[Upyoung]

Hi all,
I'm trying to simulate a simple two-phase bubble column flow using Euler model.There are one air inlet (velocity inlet) bottom and one outlet (pressure outlet) top.

In order to know elevation and shape of the air-water interface and whole velocity fields, i used a transient method. The domain is patched into water and air separately with a precise air-water interface initially.

When I calculate a 3D case, I just found that a down toward velocity (for both phases) emerged at the air-water interface at the very begining and never disapear (see pic). But when the case was changed to 2D or a VOF multiphase model, the interface would not show up.

Can anybody tell what‘s wrong it would be？

ps： I have carefully refined mesh near the interface with a high quality.

Any suggestion would be help, thank you in advance
[IMG]G:\1.jpg[/IMG][IMG]G:\2.jpg[/IMG]