J. Kissel I've taken new, more comprehensive B&K hammer response measurements of the H1SUSPRM and H1SUSPR3 cages, now that they have newly installed (what I'm calling) Venetian Baffles (see attached HAM2_NewBaffling_WithLabels.pdf for names of baffles) whose installation was finished last week LHO aLOG 39170. These baffles have pretty high-Q, low-frequency drum-head / longitudinal resonances (roughly aligned with ISI / IFO Y axis). PRM Upper: 42.38 & 46.75, 91.00 PRM Lower: 42.38 & 46.75, 75.62 PR3 Upper: 36.75, 75.6 PR3 Lower: 36.75, 83.12 My guess is that the lower frequency of the modes are the baffles bending in longitudinal in concert on the Venetian bracket, and the upper frequencies are their individual longitudinal modes. This mode-shape guess is based only on intuition, and that the lower frequency modes are seen in both upper and lower excitations. The cage's transverse modes appear to be relatively unaffected by the new baffles. I'm little surprised it hasn't stiffened up any of the transverse modes; oh well. These resonances have been identified by comparing against the history or cage resonance measurements for each of the SUS -- see the three pdfs: 2017-10-30_H1SUSPR3_CageResponse.pdf 2017-10-30_H1SUSPRM_CageResponse.pdf 2017-10-30_H1SUSPRMvsPR3_CageResonance_Comparison.pdf Note, also new with these measurements -- data out to 1.1 kHz. The former data is from LHO aLOG 6014 -- VA ON vs OFF data for H1SUSPRM and H1SUSPR3 LHO aLOG 8654 -- Former Cage Baffles on H1SUSPR3 Photos attached (and remaining HitLocations.pdf) are for historical reference for future repetition.
From Stephen and Norna
We (Stephen, Norna, Calum, Cormac) have done further experiments in the lab at Caltech to better understand the effect of the addition of the "Venetian blind" baffles ( D1700256 HSTS BAFFLE ASSY.PRM), on an HSTS and to help with the interpretation of the results seen at LHO.
A few caveats which should be noted:
a) We only have a bare structure - no vibration absorbers, hanging suspension, cables etc. attached. Also not as well dogged down as on site due to potential interference with baffle (our HSTS is not on a spacer). We have included baseline results displaying excitation of this structure without baffles mounted for comparison, see figures 2a and b described below.
b) We only used one baffle panel - so it was either attached at lower or upper position. See figure 1 for set-ups.
Basic findings
1) We show with and without damped baffle in upper position, exciting at top of structure in longitudinal (beam) direction and transverse. Basic conclusion, we do not see noticeable new resonances when exciting structure itself. See figures 2a and 2b
2) We only see extra low frequency ( ~ 40 Hz in our case) resonance introduced by baffle when directly hitting on the baffle. It is not seen when excitation is done to the structure itself. It is also only seen when hitting the baffle in its upper position, not in the lower position. See figure 3 for upper position results.
*We strongly suggest that if time permits, a test where the structure itself is hit in the longitudinal (beam) direction is done at LHO to see if this finding also holds for the PRM suspension now in situ.* From our experience we expect those low frequency peak(s) not to appear or least to be less prominent when the structure itself is excited.
3) The viton O-rings in the attachment units make a significant beneficial difference to the behaviour. Adding a baffle without viton introduces extra features which are suppressed or damped with the introduction of the viton.
See figures 4 (upper) and 5 (lower) baffle results.
4) The viton also adds some damping to the original structure resonances, apart from the first two flagpole resonances at 65 Hz (longitudinal) and 75 Hz (transverse) for our set-up. The dominantly torsional mode at ~160 Hz in our set-up shows some damping, as does the ~350 Hz feature. This can be seen particularly in the transverse results.
See figures 6 (upper) and 7 (lower).
5)We also did some investigations of different tightening levels corresponding to different levels of compression of viton O-rings within the two different flavors of attachment unit, D1700232 and D1700236. Basic result: the system is quite tolerant to different levels of tightening,with similar results over a range from hand tightened plus 1/4 turn to hand + full turn.
We will write this up more fully on the DCC at T1700473, including posting all data sets.
I have added one further set of comparison traces. In figure 8 we show the effect of including the damping O-rings in the baffle attachment units, where we are now comparing the results when hitting directly on the baffle in its upper position, rather than hitting the structure as shown in figure 4. We see again that the damping makes a significant beneficial difference.