Steady flow coupled with SYRTHES

[Brennan Sharp]

Hello,
Is it reasonable to run CS with a steady flow configuration while coupled with SYRTHES?
I have a forced-air cooling analysis. The forced-air flow is steady, which would suit CS in a steady flow configuration. SYRTHES runs only in a transient configuration.
 
When I do combine steady-flow  C_S with SYRTHES,the surface temperatures from the solid domain increase with time as
expected (steady heat input) but there is also oscillation of
temperature superimposed, from one iteration to the next in the first few time steps. This might be a consequence of C_S steady flow iterations interacting with the SYRTHES time-stepped transient solutions. 
Should I interpret this observation as an example of why it is not valid to couple a steady flow C_S simulation with SYRTHES?
Thanks,
Brennan

[Marc Sakiz]

Hello Brennan,

the Code_Saturne/SYRTHES coupling was initially designed for unsteady
flows with constant time step. A modification should be available soon
to allow variable (but uniform) time step.
 

However, all that is exchanged between SYRTHES and Code_Saturne are
temperature values and heat flux values (fluid temperature and exchange
coefficient, to be more exact). So if you're looking for a steady
result, SYRTHES and Code_Saturne do not necessarily need to have the
same time step.

We've already done cases where the SYRTHES time step was much larger
than the Code_Saturne time step, for faster convergence (in that case
we had artificially increased the time step for temperature in
Code_Saturne, using CDTVAR parameter).
 

To my knowledge, we've never coupled SYRTHES and Code_Saturne in steady
flow mode. I don't think there would be "physical" limitations to doing
so, but the steady flow mode is easily unstable, so there might be numerical limitations.
I'll see if we can try it at EDF on our basic SYRTHES/Code_Saturne test case.
 
    --Marc--
 

[Brennan Sharp]

Thanks Marc,
Thanks for that summary - it clarifies some concepts for me. I'd be interested to know the results if you can try your SYRTHES+Code_Saturne case as mentioned.
Meanwhile, the superimposed temperature oscillation I referred to appears to be due to my SYRTHES preparation, not the coupling as I first thought.
 
Cheers,
Brennan

[Marc Sakiz]

Hi Brennan,
I've just tested the steady flow mode on our basic SYRTHES coupling test case. I works all right.
But I still think instabilities could occur on more complex cases ...
 
Cheers
 
  --Marc--