Undermatching Butt Welds in High Strength Steel

2007:220 CIV

M A S T E R ' S T H E S I S

Undermatching Butt Welds in High Strength Steel

Svante Törnblom

Luleå University of Technology MSc Programmes in Engineering

Civil and mining Engineering

Department of Civil and Environmental Engineering Division of Structural Engineering

PREFACE

Preface

This is the beginning of the end of a journey that started about five years ago. In the middle of August in 2002 I packed all my belongings into a model -72 Volvo 142 that I had inherited from my grandfather. The 1000 km trip from Gotland to Luleå was a good way to see Sweden at a moderate pace (the car wouldn’t go faster than 90 km/h) and to prepare myself mentally for the coming life as a student. At this stage I did not know where this journey would end, all I knew was that I had been accepted to an open engineering programme which would give me a master’s degree.

Today, I’m fairly sure that the end of the journey will be a master’s degree in civil engineering, where this thesis is the final document.

This work was initiated by Ramböll Sverige AB.

SSAB Oxelösund financed the project and supplied material, welding, testing equipment and knowledge.

I would like to express my gratitude to the persons who made this work possible.

At first I would like to thank Professor Peter Collin who drafted me and gave me the opportunity to present this thesis in cooperation between Ramböll and LTU.

I would also like to thank Daniel Stemne, at SSAB Oxelösund, who sacrificed his time to make me understand aspects of welding in high-strength steel.

Thanks to Professor Bernt Johansson, who has listened to me and given me useful advice.

Mats Fried and Simo Hietala, at SSAB Oxelösund, have done a magnificent job, welding and fabricating the test specimens and later testing them until fracture.

A special thanks to my classmates. There have been many fruitful discussions, both related and non- related to school, during our “fika-” breaks.

Finally, Lena, thank you for all the help and support at home when I was trying to finish this thesis off during some sunny/rainy summer weeks.

Luleå, September 2007

Svante Törnblom

ABSTRACT

Abstract

The last years improvements in steel making technologies have resulted in steel grades with higher tensile strength, hardness, weldability etc. compared to conventional construction steel. This category steel gives lighter and smaller sized structures and cost reductions.

Since the weld metal is still only a cast alloy whose strength mainly depends on its chemical composition, the differences in mechanical properties between the base plate and the consumable are increasing. The question is how to choose the undermatching level between the base plate and weld metal without compromising the global strength of the joint.

In this master’s thesis, the effects of different weld geometries on the mechanical properties of undermatched welds in high strength quenched and tempered steel, are studied. Different undermatching levels have been accomplished by combining two different steels; Weldox 960 and Weldox 1100 together with two consumables; Filarc PZ6145 and Filarc PZ6149.

Two methods have been used to gather information; 30 test specimens have been manufactured and tested at SSAB’s plant in Oxelösund and previous studies in this field have been analyzed in a literature review.

Three different parameters were chosen to be studied in the laboratory tests, only one parameter was changed at a time:

• Width- to thickness relation

• Undermatching

• Relative thickness

The hypothesis was that if the weld metal is softer than the adjacent steel, it will yield before the base metal does. When the weld starts to deform, this deformation is constrained by the non-yielded surrounding metal. Tension is developed in both the width- and thickness directions, in addition to tension in the longitudinal direction due to the applied load. When the weld experiences tension in two or three material directions, the mean stress or hydrostatic stress in the weld is increased. As the hydrostatic stress is increased, by constraint, the magnitudes of the deviatoric stresses, which govern yielding, are reduced. This results in a stronger weld where increases in the applied load can be achieved.

When the width- to thickness relation was studied, five different specimen widths, from 6 mm to 96 mm, were created out of the same plate. All specimens were taken from the same coupon which means that the welding parameters were identical for all specimens. In this case the effect of the width was obvious; the ultimate strength of the joint increased from 1070 MPa (6 mm) to 1192 MPa (96 mm).

The ultimate strength of the weld metal was, in this case, 1015 MPa.

When it comes to undermatching, all test specimens in Weldox 960 steel, welded with the stronger electrode, fractured in the base metal. This in spite of that the ratio between the weld metal’s and base metal’s yield strengths was as low as 0,77. In other literature only 15 % undermatching is

recommended to achieve the full base metal strength in a joint.

Only two different relative thicknesses could be accomplished. It was still obvious that the joint with a

large weld volume had a lower strength than the one with a smaller volume.

SAMMANFATTNING

Sammanfattning

De senaste årens förbättringar i ståltillverkningsindustrin har gett stålkvaliteter med högre hållfasthet, hårdhet, svetsbarhet etc. jämfört med konventionella konstruktionsstål. Den här sortens stål ger lättare och mindre konstruktioner och kostnadsminskningar.

Eftersom svetsgodset fortfarande enbart är en legering, vars hållfasthet i huvudsak beror på dess kemiska sammansättning, ökar skillnaderna i mekaniska egenskaper mellan grundplåt och tillsatsmaterial. Frågan är, således, hur man väljer matchningsgrad mellan tillsatsmaterial och grundplåt utan att avsevärt minska den globala hållfastheten i förbandet.

I det här examensarbetet undersöks vilken inverkan olika svetsgeometrier har på de mekaniska egenskaperna hos undermatchade svetsar i höghållfasta, seghärdade konstruktionsstål. Olika undermatchningsgrader har åstadkommits genom att kombinera två olika stål; Weldox 960 och Weldox 1100, tillsammans med två olika tillsatsmaterial; Filarc PZ 6145 och Filarc PZ 6149.

Två olika metoder har använts för att samla information; dels har 30 stycken provkroppar tillverkats och dragits sönder hos SSAB i Oxelösund och dels har tidigare undersökningar inom detta område studerats i en litteraturstudie.

Tre olika parametrar valdes ut för undersökning i de praktiska testerna, enbart en parameter ändrades åt gången:

• Sambandet mellan provstavsbredd och plåttjocklek

• Undermatchningsgrad

• Relativ tjocklek (förhållande mellan svetsgodsets bredd och plåttjockleken)

Hypotesen var att om svetsgodset är mjukare än kringliggande plåt kommer det att börja flyta innan grundmaterialet gör det. Då svetsen börjar deformeras, hindras denna deformation av att kringliggande material inte flyter. Dragspänningar utvecklas i svetsens tjockleks- och längsriktingar utöver den spänning som finns i lastriktningen. När svetsen påverkas av dragspänningar i fler än två riktningar ökar medelspänningen eller den hydrostatiska spänningen. Då den hydrostatiska spänningen ökar, p.g.a inspänning, minskar de deviatoriska spänningarna som styr flytgränsen. Detta ger en starkare svets som kan belastas mer.

Då sambandet mellan provstavsbredd och plåttjocklek utreddes tillverkades fem olika provstavsbredder från 6 mm till 96 mm, ur samma plåt. Alla prov togs ur samma svetskupong, d.v.s.

att svetsparametrarna var identiska för de olika proven. Här kunde breddens inverkan tydligt påvisas;

förbandets brotthållfasthet ökade från 1070 MPa (6mm) till 1192 MPa (96mm). Brottgränsen för svetsgodset var i det här fallet 1015 MPa.

Vad gäller undermatchningsgraden så gick samtliga provstavar i Weldox 960 plåt, svetsade med den starkare elektroden i brott i grundplåten. Detta trots att förhållandet mellan elektrodens- och grundplåtens flytgränser var så lågt som 0,77. I annan litteratur rekommenderas endast 15 % undermatchning för att kunna uppnå grundplåtens hållfasthet i förbandet.

Enbart två olika relativa tjocklekar kunde åstadkommas. Det var ändå tydligt att förbandet med stor

svetsgodsvolym hade lägre hållfasthet än det med en mindre volym.

TABLE OF CONTENTS

Table of Contents

Preface... i

Abstract ... ii

Sammanfattning ... iii

Table of Contents ... iv

List of Figures ... vii

Nomenclature ... ix

1 Introduction ...- 1 -

1.1 Background...- 1 -

1.2 Aim and method (purpose) ...- 1 -

1.2.1 Approach ...- 2 -

1.2.2 Gathering information ...- 3 -

1.3 Scope and demarcation ...- 3 -

1.4 Disposition of the report ...- 3 -

2 Welding ...- 5 -

2.1 History ...- 5 -

2.2 Welding Processes ...- 5 -

2.2.1 Manual Metal Arc welding (MMA) ...- 5 -

2.2.2 Submerged Arc Welding (SAW)...- 6 -

2.2.3 Gas Metal Arc Welding (GMAW) ...- 6 -

2.2.3.1 MAG (Metal Active Gas) ...- 7 -

2.2.3.2 MIG (Metal Inert Gas) ...- 7 -

2.2.3.3 FCAW (Flux-Cored Arc Welding) ...- 7 -

2.2.3.4 TIG (Tungsten Inert Gas)...- 7 -

2.2.3.5 Highly productive GMAW with solid electrode...- 7 -

2.3 Hardening/Heat treatment...- 8 -

2.4 Weld geometry...- 10 -

2.5 Heat Affected Zones ...- 11 -

2.5.1 Single pass welds...- 11 -

2.5.2 Multi run welds...- 13 -

2.6 Weld thermal cycle of base material...- 13 -

2.7 Weldability...- 14 -

2.8 Cracking...- 14 -

2.9 Electrode Matching...- 15 -

3 Testing methods ...- 19 -

3.1 Hardness testing ...- 19 -

3.1.1 Brinell hardness testing ...- 19 -

3.1.2 Vickers hardness test ...- 20 -

3.1.3 Rockwell hardness test ...- 20 -

3.2 Tensile testing ...- 20 -

3.3 Fracture toughness testing...- 22 -

3.3.1 Background...- 22 -

3.3.2 Test specimens...- 23 -

3.3.3 Instrumentation and loading ...- 24 -

3.3.4 Fracture toughness parameters ...- 24 -

3.3.5 Charpy impact test ...- 24 -

3.3.6 Definition...- 25 -

3.3.7 Quantitative results ...- 25 -

3.3.8 Qualitative results ...- 25 -

3.3.9 Quality classes ...- 26 -

TABLE OF CONTENTS

4 Theory ...- 27 -

4.1 Static strength of welded joints...- 27 -

4.2 Influence of heat input ...- 29 -

4.3 Influence of the steel type ...- 29 -

4.4 Softer weld metal ...- 30 -

4.5 Work hardening and necking ...- 30 -

4.5.1 Definitions of stress and strain ...- 31 -

4.5.2 Relationships between true and engineering stress and strain...- 32 -

4.5.3 Necking and triaxiality ...- 34 -

5 Previous work...- 35 -

5.1 Dexter (1997) ...- 35 -

5.1.1 Introduction ...- 35 -

5.1.2 Structural requirements...- 36 -

5.1.3 Significance of strain-hardening (Y/T ratio) ...- 38 -

5.1.4 Significance of constraint ...- 40 -

5.1.5 Effect of weld defects and cracks ...- 42 -

5.1.6 Conclusions ...- 42 -

5.2 Loureiro (2002)...- 42 -

5.2.1 Experimental procedure...- 43 -

5.2.2 Results and discussion ...- 43 -

5.2.2.1 Thermal cycles...- 43 -

5.2.2.2 Microstructures ...- 43 -

5.2.2.3 Hardness...- 43 -

5.2.2.4 Tensile properties...- 44 -

5.2.3 Conclusions ...- 46 -

5.3 Fernandes et al. (2004)...- 46 -

5.3.1 Introduction ...- 46 -

5.3.2 Procedure...- 47 -

5.3.3 Results and discussion ...- 49 -

5.3.3.1 Strength and ductility of the overall sample ...- 49 -

5.3.3.2 Study of the influence of constraint in the plastic behaviour of the HAZ ...- 52 -

5.3.3.3 Influence of the HAZ dimension in the constraint effect...- 53 -

5.3.3.4 Influence of the mismatch in the constraint effects ...- 55 -

5.3.4 Conclusions ...- 56 -

6 Laboratory tests ...- 57 -

6.1 Aim and approach ...- 57 -

6.1.1 Width- to Thickness...- 57 -

6.1.2 Undermatching ...- 57 -

6.1.3 Relative thickness ...- 58 -

6.1.4 Number of specimens ...- 59 -

6.2 Experimental Procedure...- 59 -

6.2.1 Base materials...- 60 -

6.2.2 Preparation of joints...- 60 -

6.2.3 Welding processes ...- 61 -

6.2.4 Electrodes ...- 61 -

6.2.5 Welding parameters...- 62 -

6.2.6 Non destructive testing ...- 62 -

6.2.7 Static tension tests ...- 62 -

6.2.8 Macro tests...- 63 -

7 Results ...- 64 -

7.1 Mechanical tests of welded joints ...- 64 -

7.2 Mechanical tests on weld metals...- 65 -

7.3 Macro tests ...- 65 -

7.3.1 Test No.1-1 ...- 66 -

7.3.2 Test No.1-2 ...- 66 -

TABLE OF CONTENTS

7.3.3 Test No.2 ...- 66 -

7.3.4 Test No.3 ...- 67 -

7.3.5 Test No.4 ...- 67 -

7.3.6 Test No.5 ...- 67 -

7.3.7 Test No.6 ...- 68 -

8 Analysis and Discussion...- 69 -

8.1 Width- to thickness relation ...- 69 -

8.2 Undermatching...- 70 -

8.3 Relative thickness ...- 70 -

8.4 Discussion (answers to questions asked in chapter 1) ...- 71 -

8.5 Possible Sources of Error...- 72 -

8.6 Future Work...- 72 -

9 References ...- 74 -

Appendix A – Examples of Calculations ...- 77 -

Appendix B – Stress-Strain Plots ...- 79 -

Appendix C – Welding Results ...- 84 -

Appendix D – Results from Macro Tests...- 91 -

Appendix E - Specimen Photos...- 98 -

LIST OF FIGURES

List of Figures

F

1. A

... - 2 -

F

2. T

... - 4 -

F

3. E

, B

(1987). ... - 8 -

F

4. I

-I

, T

(1985). ... - 9 -

F

5. T

, W

(2002)... - 10 -

F

6. C

, W

(2002). ... - 10 -

F

7. N

S

, W

(2002). ... - 11 -

F

8. H

, B

(1995)... - 12 -

F

9. S

-

HAZ, H

(2003)... - 13 -

F

10. E

(

), D

(1994). ... - 16 -

F

11. E

(A)

(B)

(

)

(

)

, D

(1994). ... - 17 -

F

12. S

–

, D

(1994). ... - 18 -

F

13. R

,

,

, D

(1994)... - 18 -

F

14. T

, T

(1985). ... - 21 -

F

15. F

, TWI (2007)... - 23 -

F

16. E

, TWI (2007). ... - 23 -

F

17. S

-

,

C

, Y

(1997).... - 26 - F

18. H

, SSAB O

(2007). ... - 27 -

F

19. S

, S

T

(1975). ... - 28 -

F

20. E

, S

T

(1975). ... - 29 -

F

22. T

-

,

-

, K

S

(2007) ... - 31 -

F

23. L

/

-

K

S

(2007). ... - 33 -

F

24. V

, σ = ⋅ K ε

K

S

(2007)... - 33 -

F

25. N

-

. ... - 37 -

F

26. N

-

-

-

. ... - 38 -

F

27. N

-

-

-

. ... - 38 -

F

28. E

-

HSLA-80

EH36 CCT

.

D

460

... - 39 -

LIST OF FIGURES

F

30. V

-

,

, 5

. WM—

; HAZ—

-

;

BM—

. B

95%

... - 44 -

F

31. V

WM,

. H

—. B

95%

. ... - 44 -

F

32. S

-

. T

(BM

WM)

HAZ

. Y

= 400 MP

... - 48 -

F

33. S

-

. T

(BM

WM)

HAZ

. Y

= 500 MP

... - 48 -

F

34. S

-

. T

(BM

WM)

HAZ

. Y

= 600 MP

... - 48 -

F

35. N

HAZ

HAZ

. ... - 49 -

F

36. E

HAZL4

HAZ. P

: 0.1, 1.0

2.5

. ... - 50 -

F

37. N

(M

),

HAZ

. ... - 50 -

F

38. D

,

HAZ

HAZ

... - 51 -

F

39. S

HAZL6

HAZL1

. ... - 52 -

F

40. S

(

)

HAZ

HAZ

= 400 MP

... - 53 -

F

41. S

(

)

HAZ

HAZ

= 500

600 MP

... - 54 -

F

42. E

HAZ

HAZL1

HAZL6. G

HAZ

... - 54 -

F

43. E

HAZL1

HAZ

: Y

= 400 MP

,

= 0,08

Y

= 400 MP

,

= 0,16. G

HAZ

... - 55 -

F

44. J

,

-V

... - 61 -

F

45. D

. ... - 63 -

F

46. E

-

... - 69 -

NOMENCLATURE

Nomenclature

Symbols

Symbol Explanation Unit

E Young’s modulus, the modulus of elasticity [GPa]

f

Ultimate strength (σ

, R

) [MPa]

f

Yield strength (σ

, R

R

) [MPa]

Abbreviations

AWS American Welding Society BCC Body Centred Cubic

BM Base Metal

CT Compact Tension

CTOD Crack Tip Opening Displacement

CVN Charpy V-Notch

DBTT Ductile to Brittle Transition Temperature FCAW Flux Cored Arc Welding

FCC Face Centred Cubic

FL Fusion Line

FZ Fusion Zone

GMAW Gas Metal Arc Welding GSY Gross Section Yielding HAZ Heat Affected Zone

HB Brinell Hardness

HPS High Performance Steel

HR Rockwell Hardness

HSLA High Strength Low Alloy

HV Vickers Hardness

IIW International Institute of Welding

MAG Metal Active Gas

MIG Metal Inert Gas

MMA Manual Metal Arc

MSYS Minimum Specified Yield Strength NSY Net Section Yielding

QT Quenched and Tempered

SAW Submerged Arc Welding SENB Single Edge Notch Bend SMAW Shielded Metal Arc Welding TIG Tungsten Inert Gas

TIME Transferred Ionized Molten Energy TTT Time Temperature Transformation TWI The Welding Institute

WM Weld Metal

YS Yield Strength

INTRODUCTION

1 Introduction

1.1 Background

In the last two decades, there have been significant improvements in steel making technologies, both in terms of metallurgical advances and rolling and heat treatment process developments. This have resulted in “high-performance steel” (HPS) grades, which offer higher performance in tensile strength, toughness, weldability, cold formability and corrosion resistance compared to the traditionally used steel grades, IABSE (2005).

Steel grades of this category generally lead to cost reductions, smaller sized components, lightweight structures and less welding work. For example, in medium and long span bridges, the weight reduction can reach 20%. Most importantly, these new grades contribute to a sustainable environment due to improved durability properties and reduced material use, IABSE (2005).

The upgrading of steels and new design rules for weight saving in constructions request higher demands on reliable welds, in terms of strength, fatigue resistance and safety against brittle fracture.

The question is how to meet these requirements and thus how to match the weld metal with the parent metal. By adopting new rolling and cooling techniques, the difference between the mechanical properties of the parent metal and the weld metal increases. The weld metal, contrary to the parent metal, is still a cast alloy whose strength mainly depends on its chemistry, Blomqvist (1995).

By the very nature, welded joints are highly inhomogeneous. The microstructure varies in different regions of welded joints, both in the weld metal and in the heat affected zone. Welded joints are therefore considered to be the weakest link in many engineering structures. In these structures, the design and selection of degree of matching level of welded joints are essential to the structural integrity, Blomqvist (1995).

1.2 Aim and method

The aim of this master’s thesis is to study the effect of different welding procedures (heat input) and weld-geometries (specimen width and relative thickness) on the mechanical properties of undermatched welds in high strength steel. The influence of these parameters on the performance of the welded joint under tension is analysed.

Questions to be answered are:

• Can the global strength of an undermatched test specimen achieve the base plate strength?

• What effect does the specimen width have on the performance of the joint and why?

(constraint).

• How does the heat input affect the performance of the joint and why?

• How much can a transverse butt weld be undermatched without loss of strength or ductility?

• What is the effect of heat input on the hardness in the WM and HAZ?

• How does the heat input affect the yield- and tensile strength undermatching?

• Is there a level of undermatching that will induce a concentration of plastic flow in the of the joint?

weakest zone and therefore reduce the strength and ductility

• Does the ductility decrease with decreasing undermatching?

• Does the ductility increase with a reduction of the effective thickness?

INTRODUCTION

• Are there any differences in the joint performance of quenched and tempered (Weldox

he aim will be attained by analyzing previous studies in this field and by static tension tests at SSAB

k is finished, some light will hopefully be shed on how undermatched butt welds in high trength WELDOX steel perform under tension and how important parameters influences this

he different activities that have been performed during the time for this thesis are illustrated in Figure 1. This is a rough model, where some activities overlap each other.

work with him. There as also some time for a study visit to the steel plant to explain the process to me and I got to see the welding equipment and testing hall that would be used for our tests.

1100) and quenched and low temperature tempered (Weldox 960) steel?

T

in Oxelösund.

When this wor s

performance.

T

Figure 1. Approach

Problem definition

Visit SSAB Oxelosund

Redefining the problem

Literature search

Literature search Test-planning

Practical tests, SSAB Oxelosund

Results and analysis

Discussion and conclusion

1.2.1 Approach

The project started off with establishing a definition of the problem. My supervisors at LTU/Ramböll, professors Peter Collin and Bernt Johansson, presented what intentions they had and what the expected outcome was. Simultaneously, a literature search was initiated to promote my own knowledge in the area. After some weeks, SSAB’s steel plant in Oxelösund was visited. I got to meet my supervisor at SSAB, Daniel Stemne, and had the opportunity to discuss the

w

INTRODUCTION

After having discussed the work with SSAB the problem was redefined when new, important, aspects were pointed out and some different welding methods introduced. Back in Luleå my literature search was intensified and a preliminary test plan was established. The test plan described what parameters

ere interesting to study and how many tests that would be performed.

t results. The conclusions that could be drawn from these tests were presented in the end f the report.

ture and have ministered articles. Other cientific articles have been found through Google Scholar™.

1.3 Scope and demarcation

tension tests.

he results will be analyzed and compared to previous studies on similar test specimens.

he only joints studied are butt joints loaded in tension transverse to the weld axis.

he electrode strengths are decided to be:

• f = 580-680 [MPa]

tress-strain curves are to be plotted for the parent metal, the electrodes and the specimens.

5 (failure elongation) is to be analyzed.

ield- and ultimate strength is to be recorded.

is made of WELDOX

1100 & 960 with a yield strength (R

) of 100 & 960 [MPa] respectively.

1.4 Disposition of the report

ome metallurgy, heat affected zones the parent plate, weldability, cracking and electrode matching.

esting metal and welds. For example; hardness sting, tensile testing and fracture toughness testing

d to occur when varying arameters like the weld geometry, the heat input and undermatching level.

welds and a numerical study of the plastic behaviour in tension of welds in high trength steel.

w

In the beginning of April the testing could start, after this our own results were analyzed and compared to previous tes

o

1.2.2 Gathering information

Theory and information from many different sources have been used to get several opinions and a wide intake of material. The foundation of the literature search is frequently reappearing material in studies concerning undermatching butt welds loaded in tension. The literature search has been done with the help of our university library’s data base “Lucia” and by accessing TWI’s and IIW’s search engines. My supervisors have helped me to find suitable litera

s

Within this master’s thesis, a total of 30 test specimens will loaded until fracture, in static T

T T

• f

= 950-1050 [MPa]

u

S A Y

The steel plates used in the tests 1

Chapter 2 describes different welding methods, weld geometry, s in

Chapter 3 describes parameters and methods for t te

Chapter 4 describes the theory behind the phenomena that is anticipate p

Chapter 5 presents previous work in this field. Three different papers are described which discusses the structural behaviour of undermatched welds, the effect of heat input on plastic deformation in undermatched

s

INTRODUCTION

Chapter 6 describes how the practical tests were performed.

hapter 7 presents results from the tension tests

nswers the questions asked in this chapter, describes possible ources of error, looks into the future

igure 2. The structure of the report C

Chapter 8 analyses the test results, a s

F

Chapter 1 Introduction

Chapter 2 Welding

Chapter 3 Testing methods

Chapter 4 Theory

Chapter 5 Previous work

Chapter 6 Experimental procedure

Chapter 7 Results

Chapter 8 Analysis and Discussion

WELDING

2 Welding

2.1 History

Welding is a fabrication process that joins materials, usually metals or plastics, by causing coalescence. This is often done by melting the work pieces and adding a filler material to form a pool of molten material that cools to become a joint, Weman (2002). This is in contrast with, for example soldering, where a material with a lower melting-point forms a bond between the work pieces without melting the base metal.

People have been joining metals for several thousand years. The earliest examples of welding are from the Bronze and Iron Ages in Europe and the Middle East. During the medieval period advances were made in forge welding, in which blacksmiths pounded heated metal until bonding occurred. However when the Industrial Revolution, in the 19

century, took place welding was transformed. Invention of metal electrodes and the discovery of the electric arc changed the available techniques, according to Eriksson (1989).

Most welding processes require high temperatures to unify the metals. Several different energy sources can be used, including a gas flame, an electric arc, a laser, an electron beam, friction and ultrasound, Svetskommissionen (2007).

When modern welding was introduced, arc welding and its different welding methods became the largest and mostly used process. As the name indicates, the heat source is an electric arc between the base material and an electrode. The power supply can, according to Weman (2002) use either direct (DC) or alternating (AC) current and consumable or non-consumable electrodes are used. Electrical energy, transformed to heat, can generate an arc temperature up to 7000˚C.

One of the main problems concerning welding is that heated metals react with the surrounding air. To protect the hot metal from the air is therefore important in welding. There are many different methods to protect the weld puddle from the air, ranging from flux covered electrodes (giving a protective slag) to inert or active shielding gases. Sometimes all the air is removed and welding is performed in vacuum. In 1906, the Swede Oscar Kjellberg obtained a patent for the covered electrode. He used welding to repair things but was not satisfied with his results. He covered his electrode with a material that melted and produced shielding slag. The result was extraordinary good and made the foundation for one of today’s largest welding companies, ESAB (Elektriska Svetsnings AB), Weman (2002).

2.2 Welding Processes

Below is a short description of some common welding processes, Bergh (1987) and Svetskommissionen (2007).

2.2.1 Manual Metal Arc welding (MMA)

Welding with covered electrodes or MMA is also known as Shielded Metal Arc Welding (SMAW) or

stick welding. The electrode is made up of a core of steel which is covered with a flux. Electric

current strikes an arc between the base material and the consumable electrode rod. When liquefied

metal drops from the electrode core are transported by the arc into the weld pool they are protected

from the air by CO

gas formed by substances in the flux. The slag floats to the top of the puddle

where it protects the molten material during the hardening.

WELDING

Electrodes are divided into three different groups depending on the chemical composition of the slag;

acid, basic and rutile.

Acid electrodes give an even and shiny pass. The slag is easy to remove. The weld has a lower yielding point and tensile strength than rutile or basic electrodes but gives a larger possible elongation.

This kind of electrode is very rare in Sweden today.

Rutile electrodes are easy to weld with and to ignite. The risk of hydrogen cracks limits the usage to carbon steels with a yield strength less than 440 [MPa] or carbon-manganese steels.

Basic electrodes give a high quality weld in terms of strength, ductility and safety against heat induced cracks. The slag is normally harder to remove. Basic electrodes are hydroscopic and must therefore be protected from moisture.

In spite of the rather long weld times due to electrode changes and slag chipping, it is probably the most common welding process for construction steels. Equipment is relatively inexpensive and the method is well suited for both the workshop and field jobs.

Manual metal arc welding can be used for plate thicknesses larger than 2 mm.

2.2.2 Submerged Arc Welding (SAW)

SAW-welding is a highly productive, mechanized welding method that can be performed with one or several continuous electrodes. The arc(s) are burning under a layer of protective flux that melts close to the arc and produces slag on the weld. The non-molten excess flux can be recycled. The arc is not visible under the flux, hence the name.

The method is suitable for plate thicknesses above 2 mm and is often used for long horizontal welds, like for example beams.

2.2.3 Gas Metal Arc Welding (GMAW)

In this method an arc is maintained between a continuous wire-electrode and the work piece. The weld is protected from contamination by an inert or active gas. Since the electrode is fed continuously, welding speeds are greater for GMAW than for MMA. Outside welding should be avoided since the protection from the gas decreases even in light winds.

Gas metal arc welding with solid electrode is applied with short circuited metal transfer, called short- arc GMAW or with small molten metal droplets, called spray-arc GMAW.

The short-arc welding is performed with low current and voltage and the electrode wire is thin. As a result the heat input is reduced, making it possible to weld thinner material while decreasing the amount of distortion and residual stress in the weld area. The droplets form on the tip of the electrode, but instead of dropping into the weld pool they bridge the gap between the electrode and the weld pool as a result of the greater wire feed rate. This causes a short circuit and extinguishes the arc, but is quickly reignited after the surface tension of the pool pulls the molten metal bead off the electrode tip.

This process is repeated about 100 times per second, making the arc appear constant to the human eye.

Spray-arc was the first metal transfer method used in GMAW. In this variation, molten metal droplets

are rapidly passed along the stable electric arc from the electrode to the work piece. High amounts of

voltage and current are necessary together with a relatively thick electrode. This means that the

process involves high heat input and a large weld area and heat affected zone. As a result it is

generally used on work pieces of thicknesses above 4 mm.

WELDING

2.2.3.1 MAG (Metal Active Gas)

In MAG-welding the protective gas is either pure carbon dioxide or a gas mixture of 80% argon and 20% CO

. The mixture gives a softer arc, smoother weld and fewer spatters but is a bit more expensive than only carbon dioxide. It also gives higher strength and ductility to the weld. The mixture is therefore commonly used for thin plates and low alloy steels. Today’s mixtures can also contain He, O

and H

.

Plate thicknesses that can be welded varies from 0,8 mm and upwards.

2.2.3.2 MIG (Metal Inert Gas)

The equipment used in MIG-welding is basically the same as in MAG-welding. The most common protective gas is pure argon but mixtures of argon and helium are used. Helium mixtures give a hotter weld and are used for thick walled materials like aluminium and copper. MIG-welding is used for;

• stainless steels

• a lum inium

• copper

and for thicknesses greater than 1 mm.

2.2.3.3 FCAW (Flux-Cored Arc Welding)

FCAW requires a continuously fed, consumable, tubular electrode containing a flux. The welding power supply is of constant voltage or, less commonly, a constant electric current. An externally supplied shielding gas is sometimes used, but often the flux itself is relied upon to generate the necessary protection from the atmosphere. The process is widely used in construction because of its high welding speed and portability.

Welding without the externally supplied shielding gas is less sensitive to winds and can be used in outside conditions.

This method is for steel thicker than 2 mm.

2.2.3.4 TIG (Tungsten Inert Gas)

Characteristic for TIG-welding is a non-consumable electrode and possibly a separate filler material.

The electrode, which has a higher melting point than the weld temperature, is normally made of tungsten or a tungsten alloy. The protective gas is of the same type as in MIG-welding. An inert gas implies that the gas doesn’t take part in any chemical reaction in the process (burning alloys or carbonizing the weld).

This method gives an extremely clean weld of high quality because no slag is produced. It is often used in areas where the quality demands are high like nuclear-, chemistry-, airplane- and food industries. It is also applicable for metals that are hard to weld like titanium, Monel, copper-nickel etc.

Normal dimensions for TIG-welds are 0,5-6 mm.

2.2.3.5 Highly productive GMAW with solid electrode

In recent years mechanized MIG/MAG-welding with conventional solid electrode has developed towards higher productivity, Weman (2002). One of the predecessors was Canadian John Church who introduced the T.I.M.E.-method (Transferred Ionized Molten Energy). In contrast to traditional MAG- welding, a long electrode stick-out loaded with high current, is used on purpose. The resistive heating makes the electrode preheated, which allows for a higher feeding speed without increasing the current.

The T.I.M.E.-method comprises a special, patented, 4-component shielding gas. The welding

companies AGA and Linde have studied and developed the method with new gas mixtures. They call

their methods Rapid Processing™ and Linfast

.

WELDING

With higher feeding speeds the productivity increases as well, sometimes up to 20 kg molten electrode per hour. The welding speed can be doubled, compared to conventional MAG-welding, with sustained weld deposit appearance and penetration profile. Different arc types are used. The most commonly used is a type of forced short arc which can be used on ordinary welding equipment.

Yet another method to increase productivity and welding speeds is to use double electrodes. Both electrodes can be connected to the same source of current, this implies a more or less common arc (twin arc). If you instead have double sources of current, each connected to a separate electrode, it is referred to as tandem welding. The electrodes are still so close to each other that they weld in the same weld pool.

The weld speeds can be at least doubled with double electrodes, in thin plates welding speeds can be even higher.

2.3 Hardening/Heat treatment

The atoms in a metal are arranged in a unique order that can be different for different metals and metal phases. The smallest structure, where this arrangement is periodically repeated in three dimensions, is called the metal’s unit cell. One of the most common and simplest shapes found in metallic crystals is the cubic crystal system, where the unit cell is in the shape of a cube. Examples of two different unit cells are shown in Figure 3. Both are cubic. The body centred cubic (BCC) system has one lattice point in the centre of the unit cell in addition to eight corner points. The face centred cubic (FCC) system is more densely packed. Here, the atoms are placed in the corners of the cube and in the mid- point of each surface of the cube, Bergh (1987).

a) Body Centered Cubic (BCC) b) Face Centered Cubic (FCC)

Figure 3. Examples of cubic unit cells, Bergh (1987).

Steel is defined as an iron-carbon alloy with carbon content between 0,02wt% and 1,8wt%, which is the maximum solubility of carbon for austenite. Even in the narrow range of concentrations that make up steel, mixtures of carbon and iron can form into a number of different structures, or allotropes, with very different properties. At room temperature, the most stable form of iron is the BCC structure ferrite or α-iron, a fairly soft metallic material that can dissolve only a small concentration of carbon (no more than 0,02wt% at 910˚C). Above 910˚C ferrite undergoes a phase transition from BCC to an FCC structure, called austenite or γ-iron, which is similarly soft and metallic but can dissolve considerably more carbon. Ferrite is magnetic and austenite is non-magnetic. As carbon-rich austenite cools, the mixture attempts to revert to the ferrite phase, resulting in an excess of carbon. One way for carbon to leave the austenite is for cementite to precipitate out of the mix, leaving behind iron that is pure enough to take the form of ferrite and resulting in a ferrite-cementite mixture. Cementite or iron carbide is a chemical compound with the formula Fe

C. It is a hard, brittle material, containing 6,67wt% carbon, Thelning (1985).

A phase transformation from one solid into two different solids is called a eutectoid reaction, Bergh

(1987). In the Fe-C system, cementite forms in regions of higher carbon content while other areas

revert to ferrite during the eutectoid reaction. In microscope this structure has a pearl-like appearance

WELDING

and is so forth known as pearlite. Pearlite is therefore a structure that consists of alternating layers of soft, almost carbon free, ferrite and hard cementite. The eutectoid point in the Fe-C system is at approximately 0,8wt% C and 723˚C according to the iron-carbon phase diagram below.

Figure 4. Iron-Iron carbide phase diagram, Thelning (1985).

See Appendix A for an example on how to calculate the ratio between ferrite and cementite in pearlite.

Perhaps the most important allotrope is martensite, a chemical substance with about four to five times the strength of ferrite. A minimum of 0,4wt% of carbon is needed to form martensite. When the austenite is quenched to form martensite, the carbon is “frozen” in place when the structure changes from FCC to BCC, Thelning (1985). The carbon atoms are much too large to fit in the interstitial vacancies and thus distort the cell structure into a Body Centred Tetragonal (BCT) structure.

Martensite and austenite have an identical chemical composition. As such, it requires extremely little thermal activation energy to form.

The heat treatment process for most steels involves heating the alloy until austenite forms, quenching (rapidly cooling) the hot metal in water or oil, cooling it so rapidly that the transformation of ferrite or pearlite does not have time to take place. The transformation into martensite, by contrast, occurs almost immediately, due to lower activation energy.

Martensite has a lower density than austenite, so that the transformation between them results in a change of volume. In this case expansion occurs. Internal stresses from this expansion generally take the form of compression on the crystals of martensite and tension on the remaining ferrite, with a fair amount of shear on both constituents. If quenching is done improperly, these internal stresses can cause a part to shatter as it cools; at the very least, they cause work hardening and other microscopic imperfections.

At this point, if the carbon content is high enough to produce a significant concentration of martensite, the result is an extremely hard but very brittle material. Often, steel undergoes further heat treatment at a lower temperature to destroy some of the martensite (by allowing enough time for cementite and ferrite to form) and help settle the internal stresses and defects. This softens the steel, producing a more ductile and fracture-resistant metal. This process is known as tempering, which forms tempered steel, Thelning (1985).

Flanges of quenched and tempered steel can be of interest in so called hybrid beams to save weight

and costs for manufacturing. A hybrid beam is a beam with flanges stronger than the web, for example

steel quality S 420 in the web and S 690 in the flanges, Blomqvist et al. (1996).

WELDING

2.4 Weld geometry

The weld geometry is chosen considering the welding process and thickness of the work piece. It is important to choose a geometry which satisfies the required demands on strength and quality without an unnecessarily large weld volume. The welding costs increases with the volume and an increase of the heat input can cause fracture resistance- and deformation problems Weman (2002).

Figure 5. Terminology for different parts of the joint, Weman (2002).

Groove angle

Bevel angle

Root opening Root

Root face and Groove face

Welds can be shaped geometrically in many different ways. The five basic types of weld joints are (Figure 6); the butt joint, the lap-joint, the corner joint, the edge joint and the T-joint, Eriksson (1989).

Other variations exist as well – for example double-V preparation joints are characterized by the two pieces of material each tapering to a single centre point at one-half of their height. Single-U and double-U preparation joints are also fairly common. Lap joints are often more than two pieces thick, depending on the process used and the thickness of the material, many pieces can be welded together in a lap-joint geometry.

Figure 6. Common welding geometries, Weman (2002).

1a: Square butt joint

1b: Double-V preparation butt joint 1c: Single-U preparation butt joint

2: Lap-joint

3: Full open corner joint

4: Flanged edge joint

5: T-joint

WELDING

Often particular joint designs are used, almost exclusively, by certain welding processes. For example, resistance spot welding, laser beam welding, and electron beam welding are most frequently performed on lap-joints. However, some welding methods, like shielded metal arc welding, are extremely versatile and can weld virtually any type of joint, Weman (2002).

2.5 Heat Affected Zones

After welding, a number of distinct regions can be identified in the weld area. The weld itself is called the fusion zone – more specifically this is where the filler material has blended together with the base metal to a hardened weld puddle, Weman (2002). The properties of the fusion zone depend primarily on the filler material used and its compatibility with the base materials. It is surrounded by the heat- affected zone, the area that had its microstructure and properties altered by the weld. These properties depend on the base material’s behaviour when subjected to heat. The metal in this area is often weaker than both the base material and the fusion zone and is also where residual stresses are found.

In the heat affected zone of a welded joint the peak temperature has been high enough to produce solid-state microstructural changes, but too low to cause melting. The size, microstructural characteristics and mechanical properties of the HAZ are a function of the type of base metal being welded, chemical composition of the base metal, heat input and the part thickness. The transduced zone in Figure 7 is commonly divided into; a normalized zone and an over-heated zone, Weman (2002).

Figure 7. Nomenclature for boundaries and different zones according to the Swedish welding commission, Weman (2002).

Initial joint surface

Non-affected base material

Penetration zone Transduced zone

Structurally changed zone Melting boundary

Transmission boundary Boundary for structural changes

Weld (fusion zone) Weld affected base material

Weld-affected area (HAZ)

Regions that are expected in single- and multi pass HAZ are described in the next section.

2.5.1 Single pass welds

The heat affected zone of a single bead is divided into four major zones, each with its own properties, Figure 8.

Region Approximate peak temperature

Coarse-grained HAZ (CGHAZ) 1100˚C – melting point Fine-grained HAZ (FGHAZ) Ac

- 1100˚C

Intercritical HAZ (ICHAZ) Ac

– Ac

*

The temperature at which austenitic transformation is completed during the heating process

†

The temperature at which austenitic transformation commences during the heating process

WELDING

Subcritical HAZ (SCHAZ) below Ac

The coarse-grained zone corresponds to the over-heated zone in Figure 7, the fine-grained zone corresponds to the normalised zone and the intercritical- and subcritical zones correspond to the structurally changed zone.

In the coarse-grained zone embrittlement due to grain growth is the problem, Blomqvist (1995). Using a low heat input, the thermal cycle will be such that only a comparatively short time is spent above the temperature which allows grain growth. This will affect the zone in two ways. Firstly, the austenite grain growth will not be excessively large and, secondly, only a very narrow coarse-grained zone will form due to the limited extension of the area where the temperature allows grain growth. Moreover a narrow CGHAZ limits the maximal grain size possible to develop in the zone. The disadvantages associated with a low heat input is the low Δt

-value which promotes the formation of a hard and brittle martensitic structure and the appearance of weld discontinuities such as lack of fusion.

The microstructure in the fine-grained zone is normally not associated with any problems. The fine austenite grains form in the temperature regime of about 900˚C to 1075˚C and transforms during cooling to fine ferrite grains with excellent toughness according to Blomqvist (1995).

The ICHAZ is the zone which has become austenite in part after being heated to midway between the Ac

point and the Ac

point, Hamada (2003). Concentrations of austenite formers, such as carbon, occurs in the parts which have become austenite and a hardened phase can sometimes develop following cooling. This hardened phase becomes a material “notch” resulting in low toughness.

The last zone, before reaching the unaffected parent metal is the subcritical zone. This zone is difficult to distinguish from the base metal by the optical microstructure. Hardness slightly declines due to the tempering effect. Strain ageing is the most common embrittling process and can occur if the joint has been deformed plastically at high strains, Blomqvist (1995).

Figure 8. Heat affected zone of a one run weld, Blomqvist (1995).

WELDING

2.5.2 Multi run welds

Most often we are dealing with, not just one run welds, but multiple run welds, which makes it even more difficult to evaluate the mechanical properties along a cross section in the HAZ. Figure 9 illustrate the influence on the microstructure at different locations along the fusion line in a multi-run weld, due to the heat generated from the following pass or passes. The HAZ will be exposed to reheating developing new, reheated, areas, Blomqvist (1995). The relationship between the microstructure and thermal cycles is shown in Figure 9. All positions were heated by bead one to temperatures close to the melting point and transformed into the same coarse grained structure (position D). These areas are then reheated by bead two and differ from each other in the peak temperature according to the distance from the fusion line.

Figure 9. Structural distribution within multi-layer welded joint HAZ, Hamada (2003).

2.6 Weld thermal cycle of base material

An interesting fact, in most practical welding situations, is that the velocity of the arc along the work piece is much higher than the velocity of thermal diffusivity, Easterling (1992). In other words, heat flow in the welding direction is small compared to that perpendicular to the travel direction. This implies that a given slice of the material of the base plate “feels” the weld cycle as a short, intense, heat pulse. The time taken to dissipate this heat pulse within the plate is proportional to the thermal conductivity λ. On this basis, the thermal cycle for a given point in the base metal can be represented by a simple temperature-time (TTT) diagram. From experimental measurements of the weld thermal cycle it has been found that, for a given welding process, weld geometry and material, the cooling time through the range 800-500˚C is constant. The temperature of 800˚C is important in that most steels it approximately represents the Ac

transformation temperature. The use of Δt

as descriptive of a given weld is widely adopted in welding circles.

The thermal diffusivity of the base material plays a large role – if the diffusivity is high, the material

cooling rate is high and the HAZ is relatively small. Alternatively, a low diffusivity leads to slower

cooling and a larger HAZ. The amount of heat input by the welding process is important as well; a

high heat input increases the size of HAZ.

WELDING

For thick plates, Δt

is approximately proportional to the heat input Q.

Q k t

≈ ⋅

Δ

(2.1)

where k is in the range of 4-5 seconds per kJ/mm for conventional structural steels, Qiu et al. (2001).

2.7 Weldability

Most often, the quality of a weld is judged by its strength and the strength of the material around it.

Many factors influence this, including the welding method, the amount and concentration of heat input, the base material, the filler material, the flux, the design of the joint and interactions between all these factors, Weman (2002). To test the quality of a weld either destructive or non-destructive methods are used to verify that welds are defect-free, have acceptable levels of residual stresses and distortion and have acceptable heat-affected zone (HAZ) properties.

Not all metals are suitable for welding and all filler metals do not work well with acceptable base materials.

The weldability of steel is inversely proportional to a property known as the hardenability of the steel, which measures the ease of forming martensite during heat treatment. The hardenability of the steel depends on its chemical composition, with higher concentrations of carbon and other alloying elements such as manganese, chromium, silicon, molybdenum, vanadium, copper and nickel tend to increase the hardness and decrease the weldability of the material, Easterling (1992). Each of these substances influences the hardness and weldability to different magnitudes. This makes a method of comparison necessary to judge the difference in hardness between two alloys made of different alloying elements. A measure known as the equivalent carbon content is used to compare the relative weldabilities of different alloys by comparing their properties to plain carbon steel.

15 Ni Cu 5

V Mo Cr 6 C Mn

E

+

+ + + +

+

= (2.2)

A steel is considered fully weldable if E

is below 0,41. For higher E

levels the steel is denoted limited weldable, which in general means that welding should be performed with a pre-heated work- piece to reduce the cooling rate Weman (2002).

Stainless steels, because of their high chromium content, tend to behave differently with respect to weldability than other steels. Austenitic grades of stainless steel tend to be the most weldable, but they are especially sensitive to distortion due to their high coefficient of thermal expansion. Hot cracking is possible if the amount of ferrite in the weld is not controlled, to alleviate this problem an electrode is used that deposits a weld metal containing a small amount of ferrite. Other types of stainless steels, such as ferritic and martensitic stainless steels, are not as easily welded and must often be preheated and welded with special electrodes Weman (2002).

2.8 Cracking

Welding methods that involve melting metal at the site of the joint are prone to shrinkage as the heated metal cools. Shrinkage, in turn, can introduce residual stresses and both longitudinal and rotational distortion, Eriksson (1989). Distortion can pose a major problem since the final product is not the desired shape. To alleviate rotational distortion, the work pieces can be offset, so that the welding results in a correctly shaped piece. Other methods of limiting distortion, such as clamping the work pieces in place, cause the build-up of residual stress in the heat-affected zone of the base material.

These stresses can reduce the strength of the base material and can lead to failure through cold

cracking.