Framework for automatic superconducting magnet model generation & validation against transients measured in LHC magnets

Superconducting magnet circuits are a key part of particle accelerators. Simulating complex transients occurring in these circuits with dedicated models is crucial to analyze their behavior and assess the impact of failure cases. The recently developed STEAM framework from CERN tackles this challenge with a series of dedicated software tools and model libraries. These models need to be validated and evaluated against experimental transients. This thesis discusses a newly developed addition to the STEAM framework, which provides the possibility to automatically generate and validate the superconducting magnet models, and hence to significantly shorten the required time and effort to validate such models. This is achieved by conducting a parametric sweep analysis aiming at determining the unknown model parameters. The developed solution is supplementing the already existing Python programming interface within the STEAM framework for the model generation. Its usage, implementation, and integration into the framework is described. This thesis presents the application on two different use cases of LHC superconducting magnets. These achieve an excellent agreement between simulations and measurements collected during LHC operation. Furthermore, it is shown that this solution can also be used to predict the occurrence of specific behavior in the superconducting magnets such as quench-back. The application on a specific use-case on multiple quench events in the main bending dipoles of the LHC presents a way to offer simulation results to quickly and easily support the everyday operation of the accelerator. Therefore, this thesis can be seen as a guideline and presentation of the practicability, abilities, and limitations of the automated framework.