Open AccessAccuracy of time domain extension formulae of core losses in non‐oriented electrical steel laminations under non‐sinusoidal excitation
Author(s) -
Alatawneh Natheer,
Rahman Tanvir,
Hussain Sajid,
Lowther David A.,
Chromik Richard
Publication year - 2017
Publication title -
iet electric power applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.815
H-Index - 97
eISSN - 1751-8679
pISSN - 1751-8660
DOI - 10.1049/iet-epa.2016.0737
Subject(s) - waveform , eddy current , pulse width modulation , time domain , excitation , frequency domain , electrical steel , hysteresis , core (optical fiber) , materials science , mathematics , control theory (sociology) , voltage , acoustics , physics , mathematical analysis , computer science , engineering , electrical engineering , control (management) , quantum mechanics , artificial intelligence , composite material , computer vision
This study presents a comparative study on the accuracy of three iron loss prediction models. The models are based on the decomposition of core or iron losses into the hysteresis and the eddy current loss components. The time domain extensions of two frequency domain models have been used to predict the iron losses due to a number of non‐sinusoidal waveforms with and without the presence of minor loops. A third model, by Boglietti, that has been proposed recently to predict core losses for non‐sinusoidal and pulse‐width modulated (PWM) waveforms has also been studied. The unknown coefficients of each model have been determined by data fitting iron losses obtained from Epstein frame experiments for induction levels and fundamental frequencies up to 1.6 T and 2 kHz, respectively. Core losses due to PWM waveforms have been measured at various fundamental and switching frequencies in unipolar and bipolar modes. The experimentally measured iron losses have been compared to those predicted using the three models and the accuracy and applicability of each model have been discussed.