1. SRM beam position was off last week by about 3/8 inch.
When the beam is properly centered on SRM, the beam position on SRM is about 3/8" towards -X than we thought last week.
Beam centering on SRM horizontally is almost impossible to do by looking from outside HAM5, and instead it seems as if this was done in the past week by centering on the baffle hole on the back of the SRM.
Today we sent JeffK into the tube between HAM5 and HAM4 and have him look at the beam on SRM while I was holding the card right in front of the SRM. Doing it this way, it turns out that the beam is about 3/8" inch off centered on the back baffle hole in -X direction.
My guess is that the original centering was not done from the front of the SRM. I also wonder if the cage position itself is off.
TJ repositioned the baffle so the beam is centered on the back baffle, and Gerardo moved OFI such that the beam at least clears the OFI (he'll fine tune tomorrow).
2. Septum window was rotated by 120 degrees counter-clockwise and the beam was great in PIT, but of course not great in YAW.
Gerardo and Travis rotated the septum plate by 120 degrees counter-clockwise so that the beam height becomes good on OM1.
The thickest side went to South, which moved the beam further in -X direction.
The beam was blocked by the fast shutter. When I moved FS to the side, the beam was hitting somewhat to the -X side of OM1 -X damping plate adjuster screw, meaning the height was right but it was off to the side by about 36mm in -X direction.
Table below shows Koji's measurement of beam position on OM1 last week (alog 39460), beam position changes which should have been caused by septum rotation assuming that the wedge angle is 0.89 deg (measured by Koji, instead of 0.75 deg derived from the thickness measurement) (alog 13399), and the beam position measured today. If everything is right "in theory" and "measured" columns should agree. At least they don't look totally wrong.
| Last week beam pos | Change caused by Septum rotation | Change caused by good centering on SRM | Beam pos today in theory | Beam pos today measured | |
| YAW on OM1 | -11.5 mm | -20.3 mm | -9.5 mm (3/8") | -41.3 | about -36 mm |
| PIT on OM1 | +9.5 mm | -11.8 mm | NA | -2.3 | about right |
3. Rotating Septum by 180 degrees will bring the beam to a better position
Assuming that SRM is/was at the correct location (we need to check), the next question is what to do.
Unfortunately, I cannot simply move OM1 and be done with that, it seems like the beam from OM1 to OM2 will be either blocked by OM3 or very close to it. I can move OM2 but that will leave no room for the fast shutter as the OM1-OM2 will be much closer to OM2-OM3 line than before.
We can rotate the septum window by 180 degrees so the thickest side comes to the north. Assuming 0.89 deg wedge, the beam deflection is 7 mrad, so 180 deg rotation will give us 14mrad change in deflection.
The distance from the septum to OM1 is 1930mm, so this will result in the horizontal shift of 27mm without height change. The beam will be at X=-36+27 = -9mm on OM1 horizontally. Hopefully this will allow us to make room for the fast shutter without moving OM2 sideways.
Although this is apparently resolved, I still went through the exercise that Sheila suggested yesterday of figuring out the SRM beam position using A2L numbers from alog 30395, from October 2016.
Recall that for the case we actuate in angle on the lowest stage of an optic (true for SRM, not true for test masses), we can calculate the balancing coefficient alpha (original definition in 40m elog 2863) using the form derived in LHO alog 31402, alpha = (A2L * L_eul) / (A2A * A_eul).
The displacement between the center of rotation and the center of the optic is given by (alpha * conversionFactor), where the conversion factor is defined in 40m elog 2863, and for the PRM (and is the same for SRM) is calculated by Kiwamu in LHO alog 13765 as (42.2 mm / alpha).
The P2L coefficient for SRM from alog 30395 was -1.0, and the Y2L coefficient was -1.2. This means that the beam on the SRM was off by 2.0 mm in pitch and 2.4 mm in yaw when it was measured on 10 Oct 2016.
In contrast, there was concern yesterday that the beam on the SRM may have been off by about 3/8 inch, which corresponds to 9.5 mm. It sounds like today they may no longer think the beam was off so much, which is consistent with the Oct2016 measurement of the beam not being off by too much.