[Rotkohl]

Hi everyone,

I get a problem of the numerical diffusion of the phase boundary when simulating a rising bubble in quiescent liquid.

The case is that a single bubble of air (diameter=1mm) rises in a cubic container (8mm*8mm*6mm) fulfilled with water, which boundary condition of its bottom and side is set as Wall and the top the pressure-outlet. The case was simulated both in 2D and 3D and all the settings are the same. I patch the bubble in a circle or spherical region with the initial position (0.004m,0.004m,0.0025m). However, a serious numerical diffusion of the phase boundary occurs only in the 3D case (as shown in attachment Figure 1) when bubble have already rised for a while. As seen the interface contains 5 more cells, which is unreanable.

Can anyone help me to solve the problem?

Thank you!

P.S. The improtant settings of the case are presented in the attachments.