Summary: the barrel of the EY baffle is directly observed to vibrate during a wind storm at the 3.8 Hz frequency associated with scattering noise in DARM, strengthening the case against the baffle. Injections in the regions of the ETMY, ITMY and ETMX baffles produced noise in DARM when the broad-band vibration level around 4 Hz reached 3, 3 and 6 times the background vibration level, respectively. Noise was not seen at ITMX, but the other 3 should be damped.
In the spring I showed that light modulation in movies of the Manifold/Cryopump Baffle (cryobaffle), made during an injected impulse, persist as long as the noise in DARM (https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=56508 ). This points strongly towards the cryobaffle as the scattering source because not many structures would have the Q of about 1000 seen in DARM. Anamaria and others at LLO have shown that shaking near 4 Hz at the locations of 3 of the 4 cryo-baffles produce noise in DARM (links in reference above). But, until now, we haven’t been able to show that the barrel of the cryobaffle is resonanting at the ~4 Hz frequencies of the noise source in DARM. I tried using the laser vibrometer, but I couldn’t get enough light back from the baffle for it to work.
Figure 1 shows that the I was able to detect a resonance at 3.8 Hz in modulation of light from the cryobaffle during the broad-band increase in vibration that occurred during the Labor Day wind storm. In addition, a second, lower-Q resonance at 7.2 Hz is evident, also a frequency that produces noise in DARM (EY injection frequencies that produce noise in DARM are shown in Figure 2, first page).
Figure 1 also shows that I did not detect these resonances in light from the P-Cal baffle edges (I checked 3 locations). However, I do not consider this absence of detection to be strong evidence against the P-Cal periscope as being the source. This is because not all regions of the cryo-baffle produce spectra with 3.8 Hz modulation evident. The trick is to look for a region that is flashing in the movie, and take a spectrum just of that region. It seems as if there is a fairly low probability that the diffraction pattern cast by a particular region of the cryobaffle will have a deep, sharp-edged minimum at the angle of the camera.
On the other hand, the failure to observe the 3.8 Hz resonance from the P-Cal periscope, along with the observation at LLO and LHO of a ~4 Hz scattering source at ITMY, where there is no P-Cal periscope, together make it unlikely that the P-Cal periscope is the source.
Anamaria and Valera made the important discovery that ~4 Hz injections near the cryobaffle for ITMY at the LLO corner station produced significant noise in DARM (https://alog.ligo-la.caltech.edu/aLOG/index.php?callRep=53364 ). This is in addition to the locations near the cryobaffles at ETMX and ETMY. I followed up on the LLO work by looking for noise production at each of our 4 cryobaffles. Figure 2 shows that I saw noise from the same locations and with the same ordering in importance as LLO: ETMY, ITMY, ETMX and no noise seen at ITMX. The high-Q dominant resonances at the 3 problematic locations were approximately 3.8, 4.0 and 4.1 Hz, respectively.
I used broad-band shaking injections (or wind) in order to estimate the minimum vibration increase in the 4 Hz region needed to produce scattering noise in DARM (at O3b sensitivity). Figure 3 shows that I saw noise in DARM with increases in vibration over the background of 3x, 3x and 6x for ETMY, ITMY and ETMX locations respectively. I note that the velocity can be increased by the microseimic peak, which was low during these measurments. I think this means all 3 need to be fixed – a damping test is now beginning at LLO.
