Question about residual in S-A model
 

Hi,

I am recently validating my S-A turbulence model by using a plat plate case, but odd thing happened when I was running higher order spatial scheme (2th and 5th order) of inviscid term. The residual was oscillating around 1e-1 and did not converged at all.

In order to exclude the mesh problem, I redid a very simple mesh with just a viscous wall beside without round leading edge, while same thing happened.

Afterwards, to let the case simper, I repeated my simulation with just inviscid solver, unfortunately, still diverged, which confused me a lot.

Could someone please give a suggestion? what would causes the problem? really appreciated your help.

PS: my case run as below:

initial condition:
mach number =0.2 (since compressible solver)
inlet velocity =5.4 (m/s)
inflow turbulent eddy viscosity_tilde =3.0e-5/1.2
density =1.2
temperature,pressure and energy initiated as routine method.

numerical method:
5th order spatial discretization for inviscid flux
2th order spatial discretization for viscous and diffusive flux
2th order time discretization (RK SSPRK method)
CFL number 0.1 (been decreased from 0.5 when running high order scheme)