iFAST 5.3 (Slow Extraction) - December Meeting

  • Giovanni Iadarola
  • Vladimir Nagaslaev
  • +29


  • Simulation work focuses on loss reduction and spill quality
  • Need to track with crystal and through matter (PyCollimate being used and work ongoing to clean it and make it GPU-ready)
  • Interest in accurate phase-space description for TL tracking
  • Looking at Physics Informed Neural Networks for simple tracking
  • Not so easy to get beam time
  • Using MapTrack (SPS) → moving to Xsuite


  • Need to do tracking in all six dimensions, can’t approximate
  • Easier to get beam time
  • Beam dump allows operation in parallel to physics
  • PS can be used to benchmark simulation methods


  • Aim to characterise beam phase-space at extraction, to design gantries
  • Mostly using MADX, looking to use PTC (already used for gantry designs, not synchrotron)
  • Will also be using Xsuite
  • New student to join soon and to start on RFKO simulations and port to Xsuite


  • Accelerators: GSI-SIS18, SIS100, HIT
  • Interested in Nano(10-100ns), Micro (10 us), and Macro (100ms) spill structure
  • Interested in simulating beam loss on septum wires, magnetic field errors, transit time, and tune distributions
  • Using Xsuite to simulate, also MADX and Elegant (developed at APS)
  • Interest in RFKO
  • Already seeing improvements using Xsuite over MADX, at spill lengths of 1.5 seconds


  • Focussed on PS, beam loss reduction on the septum
  • Using MapTrack (Rebecca Taylor)
  • Looking at RFKO
  • Different methods of extraction, flexibility for high intensity, short spills (FLASH)
  • Interested in higher order effects
  • Can Xsuite work with fixed-field accelerators? Dynamic tune measurements?
  • Requirements: ~1 million turns in ~8 hours


  • Proton/Carbon Medical synchrotron
  • Interested in RFKO and COSE
    • Optimise SX methods and parameters for spill structure (10s spill) and intensity ripple reduction
    • Multi Energy Extraction with COSE/RFKO (RF cavities)
    • Extracted beam ripples (magnets/cavities ripple effects)
    • Beam losses
  • Mainly using MADX/cpyMad for RFKO, Xsuite for dynamic effects, BLonD
  • Looking at ripples in RFKO
  • Need 10^6 turns, 10^4 particles
  • Requirements:
    • Handling 10^4 particles for 10^6 turns
    • Noise/ripple, TbT magnet trimming, deferred expressions, and longitudinal effects from cavities
    • Recording tune change during extraction
    • Time dependency

Needs for XSuite

  • Topics touched which are already usable in Xsuite:
    • Virtual sextupole strength
    • Septum with aperture at both sides (black absorber)
    • Noise and custom RF (Pablo and GSI to make simple example to add to Xsuite doc)
  • Open points:
    • Porting PyCollimate (action for CERN-ABT)
    • Field Tracking Maps => already have 2D time dependent map, implemented in a symplectic way
      • discussion started for 3D maps (check with Gianni)
      • Make simple example and upload on Xsuite doc page
    • SX optimisation: best practice and what different labs do
There are minutes attached to this event. Show them.