[EshitaPal]

Hi all

I am trying to simulate the shell side flow of a shell and tube heat exchanger using buoyantBoussinesqSimplefoam solver. The boundary conditions are as follows:

Inlet: constant T, constant U, zerograd P
Outlet: zerograd T, zerograd U, p_rgh= 0;

I am using kEpsilon turbulence model.
The simulation diverges after few iterations (bounding epsilon diverges). If I switch off the turbulence model then the time step continuity value diverges. I have attached the log file. When I keep g=0 (i.e. no acc due to gravity) then the simuation converges perfectly. This behaves like forced convection.
Can anyone suggest what is the problem with g on?

[jherb]

I think this was already discussed in http://www.cfd-online.com/Forums/ope...roperties.html
So you probably have to modify or boundary conditions to use automatic switching between inlet and outles.

You could also try the transient BuoyantBoussinesqPimpleFoam solver.