Summary: Frequencies excited by tapping HAM6 ISI flexures while it was vented match frequencies excited in DARM by external acoustic and shaker injections.  A prototype damping clamp damped the wire resonances. We suggest development of a permanent damping clamp for the ISI suspension wires and tuned dampers for the blade springs in the 350-450 and 850-1000 Hz bands.

Ambient acoustic sound produces noise in DARM that is only a factor of 2 below our current sensitivity in the 850-1000 Hz band and not much lower in the 350-450 Hz band and several higher frequency bands (Link).  We got a chance to study HAM6 during the vent that started last week.  It turned out to be difficult to excite and monitor the resonances that dominate in vacuum when we were vented. I believe that this is because the dominant vibration coupling path to the ISI table surface during a vent is acoustic and not through the flexures, as it is in vacuum.  External shakers did not produce peaks in the geophone signals that matched the peaks in DARM. I ended up attaching a small shaker magnetically to one of the blade springs, and monitored the frequency with an accelerometer mounted close to where the flexure attached to stage 2 (see photos in Figure1).

I thought at the outset that the high-acoustic coupling bands were likely associated with modes of the ISI flexure because they lined up with ISI resonance bands characteristic of all ISIs, and all external vibration would have to be funneled through these flexures to shake the table. Figure 2 shows that I excited a number of peaks in my in-vacuum accelerometer spectra by flicking one of the 3 ISI suspension wires (number 3). The evidence that it was the wire resonances and not some other resonances that I excited by flicking the wire is strengthened by the fact that these peaks were not evident when I damped the wire by clamping a piece of Viton to it. Since flicking amplitude is not well controlled, Figure 2 shows that I went back and forth several times between clamp-on and clamp-off states. Figure 3 shows that the lowest of the excited wire resonances, 753 Hz, matches the frequency of a peak in DARM that is excited by shaking HAM6 with a shaker. Note that the peak at this frequency is not the largest feature in DARM in this region, the feature between 850-100 Hz is larger, but, resonances of the wires do seem to account for the coupling at higher frequencies.  Figure 4, for example, shows that at least two of the high frequency resonances excited by flicking the wire line up with peaks in DARM that were excited acoustically. 

Figure 5 shows that tapping on the blade spring rather than the wire produces peaks that line up with acoustically excited peaks in DARM, including the largest acoustically excited peaks in the 850-1000 Hz band. I tried clamping a small piece of Viton to the blade spring and it was not nearly as successful as the wire damping (a factor of 2 at most). Possibly a larger piece would help, but I am inclined to suggest a tuned damper like the one for the low frequency blade spring mode.

Thus I would suggest that we develop a wire clamp damper like the prototype that I made to damp the wire modes, and that we develop tuned dampers for the 350-450 Hz band and, most importantly, for the 850-1000 Hz band.

Robert S., Hugh R., Katie Banowetz, Nutsinee K.