large eddy simulation with combustion(idilat)

[chenyx]

Hello,
In code_saturne 4.0.2, I tryed idilat = 1, 2, 4 for LFR(laminar finite rate) combustion model with dynamic LES and idilat = 1 for k-omega SST URANS in a CH4&H2/air bluff body flame.
The results are quite different. In URANS, the time-average temperature in the flame are about 2000 K, but in dynamic LES, the time-average temperature in the flame are about 400-500 K with "idilat = 1", 900-1000 K with "idilat = 2" and 500 K with "idilat = 4" .
Abviously, for my case, the results that temperature is below 1000 K are all wrong.
I first doubt that it's because in one time step, the density and viscosity changed a lot so that the SIMPLEC can not solve u and p correctly, so I set a loop for convergence of density and viscosity in a time step, but the results are same with "idilat = 1".
My question is that with the same configuration, why is such a difference between LES and URANS? What should I pay more attention to when using LES compared with URANS?
Best regards,
Chen yuxuan

[Yvan Fournier]

Hello,

- v4.0 has been retired since 2019, and LES was probably not compatible with combustion in any case
-At the very least, the 2nd-order time scheme recommended for LES was made compatible with variable density only around v7.0.
- In any case, even for "simple" flows without combustion, users should pay more attention using LES than using RANS, as sensitivity to mesh quality, and to other factors influencing computation quality or order is much higher in LES.

So in any case, I recommend upgrading to v7.0 before running any additional tests.

Regards,

Yvan