Liebsch, Melvin (2022)
Inference of Boundary Data from Magnetic Measurements of Accelerator Magnets.
Technische Universität Darmstadt
doi: 10.26083/tuprints-00021144
Ph.D. Thesis, Primary publication, Publisher's Version
Text
2022-03-21-Liebsch-Melvin.pdf Copyright Information: CC BY-SA 4.0 International - Creative Commons, Attribution ShareAlike. Download (91MB) |
Item Type: | Ph.D. Thesis | ||||
---|---|---|---|---|---|
Type of entry: | Primary publication | ||||
Title: | Inference of Boundary Data from Magnetic Measurements of Accelerator Magnets | ||||
Language: | English | ||||
Referees: | Kurz, Prof. Dr. Stefan ; Russenschuck, Dr.-Ing. Stephan ; Schöps, Prof. Dr. Sebastian | ||||
Date: | 2022 | ||||
Place of Publication: | Darmstadt | ||||
Collation: | ix, 150 Seiten | ||||
Date of oral examination: | 14 March 2022 | ||||
DOI: | 10.26083/tuprints-00021144 | ||||
Abstract: | In this research, mathematical modeling is used to express the measurement process for the magneto-static field in an accelerator magnet and approaches to infer model variables from on magnetic measurements are presented. The physical relations are implied by solving a partial differential equation, for the field evaluation. With the formulation by means of a boundary value problem, measurements can be restricted to the domain boundary, reducing the effort to provide field maps in three dimensions from O(1/h3) to O(1/h2), where h is the measurement resolution. A higher order, iso-geometric boundary element method is used for the field model, which comes with benefits for the extraction of Taylor maps used for particle tracking applications, because sufficiently smooth derivatives of arbitrary order can be determined. Moreover, an indirect boundary element formulation is presented, establishing a linear relation between field and boundary data, without the need to map between Dirichlet and Neumann data, or the solution of a similar linear equation system for field evaluation. This is of advantage for the inference of boundary data from measurements. Magnetic measurements are providing voltages, or their integrals over small time windows, which are often not directly proportional to the field, or model variables. The Bayesian paradigm provides a framework to infer model variables from dependent observations, under the influence of measurement errors. In this way, uncertainty quantification is achieved effectively, as only a single realization of the measurement process is needed, without the need to execute repetitions. Moreover, an active learning algorithm is developed, used to explore the physical domain in regions with large uncertainty. In the scope of this doctoral thesis, a new three-axes Hall-probe mapper system has been commissioned and metrologically characterized. For this reason, the focus is on the application of Hall probe field mapping, for the practical realization of the above mentioned theoretical aspects. This includes the calibration of Hall effects in three dimensions, the solution of the absolute sensor position and orientation problem, as well as the derivation of a magneto-mechanical model for the quantification of measurement uncertainties due to mechanical vibrations and positioning errors. The model-based post processing using Bayesian inference is put into a general framework, which is applied to three different problems, appearing in the context of magnetic measurements. |
||||
Alternative Abstract: |
|
||||
Status: | Publisher's Version | ||||
URN: | urn:nbn:de:tuda-tuprints-211441 | ||||
Classification DDC: | 500 Science and mathematics > 510 Mathematics 500 Science and mathematics > 530 Physics 600 Technology, medicine, applied sciences > 600 Technology 600 Technology, medicine, applied sciences > 620 Engineering and machine engineering |
||||
Divisions: | 18 Department of Electrical Engineering and Information Technology > Institute for Accelerator Science and Electromagnetic Fields > Computational Electromagnetics 18 Department of Electrical Engineering and Information Technology > Institute for Accelerator Science and Electromagnetic Fields |
||||
Date Deposited: | 13 Jun 2022 12:10 | ||||
Last Modified: | 13 Jun 2022 12:10 | ||||
URI: | https://tuprints.ulb.tu-darmstadt.de/id/eprint/21144 | ||||
PPN: | 496563424 | ||||
Export: |
View Item |