I am working with parameter sweep with COMSOL. when the simulation is functioned several times. I feel the time taken to simulate is increasing that the past. can you help me to solve this?
Is the simulation transient? if so i suggest dividing the study into two steps:
Step 1/ steady state
Step 2/ transient which uses the steady state results as the initial conditions.
I am guessing that the simulation is taking longer to converge as the parameter you are changing becomes further from the real initial condition. I hope I have interpreted your question correctly. If not can you provide more detail and I will try to help if I can.
Dear Richard O'Hegarty
Thank you for the answer. But, I am working on PCF mode analyzing.
Thank you so much.
I found the answer for that. The time taken to function the simulation is depending on the number of mesh elements. Now I reduce the number of elements and I can find more fast solution for that. But less number of mesh element reduce the accuracy of simulation results. If you can please help me how to optimize this problem. Thank you so much for the answer again.
Congratulation for solving the issue. Yes. When you have smaller mesh size, you are forced to reduce the time step. That is because of the CFL number. You could run your results first by using the coarser mesh then refine it manually or using adaptive mesh. If your problem is stationary, don't even bother to run the time-dependent solver. One more thing: A good IC is important. Try to have a stationary solution then the time step is a good way to choose.
Shiuh-Hwa Shyu Hi !
can you please tell how to actually change the step size in stationary solver case ?
- Badges
- Science topic