Source term linearization

(Difference between revisions)
 Revision as of 05:55, 7 December 2005 (view source)Tsaad (Talk | contribs)← Older edit Revision as of 05:55, 7 December 2005 (view source)Tsaad (Talk | contribs) Newer edit → Line 21: Line 21: $S_P = -3T_P^{*2}$
$S_P = -3T_P^{*2}$
- =References= + ==References== #{{reference-book|author=Patankar, S.V.|year=1980|title=Numerical Heat Transfer and Fluid Flow|rest=ISBN 0070487405, Hemisphere Publishing Corporation, USA.}} #{{reference-book|author=Patankar, S.V.|year=1980|title=Numerical Heat Transfer and Fluid Flow|rest=ISBN 0070487405, Hemisphere Publishing Corporation, USA.}} #{{reference-paper|author=Murthy, Jayathi Y.|year=1998|title=Numerical Methods in Heat, Mass, and Momentum Transfer|rest=Draft Notes, Purdue University ([http://widget.ecn.purdue.edu/%7Ejmurthy/me608/main.pdf/ download])}} #{{reference-paper|author=Murthy, Jayathi Y.|year=1998|title=Numerical Methods in Heat, Mass, and Momentum Transfer|rest=Draft Notes, Purdue University ([http://widget.ecn.purdue.edu/%7Ejmurthy/me608/main.pdf/ download])}} #{{reference-paper|author=[http://webfea-lb.fea.aub.edu.lb/fea/me/CFD/ Darwish, Marwan]|year=2003|title=CFD Course Notes|rest=Notes, American University of Beirut}} #{{reference-paper|author=[http://webfea-lb.fea.aub.edu.lb/fea/me/CFD/ Darwish, Marwan]|year=2003|title=CFD Course Notes|rest=Notes, American University of Beirut}} - #{{reference-paper|author=[http://jedi.knows.it/ Saad, Tony]|year=2005|title=Implementation of a Finite Volume Unstructured CFD Solver Using Cluster Based Parallel Computing|rest=Thesis, American University of Beirut}}

Introduction

In seeking the solution of the general transport equation for a scalar $\phi$, the main objective is to correctly handle the non-linearities by transforming them into linear form and then iteratively account for the non-linearity. The source term plays a central role in this respect when it is non-linear. For example, in radiation heat transfer, the source term in energy equation is expressed as fourth powers in the temperature.
When the source is constant and independent of the conserved scalar, the finite volume method assumes that the value of S prevails of the control volume and thus can be easily integrated. For a given control volume P, we obtain
$\int_{\Omega} S d\Omega = S\Omega \,$

Picard's Method

Picard's method is the most popular method used in conjunction with the finite volume method. For a given control volume P, we start by writing the source term as
$S = S_C + S_PT_P \,$
where $S_C$ denotes the constant part of S and $S_P$ denotes the coefficient of $\phi_P$ (not the value of S at P). This allows us to place $S_P$ in the coefficients for $\phi_P$.

Let $\phi_P^*$ denote the value of $\phi_P$at the current itertaion. We now write a Taylor series expansion of S about $\phi_P^*$ as
$S = S^* + \left ( \frac {\partial S}{\partial \phi} \right ) ^* \left ( \phi_P - \phi_P^* \right )$
therefore
$S_C = S^* - \left ( \frac {\partial S}{\partial \phi} \right ) ^* \phi_P^*$
$S_P = \left( \frac {\partial S}{\partial \phi} \right ) ^*$
where $\left ( \frac {\partial S}{\partial \phi} \right ) ^*$ is the gradient of S evaluated at $\phi_P^*$.

As an illustrative example, consider $S = -T^3 + 10 \,$. Following Picard's method, we have
$\left( \frac {\partial S}{\partial \phi} \right ) = -3T^2$
$S_C = -T_P^{*3} +10 + 3T_P^{*2}T_P^* = 2T_P^{*3} +10$
$S_P = -3T_P^{*2}$

References

1. Patankar, S.V. (1980), Numerical Heat Transfer and Fluid Flow, ISBN 0070487405, Hemisphere Publishing Corporation, USA..
2. Murthy, Jayathi Y. (1998), "Numerical Methods in Heat, Mass, and Momentum Transfer", Draft Notes, Purdue University (download).
3. Darwish, Marwan (2003), "CFD Course Notes", Notes, American University of Beirut.