[Louis, Sheila, TJ, Camilla]

This is a continuation of the effort to recover PRMI.

• This morning I reset IM4, PRM, PR2, and PR3 back to their slider positions just before the x-arm gate valve was closed. 
• I aligned ITMX then ran ALIGN_IFO::PREP_FOR_MICH. Normally this would have been fine but I'd missed that Oli was in the process of setting up health check measurements on ETMX. PREP_FOR_MICH moved the ETMX suspension and moved it out of the HEALTH CHECK state, which Oli needed it in for their tests.

• Sheila edited ISC_LIBRARY.py to have set_sus_config() refrain from messing with the ETM suspensions when the flag lscparams.gate_valve_flag is True. She also updated ISC_DRMI::PREP_PRMI to set relaxed trigger thresholds for MICH when the gate valves are closed (via the same flag).

• I moved us directly to ALIGN_IFO::MICH_DARK_LOCKED and adjusted the BS. The sliders were left at P:-738, Y:338.31.

• We went to PRX and touched up the PRM. Sliders were left at P:-1181.3, Y:-451.4.

• From here we tried going to PRMI. We noticed that our triggers were very brief so I moved PRM further to maximize POPAIR. From here sliders were moved to P:-1182.3 Y:-432.4. Also while trying to acquire PRMI we noted that the beam spot on the AS AIR camera was showing a lot of motion. My initial suspicion was that it was the BS (for some reason). 

• I modified ISC_DRMI::ACQUIRE_PRMI to turn on the oplev damping on the BS. We've been doing this often when locking PRMI to help with BS motion (see 91612). The usual prescription is to turn off the oplev damping once PRMI is locked. I have not codified this in the guardian yet.

Since there was so much unexplained motion today during our attempts, I decided to take a look at the error signals and compare them to a previous PRMI lock stretch that was able to hold for much longer. I chose as reference 1471973449 (August 28 17:30 UTC) and compared it against one of our attempts today at 1472510961 (September 3 22:49 UTC). The error signals spectra for MICH and PRCL showed a lot of motion between 8 and 20 Hz. Ibrahim noted that an earthquake was rolling through around the same time. But the earthquake came through roughly an hour before this locking attempt and the ISI was back in nominal state by this time so I don't think the two are related. A comparison of the time series today and at the reference time also suggest that we may not be fully aligned onto the diodes in HAM1 and ISCT1. We might consider making sure that we're properly aligned on all the relevant diodes next week. I don't think we've taken a close look at that since pico'ing in HAM3 (91789).