Automated analysis of turn-by-turn beam trajectories following beam dumps at the LHC

The Large Hadron Collider (LHC) is a particle accelerator where two counter rotating beams circulate in opposite directions and are brought into collision at different points. The stored beam energy exceeds 400 MJ, so the machine is protected by a large number of interlock channels that provide vital information to understand failures and recover from perturbations. The Post-Mortem System organizes the data recorded during a beam abort and reconstructs the event sequence that led to it. The position of the LHC beams is monitored around the whole ring by a total of 1090 Beam Position Monitors (BPMs). In this project, we study BPMs at the time of dumps in order to detect key features. We look at invalid data and particular phenomena, such as quench heaters or missing beam-beam kick events, trying to develop and algorithm that enhances their detection. We find that the implementation of such an algorithm proves to be useful in a significant number of cases, although the establishment of the threshold values and confidence intervals remains as a subject for further study.