Page History

Attendees:

Announcements:

Agenda:

• AOB

Attendees: RL, SS, AH, KO, JB, JL, Mariano Cababie, Ray Bunker, Shilin Ray, BL, JI, SD, JY, WL, BZ, MK, SZ, ML

Announcements:

Agenda:

• Introductions of new people!
  • Mariano: COSINUS in Vienna
    • TES design that we're unsure about what to do with
  • Ray:
    • Background: DM direct detection, CDMS++
    • Led HEP QuantiSED project related to this work, now director of research at SNOLAB + Laurentian U part of G4CMP effort
  • Shilin:
    • Rising 2nd year Caltech grad student working with Sunil G.
    • Using G4CMP for qubit-based dark matter devices trying to develop with Stanford
• Office hours discussion
  • Mike needs to circle back with Eric, who has been looking at charge-surface interactions. Will talk on slack and/or in future meeting
  • Rafal O. had questions needing attention as well. 
  • Collecting fridge models is one of the LPS grant goals
    • Unsure if putting IN G4CMP or putting in separate repository
    • PNNL side has fridge model information worth collating
      • BlueFors LD400 (x2 set up differently with different cans, etc.)
    • FNAL/Oxford Proteox fridge has been simulated by Selby
    • JY: can start putting this on his plate if needed
    • Birgit: geometry implementations are Laurentian/SNOLAB responsibility
    • Plan for 2 weeks from now: anyone who has made chip or fridge geometries, please bring a single slide showing each geometry to give the Laurentian/SNOLAB people some context of where we're starting from for a geometry library
    • Mingyu: Leiden fridge geometry @ MIT (Harrington paper study)
  • Jesse:
    • Was approached to work on spin qubits, and a superconducting group approached him with some results
    • Superconducting qubit T1 dropout studies conversation
  • Selby:
    • Material package for newbies? (SLAC internal ++)
      • Mike's HEP Software Foundation Talk should have this – look on site
      • VIEWS '24 talk by Mike
      • Send Ben Loer's email to Selby (ben.loer@pnnl.gov)

      • https://docs.google.com/presentation/d/1aX2eHhikg8nZ8lnpKFrSXqpjZ3m2UL0Z8X2kB46PHBU/edit?slide=id.gd6c1fb53fc_0_0#slide=id.gd6c1fb53fc_0_0

      • Ryan: add a link to the RISQ workshop to the confluence (it's actually on Meetings & Workshops main page)
    • Running batches of jobs but no satisfying way to get output of jobs and extract info from them to analyze
      • Mike: do not need to consolidate everything into a giant root file. If you've got a root file per file with the same NTuple/TTree structure, then use TChain and that'll iterate through things
      • Mike no like S3DF's batch submission framework's pre-emption policy
      • No solution for jobs that got killed halfway through
      • Flush file very frequently and you may avoid some of this...
      • On each event, do a write to the root file and close event
      • Then at start of each event, open file for append, write data, and re-close
      • Sometimes, MOST of the 100's of jobs submitted get pre-empted
        • This is about priority – batch system kills running job if higher-priority job takes over node
        • Noah put in a hardware purchase with dedicated fraction of jobs can run at all times. Currently G4CMP doesn't have buy-in so we're piggybacking on SCDMS
      • Make NTuples from G4CMP that can then be later analyzed without having to redo the simulation
        • These are exactly TTrees and TBranches
• AOB

Attendees: RL, JB, CSW, SS, BL, SQD, ÉM, NT, JI, AH, SZ, BZ, IH, 

Announcements:

• TrackedFilmResponse making steady progress – estimating 2 weeks before it goes into code review

Agenda:

• Phonon interfacial physics (Caitlyn Stone-Whitehead) 
• AOB

Attendees: MK, CSW, RL, BL,JL, JB, MN, TN, SZ, SD, IH, AH, SS, ÉM, JY, AN, WL, IH, JI, JG

Announcements:

Agenda:

• Office-hours style question:
  • Caitlyn working on pulling and building TrackedFilmResponse – currently running into a runtime error that complains:
    • Mike recommendation: do not put setup scripts into .bashrc or .bash_profile unless you're sure you're never going to switch G4CMP versions
    • We think this might still be issues with environmental variables – possibly because on linux cluster and that can be tricky
    • Mike will create a zoom meeting today and send to Caitlyn, Ryan, Selby.
    • David: maybe we should have "common installation problems" page that outlines this – Ryan can make a confluence page for this
  • Abhiram: what were the additions to trackedFilmResponse?
    • Currently doing lit study on this: Eric's phonon modeling one and Ryan's threshold paper
    • What should I do here? Should I use trackedFilmResponse or current main branch?
      • Ryan's advice: use main branch for now to build G4CMP intuition and then migrate to trackedFilmResponse in the near-to-medium term
      • Mike's advice: since you're looking at a chip with multiple scales of feature sizes (ground plane, qubit island, etc.), trackedFilmResponse will eventually be necessary.
    • How do we do the quasiparticles response once we have the phonon hits? 
      • QuantumDeviceResponse (QDR) exists and can be used: (link)
  • Nolan:
    • Does the documentation for the phonon DOS exist? Where can I find that?
    • CSW: will put on confluence but for now link to that documentation is here: https://github.com/cstonewhitehead/G4CMP_PhononModeDOS/tree/master/PhononModeDOS_Tutorial
• Discussion/Questions: "How do we choose values in Kaplan QP?" (Selby)
  • KQP: what approximations are present in KaplanQP's implementation of QP tracking?
    • Mike: only reasonable if you have a quasiparticle trap. Not sensible with a ground plane
      • Since everything happens in 0 time, phonon that isn't going to create more QPs, might as well throw it in the trash
      • May be able to set lowQPLimit to zero and be okay, but may also trigger infinite loop
        • Need to test
      • End up throwing the quasiparticle away – it's dumped into the collected energy bin
      • Timing: KaplanQP processes things instantaneously
        • Phonons that are above a few meV – will do a lot of isotopic scattering (mode mixing) in the substrate (if larger)
        • Higher-energy phonons likely appear to travel close to the speed of sound, but not exactly, due to the isotopic scattering
      • High confidence in phonon transport
      • David: do we have diffuse scattering at sidewalls? Yes.
  • Temperature dependence on temperature dependence which affects QP density
    • Open Jira tickets
    • Temperature dependence of QP population will come out of trackedFilmResponse
    • Also, phonon thermalization: tracking phonons under assumption that they're athermal – more generally, higher energy than thermal bath
      • Do have code available in G4CMP to kill phonons below a certain energy (i.e. consistent with being part of Boltzmann Distribution for thermal energies)
      • Something fancier than just an energy threshold cut: can set a temperature, have it know the Boltzmann distribution, and then throw a dice roll to see if any given phonon should survive.
        • This is in KaplanQP
    • David: which branch is this on? It's in production, but in KaplanQP
    • Mike: have a G4CMPTrackLimiter but for charge tracks only – maybe could add a sub-process to do Boltzmann killing of phonon tracks in the substrate

• AOB

Attendees: JB, SS, AH, MK, JL, NT, BZ, SZ, ÉM, SD, SB, KO, BL, EM, 

Announcements: 

Agenda:

• Discussions of style guide prompted by questions from Ryan on TrackedFilmReponse
• Selby: questions on KaplanQP
  • Context: simulating MKID to study position-dependent phonon collection
  • Question: how do I think about the interaction between KaplanQP and the "simple" approach to phonon absorption and reflection
    • Info is in one of Ryan's emails that will be made into a new confluence page (link)
  • Philosophical question: how do we set the values in Kaplan QP?
    • Will get this on agenda for meeting 2+ weeks from now
  • KaplanQP lowQPLimit: hack used to prevent infinite loop of QPs emitting/reemitting phonons in KaplanQP.
  • Film absorption: Mike's homework hint: "quarter wave plate" – how should we think about the probability that a phonon can escape, given its wavelength relative to the film thickness?
• Israel: questions about electron/hole simulations in G4
  • Hitting issues where only see one or two valleys – will put together presentation with issues

• AOB