Reference : Energy-Based Magnetic Hysteresis Models - Theoretical Development and Finite Element ...
Dissertations and theses : Doctoral thesis
Engineering, computing & technology : Electrical & electronics engineering
http://hdl.handle.net/2268/229596
Energy-Based Magnetic Hysteresis Models - Theoretical Development and Finite Element Formulations
English
[en] Modèles Energétiques d’Hystérésis Magnétique - Développement Théorique et Formulations pour la Méthode des Elements Finis
Jacques, Kevin mailto [Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Applied and Computational Electromagnetics (ACE) >]
21-Nov-2018
Université de Liège, ​Liège, ​​Belgique
Docteur en Sciences de l'Ingénieur
254
Geuzaine, Christophe mailto
Gyselinck, Johan mailto
Vanderheyden, Benoît mailto
De Greve, Zacharie mailto
Henrotte, François mailto
Napov, Artem mailto
Kedous Lebouc, Afef mailto
Rasilo, Paavo mailto
[en] Hysteresis ; Energy-Based ; Finite Element ; Newton-Raphson ; Ferromagnetism
[en] This work focuses on the development of a highly accurate energy-based hysteresis
model for the modeling of magnetic hysteresis phenomena. The model relies on an
explicit representation of the magnetic pinning effect as a dry friction-like force acting
on the magnetic polarization. Unlike Preisach and Jiles-Atherton models, this model
is vectorial from the beginning and derives from thermodynamic first principles.
Three approaches are considered: the first one, called vector play model, relies on a
simplification that allows an explicit, and thus fast, update rule, while the two others,
called the variational and the differential approaches, avoid this simplification,
but require a non-linear equation to be solved iteratively. The vector play model and
the variational approach were already used by other researchers, whereas the differential
approach introduced in this thesis, is a new, more efficient, exact implementation,
which combines the efficiency of the vector play model with the accuracy of the variational
approach. The three hysteresis implementations lead to the same result for
purely unidirectional or rotational excitation cases, and give a rather good approximation
in all situations in-between, at least in isotropic material conditions.
These hysteresis modeling approaches are incorporated into a finite-element code as
a local constitutive relation with memory effect. The inclusion is investigated in detail
for two complementary finite-element formulations, magnetic field h or flux density
b conforming, the latter requiring the inversion of the vector hysteresis model,
naturally driven by h, for which the Newton-Raphson method is used. Then, at the
finite-element level, once again, the Newton-Raphson technique is adopted to solve
the nonlinear finite-element equations, leading to the emergence of discontinuous differential
reluctivity and permeability tensors, requiring a relaxation technique in the
Newton-Raphson scheme. To the best of the author’s knowledge, the inclusion of an
energy-based hysteresis model has never been successfully achieved in a b-conform
finite-element formulation before.
FEDO (Free Software for Electric Drive Optimization)
Researchers ; Professionals ; Students ; General public
http://hdl.handle.net/2268/229596

File(s) associated to this reference

Fulltext file(s):

FileCommentaryVersionSizeAccess
Open access
JACQUES_Kevin_thesis-20181123.pdfPublisher postprint10.01 MBView/Open

Bookmark and Share SFX Query

All documents in ORBi are protected by a user license.