A coupled multiphysics FEM model to investigate electromagnetic, thermal and mechanical effects in complex assemblies: The design of the High-Luminosity Large Hadron Collider beam screen

In the framework of the High-Luminosity Large Hadron Collider (HL-LHC) project, new beam screens will be installed by 2024 within the cold bore of the superconducting magnets. The beam screen is an octagonal shaped pipe that shields the 1.9 K magnet cryogenic system from the heat loads and damage to the magnet coils that would be otherwise induced by the highly penetrating collision debris. It also ensures that proper vacuum conditions required for the stability of the beam are met. A failure scenario of the beam screen is represented by the magnet quench, a resistive transition of the superconducting magnets, that can compromise its mechanical integrity. During a quench the magnet gradient of the quadrupole, in which the beam screen is inserted, decays from 140 T/m to about 0 T/m in 0.4 s inducing high magnitude forces in the assembly. Understanding the magnetic, thermal and mechanical behaviours of the beam screen assembly during the quench is critical to enable its effective design and operation. A numerical model, that can accurately predict the behaviours of the beam screen during a magnet quench, has been developed. Compared to the analytical formulations used to design the beam screen currently installed in the LHC, the multiphysics FEM model developed in this research introduces multiple elements of novelty and improved performance. First, self-inductance effects are accounted for and found to reduce the induced forces up to approximately 2000% at high electrical conductivity values. Second, the one-way and two-way coupling of the magnetic with the mechanical and thermal interfaces are explored and the best trade-off is defined. Third, the mechanical response of the assembly is evaluated dynamically over the evolution of the magnetic field decay rather than just in a quasi-static manner. Fourth, three dimensional geometries can also be studied enabling the design of the components to be placed along the beam axis. The model has been verified by comparison to a closed form expression showing the advantages of considering self-inductance phenomena. The mechanical integrity of the new beam screen has been demonstrated and a less conservative design has been obtained, which has permitted to relax the tight constraints on interfacing systems. Amongst other applications, the model has already been applied at CERN to support the conceptual design of an ad-hoc beam screen for the Future Circular Collider (FCC).