CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   urgent help needed (rhie-chow interpolation problem) (

Ardalan June 6, 2010 15:43

urgent help needed (rhie-chow interpolation problem)
I've got a problem with rhie-chow interpolation. In the lecture notes this interpolation is used to calculate the mass flow rates which are used in the pressure correction equation. As we know a_p and velocity field from momentum equation solution, the implementation of this interpolation is straightforward.
As I understood the rhie-chow interpolation is used to calculate velocities at cell faces. In discretized momentum equations mass flow rates at faces are also needed to calculate a_p and a_nb (convective part of coefficients). Should we also use the rhie-chow interpolation for the calculation of mass flow rates for this case? If your answer is yes, the problem which remains confusing for me is on one hand a_p calculation needs mass flow rate to be known, on the other hand calculation of mass flow rate by rhie-chow interpolation needs a_p to be known. How can I solve this problem??

arjun June 6, 2010 18:26

use AP from last iteration, start with value of AP = 1.

gerardosrez March 18, 2011 16:22

Hello Ardalan,
Just for curiosity, what kind of grid are you using? and is it a cell centred or a vertex centred scheme?

I tell you that I'm working in an unstrutured vertex centred scheme.

For your question I've read a lot about it and it's really confusing sometimes. In this case you need to remember the procedure to solve de convection - diffusion equation which is that you need the cell face velocity to know if the mass flow is entering or leaving the control volume because if you use an upwinding scheme (like I'm using) you're code then will fill the a_p or the a_nb coefficient with the mass flow value calculated at the control volume face.

In the case of momentum equation is the same thing, you need a cell face velocity to determine which coefficient will you fill with the mass flow value calculated at the face, this is to give more weight in the equation to the velocity value of the central node or it neighbours.

I propose you something give me you're email and let's share information about it, I have already developed a code in MATLAB not accurately working but I have gone far away i this subject. I'm sure we can help each other to get rid of this quickly

All times are GMT -4. The time now is 12:40.