First synchrotron injection attempt into the SuperKEKB HER

• Naoko Iida

Room: 6/2-024 - BE Auditorium Meyrin A synchrotron injection scheme for the SuperKEKB high-energy ring (HER) was implemented and experimentally evaluated. The lattice at the HER injection point was configured to provide a large horizontal dispersion of –1.6 m, and the sextupoles were optimized accordingly. Because an abort system is located near the injection point, the optics design was constrained to ensure compatibility with the abort-system requirements. Optics solutions were developed to enlarge the dynamic aperture for interaction-point β_y* values of 81 mm, 8 mm, 3 mm, and 1 mm. A systematic procedure of the injection parameters has been established using a turn-by-turn BPM in the ring. The betatron components of the injected beam was successfully removed. Using the optimized optics, synchrotron injection into the HER was successfully demonstrated, followed by the establishment of stable beam–beam collisions and the production of luminosity. As the injection repetition rate was increased, however, an unexpected degradation in injection efficiency was observed, leading to the suspension of the study. These experimental results and possible mechanisms that contributed to the observed injection degradation will be reported.

Beam Dynamics Studies and Generative Machine Learning Based Phase Space Reconstruction in the SuperKEKB Injector

• Andrea Aguirre (Autonomous University of Puebla (MX))

Room: 6/2-024 - BE Auditorium Meyrin Achieving the next luminosity milestone at SuperKEKB requires improved emittance preservation throughout the injector chain. This talk presents two complementary studies addressing this challenge. In the first, a unified simulation framework developed in OCELOT quantifies the individual and combined impact of short-range wakefields, synchrotron radiation effects, alignment errors, and dipole magnetic field imperfections on emittance growth in the SuperKEKB linac and electron beam transfer line (BTe). Accurate modelling, however, requires a realistic initial beam distribution at the electron gun, currently a key source of uncertainty. This motivates the second study, in which a generative machine learning approach based on Generalized Phase Space Reconstruction (GPSR) is developed to infer the longitudinal phase space at the gun from downstream screen measurements in the linac. The method, implemented within the differentiable CHEETAH framework, recovers complex longitudinal structures from simulated data, paving the way for future experimental validation and an extension towards full 6D phase space reconstruction