Hey folks,
in “DNS of combined forced and natural turbulent convection - N. Kasagi et. al (1997)” they make a DNS computation in a vertical plane channel. The left and right side has isothermal and noslip boundary conditions, the other walls are periodic.
The mean flow direction is upward and they control the flow with Grashof- and a Bulk-Reynoldsnumber.
I’am not sure which initial conditions I should use. Do I get turbulent flow, if I use (u,v,w,T) = (0,0,0,Tm) or should I use a turbulent velocity and a linear temperature profile as initial conditions.
If yes, how can I create an initial turbulent velocity field?
Greetings