Secondary Emission Monitor (SEM) optimisation at the CERN fixed target beamlines

Understanding the effects of ageing instrumentation on the measurement of various beam properties is key to fixing undesired behaviour. The aim of this thesis was to conduct an investigation into the unexplained observations presented by Secondary Emission Monitor (SEM) instrumentation in the North Area (NA) fixed target beamlines at European Organization for Nuclear Research (CERN). First installed in the late 1970s, last calibrated in the 1990s and still in use today, SEM monitors are ubiquitous in these beamlines and responsible for measuring quantities such as the Protons on Target (POT), the transmission efficiency and beam position, used for steering the beam. The main objectives were: to understand and separate the contributions to signal caused by foil ageing due to proton dose, changes in the vacuum conditions and additional secondary charged particles, created in interactions with the bias foils or surrounding equipment; to perform the absolute calibration of the Beam Secondary Emission Intensity monitor (BSI)s in the NA target locations; and to improve the signal to noise ratio of the Beam Secondary Emission Grid (BSG)s in the NA, thus improving the steering accuracy within the beamlines. To understand the contributions to the BSI signal from additional secondary charged particles, simulation work was conducted comparing different positions of the beam relative to the BSI geometry. The energy, vertex location, generating process and type of particle created by the simulations and their effect on the signal generation for the BSI monitors was studied. The absolute calibration measurements for the BSIs, positioned at the primary target locations in the NA, were carried out using the activation method, and as a result, new calibration factors were established for all BSIs at these locations, affecting POT measurements for all fixed target experiments in the NA and improving the reliability of the transmission efficiencies measured. The results, stability of the measurements and the impact that these have on the operation of these devices was discussed. A solution for improving the signal to noise ratio of BSGs, using autoencoders to reduce the background noise, was investigated. An autoencoder model was trained on available BSG data and tested on new data. The results and performance of this model were evaluated.