CFD Online Discussion Forums

CFD Online Discussion Forums (
-   STAR-CCM+ (
-   -   Free surface bubbling/air gap problem (

TWaung November 3, 2011 13:58

Free surface bubbling/air gap problem

I am attempting to model a flume tank (rectangular channel with free surface flow), with a wall and an actuator disk. Basically a dam with a large turbine. The attached images show a side view of the domain, with the white rectangle representing the wall, and the rectangular outline below the wall representing the extents of the actuator disk.

I am finding that initially, the water surface looks like the first image. Smooth, with no bubbles or gaps on the upstream side of the turbine. However, after running the simulation for a long time in order for the velocity through the actuator disk to reach a steady value, these bubbles/air gaps start appearing in the free surface upstream of the turbine (seen in the second image). This behaviour has also shown up even if I round the edges of the barrage.

I am primarily interested in the head difference across the wall, so this is a problem for me.

Does anyone have any idea what is going on?

The details of the sim are provided below:
Domain: 3D rectangular channel with a wall and actuator disk

Inlet BC: Mass flow inlet for water phase
Outlet BC: Pressure outlet with user function defining hydrostatic pressure
Initial Water level: Defined using a field function
Tank side walls: symmetry planes

Turbulence Model: k-epsilon

Thanks for any suggestions!

abdul099 November 5, 2011 17:51

Do you run steady state or transient? Have you checked your mesh?

TWaung November 5, 2011 19:37


I am running a transient simulation. I have a pretty fine mesh around the free surface. Do you have any mesh advice/ rules of thumb?


ping November 6, 2011 04:29

you are hiding important information in your VOF scalar plot - how well the air and water are being separated at the interface - use the default number of colours and ideally for VOF you should see non red and blue across one maybe two cells. zoom in and look at the problem area - you might need a finer mesh and/or finer timestep - also plot convective courant number and try to keep below 1.

All times are GMT -4. The time now is 17:30.