JLab, Newport News, Virginia, USA
IMP/CAS, Lanzhou, People’s Republic of China
BNL, Upton, New York, USA
Electron cooling continues to be an invaluable technique to reduce and maintain the emittance in hadron storage rings in cases where stochastic cooling is inefficient and radiative cooling is negligible. Extending the energy range of electron coolers beyond what is feasible with the conventional, electrostatic approach necessitates the use of RF fields for acceleration and, thus, a bunched electron beam. To experimentally investigate how the relative time structure of the two beams affects the cooling properties, we have set up a pulsed-beam cooling device by adding a synchronized pulsing circuit to the conventional electron source of the CSRm cooler at Institute of Modern Physics *. We show the effect of the electron bunch length and longitudinal ion focusing strength on the temporal evolution of the longitudinal and transverse ion beam profile and demonstrate the detrimental effect of timing jitter as predicted by theory and simulations. Compared to actual RF-based coolers, the simplicity and flexibility of our setup will facilitate further investigations of specific aspects of bunched cooling such as synchro-betatron coupling and phase dithering.
* M. W. Bruker et al., Phys. Rev. Accel. Beams 24, 012801 (2021)
JLab, Newport News, Virginia, USA
ODU, Norfolk, Virginia, USA
STFC/DL, Daresbury, Warrington, Cheshire, United Kingdom
We report on the development of machine learning models for classifying C100 superconducting radiofrequency (SRF) cavity faults in the Continuous Electron Beam Accelerator Facility (CEBAF) at Jefferson Lab. Of the 418 SRF cavities in CEBAF, 96 are designed with a digital low-level RF system configured such that a cavity fault triggers recordings of RF signals for each of eight cavities in the cryomodule. Subject matter experts analyze the collected time-series data and identify which of the eight cavities faulted first and classify the type of fault. This information is used to find trends and strategically deploy mitigations to problematic cryomodules. However, manually labeling the data is laborious and time-consuming. By leveraging machine learning, near real-time - rather than postmortem - identification of the offending cavity and classification of the fault type has been implemented. We discuss the performance of the machine learning models during a recent physics run. We also discuss efforts for further insights into fault types through unsupervised learning techniques and present preliminary work on cavity and fault prediction using data collected prior to a failure event.