- We see a good coherence between the central building and 'LSC_DARM' with the 3 central lines. (see image)
- We've also seen these lines are very stable and narrow in time.
- It is possible to see the frequency 462.9Hz also in the magnometers in the CEB and it is coherent with the 'LSC_DARM' signal.
- We suspect that the source of this noise is the Turbo Pumps (TP) in the CEB: we put a piezo accelerometer in 3 different positions: on the TP, on the floor 1 meter far from the TP, 3 meters far from the TP (pictures) and the signal decrease with the distance (see reference plots). We temporarily used the ENV_BS_ACC_Z channel for the piezo accelerometer.
- ~8:30-9:26 UTC piezo on floor (~1m from TP)
- ~9:26-9:30 UTC piezo on TP
- ~9:30-9:34 UTC piezo ~3m from everything
- We believe these (~450Hz) lines are not present in the end buildings due to different TP (~400Hz) at those locations. That being said the coherence between LSC_DARM and ENV_SEIS_NEB_W,N' is negligible around 400Hz.
- 'Only' the line at 456.3Hz shows two sidebands (about 1.4Hz) in 'LSC_DARM'. Paolo Ruggi suspects that its source is the rotational mode of the mirror. We're working on demodulating the 'LSC_DARM' signal, to check the coherence with 'LSC_B1p_DC'.
Due to this result, we decided to go deeper in the study. We decided to use the accelerometer that usually is on the SR flange (channel ENV_SR_ACC_Z) changing his position and taking at least 5 minutes of good data (see the attached pictures)
In the central area there are two big turbo pumps, one under the SR-BS pipe (TP1) and another under the PR-BS pipe (TP2). The two turbo pumps are different: the first has a rotational frequency around 460Hz, the second one around 480Hz.
Below there is the list the time and the position of the accelerometer:
- untill 10.00 UTC: accelerometer in his nominal position (on SR flange)
- 10.05 - 10.11: on SR-BS pipe (good data only in few moments due to the interference of other activities around the SR)
- 10.15 - 10.21: on TP1 flange
- 10.24 - 10.28: on the floor SR-WS side at 2m from TP1 and 4m from TP2
- 10.31 - 10.35: on the floor SR-WS side at 4m from TP1and 1m from TP2
- 10.40 - 10.44: on TP2 flange
- 10.47 - 10.51: on the floor SR-WN side at 1m from TP1 and about 7m from TP2
- 10.55 - 10.59: on the floor SR-WS side at 10m from TP1 and about 15m from TP2
- From 13.04 again in his nominal position (on SR flange)
The results are:
- all the peaks of the comb are moving together and that confirms that the source is just one, TP1 (first and fourth plots)
- three peaks have a nice coherence with the dark fringe (third plot)
- all the accelerometers in the CEB have the comb in their signal (second plot)
- the accelerometer on the floor at 10m from TP1 shows peaks lower (about a factor 5) than when it was at 1m. The same behavior is shown by the accelerometer on the SR flange (fourth plot)
- the peak at 480Hz (produced by TP2) doesn't show a good coherence with DARM, doesn't show any comb and the peak is very low compared with the peaks of TP1 (third plot)
In the next future we will try to understand the coupling between the TP1 and the CEB switching off all the pumps and the air conditioners in the CEB (seismic coupling) and shielding the drivers of the turbo pumps (magnetic coupling)
