[dmoroian]

Hello everybody,
It came to my attention that a simple transport equation may be written in two ways:

solve
(
<blockquote>
fvm::ddt(T)
+ fvm::div(phi, T)
- fvm::laplacian(DT, T)

);
</blockquote>
and

fvScalarMatrix TEqn
(
<blockquote> fvm::ddt(T)
+ fvm::div(phi, T)
- fvm::laplacian(DT, T)

);
TEqn.relax();
solve(TEqn == 0);
</blockquote>

What is the difference? Is there a difference?

Dragos

[grtabor]

I think the only difference is the second case includes underrelaxation on the matrix solution. Take that out and they are formally identical (I think). There might be some subtle difference in efficiency related to declaring a fvScalarMatrix explicitly vs implicitly I suppose.

Gavin