The Naval Research Lab

I worked at the U.S. Naval Research Laboratory for four summers during high school and college. I spent much of my time writing a computer program to do thermal analysis for an experiment that we put on the space shuttle. The facility I developed with the finite-element method came in handy in my job at Bell Labs the following summers.

I was working for C. H. Tsao and Jim Adams in the Laboratory for Cosmic Ray Physics. We were estimating the distribution of isotopes in cosmic rays using a pile of track detectors. To get accurate measurements, you want these plastic disks to be as close as possible to a constant temperature, so we designed an elaborate wrapping of thermal blankets. My program computed the temperature of the detectors during the year that the Long Duration Exposure Facility (including our experiment and a bunch of others) was scheduled to be in orbit. The input is the heat from solar radiation (easy enough to compute given the trajectory). On the computer I tried a bunch of configurations of blankets—it turns out that it’s possible to wrap the detectors too much—and we finally decided on a design, which the people in the lab implemented. Later on my bosses gave me the the opportunity to go to the space shuttle launch in Florida in October. But I was such a loser, I didn’t go, being too lazy to take a trip in the middle of the semester (I’d just started college).

After all that effort, I don’t think the detectors ever got analyzed. We heard that the shuttle launch a year later that was scheduled to bring the experiments back to earth was pre-empted for a military mission. Then the next shuttle launch blew up. Ultimately, I was told, they used the LDEF as target practice for the Star Wars antimissile system. At least that’s what I recall being told, but I suspected that story was too good to be true; a quick Wikipedia lookup reveals that LDEF actually was eventually retrieved, several years behind schedule. I don’t know if our plastic discs were ever etched and marked for cosmic ray events, but my guess is that that the data wouldn’t have been good for much. Really I was never clear on the purpose of any of this research, except for the general sense that cosmic rays come from exploding supernovas and they give unique information regarding the composition of the universe.

Many of my friends had summer research jobs too—there are a lot of government labs in the D.C. area. The NRL job was cool because they paid $1000. On the downside, it was a long commute—about 25 miles by car or 2 bus rides and 2 metro rides. One morning I got all the way to the lab and was driving into the entrance, when I realized I’d been asleep the entire time!

NRL is in Southwest D.C., near St. Elizabeth’s, Bolling Air Force Base, and the Blue Plains sewage treatment plant. When the wind is coming in from Blue Plains on a 90-degree summer day, you don’t want to spend much time outside.

In addition to C. H. and Jim, the cosmic ray physics lab featured a couple of old guys with Ph.D.’s, a couple of lab technicians who were really nice (one of them gave me pamphlets for an organization called the War Resisters League), and a postdoc who was pretty mean. He stepped on my glasses once! I wanted to think it was an accident but it seemed like he did it on purpose. Maybe it freaked him out, as a thirty-year-old with a Ph.D., to be working alongside a teenager who was doing essentially the same sort of work he was doing. All that physics training probably didn’t prepare him to write silly little computer programs and research articles.

We also had a couple of students—maybe they were grad students?—from the Naval Academy, who were studying physics and working on research projects during the summer. These guys were cool. One of them told us a story about a friend who had a girlfriend he was trying to get rid of, she kept bugging him etc., so he bought a custom rubber stamp and sent back one of her letters with the marking, “Return to Sender. Midshipman deceased.” Also there were a couple other summer students. We must have been working on similar projects and we were sitting in the same lab but we never ever talked about work. Usually we just sat and worked, we’d go to lunch, then at the end of the day we’d sometimes relax by having Freez-It fights. (Freez-it was a spray can of cool gas that got used in the labs, it would be sitting near the solder, the Kimwipes, etc.) There were also a cute-looking girl summer student in another lab in our building, but I’d only ever see her from the distance, and I was too much of a loser to actually go up to her or anything.

What else? As I was saying I did lots of programming. We had the ASC, the Dec-10, and one other computer I don’t remember what it was. All of them were mainframes or, at the very least, minicomputers. The ASC (short for Texas Instruments Advanced Scientific Computer) was just horrible. You had to program in something called Job Control Language and go over to another building to get the results. The Dec was more pleasant, we programmed in something Unix-like. At one point I spent a few days writing a Trojan, which the postdoc then sicced on our boss. Of course we really had nothing interesting to do with the guy’s password once we actually had it. It was the planning that was fun.

Also . . . one of the technicians told me they’d lost some of the data when roaches had eaten some of the paper tape. I saw the paper tape, it was sitting in piles in the back of the lab! Somehow I’m guessing that the fate of the free world doesn’t hinge on those numbers.

And . . . when I took the job, I had to fill out a personnel form which actually included a question on whether I had ever been a member of the Communist Party!

This all came to mind when the following came in over the Bayes email list:

Description: Applications are invited for a NRC Research Associateship
(postdoctoral research position) from those with interest and skills in
Bayesian probabilistic methods and dynamical systems. . . .

The principal investigator, Dr. Liam Healy, is building on extensive research and experience in orbit dynamics, orbit determination, space
navigation and guidance, algorithm development and scientific programming. Related areas of research in the control systems branch at NRL include relative motion dynamics, robotics, machine learning, and electrodynamic
propulsion.

About the Naval Research Laboratory: The Naval Research Laboratory is the corporate laboratory of the United States Navy and Marine Corps. It has a staff of more than 2500 scientists, engineers, and support personnel. . . .

To apply: Please see
http://nrc58.nas.edu/RAPLab10/Opportunity/Opportunity.aspx?LabCoded&ROPCDd1586&RONum·705.
You should contact the adviser before applying to understand better how your work and this research effort match.

P.S. E. T. Jaynes worked there too!

P.P.S. I sent Jim Adams an email last year telling him what I’d been working on, and in reply he gave me an update on his research (he’s now at NASA):

I [Jim] am working on some of the same things we worked on and some others. I am currently working on a probabilistic model for solar particle events. The idea is to develop a model for providing a tailored worst-case environment instead of just an extreme worst case as we did in Gelman and Adams (1984). The idea is that the user will specify the launch date, mission duration and confidence level and we will provide a worst-case that should not be exceeded during the specified mission at the specified confidence level.

Besides this I am investigating cosmic ray modulation in the outer heliosphere and I am the PI of a proposal for US participation in the Extreme Universe Space Observatory, a major experiment planned for the International Space Station (see http://jemeuso.riken.jp/en/).

I think you may remember our LDEF experiment and our TRIS experiment, both flown on the Shuttle. Since TRIS was re-flown, we flew and experiment to the Moon on Clementine, a series of flights on the Russian RESURS-1 satellites and an experiment on the Brazilian SACI-1 satellite, all before 1999. Since then it’s been balloon flights in Antarctica, northern Canada and Alaska. My team has another flight in Antarctica coming up this winter.

Also, it turns out that maybe my summer work was not such a waste of time! Jim writes:

You modeled the thermal blankets on LDEF. TRIS was a Get-Away Special experiment. It flew attached in the cargo bay while LDEF was released from the Shuttle, then retrieved on another Shuttle flight in 1991 and returned to Earth. From that experiment we learned the ionic charge state of solar energetic particles. It was quite low compared to expectations at the time, which were based on the assumption that solar energetic particles came from solar flares. Our results were consistent with a new idea at the time, that solar energetic particles were being accelerated by coronal mass ejections. We know that solar energetic particles are accelerated both ways, but the big solar particle events caused by coronal mass ejections. The LDEF results are regarded as one of the most convincing proofs that coronal mass ejections accelerate solar energetic particles.

6 thoughts on “The Naval Research Lab

  1. 1. Ah, yes, all the pretty girls we saw from a distance while we were writing code, but were too shy to approach. (sigh)

    2. The only worse storage medium than paper tape in rolls was paper tape that folded.

  2. Ah, the ASC. I never programmed it, but I once worked at the site where they were made and where they had one or two production ASCs running. I recall one technician in our area who said he had once worked on the ASC. When he started out, he supposed dropped some hardware (washers?) into the bottom of one of the racks. As he started to fish them out, the other, more experienced ASC tech said, “No, just wait.” Then he turned on the power supply (which was what occupied the bottom of the rack), there was a momentary bzz-bzz, and then silence: problem solved!

    Those things drew a /lot/ of power.

  3. I’m sorry that you missed a chance to see a shuttle launch. The one I saw (launch of the Hubble Telescope) was very impressive. Unfortunately, my wife and kids missed it because although we were all there for the “official” launch, it was scrubbed at the last minute. The kids couldn’t get a second dispensation from school for the actual launch. But I was there to talk to the press about the Hubble in my capacity as one of the Hubble science team leaders.

  4. Cosmic rays! This is not my background (I wish!), but my father turned his engineering degree to account by working as a staff member on Ray Davies’s cosmic ray project in the 1980s – it was a glamorous tale of my childhood that my dad was miles underground in a South Dakotan goldmine looking for neutrinos in tanks of dry-cleaning fluid…

  5. Pingback: Defining ourselves arbitrarily « Statistical Modeling, Causal Inference, and Social Science

Comments are closed.