# Solution of Poisson's equation

(Difference between revisions)
 Revision as of 06:12, 3 October 2005 (view source)Zxaar (Talk | contribs)← Older edit Revision as of 06:14, 3 October 2005 (view source)Zxaar (Talk | contribs) Newer edit → Line 41: Line 41: [/itex]
[/itex]
- One the coefficient and source matrix is constructed, the system could be solved with the help of iterative solvers. + Once the coefficient and source matrix is constructed, the system could be solved with the help of iterative solvers.

## Revision as of 06:14, 3 October 2005

Poisson equation occurs in many forms in CFD. Efficient and fast solution to Poisson equation is important aspect of CFD.

In CFD the Poisson equation occurs mainly in the form:

${\partial \over {\partial x_j }}\left( {\Gamma {{\partial \phi } \over {\partial x_j }}} \right) = q_\phi$

## The solution

The left hand side can be discretised in the manner explained in section related to diffusive term . The diffusive term can be broken into two explicit and implicit parts.

We have, $\vec \alpha {\rm{ = }}\frac{{{\rm{\vec A}}}}{{{\rm{\vec A}} \bullet {\rm{d\vec s}}}}$

giving us the expression:

$D_f = \Gamma _f \nabla \phi _f \bullet {\rm{\vec A = }}\Gamma _{\rm{f}} \left[ {\left( {\phi _1 - \phi _0 } \right)\vec \alpha \bullet {\rm{\vec A + }}\bar \nabla \phi \bullet {\rm{\vec A - }}\left( {\bar \nabla \phi \bullet {\rm{d\vec s}}} \right)\vec \alpha \bullet {\rm{\vec A}}} \right]$

where $\bar \nabla \phi _f$ and $\Gamma _f$ are suitable face averages.
When broken into implicit and explicit parts

$D_{implicit} = \Gamma _{\rm{f}} \left[ {\left( {\phi _1 - \phi _0 } \right)\vec \alpha \bullet {\rm{\vec A }} } \right]$
$D_{explicit} = \Gamma _f \left[ {\bar \nabla \phi \bullet \vec A - \left( {\bar \nabla \phi \bullet d\vec s} \right)\vec \alpha \bullet \vec A} \right]$

This gives us the coefficient of matrix as:

$A_l = A_l + \Gamma _f \vec \alpha \bullet \vec A$
$A_P = A_P - \Gamma _f \vec \alpha \bullet \vec A$

and for source matrix:
$S_U = S_U + q_\phi - D_{explicit} = S_U + q_\phi - \Gamma _f \left[ {\bar \nabla \phi \bullet \vec A - \left( {\bar \nabla \phi \bullet d\vec s} \right)\vec \alpha \bullet \vec A} \right]$

Once the coefficient and source matrix is constructed, the system could be solved with the help of iterative solvers.