Page 1 of 1

Steady-Compressible flow

Posted: Mon Sep 04, 2017 1:59 am
by stephio_maf
Hello,

I would like to know if in CS 5.0, it is now possible to set a Steady-compressible case (in my case steam) using the GUI. Is it possible to get a guideline?

Stephane

Re: Steady-Compressible flow

Posted: Mon Sep 04, 2017 4:45 pm
by Yvan Fournier
Hello,

There was no development in that regards to my knowledge.

Did you test / try to run ?

Regards,

Yvan

Re: Steady-Compressible flow

Posted: Fri Sep 08, 2017 6:05 am
by stephio_maf
Hello,
when using the GUI, the compressible model is not available with "steady flow" algorithm selected :cry: .

Steph

Re: Steady-Compressible flow

Posted: Wed Apr 17, 2024 1:54 am
by StandardRANSUser001
Dear Yvan and Steph,

It has been quite a few years since you have had this conversation, and I wanted to enquire whether there has been an update on this functionality. I am developing a case on CS 8.0.3, and the steady option is unavailable for the compressible solver in the GUI. Can this option be accessed without the GUI?

Any assistance on this would be greatly appreciated.

Best regards,
Sean Hanrahan

Re: Steady-Compressible flow

Posted: Thu Apr 18, 2024 9:20 pm
by Yvan Fournier
Hello,

I do not think there has been any new development in this regard, though there has been work on another compressible model which may have different constraints.

Did you try forcing the pseudo-steady algorithm using a user-defined function (I.e. cs_user_parameters) ?

Best regards,

Yvan

Re: Steady-Compressible flow

Posted: Fri May 24, 2024 9:09 am
by StandardRANSUser001
Hi Yvon,

Thanks for your help on this.

Could you please confirm that I have correctly modified the cs_user_parameters.c file.

void
cs_user_parameters(cs_domain_t *domain)
{
cs_time_step_options_t *time_opt = cs_get_glob_time_step_options();
time_opt->idtvar = CS_TIME_STEP_LOCAL;
}

For others reading this post, there is a good description of pseudo-steady algorithm in Essentials of Computational Fluid Mechanics, by Jens-Dominik Műller, page 65.

I have a question about the implementation of the pseudo-steady algorithm in Code Saturne.

Starting with your previous comments here: viewtopic.php?t=2558
You have mentioned that the pseudo-steady algorithm keeps pushing the time-stepping to a CFL number of 1. Does this mean that the CFL number defined in the GUI is not used in this calculation, or does this mean that the user defined CFL number correspond to the time-steps taken from this algorithm? Which time settings (in the GUI) are used by this algorithm?

Finally, I would just like to confirm whether CS_TIME_STEP_STEADY can be used for compressible flows without the GUI?

Best regards,
Sean Hanrahan

Re: Steady-Compressible flow

Posted: Fri May 24, 2024 9:47 pm
by Yvan Fournier
Hello,

Yes, your setting seems OK.

I am not sure the pseudo-steady algorithm matches that on your reference, as I am not familiar with the book. It would be interesting to compare.

In our case, the code simply adjusts the time step in each cell so as to try to target a CFL of 1. The reference time step defines the initialization, and as you can see setting it through the GUI, the local time step cannot go under or above a given multiplier times this reference time step.

As for the compressible case, I doubt that this algorithm is currently compatible, but do not know whether it would be easy to adapt or not.

Best regards,

Yvan