oli.patane@LIGO.ORG - posted 14:23, Tuesday 08 April 2025 - last comment - 15:32, Tuesday 08 April 2025(83818)
SR3 M1 SUS comparison between all DOFs
Jeff asked me to plot a comparison for SR3 M1 between all degrees of freedom comparing it in vacuum versus in air. I've plotted the last two measurements taken for SR3 from last August at the end of the OFI vent. One measurement was taken in air, and the other was taken in vacuum The pressure for the in vacuum measurement wasn't all the way down to our nominal, but as Jeff said in his alog at the time when we were running these measurements: "most of the molecules are out of the chamber that would contribute to buoyancy, so the SUS are at the position they will be in observing-level vacuum" (79513).
Non-image files attached to this report
Comments related to this report
Calling out the "interesting" off-diagonal elements:
D R I V E D O F
L T V R P Y
L -- nc nc meh eand YI
R T nc -- YI eand nc meh
E
S V meh YI -- meh nc YI
P
R VI esVI VI -- YI VI
D
O P esVI VI YI meh -- YI
F
Y YI nc nc nc nc --
Here's the legend to the matrix, in order of "interesting":
VI = Very Interesting (and unmodeled); very different between vac and air.
esVI = Modeled, but Still Very Interesting; very different between vac and air
YI = Yes, Interesting. DC response magnitude is a bit different between vac and air, but not by much and all the resonances show up at roughly the same magnitude.
meh = The resonant structure is different in magnitude, but probably just a difference in measurement coherence
eand = The cross coupling is expected, and not different between air and vac.
nd = Not Different (and unmodeled). The cross-coupling is there, but it doesn't change from air to vac.
I've bolded everything above "meh" to help guide the eye.
Recapping in a different way, because the plots are merged in a really funky order,
VI = L to R (pg 14),
T to P (pg 22),
Y to R (pg 33)
esVI = T to R (pg 16)
L to P (pg 20)
YI = L to Y (pg 28), Y to L (pg 27),
T to V (pg 12), V to T (pg 11),
V to P (pg 24),
P to R (pg 25),
Y to V (pg 31),
Y to P (pg 35)
What a mess!
The matrix of interesting changes is NOT symmetric across the diagonal
The matrix has unmodeled cross-coupling that *changes* between air and vac
For the elements that are supposed to be there, (like L to P / P to L and T to R / R to T), the cross coupling different between air and vacuum.
For some elements, the cross-coupling is *dramatically worse* at *vac* than it is at air.
Why is there yaw to roll coupling, and why is it changing between air and vacuum??
There's clearly more going on here than just OSEM sensor miscalibration that the Stanford team found with PR3 data in LHO:83605. These measurements are a mere 8 days apart!
The plan *was* to use SR3 as a proxy during the vent to test out the OSEM estimator algorithm they were using to improve yaw, but ... with this much different between air and vac, I'm not so sure the in-air SR3 measurements to inform an estimator to be used at vacuum.