[evcelica]

I have been struggling to get a simple phase change model to work properly at low vapor qualities (~15% and below) when running a transient model.

Higher Vapor Qualities work perfectly. I tested my material properties and the phase change in a closed test model where I added heat and watched the pressure, temperature, and vapor quality rise as expected according to hand calcs.

It is a Homogeneous Binary Mixture of N2 and N2VAP.
Material properties from an rgp file, and I also tried an equation of state with the same result.

The weird thing is if I run a steady state model where I have 0% vapor quality @ the inlet, and add heat along the length until it becomes 60% vapor quality @ the outlet, the steady state model works perfectly @ the low vapor qualities. It is only the transient model where the math explodes (quality drops and pressure explodes)

Has anyone seen this problem @ low vapor qualities.

The model is only a 1 dimensional line.
timesteps were reduced plenty CFL was 0.5, and reduced all the way down to 0.0005, still explodes after 2-3 iterations.

I tried all the following, plus more, with the same outcome:
- The in-homogeneous multi-phase model,
- Swapping equilibrium fraction and constraint on the phases
- Starting with equal 10% vapor quality everywhere in the line
- Not even adding any heat to the model, just letting it sit there.
- Starting @ slightly higher and lower than saturated Temp
- Starting my transient model from the initial conditions of the converged steady state model.

Any clue why CFX hates transient phase change @ low vapor qualities? but work fine in steady state?

I've simplified my initial conditions to get a usable solution to my problem, saying we just start with the integrated mass in the line, and use average density and equivalent vapor quality everywhere in the lines.

But I'm curious on what is causing this to blow up @ low vapor qualities in transient mode?