Machine learning based crystal collimator alignment optimization

Gianmarco Ricci,Marco D'Andrea,Mario Di Castro,Eloise Matheson,Daniele Mirarchi,Andrea Mostacci,and Stefano Redaelli
DOI: https://doi.org/10.1103/physrevaccelbeams.27.093001
2024-09-25
Physical Review Accelerators and Beams
Abstract:In the CERN Large Hadron Collider (LHC), bent crystals are used to efficiently deflect beam halo particles toward secondary collimators used as absorbers. In this crystal collimation scheme, a crystal with a length of a few millimeters can produce a deflection equivalent to a magnetic field of hundreds of Tesla at LHC top energies, improving the cleaning performance of the machine. However, crystals must be in optimal alignment with respect to the circulating beam to maximize the efficiency of the channeling process. A newly developed machine learning model automatically classifies the channeling condition of crystals using beam loss monitor signals during slow rotation of the crystal. This advancement represents a crucial step toward refining the process of identifying the optimal channeling orientation. The algorithm has been tested for the fist time in operation with Pb ion beams at the record energy of 6.8 Z TeV demonstrating its reliability. https://doi.org/10.1103/PhysRevAccelBeams.27.093001 Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI. Published by the American Physical Society
physics, particles & fields, nuclear
What problem does this paper attempt to address?