Short story:
RM2 was nicely relieved by rotating IM1 in YAW, ASC REFL sensors were centered, RM2 only used 40% of DAC range, people==happy (ASC-REFL-centered-20260605.png).
ISS QPD was nicely centered, it was easy, people==happy (iss_CENTERED.png).
Then suddenly we found that the IFO REFL beam is clipped at the bottom of the IFO REFL baffle, ISS array was totally out of whack, forward-going beam from IM2 was not centered on the retroreflection check iris between IFI input baffle and IFI HWP baffle, people!=happy.
Turns out that something bad happened to IM1 about 10 minutes after ISS was centered (IM1_bad.png). Without any change in the drive, H1:SUS-IM1_M1_OSEMINF_LL jumped by negative 172 counts causing apparent positive 1570urad jump in DAMP_P and DAMP_Y. UR changed by 12 counts, which is like O(100urad) type change.
This doesn't necessarily mean that IM1 moved that much, maybe one OSEM was bumped while we're trying to put the cover on the ISS, but the thing is that this must have caused physical rotation somehow because we see the beam moved downstream of IM1.
We tried to diagnose the IM1 but found no obvious touching, transfer functions didn't look wrong, EQ stops have gaps, magnets look as if they're OK.
You don't have to read further, below is our memo to remember what was done.
Details you don't want to remember:
We've made another YAW move (of positive 300urad) for IM1, i.e. [P,Y]=[517, -687] -> [517, -387] to relieve RM2 further.
Based on calculation, also moved IM2 and IM3 to roughly align things. IM2 [P,Y]=[0, -588.7] -> [0, -19] IM3 [P,Y]=[35.3, -415] -> [35.3, -86].
Checked the IFI output baffle and noticed that it seems to have an offset in -X direction relative to DKDP baffle. The beam was well centered on DKDP baffle, but was too much in +X direction on the IFI output baffle.
Moving IM2 further, [P,Y]=[35.3, 381] to split the difference. Distance between the forward-going beam and the parking beamdump pickoff was ok.
This resulted in large IM3 Yaw step (IM3 [P, Y] = [35, 654]) to bring the beam close to the nominal position in front of PRM. Horizontally, the beam was off by 0.5+-1mm in negative Y direction. Nominal beam height at the measurement point for PRM 159.05mm. Measured 157.1+-1mm.
IM4 Yaw moved by positive 350 ([IM4 [P,Y]=[761, -205]) to set the beam position in front of PR2. Horizontally, the beam was off by 0.5+-2mm in positive Y direction. Nominal beam height at the measurement point for PR2 168.56mm, measured 169.8+-2mm.
Centered the retro-reflection check iris to the forward-going beam between IFI input and IFI HWP baffle. Aligned PRM to center back-propagation on that iris.
PRM [P, Y] = [-1185, -270].
Centered refl asc sensors. This worked really well, RM2 DAC was only using 40% of its range after centering ASC REFL sensors. Even though this was one attempt, I'm comfortable enough to say that we'll be fine.
Moved on to center ISS array. We removed the ISS array cover from the assembly to use the input aperture tool to make things easier. Large hole aperture as well as medium was used, didn't bother to use the small aperture. We used two pico mirrors (had to use hand because no driver is connected until the IOT2L is back). Iteration between two mirrors was easy and the centering was done by 22:37 UTC.
RM1 P, Y = -196, 209 (13% of DAC range)
RM2 P, Y = 609, -602 (40% of DAC range)
Happily we put the ISS array cover back on. Unknown to us, something happened to IM1 at around 22:46 UTC.
Remaining task:
solve IM1 mystery?
shoot the pic of pico mirrors to see how much range we have.
center IM4_TRANS
take aperture photo for IFI inpjt and IM4 baffle, recenter IM4 baffle if necessary.
ground check.
I post this picture showing the separate osem trends for IM1 and also the CPS Z and RX motion of the chamber. The big step in LL IM1 osem was coincidental with noise in Z and RX of HAM2 ISI so we think it was when we were leaning into the chamber to p6ut the cover on the ISS. Later on there is a step change in the Z and RX motion, I think this was when Rahul actually stepped into chamber to screw the cover screws down. I conclude that whatever happened was when we were leaning past the IM1 from outside the chamber.
I ran the rubbing scripts from this LHO alog #77383 for IM1 but wasn't sure how to interpret the results, reference time was midnight last night. The only one that looks really different is the yaw degree of freedom.
Annamaria drove some signals through each IM1 OSEM in turn to check for rubbing.
Time of test 02:00 UTC
| Response in each OSEM | UL push | UR push | LL push | LR push |
|---|---|---|---|---|
| UL | 17.7 | 0.6 | 8 | 10 |
| UR | 0.4 | 17 | 10.7 | 7 |
| LL | 6.6 | 7.5 | 15 | 1 |
| LR | 11 | 8 | 1 | 18 |
We used an actuation of 60000 counts offset added ther the COILOUT filter bank. We can't see any obvious imbalance in the response of the four osems.
Annamaria also ran an undamped TF in pitch, shown here compared to a reference measurement from 2024. This doesn't really seem to show any obvious rubbing.
Summary: As Keita said above we are confused as to what exactly is wrong with IM1.
