The field description for the Large Hadron Collider

Abstract

The Large Hadron Collider (LHC), currently under construction at CERN, is a 27 km particle accelerator and has 1232 superconducting dipole magnets and 392 superconducting quadrupole magnets to respectively bend and focus the particle beams along their circular trajectory. Two counter rotating beams collide in four experimental insertions at a nominal centre of mass energy of 14 TeV. The LHC requires a powerful control system to correct the field variations that result from inherent properties of the superconducting magnets. If these field changes are not corrected with high speed and precision, they may jeopardize the machine performance significantly. Unfortunately, a feed-back control system that only relies on beam measurements has limited capabilities and is not sufficient to solely provide the error compensation that is needed. A system based on feed-forward control is therefore required to reduce the burden on the beam based feed-back by forecasting what the field variations will be to within a residual error comparable to beam control requirements. This thesis deals with the formulation of a static and dynamic field model as well as a set of scaling laws that together form the core of the feed-forward control system. This work also includes numerous magnetic measurements on the superconducting magnets in cryogenic conditions which enable the extraction of the parameters used in the model. The static field model is based on the reproducible magnetic effects that are dependent on the magnet excitation current. The dynamic field model is an extension of the static field model and mostly describes the behavior of the LHC during particle injection. The dynamic effects are dependent on both current and time and are not reproducible from cycle to cycle since they are dependent on the magnet powering history. Scaling laws are also formulated to provide a recalibration mechanism for the model and to extend its validity to a wider scope for the entire magnet population. The static and dynamic models as well as the scaling laws are applied on an LHC sector for the dipole magnets. The model is also tested on a sample of the main quadrupole magnets, a sample of the insertion region wide aperture magnets and on one long trim quadrupole corrector. The error obtained is within the desired tolerances and this hence demonstrates that the field model formulated is robust and adaptable to a wide range of magnet types. This dissertation also presents the development of a data acquisition system for a Hall plate based instrument that measures the fast magnetic field variations of the most important harmonics at the beginning of the particle acceleration. This new fully digitized acquisition system is shown to have a better performance than the preceding analogue system and hence provides a better foundation on which to base the scaling law for this crucial part of the LHC excitation cycle. The work presented in this thesis has been adopted by CERN as an integral part of the LHC feed-forward control system and will be used once the machine becomes operational in 2007.