No-beam-interruption Cavity Recovery

The overarching goal of this project is to experimentally demonstrate resumption of nominal beam operation within 1 second following a trip of a superconducting radio-frequency (SRF) cavity, and to demonstrate recovery of the energy reserve in the superconducting linear accelerator concomitantly to nominal beam operation. If successful, this program will provide a strong technical platform to eliminate most beam trips in the 1 – 10 minutes range, improving availability. AI/ML is an essential part for demonstrating the no-beam-interruption recovery method, for example to provide optimum rephasing of SRF cavities and fast beam loss tuning.

In high-power hadron beam operations, great care must be taken as 1) even minor deviations of cavity phases or amplitudes from their optimal settings can lead to unacceptable, uncontrolled beam losses and 2) inadequate beam loading compensation through adaptive feedforward (AFF) can likewise result in significant beam losses. In order to demonstrate the no-beam-interruption cavity recovery, the proposed work will develop a systematic technical path for 1) fast cavity operation recovery (< 1 second) at a reduced gradient in closed-loop, 2) fast resumption of beam operation with all downstream cavities rephased and with accurate estimation of AFF waveforms (< 1 second), 3) gradual restoration of accelerating gradient for the tripped cavity to its nominal level during beam operation, and 4) low beam-loss maintained during the entire cavity recovery process.

The no-beam-interruption recovery method is transformative in its founding principle that beam delivery can continue during trip recovery of a SRF cavity. The proposed no-beam-interruption SRF cavity recovery method is an iterative cavity recovery scheme that will convert >90% of the medium duration trips (minutes) to fast duration trips (< 1 sec). If successful and later fully automated, the no-beam-interruption cavity recovery method could benefit SNS availability. The method would also be adaptable and transferable to other existing or future Superconducting Linac (SCL)-based facilities.