Query on VOF for Bubble rise

Vamsi

Hi,

I have been trying for the past few weeks to run a 3D simulation in FLUENT 6.2 for a single ellipsoidal bubble rise in a tank. I measured the bubble rise velocity, after initial acceleration, from PIV. I also got the shape and size of the bubble to be an ellispoid of mean major axis 3mm, minor axes 0.8mm. I set the inital bubble shape as equivalent sphere of diameter 1.24mm and tried the following approaches using VOF in FLUENT:

1. Small domain around the bubble of 10 bubble diameters length, and cross section equal to that of tank, with bubble placed near the top. Inflow and outflow such that velocity in fluid is negative of rise velocity, and bubble velocity is rise velocity. Set the reference frame to move at bubble rise velocity.

2. Small domain of 10 bubble diameters length, and cross section equal to that of tank, with bubble rising at rise velocity, fluid inlet and outlet at 0 velocity, and moving reference frame at rise velocity.

3. Small domain of 10 bubble diameters in all three directions, with all faces described as outflows. Bubble rise velocity and moving reference frame at rise velocity.

I have used structured hexahedral cells of length, width and breadth roughly 1/5 to 1/10th of bubble dimension. Can't go smaller because of computing restrictions. I have tried using RANS, a time step such that a fluid particle may move only 1/10th of any given cell (based on maximum possible rise velocity). I want to use LES, but I get an error message when performing the initializing step.

None of these seem to work. Does anyone have experience with this type of problem? Several papers I've read are not illuminating as to what exactly was done. Please kindly help with suggestions and recommendations.

Regards and advance thanks, Vamsi