Learning Forum

[raushan79]

A_Uniform Mesh_ time 330.PNG

[Rob]

As we don't know what adaption criterion were set, what the flow looks like or what images you've posted (https://forum.ansys.com/discussion/23093/why-ansys-employees-dont-download-attachments) it's quite hard to diagnose.

[raushan79]

Hi Rob
Thank you for your reply.
I have defined a volume fraction gradient for a flow domain for the fine mesh adaptation criteria. As shown in the figure

I am defining double-level refinement
with these settings. if the simulation runs continuously then mesh adaptation is smooth as shown in the figure.

But, if I am restarting the simulation at some previous time step then the mesh adaptation is non-smooth as shown in the figure.
Please let me know if you need any other information. thank you

[Rob]

Can you post an image of the phases? I wonder if the interruption caused 1-2 steps to miss the refinement step, so the result is the same but the mesh is slightly different.

[raushan79]

thanks for your reply. here I am posting everything related to my model

solution model
solution method
Initialization
run calculation
Mesh adaptation settings

[raushan79]

[raushan79]

Mesh adaptation is non smooth and it is forming a deformed interface

[raushan79]

[raushan79]

Volume fraction plot for both phases

[raushan79]

Hi Rob Did you find any solution for the asked query.
Thankyou

[raushan79]

Hi Fluent
Extremely looking for your support.

[Rob]

Are you saving h5 format or the older version?

[raushan79]

Hi Rob
I have saved .cas.h5, .dat.h5 files and .cas and .dat files.
I have restarted the simulation with .cas.h5 file and .cas file. I found the same problem with both of the files. The mesh was not adapting smoothly.
Extremely looking for your support.

[Rob]

h5 retains the adaption register data whereas the older format would recalculate some of it.

[raushan79]

But I am getting the same problem with h5. Should I go for some other check?

[Rob]

A colleague's just spotted that the image you show on 10th has the data name as .dat.gz so the refinement is fully retained when you save & reload. With h5 the data is retained. He also recomends using the pre-defined VOF adaption (refine and coarsen) criterion as the gradients can be case dependent.

[raushan79]

I have also checked with the pre-defined VOF adaption (refine and coarsen) criterion. But, I got the same issue.

For 2 D problem, the mesh is adapting smoothly with restating the solution at any time step. The problem is with 3 D case.

[Rob]

Run the model again in 3d and make sure you save the files with .h5 format. See if the problem is repeated.

[raushan79]

I have run and find the same issue with restarting the simulations.
thank you

[Rob]

With h5? I keep coming back to this as the image you posted originally uses .dat.gz and not .dat.h5

[raushan79]

Hello Rob
I have also tried saving the files in CFF format (.h5) files as you suggested, but still, I am finding the same issue. When I am restarting transient simulation at some time step after reading the .cas.h5 and .dat.h5 file and running it for further time stepsthen mesh is adapting non uniformly in the flow domain. Please let me know if you need more information to solve this issue.

[raushan79]

Hello Rob
I have also tried saving the files in CFF format (.h5) files as, but still, I am finding the same issue. When I am restarting transient simulation at some time step after reading the .cas.h5 and .dat.h5 file and running it for further time stepsthen mesh is adapting non uniformly in the flow domain. Please let me know if you need more information to solve this issue.

[Rob]

In that case the solver is recalculating the adaption registers as part of the solution twice when you restart. The gradient data is not saved in the case file, so there must be another step as you restart the model. Instead of going straight to the calculate button on restart try using
it 20
in the TUI. This will continue the time step rather than moving onto the next one. This may mean the gradients are recalculated before the next adaption register and cycle is carried out.

[raushan79]

Hello Rob,

I apologize for the late reply. I was involved in some other project for a year. Now, came back to the same previous simulation project.

As, you have suggested to me, [In that case the solver is recalculating the adaption registers as part of the solution twice when you restart. The gradient data is not saved in the case file, so there must be another step as you restart the model. Instead of going straight to the calculate button on restart try using
it 20
in the TUI. This will continue the time step rather than moving onto the next one. This may mean the gradients are recalculated before the next adaption register and cycle is carried out.]

I did the same but the problem is still continuing.

Extremely looking for your support.