EPE Journal

European Power Electronics and Drives

ISSN: 0939-8368 (Print) 2376-9319 (Online) Journal homepage: http://www.tandfonline.com/loi/tepe20

Soft Magnetic Composite Materials Applied to

Electric Motors

Patricia Jansson

To cite this article: Patricia Jansson (2004) Soft Magnetic Composite Materials Applied to
Electric Motors, EPE Journal, 14:1, 38-42, DOI: 10.1080/09398368.2004.11463552

To link to this article: http://dx.doi.org/10.1080/09398368.2004.11463552

Published online: 22 Sep 2015.

Submit your article to this journal

Article views: 3

View related articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=tepe20

Download by: [Florida Atlantic University] Date: 17 April 2016, At: 22:14
 Patricia Jansson

Soft Magnetic Composite Materials Applied to Electric Motors

Patricia Jansson, Höganäs AB, Höganäs, Sweden

Summary
The advent of soft magnetic composites (SMC) opens the arena for innovative electrical machine designs with
three dimensional flux paths for mass production. The key to commercial success lies in interactive multi-disciplinary
co-operation with parallel development activities.

The SMC materials are based on iron powder particles with surface insulation for low eddy current loss. Insulated
particles, such as those used for dust cores have existed for many years, however these low density products with
permeability values of 150 to 200 and low saturation induction are better suited to high frequency applications. The thin
inorganic insulation of the SMC materials allows heat treatment at 500°C, which results in a partial stress relief and
permeabilities of 500 and above. High flux density requires high component density, which is achieved with high
compressibility powders. The compressibility of the powder is dependent on the surface coating, which must be minimized.
The SMC materials are suitable for large scale mass production of complex components with good tolerances, smooth
Downloaded by [Florida Atlantic University] at 22:14 17 April 2016

surfaces, no secondary operations and no material waste.

Electrical machines with two dimensional flux paths benefit from the use of laminated electrical steel as the anisotropic
properties are put to advantage. Machines such as claw pole motors, transverse flux machines etc. have not achieved
widespread industrial use, as these require a material with isotropic properties suitable for mass production. However
SMC material data alone is insufficient as the sole aspect for design focus. A new design should incorporate all aspects
of the new motor including windings, permanent magnets, reduction in the number of parts and simplification of assembly.
It is first when all the advantages of concurrent engineering are applied that full commercial success can be achieved.

Materials
Introduction

SMC materials for electrical machine applications consist of iron

powder particles, generally with a particle size distribution of less
than 150µm or in the interval 150µm to 400µm, with a thin
inorganic surface insulation, Fig. 1. These are prepared as press
ready mixes by adding various organic additives, the choice of Fig. 1: Schematic representation of the irregular shaped
which depends on the application requirements. The most usual insulated iron powder particle
choice is the lubricant Kenolube™, which ensures that parts are
released from the tooling after compacting. The press ready mix is
filled to a tool cavity, compacted with a pressure, usually in the
interval 500 MPa to 800 MPa according to the required density. A
complex multilevel component with good tolerances is ejected
from the tooling. This proceeds to the additive dependent
heat treatment which, with Kenolube™, is 500°C for 30 minutes
in air, Fig. 2.

This apparently simple process with a limited number of produc-

tion steps offers a wide variation of product property combina-
tions both magnetically and mechanically. One change in a
production parameter at any step will result in one or more pro-
perty changes in the final component. These parameters are well
documented from the materials point of view. Powder metallurgy
manufacturers with a long tradition of compacting and sintering of
components with good tolerances for the automotive industry are
familiar with statistical process control.

Powder grades and process variables

Suitable insulated powders developed for this area are

Somaloy™500 and Somaloy™550. The total loss of the SMC
materials based on Somaloy™ is compared with lamination steel Fig. 2: Schematic representation of the production process
in Fig. 3.

38 EPE Journal ⋅ Vol. 14 ⋅ no 1 ⋅ February 2004

 Soft Magnetic Composite Materials Applied to Electric Motors

These results have been obtained on ring samples (outer diameter

55 mm, inner diameter 45mm and height 5mm in accordance with
IEC standards) under sinusoidal induction using the same hyste-
risis graph instrument. Three low frequency materials are shown,
all of which are based on atomised iron powder with an addition
of 0,5 % Kenolube™, compacted at 800 MPa and heat treated at
500°C for 30 minutes in air. The first grade is ABM100.32
(< 150 µm particle size) with the first generation inorganic surface
insulation. The grades Somaloy™550 (> 150 µm – < 400 µm
particle size) and Somaloy™500 (< 150 µm particle size) have a
thinner, more resistive inorganic surface insulation and illustrate
the effect of larger particle size on the total loss. Somaloy™550
has a higher total loss and a d.c. max. permeability of 550 while
Somaloy™500 combines a lower total loss with a d.c. max. per-
meability of 500. The pure iron, 0,65 mm, laminated steel 1018
and a cold rolled 2,5 % silicon iron steel are used for comparison.
These have been stamped and stacked to form ring samples with
the same outer diameter of 55 mm, inner diameter of 45mm and Fig. 3: Insulated particles with 0,5% Kenolube™, compacted
height 5mm as the compacted powder and wound with primary at 800 MPa, heat treated at 500°C and two lamination steel
and secondary coils for testing on the same instrumentation. grades

It could be argued that the compacting of an iron powder with a

lubricant to minimise particle contact and heating to 500°C, a
Downloaded by [Florida Atlantic University] at 22:14 17 April 2016

temperature well below that of sintering where diffusion bonding

does not occur, could function for these materials. The effect of
surface insulation is illustrated in Fig. 4 where both a pure iron
powder and Somaloy™500 with 0,5 % Kenolube are compacted
at 800 MPa (density 7,34 g/cm3) and heat treated at 500°C for 30
minutes in air. The total loss is only marginally higher at 50 Hz,
but increases rapidly for the pure iron due to the higher proportion
of eddy current loss. The surface insulation withstands the heat
treatment at 500°C and maintains a low eddy current loss, [1].

The surface coating provides insulation, which reduces the eddy

current loss. Other processing routes can be applied when higher
strength is required. Some of these are compared in Table I where
all mixes have been compacted at 800 MPa. Fig. 4: Effect of heat treatment at 500°C on iron powder (Fe),
at a density of 7,34 g/cm3, compared with insulated
The addition of the lubricating binder (0,6 % LB1) permits the use Somaloy™ 500 at a density of 7,34 g/cm3.
of both conventional and warm compacting (WC) resulting in
both high green strength (G.S.) and high transverse rupture Table I: Effect of additives and processing conditions on
strength (TRS) [2]. However heat treatment is confined to a curing the density and strength of Somaloy™500 compacted at
process of 225°C to 275°C for 60 minutes in air so a partial stress 800 MPa
relief is not obtained and the permeability is lower at 200 to 300. Somaloy™500 Heat Treatment Density TRS
Changing the heat treatment atmosphere and temperature with 0,5 (g/cm3) MPa
% Kenolube™ from air to steam can be used to double the TRS
from 50 MPa to 100 MPa. This can only be applied for low 0,5 % Kenolube™ 500°C, air 7,35 50
frequency applications as it also causes an increase in the eddy 0,6 % LB1 275°C, air 7,29 100
current loss. A reduction of lubricant from 0,5 % Kenolube™ to 0,6 % LB1(WC) 275°C, air 7,31 100
0,1 % Kenolube™, using die wall lubrication, will result in higher 0,1 % Kenolube™ 500°C, air 7,40 45
density and improved magnetic properties, Table II.
0,5 % Kenolube™ 470°C, steam 7,35 100
The concept of die wall lubrication has existed for many years but
the introduction to full scale mass production is only now However there are component shapes that are suitable and are in
becoming a reality. This technique cannot be recommended for production today.
all components as yet. Components with thin wall sections and
large heights still present almost insurmountable problems. This section touches a little on the material aspects which, from
powder particle though to finished component, allow for a flexi-

Table II: Comparison of the magnetic properties of Somaloy™ 500 with 0.5% and 0.1% Kenolube™, lubricated die, after
compacting at 800 MPa and heat treatment at 500°C for 30 min. in air. Note the induction (B) at an applied field of 4000 A/m
is designated B40 and the permeability at 50 Hz, 0,5T at the max. applied field (Hmax) is designated µ@Hm.
DC Properties Frequency 50 Hz, 0,5T
800 MPa Density B40 Br µmax Hmax µ@Hm P(W/kg) at
g/cm3 T T A/m 100Hz,1,5T
SOMALOYTM500+ 7,40 1,33 0,31 581 735 532 29
0,1% Kenolube™
Ref. SOMALOYTM500 7,35 1,25 0,29 505 800 505 29
+ 0,5% Kenolube™

EPE Journal ⋅ Vol. 14 ⋅ no 1 ⋅ February 2004 39

 Patricia Jansson

bility of choice for both the magnetic and mechanical properties

combination in the final part. This flexibility may also open for
errors of material choice and/or process during the design phase.
The component shape is the first design step, which must be
considered. Interaction between the machine designer and the
materials and processing expertise ensures the construction of a
production appropriate component. This can result in an increase
of five times the productivity during compaction, which in turn
will have a positive impact on cost. Thus one component may be
larger than another and weight more but cost less. It is therefore
imperative for commercial success that an optimum machine
design, material and process choice are achieved simultaneously.

Applications
Introduction

A comparison of the measured brochure data for lamination steel Fig. 5: DC and 500 Hz hysteresis curves of the Somaloy™
and an SMC material for electrical machine applications shows a 500 material illustrate the dominance of hysteresis loss
permeability of several thousand for lamination steel compared to
500 for SMC. A comparison of the total loss not only at 50 Hz but
even at higher frequency can also show considerably lower loss
Downloaded by [Florida Atlantic University] at 22:14 17 April 2016

for some lamination grades compared to SMC materials.

A permeability of 500 is low compared to the 2000 and more that

can be achieved with lamination steels. This permeability of 500
represents the value in the final components and includes any
building factors as the samples have been compacted with the
same technique as that used for the manufacture of the final
component. It also represents the permeability in all directions.
The total loss is dominated by hysteresis loss with little or no eddy
current loss at low frequency, Fig. 5 and is isotropic.

Application of SMC shows that measured data on samples gives

reasonable correlation with calculations in an application. The
calculated results have been compared with the measured data on Fig. 6: Transverse flux machine
the prototype. The results give reasonable agreement, which
vindicate the calculation method and also indicate that use of the
loss data measured on ring samples gives good results without the
need for building factors, [3]. The correlation between SMC
measured data, application calculations and measured machine
data is confirmed.

A direct material replacement where SMC material replaces a

lamination steel component in a machine with two dimensional
flux is not to be recommended. The full potential of the SMC must
be reached by exploiting the three dimensional capability. This
should be done while exploiting all other aspects of the machine
to their full advantage.

Application examples
Fig. 7: Claw pole motor
The exploratory groundwork with three dimensional designs
using SMC materials in various machines, including the
building and evaluation of prototypes, is summarized in [4]. This
confirmed the advantages anticipated with a three dimensional
material for transverse flux (TFM) and claw pole designs, Figs. 6
& 7. More conventional motors such as a universal motor and ser-
vomotor, Figs. 8 & 9 were also evaluated. The concept of utilizing
the three dimensional component forming capability, to shorten
the length of each copper turn, thus retracting the windings while
extending the back iron and tooth tips was explored. This reduced
the end windings resulting in a shorter motor with the same power.
In the case of the servomotor the single tooth, bobbin winding
concept was applied combined with a 3D tooth shape to retract the
windings. This also reduced the amount of copper used to about
half. Thus the length of the motor was shorter, the copper losses
were halved while the power was maintained.
Fig. 8: Universal motor with SMC stator

40 EPE Journal ⋅ Vol. 14 ⋅ no 1 ⋅ February 2004

 Soft Magnetic Composite Materials Applied to Electric Motors

Fig. 9: Mk 1 servomotor armature Fig. 13: ABS for cars used to drive the oil pump, a
commercial product compliments of Aisin Seiki Co., Ltd
Table III: Servomotor comparison [5]
Laminated Powdered Powdered
Commercial Iron - Iron -
Machine (Prototype 1) pre-pressed
coils
Downloaded by [Florida Atlantic University] at 22:14 17 April 2016

(Prototype 2)
No. Poles 6 6 8
Stator Slots 18 9 12
Active length (mm) 32.5 39.0 32.8
Overall length (mm) 85 55 43.5
Slot Fill Factor 0.39 0.61 0.78
Air-gap length (mm) 0.5 0.5 0.5
Fig. 10: Single tooth with compressed bobbin winding used to Rated Power (W) 848 1445 1602
construct the Mk 2 servomotor [5]
Rated Voltage (V) 415 415 415

Further study and experimentation have lead to more improve-

ments in the servomotor design. In particular the importance of
utilising space and ensuring good thermal conductivity, [5]. The
use of compressed bobbin windings, Fig. 10, and the effect on the
thermal conductivity has been evaluated. “In moving from the
conventional winding to the hand wound bobbin arrangement of
Prototype 1 the fill factor rises from 39 % to 61 %, but with the
pre-pressed coils the fill is even higher, at 78 %” [5].

The comparison, from the commercial machine to the Mk 2 servo-

motor, is shown in Table III. Some important factors are the
reduction in overall length from 85 mm to 44 mm and the increase
in Rated Power from 848 W to 1602 W.

Measured torque versus winding temperature rise curves at rated

speed (3000 rpm) is shown in Fig. 11. The results reveal higher
Fig. 11: Temperature rise against torque at 3000 rpm [5] temperatures for the powdered iron motors at low torque due to
higher iron loss than the laminated motor. As torque builds up the
combination of lower copper loss and better thermal conduction
paths in the powdered iron motors show up markedly and they
out-perform the laminated motor by a considerable margin. Fig. 12
compares the efficiency of the original laminated machine with
prototype 2 as a function of speed at constant torques of 2 Nm and
4 Nm. The lower overall loss of the powdered iron motor is evi-
dent across the whole speed range.

The servomotor shows how the 3D component shaping capability

and the subdivision of parts can be used to advantage in the total
design concept. A material with lower permeability and higher
loss was used to build a smaller motor, which produces 1.9 times
more torque at thermal limit with the same rotor and 3.7 times
more torque per unit volume in the same frame.

This total design concept as a development platform with multi-

Fig. 12: Efficiency against speed at 4 Nm for the servomotor [5] disciplinary involvement has proved successful, as exemplified by
the servomotor.

EPE Journal ⋅ Vol. 14 ⋅ no 1 ⋅ February 2004 41

 Patricia Jansson

Large scale industrial applications benefit from this design con- References
cept approach. Aisin Seiki Co., Ltd has succeeded in designing a
new ABS motor which weights 17 % less and is 36 % shorter [1] Jansson P.: Soft Magnetic Composites – from DC to 1 MHz with
compared to the conventional motor, as shown in Fig. 13. Aisin Iron Powder, Workshop, Powder Metallurgy World Congress,
Seiki were not content with only a 3D component design, the Granada, Spain, Oct. 1998
concept to drastically reduce the low function spaces provides,
downsizing and weight reduction. Aisin Seiki have taken this [2] Jansson P.: Soft Magnetic Composites - A Rapidly Expanding
ABS motor from design concept to prototypes and is now in full Materials Group, PM2TEC’99, Vancouver, Canada, June 1999.
scale industrial production for the automotive market. [3] Jansson P.; Jack A. G.: Magnetic Assessment of SMC materials”
Twenty First Annual Conference On Properties and Applications of
Magnetic Materials, Chicago, 13th to 15th May 2002.
Conclusions [4] Jack A. G.: Exploitation of Soft Magnetic Composites for
Electrical Machines, Workshop, Powder Metallurgy World
The powder particle though the process steps to finished compo- Congress, Granada, Spain, Oct. 1998.
nent, allows for a flexibility of choice for both the magnetic and
mechanical properties combination in the final part. Interaction [5] Jack, A.G., Mecrow, B.C., Dickinson, P.G., Stephenson, D.,
between the machine designer and the materials and processing Burdess, J.S., Fawcett, J.N., Evans, T.: Permanent magnet
expertise ensures the construction of a production appropriate Machines with Powder Iron Cores and Pre-Pressed Windings,
component. It is imperative for commercial success that an opti- 1999 IEEE Industry Applications Conference, 34th Annual
mum machine design, material and process choice are achieved Meeting, Phoenix Arizona, USA. 3-7 October 1999, pp97-103.
simultaneously.

A direct material replacement where SMC material replaces a The Author

Downloaded by [Florida Atlantic University] at 22:14 17 April 2016

lamination steel component in a machine with two dimensional

flux is not to be recommended. The full potential of the SMC must Patricia Jansson has been working for Höganäs AB for more than 20
be reached by exploiting the three dimensional capability. This years, Most of this time was at R&D dept. where 15 years were devoted
should be done while exploiting all other aspects of the machine to the development of soft magnetic materials. After completing
to their full advantage. development of the first SMC materials including Somaloy500 she
moved from a position as Senior researcher to the marketing department
This total design concept as a development platform with multi- as applications manager, thus supporting application projects where
disciplinary involvement has proved successful. Somaloy materials are commercially applied.

European Power Electronics and Drives Journal

Guidelines for the preparation of a paper to be published in EPE journal

Submitting a paper Paragraphs are separated by a line space.

The papers will be submitted by e-mail (Word or .pdf), one paper Titels are in bold with level indication: size or number. They finish
copy will be sent as well, including full coordinates. E-mail without a period(.).
address: [email protected]. Mail address: EPE Secretariat, c/o Figure captions will be included in the text and will be of the kind:
VUB-TW, Pleinlaan 2, B-1050 Brussels, Belgium Fig. 1: Typical figure caption

Writing the paper The originals of the figures will be given : ink drawings, pho-
tographs, lazer print out...
The paper will be written in English. Figures can be included in the general file.

The title will be chosen short and expressive. It will be followed by Figures and photographs will be sent back after use, upon request.
the authors' names, title, function and address and by keywords.
Deadlines
The paper will start with a summary that will allow non-specialists
to understand the problem, have an idea of the state of the art, and EPE Journal is published quarterly. Submitted papers undergo a
understand the originality of the described solutions. The summary reviewing process of about 6 months. Publication dates are as
will be half a typewritten page long and refer to a figure. The follows:
figure will be particularly clear and eloquent.
Vol. 14 no 2: out 15/05/2004, normal issue with peer reviewed
The end of the paper will include a list of references and the papers
authors' curriculi and photographs.
Vol. 14 no 3: out 15/08/2004, special issue on multi-phase, multi-
When the paper is accepted motor drives

Texts are sent on floppy disks with indication of the format and a Vol. 14 no 4: out 30/11/2004, normal issue with peer reviewed
paper copy. papers

The text will be typewritten without page set-up (no columns, no Vol. 15 no 1: out 15/02/2004, normal issue with peer reviewed
styles). papers

42 EPE Journal ⋅ Vol. 14 ⋅ no 1 ⋅ February 2004