Документ взят из кэша поисковой машины. Адрес оригинального документа : http://zebu.uoregon.edu/~uochep/talks/talks04/G040391-00.pdf Дата изменения: Wed Oct 6 02:44:10 2004 Дата индексирования: Tue Oct 2 11:25:23 2012 Кодировка: Поисковые слова: п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п

PROGRESS SINCE MARCH ON S3 ENVIRONMENTAL DISTURBANCES
Dust, Acoustic, Magnetic, and LHO Seismic Mitigation and Intersite Burst Search

Robert Schofield, University of Oregon John Worden, Richard McCarthy, Bartie Rivera, Nathan Hindman, LHO AND MANY OTHERS

1

LIGO-G040391-00-W


DUST BURSTS
Dust burst in raw AS_Q time series:
12 10 8 6 4 2 0 -2 -4 -6 -8 0.2 0.22 0.24 0.26 0.28 0.3 0.32 0.34 0.36 0.38 0.4 H2:LSC-AS_Q[0][0]

After ellip("HighPass",3,3,50,50) filter:
0.8
H2:LSC-AS_Q

Signal Signal

0.6 0.4 0.2 -0 -0.2 -0.4 -0.6 -0.8 -1 0.2 0.22 0.24 0.26 0.28 0.3 0.32 0.34 0.36 0.38 0.4

Time (s)
T0=09/04/2004 05:37:00.010009 Avg=1

A generic step, filtered with the pendulum transfer function and injected onto a test mass also looks like a few cycles at 100s of Hz in an un-calibrated time series. 2


Stirred up dust produces AS_Q glitches only at dark port.
Dust (particles > 0.5um / ft^3)
1e+05

10000

2k PSL (laser table) ISCT7 (reflected port) ISCT10 (dark port)

1000

100

10

45000 15

1

50000

55000

60000

65000

70000

ASQ glitch rate (glitches/ s)

Number of seconds after Mar. 8, 2004 12:00:00 UTC
H2ASQ_4s_100f_RateHistory H2ASQ_6s_100f_RateHistory

10

5

0 45000

50000

55000

60000

65000

70000

Number of seconds after Mar. 8, 2004 12:00:00 UTC


Prototype purged cast acrylic sheet dust enclosure for science runs covering areas where beam is small and dust produces large AS_Q glitches.

No enclosure: 2.8 x 10-6 glitches/sec per particle/cubic foot. Purged enclosure: 0.062 x 10-6, a factor of roughly 50.

4


One wall of the ISCT4 table enclosure has been replaced with a HEPA filter, to keep the table clean and quickly reduce glitch rates during commissioning.

air flow

5


SUMMARY OF DUST INVESTIGATIONS 1) Dust glitches in AS_Q are produced only at dark ports, especially where the beam is small. 2) Larger than 6 sigma dust glitch rate is about 3x10-6 glitches/sec per particle/cubic foot (H1 and H2). 3) Long after entry into a clean (HEPA-vac'd table, wet moped floor) dark port enclosure, the particle count is about 1 particle/cubic foot (glitch rate 3x10-6 glitches/sec). 4) Prototype table top purged enclosure reduced glitch rate by about 50. 5) HEPA filter purging of table enclosure reduced glitch rate even more, but is noisy and should probably be used only during commissioning. 6) No individual veto found for S3 dust glitches. Epoch warning flag raised for about 24 hours after table entry. Dust glitch rates should be down to 3e-3/sec after 10 or 20 hours. Suggested S3 epoch flags: http://apex.ligo-wa.caltech.edu/~roberts/dustflags.txt 6


ACOUSTIC COUPLING
During S2 (Black: AS_Q with injected noise, Red: AS_Q, no injections)
10 10 10
H0:PEM-BSC7_MIC(REF0) -12 H0:PEM-HAM1_MIC(REF1) H1:LSC-AS_Q(REF2) H0:PEM-BSC7_MIC(REF3) H0:PEM-HAM1_MIC(REF4) H1:LSC-AS_Q(REF5) -13

m/sqrt(Hz)

10 10 10 10 10 10

-14

-15

-16

-17

-18

-19

10

2

10

3

Frequency (Hz)
*T0=12/03/2003 05:29:46 *Avg=1 BW=0.187499

Just Before S3 (see previous talks for progress with each mitigation step)
10 10
-12

)

Magnitude (m/Hz

1/2

10 10 10 10 10 10 10

-13

H1:LSC-AS_Q(REF21) H1:LSC-AS_Q(REF27) H1:LSC-AS_Q(REF29) H0:PEM-BSC7_MIC(REF89) H0:PEM-BSC7_MIC(REF30)

-14

-15

-16

-17

-18

-19

10

2

10

3

Frequency (Hz)
*T0=04/10/2003 07:05:07 Avg=1 BW=0.187499

7


Still to be done:
I. REDUCE CONTINUOUS SOURCES (factor of 3 to 5) A. Continue with plans to acoustically house or remove electronics cabinets B. Insulate pipe-feed through from mechanical room C. Insulate PSL chillers II. REDUCE COUPLING (factor of 5 for H1 & L1, less for H2) A. Clipping 1) Replace AS and REFL periscopes with V3 of new design 2) Enlarge or remove 1/2 lambda plate and polarizer in REFL path 3) Damp PSL periscopes 4) Damp mounts and dumps etc. 5) Continue testing floating legs for low f B. Backscattering from table (out of prudence - we haven't seen coupling) 1) Grouted damped rigid legs, unless interferes with clipping reduction above III. ACOUSTICALLY ISOLATE WORST COUPLING SITES A. REFL port enclosures with internal absorption kits? Reevaluate after above REFL work. 8


STEPPING BACKWARDS: Acoustic coupling increased by a factor of ten between S3 and May. Tracked to a new backscattering Blackhole beam dump.]
Red: Blackhole dump, Blue: razor blade dump
H1:LSC-AS_Q(REF20) H1:LSC-AS_Q(REF22) H1:LSC-AS_Q(REF24) H1:LSC-AS_Q(REF26) H1:LSC-AS_Q(REF28) H1:LSC-AS_Q(REF30)

10

-1

Magnitude

10

-2

10

-3

100

200

300

400

500

600

Frequency (Hz)
*T0=10/07/2004 04:52:08 Avg=1 BW=0.187499


MAGNETIC COUPLING
· During S3, pulsed heaters created side bands in AS_Q around 60 Hz · S3 PEM injections suggest that the 60 Hz peak in AS_Q can not be substantially reduced without reducing 60 Hz magnetic fields or magnetic field coupling.

Mitigation Plans
I. REDUCE SOURCES a. Eliminate 1 Hz pulsed heating (successfully tested at LHO EX and LVEA) b. Eliminate or smooth cycling of heaters (current experiment to control temperature in LVEA and EX with variable chilled water and fixed power heaters). c. Reduce 60 Hz from heaters by reducing current loop areas, and arranging so fields cancel (successfully tested reduced area heating elements). II. REDUCE COUPLING? III. INSULATE COUPLING SITES (mu-metal)? For II and III the first step is to identify coupling sites. 10


Replacing U-bent elements (shown below) with 2-pass elements and cleaning up loops in heater current distribution reduced 60 Hz magnetic field from this heater by a factor of 3.

11


Magnetic Field Coupling Sites
Searched with small magnetic field generating coil for most important coupling sites at LHO MY and EY stations Black: no injection; Blue: coil near electronics cabinets; Red: coil at vacuum chamber wall
Power spectrum
H2:LSC-AS_Q(REF0) H2:LSC-AS_Q(REF1) H2:LSC-AS_Q(REF2)

10

-2

Magnitude

10

-3

135

140

145

150

155

160

165

Frequency (Hz)
*T0=04/06/2004 05:06:58 Avg=1 BW=0.187499

Most important coupling location seems to be test mass.

12


SEISMIC COUPLING
Dewar Glitches In an attempt to reduce the glitch rate, a dewar was insulated (2 inch polystyrene foam, aluminized mylar layer, 2 more inches of foam, outer mylar layer and aluminum sheath). normal insulated

13


Red: dewar thermocouple; Black: outside temperature 40 first layer before
35 30

Centigrade

25 20 15 10

2004-07-06 12:57:36

2004-07-08 00:57:36

2004-07-09 12:57:36

2004-07-11 00:57:36

2004-07-12 12:57:36

aluminum sheath
35 30

after

25

20

15

2004-07-21 12:57:36

2004-07-23 00:57:36

2004-07-24 12:57:36

2004-07-26 00:57:36

2004-07-27 12:57:36

Got the anticipated factor of 4 reduction in diurnal variation plus direct sun shielding.


First indications are that the insulation has helped.

Night Time Seismic Spikes Peaking in the 12 to 17 Hz Band Period 5/31 - 6/30 (Before) 7/20 - 8/11 (After) Number of Event 17 2 Rate (events per night) 0.6 0.1

15


Seismic Sensitivity
S3 PEM injections showed that H1 would loose lock even with very small acoustic and seismic disturbances at the EY station. This was traced to an optical lever interface board.

Time series
2000
H0:PEM-BSC10_ACC1Z[0][0] H1:SUS-ETMY_OL4_IN1[0][0]

0

Signal

-2000

-4000

-6000

-8000 0 10 20 30 40 50 60 70

Time (s)
T0=16/07/2004 21:14:11 Avg=1/Bin=13

The bad board produced a "DC" offset every time there was an event. New board doesn't, and station is no longer so sensitive. No similar problems at other LHO out-stations. 16


INTERSITE BURST SEARCH magMon algorithm for detecting PEM bursts that are coincident between sites
· Modification of an older version of glitchMon · Coincidence before trigger generation to explore low thresholds · Can insist on coincidence between multiple channels (e.g. on 2 of 3 voltage monitors at each site) · Coincident event rates for aligned and offset LHO and LLO time series No excess over chance coincidence found in coil magnetometers:
L0:PEM-COIL_MAGX; H0:PEM-COIL_MAGX; 560-580 Hz 4th order butterworth; 5,888,200 seconds of S3

Threshold (sigma) 5.5 4.7 3.9 3.1 2.7 4

Events exceeding threshold 1

Events for offset time series

(on - off) 3 10 66 205 -730

sqrt(on+off) 2.23607 15.0997 97.1494 521.83 1145.22

119 4752 136256 655400

109 4686 136051 656130


No excess found in LLO and LHO line voltage monitors:
L0:PEM-EX_V1;;L0:PEM-EY_V1;L0:PEM-LVEA_V1;H0:PEM-LVEA2_V2; H0:PEM-MY_V1; H0:PEM-LVEA2_V1;

Threshold (sigma) 4.7 4.06 3.42 2.78 2.14 1 3

Events exceeding threshold 0 2

Events for offset time series

(on - off) 1 1 -2 250 1335 1

sqrt(on+off)

2.23607 15.3623 182.384 1913.02

117 16757 1830483

119 16507 1829148

5888200 second of S3 data; only 1 event allowed each second; only 2 signals in excess of threshold demanded from each site.

No excess found in radio channels:
H0:PEM-RADIO_LVEA; H0:PEM-RADIO_LVEA; 5888200 s

Threshold (sigma) 6.7 5.66 4.62

Events exceeding threshold 209 688 3554

Events for offset time series 236 756 3720

(on - off) -27 -68 -166

sqrt(on+off) 21.095 38 85.2877


Threshold (sigma) 4.1 3.06

Events exceeding threshold 15576 2002687

Events for offset time series 15554 2000813

(on - off) 22 1874

sqrt(on+off) 176.437 2000.87

19