To read about F's and my London trip, start here and click "newer post" to continue the story.

Thursday, December 11, 2008

In the interest of updating my memoirs - I spent Sunday evening and all day Mon-Wed in Tampa at an ASTM meeting.

I enjoyed the meeting quite a bit. Met some people I've dealt with by phone and email only - funny how, today, when I looked at emails from them their faces popped into my head. There's a lot to learn about how these standard-setting organizations are set up, and some interesting group dynamics. I had to speak just a bit b/c of the work group I chair. Everyone was very nice. Driving back and forth wasn't very nice - I live close enough to Tampa to day-trip but it meant leaving early enough to miss some of the hospitality things where the instrument vendors show their products. The next meeting, in June, is in Norfolk, VA. I can't day-trip there.

I ran into the instrument vendor who recommended me for this job. My sulfur analyzer had just started to act up when this meeting started - I discovered Sunday morning that it had lost its mind and didn't have time to fool with it. The vendor and his boss told me what to do for it, and also how to optimize it for my application, so that was worthwhile right there. This came after a technical and strangely satisfying discussion about detection limits. Perhaps you didn't know, reader(s), that when you analyze a material for a particular thing, and you don't find any, you can't report the result as zero. You have to report it as less than something, that something being your detection limit, and settling on what that detection limit is, is surprisingly complicated. For instance, you can analyze a standard that is two to five times as concentrated as what you think your detection limit is, seven times, and then multiply the standard deviation of your responses by pi. Or you can look at your signal-to-noise ratio when you run a solvent blank - but how do you know your solvent is really blank? One way to verify that is to use that solvent to make a set of standard solutions at various concentrations, plot concentration against response, and see how close your best-fit line is to going through the origin - in other words, how close your y-intercept is to zero.

I had lunch with various people every day, too. We mixed technical discussions with talking about our families, politics, a whole lot of other stuff. Yes, I had a very nice time.

I think I'm going to see if I can kick my local professional organization into a higher state of activity. I need this kind of thing.

No comments: