Jeff, Sheila
Last night when Jeff brought the ISIs back after trips due to the first earthquake, he reset the target offsets as we have discussed doing from now on. This moved ETMX ISI RZ (yaw) by 16urad on stage 1, 12 urad on stage 2, RY (pit for TMS) by 3 urad on stage 1, and 2 urad on stage 2. This is at least the third time that we have reset the target values for ETMX, and the second time we have seen a large move. (25 urad seen in RZ in alogs 10596 10602)
If we had changes of less than a urad, either our dither alingment or WFS would be able to correct for that, if we had changes of less than about 5 urad we would still be able to find the baffle PDs easily to recover the alignment. 20 urad though is difficult to recover from. I (Sheila) would prefer waiting for T240s to settle than doing a random walk in alignment over 10s of urad.
We had a look at the other optics, and they don't seem to move that much. ITMX for example had no changes larger than 100nrad last night.
It seems like the solution proposed by Brian of keeping RZ and RY for the ETMs would solve this, but here is a plot to show what is going on.
We did a model restart of all the BSC-ISIs last Tuesday. By doing that, we restored the old target values stored into the snap files.
On ETMX, these old target values were huge compare to the natural position of the ISI. That explains why, when we reset the targets yesterday after the earthquake, the new location of the ISI is so different than last week.
If you compare the location of the ISI now with the location BEFORE the model restart, you can see that the difference is way smaller (a little big on RZ and Z though).