It's been a while, mainly because things have been going crazy at school and work. Despite working long hours (I am a grad student, after all) I have been able to devote some more time to my pet project at Pegasus -- the Plasma Control System.
Essentially, what we would like to do is to enable real-time detection and correction of plasma physics parameters such as plasma current, plasma position, shape, etc. By real-time, I mean "fast on a timescale relative to the discharge length." In Pegasus, for instance, we can only keep plasmas around for a few thousandths of a second; my first ballpark correction rate will be on the order of tens of millionths of a second. I'll be doing this with a state-of-the-art acquisition and control unit coupled with software based on that used on DIII-D, one of the major research facilities in fusion research.
My job has been to assemble the hardware, get the software components up and running, and then somehow turn this PCS into a working tool for Pegasus!
This summer, I was able to get the hardware up and running, which in and of itself involves a bit of geek factor -- an embedded Linux system (the digitizer), linked as a virtual PCI card in another Linux system, itself controlled by -- wait for it -- another Linux system! The software is written in bits and pieces of four programming languages I've found so far (C, IDL, stitched together with serious bash and csh scripting). On top of that, it interfaces with an independent data acquisition, storage, and retrieval solution, MDSplus, the de facto standard of the fusion community. (It works well, but is very poorly documented.) After I get that working, I just need to patch it into our existing bunch of independent LabView codes! (A programmers nightmare/delight, depending on your point of view.)
Recently I was able to get an MDSplus system up and running, so all the pieces are now on the proverbial table. Now I "only" need to get them to all talk to each other correctly, when not doing real plasma physics or finding vacuum leaks. :)
Thanksgiving was good -- lots of food, friends, and family. I'm looking forward to Christmas, however, since by then I'll be done with this semester! Since I've passed my qualifying exam, I only need to wrap up my courses before I start serious, in-depth research on my thesis, which will likely be coming from using the completed PCS system to study physics issues on Pegasus. The upside: only a year to go, entailing much lighter semesters than I've been pulling in the past while at UW-Madison.
No comments:
Post a Comment