Characterization, control and alignment of crystal collimators of the high Luminosity Large Hadron Collider (HL-LHC)

The Large Hadron Collider (LHC), the world’s most powerful particle accelerator, is undergoing a significant upgrade to increase both its luminosity and operational lifespan. The LHC upgrade project is referred to as the High-Luminosity Large Hadron Collider (HL-LHC). A central aspect of this upgrade is the implementation of crystal collimation technology, which enhances beam loss management by using bent crystals to improve collimation efficiency for heavy-ion beams. This work aims to advance the characterization, alignment, operational supervision, and control of crystal collimators for the HL-LHC by developing novel software, algorithms, and machine learning models to streamline these processes. To address these challenges, an innovative X-ray diffractometer was designed to simplify and automate the traditionally complex task of crystal characterization. Advanced programming techniques were employed to control the diffractometer’s mechatronic components and X-ray beam, while algorithms were developed for post-processing and automation of measurements and setup. Furthermore, cutting-edge convolutional neural networks and real-time monitoring systems were designed to optimize crystal alignment and supervision during LHC operations, ensuring more efficient setup and beam collimation. This thesis also introduces a new test bench for the evaluation of a piezo-goniometer controller algorithm for crystal collimators. The overall goal of this work is to provide efficient solutions for automating particle accelerator setup and improving the performance of the LHC collimation system through advanced technologies. The findings presented in this thesis offer valuable insights into future research directions for crystal collimation systems and their integration with modern machine learning techniques.