LINAC2022 - List of Authors (Eichler, A.)

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 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)

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)

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 THPOPA26 [0.695 MB]
	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)

Paper

Title

Page

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 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)

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)

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 THPOPA26 [0.695 MB]

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)