Hai,

I am working on "flow past a square cylinder" for incompressible flows using finite difference, explicit scheme. The method is based on "SOLA". For turbulent flows, I have incorporated reynolds average equations of motion, k-equation and dissipation equation in my code. I am facing some difficult when doing finite difference for turbulent viscosity and viscous terms. Because, turbulent viscosity is also variable. Which type of differencing is to be adopted central or forward differencing?. I shall be waiting for your kind reply.

bye.........

srinivasalu,

try both method and see which works for you better

later,

Take care when you use central-diference and first order approximation. i) The artificial viscosity is the problem in the first order method. ii) oscilations are the problem in the central-difference. I prefer to use others high order approximation. By.