= External Review for interferometer alignment controls (August 2021) = * August 24, 2021 22:00-0:00 (JST) https://www.timeanddate.com/worldclock/fixedtime.html?iso=20210824T1300 * Zoom3: https://us02web.zoom.us/j/5045179604 * Minutes for discussions: https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=13224 == Purpose == From February 25 to April 21, 2020, KAGRA performed its first observing run for 4 weeks with power-recycled Fabry-Perot-Michelson (PRFPMI) congifuration. The sensitivity during the April run with GEO600, dubbed O3GK, was at the binary neutron star range of 0.5-0.7 Mpc. From July 13 to October 13, we had an intense commissioning period to try locking the full resonant sideband extraction (RSE) interferometer. We had many achievements during that period, but we have never achieved the full RSE lock yet. During the O3GK run and the RSE trial, we faced several issues related to the interferometer alignment. The sensitivity and the stability of KAGRA depended very much on the alignment, and the best sensitivity could be achieved only when the expert aligned the interferometer manually. Almost no alignment sensing and controls (ASC) loops where closed, except for a few dither alignment loops in the power recycling cavity (PRC). The purpose of this External Review is to receive comments from LIGO/Virgo interferometer experts on our plans to improve such situations. We would like to know if the issues we are trying to solve are reasonable, and if our plans are sound. We would also like to find out what are the issues we haven't identified yet. Finally, we would like to prioritize the works in the order of importance for a stable full RSE lock with ASC. == Agenda == 1. [[#agenda1|Daily alignment of the interferometer]] (Yuta Michimura, [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=13198|JGW-G2113198]]) * summary of O3-RSE trial situation and our improvement plans 1. [[#agenda2|Commissioning and simulations for alignment sensing and control]] (Yuta Michimura, [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=13198|JGW-G2113198]]) * summary of current situation and plans for commissioning towards O4, focusing on global controls using wave front sensors (WFS) and QPDs 1. [[#agenda3|Input mode cleaner alignment sensing and control report]] (Kenta Tanaka [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=13205|JGW-G2113205]], Chiaki Hirose [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=13211|JGW-G2113211]]) * status report from the site works and simulation activities == Basics of KAGRA interferometer == * f1=16.88 MHz (PRC+SRC), f2= 45.02 MHz (PRC), f2-f1=28.13 MHz * [[LCGT/subgroup/ifo/MIF/OptParam|List of optical parameters]] * [[https://doi.org/10.1103/physrevd.88.043007|PRD 88, 043007 (2013)]] == References == * Summary of Alignment Scheme of KAGRA in O3GK [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=12594|JGW-T2112594]] * Definitions for the X arm commissioning [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=8343|JGW-T1808343]] * Definitions for the DRMI commissioning [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=9573|JGW-T1909573]] ---------- <> == 1. Daily alignment of the interferometer == === 1-1. Summary of the current status === The alignment scheme used for O3GK and RSE trial is summarized graphically in [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=12594|JGW-T2112594]]. We started the daily alignment from the Xarm using the green beam. The Xarm is locked with green, and ITMX and ETMX alignments were tuned to center the beam on them, by monitoring camera (Tcam) images of the HR surfaces by our eyes. PR3 dither alignment was also done at this step to maximize the green transmission. The green injection from the back of PR2 was also tuned manually using a steering mirror (POP POM) at the back of PR2 occasionally. We then proceeded to align IR beam to the Xarm by locking the Xarm with IR, and maximizing the IR transmission with IMMT2 and PR2 dither alignment. Next, we aligned PRMI by maximizing POP90 with PRM and BS dither alignment. Somehow, when we do the Yarm alignment before PRMI, MICH fringe was not good. We then aligned Yarm with ETMY dither alignment. By monitoring camera (Tcam) images of the HR surface of ETMY by our eyes, ITMY alignment was also tuned occasionally to center the beam on ETMY. When ITMY was moved, we went back to PRMI alignment again. Note that beam position on ITMY is not checked in this scheme. Then, Yarm green beam is aligned by maximizing green transmission with SR3 dither alignment. The green injection from the back of SR2 was also tuned manually using a steering mirror (POS POM) at the back of SR2 occasionally. And Finally, the single bounce beam from ITMY is aligned to OMC by maximizing OMC transmission with OMMT2 and OMC dither alignment. Such alignment scheme was not fully automated and the interferometer alignment was not reproducible. Towards fully automated reproducible alignment, we are planninng following modifications. === 1-2. Known issues and our plans (highly uncertain) === * PRMI alignment was done before Yarm alignment * Unclear reasons. Could be related to the beam spot in ITMY. Further investigation necessary. * Beam positioning on SRM was not considered so far * We might be able to use OMMT1 transmission to monitor the beam position. We need investigation for available space. We might suffer from back scattering since we don't have an OFI before OMMT1. * Dither and SRM HR camera are also candidates, but HR camera might not be useful after mid-baffle installation. === 1-3. Known issues and our plans (less uncertain) === * IMMT1 trans QPDs were not reliable since they were on the same pylon table with ISS setup with heavy enclosure. * ISS setup will be moved to a different table * Beam centering on ITMs and ETMs was done by our eyes. * We made a script to give the position of the beam (center of the brightness) in the monitoring camera (Tcam) images of the HR surfaces. We are thinking of using this to center the beam on ITMs and ETMs automatically. * We will also restore baffle PDs on narrow angle baffles. * Actuators for steering mirrors for green steering mirrors were not included in the automation * Replace two steering mirrors with piezo and picomotor actuated ones. Use piezo for daily alignment and picomotor for off-loading the piezo feedback. ---------- <> == 2. Commissioning and simulations for alignment sensing and control == === 2-1. Commissioning status === So far, WFS and QPD loops for IMC, PRMI and FPMI are mostly closed. For PRFPMI, we didn't have much time for ASC commissioning. Also, we had ugly beam shape at POP and POP WFS were not commissioned at all so far. TRX QPDs had strange pitch and yaw coupling which needs further investigations. Achievements for different interferometer configurations are summarized below. Note that most of the loops indicated "closed" were "once closed" and were not stable enough to implement them in the guardian. Even if the loop was once closed, it was not reproducible. We suspect that linear ranges are too small and/or zero-crossing points are dependent on the interferometer alignment (for example, simulated WFS signals are distorted if we include inhomogeneity of ITMs [[https://journals.aps.org/prd/abstract/10.1103/PhysRevD.100.082005|PhysRevD.100.082005]]). We hope we can improve the situation once we establish reproducible daily alignment procedure. * IMC * IMC REFL WFS to control IP1 and MCe * UGF ~0.1 Hz * IMC TRANS DC QPD to control MCo * UGF ~0.05 Hz * https://klog.icrr.u-tokyo.ac.jp/osl/?r=13916 * IMMT1T were not used because of scattering lights when PRFPMI is RF locked * IMC ASC loops were turned off during O3GK due to excess noise * WFS and QPDs for the main interferometer * REFL and AS are commissioned relatively well and WFS DC centering loops are closed. * POP beam was too dirty and WFS are taken out * https://klog.icrr.u-tokyo.ac.jp/osl/?r=11506 * TRX QPDs had strange pitch and yaw coupling * https://klog.icrr.u-tokyo.ac.jp/osl/?r=12019 * Xarm * REFL RF17 to control SOFT/HARD closed * UGF ~50 mHz for SOFT and ~20 mHz for HARD in yaw, ~60 mHz for both HARD and SOFT in pitch * https://klog.icrr.u-tokyo.ac.jp/osl/?r=11543 * https://klog.icrr.u-tokyo.ac.jp/osl/?r=11537 * https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=10996 * Sensing matrix measurement with POP, AS, TRX not done * Yarm * TRY QPD to control SOFT closed * https://klog.icrr.u-tokyo.ac.jp/osl/?r=12325 * PRMI * REFL RF45 to control PRM and PR3 closed * UGF ~0.1 Hz * https://klog.icrr.u-tokyo.ac.jp/osl/?r=12488 * AS RF28 to control BS closed * https://klog.icrr.u-tokyo.ac.jp/osl/?r=12445 * PRMI ASC sensing matrix * https://klog.icrr.u-tokyo.ac.jp/osl/?r=14019 * DRMI * BS and PR3 dither loops using POP90 closed. * https://klog.icrr.u-tokyo.ac.jp/osl/?r=15165 * FPMI * REFL RF17 to control Differential ETMs, Common ETMs closed * https://klog.icrr.u-tokyo.ac.jp/osl/?r=11885 * https://klog.icrr.u-tokyo.ac.jp/osl/?r=11997 * AS RF28 Q to control BS closed * UGF ~30 mHz for yaw, ~80mHz for pitch * https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=11061 * https://klog.icrr.u-tokyo.ac.jp/osl/?r=12156 * PRFPMI * AS RF to control DHARD (DETM) almost closed * https://klog.icrr.u-tokyo.ac.jp/osl/?r=14027 * Sensing matrix measurements for PRM and PR3 * https://klog.icrr.u-tokyo.ac.jp/osl/?r=12802 === 2-2. Simulation status === * Optickle * FPMI, PRFPMI, SRFPMI, DRFPMI (BRSE) sensing matrix simulations with imperfections done [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=10359|JGW-T1910359]] * PRMI, DRMI simulations, implementation of f2-f1 signals not yet * FINESSE * For LSC without mirror maps, GUI for simulations with any interferometer configurations done [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=12132|JGW-T2012132]] * Working on ASC integration * Also input mode cleaner (IMC) ASC simulations and analytical calculations on going independently * We are thinking of focusing on FINESSE simulations to implement birefringence effects. [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=12222|JGW-G2012222]] === 2-3. Strategies for commissioning === Commissioning towards O3 and RSE trial mostly focused on length sensing and control, and noise hunting towards achieving the 1 Mpc goal. Stability of the lock was not considered seriously and little time was occasionally allocated for ASC, when there was some spare time. For the commissioning towards O4, we should spend more time for ASC to achieve a stable lock. We also require the suspension systems to provide user friendly alignment actuators with clear and calibrated signal paths. The alignment fluctuation of each suspension needs to be settled down to respective requirements (summarized below) with local controls. These efforts are coordinated together with the suspension commissioning team, and suspension experts are expected to be present at the control room during the interferometer commissioning period. * ITMs and ETMs (Type-A): <240 um/sec (for arm green locking), <0.44 um/sec (for central part locking), <0.1 urad RMS * BS (Type-B): <1.6 um/sec, <1 urad RMS * SRs (Type-B): <0.44 um/sec, <1 urad RMS * PRs (Type-Bp): <7.3 um/sec, <1 urad RMS * References: * K. Okutomi PhD thesis (Table 3.5) [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=9347|JGW-P1809347]] * T. Sekiguchi PhD thesis (Table 5.2) [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/DocDB/ShowDocument?docid=4155|JGW-P1504155]] * Y. Michimura et al (Section 6) [[https://doi.org/10.1088/1361-6382/aa90e3|Classical and Quantum Gravity 34, 225001 (2017)]] Our planned timeline and targets for each step are summarized below (timeline based on [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=9209|JGW-E1809209]]). Full automation of the lock acquisition and daily alignment process should also be established at each stage. * IMC (May - August 2021) * Measure the sensing matrix and compare it with analytical calculations and simulations * Laser power throughput of more than 90%. * Hold the lock of IMC with ASC for a duration of longer than 2 hours continuously * Make sure that suspension alignment drift is smaller and the length noise is smaller when the ASC loops are on. * Xarm (mid-October 2021 - mid-January 2022; ~12 weeks) * Find out the reasons for ugly POP beam * Check if REFL WFS and POP WFS give consistent sensing matrices with simulations * Align the IR and green beams to Xarm and check the positions of mid-sized baffles in PR * Yarm (Januray 2021 - February 2022; ~6 weeks) * Find out the reasons for ugly POP beam * Check if REFL WFS and POP WFS give consistent sensing matrices with simulations * Align the IR and green beams to Yarm and check the positions of mid-sized baffles in SR * FPMI (mid-January - mid-February 2022; ~4 weeks) * Center area evacuation (mid-February - mid-March 2022; ~4weeks) * Before evacuation, we at least want to check PRC/SRC Gouy phase/length, possible beam clipping (for POP), and decide SRM transmission (T=0% or T=30% 2-inch mirror) * Note that MICH could be locked in air, but we have never tried to lock PRMI in air * FPMI (March - mid-May 2022; ~9 weeks) * Investigate TRX pitch and yaw coupling issue * Check if REFL WFS, POP WFS, AS WFS and TRX QPD give consistent sensing matrices with simulations * Check the polarization content at each port when Xarm is locked/unlocked, cross check with FINESSE (birefringence) model * Mode matching ratio (including mis-alignment) of more than 90% for both arms * We adjust PR2 and PR3 positions at this stage? * Hold the lock of Xarm and Yarm in IR with ASC for a duration of longer than 2 hours continuously * SOFT and HARD controlled with REFL/AS and TRX/TRY * Make sure that suspension alignment drift is smaller when the ASC loops are on * Check the polarization content at each port, mode content at AS, cross check with FINESSE (birefringence) model * Compare them with a simple Michelson configuration to see Lawrence effect * OMC alignment commissioning * Hold the lock of FPMI in IR with ASC for a duration of longer than 2 hours continuously * CHARD controlled with REFL, DHARD controlled with AS * CSOFT and CHARD controlled with TRX and TRY * BS controlled with AS RF28 * PRMI (March 2022; ~2 weeks) * Check PRC Gouy phase and length (inconsistent results so far) * Hold the lock of PRMI in 1f or 3f signals with ASC for a duration of longer than 2 hours continuously * PR mirrors controlled with POP * BS controlled with AS RF28 * Make sure that suspension alignment drift is smaller when the ASC loops are on * PRFPMI (March - June 2022; ~13 weeks) * Hold the lock of PRFPMI with ASC for a duration of longer than 2 hours continuously * CHARD controlled with REFL, DHARD controlled with AS * CSOFT and CHARD controlled with TRX and TRY * PR mirrors controlled with POP * BS controlled with AS RF28 * Make sure that suspension alignment drift is smaller when the ASC loops are on * Towards DR * Check SRC Gouy phase and length (inconsistent results so far) * Examine the mode-hop criteria * Which mirror affects most and how much angular motions are tolerable ---------- <> == 3. Input mode cleaner alignment sensing and control report == Status of the site work and simulation activities towards achieving the IMC ASC goals discussed above will be reported by Kenta Tanaka and Chiaki Hirose. * Kenta Tanaka: IMC ASC current status [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=13205|JGW-G2113205]] * Chiaki Hirose: IMC ASC simulation [[https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=13211|JGW-G2113211]]