SIMBA: Simulations for Integrated Modeling of Beams in Accelerators

Improving accelerators through consistent, complete, and easy to use simulations.

Mission

SIMBA is a framework for performing simulations of particle accelerators that aims to be simple to use, and complete.

Our mission statement is "To create a start to end framework for consistent, transparent simulations of particle accelerators and FELs that anybody can use, and that everybody trusts."

By leveraging a standard accelerator lattice format, SIMBA is able to generate and run input files for a range of accelerator simulation codes, enabling seamless transfer of input and output distributions.

The codes currently supported by SIMBA are:

• ASTRA
• GPT
• ELEGANT
• Ocelot
• CSRTrack
• Cheetah
• Wake-T
• Xsuite

A range of other codes are also currently under active development.

Documentation

Installation

Cloning from Github

Clone SIMBA from Github:

git clone https://github.com/astec-stfc/simba.git

The package and its dependencies can be installed using the following command in the SIMBA directory:

pip install .

Install from PyPI

Alternatively, SIMBA can be installed directly from PyPI using:

pip install simba-accelerator

Optional Dependencies

In order to access the accelerator code executables, they must also inform SIMBA of their locations. See the entry on SimCodes.

Participation

We welcome contributions and suggestions from the community! SIMBA is currently under active development, and as such certain features may be missing or not working as expected. If you find any issues, please raise it here or contact Alex Brynes.

We are also happy to help with installation and setting up your accelerator lattice.

Example Lattice and Simulation

Getting started with SIMBA

Accelerator lattices in SIMBA are derived from the LAURA standard lattice format. This is a schema for providing generic descriptions of accelerator elements and layouts; see LAURA documentation.

Given that this format is designed to capture all relevant information about accelerator elements, and that it includes a built-in translator module for exporting lattice files to various simulation codes, it can be used within :mod:SIMBA for loading, modifying, writing and exporting input and lattice files for simulation codes.

Defining the Lattice Simulation

The simulation of the lattice is defined in a separate YAML file, for example CLA-Injector.def:

generator:
    default: clara_400_3ps
files:
  injector400:
    code: astra
    charge:
      cathode: True
      space_charge_mode: 2D
      mirror_charge: True
    input:
      particle_definition: 'initial_distribution'
    output:
      zstart: 0
      end_element: CLA-S02-SIM-APER-01
  Linac:
    code: elegant
    output:
      start_element: CLA-S02-SIM-APER-01
      end_element: CLA-FEA-SIM-START-01
  FEBE:
    code: ocelot
    charge:
      cathode: False
      space_charge_mode: 3D
    input: {}
    output:
      start_element: CLA-FEA-SIM-START-01
      end_element: CLA-FED-SIM-DUMP-01-START
groups:
  bunch_compressor:
    type: chicane
    elements: [CLA-VBC-MAG-DIP-01, CLA-VBC-MAG-DIP-02, CLA-VBC-MAG-DIP-03, CLA-VBC-MAG-DIP-04]
layout: /path/to/laura-lattices/CLARA/layouts.yaml
section: /path/to/laura-lattices/CLARA/sections.yaml
element_list: /path/to/laura-lattices/CLARA/YAML/

This lattice definition would produce several output files (called injector400.in, Linac.lte, and FEBE.py) for running in the ASTRA, Elegant and Ocelot beam tracking codes.

The elements are loaded from the directory /path/to/laura-lattices/CLARA/YAML/ defined above.

As this simulation starts from the cathode, the input definition is required for the first injector400 file block. An alternative method for starting is to specify input/particle_definition to point to an existing beam file.

For follow-on lattice runs, it is sufficient to define the output: start_element, which should match the output: end_element definition from the previous file block.

Running SIMBA

import simba.Framework as fw


# Define a new framework instance, in directory 'example'.
#       "clean" will empty (delete everything!) in the directory if true
#       "verbose" will print a progressbar if true
simcodes_location = "/path/to/simcodes/directory"
framework = fw.Framework(
    master_lattice="/path/to/laura-lattices/CLARA",
    directory="./example",
    generator_defaults="/path/to/laura-lattices/CLARA/Generators/clara.yaml",
    simcodes_location=simcodes_location,
    clean=True,
    verbose=True,
    )
# Load a lattice definition file. These can be found in Masterlattice/Lattices by default.
framework.loadSettings("Lattices/CLARA.def")
# Change all lattice codes to ASTRA/Elegant/GPT with exclusions (injector cannot be done in Elegant)
framework.change_Lattice_Code("All", "ASTRA", exclude=["Linac"])
# Again, but put the VBC in Elegant for CSR
framework.change_Lattice_Code("FEBE", "Elegant")
# This is the code that generates the laser distribution (ASTRA or GPT)
framework.change_generator("ASTRA")
# Load a starting laser distribution setting
framework.generator.load_defaults("clara_400_2ps_Gaussian")
# Set the thermal emittance for the generator
framework.generator.thermal_emittance = 0.0005
# This is a scaling parameter
# This defines the number of particles to create at the gun (this is "ASTRA generator" which creates distributions)
framework.generator.number_of_particles = 512
# Track the lattice
framework.track()

Example Notebooks

Some further examples on SIMBA usage can be found in the following notebooks:

• getting_started.ipynb
• beams_example.ipynb
• utility_functions.ipynb

Authors

• James Jones
• Alex Brynes
• Mark Johnson