Well, it was too good to be true. Work on the cryostat had been progressing fairly steadily and without too many mishaps. Today we had planned on closing the cryostat and starting to pump on it in preparation for cooling it. In the morning, we envisioned at about 10AM that it would take a few hours to get everything closed and we'd take an early night. I started writing this post at nearly 1AM and we're still here.
The first issue that came up was relatively simple. We had issues getting the instrument in properly due to some concentric baffling that has to nest together, but any angle between them screws the whole process up. Somehow, when we did this last time, it all just worked on the first try. This time, we had many fits and starts as the angles weren't quite right, so we had to tweak the angle with our Science Brick (a.k.a. a 26-lb tape-wrapped lead brick) and try multiple times.
Eventually we got it in...and we realized we had forgotten to plug in one of the housekeeping connectors, which required removing the instrument from the cryostat and plugging it in. After fixing that, Ilan checked wires and found that some of our fridge heater wires had opened up, meaning we wouldn't be able to cool our optics to 1K -- an unacceptable situation. So we spent the next 2 hours fixing wires.
Once we fixed all of the wiring issues, all that was left was to put the various shell lids on, seal the last o-ring lid, and pump, so we blasted some heavy metal in the high bay and got to work.
It's now 2:30 AM and we just started the vacuum pump. We pump through a restriction at first so as not to damage the delicate thermal filters via differential pressure, but eventually we will remove the restriction and let the cryostat pump directly into the pump so it will pump faster.
All in all, the problems we faced today weren't actually any worse than ones we've faced before -- it's just that much more frustrating when the end is in sight.
I'm not putting any pictures up in the post because I'm feeling a little lazy. You'll just have to click through to the pictures: