Poisson Solver

oevermann · April 12, 2006, 05:02

I'm trying to implement a standard projection method
into OpenFoam. For one of the projection steps I need to solve a pressure poisson equation with the pressure defined on the nodes of the grid. From my understanding, most solvers in OpenFoam are such that all variable are defined in cell centers and eventually interpolated onto cell faces. Is it possible to use the solve() command with a variable living on grid nodes? And do all the differential operators such as laplacian, div, etc. work on nodes?

Thanks for info

Michael

hjasak · April 12, 2006, 07:05

Hi Michael,

The tet decomposition solvers will give you a solution on vertices (and other points as well). However, if I were you I would not just jump at it.

The info you have given so far is insufficient for a proper answer. We've got a bunch of points on an unstrictured mesh and you wish to solve the Poisson equation on them. You said nothing on the method of discretisation, approximation of spatial integrals, matrix structure, accuracy etc.

Hrv

April 12, 2006, 05:02	I'm trying to implement a stan	#1
oevermann New Member Michael Oevermann Join Date: Mar 2009 Posts: 15 Rep Power: 17	I'm trying to implement a standard projection method into OpenFoam. For one of the projection steps I need to solve a pressure poisson equation with the pressure defined on the nodes of the grid. From my understanding, most solvers in OpenFoam are such that all variable are defined in cell centers and eventually interpolated onto cell faces. Is it possible to use the solve() command with a variable living on grid nodes? And do all the differential operators such as laplacian, div, etc. work on nodes? Thanks for info Michael

April 12, 2006, 07:05	Hi Michael, The tet decompo	#2
hjasak Senior Member Hrvoje Jasak Join Date: Mar 2009 Location: London, England Posts: 1,907 Rep Power: 33	Hi Michael, The tet decomposition solvers will give you a solution on vertices (and other points as well). However, if I were you I would not just jump at it. The info you have given so far is insufficient for a proper answer. We've got a bunch of points on an unstrictured mesh and you wish to solve the Poisson equation on them. You said nothing on the method of discretisation, approximation of spatial integrals, matrix structure, accuracy etc. Hrv __________________ Hrvoje Jasak Providing commercial FOAM/OpenFOAM and CFD Consulting: http://wikki.co.uk

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Poisson solver	nikwin	OpenFOAM Running, Solving & CFD	1	April 17, 2009 17:59
looking for a 3D Poisson Solver	phsieh2005	Main CFD Forum	2	March 1, 2007 20:33
Different ans of poisson eqn with different solver	Quarkz	Main CFD Forum	0	October 6, 2005 12:49
Poisson solver	Steve	Main CFD Forum	6	July 22, 2004 22:16
Solver for poisson equ.	Capi	Main CFD Forum	1	May 3, 2001 12:20