[zhekka]

Good day to all,

I'm beginner in fluent modelling and now encounter with problem:

The task is simple- there is a wall modelling (material- ammonia-liquid) with two sides (right- temperature function from x coordinate, y coordinate and time, initial temperature- 293 K and left- temperature constant 360 K)

The problem is next:

When I set material density I use boussinesq, in this case also should be difined thermal expantion coefficient (which is 0.00245 for liquid ammonia). In this case solution is not converged (laminar or other wiscous model- doesen't matter), however if I reduce this coefficient (0.00245*10^-3= 0.00000245) for the solid values (x*10^-6) all converged perfectly (except common sense)

Also I tried to use piecewise-linear function (using formula: p(t)=p0(1-b(t-t0) for density in second point but the result was the same)

I will be thankful if someone told me where the problem is.