Just before the winter break, Chuck, our advisor, and Frank, our engineer, came down to NASA Armstrong and physically installed our instrument rack on the DC8 ready for the second deployment of the ATom mission. Then first thing in January Agnieszka, my colleague, and I came down to test everything out and do the first test flight. Tests went well, and we performed a series of calibrations to monitor instrument performance. Things were going well and we even found the time and energy to go for some runs after work in the hangar.
The first test flight was about 4 hours, out and south over the pacific, then back overland and a little way over California’s central valley. We had the shake down flight (where the crew take the aircraft up with full instrument payload and literally shake it about a bit, to test that we’ve installed everything safely) the same day, so didn’t take off until the afternoon. This meant we got to see a beautiful sunset during the flight.
Whilst all out instruments worked well, a colleague of ours, Karl, has an instrument in our rack that, at high altitude, appeared to have a small internal leak. Sometimes problems like these don’t show up until you’re flying at high altitude. This is because many instruments operate at the same pressure as the air we’re sampling, so at low altitude they’re at a very similar pressure to the aircraft cabin, but at high altitude the instrument pressures are much below the cabin pressure. If there is a tiny leak, this big pressure difference can cause a flow of air from the cabin, which may show up as particles that shouldn’t be there, or just as funny flows – as was the case here.
Karl did what he could during the flight to diagnose the problem, but in the end it was too deep in the instrument to find in flight. At the end of the flight I helped him pull the instrument from the rack and take it into the lab for further work.
The next day, Karl stayed in the lab diagnosing the leak and Aga and I came back onto the plane to calibrate again and check nothing had shifted during flight. We found that we couldn’t pump down the instruments that operate at low pressure no matter how much the pumps pulled. Evidently we now had a leak. The obvious place to look was where Karl’s instrument had been plumbed in, but we had correctly closed off all the open ports there.
So then we started going systematically through the rack, isolating parts of the system to figure out where the leak could be. It was very hard to track down, but after a few hours we found it to be in the sheath flow of one of the driers (these take the moisture out of the air before we count the particles – it’s easier to make comparisons if all measurements are done dry). This is a very unexpected place for a leak, and not a component for which we normally carry a spare. We pulled the drier and temporarily replaced it with a simple tube. This fixed the leak and we were able to calibrate quickly before aircraft access finished for a long weekend with an external drier.
It had been a very intense day of chasing this problem down and rushing to get everything sorted before we lost access for four days. It felt unsatisfying to have to leave the rack with one instrument and one drier missing, and return to Colorado for a week, but it was the best we could do with the situation.
Back a NOAA in Boulder the next week I tested the drier. It was a very strange kind of leak, the nafion film, which allows water but not air to pass between the sample flow and sheath flow must have torn (it’s transparent so you can’t see this). The manufacturer had not seen this before, and wanted us to send it to them for structural testing. Meanwhile we ordered a new one (hopefully to be a spare if the first can be repaired), tested it, and shipped it back to Armstrong in California, where we will join it next week for one more test flight and then the start of the mission proper.
Meanwhile, Karl found his leak was also in a very strange place. The instrument must have been suddenly over-pressurized causing a lens to pop out of place and smash into a window deep inside the optics block. He was also able to fix this and ship back to Armstrong. We’re looking into what could have caused two such strange problems in close succession.