IAC 2024 - IA Fast Track to Search Focused AI Solutions
EV traction motor comparison - Techno Frontier 2013 - M Burwell - International Copper Accociation
1. Performance/cost comparison of induction-motor &
permanent-magnet-motor in a hybrid electric car
Malcolm Burwell – International Copper Association
James Goss, Mircea Popescu - Motor Design Ltd
July 2013 - Tokyo
2. Is it time for change in the traction motor
supply industry?
Motor-types sold by
suppliers of vehicle
traction motors *
“[Our] survey of 123 manufacturers shows far too few
making asynchronous or switched reluctance
synchronous motors... this is an industry structured
for the past that is going to have a very nasty
surprise when the future comes.” *
* Source: IDTechEx research report “Electric Motors for Electric Vehicles 2013-2023: Forecasts, Technologies, Players”
www.IDTechEx.com/emotors
2
| Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
3. The challenge for electric traction motors:
rare earth cost-levels and cost-volatility
3000
Permanent Magnet Motor Materials (“rare earths”)
Ne Oxide
Dy Oxide
2500
Dysprosium Oxide
2000
Neodymium Oxide
$ per kg
1500
Copper (for reference)
1000
$480/kg
$60/kg
$7/kg
500
0
2001 2002
2003 2004 2005 2006 2007 2008
2009 2010 2011 2012 2013
Source: metal-pages.com, Kidela Capital
3
| Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
4. Background to this work
Today, the permanent magnet motor is the leading choice for traction drives
in hybrid vehicles
But permanent magnet motors have challenges:
•
•
•
High costs
Volatile costs
Uncertain long term availability of rare earth permanent magnets
This makes alternative magnet-free motor architectures of great interest
The induction motor is one such magnet-free architecture
4
| Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
5. This presentation
The work presented here compares two equivalent 50kW tractions motors
for use in hybrid electric vehicles: a permanent magnet motor and an
equivalent induction motor
•
The main analysis has copper as the rotor cage material of an induction motor
•
Motoring and generating modes are modelled using standard drive cycles
•
Important outputs of the work, for each motor type, are:
•
•
•
5
Lifetime energy losses and costs
Relative component performance parameters, weights and costs
Top-level comments on aluminium cages are presented at the end
| Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
6. Overview of the analysis covered in this
presentation
90
)
80
70
p
60
50
Total losses in the motor
(
Permanent
magnet motor
Copper rotor
induction
motor
City driving over 120,000 miles (UDDS)
40
p
30
20
0
500
1000
Induction Motor
2510 kWh
2000 kWh
0
$220
Extra energy cost (internal combustion engine cost of $0.294/kWh)
5. Motor
Performance
1250 kWh
1100 kWh
Extra energy cost (grid price of $0.25/kWh)
1. Driving cycles
2240 kWh
1430 kWh
Combined average losses over 120,000 miles
1500
610 kWh
Aggressive driving over 120,000 miles (US06)
0
1270 kWh
Highway driving over 120,000 miles (HWFET)
10
0
$260
6. Energy Losses
& Costs
Materials per motor
Permanent magnet
motor
Copper rotor induction
motor
Weight
2. Vehicle Model
Magnetics
3. Powertrain
Model
6
Heat Flows
4. Motor Models
| Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
9.1 kg
$64
24 kg
$24
24 kg
$24
1.3 kg
$200-540
0
0
Rotor cage
9. Battery
Capacities
Cost
$31
Permanent magnets
(2011/2013 prices)
7. Inverter
Currents
Weight
4.5 kg
Steel
Permanent
Magnet Motor
Cost
Stator Copper
0
0
8.4 kg
$59
Increased inverter cost
-
0
-
$50
Total
29.8 kg
(100%)
$260-590
41.5 kg
(140%)
$200
Reduction of consumer
purchase price*
-
0
-
$150-980
8. Motor Weights
& Costs
10. Breakeven
Analysis
7. Main conclusions from this work
•
Comparing a 50kW copper-rotor induction motor to a 50kW permanent magnet motor:
•
•
-25% torque density
•
•
No rare earth metals used
+40% weight
•
+10-15% peak inverter current
However, the induction motor is a good alternative because:
•
•
It uses only $260 in extra energy over 120,000 miles
•
•
Total motor+inverter unit costs are $60-$390 less (=$150-980 lower sticker price)
Increased inverter costs are modest at ~$50/vehicle
Battery size:
•
•
•
Can optionally be increased to match increased motor losses
Unit cost savings are larger than increased battery costs up to 27kWh battery size
Using aluminum instead of copper in the rotor of a 50kW induction motor for an HEV:
•
7
Increases losses by 4%
| Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
•
Lowers torque density by 5%
8. 1. Vehicle drive cycles
Three standard drive cycles are used for the comparison of two traction
motors: a permanent magnet motor and a copper rotor induction motor.
The 120,000/10year vehicle life is assumed to be composed equally of these
three types of driving
90
Speed (miles per hour)
80
Driving
cycle
50
40
30
7.5 miles
20 mph
10.3 miles
48 mph
8.0 miles
48 mph
Aggressive
(US06)
20
10
0
0
500
1000
Time (seconds)
8
Average
speed
Highway
(HWFET)
60
Distance
City
(UDDS)
70
| Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
1500
9. 2. Vehicle Model
A standard vehicle model is used to convert drive cycle information into
powertrain torque/speed requirements.
Faero
Frolling
9
Ftraction
| Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
10. 3.1 Powertrain model
A standard two motor/generator hybrid powertrain architecture is used
•
Consists of two electrical motor/generators, MG1 and MG2 and an internal
combustion engine, all connected through a planetary gear set
•
Rotational speed of the internal combustion engine (ICE) is decoupled from
the vehicle speed to maximise efficiency
•
We analyze MG2 for performance/cost
•
We assume that MG2:
•
Has a rated power of 50kW
•
Couples to the drive wheels through
a fixed gear ratio
•
Provides 30% of motoring torque
•
Recovers up to 250Nm braking
torque
•
The ICE and brakes supply the rest
10 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
11. 3.2 Motor torques/speeds produced during driving
cycles
By applying the vehicle and powertrain models we convert the driving cycle
data into motor torque/speed data points. One data point is produced for
each one second of driving cycle
City cycle MG2 loads
(UDDS)
Highway cycle MG2 loads
(HWFET)
Aggressive cycle MG2 loads
(US06)
100
100
50
50
50
0
-50
-100
MG2 torque (Nm)
150
MG2 torque (Nm)
150
100
MG2 torque (Nm)
150
0
-50
-100
0
-50
-100
-150
-150
-150
-200
-200
-200
-250
0
1000
2000
3000
4000
MG2 Speed (rpm)
5000
-250
0
1000
2000
3000
4000
MG2 Speed (rpm)
11 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
5000
-250
0
1000
2000
3000
4000
MG2 Speed (rpm)
5000
12. 4.1 Magnetic models of permanent magnet
motor and induction motor
The two motor types were modeled for similar torque/speed performance:
same stator outside diameters, same cooling requirements but different
stack lengths
Stator OD = 270mm
Rotor OD = 180mm
Stator OD = 270mm
Rotor OD = 160mm
Stack Length = 105mm
Stack Length = 84mm
Permanent
Magnet Motor
Copper Rotor
Induction Motor
8
Poles
8
48
Stator Slots
48
-
Rotor Bars
62
12 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
13. 4.2 Reference permanent magnet motor
model
The modelled permanent magnet motor is a well-documented actual motor
used in a production hybrid vehicle.
13 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
14. 4.3 Validation of the motor performance
model
The model of the permanent magnet motor was validated against test data
from the actual motor
g
88
91
86
0
4000
5000
6000
82
80
Model data (including mechanical losses)
90
94
96
95
88
88
0
92
691
87
0
9490803
81 4
888
8
2 8
5
88
0
92 70
60
89
96 96
1000
2000
3000
4000
Speed (RPM)
88
5000
93
890 5 30
7 6 82
88 84 1
8
88
6000
14 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
80
Model Data (excluding mechanical losses)
Our analysis continues using
motor performance which
excludes mechanical losses
Test data from actual motor (including mechanical losses)
84
94
82
95
9493
92
94
91 88 87 930 9089 91 88 87 88 9590 840 89
9394
81
82
92
92
86 83 91 6
89
83
60
70
81
80
82
84
70
80
60
81
82
86 90
83
85
84
86 0 87 85 8 70 0
85
0
86
996 7
8
30
95
96
Torque (Nm)
92
93
60
70
80
81
2
83
888
456
87
50
0
90
9387
95
89
93
93
60
70
885
81
80
82
3
4
90 8867 88
92
93
91 89
89
88 87
90
92
91
Torque (Nm)
80
83 81
84 82
85
88 87 86
89
90
83 80
84 81
85 82
86
84
88 88
97 8884 8
83 80
6 1
5 2
87
85
84 86
81
83
8082
94
6
0
3000
Speed (RPM)
150
92
9
858
34
08
882
881
10
89
6097
92
90
88
81
8082
200
100
089
94
0
88
884
836
59 8
1819
0
2
96
7
0
92
91
87 89
86
p
4
0
8
96
858
Model and
actual data
correspond
well
92
91
8
90
87 9
88
84
81
8082 0 83 85 86
1000
2000
87
50
88
8 82
9
100
250
90
89
2
00
1
809
86
85
84
83 91
150
87
200
92
88
250
y
60 80
7088
81
83
2
0
300
94
Efficiency (%)
oss
89
ota
83 5 80
0
84 81
8 82
15. 4.4 Thermal Performance Comparison
Steady-state thermal analysis was used to equalize
cooling system requirements for both motors at a
118 Nm/900 rpm operating point
Permanent Magnet
Motor
Copper Rotor
Induction Motor
92%
Efficiency
88%
780 W
Stator Copper Loss
940 W
0W
Rotor Loss
230 W
0W
Stray Load Loss
140 W
100 W
Iron Loss
180 W
880 W
Total Loss
1490 W
105°C
Coolant Temperature
105°C
2.4 gallons/min
Coolant Flow Rate
2.4 gallons/min
156°C
Maximum Winding Temp
156°C
15 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
19. 6.1 Motor losses during driving cycles
From the motor models, cumulative losses during each driving cycle can be
calculated:
Cumulative losses over driving cycle (Wh)
Time (seconds)
Aggressive driving cycle losses
(US06)
Cumulative losses over driving cycle (Wh)
Highway driving cycle losses
(HWFET)
Cumulative losses over driving cycle (Wh)
City driving cycle losses
(UDDS)
Time (seconds)
Permanent magnet motor
19 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
Time (seconds)
Copper rotor induction motor
20. 6.2 Combined losses over life of the motor
The total difference in electrical running costs between the permanent
magnet motor and the copper rotor induction motor are $220-$260. Over
a typical lifetime of 120,000miles and 10 years, this is an insignificant
cost.
Total losses in the motor
Copper rotor
Permanent
induction
magnet motor
motor
City driving over 120,000 miles (UDDS)
1270 kWh
2240 kWh
Highway driving over 120,000 miles (HWFET)
610 kWh
1250 kWh
Aggressive driving over 120,000 miles (US06)
1430 kWh
2510 kWh
Combined average losses over 120,000 miles
1100 kWh
2000 kWh
Extra energy cost (grid price of $0.25/kWh)
0
$220
Extra energy cost (internal combustion engine cost of $0.294/kWh)
0
$260
20 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
21. 7. Cost of increased inverter for copper
motor induction motor
The copper rotor induction motor/generator requires 10-15% more current
to achieve maximum torque. This requires that the power electronics cost
~$50 more than for a permanent magnet motor.
Copper rotor induction motor
Speed (rpm)
21 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
Peak phase current (A)
Motoring torque (Nm)
Peak phase current (A)
Motoring torque (Nm)
Permanent magnet motor
Speed (rpm)
22. 8. Component cost comparison
The copper rotor induction motor saves between $60 (at 2013 magnet
prices) and $390 (at 2011 magnet prices) costs per vehicle. This translates
into $150-980 purchase price savings for the consumer
Materials per motor
Permanent magnet
motor
Copper rotor induction
motor
Weight
Cost
Weight
Cost
Stator Copper
4.5 kg
$31
9.1 kg
$64
Steel
24 kg
$24
24 kg
$24
Permanent magnets
(2011/2013 prices)
1.3 kg
$200-540
0
0
Rotor cage
0
0
8.4 kg
$59
Increased inverter cost
-
0
-
$50
Total
29.8 kg
(100%)
$260-590
41.5 kg
(140%)
$200
-
$150-980
Reduction in consumer
0
purchase price*
* Assumes materials-cost/consumer-price ratio = 40%
22 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
23. 9. Cost of increased battery capacity to cover
increased motor losses
Using a copper rotor induction motor can require the vehicle designer
to increase the battery size by ~7%. This would allow a customer to
perceive no difference in overall vehicle performance.
Key assumptions used in costing the required increase in battery
capacity:
•
•
•
•
•
Motor must at some time provide all motoring and braking torque in the
highway driving cycle (like a plug-in hybrid electric vehicle)
Induction motor uses 7% more motoring energy than a permanent
magnet motor
Induction motor recovers 6% less braking energy than the permanent
magnet motor
Total braking energy is 20% of the motoring energy over the driving cycle
75% of battery energy is used for motoring, 25% for auxiliary systems
(cabin conditioning, lights, radio, electronics)
23 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
24. 10. Break-even for using copper motor
induction motor
Induction motor cost savings ($)
If the designer chooses to increase battery size for a 50kW system, a
copper rotor induction motor saves total vehicle costs when the battery
size for a permanent magnet motor system is less than 27kWh
600
500
2011 break-even
Additional battery cost*
400
$390 unit cost savings
(2011 Rare Earth prices)
300
2013 break-even
200
$60 unit cost savings
(2013 Rare Earth prices)
100
0
0
10
20
30
Permanent magnet motor battery capacity (kWh)
24 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
40
* Assumes 2020 battery
pricing of $200/kWh and 7%
battery capacity increase for
copper rotor induction motor
25. Possible use of aluminum in the rotor of an
induction motor
Aluminum has only 56% of the conductivity of copper, which leads to an
inferior performance when used in the rotor of an induction motor. In a
first-pass analysis of a 50kW aluminum rotor induction motor, losses were
4% higher and power/torque densities 5% lower than the equivalent copper
rotor motor.
Aluminum rotor induction motor
96
86
0
88 89
87
86
8885
13
8884
02
0
1000
2000
88
87
85
8186
82
83
8084 70
6
3000
89
91
90
4000
88
91
90
886
06
05
4
3
8817 0
927
89
88
87
86
85
83
82
81
8084
5000
6000
60
70
80
81
82
83
84
85
86
87
88
89
90
91
92
88
60
70
81
4
3
89880
5
6
7
82
8
86
90
91
92
50
100
84
90
92
90
150
60
70
8180
82
835
84
86
887
88
89
90
90
91
60
70
80
81
82
83
84
85
86
87 8
90 89 8
91
92
91
881
884
93
06
27
58
7
60
0
8
200
92
93
94
90 91
92
100
88
90
150
Motoring torque (Nm)
90
91
92
92
250
4
8
03
6
885 0
681888
782 7
89
90
94
16 60
4
3
8 8
8057
8827 8
89
91
200
96
300
Efficiency (%)
60
70
80
81
82
83
84
85
86
87
89 88
250
90
Motoring torque (Nm)
300
82
80
Speed (rpm)
25 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
50
0
90
88
87 89
86
85
84
83
82
81 80
2000
1000
92
91
90
8889
87
86
85
84
83 80
82
81
4000
3000
Speed (rpm)
84
608
7817
88
4
3
6
5
2
89880
89
88
87
86
85
84
83
82 80
81
6000
5000
82
80
Efficiency (%)
Copper rotor induction motor
26. Main conclusions from this work
•
Comparing a 50kW copper-rotor induction motor to a 50kW permanent magnet motor:
•
•
-25% torque density
•
•
No rare earth metals used
+40% weight
•
+10-15% peak inverter current
However, the induction motor is a good alternative because:
•
•
It uses only $260 in extra energy over 120,000 miles
•
•
Total motor+inverter unit costs are $60-$390 less (=$150-980 lower sticker price)
Increased inverter costs are modest at ~$50/vehicle
Battery size:
•
•
•
Can optionally be increased to match increased motor losses
Unit cost savings are larger than increased battery costs up to 27kWh battery size
Using aluminum instead of copper in the rotor of a 50kW induction motor for an HEV:
•
Increases losses by 4%
26 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013
•
Lowers torque density by 5%
27. Thank you
For more information please contact
malcolm.burwell@copperalliance.org
Phone: +1 781 526 5027
james.goss@motor-design.com
mircea.popescu@motor-design.com
Phone: +44 1691 623305
27 | Comparison of IM & PMM in a hybrid electric car - Tokyo - July 2013