Ushiba, Hirose, K. Tanaka

## Abstract

We first tried to see if we could reproduce yesterday's results. We engaged ASC in the interferometer, used IMMT1 to the same beam position on the POPFORWARD QPD as yesterday, and got the same results.

We then tried to get an initial alignment of PRFPMI with the IMMT1 alignment set to a good alignment for PRFPMI from the beginning. As a result, we locked the PRFPMI successfully after this alignment. At this time, the transmitted power of both arms was low, but when We turned on ASC, the transmitted power recovered drastically and returned to yesterday's value.

We have already implemented this set of operations in ASC_LOCK guardian (klog24688).

Since the alignment of IMMT1 has yet to be optimal, simply because it is where we got better, we looked for the optimal alignment of IMMT1. However, we could not find it because one segment of POPFORWARD's QPD was satiated during this optimization. Therefore, to find the maximum alignment, we need to reduce the power of entering POP FORWARD QPDs by using an ND filter or centering the POPFORWARD QPDs. Alternatively, we must improve the alignment to IMMT1 by another method, such as BPC to PR2.

## What we did

• We locked PRFPMI with RF lock and engaged ASC for all DoFs of PRFPMI. Then, we fed back the POP FORWARD QPD PIT and YAW signals to IMMT1 PIT and YAW alignment so that the each beam position on the both POP FORWARD QPD1 and QPD2 was where they would have been when the PRFPMI alignment was good yesterday. This control seems to work well. When the beam position on the QPD was about the same as yesterday, we found that the transmitted power imbalance and reflected power were about the same as yesterday. This result means that these ASCs can reproduce the alignment of PRFPMI.
• After the initial alignment, ASC in PRFPMI is done by adjusting PRMI by ADS, ITM by BPC, and ADS for PIT in ETM. Then, the ADS of ETM is turned off, and ASC by WFS is used to adjust ETM. In this state, to get a good alignment of IMMT1, we need to move more than 100 urad, which is very time-consuming because we have to make the adjustment slower than the ADS of PRMI.

• For this reason, we checked to see if it would be possible to perform the entire initial alignment in the state of IMMT1 alignment at the time of good alignment of PRFPMI in advance. It would save time by eliminating the need to make significant adjustments to IMMT1 after PRFPMI lock if possible. Specifically, when aligning the IR beam for the Xarm, we aligned IMMT1 to be the beam position on the QPD of POPFORWARD, as described earlier. After that, we performed normal alignment for the other DOFs and locked PRFPMI with RF. Then, when we turned on the ASC of ETM by WFS from the beginning without ADS of ETM, the locking of PRFPMI was realized in a good alignment state.

• Hirose-san implemented this operation to ASC_LOCK guardian.

• Since the alignment of IMMT1 has yet to be optimal, simply because it is where we got better, we looked for the optimal alignment of IMMT1. However, we could not find it because one segment of POPFORWARD's QPD was satiated during this optimization. Therefore, to find the maximum alignment, we need to reduce the power of entering POP FORWARD QPDs by using an ND filter or centering the POPFORWARD QPDs. Alternatively, we must improve the alignment to IMMT1 by another method, such as BPC to PR2.