KAGRA Logbook

CAL (General)

takahiro.yamamoto - 11:57 Tuesday 30 May 2023 (25440)

Weekly calibration on May 30.

[Shingo, YamaT, with Dan (remote)]

Abstract

We measured the open loop transfer function of DARM and TFs from each stage of ETMX to DARM error signal
in this morning as the post calibration for the week of May 24.
These measurements tagged as "0530/0802"
In all measurement, there is no large difference from the measurement at just before O4a start.
At the beginning of the maintenance work, we turned off OBS_INTENT.

Quick results

Estimated actuator efficiency and optical gain are as follows.
- ETMX_TM: -3.4849e-14 m/ct (0.3% change from previous measurement klog#25349).
- ETMX_IM: -1.4452e-14 m/ct (0.4% change from previous).
- H_DARM: 254.24dB (0.7% change from previous).

Actuator efficiency of ETMX_TM was estimated from the ratio of two TFs with an excitation from pcal and from TM stage which are shown in Fig.1 and Fig.2.
At this time calibration reference is the laser power of pcal at RxPD.
Actuator efficiency of ETMX_IM was estimated from the ratio of two TFs with an excitation from IM stage and from TM stages which are shown in Fig.3 and Fig.4.
The calibration reference is the TM actuator efficiency which is estimated above.
DARM optical gain, H_DARM was estimated from the OLTF and estimated actuator efficiencies as shown in Fig.5 and Fig.6.

Relative changes from the measurement just before O4a start is less than 1%.

These measurements will be used for making offline h(t).

Images attached to this report

25440_1685415449_Screenshot from 2023-05-30 11-05-11.png

25440_1685415452_Screenshot from 2023-05-30 11-05-21.png

Comments to this report:

takahiro.yamamoto - 12:01 Tuesday 30 May 2023 (25442)

Actuator efficiency, Ha, is normalized at 10Hz as Ha * (10Hz/f)^2

Shingo Fujii - 12:31 Tuesday 30 May 2023 (25444)

I cross-checked actuator efficiency.

Estimated actuator efficiency are as follows.
- ETMX_TM: -3.4873e-14 m/ct (0.07% change from YamaT-san's result).
- ETMX_IM: -1.4165e-14 m/ct (-2.0% change from YamaT-san's result).

There is no large difference between calculations.

takafumi.ushiba - 14:24 Tuesday 30 May 2023 (25447)

Fujii-kun and Yamamoto-kun,

Is it OK that there are differences between Yamamoto-kun's and Fujii-kun's results in order of 1%?
If the same data sets were used for calculating actuator efficiency, it should be completely same if calculation and fitting errors are negligible.

Is my understaing wrong?

takahiro.yamamoto - 15:46 Tuesday 30 May 2023 (25448)

We did the cross check by using independent two methods.
There is some possibility that both ways have a local maximum problem, an unexpected systematic, uncertainty of the transfer function model, etc.
The purpose of this cross check is that our results are not contaminated such problems.

Now we face the large uncertainty of the TF model of IM stages maybe coming from the coupling between DoFs.
And estimated value of actuator efficiency strongly depends on the treatment of model uncertainty (= estimation way
Model uncertainty issue is in discussion now. It should be solved in offline h(t) and O4b products...

From the view point of the check about stupid code bugs, we should also do the cross check with completely same method (but independent coding)
In this case, two results must be same within the precision of double or float.
But we haven't finished the preparation to do it yet.

takafumi.ushiba - 16:33 Tuesday 30 May 2023 (25450)

Thanks, I got it.

takahiro.yamamoto - 18:14 Tuesday 30 May 2023 (25451)

We did also pre-calibration measurements for the week of 5/30-6/6 after all mine works and the alignment work.
Measurement time was 13:35-14:45 JST (See also klog#25449 about the timeline of maintenance).

Measurement items are same as the post-calibration in this morning.
Measured transfer functions which are tagged as "0530/1336" and residual from the TF models are shown in Fig.1-6.

New parameters of h(t) reconstruction are as follows.
ETMX_IM: -1.4188e-14@10Hz => -1.4323e-14@10Hz (~1.0% change from the last week.)
ETMX_TM: -3.4567e-14@10Hz => -3.5137e-14@10Hz (~1.6% change from the last week.)
H_DARM: 254.301dB (5.1886e12) => 254.169dB (5.1103e12) (-1.5% change from the last week.)

By using these parameters, I re-computed transfer function models at the line frequencies
in order to trace the calibration lines.
These changes are accepted in observation.snap of K1CALCS.
Channel list is shown in Fig.7 and JGW-L2314962.

After the measurements and parameter update, we requested OBSERVATION to LSC_LOCK gaurdian
and enabled OBS_INTENT flag.
Finally, we backed to OBSERVATION at 14:55 JST.

Images attached to this comment