Вы находитесь на странице: 1из 279

University of Cape Town

The copyright of this thesis vests in the author. No

quotation from it or information derived from it is to be

published without full acknowledgement of the source.

The thesis is to be used for private study or non-

commercial research purposes only.

Published by the University of Cape Town (UCT) in terms

of the non-exclusive license granted to UCT by the author.

University of Cape Town

IMPACT OF ARMATURE REWINDING ON INDUCTION MOTOR EFFICIENCY IN SOUTH AFRICA

Prepared by:

Prepared for:

Date:

Heskin Mkando Mzungu

Department of Electrical Engineering

University of Cape Town

Department of Electrical Engineering

University of Cape Town

31 August 2009

This thesis is submitted in fulfilment of the requirements for the Masters of Science in Electrical Engineering at the University of Cape Town

University of Cape Town

Declaration

I, the undersigned, declare that this thesis is my own work. All information that is not of my own has been referenced.

Sign:

University of Cape Town

Acknowledgements

I want to firstly thank God for everything because I am nothing without Him.

I also would like to express my gratitude and appreciation to the following people:

Dr. Azeem Khan (Supervisor) Dr. Paul Barendse (Co- Supervisor) Dr. Marubini Manyage

Dr. Ben Sebitosi

Mr. Chris Wozniak

Mr. Richard Okou

Miss. Kgathane Masemola

Prof. Pragasen Pillay

The Advanced Machine and Energy Systems (AMES) research group

The 2007 VCT 1st and 3

rd

year practical students.

And finally, my many wonderful family and friends for their support and love.

11

University of Cape Town

Abstract

The aim of this thesis is to evaluate the impact of annature rewinding on the efficiency of Low Voltage (LV) industrial squirrel cage induction motors in South Africa.

The efficiency of an electric motor is a measure of the effectiveness of the motor to

convert electrical power at its tenninals to mechanical power at its shaft. Although the definition is seemingly simple and straightforward, the detennination of the efficiency of an induction motor is a much-debated topic. Motor manufacturers provide efficiency data obtained through measurement and calculation according to a variety

of international standards. Several international standards exist, with each outlining

different methods and procedures for the detennination of induction motor efficiency.

Most notable among the disparities

is the treatment of stray losses. For example, the

Japanese standard JEC-37 assumes stray losses to be negligible, others such as SANS

34-2 and ASINZ 1359.5 use a fixed value, while IEEE 112, CSA 390 and lEC 34-2

of these standards were initially

prefer to make actual measurements. A number

considered. However, after preliminary laboratory-tests were perfonned, it was

observed that the IEEE

112 method B (2004) and IEC 60034-2 segregation method

(2007) appeared to be the most consistent and repeatable. The two standards were

therefore preferred and subsequently chosen for this project. The South African

standard, SANS 34-2, is available but its methods of detennining efficiency have been

found to be unsupported due

to its reference to the IEC 60034-2 (1984) which has

been abandoned and replaced. The SANS 34-2 was therefore not used in the testing.

Around the world, motor repairs have always been known to cause efficiency loss and

perfonnance deterioration. Rewind studies have been done in the United States,

Brazil, Canada, the United Kingdom and elsewhere [32] [37] [59] [63]. The results from most of these studies have however been generally inconsistent with various efficiency loss figures ranging from less than 1% to 6% [32] [37] [59] [63]. It has even been suggested that each rewind would result in a 1% loss for multiple rewinds [37]. With so many inconsistent reports and given that workmanship would vary from

111

University of Cape Town

country to country, there was an obvious need for a South African investigation be undertaken to ascertain the real situation in the country.

Three induction motor test-beds were constructed in the Electrical Machines Laboratory at the University of Cape Town (UCT). The test-beds provided flexible ways of supplying, loading and measuring the performance of a range of induction motors. Meticulous attention was paid to all aspects of the test-beds, such as motor mounting, power quality, instrumentation accuracy and calibration. This was to ensure accurate and repeatable results.

New squirrel-cage induction motors ranging from 3kW to 55kW were purchased for

testing. This range

to assess the accuracy and

industry. Several tests were performed on each motor

of motors is the most commonly found in the South African

repeatability of test results. Statistical analyses were performed on the test results for

each motor to obtain a representative average efficiency performance curve for each

motor. The motors were then sent for complete armature rewinds and retested

according to the aforementioned procedures.

of rewinds was found to be significant with

From the results obtained, the impact

efficiency drops ranging from 0.1 to 1.86%. The changes in the core and stator losses

had the biggest contribution on the change in efficiency. The increase

could be attributed to damage on the core during the winding removal process. Better

insertion of the stator winding coils while keeping the same number of effective turns

would produce lower stator conductor losses. A change in the efficiency profile was

also evident as a result of the changes in the constant and load losses.

in core losses

In addition to the above tests, the impact of motor rewind procedures was investigated

by sending two new identical 3kW motors to be rewound separately by two South African repair companies. The results showed a decrease in efficiency of 0.96% for the larger company and 1.09% for the smaller company at full load. This would appear to suggest that the different techniques and procedures used in different companies affect the losses and efficiency differently.

IV

University of Cape Town

Table of Contents

Declaration

i

Acknowledgements

ii

Abstract

,

iii

Table of Contents

v

List of Figures

ix

List of Tables

xiii

Nomenclature

xv

Chapter 1: INTRODUCTION

1

1.1. Overview

1

1.2.

1.3. Motors and Motorised systems

Background

1.3.1. Motor applications

1.3.2. Induction Motors

1.3.3. Motor repair industry

1.3.4. Impact of rewind and repair on motor induction efficiency

1.4.

I.S. Methodology

1.6.

1.7. Structure of thesis

Contributions

1.8.

Objectives

Limitations

Chapter 2: TEST BED DEVELOPMENT

2.1.

2.2. Motor selection

2.3. Laboratory capabilities

2.4. Dynamometer System

Overview

1

2

4

S

6

7

9

9

9

10

10

12

12

12

12

13

2.4 .1. Electric motor!generator dyno

14

2.S. Construction of the test beds

16

2.S.1. 3kW test bed

16

2.S.2. ISkWtestbed

18

2.S.3. 2S0kW test bed

22

2.5.4.

Couplings

27

v

2.5.5.

Bearings

28

University of Cape Town

2.6.

Instrumentation

29

2.6.1. Torque measurements

29

2.6.2. Voltage, Current, Frequency and Power measurement

36

2.6.3. Speed measurements

38

2.6.4. Resistance measurement

39

2.6.5. Temperature measurement

.40

2.6.6. Alignment

42

2.7.

Power Supply

43

2.7.1. Power quality

43

2.7.2. Laboratory Generator Supply

47

2.7.3. Power Ratings

2.8. Motor operation, safety and maintenance

2.8.1. Motor starting on

2.8.2. Motor starting on 250kW test bed

2.8.3. Safety and maintenance

3kW and 15kW test beds

2.9. Concluding remarks

Chapter 3: DATA ANALySIS

3.1.

3.2.

3.3. Repeatability

3.4. Regression Analysis

3.5. Uncertainty or error Measurement

3.6. Concluding remarks

Overview

MATLAB

Chapter 4: INDUCTION MOTOR EFFICIENCY DETERMINATION

4.1.

4.2. Definition of Induction Motor Efficiency

Overview

48

48

.48

.49

50

52

53

53

53

53

56

57

62

63

63

63

4.2.1. Induction Motor Losses

63

4.2.2. Motor Efficiency Profile

66

4.3.

Determination of induction motor efficiency

69

4.3.1. Motor efficiency test standards

69

4.3.2. Comparison of Standards

72

4.3.3. Determination of

72

4.3.4. General Differences

80

VI

4.4.

Preliminary test results

88

University of Cape Town

4.4 .1.

Comparison

of test

standards

88

4.4.2.

Comparison of test methods

93

4.5.

Concluding remarks

 

95

Chapter 5: REPAIR AND REWIND OF INDUCTION MOTORS

96

5.1 . Overview

96

5.2.

MotorFailure

96

5.2.1.

Motor Stresses

 

97

5.3.

Motor repair procedures

97

5.3.1. Inspection, dismantling and testing

98

5.3.2. Winding removal and core processing

98

102

103

103

104

104

105

105

106

107

107

112

118

Chapter 6: IMPACT OF REWINDING ON MOTOR EFFICIENCY: TEST

RESULTS

119

119

120

119

5.3.3. Winding and varnishing

5.3.4. Assembling of motor

5.3.5. Final testing

5.3.6. Impact of poor repair procedures on induction motor loss

5.4. Motor rewind procedures and practices in South Africa - Case Study

5.5. Rewind Company Profiles

5.5.1. Company A

5.5.2. Company B

5.6. Rewind Procedures

5.6.1. Company A

5.6.2. Company B

5.7. Concluding remarks

6.1.

Overview

6.2. Test Methodology

6.3. 3kW motor results

6.4. 7.5kW motor results

122

6.5. llkW motor results

124

6.6. 15kW motor results

126

6.7. 22kW motor results

127

6.8. 37kW motor results

129

6.9. 45kW motor results

132

VB

6.10.

55kW motor results

134

University of Cape Town

6.11.

Motor loss and efficiency discussion

135

6.11.1. Motor loss discussion

136

6.11.2. Efficiency discussion

141

6.12.

Comparison of rewind procedures

143

6.12.1. Company A loss results

144

6.12.2. Company B loss results

145

6.12.3. Efficiency comparison of rewound motors

147

6.13.

Concluding remarks

147

Chapter 7: MOTOR REPAIR VERSUS REPLACEMENT ECONOMICS

148

7.1. Overview

148

7.2. Options available when motors faiL

7.2.1. Repair versus Replace Decision

7.3. Motor Systems

7.4. Case Examples

7.4.1. Scenario 1 ~ Initial purchase costs

7.4.2. Scenario 2 ~ ACO

7.4.3. Scenario 3 ~ SPP

7.4.4. Scenario 4 ~ LCC

7.5. Concluding remarks

Chapter 8: CONCLUSIONS AND RECOMMENDATIONS

8.1.

8.2. Recommendations for Future Work

Conclusions

REFERENCES

APPENDIX

Vlll

148

149

155

155

156

156

157

157

158

159

159

160

162

169

University of Cape Town

List of Figures

Figure 1-1: Electricity consumption and demand in South Africa (data from 2006) [3]

 

2

Figure 1-2:

Distribution of motors according

to size

3

Figure 1-3: Estimated motor applications [5]

5

Figure 1-4: Example of an Induction motor

5

Figure 2-1:

Schematic of Generator Set

14

Figure

2-2:

3kW test bed

16

Figure 2-3: Cradled DC dynamometer

 

17

Figure 2-4: 4-Quadrant 16kVA DC Drive

Figure 2-5: 15kW test bed

Figure 2-6: 15kW test bed: transducer system with C-channels

Figure 2-7: C-channels distortion

Figure 2-8: 15kW test bed with steel base

Figure 2-9: Adjustable base design

Figure 2-10: Resistor banks

Figure 2-11: 250kW test bed

Figure 2-12: Deep tooth and groove contact [19]

Figure 2-13: Pulley belt variables

Figure 2-14: Force deflection apparatus

Figure 2-15: HBM inline torque transducer

Figure 2-16: 500V 4-quadrant DC drive

Figure 2-17: Motor couplings

Figure 2-18: Deep-groove ball bearings

Figure 2-19: Torque determination from DC Dyno loss

18

19

19

20

20

21

22

23

24

24

26

26

27

28

28

30

Figure 2-20: Mounted of a 100kg Load cell

31

Figure 2-21: Torque Calibration

32

Figure 2-22: Linearity of good, rigid transducer mounting

32

Figure 2-23:

Poor mounted of Loadcell

33

Figure

2-24:

Error in linearity due to poor transducer mounting

33

Figure 2-25: Wheatstone bridge circuit [25]

34

ix

University of Cape Town

Figure 2-26: Inline Torque calibration using the lever-arm-method

35

Figure 2-27: Transducer linearity

36

Figure 2-28: WT1600 Yokogawa power analyzer

36

Figure 2-29:

Current clamps

37

Figure 2-30: Contact/Photo tachometer

38

Figure 2-31: 30 teeth gear with a proximity sensor

38

Figure 2-32:

Yokogawa galvanometer

39

Figure 2-33: Thermocouple installation

41

Figure 2-34: Pico logger measuring thermocouples

.42

Figure 2-35: Shaft alignment using a clock dial

42

Figure 2-36:

Alignment of pulley system using the clock dial

.43

Figure 2-37: Eskom Supply Voltage THD

Figure 2-38: Eskom Supply Voltage unbalance

Supply voltage magnitude

Figure

Figure 2-40: Eskom Supply voltage frequency

Figure 2-41: Eskom Supply impact on repeatability

Figure 2-42: Eskom Supply impact on motor losses Morning (red), Afternoon (blue)

46

.46

.44

.44

.45

.45

2-39:

Eskom

Figure 2-43: Schematic of Generator Set

Figure 2-44: Star-delta starter

Figure 2-45: 250kW test motor

Figure 2-46: Protection for the 250kW test beds rated at 250A

Figure 2-47: Test area security gate

Figure 2-48: DC dyno commutator cleaning

Figure 3-1: Repeatability of the 22kW efficiency over five test cycles

Figure 3-2: Repeatability of the 22kW losses over five test cycles

Figure 3-3: Motor testing system [69]

47

.49

.49

50

51

51

55

56

60

Figure 4-1 : Average motor losses in 4-pole Induction motors

64

Figure 4-2:

Typical

loss components of an induction motor plotted against load

66

Figure 4-3:

Typical

efficiency profile of an Induction

67

Figure 4-4: Typical efficiency and loss curves of an induction motor

68

Figure 4-5:

Eh-star test circuit

75

Figure 4-6:

F&W and core loss separated graphically

76

Figure 4-7: SLL correction

78

x

University of Cape Town

Figure 4-8: Random error in calculated SLL [49]

79

Figure 4-9: Equivalent circuit according to the IEEE [II]

81

Figure 4-10: Calculated SLL from the IEEE, IEC and JEC

90

Figure 4-11: Impact of temperature correction on Stator loss

92

Figure

4-12:

Impact

of temperature

correction on

Rotor loss

92

Figure

4-13:

7.5kW test comparison

 

93

Figure 4-14:

llkW test comparison

94

Figure 4-15:

15kW test

comparison

94

Figure 5-1: Loop test wiring diagram

100

Figure

5-2:

Company A's

workshop

105

Figure

5-3:

Company B's

workshop

106

Figure 5-4: Company A's Controlled-temperature oven

Figure 5-5: Commercial core tester from Henry du Preez & Associates

Figure 5-6: Company A's winding area

Figure 5-7: Company A's dipping tank

Figure 5-8: Curing oven

Figure 5-9: Reassembled motor 55kW awaiting final tests and painting

Figure 5-10: Coil removal using blow torch

Figure 5-11: Old removed windings

Figure 5-12: Core testing facilities

Figure 5-13: Hand winding with counter

Figure

Figure

Figure 5-16: Company B's curing oven

Figure 5-17: Company B's variable voltage supply control panel

Figure 6-1: 3kW motor efficiency change

Figure 6-2: Change in losses of the 3kW motor

Figure 6-3: 7.5kW motor efficiency change

5-14:

5-15:

Insertion of the windings on the 3kW motor Company B 's varnish dip tank

l08

108

110

111

111

112

113

113

114

115

115

116

117

118

120

121

122

Figure 6-4: Change in losses of the 7.5kW motor

123

Figure

6-5:

11 kW motor efficiency change

124

Figure 6-6: Change in losses of the l1kW motor

125

Figure 6-7: 15kW motor efficiency change

126

Figure 6-8: Change in losses of the 15kW motor

127

Figure 6-9: 22kW motor efficiency change

128

Xl

University of Cape Town

Figure 6-10: Change in losses of the 22kW motor

129

Figure 6-11: 37kW motor efficiency change

130

Figure 6-12: Change in losses of the 37kW motor

131

Figure 6-13: 45kW motor efficiency change

132

Figure 6-14: Change in losses of the 45kW motor

133

Figure 6-15: 55kW motor efficiency change

134

Figure 6-16: Change in losses of the 55kW motor

135

Figure 6-17: Stator Copper losses percentage change

136

Figure 6-18:

Rotor Copper losses percentage change

137

Figure 6-19: Change in motor characteristics due to motor rewinding

137

Figure 6-20: Core losses percentage change

139

Figure 6-21: Stray Load losses percentage change

Figure 6-22: F&W loss percentage change

Figure 6-23: Impact of Armature rewinding

Figure 6-24: Shift in peak

Figure 6-25: Load-dependent and load-independent losses

Figure 6-26: Impact of winding at Company A

Figure 6-27: Percentage loss change of the 3kW rewound in Company A

Figure 6-28: Percentage change due to Company B Rewind

Figure 6-29: Percentage loss change of the 3kW rewound in Company A

Figure 6-30: Damage to the slot teeth due to winding extraction

Figure 7-1: Replace-Repair decision model [EASA]

Figure 7-2: Initial costing of rewind and replace

Figure 7-3: Effects of varying variables

Figure 7-4: Typical motor system

xu

140

141

141

142

143

144

145

145

146

147

148

150

154

155

University of Cape Town

List of Tables

Table 1-1: Eskom target generation mix [1]

1

Table 1-2: Estimated market demand for electric motors in SA

.4

Table 2-1: Selected motors for project

12

Table 2-2: Test bed power ratings

14

Table

2-3:

Belt

Installation

data

25

Table 2-4:

DC dyno output power method vs reaction torque method

30

Table 2-5: Accuracy of Power Analyzer

37

Table 2-6: Accuracy of resistance calculation

40

Table

2-7:

Standards Power Quality Limits

.43

Table 2-8: Generator Power Quality Limits

Table 2-9: Test bed power ratings

Table 3-1: Repeatability STD for each motor

Table 3-2: Reproducibility of a 7.5kW motor on the 15kW test bed

Table 3-3: REE, Uncertainty's influence and importance for llkW efficiency

Table 3-4: REE, Uncertainty's influence and importance for 37kW efficiency

Table 4-1 : International induction motor testing standards

Table 4-2: Harmonized standards and available methods

Table 4-3: Available methods in each of the standards [10] [11] [42]

Table 4-6: The instrumentation requirements of the three standards [10] [11] [42]

Table 4-7: Comparison

standards

Table 4-8: SLL as a percentage of input power

Table 4-9: Core loss values of IEEE and

Table 5-1: Distribution offailures in Induction Motors [32]

Table 5-2: Impact of rewind and repair procedures on induction motor loss

48

.48

54

59

61

61

70

72

80

86

of full-load efficiencies of motors as determined by several

89

90

91

96

104

Table

5-3:

Rewind company profiles

107

Table

5-4:

Decision based in core testing results

109

Table 6-1: Other changes due to motor rewind

121

Table 6-2: Other changes due to motor rewind

123

Table 6-3: Other changes due to motor rewind

125

Table 6-4: Other changes due to motor rewind

127

Xlll

University of Cape Town

Table 6-5: Other changes due to motor rewind

129

Table 6-6: Other changes due to motor rewind

131

Table 6-7: Other changes due to motor rewind

133

Table 6-8: Other changes due to motor rewind

135

Table 6-9: Core tester results

138

Table 6-10: Efficiency comparison of Company A and Company B

147

Table 7-1: Repair/Replace Decision table

156

Table 7-2: ACO of l5kW motOf

156

Table 7-3: spp for the 15kW at 50 - 100% loading

157

XIV

University of Cape Town

Nomenclature

DME:

DSM:

ERC:

UCTML:

Dyno:

MUT:

DC:

AC:

SANAS:

IEC:

IEEE:

SABS:

SANS:

CSA:

JEC:

SG:

THD:

SLL:

Department of Minerals and Energy Demand Side Management Energy Research Centre University of Cape Town's Machine Laboratory Dynamometers Motor under test Direct current Alternating current South Africa National Association of Standards

International Electrotechnical Commission

Institute of Electrical and Electronic Engineers

South African Beareu of Standard

South African National Standard

Canadian Standards Association

Japanese Electrotechnical Commission

Strain gauge

Total harmonic distortion

Stray load loss

xv

University of Cape Town

University of Cape Town

Chapter I: INTRODUCTION

1.1. Overv iew

Thi, "hapler gi,e"; rdeyam background imo the researeh and nutline the 'truclurc of the the,i,_

1.2. Background

E,kom generate< 95·;' of South Africa's (SA) electricity end 4~'Y. of lhe e leclricity u""d in Africa. II al,o Own' and operate, the South Atr",,"n national tr"llSmi"i,,., system Ill . An in,talled "apacily of 4, O:l7MW (with maximum capacity of 38

744MW) is ochieved thru\lgh " yeneTdlion mi, of pumped "Grage, eoaL hydro.

mlcl . "r ond g",,",'dic,d. The co"lrihulinn from each source is shown in TobIe J-1 with

co"1 being the 1"",.,( cnmrihulor.

r ,bl , I-I, '_'''on1t>r~" 1('"o,,";no m], III

 
 
 

C,,.I

   
 

,'1)],·

1

Only,

for pe>k &l'PP/Y whell "'cOcO

1l)"m,'Nud

 

,

 

IH

:!i '.•

Y. mp

i

&

1:'"

4%

10"

up

",)<1t

P' t\lrtH n,

Only,,,, r",p"'"

,"pply wile" ,,«d,d

hop"'''

   

:!%-

11 \ ;,

The demand and cOllSumption nf ek"lricily per 'e,'lor m Soulh Africa is sho"", in

Fi~'Ure 1-1.

The

ind\lstriol

,",clnr ho,

Ihe

l"rge.,;1

cnn,umplion

and

demand

of

electricity, up to J 7.7% consumption ond J~ .~% dL.",and. followed by re,idemial and

then mining.

University of Cape Town

University of Cape Town In 2006, SA experienced an incre"o in eloclricily demand. eHu>od by political
University of Cape Town In 2006, SA experienced an incre"o in eloclricily demand. eHu>od by political

In 2006, SA experienced an incre"o in eloclricily demand. eHu>od by political and

~cotllllllic fact''':', This resulted in the dem'Uld exceeding the supply [2] and

Th~ problem w""

,ub""lU<01\[ f""c~d outage, and interruptions around the counl"

al"" cumpound<Od by othel' factors such as t1>" rising and volatile fu~1 price" global

pl'l-.blem, hav~ highlighted

the nc"d I,,, "n~"1"gy effiei"ney and hH~ """.lerat"d rh" drive IO"

"nergy·dlici"n l,

teehnolo~e, "nd pmelie,",

w.rming C''''C"TTL', "nd <lepcT1d"nc~ on r,,,:e ign oil. The,e

n!

The Department 01 Minerai> and En"rgy (Dl>!E) ",I C<)n1l'reh"mive largel' for en.r~y

~mci~"cy improvements to ease the pre"tKe on EskOlll, The tar~et set. by th e DME.

for 2015 i, " rotal ~!ectrical energy reducrion of 12%. to be achieved as foUow>:

industri"l and mining (15'1.), powel' generation (15%), commercial and I"lblic huilding' (I 5%), the rc,id"ntial sectOr (lO'%) and the trdn,;pnrt =tOr (9'1. ) [3]_

1.3. Mo ton an d 'lolori~NI sys tems

MOlors "",I mOll)riseci sysl~ms around lhe world are ~3tilllated to account for 40',1, of

global ekelrie en"rgy comumption [4]

6(1"1. of 'he u)tal Lnciusrriallmining included) energy u,", and atoom 57%. of th~ peak

In South Ali'ica, motor, account fol' up to

d""mn{1 gL'Tlemtion 141, According to E,kom Demand Side \\,na~"""'T" (DSM) I I I, "n e>timated total of 100 000 m~to,,; are opeMin~ in the In{lu"ri,1 He~tOf, and

consume up to 10 GW of eleelricity (the C<luival ent to tbe thou 1.6-million h"",el\(1l<t'l

power oonsumption of more

Figure 1_2 ,I;ow, the dimibulion of motor, according to size_ The data "as gallli-ttd hy E'kom DHM in a mOlor ,ur.-. y conducted in 2007_

r

,, ,,

,

,

, , , j , , , , , , • , ,- ,, !
,
,
,
j
, ,
, ,
,
,
• ,
,-
,,
!
• ,-
-
~~~~~~~~~~#'$###'#"~4;'#
~'
~
~o '~~.'
".~
~~
,,~ ,~
I' '" ~'I>- & of ~,~ ~ ,
'"
"
'
,.
"
.0 '
.f
• .
"
,0',"-
."'
D"ta from Somb Africa', largest 01010" dislrihulO< ,llQw, tbat motors ranging from
0,18 to 90kW are the moM COmmon ,ize; and {Iominale about 97% of total electric
motor market ,hare
in &>ulh Afrie" 151, Table 1-2 ,how. the e'timated demand pel'
molor ratin~ in SA.
University of Cape Town

~

University of Cape Town

T.hk 1·2: ["III1.trd markot d<m•• d '0<.I.<t!'!, IlIOro<l ,. SA

3.1. 1

"

"

"

1' 00
1' 00
IlIOro<l ,. SA 3.1. 1 " " " 1' 00 " " Tot.1 UoJ<> .\1010r applic"lions
IlIOro<l ,. SA 3.1. 1 " " " 1' 00 " " Tot.1 UoJ<> .\1010r applic"lions
IlIOro<l ,. SA 3.1. 1 " " " 1' 00 " " Tot.1 UoJ<> .\1010r applic"lions

"

"

,. SA 3.1. 1 " " " 1' 00 " " Tot.1 UoJ<> .\1010r applic"lions Th~

Tot.1 UoJ<>

.\1010r applic"lions

Th~ key induslri"" in which electric motors operate in SA are [5]:

P~trochL,"ical

\1ining

Pulp ""d Papo:r

Iron "00 Stc"l

Sugar TIle motor, peJionn varying fitnctiuns r"nging from ,imple conditioning applicatioos to dri\'ing the mills lI"'t crush millions of tOIlS of coal for power station" A sUl",ey Jone by 1"" Energy R~search Cen[J'e (ERC) found that the application of motor, i, di"ribll1ed according to figm e 1·3 ,

University of Cape Town

MIS(e 2S%
MIS(e
2S%

Fans. 2~'\

University of Cape Town MIS(e 2S% Fans. 2~'\ In dllellO/l mOl"", (sqw JTc1 cab"') ~fC tbc
University of Cape Town MIS(e 2S% Fans. 2~'\ In dllellO/l mOl"", (sqw JTc1 cab"') ~fC tbc

In dllellO/l mOl"", (sqw JTc1 cab"') ~fC tbc la r gcst 5JIl~C

_ Pumps

I - FllrlS

DGompo"eS$Ot'S

.-

IDCon~

,,"" o f clc<:mc'l y '"

,ndus1ry '" SImlh Africa and ;,round lbe "'.odd , This IS bec ause mducnon mo~.

cros.<·

nnd

Induct,on motOOl arc electromechan,cal dc"c.' that ~o",i" or a c"', and b~j"nccd

,OtO< and a fixed Slater <x"'taining ,o"duc'1\'c ""ndings that

induced produce ~

of 'InOOcrn" ",dumy [6] [7]. Squirrcl,c~ie

:<:lIonal

~rc klw

,".w ,hu

n

,

the

n in r;SUf{;

"'o

hon.,'!i

1·4. an: sl"'ple. , '=t;10 and robust rn. chtne

hen

,""snel,e field 1""1 spin. the rolOr.

,--
,--
an: sl"'ple. , '=t;10 and robust rn. chtne hen ,""snel,e field 1""1 spin. the rolOr. ,--

,

University of Cape Town

Induction mOlars therefore ho' e the lorgest jlOIemial for energy saving, with oppomlllities ffinging trom the motor user to the utility ,upplying i[ with powe'-

1.33. '.I'Hm· r~p"ir industry

Mutor rqlllir ar~1 rewind in South Africa have i:>cen llIc,eosing uver the p.st t"w years [1ll" " 0 ,"sui I of increo,ed P,"",U," on indu,lry co",ed by 181;

• An in",ease in the intensity of ntilllng activities, as well as the establishment of new mines in SA_ With umlergroluld mining being the mOre Cummon tw e in Soulh A frica, the depth and si".e 0 f thes<; mir"" lead tu mure u,e of mowrs

and th~Tcfore incre"",,,,, in the need fur more mutor repmr.

• An inere"e ill infrn"trucull-e "uch

os rooo,. ,oilways. th e Gammin project and

hOllSlllg in preparation for the Fifa 2010 soccer World Cup_

• Eskom proje<:ts such

as the supply ca]Xlcity expansion (commi"ioning of

m()[hballed power swions: Camden. Groolvkl .nd Komali)

tt;qum: the rc-pair of electric moto",_

Th",e pmjecis

• Ihe sudden increo "" in industry and um"lllsa,ion of countr;", such as Chin.

ond Indi" has IN to a ,Iecreas e m the d e li,""lY of motors due to 'he press UTe On

the world's economy.

In 2005, it wa, rc

I{ ~.~illion on mOlur rqlll'" I~ 1_

,ortcd

th.1 the mining industry 1lI S<.)uth A fnc" "pent approximmely

De'pite havin~ "uch " larMe rev.-ir~1 in,lu'iry in South Africa. it is no< known how

variou s

thesi , ai m' to inve,rigate the imp"et of rev.-inds on inductionllloto, efficiency

rewinding techni4ue , u""d by motor repai re r>; an"ct motor e!1lciency_ Thi s

University of Cape Town

1.3.4. Impact of rewind and repair on motor induction efficiency

It is widely known internationally that repairs affects induction motor efficiency. A review of research into quantifying this impact on efficiency shows various ranges of efficiency loss.

Ontario Hydro conducted a study in 1991 on electric motor rewinds to determine the effects of rewinds/repairs on electric motor efficiency. Nine 15kW (20Hp) standard efficient motors were damaged then randomly sent to electric motor rewind shops in Canada. The average loss of efficiency was about 1.1 %, with the greatest reduction around 3.4%. The efficiency losses were attributed to reduction in conductor cross sectional area, core damage resulting in hot spots, and reduced insulation systems.

[59]

A similar study was performed by BC Hydro in 1993 on ten energy efficient motors.

The results were different from the study done by the Ontario Hydro. The average

loss was 0.5%. The greatest losses were found to be due to increased bearing losses,

due to friction, whilst there was little or no increase in core losses observed. The

higher friction losses were due to bearing replacement with lower quality bearings

[59].

In 1994, J. C. Hirzel of General Electric found efficiency losses ranging from 1.9 to

6%. These results were from four motors ranging from 10 - 200Hp. [62]

The Canadian Electrical Association (CEA) undertook a controlled electric motor

repair study in 1995. Several repaired motors were burnt out and underwent the same

repair process, including burnouts, dip and bake. Little or no reduction in efficiency

was found when the motors went through a controlled repair process. Repeated

rewinds (after three rewinds) also showed little to no reduction in efficiency. [59]

In 1998, Dreisilker Electric Motors and the University of Illinois performed a study on the impact of burnout temperatures on electric motors. Temperatures above 320'C were found to cause the stator frames to distort. This was observed on both aluminium and cast iron frames. [59]

7

University of Cape Town

In 2003, the Electrical Apparatus Service Association (EASA) and Association of Electrical and Mechanical Trade (AEMT) undertook a larger research project to detennine the impact of repair and rewind on induction motor efficiency. The objective was to prepare a 'Good Practice to Maintain Efficiency' document. Twenty- two new motors ranging from 37.5 to 225kW with two smaller motors of 5.5KW were tested before and after rewinding. Motors were tested at the University of Nottingham in the UK. Six motors (75-112kW) averaged losses of 0.4%, ten motors (45-150kW) averaged losses of 0.03% and five motor (75-150kW) averaged increases of 0.5%. The report concludes by stating that motor efficiency can be maintained when 'good' repair practice is followed. [32]

In 2006, Wenping Cao et al perfonned a comprehensive rewind study involving 23

motors ranging from 5.5 to 225 kW. Cao's results showed that rewinding could

impact on motor efficiency and perfonnance. Careful control

of the stator winding

design while keeping the same number of effective turns will produce a lower stator-

conductor loss. This will offset any slight increase in the core, windage, and friction

losses. Even repeated rewinds do not cause an appreciable change to the motor

efficiency on average [37].

In 2007, Bortoni et al conducted a study in Brazil on a set of eight low-voltage motors

with ratings from 2.2 to 11 kW (3 - 15 hp). The study was to detennine the influence

of rewinding on motor efficiency. The results showed that motor efficiency had a

small increase after repair with full-load efficiency increase between - 0.9 and +2.0%

respectively. The overall quality of the repair shop was found to have an influence on

the impact of the repair on efficiency [63].

Sasol's study into the effects of repairs on motor efficiency found an average drop of 3%. Motors were tested randomly after being repaired [64].

With such diverse conclusions of the effects of repairs on motor efficiency; it is evidently difficult to rely on such results. The rewind procedures followed at a repair shop appears to be the most consistent cause of motor efficiency loss.

8

1.4.

Objectives

University of Cape Town

The main objectives of this thesis are:

• To build test beds that can be used to accurately and reliably test induction motors ranging from 3 - 55kW in South Africa;

• To analyse the factors that influence induction motor efficiency and

• To examine the effect of rewinding on the efficiency of induction motors by profiling the efficiency loss or gain after rewinding process

1.5. Methodology

The methodology associated with this thesis include:

• Efficiency testing of motors of various kW ratings before rewind to determine

the actual efficiency which may differ from the nameplate value;

Liaise with motor repairers to determine the current rewinding practices that

are being employed;

• Performing single armature rewinds on the entire range of motors and

• Retesting the motor efficiency of the rewound motors to determine the impact

of the rewinding process or techniques.

1.6. Limitations

The study is limited to induction motors with squirrel cage rotors because the

efficiency standards are based on these motors, and they are most commonly

used in South African industries.

New motors of capacities ranging from 3kW to 55kW have been chosen for the rewind project. Induction motors in this range are popular in South Africa as shown in Figure 1-2. This range also has the greatest potential for energy savings due to their inherently lower efficiencies compared to larger motors.

9

University of Cape Town

• The motor repairers used in this thesis are all located in the Western Cape regIOn.

1.7. Structure of thesis

The structure of each chapter begins with a brief overview followed by a literature review of the particular topic (in most chapters) and then the results are presented.

Chapter 2 describes the methodology followed in the construction of the test beds and motor testing.

Chapter 3 presents details of the data analysis that was used in all recorded data

Chapter 4 presents literature and results on motor efficiency and the determination of

motor efficiency according to international standards.

Chapter 5 outlines the repair and rewind procedures and how these could impact on

motor efficiency. A review

companies is presented.

of the rewind procedures of two South African repair

Chapter 6 presents the analyses

discussed in chapter 5.

of the results from the repair and rewind testing

Chapter 7 discusses motor economics scenarios.

Chapter 8 presents the conclusions and recommendations.

1.8. Contributions

• Mzungu, H.M.; Manyage, M.l; Khan, M.A.; Barendse, P.; Mthombeni, T.L.; Pillay,P., 'Application of induction machine efficiency testing standards in South Africa', Electric Machines and Drives Conference, 2009. IEMDC '09. IEEE International, 3-6 May 2009 Page(s):1455 - 1462

10

University of Cape Town

• Prof. P. Pillay, Dr. M. Manyage, Dr. A Khan, Dr. P. Barendse and Mr. M. Mzungu, 'Induction Motor Efficiency and Rewind Study', Eskom Demand Side Management Research Report

• H. M. Mzungu, P. Barendse, M. A Khan and M. Manyage, 'Determination of Effects on Induction Motor Efficiency, ICUE Conference', May 2008

• M.A Khan, D. Pati and H.M. Mzungu, 'Comparison of a 3kw Standard and High Efficiency Induction Motor', Vector Journal, June 2008

• Mzungu, H.M.; Sebitosi, AB.; Khan, M.A;. 'Comparison of Standards for

Determining Losses and Efficiency of Three-Phase Induction Motors',

Power Engineering Society Conference and Exposition in Africa, 2007.

PowerAfrica '07. IEEE16-20 July 2007 Page(s): 1 - 6

• H. M. Mzungu, A B. Sebitosi, M. A Khan, 'Comparison of Standards for

Determining Losses and Efficiency of Three-Phase Induction Motors',

Vector Journal, July 2007

11

University of Cape Town

Chapter 2: TEST BED DEVELOPMENT

2.1. Overview

This chapter describes the methodology that was followed through the entire project. Each section presents a detail process from the motor selection to the construction of test beds and finally, the treatment of the collected data.

2.2. Motor selection

Motors ranging from 3-55kW were selected for the rewind project. This range has the largest motor population and demand of motors in South Africa as shown in Figure

1-2 in Chapter 1. The criteria set for the motors were:

• Standard efficiency motors, 4 Pole, 1500rpm, IP 55 protection

• Foot mounted with standard frames sizes, 50Hz, 380V Delta connection

• Duty cycle: SI, Insulation class F

Table 2-1 shows the actual WEG motor sizes and the ratings for each tested motor.

Detailed catalogues for each motor are found in Appendix

A.

Table 2-1: Selected motors for project

Motor size

Speed

Current

PF (p.u.)

Frame

(kW)

(rpm)

(A)

3

1390

6.63

0.84

100L

7.5

1450

12.1

0.87

132M

II

1455

22.9

0.83

160M

15

1455

30

0.85

160L

22

1465

42.6

0.85

180L

37

1480

69.1

0.88

225S/M

45

1475

83.2

0.88

225S/M

55

1475

99.4

0.9

250S/M

2.3. Laboratory capabilities

All testing were performed at the University of Cape Town's Machine Laboratory (UCTML). Three test beds were constructed with great emphasis placed on precision

12

University of Cape Town

and accuracy in order to ensure reliable results with high repeatability (see Section

2.S).

The UCTML is unique in that it has a flexible distribution system and can provide both AC and DC supplies. The lab is rated at SOOkVA and has a SOOkVA substation supplied with IIkV. The IIkV supply is stepped down to SOOV through a three- phase transformer. The lab contains two SOOV, 2S0kW 4-quadrant DC drives connected to two separate 2S0kW, 980rpm DC motors; a 6.6kV, S20kW, S7A 980RPM alternator, a 7SkW AC variable speed drive, 3 and I2kW DC motor capable of running up to 200% of full-load rated conditions; and three I6kW 400V 40A DC drives.

2.4. Dynamometer System

Efficiency testing requires loading of the test motors and the measurement of both

input and output power [10] [11] [12]. Electrical input can be easily measured with a

power analyzer. Mechanical power on the other hand requires the measurement

both the shaft speed and torque as shown in Equation 2.1 [13].

of

Power = Speed x Torque

(2.1)

Dynamometer (dyno) systems are instruments used to place a controlled mechanical

load on devices such as motors. The test system allows for the measurement of torque

and rotational speed, which are used to calculate the output power exerted by motors

"are to motors and motor

(or engine or other rotating prime mover). Dynamometers

drives as oscilloscopes are to electronics - a basic test instrument" [13].

There are different types of dynamometer systems that are used in motor testing.

These are:

• Pronydyno

• Eddy current brake

• Water brake dyno

• Electric motor/generator dyno

13

University of Cape Town

2.4.1. Electric motor/generator dyno

The electric motor/generator type dyno system is the most widely used loading technique and was therefore used on all three test beds with DC generators to dissipate power. The 3kW and l5kW test beds used the cradled type dyno (see sections 2.3.3 and 2.3.4) and the 250kW test bed used the absorption dyno (see section 2.3.5). Table 2-2 shows ratings of the DC generators for the three test beds.

Table 2-2: Test bed power ratings

TestBed

DC Motor type

DC Dynamometer Radng

Inducdon Motors tested

3kW

Separately Excited

3kW

3kW

l5kW

Separately Excited

l2kW

7.5, 11, l5kW

250kW

Separately Excited

250kW

22,37,45, 55kW

An

operates with an adjustable-speed drive. The dyno can be either an AC or DC

machine operating as a generator. The dyno is driven by the test motor and converts

the

electric motor/generator dyno system is a specialised dynamometer type that

mechanical energy into electrical energy [14].

Loading of the test motor can be achieved through armature current control of the DC

generator (Equation 2.2). This can be explained by looking at the separately excited

motor circuit in generating mode as shown in Figure 2-1 [15]. The field is excited

DC

separately by a DC current and its rotor is driven by the test motor at shaft speed rom.

and its rotor is driven by the test motor at shaft speed rom. + Ia If

+

Ia

If

Va

and its rotor is driven by the test motor at shaft speed rom. + Ia If
and its rotor is driven by the test motor at shaft speed rom. + Ia If

R

Figure 2-1: Schematic of Generator Set

14

University of Cape Town

Tl'lIe IOrquc dr:\"doptd by Ih~ an"'l ~r~ " I 110<' 1)(' m o'",. ,, 3 funct IOn of lilt Ilu~ /Iud

nil br COlllmlled by adJ u , un g

3nn~lul'l: ~uon-nl as """"

In EqU311<)11 22 [ I S) . Tot ,

lhe ~nn311lt~,-urt'nll or Ih~ firi<! flu~

l h e ~nn311lt~ ,-urt'nll or Ih~ firi<! flu~ (1,11 Wh.",. '1 - '\c"elorcd \(lf4UC F

(1,11

Wh.",.

'1 - '\c"elorcd \(lf4UC F - field flll~

I, " ."nature currenl

\(lf4UC F - field flll~ I, " ."nature currenl The n ,~ed a""'l~re , IIa &"

The n ,~ed a""'l~re ,

IIa &" b3<~ <

f(I-;.) ;$II"'><I",1

rIM flux and sha fl~ .

(Of

Ih" 1/I"' J<!u"" o f thr ;umalU. " "" . I1Ul '

a nd , he Sum o f I~ armar~IT a nd ~~lcm31

res"t """.' . E,

"",.OII

2 J 1151 .how~ II

, 'cla'">n!'hir

2 J 1151 .how~ II , 'cla'">n!'hir C, " genernted I)C VOIt.lg~ R, - ImU'uIT
2 J 1151 .how~ II , 'cla'">n!'hir C, " genernted I)C VOIt.lg~ R, - ImU'uIT

C, " genernted I)C VOIt.lg~

R, - ImU'uIT =",al1C~

I! - olemal "",;,ta

e<:

(l.ll

~ -.n.fl"pecd

By le.mongmg Equauons 2 2 a nd 2 J. i1 n n be """" tloa. t he ""' P'" c~n he .""1".1

by ftdJlI5l1 ng .11M th~ flux, thr U ltnlal re SlS/al1Ce (R ) or the <

rTrol

<1

T

(2.4)

('u n""1 eont"'] ;< normally u,ed due 10 the WIde

co"lml

range. This ,onl.ol

I'

adllc,~ th.n",gh OOfl \'M~

hral lng Contml•.

wch

11<

f<lu

-qu~r3.l\t I.I C

dn"~ wuh lege", ra,, 'e

"

University of Cape Town

2.5. Co ns l ruc li o n o f th e tes t heds

Accordin:; to Wallace et "1 IYI, the elemenl, for" hig,ltly functional an<1 emcient [e<ting lab"'"tol)' and the production ofooth accurate and reliable result< arc'

• industrial relevant power rating capability;

, ",,'ide rangc of 'el'vice "ollages, with ability to <imulate under., Over· and

unh"lanc~d vollageS:

, Wide ,"n:;e of machine 'peed,;

, Mouming of" wide range ufmadline trame ,izes;

, Ability to test both motors and generatOl~:

Ahility to test po\\ el' electronic converters:

, Abilily W mak~ accuralc mCa,urmLen"

These dement' were du,ely applicd durin:; Ihe cOn>lructlon of the le;1 heds .

2.5.1. 3kW te,t bod

The 3kW t"t t>c<! wa, dc,igned 10 "ccummoda [e mOIOr>; ranwng from 3kW and

belu,, _ II wa, c"",truclcd OIl a nat ""lid h"se wiLb lracb_ Figurc 2_1 ,how, the

complele le't hed.

wiLb lracb_ Figurc 2_1 ,how, the complele le't hed. The DC generalor used as a dyn"momeler

The DC generalor used as a dyn"momeler i, ra1e<1 al 3kW. ['hi, machine" cr"d led OIl " duuble bearin:; :;imhle Sil1.lcluTe and could he Ioa<k:d 10 200 ' ,. or ie;; r" ling. Fi~ure

University of Cape Town

3 ,how, the cradled dynamometer 'ys tem _A 501:!,- load cell is used to measure torque (,ec Sect;"" 2,4. I) "nd nol'l' thc tx: mr>lOr from TIKl:ing_

I) "nd nol'l' thc tx: mr>lOr from TIKl:ing_ 1.5.1.1 . .-\101M ,,,,,Ier le.,-I loadin/: II 16kVII,

1.5.1.1 .

.-\101M ,,,,,Ier le.,-I loadin/:

II 16kVII,

coolIol the current of the dyno_ A multi-tum potentiometer "linwed ""cura!e current

,gOV. 35A, four-quadrant. DC dlin , shown in figure 2-4, wa, used to

comml and therefore torque control

University of Cape Town

University of Cape Town 2.5,2, 15kW te,t bod n", 15 kW te,t bed wa, des igned

2.5,2, 15kW te,t bod

n", 15 kW te,t bed wa, des igned to accommodate motors from 7 .5k W to 15kW The

beJ wa, fiueJ wilh a 15kW DC dynamometer thm could be lo.lded '" 200% of ic,

ralm ~. Three configuration, w ere aUempted;

• In line LUrtlue tnmsJucer ,y,t~,.\ witb a ( ;'c hanne l, ba", (figure 2.5),

• Cra{Iic<IDyno 10.", ," sy'tem tran«lucer wilb a (' -c hannel, ba,e (Fi gure 2-1i)

t!ml

• Cradled D)lIO - force 'ys lem tnm",lucer wiLh" ",lid 'tee! b" ,e (Figure 2-7)

2.5.2.1. Inlin~ turque lrull.«/ucer ~';rll C-cllann~[.

The te-,;I be d "'t! , de s il,med . nd conSlruct.d LO accommC>ciate an inl1l>e tor~uc

transducer. The moWr, wen: mounu:d on C_dWLDcI, ""J (I" lransdu"cr "'" ' ]nounl",1 on two be;"'n~ ,Wnd, '" ,1"-,,,", in Fi~ure 2_5. Tbi, mounlin~ pre",nu:d a number of

ch"lienge" The" li gmncnL or tbe lnn l: ,hafl wa< Jifficull a nd therefi>re high vibrations

were pre&ent dur ing testing. The yibrmions led

tn " number of premature failures of

University of Cape Town

the hearin~'. The ralinM 10 the inline lot4ue Iran;,ducer i< 2(~~m (or 2kNm). WhlCh

result e d in it

tem I1"'rfnn""d failed with calculaled negative cnm;lalinn [ac1<>« (dcL1ih in Ch"rt~.,.

3).

being operated"t US low e ,t range rosulting to brge errors [16J_ All the

US low e ,t range rosulting to brge errors [16J_ All the 1.5.1.], Cradled dYlw -

1.5.1.], Cradled dYlw - force tru

Tfw M UT and DC dYTlo wc'" moumed directly ",·ith C-ch"nnels mountings. The dj110

"'"' cradled (see fi~'U", 2-6), similar to the JkW dj110 system. with a I OOkg load cell

holding the d)-1Io limn rocking_

Jucer

.\yMem with C·d,annel.'

to the JkW dj110 system. with a I OOkg load cell holding the d)-1Io limn rocking_

University of Cape Town

II wa, found th at the C-dwnn el type moWr mOUllting' dis!011ed when boll' woro , ighten cel (rigure 2 -n Thi, di,t011ion led to what i. known a, -",II fnot' [17]. M()[o1'<

with ",fi [!lOt move during operating. Thi, movement resul" in vibrations, which leaJ In increased motor lo,ses. and e\ enuwlly to premature bearing milure

l'_'OaJu)CI d"'o<tLon dunn~ bolt I [igtltening Je'"I["l~ In ,,""'"
l'_'OaJu)CI d"'o<tLon dunn~ bolt
I
[igtltening Je'"I["l~ In ,,""'"
mJ,,"IL~n")clu ,nd "Ix.tioJ>O
mJ,,"IL~n")clu ,nd "Ix.tioJ>O Various C-channels were used for the dinen:n ' MUT. Tbis

Various C-channels were used for the dinen:n ' MUT. Tbis m""nt ~xh"u,ti\"e motor

overhauls eluring MUT changeover,. The,e di,advantage, resulled in abandoning the

C-c il , TlTl ol, anel n:<.lc,ignmg tho te,t beJ mountings.

1.5.1.3.

Cradled dJnll _flm'~ IramduNr .'y.llcm with .,Ieel bu.'~

Sol id still ba,e, We"'" m""hin~d for fx,\il (h. MUT .I1J Ihe DC Jyno. Tili, can be "'en

itt rigure 2-8 with a mOl"Hed llkW motor. The bas" "'OS d~'ign"d to accOllllllooate

.

11 to,IOO mOl"", (7.5 - 15~W) Wi lhout any mounting ovemaul

d~'ign"d to accOllllllooate . 11 to,IOO mOl"", (7.5 - 15~W) Wi lhout any mounting ovemaul '"

'"

University of Cape Town

Figure 2·'1 ,11<"", Ihe adju",aI>lc hase design, which can accommo<imc li11tM IC't motors by shifting the two ,i,k ,reel"

li11tM IC't motors by shifting the two ,i,k ,reel" Fi~"" 1-9, .~dj",'.bk b."

Fi~"" 1-9, .~dj",'.bkb." ""';~.

The ha,., "1;.0 provided helter MlJI' mounting and this in tum signiflclint1y re,lueed

mol",- vibralion,

1.5.1

1.

MUT /",,,/inf(

Loading of the MlJI' on the 15kW test bed was ochieve,1 with li combinmion of

re,is tor bank and current control of the same DC drive used in the 3kW lest bc<i

(figure 2-10). The resistor hanks were used to load the 15kW test moWr liNlve rat«i

load, as required by the testing standard. The hi ghe't loading achi eved Of! tI", 15kW

molo.- wa, 125% of"ICd load.

University of Cape Town

,:.

·

·

·

.-

!

I

\

,.

-"

.'

,

2.5.3, !SDkW tOM bed

' In.,

25 (JkW

1,"'1

f>c~1 "'

,

JC'"~,,,,(! n" th e

IJ C rMI .', na,c

plate (~'i~u," 2-1 1) ,

,, '

til e

I'll "

',, 'e,a ll

b. ", p i "'" j, n~id. nat and fonn> a S"\g1 ~ pl~ ,,~ Tho .<ha ntag~ of Ill; ,

i\rucl~1\: noma;!1, Mx'u,

,l y p< "" "oned nnd f'<<>, ' ,d."

rhe 2.W~W ,~" h.,d ."",,;trod"" 0<1 ,I." N~

pi",

~'Kld damr'n~ "r , -in,.lin.,

r' 7]

,Iw,"'" ;n Fi~u", 2_11

MOblrs

.al~d frul n 2l h. 1<> 55 k W wen; «-so," "" Ih" I~ brd

"

University of Cape Town

University of Cape Town The 250kW DC dYIMmometcr W", mounted on four hase plates and the

The 250kW DC dYIMmometcr W", mounted on four hase plates and the !>ILj on a

tIm plme a, shown in Fi~ure 2-1 I , The molars were tightly r,,,leood down u;.ing high-

tension hoUs_ T1,e coupling of the dynomometer and the MUT W"' done lhrough a

pullcy ,y,lem The pulley ,y'l~"" eom'ert",1 lhe 1000 rpm 'p"ed rating of the DC

dy 1>O 10 Ihe 15()() rpm ,pc",1 raling oflhe testmolo","_

1.5

1.1.

Pu/lq,},,\fem

Th e pulley system was designed for a ge"r ",rio ,,1'2;) [l'Ki. m"'imum power tran,fcr

of90kW_ The pulley sys tem brought ahout " munher of ad""nt'g e , in LCTMt

maximizeJ the limited floor space and

molors easy

made the remon,1 and adjustment of the Ie,t

U

i) Pulley BellS

Hcll>. fTOm Gatc, Synchrouou, Hell" were uscd hcc",,;.e or the dcep tooth and groove contact a, ,",own in FilP're 2-12, This deep contact pTOvidcd ,uperior 10.J_oMrying

University of Cape Town

strenglh. reduced no;,e, reduced viilralLoll, and accurale liminy and ,ynchronizmion.

with no io,-, "r lDnjue carryiny cap"ililiL y [I HI

no io,-, "r lDnjue carryiny cap"ililiL y [I HI ii) hmul/m;o" When in'ta lillll: a

ii)

hmul/m;o"

When in'ta lillll: a .ynchrun"", belt, the right len,in" i, required to Jchie". optimum

p"rtormJn"~ 1191. If til<: l~miun i. 100 low the belt's teeth will jump (al,o callffi

ratcli<:tingl ,luling hig,h luad op:.:raling cunditions. If til., tension is too high the I~ll"

life will he ,.educed and possible damage COllld be 1IlCll,.,.ed to ~rinl:', 'han, a[ld

other drive component'_ [19]_ The 1'ecommended inSlalia!;", lm,inn "an bc ~"timat<:d

in tw" ways. Calculating the deflecling fi)l'ce, (lC<jualion 2.5) "nd oaJcu]aling Ihe

,"cormncndcd!:>e1L di,l"lIce deflection (f":quali(J!'l 2_7) [20 1_ Figure 1"13 Ixlow ,hows

lhe variable. used in hju"tinn, 2.5 10 2.7.

2_7) [20 1_ Figure 1"13 Ixlow ,hows lhe variable. used in hju"tinn, 2.5 10 2.7. D
2_7) [20 1_ Figure 1"13 Ixlow ,hows lhe variable. used in hju"tinn, 2.5 10 2.7. D

D

2_7) [20 1_ Figure 1"13 Ixlow ,hows lhe variable. used in hju"tinn, 2.5 10 2.7. D

;

2_7) [20 1_ Figure 1"13 Ixlow ,hows lhe variable. used in hju"tinn, 2.5 10 2.7. D
2_7) [20 1_ Figure 1"13 Ixlow ,hows lhe variable. used in hju"tinn, 2.5 10 2.7. D

Where,

(l.5)

~ - 'l~ n "'I; lin l: force ( N)

I' = uJn!mm~ p""e' (1c\V)

k -

f,() fi\r

"mum

;R.OlalklllO(' 1I."'~;l\n

2~ f" r m""mUm mst~lI"u"" ".",

'"

mst~lI"u"" ".", '" " = Pilch X N X rpm 60 UOO P'I~h - , .:"

" =

Pilch X N X rpm

60 UOO

".", '" " = Pilch X N X rpm 60 UOO P'I~h - , .:" 111

P'I~h -

, .:" 111

«" 1-13,· (mnl)

rpm - sPffil ordn~ D (iff rigut~ 2·ll)

(1.&)

Rn,w

1'1 - number

01 groO\-e'5

"

",/rJ hrlr "rprrli<l~

on d

,,, 0 ( Ott figu~ 2 I J)

s

(1.7)

,.- 50

,.- 5 0
 
 

r

- he ll d Cn C,' Ti <.>n (mml

~ - belt S9"n Itn~th(111111)

University of Cape Town

" j "blo 2·~ ! h oWI the c"k;ula 'e<l dllta tor 11", ~ lt Ill~J llll{jOO

Tobl, I

\.

11,11 1.".11"", " .I,,,

 

PIMIII_j l" '"

0

-c:-'

-

 

"'.,

 

,

 

SI_I

---

-(-,

J

     
 

.

j

"

"

.11'

Th~ dcAccl10n of,~ !:>ell /9.2 mon) aoo ","~,mum forces (328 :-I)"~ both usN III

{he b(:,ll ,n.olaU;Ii'''''' ripn: 1·1 J 1Bo"5 .he 8pp.tr'i1IUS u

fu",~.

m

to measu

,

tbl- deflooion

University of Cape Town

University of Cape Town 1.5.1.1. lim/"~ trtlJl.,d"ur mountinc An HB!>.1 inlin. torqu~ lrnnsduc","

1.5.1.1.

lim/"~ trtlJl.,d"ur mountinc

An HB!>.1 inlin. torqu~ lrnnsduc"," was u&td to ""'a,un; tor<IU~011 the 250kW te' l f>c-<i

(",e Section 2+1 fo1' lbe o~ration). Wilh a cakulat<Xi maximum 1'adial fon;c or

520S"< on the tOl'(I"" transducer from Ih, pulley." stn letur<: WiI.S Ms igned!o wit h'ta nd

thi< forcc. Th<: U""~noduc", ,tru<;l ur. is slKlwn in Figm" 2- 15

igned!o wit h'ta nd thi< forcc. Th<: U""~noduc", ,tru<;l ur. is slKlwn in Figm" 2- 15

University of Cape Town

1

L1.3,

NUT I(lad;~g

The If>ading of the MUT f>n the 250kW test bed i, achieved Ihn>ugh current control

"oing the 5O(1V, 4-<tua<!r"TH, DC drive shown in figure 2-16.

4-<tua<!r"TH, DC drive shown in figure 2-16. 2.5.4. Conpling. F""ner laper lock coupling, were

2.5.4. Conpling.

F""ner laper lock coupling, were used 0" alilhe mot"" f>n the different test beds as

,1>0"" in Figure 2· I 7. The coupling, come in lwf> part'_ the bushing lhal goes onto

quick and easy

installation. The straight edge and machi!led outside otfer quick ~lignm""t. The

rubber. which js wedged between the couplings, acconunodme, millOr misalignment

tho ,han "rod the co"pli"~ Ih"1 goes ove r il The coupli"g.' allow

[21 ].

University of Cape Town

University of Cape Town l ~~ ~~ . ~a r; , In ~II ' hC' Inl

l

~~

~~

.

~a

r;

,

In ~II ' hC' Inl bf'd mOlOJ m""''''Bp lIS

sh" m Figure 2· 18. "fh<-« beannl\S "0:1\' cho:.en bec.au;;e 1hty "a,·~ ttduo:f<j

Sh,eldf'd d~""I"H}\" ball bc·:onnl\S "~r~ u

:d

·"

M~II""31Iriclloo and ",ppon ""''' ro~!oal and ~ laJ I""d. [21] 0I1>ef ad\ · "" t.gcl of

'he d""ejl.j;' ''''''''

m"",.""'"ICC

h<:""n~' "rc .h~ ."~pk d

"j;,gol

. 1o<1~ >cr,·icc

H"

and ,)'"

"~'C or

"rc .h~ ."~pk d "j;,gol . 1o<1~ >cr,·icc H" and ,)'" "~'C or

University of Cape Town

2.6. In st rume ntat ion

Instrumentation plays an il11(Xlrtonl rule in dctc-rmining cmciencr aocur"teiy .nd wilh

hiMh "-1lCatability [22 j. Efllciency testing requires the following m.jor quontities to be

l11co,un;d accurately: Torquc. Speed. Voltage, Current. Po" Re,i,loncc

Te mperarure .nd

,r.

AliMlll11Cllt is Jlso ffilOtber crucial m easurement that play> a part illihe accuracy and

n;pcat.bility

me",urem e nt will be d iscussed ><:pa," [e ly.

each

of the

d"ta.

The

accuracy.

calihration anJ

melhod used

for

2.6.1. Torqu. me••ureruen!.

ro rquc i, Ihc m,,,t crllciol mechanical quamity to be measured in a te,t [16]. The

prc"", mc.,urcmcnt of torque is the most demanding aspecl to the users of Ie&!

bench"

outpnl power (Equation 2.1) a nJ ha, been foond [0 le a J 10 lorgc C'IT{lr.; in erfioic'Y\cy

delennination, ifmeasured

and manufacrurers [i~l. Tor""e i, useJ in lhc Jclerminalion of lhe mo [o'

incorrcc t ly 1221.

T1\ree lor""e measun;mcnl mclhtxi< wcn: te'ted ond used . These ore discussed below.

1.6.1.1.

r"''Iue j'mn th~ output IIf Ih~ DC dynonlmneler

Rearranging Eyualion 2.1 shows thai torque c.n be calculated hy dividing the output

powcr of the -"1UT by ,peed. The OUtpUI (Xlwer of thc MU'I can be caloulated by

oJdi ng the klSses of n,e DC d;110 to the DC J yno OOlpul powcr. ~i1'UTC2·1 ~ illu'tTolcs

thi, pT()cedllre. 11\i, melhod was tcst eJ on th c 3k W te,t bcd using a 3kW motor.

University of Cape Town

.\\uTOlTPL:T

IX' Ol]lpul + TX: 10'"

Il C LOSS

Wiuding 10",

Field loss

Moch"nic"lloSJ;

C LOSS Wiuding 10", Field loss Moch"nic"lloSJ; oc_ DC Ot']'pt']' POWER The ""ull,

oc_

LOSS Wiuding 10", Field loss Moch"nic"lloSJ; oc_ DC Ot']'pt']' POWER The ""ull,

DC Ot']'pt']'

POWER

oc_ DC Ot']'pt']' POWER The ""ull, d""",mented to.md are showu

The ""ull,

d""",mented

to.md are showu in Table 2-4. Large dinereuce in lIlo m"",ured 10000"O W'" !'{Hmd

especially on the lower loading points.

from Ihi, method w",,, compared to lbe m"re accu"He "r~l wcll-

metbod meas\lfing reoctiou torque (oce Section 2,4. 1,2). The """lts

torque (oce Section 2,4. 1,2). The """lts Th" <11~lhOll W"'

Th" <11~lhOll W"' "b""dor",<1 d"e 10 Ihe i.rge enors Ihm were found to lead 10 I,,,,,r

<I~lennin"tion of e I'liciency. The """'" 01" Ihe "'ITors j" the inaccurale deLel'min"ti"Tl of

the DC dyno losses, Details of tl,., calculmions pmCedlU., is L()uud ul App"n<lix fl.

2.6.1.2. Reaction /orqlle ",~as"r~m"nt

Mcasuring th" tmqLLc by m"asurin~ Ihe ,c"ctioTl «!TC" " "commonly u,ed <11et11o<1 of

lOrqllC m"aw,em"n~ c.,peciall}, in "ngiu" ",,,ing Ilili. Th" jl<1Tlcip1c u,e<l i, Ihat the

torque 01 the "llO L' equa l ' tbe reoctioo torqno "Ldyno 'Mm . ' 11, ,, dino is fro: to rolate

hul i, hold", r<",ili(>r\ h)' a lo"d ceiL Figure 2·20 ,how, 11", 15kW "r".U~d dyno and

th" load ceiL

30

University of Cape Town

TI", lo",e acling on Ihe 'Ialor of Ihe d,llo is lranslated to an elccnic signal by a 10<><1 cell (or forcc lransducCT) rIlc load cell. u.<;;ng strain gaugcs i n a \,'healStrtIlC bridge

circuit. is rigidly all""hed to thc d}llo so Ihat any lorcc on thc stator is mc"surcd by Ihc trallSih",er. The volwgc ,ig"ai from thc Iran s<luccr is Im",mi lted 10 a signal

"mpliller ""d Ihen to a display, This wrqu e measurement method W"' use<1 on lhe 3kW "nd 15kW {. ,{ bed"

2.6,}'1• .4ccuracy and Calibration (chanRe the numbering

w nextl~"e/ 1.6.1.2.1)

rile d;s"dn,lllagc found in the re""tion torquc measur~'\1lenl is lhe nce<1 t<Jr ~omplex r""ch""ic"l arrJngemen{s, Poor m''''h"nical arrangement leads 10 \'ibmlions. which

leoo w '!lick. ring· of the last digit' on the display_ The las, of th ese la't digit< low ers

thc torqu e me",uremellt accuracy [23]. Poor mmulting can b. detected thwugh

calibration,

[23]. Poor mmulting can b. detected thwugh calibration, Inc 3kW "ml 15kW lest bcd, w"' done

Inc 3kW "ml 15kW lest bcd, w"' done by loading"

le,'cr "nn with known wei;1>t,", shown in Figure 2-21 [16J, The w. ights used were

"cighed with a c,li~ml.d ,"ale_ The known applied fo1'c. (ami I"",,"fore torqu e) i, used to calibmte Ihc 10a,,1 cell,

C.iibra{ion rtf thc load coil on

University of Cape Town

University of Cape Town 1 1·11: I ~ (Olil> H T1,~ quahly or tJ)~ lood cdl

1

1·11: I

~

(Olil>

H

T1,~ quahly or tJ)~lood cdl and lh~ dY'10 moum"'l! " rrll""lcd m lh" lin~"nl) ",f lhi:

f<)l'w .woqu, tho 1""1 "" _ Hgu

2·22

""'W~ l~ .nlibr.l ion

of Ill"

I"" d

,-en 'm

tl",

I S kW lell b<cd. The weight '

",etC l <>.ldo!<! Utld u"lo~d<d_ Am ino' hY"ic""S1>

"all b<:

~",'''. "hid, n:p"""nls Ii" inh ."" i bo!b a " i,> u! ,,, ,,I h,n,-, .rror of ~h" lo a d coli [ 2 4 J.

• • , " '. " • •
,
"
'. "

"

"

• , " '. " • • • • " " Poor muun"n~ as " ,,,,

Poor muun"n~ as " ,,,,

=<I.arlit-.,

can be dele-'Ied duong ""Iibralion

f,!:."" 2.)2

bel"" >I,,,," > ,"" au""".', tiI'lI nmun""g ~1I.mpl_

University of Cape Town

University of Cape Town n ,~ l.ll. r , , ,l l ~ Although cOlldmon' a"d

n

,~ l.ll. r ,

,

,l

l

~

Although

cOlldmon' a"d

clearly i11

[be !tIou '~ "1i "!' ngld ~,"'ugh. Ihe

Ih~1 r,,'uh.

In

~ Curve

<lr:lI,'<I

",

~'\;UfC 2·2" .

in [he

I"wee h ., h<:nd du n nl: I"~h IQ"I""

hliki cell ,nounlmg I;"

,nl)".

to ,i~.,.,ific"nl em,,- ;n th .

Til;. "

torque

",h;eh 1<:'1(\,

m~

\u""l\"

n\ '}'SIems

I

I

,

.00' '"0_ •• ,,,, ,,,"

,

I I , .00' '"0_ •• ,,,, ,,," , • • ,~ - . - '.---

,~

-

.

- '.---

~

• • • • • •
• •

-

I I , .00' '"0_ •• ,,,, ,,," , • • ,~ - . - '.---

.\.1

University of Cape Town

1. 2.6.1

Inline ""que ",~a,'ure",em

(Numbering should I>e 2. 6.1.1. updau Ih~ reM)

Mea,uring the

This i,; tile mo,; l accurate ",ethou or

Lranwucer ra l eu al 2kN wil h an OULl'ul ,i~"l bed U'i~un; 2.5 in Scction2,},5).

in li ne torque l'eq ui ,,,,; ilie

m~a'u"'ment or lo~uc on !he rotati ng ,ha l\.,_

mline

mC",uri ng IOr'!ue l16]. A HEM TI

of I .5m ViV wa, used on th e 2';.() k V"- tcst

rhc lransducc,- h", " rotor an d " ,tatOI, TIle tOl'que ,ign.1 '" l'wduced by the dastic dCrOTm"lion or Ihe rotor. Thi, is done using '<rain g.uge,; (S(;,). SG s me"""e dctonn" lion with a change in resi'tance pl'opo[tional LO the in duced stmin l25J, The SG, in the IIM Il tron , dueer are conne~""d In WiLc"hlone bridge circul1 wi th temperature compemalion S(;, Thi,,, " ,uwn in Fib~"e 2-25 below,

, , • o • , •
,
,
o
, •

,

,uwn in Fib~"e 2-25 below, , , • o • , • , 1 hc ,
,uwn in Fib~"e 2-25 below, , , • o • , • , 1 hc ,

1 hc , uW iy >o lt" ~e (O,5V - 12V) "nd "'e"s1U'Cmenl ,ign"ls are tron s m itted from tile ,,"u r 10 the , t" tor Q"ou~ ,li>c, ,lip rin~s "nd ~raphite cmbon bru,ll.,. , TIle hrus h

are he ld

[l'an'>ducer to equal i,e tile potential be tween lhe l'otor .nd "alor_ il l" inpUI n;,isl"ncc

is 350 Ohms,

down b y pressme spli ng',

125 1

An extra slip ring

is indudeJ in

the

HHM

University of Cape Town

Th~ inline wrque lransducer oper.le, 10~~lhc"T with a .<kH, carrier-frequency amplifier [26]. The amplifier satisfies two condiliom;

• II "mpli!", Ih , ,mall ,i~'11al (7.5 mV al 1.5m\'iV w,th an excitmion of 5\') inlo" ,uilJble 'iglral of± I0\/

• II tiIl~", hru,h noi"" from Ih, , ib'11a l in md,r to isol",e the componem ttequency !hm i, phy,ic"lly rolev"nl (5 kHz]

The amplifLed signal i, fed into the power analyzer w~

<lp!ion, are availahle. l'ho opcr",ion document, for the TI are found in the AppendIx

a,",raging ;!]1<i ml~ring

-e

,

Au"raq and Calihratit", (chuA ",,,nhering)

2.~.1.4.1.

The Iwnsducer ~"' a O,J% occuwcy, C"libralion "'"" done u, ing the lever·arm_

method [25]. A 2m bar was atlached on one e nd of!lle torqu~ transducer and blockeJ

on the other 'ide. Weight' (weighed by a calibrated scale) were looo ed and Unlo.lded

to chwk fm lin,"rily. The C4uation of []\c ca li b,,"ion li ne was u"d to determine the

re4 uircd (or4 Ue, The c"libr. lion ",(uP i. , hown in FilruTC 2_26

li ne was u"d to determine the re4 uircd (or4 Ue, The c"libr. lion ",(uP i.

University of Cape Town

Figure 2-27 shows a graph of the applied force versus the Olltput volmg~ oflhe torque tran,ducer. The tmnsducer was zeroed Ihroogh the amplifier before loading {he bar I'his compen"'te< for the weigh, of'he bar it;,elf and enSUreS ""curacy or the