Magnetic measurements of MCBC and MCBY orbit correctors under special cycling conditions

In the LHC, in addition to normal beam operation, the MCBC and MCBY orbit-corrector magnets are used for deflecting the beam during the van der Meer (vdM) scans used for calibrating the absolute luminosity scale. Recent studies by ATLAS and CMS strongly indicate that the beam displacements induced by the closed-orbit bumps used in such scans are affected by non-linearities of magnetic origin, that if left unaccounted for, can bias the absolute luminosity scale by a significant fraction of the total luminosity-uncertainty budget. In this paper, we explore experimentally the extent to which superconductor magnetization effects may account for the observed non-linear response of these closed-orbit bumps. The measurements are non trivial since the MCBC and the MCBY are used in a low field range, with special current cycles. At higher field levels, the saturation of the iron yoke can be a source of non-linearity, and the cross-talk from the adjacent aperture could affect the field quality. This document presents the results of two magnetic measurement campaigns carried out on MCBC and MCBY corrector magnets from December 2020 to February 2021. A first set of test cycles has been performed to assess the impact of the above-mentioned effects. In addition, since the effect of superconductor magnetization depends on the powering history, a set of special test cycles, emulating the actual powering of the magnets for the vdM scans, have been performed. The measurement results show that, under powering conditions similar to the vdM scans, the magnet has a non linear component in the transfer function due to hysteresis induced by the superconductor magnetization of the order of 0.05% of the nominal field of the correctors. We also characterized the transitions of magnetization branches due to a change of ramp rate in the magnets, occurring during the vdM scans, showing that a change of branch requires a current swing of the order of 10 A. Finally, the excellent reproducibility of the measured field integrals from one scan to the next suggests that the apparent percent-level variations of the measured beam-displacement scale observed during vdM sessions cannot be attributed to the steer- ing correctors. Depending on the magnet and test cycle considered, the maximum difference in linear coefficient is 7 µTm/A, 0.021% with respect to the mean value, while the minimal difference is below 1 µTm/A, which is close to the measurement precision.