Analysis and Design of Non-Rare-Earth Traction Motor and Drive
14th January 2015
Prof. Patrick Luk, Dr Weizhong Fei, Cranfield University
Prof. Pickert Volker, Dr. Chris Morton, Newcastle University
Prof. Keith Pullen, Dr. Niall McGlashan, City University London
Objectives
2
• “The primary aim of this project is to showcase a high performance and low cost traction drive and has a sustainable supply chain that is critical for the uptake of the EV market”. Specific objectives are:
1. Develop a high performance ferrite motor with full functional integration with its converter.
2. Develop a power converter for the ferrite motor. 3. Mechanical and thermal integration of the motor and the controller with
the following targets.
• $12/kW; 1.2 kW/kg; 3.5 kW/L,
• efficiency 93% (10%-100% speed at 20% rated torque)
• All data is based on 70°C inlet temperature at 8 l/min water/glycol 50/50 mix
3
Design Rationale • Flux focusing to maximize PM torque --> high rotor pole number (>6)
• Saliency to boost reluctance torque--> multilayer interior magnets
• High power density --> high rotational speed
• Limited switching frequency --> low rotor pole number, low rotational speed
• Rotor integrity limitation -->low rotational speed, simple rotor structure
Sufficient PM air-gap flux density
(flux focussing)
Rotor mechanical
integrity (stress
limitation)
Rotor pole number
Aspect ratio
(rotor inertia)
Rotor complexity
Efficiency of converter
(PWM frequency)
Estimated real power 20kW
Rotational speed 10,000rpm (rated); 20,000rpm (maximum)
Efficiency >93%
Nominal Bar Bus Voltage 300V
Ambient temperature 60 degrees
Pole Pairs 4
Cooling Water cooled
4
Magnet Layer Number(Initial)
One layer
Two and three layers can deliver much higher torque than one layer, two and three layers deliver almost the same torque
Two Layer
Three machines delver almost same torque ripple
Three Layer
The demagnetization decreases as the magnet layer increases, two layer is chosen for compromise between rotor complexity and performance
5
Integrated optimisation and design
Start Stator
parameters
Thermal
analysis
Rotor
geometric
optimizationDemagnetization
analysis
YES
NORotor stress
analysis
NO
Finalise
design
YES
6
Rotor Axial Slot Optimization
One slot Three slots Five slots Ten slots
19
21
23
25
27
29
31
0 15 30 45 60Rotor Position (Degree)
Ele
ctro
ma
gn
etic
To
rqu
e (N
.m) one slot three slot five slot ten slot
22
23
24
25
26
1 3 5 10Rotor Axial Slot
Av
era
ge
To
rqu
e (N
.m)
8
8.2
8.4
8.6
8.8
9
1 3 5 10Rotor Axial Slot
P-P
To
rqu
e R
ipp
le(N
.m)
Ten-slot configuration is chosen for the final design
7
Stator Slot Comparison
30 slots 36 slots 42 slots 48 slots
17
19
21
23
25
27
0 15 30 45 60Rotor Position (Degree)
Ele
ctro
ma
gn
etic
To
rqu
e (N
.m) 30 slots 36 slots 42 slots 48 slots
20
21
22
23
24
30 36 42 48 Stator Slot Number
Aver
ag
e T
orq
ue
(N.m
)
0
1
2
3
4
5
6
7
8
30 36 42 48 Stator Slot Number
P-P
Torq
ue
Rip
ple
(N.m
)
8
Stator Slot Comparison
6
7
8
9
10
11
12
13
14
0 15 30 45 60Rotor Position (Degree)
Ele
ctro
ma
gn
etic
To
rqu
e (N
.m) 30 slots 36 slots 42 slots 48 slots
26
28
30
32
34
36
38
40
0 15 30 45 60Rotor Position (Degree)
Ele
ctro
ma
gn
etic
To
rqu
e (N
.m) 30 slots 36 slots 42 slots 48 slots
34
36
38
40
42
44
46
48
50
52
54
0 15 30 45 60Rotor Position (Degree)
Ele
ctro
ma
gn
etic
To
rqu
e (N
.m) 30 slots 36 slots 42 slots 48 slots
Half rated current One and half rated current Twice rated current
0
5
10
15
20
25
30
35
40
45
50
55
0 0.5 1 1.5 2Current (pu)
Av
era
ge
To
rqu
e(N
.m)
30 slots 36 slots 42 slots 48 slots
0
3
6
9
12
15
0 0.5 1 1.5 2Current (pu)
P-P
To
rqu
e R
ipp
le(N
.m)
30 slots 36 slots 42 slots 48 slots
0
5
10
15
20
25
30
35
40
45
50
55
0 0.5 1 1.5 2Current (pu)
Op
tim
al
Cu
rren
t A
ng
le (
Deg
ree)
30 slots 36 slots 42 slots 48 slots
9
Stator Slot Comparison
0
0.2
0.4
0.6
0.8
1
1.2
30 36 42 48Current (pu)
Dem
ag
net
iza
tio
n (
%)
22.5828081
1.73327075
21.5602789
17.0611506
0
5
10
15
20
25
30 36 42 48Current (pu)
Dem
ag
net
iza
tio
n (
%)
Rated current One and half rated current Twice rated current
30 slots 36 slots 42 slots 48 slots
By considering the average torque output, torque ripple, and overload capability, 36
stator slots are chosen for the prototype machine.
0
0.05
0.1
0.15
0.2
0.25
0.3
30 36 42 48 Current (pu)
Dem
ag
net
izati
on
(%
)
10
Final Design Parameters
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0 30 60 90
Rotor Position (Degree)
Co
gg
ing
To
rqu
e (N
.m)
-100
-80
-60
-40
-20
0
20
40
60
80
100
0 30 60 90
Rotor Position (Degree)
Ph
ase
Ba
ck E
MF
(V)
Stator Slot Rotor pole Turn per Coil Coil per Phase R (phase) Ld Lq
36 8 2 12 (In series) 8.7mΩ 0.21mH 0.55mH
Stator outer diameter Stator inner diameter Air gap Stator stack length
144mm 96.4mm 0.5mm 86.5mm
Rotor stacks Rotor axial slots Magnet volume
6.5mm*11 1.5mm*10 206.4cm3
11
High Fidelity Machine Model Initialize rotor position �N, N=0
Initialize stator current IdN, N=0
Initialize stator current IqN, N=0
Perform nolinear FEA at rotor �N
and load current IdN and IqN
Fix permeability for each
element in the FEA model
Set stator currents to zero Set PM residual flux density as zero
Perform linear FEA to compute
PM flux linkage
Perform linear FEA to compute
winding inductances
All IqN,
Completed?
No IqN= IqN+1
All IdN,
Completed?
Yes
IdN= IdN+1No
All �N,
Completed?
Yes
�N= �N+1No
Stop and save PN
flux linkage and
inductance matrice
Yes
High Fidelity Machine Model based on the extracted PM flux linkage, inductance matrice,
end winding leakage, and reluctance
Circuit simulator with high fidelity power switch model and control toolbox
•Machine design validation •System level simulation •performance prediction •Controller design •Demagnetization assessment and prevention
12
Control Structure MTPA
PI
PISVPWM
Power
Module
PMSMSensor
udr
uqr
id
iq
wm
abcib
iadq
q
wmr
iqr
+-
+-
+-
idr
13
Portunus with High fidelity model
Rotor Slice Design • The principle design
problem faced is to contain the magnets and stator lamination wedges against centrifugal action.
• Recap: Tried various geometries using a supporting peg which proved adequate when rotor diameters were low.
Rotor Slice Design -Cont • Magnetic analysis
showed that rotor diameter and number of poles needed to increase
• The result was unacceptable stress levels in the peg type arrangement even after optimisation.
• A pin supported design was suggested to reduce the stress in the lamination steel to an acceptable level
Rotor/Shaft Mechanical Features
17
Final Rotor Stress Analysis
Motor Cooling Design - Model • A 1.5D, transient heat transfer
model was developed to estimate radial temperature distribution in the motor.
• The model allows for heat generation in stator, rotor and the windage gap.
• The composite rotor design is modelled as two overlapping fins and the stator by considering winding and slot conduction in parallel.
Outer fins with
convective HT
Outrer casing
Stator – no slots
Two path model
windings + slots
Air gap with
windage model
Outer rotor slices
Overlapping fin
model of rotor and
shaft
Hollow shaft
cooled with oil in
the bore
Heat
Flow
Motor Cooling Design - Results • Assumptions: 1. Average coolant T =
72.5⁰C. 2. Motor Nominal Power
20 kW. 3. Combined loss in motor
of 5%: 4% shared amongst stator components; 1% in rotor.
4. Windage loss and heat transfer from: ESDU 07704 and Howey et al 2012.
5. hcooling fin based on flat plate convection model.
6. hoil based on thin film conduction model.
Steady State 20 kW 50 kW for five minutes after steady state at 20 kW
Overall Package Design
Key Features of Final Design • Nominal rating: 20 kW at 10,000 rpm.
• Max speed 20,000 rpm.
• Fully subcritical machine design.
• Water cooled aluminium body with integrated cooling of semiconductors. a) Fins on motor barrel arranged in a multi-start helical format
b) Semiconductor cooled by fined heat exchanger aligned with main fins.
• Oil cooled rotor and bearings. a) Oil lubricated/cooled bearings fed with oil through hollow shaft arrangement.
b) Two spring preloaded, deep groove bearings used to support main shaft.
c) No cooling fan required.
Future works • Prototype manufacture and assembly (by End of March)
• Rotor integrity test at high speed in City University London
• Improve high fidelity machine model based on 3-D FEA results, and temperature effects
• Full performance profile evaluation of the machine
• Demagnetization analysis to derive current vector limit for each rotor position, embedded in the final controller design to prevent demagnetization
• Develop more sophisticated control algorithm based on high fidelity model to improve overall performance
• DSP programming and machine on load testing
22