Solving point implicit schemes in openFoam

srinath · October 12, 2008, 12:06

Hello

Is it possible to use openFoam to solve point implicit schemes?

I want to solve
A_i*U_i = B_i, over all cells.
A_i is a 6*6 matrix.
The coupling between the cells is brought about by the way B_i is computed.
Is it possible to solve this on openFoam?

regards
Srinath

hjasak · October 12, 2008, 14:14

Hello Srinath,

Yes: make a loop through all cells, in each cell make a 6x6 dense matrix and invert locally.

There are other ways as well, but this is easiest for the moment.

Enjoy,

Hrv

srinath · October 13, 2008, 01:50

Hello Hrvoje

I need to make a 5*5 matrix(not 6*6 sorry for the typo)
The entries will be functions of u,v,p etc.

So do you suggest i just define a new class, for any general square matrix, and write my own Gauss Siedel routine?
I think the volTensorField is restricted to a sz of 3*3

Regards
Srinath

October 12, 2008, 12:06	Hello Is it possible to use	#1
srinath Member srinath Join Date: Mar 2009 Location: Champaign, USA Posts: 91 Rep Power: 17	Hello Is it possible to use openFoam to solve point implicit schemes? I want to solve A_iU_i = B_i, over all cells. A_i is a 66 matrix. The coupling between the cells is brought about by the way B_i is computed. Is it possible to solve this on openFoam? regards Srinath

October 12, 2008, 14:14	Hello Srinath, Yes: make a	#2
hjasak Senior Member Hrvoje Jasak Join Date: Mar 2009 Location: London, England Posts: 1,907 Rep Power: 33	Hello Srinath, Yes: make a loop through all cells, in each cell make a 6x6 dense matrix and invert locally. There are other ways as well, but this is easiest for the moment. Enjoy, Hrv __________________ Hrvoje Jasak Providing commercial FOAM/OpenFOAM and CFD Consulting: http://wikki.co.uk

October 13, 2008, 01:50	Hello Hrvoje I need to make	#3
srinath Member srinath Join Date: Mar 2009 Location: Champaign, USA Posts: 91 Rep Power: 17	Hello Hrvoje I need to make a 55 matrix(not 66 sorry for the typo) The entries will be functions of u,v,p etc. So do you suggest i just define a new class, for any general square matrix, and write my own Gauss Siedel routine? I think the volTensorField is restricted to a sz of 3*3 Regards Srinath

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Physical Reason for stability of Implicit Schemes?	radhakrishnan	Main CFD Forum	26	October 3, 2023 23:05
Euler Equation Implicit Upwind Schemes	felixrieper	OpenFOAM Running, Solving & CFD	1	February 9, 2014 01:58
Schemes for solving Euler equations	shuo	OpenFOAM Running, Solving & CFD	0	December 10, 2006 05:16
CFL condition for implicit CD schemes	turb	Main CFD Forum	4	August 16, 2005 13:56
Implicit equation solving	Dr B.M. Smith (Smith)	OpenFOAM Running, Solving & CFD	2	December 10, 2004 06:36