PM Generator Characteristics for Oscillatory Engine Based

Portable Power System

A. Zachas, L. Wu, R.G. Harley & J.R. Mayor

Grainger CEME Seminar 5 March 2007

Slide 1 of 26

Sponsored by Powerix Technologies under contract to

DARPA DSO

Overview

Introduction Research Objectives Portable Power System Oscillatory Motion Generator Model & Waveform

Characteristics Initial Power Estimate Influence of Generator Parameters Experimental Results Conclusion

Slide 2 of 26

Introduction

Need for lightweight portable power system in remote locations

Meso-scaled internal combustion swing engine (MISCE) developed to operate from several fuel sources

Oscillatory motion as opposed to rotational motion

Characteristics of a surface mount PM generator determined for this motion

Slide 3 of 26

Research Objectives

Understand the oscillatory motion Model oscillatory motion in FEA package Determine characteristic waveforms for

oscillatory motion Evaluate effect of generator parameters

on the characteristic waveform Use waveforms to estimate output power

from generator Experimental validation of simulation

results

Slide 4 of 26

Portable Power System

MICSE Power Generation System technology is a synergy of two novel energy conversion devices, micro-swing engines and swing-optimized PMAC swing-generators

Micro Internal Combustion Swing Engine (MICSE) converts high specific energy liquid fuels to oscillatory mechanical power (chemical-mechanical)

MICSE systems are internal combustion engines with four chambers separated by an oscillating swing arm

Mechanical-to-electrical power conversion via a direct-coupled swing-optimized permanent magnet AC induction generator

MPG systems are adaptable to a wide range of practical fuels, including butane/propane and JP-8

MPGs can be based on two-stroke and four-stroke MICSE designs and enables application-specific tailoring of the power system

1 2

34

Slide 5 of 26

Initial Swing-engine Testing In-chamber static combustion testing

matched earlier calorimeter testing with cold-start quench losses <40%

Early testing with the swing-engine revealed significant seal failures resulting in chamber-to-chamber leakage

1.77 1.78 1.79 1.8 1.81 1.82 1.83

-2

0

2

4

6

8

10

12

14

16

18

Time (sec)

Pre

ssure

(psig

), V

oltage (

V),

or

Positio

n (

deg)

Cham 1Cham 2Cham 3Cham 4Spark 13Spark 24SA position

-10

0

10

20

30

40

50

60

70

80

90

0 0.02 0.04 0.06 0.08 0.1

Time From Spark (sec)

Pre

ssu

re (

psi

g)

MTD Chamber Test 2

MTD Chamber Test 3

Calorimeter Chamber

Slide 6 of 26

MICSE SummaryMTD-1Ex 350W MICSE System

Power Level (W) 350

Thermal Efficiency 20%

Quenching Loss 20%

Trapping Efficiency 20%

Burned Mass Fraction 80%

Chemical to Mechanical Efficiency 2.2%

Size (LxWxH)3.3“ x 4“ x

6"

Total mass (with active valve-train) (g) 1213

Valve-train power draw (W) 20

Slide 7 of 26

Oscillatory Motion

Harmonic

Hz %

1 60 89.7

3 180 7.5

5 300 1.5

7 420 0.7

( ) cos(2 )ˆm t ft

Slide 8 of 26

Generator Model

Modeled with 2D finite element package and used a transient solver to apply motion to the rotor

No load back emf and flux linkage determined by the software

A1+A1-

A2-

A2+A3+

A8+

B1+

B8+

B2-B1-

C7-

C8-

C1+C8+

SOUTH

NORTH

Slide 9 of 26

FEM No Load Results

Slide 10 of 26

Initial Power Estimates

Performed FFT analysis on no load back EMF to find dominant harmonics

Estimated winding resistance and winding inductance using machine geometry

Connected no load back EMF as the source to a standard 6-diode bridge rectifier and a resistive load

Slide 11 of 26

V-

V+

V b 4

V c 4

V a 4D 1 D 5D 3

D 6D 2

R lo a d

3 1 5

C d

3 0 0 m

I C = 3 9 3

D 4

L a 0 . 0 1 3 5 8 1 3 6 7

L b 0 . 0 1 3 5 8 1 3 6 7

L c 0 . 0 1 3 5 8 1 3 6 7

V b 1

V a 3

V c 1 R c

4 . 9 1 3 5 9 9 3 0 6

R a

4 . 9 1 3 5 9 9 3 0 6

R b

4 . 9 1 3 5 9 9 3 0 6

V b 2

V c 2

V b 3

V c 3

C 3 5 . 0 9 5 5 u

R 3 1 e 9

C 2 6 . 1 6 5 5 u

R 4 1 e 9

C 1 5 . 0 9 5 5 u

R 5 1 e 9

V a 2 V a 5V a 1

V b 5

V c 5

V a 6 V a 7

V b 6 V b 7

V c 6 V c 7

FFT (Ea)FFT (Eb)FFT (Ec)

Initial Power Estimates

Maximum current density of 6 x 106A/m2

% of input power used for additional losses (friction, windage, hysteresis, armature reaction, etc)

Output power of about 417 W

Slide 12 of 26

Case

PLoad (W)

VLoad ILoad RLoa

d

Pin (W)

Papp (W)

Iarm

s

Ibrm

s

Icrm

s

1 497.7108.1

54.60

223.5 537.8 632.35

5.06

5.08

4.81

2 489.2108.3

54.51

524.0 527.2 622.30

4.98

5.00

4.74

3 481.0108.56

4.431

24.5

518.0612.4

54.89

4.92

4.67

4 473.1108.7

54.35

025.0 508.9 602.80

4.81

4.84

4.61

Stray Load 5.18 W

Friction & Windage 5.18 W

Copper 37.00 W

Stator Core

16.25 WArmature Reaction 36.26 W

Loss Breakdown: Case 3 RLoad = 24.5 Ω

Input Power 518 W

Total Losses 100 W

Average Power

418 W

Efficiency 81%

Generator Parameters

Magnet pitch to coil pitch• full pitch to maximize induced back EMF

Slide 13 of 26

Generator Parameters

Tooth and stator thickness and material• maximize material utilization through FEA studies• minimize core loss effects due to increased oscillation

frequency• carried out extensive material study• considering fabrication & availability, fine non-oriented

electrical steel (Si Fe) was selected

Slide 14 of 26

MaterialSaturation

Flux Density

Core Loss at 400Hz (@

1T)

M19 – 26 Gauge (0.47 mm) 1.7 T (17 kG)

24.48 W/kg

Cogent NO 005 (0.12 mm) 1.8 T (18 kG)

11.8 W/kg

Metglas™ 2605C0 (0.023 mm)

1.8 T (14 kG)

6.0 W/kg

Hiperco® 50 2.2 T (22 kG)

17.64 W/kg

Generator Parameters

Number of magnet poles• more poles produce “conventional shape at peak speed

60 slot 20 pole rotational

60 slot 20 pole oscillation

Slide 15 of 26

Generator Parameters

Number of magnet poles• more poles produce “conventional shape at peak speed

18 slot 6 pole oscillation

30 slot 10 pole oscillation

Slide 16 of 26

Generator Parameters

Magnet thickness• large compared to airgap, yet try not waste material

1mm magnet thickness

2mm magnet thickness

Slide 17 of 26

Generator Parameters

Magnet thickness• large compared to airgap, yet try not waste material

3mm magnet thickness

3.5mm magnet thickness

Slide 18 of 26

Generator Parameters

Cogging torque• reduce by skewing of the magnets• Stagger magnets to create skew• 3 steps of 4o to give 12o skew

Rotor Starting position• found to have little or no effect if the number of

poles and swing angle were large Rotor Speed

• increased through use of a 1:4 gearbox by maintaining the same swing frequency but covering a larger swing angle

Slide 19 of 26

Swing-optimized PMAC Prototype

MTD-1G1 MTD-1G2Stator outer diameter (mm) 62Rotor outer diameter (mm) 32.5Axial length (mm) 40Rotor Inertia (kg.m2) 2.615 x 10-5

Air Gap (mm) 0.25Number of stator slots 30Number of poles 10Winding AWG 22 bifilarNo. turns per phase 130,130,13

0140,130,140

Phase Resistance at 100oC (Ω)

0.681 0.61,0.58

Max. RMS current (A) 5Mass (kg) 1.4

Thermo-mechanical design optimization studies resulted in integrated cooling fins and to allow >6A/m2 current densities

Stator windings were potted with thermally conductive epoxy improve winding thermal management

Two 450W PMAC swing-optimized generators were fabricated with different winding configurations for maximum copper fill factor

Stator ring and spider laminates

Slide 20 of 26

Experimental Validation 1

Four-bar linkage built to simulate MICSE motion – converts rotational motion to oscillatory motion

Gearing provides 4x increase in speed and amplitude compared to direct drive

Slide 21 of 26

A

B

C

Front

Rotation

Oscillation

Experimental Results 1

Slide 22 of 26

MTD 1G2: Maxwell 2D Back EMF vs Time

-30

-20

-10

0

10

20

30

0 0.02 0.04 0.06 0.08 0.1 0.12

Time (sec)

Vo

lta

ge

(V

)

Phase A Maxwell

Phase B Maxwell

Phase C Maxwell

MTD-1G1 No-load 8.4Hz

-30

-20

-10

0

10

20

30

0.000 0.020 0.040 0.060 0.080 0.100 0.120

Time (s)

Vo

lta

ge

(V

)

Phase A

Phase B

Phase C

MTD-1Gx Model Validation

Simulated Back EMF at 8.4Hz

Measured Back EMF at 8.4Hz (MTD-1G1)

Slight differences in model & prototype

Mismatch between simulation & test frequencies

Approximate velocity profile 2D FEA model without skew Mechanical slip and

vibration present in four-bar linkage

Load Oscillatory Performance

Test

Freq.

Vrms

Power

1 2.15 2.5 3.8

2 4 5.0 15.1

3 8.66 10.3 63.2

4 16 18.9 213.5

55 ~700

Oscillatory Power Testing

Actual power measured at frequencies up to 16Hz Estimated power at 55Hz is >700W based on FEA

simulation

Slide 23 of 26

Experimental Results 2

Slide 24 of 26

MTD-1Gx Model Validation

Simulated Back EMF at 16Hz

Measured Back EMF at 16Hz (MTD-1G1)

No load back EMF Generator coupled directly

to MICSE and operating at approximately 16Hz oscillation frequency

Conclusion

Ultra portable power delivery system has been introduced

Approximate velocity profile for oscillatory motion for use in FEA has been determined

Influence of generator parameters has been evaluated

Characteristic no load waveforms presented Simulations have been validated with

experimental and test data

Slide 25 of 26

Questions

Slide 26 of 26