Reports 1-1 of 1 Clear search Modify search
LAS (bKAGRA laser)
takahiro.yamamoto - 21:14 Friday 02 June 2023 (25486) Print this report
ISS DOWN before lockloss again

[Ushiba, YamaT]

Abstract
ISS was down before lockloss again at 17:07:07 JST.
This is a same situation as klog#25461.
Local seismic disturbance in human band (3-10Hz) seemed to be related to ISS down.
 
Details
Figure 1 shows some time series signals around lockloss.

The leftmost column shows the guardian state numbers.
At first, ISS guardian went to DOWN (top panel of the leftmost column).
As the next, SYS_SDF guardian detected the EPICS record changes in the DOWN state of ISS guardian (the 2nd top panel).
After then, IFO guardian left OBSERVING state because of SDF differences (the 3rd top panel)
After a few samples in 16Hz, LSC_LOCK and IMC guardian went to DOWN state (the 4-5th top panels).
Finally IO guardian went to DOWN because of IMC guardian left LOCKED state (the bottom panel).
This timeline is same situation as klog#25461.

About 5 seconds before ISS went to DOWN, seismometers detected the disturbance
in the frequency band of 3-10Hz (top panel in the middle column).
CARM feedback signals and TRANS QPD signals at MCE and IMMT1 was increasing
in conjunction with this seismic disturbance (the 2nd-4th top panels in the middle column).

The laser power fluctuation could be detected at PMC output
though no excess was seen at fiber output (the 1st-2nd top panels in the right column).
And also, IMC output power didn't show the any excess corresponding to the seismic disturbance (the 3rd top panel).
This fact seems to be inconsistent with the fact that the TRANS QPDs at IMMT1 and MCE shows large fluctuation.
Is this just a SNR problem...?


Though I hadn't focused on the seismic disturbance in klog#25461,
I could confirm that there was a seismic disturbance around 3-10Hz around also previous ISS DOWN as shown in Fig.2.

The problem that IFO guardian speaks in unintentional situation should be fixed in the next maintenance day.
Independently of this, we need to some kind of improvement (optics, control scheme, etc.?) is required to improve the situation that
ISS is easily affected by the local seismic disturbance in the frequency band of human activity.

Images attached to this report
Comments to this report:
takahiro.yamamoto - 1:30 Sunday 04 June 2023 (25494) Print this report
There were some lockloss with similar situation even if ISS went to DOWN at the same time (in 16Hz) with LSC_LOCK.
(Excluding the Earthquake and the intentional lockloss due to the maintenance, 20-25% of lockloss could be classified as this case(?).)

Attachments shows the 5 lockloss cases which has coincident with the large disturbances around human activity band.
(The 2nd and the 5th figures are same time as Fig.1 and Fig.2 in klog#25486, respectively.)
In all cases we can see the human activity band disturbance (the 6-7th row of middle column) a few second before the lockloss.
At the same time TRANS QPD of IMMT1 (bottom row of the most two left columns) and MCE and TypeC, Bp oplevs (most two right columns) show large fluctuations.
In some cases, MICH/PRCL error signals (the 3rd and 4th row of the most left column) also shows large disturbance.

Because we can see too many excesses in the human activity band, it may be better to consider the accidental coincidence rate.
Anyway, it might be better to investigate the relation between the presence (absence) of the large fluctuation on each optics signals
and RMS of seismic motion in the human activity band.
Images attached to this comment
tomotada.akutsu - 12:36 Sunday 04 June 2023 (25498) Print this report
  • The time delay ordering among SEIS_IXV, EXV, and EYV are not always the same... so the fluctuation source might not be always the same.
  • It might be better to check if the suspensions fluctuations would be consistent with these human-activity seismic motions. Are they explained with the measured transfer functinos?
takaaki.yokozawa - 11:18 Monday 05 June 2023 (25507) Print this report
Their time series / BLRMS 1-3 Hz / BLRMS 3-10 Hz
Images attached to this comment
tatsuki.washimi - 11:38 Monday 05 June 2023 (25506) Print this report

I checked the seismometer signal. It looks typical earthquake waveform because

  • the signal is almost the same timing and amplitude for all areas.
  • the shape of the waveform looks P-wave and S-wave
  • the arrival time was delayed 1~2 seconds at the EYV compared with IXV/EXV. -> coming from a south-south-east direction?

So I checked the Hi-net data at Kamioka, Furukawa, Takayama, and Yatsuo.
A similar signal is also observed in all points and earlier and larger in the Southan area.

So I can conclude it was an earthquake.

Images attached to this comment
tatsuki.washimi - 13:41 Monday 05 June 2023 (25511) Print this report
Images attached to this comment
tomotada.akutsu - 14:34 Monday 05 June 2023 (25512) Print this report

> So I can conclude it was an earthquake.

Thanks, a nice analysis.

Search Help
×

Warning

×