Deformed Bubbles

Using some computed data from Prof. John McLaughlin and Dongying Qian at Clarkson University.

We are interested in determining the criterion for bubble breakup in turbulent flow. We are using the lattice Boltzmann method to compute the behavior of a bubble in a homogeneous turbulence field under zero-gravity conditions. In a typical run, we create a single bubble in the middle of the computational box. Then, we turn on a random stirring force that creates turbulence. If the turbulence is strong enough, the bubble breaks up into one or more smaller bubbles. Even if the bubble doesn't break, it develops into very complicated shapes if the turbulence is strong.

This work was supported by the Engineering Research Program of the Office of Basic Energy Sciences at the Department of Energy under grant DE-FG02-88ER13919. It was also supported by a grant from DuPont.

(For background into the vis tools being used here, refer to this page.)

surfArea = 3917.12371075

surfArea = 3942.01223861
As long as there is just one bubble, the vtkSurfaceReconstructionFilter works quite well. (There are a couple anomalies on the finer mesh -- a tiny hole and a tiny bump). If your browser supports VRML files, you can download the VRML data (~1M).

0-10: Left-to-right, top-to-bottom:

Surface area calculations:

0) 3275.04006944
1) 3408.17386152
2) 3495.83317411
3) 3655.28416597
4) 3534.66402797
5) 4027.3778629
6) 4140.53521925
7) 4356.96326184
8) 4643.39795774
9) 5505.60795537 (bogus due to split across periodic BCs)
10) 4956.31038353 (bogus due to split across periodic BCs)

To insert these into a report, here is a (gzipped) .eps file containing a tiled image (4x3 images), and a .tex file showing how to display it: