JEF meeting, November 4, 2022

From GlueXWiki
Jump to: navigation, search

Meeting time and location:

  • 9:30 am EST (JLab time) remote only
  • Zoom meeting ID: 161 050 9142 Passcode: 004965

Agenda

  • Preparation for the FCAL2 installation (coordination with Tim's group) -- Sasha
  • Light Monitoring System -- Arshak (SORRY, I probably will not be able to attend meeting in person, because of my owl shift)
    • 2_dimensional plot for plexi uniformity tests, I have scanned 5 lines {X0(center); X0(+-)20cm, X0(+-)39cm; Y0(center); Y0(+-)10cm, Y0(+-)19cm} for each X(80cm) and Y(40cm) directions. From this plot it can be easily seen that on the corners and edges the light amount is higher than on the central part, for the corners part the issue is solved by just removing the fibers which was attached exactly on the corners, for the edges still the plan is to attach optical attenuator to reduce the light amount https://userweb.jlab.org/~aasatury/Fcal_LMS/2d_1.gif
    • Cosmic tests setup for cross-talks
    • To check the cross-talks between modules, it's been constructed 3x3 matrix from PBWO4 blocks, also for this tests i made the same shape plexi as the actual one with smaller dimensions (62x62mm^2) which will cover the front face of the prototype. by the following link it is only the setup pictures. The plan is: 1) check with cosmics, 2) develop LED+Fiber setup (respectively to small dimensions) and trigger from LED, and final test if we will still need it is 3) move the setup to the hall and test it under beam conditions. I'll keep you updated as soon as there will be some good results. I will need to calibrate the modules, with more statistics to be able to continue cross talk tests see picture "cosmic_results.png" under the following link. https://userweb.jlab.org/~aasatury/Fcal_LMS/cosmics/

Minutes

Attendees: Olga, Sasha, Phil, Zisis, Tegan, Azizah, Axel, Jon, Joerg, Tim B., S.T.

  • Azizah presented an update on cluster classification for FCAL-2 using TMVA
    • Shower reconstruction using default algorithm
    • Training using simulated ω→π+π-π0 events
      • Several training variables show clear differences between signal and background events
      • Background timing slower for lead tungstate compared to LG
    • Tried several different ML algorithms: MLP, BDT, HMatrix, Fisher
      • Want ROC curve to be as close to the right top corner as possible
        • MLP and BDT work better than HMatrix and Fisher
      • No indication of over-training
    • Looked at insert region and outer LG region separately and together
    • Next step look at performance for η channels
  • Received 20 more crystals from physics division
    • Expect 45 by end of the year + 75 more on order
    • Wrapping by detector support group continues
  • 1524 light guides glued, wrapping continues
  • Brazing of brass strap assemblies has resumed, 250 remaining
  • Total number of fabricated modules = 1318, goal is 1400 by the end of the year
  • Need to dissasemble/refurbish COMCAL
  • Have regular FCAL2 installation meetings
    • Discussed cabling, patch panels for insert
    • Discussed possible problems with LG blocks (radiation damage, holes in wrapping, ...)
      • Many HV set at maximum value --> ~400 PMTs need to be replaced?
    • Sasha put together extensive task list
      • Possible tasks for students in spring/summer of next year, estimate 11 students (3 months each)
  • GWU to work on PMT divider tests
  • Detector support group will continue their participation
  • Arshak continuing to check LMS uniformity and starting to work on checking cross talk