[CneSpaulding]

Hi gentlemen,

I'm stuck for 2 weeks on my case, I really don't know how to solve my problem.
That's why I'm writing those few words to ask for some help.

My domain is a 3D rectangle wich represent the 1/4 of the real domain ( I'm using symmetry planes).

I have 2 regions (The rectangle is cut in the middle), wich are glass on the bot and water+air in the second one).

I have a thermal source in the corner of the middle of my domain ( between the 2 symmetry planes ).

I have to modelize a 1/4 air bubble in the corner WICH MUST:
-Stay in the corner
-Keep the same geometry
The bubble is surrounded by water.
The goal is to modelize natural convection in the air&water induced by Boussinesq & Marangoni

My initialization seems correct but in the first iterations the bubble decomposes.. ( I think this is a surface tension problem)

I used for this simulation:
VOF Model // Eulerian Multiphase // Segreted Flow&Temperature // Laminar
Water: Constant density, Laminar
Air: Ideal gas (Error solver diverged), Laminar

I tried Steady and Unsteady models but I have the same behavior.
Surface tension is a function of Temperature.

I did not found any topics for Marangoni problems...
If you have any advices or if you studied the same case do not hesitate to reply!


PS: You can find 2 Screenshots.
http://cjoint.com/14ju/DGdlxnWyrEP.htm
http://cjoint.com/14ju/DGdlxSpxjK6.htm

[me3840]

Your interface between the gas and the water looks to be far too coarse.