Author: Branlard, J.
Paper Title Page
MOPOPA02 Identification of the Mechanical Dynamics of the Superconducting Radio-Frequency Cavities for the European XFEL CW Upgrade 76
 
  • W.H. Syed, A. Bellandi, J. Branlard, A. Eichler
    DESY, Hamburg, Germany
 
  The European X-Ray Free-Electron Laser (EuXFEL) is to-date the largest X-ray research facility around the world which spans over 3.4 km. EuXFEL is currently being operated in a pulsed mode with a repetition rate of 10Hz. One upgrade scenario consists of operating the EuXFEL also in a Continuous-Wave (CW) mode of operation to improve the quality of experiments. This upgrade brings new challenges and requires new algorithms to deal with controlling a stable accelerating field inside the Superconducting Radiofrequency (SRF) accelerating cavities and keeping them on resonance in this new mode of operation. The purpose of this research work is to identify the mechanical dynamics of the cavities which will facilitate the development of the resonance controller for the CW upgrade. To this extent, experiments were conducted at a test bench. For the first time, in this work, two different types of spectrally rich excitation signals: multi-sine and stepped-sine are used to excite the mechanical dynamics of the cavities using the piezo actuator. After the analysis of experimental data, mechanical modes are successfully identified and will be used to design the controller.  
poster icon Poster MOPOPA02 [0.687 MB]  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-LINAC2022-MOPOPA02  
About • Received ※ 23 August 2022 — Revised ※ 25 August 2022 — Accepted ※ 26 August 2022 — Issue date ※ 01 September 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
MOPOPA03 Beam-Transient-Based LLRF Voltage Signal Calibration for the European XFEL 80
 
  • N. Walker, V. Ayvazyan, J. Branlard, S. Pfeiffer, Ch. Schmidt
    DESY, Hamburg, Germany
 
  The European XFEL linac consists of 25 superconducting RF (SRF) stations. With the exception of the first station which is part of the injector, each station comprises 32 1.3-GHz SRF TESLA cavities, driven by a single 10-MW klystron. A sophisticated state-of-the-art low-level RF (LLRF) system maintains the complex vector sum of each RF station. Monitoring and maintaining the calibration of the cavity electric field (gradient) probe signals has proven critical in achieving the maximum energy performance and availability of the SRF linac. Since there are no dedicated diagnostics for cross-checking calibration of the LLRF system, a procedure has been implemented based on simultaneously measuring the beam transient in open-loop operation of all cavities. Based on methods originally developed at FLASH, the European XFEL procedure makes use of automation and the XFEL LLRF DAQ system to provide a robust and relatively fast (minutes) way of extracting the transient data, and is now routinely scheduled once per week. In this paper, we will report on the background, implementation, analysis methods, typical results, and their subsequent application for machine operation.  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-LINAC2022-MOPOPA03  
About • Received ※ 13 August 2022 — Revised ※ 23 August 2022 — Accepted ※ 14 September 2022 — Issue date ※ 27 September 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
THPOPA15 Anomaly Detection Based Quench Detection System for CW Operation of SRF Cavities 775
 
  • G. Martino, A. Bellandi, J. Branlard, A. Eichler, H. Schlarb
    DESY, Hamburg, Germany
  • S. Aderhold, A.L. Benwell, D. Gonnella, S.L. Hoobler, J. Nelson, R.D. Porter, A. Ratti, L.M. Zacarias
    SLAC, Menlo Park, California, USA
  • L.R. Doolittle
    LBNL, Berkeley, California, USA
  • G. Fey
    Hamburg University of Technology, Hamburg, Germany
 
  Funding: This work is supported by DASHH (Data Science in Hamburg - HELMHOLTZ Graduate School for the Structure of Matter) under Grant No.: HIDSS-0002.
Superconducting radio frequency (SRF) cavities are used in modern particle accelerators to take advantage of their very high quality factor (Q). A higher Q means that a higher RF field can be sustained, and a higher acceleration can be produced in the cavity for length unity. However, in certain situations, e.g., too high RF field, the SRF cavities can experience quenches that risk creating damage due to the rapid increase in the heat load. This is especially negative in continuous wave (CW) operation due to the impossibility of the system to recover during the off-load period. The design goal of a quench-detection system is to protect the system without being a limiting factor during the operation. In this paper, we compare two different classification approaches for improving a quench detection system. We perform tests using traces recorded from LCLS-II and show that the ARSENAL classifier outperforms a CNN classifier in terms of accuracy.
 
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-LINAC2022-THPOPA15  
About • Received ※ 24 August 2022 — Accepted ※ 25 August 2022 — Issue date ※ 23 September 2022  
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
THPOPA21 Narrow Bandwidth Active Noise Control for Microphonics Rejection in Superconducting Cavities at LCLS-II 785
SUPCPA06   use link to see paper's listing under its alternate paper code  
 
  • A. Bellandi, J. Branlard
    DESY, Hamburg, Germany
  • S. Aderhold, A.L. Benwell, A. Brachmann, J.A. Diaz Cruz, D. Gonnella, S.L. Hoobler, J. Nelson, A. Ratti, L.M. Zacarias
    SLAC, Menlo Park, California, USA
  • J.A. Diaz Cruz
    UNM-ECE, Albuquerque, USA
  • R.D. Porter
    Cornell University (CLASSE), Cornell Laboratory for Accelerator-Based Sciences and Education, Ithaca, New York, USA
 
  LCLS-II is an X-Ray Free Electron Laser (XFEL) under commissioning at SLAC, being the first Continuous Wave (CW) hard XFEL in the world to come into operation. To accelerate the electron beam to an energy of 4 GeV, 280 superconducting cavities of the TESLA types are used. A Loaded Q (QL) value of 4x107 is used to drive the cavities at a power level of a few kilowatts. For this QL value, the RF cavity bandwidth is equal to 32 Hz. Therefore, keeping the cavity resonance frequency within such bandwidth is imperative to avoid a significant increase in the required RF power. In superconducting accelerators, resonance frequency variations are produced by mechanical microphonic vibrations of the cavities. One source of microphonics noise is rotary machinery such as vacuum pumps or HVAC equipment. A possible method to reject these disturbances is to use Narrowband Active Noise Control (NANC) techniques. Such a technique was already tested at DESY/CMTB and Cornell/CBETA. This proceeding presents the implementation of a NANC controller in the LCLS-II Low Level RF (LLRF) control system. Tests on the rejection of LCLS-II microphonics disturbances are also presented.  
poster icon Poster THPOPA21 [1.843 MB]  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-LINAC2022-THPOPA21  
About • Received ※ 24 August 2022 — Revised ※ 30 August 2022 — Accepted ※ 02 September 2022 — Issue date ※ 26 September 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
 
THPOPA26 Machine Learning Assisted Cavity Quench Identification at the European XFEL 798
THOPA09   use link to see paper's listing under its alternate paper code  
 
  • J. Branlard, A. Eichler, J.H.K. Timm, N. Walker
    DESY, Hamburg, Germany
 
  A server-based quench detection system is used since the beginning of operation at the European XFEL (2017) to stop driving superconducting cavities if they experience a quench. While this approach effectively detects quenches, it also generates false positives, tripping the accelerating stations when failures other than quenches occur. Using the post-mortem data snapshots generated for every trip, an additional signal (referred to as residual) is systematically computed based on the standard cavity model. Following an initial training on a set of such residuals derived from quench as well as non-quench events, two independent machine learning engines analyze routinely the trip snapshots and their residuals to identify if a trip was indeed triggered by a quench or has another root cause. The outcome of the analysis is automatically appended to the data snapshots and distributed to a team of experts. This constitutes a fully deployed example of machine-learning-assisted failure classification to identify quenches, supporting experts in their daily routine of monitoring and documenting the accelerator uptime and availability.  
slides icon Slides THPOPA26 [0.695 MB]  
poster icon Poster THPOPA26 [0.975 MB]  
DOI • reference for this paper ※ https://doi.org/10.18429/JACoW-LINAC2022-THPOPA26  
About • Received ※ 19 August 2022 — Revised ※ 24 August 2022 — Accepted ※ 01 September 2022 — Issue date ※ 01 September 2022
Cite • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)