Effective methods for optimal design of induction coils on example of surface hardening

verfasst von: Yuliya Pleshivtseva, Marco Baldan, Anton Popov, Alexander Nikanorov, Edgar Rapoport, Bernard Nacke
Abstract: Purpose: This paper aims to describe main ideas and demonstrates results of the research activities carried out by the authors in the field of optimal design concepts for induction heater for surface hardening. The main goal of the research studies is the application of different optimization methods and numerical finite element method (FEM) codes for field analysis to solve the optimal design problem that is mathematically formulated in terms of the one of the most important optimization criteria for surface hardening technology, e.g. maximum temperature uniformity within the hardening surface layer. Design/methodology/approach: Evolutionary algorithm based on Adaptive Gaussian Process-Assisted Differential Evolution for MEMS Design Optimization (AGDEMO) and alternance method of parametric optimization based on optimal control theory are applied as effective tools for the practice-oriented problem for optimization of induction heater design based on non-linear coupled electromagnetic and temperature field analysis. Different approaches are used for combining FEM codes for interconnected field analysis and optimization algorithms into automated optimization procedure. Findings: Optimization procedures are tested and investigated for optimal design problem solution on the examples of induction hardening of steel cylindrical billet. Practical implications: Solved problems are based on the design of practical industrial applications. The developed optimization procedures are planned to be applied to the wide range of real-life problems of the optimal design of different electromagnetic devices and systems. Originality/value: This paper describes main ideas and results of the research activities carried out by the authors in the field of optimal design of induction heaters for hardening based on numerical coupled electromagnetic and temperature field analysis. The implementation of the automated procedure that combines a numerical FEM code for coupled field analysis with an optimization algorithm and its subsequent application for designing induction heaters makes the proposed approach specific and original. This paper also demonstrates that different optimization strategies used (evolutionary algorithm based on AGDEMO and alternance method of optimal control theory) are effective for real-life industrial applications for optimization of induction heaters design.
Organisationseinheit(en): Institut für Elektroprozesstechnik
Externe Organisation(en): Samara State Technical University
Typ: Artikel
Journal: COMPEL - The International Journal for Computation and Mathematics in Electrical and Electronic Engineering
Band: 39
Seiten: 90-99
Anzahl der Seiten: 10
ISSN: 0332-1649
Publikationsdatum: 07.01.2020
Publikationsstatus: Veröffentlicht
Peer-reviewed: Ja
ASJC Scopus Sachgebiete: Angewandte Informatik, Theoretische Informatik und Mathematik, Elektrotechnik und Elektronik, Angewandte Mathematik
Elektronische Version(en): https://doi.org/10.1108/COMPEL-05-2019-0216 (Zugang: Geschlossen)

BibTeX

@article{46022b3471284660a4eaf4c2d5d478e8,
title = "Effective methods for optimal design of induction coils on example of surface hardening",
abstract = "Purpose: This paper aims to describe main ideas and demonstrates results of the research activities carried out by the authors in the field of optimal design concepts for induction heater for surface hardening. The main goal of the research studies is the application of different optimization methods and numerical finite element method (FEM) codes for field analysis to solve the optimal design problem that is mathematically formulated in terms of the one of the most important optimization criteria for surface hardening technology, e.g. maximum temperature uniformity within the hardening surface layer. Design/methodology/approach: Evolutionary algorithm based on Adaptive Gaussian Process-Assisted Differential Evolution for MEMS Design Optimization (AGDEMO) and alternance method of parametric optimization based on optimal control theory are applied as effective tools for the practice-oriented problem for optimization of induction heater design based on non-linear coupled electromagnetic and temperature field analysis. Different approaches are used for combining FEM codes for interconnected field analysis and optimization algorithms into automated optimization procedure. Findings: Optimization procedures are tested and investigated for optimal design problem solution on the examples of induction hardening of steel cylindrical billet. Practical implications: Solved problems are based on the design of practical industrial applications. The developed optimization procedures are planned to be applied to the wide range of real-life problems of the optimal design of different electromagnetic devices and systems. Originality/value: This paper describes main ideas and results of the research activities carried out by the authors in the field of optimal design of induction heaters for hardening based on numerical coupled electromagnetic and temperature field analysis. The implementation of the automated procedure that combines a numerical FEM code for coupled field analysis with an optimization algorithm and its subsequent application for designing induction heaters makes the proposed approach specific and original. This paper also demonstrates that different optimization strategies used (evolutionary algorithm based on AGDEMO and alternance method of optimal control theory) are effective for real-life industrial applications for optimization of induction heaters design.",
keywords = "AGDEMO, Alternance method, Computational electromagnetics, Design optimization methodology, Evolutionary algorithm, Field analysis, Induction heating, Optimal design",
author = "Yuliya Pleshivtseva and Marco Baldan and Anton Popov and Alexander Nikanorov and Edgar Rapoport and Bernard Nacke",
note = "Funding Information: In this work, different FEM-based software and optimization approaches have been used for solving the same problem, i.e. an inductor design for hardening applications. Obtained results differ only by negligible quantities owing to small discrepancies in the FEM modelling. The solution got with the evolutionary algorithm confirms the prevision of the theory that was used in the alternance method. In both cases, the deviation from the desired temperature is only 2°C that completely satisfies the strictest technological requirements. This work is partly supported by Russian Foundation for Basic Researches (Projects №19-08-00232 and №18-08-00048). ",
year = "2020",
month = jan,
day = "7",
doi = "10.1108/COMPEL-05-2019-0216",
language = "English",
volume = "39",
pages = "90--99",
journal = "COMPEL - The International Journal for Computation and Mathematics in Electrical and Electronic Engineering",
issn = "0332-1649",
publisher = "Emerald Group Publishing Ltd.",
number = "1",
}