Discussion Forum

Hi everyone,

I have set up a 2D model using General Heat Transfer, WC Navier-Stokes and DC.
Gas is heated in a confined domain by current, and convection cells build up.
Because the volume of the gas is constant, p has to fulfill the additional constraint
that the subdomain integrated equation -p_integrated*one_over_T_integrated+p0/T0=0, i.e. mean pressure has to rise
if the mean temperature does. I have written this contraint in to the "constr" field of a point.
If I start the model using the solver "Time-dependend segregated" I can watch the
gas heat up and the pressure rises to fulfill the ideal gas law. That's fine.
As I'm using DC current for testing the model, pressure, temperature and velocity field
will become stationary after the system has heated up, so "Stationary segregated" should
come to rest as well after some time. But in this case pressure drops down to zero and the
solver stops, as if the constraint is not evaluated any more.
Using "Time-dependend segregated" everything is fine.

I'd appreciate any clue...