Professional Documents
Culture Documents
Department of Energy
FreedomCAR and Vehicle Technologies, EE-2G
1000 Independence Avenue, S.W.
Washington, D.C. 20585-0121
FY 2005
Prepared by:
Submitted to:
June 2005
ORNL/TM-2005/86
Prepared by the
OAK RIDGE NATIONAL LABORATORY
Oak Ridge, Tennessee 37831
managed by
UT-BATTELLE, LLC
for the
U.S. DEPARTMENT OF ENERGY
Under contract DE-AC05-00OR22725
This report was prepared as an account of work sponsored by an agency of the
United States Government. Neither the United States Government nor any
agency thereof, nor any of their employees, makes any warranty, express or
implied, or assumes any legal liability or responsibility for the accuracy,
completeness, or usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately owned rights.
Reference herein to any specific commercial product, process, or service by trade
name, trademark, manufacturer, or otherwise, does not necessarily constitute or
imply its endorsement, recommendation, or favoring by the United States
Government or any agency thereof. The views and opinions of authors expressed
herein do not necessarily state or reflect those of the United States Government
or any agency thereof.
CONTENTS
Page
INTRODUCTION................................................................................................................... 1
BACKGROUND ..................................................................................................................... 1
CONCLUSIONS ..................................................................................................................... 37
REFERENCES........................................................................................................................ 38
DISTRIBUTION ..................................................................................................................... 39
ii
LIST OF FIGURES
Figure Page
iii
LIST OF FIGURES (contd)
Figure Page
iv
LIST OF FIGURES (contd)
Figure Page
25 Relationship between magnet material, yI_1, speed, and advance for the motor
having dual-layer U-shaped IPMs:
(a) Minimum magnet material corner view............................................................... 31
(b) Advance contour view ......................................................................................... 31
26 Relationship between magnet amount, yI_1, speed, and power for the motor
having dual-layer U-shaped IPMs:
(a) Minimum magnet material corner view............................................................... 32
(b) Maximum magnet material corner view.............................................................. 32
27 Relationship between magnet amount, yI_1, speed, and percentage of
reluctance torque for the motor having dual-layer U-shaped IPMs .......................... 33
28 CPSR as a function of magnet amount, yI_1, for the motor having dual-layer
U-shaped IPMs .......................................................................................................... 33
29 Speed and normalized cost and torque for the motor having dual-layer
U-shaped IPMs:
(a) View from low-cost corner .................................................................................. 34
(b) Torque contour plot ............................................................................................. 34
30 Performance comparison of IPM motor having double-layer U-shaped
magnets with surface-mounted PM motor having only PM torque:
(a) Torque performance comparison ......................................................................... 35
(b) Power performance comparison .......................................................................... 35
31 Performance comparison of four motors with varying degrees of reluctance:
(1) surface-mounted PM motor producing only PM torque; (2) inset PM/reluctance
motor with magnet pitch, M = 125; (3) V-shaped IPM motor with M = 116;
and (4) double-layer U-shaped IPM with yI_1 = 22.4 mm
(a) Torque performance comparison ......................................................................... 36
(b) Power performance comparison .......................................................................... 36
v
INTRODUCTION
The international research community has lately focused efforts on interior permanent magnet
(IPM) motors to produce a traction motor for hybrid electric vehicles (HEV). One of the
beneficial features of this technology is the additional torque produced by reluctance. The
objective of this report is to analytically describe the role that reluctance plays in permanent
magnet (PM) motors, to explore ways to increase reluctance torque without sacrificing the torque
produced by the PMs, and to compare three IPM configurations with respect to torque, power,
amount of magnet material required (cost), and percentage of reluctance torque. Results of this
study will be used to determine future research directions in utilizing reluctance to obtain
maximum torque and power while using a minimum amount of magnet material.
BACKGROUND
Once a PM machine is built, the strength and number of magnets in the rotor and the number of
poles and turns in the coils of the stator remain constant; thus, the amount of PM-generated
magnetic flux linked by the coils of the stator remains fixed. As a result, the back-electromotive-
force (back-emf) voltage induced by the PMs increases linearly with the speed of the rotor.
As the rotor speed increases the back-emf voltage rises, which results in a rapid reduction in the
available voltage, (the difference between the supply voltage and the back-emf). When there is
no longer any voltage available to drive current into the stator, the maximum speed has been
reached.
Reluctance is a property used in magnetic circuits that accounts for the ratio of geometric length
to area and the magnetic properties of a medium, such as permeability, , in which a magnetic
flux flows. The expression for reluctance is R = l /( A ) , where l is the length of the flux path
and A is the area normal to the flux. Reluctance relates magnetic flux to magnetomotive force in
a manner similar to the way that resistance and inductance relate current to voltage in an electric
circuit. Since the externally controllable measurement parameters of a motor are electric, it is
customary to study the performance of motors in terms of electric rather than magnetic
quantities. The relationship between magnetic reluctance and electric inductance is complex
except for the simplest of cases, like that of a coil wrapped with Nt turns of wire around a
magnetic path of reluctance, R . This simple relationship is
L = Nt2/R. (1)
As Eq. (1) shows, electric inductance is proportional to the inverse of magnetic reactance.
Consequently, high inductance corresponds to low reluctance.
Two distinct reference axes exist in most motors, the direct axis (d-axis) and the quadrature axis
(q-axis). They are perpendicular in terms of electric degrees. Physically, they correspond to those
rotor positions that yield the maximum and minimum amounts of magnetic flux linked in the
stator coils. In PM motors, the d-axis corresponds to the center of a rotors PM, while the q-axis
corresponds to the midpoint in the space separating a PM from its closest PM of different
polarity. These axes are identified later in this report in Fig. 3 through Fig. 5.
1
Figure 1 shows the electric root-mean-square (RMS) phasor diagram representative of all PM
motors driven by sinusoidal voltages with the stator current in phase with the PM motors back-
emf. Note that it is customary to show RMS phasor diagrams with the rotor in the d-axis position
because the PMs magnetic flux linked by the stator coils is at a maximum, thus generating the
highest back-emf. The meaning of the symbols used in this figure, and in the rest of this report,
are as follows:
q-axis
Er = LI
V
R*I
Epm = pm
I
r pm
d-axis
The stator current, I, is in phase with the back-emf, Epm, induced by the PM.
The stator reaction magnetic flux, r, is perpendicular to the PM flux.
The stator reaction back-emf is perpendicular to the stator current.
The power input is: Pin = I V cos() = I (Epm + I R)
The Power output is: Pout = I Epm
The tip of the back-emf, Epm, moves up as the speed increases causing the
current to diminish. The tip of the terminal voltage, V, moves clockwise on the
circumference until it reaches the summit at the maximum speed:
max= V/pm
2
As shown in the extensive caption of Fig. 1, higher-strength PMs reduce the maximum speed of
operation.
Since locomotion applications tend to prefer motors with wide ranges of speed to eliminate the
need for or reduce the complexity of gear-boxes, it is important to find ways to increase the
speed range of PM motors. One obvious possibility is to use a multilevel voltage source where
available voltage would increase with rotor speed. Drive costs, already the major component of a
PM drive system, may limit this approach to two stages, with the higher stage enabled at the
higher speeds. Presently, improvements in stator and rotor topologies together with control
approaches have enabled better overall performance/cost potential.
Comment: I think motor cost/kW are higher than drive (inverter) costs/kW
In PM motors, optimal torque production occurs when the back-emf voltage and the stator
current are in phase. When they are not, the current component normal to the back-emf produces
a magnetic field that opposes or reinforces that of the PM, depending on its direction. The time-
dependent linked flux then includes stator reaction components in both the d- and the q-axes,
such that
d = pm + rd = pm + Ld id (2)
q = rq = Lq iq ,
where
iq is the current component in-phase with the PMs back-emf (q-axis direction)
which is the primary torque producing component of current,
id is the current component normal to the back-emf (d-axis direction) which is the
flux producing component of current,
rd = Ld id represents the flux linkages created by the stator currents id component,
rq = Lq iq represents the flux linkages created by the stator currents iq component,
Ld and Lq are the inductances when the rotor is in the d-axis and q-axis positions.
Equation (2) shows that when id is negative, the reaction field has a d-axis component opposing
that of the PM, thereby weakening the flux. This in turn causes a reduction in the q-axis back-
emf voltage; consequently, higher Iq currents at the same or higher speeds for the same total
current would be possible, thus extending the motors operating speed range. This extension is
achieved at the cost of additional ohmic losses in the stator, but it reduces the angle, , between
the stators voltage, V, and current, I, thus improving the power factor.
Figure 2 shows the RMS phasor diagram describing the operation of a generic PM motor driven
sinusoidally with flux-weakening since the current is in the negative d-axis direction. In it, the
meaning of the symbols is the same as for Fig. 1, with terms added that represent d- and q-axis
components Id, Iq, Ld, Lq, rd, rq, Erd, Erq of I, L, r, and Er, respectively.
3
q-axis
Erd = Iq Lq
Id R
Erq= Id Ld
Iq R
V
Epm = pm
I Iq
rq
Id pm
d-axis
Er rd
The power input is: Pin = I V cos()
= Iq (Epm + Iq R- Id Ld ) + Id (Id R +Iq Lq )
= Iq Epm + (Iq2+ Id2 )R + Id Iq (Lq - Ld )
The power output is: Pout = Iq (Epm - Ld Id ) + Id (Lq Iq )
= Iq (Epm + Id (Lq Ld) ) = Iq (Epm + Id Ld ( 1) )
For surface mounted PMs, where Lq = Ld, Id in the negative direction weakens the
magnet so that the motor may be driven at higher speeds. However, there is no increase
in output power, while the input power must increase to supply the additional resistance
heat loss.
For IPM motors Lq >Ld thus if Id is along the negative d-axis as shown above, the new
term introduced by the presence of Id has a positive power.
As discussed in the caption of Fig. 2, in addition to flux weakening, extra mechanical power is
produced when the saliency ratio, defined as
= Lq/Ld , (3)
Characteristic Current
The magnitude of the d-axis current needed to completely cancel the PM magnetic flux is often
referred to as the motors characteristic current. In the open literature the characteristic current
is also identified as Ix (3). By setting d = 0, in Eq. (2) we obtain
4
Surface-mounted PMs, for which Lq = Ld, operated with an amount Ich of d-axis current in the
negative direction, would have no speed limit since the back-emf induced by the PM would be
zero. Unfortunately, the inductance, Ld, is small in PM motors; consequently, the magnitude of
the characteristic current is large, showing that large currents cause only small amounts of flux
weakening.
The power output for a PM with an Id equal to the characteristic current, Ich, in Eq. (4) injected in
the negative direction would be
IPM motors differ in that they have the PMs inserted in the body of the rotor. This not only
provides mechanical support for the magnets and protects them from the environment and from
demagnetization, but also makes the inductance of the rotor position dependent, so that Ld < Lq.
Consequently, when an Id current is inserted in the negative direction shown in Fig. 2 to weaken
the flux, extra mechanical power is generated in the motor. IPMs are thus considered as hybrids
of PM and reluctance technologies. In addition, in IPMs the boundary between the rotor poles
and the air-gap is high-permeability silicon steel. It allows phase-advance to give the PM flux a
tangential component within the rotor pole, thus skewing the air-gap flux distribution to such a
degree that stator teeth are saturated. This allows for an externally controlled variation in
reluctance that is not possible in surface-mounted PM motors.
has two components. In IPMs, the reluctance term is not zero and contributes to the motors
output.
Equation (7) shows that the total torque output can be maintained while reducing the PM torque
by compensating with an increase in the reluctance torque. When the strength of the PM flux is
reduced, pm and the magnitude of the characteristic current are also reduced. This makes flux-
weakening by insertion of negative current in the d-axis less costly and more practical.
Solving Eq. (6) for Epm, substituting Epm = pm , and taking the partial derivative of pm with
respect to for fixed Pout, , Id, Iq, and Ld, we obtain the relationship between the change of PM
strength needed and a change in the degree of saliency,
pm = - Id Ld . (8)
5
That is, an increase of saliency would result in a decrease in magnet strength required
proportional to the product of Id Ld, the values of the d-axis inductance and d-axis current
component.
Id
pm= k e - , (10)
I ch
where k is an integration constant. This is an important relationship since it shows that the
magnet strength needed diminishes exponentially with the degree of saliency and with the
amount of d-axis current injected relative to the value of the characteristic current. Therefore,
decreasing the value of the characteristic current and increasing the saliency are clear goals for
improving PM motor performance.
(a) Equation (4) shows that to reduce the magnitude of a motors characteristic current, one
may (1) reduce the magnitude of the PM flux links, and/or (2) increase the value of the d-
axis inductance.
(b) Equation (6) shows that IPM motors are superior to surface-mounted motors for flux
weakening since under flux-weakening conditions they provide extra mechanical power
output because their saliency is larger than unity.
(c) Equation (8) shows the desirability of increasing (1) the saliency, and (2) the d-axis
inductance in order to reduce the amount of PM flux linkage needed by a motor. This
would further enhance the motors flux-weakening capabilities and reduce the cost of
PMs while maintaining the motors power level.
(d) Equation (10) shows that for equal motor power output, the magnet strength needed
diminishes exponentially with the degree of saliency and with the amount of d-axis
current injected relative to the value of the characteristic current. Therefore, decreasing
the value of the characteristic current and increasing the saliency should be clear goals for
improving PM motor performance.
(e) Equation (3) shows that to increase the saliency ratio, , one should increase the
quadrature inductance, Lq. One could decrease the d-axis inductance, Ld, but it would be
counterproductive because of (a) and (c) above, which call for an increase in Ld.
6
(f) If Ld is increased as recommended by (a) and (c) above, then a proportional or larger
increase should be sought for Lq in order to maintain or augment the saliency ratio as
recommended by (d).
Please note that from Eq. (1) these findings can be rephrased in terms of reluctance merely by
replacing the word inductance with the word reluctance together with the inverse of its
preceding verb, increase or decrease. For instance, increase the inductance, Ld, would be
replaced by decrease the reluctance, Rd.
Ways to reduce the PM flux linked by the stator coils include (1) decreasing the strength or
volume of the PMs, (2) increasing the magnetic flux bypassing of the stator coils, and
(3) reducing the number of coils and/or span of the stator coils.
Ways to create saliency include (1) replacing PM material in the surface-mounted PM motor
with silicon steel, as in the case of the inset PM motor; (2) embedding the PMs in the rotor, as in
IPM motors; and (3) placing magnetic flux barriers, as in the multi-layered IPM designs.
We have investigated three PM motors that exhibit varying amounts of reluctance. The inset PM
was assessed for the impact of trading PM for iron; the V-shaped single-layer IPM was assessed
for the impact of embedding the PMs in the iron and of trading PM for iron; and the U-shaped
multilayered IPMs were assessed for the effect of changing the amount of PM material in an
embedded PM design. For each design, the performance vs. amount of PM material was studied
to determine the optimal configuration. Parametric computations of performance were performed
using the SPEED Consortiums brushless direct current (BLDC) computer simulator for each
design. The topologies, methodologies, and results obtained are presented and discussed in the
next section.
Three IPM motor configurations, which exhibit different reluctance properties, have been
parametrically characterized for comparison of torque, power, and amount of magnet material.
Figures 35 show two variations of their structures with two different distributions of PM and
iron.
Figures 3(a) and 3(b) correspond to the inset type motor, whose magnets are flush with the rotor
surface. In the parametric calculations for this motor, the angular pitch, Beta M (M), of the PM
is varied from a low value of 5o to a high value of 175 o in 10o increments. The degrees are with
respect to one complete electrical cycle, which includes a north and a south pole. In Fig. 3(a)
one quadrant represents a 360o electrical cycle, which includes 12 stator teeth. The angular pitch,
M, is the angle in electrical degrees subtended by the magnet. In Fig. 3(a) this includes four
teeth, which leads to 360o 4/12 = 120o. Note that as the pitch of the magnets decreases, the
width of the iron web between the magnets increases. Also note that for the highest pitch, the
expected behavior is that of a surface-mounted PM motor.
7
q M q
M
d d
(a) Low magnet fraction, M = 120o. (b) High magnet fraction, M = 175o.
Fig. 3. Inset surface mounted PM motors.
Figures 4(a) and 4(b) correspond to the V-shaped IPM motor. In the parametric calculations for
this motor, the lower radial position of the PMs is kept constant while the angular pitch of the
PM at the rotors surface is varied from a low value of 50o to a high value of 175o in 10o
increments. Note that as the pitch of the magnets decreases, the width of the iron web between
the magnets increases. In this design we have reluctance variations on account of the iron web, as
in the inset type motor, but in addition we have the soft-PM effect of iron pole-caps that allow
control of the shape of the distribution of flux crossing the air-gap.
q q
M M
d d
(a) Low magnet fraction, M = 50o. (b) High magnet fraction, M = 160o.
Fig. 4. V-shaped IPM motors (type 4).
Figures 5(a) and 5(b) correspond to the dual-layer U-shaped IPM motor. In the parametric
calculations for this motor, the angular pitch of the PM at the rotors surface and the lower radial
position of the PMs are kept constant, while the length of the lower magnets in the two layers of
U-shaped PMs is varied proportionally. The length of the innermost layers magnet, yI_1, is
8
varied from a low value of 2.4 mm to a high value of 17 mm in 0.73-mm increments. Note that
the width of the iron web at the air-gap boundary remains constant, while the amount of
magnetic material increases with the value of yI_1.
q q
yl_1 d yl_1
(a) Lowest magnet fraction, yI_1 = 2.4 mm. (b) Highest magnet fraction, yI_1 = 17 mm.
Fig. 5. U-shaped IPM motors (type 6).
Each motor topology was explored using the BLDC motor simulation software, a product of the
SPEED Consortium. All of the motors have the same stator and overall rotor dimensions. The
same operating conditions, such as max current and voltage, were used for all to provide the
basis for comparative analysis.
The simulation software computed the PM and reluctance torques, power output, current peak
and RMS, back-emf, etc., characterizing the performance of the motors. The analysis excluded
thermal flux and mechanical stress analysis, although these would be of interest for further study.
For each of the three topologies, parametric studies were performed which nested a range of
advance angles within a range of speeds for each magnet configuration. The magnet-to-iron ratio
was varied by means of the PM angular pitch parameters, M or yI_1, as appropriate. Figure 6
shows how the amount of PM material varies with the varying parameter, M or yI_1, in each
motor type. Note that Beta M = M in the figures. There is a large difference between the three
topologies. At each speed there is an optimal phase advance, which determines the flux-
weakening current component, for torque production that is not necessarily the same as for
optimum motor efficiency. The results presented here correspond to the phase advance that
computed the highest torque.
9
5000
4594
4500
4069
4000
3806
3500 3544
3281
3000
Area [mm^2]
1706
1500
131
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
, yI_1
BetaM,
yI_1
The results from the analysis are shown graphically in Figs. 714 for the inset PM motor
(InsRel), Figs. 1522 for the V-shaped IPM (type 6) motor, and Figs. 2331 for the dual-layer U-
shape IPM (type 4) motor. Based upon these results, conclusions and recommendations for
further research are made.
InsRel Motor
Figures 7(a) and 7(b) show the curves of torque and efficiency vs. M for the full range of rotor
speeds. As mentioned earlier, the points in the figure correspond to the phase advance that yields
the peak torque for the speed and under consideration. Figure 7(b) is a magnified version of
Fig. 7(a) to better show the M dependency. The curves show a parabolic-like dependence on the
change in the magnet pitch value. Each parabolic curve describes the dependence on M at the
constant speed to its left in the x-axis. For example, at 100 rpm the torque squares and the
efficiency diamonds represent values of M from 5o to 175o in steps of 10o from left to right.
Note that as speed increases, the peaks move toward higher values of M.
10
700 100
90
600
80
500 70
60
Efficiency [%]
Torque [Nm]
400
Torque
50
Efficiency
300
40
200 30
20
100
10
0 0
10 00
11 00
11 0
12 00
12 00
13 00
13 00
14 0
14 00
15 0
0
16 0
21 0
26 0
31 0
36 0
00
46 0
51 0
00
61 0
00
71 0
00
81 0
86 0
91 0
00
10 00
0
11 0
10
60
60
10
0
0
0
0
0
0
0
0
0
0
10
60
1
6
6
1
6
1
1
41
56
66
76
96
Speed [RPM]
700 100
90
600
80
500
70
60
Efficiency [%]
Torque [Nm]
400
Torque
50
Efficiency
300
40
200 30
20
100
10
0 0
100 600 1100 1600 2100
Speed [RPM]
(b) Expanded view for a reduced set of speeds for the inset PM motor.
Fig. 7. Inset PM motor torque and efficiency as a function of speed for a range of magnet pitches.
Figures 8(a), 8(b), and 8(c) display in classical formats the same information relationships
between M, speed, and torque as Fig. 7. Figure 8(c) in particular facilitates the determination of
the value of M required to achieve a level of torque through a given speed range. For instance,
for a torque output above 50 N-m over the entire speed range, the minimum amount of PM
material needed corresponds to M = 95.
11
Speed Variation Upon Torque
700
600
BetaM_5
BM_5
BM_15
BetaM_15
500 BM_25
BetaM_25
BetaM_35
BM_35
BM_45
BetaM_45
BM_55
BetaM_55
400
BM_65
BetaM_65
Torque [Nm]
BM_75
BetaM_75
BM_85
BetaM_85
300
BetaM_95
BM_95
BetaM_105
BM_105
BetaM_115
BM_115
200
BetaM_125
BM_125
BetaM_135
BM_135
BetaM_145
BM_145
100 BM_155
BetaM_155
BM_165
BetaM_165
BM_175
BetaM_175
0
100
600
1100
1600
2100
2600
3100
3600
4100
4600
5100
5600
6100
6600
7100
7600
8100
8600
9100
9600
10100
10600
11100
11600
12100
12600
13100
13600
14100
14600
15100
-100
Speed [RPM]
700
650
600
650-700
550 600-650
500 550-600
450 500-550
400 Torque [Nm] 450-500
400-450
350
350-400
300
300-350
250
250-300
200 200-250
150 150-200
100 100-150
50 50-100
0 0-50
-50 -50-0
BetaM_165
_165M
100
1100
BetaM_125
M_125
2100
3100
4100
5100
BetaM_85
M_85
6100
7100
8100
M [deg] M_45
10100
11100
12100
BetaM_5
13100
Speed [RPM]
M_5
14100
15100
Fig. 8. Relationship between torque, magnet pitch, M, and speed for the inset PM motor.
12
Torque vs Speed vs BetaM
BetaM [deg]
BM [deg]
BM_175
BetaM_175
BM_165
BetaM_165 Torque [Nm]
BM_155
BetaM_155 600-650
BM_145
BetaM_145 550-600
BM_135
BetaM_135
500-550
BM_125
BetaM_125
450-500
BM_115
BetaM_115
400-450
BM_105
BetaM_105
350-400
BM_95
BetaM_95
BM_85
BetaM_85 300-350
BM_75
BetaM_75 250-300
BM_65
BetaM_65 200-250
BM_55
BetaM_55 150-200
BM_45
BetaM_45 100-150
BM_35
BetaM_35 50-100
BM_25
BetaM_25 0-50
BM_15
BetaM_15 -50-0
BM_5
BetaM_5
15100
14100
13100
12100
11100
10100
9100
8100
7100
6100
5100
4100
3100
2100
1100
100
Speed [RPM]
Fig. 8. Relationship between torque, magnet pitch, M, and speed for the inset PM motor (contd).
Similarly, the complete relationship between M and speed upon advance and upon power is
shown below in Figs. 9(a), 9(b), 10(a), and 10(b), respectively.
The percentage of torque produced by reluctance is shown in Figs. 11(a) and 11(b).
The constant power speed ratio (CPSR) may be found using Figs. 8 and 10. From Fig. 8, the base
speed is estimated as the speed when torque delivery begins to drop, which appears to be about
1100 rpm. From Fig. 10, the intersection of a power plane, whose value is the rated power with
the surface, may be used to determine the maximum speed for which the motor is capable of
delivering rated power by observing where the power falls below that plane. The CPSR is the
ratio of the maximum useful speed to the base speed and is determined for each value of the
magnet pitch, M. For a threshold value of 40 kW, the CPSR as a function of M is shown in
Fig. 12. In it, one can see that a CPSR = 10 can be obtained when M = 125o.
13
Advance vs Speed
120
100
80
Advance [deg]
100-120
80-100
60 60-80
40-60
20-40
40
0-20
20
100
1100
BetaM_165
M_165
2100
3100
4100
BetaM_125
5100
M_125
6100
7100
BetaM_85
M_85 8100
9100
10100
BetaM_45
M_45
11100
BetaM [deg]
M [deg]
Speed [RPM]
12100
13100
BetaM_5
M_5
14100
15100
Advance vs Speed
120
100
80 100-120
Advance [deg]
80-100
60-80
60 40-60
20-40
0-20
40
100
3100
6100 20
Speed [RPM] 9100
12100 0
BM_5
BetaM_5
BM_25
BetaM_25
BM_45
BetaM_45
BM_65
15100
BetaM_65
BetaM_85
BM_85
BetaM_105
BM_105
BM_125
BetaM_125
BM_145
BetaM_145
BM_165
etaM_165
BetaM [deg]
(b) Maximum magnet corner view.
Fig. 9. Relationship between magnet pitch, M, speed, and advance for the inset PM motor.
14
Power
Powervs
vs.Speed
Speedvsvs.
BetaM
BM
80
70
60
Power [kW]
50
40 70-80
60-70
30 50-60
40-50
20
30-40
10 20-30
10-20
0
0-10
15100
13600
BBetaM_165
M_165
12100
10600
M_125
BBetaM_125
9100
M_85
BBetaM_85
7600
6100
M_45
BBetaM_45
4600
Speed [RPM]
3100
BMBetaM
[ deg][deg]
B M_5
BetaM_5
1600
100
(a) Minimum magnet corner view.
Power
Power vs. Speed
vs Speed vs. BM
vs BetaM
80
70
60 70-80
60-70
50
Power [kW]
50-60
100 40-50
2100 40
30-40
4100 20-30
30
6100 10-20
Speed [RPM] 8100 20 0-10
10100
10
12100
0
14100
BM_5
BetaM_5
BM_25
BetaM_25
BM_45
BetaM_45
BM_65
BetaM_65
BM_85
BetaM_85
BM_105
BetaM_105
BM_125
BetaM_125
BM_145
BetaM_145
BM_165
BetaM_165
BM_ [deg]
BetaM [deg]
Fig. 10. Relationship between M, speed, and maximum power for the inset PM motor.
15
%Reluctance Torque vs Speed vs BetaM
100.00%
0.9-1
90.00%
0.8-0.9
80.00% 0.7-0.8
70.00% 0.6-0.7
%Reluctance [%] 0.5-0.6
60.00%
0.4-0.5
50.00%
0.3-0.4
40.00%
0.2-0.3
30.00%
100 0.1-0.2
20.00% 0-0.1
2600
10.00% -0.1-0
5100
0.00% -0.2--0.1
-10.00% 7600
-20.00% 10100 Speed [RPM]
BM_165
BetaM_165
BM_145
BetaM_145
BM_125
12600
BetaM_125
BM_105
BetaM_105
BM85
BetaM_85
BM65
BetaM_65
15100
BM45
BetaM_45
BM25
BetaM_25
BM5
BetaM_5
BM [deg]
BetaM [deg]
%Reluctance Torque
%Reluctance Torque vsvs. Speed
Speed vs. BM
vs BetaM
0.9-1
100.00% 0.8-0.9
90.00% 0.7-0.8
0.6-0.7
80.00%
0.5-0.6
70.00%
0.4-0.5
%Reluctance [%]
60.00%
0.3-0.4
100 50.00% 0.2-0.3
2100 40.00% 0.1-0.2
4100 30.00% 0-0.1
6100 20.00% -0.1-0
10.00% -0.2--0.1
8100
Speed [RPM] 0.00%
10100 -10.00%
12100 -20.00%
BM5
BetaM_5
BM25
BetaM_25
BM45
BetaM_45
14100
BM65
BetaM_65
BM85
BetaM_85
BM_105
BetaM_105
BM_125
BetaM_125
BM_145
BetaM_145
BM_165
BetaM_165
BM [deg]
BetaM [deg]
16
CPSR at 40kW
14
12
10
InsRel_CPSR_40kW
CPSR
0
5 15 25 35 45 55 65 75 85 95 105 115 125 135 145 155 165 175
-2
BM
BetaM
Fig. 12. CPSR as a function of magnet pitch, M, for the inset PM motor type.
Assuming the magnet cost is proportional to the amount of magnet, and then the relative cost
which may be used for cost comparisons, is the magnet area divided by the maximum magnet
area. Using Fig. 6 and the U-shaped type 6 motor as an example, maximum magnet area is
2892 mm2, which occurs for yI_1 = 17mm and would have a normalized cost of 1. The minimum
magnet area in this example is 2366 mm2, which occurs at yI_1 = 2.4mm and has a normalized
cost of 0.82. Figures 13(a) and 13(b) show graphically the cost for producing torque for each M
for the inset PM motor type over the entire speed range. Figure 13(b) shows that, in order to
produce torque over the full speed range, the normalized cost incurred is between 0.59 and 1.0,
which corresponds to values of M between 95 and 175.
17
Normalized Torque and Cost
0.9
0.9-1
0.8
0.8-0.9
0.7 0.7-0.8
0.6-0.7
0.6 0.5-0.6
0.4-0.5
Normalized Torque 0.5
0.3-0.4
0.4 0.2-0.3
0.1-0.2
0.3 100 0-0.1
0.2 2100
4600
0.1
6600
0 8600
Speed [RPM]
1.00
0.94
10600
0.89
0.83
0.77
0.71
0.66
0.60
12600
0.54
0.49
0.43
0.37
0.31
0.26
15100
0.20
Normalized Cost
0.14
0.09
0.03
Normalized Cost
0.03
0.09
0.14
0.20 Normalized Torque
0.26 0.9-1
0.31 0.8-0.9
0.37
0.7-0.8
0.43
0.6-0.7
0.49
0.5-0.6
0.54
0.4-0.5
0.60
0.3-0.4
0.66
0.2-0.3
0.71
0.1-0.2
0.77
0-0.1
0.83
0.89
0.94
1.00
15100
14600
14100
13100
12600
12100
11600
11100
10600
10100
9600
9100
8600
8100
7600
7100
6600
6100
5600
5100
4600
4100
3600
3100
2100
1600
1100
600
100
Speed [RPM]
Fig. 13. Relationship between normalized torque, speed, and relative magnet cost
for the inset PM motor.
18
Figure 14(a) and 14(b) show a comparison of the torque and power vs. speed curves for the M
values considered to be the best for the inset PM with those for an equivalent surface-mounted
PM that has no reluctance torque and is represented here by an inset PM with M = 180o.
Zero Reluctance InsRel Motor
700
600
500
Torque [Nm]
400 BetaM_180
BM_180
BM_105
BetaM_105
BetaM_115
BM_115
300 BM_125
BetaM_125
200
100
0
100
600
1100
1600
2100
2600
3100
3600
4100
4600
5100
5600
6100
6600
7100
7600
8100
8600
9100
9600
10100
10600
11100
11600
12100
12600
13100
13600
14100
14600
15100
Speed [RPM]
70
60
50
40
Power [kW]
BetaM_180
BM_180
BM_105
BetaM_105
BM_115
BetaM_115
30 BM_125
BetaM_125
20
10
0
100
600
1100
1600
2100
2600
3100
3600
4100
4600
5100
5600
6100
6600
7100
7600
8100
8600
9100
9600
10100
10600
11100
11600
12100
12600
13100
13600
14100
14600
15100
Speed [RPM]
Fig. 14. Comparison of selected optimal (magnet pitches, M = 105, 115, and 125o with highest CPSR) inset
PM motor performance with surface-mounted PM motor (M = 180o).
The surface-mounted PM not only has lower peak torques at low speeds, but also has a smaller
speed range.
19
V-Shaped IPM Type-4 Results
Figure 15(a) and 15(b) show the curves of efficiency and torque for the IPM type-4 motor
topology. The curves have a V or inverted parabola shape, again due to the variation in M.
Note the drastic drop in efficiency in the 100-rpm range. This should be considered a limiting
factor in determining the effective operating speed range. As shown previously for the inset PM
motor, the torque vs. speed for each M is plotted in Fig. 16.
BetaM Variation
B Variation UponTorque
Upon Torque and
and Efficiency
Efficiency
M
800 100
700 90
600 80
500 70
Torque [Nm]
Torque
400 60
Efficiency
300 50
200 40
100 30
0 20
10 00
11 00
11 00
12 0
12 00
13 00
13 00
14 00
14 0
15 00
0
16 0
00
26 0
31 0
36 0
41 0
00
51 0
56 0
61 0
00
71 0
76 0
81 0
00
91 0
96 0
10 00
0
11 0
60
10
10
0
0
0
0
0
0
0
0
0
0
0
0
0
10
60
1
6
1
1
6
1
6
6
21
46
66
86
Speed [RPM]
Efficiency [%]
800 100
90
700
80
600
70
500
60
Toruqe [Nm]
Torque
400 50
Efficiency
40
300
30
200
20
100
10
0 0
100 600 1100 1600
Speed [RPM]
Fig. 15. V-shaped IPM motor torque and efficiency as a function of speed for a range of magnet pitches.
20
Speed
Speed VariationUpon
Variation UponTorque
Torquefor
forEach
EachBetaM
BM
800
BM_50
BM_50
BetaM_50
BetaM_56
BM_56
BM_56
700 BetaM_62
BM_62
BM_62
BetaM_68
BM_68
BM_68
BetaM_74
BM_74
BM_74
600 BM_80
BetaM_80
BM_80
BM_86
BetaM_86
BM_86
BM_92
BetaM_92
BM_92
500 BM_98
BM_98
BetaM_98
Torque [Nm]
BM_104
BetaM_104
BM_104
BM_110
BM_110
BetaM_110
400
BM_116
BM_116
BetaM_116
BM_122
BM_122
BetaM_122
300
BM_128
BM_128
BetaM_128
BM_134
BetaM_134
BM_134
BM_140
BM_140
BetaM_140
200 BM_146
BM_146
BetaM_146
BM_152
BM_152
BetaM_152
BM_158
BM_158
BetaM_158
100 BM_164
BetaM_164
BM_164
BM_170
BM_170
BetaM_170
10 0
11 0
11 0
12 0
12 0
13 0
13 0
14 0
14 0
15 0
0
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
00
0
0
10
60
10
60
10
60
10
60
10
60
10
10
60
11
16
21
26
31
36
41
46
51
56
61
66
71
76
81
86
91
96
10 Speed [RPM]
800
700
600
500
700-800
600-700
Torque [Nm] 400
500-600
400-500
300
300-400
200-300
200 100-200
0-100
100
BBetaM_170
M_170
0
BBetaM_140
M_140
100
1600
3100
BetaM_110
BM_110
4600
6100
7600
BM_80
BetaM_80
BBetaM [deg]
M [deg]
9100
10600
12100
15100
Fig. 16. Relationship between magnet pitch, M, speed, and torque for the V-shaped IPM motor.
21
Torque vs.
Torque Speed vs.
vs Speed BM
vs BetaM
800
700
600
700-800
500 600-700
500-600
Torque [Nm] 400 400-500
300-400
300 200-300
100
1600 100-200
200 3100
4600 0-100
6100
100 7600
9100 Speed [RPM]
0 10600
BM_50
12100
BetaM_50
BM_62
BetaM_62
BM_74
BetaM_74
BM_86
BetaM_86
13600
BM_98
BetaM_98
BM_110
BetaM_110
BM_122
BetaM_122
15100
BM_134
BetaM_134
BM_146
BetaM_146
BM_158
BetaM_158
BM_170
BetaM_170
BM [deg]
BetaM [deg]
Fig. 16. Relationship between magnet pitch, M, speed, and torque for the V-shaped IPM motor (contd).
Figures 16(a), 16(b), and 16(c) show the same informational relationship between M, speed, and
torque as Fig. 15 in classical formats. Figure 16(c), in particular, facilitates the determination of
the amount of magnet required to achieve a level of torque through a given speed range. For use
of the complete speed range with a torque output of above 500 N-m, M must be at least 116.
Similarly, the complete relationship between M and speed upon advance and power is shown
below in Figs. 17(a), 17(b), 18(a), and 18(b), respectively.
22
Advance
Advance vsvs. Speed
Speed vs. BM
vs BetaM
100
95
95-100
90
90-95
85
85-90
80 80-85
75 75-80
70-75
70
Advance [deg] 65-70
65
60-65
60 55-60
55 15100 50-55
50 13600 45-50
12100
45 10600 40-45
9100 35-40
40 7600
6100
35 Speed [RPM]
4600
BM_170
BetaM_170
BM_158
BetaM_158
3100
BM_146
BetaM_146
BM_134
BetaM_134
BM_122
1600
BetaM_122
BM_110
BetaM_110
BM_98
BetaM_98
100
BetaM_86
BM_86
BM_74
BetaM_74
BM_62
BetaM_62
Advance_50
BM_50
Advance_50
BM [deg]
BetaM [deg]
Advancevs
Advance vs.Speed
Speedvsvs. BM
BetaM
100
95 95-100
90 90-95
85 85-90
80-85
80
75-80
75
70-75
70 65-70
Advance [deg]
65 60-65
60 55-60
15100
13600 55 50-55
12100 45-50
10600 50
40-45
9100 45
7600 35-40
6100 40
Speed [RPM]
4600 35
3100
Advance_50
Advance_50
BM_62
BetaM_62
BM_74
BetaM_74
BM_86
1600
BetaM_86
BM_98
BetaM_98
BM_110
BetaM_110
BM_122
BetaM_122
100
BM_134
BetaM_134
BM_146
BetaM_146
BM_158
BetaM_158
BM_170
BetaM_170
B M [deg]
BetaM [deg]
Fig. 17. Relationship between magnet pitch, M, speed, and advance for the V-shaped IPM motor.
23
Power vs.Speed
Power vs Speedvsvs. BM
BetaM
70
65
60
65-70
55 60-65
55-60
50 50-55
45-50
45 Power [kW] 40-45
35-40
40
30-35
35 25-30
20-25
30
BM_170
BetaM_170
25
BetaM_140
BM_140
BetaM_110
BM_110 20
BM[deg]
BetaM [deg]
15100
13600
BetaM_80
BM_80
12100
10600
9100
7600
6100
BetaM_50
BM_50
4600
3100
1600
100
Speed [RPM]
BM [deg]
BetaM [deg]
BetaM_170
BM_170
BetaM_164
BM_164
BetaM_158
BM_158
BetaM_152
BM_152
BetaM_146
BM_146
BetaM_140
BM_140 Power [kW]
BetaM_134
BM_134 65-70
BetaM_128
BM_128 60-65
BetaM_122
BM_122 55-60
BetaM_116
BM_116 50-55
BetaM_110
BM_110 45-50
BetaM_104
BM_104 40-45
BetaM_98
BM_98 35-40
BetaM_92
BM_92 30-35
BetaM_86
BM_86 25-30
BetaM_80
BM_80 20-25
BetaM_74
BM_74
BetaM_68
BM_68
BetaM_62
BM_62
BetaM_56
BM_56
BetaM_50
BM_50
100 1600 3100 4600 6100 7600 9100 10600 12100 13600 15100
Speed [RPM]
Fig. 18. Relationship between magnet pitch, M, speed, and maximum power for the V-shaped IPM motor.
The contour plot in Fig. 18(b) clearly shows the general trend of higher power (dark blue)
extending into higher speed ranges over the midrange of the M variation. Figure 19 shows the
percentage of torque produced by reluctance.
24
% Reluctance
% Reluctance Torque
Torque vs vs. Speed
Speed vs vs. BM
BetaM
110.00%
105.00%
100.00% 1.05-1.1
1-1.05
95.00%
0.95-1
90.00% 0.9-0.95
85.00% 0.85-0.9
15100 80.00% %Reluctance [%] 0.8-0.85
13600 0.75-0.8
75.00%
12100 0.7-0.75
10600 70.00%
9100 0.65-0.7
65.00% 0.6-0.65
7600
Speed [RPM] 6100 60.00% 0.55-0.6
4600 55.00% 0.5-0.55
3100
50.00%
1600
BM_50
BetaM_50
BetaM_62
BM_62
BM_74
BetaM_74
BM_86
BetaM_86
BM_98
BetaM_98
100
BM_110
BetaM_110
BM_122
BetaM_122
BetaM_134
BM_134
BetaM_146
BM_146
BetaM_158
BM_158
BetaM_170
BM_170
BM [deg]
BetaM [deg]
Figure 20 shows the CPSR for 40 kW. The CPSR peaks at a value of 12 for a value of M
between 110 and 122; a value of M = 77 would be enough for a CPSR of 10.
CPSR at 40kW
14
12
10
8
CPSR
Interp_40kW
40kW - 2nd Order
6
0
50
56
62
68
74
80
86
92
98
4
0
6
2
8
4
0
6
2
8
4
0
10
11
11
12
12
13
14
14
15
15
16
17
BM
BetaM
Fig. 20. CPSR as a function of magnet pitch, M, for the V-shaped IPM motor.
As with the InsRel motor, the relationship between normalized cost, torque, and speed is shown
in Figs. 21(a) and 21(b).
25
Surface Plot of Normalized Cost
0.95-1
1 0.9-0.95
0.95
0.85-0.9
0.9
0.85 0.8-0.85
0.8 0.75-0.8
0.75
0.7 0.7-0.75
0.65 0.65-0.7
0.6
0.55 0.6-0.65
0.5 Normalized Torque 0.55-0.6
0.45
0.4 0.5-0.55
0.35 0.45-0.5
0.3
0.25 0.4-0.45
0.2
0.15 0.35-0.4
0.1 0.3-0.35
0.05
0 0.25-0.3
0.2-0.25
15100
13600
1.00 0.15-0.2
12100
0.1-0.15
10600
0.99
9100
0.05-0.1
0.90
7600
0-0.05
6100
0.76
Speed [RPM]
4600
0.63
3100
Normalized Cost
0.50
1600
100
Normalized Torque
Normalized Cost
0.95-1
1.00
1.00 0.9-0.95
0.99 0.85-0.9
0.99 0.8-0.85
0.99 0.75-0.8
0.98
0.7-0.75
0.98
0.94 0.65-0.7
0.90 0.6-0.65
0.87 0.55-0.6
0.83 0.5-0.55
0.79
0.45-0.5
0.76
0.72 0.4-0.45
0.69 0.35-0.4
0.66 0.3-0.35
0.63 0.25-0.3
0.60
0.2-0.25
0.56
0.53 0.15-0.2
0.50 0.1-0.15
100
600
1100
1600
2100
2600
3100
3600
4100
4600
5100
5600
6100
6600
7100
7600
8100
8600
9100
9600
10100
10600
11100
11600
12100
12600
13100
13600
14100
14600
15100
0.05-0.1
0-0.05
Speed [RPM]
Fig. 21. Relationship between normalized torque, speed, and relative magnet cost for V-shaped IPM motor.
Figure 21(b) shows that the cost can vary from 100% to 50% and that the highest torque region
(red, orange, purple, and blue) over the entire speed range appears to be centered at about 83%
cost.
26
Finally, Figs. 22(a) and 22(b) compare the torque and power vs. speed of the type-4 and the
surface-mounted PM. As with the inset PM, the surface-mounted PM has an initial torque
roughly half that of the motor with V-shaped magnets. It marginally increased above the motor
with V-shaped magnets before falling below at speeds above 7600 rpm. This illustration shows
the positive impact of reluctance torque generation over the complete operating range.
Torque vs Speed Comparison
500
450
400
350
300
Torque [Nm]
BetaM_104
BM_104
BetaM_110
BM_110
250
BetaM_116
BM_116
BetaM_180
BM_180
200
150
100
50
0
100
600
1100
1600
2100
2600
3100
3600
4100
4600
5100
5600
6100
6600
7100
7600
8100
8600
9100
9600
10100
10600
11100
11600
12100
12600
13100
13600
14100
14600
15100
Speed [RPM]
70
60
50
40
Power [kW]
BM_104
BetaM_104
BM_110
BetaM_110
BM_116
BetaM_116
30 BetaM_180
BM_180
20
10
0
100
600
1100
1600
2100
2600
3100
3600
4100
4600
5100
5600
6100
6600
7100
7600
8100
8600
9100
9600
10100
10600
11100
11600
12100
12600
13100
13600
14100
14600
15100
Speed [RPM]
27
U-Shaped Dual-Layer IPM Type-6 Results
In this geometry, the parameter determining the amount of magnet material, M, is replaced with
the variable, yI_1, which is the length of the magnet in the bottom layer of the innermost part of
the U. The width of the silicon steel web at the air-gap boundary is kept constant. Thus, the
only reluctance changes produced by the change of yI_1 are associated with the positioning of
internal flux barriers. The main effect of yI_1 variation is the amount of PM material.
Figures 23(a) and 23(b) show that in the U-shaped dual-layer IPM type-6 motor, the variation in
PM positioning and amount of magnet had little impact in the torque produced, as indicated by
the horizontal segments in Fig. 23(a) and, especially, Fig. 23(b). Note the sharp roll-off of
efficiency as speed increases. This efficiency roll-off may determine the optimal length of the
innermost PM in this IPM type.
BetaM
B Variation
M Variation Upon
Upon Torque
Torque andand Efficiency
Efficienciy
600 100
90
500
80
70
400
60
Efficinecy [%]
Torque [Nm]
Torque
300 50
Efficiency
40
200
30
20
100
10
0 0
10 00
11 00
11 00
12 00
12 00
13 0
13 00
14 00
14 00
15 00
0
16 0
21 0
26 0
31 0
00
41 0
46 0
51 0
56 0
61 0
66 0
00
76 0
81 0
86 0
00
96 0
10 00
0
11 0
0
0
0
0
0
0
0
0
0
0
0
0
0
60
10
10
60
1
6
1
6
1
1
6
1
6
36
71
91
Speed [RPM]
Fig. 23. Torque and efficiency as a function of speed over a range of magnet pitches for
the motor with dual-layer U-shaped IPMs.
28
Zoom of Torque and Efficiency vs Speed
Magnification of Torque and Efficiency vs Speed
600 100
90
500
80
70
400
60
Efficiency [%]
Torque [Nm]
Torque
300 50
Efficiency
40
200
30
20
100
10
0 0
100 600 1100 1600
Speed [RPM]
Fig. 23. Torque and efficiency as a function of speed over a range of magnet pitches for
the motor with dual-layer U-shaped IPMs (contd).
Figures 24(a) and 24(b) show the relationship between yI_1, speed, and torque for this dual-layer
U-shaped PM motor. The tightly packed set of torque vs. speed curves confirms that the torque is
independent of changes in length of the U-shaped lower magnets.
Figures 25(a) and 25(b) are 3D views of the relationships between yI_1, speed, and phase
advance for the dual-layer U-shaped PM motor type. Again, the values are nearly independent of
variations in yI_1.
Figures 26(a) and 26(b) are 3D views of the relationships between yI_1, speed, and power output
for the dual-layer U-shaped PM motor. Again, the values are nearly independent of variations in
yI_1.
29
Speed Variation Upon Torque for Each yI_1
600
yI_1=2.4
yI_1=3.13
500 yI_1=3.86
yI_1=4.59
yI_1=5.32
yI_1=6.05
400 yI_1=6.78
yI_1=7.51
yI_1=8.24
Torque [Nm]
yI_1=8.97
300 yI_1=9.7
yI_1=10.43
yI_1=11.16
yI_1=11.89
200 yI_1=12.62
yI_1=13.35
yI_1=14.08
yI_1=14.81
100
yI_1=15.54
yI_1=16.27
yI_1=17
0
100
600
1100
1600
2100
2600
3100
3600
4100
4600
5100
5600
6100
6600
7100
7600
8100
8600
9100
9600
10100
10600
11100
11600
12100
12600
13100
13600
14100
14600
15100
Speed [RPM]
600
500
400
500-600
Torque [Nm]
400-500
300 300-400
200-300
100-200
0-100
200
100
yI_1=17
yI_1=13.35
0 yI_1=9.7
100
1100
2100
3100
4100
yI_1=6.05
5100
6100
7100
8100
9100
10100
11100
yI_1=2.4
12100
13100
14100
15100
Speed [RPM]
Fig. 24. Relationship between magnet amount, yI_1, speed, and torque
for the motor having dual-layer U-shaped IPMs.
30
Advance vs Speed vs yI_1
120
100
80
Advance [deg]
100-120
80-100
60 60-80
40-60
20-40
40 0-20
20
yI_1=17
yI_1=14.08
0
yI_1=11.16
15100
14100
13100
yI_1=8.24
12100
11100
10100
9100
8100
yI_1=5.32 7100
6100
5100
4100
3100
yI_1=2.4
2100
1100
Speed [RPM]
100
yI_1=17
yI_1=16.27
yI_1=15.54
yI_1=14.81
yI_1=14.08
yI_1=13.35
yI_1=12.62 Advance [deg]
yI_1=11.89
100-120
yI_1=11.16
80-100
yI_1=10.43
60-80
yI_1=9.7
40-60
yI_1=8.97
20-40
yI_1=8.24 0-20
yI_1=7.51
yI_1=6.78
yI_1=6.05
yI_1=5.32
yI_1=4.59
yI_1=3.86
yI_1=3.13
yI_1=2.4
100
1100
2100
3100
4100
5100
6100
7100
8100
9100
10100
11100
12100
13100
14100
15100
Speed [RPM]
Fig. 25. Relationship between magnet material, yI_1, speed, and advance
for the motor having dual-layer U-shaped IPMs.
31
Power vs Speed vs yI_1
60
50
40
Power [kW]
50-60
40-50
30
30-40
20-30
10-20
20
0-10
10
yI_1=17
yI_1=14.08
yI_1=11.16 0
15100
yI_1=8.24
14100
13100
12100
11100
10100
9100
yI_1=5.32
8100
7100
6100
5100
4100
yI_1=2.4
3100
2100
1100
100
Speed [RPM]
60
50
40
Power [kW]
30 50-60
40-50
20 30-40
20-30
10 10-20
0-10
0
100
yI_1=2.4
1600
3100
yI_1=5.32
4600
yI_1=8.24
6100
7600
yI_1=11.16
9100
10600
yI_1=17
13600
15100
Fig. 26. Relationship between magnet amount, yI_1, speed, and power
for the motor having dual-layer U-shaped IPMs.
Figure 27 shows the extreme contribution of reluctance at high speeds. Note that at these same
high speeds this motor exhibits low efficiency, as shown earlier in Fig. 23.
32
%Reluctance vs Speed vs yI_1
8000%
7000%
6000%
%Reluctance [%]
5000%
70-80
4000% 60-70
50-60
3000% 40-50
30-40
2000%
20-30
1000% 10-20
0-10
yI_1=17 0%
15100
yI_1=14.08
13600
12100
yI_1=11.16
10600
9100
yI_1=8.24
7600
6100
yI_1=5.32
4600
Speed [RPM]
3100
yI_1=2.4
1600
100
Fig. 27. Relationship between magnet amount, yI_1, speed, and percentage of
reluctance torque for motor having dual-layer U-shaped IPMs.
Figure 28 shows that the CPSR for the motor topology is unaffected by changes in yI_1;
consequently, the minimum value of yI_1 is best.
CPSR at 40kW
3.8
3.6
3.4
3.2
CPSR
CPSR_40
3
40kW - 2nd Order
2.8
2.6
2.4
2.2
2
2.4 3.13 3.86 4.59 5.32 6.05 6.78 7.51 8.24 8.97 9.7 10.4 11.2 11.9 12.6 13.4 14.1 14.8 15.5 16.3 17
yI_1 [mm]
Fig. 28. CPSR as a function of magnet amount, yI_1, for the motor having dual-layer U-shaped IPMs.
33
Figures 29(a) and 29(b) show the normalized costs of the magnet and torque generated at each
speed.
Normalized Cost vs Speed
1.00
0.95
0.90 0.95-1.00
0.85 0.90-0.95
0.80 0.85-0.90
0.75 0.80-0.85
0.70 0.75-0.80
Normalized Torque
0.65 0.70-0.75
0.60 0.65-0.70
0.55 0.60-0.65
0.50
0.55-0.60
0.45
0.40 0.50-0.55
0.35 0.45-0.50
0.30 0.40-0.45
0.25 0.35-0.40
0.20 0.30-0.35
0.15 1.00
0.25-0.30
0.10 0.96
0.05 0.20-0.25
0.00 0.93 0.15-0.20
100
0.05-0.10
5100
6100
0.85
7100
8100
0.00-0.05
9100
10100
11100
12100
0.82
13100
14100
15100
Speed [RPM]
Normalized Cost
1.00
0.99 Normalized Torque
0.98 0.95-1.00
0.97 0.90-0.95
0.96 0.85-0.90
0.95 0.80-0.85
0.95 0.75-0.80
0.94 0.70-0.75
0.93
0.65-0.70
0.92
0.60-0.65
0.91
0.55-0.60
0.90
0.50-0.55
0.89
0.88 0.45-0.50
0.87 0.40-0.45
0.86 0.35-0.40
0.85 0.30-0.35
0.85 0.25-0.30
0.84 0.20-0.25
0.83 0.15-0.20
0.82 0.10-0.15
15100
14100
13100
12100
11100
10100
9100
8100
7100
6100
5100
4100
3100
2100
1100
100
0.05-0.10
0.00-0.05
Speed [RPM]
Fig. 29. Speed and normalized cost and torque for the motor having dual-layer U-shaped IPMs.
34
Finally, Figs. 30(a) and 30(b) show the toque and power vs. speed curves for this topology and
for the surface-mounted IPM, which is the inset PM motor with M = 180. The type-6 motor
has significantly higher torque up to 1600 rpm, but its speed range is lower. As shown by
Fig. 30(b), the performance computed for this motor type is quite inferior to that of the surface-
mounted PM motor.
Torque vs Speed Comparison
600
500
400
Toruqe [Nm]
yI_1=2.4
yI_1=3.13
300
yI_1=3.86
BetaM_180
200
100
0
100
600
1100
1600
2100
2600
3100
3600
4100
4600
5100
5600
6100
6600
7100
7600
8100
8600
9100
9600
10100
10600
11100
11600
12100
12600
13100
13600
14100
14600
15100
Speed [RPM]
70000
60000
50000
40000
Power [kW]
yI_1=2.4
yI_1=3.13
yI_1=3.86
30000 BetaM_180
20000
10000
0
00
00
00
00
00
00
0
00
00
00
00
00
00
00
00
00
10
1
11
21
31
41
51
61
71
81
91
10
11
12
13
14
15
Speed [RPM]
Fig. 30. Performance comparison of IPM motor having double-layer U-shaped magnets
with surface-mounted PM motor having only PM torque.
35
Comparison of Best of Each Topology
For reference and comparison, the torque and power plots for what we have chosen as the best in
each of the three motor types have been plotted in Figs. 30(a) and 30(b). For completeness, the
curves for the surface-mounted PM have also been included.
Torque Comparision for All Four Motor Types
700
600
500
Torque [Nm]
200
100
0
0
0
00
00
00
00
00
00
00
00
00
0
10
10
10
10
10
10
10
11
21
31
41
51
61
71
81
91
10
11
12
13
14
15
Speed [RPM]
70
60
50
40
Power [kW]
Surface Mount
Inset PM
V-Shape Type-4
30 U-Shape Type-6
20
10
0
0
0
00
00
00
00
00
00
00
00
00
10
10
10
10
10
10
10
11
21
31
41
51
61
71
81
91
10
11
12
13
14
15
Speed [RPM]
Fig. 31. Performance comparison of four motors with varying degrees of reluctance: (1) surface-mounted
PM motor producing only PM torque; (2) inset PM/reluctance motor with magnet pitch, M = 125;
(3) V-shaped IPM motor with M = 116; and (4) double-layer U-shaped IPM with yI_1 = 22.4 mm.
36
Figure 31(a) shows that the torque at low speeds is significantly higher for all the IPM motors
than for the surface-mounted PM. The slopes of the curves in Fig. 31(b) show the impact of
reluctance on lowering the rate of power drop-off with speed.
The best of the motors with inset and V-shape magnets have very similar performances. The U-
shape type-6 motor has lower performance and shorter speed range, but comparison to the others
may not be warranted because its base geometry is not necessarily optimal. The intended goal
was to evaluate the impact of changing the amount of PM while keeping the geometry at the air-
gap surface constant. There is a need to continue studies in investigating the effect of changing
the amount of PM closer to the periphery.
CONCLUSIONS
Through our preliminary analysis, we have shown the benefits of properly controlling reluctance
and the need for saliency to further increase reluctance torque.
ORNL has investigated, by means of parametric studies, the performance of three PM designs
which represent progressive approaches of using reluctance to improve the performance of PM
motors for locomotion. The results obtained show the need for detailed finite-element
simulations and indicate the direction of further research.
As expected, due to the reluctance torque component, all three types of IPM motors studied have
significantly higher torque at low speeds than a similar surface-mounted PM motor (Fig. 31(a))
The rate of power drop as speed increases is lower in all the IPMs (Fig. 31(b)), as expected
because they have weaker magnet strength, which reduces the value of max (Caption Fig. 1).
The inset PM introduces reluctance torque, causing saliency in the rotor by replacing some of the
PM material in its surface with iron (Fig. 3). This study concludes that the optimal amount of PM
material for a CPSR of 10 at 40 kW corresponds to an angular pitch of M = 125 (Fig. 12). This
is 70% of the amount of PM material in a similar surface-mounted PM motor, which would
result in lower cost for the motor in addition to improved performance.
The embedded V-shaped PM motor immerses the PMs into the rotors iron (Fig. 4), creating
soft-iron magnetic poles in addition to introducing saliency. It is concluded that the angular
magnet pitch for optimal torque and speed range is 105 (Fig. 18(b)). Then the motor can operate
over the whole speed range at powers above 50 kW. Note that for a CPSR = 10 at 40 kW, the
motor would only require a magnet pitch of M = 77 (Fig. 20 using the 2nd order line fit), which
would be more economical.
The dual-layer U-shape PM motor (Fig. 5), in addition to creating soft-iron magnetic poles and
saliency in the rotor, introduces multiple barriers to increase anisotropy in the magnetic path,
thereby enhancing saliency. ORNL has not yet determined the operational benefits expected
from the added design complexity. The intended goal was to evaluate the impact of changing the
amount of PM in the inside while keeping the geometry at the air-gap surface constant. The
conclusion is clear: there is no impact. As Fig. 24(a) shows, performance was essentially
unaffected by changes in the length of the lower magnets. As a result, the obvious optimal
37
geometry was the one with the shortest magnet possible. Comparison of power and torque
outputs to those of the other optimized motor types is not warranted because the base geometry is
not necessarily optimal. It is also possible that the computer model does not simulate this multi-
layer type of motor properly. Further studies will investigae this topology with more detailed
variations of the width of the iron web between the PMs, as in the other two types, and using
finite elements to study the impact of the multiple flux barriers on reluctance.
Although the performance curves of the inset and V-shaped types differ on most of the
parametric calculations, the performance of each with the best combination of parameters, which
is shown together in Fig. 31(b), is strikingly similar. Because of the air-gap flux skewing
capability inherent to the IPM magnetic poles, it is expected that the V-shaped IPM will have
higher output at the highest speeds. This will require further detailed analyses using finite-
element computations.
REFERENCES
38
DISTRIBUTION
Internal
1. D. J. Adams 5. M. Olszewski
2. E. C. Fox 6. P. J. Otaduy
3. L. D. Marlino 78. Laboratory Records
4. J. W. McKeever
External
39