Particle counts taken during the installation of the BSC1 door.

Day's Activities

• early morning:  Begin Y-beam Manifold Spool install
• 8:37-10:30:  At EX, rotate SEI seismometer & move PEM Seismometer (Jeff K)
• 8:38:  Y1 Beamtube accelerometer work this morning (Paul/Jordan)
• 9:18:  Viewport survey begins (Gerardo)
• 9:25:  Students taking stroll to MY (temp 86degF)
• 9:49:  opening GV14, and pressure should return to normal within a few hours (Kyle)
• 10ish-lunch:  Starting HAM5 incursion
• 10:10:  Clean up work (Travis)
• 10:20:  HAM5 ISI check (Hugh)
• 10:28:  Nathan heading into optics lab.  Plans to not run overnight measurements during weekend
• 11:30:  Betsy heading out to BSCs
• lunchtime:  Started viewport work at X-Beam Manifold Spool (Gerardo, Doug, Jason)
• 12:10:  BSC1 work (Arnaud)
• 1:20:  ring heater pin outs (greg, filiberto)
• 1:25:  EX accelerometer/microphone work (Paul/Jordan)
• 3:00:  Landry tour
• 3:40:  recompile  etmx isi & also start/restart (Jeff K)
• 3:45:  Tour for Rick & UTB crew
• 3:50:  DAQ restart (Jeff K)

A clean set of top-to-top TFs were previously taken for the ETMX (QUAD) suspension (see LHO aLOG entry 13134), to complete Phase 3a power spectra have been taken with damping loops both ON and OFF for all stages.

These power spectra measurements have been compared to previous in-vacuum measurements (allquads_2014-08-01_Phase3a_H1ETMX_ALL_Spectra_D*.pdf). The plot key is as follows:-

Black Dashed Line = Expected Sensor Noise
Blue Solid Line = H1SUSETMX 2014−01−28_1600, Phase 3b (in-vacuum)
Green Solid Line = H1SUSETMX 2014−08−01_1400, Phase 3a (in-air)

Summary:

Noise floors for ETMX are consistent with previous measurements. However, the LF and RT channels see features in the V DOF for both chains at ~7.0 Hz, ~15.5 Hz and 26.94 Hz, which were not present in previous measurements. Such features were also seen in recent TMSX power spectra too (see LHO aLOG entry 13151), and could be environmental, due to purge air pumps, clean rooms etc. and should be further investigated to ensure it's not baffle related.

It should also be noted, that the ETMX L2 UL channel exhibits excess sensor noise, as outlined in a recent aLOG entry (see LHO aLOG entry 13065). 

All data, scripts and plots have been committed to the sus svn as of this entry.

Particle counts taken during the removal and reinstallation of the north HAM5 door.

Particle counts taken during the installation of the Y-Arm Spool

BSC2

After respraying the BS optic with FC yesterday, we successfully pulled the FC sheets from both HR and AR surfaces of the the BS optic.  After N2 blowing and some spot cleaning, we estimate particulate concentrations of 5-10 particles per square inch.  We then reinstalled the elliptical baffles which had been removed for FC application, unlocked the SUS, and Jim unlocked the ISI.  Quick TFs on the SUS were ran and approved, witness plate situated, and we moved to BSC1. 

BSC1

The FC sheet removal from the external surfaces of ITMy went without issue.  After N2 blowing, we estimate particulate concentrations of 5-10 particles per square inch on the external surfaces of both optics, and a factor of 2-3 higher on the internal surfaces.  We unlocked both chains, Jim unlocked the ISI, quick TFs on the SUS ran and approved, witness plate and 1" optic installed, Betsy wiped her way out of chamber, and we handed it over to Apollo for door installation.

A clean set of TFs were previously taken for TMSX (TMTS) suspensions (see LHO aLOG entry 13140), to complete Phase 3a power spectra have been taken with damping loops both ON and OFF for all stages.

These power spectra measurements have been compared to previous in-vacuum measurements (alltmtss_2014-08-01_Phase3a_H1TMSX_ALL_Spectra_D*.pdf). The plot key is as follows:-

Black Dashed Line = Expected Sensor Noise
Blue Solid Line = H1SUSTMSX 2014−05−09_1800, Phase 3b (in-vacuum)
Green Solid Line = H1SUSTMSX 2014−08−01_1100, Phase 3a (in-air)

Summary:

Noise floors for TMSX are consistent with previous measurements. However, it should be noted that both LF and RT channels see features in the V DOF at ~7.0 Hz, ~15.5 Hz and 26.94 Hz, which were not present in previous measurements. These could be environmental, due to purge air pumps, and clean rooms etc. and should be further investigated to ensure it's not baffle related.  

All data, scripts and plots have been committed to the sus svn as of this entry.

[Gary T, Danny S, Matt H, Stuart A, Norna R]

During the close-out work of the BSC9 chamber (see LHO aLOG entry 13085) we were able to further investigate the source of the excess noise that had previously been observed on the ETMX (QUAD) L2 (PUM) UL channel (see LHO aLOG entry 13047).

Previously, I had swapped field cables at the Satellite Box between ETMX L1 and L2 stages running to the chamber. This indicated that the noise was downstream, i.e. with the field cables themselves or in-chamber. We repeated this test, but this time swapping field cables at the air side of the vacuum flange. Further measurements then eliminated the field cables and confirmed the issue resided in-chamber.

Gary inspected and re-seated the in-chamber connectors from the vacuum flange to the OSEM, but this offered no improvement. Finally, the L2 UL AOSEM (s/n #270) was removed and a spare unit connected up in it's place, but again, unfortunately this offered no improvement.

This indicates that the excess noise is most likely a in-vacuum harness problem, which would be invasive to attempt to replace. Therefore, it was decided to generate an integration issue and proceed with the close-out of the chamber, with the excess sensor noise on this channel.

Gerardo, JeffB, Mike, Hugh, Travis, Appollo, Kiwamu,

We have removed two irises from the output faraday isolator (OFI). They were originally meant for a temporary tool for aligning the OFI and we simply forgot to remove them. So we opened up the HAM5 north door and locked the north side of the ISI. Then we took out the two irises -- one at the input side and the other at the output side. Note that we left the irises' holders attached. Also, Gerardo replaced the holder at the output side with a clean one because the one we had was not class-A compatible.

After unlocking the ISI, Hugh checked the balance of the ISI and he did not see a significant change. So we did not have to put a counter weight as expected.

We revisited the HMA5 installation exit check list, performed those activities and found nothing special. So it is good.

Some pictures will be uploaded later.

J. Kissel

I've aligned the GND T240 to the global IFO coordinate system (see LHO aLOG 12930 for discovery of mis-alignment). Further, in preparation for the UWash tiltmeter install, I've returned the PEM Guralp floor seismometer to inside the stubs of the PEM stay-clear zone.

--------
Details

Alignment / Leveling:
PEM's Guralp - The instrument doesn't have any holes for an alignment rod, so I had to align the instrument by eye. I've pointed the "North" direction (alignment nub near the the bottom of the instrument) in the IFO's +X direction (as it was already, in its previous location). I've also re-leveled the instrument, and re-covered it in its thermal shield. See first three pages of attached pictures. 
GND T240 - I used a brand new alignment rod "swizzle stick" (thanks Hugh!) to align the "North" direction in the IFO's +Y direction. See pgs 6-8 of attached pictures. (Hugh gave me a borehole micrometer with rounded tips to better measure the inner-diameter of the T240 alignment rod hole. It measures 0.315 [in] = 8.001 [mm]; consistent with what had been previously measured using flat-edged calipers; see pg 4-5 for pictures.)

Quantitatively confirming that the instruments are co-aligned with each other, and aligned well with the global IFO coordinates is difficult, at best, since there's no good alignment reference in this area. In attempts to quantify it, I've attached the coherence between each of the ground instruments and the T240s on the ISI in chamber. At the time, the platform had ST1 isolated and ST2 watchdog tripped (and I didn't check the status of HPI, but presumably if ST1 was isolated, HPI was robust isolated). That the platform was isolated sort of ruins the analysis, but it's something. Will ask the seismic team for advice.

Centering:
After moving/rotating the T240, the output signals remained saturated for ~5-10 minutes; not un-expected -- I've had the same experience with STS2s I've commissioned. Anxious to see if the instrument was saturated or rail/broken because of the move, I centered the instrument using the "Auto Z" button on the 4th chassis down (U18 of SEI-C1 rack -- see pg 9-11 of pictures), and the instrument went to zero within a second of the button push, but then veered off into lala land again shortly after. Waiting a few more minutes, and the signals came back into the ADC range. It still did not settle to zero mean within another 5 minutes, so I zeroed the instrument once more. Time series behavior was the same, so it just takes many many minutes to really settle. Again, no surprises here.

Calibration:
Sensor              Calibration
ISI T240s     1e-9      [(m/s) / (nm/s)]
GND T240s     1e-9      [(m/s) / (nm/s)]
PEM Guralp    7.6e-8    [(m/s) / ct]
pem.ligo.org suggest the calibration is "0.0076 [(um/s) / ct]," or 7.6e-9 [(m/s) / ct]. However, plotting this against the well-known (by my at least) calibrations of the T240s, the quoted PEM calibration appears low by a factor of 10. Hence, my calibration above. 

ITMy

• X axis position: -200.8 mm*
  • Target: -200.0 mm
  • Error: -0.8 mm
  • Tolerance: ±1.0 mm
• Y axis position: 4982.7 mm
  • Target: 4983.1 mm
  • Error: -0.4 mm
  • Tolerance: ±3.0 mm
• Z axis position: -79.2 mm
  • Target: -79.8 mm
  • Error: +0.6 mm
  • Tolerance: ±1.0 mm
• Pitch: 366 µrad down**
  • Target: 340 µrad down (352 µrad down FC correction - 13 µrad up target pitch)
  • Error: 26 µrad down
  • Tolerance: ±50 µrad
• Yaw: 17 µrad CW
  • Target: 0 µrad
  • Error: 17 µrad CW
  • Tolerance: ±50 µrad

CPy

• Pitch: 833 µrad down***
  • Target: 1.296 mrad down (to match the ITMy AR surface assuming zero ITMy pitch error; the ITMy has a vertical wedge of 0.075°, or 1.309 mrad down)
  • Error: 463 µrad up
  • Tolerance: ±1.4 mrad
• Yaw: 31 µrad CCW
  • Target: 0 µrad (to also match the ITMy AR surface)
  • Error: 31 µrad CCW
  • Tolerance: ±1.4 mrad

*** This is the actual pitch of the CPy.  The listed error is relative to where the CPy needs to be to be parallel with the ITMy AR surface, assuming the ITMy has zero pitch error; this does not include any FC correction done on the ITMy.  Since we cannot verify the ITMy pitch after FC removal, we also cannot verify the CPy pitch relative to the ITMy AR surface after FC removal.  We have to assume that E1200791 is correct (we have only been able to test this one time at the ETMy, which moved upward by 442 µrad, but one data point does not a trend make) and that the ITMy will pitch up by ~352 µrad upon FC removal.  Therefore our expected relative pitch between ITMy and CPy upon FC removal is 490 µrad up (14 µrad down ITMy HR pitch + 1309 µrad down wedge - 833 µrad CPy pitch = 490 µrad up relative pitch).  As stated above we can weigh the FC to get some idea of the pitch change.

ACB (relative to the ITMy)

• Lateral position error: -0.3 mm
  • Tolerance: ±2.0 mm
• Vertical position error: +1.2 mm
  • Tolerance: ±2.0 mm

This completes IAS alignment of ITMy; our equipment has been removed from the Y manifold spool area to allow for the Cryopump Baffle installation.  This also ends the IAS alignment of the aLIGO H1 interferometer.

Summary: Eight years of wind data at 3 LHO stations were analyzed. Fifteen percent of hours had wind speeds exceeding 10 m/s, the speed at which displacement and tilt from wind start to significantly increase the seismometer signals. Nearly 25% of hours in April, the month with the highest average wind speed, have winds exceeding 10 m/s. Data are also shown for individual stations; readings for EY are the highest.

Introduction: The analysis of wind data is important because wind speeds over 10 m/s increase ground motion and thus affect interferometer performance. This increase in ground motion is caused by wind interacting with the topology of the site and, most importantly, the surfaces of buildings. In addition to producing displacements, wind blowing on the buildings can tilt them, which can produce spurious acceleration signals from seismometers.

Methods: Eight years of data between 2004 and 2012 were analyzed for patterns across the years. The data was extracted from DataViewer for the Corner Station (CS), End Station X (EX), and End Station Y (EY). The "Maximum Channel" was selected of the "Hourly Trend" setting within DataViewer. This time period was selected because it was not missing more than 58 days of data at a single time in the series and only a total of 218.7 days were missing from the entire data source. This amounted to about 8% of the data missing from the entire data set. The data was considered missing if DataViewer did not provide it.

Results: The first figure shows that 15% of the hours in the 8 years analyzed exceeded 10 m/s, which significantly increases ground motion. The second figure shows that, during each of the spring months, the wind was greater than 10 m/s in more than 15% of the hours, peaking at 24% in April. The third figure shows that the average hourly maximum wind speed varied the most from year-to-year in the month of February. The fourth figure includes statistics for the individual stations. Averages for EY are higher than for other stations. For example, 27% of April hours exceed 10 m/s at EY while the average for all stations was 24%.

Margarita Vidrio, Robert Schofield

Calum, Rich, Matt

HAM4

The soft cover on the South door was not billowing at all. Despite the obviously low purge air and Matt working in HAM4, the initial particle counts were low. Before starting work, I wiped down everything I could reach with a pre-wetted wipe. Calum and I inspected the optic with a green flashlight (not to be confused with the Green Lantern, D1400060). For both faces of SR2, we estimated 25-100 particles per square inch depending on the region. BEFORE applying FC, we found a ~1 inch piece of leftover FC from the alignment layer on the HR side, and what appears to be a long strand of FC on an earthquake stop. There were numerous glove prints and leftover FC on both faces near the barrel. Because the HSTS legs are warped from welding, we had to use 2 washers on a side of the spray cone bracket to keep the cone on its intended track. Not surprisingly, the particle counts spiked while spraying FC. The "CAL Alarm" popped up on the count taken while spraying. I need to look up the settings for the handheld particle counters to check what that means. The FC appeared much thinner and had more bubbles than usual, but there were less cobwebs than we observed on Saturday. We moved on to HAM5 to let the FC dry.

Matt returned later to retract then remove the spray cone. The wipe on the end of the cone stuck against the optic when the cone was retracted. This has never happened before at either site, not even when a cone was left on overnight. I used the same procedure (E1300017) as always. Matt pulled very gently on the wipe, and it separated from the optic. I painted 3 layers of FC around the edges and added a PEEK tab. The painted crust was much darker in color than the sprayed FC.

HAM5

Like HAM4, there appeared to be very little purge air. Spraying SR2 with FC only a few minutes earlier likely caused the high initial particle count in HAM5. SR3's HR side had 10-30 particles per square inch and a big streak of FC from the alignment layer plus small bits of FC near the barrel. SR3 was sprayed on Saturday. A modified mix with slightly more thinner was used for the 1st layer in an attempt to reduce cobwebbing. It worked, but there was more backsplash (contained by the cone). The standard 1:1 mix ratio of FC and thinner was used for the 2nd and 3rd spray layers.

On Sunday, we removed the cone and painted the crust on SR3. The dried FC spray looked uneven and runny.

BSC2

The beamsplitter was absolutely filthy, and had over 50-100+ particles per square inch. There were tiny circles all over the HR side that looked like rings left by evaporation. There were noticeably more fibers on the AR side. Both faces had glove prints and lefover FC near the barrel. I didn't get any good pictures unfortunately.

I sprayed both faces on Friday, and painted the crusts on Sunday. Like SR3, the FC on the BS was extremely runny.

I'll add photos soon.