CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Running, Solving & CFD (
-   -   Estimation of the flow coefficient of an intake valve (

lucchini February 4, 2006 09:57

Hi All, I'm asking for some
Hi All,

I'm asking for some advices about a CFD calculation I would like to perform. I need to calculate the flow coefficients of the intake valve of a pent-roof spark-ignition engine.
I created the mesh with commercial mesh generator and I converted it correctly. I defined three patches:

- walls: they are the cylinder walls, the port walls, the valve walls
- inlet: where the air comes in
- outlet: where the air goes out


1) Which solver can I use? I was thinking about sonicTurbFoam. But I am not sure it can be the right choiche because the pressure difference experimentally imposed is usually about 0.1 bar. May I consider the fluid as incompressible?

2) If I use a sonicTurbFoam, what boundary conditions should I impose for pressure and velocity at inlet and outlet patches?

I was thinking:

inlet: fixedValue for velocity and zeroGradient for pressure
outlet: zeroGradient for velocity and fixedValue for pressure (set to the experimental value of 0.9 bar).

with this setup I get oscillations in the pressure field and I don't reach a steady solution as I would.

How can I get rid of these oscillations? Should I use pressureTransmissive instead for the pressure boundary condition for the outlet?

I really appreciate any suggestion/comments and would like to know if someone else foam-user has experiences about flow-coefficients estimation for engine valves.

Thanks a lot.



fedegavo February 5, 2006 06:37

grazie tommaso..
grazie tommaso..

eugene February 6, 2006 07:46

Unless your flow is transonic,
Unless your flow is transonic, I would use rhoTurbFoam. And yes, a pressureTransmissive outlet is necessary to prevent wave reflection. (Please let me know how it [pressureTransmissive] performs, the last time I tried to use it I ran into some problems, which I didnt have time to investigate.)

fabianpk February 7, 2006 11:11

Ah finally Tommaso I can help
Ah finally Tommaso I can help you!

I use the pressureTransmissive alot, it's very efficient at damping out oscillating waves in a wedge, especially if there's a very big pressure wave coming from a small nozzle into the wedge.
Nevertheless, remember to that lInf is a compromise between how reflective you can allow it to be and how important it is for the pressure to remain at pInf. A little trial and error is appropriate I think.
BTW How high is your estimated Mach number?


All times are GMT -4. The time now is 14:36.