[Benzaa]

When do you know that a system of coupled equation will generate a stable solution??

I've noticed that the source term plays a huge part during the solution of the differential equation. If it is too large, then there is a good chance that the system will be difficult to solve (or the solution will diverge).

My problem is too define how large can it be??

For example, if i am calculating temperatures, the source term can be in the order of thousands, whereas with concentration, the source term is usually a fraction.

I know that if I split the source term into Sc and Sp (like Patankar suggests) that is a good way to improve the stability. But in some cases, this is not enough.

Could anyone share some of his/hers experiences with this issue?