[lablan]

Hello All,

I am currently trying to model a converging diverging rectangular duct with supercritical carbon dioxide as the working fluid. I am using a RGP table as the equation of state which includes metastable properties and as such instead of the saturation line included in the table, the spinodal line is included instead. With the boundary conditions specified, the flow at the throat should go into the metastable region, but not reaching the spinodal limit.

My boundary conditions are mass flow inlet and pressure outlet and I am using the total energy solver.

I am facing problems with getting the solution to converge, its stalling at around between 2e-3 and 2e-4 and has oscillating values for the pressure monitors.

I have tried a few different time scales and also messed with some of the expert parameters, but nothing as really helped in reaching convergence.

I ran a similar case in which had the saturation included and was able to reach convergence with that one, but had to use a very small time step.

I am using this case as initial conditions to the one where metastable properties are included.

I believe the instability is due to the variations in the fluid properties when nearing the spinodal limit.

Does anyone have an suggestions for me to try that might aid in the convergence?

Thanks!

Lauren

[ghorrocks]

There is a general FAQ on this: http://www.cfd-online.com/Wiki/Ansys...gence_criteria

But in this case, as you say the issue is almost certainly caused by your complex material condition. To make this converge you will need:
* Small time steps
* Good initial condition
* Very good mesh quality (this is critical, any improvement in mesh quality will have a major benefit)
* Double precision solver