Hi all,
I am trying to model temperature gradient focusing via joule heating. Electric current was applied at a micro-channel filled with electrolyte. The current then induced joule heating, an electrophorectic flow and electroosmosis opposing each other. Due to the temperature dependent material properties of the electrolyte, a high concentration region would be created.
In comsol, I used electric currents (ec), creeping flow(spf), heat transfer in fluid (ht), transport of diluted species (chds) and joule heating. Everything went fine until I introduced a Poisson equation to solve for electric double layer potential. Then every time I run it comsol give me an error of:
"Recoverable error in residual, but solver could not recover.
Time : 0.02179981836602723
Last time step is not converged."
I am new to comsol. So any help would be appreciated. Thanks.
Comsol uses "finite element" in order to solving problems. maybe this method is not so strong to solve your problem. You can try "Fluent" software which uses the "finite volume" that is so stronger than finite element.
Thank you Franco, I have used a very fine mesh to model, it helps to carry the model to a slightly later timestep, but that's all it got to, I will try to make a even finer mesh at the junction where it usually has large changes in value.
Thanks Aref, I will consider switching if nothing works.
Hi Chun.
Try working on the mesh.
Keep in mind that the finer the mesh is, the more difficult is to obtain a result immediately. Your work should consist in a progressive mesh refinement procedure.
Refinement means both an increase in the number of elements and a mesh improvement in the critical points of your domain in order to help convergence of numerical methods.
I usually start with a coarse mesh to then play a progressive refinement giving as starting point of the simulation results obtained with the coarser one.
Hi Sabine and Pacolo, thank you for you help, I think the error has something to do with no slip BC inside a micro-channel. Anyway, I got around the error by using comsol in-built function for electroosmosis instead of solving the PDE. Thank you very much for helping.
I have the same error when solving dbd plasma actuator with voltage higher than 1200 V
i use extermely fine mesh but didnt help to much..my system domain is 1*0.5 mm
I found a different solution You may increase the "Relative Tolerance" in the "time dependent solver" node or any relevant node in the "Solver Configuration". I think this leads to more tolerable divergence in the residuals and make the solution on going without that error. This option worked for me very well.
- Badges
- Science topic