Learning Forum

[FaizanYounas]

Hello there. I'm conducting a simple case of pool boiling. I have considered a simple rectangular domain with the bottom wall as super-heated. I'm using the VOF method with the Lee model. The results are normal and what I was expecting. But, the problem I observe is the continuity is not converging. The key point to mention here is when I change the scale of my geometry (2cm by 1cm) the continuity does converge. But, when I change the geometry units to mm (2mm by 1mm) this problem arises.

Does anyone have an idea what could be the reason for this behavior and especially how to avoid this?

[Amine Ben Hadj Ali]

More important to check if you mass imbalances if the domain is closed. Also judge the convergence of every time step by monitoring key quantities on top of residuals and imbalanes.

[FaizanYounas]

No, my domain is not closed. There's a pressure outlet boundary condition at the top. Also, if I initialize the solution at the subcooling temperature the continuity does converge. But, only till the vapors starts to form. As soon as the nucleate boiling starts, the continuity diverges.

[Amine Ben Hadj Ali]

Even if it is open you can easily draw mass imbalance in the full system considering what comes in/out of the outlet. Moreover be aware that your result are depending on the Lee Coefficient you are using.

[Rob]

What time step are you using?

[FaizanYounas]

My time-step size is 1e-7s (as it is a 2mmx1mm domain) to keep the courant number under 0.5 or 1.

[Rob]

Drop the time step an order or two in magnitude and see what happens. If the boiling rate is relatively high relative to the cell size you're creating an awful lot of gas when you change phase relative to the mesh.

[Amine Ben Hadj Ali]

.