Accelerator physics codes

For many applications it is sufficient to track a single particle through the relevant electric and magnetic fields. Old codes no longer maintained by their original authors or home institutions include: BETA,[3] AGS, ALIGN, COMFORT, DESIGN, DIMAD, HARMON, LEGO, LIAR, MAGIC, MARYLIE, PATRICIA, PETROS, RACETRACK, SYNCH,[4] TRANSPORT, TURTLE, and UAL. Some legacy codes are maintained by commercial organizations for academic, industrial and medical accelerator facilities that continue to use those codes. TRANSPORT, TRACE 3-D and TURTLE are among the historic codes that are commercially maintained.[5]

Major maintained codes include:

The self interaction (e.g. space charge) of the charged particle beam can cause growth of the beam, such as with bunch lengthening, or intrabeam scattering. Additionally, space charge effects may cause instabilities and associated beam loss. Typically, at relatively low energies (roughly for energies where the relativistic gamma factor is less than 10 or so), the Poisson equation is solved at intervals during the tracking using Particle-in-cell algorithms. Space charge effects lessen at higher energies so at higher energies the space charge effects may be modeled using simpler algorithms that are computationally much faster than the algorithms used at lower energies. Codes that handle low energy space charge effects include:

• ASTRA [8]
• Bmad
• CST Studio Suite [28]
• GPT [29]
• IMPACT [30]
• mbtrack [31]
• ORBIT,[32] PyORBIT[33]
• OPAL [34]
• PyHEADTAIL [35]
• Synergia [36]
• TraceWin [37]
• Tranft [38]
• VSim [39]
• Warp [40]

At higher energies, space charge effects include Touschek scattering and coherent synchrotron radiation (CSR). Codes that handle higher energy space charge include:

• Bmad
• ELEGANT
• MaryLie
• SAD

When two beams collide, the electromagnetic field of one beam will then have strong effects on the other one, called beam-beam effects. So called "weak-strong" simulations model one beam (called the "strong" beam since it affects the other beam) as a fixed distribution (typically a Gaussian distribution) which interacts with the particles of the other "weak" beam. This greatly simplifies the simulation. A full "strong-strong" simulation is more complicated and takes more simulation time. Strong-strong codes include

• GUINEA-PIG[41]

An important class of collective effects may be summarized in terms of the beams response to an "impedance". An important job is thus the computation of this impedance for the machine. Codes for this computation include

• ABCI [42]
• ACE3P [43]
• CST Studio Suite [44]
• GdfidL[45]
• TBCI [46]
• VSim [39]

To control the charged particle beam, appropriate electric and magnetic fields must be created. There are software packages to help in the design and understanding of the magnets, RF cavities, and other elements that create these fields. Codes include

• ACE3P [43]
• COMSOL Multiphysics [47]
• CST Studio Suite [48]
• OPERA [49]
• VSim [39]

Given the variety of modelling tasks, there is not one common data format that has developed. For describing the layout of an accelerator and the corresponding elements, one uses a so-called "lattice file". There have been numerous attempts at unifying the lattice file formats used in different codes. One unification attempt is the Accelerator Markup Language, and the Universal Accelerator Parser.[50] Another attempt at a unified approach to accelerator codes is the UAL or Universal Accelerator Library.[51]

The file formats used in MAD may be the most common, with translation routines available to convert to an input form needed for a different code. Associated with the Elegant code is a data format called SDDS, with an associated suite of tools. If one uses a Matlab-based code, such as Accelerator Toolbox, one has available all the tools within Matlab.

There are many applications of particle accelerators. For example, two important applications are elementary particle physics and synchrotron radiation production. When performing a modeling task for any accelerator operation, the results of charged particle beam dynamics simulations must feed into the associated application. Thus, for a full simulation, one must include the codes in associated applications. For particle physics, the simulation may be continued in a detector with a code such as Geant4.

For a synchrotron radiation facility, for example, the electron beam produces an x-ray beam that then travels down a beamline before reaching the experiment. Thus, the electron beam modeling software must interface with the x-ray optics modelling software such as SRW,[52] Shadow,[53] McXTrace,[54] or Spectra.[55] Bmad[10] can model both X-rays and charged particle beams. The x-rays are used in an experiment which may be modeled and analyzed with various software, such as the DAWN science platform.[56] OCELOT [57] also includes both synchrotron radiation calculation and x-ray propagation models.

Industrial and medical accelerators represent another area of important applications. A 2013 survey estimated that there were about 27,000 industrial accelerators and another 14,000 medical accelerators world wide,[58] and those numbers have continued to increase since that time.[59] Codes used at those facilities vary considerably and often include a mix of traditional codes and custom codes developed for specific applications. The Advanced Orbit Code (AOC)[60] developed at Ion Beam Applications is an example.

• List of codes from UCLA Particle Beam Physics Laboratory Archived 2018-07-17 at the Wayback Machine
• Comparison of Accelerator Codes