Speaker
Description
Ultrashort bunch length and longitudinal beam profile measurements are essential for characterizing electron beams in short-pulse accelerators, including free-electron laser facilities and novel accelerator concepts such as EuPRAXIA. A single-shot longitudinal diagnostics approach based on broadband spatial imaging of coherent transition radiation (CTR) in the THz region, combined with data-driven analysis, is presented. The diagnostic system has been installed at the MAX IV Short-Pulse Facility (SPF), where extensive experimental data have been collected.
The broadband imaging system was designed using optical simulations to study sensitivity to longitudinal profile properties. Experimental CTR images were acquired across a range of machine settings and correlated with reference measurements from a transverse-deflecting cavity. Machine learning techniques (i.e., convolutional neural networks (CNN), ensemble models, transfer learning) were subsequently applied to the CTR image data to infer bunch length and reconstruct longitudinal beam profiles in the femtosecond range. This contribution describes the optical system design, experimental setup, and data acquisition, and presents results demonstrating the method's capability to retrieve longitudinal beam parameters on a single-shot basis.
Simulations indicate that spatial CTR imaging, combined with machine learning, provides a promising, compact, and reliable approach for single-shot longitudinal beam diagnostics, with potential applicability to high-repetition-rate and short-pulse accelerator facilities, for which an experimental demonstration is underway.
Funding Agency
EU’s Horizon Europe research and innovation no. 101073480, UKRI guarantee funds, and no. 101188004; Cockcroft Institute core grant no. STFC ST/V001612/1, AWAKE-UK phase II no. STFC ST/X005208/1.
| I have read and accept the Privacy Policy Statement | Yes |
|---|---|
| Supervisor's name | Joseph Wolfenden & Carsten Welsch |
| Supervisor's email | welsch@liverpool.ac.uk |
| Paper submission software | LATeX |