Good news is, we were able to see the reflection of LSC POP_A sensor and adjusted the beam dump.
Bad news is, I ended up finding many, many ghost beams in REFL path. We started installing beam dumps but the only remaining was the ghost beam from the first 90:10 hitting the inside barrel of the first 50:50.
We ran out of beam dumps for the day so we need to continue tomorrow.
This means that we need to stay in laser hazard in the morning. Fil also need to do more ground check in the morning. Hopefully the door team could start in the afternoon, but we need to talk tomorrow.
Here are some pictures of the beam dumps we installed last night to dump ghost beams originating from the beam splitters on the REFL path. This is the status as of Tuesday night, Keita and Ryan went in this morning to finish up this work.
Some of the earlier pictures are out of date now since we took the beam dump we installed near REFLAIR and used it to dump the original ghost beam upstream, removing the need for the REFLAIR beam dump.
Continued the work on Wednesday morning with Ryan. See 1st attachment, that's cartoon showing which additional beam dumps were installed. Circled in green are for ghost beams in REFL path.
It took some time but we figured out that, not surprisingly, the strongest ghost beam was the AR reflection of the 1st 90:10 (M14) reflected by the 90:10 itself, going into the REFL path direction. On the 1st 50:50 (M17), that ghost beam was half reflected and hit the back of WFSB and half transmtted and hit the inner barrel of M17 and came out as a funny beam, going to the next 50:50 and reflected. I don't have any "before" picture but Georgia might.
We inserted a beam dump with only one plate ("A"), then installed a V-shaped one ("B") to receive the reflection by A. 2nd picture shows the ghost beam and the main beam on a viewer card in front of the "A" beam dump. The ghost beam is not clipped by the beam dump edge but is very close, but I wanted to make sure that the main beam is not clipped. The 3rd picture shows the ghost beam caugut by the beam dump. Here the translucent edge of the card is aligned with the edge of the glass plate, so there is some room before it's clipped. 4th picture shows the V-shaped beam dump at "B" position.
There were also two ghost beams we couldn't identify exactly, but we installed two beam dumps ("C" and "D"). See 5th and 6th picture. It might have been possible to just use one for both beams but the beams were separated enough that we couldn't comfortably do it at "D" location, and "C" location was not really ideal for precise positioning of the beam dumps due to the fact that it's close to the center of the table and we had to reach deep into the chamber.
BTW, in the 5th/6th pic you can also see that the 90% beam on the high power beam dump is super bright and scatter very strongly. Yesterday Georgia and I were worried if it was hitting the metal part, but after moving the beam down by a few mm on the beam dump (there wasn't any change in the brightness) we concluded that the beam was/is hitting the SiC plate and it's the nature of the plate.
The rest of the pictures show the newly installed beam dumps (as well as the old ones).
All in all, in this incursion we installed 8 new beam dumps. Only one of them (the one to receive the transmission of M2 corner mirror which is now 50:50) was planned to be installed. One was installed as we were able to see the ghost beam of M2. Four were installed to catch ghost beams that comes from upstread of M2. Two were installed because we saw two ghost beams from the main septum viewport for the PSL beam, which landed on and were reflected by the newly installed nozzle baffle (alog 63587), and we wanted to properly damp these reflections on the table.