CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Main CFD Forum (https://www.cfd-online.com/Forums/main/)
-   -   Mathematical derivation of SIMPLE etc. (https://www.cfd-online.com/Forums/main/127701-mathematical-derivation-simple-etc.html)

Za-ck December 18, 2013 03:28

Mathematical derivation of SIMPLE etc.
 
Hi,

are the any mathematical derivations/proofs which show that the algorithms like SIMPLE/SIMPLER/PISO are working? I can only find the original papers but I don't like to buy them till I know that this mathematical stuff is included.

Best Regards,

Za-ck

RodriguezFatz December 18, 2013 09:38

I don't think there are any proofs. Like all other non-linear solvers (at least that I know) I think you can not show that they converge for every case. As far as I know, this (showing convergence) only works with really simple cases, such as linear problems.

levivad December 18, 2013 10:40

I don't know about the specific algorithms above, but YOU CAN show stability and convergence for nonlinear solvers.
In fact I'm pretty sure you must show stability and convergence of any new solver.
CFX for example is unconditionally stable///

RodriguezFatz December 19, 2013 04:13

Quote:

Originally Posted by levivad (Post 466929)
CFX for example is unconditionally stable///

This statement does not make any sense. CFX is a software tool with tons of different numerical methods.
Please go to CFX and set under relaxation factor for pressure to 100 in SIMPLE. See what happens.

FMDenaro December 19, 2013 04:49

Quote:

Originally Posted by Za-ck (Post 466872)
Hi,

are the any mathematical derivations/proofs which show that the algorithms like SIMPLE/SIMPLER/PISO are working? I can only find the original papers but I don't like to buy them till I know that this mathematical stuff is included.

Best Regards,

Za-ck


Could you better formulate your question? what do you mean with "are working"? Discrete formulations that are consistent and stable must always work ...


All times are GMT -4. The time now is 19:49.