J. Kissel, R. Short, B. Weaver

Before, during, and after today's locking, or "muting," of the H1SUSRM1 and H1SUSRM2 (both HTTSs in HAM1) blade springs with there own vertical EQ stops, ala E2200181 (see blow-by-blow procedure of how it went down for us in LHO:63552), Ryan and I gathered more transfer functions of the REFL WFS A and B's DC "low frequency" QPD-like PITCH output w.r.t. the table top intertial sensors (the TT L4Cs), replicating the measurement from LHO:63474, this time in air.

Attached are the results, over three frequency spans:
    (Attachment 1) 3 to 50 Hz -- the region where we've seen and expected to reduce the TT motion to WFS DC Pitch coupling like LLO's coupling in LLO:59677.
    (Attachment 2) A broader frequency range from 3 Hz up to 100 Hz.
    (Attachment 3) A zoom in on the 50 to 100 Hz region.

In (1) -- and this is the format of all the attachments -- I show for WFS A DC on the left, and WFS B DC on the right,
    - BLACK : The "before" measurement, taken this morning prior to any locking of blades -- this matches the in-vacuum coupling seen in LHO:63474
    - GREEN: After locking down RM1's blades ONLY to the ~0.5 mm deflection levels. Here, interestingly, we remove almost all of the broad-band coherence between Z table motion and the WFS DC Pitch signal, and leave only (presumably RM2's) blades. After investigating as much as we could mechanically, we couldn't find much of anything that's markedly different between RM1 and RM2. We can wave our hands all day about eddy current damping magnets, but 
        * The window to view the spacing between eddy current damping magnets and the nearby magnetic current damping providing device (not sure of the material) is too small and awkward to make a good visual comparison, and
        * Given that the coupling is so identical between H1 and L1 I would be surprised if this was something "human adjustable" (they didn't take this "in between" measurement identifying *which* of the two dominated the coupling)
    - BLUE: After locking down both RM1 and RM2 blades. One can see that the last of the coherence in this region -- at the free-swinging vertical blade mode frequency of 6 Hz disappears.
    - RED: After pushing down the RM1 blades a bit further to the recommended ~1 mm. We so no change, and we continue to see no coherence in this frequency region.

Great success!

However, I made the mistake of looking at this transfer function over a broader frequency region.

We found immediately, when comparing "No Blades Locked" to "Only RM1 Blades Locked" that the frequency region between 65 and 100 Hz became *more* coherent, rather than *less* coherent, and in a complicated mix of sharp frequency dependent features mixed with a flat-ish increase in magnitude. One can see -- particularly in (Attachment 3) -- that 
    * there was coherence in this region prior to locking the blades, and 
    * locking the blades has made that coupling strong / more coherent.
This feature must *also* be related to RM1 in particular, since it 
    (a) got worse after initially locking RM1 at the ~0.5mm deflection, 
    (b) didn't change after locking RM2, and
but interestingly, it didn't change after locking RM1 further.

Not expecting this, we weren't prepared for it:
    (A) Just like the low-frequency broad-band coupling, we don't have an idea as to what's causing this 65-80 Hz coupling,
    (B) We there for weren't armed with a solution, and
    (C) We expect that at *this* high a frequency, we can sufficiently roll-off this sensor noise in the CHARD loops enough that it won't impact CHARD and therefore DARM,
so with that we declare 
    - "VICTORY...ish... right?" 
    - "we'll figure that out later, if we need to" and 
    - to move on with the rest of the tasks in chamber. 

A guess -- not the only, and certainly not the best guess -- is that it's some feature that *was* vertically isolated out of this transfer function and now is no longer vertically isolated out. 
We again pointed fingers at the eddy current damping, but it could also be a shorting of the squashed O-ring that serves as a Q reducer between the two metal components that form the blade.
But really we don't know, and we've convinced ourselves that we don't need to care.

Details of the measurement methods and a few photos to come in the comments.