Learning Forum

[ssyadav]

Dear Abenhadj

Vineeth CNS was my student, thank you very much for guiding us so far. We are able to perform some simulations now but the solver diverges suddenly.

We found that gradually increasing the pressure at inlet and decreasing at the outlet from the mean initial pressure is the best way to proceed. The solver diverges when phase change starts some where in the nozzle.

I was going through the paper "Non-equilibrium condensation of supercritical carbon dioxide in a converging-diverging nozzle" by Alireza Ameli which mentions that

"The Gibbs free energy change of the vapor near the critical point is very small which leads the equation
to have an infinite value and causes error in the solver. In order to prevent this numerical instability, the
solver clips the critical radius to a certain amount, as a default 1 mm. This extremely large default
clipping value decreases the accuracy of the simulation and results in non-realistic droplet sizes. The
problem is that it is not possible to change or modify the formula in the Ansys CFX by the user. To
sidestep this problem, the critical radius was calculated according to the equation (1) in the post
processing stage. It was observed that the critical radius, see Figure 4, changes very moderately also in
the nucleation zone. Therefore, the radius of one nano meter has been estimated to be realistic for the
critical droplet size in this study and in the calculations the critical radius is clipped, by a CEL, to one
nano meter."

We are also trying to clip the critical radius but do not know the variable name in for Liquid CO2 critical radius.

Please help us again regarding how we can limit the critical radius of the liquid phase to say 1 nano meter using CEL expression.

Thanks again for the help.

Regards