CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   Jamesonís artificial dissipation in Euler Eq's (

Fereidoun-e-Jam April 4, 2005 07:51

Jamesonís artificial dissipation in Euler Eq's
Dear Friends,

I am trying to simulate the flow field over a 2d bump, i.e. 2d Euler equations. I have a question that how should I calculate the artificial dissipation terms on the first two rows of the cells near the wall (especially the first row over the wall). Also I need the value of the pressure on the wall for boundary condition, is it possible to extrapolate the unknown wall pressure from inside the field?

Your attention and help will be highly appreciated,

With the best regards,

F. Jam

daniel April 8, 2005 18:03

Re: Jamesonís artificial dissipation in Euler Eq's
I have used the Jameson-Swanson-Turkel scheme in a finite volume formulation, and I implemented the bc by rows of extra cells: the internal cell were updated by Runge Kutta, and then the extra cell were changed imposing the BC., thus, the extra cells don't require any spatial derivation, while for the row of cells adiacent to the extra cells I simply didn't introduce the dissipation term. It worked fine. For the pressure into Eulerian field, just update it by dp/dr=rho*Vt^2/r. This equation is exact, But many times the extrapolation is good as well. Hope this can help

Fereidoun April 9, 2005 10:52

Re: Jamesonís artificial dissipation in Euler Eq's
Dear Sir,

Thanks for your comments. I wrote my code. But it does not work because in about 50 first iterations it diverges. I applied two ghost cells below the lower wall and two above the upper wall, also two before the inlet and two after outlet for calculating artificial dissipation terms. But for pressure on the wall I used extrapolation. My inlet boundary condition is supersonic. And outlet also is supersonic. So for inlet I set free stream values for my conservative variables and for the outlet I used extrapolation from the field. I do no know where is my mistake I checked my code many times but it does not converges at all. If you have a FVM code in FORTRAN for this problem (flow in a channel over a circular bump) would you send it for me please?

With the best regards,

F. Jam,

All times are GMT -4. The time now is 19:12.