CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   FVM in 1-D spherical coordinates (http://www.cfd-online.com/Forums/main/83977-fvm-1-d-spherical-coordinates.html)

 lost.identity January 17, 2011 10:06

FVM in 1-D spherical coordinates

I'm trying to obtain discretised equations in 1-D using FVM. It seems to me that there are two ways of doing it,

1) Firstly for the general momentum equation written in vector form (including divergence operators) I apply FVM (i.e. integrate over a control-volume and apply Gauss's theorem). I then write the resulting equation in 1-D spherical coordinates.

2) I write the momentum equation in 1-D spherical coordinates and I have extra geometric source terms compared with the Cartesian case. I then apply FVM (integrate over the volume). This is actually more like finite difference method.

I find the difference between the two discretisation is the extra geometric source terms. I'm not sure which method is more correct but I thought that in 1-D both FVM and finite-difference should yeild the same discretisation. Could anyone suggest which method is best?

Thanks.

 All times are GMT -4. The time now is 16:06.