[ROY4]

Hello,

I'm modeling a bubble column and in a paper I read that they have used SIMPLE algorithm for pressure velocity coupling.I studied the CFX guides and did not find any thing about SIMPLE algorithm and how to bring it into calculations. There is just a Rhie-chow method. Can any one please help me what should I do?

By the way I wanted to ask what will happen if I do not set any thing for tabs that I brought in attached pictures? Will it make my calculations wrong? or CFX will use any special default for it?


Thanks in advance

[ghorrocks]

SIMPLE refers to the solution technique, especially how the pressure/ass equation is coupled to the momentum equations. SIMPLE commonly used on many solvers, but is very old. It solves them sequentially.

CFX uses a coupled technique to link the pressure/mass equation to the momentum equation where the pressure/mass equation is solved simultaneously with the momentum equations in one big matrix. This requires more memory to solve, but for most cases converges far faster. And memory is cheap these days so for most cases the coupled solver is far superior.

CFX has no option to change the coupling. You cannot run CFX in SIMPLE mode, the coupled solver is hard-coded into the software.

[ROY4]

Originally Posted by ghorrocks

SIMPLE refers to the solution technique, especially how the pressure/ass equation is coupled to the momentum equations. SIMPLE commonly used on many solvers, but is very old. It solves them sequentially.

CFX uses a coupled technique to link the pressure/mass equation to the momentum equation where the pressure/mass equation is solved simultaneously with the momentum equations in one big matrix. This requires more memory to solve, but for most cases converges far faster. And memory is cheap these days so for most cases the coupled solver is far superior.

CFX has no option to change the coupling. You cannot run CFX in SIMPLE mode, the coupled solver is hard-coded into the software.

Dear ghorrocks,
Thanks for your fast answering, is Rhie-Chow the method of coupling that you mentioned above?
and I understood that my pictures where not attached, so I attach them again here, Does it make any problem if I do not set any thing in these tabs? will CFX use its own defaults if I do not set any thing?
and my problem is Eulerian multiphase by the way

[ghorrocks]

The golden rule is: Unless you have a very good reason to change the defaults just leave it the way it is. So unless you know why the defaults do not work for your case then just leave them alone.

Rhie-Chow is the method used to avoid check-boarding of the pressure-velocity coupling due to the odd and even nodes being uncoupled. This is completely different to pressure-velocity coupling method, and Rhie-Chow is applicable to both SIMPLE and Coupled solvers. Rhie-Chow is discussed in most introductory CFD textbooks, I recommend you have a read to understand the background.

[ROY4]

Originally Posted by ghorrocks

The golden rule is: Unless you have a very good reason to change the defaults just leave it the way it is. So unless you know why the defaults do not work for your case then just leave them alone.

Rhie-Chow is the method used to avoid check-boarding of the pressure-velocity coupling due to the odd and even nodes being uncoupled. This is completely different to pressure-velocity coupling method, and Rhie-Chow is applicable to both SIMPLE and Coupled solvers. Rhie-Chow is discussed in most introductory CFD textbooks, I recommend you have a read to understand the background.

I will read. Thank you so much.
I have left the settings on its own defaults and did not change them. But as my results are not getting validated for sparger region, I thought that maybe I need to set some thing in mentioned tabs.

[ghorrocks]

This FAQ may help: https://www.cfd-online.com/Wiki/Ansy..._inaccurate.3F

The inaccuracy is much more likely to be caused by:
* inappropriate physical models
* Too coarse mesh, or too low quality
* Too large time step (if transient)
* Too loose convergence