Discussion Forum

First thing: I didn't notice a mesh anywhere in your model. Normally (at least in my experience) you have to have a mesh to execute a finite element model. Other than that, I think your boundary conditions may have some issues. But... one step at a time.

I ran it with a default mesh and it didn't seem like it was converging.

I would run the electrical problem by itself to make sure that converges OK.

First thing: I didn't notice a mesh anywhere in your model. Normally (at least in my experience) you have to have a mesh to execute a finite element model. Other than that, I think your boundary conditions may have some issues. But... one step at a time.

Thanks for the comment,

I used physics-controlled mesh, with finer element size(which was irrelavent with computing result.). but I couldn't upload the MPH file containing the mesh due to 5-MB size limit. And, I also checked boundary conditions, for both Electric Currents and Heat Transfer in Solids, by trunning off the Multiphysics module and running each physics solely. Which resulted in convergence. (it did not shown the temperature rise by joule heating of course.) It would be so grateful if you take a look once again.

Thanks.

1. Eliminating your specification of the mesh entirely is not necessary to clear the mesh. Instead, you should be able to get your file below 5MB by choosing "Clear All Built Meshes" and "Clear All Solutions" from the ribbon at the top. If you then post that .mph file, others here will be better equipped to find the errors.
2. I noticed different boundary conditions on those two similar rectangles at the top of your model. You might want to take a closer look at those.

There are two problems with the model. 1. There is no mesh setting. Please make mesh node and setting mesh. 2. I think this model shows increasing temperature. So Stationary study not converged. please test this model time dependent study.

Thanks.

1. added mesh
2. changed from normal to next finer mesh to remove warning message about too-small surface area mesh
3. change solvers to all direct using mumps
4. added some plots of temperature at the outer surfaces
5. reduced damping factors on all variables to 0.5, and set the pseudo time stepper on to slow things down to allow for debugging
6. saw that the temperature was eventually going very high
7. decided to reduce the current input at the terminal surface by a factor of 0.5 using an added reducing parameter
8. came to good steady state
9. removed pseudo time stepper
10. added some plots of T, V, and current
11. removed the transparency graphics
12. suggest checking your consistency on which surfaces receive the convective boundary, looks like one of your ground surfaces receives it, and the other doesn't 13.attached the changed model file
13. leave a good tip at the table