[luckyluke]

I want to make it clear how one certain scheme (e.g. Gamma201) is called by "div(phi, gamma)".

I got the route as following:
<1>fvm::div(phi, gamma)
<2>.fv::convectionScheme<type>::New
(
vf.mesh(),
flux,
vf.mesh().divScheme(name)
)().fvmDiv(flux, vf)
<3>.return cstrIter()(mesh, faceFlux, schemeData)
I cannot understand the '()' operator in step <2> and step <3>. And the member function "fvmDiv(~~)" of convectionScheme class is a virtual one. I am confused.

Jasak,weller, or anyone else, please give me a detailed explanation for my stupid question?

By the way, what should I do if I want use a new scheme in my solver?

[mattijs]

1. looks up the scheme to be used from your fvSchemes dictionary
2. instantiates a scheme with this name

Then fvmDiv uses this scheme to do the discretisation.

Have a look at e.g. linearUpwind.H and .C

Just create a copy of those files and rename them and replace all 'linearUpwind' with your scheme's name. Compile & link in the .C file and you will have your new discretisation scheme.