Dear Obiajulu, good evening.

You can optimize your transient problem with time advancing methodology. But I really suggest you to increase your time step (for 1s-10s). In your problem the phenomena involved do not require (usually) so small time steps.

For convergence problem, you should choose an implicit method, since your time step is not so small and your mesh can be coarse.

I hope to be helpful.

Hi Guilherme,

Thanks very much for the prompt response. May I please understand what exactly you mean by time advancing methodology. I'm still experiencing convergence problem even with the coarse mesh. For a natural convection problem, shouldn't it be necessary to use finer meshes at the wall? Or doesn't that matter? I should also mention that there is a source of heat within the tank (a phase change material). So there are two natural convection phenomena. First is heat transfer from the PCM to the fluid in the tank and the second is heat transfer from the walls of the tank to the ambient. I'm using the multiphase, energy and solidification and melting models in Fluent.

Dear Obiajulu,

In your problem the mesh next to the heat source boundaries are crucial. You must refine this region due to boundary layer phenomena in there.

The time adaptive time-step methods change the time step in each phase of the simulation. So you can reduce your simulation time when the steady-state is approached. But I still think if you increase the time step, your simulation will keep stable and relevant.

Best Regards.

Thanks Guilherme,

I managed to model 2 systems (water tank and phase change material)  independently and i achieved convergence in both cases even without using refinement near the walls. If the solution converges without refinement near the wall, is the result still somewhat reliable? Now, even though simulating both systems independently seems to work, when i couple them (putting the phase change material in tank), the simulation does not converge. What could likely be the problem here?

Best Regards

Guilherme B. Lopes Jr

sir, I have also a problem with heating water in a tube, in the transient thermal module of Ansys. the problem is that when I fine the mesh the results are not converging, and at coarse mesh the results comes quite good