" Not adequatly refined mesh" problem not resolved

metamaterials
heat
mesh

#1

Hello,
I have followed the forum advice to increase the value of max refine steps, but after working for about 20min, the same error happens. As an input I am using the absorption from the optical simulation. My geommetry in both simulations are two resonators, but absorption group includes them both.


#2

Hi Roberto,

Sometimes depending on the structure under simulation, even a large number for “max refine steps” may not be enough. For example, if you have a very small “min edge length” so that the solver has to refine the mesh a lot to reach that value. Or if you have some very thin but long geometries that require edge lengths that are even smaller that the “min edge length”. In such cases the solver is forced to use edge lengths even smaller than the “min edge length” value and the meshing can take a lot of iterations. One additional step you may take is to reduce the triangle quality in your mesh which will reduce the number of iterations during the meshing. This can be done from here,


#3

Another point to mention here is that even if the solver gives you the warning (mesh not adequately refined) it still does not imply that the simulation is bound to fail. You may still get a convergence provided that other settings are okay. If you would like to share your file then I can take a look to see if there is some way to get the simulation to converge.


#4

Thank you very much. I tried to reduce the traingle quality, but it 's still not working. How to upload my file, even in the layout mode it is a bit bigger (7343KB) than the limit?


#5

sorry here is the file
https://we.tl/97mc0oEtpT https://we.tl/97mc0oEtpT


#6

Hi Roberto,

Thanks for sharing the file. As I guessed earlier, the problem in your simulation is not coming from an inadequate mesh. In fact if I calculate the mesh as it is, I do not see any warning saying the mesh is not fine enough. The problem in your simulation setup right now is that there is no reference temperature. You need at least one thermal boundary condition providing some reference temperature to the simulation (so that the solver can converge on a unique solution). When I apply a fixed temperature boundary condition at the bottom of the solver region (-z) and set the temperature to 300, the simulation runs fine and the result looks reasonable.

Cheers!

Ahsan


#7

Thank you so much! I don’t know how I skipped that part during the Photothermal heating tutorial! Thank you once again!