University of Wollongong

Research Online
University of Wollongong Thesis Collection University of Wollongong Thesis Collections

1994

Process monitoring and control during spot

welding of metallic coated steels
Stephen William Howes
University of Wollongong

Recommended Citation
Howes, Stephen William, Process monitoring and control during spot welding of metallic coated steels, Master of Engineering (Hons.)
thesis, Department of Mechanical Engineering, University of Wollongong, 1994. http://ro.uow.edu.au/theses/2524

Research Online is the open access institutional repository for the

University of Wollongong. For further information contact the UOW
Library: [email protected]
PROCESS MONITORING AND CONTROL
DURING SPOT WELDING OF METALLIC
COATED STEELS

A thesis submitted in fulfilment of the

requirements for the award of the degree

MASTER OF ENGINEERING HONOURS

from

UNIVERSITY OF WOLLONGONG

by

STEPHEN WILLIAM HOWES

Bachelor of Metallurgy

DEPARTMENT OF MECHANICAL ENGINEERING

1994
This is to certify that this work has not been submitted
for a degree to any other university or institution.

Stephen H o w e s
This thesis is dedicated to my wife Patricia
and my son Adrian.
 ABSTRACT

Metallic coated steel sheet can be very aggressive towards copper alloy, resistance spot

welding electrodes. These steels accelerate the electrode degradation process far in excess of

that experienced w h e n spot welding uncoated steels through an accumulation of the coating

around the electrode tips, changing the geometry of the electrode and through an erosion

process on the tip face in contact with the workpiece. The rate of electrode degradation is

closely linked to the chemistry of the metallic coating and can be very severe, particularly

w h e n welding steels with coatings high in aluminium or w h e n inappropriate welding

conditions are used.

High volume welding industries employing these steels are n o w faced with the need to

more frequently monitor weld quality and electrode condition in order to sustain structural

integrity of their product. Being very time consuming, the traditional shop floor, destructive

techniques for assessing weld quality, are no longer considered appropriate. This has

provided the impetus for researchers to develop new, non-destructive techniques which are

easily implemented and which remove operator judgement as to the condition of electrodes.

This work examines the feasibility of such a technique. The Adept (Advanced Grey

Scale) Vision System has been assessed as to its ability to view and accurately measure the

impression left by spot welding electrodes on the surface of a zinc-iron rich metallic coated

steel during a simulated production run and to equate these measurements to weld nugget
size and tensile shear strength.

T h e results of this work appear promising in that the vision system has identified

electrode wear patterns and has successfully related these to both nugget size and tensile

shear strength, these being paramount to engineers charged with design. The accuracy of

the vision system has been thoroughly checked through the use of other techniques including

carbon paper impression measurements of the degrading tips, profilometry on the

impression left by the indenting electrode and ultrasonic and X-ray measurement of the weld

prior to final destructive peel testing to measure actual weld nugget size. All techniques

i
indicate trends that support the vision systems future capability as a new, non-destructive
weld quality monitoring system.
 A CKNO WLEDGEMENTS

T h e author gratefully acknowledges Dr Elias Siores, w h o provided the initial

encouragement to undertake this work and for his continuous guidance and support during

the period over which it was conducted. Further to this, he encouraged submissions of four

technical welding papers related to the topic, to both international conferences and journals

thereby building on the authors recognition in the welding arena and in particular the field of

resistance spot welding, an area in which the author has some 15 years experience while

working at B H P Slab and Plate Products Division Welding Development Department and

Sheet an Coil Products Division, Research and Technology Centre, Port Kembla, Australia.

Since these submissions were m a d e during the course of the degree, they are presented in

full as Appendices 1 through to 5. The author also acknowledges Dr A Basu w h o provided

guidance during the final stages of the thesis.

The author also wishes to acknowledge the invaluable assistance provided by the

following individuals during various stages of the work:

Dr John Lake, B H P Wire Products Division (formerly of B H P Sheet and Coil Products
Division) for early discussions regarding the scope of the work;
D r E w a n Maddrell, formerly of B H P Sheet and Coil Products Division, Research and
Technology Centre, for instruction in the operation of the surface profilometer
equipment;
M r Jeff Roberts of the Illawarra Technology Corporation for instruction in the operation
of the visual analysis equipment;
D r B o b Harrison of Australian Nuclear Science and Technology Organisation for
assistance with ultrasonic examination of weld samples;
M r Ted O w e n of B H P Slab and Plate Products Division for assistance with radiographic
examination of weld samples;
M r s Cecily M u n r o of B H P Sheet and Coil Products Division for her assistance in
preparation of this document.

Special thanks are due to the management of BHP's Research and Technology Centre,

Port Kembla, Australia and in particular Dr Robert Hobbs (General Manager), for providing

m e with the opportunity to undertake this work and finally to m y wife Patricia and son

Adrian, w h o without their love and understanding, this thesis would not have been possible.

iii
 TABLE OF CONTENTS

Abstract

Acknowledgements

Table of Contents

List of Figures

List of Tables

1.0 INTRODUCTION

2.0 LITERATURE SURVEY

2.1 Resistance Spot Welding

2.2 Monitoring and Control Devices

2.3 Electrical Parameters — General

2.3.1 Welding current

2.3.2 Voltage

2.3.3 Time

2.3.4 Resistance

2.3.5 Monitors for welding voltage and current

2.3.6 Resistance measurement and monitoring of dynamic

2.4 Nugget Temperature

2.5 Infra-Red Radiography

2.6 Weld Expansion and Contraction

2.7 Acoustic Emission

2.8 Ultrasonic Signals

2.9 N e w Monitoring Approaches

3.0 EXPERIMENTAL P R O C E D U R E S A N D PRELIMINARY RESULTS 35

3.1 Materials Used in this Study 35

3.1.1 Metallic coated steel sheet 35

3.1.2 Electrode selection 35

3.2 Preliminary Experimental W o r k 37

3.2.1 Identification of optimum weld parameters 37

3.2.2 Electrode life testing procedure 44

3.2.3 Electrode tip life test conditions for weld parameter study 46

3.2.4 Description of the welding machine 47

3.2.5 Electrode tip life test results — weld parameter study 48

3.3 Description of Profilometer and Vision Sensing Equipment 55

3.3.1 Profilometer 55

3.3.2 Vision sensing 5§

4.0 EXPERIMENTAL RESULTS 62

4.1 Weldability Lobe 63

4.2 Tracking of Electrode Tip Degradation through Carbon Paper 65

Impressions
4.3 Tracking of Weld Impression Diameter through Profilometry 69

4.4 Tracking of Weld Impression through Vision Sensing 73

4.5 Measurement of Nugget Diameter through X-Ray Analysis 74

4.6 Measurement of Nugget Diameter through Ultrasonic Technique 76

4.7 Measurement of Nugget Diameter through Conventional

Vernier Calliper Method 09
4.8 Tensile Shear Strength Testing of Weld Nuggets 84

5.0 DISCUSSION OF RESULTS 88

6.0 CONCLUSIONS
109

v
7.0 FUTURE WORK

REFERENCES

APPENDICES
LIST O F FIGURES

I. The resistance spot welding process 18] 5

2. Weld metal expulsion during resistance spot welding 7

3. Typical weld lobes for 1.2 m m uncoatcd and galvanized steels [21 9

4. Truncated cone and radiused spot welding electrode tip profiles 10

5. Peel test for determining weld nugget diameter and penetration [62] 12

6. The hammer and chisel test [8] 12

7. Typicalrelationshipbetween monitor count and nugget diameter [10] 14

8. Schematic of dynamic electrical parameters during spot welding of mild steel [8] 17

9. The relationship between weld diameter and voltage drop during resistance 18
spot welding [8]

10. The relationship between nugget diameter and the area under the voltage curve 19
produced during resistance spot welding [8]

11. Andrews'model of the simplest form of a dynamic resistance monitor [31] 23

12. A typical dynamic resistance curve for a single pulse resistance spot weld on 23
an uncoated mild steel [32]
13. Positioning of an infra-red emission detector shown as a fibre optic cable 26
light guide' for monitoring of spot weld quality. The upper electrode is
not shown [9]
14. The positioning of dual infra-red emission detectors outside of the welding 26
electrodes and at 45° to the surface of the workpiece [8]
15. Typical results from an infra-red monitoring system showing the 27
relationship between trace amplitude and weld shear strength [8]

16. A characteristic curve of electrode movement (distance designated by 'S') 28

during a single pulse spot weld observed by Stiebel et al [36]
17. Schematic output data of acoustic emission during resistance spot welding [8] 30

18. Schematic of current input into a resistance spot weld [8] 4\

19. Finite element analysis used to measure voltage points within a truncated 43
cone electrode profile (half-section shown) [54]

20. The ratio of current density (J) at electrode face edge (or shoulder) to current 43
density at face centre in a truncated cone electrode profile [54]
21. Positioning of test strips into the welder for electrode tip life testing 46

22. General view of resistance spot welding facility including data acquisition 48
system and lathe for machining electrodes

vii
23. Tip life test results for electrogalvani/.cd, Z I N C A N N E A L (G2 ZFIOO) AI-45% 50
zinc-coated and Z n - 5 % Al-coated steels

24. Tip life test results for hot dipped galvanized steel 51

25. Profilometer pick-up ami and stylus drawn across indented resistance spot 55
weld surface

26. Components of the Taylor Hobson Form Talysurf 120 profilometer 56

27. Schematic of vision system experimental set-up 59

28. Resistance spot weldability lobe for 0.8 m m thick Z I N C A N N E A L ® G 3 N Z F 8 0 65

29. Carbon paper impressions taken during the course of an electrode tip life test 67
on Z I N C A N N E A L G 3 N Z F 8 0

30. Average tip diameterresultsobtained from carbon paper impressions 68

31. A transverse profilometer trace output (weld 3800, transverse direction) 70

showing points of reference for weld impression diameter measurement (x-x)

32. Profilometer trace showing areas rejected as possible reference points (arrowed) 71
for measurement of weld impression diameter (weld 3000, longitudinal direction)
3 3. Average profilometer impression diameter results 72

34. Average vision system impression diameter results 73

35. Representative X-ray prints of welds taken throughout the initial electrode 75
life test
36. Average nugget diameter results from X-ray analysis 76

3 7. Ultrasonic C-scan image for Weld Number 25 .initial electrode tip life test 79

3 8. Ultrasonic C-scan image from Weld Number 2500, initial electrode tip life test 80

3 9. Ultrasonic C-scan image from Weld Number 4600, initial electrode tip life test 81

40. Average nugget diameter results from inner and outer boundaries measured 82
from ultrasonic C-scans
41. Average nugget diameter results measured via vernier callipers after peel testing 83

42. Tension-shear test specimen configuration showing location of spot weld [7] 84

43. Average vision system impression diameter results, second electrode tip life test 85

44. Average tension-shear test results g^

45. Schematic showing the measurement positions through the carbon paper 94
impression method, profilometer and vision system

46. Longitudinal profilometer trace of the 350th weld showing positions 96

corresponding to electrode wear regions
47. Transverse profilometer trace of the 350lh weld showing position corresponding 97
to 'pin-hole cavity on the electrode face
48. Optimum vision system welding window as a function of tensile shear strength 105
49. Optimum weld nugget diameter operating under the vision system 106

IX
 LIST OF TABLES

Types of metallic coated steels 8

Composition and properties of Z I N C A N N E A L G 3 N Z F 8 0 steel 36

Composition and properties of Z I N C A N N E A L G 2 ZF100 steel 38

Composition and properties of Z I N C F O R M G 3 0 0 A 0 6 Z275 (free zinc-coated, 38

hot-dipped galvanized steel)

Composition and properties of Z I N C S E A L C A 3 S N - E E Z 1 0 39

(electrogalvanized steel)

Composition and properties of zinc-5% aluminium coated steel (Galfan) 39

Composition and properties of Z I N C A L U M E G 3 0 0 A Z 150 (aluminium-45% 40

zinc-coated steel)

Electrode tip life test results — weld parameter study 49

Current level recordings for welding lobe construction 65

Wear-in, half-life, maturity and termination zones for the carbon paper, 98
profilometer and vision system measurement techniques
CHAPTER 1.0

INTRODUCTION

1
1.0 INTRODUCTION

The research presented in this work has been driven by an inherent problem faced by

high volume welding industries using resistance spot welding to join metallic coated, sheet

steel components. This problem is the need for frequent line stoppages in order to replace or

redress worn electrodes which have been rapidly degraded through the interaction of the

various metallic coatings with the working face of copper-based electrodes. Without

constant monitoring of electrode condition, weld quality suffers.

W e l d quality itself has been traditionally monitored via destructive techniques

however, m a n y researchers have offered methods for non-destructive monitoring. S o m e of

these are suitable for use while welding metallic coated steels but all of the techniques require

s o m e form of contact with the workpiece. This research presents an effective alternative

based on vision sensing of weld quality in real-time and as such does not rely on physical

contact.

Prior to undertaking any experimental work to assess the potential for a resistance spot

weld quality monitoring and control system to be vision based, an intensive review of the

literature was undertaken.

In the initial phase of this review which is covered in Chapter 2.0, the basics of the

resistance spot welding process is discussed with emphasis on the problems associated with

a change from welding uncoated sheet steels to welding various types of metallic coated

sheet steels. The general mechanisms for accelerated electrode degradation are described,

along with a brief description of current shop floor, destructive weld quality monitoring

procedures. Chapter 2 then covers in detail the gamut of non-destructive weld quality

monitoring procedures which have been developed over approximately the last two decades,

some of which are in current use.

Chapter 3.0 outlines the initial experimental work conducted in order to identify

appropriate welding conditions for Z I N C A N N E A L (zinc-iron rich) type metallic coated

steels prior to conducting vision system monitored, electrode tip life tests. The effects of

electrode tip geometry (tip diameter and included angle on truncated cone tips) and the

2
inclusion of welding current upslope into the welding schedules for this steel type and hot-

dipped galvanized, electrogalvanized, zinc-5% aluminium coated and aluminium-45% zinc

coated steels are also studied. Chapter 3.0 also describes in detail, equipment used in this

thesis, including the spot welding equipment and its data acquisition system, the Rank

Taylor Hobson F o r m Talysurf for profilometry measurements of weld surfaces and the

Adept vision sensing equipment.

Chapter 4.0 outlines the procedure adopted for monitoring of various parameters

during the course of electrode tip life testing on 0.8 m m Z I N C A N N E A L G 3 N ZF80.

Results are presented in both tabular and graphical form for carbon paper impression

measurements taken from the degrading electrode tips, profilometry measurements of

impression width left on the sheet surface by the electrodes, vision system measurements of

the impression width, X-ray and ultrasonic measurements of nugget diameter and peel and

tensile-shear testing on welds.

Chapter 5.0 discusses the trends that were evident from the measurements taken and

examines the relationships between each of these and those of the vision system. F r o m

these relationships, the optimum vision system welding window is defined, the boundaries

of which are linked to weld nugget size and tensile-shear strength.

Chapter 6.0 draws firm conclusions from the results of this thesis, while Chapter 7.0

offers directions for future work, including improvements to the experimental vision system

set-up to provide more appropriate illumination to the test piece and the incorporation of

microprocessor controls to establish a closed-loop feedback, weld-quality monitoring and

control system.

3
 CHAPTER 2.0

LITERATURE SURVEY

4
2.0 LITERATURE SURVEY

2.1 Resistance Spot Welding

Resistance spot welding belongs to a group of resistance welding processes including

projection welding and seam welding. T h e heat to form a weld via this process, is generated

by resistance to the flow of electrical current through the parts being joined. In the case of

resistance spot welding, a pair of electrodes held by the welding machine conduct the

electrical current to the workpiece under a force usually supplied by an air actuated cylinder

as shown in Figure 1. Since the electrodes are m a d e of a copper alloy and are mostly water

cooled, heat is quickly extracted away from the surfaces in direct contact with the electrodes,

thus avoiding the possibility of the electrodes being welded to the sheet surface. T h e

greatest resistance, /?/, to the flow of current takes place at the faying surface (sheet to sheet

interface) and is constrained by the electrode contact area. Heat is not as easily extracted

from this area and so a weld is formed.

Weld Temperature

Figure 1: T h eresistancespot welding process [8]

Rj = contact resistance of the faying surface

/?2 = contact resistance of electrode/material surfaces
R3 = electrical resistance of the material.

5
 While rapid developments are being m a d e in other welding and adhesive bonding

processes, it is the resistance spot welding technique that remains the most widely used sheet

metal joining technique. Its ease of operation, cost effectiveness and flexible nature in that it

can be integrated with automated and robotic facilities, makes it desirable, particularly for

high volume welding such as that found in the automotive and domestic/household appliance

manufacturing industries. In addition to the author's experiences in thisfield,it has been

noted that at the shop floor level this process is one of the most widely ill-used joining

techniques. This is influenced by the m a n y process variations available, the industrial

environment in which this process operates, the material surface quality (metallurgical

cleanliness, chemical composition) and characteristics (surface roughness, oiliness) of the

steel sheet, the type of the joint geometry and configuration, electrode type and geometry,

welding machine type and capabilities and operator performance. Each of these effects can

affect weld quality levels.

While m a n y of these can be controlled through good management of steel processing

lines and shop floor welding process, it is the operators' abuse of the procedure that is

difficult to control. This is exacerbated in cases where weld monitoring and control devices

are not in place. Proper maintenance of electrodes is a critical issue in the resistance spot

welding process and without suitable electrode tip redressing tools and standard operating

procedures in place, s o m e operators employ rough cut files and emery cloths to repair

damaged or worn electrode surfaces. This normally manifests itself as poor weld quality

since electrodes invariably become misaligned and their working faces irregular in shape, if

mistreated.

Also, there exists a mentality in high volume welding industries that resistance spot

welds, which show visual evidence of metal expulsion during welding, are good welds and

so welding currents are set exceptionally high. Expulsion is the term given to the spraying

of molten metal out from either the faying surface or the point of contact between the

electrode face and the sheet surface (Figure 2). This usually indicates that the welding

current setting is too high and can occur through the action of the electrode force and the

electromagnetic fields generated during welding. T h e author has noted a predominance of

6
expulsion in the automotive industry, in particular where both pedestal and hand held,

umbilical supported (roof mounted) spot welding machines are employed. While it can be

shown that at welding currents just beyond the expulsion point level, full strength welds

can be achieved [1], there seems to be little regard for the effect that expulsion has on

electrode life. It would be far better to control welding current so that it maintains a leve

below the expulsion point which would not only optimise weld quality but also maximise

electrode working life through reducing material loss from the electrode. As a result, down-

time in order to replace or redress damaged or worn electrodes would be minimal.

The use of excessive welding currents is not as predominant in the appliance industry

as there is usually a need for tighter control on the degree of weld indentation, certainly

where components are exposed to scrutiny of the consumer eye. Aesthetic appearance of

welds in this industry appears to be far more important than in the automotive arena where

the vast majority of welds are hidden from view.

Consumer demand for quality and reliable products has given rise to an expanding

range of metallic coated steels being developed, each with their own resistance spot

7
wcldability characteristics. A s can be seen from Table I, there is quite a broad range of

coated steel types available in the market-place, presenting the fabricator with even further

problems not normally encountered during spot welding of uncoated steels.

Table 1 — T y p e s of Metallic Coated Steels

Steel Type Coating

Hot Dipped Galvanized : zinc

: zinc-iron
: aluminium
: zinc-iron + iron-zinc electrogalvanized
flash coat
: zinc-55% aluminium
: zinc-5% aluminium
: lead-tin

Electrogalvanized : zinc
: zinc-iron
: zinc-nickel
: zinc-nickel + iron-zinc flash coat
: zinc-iron + iron-zinc flash coat
: zinc-nickel + iron phosphorus flash coat
tin
chrome
zinc-iron + iron-phosphorus flash coat

Others : electrogalvanized zinc-nickel + organic overlay

: metallic/organic film
(coil coated spot weldable primers)

It has been recognised by others and also from the authors o w n research, that the

various metallic coated steels can react in different ways to certain welding parameter

changes [2] creating some confusion for the sheet metal fabricator. Only an experienced

operator is capable of selecting optimum welding parameters for a particular steel type.

Studies have shown for example that electrode geometry [2], electrode chemistry [3,4] and

the introduction of weld current ramping techniques into the welding schedule [2,5] can have

dramatic effects on electrode life and while many thousands of spot welds can be produced

on uncoated steels without redressing or replacing the electrodes, as little as 250 welds may

only be possible on some metallic coated steels.

The resistance welding behavioural difference between uncoated and metallic coated

steels is best explained in the following way.

8
 A s stated previously, in any resistance welding operation (be it spot, projection or

scam welding), the formation of the weld is dependant on the heat generated due to the

electrical resistance of the materials being joined. The total heat generated (H) in joules is

given by the expression:

H = I2Rt (1)

where, / = welding current in amperes

R = sum of the electricalresistancesin the system in ohms. This includes

resistances of the electrodes, contact surfaces and metal substrates..

t = time for which current is allowed to flow (seconds or cycles).

In order to maintain heat for the welding process, any decrease in electrical resistance

requires an increase in welding current and/or time. The overall electrical resistance and

contact resistances during spot welding of uncoated steels are usually relatively high in

comparison to those experienced during welding of metallic coated steels. There is a need

then to use higher welding currents/times to produce weld sizes of similar diameter and

quality to those produced on uncoated steels of the same thickness as shown in Figure 3.

Expulsion
16

UJ
cn
U

UJIO
Uncoated " ^ ^
Minimum nugget diameter

J I I L J L
2 4 6 8 10 12 14 16 18 20
W E L D TIME (Cycles)

Figure 3: Typical weld lobes for 1.2 m m uncoated and galvanized steels [2]

9
 T h e heat generated at the electrode-sheet interface provides the driving force for

alloying to occur between the metallic coating of the sheet and the copper electrodes. This

alloying behaviour has two effects on weldability. Firstly, there is a tendency for alloy to

build up around the periphery of the electrode tip. This occurs when using either truncated

cone or radiused electrode profiles as shown in Figure 4. Because of this, as welding

progresses, the electrode tip diameter effectively increases leading to a decrease in current

density and subsequent production of smaller and smaller welds of inferior quality. T h e

radiused electrode would not normally be recommended for spot welding coated steels as

even the smallest amount of alloying rapidly changes the contact area of the electrode. These

are being used however, where partfit-upis a problem, such as where two curved surfaces

meet, and are usually used in conjunction with hand held spot welding guns which are

employed in a number of different orientations in a production line.

Included Angle

Diam.

TRUNCATED RADIUSED
Figure 4: Truncated cone and radiused spot welding electrode tip profiles

Secondly, direct alloying with the electrode face causes erosion or pitting of the face,

thus constantly changing the contact area between the electrode and the sheet surface. It has

been shown by the author that the wear rate of electrodes can be directly related to the

composition of the metallic coating [6]. While employing Resistance Welder Manufacturers

Association ( R W M A ) Class 2 ( C u - 0 . 8 % C r ) truncated cone electrodes to weld metallic

coated steels high in coating aluminium, the erosion rate is rapid. However, the erosion rate

is less rapid during welding of metallic coated products which arerichin iron. Alloy layers

10
built up on electrodes used to weld the latter product, tend to be more resistant to

delamination from the electrode surface. O n the other hand, pockets of aluminium in high

concentrations in built up alloy layers can be easily stripped away from the electrode surface

when it is pulled away from the sheet. This is attributed to the brittleness and relatively high

hardness of these pockets. A s the erosion process continues, the electrode working face

becomes larger in diameter. Without periodically increasing the welding current to

counteract the associated drop in current density, weld quality will naturally deteriorate and

size will decrease.

Monitoring of spot weld quality in the past, meant laborious, time consuming and

often very costly destructive testing of welds, most of which have tests proved to be

unreliable indicators of welding performance. Weld nugget diameter, measured via vernier

callipers after peel or h a m m e r and chisel tests are performed on selected welds, is generally

accepted as the most appropriate w a y of examining weld quality [7]. The peel test relies on

peeling a weld apart and determining the nugget size (sometimes referred to as a plug) which

is pulled from one side of the joint, as shown in Figure 5. Nugget sizes must be no less

than (in most cases) 8 0 % of the original electrode tip diameter employed in order to be

considered satisfactory welds. Thin gauge steel sheet however, during peel tests usually

pull with a tail remaining on the nugget as shown in Figure 5. This makes accurate

measurement of nugget size difficult since the callipers cannot easily reach the edge of the

nugget. Even if a weld under testing has not reached the m i n i m u m size requirement,

isolating the parameter at fault can be quite difficult. The h a m m e r and chisel test relies on

driving a chisel between two close welds as shown in Figure 6. The welds are considered to

be of good quality if nuggets do not pull out of the sheet under the action of the chisel.

M a n y still use these quality checking procedures, but there has been an ongoing search

for more cost effective and more stringent quality control procedures in recent times. A s a

result, there has been an upsurge in the development of resistance welding monitoring and

control devices. These, fundamentally provide routine checking of the basic welding

parameters and condition of the welding equipment and can also adjust selected process

input parameters including welding current, through feedback control loops. With the aid of

ll
these devices, electrode life can be optimised and weld quality assured after suitable welding

conditions have been established.

Figure 5: Peel test for determining weld nugget diameter and penetration [62]

Figure 6: The h a m m e r and chisel test [8]

12
2.2 Monitoring and Control Devices

The literature offers a number of definitions for weld process monitoring all of which

relate to acceptable weld quality control. Dickinson [8] defines the monitoring process as

"one which gives information about the quality of the product". Beatson [9] states that "any

simple monitor or correction system must look at some phenomenon or characteristic which

occurs or changes as a result of the start and growth of a weld and whose output bears a

definite relationship to the weld diameter, this usually being taken as a production criterion

of weld quality".

A control or correction system can be defined as a unit which monitors one or more

critical parameters during the weld cycle in an attempt to maintain weld quality under varying

external conditions by automatically processing these parameters and using the output to

automatically adjust the welding conditions.

Available resistance welding monitors do not solve welding problems, they simply

highlight them by providing a track record of every spot weld made. It is then up to the

operator to recognize any anomalies present and in the absence of any control devices,

manually adjust the process accordingly.

The characteristics of the ideal monitor and the exact parameter which gives the best

indication of weld quality, has been debated by many, however, most systems display the

magnitude of the monitored parameter on some form of analogue or digital measurement.

Rivet et al [10] shows the idealrelationshipbetween monitor count (the critical parameter,

indicated by meter deflection) and weld quality (nugget diameter) in Figure 7. H e states that

the relationship should be relatively linear, should be similar for variation of all those

parameters which could affect weld quality in production such as mains fluctuations,

electrode wear and current shunting, and should have as narrow a scatter band (shown as

line A B on Figure 7) as possible to enable prediction of weld size.

13
 (J
oc

I
Z
O

N U G G E T DIAMETER

Figure 7: Typical relationship between monitor count and nugget diameter [10]

The literature also indicates a number of process parameters which have been used in

monitoring of the spot welding process, most of which can be used in a closed loop

feedback control system [8]. They all provide a non-destructive means for assessing weld

quality and either totally eliminate, or at least reduce, the need for destructive testing of

welds. These parameters include:

a) electrical parameters

b) nugget temperature

c) infra-red radiography

d) weld expansion and contraction

e) acoustic emissions and

0 ultrasonic signals (reflection, attenuation and velocity).

Each technique is discussed below.

14
2.3 Electrical Parameters — General

Monitoring and control of electrical parameters has received the greatest amount of

attention and the devices manufactured for this purpose have been considered the most

successful [11], This is mainly attributed to the ease of attachment to the welding machine

of such devices to measure current, voltage and time. N o sophisticated sensing equipment is

required to examine these parameters. There are however, some reported disadvantages,

these being: a) their general confinement to uncoated mild steels, since there have been poor

correlations between weld size and voltage on metallic coated or uncoated materials, b)

degradation of electrodes can cause misleading results, and c) the vulnerability of voltage

clips due to their need to be in close proximity to the weld zone.

2.3.1 Welding current

AC spot welding machine controls do not have facilities to control the current directly.

T h e machine can only control the amplitude and shape of the voltage produced at the

secondary transformer [12]. This voltage applied across the transformer's total secondary

impedance (comprising of resistance and inductance) including the machine throat and

workpiece, causes the current to flow. Therefore, the welding machine controls the current

indirectly by the transformers secondary impedance.

Current can be adjusted on most spot welders through a primary tapping switch which

adjusts the current level by changing the ratio of primary to secondary turns. Another

current control is the "phase shift" or "heat" setting dial. The heat value (or raw value) of the

current can be altered through this control.

Because the transformer of spot welding machines is connected directly to the power

line by the control circuits, current and therefore weld quality can be affected by voltage

fluctuations. Without suitable monitoring equipment, this cannot be detected.

Welding current can also be affected through "current upslope" and "downslope"

controls which are available on more modern spot welding machines. These can be used to

15
alter the phase shift automatically from s o m e small value of current to the preset phase shift

value over a nominated period of time (upslope). In summary, welding current is affected

by line voltage fluctuations, phase shift, transformer tapping and slope controls.

Appleby [13] states that the accurate measurement of rms value of weld current is one

of the most difficult welding parameters to measure, but at the same time it is one of the most

important in determining weld quality and reliability.

The literature details two procedures for measurement of spot welding current, through

the use of a toroidal coil [13] or through Hall-effect probes [14].

The toroidal coil placed around the primary or secondary of the spot welding machine

is the more commonly used method. This measures voltages induced by the changing

magnetic fields surrounding the weld current carrying conductors of the welder and is only

useful for measuring A C currents. T h e voltage signal can be electronically integrated to

provide an indirect measurement of the welding current as indicated by Faraday's L a w :

V = dl/dt A cos 6 (2)

where, V is the measured voltage, dl/dt is the change in current with respect to time, A is the

area of the toroidal loop and 9 is the angle of the loop to the magneticfield.The output is

displayed as a peak current on a meter calibrated in rms units. A portable, battery operated

electronic weld current meter has been developed by British Federal W o r k s [14]. Studies

have shown, however, that with this method, current values can change with position and

orientation of the toroid. Errors as large as 5 % have been noted [14]. A similar machine is

also available from Hirst Electronic Industries Limited [15].

T h e Hall-effect probe is recognised as an accurate but expensive technique for

measuring current. This technique measures the voltage across a semiconductor due to

surrounding magnetic fields. It is therefore an indirect method for measuring welding

current. Equation (2) also applies to the Hall-effect probe but the value of A is different.

The technique is also sensitive to changes in orientation and position of the probe.

16
2.3.2 Voltage

Voltage measurements can be used to measure weld quality. These can be obtained by
attaching voltage probes to either top or bottom surfaces of the materials being welded or to
both spot welding electrodes (upper and lower). Voltage traces as shown in Figure 8 can be
obtained [8]. M a x i m u m voltage has been related to the growth of the spot weld nugget in
plain carbon steels [16,17] and ultimately, nugget tensile shear strength (tensile strength
increases with m a x i m u m voltage). Similarly, nugget diameter can be related directly to the
voltage drop at expulsion point, occurring after the peak voltage [18]. This system is
reported to be reliable only when welding mild steel, stainless steel, titanium and fully iron-
alloyed galvanized steels, normally referred to as galvanneal. Figure 9 shows the typical
relationship between voltage drop and nugget diameter.

Figure 8: Schematic of dynamic electrical parameters during spot welding of mild steel [8]

17
 REDUCTION IN ELECTRODE VOLTAGE, V%
Figure 9: T h erelationshipbetween weld diameter and voltage drop duringresistancespot
welding [8]
O interface failure
• nugget pull-out
• expulsion

The system is not suited for monitoring spot welding when using free zinc coated (hot-

dipped galvanized) steels [18,19] and aluminium [18], since there is no drop in voltage up to

the expulsion point for these materials. T h e voltage drop system could be employed to

obtain full size, expulsion welds infreezinc coated steels, but there is the disadvantage of

accelerated electrode wear due to the fact that electrode sticking to the sheet surface is more

prominent at the higher current levels required to produce the weld expulsion condition.

Dickinson [3] has also reported that a relationship can be found between nugget

diameter and heat input into the weld (area under the voltage-time curve shown in Figure 8).

This relationship depicted schematically in Figure 10 can be used to pin-point a weld

termination point, ie welding can be stopped prior to the nugget reaching the m i n i m u m

nugget diameter. Ganbowski et al [20] have suggested that this method could be used

during spot welding of metallic coated steels.

 N U G G E T DIAMETER

Figure 10: The relationship between nugget diameter and the area under the voltage curve
produced during resistance spot welding [8]

2.3.3 Time

Welding time in Equation (1) is simply controlled by the number of cycles of 50 Hz

(Australia) or 60 H z current allowed to flow through the materials being welded.

2.3.4 Resistance

There are a number of variables controlling resistance heating in spot welding. The

conductivity of the material, surface condition of the material, material thickness and

electrode condition and diameter are the most important. Bourtorwick [21] points out that

electrode force, electrode alignment with respect to each other or with respect to the

workpiece itself, and poorly made electrical connections can also change resistance and

therefore affect weld quality.

T h e total resistance of the workpiece is governed by the material, electrical

conductivity and thickness, and also the resistance at the contact surfaces between the

electrode and sheet and the faying surfaces. The electrode diameter determines the area

through which the currentflowsand hence the resistance that the workpiece presents.

19
 W h e n the electrodes are brought into contact with the workpiece and current is applied,

there is an initial high contact resistance at the electrode/sheet interface. This is attributed to

the presence of high spots on the electrode and/or sheet surface. High power (PR) is

generated at these points and they are brought rapidly to a temperature at which they can

collapse. T h e process continues, with the electrodes then, bedding into the sheet surface and

spreading the heat more evenly over the surfaces of the sheet. If either electrode, or sheet

surface, is badly contaminated or oxidized, initial surface contact resistance will be abnormal

and this will affect the bedding in process and subsequent weld quality.

2.3.5 Monitors for welding voltage and current

The Welding Institute (TWI) in UK has developed a simple control unit for measuring

and controlling voltage during resistance spot welding [22]. The voltage signal is received

from two sensing leads attached to the arms of the welding machine and is fed through a

filter to minimise induction and noise pick-up. The signal is then passed on to a voltage

drop comparator so that w h e n the voltage drop across the electrode decreases from the

m i n i m u m level by an amount greater than the preset level, the welding current is terminated.

Special controls have been inbuilt so that sudden drops in mains voltage do not prematurely

cease the welding current. A weld timer also functioning in parallel with the control system

ensures that welding can be terminated after a preset time if the voltage control does not

operate. T h e system is said to be suitable for both pedestal and portable gun welding

machines.

Johnson [23] states that several monitor and automatic correction systems have been

commercially available for resistance, spot, seam and projection welding but the most

c o m m o n in use are those which control current and voltage automatically. Automatic voltage

regulators can also be employed to compensate for mains voltage fluctuations as is automatic

current regulators for cases where impedance in the throat of the welding machine changes

during welding. O n e U K manufacturer (not identified) has built a unit that has combined

voltage and current regulators which can automatically adjust phase-shift. M a x i m u m

20
response time for voltage or current control to within 5 % of the final value is stated as 3

cycles (60 Hz).

Johnson [23] also mentions that some commercially available monitors work on the

system of indicating weld quality by integrating either the voltage drop above a m i n i m u m

threshold value across the welding electrodes during welding, or the welding current.

Accept and reject levels are established firstly through welding trials and performing peel

tests to determine nugget diameter. The output of the monitor is displayed on a meter which

operates accept orrejectindications.

The Digimetrics N D T - 7 0 0 direct A C / D C weld monitor developed in the United States

of America [24] is a state of the art micro-processor based system for non-destructive testing

of resistance spot welds. This unit detects changes in material composition, electrode wear

or pressure, current,resistanceand weld cycles. Only two pick-up cables are required, one

for voltage (connected across the electrode holder) and one for current (a current coil placed

around one of the welding cables). The operator needs to conduct some welding trials in

order to determine the materials optimum welding schedule ie. m i n i m u m and m a x i m u m

current levels, then the monitor is adjusted to set "peak amps", "amp seconds" and "volt

seconds". A poor weld is indicated by an activated L E D and this suggests that the unit

provides only monitoring and not feedback control.

A number of similar machines have also been developed as a result of co-operation

between The Welding Institute and the A R O Machinery C o m p a n y Limited ( U K ) under the

"Spotrite" trade n a m e [25,26]. These give both visual and audio alarms when welds are

outside quality control preset limits.

Variations in voltage and current among other changing parameters are also accounted

for with the British Federal, Weldmatic M P 1 6 P unit, incorporating microprocessor

technology [27].

21
2.3.6 Resistance measurement and monitoring of dynamic resistance

Resistivity can be used to detect poor weld quality. The output from resistance

measurement, usually digital, is considered easy to interpret and the instrumentation easy to

apply. Resistivity is computed by injecting a current into the weld area, measuring the

resulting voltage drop and converting that drop to a resistivity reading. It has also been

shown that increasing weld shear strength can be directly related to a decrease in resistivity
in the weld fusion zone [28].

The only problem noted with this test is that cold welds m a y be detected as good

welds. This can be overcome by impacting the weld prior to testing in order to break the

weld and thus cause the resistivity measurement to be high.

Monitoring of weld quality through the aid of dynamic resistance curves is a well

accepted method and isfrequentlymentioned in the literature. Dynamic resistance is the term

given to the varying electrical resistance of the weldment during a spot weld [29] and refers
to the quotient:

Vw/Iw = Rw Q)

where, Vw and Iw represent instantaneous values of weld voltage and welding current,

respectively [30]. Andrews [31] shows schematically the simplest form of a dynamic

resistance monitor in Figure 11. T h e voltage output from such a device which is a

representation of the dynamic resistance of the weld, is fed into a recorder to obtain a trace

which can be compared to a trace for a sound weld. If the recorded trace falls outside a

prescribed limit, the weld is rejected and the parameters adjusted until sound welds are once
again produced.

22
 Weld Rectification
Amps Amplifier

oCM>C>C>1
Isolation IntegratlonVarlable
Torold Gain
Divider Low Pass
Filter
"Smp3

Tip Volts

)—uy
Isolation Variable Gain Rectifier

Figure 11: Andrews' model of the simplest form of a dynamic resistance monitor [31]

Many dynamic resistance curves or traces for different materials are shown throughout

the literature, but they are all basically of the configuration shown in Figure 12, which is

said to be "typical" of a single pulse weld on an uncoated mild steel [32]. A brief description

of the various curves and their stages follows.

Figure 12: A typical dynamic resistance curve for a single pulse resistance spot weld on an
uncoated mild steel [32]

23
Stage 1

Resistance is high due to the surface roughness of the electrodes and material being

welded. The resistance is lowered rapidly as the electrodes bed in to the sheet surface and

the contact area increases. The resistance reaches a minimum when the contact is greatest.

This is shown by the valley in Figure 12.

Stage 2

A sresistivityincreases through heating of the workpieces, theresistancecurverisesto

a peak. At the peak, the faying surface is nearing the molten state.

Stage 3

O n reaching the peak of the curve, the electrodes move into the sheet surface and the

current path becomes shortened. The weld nugget begins to grow such that the path over

which the current flows is also broader. Subsequently, the resistance curve falls off. If

expulsion has taken place, then the resistance curve falls off quite sharply and this can be
easily recognised.

T o control weld quality, the operator can either stop the welding procedure w h e n a

particular drop in resistance is seen relative to the resistance at the peak of the curve

(uncoated steels), or stopping at a time beyond the valley and the peak of the curve. The

latter method is more commonly employed with galvanized steels.

There is no information in the literature to suggest that this method is being used in a

closed loop feedback control m o d e during spot welding, but it is being utilised widely as a

quality monitoring system.

2.4 Nugget Temperature

During the growth of a spot weld nugget, the surrounding sheet metal is heated.

Various methods have been employed to relate the temperature of the heated surfaces to the

temperature at the centre of the nugget, thus providing an indication of whether or not full

24
growth has occurred. Thermocouple attachment to either the sheet surface in close proximity

to the weld [33], or to one of the electrodes, were the earliest methods used. Data was

usually presented as a graph of temperature versus time.

These techniques were not acceptable to production since considerable time was

necessary to install the thermocouple correctly. Additionally, electrode temperature is

dependent upon the current level and welding time, each of which cause variations in

electrode temperature gradients. This could make control of spot weld quality difficult.

2.5 Infra-Red Radiography

Measurement of sheet surface temperature radiation to indicate weld quality has also

been considered for m a n y years. Continuous monitoring of the intensity of infra-red

radiation emitted from the sheet surface immediately adjacent to the electrode has

theoretically been related to the weld nugget size and the tensile shear strength [9]. The

advantages of this technique over the thermocouple attachment method are twofold. Firstly,

there is no need for the detectors to be in direct contact with the heat source. S o m e detectors

can be mounted permanently to the sides of a welding machine, well out of the operators

way. Secondly, the response time of a suitable detector element can be very short and the

process time presents no limitations.

Infra-red emission from the sheet surface can be sensed through either afibreoptic

cable [9,32] mounted in the welding electrode and transmitting radiation to an external

detector as shown in Figure 13, or through infra-red detectors focused at spots on the sheet

surface immediately adjacent to the electrode [9,10]. Sensing through thefibreoptic method

is considered impractical for most applications because of the interference to electrode

construction and because electrode wear m a y produce misleading results. The second

method is deemed more appropriate. Researchers have employed both one and two detector

units positioned at a 45 degree angle to the surface of the work [34]. W h e n two detectors

are used, these are placed on opposing sides of one of the electrodes as shown in Figure 14.

While this method is more practical, the positioning of delicate sensors in close proximity to

25
the weld zone could m a k e it difficult to use, particularly w h e n m a n y different types of

components of various shapes need to be welded.

Material

Electrode

LIGHT
GUIDE

Figure 13: Positioning of an infra-red emission detector shown as afibreoptic cable 'light
guide' for monitoring of spot weld quality. T h e upper electrode is not shown [9].

yv /.' / "~\ .\< z-LIGHT

LIGHTPIPE
PIPE
SILICONE FILTER
PHOTOCELL

Figure 14: The positioning of dual infra-red emission detectors outside of the welding
electrodes and at 45* to the surface of the workpiece [8]

26
 Typical results of the infra-red monitoring system arc presented in Figure 15. The
amplitude of the detector trace from small nuggets is low. These nuggets exhibit low shear
loads. For acceptable nuggets, the amplitude of the trace is intermediate while a large
amplitude is observed for welds which expel.
Regardless of the technique used, infra-red monitoring suffers from the disadvantage
that dirt and fumes m a y cause spurious results and that oxides on the sheet surface may
cause some variations in surface emissivity. For this, it would be even less appropriate to
employ this technique during spot welding of zinc coated steels because of fumes generated
through the volatilization of the zinc. The inherent problems associated with temperature
monitoring have resulted in this technique not being widely accepted.

IK \A

0 1 2 3 4 5 6 7 8 9 10
TRACE AMPLITUDE RECORDED FROM INFRARED DETECTOR

Figure 15: Typical results from an infra-red monitoring system showing the relationship
between trace amplitude and weld shear strength [8]

2.6 W e l d Expansion and Contraction

In the 1950s, researchers in the U K , U S S R and U S A indicated the possibility of

measuring the electrode movement as a result of resistance heating causing metal expansion
and relating this back to weld nugget dimensions [10]. There are two opposing forces

27
during spot welding. Firstly, there is the natural inward movement of the electrodes due to

the applied force from the welding machine and the fact that as the material becomes molten,

the electrodes are more free to m o v e towards each other. Opposing this movement is the

force generated on the electrodes through thermal expansion of the weld and surrounding

material [35]. The degree of separation of the electrodes can be related to the volume of

molten metal and can therefore be used as an indication of nugget growth. Electrode

separation is monitored by measuring the piston rod travel of the air cylinder in the welding

machine via displacement transducers.

Figure 16 shows the characteristic curve of electrode movement during the formation

of a spot weld as observed by Stiebel et al [36]. A s well as the welding force exerted by the

machine mechanics and thermal expansion of the steel work pieces causing electrode

movement, Stiebel et al also points out that there is a contributing factor from thermal

expansion of the copper electrodes. A brief description of the events that take place in regard

to electrode movement can be described with reference to Figure 16.

Figure 16: A characteristic curve of electrode movement (distance designated by's') during
a single pulse spot weld as observed by Stiebel et al [36]

T h e heat generated, initially at the faying surface w h e n the current is applied, causes thermal

expansion of the workpieces and electrodes. The electrodes are forced to m o v e apart and

this is shown by therisingdisplacement curve (point 2). A series of thermal events then

takes place resulting finally in a slowing d o w n of the outward separation of the electrodes

(point 3), until a force equilibrium state is reached. The electrodes at this point are stationary

(point 4) and subsequently they begin to m o v e together (point 5) as the temperature of the

28
fusion zone increases. O n terminating the welding current, the fusion zone cools and

therefore contracts with the electrodes following this movement under the action of the

welding force (point 6). The final separation of the electrodes is less than the initial

separation because of the indentation into the sheet surfaces. Stiebel et al concluded from

this type of analysis on a large number of spot welds, that good quality welds are assured if

the electrodes m o v e into the sheet surfaces after the point of force equilibrium.

Johnson [23] has suggested that a good weld will expand at an optimum rate and to an

optimum level. W e l d current control can then be adjusted to track these optimum levels.

Welds m a d e with high heat input are associated with rapid expansion and weld expulsion

causing rapid electrode closing. Welds made with too low a heat input have slow expansion

and do not attain full electrode separation. Johnson et al [37] have based a monitoring

device on this principle. Monitoring and control devices based on thermal expansion

systems have also been established by Janota [38,39] and Taylor et al [40]. From these

works it is evident that there are some drawbacks with the use of thermal expansion as a

weld quality monitoring device. The method can normally only be used in conjunction with

pedestal (free-standing) welders which are quiterigidand which have lowfrictioncylinders.

It is not adaptable to gun welding machines and is not suitable on welds close to edges

where partfit-upis a problem. In this case, excessive weld expulsion could take place

resulting in misleading signals in the load control systems.

2.7 Acoustic Emission

Acoustic emission is said to arise from the transient elastic energy that is

spontaneously released when materials undergo deformation, fracture, or both [35]. There

are a number of stages in the formation of a spot weld during which acoustic signals are

emitted all of which can be detected by a piezo electric transducer mounted to the electrode

holders of the welding machine. Signals detected include those from initial electrode contact

and at electrode separation, expansion, recrystallization, solid state transformations, plastic

deformation, melting, vaporization, solidification, and on occasions, weld cracking. A

29
typical acoustic emission trace is shown schematically in Figure 17. Because the amplitude

is highest at the expulsion point, it has been possible to determine the weld at the onset of

expulsion through feedback current control. Expulsion represents the point of m a x i m u m

nugget growth. Havens [41] points out that electrode mushrooming (plastic deformation of

the tip face) is reduced through this method because electrode heating is lowered. This

would only lead to improved electrode life. The method has been successfully used on both

metallic coated and uncoated materials; however, a major drawback of the procedure lies in

the fact that signal processing can become cumbersome when multiple spot welds are made.

W h e n using this procedure on metallic coated sheet steels, the signal emitted when the zinc

coating boils has been employed to control the welding process [42],

Figure 17: Schematic output data of acoustic emission during resistance spot welding [8]

Acoustic emission adaptive control is finding use in the automotive industry with

General Motors ( G M ) in the United States leading the w a y [43]. G M ' s Wilmington

Delaware plant is using an acoustic feedback control unit developed by the Trodyne

Corporation of Teterboro, N e w Jersey to control resistance spot welds on galvanized steel,

valance panels for their Chevette vehicle. The system is being used along with the resistance

drop method described previously to improve the overall weld quality.

30
2.8 Ultrasonic Signals

Ultrasonics is high frequency sound above the range of hearing (20-25 k H z per

second). Monitoring ultrasonic signals emitted during the formation of a weld is another

method in use to examine weld quality. Basically, the weld pool behaviour, comprising of

both melting and solidification, affects the amplitude of the ultrasonic signals emitted from

the weld, ie the ultrasonic sensor placed in the vicinity of the weld pool can act as a

temperature probe. It is possible then to relate indirectly the changes of the ultrasonic signal

to the time of cooling of the weld nugget to some predefined temperature [35].

There are a number of ultrasonic techniques mentioned in the literature for spot weld

monitoring. The most c o m m o n procedure involves mounting a transmitter crystal on the

water cooled opening of one of the spot welding electrodes and placing a receiver on the

other [9]. Ultrasonic signals are then passed through the weld zone as welding takes place.

Nugget growth can be monitored through observation of the change in signal height during

the weld time. British Aerospace in conjunction with the National N D T Centre ( A E A ) , U K ,

has developed a technique based on this principle for monitoring spot weld quality in

titanium steels [44]. Through changing weld pitch (distance between welds), electrode tip

diameter, plate surface condition and plate thickness, correlations between parameter

variations and ultrasonic signals were studied. Ultrasonics easily distinguished nugget size

and electrode tip diameter. It w a s reported that these correlations can be used as a basis for

producing a feed-back system.

Hain [45] has developed a method which generates short bursts of high energy which

are transmitted via a probe tip through the weld zone. Echoed signals of decreased intensity

reflected off the front and back surfaces of the weld are received by the probe and are

displayed on a cathode ray tube screen as vertical spikes. The thickness of the weld is

indicated by the separation between two successive spikes, as this distance represents the

time for the signal to travel between thefrontand back weld surfaces.

31
 Allen [46] reports that Budd Canada Incorporated have developed a technique k n o w n

as the Water C o l u m n Pulse-Echo Technique, which has been used in various applications

including thickness measurement, flaw detection, materials differentiation and spot weld

quality testing. T h e technique is unique in that a contained column of pressurized water is

sustained between the crystal of the transducer mounted within one electrode and the object

being tested rather than the usual method of placing gel, oil or water directly onto the object

surface. T h e transducer is both the transmitter and receiver with the returned or echoed

signals entering a flaw detector unit. From there the signal is amplified, rectified and sent to

an oscilloscope display. Interpretation of data is similar to that previously described.

Ultrasonic monitoring does have some problems, most of which need to be overcome

before the technique is fully accepted. These include:

i) the need for good surfaces to ensure adequate coupling,

ii) lack of signal resolution from thin material welds,

iii) the need to be able to interpret the oscillographic traces in real-time,

iv) the inability of thin and quite often fragile transducers to cope with the rigours of

high volume production welding and

v) extraneous signals sometimes emitted from zinc surfaces.

2.9 New Monitoring Approaches

The work in this thesis evaluates further non-destructive, weld quality monitoring

techniques not yet examined. These include profilometry (tactile sensing) and machine

vision (remote sensing). T h e basic approach taken to the study was to collect resistance spot

weld samples during electrode tip life testing on a c o m m o n l y used metallic coated steel type

and then to establish the relationships existing between weld nugget size, measured through

various means and the diameter of the weld impression m a d e on the sheet surface, measured

via the profilometer and vision system.

32
 T h e objective was to establish the feasibility of utilising the vision system only for on-
line weld quality monitoring and if possible, at a later stage incorporate this into a feedback
control loop to maintain quality in almost real-time. The profilometer was used essentially as
a check on measurements obtained through the vision equipment. Profilometry would be far
too slow for consideration as a monitoring system and would not be readily accepted as an
on-line monitoring system.

33
 CHAPTER 3.0

EXPERIMENTAL PROCEDURES
AND
PRELIMINARY RESULTS

34
3.0 EXPERIMENTAL PROCEDURES A N D PRELIMINARY RESULTS

3.1 Materials Used in This Study

3.1.1 Metallic coated sheet steel

The steel sheet used in this study was BHP Steel — Sheet and Coil Products

Division's Z I N C A N N E A L ® * G 3 N Z F 8 0 of 0.8 m m base thickness. This is a galvanneal

type product processed along the conventional hot-dipped galvanized steel route but the zinc

coating on the strip after emerging from the zinc bath, is kept molten for a short time by a

furnace placed around the strip. This gives the coating on the strip time to completely alloy

with the iron in the substrate. The strip is then cooled in a similar way to normal zinc

spangled finished materials. Unlike the traditional hot-dipped spangle zinc coating, which

generally has a relatively rough surface and can be a problem for producing high quality

paint finishes, the Z I N C A N N E A L surface is considerably smoother and spangle free,

facilitating an excellent surface for paint adhesion and finish.

D u e to its iron rich coating, this type of product is generally characterised by excellent

resistance spot weldability and is employed widely in the Australian automotive industry as

inner door and floor panels, wheel housings,firewalls,side sill outers and other areas of

application.

Within the product name, "G" represents a galvanized steel product, "3" designates the

material's level of formability (in this case a high level), "N" a non-ageing grade, "ZF"

represents an alloyed coating of zinc and iron and "80" the total coating mass in grams per

square metre (nominally 40 grams per square metre each side).

The composition and properties of this steel, obtained through submission of samples

to chemical and mechanical test laboratories at B H P Steel — Slab and Plate Products

Division and Sheet and Coil Products Division, are given in Table 2. The product data sheet

for this steel is shown in Appendix 6.

* ZINCANNEAL is a registered trade mark of BHP Steel (JLA) Ply Ltd

35
 Tabic 2 — Composition and Propcrtlei of Z I N C A N N E A L G3N ZF80 Sttcl

Substrate Chemical Contposilion_£%i'

C p Mn Si s Ni Cr Mo Cu sol Sn Nb Ti V total Ca N O
Al B
.07 .017 .17 .010 .013 .022 .015 .005 .012 .054 .006 .008 .086 .003 .0003 .0005 .0054 .0055

" Couting Composition (%)

Al Pb Fc Zn
Top Surface 0.16 0.05 9.33 90.46
Bottom Surface 016 0.06 9.75 90.00
(Zn obtained by difference)
Average Coating Mass
(tJmh

Top Surface 43
Bottom Surface 43

Tensile Properties
Test Direction Hardness PS TS UE TE
(HR30T) (MPa) (MPa) (%)
longitudinal 42 159 322 22.9 39
transverse 42 170 325 20.0 37
m

3.1.2 Electrode selection

While various resistance spot welding electrode compositions are employed

throughout the world, eg Cu/Cr, Cu/Zr, Cu/Cr/Zr and AI2O3 dispersion strengthened, the

Resistance Welder Manufacturers Association Class II (Cu/0.8%-1.0%Cr) electrode

composition was used in this study due to its commonality throughout Australian sheet metal

fabrication industries. This alloy is also referred to as J47 alloy, the technical data sheets of

which are shown in Appendix 7 and Appendix 8. Refer only to data on J47 rod.

For a prolonged electrode life, a truncated cone profile was chosen. However, due to

conflicting arguments presented in the literature, it was necessary to examine the effects of

various electrode geometries for resistance spot welding on the Z I N C A N N E A L product.

Optimum conditions were then applied during electrodetiplife testing during which specific

data and weld samples were collected for submission to the profilometer and vision systems.

Further to examining the Z I N C A N N E A L product,freezinc coated, hot-dipped galvanized

steel, electrogalvanized steel, zinc-5% aluminium coated (Galfan) and aluminium-45% zinc

coated (sometimesreferredto as zinc-55% aluminium coated) steel were also examined.

36
3.2 Preliminary Experimental W o r k

3.2.1 Identification of optimum weld parameters

There has been some argument presented in the literature over the effects of weld

current upsloping, electrode tip diameter and electrode tip angle (to be referred to as included

angle), on the resistance spot weldability of metallic coated steels. It is apparent that certain

claims have been made after trials have been conducted on one material type only and that the

observed effects have been assumed to hold true for other metallic coated steel types. In

order to examine these effects, a short study involving tip life testing was conducted on

another galvanneal product, namely Z I N C A N N E A L G 2 ZF100 of 0.6 m m base thickness,

also a B H P Steel product. The actual composition and properties of this are shown in Table

3, while the product data sheet is shown in Appendix 9. It was possible to use this material

for this trial since trends only were being identified, eg does decreasing the tip included

angle improve electrode life? From theresults,some insight into selection of these more

critical parameters for welding galvanneal type steels would be possible and these could be

employed during weld trials on the Z I N C A N N E A L G 3 N Z F 8 0 material, more the subject of

this thesis. Material properties of the hot-dipped galvanized, electrogalvanized, zinc-5%

aluminium coated and aluminium-45% zinc coated steels also of 0.60 m m base thickness are

presented in Tables 4, 5,6 and 7 respectively.

37
 Tabic 3 — Composition and Propcrtlei of Z I N C A N N E A L G2 Z F 1 0 0 Steel

Substrate Chemical Composition (*)

C P Mn Si S Nl Cr Mo Cu sol Sn Nb Ti V total Ca N O
Al B

.055 .016 .21 .010 .007 .025 .015 .003 .013 .041 oo: <.005 <.003 <003 <.0003 < 0 0 0 5 0042 .0027

Coatina ComDosition <%)

Al Pb Fe Zn

Top Surface 0.16 0.05 9.35 90.44

Bottom Surface 0.16 0.06 9.35 90.44
(Zn obtained bv difference)
Average Coating Mass
(f/rnh

Top Surface 50
Bottom Surface 60

Tensile ProDerties
Test Direction Hardness LYS TS YE UE TE
(HR30T) (MPa) (MPa) (%) TO (%)

longitudinal 62 330 410 2.8 18.5 32

transverse 61 350 410 4.3 17.6 30

Table 4 — Composition and Properties of Z I N C F O R M * G 3 0 0 A06 Z275

(free zinc-coated, hot-dipped galvanized siteel)

Substrate Chemical Composition (%)

C P Mn Si S Ni Cr Mo Cu sol Sn Nb Ti V total a N 0
Al B

.050 .021 .21 .015 .006 .020 .025 .003 .012 .046 .oo: .005 .003 .005 <.003 <.0005 .0024 .0030

Coatine Composition (%)

Al Pb Fe Zn

Top Surface 0.53 0.215 0.31 98.90

Bottom Surface 0.53 0.2<5 0.31 98.90
fZnotMost bv difference)
Average Coating Mass
(R/m2)

Top Surface 165

Bottom Surface 160

TensiltPjoce,rues
Test Direction Hardness LYS TS YE UE TE
(HR30T) (MPa) (MPa) (%) (%) (%)

longitudinal 60 378 404 11.2 21.9 33

transverse 61 378 405 12.2 20.5 31
 Tabic 5 — Composition and Properties of ZINCSKAL* CA3SN-E FZ10
(electrogalvanized steel)

Substrate Chemical Composition (%)

C P Mn Si S Ni Cr Mo Cu sol Sn Nb Tl V total Ca N O
Al B

.070 .016 .23 .015 .015 .017 .013 .004 .014 .038 .002 .005 .003 <.003 <.0003 < 0 0 0 5 .0019 .0034

Coatina Composition (%)

Al Pb Fe

Top Surface <0.01 <0.0l 0.23

Bottom Surface <0.01 <0.01 0.28
(Zn obtained bv difference)
Average Coating Mass
(a/m2)

Top Surface 5
Bottom Surface 6

Tensile Properties
Test Direction Hardness PS TS UE TE
(HR30T) (MPa) (MPa) (%) (%)

longitudinal 48 195 340 22.8 39

transverse 43 190 335 22.7 39

Table 6 — Composition and Properties of Zinc-5% Aluminium-Coated Steel

(Galfan)

Substrate Chemical Composition (%)

C Ni Mo Cu sol Sn Nb Ti V total Ca N
P Mn Si S a Al B
O

.050 .010 .31 .015 .010 .014 .026 .007 .019 .062 .OW <.005 <.002 <.002 <.0003 <.0005 .0045 .0039

CoatinE Cornposition (%)

Al Fe Zn

Top Surface 5.9 0.1 93.92

Bottom Surface 5.9 0.1 i 93.92
fZnot tained bv difference)
Average Coating Mass
(g/m2)

Top Surface 147

Bottom Surface 128

Tensile Prooerties
Test Direction Hardness LYS TS YE UE TE
(HR30T) (MPa) (MPa) (%) TO TO
longitudinal 55 355 399 6.3 19.9 32
transverse 56 389 406 14.4 18.8 33

* ZINCFORM, ZINCSEAL and ZINCALUME areregisteredtrade marks of BHP Steel (JLA) Pty Ltd

39
 Table 7 — Composition and Properties of ZINCAI.UME G300 AZI50
(aluminium-45% ilnc-coatcd s(ccl)

Substrate Chemical Composition (%)

C P Mn Si S Ni Cr Mo Cu sol Sn Nb Ti V total Ca N O
Al B

.055 Oil .23 .015 .014 .026 .022 .004 .021 .054 .003 .007 .003 <O02 < 0 0 0 3 <.O0O5 ,0025 .0027
I Coating Composition {%)
1 Al Fe Si Zn

Top Surface 52.8 1.37 0.47 45.30

Bottom Surface 52.8 1.37 0.47 45.30
(Zn obtained by difference)
Average Coating Mass
(a/m 2 )

Top Surface 75
Bottom Surface 75

Tensile Prope rties

Test Direction Hardness LYS TS YE UE TE
(HR30T) (MPa) (MPa) (%) TO TO
longitudinal 59 358 415 5.1 17.8 27
transverse 58 391 415 12.3 18.1 27

Background information on the effect of current upslope

Upslope control is the gradual increase of the weld current magnitude during the

initial period of current flow in each weld as shown schematically in Figure 18. The rate of

current change during this initial period and the time interval involved are readily adjustable

on control panels of spot welders equipped with upsloping facilities. This technique assists

in melting and displacing the metallic coating, thus enabling the spot welding electrodes to

gradually seat before the m a x i m u m welding current is reached. This is believed to minimise

the heating between the electrode and workpiece thereby retarding alloying diffusion rates,

which are heat dependent, between the metallic coating and the copper in the electrodes.

Williams [47] and Freytag [48] have concluded, from laboratory studies on hot-dipped

galvanized steel, that there was no advantage in using the current upsloping technique. They

argue that zinc coated steel possesses a natural inbuilt upslope condition resulting from the

longer time to form a weld nugget because of the need to remove zinc from the weld zone.

T h e addition of current upslope would only result in inefficient heating and the necessity for

m u c h longer weld times.

40
UJ
oc
WELD CURRENT
DC
O TEMPER CURRENT,

\mm -A*/ —1^ f — ^t—[^ I tm\m « fj* *—»

UPSLOPE DOWNSLOPE TEMPER H O L D TIME

SQUEEZE TIME WELD TIME OFF TIME

(QUENCH)

Figure 18: Schematic of current input into a resistance spot weld [8]

Similarly, tests conducted by Schueler [49] on hot-dipped galvanized steel have

indicated that current upsloping would not improve electrode life since he observed that the

electrode to electrode resistance is lower during the early part of the weld heating period.

This would indicate that higher current is required before zinc can be removed from the weld

faying surface through volatilization. In contrast to this, Gedeon et al [50] showed that,

w h e n using truncated cone electrodes, upsloping (and downsloping) widen the weld lobes

of hot-dipped galvanized steels which have a free zinc layer. Johnson [51] also observed

that upsloping is beneficial during spot welding of galvanized plain carbon steel sheet in that,

not only tip life is extended, but welds also became more consistent. There is no

information in the literature on the effect of current upslope on the resistance spot weldability

of Zn-Fe alloy, metallic coated steels such as Z I N C A N N E A L .

Background information on the effect of electrode tip diameter

Selection of the electrode tip diameter (refer to Figure 4) depends upon the weld size

desired and the resulting strength of the weld, on the sheet thickness and in many cases

41
design of the component. In the latter case, tip diameters (and shape) m a y be constrained to

fit with m i n i m u m component flange widths for example.

Gould [52] demonstrated that w h e n resistance spot welding hot-dipped galvanized

steel, increasing the electrodetipdiameter results in a decrease in electrode tip life. O n the

other hand, Nealon and Lake [53] have demonstrated that an increase in tip diameter extends

the life of electrodes used to weld zinc-55% aluminium coated steel. It is possible then that

the various steel types react in different ways to changes in tip diameter. The usual electrode

tip diameter selection criteria for truncated cone profiled electrodes is based on 5/t where t is

the thickness of a single sheet to be welded. Again, there is no literature information on the

effects of tip diameter w h e nresistancespot welding Zn-Fe alloy coated steels.

Background information on the effect of electrode tip angle

The truncated cone electrode is generally accepted as the optimum configuration for

spot welding zinc coated steels. In the past, a 120 degree included angle (refer to Figure 4)

has been used on electrodes for spot weldability studies at BHP's Research and Technology

Centre, while a majority of U S A research establishments and auto manufacturers employ an

included angle of 90 degrees for spot welding zinc coated steels up to approximately

1.50 m m single sheet thickness.

Bowers and Eager [54] conducted a study to determine the effect of non-uniform

current distribution across the electrode face as a factor influencing electrode wear. The

study of various electrode geometries revealed specific trends between the current density

profile at the electrode face and the shape of the electrode. The effect of geometry on current

distribution was determined through computer modelling, which was employed to produce a

voltage m a p of each electrode shape as shown in Figure 19. F r o m the voltage map, the

current density profile at the electrode face was calculated. The ratio of the current density

at the electrode shoulder (face edge or periphery) to the density at the face centre was found

to decrease with increasing included angle as shown in Figure 20. A s the current density

distribution approaches unity, electrode erosion/wear became more evenly

42
distributed across the face, thus promoting longer electrode life. These results suggest that

the 9 0 degree included angle tip should outperform the 120 degree included angle tip. O n

the other hand, in Europe, the general belief is that the 120 degree included angle improves

tip cooling by dispersing heat over a larger electrode area and therefore improved electrode

life [55].

Greenwood [56] and Nied [57] have also modelled the temperature distribution during

the spot welding process and have recognised that most of the current that enters the weld

does so near the shoulder of the electrode, thus causing a current concentration at that point.

Greenwood's model of temperature generated in spot welding only takes into account

electrode geometry. Nied however, has taken into account the deformation of the sheet

w h e n electrode pressure is applied and noted that current localization m a y be due, at least

partially, to higher pressures at the electrode shoulder. Nied also noted that heat production

is a function of current density and because this could be highest at the periphery of the

electrode, it can be assumed that alloy growth/wear is greatest at this region.

a.76 L 4 —
9
C A
•
0 A
3
a.sa
r\
A
9
0>
2
• A
B.2S -3 A
O
H
1 A
H
<
QC
e ee ' 1 ' 1 '1
19 38 60 88
8 00 8.18 8.28 8.38
Scale in inches Electrode Included Angle
Figure 19: Finite element analysis used to Figure 20: The ratio of current density (J)
measure voltage points within a truncated cone at electrode face edge (or shoulder) to
electrode profile (half-section shown) [54] current density at face centre in a truncated
cone electrode profile [54]

43
3.2.2 Electrode life testing procedure

In order to assess the effects of including current upslope in the welding schedule and

changing the electrode tip diameter and included angle, the materials were subjected to

electrode tip life testing. Although there are m a n y minor variations on the way in which

electrode tip life testing is performed, this type of test is recognised world-wide as a

resistance spot weldability, qualification test.

A n electrode life test identifies the point at which spot welding electrodes fail to

produce a weld nugget of acceptable size during continuous welding at preset conditions and

without any change to welding current in order to compensate for growth in tip diameter

through wear or alloy build-up. T h e procedure adopted for electrode tip life testing in this
study was as follows:

i) Strips from the test material were sheared to 40 mm wide, transverse to the rolling

direction of the material. In this case, the sheets provided for this work were 1 m

wide, thus the test strips were 40 m m wide by 1 m long;

ii) Electrodes were machined with the specified tip diameter (designated by sheet

thickness) and included angle. Machining of electrodes used in this work was

carried out on a bench lathe with diameters being machined to an accuracy of

±0.05 m m ;

iii) A pair of electrodes were placed in the upper and lower electrode holders of the

welding machine, bedded into place by applying air to the upper cylinder of the

machine, and bringing the electrode faces together under a small force. Electrodes

were then checked for accurate alignment of thetipfaces.

iv) Alignment was achieved by closing the electrodes down onto a sheet of carbon

paper inserted between two sheets of blank paper. This was done without the

44
 application of welding current. The diameter of the imprint left on the blank paper

from the carbon paper was then measured in two directions with a graduated

eyepiece for eccentricity. The eyepiece used in this work was a Peak eyegauge,

graduated to 0.05 m m and with a 7 x magnification;

v) Two strips were placed between the spot welding electrodes as shown in Figure 21

and after setting a weld counter to zero, setting "electrode squeeze time", "weld

time" and "off time" on the welder control unit, thefirstweld was m a d e 20 m m

away from one end of the strips. This avoided any strip edge effects such as

excessive edge build-up of metallic coating, a c o m m o n occurrence in the

processing of metallic coated steel sheet. Welds were then m a d e d o w n the

centreline of the strips with a separation of approximately 10 m m . The last 20 m m

of the strips were also avoided;

vi) During the initial few welds, welding current was gradually increased until a weld

expulsion point was reached. The welding current was then turned back to a point

just below this expulsion point level for the next weld. T h e remainder of the

electrodetiplife test was then conducted at this level;

vii) Periodically, a spot weld was removed from the strips by shearing the strips either

side of the weld. These welds were then subjected to a peel test by gripping the

sample in a vice and prying the weld apart torevealthe weld nugget;

viii) The weld nugget was measured in two directions 90 degrees apart across the

nugget diameter with vernier callipers and the average diameter was recorded. This

process was then continued periodically until a point was reached at which the

average weld nugget size reached a diameter equal to 8 0 % of the original electrode

tip diameter employed in the test. At this point, the effective electrode tip life was

said to be over. A m i n i m u m of three electrodetiplife test runs were performed

45
 usingfreshlymachined electrodes for each run and the average result was recorded

as electrode tip life for the particular electrode/material/welding parameter

combination. The greater the electrode life, the more weldable the material.

T h e 8 0 % electrode tip diameter cut-off point for the m i n i m u m nugget size is a

generally accepted world-wide criterion used by the majority of researchers for tip life

testing.

Figure 21: Positioning of test strips into the welder for electrode tip life testing

3.2.3 Electrode tip life test conditions for weld parameter study

Electrode tip life tests for the weld parameter study were conducted at an electrode

force of 2.0 k N , 7 cycles (50 H z ) weld time, electrode squeeze time of 25 cycles and an off

time of 40 cycles. The machine delay in removing the upper electrode from the sheet surface

meant that there was effectively a 3 cycle electrode hold-time.

46
 Truncated cone electrode configurations of 120 and 90 degree included angles were

tested at tip diameters of 5 m m , 5 m m plus a 4 cycle current upslope in the welding schedule

and 7 m m . The 4 cycle current upslope period was arbitrarily chosen as a time suitable for

use in high volume production welding and is in line with that used by Lynch and Goodwin

[58] in looking at the effects of current upslope on the weldability of zinc-5% aluminium

coated steel. Welding time and force were selected from in-house welding schedules

developed for hot-dipped zinc-coated steel sheet.

3.2.4 Description of the welding machine

A microcomputer controlled, 120 kVA, Schlatter Selecta, pedestal, resistance spot

welding machine, was employed in all welding experimental work.

The welder was connected to an I B M X T computer equipped with a Data Translation

D T 2 8 0 1 - D data acquisition board which is capable of monitoring the weldtipforce, welding

voltage and welding current throughout each weld of the tip life tests. The tip force signal

was generated by a piezo-electric transducer, the tip voltage was obtained directly from a set

of contacts fixed to the welder electrode holders, and the current signal was developed

across a k n o w n low resistance in series with the workpiece.

The software for the computer is written in Microsoft compiled B A S I C [59]. This

system provides m e n u driven facilities for monitoring, displaying, archiving and producing

hard copies of graphical or tabulated data. A block schematic of the welder and ancillary

equipment is provided in Appendix 10. A general view of the equipment including welder

and control unit, computer facility and lathe for machining of electrodes is shown in

Figure 22.

47
Figure 22: General view of resistance spot welding facility including data acquisition system
and lathe for machining electrodes

3.2.5 Electrode tip life test results — weld parameter study

Table 8 shows actual tip life test results while average results from Table 8 are

represented graphically in Figures 23 and 24.

48
 Table ti — Electrode Tip Life Test Results — Weld Parameter Study

ZINCANNEAL Hot-Dipped Gulvtmizcd

Tip Angle 120' 90* 120' 90*
Test No. 1 2 3 1 2 3 1 2 3 1 2 3

5 m m tip 6131 4902 6008 4795 5213 6096 550 575 575 520 700 700
Ave 568() Ave 536*1 Ave 56"! Ave 640

7 m m lip 2003| 2313 1932 2197 3190 2553 690 650 550 860 810 600
Ave 208 Ave 2646 Ave 630 Ave 757

5 m m tip 7327 7985 7692 8231 7902 7848 700 620 550 700 675 630
plus upslope ve766 7 ve7993 Av62] Ave 668
1 I 1 I
1 Electrogalvanized Zinc-5% Aluminium
Tip Angle 120* 90* 120' 90*
Test No. 1 2 3 1 2 3 1 2 3 1 2 3

5 m m tip 6093 4849 5055 3498 3982 3042 310 280 345 350 300 310
A .ve 5332 Aive 35C7 Ave 312 Ave 320

7 m m tip Noresult 650 710 800 1050 840 1010

Ave 720 Ave 967

5 m m tip 5524 5821 6114 3385 3628 3572 551 675 1 700 450 610 567
plus upslope /w e 58:>0 /jve 35: Ave 642 Ave 542
1 1 f 1 1 ! 1
Aluminium-45% Zinc
Tip Angle 120' 90*
Test No. 1 2 3 1 2 3

5 m m tip 230 270 250 340 360 280

Ave 2.>0 Ave y.»7

7 m m tip 1050 975 920 1380 1000 1570

Ave 9152 \vel3 17
4

5 m m tip 225 275 240 325 345 270

plus upslope Ave 2 47 Ave 3 13

1 1 1 1
 FIGURE 23. AVERAGE TIP LIFE TEST RESULTS
KEY: A:6mm TIP, B:7mm TIP, C: 6mm TIP plus UP8L0PE

ELECTROGALVANIZED ZINCANNEAL
E>3 120% 0O°tlp K * 3 i20°tip ^ »o%

NO RESULT

B C
TIP CONDITION TIP CONDITION

AI-45%Zn COATED Zn-5%AI

E 2 120°tlp ^ wftlp K * 3 120°tlp M «(AiP

1400 1400

1200 1200

1000 1000
Ul
IL
Ul 800 UJ 800
Q o
O
600 o
o 600
I a
a 400 Ul 400
ui

200 200

B C B C
TIP CONDITION TIP CONDITION

50
FIGURE 24. AVERAGE TIP LIFE TEST RESULTS

HOT DIPPED GALV.

&S3 120% WrJp

1400

1200

1000

ui 800
a
o 600
E
d
Ul 400

200

A B C
TIP CONDITION

51
Electrode tip diameter

Increasing the tip diameter from 5 m m to 7 m m proved difficult with the

electrogalvanized steel w h e n all other parameters were fixed. Weld sizes were well below

the minimum level of 8 0 % of the electrode tip diameter even at the start of the life test and at

a current level just below the expulsion point. A number of attempts were made with 7 m m

diameter fresh electrodes and in each case poor weld sizes were developed. Irregular

expulsion from the sheet outer surfaces was thought to be the major cause for this

inconsistency. It could be that the small zinc coating mass of approximately 5 g/m 2 either

side is quite sensitive to even small electrode misalignment. That is, if the electrode is not in

firm contact with the strip surface over the full 7 m m tip diameter, then the electrogalvanized

coating is easily expelled. Energy expended in doing this would m e a n less energy was

available for weld nugget development at the faying surface.

D u e to this effect, n o results were obtained for electrogalvanized steel at 7 m m tip

diameter. A result m a y be obtained if electrode force is increased to 2.5-3.0 k N to assist the

electrode bedding in process. A result m a y also be obtained if electrode tip diameter is

decreased to 6 m m while leaving the electrode force at the original setting of 2.0 kN.

However, these suggestions could be the subject of future work.

Increasing the tip diameter from 5 m m to 7 m m w h e n welding Z I N C A N N E A L

resulted in a fairly substantial decrease in electrode life (approximately 2500 welds)

regardless of electrode tip included angle. Surface expulsion was not observed during

welding of mis material at 7 m m tip diameter as was observed with the electrogalvanized

steel. It is thought that even if small electrode misalignment had taken place, the heavier

coating mass of Z I N C A N N E A L (100 g/m 2 total) would be more difficult to volatilize and

remove prior to the electrode bedding in to the sheet surface. The majority of energy is

therefore transferred quickly to the faying surface for nugget development.

Both the 120 degree and 9 0 degree included angle tips produced approximately a

4-fold increase intiplife with the increase intipdiameter during welding of the aluminium-

4 5 % zinc coated steel. T h e zinc-5% aluminium coated steel reacted similarly to the

52
aluminium-45% zinc coated steel producing a 2.5-fold and 3-fold increase respectively for

the 120 degree and 90 degree electrode tips. The results indicate that even low levels of

aluminium in the coating can rapidly degrade the electrode face. It is envisaged that an

electrode degradation mechanism, similar to that found for the aluminium-45% zinc coated

steel [6] also applies to electrodes used to weld the zinc-5% aluminium coated steel. That is,

the aluminium-rich areas in the outermost alloy layer formed on the electrode are hard and

brittle and are quite easily stripped away from the surface, thus leaving voids.

There was little improvement in electrode life for increased tip diameter when welding

the hot-dipped galvanized steel. This is in direct contrast to the results obtained by Gould
[52].

Electrode tip angle

Reducing the electrode tip included angle from 120 degree to 90 degree had a marked

effect in reducing the life of electrodes to weld the electrogalvanized steel. This reduction

occurred for both conditions of 5 m m tip diameter with and without the introduction of

current upslope. In contrast, there w a s little variation in results obtained between the two

included tip angles over the various conditions examined for welding of Z I N C A N N E A L ,

and there was some improvement in electrode life with decrease in includedtipangle during

welding of the aluminium-45% zinc coated, zinc-5% aluminium coated and hot-dipped

galvanized qualities. T h e only exception to this w a s a decrease in tip life w h e n welding the

zinc-5% aluminium coated steel with a 5 m m electrode and incorporating the current upslope

period. This result, however, m a y be attributed to local variations in surface characteristics

on the product used in this study. In general, the theories put forward by Bowers and Eager

[54] hold true for only some metallic coated steel types.

53
Current upslope

Current upslope was found to be of benefit when welding the ZINCANNEAL and

zinc-5% aluminium coated steels only. Introducing the four cycle current upslope period

into the welding schedule on Z I N C A N N E A L increased the electrode life of a 5 m m tip,

regardless of included angle, by approximately 2500 welds. It is possible that this

improvement m a y be improved further at increased current upslope periods. There was no

significant improvement in tip life observed when using current upslope for welding the

aluminium-45% zinc coated and hot-dipped galvanized steels regardless oftipincluded angle

employed.

These results present quite a complex picture of the benefits of the parameters

examined on these materials. M u c h of the work overseas has concentrated on only zinc-5%

aluminium coated steel (current upslope effects) and on hot-dipped galvanized steel

(electrode geometry effects). A more complete picture is now in hand, however, the fact that

not all metallic coated steels respond to a particular parameter change in the same way, only

highlights the need for further work in order to understand the mechanisms of electrode

degradation fully.

Although tip life could be markedly improved by using current upslope when welding

Z I N C A N N E A L , discussions during the course of this thesis with automotive and appliance

industries revealed that m a n y did not have this facility on their spot welding machines. It

was felt then that this welding parameter should not be included in subsequent experimental

work on the Z I N C A N N E A L G 3 N ZF80. The benefit was highlighted, however, in a

technical paper submission to the Australasian Welding Journal, Fourth Quarter 1992 (see

Appendix 3). Since the majority of materials benefited from the lower included tip angle and

that this was not detrimental totiplife when welding the Z I N C A N N E A L , this angle would

be adopted in subsequent experimental work. A tip diameter of 5 m m would also be adopted

for the slightly thicker (0.8 m m ) Z I N C A N N E A L G 3 N Z F 8 0 material since there was a

definite trend of decrease in tip life with the use oftipdiameters larger than the recommended

sizes.

54
3.3 Description of Profilometer and Vision Sensing E q u i p m e n t

3.3.1 Profilometer

The profilometer employed in this study was a Rank Taylor Hobson Form Talysurf

120 machine. The profilometry technique relies on dragging a fine tipped stylus across the

surface to be measured, in this case the indented weld surface after removal from the weld

as shown in Figure 25, to map the surface profile of the specimen. Correlation of the

vertical and horizontal movements of the stylus provides a direct record of the surface pr

limited in resolution primarily by the geometry of the stylus tip dimensions [60].

Figure 25: Profilometer pick-up arm and stylus drawn across indented resistance spot weld
surface

The components of the profilometer system are indicated in Figure 26 and a brief description

of each follows.
 Figure 26: Components of the Taylor Hobson Form Talysurf 120 profilometer

The 120 mm Traverse Unit (\ 1

The traverse unit houses the motor, gearbox and associated circuits for traversing the

pick-up across the surface of the weld specimen. A straightness datum is incorporated

enabling measurements of up to 120 m m long to be made without reference to an external

straight line datum. The unit is mounted on a column with a w o r m gear allowing vertical

movement for height adjustment of the stylus arm over the specimen. Appendix 11 provides

technical details of the 120 m m traverse unit.

The Stylus Arm m

The stylus arm employed contained a standard conisphere diamond (type 112/2009)

with a tip radius of 1.5-2.5 jam (60-100 u m ) . The depth of recess measurable from the

56
stylus is 0.76 m m (0.03 in). This adequately covers any impression depth left during the

adopted spot welding technique.

Epoxv Granite Base (3)

The base is an epoxy granite construction on anti-vibration mounts. This provides a

firm support for the column and the test specimen.

The Electronic Interface Module (4)

The electronic interface module forms the link between the measuring system and the

host computer, providing data link and interfacing circuits for the system and the power

supply to the traverse unit. The module also includes a thermal sensitive printing facility for

producing hard copies of the specimen surface profile.

The Computer (5)

The computer performs the control and calculation functions. This is an Olivetti M300

P C with high resolution V G A graphics (full colour). T h e software is mouse driven. The

computer contains a single 1.4 M B 3.5 in floppy disk drive and a 60 M B hard disk.

Motor Controller (6)

The motor controller is a Taylor Hobson CD4 controller. This provides the interface

between the computer and motor operation for the traverse unit.

57
3.3.2 Vision sensing

W h e n used with robots, vision systems can recognise and locate objects which are

randomly placed in a work area. The robot can then pick them up for further processing. In

addition to this facility of providing robotic guidance, vision systems can be utilised to

measure component positioning and dimensions [61]. It is the latter feature that is of interest

in resistance spot welding monitoring and control, lf a vision system placed in close

proximity to a resistance spot welding machine can measure the diameter of a weld

impression left on the workpiece surface and relate this dimension to the actual weld nugget

size, then it is possible to utilise vision sensing for on-line monitoring and weld quality

control. In this set-up, if a weld impression is sensed to be wider than a preset value, then it

will be recognised that the electrode working face is worn to the level at which current

density is insufficient to produce weld nuggets above the m i n i m u m diameter requirement.

The microprocessors can then be employed to adjust the welding current level automatically

in order to continue welding at optimised conditions.

In order to assess this vision sensing concept, an A D E P T V I S I O N A G S (advanced

grayscale system) was employed to monitor welds taken during electrodetiplife testing.

The system essentially comprises of a camera (Pulmic C C D , T M - 5 4 0 ) mounted on a

tripod directly over the weld sample, a monitor (NEC/Multi Sync 30) and a controller or

central processing unit. A schematic of the system is shown in Figure 27.

T h e Adept Vision A G S uses a solid-state camera with a rectangular C C D (charged

couple device) sensor array. The output of the camera is a standard RS-170 video signal: 60

frames/second, with 2:1 interlace. The vision system grabs a single camera frame of data (in

1/30* second), digitizes it into greyscale and edge or binary images, and then processes the

image accordingly.

58
 VISION SENSING (REMOTE SENSING)

LIGHT SOURCE

WELD SAMPLE

Figure 27: Schematic of vision system experimental set-up

In binary m o d e , the system extracts boundaries from the camera image by a process

called thresholding, into black regions and white regions. Thresholding is the process of

considering all shades of grey darker than a specified level to be black and all lighter shades

of grey to be white. In the greyscale system, 'edges' are processed. A n edge is a region

where the brightness is changing very rapidly (in this case the edges of the weld

impression). These edges are extracted from the camera image by computing a cross-

gradient (or Sobel) edge operator in each 3 pixel high x 3 pixel wide portion of the image.

The vision system applies the edge operator at all the pixel locations in the image and then

thresholds the magnitude of the edges so that weak edges disappear. This process is

necessary for data reduction because almost all of the pixels have a non-zero edge strength in

their neighbourhood. All the pixels with edge magnitudes above the threshold are the edges

that are processed further. Subsequent processing of edges is like binary image processing.

The regions of high contrast are processed like black and white regions.

Greyscale vision was employed in this work because of its advantages over binary

vision. Interior surface features of objects are visible with greyscale vision when an object's

surface appears to have different brightness. This is a critical feature in identifying the edges

of the weld indentation or impression. The interior of the impression appears darker than the

59
outside sheet surface and is therefore depicted on the monitor as a well defined circle through
which line measurements can be taken over two impression diameters, 90 degrees apart.
T h e greyscale vision system also does not rely heavily on uniform lighting. The
contrast between surfaces does not change as m u c h as the absolute brightness of the surfaces
when lighting gets brighter or dimmer. Lighting sometimes may change over time or could
be non-uniform over the area of interest. If the contrast between the area of interest and its
background is not high, then non-uniform lighting over the area of interest can make binary
image processing impossible.
In the experimental set-up a desk lamp was used as the light source. The 60 watt
globe provided sufficient lighting for a clear image production. The camera was mounted
perpendicular to the weld surface and focused onto the impression. This focus remained
unchanged for all welds examined to avoid having to recalibrate the vision system.

60
 CHAPTER 4.0

EXPERIMENTAL RESULTS

61
4.0 EXPERIMENTAL RESULTS

The following is a broad outline of the experimental procedure adopted in order to

assess the feasibility of employing vision sensing as a newresistancespot welding quality
monitoring and control device.

(a) A weldability lobe was constructed to characterise the resistance spot weldability of
Z I N C A N N E A L G 3 N ZF80.

(b) An electrode tip life test was conducted on this material at selected welding conditions
determined from the preliminary experimental work and from the weldability lobe
assessment.

(c) During the course of the tip life test, degradation of the electrode tips was tracked via
carbon paper impressions taken at selected intervals.

(d) Weld samples were taken at near identical intervals for

(i) profilometer measurement of weld impression diameter;

(ii) vision measurement of weld impression diameter;

(iii) X-ray measurement of internal weld nugget diameter;

(iv) ultrasonic measurement of internal weld nugget diameter, and

(v) peel testing for manual measurement of weld nugget diameter with vernier
callipers.

62
(c) Since tensile shear strength is of prime importance, a second tip life test was conducted

and further samples were collected for

(i) vision measurement of weld impression diameter, and

(ii) tensile-shear testing of weld nugget strength.

(0 Relationships were then studied between the vision system results and those obtained

from alternative measurement techniques for monitoring actual weld nugget size and

strength,tipdegradation and weld impression diameter.

4.1 Weldability Lobe

In order to characterise the resistance spot weldability of the 0.8 mm ZINCANNEAL

G 3 N Z F 8 0 and to select a suitable welding current level for electrode life testing on this

material, a weldability lobe was constructed.

A lobe is constructed over a range of weld cycletimesand determines the current range

over which weld nuggets of a satisfactory size can be produced. T w o lines are normally

displayed, a line representing the current level at which m i n i m u m weld nugget size is

achieved (generally 8 0 % of the electrode tip diameter chosen) and a linerepresentingthe

current level at which metal expulsion will occur. A third line, referred to as 'nominal' is

sometimes shown which represents the current level at which a mid-sized weld nugget is

achieved.

The following conditions were employed to construct the lobe using R W M A Class II

truncated cone electrodes of 5 m m tip diameter and 90 degree included angle.

63
 Electrode squeeze time: 25 cycles (50 Hz)

Weld times: 5, 7,9,11 and 13 cycles

Electrode hold time: 10 cycles

Welding force: 2.5 k N

The electrode squeeze time was the time necessary for the pneumatic cylinders of the

welding machine to reach the preset welding force level of 2.5 kN. The welding force level

was set at 0.5 k N higher than that employed in the preliminary work on 0.6 m m thick

material to account for the slight increase in material thickness of the Z I N C A N N E A L G 3 N

ZF80.

At each set weld cycle time, current levels to produce m i n i m u m weld size and

expulsion point were determined by averaging the current recordings of three welds. Each

current recording in itself was an average weld current (kA R M S ) calculated by the computer

program [59] over the duration of an individual weld period. The results obtained directly

from the computer data acquisition system are shown in Table 9, while Figure 28 shows the

plotted weld lobe. Note that minimum weld size was determined through peel testing and

measurement of average nugget diameter, while the expulsion point was determined by

further increasing the current level from the minimum point until there was visual evidence

of expulsion.

64
 Table 9 — Current Level Recordings for Welding Lobe Construction

Weld Time Minimum Weld Nugget Diameter Expulsion Current Level Lobe Width
(cycles) Current Level (ave kA R M S ) (k amps)
(ave kA R M S )

5 7.40, 7.33. 7.33 8.17, 8.34, 8.29

ave 7.35 ave 8.27 0.92
7 6.93, 6.87, 6.97 7.76, 7.80. 7.77
ave 6.92 ave 7.78 0.86
9 6.52, 6.50, 6.57 7.41, 7.50. 7.67
ave 6.53 ave 7.53 1.00
11 6.39, 6.30, 6.35 7.05, 7.15. 7.14
ave 6.34 ave 7.11 0.77
13 6.23, 6.24, 6.26 6.79, 6.81, 6.84
ave 6.24 ave 6.81 0.57

mln. current expulsion

E
<
St.

c
9
O
c

7 9 11
Weld Time (cycles)

Figure 28: Resistance spot weldability lobe for 0.8 m m thick Z I N C A N N E A L G 3 N ZF80

4.2 Tracking of Electrode Tip Degradation Through Carbon Paper

Impressions

Utilising the weldability lobe, a welding current/welding time combination was

selected to produce an optimum weld nugget size. This condition was then employed for an

electrode life test (using the 5 m m tip diameter, 90 degree included angle electrode) on the

Z I N C A N N E A L G 3 N ZF80 during which electrode tip growth and wear was tracked and

65
weld samples were collected for submission to the profilometer and vision sensing

equipment, and for X-ray and ultrasonic examination.

The weld time chosen was 9 cycles, representing the midpoint of the lobe length and a

current level just below that which resulted in metal expulsion. Welding force and squeeze

time as per those employed to construct the lobe were used for the life test and an off time of

40 cycles (time between welds) was used to simulate high productivity welding. At no time

during the course of the electrode tip life run was the current adjusted.

Degradation of electrodes through alloying and erosion was initially tracked via the

carbon paper impression method described in Section 3.2.2 for selected welds over the

effective life of the electrodes. The life of the electrodes was deemed to be over after

completing 4600 welds (ie weld size had reached 4 m m in diameter). Figure 29 shows the

impressions obtained.

The diameter of the electrode tip impressions was measured in two directions

90 degrees apart using the Peak eyegauge, the results of which are shown in Appendix 12.

Average results from these are presented graphically in Figure 30.

66
 SO -7S

Q o e 9 9 •
•
too ns /50 17? 2oo 2SO Joo

% % « 9 9 m •
SSo *too 4 So SOd sso eSOO 4>SO

• • • • • • •
loo 7So 2oa 2So <fco 9sa /ooo

II0O- iZOO /30o /too fSoo /rn-rnOd /70t

*
# * * 0 m »
/ffoo /ft>o m-lOOO suoo SLmmOO 23CO 24.00

© 9 9 % •

:asoo
e
2.600 X?oo 2-300 .3900 3COtO 3?<>3

• *, • t ® ® Q
34 Oo 3600 3SOO 4ooo AZoa 44CO 46GO

Figure 29: Carbon paper impressions taken during the course of an electrodetiplife test on
Z I N C A N N E A L G 3 N ZF80

67
E

UJ 6
LU
**t tt t
2
Q ,,?• t '••
UJ % M
t t
QC t
UJ
> • t»
< t t*.

' i ' ' ' ' ' ' ' i ' i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i

1000 2000 3000 4000 5000

NUMBER OF WELDS

Figure 30: Averagetipdiameter results obtained from carbon paper impressions

68
4.3 Tracking of W e l d Impression Diameter T h r o u g h Profilometry

During the initial electrode tip life test, weld samples were removed for submission to

the profilometer in order to monitor the change in weld impression diameter. For the

purpose of this study, weld identification is identical to that used in the carbon paper

impression check on electrode degradation, however, actual weld samples were taken

immediately after each carbon paper impression test was taken. A scanning length of 10 m m

was chosen for the profilometer to cover both a 'run-on', 'run-off section on the parent

strip and all envisaged growth of the electrode. T w o profilometer traces were taken on each

selected weld. These were taken normal to each other and the average weld impression

diameter recorded. Traces taken parallel with the strip length were referred to as

'longitudinal' while those taken at 90 degrees to the longitudinal were referred to as

'transverse'.

A typical trace output from the profilometer is shown in Figure 31. The relatively

horizontal portions of the trace outside of the two major peaks represents the undamaged

surface of the test material. The undulating valley region between the two peaks represents

the portion of strip indented during spot welding, by the truncated cone electrodes, hence the

slightly tapered side faces leading up to the high peaks. The peaks (and those shown in

Figure 31 are just two points on the periphery of a circle) represent the highest points of

material build-up on the strip surface adjacent to the indented surface.

The horizontal distance between the two highest peaks was chosen for measurement of

weld impression diameter (shown as x-x) as the apex of each peak was the clearest and most

consistent reference point for measurement on each recorded trace. The peak to peak

distance was obtained via the distance measurement software in the computer facility. Valley

distance could not be chosen as a measure of weld impression diameter as there was some

degree of uncertainty over points of inflection that could be considered as consistent

reference points. There was sufficient yariation in the profile shapes over the electrode life

in this region due to the electrode degradation process. A measurement between the outside

of the m a x i m u m peaks near the sheet surface w a s also not chosen since it was difficult on

69
 ! !

R i

•a x

88

8*b
oo «

Is3
r§
B-'g
§J
*t
8£
S's
< a
<o £
tt ©
a J3
E

• •

OlIKIHI lOM <:|Nin.l l l l i O H H n «

BOO
.... et J
<M«
BOO 0 1 "V "Ml»'d

«H<- « • H I u » » u | a j | t inA
• * i r> .„3 .l.t BOO ni "I"". I'MH»(1

-3 B/. , I j S . y | • |<IU) | IO||

m
JO • „ , f, mjt\ BOO WW » I « ' S | • lin » i ...||
111 ...«.< .in ii'KHl lUWU'll

70
 \ \ i
• • • • •
t 3 S S I

tt

s
a

osiNiaj j ON siNiod n i * w j n« ONIHUUII

M M
see Bl P<*3 "I IJ» 'd
MM. BOB B- >-"»>S •! I JiJd
B MM' . IOIHJ|j|u6.u l«»l>'»A
1 *1 «
'c .nd Mn B B S •I »»S I »'l .-"•is
M j... I i « J I j I u t . y | « 1 '. o / i . f n
1 m I *> B Bl
M n n
ll" .nd BBII IVBs I •'S |»»-*o/| 10||
llllil'l.l ' H I IMiiu 11 IMUi.:i

71
many traces to clearly identify points of inflection over which to measure. A n example of
these uncertain measurement reference points can be seen on the second trace, Figure 32.
Points which were rejected as reference points for measurement of weld impression diameter
are shown arrowed.
Appendix 13 details results from peak to peak measurements (weld impression
diameters) throughout the initial tip life test. Averages of longitudinal and transverse peak to
peak measurements for each weld tested are presented graphically in Figure 33.

100J^ 2000 3000 4000 5000

NUMBER OF WELDS

Figure 33: Average profilometer impression diameter results

72
4.4 Tracking of W e l d Impression Diameter T h r o u g h Vision Sensing

The Adept AGS vision system was used on the same welds subjected to profilometry

measurements. T h e line measurement function w a s employed to measure impression

diameter again in the longitudinal and transverse directions but in pixel spacings which were

later converted to distance in millimetres. The circular function was also employed to

determine the percentage of edges found on the periphery of the impression. Appendix 14

details results from the vision system analysis. Averages of longitudinal and transverse

impression diameter for each weld tested are presented graphically in Figure 34.

y :
. WEAR-IN HALF-LIFE MATURITY TERMINATION
I

E 8
E
cc
• •
t
t

,• V M .
|

9
• -•"•] ttt * • • •

LU 9 9
h 7 L . < * ,<| •« l •

LU
5 t • •
<
0
•

UJ 6 - 9
0 •••
< •
cr
UJ •
> 5
<
(

* i i i i i • • .i i 1 i i i i i i i i i _ i • i i i i i < i i i < i i i i i i i i i i i i i i i i

1000 2000 3000 4000 5000

NUMBER OF WELDS

Figure 34: Average vision system impression diameter results

73
4.5 M e a s u r e m e n t of Nugget Diameter T h r o u g h X - R a y Analysis

In an effort to obtain a more accurate indication of nugget diameter prior to peel testing

of the selected welds and manual measurement of the diameter through vernier callipers, X-

ray and ultrasonic examination of the welds were conducted. For X-ray examination, a fine

grain film type D 4 was used. Focus to film distance was 760 m m . Emitter current was
4 m A and emitter voltage 85 k V A .

Figure 35 shows representative X-ray points from actual negatives taken throughout

the life of the electrodes. Longitudinal and transverse nugget diameters were measured

using the Peak eyegauge after placing the negatives on a light box to illuminate the X-ray.

Appendix 15 details the results of these measurements. Averages of longitudinal and

transverse nugget diameters are presented graphically in Figure 36.

74
Figure 35: Representative X-ray prints of welds taken throughout the initial electrode life
test

75
 9

8
I
vw
GC
UJ .... ••••• • , , •
2
<
Q
UJ 6
0
<
II
UJ
I 5
' ' ' ' ' ' ' ' ' I ' ' ' ' ' ' ' ' ' I ' 1 ' ' ' ' ' I ' I t I ' ' I I I I I I I I I I I I 1 I.J.

1000 2000 3000 4000 5000

NUMBER OF WELDS

Figure 36: Average nugget diameterresultsfromX-ray analysis

4.6 M e a s u r e m e n t of Nugget Diameter T h r o u g h Ultrasonic Technique

Ultrasonic testing can be employed to detect a number of discontinuities in weldments

including cracks, laminations, shrinkage cavities, porosity, slag inclusions, incomplete

fusion or bonding and incomplete joint penetration. In the case of this work the technique

was employed to measure the fused, resistance spot weld boundary at the faying surface in

an attempt to predict weld nugget size.

76
 In this non-destructive measurement technique, beams of high frequency sound waves

are directed into the test piece on a predictable path. W h e n a discontinuity lies in the path of

the beams, reflection occurs which can be detected and analysed on a cathode ray tube.

The equipment used in this work was supplied by the Australian Nuclear Science and

Technology Organisation ( A N S T O ) at Lucas Heights Research Laboratories, N S W ,

Australia. The flaw detector was a Krautkramer USIP 12 unit with a D T M 12 digital

thickness meter. T o emit the sound waves, an Accu-Tron 75 M H z transducer with a focal

length of 12.7 m m was employed. The scanning tank (tank in which sample sits immersed

in fluid) was constructed by A N S T O . A personal computer controlled movement of the

transducer which had a minimum step size of 0.12 m m .

Amplitude of ultrasonic signals returned from each sample was recorded on a 256

colour/64 grey scale P C display screen, each amplituderepresentedby a colour.

The procedure adopted was to immerse each weld sample, which was previously X-

rayed, into the water tank and onto a levelling jig. In order to avoid rusting through ingress

of water at the weld interface prior to peel testing, each weld sample was protected with a

silicon sealant around the edges of the steel strip. The sealant was allowed a number of days

to fully cure prior to sample immersion. After the weld sample was in place on the levelling

jig, the ultrasound transducer was focused on the surface of the sample and normalised to

the surface to obtain a m a x i m u m return echo from the reflected sound waves. The

transducer was then lowered so that the focal point was at the centre of the spot weld

thickness. This positioning allowed a 20 m m x 20 m m C-scan to be possible with the spot

weld being approximately central in the image recorded by the personal computer. The C-

scans were conducted with a step size of 0.24 m m . Actual scan size was 19.92 m m x

19.92 m m .

Images from C-scans were stored on floppy disc for subsequent measurement of

nugget size, as depicted by the weld boundary in each image, in both longitudinal and

transverse directions.

Figures 37, 38 and 39 show representative ultrasonic C-scan images from selected

welds taken during the initial electrode tip life test. It was evident from the images that two

77
boundaries could be measured which could reflect the weld nugget, an outer circular

boundary designated in red and an inner boundary designated by white which, during the

course of the electrode life, changed shape considerably and became more difficult to define.

Appendix 16 details longitudinal and transverse measurements obtained from both

boundaries while average results are presented graphically in Figure 40.

78
^lfSv3&rvSw^m0^y^^MV'X*i-^.y1>,^S^
2
o o • a o a
l^nr
a a Ta
a a a a a a a a a
ai a. r-. *-B in T pi n TH r% •« in
•tf

79
a
a

el
a
ri

n
a
--i
a

80
 £
a a a
£
Q n m i x - o w i T e n c . ^ a ^ T ffl

"W J L L „ ,IIU1

a
a
pi
d
a
pi

a 2=

- . . . | 1 .
T — r - i — i " ' i ,ri—r~r~
a a a a • a a a
1^ Ml Ul
a a a a a a a a
D i o i t s i i u i V n r J s

:V'J

81
 o Inner t outer

8
E t
• • • • •
§ 7
QC
L* -•••»• rv * • i
tt t t t t
UJ Ot
0
< 5.^0
Q o
UJ A
(5 * P o\ 0
<
UJ 3
>
<
2
:
I I I I 1 I I I I I I ' • I I I I • ' I < ' I ' ' I ' ' ' I ' ' I I ' I ' ' I I ' ' ' ' ' I ' ' I

1000 2000 3000 4000 5000

NUMBER OF WELDS

Figure 40: Average nugget diameterresultsfrominner and outer boundaries measured from
ultrasonic C-scans

4.7 M e a s u r e m e n t of N u g g e t Diameter Through Conventional Vernier

Calliper Method

O n completing weld measurements through profilometry, vision sensing, X-ray and

ultrasonic techniques, welds were subjected to peel testing withresultingnuggets being

measured again in both longitudinal and transverse directions with a vernier calliper.

Appendix 17 details results obtained through this technique, while averages of longitudinal

82
and transverse weld nugget dimensions from each peel test are presented graphically in

Figure 41.

9F

*-«i
8
E
£ 7
n
UJ ?

< 5 t t# ' t #
5 • •••
•• • •

o4
<
OC m.
UJ 3
>
<
2
4 Mill
. . . . . t . . . t t • • • • 1 1 1 1 • i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
_J
0 1000 2000 3000 4000 5000

NUMBER OF WELDS

Figure 41: Average nugget diameter results measured via vernier callipers after peel testing

83
4.8 Tensile-Sheur Strength Testing of Weld Nuggets

The tensile-shear strength of resistance spot welds is of prime importance, particula

in the automotive industry where spot welds are constantly subjected to stresses imparted

during vehicle movement and could be subjected to severe impact loading during vehicle

collision. Tensile-shear tests are used for quality assurance testing of production welds

(along with conventional peel testing) due to the fact that they are relatively inexpensive to

perform. Rather than destructively testing a formed and welded component, the tensile-

shear test can be conducted on specially prepared samples as shown in Figure 42. Weld

strength can be established at any time during the course of production welding, although

this is time-consuming, and can be used as a direct check on electrode degradation if so

desired.

The test is performed in a load frame and in the case of steel sheet thickness less than

approximately 1.0 m m , eccentric loading on the weld causes bending and rotation of the

weld to take place. This results in failure occurring around the periphery of the weld nugget.

DIRECTION OF
EDGES A S SHEARED
ROLLING (PREFERRED)

L
SPOT-WELD CENTERED
AS S H O W N

Figure 42: Tensile-shear test specimen configuration showing location of spot weld [71

84
 In order to assess the vision system's ability to monitor reducing tensile-shear strength
as a direct result of electrode degradation, a second electrode tip life was conducted on the
0.8 m m Z I N C A N N E A L G 3 N Z F 8 0 using identical conditions as those employed in the
initial electrode life test run. During the course of this second test, tensile-shear test samples
were prepared at selected weld intervals, the weld impression diameters being assessed on
these prior to subjecting the samples to testing on a load-frame at BHP's Research and
Technology Centre.
Tensile-shear samples were prepared in accordance with Figure 42, each half being 50
m m x 150 m m with an overlap region of 50 m m .
With the inherent variability which is c o m m o n between individual electrode tip life
tests, it was surprising to obtain an electrode life on this second run of 4601 welds. This
similarity in result to the initial run (4606 welds) indicates the exactness of conditions
applied to each of the two runs.
Vision systemresultsare tabled in Appendix 18 while average results are presented
graphically in Figure 43.
9 r

E 8
E
rr
i-
UJ
5
<
5
LU
CD
<
CC
LU
>
<

1000 2000 3000 4000 5000

NUMBER OF W E L D S
Figure 43: Average vision system impression diameter results, second electrode tip life test

85
 Tensile-shear testing on welds was conducted on an Instron load frame, Model 1341,
servo-hydraulic, with a loading capacity of 100 kN (resolution of 1 N). Cross-head speed
for testing was 3 x 10"3 mm/second. Tensile-shear test results are tabled in Appendix 19,
while average results are presented graphically in Figure 44.

CB
Q. 500
I
0
Z
UJ • • • t . •. t

cc
h
•;
(0 t t
cc 400
<
UJ
X
(0
UJ
.J
CO
z
UJ
• ' i ' ' ' ' ' • ' ' ' < ' i i i i i i i i i i i i i i i i i i
h 300 i i i i i i i i

lil
1000 2000 3000 4000 5000
0
<
cc NUMBER OF WELDS
LU
>
<

Figure 44: Average tensile-shear test results

86
 CHAPTER 5.0

DISCUSSION OF RESULTS

87
5.0 DISCUSSION OF RESULTS

In order to study trends that m a y exist during the effective working life of a resistance

spot welding electrode and to have these trends also recognised by the proposed vision

system, a metallic coated steel needed to be chosen which imparted extended life to the

electrode. BHP's Z I N C A N N E A L G 3 N Z F 8 0 is one such steel.

T h e additional heat treatment process imparted to Z I N C A N N E A L (and other

galvannealed steels), immediately after it exits the galvanizing, metal coating pot, generates

an iron-rich zinc coating (approximately 9.3-9.8% Fe) that is ideally suited to the resistance

spot welding process. Under optimum welding parameters, industries employing spot

welding as the preferred sheet metal joining technique, can take full advantage of this coating

by w a y of minimising the frequency of stoppages required to replace or redress worn

electrodes.

A s a direct result of the high concentration of iron in coating, when compared to other

metallic coated steel types examined in this study (see Tables 4, 5, 6 and 7), contact

resistances between the electrode working face and at the faying surface during welding, are

generally lower for Z I N C A N N E A L . This translates into a welding lobe which encompasses

a lower welding current band than that seen on free zinc-coated, hot-dipped galvanized steels

and on aluminium-zinc coated steels when spot welded under identical conditions (Appendix

5). T h e lower heat input requirement to produce welds of acceptable size, retards the

deterioration of the electrode caused through accumulation of the metallic coating both

around the periphery of the electrode tip and across the actual working face in direct contact

with the workpiece. Through this retardation, current density (applied welding

current/contact area) can be maintained at a higher level for longer periods with subsequent

weld size being also kept larger.

Studies of electrode degradation mechanisms by Finlay, Samandi, and H o w e s

(Appendix 5) and by H o w e s and Lake [2] haverevealedthat the iron-rich metallic coating of

Z I N C A N N E A L is transferred during welding onto the electrode face in distinct and uniform

88
alloy layers. These layers are quite different in comparison to layers formed on electrodes

used to weld other coated steel types.

Through scanning electron microscopy, elemental compositions of alloy layers have

been determined. In the case of Z I N C A N N E A L , the composition of the very stable layers

formed on the electrode were found to be constant through the layer thickness. A total of

four layers was evident. T h e innermost layer was found to be a ductile beta (P) brass of

approximate composition 5 2 % zinc, 4 8 % copper. Immediately adjacent to this lies a brittle

g a m m a (y) brass layer of 6 3 % zinc, 3 6 % copper. Separating the outermost layer of brittle

zinc oxide and the inner brasses is a zinc-iron alloy layer of varying composition, depending

on the electrode composition used. This layer is quite thick (30-60 u.m) when Cu/Cr or

Cu/Cr/Zr electrodes are used and is quite thin (5 uni) and severely cracked when Cu/Zr

electrodes are used.

The nature of the zinc-iron alloy layer produced on the Cu/Cr and Cu/Cr/Zr electrodes

imparts a resistance to fracture and would not be easily eroded away. In contrast, the

cracking observed through this layer on Cu/Zr electrodes would promote erosion of the tip

face. W h e n an electrode is removed from the workpiece, afractureline must occur across a

plane of greatest weakness. Since the layers formed on Cu/Cr and Cu/Cr/Zr electrodes are

uniform and crack-free, it can be expected that thisfractureline would also be uniform and

relatively parallel with the electrode face. The electrode face after each removal would be

relatively smooth in comparison to that which would occur with the Cu/Zr electrode with its

cracked zinc-iron layer. These differences are clearly evident in the photomicrographs

presented in Appendix 5. T h e smoother electrode surface after each weld on

Z I N C A N N E A L , using either the Cu/Cr or Cu/Cr/Zr electrode, would m e a n that with each

successive weld, a more uniform current density across the electrode face is maintained.

Nied [57], as mentioned in 3.2.1, has demonstrated the importance of uniform current

density in maximising electrode life.

In contrast, the irregular surface produced on the Cu/Zr electrode when welding

Z I N C A N N E A L would be expected to promote non-uniform current density across the tip

face and could generate an out-of-control, electrode wear pattern. High spots on the

89
irregular face c o m e into contact with the sheet surface initially during each weld, with the

passage of welding current being localised through these. Further alloy will build up at

these locations and will freeze during the "hold time'. These points on the face are then more

susceptible to stripping away from the electrode w h e n the electrode is removed from the

workpiece. This erosion process repeats itself with each successive weld with high spots

n o w at n e w locations coming into contact with the sheet surface. The irregular or eroded

surface constantly changes and the wear is far more rapid d o w n the axis of the tapered face

of the electgrode. The truncated conical configuration of the electrodes means that an

increasing, electrode tip surface area is being presented to the sheet with each weld. This

coincides with a gradual reduction in current density and therefore the formation of smaller

and smaller welds at the faying surface.

T h e electrode tip life data presented in Appendix 5 hasreflectedthis erosion analysis

with the Cu/Zr electrode composition giving the least electrode life w h e n welding the

Z I N C A N N E A L (shown as galvanneal) steel. While the Cu/Cr/Zr electrode, with its

combined attributes of high hardness and good electrical conductivity, proved to be the best

performer in regard to.electrode life, the availability and lower cost of the Cu/Cr electrode

and its current extensive use in Australia, m a d e it the obvious choice for the purpose of this

study.

T h e initial experimental work on Z I N C A N N E A L G 2 Z F 1 0 0 has proved beneficial in

establishing appropriate welding conditions on zinc-iron metallic coated steels and has

clearly demonstrated the superior spot weldability of the metallic coating. At the same base

steel thickness of 0.60 m m and identical welding conditions, the Z I N C A N N E A L coating

gaveriseto electrode life even higher than that obtained w h e n welding the electrogalvanized

steel which had a total coating mass of only 11 g/m 2 (Table 3). The higher concentration of

iron in the coating of the Z I N C A N N E A L , combined with the low level of aluminium is

thought to be responsible for the improvement.

Electrode life was, under most conditions examined, enhanced through the use of a

lower electrode tip included angle. Only the electrogalvanized steel consistently produced

lower electrode tip life with thereductionof included angle from 120 degrees to 90 degrees.

90
Electrogalvanized steel has been observed by the author to be more susceptible to metal

expulsion from the sheet surface during spot welding. This is most likely to be a function of

the low coating mass. It could be that the 120 degree included angle tip, shields the alloy

being accumulated around the periphery of the electrode as it indents the sheet surface,

thereby minimising the tendency to expel the coating outwards, away from the indenting tip.

T h e sharper angle of incidence subtended between the sheet surface and the tapered face of

the 9 0 degree included angle tip, means that less shielding is available to the accumulating

alloy. Since the current density would be higher, at a preset current input, with the 90

degree included angle tip, due to the less electrode surface area being in contact with the

sheet surface, then it could be expected that the accumulating alloy would be at a higher

temperature. In the case of the 90 degree included angle tip then, surface expulsion, with the

low coating mass, would occur more easily.

It is understood that the coating building up around the electrodetipperiphery could be

a combination of molten coating not alloyed with the copper electrode, and this would be

furthest away from the electrode, and coating which has alloyed directly with the copper

electrode. A n y surface expulsion then would mean that some electrode material is being

removed or eroded from the periphery. It is this action that is responsible for the

behavioural difference between the electrogalvanized steel welded with the 9 0 degree

included angletipand the other steels examined in the initial part of this study.

In order to obtain m a x i m u m tip life test results on the Z I N C A N N E A L G 3 N ZF80,

accurate electrode tip alignment to the sheet surface was necessary. Accurate alignment

would ensure uniform current density across the electrode face and therefore a more uniform

wear pattern. The carbon paper impression method used in this work provided a check on

initial electrode alignment and snapshots of the electrode wear pattern during the course of

electrode tip life testing.

T h e carbon paper method has indicated an intricate electrode wear and recovery

procedure throughout the course of the electrode tip life test. T h e electrode impressions

shown in Figure 29, together with the impression diameter results of Appendix 12, indicate

that no electrode contact face erosion had taken place on weld 1. Alloy had accumulated,

91
however, around the periphery of the electrode, as indicated by the average impression

diameter of 5.30 m m . This accumulation was uniformly dispersed around the periphery of

the original 5.0 m m electrode as the impression was found to be circular. It should be noted

that there m a y be some minimal contribution to impression diameter through the carbon

paper wrapping around the taper of the electrode. This is thought to occur since the carbon

paper could be seen to distort vertically upwards around the upper electrode during the

taking of each impression. The distortion is considered to be constant, however, since

electrode force remained constant throughout the tip life test.

B y the time weld 50 was reached, two features relating to the electrode wear were

apparent. Firstly, the impression indicated that a major cavity had formed at approximately

the 1 o'clock position and secondly, through measurement of the impression diameter, it

was found that the electrode diameter had become smaller (5.05 m m ) . The cavity formed,

slowly repaired itself by the time the 250 t h weld w a s made. The impression diameter results

showed that from the 50 t h to the 200 th weld there was some growth in electrode diameter.

This growth could be a combination of alloy accumulation around the tip periphery and

gradual weld by weld erosion at the electrode face increasing the tip diameter by wearing the

electrode d o w n through its central axis. Since the electrode is tapered (90 degree included

angle), naturally the diameter would be observed to increase. It is this second mechanism

that would have been responsible for repair of the cavity. Minor face erosion would have

occurred across the electrode face still coming into contact with the sheet surface, the only

region not in contact being the area of the cavity. T h e erosion continues until the full depth

of the cavity is reached, at which point the electrode face is once again in uniform contact

with the sheet. There was another major cavity formed by the 75 th weld at the 6 o'clock

position, however this was quickly repaired as there was no sign of it by the 100th weld.

Removal of the accumulated alloy around thetipperiphery is the only explanation for

the reduction in the diameter of the 50 t h weld. It could be argued that erosion at the

periphery could be responsible, however the 5 0 t h weld was still quite circular. Other

impressions did indicate that peripheral erosion did occur at times (impressions 1600 to

2400) but in the case of the 50 th weld, there is more likelihood that either expulsion of the

92
accumulated alloy had occurred, and as indicated earlier, this is more a feature of metallic

coated steels with very light coating masses, or it could be that through the action of the

electrode indenting the sheet surface, the alloy had been sheared off.

M i n i m u m impression diameter was reached at around the 2200 th weld. At this point,

peripheral erosion was the main feature. In general, electrode growth occurred from this

point on until the 4600 t h weld. S o m e electrode recoveries did occur where again, alloy

could have been sheared away bringing the electrode diameter back smaller. Other cavities

also began to form which were repaired through the process described previously. At the

4600 t h weld, electrode growth was such that the current density was n o w insufficient to

sustain a weld nugget (measured through peel testing) above the m i n i m u m requirement of

4.0 m m ( 8 0 % of the electrode tip diameter). Exacerbating the growth in tip diameter to

5.90 m m and its subsequent effect on weld size, was the large central cavity at the 4600 t h

weld. Without this region being in contact with the sheet surface, there was little chance for

uniformresistanceheating at the faying surface.

T h e electrode growth and recovery process appears quite cyclic as indicated in Figure

30, but non-uniform in frequency. This cyclic behaviour, together with the major reduction

in electrode diameter at around the 2200 t h weld, served as key focal points for determining

the ability of the vision system to identify major changes to the electrode and subsequent

weld size.

T h e profilometer technique was used as a preliminary check on measurements obtained

through the vision system, as was also the carbon paper impression method. Although the

profilometer technique measured peak to peak distance across the impression left by the

electrode on the sheet surface, as shown in Figure 45, it indicated a similar growth and

recovery pattern that w a s found from the carbon paper impressions. A major reduction in

electrode tip diameter was still detected in the vicinity of the 2000 t h weld, as shown in

Figure 33.
T h e only difference occurring between these two measurement systems was the overall

increase of approximately 0.5 m m on each diameter measured from the profilometer traces

over that measured from the carbon paper impressions. This was not an unexpected result

93
since the carbon paper measured the tip face diameter (Figure 45) while the profilometer
measured across the highest points on the sheet surface. These peaks are located on coating
that had been forced radially outwards from the tapered electrode as it indented the sheet
surface.

VISION SYSTEM

PROFILOMETER

— ^
WELD NUGGET
STEEL
3 SHEET

CARBON
IMPRESSION

Figure 45: Schematic showing the measurement positions through the carbon paper
impression method, profilometer and vision system

The cavities found on the electrode face through the carbon paper technique were
reflected on the indented sheet surface and were accurately revealed through the profilometer
traces.

At the 350 th weld, for example, the carbon paper impression of the electrode face
showed that a small cavity had developed at the 3 o'clock position (Figure 29). Although

94
difficult to sec, the carbon paper impression also showed a 'pin-hole' cavity jusl to the right
hand side of the central position. For the profilometer technique to be considered accurate, it
would need to detect both of these cavities on the sheet surface, which in reality is the mirror
image of the electrode face. Since longitudinal and transverse traces were taken through the
centre of each indentation, in the case of the 350 th weld, the longitudinal trace should reveal
both cavities while the transverse trace should reveal only the smaller 'pin-hole' cavity.
Obviously, it would be quite easy to miss the 'pin-hole' through the profilometer technique
since the diameter of the stylus covers only a small area as it is dragged across the
indentation. Figure 46, the longitudinal trace from the 350 th weld and Figure 47, the
transverse trace from the same weld show that for this weld at least, the cavities were not
missed. If multiple traces were taken, a complete 3-D picture could be obtained of the
cavities as well as the orientation of the electrode face to the sheet surface.

95
 8 «
s •
3
a
IS 3
m.

h
X
. i .. . H > t l f t H C u il b u . ;

O
•a
IA

awc>
no
-C
W3
2 C
t/5
rn)
a o5b
i/-> H
cn s
•5
o •8
8 b
brI-3I "0
go
<u bO
6 c
o =5
E3 c
2D.8. ss
-a
c go
O
•a
3
T3
Mi

5?
p

tu

E i E E E E E i E E E
3 3 3 3 3 3 3 3 3 3 3
a a a a a 8 a a a a a
a a a a a a a a a a a
N a 03 it v fl .N IS «
. »

rl 1 | N I JW LON '.111-1 I" 104,- ' 1H ">NIN»«».I\

at f J3 1 i ° 'd
"1 ' B B 0 0- i 11* : ~ •' 1
I r. i •» , 1 ' 1 '
i* BBS
•1 ei *. » •

i • i i * ' '1 '

mv. 1 »• )
M i. . •!
• M l '1 1' 11.1
96
 I ! 1 1 1 f I
• a m • i • i • i • 1 • i• i• » |
•
a
s
n N I« si
M
.
i
M
•
si
«
? si i
m
i
i
m
i
i
S
.
. .

c
o

Figure 47: Transverse profil ometer trace of the 350th weld showing po
;c
J "v5

corresponding to 'pin-hole' cavity on the electrode face

J ^
I

>

E E E E E E E E E E E
3 3 3 3 3 3 3 3 3 3 3
a a a a o s s a a a s
sa • s a s a a s a a s
i in a i A B i n » i n a i f t 8
f •n N M M M * 1 M M M
• . * I I I
*
i.. oaiNiad I O N sjsiod 3~i i joad m w i ONiNawn
; -«. i
!•
1
*• ~ «"• B O B BT Pu3 »Ild>OJ el
» »» B O B B »J«»S »I!iOJd
i B 00Z * I O I » « 3 | J)l u E « u I « O U J » H
JO l]iMp J"d mn BBB BS »l»JS l«'UJ»ll 1
0 0Z * 'Jei i r M j i u B i y |«)JO7|JOH 1
JO » 1 r- 1 p-ind jn 000 BBS <M»3S I " 1 J O " '°H
'
3~l 1 IC-t i m i J I O O U 3±3-|dUOO 97
 T h e single longitudinal and transverse traces taken on each weld during this work did

provide some indication of the extent to which the electrode was indenting the sheet surface,

as well as s o m e insight into its orientation with respect to the sheet surface. Electrode

orientation is evident from the degree of slope on the valley region between the two highest

points on each trace. T h e depth and width of the indentation and orientation play a major

role in explanation of the vision system results.

T h e average impression diameter results s h o w n in Figure 34, obtained through the

vision system working on greyscale m o d e , clearly indicate that the system can detect the

changes that take place as an electrode wears. For this initial electrode tip life run, the vision

system has indicated four zones through which the electrode passes. Each of these zones

designated as 'wear-in', half-life', 'maturity' and 'termination' could also be identified, to

s o m e degree, on graphs showing results of carbon paper impression diameters (Figure 30)

and profilometer impression diameters (Figure 33).

T h e zones are by no means distinct, ie they do not have a particular weld at which they

start and finish, with the exception of the veryfirstweld m a d e at the start of the tip life test

and the very last weld depicted as the end of the effective working life of the electrode. T h e

change from one zone to the next is quite gradual in this particular case and can take place

over a few hundred welds if results from the carbon paper impression technique,

profilometer technique and vision system are collectively examined, as shown in Table 10.

Table 10 — Wear-In, Half-Life, Maturity and Termination Zones for the Carbon Paper,
Profilometer and Vision System Measurement Techniques

Carbon Paper Technique Profilometer Technique Vision System

Wear-in 0 - 1200 welds 0 - 1000 welds 0 - 1000 welds

Half-life 1200 - 2200 welds 1000-2100 welds 1000 - 2000 welds
Maturity 2200 - 4000 welds 2100 - 4000 welds 2000 - 4000 welds
Termination 4000 ~ 4600 welds 4000 - 4600 welds 4000 - 4600 welds

T h e electrode wear-in period is in general characterised by a rapid increase in tip

diameter through alloy building up around the electrode tip working in combination with

electrode face erosion. Current density would berelativelyhigh during this initial period but

98
decreases as the build-up progresses. After approximately 1000-1200 welds, the alloy

build-up around the top has reached a level at which it can no longer be supported by the

tapered geometry and an electrode recovery phase begins. During this phase, the build-up is

removed. Electrode face erosion continues but at a m u c h slower pace than the rate of

removal of alloy from around the tip. T h e recovery process continues until the electrode

reaches its half-life stage (2000-2200 welds). At this point, the electrode tip diameter has

recovered to a level where the increased current density begins to dictate that once again,

rapid electrode growth should occur and at a similar rate, if not faster than that experienced

during the wear-in period. T h e rate of growth appears to stabilise up to the point of

termination with minor electrode recovery taking place on a number of occasions. After

approximately 4000 welds, the extent of alloy build-up and erosion is such that the current

density can no longer sustain welds above the pre-selected m i n i m u m value. At 4600 welds,

electrode life is considered as over with no further recoveries taking place.

It is interesting to note that the initial value recorded at the commencement of thetiplife

test from the vision system is well below that of the 5.0 m m original tip diameter and yet

after the wear-in period, the average diameter is approximately 1.5 m m larger than those

diameters recorded by the previous two methods. This anomaly can be explained through

the vision system's characteristic to measure from reference points based on lighting

contrast.

There is evidence from the profilometer traces to indicate that the depth of impressions

left on the sheet surface during welding by the indenting electrode, is greater at the

commencement of the electrode wear-in period. With little electrode face erosion and alloy

growth taken place at this time, current density is high. The higher the current density, the

greater the volume of molten metal generated at the faying surface and therefore the greater

the opportunity for the opposing welding electrodes to m o v e towards each other. It is this

inward m o v e m e n t of the electrodes attempting to displace the molten interface, that is

responsible for the high level of indentation at the commencement of the electrode life test.

Making this indentation appear deeper to the vision system is the peak (or ridge)

formed around the electrode on the sheet surface. The greater the indentation depth, the

99
higher the peak. Figure 46, a profilometer trace of the 350 th weld shows one such example

of a deeply indented sheet surface in combination with a high surface peak. Because of the

vision system's requirement to be situated directly over the impression, the lighting source,

in this case a single lamp, was placed close by but on one side of the impression. It is

thought that due to the depth of impression and the surface peak, shadows have been cast

across a portion of the base of the impression. Since the vision system relies on depicting

boundaries between zones of differing contrast (grey scale monitoring), the final

measurement m a y have been smaller than the true diameter of the impression base (the weld
diameter).

A s welding proceeded, the electrode diameter grew through the mechanisms

previously explained. Because the applied electrode force was unchanged throughout the

life test, the depth of impression became less and less and the surface peak subsequently

became gradually less predominant. T h e impression diameter increased as a function of the

growth of the electrode tip. The vision system is n o w no longer influenced by shadows

across the base of the impression but depicts boundaries of high contrast closer to the

periphery of the impression at the sheet surface.

For comparing the results from the vision system and profilometer, it is evident that

the region of highest contrast in the latter period of electrode life testing (beyond 500 welds)

is located around the outside of the surface peak. The vision system would therefore always

record a large diameter than the profilometer which recorded distance between the apex of

the peak across the impression.

The circular function results for the vision system (see Appendix 14) revealed that, on

average 8 1 % of the weld impression boundary could be depicted. The variation in circular

functionresultsover the course of thetiplife test m a y be attributed to the shadowing effect.

X-ray results revealed the smallest degree of scatter of any of the measurement

procedures adopted in this work. A n unexpected outcome of theseresults,however, is the

increasing diameter of the weld nugget over the tip life test (see Figure 36). While the

results clearly showed the oscillating behaviour associated with periodic electrode recovery

and the radical shift that occurred at around 2000 welds, the increasing diameter contravenes

100
the expected result. The weld diameter should decrease with increasing electrode tip

diameter as was seen from the peel test results shown in Figure 41. It is thought that a

contribution had been made to the X-ray diameter results by a lightly fused zone around the

actual weld nugget, the diameter of which increases with increasing tip diameter. This

lightly fused zone, seen as a dark ring around each weld X-ray (Figure 35) is likely to be

just the metallic coating bonding at the faying surface and being of low strength does not

contribute to the weld nugget size revealed during peel testing.

Although the C-scan, ultrasonic test results followed the expected trend of decreasing

weld size with increasing tip diameter (see Figure 40), measurement of nugget diameter

through this process relied heavily on interpretation of the images presented.

While all of the welds in the wear-in period produced C-scan images with well defined

boundaries, these boundaries became less defined as the electrode matured. After the wear-

in period, it became necessary to estimate the location of the boundary in many instances, as

the white regions, ie the regions of greatest fusion became scattered over the weld image.

Figure 37 (Weld N u m b e r 25), for example, clearly showed a well-defined white boundary

which could be easily measured. In contrast, Figure 38 (Weld Number 2500) and Figure

39 (Weld N u m b e r 4600) relied on the interpreter to define the boundary, thereby introducing

a variable into the measurement. This explains the degree of scatter in theresultsshown for

the inner boundary in Figure 40. Unfortunately it is this inner boundary result that

represents the true weld nugget. D u e to the scatter however, these results cannot be relied

upon to indicate the change in weld nugget diameter over thetiplife test The outer diameter

results have far less scatter and like the X-ray results, indicate a major recovery at around

2000 welds. These results are again higher than the actual peel test results s h o w n

graphically in Figure 41, indicating as in the X-ray results, some contribution

(approximately 1.5 m m ) from a lightly fused zone around the nugget. In general then,

neither the X-ray method or ultrasonic method provided an accurate picture of the weld

nugget diameter. They did, however, support thefindingsof the vision system in that they

depicted the major changes that took place over the course of the tip life test

101
 O n examining the results from peel testing, three distinctive peaks (or rises in weld

nugget diameter) could be seen. These occurred at around 1000 welds, 2000 welds and

3000 welds. It could be expected that if the electrode had recovered at these points during its

service life, then the impression left on the sheet surface would be decreased and the vision

system would see these. This was in fact the case as Figure 34, a graphical representation of

the vision systemresults,shows a decrease in surface impression diameter at 1000, 2000

and in the vicinity of 3000 welds. The rapid rise in impression diameter after 2000 welds

which continued up until approximately 2500 welds was also clearly associated with a

marked drop in weld nugget size from peel testing over the same period.

T h e oscillation in average weld impression diameter as seen by the vision system was

also evident in the results of the second tip life test (Figure 43) but the amplitude of the

oscillation was less than that seen in thefirstrun (0.6 m m first run, 0.1 m m second run).

Major electrode tip diameter recoveries were not a feature of the second run, nor was the

initial rapid increase in impression diameter. Instead, the impression diameter grew steadily

to the point of termination. In contrast to the initial run, the electrode used in the second ran

appears not to have indented the sheet surface heavily over thefirstfew hundred welds. It

m a y be that during the initial few welds where optimum weld current was being established,

a welding current slightly lower than that employed in the initial run was chosen, ie further

below the expulsion point. If the optimum current was lower in the second run, then it is

feasible that less indentation would take place. With less indentation, there would be less

height on the peaks formed around the indented sheet surface and therefore the vision system

would not be reading diameters as large as that seen in the initial run. The actual recorded

difference between the initial and second run, not taking into account the wear-in period of

the initial run, is approximately 1.5 to 2.0 m m . Other variables such as quality of the

machined electrode face (surface roughness) and coating mass variations across or along the

rolling direction of the Z I N C A N N E A L sheets used in the second run m a y have played some

minor role in generating the difference.

Direcdy associated with the increase in impression diameter over the tip life test was a

decrease in tensile shear strength of the welds as shown by the results in Figure 44. The

102
reduction in shear strength occurs as a result of the increasing tip diameter progressively

lowering the current density and therefore production of smaller and smaller welds.

Noticeable reduction in weld tensile shear strength followed directly by an increase

occurred in the vicinity of the 3700 t h , 3200 th , 2200 th , 1200th and 500 t h welds and these

deviations were detected, although not as visually pronounced by the vision system.

Through these results then, it is evident that the vision system can be employed to predict a

drop in weld size and an associated reduction in tensile shear strength.

A n optimum welding w i n d o w for the Z I N C A N N E A L used in this study, based

entirely on the vision system, can be determined knowing the optimum band for tensile shear

strength.

If it is considered that the most stable zone during an electrode life test on this steel

type occurs between the 1000 th and 4000 t h weld, ie between the commencement of the half-

life zone and end of the maturity zone (Figure 34), then from the average tensile shear test

results presented in Figure 44, the optimum tensile shear band is 417-453 M P a . By

applying this band to Figure 48, a plot of average vision system diameter results of the

second tip life run versus tensile shear strength results, an optimum vision system welding

window can be identified at between 6.13 m m and 6.35 m m . While the vision system sees

impression diameters within this window, the electrodetipscan be considered as suitable for

the continuation of welding. Beyond the 6.35 m m diameter level, the electrodes are in the

termination zone and will have to be either replaced or redressed or microprocessor control

could be employed to increase the welding current.

Note that the m i n i m u m tensile strength for 0.80 m m Z I N C A N N E A L G 3 N is 280

M P a . In order to still pull a slug at this m i n i m u m tensile strength of the parent strip, Figure

48 indicates that the vision system will be seeing an impression diameter of 7.16 m m .

In an attempt to relate the optimum vision system welding window to actual weld

nugget size, vernier callipers were employed to measure the nugget diameters from the

tensile shear test specimens after they were pulled. This was unsuccessful in that the vernier

callipers could notfitunder the extensive tails pulled with each weld nugget. Since both

electrode tip life runs produced a similar number of welds (4600 and 4606),resultsfrom

103
peel testing and vision system results from the initial run were plotted against each other over
which the optimum vision system welding window was placed to determine nugget size.
The results shown in Figure 49 indicate that weld nuggets of between 5.26 m m and 5.40
m m diameter will be produced with the vision system working within the optimum welding
window. Thisrepresentsan increase of 5.2% to 8.0% over the initial electrode tip diameter
of 5.0 m m , not unreasonable since the molten zone forming the weld nugget is generally not
just constrained at the faying surface over the contact area of the electrode face but isfreeto
m o v e outwards some small distance between the sheets being joined.

104
 8.00
REGRESSION EQUATION
7,80 Y-0.0059X • 0.814

R*- 0.4514
7.60

7.40 r

7.20 IINIMUM TENSILE STRENGTH

7.00

6.80

6.60

6.40
VISION SYSTEM WELDING WINDOW
6.20

6.00 • • •

5.80
. i i i i i i i i » i i i i i il i i i i i » i I. i i i i
5,60 ' "'
280 320 360 400 440 480

TENSILE S H E A R S T R E N G T H (Mpa)

Figure 48: O p t i m u m vision system welding window as a function of tensile shear strength

105
9.00

8.50
• •

8.00 s •
7.50

7.00

6.50
• ••X
VI810N SYSTEM WELDINQ WINDOW \

6.00 •

5.50 "

5.00 -
r

REGRESSION EQUATION
4.50 - Y--0.413iA 275Q3X • 1M78
•

R*- 0.2738
• i i . • • i • • 1 i • • i • i • • • 1 • • i i i . • i • 1 • • •_, _._! 1 1 1 1
4.00
6

Av. NUGGET DIAMETER (mm)

Figure 49: Optimum weld nugget diameter operating under the vision system

106
 The vision system is not designed to operate in real time such as the dynamic
resistance method or weld expansion and contraction method described in the Introduction.
Instead it relies on examining the weld made previously and offers the major advantage of
allowing the weld, time to cool and contract prior to measurement of the impression
diameter. This ensures a higher degree of accuracy in relaying and predicting weld nugget
size.

107
CHAPTER 6.0

CONCLUSIONS
6.0 CONCLUSIONS

The feasibility of the Adept Vision system to monitor and control resistance spot weld

quality on metallic coated steel sheet during simulated production welding, has been

assessed during the course of electrode tip life testing. T o provide the vision system with a

sufficient number of welds over which trends in weld quality and hence electrode condition

could be depicted, Z I N C A N N E A L , a zinc-iron-rich metallic coated steel, used primarily in

automotive and appliance industries, has been welded using commonly employed Cu/0.8%

Cr truncated cone electrodes. Theresultsof this work haverevealedthe following.

1. During an initial weld parameter study conducted using electrogalvanized steel,

Z I N C A N N E A L , A l - 4 5 % zinc coated steel, Z n - 5 % Al coated steel and hot-dipped

galvanized steel, only electrodes used to weld the Z I N C A N N E A L and Z n - 5 % Al coated

steels benefited to any extent through the addition of current upslope into the welding

schedule.

2. In general, a truncated cone electrode tip profile with an included angle of 90 degrees

rather than 120 degrees produced an increase in electrode life with the exception of

electrodes used to weld the electrogalvanized steel.

3. An increase in electrode tip diameter over the recommended size was detrimental to

electrode tip life when welding electrogalvanized and Z I N C A N N E A L steels.

4. Measurements taken from carbon paper impressions of an electrode tip used to weld

0.80 m m thick Z I N C A N N E A L G 3 N Z F 8 0 steel showed that the electrode diameter

oscillated over the effective life of the electrode.

109
5. Profilometer traces taken across the indented sheet surface over the course of the

electrode life indicated similar trends to that of the carbon paper measurements and that

electrode indentation was greatest at the commencement of welding.

6. The vision system accurately depicted the oscillating trends associated with electrode tip

diameter recovery, the actual measurements being closely associated with weld

illumination and depth of electrode impression on the sheet surface.

7. X-ray and ultrasonic C-scans taken from welds prior to peel testing failed to produce

accurate weld nugget sizes but assisted in establishing confidence in the vision systems

ability to depict changing weld quality.

8. The vision system results correlated fairly well with tensile shear strength of the weld

nugget and if maintained through feedback control, to produce impression diameters of

between 6.13 m m and 6.35 m m , optimum tensile shear strength would result.

9. The optimum vision system welding window produced weld nugget diameters between

5 . 2 % and 8.0% greater than the 5.0 m m electrodetipdiameter employed in this study.

110
CHAPTER 7.0

FUTURE WORK

111
7.0 FUTURE W O R K

During the course of this work and from the author's previous experience, it has been

noted that on separation of the welding electrode from the workpiece after the weld cycle has

been completed, a relatively smooth fracture plane is left on the sheet surface of

Z I N C A N N E A L . The vision system assessed appears to have little difficulty in visualising

the base of the impression left by the indenting electrode due to this surface. T h e only

interference in obtaining an accurate measurement across the base is the surface ridge or peak

on the sheet surface casting shadows, the contrast changes being points that the vision

system focuses on. In the case of A l - 4 5 % Z n coated steel, electrode/workpiece separation

planes are quite rough as brittle alloy pockets formed on the electrode working face are

stripped off and are frozen on the sheet surface. Even if the electrode indentation depth is far

less towards the end of the electrode life, accuracy of the vision system m a y be hampered by

the roughened weld surface. It is proposed then, that the vision system be employed on this

steel type in order to assess its effectiveness in predicting weld nugget size.

T h e lighting source appeared to play an important role in the vision system's ability to

very accurately monitor weld impression size and hence predict nugget size. B y positioning

a number of lamps around the vision system camera, more accurate results m a y have been

achieved. Fibre optic lamps attached to the camera m a y provide better illumination as these

would shed light in a more normal direction to the weld surface thereby minimising or totally

illuminating shadows. Future work should then include an assessment of more efficient

lighting of the vision zone. This m a y also address the problem of welding the A l - 4 5 % Z n

coated steel.

Also, better illumination can be provided by employing a low power laser beam, since

monochromatic light would help bring out the details of the weld nugget area and its

periphery w h e n focused on the weld spot.

A grey scale camera and frame grabbing facility are adequate for extracting relevant

features from the weld nugget impressions thus alleviating the need for expensive colour

image processing hardware and software. Grey scale images are also m u c h simpler to

112
process in real-time. Indeed with the advent of the new pcndium microprocessor and its

immediate predecessor, the grey image processing: task of single frames containing spot

weld nugget impressions can be achieved in milliseconds. This real-time capability allows

sufficient time for the pre-programmed microprocessor to alter the combination of process

input parameters, if needed, in between two successive spot welds in order to obtain the

desired spot size that provides the needed strength and appearance. This way, the process

will conform to the workpiece following predetermined acceptable spot weld standards that

have been evaluated through this thesis.

In this work the correlation between spot size, measured using machine vision and

weld strength w a s established. A method, or rather a procedure for feeding back the

monitored process output to the system and adjusting the input parameters accordingly, is

necessary. In this situation, adaptive process control can be achieved by selecting a new set

of primary process input parameters, ie parameters that their magnitude can be changed

during the process, eg weld force and current levels. Selection of these parameters will

depend on their cumulative effect on weld nugget size, since they are coupled. Closed loop

control can then materialise by simply comparing the size of afreshlym a d e weld nugget to

the desired one stored in the microprocessor's m e m o r y as an area array having certain

dimensions, ie radius (or diameter), that maximises weld strength. Image differentiation

algorithms already exist, especially for regular shapes, as is the case of weld nuggets, and

are capable of image subtraction. This subtraction is usually performed by superimposing

the actual and desired images and the difference in magnitude of their respective radii can be

correlated to the required change of magnitude of process input parameters.

Subsequent to this work, a control algorithm with a stepped function needs to be

developed in order to increase or decrease the input parameters as appropriate and the system

to perform the next spot weld with the aim of maximising its quality characteristics,

specifically strength and appearance. It is important to note however, that the control system

needs to observe the life cycle of the truncated cone electrode and distinctregionsthat have

been identified in this thesis. In other words, it will be pointless to attempt to alter any

process input parameters during Stage 1 (wear-in) and Stage 4 (termination) for obvious

113
reasons. Therefore, efforts to adaptivcly control the process should concentrate in Stages 2

and 3 (half-life and maturity respectively), where the useful life of the electrode lies. It is

also imperative to set tolerance limits that are acceptable within these stages and with regard

to the m i n i m u m and m a x i m u m desirable weld nugget size. These limits can be easily

obtained from the data presented in this thesis,relatingweld nugget size and strength in the

designated Stages 2 and 3. In addition, the control system must be able to distinguish weld

nugget size variations due to material imperfections. Such imperfections will depict larger or

smaller weld nugget sizes to the ones obtained under controlled laboratory conditions. T o

overcome this challenge, control software that accounts for such variations can be developed

using thresholding or quality band techniques which can be incorporated in the form of

simple subroutines.

Following the above mentioned proposed future work, a machine vision system

capable of monitoring and controlling the spot welding process by measuring the direct

outcome of the process prior to performing the next spot weld can be fully developed and

applied to any industrial environment that warrants the employment of automated spot

welding facilities. However, the harshness of this environment m a y necessitate the need for

maintenance, ie cleaning of the machine vision lens protector.

Finally, the concept of controlling the spot welding process using machine vision can

be taken further and applied also for path following and positioning of the electrode surfaces

parallel to parent material surfaces of the lap joint, eg contoured paths encountered on car

bodies. Additional control algorithms for path generation and following are required in this

case, together with a decision-making ability or to the exact location of the spot weld in the

path and its strategic importance at that location. However, to accomplish these high levels

of intelligence tasks, an artificial intelligence system is necessary. It is proposed, that in this

case, a neural network is used to provide spot weld pattern recognition and subtraction, ie

low intelligence level decisions, whilst an expert system with real-time inference engine

provides the high intelligence level decisions including path following and spot weld

placement. Such a hybrid intelligence system is feasible and there are already software

packages available that can help develop the autonomous spot welding tasks.

114
REFERENCES

[1] Nealon, C.S., Lake, J.S.H. Conditions for acceptable resistance spot welding of

Z 1 N C A L U M E coated qualities. B H P Research and Technology Report N o 912,

November 1986.

12] McGregor, G. Resistance spot welding of metallic coated steels. Paper presented at

the Australian Welding Institute and Australian Institute of Metals Seminar, Resistance
Welding, Adelaide, 1982.

[3] Assel, M.D., Nadkarni, A.V. Non-dressing dispersion strengthened copper electrode

designs for galvanized steel welding. Conference Paper N o 9, Sheet Metal Welding

Conference III, American Welding Society, Detroit, October 1988.

[4] Kimchi, M., Gould, J.E. The evaluation of resistance spot welding electrode

materials for welding galvanized steels. Conference Paper N o 8, Sheet Metal

Welding Conference m , American Welding Society, Detroit, October 1988.

[5] Howes, S.W. The resistance spot weldability of Weirton Steel Corporation Z n - 5 %

aluminium coated steel. B H P Steel Research and Technology Laboratory Report N o

1060, July 1989.

[6] Howes, S.W., Lake, J.S.H. Degradation of electrodes whenresistancespot welding

metallic coated sheet steels. Conference Paper N o 10, Sheet Metal Welding

Conference UJ, American Welding Society, Detroit, October 1988.

[7] A W S Welding Handbook, Welding Technology, Eighth Edition Volume 1.

[8] Dickinson, D.W. Welding in the automotive industry. Report on AISI Project N o

1201-4O9C, Republic Steel Corporation, August 1981.

[9] Beatson, E.V. A n introduction to quality control systems in resistance welding.

Paper presented at the Resistance Welding Control and Monitoring Seminar, T W I

1977.

[10] Rivett, R.M., Johnson, K.I. Quality control for resistance welding — A review.

Conference Paper N o 4, Sheet Metal Welding Conference I, American Welding

Society, Detroit, 1984.

115
[11] Tsai, C.L., Dai, W.L., Dickinson, D.W., Paritan, J.C. Analysis and development of

a real-time control methodology in resistance spot welding. Welding Research

Supplement, December 1991, pp 339-s to 351-s.

[12] Natzler, P.J. The monitoring and in-process control of resistance welding.

Advanced Welding Systems. Proceedings of 1st International Conference, London,

November 1985, Paper N o 48, pp 513-519.

[13] Appleby, G.W. Recent developments inresistancewelding controls to improve

standards of welding quality and reliability. Conference Paper N o 23, Welding

Institute Conference on Steel Sheet and Strip Welding, Kenilworth, March 1972.

[14] Gedeon, S.A., Sorenson, C D . , Ulrich, K.J., Eager, T.W. Measurement of

dynamic electrical and mechanical properties of resistance spot welds. Welding

Journal, December 1987, pp 378-385.

[15] Publicity Brochure, Hirst weld current meter M K V I , Hirst Electric Industries Ltd,

Sussex, England.

[16] Nakata, S., Nishikawa, M., Kurozumi, Y., Okuna, H. Monitoring of spot weld

quality by measurement of voltage between electrodetips.Transcript of The Japan

Welding Society, April 1979, pp 58-66.

[17] Nakata, S., Nishikawa, M., Horie, J., Kawasaki, M . Quality assurance in spot

welding by detecting voltage between electrode tips. JJW Document 111-639-80,

1980.

[18] Johnson, K.I. Automatic spot weld correction. Metal Construction, February 1977,

pp 70-71.

[19] Rivett, R.M. In-process spot weld control. Metal Construction, M a y 1980, pp 230-

234.

[20] Ganbowski, F.J., Williams, N T . Advantages in resistance spot and seam welding

of zinc coated steel strip. Sheet Metal Industries, November 1972, pp 692-705.

[21] Bourtorwick, M.J. Weld quality monitoring of resistance welding. Conference

Paper N o 14. Sheet Metal Welding Conference HI, Detroit, October 1988.

116
[22] Broomhead, J.H.W., Dony, P.H. Resistance spot welding quality assurance.

Welding and Metal Fabrication, July 1990, pp 309-314.

[23] Johnson, K.I. Resistance welding quality control techniques. Metal Construction

and British Welding Journal, M a y 1973, pp 176-181.

[24] Digimetrics NDT-700 Publicity Brochure, N.H. Elliot Electronic Enterprises Pty Ltd,

N S W , Australia.

[25] Spotrite 2030 Monitor Control Unit. Publicity Brochure, A R O Machinery C o Ltd,

Castelnau, London.

[26] Spotrite 2043 Monitor Control Unit. Publicity Brochure, A R O Machinery C o Ltd,

Castelnau, London.

[27] Weldamatic M P 1 6 P Publicity Brochure, British Federal Ltd, Castle Hill Works, West

Midlands.

[28] Hain, R. Resistivity testing of spot welds challenges ultrasonics. Welding Journal,

M a y 1988, pp 46-50.

[29] Bhattacharya, S., Andrews, D.R. Significance of dynamic resistance curves in the

theory and practice of spot welding. Welding and Metal Fabrication, September

1974, pp 296-301.

[30] Andrews, D.R., Bhattacharya, S. Resistance weld monitoring for production. Metal

Construction and British Welding Journal, M a y 1993, pp 172-175.

[31] Andrews, D.R. Spot and projection welding quality assurance. Sheet Metal

Industries, July 1978, pp 781-784.

[32] Hawkins, I.L. Superior quality production spot welding using adaptive control.

Conference Paper N o 22. Sheet Metal Welding Conference I, Detroit, November

1984.

[33] Beevers, A. Temperature measurement to determine post weld heat treatment

conditions for spot welds. British Welding Journal, April 1963, pp 173-183.

[34] Snee, R.K. Infrared monitoring of resistance spot welding. Metal Construction and

British Welding Journal, April 1972, pp 142-148.

117
[35] Broomhead, J.H.W., Dony, P.H. Resistance spot welding quality assurance.

Welding and Metal Fabricator, July 1990, pp 309-314.

[36] Stiebel, A., Ulmer, C , Kodrack, D., Holmes, B.B. Monitoring and control of spot

weld operations. Conference Paper N o 4. Sheet Metal Welding Conference II,

Detroit, 1986.

[37] Johnson, K.I., Hannah, M.D., Roswell, S.L., Dinsdale, W . D . Properties of

splashed resistance welds. Metal Construction and British Welding Journal,

November 1973, pp 401-406.

[38] Janota, M . Adaptive system of resistance welding control. Conference on Advances

in Welding Processes, Harrogate 1978.

[39] Janota, M . Adaptive control for assurance of quality inresistancewelding. Welding

Research Institute, Batislava, Czechoslovakia. Source unknown.

[40] Taylor, J.L., Xie, P. A new approach to the displacement monitor in resistance spot

welding of mild steel sheet. Metal Construction, February 1987, pp 72-75.

[41] Havens, J.R. Controlling spot welding quality and expulsion. S M E Paper A D 7 6 -

279, 1976.

[42] Vahaviolos, S.J. Applications of acoustic emission to factory automation and process

control. Materials Evaluation, December 1984, pp 1650-1655.

[43] Wrigley, A.L. G M using acoustic-type adaptive welding controls. American Metal

Market/Metal working News, March 21st, 1977, p 18.

[44] Bull, C.E., Stacey, K.A., Calcraft, R. On-line weld monitoring using ultrasonics.

British Journal of N D T , Vol 35, N o 2, February 1993 pp 57-64.

[45] Hain, R. Resistivity testing of spot welds challenges ultrasonics. Welding Journal,

M a y 1988, pp 46-50.

[46] Allen, L.A. Ultrasonic inspection of resistance spot welds. Conference Paper N o 3.

Sheet Metal Welding Conference II, Dearborn, Michigan, 1984.

[47] Williams, N.T. Recent developments in the resistance welding of zinc coated steels.

British Steel Corporation Report SM/881/D.

118
[48] Freytag, N.A. Spot welding galvanized steel. 7th International Conference on hot-

dip galvanizing, Paris 1964.

[49] Schueler, A.S. Welding steels with thin metal coatings. Welding Journal, August

1963, pp 378-384.

[50] Gedeon, S.A., Schrock, D., LaPointe, J., Eager, T.W. Metallurgical and process

variables affecting theresistancespot weldability of galvanized sheet steels. S A E

Technical Paper Series 840113, February 1984.

[51] Johnson, I.W. Spot welding of carbon steel. Welding Journal, March 1960, pp 89-

96.

[52] Gould, J.E. The effect of welding time and electrode diameter on electrode life when

resistance spot welding galvanized steel. Conference Paper N o 12, Sheet Metal

Welding Conference II, American Welding Society, Detroit, October 1986.

[53] Nealon, C.S., Lake, J.S.H. Resistance spot welding tip lives for Z I N C A L U M E

coated qualities. B H P Steel Research and Technology Report N o 950, January 1987.

[54] Bowers, R.J., Eager, T.W. Non-uniform current distribution in spot welding.

Conference Paper N o 15, Sheet Metal Welding Conference II, American Welding

Society, Detroit, October 1986.

[55] Lake, J.S.H. B H P Sheet and Coil Products Division overseas trip report, M a y 1989.

[56] Greenwood, J.A. Temperatures in spot welding. British Welding Journal 8 (6),

1961, pp 316-322.

[57] Nied, H.A. The finite element modelling of the resistance spot welding process.

Welding Journal, 63 (4) 1984, pp 123s-132s.

[58] Lynch, R.F., Goodwin, F.E. Galfan coated steel for automotive applications, S A E

Technical Paper Series 860658 (Feb 1986).

[59] Taylor, R., Morrison, B.R. Microcomputer controlled spot welder instrumentation.

B H P Sheet and Coil Products Division, Research and Technology Centre, Report N o

910, September 1989.

119
[601 Maddrell, E.R. A review of the techniques and applications of surface roughness
measurement. B H P Sheet and Coil Products Division, Research and Technology
Centre, Report N o 1039, January 1991.

[61] Adept Vision Reference Guide, Version 10.1. Adept Technology Inc. San Jose,
California U S A , February 1991.

[62] Welding Zinc Coated Steel. Handbook of the Australian Zinc Development
Association/John Lysaght (Australia) Limited.

120
APPENDICES

1. Howes, S.W. Electrode Erosion During Resistance Welding of Metallic Coated

Steels. Paper presented at the Institute of Engineers Australia, Fifth International

Conference on Manufacturing Engineering 1990, Wollongong 11-13 July 1990.

2. Howes, S.W., Siores, E. Resistance Weld Design and Product Quality. Paper

presented at the International Mechanical Engineering Congress, Sydney 8-12 July

1991.

3. Howes, S.W. Improving Resistance Welding of Coated Steels. Paper presented in

the Australasian Welding Journal, Fourth Quarter 1992.

4. Howes, S.W., Siores, E. Resistance Welding — Refurbishing a Mature

Technology. Paper presented at the W T I A Conference, University of Wollongong,

1993.

5. Finlay, R., Samandi, M., Howes, S. Optimising Resistance Spot Weldability of

Metallic Coated Steels Through Electrode Chemistry. Paper presented at combined

W T I A / A I N D T Conference, Melbourne, October 1994.

6. Product Data Sheet for Z I N C A N N E A L G3N.

7. Technical Data Sheet 1 for J47 Copper Alloy.

8. Technical Data Sheet 2 for J47 Copper Alloy.

9. Product Data Sheet for Z I N C A N N E A L G2.

10. Block schematic of B H P Research and Technology Centre's Resistance Spot Welder

and ancillary equipment

11. Technical specification for Form Talysurf 120 m m Traverse Unit.

12. Electrode Tip Carbon Paper Impression Diameter Results.

13. Profilometer Weld Impression Diameter Results — Initial Electrode Tip life Test

14. Vision System Weld Impression Diameter Results - Initial Electrode Tip Life Test

15. X-Ray Weld Nugget Size Results — Initial Electrode Tip Life Test

16. Ultrasonic C-Scan Weld Nugget Size Results — Initial Electrode Tip Life Test

17. Peel Test Results Measured Through Vernier Callipers.

121
18. Vision System Weld Impression Diameter Results - Second Electrode Tip Life Run.
19. Tensile Shear Test Results.
APPENDIX I

123
 Tr* Institution ol Engineer* Auttrata
Fifth International Conference on
Manufacturing Engineering iggn
Wollongong 11 -13 July 1890

Electrode Erosion During Resistance

Welding of Metallic Coated Steels
SW HOWES
Research Officer, B H P Coated Products Division.
Research & Technology Centre, Port Kembla

S U M M A R Y Rapid erosion of electrodes is an Inherent problem facing sheet metal fabricators engaged in the resistance
ipot welding of metallic coated steels. The rate of degradation is closely linked to the composition of the metallic
coating of each steel type but can be controlled to some extent through optimisation of welding parameters. Some
parameters are found to be unique to a particular material. This study examines the influence of electrode tip geometry,
and current upsloping on the weldability of AI-*»5% Zn coated and galvanized steels. Reducing electrode tip included
angle was associated with an Improvement in electrode life on both steels. A n increase in tip diameter produced a
marked increase in electrode life on the AI-*»5% Zn coated steel but not the galvanized steel. The addition of current
upsloping to the welding schedule of these steels did not improve their weldability.
I. INTRODUCTION taken in order to reduce the frequency of electrode
redressing.
The formation of a weld in any resistance welding
operation is dependent on the heat generated due to the
electrical resistance of the materials being joined. The
total heat generated (H) in joules, is given by the
equation:
H = l2Rt

where 1 = welding current (amperes)

R = electrical resistance (ohms)
t = time for current flow (seconds or cycles).
To maintain sufficient heat for welding, any decrease in
electrical resistance requires an increase in welding
current and/or welding time. The overall electrical
resistance and contact resistances during resistance spot
welding of uncoated steels are usually quite high in
comparison to those experienced during welding of
metallic coated steels. This is represented
schematically in Figure I. There is a need then to use
higher welding currents/times to achieve weld sizes of
acceptable quality similar to those produced on uncoated
steels of similar thickness. This is easily achieved but at
great expense to electrode life. The heat generated at
the electrode sheet surface interface provides the
driving force for alloying to occur between the metallic
coating of the sheet and the copper electrode. This
causes erosion or pitting of the electrode face, an initial
rapid increase in electrode tip diameter due to build-up
of alloy at the periphery of the face and a subsequent
deterioration in weld quality as welding progresses [ll
High volume Industries joining large tonnages of metallic
coated steels are then faced with having to frequently
stop production at welding stations in order to redress or
replace worn electrodes. Also indicated in Figure 1 is
the much smaller temperature difference between the Figure I Schematic representation of resistance spot
electrode/material interface and faying surface welding showing the points of electrical resistance and
(material/material interface) for the metallic coated the resulting temperature profile
steeL This translates to a narrow welding range for R j - contact resistance of faying surfaces
these steels and a higher chance of metal expulsion at R 2 - contact resistance of electrode/material interface
the surface. R 3 - electrical resistance of material
To improve the weldability of these steels and to make
them more acceptable to fabrication industries, it is This study examines the effects of welding current
necessary to examine resistance spot welding parameters upslope, electrode tip diameter, and electrode tip angle
and to determine their effects on electrode life. It is on the wettability of Al-45% Zn coated and hot dipped
important also to determine if the various metallic galvanized ( H D G ) steels.
coated steels react similarly to any one parameter or a
change in parameters. Corrective steps can then be
2. TABLE I

Voidability of sheet steels is defined in two ways in this C O A T I N G COMPOSITION

study, through weld lobe current range and electrode life
test results. A welding lobe represents the range of Al Fe Pb Si Coating
suitable weld time/current combinations under which (%) (%) (%) Mass
satisfactory welds may be made. The wider tne lobe, the (g/m 2 )
more weldable the product. Wide lobes are beneficial to
Al-45% Zn 52.3 1.37 0.47 130
the operator since they permit a much broader range of
HDG 0.33 0.3I 0.26 3I7
currents to be utilized during the set-up period, weld
quality is more difficult to control with narrow lobes 4. EXPERIMENTAL PROCEDURE
since even small power fluctuations m a y generate
welding currents outside the acceptable range, resulting Welding was conducted on a 120 kVA pedestal ipot
in the production of interior welds. Figure 2 shows welder using R W M A Class II Cu/Cr truncated cone
typical weld lobes for uncoated and galvanized steels of electrodes, machined to within +0.03 m m of the
equivalent thickness. The position and width of these nominated tip diameter. Welding lobes were constructed
lobes can vary with material thickness, electrode force, using an electrode squeeze time of 23 cycles, weld times
surface condition of the workpiece, electrode condition, of 3, 7, 9, 11 and 13 cycles (30 Hz), a hold time of IO
test method and measurement of weld sizes and, in the cycles, electrode force of 2.0 kN and electrode tip
case of metallic coated steels, the composition of the configurations of:
coating. Minor variation m a y also be attributed to steel
base chemistry. (i) 5 mm tip diameter and 90* included angle;
16 (ii) 3 m m tip diameter and 105* included angle;
_E»>ULSION
14 (iii) 5 m m tip diameter and 120* included angle, and
(iv) 5 m m tip diameter, 120* included angle and a 4
cycle current upslope.

i;: M H NUGGET
DiA^cTen
Electrode life tests were conducted at 2.0 kN electrode
force, 7 cycles weld time and a weld current just below
expulsion level for electrode configurations of 120* and
90* included angles and:
O 2 4 6 8 10 12 14 16 18 20 (i) 3 mm tip diameter;
(ii) 5 m m tip diameter and 4 cycle current upslope, and
WELD TIME (CYCLES) (iii) 7 m m tip diameter.
Figure 2 Typical welding lobes for 1.2 mm uncoated Nugget size was measured using a vernier caliper and the
and galvanized steels test terminated when the minimum diameter was
reached. A number of tests were performed at each
The effective life of an electrode is defined by the condition and the average electrode life calculated.
number of welds it can produce before minimum size
welds are made. There is no world-wide criterion for the 5. THE EFFECTS OF WELDING PARAMETERS ON
minimum weld size, however, the most commonly E L E C T R O D E TIP E R O S I O N
accepted is that weld size must be no less than 8 0 % of
Welding lobes for Al-45% Zn coated and HDG qualities
the original electrode tip diameter. Electrode life is
are shown in Figures 3 and 4 respectively. The key to
always increased with increasing weld lobe width. A s
these figures is shown after Figure 5. Table U shows
welding progresses and the electrode tip face is enlarged
weld lobe current details at a welding time of nine
through erosion and alloy ouild-up at the periphery, the
cycles for these steels. This time represents the
current density decreases due to the applied current
midpoint of the welding times used in construction of the
acting over an ever increasing surface area. A point is
lobes.
eventually reached when the minimum weld size is
produced. The number of welds needed to reach this
minimum level is increased with increasing lobe width.
Electrodes must possess high electrical and thermal
conductivity, high hardness at ambient and operating
temperature and, most importantly, resistance to
alloying to achieve long life. The most commonly
employed electrode composition in Australia is the
Cu-l% Cr (nominal) alloy, selected because of its
relatively low cost and sufficiently high hot strength.
There are numerous other alloys which provide longer
electrode lives. These cost considerably more however
and, in the case of the AI2O3 dispersion strengthened
electrodes, while life can be extended fourfold, the cost
is four times as great as for the Cu-lX Cr electrode.
While there is obviously an economic advantage in using 8 10 12
this type of electrode due to the reduced downtime for WBJ3 TtvtH Ccyctoei
redressing, most fabricators still insist on using the more
readily available electrode compositions.
3. MATERIALS
The samples which were used for this investigation were
all produced from 0.6 m m thick aluminium killed low
carbon steel base. The coating compositions and coating Figure 3 Welding lobes for 0.6 m m Al-45% Zn
masses are listed in Table L coated steel
 rapid electrode degradation (2). It is generally ttrttH
m a t the uuncateo cone electrode profile (Figure J) *an
an included angle between %' and 140*, is the optin-.um
shape lor weloing metallic coated steels [)] and was
therefore used in this study.

8 10 14
WELD TIME (cyclM)

RADIUSED TRLNCATBD PIM=LE

Figure 4 Welding lobes for 0.6 m m H D G steel

Key: Figure 5 Resistance spot welding electrode tip profiles

120* included angle electrode tip
103* included angle electrode tip Previous research on the truncated cone profile has
90* included angle electrode tip suggested that the ratio of current oensity at the
120* included angle electrode tip plus electrode face edge to that at the face centre is a strong
current upslope function of the cone angle I H Figure 6 indicates that
the lower the included angle, the more uniform the
current distribution across the electrode face. It follows
T A B L E 11
that alloying across the face and hence erosion woula
vVELD LOBE CURRENT DETAILS AT 9 CYCLES also be more uniform. This should retard the electrode
degradation process and extend electrode life since a
non-uniform current distribution would more rapidly
Electrode nl-45%Zn HDG change tne shape of tne electrode.
tip angle Coated OJPttiKT DENS'TY VS ITJCUJDeD ANGLE

V.P 10.3 10.5

120* R 9.7-10.9 9.6-11.3

105*
\7
V.P
R
1.2
10.0
9.4-10.6
1.7
10.0
9.4-10.6
*I
*
W 34-
W 1.2 1.2 si
MP 9.7 9.7 •

90* R 9.0-10.4 9.0-10.3 o

W 1.4 1.3 1

120* plus
MP
R
9.5
8.9-10.1
10.5
9.9-11.1
I' 0
.,.., _J
30
.
60 M 120 ISO
upslope 1.2 1.2
MO_UDED CONE ANGLE (degress)

Key:
Figure 6 Variation of the ratio of current density (3)
MP - welding lobe midpoint in k amps the electrode face edge to the current density at the
R - welding lobe range in k amps face centre with included angle in truncated cone
W - welding lobe width in k amps electrodes. Data from Bowers and Eager 1986 [21
Results from this study showed that changing electrode
5.1 Electrode Tip Geometry tip included angle had little effect on weld lobe width,
but did affect both lobe position and electrode life. The
The search for an electrode shape which exhibits long results of electrode tip life tests are shown in Figure 7.
life, has good mechanical strength and is easily Decreasing the included angle pushed the welding lobe to
maintained is still continuing. From a previous study at lower current levels for both Al-45% Zn and H D G steels
the B H P Research and Technology Centre, the life of the (Figures 3 and 4 and Table II). Electrode life was also
extended tip or pimple tip electrode shown in Figure 5, found to improve to varying degrees with lowering of the
when welding steel with from 5 to 100% aluminium in included tip angle. In conjunction with lowered included
the coating was found to be superior over that of the angle, the use of the larger tip diameter (7 m m )
more commonly used truncated cone electrode- It could produced the best improvement in electrode life on each
also be expected to perform in a similar manner on any steel type-
other metallic coated steel, however, this profile lacks Tip diameter selection usually depends only on the size
mechanical strength and can create welding difficulties of weld required, which can De limited by the design of a
if misalignment occurs. This eliminates the extended tip product such as flange widths and on sheet thickness.
electrode from industrial use. When electrode alignment The results of this work however, have shown that the
or part fit-up is a problem, radiused electrodes can be composition of the metallic coating also plays arolein
used. Zinc build-up on these is quite rapid however, the selection process.
changing the shape of the electrode face leading to
 Zn HOG

1400
1200
U! 1000
800
600
i-
•

400 •

200 ' •EvH

0
A B C A B C
TIP CONDITION TIP CONDITION
KEY A:6mm TIP :90° INCLUDED ANGLE
B:7mm TIP
:120° INCLUDED ANGLE
C:5mm TIP (4 CYCLE UPSLOPE)
Figure 7 Electrode tip life test results

In this study, the 120* and 90* included angle tips for Al-45% Zn coated steel, however, was shifted to a
produced approximately a 4-fold increase in tip life with slightly lower current range, offering the possibility of
the increase in tip diameter from 5 m m to 7 m m for some saving in power cost to tne faoricator.
welding Al-45% Zn coated steel (Figure 7). For the H D G
steel, there was little difference in electrode life for 6. COMMENTS
increased tip diameter.
Extended electrode life on both steels with reduction in
5.2 Welding Current Upslope electrode tip included angle m a y oe attriouted to more
uniform erosion over the electrode face and possibly
Upslope control is the gradual increase of the welding more easily detached alloy buila-up at the electrode
current magnitude during the initial period of current periphery. This could be due to the higher angle of
flow in each spot weld (Figure 8). The rate of current incidence made Detween tne electrode shoulder and
change and the time interval involved are readily sheet surface. By periodically removing the build-up,
adjustable on control panels of spot welders with this current density is kept high and wela size maximised for
facility. Upsloping is said to assist in melting and the surface condition of the electrode at any given time
displacing the metallic coating, enabling tne spot during welding, nlso for a given amount of electrode
welding electrodes to gradually seat before m a x i m u m erosion increase in tip diameter is less for an included
current is reached [51 This is believed to minimise the angle of 90*, so the decrease in current density is less
heating between the electrode and workpiece and than it would be for an included angle of 120*. O n the
therefore minimise the rate of alloying between the strength of the results presented, preference should be
metallic coating and the copper electrodes. However, given to the use of the 90* electrode tip.
results from weld lobe construction and electrode life Research is continuing into possible causes for the
tests showed that not all metallic coated steels benefit difference in welding performance of n D G steel over
from upsloping. Al-45% Zn coated steel with increasing tip diameter.
Certainly, no oenefit can be seen in using larger than
currently recommended tip diameters for welding h D G
steels. Employing larger Diameters, while not affecting
electrode life, would only result in the need to use higher
welding currents at cost to the faoricator. It is highly
recommended however, that for the Al-45% Zn coated
steel, consideration oe given to the use of slightly larger
tip diameters if the design of the components to be
joined allows it. Special care should then be taken to
accurately machine and align the electrode faces, as
obtaining uniform electrode/sheet surface contact
becomes more difficult with increasing tip diameters.
TIME 7. CONCLUSIONS
Optimum conditions for resistance spot welding of
Figure 8 The location of current upslope (u/s) in the Al-45% Zn coated and H D G steels have been established
resistance spot welding schedule through the construction of welding lobes and electrode
tip life tests. The results have shown that:
Lobe width did not change with the addition of current
upslope for Al-45% Zn coated steel (Figure 3) but (i) electrode life is greatest for the Al-45% Zn coated
narrowed for H D G (Figure 4) steel and there were no steel welded with electrodes of 7 m m tip diameter
appreciable changes in electrode life. The welding lobe and an included angle of 90*;
(I() the use\-ni tru,ncatevi cone electrode* with an *. REFERENCES
included a n g l e o f 1(J" iill#ri " improvement in
weldability over the 120* electrode profile on both I. Williams, N.T. "Recent Developments in U*
steels; Resistance Welding ol 2inc Coated Steels". Bntjvr.
Steel Corporation Report No. SM/8S1/D.
(ii!) the use of larger up diameters proves beneficial in
improving the life of electrodes used to weld 2. Freytag, N.A. "Spot Welding Calvanized Ste*.-.
M - U 3 % Zn coated steel but not H D G steel; Paper presented at the 7th IntcrnatioMi
Conlerence on Hot Dip Galvanizing, Paris 1964.
(iv) the addition of current upslope to the welding
schedules did not improve the weldability of these ). Schueller, A.W. 'Welding Steels with Thin Metal
steels. Coatings", welding Research Supplement, August
1963.
J. ACKNOWLEDGEMENTS
i*. Bowers, RJ. and Eager, T.W. "Non-Uniform
The author wishes to thank the management of BMP Current Distribution in Spot welding". Conlerence
Steel Coated Products Division for permission to publish Proceedings, Sheet Metal Welding Conference II.
the information contained in this paper and Dr J.S.H. Detroit, U S A October 1986.
Lake for his suggestions during its preparation.
3. Lynch, R.F. and Goodwin, F.E. "Galfan Coated
Steel for Automotive Applications". SAC
Technical Paper Series 860658, February 1986.
APPENDIX 2

124
 international Mechanical
Engineering Congress Sydney
8-12 July 1991

Resistance Weld Design

and Product Quality
S.W. H O W E S
Senior Research Officer, BHP Coated Products Division
E. SIORES
Mechanical Engineering Department, The University of Wollongong

ABSTRACT

Resistance welding is the most widely used joining process which is currently experiencing a resurgence in today's ouality-
conscious environment T h e process appeals to production engineers, designers and equipment specifiers in m a n y industries
because of its process record of consistent quality production at high speeds. For this it is considered to be the most cost-
effective w a y of joining light-gauge sheet metal and especially high strength coated steels and aluminium. L o w volume
industries (eg aerospace) employ resistance welding together with sensing means for both monitoring and feedback control
purposes. O n the other hand mass production industries (eg automotive) have yet to implement such adaptive control facilities.
Recently, however, the strive for increasingly higher productivity and quality levels has forced mass production industries to
turn to such in-process control facilities. These requirements have led to the development of sophisticated but simple to use
microprocessor controlled systems capable of both monitoring and controlling the process input and output parameters during
the welding cycle. This article describes the various approaches and equipment available today. The aim is to encourage
medium size rnanufacturing industries in Australia to implement such alternative cost effective production solutions for the sake
of their survival in the market place.

1. I N T R O D U C T I O N

The increasing demand of consumers for quality T h e simple shop floor tests performed on spot welds
welded products has seen a resurgence in the resistance spot are no longer considered a satisfactory means by which weld
welding technique with n e w and innovative methods being quality can be examined and maintained. Tests such as the peel
developed to ensure spot weld integrity. This technique, and chisel test are considered by m a n y to be unreliable
normally used to join thin gauge sheet steel has been ill-used indicators of weld quality and require time-consuming and
for m a n y years, more than likely because of the relatively costly destructive testing in order to obtain a result Material
forgiving nature of uncoated materials. Even now, it is not on which spot welds are found to fail by face-fracture mode for
u n c o m m o n to observe sheet-metal fabricators, involved in high example, can be shown to possess quite a high Ductility Ratio
volume welding industries, spot welding with excessively high (D.R.) this being a more accurate measure of weld strength
welding currents with subsequent weld metal expulsion taking (notch sensitivity of spot welds). This test too however, is
place. While this is a visual sign that at least a weld is quite time consuming both in test sample preparation and
occurring, there appears to be little regard for the damage testing time and has not been widely adopted.
caused to electrodes. If a watchful eye is not kept on electrode Low volume industries such as the aerospace industry
condition then as welding progresses weld quality will have for s o m e time realised the benefits of employing
deteriorate. resistance spot weld monitoring and in m a n y cases feedback
Exacerbating this effect is the conversion by many to control devices to streamline production welding without the
the range of metallic coated steels with their inherent anti- need for destructive testing of weld joints. Recemly however,
corrosion properties. It has been shown that alloying takes the strive for increasingly higher productivity and quality levels
place between the metallic coating of the steel and the copper has forced higher volume welding industries such as the
electrodes and depending o n the composition of the coating, automotive and appliance industries to do likewise.
the degradation of the electrodes through erosion of brittle alloy
layers formed, can be quite rapid (1). T h e rate of diffusion of This paper describes the various approaches and
alloying elements into the electrodes is heat dependent It is equipment available today with the hope that it will aicourage
therefore even more important to optimise the spot welding m e d i u m sized manufacturing industries in Australia to
Eammeters to extend the electrode life when using these steels. implement such alternative, cost effective production solutions
: is also worthy to note that each metallic coated steel can have for the sake of their survival in the market place.
its o w n unique set of welding parameters and that a parameter
such as current upslope m a y not be beneficial during welding
of some steels (2). Current upslope, available on most modern
spot welding machines is simply used to ramp the welding
current up to a rr»«T"f"""' point and is said to aid the electrodes
to bed-in to the sheet surface (3).
 These parameters include:
I, PROCESS MONITORING AND CONTROL
SYSTEMS (i) electrical parameters
There are a number of definitions for monltcrina but all
(ii) nugget temperature
basically relate back to weld quality. Dickinson (4) defines the
monitoring process as "one which gives Information about the (iii) infrared radiometry
quality of the product". Beatson (5) states "any sample
monitor must look at some phenomenon or characteristic which (iv) weld expansion and contraction
occurs or changes as a result of the start and growth of a weld
and whose output bears a definite relationship to the weld (v) acoustic emissions, and
diameter, this usually being taken as a production cirterion of
weld quality". (vi) ultrasonic signals
A resistance welding monitor does not solve welding
problems, it simply highlights them by providing a track record
of every spot weld made. It is then up to the operator to
recognise any anomaly and In the absence of a control device, 3. ELECTRICAL PARAMETERS
adjust the process accordingly. The characteristics of the ideal
In any resistance welding process, the formation of a
monitor and the exact parameter which gives the best Indication
weld relies onresistanceheating. The total heat generated (H)
of weld quality, has been debated by many, however, most
in joules is given by the expression:
systems display the magnitude of the monitored parameter on
some form of meter. T h e ideal relationship between monitor
count (the critical parameter indicated by meter deflection) and H-I^Rt
weld quality (nugget diameter) is shown in Figure 1. T h e
relationship should be relatively linear, should be similar for where I is the welding current (amperes),
variation of all those parameters which could affect weld
quality in production such as mains fluctuations, electrode R is the sum of the electrical resistances
wear and current shunting and should have as narrow a scatter (ohms) in the system (material/material
band as possible (line A B on Figure 1) to enable prediction of interface, material/electrode interfaces
weld size. and bulk material), and
With regard to a 'control or correction system', this is t is the time for current flow in seconds or
defined as a unit which monitors one or more of the critical cycles (50 or 60 Hz).
welding parameters during the weld cycle in an attempt to
maintain weld quality under varying external conditions by The electrical parameters can be measured through various
automatically processing these parameters and using the output means. Maintaining the welding current (I) is extremely
lo autjomatically adjust the welding conditions. important for good quality welds. Current density can fall on
quite rapidly with wear of welding electrodes and/or alloy build
up around the electrode face in the case of metallic coated steels
or mushrooming of the face (plastic deformation caused
through heat) observed w h e n welding both uncoated and
metallic coated steels. Current can be measured indirectly
through measurement of voltage by either a toroidal coil placed
around the primary or secondary of the welding machine,
which is the most c o m m o n method or through the more
accurate Hall-effect probe, which measures voltage across a
Z semiconductor due to the surrounding magnetic fields.
O Instrumentation m e n electronically integrates the voltage signal
o to obtain current as indicated by Faradays Law:
V = ^Acos8
cc
o where V is measured voltage.
\-
dl
z dt
is change in current withrespectto
O time,

s NUGGET DIAMETER
is the area of the toxoid loop (differe
for Hall-effect method), and

8 is the angle of the loop to the magnetic

field.
Figure 1 Schematic representation of a typical
monitor count/nugget diameter The output is normally displayed as a peak current on a
relationship meter calibrated in rms units. Automatic 'current steppers' can
then be employed to adjust welding current w h e n a certain
A number of process parameters have been identified as predefined m i n i m u m level has been reached however, most
suitable for mora taring weld quality, most of which can also be sheet-metal fabricators do not employ steppers. In most cases,
employed in closed loop feedback control They all provide a a current meter is used only to initially set up a welding
non-destructive means for checking welds and either totally schedule and to periodically monitor the current
eliminate or at least reduce the frequency of destructive testing.
 Direct readings of voltage are obtained through voltage There are several unlu developed in the U K and U S A
probes attached to either top and bottom surfaces of the for both monitoring and automatically correcting either current
materials being welded or to both spot welding electrodes. or voltage. The D I G I M E T R I C S NDT-700 developed In the
M a x i m u m voluge has been related to the growth of spot weld U S A and S P O T R I T E developed by the British Welding
nuggets In plain carbon steels (6) and ultimately nugget tensile Institute and A R O Machinery C o Ltd London are Just two
shear strength (tensile strength Increases with m a x i m u m examples of these microprocessor uniu.
voluge). Nugget diameter can also be related directly to the
voluge drop to expulsion point occurring after the m a x i m u m or Measurement of dynamic resistance appears to be one
peak voluge (7). Weld quality monitoring through this method of the most widely accepted procedures as a means to control
n not suiuble for aluminium or free zinc coated steels since spot weld quality. Dynamic resistance li the term given io the
there is no drop in voluge up to the expulsion point for these varyingelectrical resistance of the weldment during a spot
materials. A more appropriate method using measurement of weld. The simplest form of a dynamic resistance monitor is
heat input u shown by the area under a voluge-time curve in shown in Figure 4 (9) while a typical output from spot welding
Figure 2 (3) has been suggested for metallic coated steels (8). on uncoated mild steel is shown in Figure 5 along with a
The relationship shown in Figure 3 (3) between heat Input and schematic representation of the nature of the sheet/sheet surface
nugget diameter can be used to pinpoint a weld termination during a single pulse weld. Through feedback control, the
point ie welding can be stopped prior to the nugget reaching a monitor will reject the weld and adjust the welding parameters
minimum nugget diameter. if the trace falls outside of that prescribed as an optimum trace.
T o control weld quality, welding can be stopped when a
particular drop in resistance is seen relative to theresistanceat
the peak of the curve (used for uncoated steels) or stopping at a
time beyond the peak which is proportional to thetimebetween
the valley and the peak of the curve. The latter method is more
commonly employed on galvanized steels.
toctiftciiia,

IkOlMfcO InMftlion VuitM* m— > » •

Umm.

1
i~X^—O stuitbl* C M I lUdilitf

TIME
Figure 4 A n d r e w s et al (9) model of the
simplest form of a dynamic resistance
monitor.
Figure 2 Schematic of d y n a m i c electrical
parameters during spot welding of mild
steel.

i
>
>
K
3
U
Ol
a
S e
ACCEPTABLE
> forciuTiNa
RANOE

MIN MUOGCT OU
<
•
NUSGCT DIAMETER

Figure 3 T h e relationship between weld nugget Figure 5 A typical dynamic resistance curve for
diameter a n d the area under the voltage a single pulse resistance spot weld on
curve produced during resistance spot uncoated mild steel
welding.
4. NUGGET TEMPERATURE Figure 7 (10), along with a brief description of the events that
lake place. Good welds are said to expand at an optimum rate
There have been attempts to relate the temperature of and to an optimum level. Weld current control can then be
the heated surfaces to the temperature at the centre of the adjusted to track these optimum levels. T h e main
nugget providing an Indication of whether full growth had disadvantages with this method are that it can only be applied
occurred. Thermocouples attached to either the sheet surface to pedestal and not gun welding machines because machine
or electrodes provided graphical data of temperature vs time. rigidity is required and it is not suiuble on welds close to edges
Attachment of the probes is time consuming however. This of where pan fit-up is a problem
together with the fact that electrode temperature is dependent
upon the current level and welding time, each of which causes
variations in electrode temperature gradients has given rise to
the non-acceptance of this method in high volume welding
esublishments.

S. INFRARED RADIOMETRY
Monitoring of Infrared radiation intensity emitted from
the sheet surface adjacent to welding electrodes is more suitable
than the temperature method as there is no need to have
detectors in direct contact with the heat source. The intensity
has been shown to be related to weld nugget size and tensile
shear strength (5). Detectors employed are fibre optic cables
mounted either in the welding electrode and transmitting
radiation to an external detector or through infrared detectors
focused at spots on the sheet surface immediately adjacent to Figure 7 A characteristic curve of electrode
the electrode. This is considered to be the more appropriate movement(s) during a single pulse spot
method. Typical results from this method are presented in weld.
Figure 6 (4). The amplitude of the detector trace from small
nuggets is low. These nuggets possess low shear loads. High Area 2-3 Thermal expansion of workpieces causing
amplitude traces indicate weld metal expulsion while electrodes to move apart
uuermediate trace height represents accepuble welds. Tjifrared Point 4 Force equilibrium obtained, electrodes
monitoring systems suffer from the major disadvantage of stationary.
being affected by dirt and fumes essentially eliminating the
method for use with metallic coated steels. Area 4-5 Temperature of fusion zone increases,
electrodes moving together.
Area 5-6 Termination of weld current cooling of fusion
zone leading to contraction, electrodes moving
together.

7. ACOUSTIC EMISSION
az
There are a number of stages in the production of a spot
weld during which acoustic signals are emitted, all of winch
can be detected by a piezoelectric transducer mounted to the
electrode holders of the welding machine. Initial electrode
contact and separation, expansion, recrystaQization, solid state
transformations, plastic deformation, melting, vaporization,
solidification and on occasions weld cracking can all be
detected. A schematicrepresentationof an acoustic emission
trace is shown in Figure 8 (4). Because the signal is highest at
TRACE AMPLITUDE RECORDED FROM INFRARED DETECTOR
weld expulsion (maximum nugget growth), it is possible to
terminate the weld just prior to the onset of expulsion through
feedback control, controlling current There is an obvious
Figure 6 Typical results from an infrared advantage in this of maximised electrode life. General Motors,
monitoring system showing the Wilmington Delaware plant in the U S A has led the way in
relationship between trace amplitude utilizing a system developed by the Trodyne Corporation of
and weld shear strength. N e w Jersey to control resistance spot welds on galvanized
steels valence panels for their Cbevette vehicle.
6. WELD EXPANSION AND CONTRACTION

There are two opposing forces during spot welding, the

natural inward movement of electrodes due to tne applied force
from the welding machine and the fact that as the material
becomes molten, the electrodes are free to m o v e in closer to
each other and opposing this is the force generated o n the
electrodes througn thermal expansion of the weld and
surrounding material. T h e degree of separation of the
electrodes, tracked through displacement transducers can be
related to the volume of molten metal and can therefore be used
as an indication of nugget growth. A typical trace is shown in
 REFERENCES
IX PUL HON
/ I Howes, S.W. and Lake, J.S.H. "Degradation of
Electrodes When Resisunce Welding Metallic Coated
Sheet Steels". Sheet Metal Welding Conference m
Detroit USA, October 1988. Paper No. 10.
2 Howes, S.W. "Electrode Erosion During Resistance
Welding of Metallic Coated Steels". The Fifth
International Conference on Manufacturing Engineering
1990. The University of Wollongong, July 1990, Vol
Ll'iHiii iii""-" iiti.iiiiiitiinuuiiii •TIME

U 3 Lynch, R.F. and Goodwin, F.E "Galfon Coated Steel

SL'.CTROOl / POST WELO » i S T -... 0 ILICTROOI for Automotive Applications". S A E Technical Paper
a*MP0ov»N ,-.'. / NUOOIT Series 860658, February 1986.
IT««T J FORMATION
WCLO TIME
4 Dickinson, D.W. "Welding In the Automotive
Industry". Report on AISI Project N o 1201-409C,
Republic Steel Corporation, August 1981.
Figure 8 Schematic output data of acoustic 3 Beatson, E.V. "An Introduction to Weld Quality
emission during resistance spot Control Systems in Resistance Welding". Paper
welding. presented at the Resistance Welding Control and
Monitoring Seminar, TWI1977.
6 Nakata, S., Nishikawa, M., Kurozumi, Y., Okuna, H.
"Monitoring of Spot Weld Quality By Measurement of
Voluge Between Electrode Tips . Transcript of The
8. U L T R A S O N I C SIGNALS Japan Welding Society, April 1979.
Ultrasonics is high frequency sound above the range 7 Johnson, I.K, "Automatic Spot Weld Correction".
of human hearing (2Q-25kHz per second). Spot weld pool Metal Construction, February 1977.
behaviour comprising of both melting and solidification has an
8 Ganbowski, F.J., Williams, N.T. "Advances In
effect on the amplitude of ultrasonic signals emitted from the
Resisunce Spot and Seam Welding of Zinc-Coated
weld. A n ultrasonic sensor placed in the vicinity of a weld
Steel Strip". Sheet Metal Industries, November 1972.
pool can therefore act as a temperature measuring device. It is
possible then torelateindirectly, the changes of the ultrasonic 9 Andrews, D.R., Batuchrya, S. "Resisunce - Weld
signal to the time of cooling of the weld nugget to some Monitoring For Production". Metal Construction and
predefined temperature (11). British Welding Journal, May 1973.
The most common method involves mounting a
transmitter crystal in the water cooled opening of one of the 10 Stiebel, A., Ulmer, C, Kodrack, D., Holmes, B.B.
spot welding electrodes and mounting a receiver in the other. "Monitoring and Control of Spot Weld Operations".
Ultrasonic signals are then passed through the weld zone as Sheet Metal Welding Conference II, Detroit USA,
welding takes place. Nugget growth is then monitored through 1986.
observation of the change in signal height during the weld time 11 Broomhead, J.H.W., Dony, P.H. "Resisunce Spot
(5). Welding Quality Assurance". Welding and Metal
Fabrication, July, 1990.
Ultrasonic monitoring does have some drawbacks most
of which will have to be overcome before the method is fully
accepted. These include the need for good surfaces to ensure
adequate coupling, lack of signal resolution from thin material
welds, the need to be able to define from the traces a good
from a poor weld, the inability of rather fragile transducers to
stand up to high volume welding and confusion of signals
from extraneous signals emitted during welding of zinc coated
steels.

9. CONCLUSIONS

The techniques described may be used for on-line

monitoring of the resisunce welding process. Perturbations
during the process can be detected employing these sensing
devices and corrected in real-time using adaptive control
techniques. The aim of such systems is to ensure that the
process parameter inputs are regulated adequately during the
course of welding thus, producing therequiredquality output
This technology is now becoming commercially available. It is
anticipated that these systems will penetrate production lines in
the future and form an integral part of the automated/robotic
welding cells in various manufacturing industries.
APPENDIX 3

125
 PROCESSES

Improving resistance
welding of coated steels
When spot welding metallic coated steel sheet, electrode
life Is dramatically shortened compared with spot
welding uncoated steel. However, Stephen Howes,
working In the Research and Technology Centre of BHP
Sheet & Coil Products Division at Port Kembla, has
shown that by careful choice of welding parameters, In
particular electrode tip diameter, electrode life can quite
easily be increased four-fold or more.
C o n s u m e r d e m a n d for increased this steel can be markedly improved.
corrosion protection in manufac- By Stephen Howes A similar erosion process takes place
tured articles has forced m a n y on electrodes used to weld other galva-
sheet metal fabricators, including nised steel types — such as free zinc
those in the automotive and appliance coated, hot dipped galvanised, fully iron
industries, to move towards using m e - alloyed (galvanneal), hot dipped galva-
tallic coated sheet While m a n y ad- nised and electrolytically zinc coated
vances have been made in joining sheet steel — but the small amount or total
steel, such as laser welding, clinching absence of aluminium in the steel coat-
and adhesives, resistance welding re- ing of these steels means the erosion
mains the most popular form of joining. rate is m u c h slower. T h e rate of elec-
It is a relatively forgiving process, and trode degradation is most rapid for Zn-
Figure 1. Resistance spot welding
small deviations from o p t i m u m settings 55 per cent Al coated steel followed by
will not create disastrous results. It is electrode tip profiles. free zinc coated, hot dipped galvanised
also still economically viable. steel and is least for galvannealed and
redressing or replacement is necessary
However, converting from uncoated electrolytically zinc coated steels. T h e
All zinc coated steels will rapidly de-
steels to metallic coated steels requires high iron content of the galvannealed
grade the electrodes to varying degrees
re-education of both shop floor per- steel coating contributes to the lowering
depending on the composition of the
sonnel and design engineers. T o achieve of the electrode degradation rate when
steel coating. T h e increase in welding
the quality of welds already possible welding this steel type.
time and current required to spot weld
with uncoated steels, resistance spot In addition to this erosion process is
galvanised steels provides the driving
welding schedules need to be changed. the gradual build-up of alloy around the
force for alloying to take place between
The Sheet and Coil Products Divisions periphery of the electrode tip. A s the
the copper electrodes and the metallic
R T C is examining resistance spot weld- build-up progresses, the tip diameter of
coating of the steel.
ing to improve the weldability of its hot the conventionally used truncated cone
T h e alloy layers developed for the
dipped and electrolytically coated galva- electrode (Figure 1) increases. In the
electrode working faces are complex in
nised steels. absence of automatic current stepper
nature but it is clear that the presence of
The RTC is also looking at what devices, the current density progres-
aluminium, even in small quantities, in
causes electrode degradation during sively decreases and again weld quality
the metallic coating composition can
spot welding. A 120 k V A pedestal spot deteriorates. T h e use of radiused or
have a detrimental effect on the life of
welder linked to a computer which m o n - domed electrodes should be avoided on
electrodes. Aluminium can form brittle
itors each weld has been used, providing galvanised steels as the rate of diameter
pockets on the electrode face and these
insights into the effects of various weld- change with alloy build-up is most rapid
pockets can be stripped from the face
ing parameters. with this type of electrode. They should
leaving voids on retracting the electrode
only be employed where part fit-up is
Electrode degradation away from the material being welded.
quite poor.
M a n y thousands of spot welds can be This erosion process continues with
each weld and weld quality progressively Parameters under study
achieved from the one set of electrodes
deteriorates. This is typical of Zn-55 per T h e parameters — and their effects
when employed to join uncoated steel
cent Al coated, hot dipped galvanised on spot weldability — under study at
sheet but a change to metallic coated
the R T C are electrode tip diameter,
steel sheet m a y m e a n as little as 200 steel, but it will be seen that through
careful selection of welding parameters, current upslope and electrode tip angle,
welds from a pair of electrodes before
the life of electrodes employed to weld all of which have been the subject of
some controversy.
AUSTRALASIAN WtlDING JOURNAt, fOUTCTH QUARTER 1992
 PROCESSES

The selection of electrode tip diam- ies of current distribution ncnmtt an machined electrodes and the average
eter dependa on a number of factors, electrode face suggest that aa electrode result was determined. T h e end point of
including the size of the weld required, angle gets smaller, tip life improves. each life test was determined by pe-
the sheet thickness and the design of Bowers and Eager found that current riodically measuring the weld nugget
the component being welded — eg, density was progressively more uniform size across major and minor axes and
minimum flange widths m a y be desir- across the working face of the electrode calculating the average diameter. W h e n
able. with decreasing angle, and hence elec- the nugget size diminished to a diam-
Current upslope is the gradual in- trode wear was more uniformly spread eter equal to 80 per cent of the original
crease of the weld current magnitude across the face. It will be seen, however, electrode tip diameter employed, the life
during the initial period of current flow that this is not the case for all metallic of the electrode was deemed to be over.
in each weld. Introducing this into the coated steels. The four cycle current upslope was ar-
welding schedule assists in melting and Experimental procedure bitrarily chosen as an acceptable time
displacing the metallic coating, enabling T h e materials examined in the R T C for high volume welding. Determining
the electrodes to gradually bed into the study were all of a 0.60 m m base thick- the expulsion point for each life test
sheet surface before m a x i m u m current ness. T h e electrodes used were R W M A took approximately 10 to 15 welds.
is reached. This technique is believed to Class II (Cu-0.8 per cent Cr) truncated These were included in the weld count.
reduce alloying rates between the me- cone type. T h e effect of parameter Results
tallic coating and electrode as electrode/ change on weldability was measured The results are presented graphically
workpiece heating is minimised. While through electrode life testing conducted in Figure 2. While maintaining the same
the truncated cone electrode profile is at 2.0 k N electrode force, seven cycles electrode force and welding time, a
generally accepted as the optimum con- (50 H z ) weld time and a weld current change in tip diameter from 5 to 7 m m
figuration for spot welding galvanised just below the expulsion level for elec- when welding the galvannealed and elec-
sheet, there has been considerable inter- trode configurations of 120 and 90 trogalvanised materials produced det-
est in the effect of electrode angle on (shown as x in Figure 1) and (i) 5 m m rimental results. In particular, the elec-
weldability. Bowers and Eager (Non tip diameter, (ii) 5 m m tip diameter and trogalvanised material could not be
Uniform Current Distribution in Spot four cycle current upslope and (iii) 7 welded successfully at a 7 m m tip diam-
Welding. Conference Proceedings, Sheet m m tip diameter. eter. Weld nugget sizes were well below
Metal Welding Conference II, Detroit, A number of life tests were con- the minimum level of 80 per cent of the
US, October 1986) through their stud- ducted at each setting using freshly electrode tip diameter even at the start

CHECK-OUT E * FOR PERFORMANCE

E* FOR VALUE
IE* FOR RELIABILITY
F r o m C E M — o n e of Italy's leading
manufacturers of welding a n d
plasma cutting equipment — a full
range of models designed to cover
every need — from panel repair
shops to m e d i u m a n d h e a v y
fabrication shops.
Value-engineered throughout to
CEMIG 35 POWER PLASMA provide m a x i m u m performance per
3-phase MIG unit with separate wire kilo.
feeder.
Automatic wire speed control — plus
PL5S Offering proven relaibility.
• Uses compressed air as plasma
interval timer.
gas for high economy.
350A at 5 0 % duty cycle. TO CHECK-OUT CEM ON
• Cuts mild steel u p to 1 6 m m .
• Cutting current — 50A. PERFORMANCE A N D PRICE
Call n o w for your nearest supplier:
CALL NOW!
Broadway Welding Supplies Unit 8/400
21 Meeks Road 23 Market Street Cnr Kerry & Beaudesert Grand Junction Rd
Marrickville Sunshine Rd, Coopers Plains Mansfield Park
Sydney Melbourne Brisbane Adelaide
Phone (02) 517 2133 Phone (03)311 8011 Phone (07) 875 1122 Phone (08)3472003
Fax (02) 516 2161 Fax (03) 311 8799 Fax (07) 274 0014 Fax (08)3470576

AUSTRALASIAN WE10ING JOURNAL, FOURTH QUARTER 1992

 PROCESSES

cent improvement in electrode life when

usinK the name electrode lip diameter.
HOT DIPPED GALV. Zn-55% Al COATED N o explanation can be offered as Ui why
this is the only one of the material* to
iss i2o-i.p ma vo*up IW 120'Hp Hi 90'iip benefit from upslope but It could be
associated with the relatively high iron
1400 1400 content and low aluminium levels in the
coating of this steel type.
1200 1200 It is possible that the four cycle time
period chosen may not be optimum and
1000 that even greater improvements can be
1000
achieved with longer times. Productiv-
ity considerations would have to be tak-
800 800 en into account when selecting upslope
timing for use on the shopfloor.While
600 600 the introduction of upslope was not
detrimental to electrode life when weld-
400 400 ing the remaining materials, the results
indicate that there would be no benefit
200 gained from employing it.
Bowers and Eagers findings were
generally found to hold true in regard to
A S C A B C lower electrode angles being the better
TIP CONOITION TIP CONOITION performers except when welding the
ELECTROGALVANIZED GALVANNEALED electrogalvanised steel. In this case,
electrode life was found to deteriorate
120*tip vO'tip l20Mip 90'lip considerably when changing from elec-
trodes with an included angle of 120 to
9000 90 deg. This anomaly has no obvious
8000 explanation. It is obvious that during a
lifetest,nugget size oscillates as the
7000 alloy builds up around the periphery of
the electrodes, reducing current density
„ 6000 and then shearing away when it can no
t longer be tolerated by the electrode.
1 5000 Current density then rises again and
o weld size recovers.
o
It is conceivable that electrodes with
I 4000
t lower angles remove the build-up more
3000
easily due to the sharper angle of in-
2000 cidence with the sheet surface, and
1000 hence have longer life. This could ex-
plain w h y pimple tipped electrodes
(Figure 1) with vertical sides are far
superior to conventional truncated elec-
TIP CONOITION TIPCONOmON trodes. It is unfortunate, however, that
KEY: A:5mm TIP, B:7mm TIP. C.-5mm TIP & U/SIOPE. this configuration is not readily adapt-
Figure 2. Electrode lite test results. able to production welding as it is easily
damaged during set up and lacks
of thejUe test and at a weld current just force was increased to 2.5 to 3.0 k N to mechanical strength.
below the expulsion point. Irregular ex- assist the electrodes to bed in, then far Discussion
pulsion from the sheet outer surfaces is better results would be achieved. Studies are continuing an the inter-
thought to be the major cause of this Similarly, a decrease in tip diameter action of metallic coatings with other
inconsistency. It could be that the small to 6 m m would produce better results. electrode compositions including Cu/Zr,
sine coating mass of approximately 5 g/ There was only a marginal improvement Cu/Cr/Zr and aluminium oxide disper-
m' on either side is quite sensitive to in electrode life with increasing tip di- sion strengthened electrodes — all of
even small electrode misalignment. ameter during welding of the hot dipped which are currently available in the
That is, if the electrodes are not in firm galvanised material but the same tip marketplace — and with other experi-
contact with the sheet over the full 7 diameter change produced over a four- mental compositions. Many of these
m m diameter (and electrode alignment fold increase in electrode life during studies will take place at the recently
becomes increasingly more difficult welding of the Zn-55 per cent Al coated established Cooperative Research Cen-
with increasing diameter), then the zinc material. This increase was observed tre for Materials Welding and Joining,
coating is easily blown away. Energy regardless of the electrode tip angle using the facilities of the CSIRO Div-
expended in doing this would mean less employed. ision of Manufacturing Technology, the
energy was available for weld nugget Current upslope was found to be University of Wollongong and BHP's
development at the faying surfaces beneficial only when welding the galvan- Technology Cemtre at Port Kembla.
(sheet/sheet interface). If the electrode nealed material. This produced a 30 per Stephen Howes is senior research office
AUSTRALASIAN WEIDING JOURNAl, FOURTH QUARTER 1992 at BHP Steel, Sheet & Coil Products
Division, Port Kembla.
APPENDIX 4

126
 RESISTANCE W E L D I N G — REFURBISHING A M A T U R E T E C H N O L O G Y

by Stephen Howes
B H P Steel — Sheet &. Coil Division — Research & Technology Centre
and
Elios Siores
University of Wollongong — Department of Mechanical Engineering

ABSTRACT when spot welding metallic coated surfaces (galvanized

steels), higher welding currents and welding times must
Resistance welding is the most widely used joining be employed to produce welds of similar size to those
process for sheet steel and is currently experiencing a being produced when uncoated steels of equal thickness
resurgence in today's quality conscious environment. are joined. These higher currents and times provide the
The fact that it is relatively simple in operation is a cost driving force for alloy formation and build-up around the
effective technique and can be quality controlled even at periphery of the electrode, consequently affecting current
the high welding rates employed in industries, such as the density and promoting the gradual erosion of the
automotive and appliance industries, makes it appealing to electrode's working face. The rate of this electrode
production engineers, designers and equipment degradation is dependent not only on parameter selection
specifiers. The increasing trend towards employment of but also on the chemical composition of the metallic
metallic coated steels however, presents problems not coating of the steel sheet*2), the electrode composition*3-4)
normally encountered in the past when resistance welding and geometry*5).
uncoated steels. Accelerated electrode degradation is the In the past it has been standard practice to monitor
major concern and if left unchecked can result in inferior weld quality using destructive testing methods such as the
weld quality. The drive for higher productivity and peel, and hammer and chisel test. These are considered to
quality levels has also forced high volume production be time consuming, cumbersome and unproductive
industries to turn to in-process control facilities. These techniques. When weld nugget sizes reach the predefined
requirements have led to the development of sophisticated minimum (approximately 8 0 % of the electrode tip
but user friendly microprocessor controlled systems diameter employed), electrodes are redressed either by
capable of both monitoring and controlling the process special redressing tools or in most cases byfilingthe
input and output parameters. This paper concentrates on electrode working faces to the flat position again. This
developing two distinct techniques for monitoring weld latter practice is not conducive to the production of high
nugget size during resistance spot welding of metallic quality welds as it often leads to misalignment of
coated steels. These are surface profilometry (tactile electrode faces. While the above quality monitoring
sensing) and machine vision (remote sensing). The aim procedures are still widely used, they can be grossly
of this work is to encourage medium sized manufacturing inaccurate and produce misleading results.
industries in Australia to implement such alternative cost During peel testing, thin sheet steels, in particular,
effective production solutions for gaining enhanced have a tendency to pull spot welds that have a tail.
product quality through on-line process optimisation. Because of this, vernier callipers cannot record a true
1. INTRODUCTION measure of weld nugget diameter. Through this
The resistance spot welding process in general, is a measurement technique, welding could be allowed to
relatively forgiving joining technique with relatively small continue even though the minimum nugget size has
parameter changes not creating disastrous results. This is already been reached. Too frequently, damage is caused
certainly the case for uncoated sheet steel where welding to the welds using the hammer and chisel technique thus
lobe widths can be quite broad and there is some degree making nugget size measurement again difficult to
of freedom in parameter choice. However this does not observe. It is also worth noting that rather than destroy a
apply to the growing number of metallic coated steels. component, some practitioners use scrap sheet to
Welding parameter selection can be critical for optimum determine or assess weld size. Placing scrap in the
electrode life and some parameters are unique to certain welder of different size to the actual component can lead
metallic coated steels'". Welding lobes are to inaccuracies as the effect of current shunting from other
characteristically narrower for this group of steels thus welds in the actual component needs to be accounted for
requiring more stringent quality monitoring and control. as well as the effect of placing various dimensions of
The tighter the lobe, the easier it is to exceed acceptable scrap in the throat of the welder.
conditions. Low volume high technology industries such as the
Due to the lower contact resistance encountered aerospace industry have for some time realised the benefit
of employing resistance spot weld on-line monitoring collect resistance spot weld samples during a single
devices and Tn m a n y cases feedback control devices to electrode life test on a metallic coated steel and then to
sircnmlinc production welding without the need for establish the relationship existing between weld surface
destructive testing of weld joints. In recent times, the indentation or impression diameter measured by the two
drive for increasingly higher productivity and quality systems and actual weld nugget diameter, Carbon paper
levels has also forced higher volume industries such as Imprints of the electrode tip lace were also taken to track
ihe automotive and appliance industries to follow suit. electrode wear. The objective is to establish the feasibility
Sonic of these are being used successfully overseas, but of utilising the vision system only for on-line weld quality
many arc unsuitable for metallic coated sheet steels for monitoring and at a later stage incorporate this into a
one reason or another as discussed below. feedback control loop to maintain weld quality In real
2. ON-LINE MONITORING TECHNIQUES time. The profilometer and carbon paper imprints were
DURING SPOT WELDING used essentially as a check on measurements obtained
through the vision equipment.
On-line monitoring techniques during spot welding The profilometer was a machine in use at BHP's
have taken m a n y forms but all basically look at some Research and Technology Centre, Port Kembla. The
phenomenon or characteristic which occurs or changes as vision system w a s an advanced grey scale facility
a result of the start and growth of a weld and whose [irovided b y the Illawarra Technology Corporation,
output bears a definite relationship to the weld diameter, ocated at the University of Wollongong.
(his being taken as a production criterion of weld
The profilometer works on monitoring vertical
quality*6). A n u m b e r of process parameters and
movements of a stylus being dragged across a surface, in
techniques have been identified as suitable for monitoring
this case, the indented spot weld surface left after
weld quality all of which provide a non-destructive means
removing the electrode from the sheet surface.
for checking welds. These parameters include:
Correlation of horizontal and vertical movements of the
(a) electrical parameters (voltage, current, resistance stylus then provides a direct record of the surface profile
and energy) limited in resolution primarily by the geometry of the
(b) nugget temperature stylus tip. The technique is quite slow in comparison
(c) infra-red radiometry with non-contact measuring systems but it does allow for
(d) weld expansion and contraction a more critical look at the shape of the welded surface if
(e) acoustic emission, and needed. Measurements of indentation diameter were
(f) ultrasonic signals (reflection, attenuation and taken in two directions and the average diameter was
velocity) recorded.
Monitoring of electrical parameters has proved to be The vision system comprised of a CCD (charged
one of the most successful however these systems are couple device) camera, mounted on a tripod and placed
generally used only on uncoated steels and for sensing the directly over the welds for visual imaging on a monitor.
input variables of the process. Tsai et al*7) have reported Again, measurements in two directions were taken with
that resistance or voltage characteristics of coated mild the aid of inbuilt image analysis and processing software
steel sheet have poor correlations to weld size. and the average diameter was recorded.
Attachment of thermocouples to the workpiece for 3.1 Experimental Procedures
measuring weld temperature, not only interrupts the
welding procedure, but can b e inaccurate due to The material chosen for this study was a
temperature gradient variations in the electrodes resulting galvannealed product of 0.80 m m base thickness and a
from periodic changes in current level and weld time*8). coating mass of 80 g/m2. Composition of the base steel
Also zinc fumes generated during welding of zinc coated and coating are given in Table 1. The product is a hot
steels con cause spurious results when infra-red detectors dipped galvanized steel which, on leaving the coating
are employed to monitor process output. Displacement bath, enters a furnace which keeps the zinc coating molten
transducers used to monitor weld expansion and for s o m e time to promote alloying of the zinc with iron in
contraction phenomena and therefore weld volume, can the substrate. This type of material,richwith iron in its
only be employed o n rigid pedestal type welding coating, is generally characterised by excellent resistance
machines. Acoustic emission and ultrasonic based spot weldability and is widely used in the automotive
monitors have had s o m e success on coated steels*^) but industry for inner door and floor panels, wheel housings,
sensors are expensive and quite fragile. Incorporation of firewalls, and other areas of application. A typical
these sensors into the welding systems usually requires welding lobe for this material is shown in Figure I.
expensive redesign of the electrode assemblies. The electrodes employed were of the RWMA Class
3. NEW APPROACHES USING II (Cu/0.8% Cr) truncated cone type machined to an
included angle of 90* and a tip diameter of S.O m m .
PROFILOMETRY AND MACHINE
VISION TECHNIQUES
This work evaluates two further non-destructive,
weld quality monitoring techniques, profilometry and
vision sensing. T h e basic approach to the study was to
 T A B L E I — C H E M I C A L C O M P O S I T I O N O F TEST M A T E R I A L

Base Chemistry

c P Mn Si s Ni Cr Mo Cu SolAI Sn Nb Ti
.07 .017 .17 .010 .013 .022 .015 .005 .012 .054 .006 .008 .086

Coating Composition

%A1 %Pb %Fe %Zn

Top surface .16 .05 9.33 90.4
Bottom surface .18 .06 9.75 90.0

A 120 k V A pedestal welder was used interfaced 3.2 Results Obtained

with a computer for continuous monitoring of process
input parameters including voltage, welding current, The electrode useful life was considered to be over
electrode force and energy input. when weld nugget diameterreacheda size equivalent to
8 0 % of the original electrode tip diameter, ie when the
nugget size diminished to a diameter of 4.0 m m . This
occurred after approximately 4600 welds were made on
strips, 1 m long by 40 m m wide. Results from carbon
paper impressions left by the electrodes are presented
graphically in Figure 2. Degradation of the electrodes
took place rapidly up to approximately 500 welds.
During the period up to the failure point, electrode
diameter measurements oscillated between a low of
approximately 4.3 m m and a maximum of 5.9 m m , the
period of oscillation being approximately equal to 500
welds. After 4400 welds, electrode diameter grew
rapidly to 5.9 m m , at which point nugget sizereachedthe
minimum acceptable diameter.
W»id Tim* (cyclit)

Figure 1: Weld lobe 0.8 mm ZINCANNEAL® G3N

ZF80
Monitoring of welding conditions, electrode wear
and the collection of weld samples for submission to the E
profilometer and vision equipment took place during an 5
electrode life test run. Welding took place at a current just
below the metal expulsion point and at a weld time of nine
cycles (50 Hz). A n electrode force of 2.5 k N was
employed with a squeeze time of 25 cycles and an off
time of 40 cycles (time intervals between welds).
Welding current was not adjusted during the run and no 1000 2000 3OO0 4000 5000

electrode conditioning procedures were applied prior to Numb*r of W*Mi

commencement of the life test. At selected points
throughout the test, electrode wear was tracked by placing Figure 2: Carbon paper impression diameter results
carbon paper between the electrodes and applying the
weld force without the welding current turned on. Tip 3.2.1 Profilometer Measurements
diameter was then measured from the impressions left by
the carbon paper using a graduated eyepiece. Weld The stroke of the stylus was adjusted to scan
samples taken at these points were then used to collect 10 m m across each weld impression in both the
indentation profile measurement using the longitudinal and transverse directions. A typical trace is
abovementioned two pieces, of apparatus, namely shown in Figure 3 (weld 450 transverse direction).
profilometry and vision sensing, prior to undergoing a
peel test for manual measurement of weld size using
conventional vernier callipers.
 I
LL
nil it
II li
; l:1: *:
1
Hi i
m i* •
mS • « k

1 --1* 1«
I •• •«
M
ii": :i i *"
u;: a I — • -
Figure 3: Typical profilometer trace

Peak to peak impression diameters were measured from 'maturity' phase and a 'termination' phase.
both longitudinal and transverse directions on each weld
,INN|H
and the average diameter for each was calculated. T h e
variation in average impression diameter over the duration
of the life test is presented graphically in Figure 4. _ •
E •• • • •
E,
7
I
s
E
>
<

1000 2000 3000 4000 5000

Number ol Welds
Figure 5: Vision system diameter results
1000 2000 3000 4000 5000
The vision system appeared to provide the clearest
Humb.f el W.ldt indication of these phases with a wear-in phase between
0-750 welds, half-life between 750-2000 welds, maturity
Figure 4: Profilometer impression diameter results between 2000-4000 welds and termination between 4000-
4600 welds.
3.2.2 Vision Measurements
During wear-in, a rapid acceleration of tip diameter
Longitudinal and transverse weld impression growth is observed as alloy is quickly accumulated
measurements were recorded in pixel spacings and then around the electrode tip periphery. This acceleration can
converted to distances in millimetres. T h e average also be observed from carbon paper impressions of the
diameter for each impression w a s then calculated. T h e electrode working face (wear-in period of 500 welds) and
results again are represented graphically in Figure 5. from profilometer measurements (wear-in period of 300
welds). T h e most consistent feature in this case,
4. DISCUSSION AND CONCLUDING regardless of the system applied is the point of half-life, ie
REMARKS the point at which the average tip diameter reduces to
Over the effective life span of an electrode used to some m i n i m u m point. This was recorded at 2200 welds
resistance spot weld galvanneal of the type used in this using carbon paper imprints, 2100 welds using the
study, distinctive electrode life phases could be identified. profilometer and 2000 welds using the vision system.
Depending o n the monitoring or measurement system Maturity is represented by the period of recovery
applied, the position of these phases, although all after this half-life m i n i m u m and runs for approximately
recognisable, vary to some extent. These phases can be the same duration as the half-life. Even in this period,
referred to as a 'wear-in' phase, a 'half-life' phase, a cyclic growth and recovery oscillations take place as is
observed over the half-life period. These oscillations are REFERENCES
thought lo be associated with removal and build-up of
alloy at the electrode tip. (I) Howes, S.W. Electrode Erosion during Resistance
Welding of Metallic Coated Steels. 5th
While it is evident that there is some scatter between International Conference on Manufacturing
the systems, it is difficult to determine the system which Engineering. The Institute of Engineers, Australia.
is the most accurate in predicting actual weld nugget size University of Wollongong II-13th July. 1990.
as each has unique measurement difficulties.
(2) Howes, S.W., Lake, J.S.H. Degradation of
When using the carbon paper method, any alloy Electrodes when Resistance Spot Welding Metallic
build-up at the tip periphery would also be imprinted and Coated Sheet Steels. Paper N o 10, A W S Sheet
therefore would be included in the measurement of tip Metal Welding Conference III, Detroit, U S A ,
diameter. However, these imprints are taken in the October 1988.
absence of welding power and it is possible that with the
power applied to make subsequent welds, the build-up (3) Assel, M.D., Nadkarni, A.V. Non-Dressing
could be removed. A false indication then of resulting Dispersion Strengthened Copper Electrode Designs
weld nugget size m a y be recorded. The profilometer on for Galvanized Steel Welding. Paper N o 9, A W S
the other hand can only measure peak to peak heights Sheet Metal Welding Conference 111, Detroit, U S A ,
across the weld impression diameter left on the sheet October 1988.
surface and therefore does not record the actual welding (4) Kimchi, M., Gould, J.E. The Evaluation of
face diameter. Similarly, the vision system relies on Resistance Spot Welding Electrode Materiols for
measurements taken across the line of highest contrast. Welding Galvanized Steels. Paper N o 8, A W S
This occurs on the outer periphery of the weld impression Sheet Metal Welding Conference N o III, Detroit,
and so once again, a false impression of actual nugget U S A , October 1988.
size may be recorded.
(5) McGregor, G. The Resistance Spot Welding of
Further light will be shed on the accuracy of these
Metallic Coated Steels. Paper presented at
techniques once ultrasonic measurements and X-ray
Australian Welding Institute and The Australian
measurements on the welds have been conducted and
Institute of Metals Seminar, Resistance Welding,
actual weld nugget sizes have been measured with
Adelaide 1982.
callipers after peel testing. Further research will also
attempt to correlate the size and shape of the nugget (6) Beatson, E.V. An Introduction to Quality Control
measured using the vision sensing and spot weld shear Systems in Resistance Welding. Paper presented at
strength during the abovementioned electrode life phases the Resistance Welding Control and Monitoring
as this is a critical requirement in design of m a n y spot Seminar, T W l 1977.
welded components.
(7) Tsai, C.L., Dai, W.L., Dickinson, D.W., Papritan,
In summary, from these early results, it appears that
J.C. Analysis and Development of a Real-Time
the vision system could be used for on-line measurements
Control Methodology in Resistance Spot Welding.
of electrode degradation and prediction of the useful life
Welding Research Supplement pp 339-S to 351-S,
of electrodes. The ultimate aim is to develop a vision
December 1991.
system that is capable of predicting electrode life and
bond strength using image analysis techniques during the (8) Beevers, A. Temperature Measurements to
process ofresistancespot welding. Used in conjunction Determine Post Weld Heat Treatment Conditions
with automatic electrode tip dressing tools, this will be a for Spot Welds. British Welding Journal, pp 173-
useful tool in removing the need for operator judgement 183, April 1963.
while at the same time minimising downtime in order to (9) Gedeon, S.A., Sorenson, CD., Ulrich, K.J.,
replace or redress electrodes. Productivity and weld Eager, T.W. Measurement of Dynamic Electrical
quality will then be enhanced. and Mechanical Properties of Resistance Spot
Welds. Welding Journal, pp 378-385, December
1987.
APPENDIX 5

127
OPTIMISING RESISTANCE SPOT WELDABILITY
OF METALLIC COATED STEELS
THROUGH ELECTRODE CHEMISTRY

Ross Finlay*, M . Samandi* and S. Howes +

•University of Wollongong, NSW, AUST.

+BHP STEEL, Sheet and Coil Products Division,

Port Kembla, N S W , AUST.
ABSTRACT
Resistance spot welding is the preferred joining technology in establishments producing high
volume manufactured goods such as automobiles and uppliunces. Consumer demand for longer
life products and extended warranties has resulted in steel suppliers developing a gamut of
metallic coated steels with various coating compositions for use within these industries.
Consequently, manufacturers n o w require a much broader knowledge on the optimum welding
parameters for each steel type. This work examines the resistance spot weldability aspecLs of
three c o m m o n metallic coated sheet steels namely aluminium - 4 5 % zinc coated, hot-dipped
galvanised and galvanneal steels using commercially available electrode chemistries such as
copper chromium (Cu/Cr), copper chromium zirconium (Cu/Cr/Zr), and copper zirconium
(Cu/Zr). T h e aggressive nature of the erosion process encounicrcd for aluminium-45% zinc
coated and hot dipped galvanised steels masks any influence the alloying elements have on
electrode life. The benefit of alloying elements is apparent in life tests for galvanneal steel
where the erosion process is more gradual. The composition of the Cu/Cr/Zr electrode is a
compromise between the high hardness of Cu/Cr electrode composition and the increased
conductivity of the Cu/Zr electrode composition and proved to be the most effective in obtaining
m a x i m u m life.
1. INTRODUCTION
1.1 BACKGROUND
Resistance spot welding is a method of joining sheet steel by the passage of a high current
through the workpieces to be joined, via a pair of clamping elecuodes resulting in a weld nugget
at the point of highest resistance. This occurs at the sheet to sheet interface or 'faying surface'.
Electrodes must possess high electrical conductivity to minimise electrode heating, and high
thermal conductivity to extract heat from the contact area between the electrode tip and the
workpiece. Electrodes should also exhibit high hot strength to resist softening caused by high
welding forces, temperatures and compressive stresses. In addition to maximise electrode life,
no alloying reaction should occur between electrode and sheet metal.
The significant increase in the usage of metallic coated steels has raised new concerns over
electrode deterioration. The increased welding currents and times required to weld metallic
coated steels over those employed on uncoated steels results in alloying of the metallic coating
with the copper electrodes. This alloying frequently causes non uniform pitting of the electrode
face as well as electrode edge alloy build up. Generally, the shape, dimensions and surface
condition of the electrode tips deteriorate when welding these steels. Current density and
distribution during welding are greatly affected by such variations in electrode condition which
affects nugget development. The tip life of electrodes used to weld metallic coated steels may
be as little as 250 welds, while electrodes used for welding uncoated steels m a y exhibit a life in
excess of 50,000 welds. The composition, geometry, operating temperature, welding force and
tip condition of the electrode plays a critical role in the spot welding of metallic coated steels,
since the alloying reactions that occur between the coating and the electrode ultimately
determine electrode life.
In an effort to combat the rapid deterioration of electrode tips, numerous new electrode mate
have emerged to weld metallic coated steels. These include Cu/Cr, Cu/Zr, Cu/Cr/Zr and
aluminium oxide dispersion strengthened copper.
1.2 WELDABILITY TESTS
The weldability of sheet steel is altered by changes in welding parameters and electrode set
and can therefore be characterised by a welding lobe which displays the range of welding time
and current settings within which satisfactory welds may be made.
 Kxpulslon
16

2 M
•m^*
mmt

e
E "
9 Minimum Nugget
u Dlunivtcr

"5

6 8 10 12 14 16 18
Weld Time (cycles)

Figure 1 Typical welding lobes for 1.2mm thick uncoated and galvanised steels [1].

The lower curve represents the conditions at which a specified minimum nugget diameter will
form while the upper curve represents the conditions at which metal expulsion will occur. A
wider lobe is more useful since it allows a greater range of currents to be used. Welding lobes
which display low current values are also beneficial in that, lower currents result in lower power
consumption and hence reduced welding costs. In addition, lower heat input reduces the rate of
electrode degradation. Generally welding lobes of metallic coated steels are higher than those of
uncoated steels due to the increased currents and welding times because of the lower contact
resistances. The exact position and size of welding lobes is dictated by electrode geometry,
electrode tip condition, electrode force, sheet and coating thickness, coating composition, test
method and analysis of results.
An electrode life test aims to define the point at which weld quality deteriorates to an
unacceptable level, ie., w h e n the electrode needs to be redressed or replaced. Although there is
no universal standard, it is c o m m o n l y accepted that a weld nugget diameter which falls below
8 0 % of the initial electrode tip diameter is unacceptable.

1.3 ELECTRODE COMPOSITION

The most commonly used electrode composition in Australia is Cu-0.8%Cr (J47 Alloy) since it
exhibits the best all round properties. There is however, a growing tendency to use different
types of electrode material to increase electrode tip life. The traditional Cu/Cr material has high
hardness and moderate electrical conductivity. T h e Cu/Zr material has lower hardness but
increased electrical conductivity. T h e alloy addition is about 0.3 w t % Zr which is present as
small precipitates of a Cu-Zr intermetallic. The zirconium is thought to have the effect of
inhibiting pick up on the electrode surface as well as reducing the grain boundary penetration
[2].The Cu/Cr/Zr material has a hardness value and electrical conductivity which lies midway
between Cu/Cr and Cu/Zr. T h e alloy addition is about 1.0 w t % Cr and 0.05 w t % Zr which are
present as small intermetallics.
Aluminium oxide dispersion strengthened copper has a higher strength than the standard spot
welding electrode materials which reportedly lessens electrode tip deformation. Dickinson [3]
and Jud [4] report electrode lives of four times that of standard electrodes at elevated
temperatures. The major disadvantage with this type of electrode is a four to five times increase
in cost over Cu-Cr electrodes. Refractory based metal inserts including powder composites of
molybdenum copper or tungsten copper containing a discontinuous refractory phase in a copper
matrix arereportedlo increa.se electrode life, although at substantially greater cost |5.6|. Sailo el
al |71 trialed a compound electrode with an insert. The core of the compound electrode was
made from a conventional chromium containing copper base alloy and was covered with pure
eopper of lower strength. This type of electrode was found lo last twice as long as conventional
eopper base alloy electrodes, used to weld zinc vapour deposited sheet sieel. The electrode tip
was characterised by a convex wear profile which maintained u relatively small face diameter
and extended electrode life.
Many of the reported improvements in tip life due to changes in electrode material are fo
particular conditions and as a result they are rarely achieved in other environments, especially in
industry (8J.
2. EXPERIMENTAL PROCEDURE

2.1 ELECTRODES

The electrodes used in this program were 90-degree (included angle) truncated cone type
electrodes each with a tip machined to 5.0 ± 0.1mm diameter. This type of electrode was
utilised for both weld lobe construction and electrode life tests. The composition and properties
of the electrodes tested are typically
ALLOY HARDNESS ELEC. COND.
Cu - 0.8%Cr 174Hv20 80% IACS
Cu - 0.3%Zr 152HV20 90% IACS
Cu - 1.0%Cr - 0.05%Zr 160Hv20 85% IACS
Table 1. Electrode alloy composition and properties. Electrical conductivities expressed
percentage of the conductivity of the International Standard for annealed copper (I.A.C.S.)

Induded
Angle

Figure 2. Spot welding electrode and electrode tip configuration.

Electrode tips were aligned to maintain consistent current density across the working
electrode. Carbon paper imprints of the actual contact area were obtained by bringing the
electrode tips into contact with the carbon paper in between.
22 MATERIALS
The materials used for this work were 0.8 mm base thickness, metallic coated steels, su
B H P S T E E L , Sheet and Coil Products Division. Three classes of coated steels were
investigated, namely;
1. hot dipped galvanised
2. aluminium-45% zinc coated (hot dipped)
3. galvanneal (zinc/iron alloyed)
Chemical compositions and coating weights for each steel are presented in Table 2. The
thicknesses are not consistent for all three samples. Kimchi[9] reports that differences of 2:1 in
thickness m a k e only a small difference lo electrode life. It is however, acknowledged that
increased coaling weight has a detrimental effect on electrode life and since hot dipped
galvanised and aluminium-45% zinc coaled steel products are generally heavier in coating
weight lhan galvanneal steel, the electrode tip life is indeed shorter with them.
2.3 SPOT WELDING APPARATUS
Weld lobe construction and electrode life testing was conducted on a Schlatter 120 kVA pedes
style, spot welder. The current values (RMS), electrode force, voltage and energy are monitored
via an I B M personal computer which is interfaced with Ihe welding machine. The welding
machine controller settings employed throughout this work were: squeeze time 25 cycles, hold
lime 3 cycles and an off time period between welds of 40 cycles. Squeeze lime is the time
allowed for the electrodes to reach m a x i m u m clamping force. Hold lime is ihe time for which
electrodes remain in contact with the workpiece afler the current is turned off. The clamping
force on the electrodes was 2.5 ± O.lkN.
2.4 WELD LOBE CONSTRUCTION

Triplicate tests were performed for each electrode material. The lobes are constructed betwe
and 13 cycles (1 cycle/sec = 1Hz) at 2 cycle increments. This represents the most practical and
economical welding schedules for the steels to be used. At each weld time the current is stepped
up in increments of approximately 0.2 - 0.3kA until the m i n i m u m acceptable weld nugget
diameter is reached. The destructive peel tesl is used to assess average weld nugget diameter.
Once the first minimum nugget diameter has been recorded two subsequent tests in four at the
same current level must also display a weld nugget greater lhan or equal to the acceptable
minimum. The recorded current value on the lobe is the average of the three current readings.
The weld lobes plotted represent the average of three repeated lobes. The point of expulsion, for
each weld time is located by increasing the weld current in increments of 0.2 - 0.3kA until
molten metal is blown from the weld interface. Then at the same current level two subsequent
tests in four must demonstrate expulsion. The recorded current value is the average of these
three readings.
2.5 ELECTRODE LIFE TEST
An intermediate weld time of 7 cycles was chosen for all life tests as it produces optimum
performance in each electrode. Long weld times create higher tip temperatures and increase the
amount of coating metal diffusion into the electrode, while short weld times result in a more
severe heating and cooling cycle increasing alloy build up which reduces current density. To
avoid expulsion, which results in rapid electrode wear, the welding current was set just below
the expulsion level for each steel sample. Triplicate life tests were carried out for each material.
Each life test is carried out by spot welding strips l m long by 4 0 m m wide. Electrode
degradation was monitored through peel tests conducted at 100 weld intervals. As the weld
nugget diameter approaches 8 0 % of the original tip diameter, peel tests are conducted every 40
welds. Termination of the electrode life test is established once the averaged weld nugget
diameter falls below the acceptable minimum.
2.6 METALLOGRAPHIC EXAMINATION
After 1000 welds three electrodes were sectioned along their axis, mounted and prepared
metallographically. The alloy layers which had developed on the electrode tips were examined
using a Leica S440 scanning electron microscope ( S E M ) and then analysed using energy
dispersive spectroscopy (EDS). The analytical technique was standardless semi-quantitative
analysis. Contaminants and trace elements were ignored and the results were normalised to a
total of 1 0 0 % .
3. RESULTS
3.1 WELD LOBES
Triplicate tests were performed to eliminate any unrelated scalier points. Figure 3 shows the
weld lobes for hot dipped galvanised steel for each of the electrode compositions. T h e current
values are an average of the three tests. Figures 4. and 5. display ihe same information for
aluminium-45% zinc coated and galvanneal steel respectively. The lobes appearing in Figures 3,
4, & 5 all exhibited the conventional shape, ic. decreasing weld time required an increase in
current and vice versa. Il was observed that the weld lobes exhibited a high degree of scatter. In
some cases the scalier was over a range of approximately lkA.
The weld lobes constructed for hot dipped galvanised and aluminium-45% zinc coated steel
were almost identical in terms of lobe position, width and current range (Figures 3 & 4). This is
to be expected as both steels undergo similar hot dip coaling processes. The weld lobes for
galvanneal steel however, lie approximately lkA below those of both aluminium-45% zinc
coated and hot dipped galvanised. This coating is fully alloyed and has a high iron content
which reduces the current required. This improves weldability of ihe product as lower currents,
reduce power costs and decrease the amount of thermal cycling which reduces the rate of metal
diffusion and alloy build up. The weldable current range for galvanneal sleel is very narrow,
approximately 0.75kA, compared to approximately 1.2kA for hot dipped galvanised and
aluminium-45% zinc coated steels. Comparison between the alternative electrode compositions
show discernible difference in selecting one composition over another in terms of power
consumption or welding times.
3.2 ELECTRODE LIFE DATA
The electrode life results shown in Figure 6 indicate that the life of electrodes used to weld
galvanneal steel exceeds the life of both aluminium-45% zinc coated and hot dipped galvanised
steels by a factor of 10 and 5 respectively. In addition, the life of electrodes used to weld
aluminium-45% zinc coated and hot dipped galvanised steels is consistent regardless of the
electrode composition. From these results it is obvious that no clear trend exists between the
alternative electrode compositions for aluminium-45% zinc coated and hot dipped galvanised
steels. These results provide no encouragement to utilise the alternative electrode compositions.
However, the life of a Cu/Cr/Zr electrode used to weld galvanneal steel exceeds that of Cu/Cr
2 5 % and Cu/Zr by 3 6 % . From an industrial viewpoint this is quite important as galvanneal steel
is used extensively in automobile panel assembly and any improvements in electrode life will
ultimately lead to reduced costs on the production line.
3.3 ELECTRODE TIP EXAMINATION
Electron micrographs from cross sections of each electrode composition after 1000 welds using
galvanneal steel are shown in Figures 7 to 9. The elemental composition is shown in Table 3
and is the average of at least three analyses. The alloy layer boundaries are well defined and the
composition of each layer is constant throughout the thickness of the layer. Figures 7,8 & 9
show the development of the alloy layers on the high conductivity copper substrate. The
innermost layer is a ductile beta (B) brass of approximate composition 5 2 % zinc, 4 8 % copper.
The adjacent layer is a brittle g a m m a (y) brass, 6 3 % zinc, 3 6 % copper. A n iron zinc alloy layer
of varying iron composition for each electrode composition separates the brasses from a very
porous, brittle zinc oxide surface layer. Moritafll] conducted a similar study and by using X-
ray diffraction identified the outer brass layer as a hard and brittle g a m m a (y) brass (Cu 5 Zn 8 )
and the inner brass layer a ductile beta (B) brass (CuZn). The surface layer was zinc oxide,
adjacent to a porous iron zinc alloy (T phase: Fe 3 Zn 7 ). The (B) and(y) brasses are stable for
each electrode composition except Cu/Zr which displays cracking in the (y) brass layer which is
arrested at the brass interface. All samples exhibit alloy layers which are l5-20|jm thick. The
Cu/Cr and Cu/Cr/Zr electrodes bolh exhibit a continuous, stable iron zinc alloy layer which is
relatively thick between 30 and 60um. T h e Cu/Zr sample has a very ihin (5iUin) discontinuous
iron zinc alloy layer which, judging by the cracking in ihe (y) brass layer, provides little or no
mechanical protection to the underlying brass layers. This irend is continued in the zinc oxide
surface layer. The Cu/Cr and Cu/Cr/Zr electrodes display stable zinc oxides which are ~6()|itn
thick; ihe zinc oxide on the Cu/Zr electrode appears fractured, unstable and thin (30pm).

4. DISCUSSION
4.1 WELD LOBES
It was observed that the degree of scalier on the weld lobes of galvanneal steel is greater than
either hot dipped galvanised and aluminium-45% zinc coated steels. This is attributable to the
narrowness of the welding w i n d o w for galvanneal steel compared to hot dipped galvanised or
aluminium-45% zinc coated steels. A wide window accommodates small fluctuations in power
and machine set up (for a given welding time a larger range of current values will still produce
an acceptable weld) while a narrow w i n d o w will not. Even though great care was taken to
ensure that electrode tip diameter was maintained at 5.0 ± 0.1mm, alignment was exact, and
weld force w a s 2.5 ± O.lkN, scatter up to l.OkA was observed. Reproducing ihe results was
extremely difficult. A s a multi parameter system it is extremely difficult to maintain all
variables exactly the same and hence this amount of scatter was not unexpected. Realistically,
weld lobes only serve as a general guide to the weld time and current combinations which will
produce a satisfactory weld.
Figures 3,4, & 5show that the weld lobes of galvanneal steel, for each electrode composition lie
approximately l k A below that of both aluminium-45% zinc coated and hot dipped galvanised
steels. D u e to the nature of the coating, the current levels for producing acceptable welds are
higher for hot dipped galvanised (free zinc coating) and aluminium-45% zinc coated (Al-45%Zn
coating) steels than for galvanneal (zinc/iron coatings) steel. Zinc/iron coalings display
increased electrical resistivity, higher melting point and reduced shunting of the alloyed coating
and as a result require less current. The situation is further complicated by the effect of coating
weight. Natale [10] has explained that increased shunting at both the electrode/sheet interface
and the faying plane, required increased currents as zinc coating weight increased. While
Kimchi[9] has stated that differences of 2:1 in coaling weight make only small differences to
current values and electrode life, it is inevitable that the current values required to weld
galvanneal would be increased if the coating weight was increased.
4.2 ELECTRODE LIFE DATA

The improvement in electrode life achieved when welding galvanneal steel is largely due to the
coating composition. T h e improvement in electrode life w h e n welding galvanneal steel for all
electrode compositions results from higher electrical resistivity of the coating (lower welding
current), higher melting temperature of the coating (reduced electrode alloying) and lower
thermal conductivity than unalloyed zinc coatings.
The extremely poor performance of aluminium-45% zinc coaled and hot dipped galvanised
steels also relates to the composition of the coating and is primarily due to the addition of
aluminium to the zinc coating to inhibit the growth of iron-zinc intermetallics during the hot
dipped galvanising process and to improve adherence between the coating and the steel. Its
presence, even in small quantities in the alloy layers which grow on the electrode tip as welding
proceeds, alters the weldability of the coated steel and severely reduces electrode life. Brittle
pockets containing aluminium form in the alloy layers which are easily broken away leaving a
pitted electrode face. This is especially noticeable on electrodes used to weld aluminium-45%
zinc coated steel and to a lesser extent hot dipped galvanised steel which displays a similar
erosion process, although the erosion rale is m u c h slower due lo the m u c h smaller amount of
aluminium in ihe coating. Furthermore, the increased eurrenl requirements due lo the increased
coating weighi combined with increased zinc pick up at the sides of the electrodes, results in
reduced electrode life. Galvanneal steel has a similar aluminium content to hoi dipped
galvanised steel yel the life is m u c h longer. It is believed thai the high iron content of the
galvanneal coating contributes to the lowering of the electrode degradation rate when welding
ihis sleel type. It is suggested that iron m a y have a stabilising influence over the diffusion of
both zinc and aluminium into the copper, or of the copper into the complex alloy layers.
Alloying elements are added to the copper electrode material to improve either the mechanical
electrical properties of the electrode, usually at ihe expense of the olher. The chromium is added
to increase hardness which improves mechanical strength and increases tip life, although
electrical conductivity is reduced. Zirconium is added to increase electrical conductivity,
although mechanical strength and hence tip life is reduced. In ihe ease of aluminium-45% zinc
coated and hot dipped galvanised steels the reaction which occurs during welding between the
copper electrode and the metallic coating is so aggressive lhat electrode life is severely reduced.
Hence, no trend is observed between each of the alternative electrode compositions. The
degradation process for the electrodes used to weld galvanneal steel is more gradual and
therefore the effect of additional alloying elements is more pronounced. The high hardness of
the Cu/Cr electrode would appear to have a slightly stronger effect on improving electrode life
than the increased conductivity of the Cu/Zr electrode which displays a shorter life. The
Cu/Cr/Zr electrode however, which is a compromise of the high hardness of the Cu/Cr electrode
and the increased conductivity of the Cu/Zr electrode displays the longest electrode life. This
Cu/Cr/Zr composition offers the best all round properties and hence the longest life.
4.3 ELECTRODE TIP DEGRADATION

The formation of alloy layers on the tips of electrodes used to weld hoi dipped galvanised,
galvannneal and aluminium-45% zinc coated steels has been studied by Howes[12J. H e reported
that there w a s a concentration of aluminium and iron in the electrode alloy layers for each steel
type, although the iron was not distributed uniformly across the tip face except for the
galvanneal steel. It was concluded that the areas of rapid erosion are those where there is
substantial aluminium but little iron present in the alloy layer. This occurred for the galvanised
and in particular for A l - 4 5 % Z n steels, but not for the galvanneal steel. The galvanneal coating
has a lower aluminium content than hot dipped galvanised steels and increased iron content
which is believed to have a stabilising influence over diffusion and the formation of
intermetallics. T h e interest in this work lies in the variation of electrode life for each electrode
composition used to weld galvanneal steel.
The cross sections of each electrode composition (Figures 7 to 9) show equally well developed
brass layers of similar composition. The variation lies in the thickness and composition of the
iron zinc alloy layers. T h e iron zinc alloy layer for the Cu/Zr (Figure 9) electrode is very thin,
discontinuous and fractured whereas for Cu/Cr (Figure 8), and Cu/Cr/Zr (Figure 7) electrodes
the iron zinc alloy layer is stable and relatively thick. Moritafl 1J found a similar result when
spot welding galvanneal steel and observed that when the iron zinc layer is thick electrode life is
relatively long. It is possible that iron in the alloy layers m a y influence electrode degradation by
affecting the mechanical properties of the alloy layers, specifically its resistance to fracture.
Alternatively, iron m a y have a stabilising influence over diffusion of zinc into copper or of
copper into the complex alloy layers. It is apparent from Figure 9 thai when the iron zinc alloy
layer is thin it provides no mechanical protection to the (y) brass layer underneath which appears
cracked and spalled. A s well as the variation in thickness of the iron zinc alloy layer is the
variation in the composition of this layer. There is a corresponding increase in electrode life
with an increase in iron content in the iron zinc alloy layer, (Table 3).
5. CONCLUSIONS

This work examined the characteristics of three electrode materials used for resistance spot
welding metallic coated steels. The electrode materials evaluated were: Cu/Cr, Cu/Zr, and
Cu/Cr/Zr. Steel samples used were galvanneal, aluminium-45% zinc coaled, and hot dipped
galvanised. The electrodes were evaluated using standard weld lobe construction technique and
conveniional electrode life tests. Results from ihese tests were analysed and the conclusions are
summarised below.
(1) The weld lobes of galvanneal sleel (for each electrode composition) were observed lo lie
approximately lkA below those of both aluminium-45% zinc coated and hot dipped galvanised
steels. The galvanneal coating has increased electrical resistivity, higher melting point and
reduced current shunting which reduces the current needed to obtain a satisfactory weld.
(2) In the conventional electrode life test for the aluminium-45% zinc coaled and hot dipped
galvanised steels there is no clear indication that utilising the alternative electrode compositions
will improve electrode life. In the case of galvanneal steel the Cu/Cr/Zr electrode displayed a
marked improvement over ihe Cu/Cr and Cu/Zr electrode compositions.
(3) The aggressive nature of the erosion process encountered for aluminium-45% zinc coated
and hot dipped galvanised steels masks any influence the alloying elements m a y have on
electrode life. The benefit of alloying elements becomes apparent in life tests for galvanneal
steel where the erosion process is more gradual. The composition of the Cu/Cr/Zr electrode is a
compromise between the high hardness of Cu/Cr electrode composition and the increased
conductivity of the Cu/Zr electrode composition and proved to be the most effective in obtaining
m a x i m u m life.
(4) The inferior life performance of the Cu/Zr electrode compared to the Cu/Cr and then the
Cu/Cr/Zr electrode is due to the lower concentration of iron in the iron zinc alloy layer and the
reduced thickness of this layer.
It is clear that limited improvement can be achieved by altering the bulk composition of
electrodes. Further addition of alloying elements to the base electrode is impractical because of
the detrimental effect on other behaviour. However, deposition of thin, adherent coatings on the
electrodes appears to be a logical w a y to modify the surface properties without compromising
the bulk properties of the electrodes. This aspect is now being investigated and the suitability of
a range ofrefractorycoatings, deposited by physical vapour deposition technology, is currendy
under examination. T h e early indications are that these coalings could significantly improve
weldability and electrode life. The results of this investigation will be published elsewhere.
6. ACKNOWLEDGMENTS
This project has the support of BHP Research Port Kembla Laboratory. Substantial assistance
has been provided with the allocation of personnel to supervise the project and the use of
extensive facilities located at the site.
7. REFERENCES
I. HOWES, S.W. Electrode Erosion During Resistance Welding of Metallic Coated Steels.
The Fifth International Conference on Manufacturing Engineering 1990, The University
of Wollongong, July 1990.

2. NAT ALE, T.V., A Review of ihe Resistance Spot Welding Behaviour of Galvanised
Steels. Sheet Metal Welding Conference III, Detroit, U S A , October 1988, Paper No. 1.

3. DICKINSON, D.W., Welding in the Automotive Industry. A1S1 Report SG81 -5.
August, 1981.

4. JUD, R.W., Joining Galvanised and Galvannealed Steels. SAE Technical Paper Series,
No. 840285, 1984.

5. LAVERY, R.C. WILLIAMS, NT., Resistance Spot Welding of Galvanised Steels.

Sheet Metal Industries, March, 1970. pp201.

6. KEY, J.F., COURTNEY, T.M., Refractory Metal Composite Tips for Resistance Spot
Welding of Galvanised Steel. Welding Journal, Supplement June 1974, pp261 -266.

7. SAITO, T., NISHI, T., TAKAHASHI, Y., Electrode Tip Life in Resistance Spot
Welding of Zinc and Zinc Alloy Coated Sheet Steels. Nippon Steel Technical Report
No. 37 Apr. 1988.

8. NEAL'ON, C.S., LAKE, J.S.H., Resistance Spot Welding Tip Lives for Zincalume Coated
Qualities. .BHP Steel Research and Technology Centre, Research Report No. 950,
January 1987.

9. KIMCHI. M., GOULD, J. Effects of Coating Weight on the Resistance Spot Weldability
of Galvanised Steel. S A E Technical Paper Series, No. 860535, 1986.

10. NAT ALE, T. V.., A Comparison of the Resistance Spot Weldability of Hot Dip and
Electrogalvanised Sheet Steels. S A E Technical Paper Series No. 860435,1986.

11. MORTTA, A., INOUE, S., TAKEZOE, A, Spot Weldability of Zinc Vapour
Deposited Steel Sheets. Journal of Metal Finishing Society, Japan. Vol. 39.,
No. 5., pp270-275.

12. HOWES, S.W., LAKE, J.S.H., Degradation of electrodes when resistance welding meta
coated sheet steels. B H P Steel Research and Technology Centre, Research Report
No. 1002, December 1989.
 TABLE NO. 2.

C O A T I N G COMPOSITION

Al Fe Pb Si Coating
(%) (%) (%) (g/m2)
Al-45%Zn 56.7 1.66 1.39 150
Galvanised 0.30 0.13 0.14 275
Galvanneal 0.17 9.02 0.08 80

TABLE NO. 3
COMPOSITIONAL ANALYSIS

Copper Chromium Electrode: Life 4880

Cu (B)brass (y)brass Fe-Zn ZnO

Cu 99.4 51.7 38.4 5.2 5.5
Zn 47.6 59.7 42.4 58.2
Fe 0.4 1.7 51.6 33.8
Al 0.6
Cr 0.6 0.3 0.2 0.9 1.9

Copper Chromium Zirconium Electrode: Life 5990

Cu (B)brass (y)brass Fe-Zn ZnO

Cu 99.6 51.5 33.2 1.8 4.1

Zn 47.9 64.8 39.4 60.9
Fe 0.3 1.4 58.2 31.8
Al 0.9
0.3 0.2 0.5 0.4 0.1
Cr 0.1 0.1 0.1 0.1
..0.1
Zr
Copper Zirconium Electrode: Life 4410

Cu (B)brass (y)brass Fe-Zn ZnO

Cu 99.3 53.2 36.6 18.6 2.5

Zn 46.5 62.4 61.2 83.8
Fe 0.1 0.8 20.0 12.4
Al 0.2 0.1 0.9
0.2 0.1 0.1 0.4
Zr 0.7
 W E l . n L O B E S : M O T D I P P E D G A L V A N I S E D Cu/Cr, Cu/Cr/Zr, Cu/Zr

11
CuCr Weld Current (Min)
CuCr Weld Current (Exp)
io - CuZr Weld Current (Min)
CuZr Weld Current (Exp)
3 CuCrZr Weld Current (Min)
e CuCrZr Weld Current (Exp)

£
9 8-
u
2
"55 7-

-r -r —r- -|
6 8 10 12 14

Weld Time (cycles)

Figure 3 Weldability lobes for different electrode compositions used for spot welding of
hot dipped galvanised steel

W E L D L O B E S : A L - 4 5 % Z N C O A T E D Cu/Cr, Cu/Cr/Zr, Cu/Zr

11 n
CuCr Weld Current (Min)
CuCr Weld Current (Exp)
10- CuZr Weld Current (Min)
CuZr Weld Current (Exp)
CuCrZr Weld Current (Min)
9- CuCrZr Weld Current (Exp)
e
v
u
u
S
U
7-

—r-
6 8 10 12 14

Weld Time (cycles)

Figure 4 Weldability lobes for different electrode compositions used for spot welding of
Al-45% Zn steels.
 W E L D LOIIES: GALVANNEAL Cu/Cr, Cu/Cr/Zr, Cu/Zr

11 -\
-O- CuCr Weld Current (Mln)
CuCr Weld Current (Exp)

10- CuZr Weld Current (Min)

CuZr Weld Current (Exp)

2 CuCrZr Weld Current (Min)

9- CuCrZr Weld Current (Exp)
B
V
fc
9
8-
2

-~t— "T-
6 10 14
12

Weld Time (cycles)

Figure 5 Weldability lobes for different electrode compositions used for spot welding
galvanneal steel.

ELECTRODE LIFE TEST RESULTS

6000

5000-

4000-
•J E3 Cu/Cr
va
3000- U Cu/Zr
8
m Cu/Cr/Zr
2000-

1000-

GALVANNEAL AL45%ZN GALVANISED

C O A T E D STEEL

Figure 6 Electrode life data for all electrode compositions and all coated steels.
Figure 7 Electrode alloying during resistance spot welding of galvanneal steel using
Cu/Cr/Zr electrode.

_
•Cu/Cr substrateH
•
•Fe-Zn layerB beta brass '.UIM.IJ.W--
% /
w
y^*

> f r
fmQ&'iMBb&Ui
•'hm&mWm-il &MS9BU0Ukti»Ji

?0i Jffe&-rSf:
Rsi »^Jj2?rtr£wrVyL£2rV(
Jir •/• •ir "JV** & !

in i V ^ / z i /%' 'J^l
M L >ISK P B
n fniTrWT* tf ii r*ni all * > » v/. J* - 3 r v *
•
feKif3^1b&T*Qr^T * J

rA^^TJfcfJpSviRKPjt
y
V>rf
V.il"J •' >• \»J.*' • *' -VJ

"•r....^.:,.^/;^ !®ssfflsr«ra!i
Zn-oxlde surface layer
•

EHT=20.00 kV WD= 24 mm Mag=400.31 X

30u | 1 Photo No.=l Detector^ QBSD

Figure 8 Electrode alloying during resistance spot welding of galvanneal steel using
C u / C r electrode.
 u/Zr substrate

e-Zn layer lM#H#i arama brass

Zn-oxide surface layer

EOT =20.00 kV 25 run Mag=400.31 Y

Photo No. Detector= QBSD

Figure 9 Electrode alloying during resistance spot welding of galvanneal steel using
C u / Z r electrode.
 APPENDIX 6
PRODUCT DATA SHEET FOR ZINCANNEAL G3N

128
 Steel
ZINCANNEAL® G3N
Revision 2. May 1991
This literature supersedes all previous Issues

G E N E R A L DESCRIPTION TYPICAL USIS

Z I N C A N N E A L G 3 N — matt hot-dipped zinc- Automotive components.
iron alloy coated drawing steel guaranteed
non-ageing and free from stretcher strain
suitable for direct on painting. S o m e
powdering of the coating m a y occur with
severe deformation.

Standard: Australian Standard 1397

DIMENSIONS
For normal supply product
Preferred thickness, m m Width range, m m Sheet length, m m
Strip Sheet
0.50.0.55.0.60.0.65.0.70,0.75 250-1380 610-1380 1085-4870
0.80.0.85,0.90,0.95.1.0.1.11.2 250-1525 610-1525 1085-4870
Fo» m a x i m u m widths leu than 1200 m m . any width between half maximum
and 610 m m Is available by negotiation only.

M E C H A N I C A L PROPERTIES CHEMICAL COMPOSITION

Steel base Guaranteed Typical Guaranteed maximum, %
(Basis 0.80 m m thick)
Transverse tensile Carbon(C) 0.08
Yield strength. MPa — 130-170 Phosphorus(P) 0.02
Tensile strength. MPa — 280-320 Manganese(Mn) 0.40
Elong on 80 mm, % 32 min 37-46 Sulphur(S) 0.025
Hardness, HR30T 55 m a x 38-48
180° transverse bend
(L axis) Ot Ot
Uniform elongation. % — 20-24
Plastic strain ratio
rO — 1.3-1.8
r90 — 1.3-2.1
r45 — 1.2-1.6
Coating: Refer Table 3. General Information for bend test performance.

SUPPLY CONDITION FABRICATING

PERFORMANCE
Normal Optional* Mfmlted to 5-excellent
Coating class ZFJOO — Bending 5
Surface finish Smooth, matt — Drawing 5
Surface treatment Phosphated Unphosphated Pressing 4
(oiled) Roll-forming 5
Tolerance class Welding 5
Thickness AT — Painting 5
Width AW BW
Length AL BL
Flatness AF BF
Oiling Not oiled —
Branding Branded Not branded
tOptional supply conditions and coating classes m a y be subject to size range restrictions.
Z I N C A N N E A L 0 is a registered trade mark of John Lysaght (Australia! Limited.

Please ensure you have the current data sheet for this product.

Sheet & Coll Products

John Lysaght (/VutraM Limited |A£.N. 000 011 0S8| Copyright © 1991 John Lysaght (Australia) Limited
The Broken Hill Proprietary Company Umlied |AJCJ\|. 004 028 077)
AuSMltan Iron and Steel Fly Ud IA.C.N. 000 01162S|
numtnlltn mVUt mduflflei « y Ud (AJC.N. 004 4K> 833)
SYDNIYl |02] 708 8888 M I L S O U R N f J (03) SH* 2232 M I M A N d (07) 345 0SS5 AOILAJOIi (OB) 243 7333 K R T M t fCT) 330 0444
 APPENDIX 7
TECHNICAL DATA SHEET 1 FOR J47 COPPER ALLOY

129
TECHNICAL
DATA SHEET
NUMBER
1 JAMES COPPEU LEE PTY. LTD.

KM
If a Chromium Copper alloy, conforming to R W M A class (2)

specification. This rod material has a (2) stag* heat treatment

procedure, coupled with a work hardening process designed

specially for spot welding tips.

High Conductivity 8 0 - 8 5 % IACS

ROD gives you Rockwell hardness 75-80B
Excellent bending properties

When these features are combined, then you have:

Longer life (more welds between dressing)
Single and double bends are possible.
Uniform high quality metal.

W e are the actual manufacturers and distributors o f — \ f l " •.

ALWAYS REMEMBER That means w e alloy, cast, hot roll, draw, and heat
treat the rod, all under our o w n personal supervision

Perhaps you require cast tips or some special type of clamp

with cast in water tubes, castings up to 300-400 lb. have
been undertaken successfully.

PHYSICAL
Properties of Conductivity 75 V. IACS

j castings Rockwell hardness 65B

.47

TRY NOW and test it for yourself

47.
For further information JAMES COPPELL LEE PTY. LTD.
please contact: 500 La Trobe Street, Melbourne. Phone: 328 26.
 APPENDIX 8
TECHNICAL DATA SHEET 2 FOR J47 COPPER ALLOY

130
TECHNICAL
DATA SHEET
NUMBER
2 JAMES COPPELL LEE PTY. LTD.

COPPER 99.5%
CHROMIUM 0.5% Min.

Coeff. of thermal expn. 25-300*C. 17 x 10-*

1 • I
physical Thermal conductivity C. G. S. Units 0.75

properties Electrical conductivity % I.A.C.S. 80-85

Youngs Modulus "E" Ib./sq. In. 17 x 10*

Machinabiliry rating. Free-cutting brass = 100 80

Resistance to softening under stress-free conditions 500 # C

0.1% Bongo
U.TA
M O O f STIESS Toni/tq. In % HAtONESS
mechanical Toni/tq. In on 3 Int.

properties Drawn tar and tod 25-30 30-35 2O-30 135-160

Forging>>H«o> Tr»ot»d 16-20 22-28 20-30 110-135

.Cold
- - - "1" Work
21-27 25-30 18-25 135-160

Caitingt-Hw! Tr.at.d 15 20-24 20-26 110-135

Inch.i Diam.t.r Weight Ib./ff. (achat Diamtlm* w.Johi a./ft.

1/4 .189 7/8 2.38

drawn 5/16 .296 1 . 3.03
sections 3/8 .426 1-1/8 3.83

1/2 J57 1-1/4 4.73

5/8 1.18 1-3/8 5.72

3/4 1.70 1-1/2 JL. 6.81

lings, and castings on application

ALL OTHER SIZES, shaf>es, rectangles, for{ ( m
JS
JAMES COPPELL LEE PTY. LTD.
500 La Trobe Street, Melbourne. Phone: 328 2651
 APPENDIX 9
PRODUCT DATA SHEET FOR ZINCANNEAL G2

131
 Steel
ZINCANNEAL® G2
Revision 4. May 1991
This literature supersedes all previous Issues

G E N E R A L DESCRIPTION TYPICAL USES

Z I N C A N N E A L G 2 — matt hot-dipped Acoustic ceiling tiles louvres, door frames,
zinc-iron alloy coated commercial forming exterior signs, commercial refrigerators and
steel suitable for direct-on painting, and freezers.
guaranteed for Pittsburg lock-seaming.
S o m e powdering of the coating m a y
occur with severe deformation.

Standard: Australian Standard 1397

DIMENSIONS
For normal supply product
Preferred thickness, m m Width range, m m Sheet length, m m
Strip Sheet
0.40, 0.42.0.45.0.48 250-1300 610-1300
0.50,0.55.0.60.0.65, 0.70, 0.75 250-1525 610-1525
0.80,0.85.0.90.0.95, 1.0,1.1 250-1525 610-1525
1.2. 1.3. 1.4, 1.5. 1.6 250-1525 610-1525 1085-4870
1.7 250-1435 610-1435
1.8 250-1355 610-1355
1.9 250-1285 610-1285
2.0 250-1220 610-1220
W h e r e m a x i m u m widths are less than 1200 m m . any width between half m a x i m u m
and 610 m m is available by negotiation only.

MECHANICAL PROPERTIES CHEMICAL COMPOSITION

Steel base Guaranteed maximum, %
(Basis 0.80 m m thick|
Transverse tensile Carbon(C) 0.10
Yield strength, MPa 260-360 Phosphorus(P) 0.025
Tensile strength, MPa 340-440 Manganese(Mn) 0.45
Elong on 80 mm, % 27 min 30-40 Sulphur(S) 0.03
Hardness. HR30T 50-60
180° transverse bend
Ot Ot
[L axis)
Pass Pass
Pittsburgh lockseam
Coating: Refer Table 3. General Information for bend test performance.

SUPPLY CONDITION FABRICATING

PERFORMANCE
Normal Optional* 1-llmlted to 5-excellent
Coating class ZF100 — 5
Bending
Surface finish Smooth, matt — 3
Drawing
Surface treatment Phosphated Unphosphated (oiled) 3
Pressing
Tolerance class 5
AT Roll-forming
Thickness 5
BW Pittsburgh lock-seaming
Widtfi AW Welding
5
Length AL BL 5
Painting
Flatness AF BF
Oiling Not oiled Not branded
Branding Branded
fOptional supply conditions and coating classes m a y be subject to size range restrictions
Z I N C A N N E A L * is a registered trade mark of John Lysaght |Australia| Limited.

Please ensure you have the current data sheet for this product

Sheet & Coll Products

John Lyiaght |Au«/alia| Umtted |A C M . 000 011 058| Copyright © 199! John Lysaght {Australia} Limited
The Broken HJU Proprietary Company Limned (AJC.N. 004 028077|
Ausralian Iron and Steel Ply lid (AJCJM. 000 019 62S|
/Vusnllan Wire m d u a r m Ply Ltd (AJC.N. 004 410 833|
fYONCYI |02J 708 8888 M I L B O U R N I i (03) 586 2222 M I S B A N b (07| 34$ 0555 A O * 1 _ A J O « J (08) 243 7333 PMTm|09| 3300644
 APPENDIX 10
BLOCK SCHEMATIC OF BHP RESEARCH AND TECHNOLOGY CENTRES
RESISTANCE SPOT WELDER AND ANCILLARY EQUIPMENT

Schlatter
Spot Welder
Charge
Amplifier

Force

Xs Current

pffi^ voltage
Flux Sensing
Coil

Passive Passive
Divider Divider
Display Monitor J
( r 2) i r 2)

Printer

Key: A D C — Analogue to Digital Converter

PGIA — Programmable Gain Instrumentation Amplifier
 APPENDIX 11
TECHNICAL SPECIFICATION FOR FORM TALYSURF 120 mm TRAVERSE UNIT

Traverse length: 120 m m (4.72 in) O/travcl 1.0 m m

((0.04 in) at cither end)

Straightncss of Traverse Within 2 parallel lines spaced 0.5 urn over 120 mm
(column mounted):
(20 LI in over 4.72 in).
and
0.25 u m (10 LI in) over any 60 m m (2.36 in)
traverse

Slraightness of Traverse Within 2 parallel lines spaced 1.0 urn over 120 mm
(free standing) (40 u in over 4.72 in)
and
0.5 urn (20 u in) over any 60 m m (2.36 in) traverse

Traverse speeds —
Distance to achieve stable measuring speed: 0.2 m m (0.008 in) max
Measuring speeds: 1.0 and 0.5 mm/s ± 5 %
(0.04 and 0.02 in ± 5 % )

Instantaneous speed fluctuation:

Reverse speed: 10 mm/s (0.4 in/s) max

Movement speeds (±15%) 0.5 mm/s (0.0197 in/s)

1.0 mm/s (0.0394 in/s)
2.5 mm/s (0.0984 in/s)
5.0 mm/s (0.1969 in/s)
10 mm/s (0.3937 in/s)

Horizontal stylus position (along 'Y' axis) relative 52.5 m m (2.06 in) nominal
to mounting face of carriages

Traverse end stops: Fixed optical switches

Simple traverse position scale: 1 mm and 0.05 in divs

Overall dimensions (WxDxH) 390 mm x 100 mm x 150 mm

(15.6 x 5 x 7.7 in)

Clearance under tip of standard stylus and the75 mm

plane
of the feet of a free standing unit:

Computer I/O: ELM or terminal module

Weight approx 11.5 kg (25.3 lb)

 APPENDIX 12
ELECTRODE TIP CARBON PAPER IMPRESSION DIAMETER RESULTS

Weld Impression Diameter Average Weld Impression Diameter Average

Number (mm) (mm) Number (mm) (mm)
I 5.30/5.30 5.30 1300 5.50/5.20 5.35
50 5.10/5.00 5.05 1400 4.80/5.10 4.95
75 5.10/5.00 5.05 1500 4.50/5.00 4.75
100 5.50/5.00 5.25 1600 4.50/4.60 4.55
125 5.50/5.00 5.25 1700 4.70/4.70 4.70
150 5.30/5.50 5.40 1800 4.50/5.00 4.75
175 5.30/5.30 5.30 1900 4.50/4.60 4.55
200 5.50/5.30 5.40 2000 4.20/5.00 4.60
250 5.10/5.20 5.15 2100 4.20/4.80 4.50
300 5.40/5.40 5.40 2200 4.50/4.10 4.30
350 5.40/5.40 5.40 2300 4.20/5.50 4.85
400 5.50/5.50 5.50 2400 5.60/4.30 4.95
450 5.40/5.50 5.45 2500 5.80/5.50 5.65
500 5.80/5.60 5.70 2600 5.50/5.70 5.60
550 5.50/5.70 5.60 2700 5.50/5.80 5.65
600 5.50/5.80 5.65 2800 5.50/5.90 5.70
650 5.30/5.40 5.35 2900 5.60/5.80 5.70
700 4.90/6.80 5.85 3000 5.50/5.50 5.50
750 4.70/5.50 5.10 3200 5.10/5.40 5.25
800 5.50/4.80 5.15 3400 5.50/5.90 5.70
850 5.40/5.50 5.45 3600 5.00/5.90 5.45
900 5.50/5.50 5.50 3800 5.40/5.90 5.65
950 5.30/5.50 5.40 4000 4.50/5.90 5.20
1000 5.50/5.50 5.50 4200 5.50/6.00 5.75
1100 5.50/5.20 5.35 4400 5.80/5.90 5.85
1200 5.80/5.80 5.80 4600 5.80/6.00 5.90

134
 APPENDIX 13
PROFILOMETER WELD IMPRESSION DIAMETER RESULTS — INITIAL
ELECTRODE TIP LIFE TEST

Impression Diameter ( m m )
Weld Number Longitudinal Transverse Average Diameter ( m m )

1 6.082 6.082 6.082

25 6.287 6.194 6.241
50 5.541 5.690 5.616
75 5.616 5.653 5.635
100 5.690 5.672 5.681
150 5.783 5.746 5.765
175 5.858 5.690 5.774
200 5.746 5.727 5.737
250 5.895 5.933 5.914
300 5.970 5.802 5.886
350 5.541 5.877 5.709
400 4.944 6.045 5.495
450 4.944 5.970 5.457
500 5.019 5.821 5.420
550 4.963 6.194 5.579
600 6.045 6.082 6.064
650 4.981 6.045 5.513
700 5.317 6.082 5.700
750 4.907 6.119 5.513
800 5.354 6.082 5.718
850 5.429 6.306 5.868
900 5.336 6.175 5.756
950 5.093 6.306 5.700
1000 5.970 6.679 6.325
1100 5.205 6.492 5.849
1200 5.317 6.362 5.840
1300 5.205 6.324 5.765
1400 5.485 6.306 5.900
1500 5.224 6.362 5.793
1600 5.727 6.436 6.082
1700 5.933 5.354 5.644
1800 4.888 5.392 5.140
1900 4.683 5.504 5.094
2000 5.168 4.888 5.028
2100 4.478 5.168 4.823
2200 5.429 4.366 4.898
2300 5.485 4.533 5.009
2400 4.795 6.623 5.709
2500 5.746 5.373 5.600
2600 6.045 6.530 6.288
2700 6.138 6.492 6.315
2800 6.119 6.455 6.287
2900 6.250 6.586 6.418
3000 6.231 6.287 6.259
3200 6.287 6.642 6.465
3400 6.287 6.735 6.511
3600 6.213 6.791 6.502
3800 6.175 6.548 6.362
4000 5.429 6.511 5.970
4200 5.914 6.828 6.371
4400 4.981 6.418 5.700
4600 6.306 6.921 6.614
 APPENDIX 14
VISION SYSTEM WELD IMPRESSION DIAMETER RESULTS — INITIAL
ELECTRODE TIP LIFE TEST

1v
Veld
dumber
1 mprcssion Diameter (Pixel Spacing) Average Diameter Average Diameter Circular Function
Longitudinal Transverse (Pixel Spacing) (mm) (%)
1
I 53.641 50.609 52.075 4.470 95
25 60.516 60.547 60.532 5.196 85
50 55.266 56.344 55.805 4.790 98
75 56.484 58.719 57.602 4.944 69
100 59.797 56.000 57.899 4.970 57
125 56.844 56.156 56.500 4.850 92
150 87.891 76.250 82.071 7.045 99
175 68.031 65.859 66.945 5.746 91
200 67.797 66.672 67.235 5.771 51
250 68.000 63.125 65.563 5.628 84
300 73.797 67.031 70.414 6.044 88
350 65.391 67.453 66.422 5.701 74
400 66.344 87.432 76.888 6.600 80
450 77.469 86.297 81.883 7.029 75
500 78.828 86.203 82.516 7.083 69
550 82.891 81.516 82.204 7.056 64
600 80.156 87.313 83.735 7.188 58
650 95.922 86.125 91.024 7.813 75
700 74.313 80.531 77.422 6.646 57
750 73.609 101.016 87.313 7.495 60
800 82.547 97.594 90.071 7.731 86
850 72.938 91.391 82.165 7.053 90
900 77.016 78.109 77.563 6.658 85
950 78.984 88.906 83.945 7.206 67
1000 78.313 100.297 89.305 7.666 80
1100 81.688 98.797 90.243 7.746 72
1200 90.953 98.219 94.586 8.119 84
1300 88.563 94.141 91.352 7.841 97
1400 87.359 90.969 89.164 7.654 84
1500 84.625 95.594 90.110 7.735 82
1600 86.469 95.703 91.086 7.819 82
1700 89.609 92.219 90.914 7.804 82
1800 86.922 92.063 89.493 7.682 89
1900 79.594 86.578 83.086 7.132 88
2000 81.234 83.516 82.375 7.071 100
2100 83.547 88.516 86.032 7.385 95
2200 75.500 90.766 83.133 7.136 90
2300 77.391 93.781 85.586 7.346 63
2400 82.859 94.094 88.477 7.595 47
2500 86.797 98.484 92.641 7.952 86
2600 83.688 98.109 90.899 7.802 83
2700 89.844 91.813 90.829 7.796 85
2800 91.344 90.359 90.852 7.798 98
2900 90.828 100.375 95.602 8.206 96
3000 92.219 96.922 94.571 8.118 97
3200 87.313 98.813 93.063 7.988 87
3400 98.875 95.219 97.047 8.330 83
3600 94.297 101.875 98.086 8.419 85
3800 89.609 96.422 93.016 7.984 59
4000 88.359 93.672 91.016 7.813 96
4200 85.984 97.750 91.867 7.886 91
4400 86.063 95.391 90.727 7.788 97
4600 j 86.781 95.438 91.110 7.821 87

Note: Conversion 1 m m = 11.65 pixels

 APPENDIX 15
X-RAY WELD NUGGET SIZE RESULTS — INITIAL ELECTRODE TIP LIFE TEST

Weld Number Longitudinal Nugget Transverse Nugget Average Nugget Diameter

Diameter ( m m ) Diameter ( m m ) (mm)

1 6.50 6.80 6.65

25 6.80 6.80 6.80
50 6.20 6.20 6.20
75 6.10 6.20 6.15
100 6.10 6.10 6.10
125 6.10 6.10 6.10
150 6.10 6.10 6.10
175 6.20 6.10 6.15
200 6.20 6.10 6.15
250 6.20 6.30 6.25
300 6.40 6.40 6.40
350 6.40 6.40 6.40
400 6.40 6.50 6.45
450 6.20 6.50 6.35
500 6.40 6.30 6.35
550 6.50 6.40 6.45
600 6.40 6.60 6.50
650 6.50 6.50 6.50
700 6.50 6.60 6.55
750 6.30 6.20 6.25
800 6.40 6.60 6.50
850 6.40 6.60 6.50
900 6.30 6.40 6.35
950 6.50 6.50 6.50
1000 6.40 6.90 6.65
1100 6.50 6.60 6.55
1200 6.30 6.40 6.35
1300 6.40 6.50 6.45
1400 6.30 6.40 6.35
1500 6.50 6.60 6.55
1600 6.50 6.50 6.50
1700 6.40 6.50 6.45
1800 6.70 6.30 6.50
1900 6.60 6.50 6.55
2000 6.60 6.50 6.55
2100 7.00 6.80 6.90
2200 7.00 6.80 6.90
2300 7.00 6.70 6.85
2400 7.00 6.80 6.90
2500 7.50 6.90 7.20
2600 6.90 7.00 6.95
2700 7.20 6.90 7.05
2800 7.00 7.00 7.00
2900 7.00 6.90 6.95
3000 7.00 7.00 7.00
3200 7.00 7.10 7.05
3400 6.80 7.00 6.90
3600 6.70 7.00 6.85
3800 6.80 7.10 6.95
4000 7.30 7.10 7.20
4200 7.50 7.20 7.35
4400 7.50 8.00 7.75
4600 1 7.60 7.30 7.45
 APPENDIX 16
ULTRASONIC C-SCAN WELD NUGGET SIZE RESULTS — INITIAL ELECTRODE
TIP LIFE TEST
Weld N u m b e r -ongiludinal/Transvcrsc Average Nugget Longitudinal/Transverse Average Nugget
Nugget Diameter ( m m ) Diameter ( m m ) - Nugget Diameter ( m m ) Diameter ( m m ) -
- Outer Outer - Inner Inner

1 N o result N o result N o result N o result

25 7.44/7.44 7.44 5.04/4.80 4.92
50 6.24/6.48 6.46 5.04/4.80 4.92
75 6.48/6.26 6.36 4.32/3.84 4.08
100 6.24/6.48 6.36 4.80/4.80 4.80
125 6.48/6.48 6.48 5.28/5.76 5.52
150 7.20/6.72 6.96 4.56/5.28 4.92
175 6.24/6.24 6.24 1.68/2.16 1.92
200 6.48/6.48 6.48 -/- -
250 6.48/6.48 6.48 5.04/4.80 4.92
300 6.24/6.48 6.36 4.80/4.80 4.80
350 5.76/5.76 5.76 4.342/3.36 3.84
400 6.72/6.72 6.72 5.28/4.80 5.04
450 6.48/6.48 6.48 4.80/3.60 4.20
500 5.52/6.24 5.88 -/- -
550 5.76/6.48 6.12 -/- -
600 6.48/6.48 6.48 3.84/4.08 3.96
650 7.20/7.44 7.32 4.56/3.12 3.84
700 5.76/6.48 6.12 -/- -
750 5.28/5.76 5.52 -/- -
800 5.28/6.96 6.12 4.80/5.52 5.16
850 6.48/6.24 6.36 5.04/5.76 5.40
900 5.52/6.48 6.00 5.28/5.52 5.40
950 5.76/6.72 6.24 4.80/5.76 5.28
1000 6.48/8.16 7.32 -/- -
1100 6.48/6.96 6.72 5.76/7.20 6.48
1200 6.48/7.20 6.84 2.64/2.40 2.52
1300 6.96/7.20 7.08 -/- -
1400 6.96/6.96 6.96 1.20/0.72 0.96
1500 N o result N o result N o result N o result
1600 5.04/6.24 5.64 -/- -
1700 7.20/7.68 7.44 -f- -
1800 6.24/6.72 6.48 -/- -
1900 6.72/6.96 6.84 5.76/6.24 6.00
2000 5.28/5.52 5.40 -/- -
2100 5.28/5.28 5.28 1.20/1.44 1.32
2200 6.48/6.00 6.24 6.24/5.04 5.64
2300 5.52/5.58 5.55 -/- -
2400 6.96/6.96 6.96 4.32/4.80 4.56
2500 6.00/6.24 6.12 -/- —
2600 6.24/6.72 6.48 4.32/4.08 4.20
2700 6.24/5.76 6.00 4.56/5.76 5.16
2800 5.28/6.72 6.00 2.88/3.12 3.00
2900 7.20/6.96 7.08 -/- —
3000 6.24/6.48 6.36 4.56/4.08 4.32
3200 6.96/6.72 6.84 4.80/5.04 4.92
3400 6.24/7.20 6.72 4.80/4.32 4.56
3600 6.48/6.72 6.60 4.32/4.08 4.20
3800 5.76/6.48 6.12 3.36/4.80 4.08
4000 5.52/6.72 6.12 -/- -
4200 5.76/6.48 6.12 3.36/4.56 3.96
4400 5.28/6.48 5.88 -/- —
4600 5.76/6.48 6.12 3.84/2.88 3.36

138
 APPENDIX 17
PEEL TEST RESULTS MEASURED THROUGH VERNIER CALLIPERS
Weld Number Longitudinal Nugget Diameter Transverse Nugget Diameter Average Nugget Diameter
(mm) (mml (mm)

1 5.54 6.10 5.82

25 5.46 5.92 5.69
50 5.32 Tail excessive 5.32
75 4.90 5.24 5.07
100 4.92 5.42 5.17
125 4.60 4.90 4.75
150 4.64 5.36 5.00
175 5.16 4.42 4.79
200 4.58 4.98 4.78
250 4.34 4.66 4.50
300 4.66 5.22 4.94
350 4.58 5.18 4.88
400 4.54 5.36 4.95
450 4.14 4.82 4.48
500 4.52 5.00 4.76
550 4.46 5.40 4.93
600 4.12 4.98 4.55
650 4.18 4.88 4.53
700 4.44 5.22 4.83
750 4.26 5.02 4.64
800 4.38 4.98 4.68
850 4.62 5.48 5.05
900 4.26 4.92 4.59
950 4.42 5.22 4.82
1000 5.44 5.28 5.36
1100 5.14 4.98 5.06
1200 4.82 5.82 5.32
1300 4.74 5.56 5.15
1400 4.58 4.64 4.61
1500 4.18 4.24 4.21
1600 4.06 4.82 4.44
1700 3.84 4.38 4.11
1800 3.80 4.38 4.09
1900 4.44 4.04 4.24
2000 4.56 4.40 4.48
2100 3.68 4.42 4.48
2200 3.64 4.08 3.86
2300 3.24 4.28 3.76
2400 3.20 3.84 3.52
2500 2.44 2.74 2.59
2600 3.54 4.10 3.82
2700 4.20 4.42 4.31
2800 4.00 3.84 3.92
2900 4.36 4.50 4.43
3000 4.68 4.92 4.80
3200 4.54 5.10 4.82
3400 4.24 5.00 4.62
3600 4.42 5.50 4.96
3800 4.82 4.72 4.77
4000 4.16 5.28 4.72
4200 2.50 2.86 2.68
4400 3.34 3.04 3.19
4600 3.24 4.00 3.62
 APPENDIX 18
VISION SYSTEM WELD IMPRESSION DIAMETER RESULTS — SECOND
ELECTRODE TIP LIFE RUN
Weld Number Impression Diameter (Pixel Spacing) Average Diameter Average Diameter
Longitudinal Transverse (Pixel Spacing) (mm)

1 446.625 439.750 443.188

2 451.734 429.1456 440.445 6.211
3 434.266 446.312 440.289
249 427.469 405.687 416.578
250 422.937 430.172 426.555 5.981
251 425.344 438.250 431.797
499 418.391 431.609 425.000
500 409.500 438.219 423.860 5.983
501 416.906 435.844 426.375
749 420.547 429.828 425.188
750 416.734 446.141 431.438 5.996
751 412.094 430.672 421.383
999 437.812 431.187 434.500
1000 420.266 446.328 433.297 6.117
1001 429.906 442.234 436.070
1249 423.594 429.047 426.321
1250 412.734 435.500 424.117 6.027
1251 426.625 442.047 434.336
1499 428.672 447.812 438.242
1500 441.734 439.000 440.367 6.175
1501 423.672 451.641 437.657
1749 440.812 439.422 440.117
1750 430.656 434.937 432.797 6.230
1751 455.359 454.984 455.172
1999 424.937 443.641 434.289
2000 443.844 450.109 446.977 6.186
2001 432.922 441.641 437.282
2249 438.547 446.172 442.360
2250 435.016 458.359 446.688 6.236
2251 428.594 451.969 440.282
2499 453.375 458.734 436.055
2500 447.625 . 471.922 459.774 6.338
2501 439.859 470.547 455.204
2749 433.219 457.234 445.227
2750 447331 453.375 450.453 6.382
2751 470.734 458.641 446.688
2999 458.453 451.781 455.117
3000 435.859 452.469 444.164 6.333
3001 451.453 449.812 450.633
3249 442.937 450.859 446.898
3250 450.656 450.922 450.789 6.338
3251 454378 452.062 453.320
3499 467331 464.750 466.141
3500 444394 456.156 450.375 6.464
3501 468316 454.219 461.368
3749 458.687 458.641 458.664
3750 446.656 456.203 451.430 6.382
3751 436.406 464.031 450.219
3999 464.719 458.312 461.516
4000 464.344 466.578 465.461 6.508
4001 459.187 461.234 460.211
4249 466.672 447.719 457.1%
 APPENDIX 18 (Cont)

Weld Number Impression Diameter (Pixel Spacing) Average Diameter Average Diameter
Longitudinal Transverse (Pixel Spacing) (mm)

4250 471.641 465.391 468.516 6.529

4251 465.531 466.297 465.914
4499 466.234 468.687 467.461
4500 465.094 452.875 458.985 6.521
4501 467.656 459.297 463.477
4599 485.312 458.641 471.977
4600 472.000 453.609 462.805 6.605
4601 481.906 464.344 473.125

Note: Conversion 1 m m = 71.05 pixels

141
 APPENDIX 19
TENSILE SHEAR TEST RESULTS
Weld Number Tensile Shear Strength Ave Tensile Shear Strength
(MPa) (MPa)

1 413
2 420 417
3 418
249 463
250 450 457
251 459
499 439
500 435 436
501 435
749 453
750 451 451
751 449
999 459
1000 445 453
1001 455
1249 423
1250 410 417
1251 418
1499 459
1500 450 453
1501 449
1749 452
1750 445 448
1751 447
1999 446
2000 438 447
2001 456
2249 435
2250 442 436
2251 430
2499 449
2500 453 451
2501 450
2749 439
2750 456 447
2751 446
2999 440
3000 445 445
3001 440
3249 418
3250 417 412
3251 401
3499 421
3500 436 421
3501 406
3749 410
3750 419 411
3751 403
3999 413
4000 420 417
4001 418
4249 406
4250 391 401
4251 406
4499 396
 APPENDIX 19 (Cont)

Weld Number Tensile Shear Strength Ave Tensile Shear Strength

(MPa) (MPa)

4500 384 394

4501 402
4599 395
4600 374 385
4601 386

143