Learning Forum

[mechE]

I am trying to simulate the impact and subsequent freezing of a water microdroplet (50 micron in diameter) on a thin sheet of ice resting on a cold wall. I am using the VOF multiphase and Solidification & Melting models in ANSYS Fluent to model the problem. The VOF model is used to simulate the air and water phases and the solidification & melting model is used to simulate freezing of water into ice.  

To set up the problem I patch in a thin sheet of ice (20 micron thick and 150 microns high) along the bottom wall. The temperature in the entire domain is initialized to -35C. The liquid fraction of water is 0 everywhere in the domain, as expected. Under these conditions I would expect the ice to stay stationary. However, upon starting the simulation I observe the ice sheet starting to deform and eventually turn into a droplet. I suspect that this problem is being caused by surface tension. I tried defining surface tension as a function of temperature (0 N/m below -1C, 0.072 N/m above 0C and linearly increasing between -1 and 0 C). This stops the ice sheet from moving but causes numerical issues when I try to simulate the droplet motion (the droplet shape is not as expected, and the liquid fraction doesn’t make sense). Does anyone how to prevent surface tension from moving the solid fraction in a VOF + solidification & melting problem for microscale geometries? Do I need to define surface tension in a different way? I have tried using both the Continuum Surface Force and Continuum Surface Stress models, and neither seem to work. I have also tried playing with the mushy zone parameter, but that doesn’t seem to help either. I have tried everything I can think of and would really appreciate any help on this issue. Thank you!   

Ice wall at times t = 0 microsecond

Ice wall at times t = 25 microsecond

Ice wall at times t = 50 microsecond

Liquid fraction of ice wall is always 0

 

Numerical issues while simulating droplet freezing caused by using temperature-variant surface tension coefficient  

 

[Rob]

The solidification model works by fixing the flow in the solid part. However, at very small scales the surface tension force becomes very significant and can cause this problem. Adjusting the surface tension should work, but you may want to adjust both the mushy values and range you're changing the surface tension over.