iFAST 5.3 (Slow Extraction) - March Meeting

Tuesday 14 Mar 2023, 16:00 → 17:00 Europe/Zurich

Announcements

Peter Forck

• If somebody from the WG wants to publish a paper related to the studies done in the context of this WG, especially IPAC23, please add the iFAST funding acknowledgement. See email sent to the mailing-list, or here:
• This project has received funding from the European Union’s Horizon 2020 Research and Innovation programme under GA No 101004730
• IPAC23: PF is hoping to write an overview paper - not just simulations, but pointing out all the work iFAST 5.3 is doing, as there will be an iFAST meeting
• Check the Indico pages for everyone involved and collate a list

Francesco Velotti

• ABP colleagues are almost done porting matter-particle interaction from sixtrack to Xsuite - beta versions are available
• Field Tracking maps - some examples available, 3D/time-varying will take more time but is in pipeline. Maps extraction from field map can be envisaged after this is completed.
• Gianni can be invited again to discuss new developments or deployments - please send any requests
• Work ongoing to port pyCollimate to Xsuite too.

Slow Extraction Simulations for NIMMS Machines

R. Taylor, E. Benedetto

• Most things involve tracking maps, so techniques are transferrable to Xsuite

NIMMS

• Static simulations: MAD-X PTC
  • Tracking all elements
• Dynamic Simulations: maptrack
  • Developed by F. Velotti and P Arrutia at ABT
  • Lattice of PIMMS is cut into sections, so can focus on important elements
  • Complete flexibility over any method
• maptrack modules
  • All in python, own maptrack modules can be written and incorporated
  • Dynamic elements especially

Recommendations for best practices

• Parallelisation
  • Even if maptrack doesn’t have it’s own parallel tool, cython parallel can be used with Multiprocessing Pool
  • Parallelise by particle number or variable
  • Sympy parallel branch of maptrack performs better than Cython branch
• Aperture limits
  • maptrack ES has a complex aperture - using masked arrays
  • Masked arrays avoids conditionals/loops/indexing for septa
  • Pycollimate is also being implemented in Xsuite
• Memory handling
  • 10,000 particles over 1,000,000 turns is ~10GB
  • memmap objects ensures it only loads the chunks of data required for the function - as opposed to CSV files etc
• Tune Calculations
  • pyNAFF library used to calculate fundamental frequencies
• Applications
  • MapTrack is very flexible - NIMMS has a variety of designs and upgrades
  • SEEIST normal-conducting, SEEIST superconducting, Baltic Helium normal-conducting
  • MapTrack lattices can be used with interchangeable extraction methods making comparisons very easy
• RFKO in maptrack
  • Amplitude modulation, split in to four regions with different amplitude functions
• Side Topic: Fixed Field Accelerator
  • Extraction performed in Zgoubi as Xsuite isn’t performant with multiple closed orbits
• Xsuite in NIMMS
  • Exploring Slow Ex simulations
  • Once optics has been finalised for compact rings (mainly helium synchrotron), moving to Xsuite for precise tracking

Example of using transverse damper to drive SX in Xsuite

T. Bass, Philipp Niedermayer

• TB showed an example on how to use the transverse damper (exciter) to drive slow extraction in the PS
• The work is based on the contribution from Philipp Niedermayer to add this element to Xsuite
• TB will add this notebook to the list of examples available on the Xsuite documentation.

16:00 → 16:10 Welcome

Speaker: Francesco Maria Velotti (CERN)

16:10 → 16:35 Status with simulations

Speaker: R Taylor (Imperial College London)

NIMMS_Simulation_Slides

16:35 → 16:50 Transverse Excitation for RFKO in Xsuite

Speaker: Thomas Bass (University of London (GB))

2023-03-14-iFAST-5-3.pptx