[titio]

Hi foamers,

I am having problems with QUICK to simulate the flow in a T-Mixer geometry, with two inlets and two outlets, for a complex fluid where the normal stresses have to be taken into account.

I manage to do the simulations with gamma, upwind, minmod and SFCD, but when I use Quick, I does not work. In particular the simulation blows. Looks like I am using an unbounded Quick method.

How can I use a bounded Quick method? I am testing several methods to see which one is better.

Regards,

António Martins

[santos]

According to the User Guide, the OpenFOAM implementation of QUICK is supposed to be bounded. Like you, I have had problems with this scheme where my unsteady flow simulations diverged randomly. limitedLinearV behaved much better.

Regards,
José Santos

[alberto]

QUICK might lead to instabilities due to structure of the interpolation formula (signs of the terms).

I would let QUICK go since it has no advantage on unstructured grids (It is at best second order, and being upwinded, it is quite dissipative).

You might want to give a try to linearUpwind (and linearUpwindV),

maybe with the cellLimited option:

div(phi,U) Gauss linearUpwindV cellLimited Gauss linear 1;

IMHO, it usually gives pretty good results in many practical applications.

Best,

[maddalena]

Hello Alberto, Hello Josè Santos,
I am looking for some good references on numerical schemes for CFD application. Somewhat that help me to understand what the influence of the different schemes is on the solution, and what is best for specific applications. Any suggestions? Thank you,
maddalena

[santos]

Dear Maddalena,

I would suggest a good CFD book, being my personal option: "Computational Methods for Fluid Dynamics" by Ferziger and Peric.

Regards,
Jose Santos