Hi, Friends;

Has anyone used Van Leer's vector splitting in Cylindrical coordinates? How to split the flux terms in the R-direction which contains [R*rho*Vr, R*rho*Vx*Vr, etc...]? I made a simple extension of its original formular by multiplying an R in front of F+ and F-. It seems not right . Any comments? references? experience? Thank you in advance.

H2

The formulas for Van Leer's Flux Vector Splitting are usually written in a coordinate independent, finite volume format. If you haven't seen those yet, simply visit my website at www.cfd4pc.com where you can find the 2-D version. Extension to 3-D is straightforward.

You should have no problem working in a finite volume format, whether cylindrical, spherical or fully 3D, the equations are very simple.

There is really no point deriving them for the cylindrical or spherical equations. You are just giving yourself a hard time by doing so.

Let me know if you need more help!