Nonlinear convergence acceleration of magnetic field computation

Shuangxia Niu; Siu Lau Ho; Weinong Fu

doi:10.1109/INTMAG.2015.7156784

Nonlinear convergence acceleration of magnetic field computation

Shuangxia Niu, Siu Lau Ho, Weinong Fu

The Hong Kong Polytechnic University

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

2 Citations (Scopus)

Abstract

Finite element method (FEM) has been widely used to analyze the performance of electric devices. Due to the nonlinear permeability of iron materials, Newton-Raphson (N-R) method is usually used in FEM for nonlinear iteration. To accelerate the nonlinear convergence, one effective technique to accelerate the nonlinear convergence is proposed. The basic idea is to use different relaxation factors in different material regions in electric devices. The difference between this proposed method and conventional method is that in conventional method only one relaxation factor is used for all material regions. With the proposed method the total computing time of iteration process of FEM can be reduced effectively. The proposed algorithm is tested on a magnetic field problem of a highly saturated transformer to showcase its effectiveness.

Original language	English
Title of host publication	2015 IEEE International Magnetics Conference, INTERMAG 2015
Publisher	IEEE
ISBN (Electronic)	9781479973224
DOIs	https://doi.org/10.1109/INTMAG.2015.7156784
Publication status	Published - 1 Jan 2015
Event	2015 IEEE International Magnetics Conference, INTERMAG 2015 - Beijing, China Duration: 11 May 2015 → 15 May 2015

Conference

Conference	2015 IEEE International Magnetics Conference, INTERMAG 2015
Country/Territory	China
City	Beijing
Period	11/05/15 → 15/05/15

Keywords

Electric device
finite element method
magnetic field
nonlinear
relaxation factor
transformer

ASJC Scopus subject areas

Surfaces, Coatings and Films
Electronic, Optical and Magnetic Materials
Electrical and Electronic Engineering

More information

10.1109/INTMAG.2015.7156784

Cite this

@inproceedings{fa79793db0234b439c0470ae678448c4,

title = "Nonlinear convergence acceleration of magnetic field computation",

abstract = "Finite element method (FEM) has been widely used to analyze the performance of electric devices. Due to the nonlinear permeability of iron materials, Newton-Raphson (N-R) method is usually used in FEM for nonlinear iteration. To accelerate the nonlinear convergence, one effective technique to accelerate the nonlinear convergence is proposed. The basic idea is to use different relaxation factors in different material regions in electric devices. The difference between this proposed method and conventional method is that in conventional method only one relaxation factor is used for all material regions. With the proposed method the total computing time of iteration process of FEM can be reduced effectively. The proposed algorithm is tested on a magnetic field problem of a highly saturated transformer to showcase its effectiveness.",

keywords = "Electric device, finite element method, magnetic field, nonlinear, relaxation factor, transformer",

author = "Shuangxia Niu and Ho, {Siu Lau} and Weinong Fu",

year = "2015",

month = jan,

day = "1",

doi = "10.1109/INTMAG.2015.7156784",

language = "English",

booktitle = "2015 IEEE International Magnetics Conference, INTERMAG 2015",

publisher = "IEEE",

address = "United States",

note = "2015 IEEE International Magnetics Conference, INTERMAG 2015 ; Conference date: 11-05-2015 Through 15-05-2015",

}

TY - GEN

T1 - Nonlinear convergence acceleration of magnetic field computation

AU - Niu, Shuangxia

AU - Ho, Siu Lau

AU - Fu, Weinong

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Finite element method (FEM) has been widely used to analyze the performance of electric devices. Due to the nonlinear permeability of iron materials, Newton-Raphson (N-R) method is usually used in FEM for nonlinear iteration. To accelerate the nonlinear convergence, one effective technique to accelerate the nonlinear convergence is proposed. The basic idea is to use different relaxation factors in different material regions in electric devices. The difference between this proposed method and conventional method is that in conventional method only one relaxation factor is used for all material regions. With the proposed method the total computing time of iteration process of FEM can be reduced effectively. The proposed algorithm is tested on a magnetic field problem of a highly saturated transformer to showcase its effectiveness.

AB - Finite element method (FEM) has been widely used to analyze the performance of electric devices. Due to the nonlinear permeability of iron materials, Newton-Raphson (N-R) method is usually used in FEM for nonlinear iteration. To accelerate the nonlinear convergence, one effective technique to accelerate the nonlinear convergence is proposed. The basic idea is to use different relaxation factors in different material regions in electric devices. The difference between this proposed method and conventional method is that in conventional method only one relaxation factor is used for all material regions. With the proposed method the total computing time of iteration process of FEM can be reduced effectively. The proposed algorithm is tested on a magnetic field problem of a highly saturated transformer to showcase its effectiveness.

KW - Electric device

KW - finite element method

KW - magnetic field

KW - nonlinear

KW - relaxation factor

KW - transformer

UR - http://www.scopus.com/inward/record.url?scp=84942474095&partnerID=8YFLogxK

U2 - 10.1109/INTMAG.2015.7156784

DO - 10.1109/INTMAG.2015.7156784

M3 - Conference article published in proceeding or book

BT - 2015 IEEE International Magnetics Conference, INTERMAG 2015

PB - IEEE

T2 - 2015 IEEE International Magnetics Conference, INTERMAG 2015

Y2 - 11 May 2015 through 15 May 2015

ER -

Nonlinear convergence acceleration of magnetic field computation

Abstract

Conference

Keywords

ASJC Scopus subject areas

More information

Other files and links

Fingerprint

Cite this