User login


You are here

A new type of bubble raft--challenge for clever students

17 years ago, while a postdoc at IBM meant to be doing other things, I thought about the following. Then recently I visited Ali Argon at MIT, and we discussed conventional bubble rafts and how useful they had been in studies of some problems in mechanics...such as of defects and so on.

So, here is the concept, motivated as it was by thinking about fullerenes in 1990.

Consider two concentric spheres, with a gap between them. The inner sphere has radius R1, and the outer spherical shell has radius R2. It can of course have some thickness, and should be transparent. The gap between the inner sphere (which can also be a spherical shell) and the outer sphere's inner surface, is thus R2-R1.

The experimental challenge in configuring a "bubble raft" experiment in which bubbles are formed and maintained in the region between these 2 spheres, is that there should be no physical contact between inner and outer sphere. The inner sphere needs to "hover" insider of the outer sphere.

I had played around with cutting a plastic transparent ball in two hemispherical shells, and attaching each with pins to an inner sphere, and then applying glue to join them. Alas, the pins I used to attach them were contact lines for the bubbles that I ultimately injected, thus messing up the experiment that one would like to have, which is a "3-d bubble raft" covering any topology (here, nested spheres), without any physical objects in the space where the bubbles are formed and where they evolve. Of course I was supposed to be doing other things during my postdoc, so had limited time and it seems to have become only more limited since then.

I have little doubt there is some clever student who can launch this new field of 3d bubble raft studies. Suppose 32 bubbles are injected through a small hole in the outer sphere. What structure do you think will form as the 'minimum energy' structure? If one uses nested cylinders but can deform the outer cylinder (imagine simply pushing in with your finger or perhaps a sharper object, or attaching "handles" so that you can push in or pull out) to give local positive, or negative, curvature, what will the final bubble configuration look like? How to make a system or set of experimental systems so that any topology could be explored, rather than simply flat surfaces?

The problem with this is that it is simply too much fun for a professor to busy himself with. I have more mundane tasks like writing proposals, entertaining my baby boy with his blocks, and so on.

So, this is Rod Ruoff's March 2007 Challenge. Maybe an iMechanica subscriber will write in that such 3-d bubble rafts have already been explored. My literature search 16 years ago did not turn anything up...2-d bubble rafts were studied on photocopy machines and "pictures" were snapped as a function of time to watch the evolution of the bubble ensemble, and so on. So, my literature searching on this is also dated.  



Zhigang Suo's picture

Intriguing ideas, Rod!  Two thoughts come to mind.  You are probably familiar with the following papers, but I'll list them here for the students who might want to follow through your suggestions.

A Nature paper (2002 vol:418, pg:307) by Subra Suresh and coworkers mentioned that they could fit the interaction between bubbles with a potential, which scales with that of atomic potential, so that they can do molecular dynamic simulations.  Thus, both a bubble raft and a molecular dynamic simulation become models of real atoms.  I wonder if someone can run a molecular simulation of your experiment:  packing in a curved space. 

Also, you might have seen the work Spaepen, Weitz and colleagues on using colloids to study packing of spheres under various constraints.  The use of colloids allow them to visualize various defects.  They call colloids their analog computer.  Here is a paper in Science that they published. 

Thanks for your comments Zhigang. I was aware of the work by Spaepen, Weitz, and colleagues, and also their reference to natural systems that have interesting packing of objects on, e.g., spherical or quasi-spherical surfaces. But I hadn't seen the article by Suresh and coworkers so will read it and perhaps contact them to see if they would like to consider the nested sphere case first in MD. While it seems reasonable to me that 32 bubbles would be like the truncated icosahedron shape of the C60 molecule (that is, that there might be 12 pentagons and 20 hexagons, with all of the pentagons isolated)...maybe they won't, and either way it is interesting in my opinion... 

I did in one manuscript refer to my own studies of using superglue to attach precision ball bearing spheres to a larger sphere, in a paper published in CPL:

 R. S. Ruoff, T. Thornton, and D. Smith, Density of fullerene-containing soot as determined by helium pycnometry. Chem. Phys. Lett., 186, 456-8 (1991).

I wanted to better understand how fullerenes of different sizes might conceivably pack together, and this led me also to some published work on the packing of spheres on spheres (referenced in the above manuscript) as well as these experiments with some glue and some precision diameter metal spheres.

Subscribe to Comments for "A new type of bubble raft--challenge for clever students"

Recent comments

More comments


Subscribe to Syndicate