[paulzhang]

Hi All,

I am trying to simulate a long tunnel flow case, in which the flow stream conditions are:
rho=0.3kg/m^3, T=1000K, p=10,000Pa, V=0.3 m/s. For testing, it is just a laminar case. I am developing a compressible flow code, so no such divergence free conditions at all. The density is changing.

My question is for such low Mach=1.e-3, Re~1, which suitable convective flow scheme I should use? I tried Roe and Vanleer. The result is not satisfying .


Thanks a lot.
Paul

[FMDenaro]

Originally Posted by paulzhang

Hi All,

I am trying to simulate a long tunnel flow case, in which the flow stream conditions are:
rho=0.3kg/m^3, T=1000K, p=10,000Pa, V=0.3 m/s. For testing, it is just a laminar case. I am developing a compressible flow code, so no such divergence free conditions at all. The density is changing.

My question is for such low Mach=1.e-3, Re~1, which suitable convective flow scheme I should use? I tried Roe and Vanleer. The result is not satisfying .


Thanks a lot.
Paul

I don't think the problem is in the convective scheme but in the ill-conditioned system... you need to use a suitable low-Mach formulation

[paulzhang]

Originally Posted by FMDenaro

I don't think the problem is in the convective scheme but in the ill-conditioned system... you need to use a suitable low-Mach formulation

I am not quite sure if I got your point. You mean I have to change my governing equations to make it work at low-Mach flow?

Thanks again.
Paul

[FMDenaro]

Originally Posted by paulzhang

I am not quite sure if I got your point. You mean I have to change my governing equations to make it work at low-Mach flow?

Thanks again.
Paul

In some formulations the equations are rewritten by using a pre-conditioner operator ... the problem with the low-Mach number is the stiff pressure condition you encounter (sound speed tends to infinity).
Many CFD textbook explain the problem