CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Main CFD Forum (https://www.cfd-online.com/Forums/main/)
-   -   Flux vector splitting (https://www.cfd-online.com/Forums/main/1652-flux-vector-splitting.html)

jypark December 15, 1999 01:25

Flux vector splitting
 
hi everybody!

I calculate the flow in compressor with van-leer's FVS at rotating coordinates.

I have converged solution but the mass is not same between inlet and exit.

so I think that there is something wrong in FVS.

when I used FVS, I calculated the relative Mach number.

then the sonic speed is obtaind from density and pressure.

I want to know that whether the sonic speeds are same in rotating coordinates and absolute coordinates.

and If anyone know the resaons of difference in mass between Inlet and exit, please let me know.

thanks.


John C. Chien December 15, 1999 02:25

Re: Flux vector splitting
 
(1). Refine the mesh until the solution is mesh independent. (2). Change the method, and try the same procedure again until the mesh independent solution is obtained. (3). Change the flow condition, so that the temperature and velocity are change, repeat the same calculation again. (4). Make sure that you change the temperature at least 10 times larger, and run additional calculations. (5). Then you should get the answer to your question. (6). Next time, make sure to mention the Mach number at the inlet, laminar or turbulent, or inviscid, Reynolds number, mesh size. If you still don't have the answer, come back here and tell us what you have done.

leestein December 15, 1999 17:58

Re: Flux vector splitting
 
Hi, Park,

it is well-known that the FVS of van Leer is very good for Euler flows, but has big problem in boundary layer. So, this scheme is not good for viscous flow. You can try either AUSM (new version of van Leer) or Roe.

Good luck!


All times are GMT -4. The time now is 06:27.