[daniel_mills]

For a Steady State SIMPLE solver how hard would it be to implicitly solve for p, then explicitly relax p, then update the fluxes based on the explicitly relaxed value of p, instead of implicitly solving, then updating fluxes and then explicitly relaxing?

[hjasak]

Very easy: just swap the statements.

Unfortunately, it would also be not too clever: if you update the fluxes after pressure under-relaxation, you will violate continuity.

Hrv

[daniel_mills]

Thank you for your quick response. I am a CFD newbie. I was just going off the cfd-online description of SIMPLE in which it states the following.

1.Set the boundary conditions.

2.Compute the gradients of velocity and pressure.

3.Solve the discretized momentum equation to compute the intermediate velocity field.

4.Compute the uncorrected mass fluxes at faces .

5.Solve the pressure correction equation to produce cell values of the pressure correction .

6.Update the pressure field: where urf is the under-relaxation factor for pressure.

7.Update the boundary pressure corrections .

8.Correct the face mass fluxes:

9.Correct the cell velocities: ; where is the gradient of the pressure corrections, is the vector of central coefficients for the discretized linear system representing the velocity equation and Vol is the cell volume.

10.Update density due to pressure changes.

My problem is I actually get greater stability with the relaxation value for p set to 1 early in my start up. Than I do under-realxing it. I was just wondering if this had somthing to do with the fluxes of phi?

[daniel_mills]

I might as well include the link.

http://www.cfd-online.com/Wiki/SIMPLE_algorithm

[daniel_mills]

Are steps 6,7, and 8 for SIMPLE_algorithm on the cfd-online Wiki in the wrong order or am I just mis-interpreting the way it is laid out?

[daniel_mills]

Can someone please answer a question for me?

When I am solving for an unsteady swirling flow using rhoSimpleFoam why does setting my relaxation for pressure to around 0.9 actually stabilize the unstable pressure oscillations that develop when the flow field starts to swirl?

I would think the more I lower the pressure relaxation the more stable it would be, but the exact opposite is true. Why??

My relaxations are as follows:

rho 0.1
p 0.9
U 0.7
h 0.7

This works well and gives me fairly similar numbers to fluent.

[msrinath80]

Normally step 6 (Update the pressure field: where urf is the under-relaxation factor for pressure.) is done at the last, but I guess it would not make any difference as the extrapolation of pressure correction at the boundary(s) or the updating of mass fluxes have nothing to do with the corrected pressure [at that point in time]. It is only in the next outer iteration that the corrected pressure itself is used for computation of the gradient (source term).