In this paper, an interior permanent magnet synchronous motor (IPMSM) with a stator core made of amorphous magnetic material (AMM) is presented. The IPMSM is driven by a voltage source three-phase inverter with classical pulse width modulation (PWM) control. The core losses under no-load condition are measured by experiment and compared to an equivalent IPMSM with a stator core made of NO steel. Under these conditions, the core losses are influenced by the stator, rotor and magnet shapes but also by the PWM carrier signal that implies a high frequency harmonic in the magnetic flux density. It is demonstrated that the AMM can reduce the core losses by about 56 %.

Amorphous magnetic materials (AMM) offer low iron loss compared to conventional non-oriented (NO) silicon steel. This is especially true in high frequencies and the AMM can then bring substantial improvement for high speed motor applications.1,2 Moreover, AMM is also successfully used for radial-field switched reluctance motor (SRM).3 In Ref. 3, a core loss reduction of about 63 % is found at 8500 rpm, compared to the same machine using low loss silicon steel. The proposed SRM’s efficiency is competitive with a permanent magnet synchronous motor (PMSM). Another application in Ref. 4 shows that an induction motor with an AMM stator core offers a reduction of the core losses by almost 50 % at no-load and sinusoidal-supplied voltage. Taking advantage of the manufacture ease, AMM is also often used for axial-field PMSMs.1,5–7 Axial-field PMSMs are however not suited for all applications and in Ref. 8, a radial-field synchronous motor with surface-mounted magnets (SPMSM) is developed. Only the stator teeth are made of AMM but the authors still measure a core loss reduction of about 40 % at 5500 rpm.

In a previous work,9 a radial-field synchronous motor with buried magnets (IPMSM) and a full stator made of AMM was proposed. Its core losses under no-current condition were measured. The no-current condition represents the state where the rotor is driven by an external motor and the stator windings are in open-circuit. A reduction by about 50 % compared to the same motor using only NO silicon steel was obtained. The same type of motor is also proposed in Ref. 10 but the authors focus on using AMM to increase the power density using geometrical optimization. In other words, the core losses are not specifically investigated in Ref. 10.

This paper is based on the previous work9 and this time, the core losses are measured under inverter-fed and no-load condition. Compared to the no-current condition in Ref. 9, the inverter pulse width modulation (PWM) is responsible for high frequency harmonics in the magnetic flux density that increase the core losses.11–13 

Two IPMSMs with identical geometries have been manufactured. The first, hereafter called NO-IPM, has a laminated stator made of standard NO silicon steel sheets, and the second, hereafter called AMM-IPM, has a laminated stator made of AMM. The material characteristics are described in Table I.

TABLE I.

Material characteristics.

NO AMM
Reference  35H300  2605HB1M 
Composition  Fe-Si  Fe-Si-B 
Thickness (μm)  350  25 
Saturation magnetic flux density (T)  2.12  1.63 
Resistivity (μΩ.m)  0.5  1.20 
Iron loss density at 50 Hz and 1 T (W/kg)  1.119  0.171 
Iron loss density at 50 Hz and 1.5 T (W/kg)  2.419  0.400 
NO AMM
Reference  35H300  2605HB1M 
Composition  Fe-Si  Fe-Si-B 
Thickness (μm)  350  25 
Saturation magnetic flux density (T)  2.12  1.63 
Resistivity (μΩ.m)  0.5  1.20 
Iron loss density at 50 Hz and 1 T (W/kg)  1.119  0.171 
Iron loss density at 50 Hz and 1.5 T (W/kg)  2.419  0.400 

An axial cross-section view of the proposed motor is illustrated in Fig. 1 and its geometrical characteristics are listed in Table II.

FIG. 1.

Quarter axial cross-section view of the motor. The cross-section of the full motor has horizontal and vertical reflection symmetry.

FIG. 1.

Quarter axial cross-section view of the motor. The cross-section of the full motor has horizontal and vertical reflection symmetry.

Close modal
TABLE II.

IPMSM geometrical characteristics.

Poles / Slot number  8/12 
Radius of stator core Rso  64 mm 
Radius of rotor core Rr  37 mm 
Air gap g  1.25 mm 
Yoke width Wy  9.2 mm 
Tooth width Wt  10 mm 
Magnet length LPM  20 mm 
Magnet thickness WPM  2 mm 
Core length Lc  47 mm 
Winding method  Concentrated 
Number of winding turns  37 
Poles / Slot number  8/12 
Radius of stator core Rso  64 mm 
Radius of rotor core Rr  37 mm 
Air gap g  1.25 mm 
Yoke width Wy  9.2 mm 
Tooth width Wt  10 mm 
Magnet length LPM  20 mm 
Magnet thickness WPM  2 mm 
Core length Lc  47 mm 
Winding method  Concentrated 
Number of winding turns  37 

The non-crystalline amorphous alloy ribbon is made using rapidly quenched casting technology. The ribbon is then submitted to an annealing step at 400°C under longitudinal magnetic field in order to reduce its specific iron loss, relieve the internal stress and improve the magnetic isotropy.14,15 The ribbon has a high tensile strength of 2100 MN/m2 but has a low out-of-plane shearing strength, mostly due to its thinness. As a consequence, special techniques are adopted for the stator core manufacture. The amorphous sheets are first stacked and cut into the stator shape using a wire-cut technique. The cut stack is then placed into a vacuum box and immersed into resin for lamination coating and bonding. The stack is then submitted to high pressure in order to make the resin fill the interlaminations. Finally, the bonded stack is heated at a temperature of 180°C for 150 min. This heating step is part of the core manufacture process to harden the resin and obtain a good quality coating. Even though annealing of AMM core has been tried by some authors to improve its overall magnetic characteristic,2 the heating step performed here is not a strictly speaking annealing as the temperature remains quite low and the magnetic characteristic is assumed almost unchanged.

The stacking factor of the AMM stator core is 0.86 and that of the NO stator is 0.99.

Each IPMSM is star-connected. In order to compare the back-EMF of the two motors, each IPMSM is rotated by an external brushless DC motor and the IPMSM’s windings set in open circuit. The voltage measured between one of the phase and the neutral point gives the back-EMF. The measured back-EMF waveforms of both motors have been found very similar. The peak value of the fundamental of the back-EMF has been measured from 750 rpm to 3000 rpm and the back-EMF constant has been estimated by a linear regression on the experimental points. The NO-IPM has a back-EMF constant of 26.9 V/krpm while the one of the AMM-IPM is 26.6 V/krpm. Even though the magnetic flux density saturation point of the AMM is lower than for NO (Table I), the back-EMF constants are very similar because the magnetic saturation is almost never reached in both cases. It could however have an impact at high current under load condition but this point is not discussed in this paper.

The measured total IPMSM core losses are the hysteresis and eddy-current losses in the magnets, rotor core and stator core.

The experimental test bench of the no-load test is illustrated in Fig. 2. During this test, the IPMSM is driven by a three-phase voltage source inverter with IGBTs. A standard vector control using a fixed frequency PWM is implemented for the speed control. The d-axis current reference is zero and the PWM carrier frequency is 10 kHz. The inverter input voltage Vdc is provided by a DC power source and tuned to obtain a PWM modulation ratio of approximately 1. No load is connected to the IPMSM rotor shaft but an encoder is still needed to provide a feedback on the rotational speed.

FIG. 2.

Experimental test-bench of the no-load test.

FIG. 2.

Experimental test-bench of the no-load test.

Close modal

The IPMSM input active electrical power P is measured by a power analyzer using the equation

(1)

where iu,v,w are respectively the u, v, and w phase currents, vu,v,wn are respectively the u, v, and w phase to neutral voltages and n is the sample number in an integral number of electrical periods. The measurement sampling frequency is 2.2 MHz. In order to calculate the copper losses, the rms current has also been calculated from the sampled current waveform using the equation

(2)

where x can be u, v or w depending on the considered phase.

The core losses under no-load can then be calculated using

(3)

where Rs is the phase to neutral winding resistance, Pmech and Penc are the mechanical losses of the IPMSM and the encoder respectively. The mechanical losses are previously measured by substituting the rotor with magnets by a rotor without magnets, but with the same shape and identical bearings. The IPMSM without magnets is then rotated by an external brushless DC motor and a torque meter is used to measure the torque needed to rotate the shaft at a given speed. The obtained torque represents the mechanical losses.

According to the Steinmetz equation, the iron losses mainly depend on the peak value and frequency of the magnetic flux density. Consequently it is interesting to measure the magnetic flux density during operation. One B coil of 10 turns is placed on each of the four teeth of the u phase coils (as illustrated in Fig. 1). The B coils provide a voltage value VB used to calculate the average magnetic flux density B through the tooth cross section area St. The corresponding equation is

(4)

where NB is the number of turns of the B coil and St is given by

(5)

The magnetic flux density in the four teeth has been measured in both motors at a speed of 750 rpm. The comparison is illustrated in Fig. 3. There is a little deviation between the peak values in each tooth, probably due to a misalignment of the rotor. The average peak value among the four teeth is 0.907 T for the NO-IPM stator and 0.883 T for the AMM-IPM stator. The peak values are very similar despite the difference in the magnetic saturation of the materials. Moreover, since the stator and rotor shapes are the same, the waveform shapes are also similar. The two stators are subject to almost similar conditions in terms of magnetic flux density variations but, as will be seen in the next section, the core losses of the AMM-IPM will be sensibly lower than for the NO-IPM.

FIG. 3.

Magnetic flux density through the teeth cross-section (four u phase teeth). a) NO-IPM b) AMM-IPM.

FIG. 3.

Magnetic flux density through the teeth cross-section (four u phase teeth). a) NO-IPM b) AMM-IPM.

Close modal

The core losses have been measured at 750 rpm, 1500 rpm, 2250 rpm, and 3000 rpm. It should be noted that the measurement of the mechanical losses using the torque meter is rather delicate because the torque to measure is low and the measurement subject to uncertainties. Fig. 4 reports the core losses of both IPMSMs under no-load and compares them with the core losses under no-current condition.9 The core losses under no-load condition are superior because of the effect of the PWM carrier frequency on the magnetic flux density time harmonics.12,13 When the NO stator is used, this increase is 22 % in average, against 9 % for the AMM stator. The core losses caused by the carrier frequency in the AMM-IPM are then smaller than in the NO-IPM, which confirms the benefit of using AMM at high frequencies (carrier frequency of 10 kHz and fundamental frequency of 200 Hz at 3000 rpm). Overall, the average decrease obtained by using AMM instead of NO in the stator core is about 56 % in no-load condition. This is 6 % more than the decrease obtained in no-current condition.

FIG. 4.

Core loss comparison between NO-IPM and AMM-IPM.

FIG. 4.

Core loss comparison between NO-IPM and AMM-IPM.

Close modal

In this paper, it is found that replacing the NO stator core of the proposed IPMSM by an AMM stator core of the same shape leads to a reduction of the core losses by about 56 % under inverter-fed and no-load condition. This result confirms the benefit of using AMM to reduce the motor core losses. Future research will concentrate on evaluating the core losses for different values of the PWM carrier frequency.

This work is partially supported by the Ministry of Education, Culture, Sports, Science and Technology program, Japan, for private universities and by KAKENHI (26420259).

1.
Z.
Wang
,
Y.
Enomoto
,
R.
Masaki
,
K.
Souma
,
H.
Itabashi
, and
S.
Tanigawa
, inProc. IEEE 8th Int. Conf. on Power Electronics and ECCE Asia, Jeju, Korea, 2011, pp. 1940-1945.
2.
R.
Kolano
,
K.
Krykowsky
,
A.
Kolano-Burian
,
M.
Polak
,
J.
Szynowski
, and
P.
Zackiewicz
,
IEEE Trans. Magn.
49
(
4
),
1367
(
2013
).
3.
A.
Chiba
,
H.
Hayashi
,
K.
Nakamura
,
S.
Ito
,
K.
Tungpimolrut
,
T.
Fukao
,
M. A.
Rahman
, and
M.
Yoshida
,
IEEE Trans. Ind. Appl.
44
(
3
),
699
(
2008
).
4.
M.
Dems
and
K.
Komeza
,
IEEE Trans. Ind. Electron.
61
(
6
),
3046
(
2014
).
5.
Z.
Wang
,
Y.
Enomoto
,
M.
Ito
,
R.
Masaki
,
S.
Morinaga
,
H.
Itabashi
, and
S.
Tanigawa
,
IEEE Trans. Magn.
46
(
2
),
570
(
2010
).
6.
G. S.
Liew
,
W. L.
Soong
,
N.
Ertugrul
, and
J.
Gayler
, inProc. 20th Australasian Universities Power Engineering Conf., Christchurch, New Zealand, 2010, pp. 1-6.
7.
Z.
Wang
,
R.
Masaki
,
S.
Morinaga
,
Y.
Enomoto
,
H.
Itabashi
,
M.
Ito
, and
S.
Tanigawa
,
IEEE Trans. Ind. Appl.
47
(
3
),
1293
(
2011
).
8.
Y.
Enomoto
,
M.
Ito
,
H.
Koharagi
,
R.
Masaki
,
S.
Ohiwa
,
C.
Ishihara
, and
M.
Mita
,
IEEE Trans. Magn.
41
(
11
),
4304
(
2005
).
9.
S.
Okamoto
,
N.
Denis
,
Y.
Kato
,
M.
Ieki
, and
K.
Fujisaki
, Core Loss Reduction of an Interior Permanent Magnet Synchronous Motor Using Amorphous Stator Core,” accepted for publication inIEEE Trans. Ind. Appl.
10.
T.
Fan
,
Q.
Li
, and
X.
Wen
,
IEEE Trans. Ind. Electron.
61
(
9
),
4510
(
2014
).
11.
K.
Yamazaki
and
Y.
Seto
,
IEEE Trans. Ind. Appl.
42
(
4
),
1045
(
2006
).
12.
K.
Fujisaki
,
R.
Yamada
, and
T.
Kusakabe
,
IEEJ Trans. Ind. Appl.
133
(
1
),
69
(
2012
).
13.
N.
Denis
,
K.
Fujitani
,
Y.
Kato
,
M.
Ieki
, and
K.
Fujisaki
, inProc. Int. Conf. Electrical Machines and Systems (ICEMS), Pattaya, Thailand, 2015, pp. 1-7.
14.
S.
Ueno
,
T.
Todaka
, and
M.
Enokizono
,
IEEE Trans. Magn.
47
(
10
),
3188
(
2011
).
15.
H.
Fujimori
,
H.
Yoshimoto
,
T.
Masumoto
, and
T.
Mitera
,
J. of Applied Physics
52
,
1893
(
1981
).