PSL – The leaking coolant manifold was replaced yesterday. Coolant has been circulating since then with no apparent problems. Today efforts will concentrate on getting the PMC relocked. 

SEI – The new hardware has been installed and tested. The crew is closing out some final cleanup tasks; then will run a final set of transfer functions. If everything checks out OK, the plan is to hang the last two doors after lunch.

Vac – Pumping will start on HAM6 as soon as the doors are back on. Gate Valves should be opened on Monday. 

FRS – Reviewed and closed out completed FRS tickets.   

The output of the 4 flow sensors - one per laser head - from the past 24 hours is attached.
The sensor for head 1 was replaced yesterday.  Not obvious to me why it did not settle down
as rapidly as the other 3 sensors.  But we will keep an eye on it.

The crystal chiller water level was lower this morning than what it was yesterday afternoon.
This might be due to the air bubbles working its way out of the plumbing.  We will continue
to monitor for leaks.

J. Kissel

As per this April HAM6 vent's plan, E1600092, I've checked that the SUS have not been affected by this week's vent activities by measuring all DOFs of all SUS in the chamber -- the OMC and OMs 1 through 3. Attached are the results of the transfer functions compared against the last in-vacuum results. All SUS are A-OK; SUS is a go for chamber close out.

New data sets live here:
/ligo/svncommon/SusSVN/sus/trunk/OMCS/H1/OMC/SAGM1/Data/2016-04-08*.xml
/ligo/svncommon/SusSVN/sus/trunk/HTTS/H1/OM1/SAGM1/Data/2016-04-08*.xml
/ligo/svncommon/SusSVN/sus/trunk/HTTS/H1/OM2/SAGM1/Data/2016-04-08*.xml
/ligo/svncommon/SusSVN/sus/trunk/HTTS/H1/OM3/SAGM1/Data/2016-04-08*.xml
Note, I've improved the OM templates by tuning the drive and adding a boost to the low-frequency drive. The OMC's template didn't need it.

J. Kissel, J. Warner

Based on the results of B&K hammering the H1 ISI HAM6 blades before and after installing the new blade tip and flexure dampers, I've retuned and re-installed the pre-existing tuned-mass dampers. Lots of details, results, discussion and techniques below. As Jim mentions in LHO aLOG 26482, this concludes the installation of the HAM-ISI damping mechanisms for H1 ISI HAM6. 

Thanks to all who've have helped!

We will close up the chamber tomorrow as planned.

Details
----------------------
Blade Characterization

Over the past few days, as we've been installing the new Blade Tip Dampers (D1500469)
 and Flexure Dampers (D1500200) -- see LHO aLOGs 26482, 26473, 26456, and 26452 -- we've been gathering before-and-after B&K transfer function of the blade-spring & flexure system without the eLIGO tuned-massed dampers.

The results of the this characterization are shown in the attachment 2016-04-07_H1ISIHAM6_DampingInstall_BandKResults_BladeSpring_UndampedvsDamped.pdf. (Recall, for your convenience, which blade is which from the attachment HAM6ISI_Sensor_and_Blade_Names.pdf).

As expected, all three blades' first bending mode has moved down from ~153 Hz to ~139 Hz due to the extra mass of the blade-tip damper. Note, sadly, only Corner 3's "after" measurement has any coherence above ~300 Hz because we didn't starting using the B&K hammer's hard tip until that last corner. That means we don't have a quantitative metric for how much the dampers reduced each blade's higher frequency modes we expected to damp (at ~450 Hz and ~900 Hz) -- but at least we can see the end result from corner three. However, we have no reason to believe that these dampers aren't working at these frequencies as expected. Once pumped down, we'll make a comparison between the GS13 ASDs before and after the vent, and that'll be our final assurance that all has gone well.

We definitely need to use the hard-tip on the B&K hammer for all future installs and characterization of ISI damping mechanisms.

Tuned Mass Damper Tuning

Once we'd identified the new first bending mode frequency, the smaller, narrow-band tuned mass dampers needed retuning. I took them into the optics lab, set up a crude shadow sensor system with a HeNe laser and On-Trac QPD, hooking up the output to an SR785. See attachment 2016-04-07_TMDTuningPics.pdf for pics of the setup. 

The tuning itself was pretty easy. I went in armed with a set of calipers to try to a priori predict the increase in length needed to decrease the resonance to the desired frequency, but the tightening of the restraining bracket turned the frequency tuning to a guess-and-check game regardless. Pre- and Post measuring after tuning was finished showed an increase in length of 0.6+/-0.015 [in] to 0.66+/-0.015 [in]. 

A comparison of the resonant frequency and Q is shown explicitly in 2016-04-07_H1ISIHAM6_BladeTMD_Tuning.pdf. I was able to preserve the Q of about 10, and moved the frequency down to 139 [Hz] as desired. Given that I knew the frequency of each blade's first bending mode, and the resonance of each tuned mass damper, I matched each blade to the appropriate TMD and they're thus annotated in the legend by on which blade they were installed.

Tuned Mass Damper Installation

Nothing much to report here, the installation went smoothly. With the TMD in hand, I reached through the outer wall and "picture" access panel to secure the magnetically attached device to the blade. It took a few pictures and adjustments to get the final location "right" as shown in 2016-04-07_TMDInstalledPics.pdf (though they probably don't *have* to be so precisely oriented and located). Installing the TMD on Corner 1's blade was the "most difficult," in that it's a little uncomfortable and awkward climbing in chamber, sitting down, and reaching into the ISI, but it's totally doable even for someone of my stature.

To help identify potential scattering sites.

Miriam, Evan G, Evan H

*SUMMARY*

We investigate the possibility that there occurs a saturation or a slew rate problem somewhere in the analog electronics of the sensing chain causing (or resulting in) a blip glitch. This investigation finds no issue with either of these for a sample of the four loudest blip glitches in O1 data.

*DETAILS*

We investigate four different times: 1128085613.28, 1128221842.47, 1128264648.20, 1130156793.54

1. Get the data from OMC-DCPD_A_OUT_DQ and OMC-DCPD_B_OUT_DQ. These data have undergone some digital filtering to undo what the analog filtering did before the ADC. It is our best estimate of the photocurrent from the OMC DCPD. The units are in mA.

2. Convert the data into Volts (V=IR), where R=400 Ohms (see DCC:D060572) 

3. Following the schematic of the in-vacuum OMC DCPD, there are two zpk filters, each with one zero at 8Hz and one pole at 80Hz (we ignore the 15.9kHz pole, as it is far out of the frequency band of interest), and there is a gain of 2 at the differential output. We apply this filtering to the DCPD data (in V). The resulting time-series has to be within +/-15 V to not saturate. In the selected times of blip glitches, we obtain voltages that are well within this bound (see attached document, "preamp" plots) 

4. Continuing with the signal chain, the whitening chassis (DCC:D1001530). The whitening gain setting has a nominal value of 0 dB. In addition, only a single whitening stage is applied with one zero at 1Hz and one pole at 10Hz. The resulting time-series has to be within +/-15 V to not saturate the whitening op-amp, and within +/-20 V to not saturate the ADC. In the selected times of blip glitches, we obtain voltages that are well within these bounds (see attached document, "whitch" plots).

5. Typical slew rates for the op-amps used in these circuits are 2.5 or 5 V/us. None of the times selected show such high slew rates.

6. Note that the plots in the attachment are dominated by the low frequency content of the signal, that has an amplitude much greater than blip glitches. Only after whitening the time-series, the blip glitches become dominant.

*CONCLUSION*

The analog electronics in the sensing chain do not appear to be responsible for blip glitches. 

HV cable was pulled in from VEA mechanical room to X2-8 beam tube by Richard and Gerardo this morning. Connectors were terminated at both ends this afternoon.

J. Oberling, P. King

Spent today installing and testing the new HPO water manifold.  The mainfold is now installed and the valves for the MOPA, PWR, and Laser Head water circuits are set to provide the proper flow rates.  While we were doing this we figured it was a good idea to implement ECR E1500408; this ECR is for installing valve on the outlet side of the Laser Crystal water circuit.  This is now also complete.  A picture of the new water manifold and valve is attached.  This completes this ECR.

Once everything was installed and verified that it wasn't leaking, We noticed the flow sensor for laser head #1 (part of the Laser Head water circuit) had stopped reading.  Opening up the HPO box we saw that the turbine was not spinning.  We drained the circuit and removed and replaced the flow sensor.  Before installing the new flow sensor, Peter found an ~1/2" piece of black plastic (looked like a piece leftover from cutting the threads on the main water manifold.  Recall we pulled out a bunch of these yesterday while prepping the new manifold.) sitting in the output side of the flow sensor (in the small black manifold that sits above the HPO resonator cavity).  We installed the new flow sensor, turned on the chillers, and watched for leaks.  After several minutes no leaks were observed, so we left the crystal chiller running overnight to test the system (the diode chiller is still OFF).  The HPO and FE lasers are both OFF.  Will continue in the morning.

Activity Log:All Times in UTC (PT)

15:00 (08:00) – Start of shift

15:41 (08:41) Kyle – Going to End-Y mechanical room

15:52 (08:52) Gerardo – Going into the LVEA looking for cables

15:57 (08:57) Jim & Corey – Going to HAM6

16:05 (09:05) Jeff K. – Going to HAM6

16:10 (09:10) Peter & Jason – Going into the H1 PSL enclosure

16:15 (09:15) Gerardo – Out of the LVEA

16:22 (09:22) N2 delivery to the X-Arm

16:49 (09:49) Richard & Gerardo – Going to End-X

17:10 (10:10) N2 delivery to Y-Arm CP4

17:30 (10:30) Kyle – In LVEA near BSC4 (WP #5817)

18:15 (11:15) Jeff K. – Out of LVEA

18:16 (11:16) Ed – Terminating cables at HAM1, Vertex, and HAM6  

18:32 (11:32) Richard & Gerardo – Back from End-X

19:36 (12:36) Filiberto – Going to End-Y mechanical room

20:00 (13:00) Jeff K. – Going into LVEA then Optics lab

20:01 (13:01) Jim – Going to HAM6.

20:15 (13:15) RO alarm – John notified. No alarm – Dave doing burt restore to fix alarm handler

20:41 (13:41) High dust counts at HAM6. Stand down ISI work while Kyle & Jeff were sealing up East Door curtains for door install

21:20 (14:20) Krishna & Michael – Going to End-Y for BRS

21:39 (14:39) Bob's Heating on site to work on VPW HVAC

22:10 (15:10) Kyle, John, Chandra, Jeff – Reinstall HAM6 east door.

23:03 (16:03) Jeff K – Going to HAM6 to install tuned mass dampers

End of Shift Summary:

Title:04/07/2016, Day Shift 15:00 – 23:00(08:00 – 16:00) All times in UTC (PT)

Support:

Incoming Operator:

Shift Detail Summary:Work on HAM6 is going well and should wrap up tomorrow. East HAM-6 Door has been reinstalled. 

Michael, Krishna

We did more sensor correction (SC) tests today with mildly windy conditions. This was similar to the tests reported in 26455. However, we had made one important mistake in that test - the CPS value reported was before the blend filter input which is measured after sensor correction. This was wrong, it needs to be measured before SC (using H1:ISI-ETMY_ST1_SCSUM_CPS_X_IN_DQ). This is a more accurate measure of the local table motion.

As before, the four configurations were: a) 90 mHz blends, b) 90 mHz blends and SC using BRS, c) 45 mHz blends, d) 45 mHz blends and SC using BRS and we also tested e) 90 mHz blends and SC without using BRS.

The two measurements for End Y (BRS-2) and End X (BRS-1) are shown in the attached pdfs. We saw very similar looking results in both.

The wind-speed varied between 5-20 mph with a rough average of ~12 mph for EY and ~9 mph for EX. First page shows the Stage 1 T240 motion and as before 45m+BRS_SC gives the best performance, with 45m, 90m+BRS_SC or 90m+SC configurations giving roughly the same performance. The next page shows the RMS of the CPS signals before the sensor correction. 90m has the least CPS motion with 90m+BRS_SC being the next best. SC using BRS and not using it (with or without tilt-subtraction), can be compared by using the blue and yellow/orange lines respectively. The third page shows the ground motion during each configuration and the corresponding tilt-subtracted signal.

Some comments:

1) The sensor correction filter used here (Mitt_SC) is a broadband filter going down to 30 mHz. We may want to tune it for 40-50 mHz instead, to reduce the low-frequency motion.

2) Based on the results so far, the best configuration for 0-10 mph wind-speeds (LOW wind) might be 45 mHz blends+BRS_SC which gives the best isolation (a factor of ~50-100 at the microseism!). And for winds above 10 mph (HIGH winds), 90 mHz +BRS_SC might be best. This is the same configuration that was followed for O1 but with the addition of SC using BRS.

This afternoon, we recentered the DC position of the beam-balance. Due to the process, the amplitude is very high so we are using this opporunity for a Q measurement.

Corey, JeffK , Jim

This morning, while Corey put doors on the south side of the chamber, Jeff and I set up the B&K and got before measurements for the last corner. I then installed the last of the new damping elements, and we got the final B&K measurements for the ISI. Corey then helped me close up the last corner. After lunch, Corey and I went back out, unlocked and floated the ISI, while Jeff took the the old TMD's to the lab for retuning.  Jeff will probably be logging results for TMD's and B&K soon. I'll add that we used the "hard" tip for the B&K for this mornings measurements. We should have started with that, it worked a lot better than hte soft tip, with coherence out to 2khz. Jeff says he will hang out and reinstall TMDs later tonight.

Per Hugh's suggestion, I looked at the ISI's operating CPS position over last week (before the vent) and rebalanced the table as close as I could get. Looking at the CPS in the cartesian basis made this very easy as any mass we added only changed RX,RY and Z. I'd heard this suggestion from others in the seismic group, but this was the first time I'd tried it. Recommened. 

Currently, the vacuum group is putting on the East door on HAM6. This will still allow us to access every thing we need to on the ISI tomorrow. 

Tomorrow, we still need to unlock the ISI again, do a sweep of the chamber, pull the septum cover and do close-out tf's. We should also check the beam diverter still actuates and that the OMC survived the rough-housing it got while Corey and I rebalanced the table.

S&K electric converted single receptacles to duplex today which allows us to power fans and the HAM1 and HAM6 ion pump controllers (without extension cords!!!)

Chart updated to include both high power beam dumps.

For reference here is the last HAM6 pumpdown in June/July 2015.  From atm to stable turbo running was ~3.5 hours.

This log is duplicated in LLO alog 25563.

Summary

Further to my alog 25932, I have calculated approximate TCS power levels for O2, this time including the effect of the ring heaters. The results are presented in tabular form.

The bottom line:

1. the RHs are used predominantly to correct the surface deformation on the test masses.
2. The CO2 lasers are used to correct the residual lens in the test mass after the RHs have been applied. 
3. We might need to commission annular heating on CO2Y at LHO.
4. ETMY_RH at LLO requires a larger increase in power due to the relatively high absorption in ETMY

Calculation:

Surface curvature

To correct the surface curvature errors, we use the RH. It is straightforward to show that the required change in RH power is given by:

dP_RH = - (absorb*dParm) * dSD/dP_self / (dSD/dP_RH),

where 'absorb' is the absorption in the surface of the optic, dParm is the change in the power in the arm, dSD/dP_self (H1:TCS-SIM__SURF_DEFOCUS_SELF_GAIN) is the steady-state change in the surface deformation in diopters per Watt absorbed power, and dSD/dP_RH (H1:TCS-SIM__SURF_DEFOCUS_RH_GAIN) is the steady-state change in the surface deformation per Watt power from the RH. These values are available on the TCS SIM MEDM page.

Substrate thermal lens

The same calculation can be done for the CO2 laser power:

dP_CO2 = - [ (absorb*dParm) * dS/dP_self + dP_RH * dS/dP_RH ] / (dS/dP_CO2),

where dS/dP_self, dS/dP_RH, dS/dP_CO2 are the change in the substrate lens defocus per Watt for self-heating, RH and central heating CO2 laser respectively. Note that the RH power has, nominally, been fixed by correcting the surface curvature in the previous step.

Note the distinction between the responses for the surface deformation, dSD/dP and the substrate lens, dS/dP.

LHO settings:

Based on the best estimates for the absorption in the test masses, the changes in actuator settings are:

Obviously, we can't apply negative power with central heating on CO2Y. The reason CO2Y was set to 0W for O1 was because it wasn't working during that science run - therefore, this is probably not the best operating power for CO2Y at 100kW of arm power.

• A recommended commissioning activity would be to optimize CO2Y at 100kW in the arms and then recalculate the required TCS at 200kW in the arms.
• If we find we do need more negative lensing in the substrates of the ITMs we can do one of two things:
  • Use more RH power and move the ITM surface curvature away from the nominal operating point
  • Or, commission an annular heating mask with the CO2 lasers.

LLO settings:

Notice the relatively large increase required for ETMY - this is due to our measurements indicating an absorption on there of the order of 1.6ppm, or about 5 or 6 times larger than the average of the other optics.

Krishna, Michael

Due to the slow wind day, we decided to take a look at an 6.7 magnitude earthquake that struck near Australia last night. All three STS seismometers (ETMX, EMTMY, and ITMY) and the BRS picked it up very clearly. The first plot in the attached pdf shows the time series of both the z component of the STS at ETMY and the BRS2 also at ETMY. The BRS shows clear coherence with the STS at the earthquake frequencies. The second plot shows the ASD of the same two signals along with the possible acceleration coupling of the BRS assuming a d value of 0.5 microns. The BRS signal is a factor of two or three above the possible acceleration coupling around the earthquake which suggests that it is seeing the real tilt component of the Rayleigh waves of the earthquake.

Along with studying how the BRS sees earthquakes, we used z components of the STS seismometers to estimate the velocity and direction of the primary wave. By measuring the phase between the seismometers at the end stations and the seismometer at the corner station, we were able to get the time delay of each arm. Using this we get an estimate for the velocity of 4700 +- 800 m/s and the angle of 71+-10 degrees from the X-arm. The third and fourth plots in the pdf show how the angle and velocity measurements change with different number of averages in the phase calculations. The fifth and sixth plots show how the measurements change over 125 second chunks of time.

Earthquake info: http://earthquake.usgs.gov/earthquakes/eventpage/us20005fsi#general

Edit: We realized that yesterday's calculations of the acceleration coupling had an extra factor of g so the BRS signal is actually a factor of 30 above the possible acceleration coupling. The first plot in the Update pdf shows the updated ASD of the z component of the STS at ETMY, the BRS, and the correct acceleration coupling. Also, using the ratio of the z component of the seismometer and the tilt signal we were able to get another estimate for the velocity which is consistent with the previous analysis.

We also decided to take a look at how the different STS seismometers reacted to the earthquake to ensure they were calibrated correctly. The second plot of the Update pdf shows the ASD of the z component of all three seismometers. The final four plots show the magnitude of the transfer function at the earthquake's primary frequency between the ETM seismometers and the ITMY seismometer with varying number of averages and over time of the earthquake. This suggests that the ETMX seismometer's calibration is ~7% less than ITMY and ETMY's calibration is ~15% less. Coincidentally, we had to add a factor of 0.85 to the BRS to get the gain to match the ETMY seismometer.

J. Warner, N. Kijbunchoo, C. Gray, J. Kissel, K. Kawabe

We've made good progress today. Two of the corner's new damping systems are completely installed and characterized (analysis of data to come). That's Corner 1 and Corner 2's blade tip dampers, and the Horizontal and Vertical GS13 can dampers on Corner 2 and Corner 3. We've also re-installed the hardest of the inner and outer walls on Corner 1.

Schedule Status:
On the docket tonight:
- Process and post today's B&K hammer results
- Re-tune tune the smaller Tuned Mass Dampers

Up for tomorrow:
- Characterize the free blade and free cans again (it's quick, and we want to be consistent)
- Install of Corner 3's blade tip damper (the last one)
- Characterize the blade damper
- Install of Corner 1's GS13 can dampers (horizontal and vertical; the last one)
- Characterize the cans
- Re-install all of the walls and trim mass
- Install newly-retuned tuned mass dampers on all blades
- Float and balance the table

For Friday:
Begin close out checklist.

This puts us maybe ~1/2 a day behind schedule from E1600092. We'll discuss if we've made it back on schedule by lunch tomorrow.

Attached are some pictures of the days work, both of the damper installation, as well as some publicity photos of the stuff on top of HAM6. To entice you, I've attached the best pictures from both .pdfs in their raw form.

Stay tuned for characterization results of the two corner's blade tip dampers. I can tell you right now that we have too little coherence with the B&K up at ~1-2 kHz, so there won't be much to show on the Can Dampers. However, we've audibly compared damped vs. undamped cans with impacts from a metal wrench and there's a satisfactory "clunk" instead of multi-tone "ding."

The HPO survived overnight without any problems.

This morning the HPO was turned off to allow for testing of the pre-modecleaner setup
with the front end laser.  After that was completed the HPO was turned back on.  After
being on for ~20 minutes, a leak developed in both flow meters leading to turning the
HPO off (the flow meters being for the power amplifier and the power meters).  Replacement
of the water manifold is in progress.  Fortunately no damage was done to the HPO.

Examination of the MOPA flow meter showed a large rust stain on the flow meter body.
In retrospect this partly explains why in recent weeks the topping up of the crystal
chiller was occurring more frequently.  In the short time between turning off the laser(s)
and the chillers, the flow meter for cooling the power meters leaked water profusely.
Examination of that flow meter showed a break in the body.

It should be noted that the flow meters, which are part of the water manifold, are located
on the floor under the PSL table and are typically not inspected every day.

The spare water manifold is being prepped for installation.  In doing so it was noticed
that there were a number of material strands left from when the threads were tapped.  As
many of these were removed as possible, thereby reducing the possibility of one of the
coming loose and jamming up the plumbing else where in the system.



Jason, Peter