variables T and T2. This might also be the reason for your result (twice the temperature (T) in the inner cylinder).
--
Best Regards
Phillip
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thank You Philpp. Yes that works so far. Heat now is flowing between the two domains.
I also want to include thermal expansion resulting from the application of a boundary heat source.
Therefore I added the "Structural Mechanics Module" and chose "Solid Mechanics" for the whole system consisting of copper and water. By an error message I was asked to add the physical data of water (Young's modulus, Poisson ratio and Coefficient of thermal expansion) manually. But probably I should not use te "Solid Mechanics" node for the water domain?? But what should I use for the water reacting to and transmitting heat and pressure?? Another COMSOL module?
In the end I get an error message:
Undefined value found.
- Detail: Undefined value found in the stiffness matrix.
There are 1820 equations giving NaN/Inf in the matrix rows for the variable comp1.u.
at coordinates: (-0.0866029,0.0499994,0.1), (-0.0500011,0.0866019,0.1), (-0.0707113,0.07071,0.1), (-0.1,0,0.1), (-0.0965927,0.0258816,0.1), ...
There are 1820 equations giving NaN/Inf in the matrix rows for the variable comp1.v.
at coordinates: (-0.0866029,0.0499994,0.1), (-0.0500011,0.0866019,0.1), (-0.0707113,0.07071,0.1), (-0.1,0,0.1), (-0.0965927,0.0258816,0.1), ...
There are 1820 equations giving NaN/Inf in the matrix rows for the variable comp1.w.
at coordinates: (-0.0866029,0.0499994,0.1), (-0.0500011,0.0866019,0.1), (-0.0707113,0.07071,0.1), (-0.1,0,0.1), (-0.0965927,0.0258816,0.1), ...
and similarly for the degrees of freedom, NaN/Inf in the matrix columns.
I also want to include thermal expansion resulting from the application of a boundary heat source.
Therefore I added the "Structural Mechanics Module" and chose "Solid Mechanics" for the whole system consisting of copper and water. By an error message I was asked to add the physical data of water (Young's modulus, Poisson ratio and Coefficient of thermal expansion) manually. But probably I should not use te "Solid Mechanics" node for the water domain?? But what should I use for the water reacting to and transmitting heat and pressure?? Another COMSOL module?
In the end I get an error message:
Undefined value found.
- Detail: Undefined value found in the stiffness matrix.
There are 1820 equations giving NaN/Inf in the matrix rows for the variable comp1.u.
at coordinates: (-0.0866029,0.0499994,0.1), (-0.0500011,0.0866019,0.1), (-0.0707113,0.07071,0.1), (-0.1,0,0.1), (-0.0965927,0.0258816,0.1), ...
There are 1820 equations giving NaN/Inf in the matrix rows for the variable comp1.v.
at coordinates: (-0.0866029,0.0499994,0.1), (-0.0500011,0.0866019,0.1), (-0.0707113,0.07071,0.1), (-0.1,0,0.1), (-0.0965927,0.0258816,0.1), ...
There are 1820 equations giving NaN/Inf in the matrix rows for the variable comp1.w.
at coordinates: (-0.0866029,0.0499994,0.1), (-0.0500011,0.0866019,0.1), (-0.0707113,0.07071,0.1), (-0.1,0,0.1), (-0.0965927,0.0258816,0.1), ...
and similarly for the degrees of freedom, NaN/Inf in the matrix columns.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Dear Hans,
When there are two physics nodes using Temperature as a variable you can overwrite the “T2” in the second one with T. That way they use the same degrees of freedom. Also instead of using separate fluid and solid heat transfer physics consider using the “Conjugate Heat Transfer” which automatically handles both solid and fluid.
Nagi Elabbasi
Veryst Engineering
When there are two physics nodes using Temperature as a variable you can overwrite the “T2” in the second one with T. That way they use the same degrees of freedom. Also instead of using separate fluid and solid heat transfer physics consider using the “Conjugate Heat Transfer” which automatically handles both solid and fluid.
Nagi Elabbasi
Veryst Engineering
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
One more thought. If you do not expect the change in solid geometry due to thermal expansion to affect the flow I suggest solving the heat transfer problem first then setting up an structural analysis that uses the temperature from the first solution to calculate thermal expansion and thermal stresses.
Nagi Elabbasi
Veryst Engineering
Nagi Elabbasi
Veryst Engineering
Please login with a confirmed email address before reporting spam
Posted:
9 years ago
Hello Mr. Phillip,
Thank you for your help. I am doing simulation for electrochemical cell joule heating. I have my electrodes and electrolyte in an 2D axysymmetric model. I take the joule heat and I am applying heat transfer as you have suggested. It works fine for me. However, within the combined geometry, I wanted to add convective heat transfer at the wall of electrode and electrolyte (within geometry) which is inaccessible being internal boundary. PFA image showing inaccessible boundary. What kind of Physics do you recommend me to use?
Thank you,
Ketaki.
Thank you for your help. I am doing simulation for electrochemical cell joule heating. I have my electrodes and electrolyte in an 2D axysymmetric model. I take the joule heat and I am applying heat transfer as you have suggested. It works fine for me. However, within the combined geometry, I wanted to add convective heat transfer at the wall of electrode and electrolyte (within geometry) which is inaccessible being internal boundary. PFA image showing inaccessible boundary. What kind of Physics do you recommend me to use?
Thank you,
Ketaki.
Attachments:
Please login with a confirmed email address before reporting spam
Posted:
9 years ago
Dear Ketaki,
thats right, you can not convective heat transfer to an "inner" boundary. Convective heat transfer boundary condition is meant to be used if you do not actually want to simulate the fluid outside where the convection takes place.
However, if you are modeling this outside part, there is no need to add the boundary condition because the convective heat flux is computed correctly.
--
Best Regards
Phillip
thats right, you can not convective heat transfer to an "inner" boundary. Convective heat transfer boundary condition is meant to be used if you do not actually want to simulate the fluid outside where the convection takes place.
However, if you are modeling this outside part, there is no need to add the boundary condition because the convective heat flux is computed correctly.
--
Best Regards
Phillip
Please login with a confirmed email address before reporting spam
Posted:
9 years ago
Mr. Phillip,
I am grateful for your support. I am sorry I did not explain my problem properly. I am not simulating the outside part and I am giving convective heat transfer at those boundaries as you correctly said so.
Whereas I want to add conjugate heat transfer to the fluid inside my geometry where it comes in contact with the cathode (solid wall which is also part of my component). I want to add the conjugate heat transfer at the inner boundary of cathode and fluid and simulate that simultaneously with the overall conduction in component. (see attached image).
I have added separate 'conjugate heat transfer' but not sure how to couple them to run it simultaneously at each time step.
Also, temperature of fluid is going above the boiling point even if I add heat transfer with phase change.
Can you help me with these two issues?
Thank you very much.
Sincerely,
Ketaki.
I am grateful for your support. I am sorry I did not explain my problem properly. I am not simulating the outside part and I am giving convective heat transfer at those boundaries as you correctly said so.
Whereas I want to add conjugate heat transfer to the fluid inside my geometry where it comes in contact with the cathode (solid wall which is also part of my component). I want to add the conjugate heat transfer at the inner boundary of cathode and fluid and simulate that simultaneously with the overall conduction in component. (see attached image).
I have added separate 'conjugate heat transfer' but not sure how to couple them to run it simultaneously at each time step.
Also, temperature of fluid is going above the boiling point even if I add heat transfer with phase change.
Can you help me with these two issues?
Thank you very much.
Sincerely,
Ketaki.
Attachments: