Abnormal increase in linear solvers elapsed time for LES WALE

Questions and remarks about Code_Saturne usage
Forum rules
Please read the forum usage recommendations before posting.
Post Reply
Mohammad
Posts: 44
Joined: Thu Oct 25, 2018 12:18 pm

Abnormal increase in linear solvers elapsed time for LES WALE

Post by Mohammad » Mon Dec 02, 2019 11:31 am

Hello,

I'm simulating a duct flow using LES WALE model with code_saturne 5.0.9. (parallel computation with 4 cores)
There's something strange in simulation's elapsed time for each iteration.

The linear solvers elapsed time increases dramatically from about 2 seconds at iteration #2 to about 70 seconds at iteration #100 and follows the same procedure which makes the simulation impossible!
I have attached the timer_stats.csv file.
timer_stats.csv
(11.19 KiB) Downloaded 3 times

I'm calculating some time moments. It may increase the elapsed time in comparison with a case without time moments calculation, but it should not increase the elapsed time after each iteration!

This problem only occurs when using LES WALE model. There's no problem with dynamic Smagorinsky model.

Its really strange.

Best regards,
Mohammad

Yvan Fournier
Posts: 2791
Joined: Mon Feb 20, 2012 3:25 pm

Re: Abnormal increase in linear solvers elapsed time for LES WALE

Post by Yvan Fournier » Mon Dec 02, 2019 2:22 pm

Hello,

Is the code otherwise behaving well, or could it be diverging (in whch case you would probably observe an increase in the Courant number in the run_solver.log) ?

Regards,

Yvan

Mohammad
Posts: 44
Joined: Thu Oct 25, 2018 12:18 pm

Re: Abnormal increase in linear solvers elapsed time for LES WALE

Post by Mohammad » Wed Dec 04, 2019 4:12 pm

Hello,

Thank you very much dear Yvan, Your replies are always helpful.

The problem caused by diverging and solved.

Regards,
Mohammad

Post Reply