Virgo Logbook

A preliminary identification of the drum modes of the new mirrors was already tried by Michele, but this was based on limited data and some guesses. The experiment of changing the temperature of the ovens makes the identification easy for the input mirrors. About 3.5 days of data was used, from Thursday till the last lock of the 'weekend run'. On Thursday, the correction sent to the heaters was set to maximum for the WI and set to zero for the NI, while on Friday both corrections were set to an intermediate value. Fig 1 shows the effect on the temperature of the bottom flange (Em_TEtest*) and on the sensors installed on the safety structure (Em_TETO*). To analyze the data, the dark-fringe was first demodulated at the frequencies of several interesting peaks, down-sampled to 2 Hz and saved to disk, this takes about 10 hours. After this step, narrow-band spectrograms with mHz resolution can be calculated in about 10 seconds.

Fig 2 shows a very clear correlation between the thermometer of the NI and lines at 3964, 5672 and 5676 Hz. Similarly, fig 3 shows that lines at 3939, 5642, 5652 and 5660 should be attributed to the WI. The dip in frequency around 70 hours, which is not seen by the thermometer, is caused by the switch-off of the WI TCS.

Fig. 4 shows the behavior of the remaining modes, which can be clearly divided in two groups. Assuming Michele's asumption about the end mirrors is right, the lines at 4026 and 5707 Hz belong to the NE and those at 3969, 4015 and 5694 Hz to the WE. Note that the ones attributed to the WE show pretty large excursions in temperature compared to the NE, maybe due to a much higher absorption.

Fig 5 and 6 show the group-portraits of the butterfly and drum-mode families. I'll leave the explaining of the mode-splitting to the experts.