Mathematical Modeling, Simulation and Optimization for CERN’s Quench Protection System
Develop a scalable, modular co-simulation framework to analyze quench behavior in superconducting magnets and ensure the stability and accuracy of simulations through rigorous mathematical modeling and analysis.
Image copyright: TU Darmstadt
Involved researchers
- Idoia Cortes Garcia (TU Eindhoven)
- Sebastian Schöps (TU Darmstadt)
Partners from industry / Research lab
- Lorenzo Bortot (CERN, Switzerland)
- Matthias Mentink (CERN, Switzerland)
Problems description
Superconducting magnets at CERN’s LHC require robust quench protection to prevent damage from loss of superconductivity. Simulating these events is a complex multiphysical, multiscale, and multirate problem. Traditional monolithic simulation approaches are inefficient, necessitating a modular framework capable of coupling heterogeneous models across time and physical domains.
Goals
Develop a scalable, modular co-simulation framework to analyze quench behavior in superconducting magnets and ensure the stability and accuracy of simulations through rigorous mathematical modeling and analysis.
Methods
Hierarchical modeling; Time domain finite element method; Co-simulation via waveform relaxation; Differential-algebraic equations (DAEs); Structural analysis of DAE index
Outcome
The collaboration created the STEAM software platform, enabling accurate and efficient simulation of quench events in LHC magnets; adopted by CERN and other international labs for design and safety studies.
Reference
Lorenzo Bortot, Bernhard Auchmann, Idoia Cortes Garcia, Alejando M. Fernando Navarro, Michał Maciejewski, Matthias Mentink, Marco Prioli, Emmanuele Ravaioli, Sebastian Schöps, and Arjan Verweij. STEAM: a hierarchical co-simulation framework for superconducting accelerator magnet circuits. IEEE Transactions on Applied Superconductivity, 2018. ARXIV: 1801.08957, doi:10.1109/TASC.2017.2787665. Eingereicht von: Sebastian Schöps [04.05.25]
