1 of 8

Update on Start-to-End Modeling Efforts at SLAC

Claudio, Sanjeev, Nathan, Alex, Doug, Robert, Spencer

January 11, 2024

2 of 8

Start-to-End Modeling Challenge

2

  • Start-to-End (S2E) simulations of FACET-II include:
    • Injector (GPT/Impact)
    • Linac (Lucretia/BMAD)
    • Plasma (OSIRIS/QuickPIC/QPAD)
    • Dump line (Lucretia)
  • We need a framework for linking the simulations.
    • Ongoing effort at SLAC.
  • Comparison of machine measurements with beamline simulations is a lot of work!
    • Glen White provided a lot of this effort for FACET.
    • Nathan and Claudio stepping in to fill the gap.
    • Leverage expertise of Auralee’s team to build out infrastructure.
  • We are creating a design document to specify S2E needs: FACET S2E Design Doc

​

Complete S2E modeling of FACET beamline is a multi-person effort.

3 of 8

Framework and Design Document (Claudio)

3

  • Goal of the design document is to describe S2E codes and simulation infrastructure and methods.
  • Approach is to split S2E simulations in different chunks, simulated with different codes, and transport particles between codes via openPMD.
  • Our S2E modeling effort will leverage/benefit from tools developed for LCLS e.g. LUME python wrapper for multiple codes.
  • We are planning on running simulations on SLAC hardware. Current S2E sims are being run on local PC (glenbox). Future simulations will likely run on SLAC’s HPC cluster S3DF.
  • Current codes used to model the injector (GPT) and linac (Lucretia) may not be optimal choices going forward for a number of reasons.
  • Short term goals for S2E modeling effort is to evaluate viability of other codes by:
    • Running a demo simulation of injector with IMPACT, compare with GPT
    • Running a demo simulation of the linac with Bmad, compare with Lucretia
  • Long-term goals of S2E modeling effort are to:
    • Maintain a repository of beam/lattice files reflecting the most up-to-date beamline and baseline beam configurations.
    • Provide a ‘master script’ for running cathode-to-dump simulations for users for the baseline beam single and two bunch beam configurations.
  • S2E simulation effort will update on progress monthly at internal meeting with AARD/ops team.

4 of 8

Progress on the Injector (Nathan)

4

  • Working to remove “magic numbers”
    • Goal is to to be explainable whenever possible but at least systematic if not
  • Parameter scans of both physical and virtual injector settings has enabled the creation of transform functions
    • Allow us to relate the value read off a PV to a variable assigned in Lucretia/GPT
  • Permits optimization at new operating parameters (notably charge, Schottky phase, and laser pulse specs)
  • More details from 2023-09-27 talk

Naive transform

Best linear transforms

5 of 8

Framework for OpenPMD (Sanjeev)

5

Define Input Beam

openPMD

GDF

GPT Simulation

GDF

openPMD

Lucretia Simulation

Matlab

openPMD

Beam Particles Repository

Simulation example for FACET-II Injector

6 of 8

Lucretia to PIC (Alex)

  • Handshake scripts available:
    • Lucretia to HiPACE++
    • Lucretia to FBPIC
    • Lucretia to QuickPIC (by UCLA)
  • QuickPIC and HiPACE++ containerization for S3DF

​

  • Work together with Mason Stobbe, Robert Holtzapple and Doug on matching studies in the gas jet
    • Quadrupole scans
    • Optimization loop
    • Might include sextupoles in the future

​

  • Work on Lucretia-2-PIC simulations for staging
  • Some Lucretia-2-PIC simulations with collimators

6

7 of 8

Simulations of single bunch energy depletion with Li oven (Doug)

  • A short term goal for E300 will be optimizing single-bunch drive-to-wake energy transfer efficiency in a 40cm long Li oven
  • Simulations of the presently available beam conditions will help to guide the experimental program

​

  • Beam conditions at end of 2023 run:
    • Bunch charge: 1-1.5 nC
    • Bunch size: Roughly 20x20x20 µm^3 (sigma_x, sigma_y, sigma_z)
    • Incoming Twiss: beta = 50 cm → emit_n ~ 20 µm
  • ​

DAQ 4797 - 11/17/2023

Li oven, 5 Torr, TC4=835C

qb=1nc, σ~40x40x15 um^3

8 of 8

E300 Simulation Requests (Robert)

8

  • Simulation setup (+beam size, beta, emittance on previous slide):
    • 40cm Lithium oven longitudinal density profile
    • Plasma density 3.5X10^16 cm^-3
    • Beam ionized
    • Beam waist located according to https://doi.org/10.1103/PhysRevAccelBeams.23.011302
  • Requested outputs:
    • Macro particles at simulation end
    • Maximum single particle energy loss
    • Percentage of beam energy lost to the wake
    • Movie of the beam travelling through the wake and the evolution of the longitudinal phase space
  • Initial parameter scans:
    • Plasma density - “easiest” experimental knob
    • Bunch length - should improve energy loss
  • Next steps depend on results: scan beta, scan waist location etc.