MAD-X

MAD-X
Name	MAD-X
Developer	CERN
Released	1990s
Programming language	FORTRAN, Python
Operating system	Linux, macOS, Windows
Genre	Accelerator physics, beam optics
License	CERN Open Source

MAD-X MAD-X is a computer program for charged-particle beam optics and accelerator lattice design, widely used in particle physics and accelerator engineering. It provides simulation, matching, and tracking capabilities for synchrotrons, transfer lines, beamlines, and storage rings, enabling design work for facilities and experiments. MAD-X integrates with tools for beam dynamics, optics measurement, and control-system commissioning.

Overview

MAD-X is employed by accelerator laboratories, research institutions, and universities to model beam transport and optics for machines such as synchrotrons, colliders, and light sources. Major users include CERN, DESY, SLAC National Accelerator Laboratory, Brookhaven National Laboratory, and Fermilab. The codebase interacts with software ecosystems like ROOT (software), Python (programming language), MATLAB, Geant4, and SixTrack for end-to-end simulation, data analysis, and beam loss studies. MAD-X supports lattice elements, matching algorithms, Twiss parameter calculations, and matrix formalism relevant to projects like the Large Hadron Collider, European XFEL, and Swiss Light Source.

History and development

MAD-X evolved from earlier generations of accelerator design codes used in design efforts at SLAC National Accelerator Laboratory and CERN during the late 20th century. Development involved collaborations among accelerator physics groups at institutions such as KEK, GSI Helmholtz Centre for Heavy Ion Research, TRIUMF, INFN, and Oak Ridge National Laboratory. The project incorporated ideas from predecessor tools developed for projects like the SLC, LEP, and early synchrotron light sources. Over time, contributions from researchers affiliated with Imperial College London, MIT, University of Oxford, TU Darmstadt, and University of Manchester expanded functionality, while software engineering practices from organizations like Eclipse Foundation and GNU Project informed packaging and licensing approaches.

Design principles and functionality

MAD-X is grounded in linear and nonlinear beam optics theory, implementing transfer matrices, Lie algebraic methods, and symplectic tracking appropriate for machines such as Proton Synchrotron, Spallation Neutron Source, and Diamond Light Source. It provides commands for lattice element definition, sequence construction, beam matching, and optics reports used by accelerator physicists at facilities including CERN Super Proton Synchrotron, ALBA Synchrotron, and Paul Scherrer Institute. The tool emphasizes reproducibility, scriptability via Python (programming language) and native input files, and interoperability with measurement systems like EPICS and control-room software stacks at European Organization for Nuclear Research. MAD-X implements matching routines used in commissioning campaigns for projects such as ITER neutral beam lines and injector chains at HERA and RHIC.

Applications and use cases

MAD-X is applied in lattice design for high-energy machines, low-energy beam transport, injector studies, and optics correction workflows used in experiments at CERN, DESY, Fermilab, Brookhaven National Laboratory, and Los Alamos National Laboratory. It supports optics design for light source upgrades at facilities like MAX IV Laboratory, SOLEIL, and ESRF and is used in conceptual studies for proposed projects such as Future Circular Collider and International Linear Collider. Other applications include beam-based alignment procedures, collimation studies tied to LHC hardware, transfer line matching for experiments at PSI, and teaching curricula in accelerator physics at universities like University of California, Berkeley, University of Manchester, and University of Oxford.

Implementation and architecture

MAD-X combines a core written in legacy languages with modern bindings and extensions; historically the codebase uses FORTRAN for numerical kernels and connectors to Python (programming language) for scripting and automation. The architecture supports modular definition of elements—quadrupoles, dipoles, sextupoles, RF cavities—and sequence concatenation used in modeling machines like LEIR and ISOLDE. Data interchange formats allow integration with HDF5-based workflows, ASCII-based lattice files, and visualization via ROOT (software) or third-party plotting libraries. Interfaces exist for external tracking engines such as SixTrack and particle-matter interaction tools like Geant4, facilitating multi-physics studies.

Performance and benchmarking

Performance of MAD-X depends on lattice complexity, element count, and tracking resolution; benchmark comparisons are often performed against tools like ELEGANT, OPAL (accelerator code), and PLACET for accuracy and computational cost in tasks such as long-term tracking for dynamic aperture studies and optics matching for beam emittance control. Optimizations include compiled numerical kernels, parallelized post-processing with OpenMP or job distribution on clusters managed by Slurm Workload Manager and batch systems at centers like CERN IT and NERSC. Validation campaigns compare MAD-X predictions with beam measurements from machines such as ALICE (detector), LHCb, and injector test beams at DESY II.

Licensing and community

MAD-X is distributed under an open-source license maintained by CERN, encouraging contributions from a global community of accelerator physicists at institutions including CERN, DESY, SLAC National Accelerator Laboratory, Fermilab, and numerous universities. Development and user support occur through mailing lists, workshops at conferences like IPAC, PAC (Conference), and community events organized by EuCARD, HEP Software Foundation, and national laboratories. Commercial vendors offering accelerator design services and instrumentation, for example those collaborating with Thales Group or Siemens, also interface with MAD-X outputs in project workflows.

Category:Accelerator physics software Category:CERN software