Efficient modeling of coil filament losses in 2D

Leena Lehti, Janne Sami Keränen, Saku Suuriniemi, Timo Tarhasaari, Lauri Kettunen

    Research output: Contribution to journalArticleScientificpeer-review

    2 Citations (Scopus)

    Abstract

    Purpose: The authors aim to search for a practical and accurate way to get good loss estimates for coil filaments in electrical machines, for example transformers. At the moment including loss estimations into standard finite element computations is prohibitively expensive for large coils. Design/methodology/approach: A low-dimensional function space for finite element method (FEM) is introduced on the filament-air interface and then extended into the filament to significantly reduce the number of unknowns per filament. Careful choice of these extensions enables good loss estimate accuracy. The result is a system matrix assembly block that can be used verbatim for all filaments, further reducing the cost. Both net current and voltage per length of the filament are readily available in the problem formulation. Findings: The loss estimates from the developed model agree well with traditional FEM and the computation times are faster. Originality/value: To produce accurate loss estimates in large coils, the low-dimensional function space is constricted on the filament boundaries. The proposed method enables electrical engineers to compute the ohmic losses of individual conductors.
    Original languageEnglish
    Pages (from-to)1631-1642
    JournalCOMPEL: The International Journal for Computation and Mathematics in Electrical and Electronic Engineering
    Volume32
    Issue number5
    DOIs
    Publication statusPublished - 2013
    MoE publication typeA1 Journal article-refereed

    Keywords

    • coil modeling
    • electric coils
    • electrical equipment
    • FEM
    • finite element analysis
    • winding loss estimate

    Fingerprint

    Dive into the research topics of 'Efficient modeling of coil filament losses in 2D'. Together they form a unique fingerprint.

    Cite this