Virgo Logbook

Injection system (General activities)

chiummo, DeRossi, Fiori, Romero - 21:14 Tuesday 08 October 2019 (47156)

Laser bench noise injections

This afternoon we injected some mechanical noise to the laser bench by using the small shaker installed on the leg of the bench (see 47083). The mechanical effect was monitored by means of the accelerometer on the Laser bench, recently moved from vertical to horizontal (see 47078).
The noise injection parameters are reported below. In the attached plot the effect of the largest injected line at 20Hz, where there eems to be a clear resonance, is clrearly visible on both PRCL and DARM.

Analysis and projection of quiet condition will be performed shortly, but it seems likely that the bump in the quiet DARM at that frequency is correlated to the Laser bench resonance. If confirmed, mitigation actions will be studied.

Injection times:

quiet condition
GPS = 1254588112 08 Oct 2019 16:41:34 UTC (08 Oct 2019 18:41:34 CEST)

small shaker injecting broad noise freq = 15-25Hz; amplitude: 0.0001
GPS = 1254588756 08 Oct 2019 16:52:18 UTC (08 Oct 2019 18:52:18 CEST)

small shaker injecting a line freq= 20Hz
GPS = 1254589862 08 Oct 2019 17:10:44 UTC (08 Oct 2019 19:10:44 CEST)

small shaker injecting a line freq= 20Hz, amplitude x5
GPS_start = 1254590327 08 Oct 2019 17:18:29 UTC (08 Oct 2019 19:18:29 CEST)
GPS_end = 1254590891 08 Oct 2019 17:27:53 UTC (08 Oct 2019 19:27:53 CEST)

Images attached to this report

Comments to this report:

Tringali, Fiori, Paoletti - 12:35 Wednesday 09 October 2019 (47165)

In the attached plot , we show that during the shaking of laser bench, similar bumps are seen also in BsX channels.

Images attached to this comment

romero - 13:56 Monday 21 October 2019 (47293)

An analysis of the data has been performed. For the quiet time in the entry (1254588112), a noise projection has been done into PRCL. We have taken the accelerometer in the LB: 'V1:ENV_LB_ACC_Z' and applied a high-pass filter with frequency cutoff at 4.5Hz (see AccFiltQuiet.jpg). The reason is that it's not very accurate for low frequencies, so we have taken separately the Guralp velocimeter: 'V1:ENV_CEB_SEIS_W' and applied a low-pass filter with frequency cutoff at 2.5Hz (see GuralpFiltQuiet.jpg). Then, after adding both signals (see SumQuiet.jpg), we projected it onto PRCL (see NoiseProj.jpg).

We see that the 20Hz bump and the 40Hz bump are fitted with the signal from the LB with a coupling constant of 0.5. Consequently, we already have a clear source for these bumps: scattered light from the laser bench.

Following the same procedure, a fit of the accelerometer data has been done to hrec during the injection time: from 1254590327 to 1254590891 (see FIT.jpg)

Images attached to this comment