Monday, August 22, 2011

What I Actually Did

I leave Santiago Wednesday night, so these are my last few days of working. Mostly it'd wrapping stuff up and polishing, but I got that done two weeks ago. Nowadays it's busywork. I figure that it's high time to explain what I actually did all summer.

Take a metal. All of the atoms and stuff in the metal are arranged in a neat periodic lattice. Now squish atoms together, twist them around, and generally screw up the orderliness of the lattice. This is called a defect. A dislocation, to be more precise. Defects play a major role in the properties of materials. It pays to know many there are the ratios of the various types. Until recently the only ways to do this were 'guess' and 'zap it with x-rays'. My professor has come up with a theoretical method of finding defects with ultrasound. Ostensibly my job was to confirm or reject this theory

In order to confirm it we have to compare it to a method with known validity. Before I came in the professors decided 'guessing' was too vague and that crystallography was the way to go. The idea is that x-rays scattered off a crystal would form bright, intense peaks. Defects would cause the peaks to be a little dimmer and a little wider. This is called broadening. Measure the amount it broadens and you're done. Simple, right?

Not really. Due to QM effects it's impossible to get a single, nice peak without a particle accelerator. You can only get double peaks, where the two overlap so much it's impossible to tell where one ends and the other begins. My professor had no idea how to separate them and brought me on for the sole purpose of doing it. In MATLAB. Which I didn't know.

It took me six weeks of teaching myself MATLAB and running through various dead ends, but eventually I figured out how to pull it off. I still finished it ahead of schedule. It's a bit of a hollow victory, though, since I feel that if I knew MATLAB from the start I could have done it in two. Ah well. This is why I'm an undergrad and not a professor.

Overall I think it was worth it. I was able to contribute significantly to the project, got things done faster than expected, and impressed my professor with what I pulled off. They haven't compared my results to the theory yet, or if they have they have yet to tell me. If it succeeds, hooray! If not, at least I learned MATLAB. It's a pretty great language.

Since then I've been writing documentation and puttering around with a simulation of a triangular lattice. Two more days and I'm done. Then comes Patagonia. The capstone of the summer.

1 comment: