Comparison of full and reduced potential formulations for low-frequency applications

Author

Bedrosian, Gay ; Chari, M.V.K. ; Joseph, James

Author_Institution

General Electric Corp. Res. & Dev., Schenectady, NY, USA

Volume

29

Issue

2

fYear

1993

fDate

3/1/1993 12:00:00 AM

Firstpage

1321

Lastpage

1324

Abstract

The authors examine five alternative finite element formulations of magnetostatic problems: (1) full vector potential, (2) reduced vector potential, (3) reduced scalar potential with source integral over all volume, (4) reduced scalar potential with source integral over permeable volumes only, and (5) reduced scalar potential with surface source integral. They show that (3) is more prone to numerical errors than the other methods, and that (4) and (5) are competitive in accuracy with (1) and (2) in air regions but are easier to implement in three dimensions. However, all three scalar formulations suffer from cancellation errors in iron regions

Keywords

error analysis; finite element analysis; magnetic fields; magnetostatics; vectors; accuracy; air regions; finite element formulations; full vector potential; iron regions; low-frequency applications; magnetostatic problems; numerical errors; permeable volumes; reduced scalar potential; reduced vector potential; source integral; surface source integral; three dimensions; Design optimization; Finite element methods; Integral equations; Iron; Machinery; Magnetic fields; Magnetic flux; Magnetostatics; Maxwell equations; Permeability;

fLanguage

English

Journal_Title

Magnetics, IEEE Transactions on

Publisher

ieee

ISSN

0018-9464

Type

jour

DOI

10.1109/20.250641

Filename

250641