[Davide_sd]

Hi all,

i'm doing a simulation with simpleFoam. The turbulence model is k-omega SST.
After 11k iteration, the iterative process slowdown... For instance, before iteration 11k the pc was able to do 1000 iters/min. After 11k iterations, there was at most 20 iters/min.

You can see a plot of the residuals in the attachment.
In the attachment there is also the fvScheme and fvSolution...

What did i do wrong?

[Antimony]

Hi Davide,

May I ask what is the problem/geometry that you are simulating and if you could show any results that you got?

The sudden drop in the residuals seems a little weird. Is your case an extremely simple model or something?

Regards,

Antimony

[Davide_sd]

Hi Antimony, thank you for your reply...

i'm simulating a hill.
Mesh lenght: 1.65m
Mesh height: 0.24m
Hill height: 0.06m

The flow is water.

The mesh is essentialy identical to the one i've used in Fluent (i'm doing a comparison: openFoam vs Fluent).
First node height: 0.55e-3 m
50 cells in vertical direction, 336 cells in the horizontal direction.

About 3 hours later of the start of the slowdown, i've stopped the simulation (you can see the residuals plot in the attachment).

I've posted a line-plot where it is supposed to be the recircuation region, and actually OF is able to catch the recirculation... I need to compare the simulation result with the experimental data.

If you need to check the case, i can do an upload.

To run the simulation, i have used the following commands:
pyFoamDecompose.py . 4
pyFoamPlotRunner.py --autosense-parallel --progress --hardcopy simpleFoam

[Davide_sd]

i've compared the results with fluent results: the velocities are almost identical (i still have to compare the turbulent kinetic energy).
still i need to figure it out what is cause of the slowdown...

[Antimony]

Hi,

From what you have posted, it seems like your solution has pretty much converged, especially after 12000 iterations (The residual you get seems to be in the range of machine error of 1e-14) Any particular reason you are interested in the speed of the iterations? How many iterations are you planning to run it till? Since you have stated that you are using simpleFoam, which is a steady state solver, I would think that your simulation is basically done and you would just need to postprocess the result.

Regards,

Antimony

[Davide_sd]

of course my simulation is done, but the problem remain unsolved:

10 minutes to do the first 10500 iterations + 1 hour to do the other 1500 iterations = 70 minutes simulation time to do 12000 iterations...

don't get me wrong, i'm happy with the results, but if i can't understand the reason of the slowdown, probably i'll never switch to OF.... time = money, to me.