Markku Pirinen
THE EFFECTS OF WELDING HEAT INPUT ON THE USABILITY OF HIGH STRENGTH STEELS IN WELDED STRUCTURES
Acta Universitatis Lappeenrantaensis 514
Thesis for the degree of Doctor of Science (Technology) to be presented with due permission for public examination and criticism in Auditorium 1381 at
Lappeenranta University of Technology, Lappeenranta, Finland, on the 25th of May, 2013, at noon.
2 Supervisor Professor Jukka Martikainen
Faculty of Technology
Department of Mechanical Engineering Lappeenranta University of Technology Finland
Reviewers Professor Victor Karkhin
Department of Welding and Laser Technologies St.Petersburg State Polytechnical University 29 Polytechnicheskaya, St. Petersburg 195251 Russia
Professor emeritus Algirdas Bargelis
(Honorary Doctor of Lappeenranta University of Technology) Faculty of Mechanical Engineering and Mechatronics Department of Manufacturing Technologies
Kaunas University of Technology Kęstučio St. 27, LT-44025 Kaunas Lithuania
Opponents Professor Victor Karkhin
Department of Welding and Laser Technologies St.Petersburg State Polytechnical University 29 Polytechnicheskaya, St. Petersburg 195251 Russia
Professor emeritus Algirdas Bargelis
(Honorary Doctor of Lappeenranta University of Technology) Faculty of Mechanical Engineering and Mechatronics Department of Manufacturing Technologies
Kaunas University of Technology Kęstučio St. 27, LT-44025 Kaunas Lithuania
ISBN 978-952-265-399-4 ISBN 978-952-265-400-7 (PDF)
ISSN-L 1456-4491 ISSN 1456-4491
Lappeenrannan teknillinen yliopisto Yliopistopaino 2013
i ABSTRACT
Markku Pirinen
The effects of welding heat input on the usability of high strength steels in welded structures.
Lappeenranta 2013
174 pages plus 4 appendices at 4 pages Acta Universitatis Lappeenrantaensis 514 Diss. Lappeenranta University of Technology ISBN 978-952-265-399-4
ISBN 978-952-265-400-7 (PDF) ISSN-L 1456-4491, ISSN 1456-4491
High strength steel (HSS) has been in use in workshops since the 1980s. At that time, the significance of the term HSS differed from the modern conception as the maximum yield strength of HSSs has increased nearly every year. There are three different ways to make HSS. The first and oldest method is QT (quenched and tempered) followed by the TMCP (thermomechanical controlled process) and DQ (direct quenching) methods.
This thesis consists of two parts, the first of which part introduces the research topic and discusses welded HSS structures by characterizing the most important variables. In the second part of the thesis, the usability of welded HSS structures is examined through a set of laboratory tests.
The results of this study explain the differences in the usability of the welded HSSs made by the three different methods. The results additionally indicate that usage of different HSSs in the welded structures presumes that manufacturers know what kind of HSS they are welding. As manufacturers use greater strength HSSs in welded structures, the demands for welding rise as well.
ii
Therefore, during the manufacturing process, factors such as heat input, cooling time, weld quality, and more must be under careful observation.
Keywords: high strength steel, usability, heat input, cooling time, high strength steel filler metal
UDC 678.029.43:621.791:624.078.45:624.014.2
iii ACKNOWLEDGEMENTS
This thesis has been carried out in the Department of Mechanical Engineering at Lappeenranta University of Technology.
I would like first to thank Professor Jukka Martikainen for his guidance throughout this process. Your support in the major point of my work gave me bottom line that I can clarify in this journey.
I want to express my utmost gratitude to Dr. Paul Kah, Dr. Mika Lohtander and Professor Timo Björk. You have given me a positive example to follow and great advice to help me to finish this thesis. Timo, you always supported me in my endeavors despite that fact that you were often very busy.
I offer my sincere thanks to my colleagues for their friendly support and for our pleasant working atmosphere. Special thanks go to Harri Rötkö, Antti Heikkinen, Antti Kähkönen and Esa Hiltunen. You have done great work in the laboratory during test processes. I also wish to thank the department secretaries, Ms. Kaija Tammelin and Anna-Kaisa Partanen, for all their support in administrative issues. I also cannot forget the work of all the steel structures laboratory staff. You are all professional and I am proud that I have had opportunity to research with you.
There are also many other people from Lappeenranta University of Technology that have not been mentioned, but I believe they know their contribution to this dissertation. Thank You.
I thank my proofreader Miss Jennifer Riley. You have worked hard to correct my thesis into flowing English.
Despite the distance between our homes, my children, their spouses, and my grandchildren are always on my mind. Your comments and lovely support during this process have been the power which has seen me through this work.
iv
My dearest Pirjo- thank you for your affection and patience during this journey.
Without you, this never would have been possible.
v CONTENTS
ABSTRACT
ACKNOWLEDGEMENTS TABLE OF CONTENTS
LIST OF ABBREVIATIONS AND SYMBOLS
Standards ... xii
1. INTRODUCTION ... 14
1.1. Background ... 14
2. STATE OF ART ... 16
2.1. What is HSS? ... 17
2.2. Effects of alloying elements in HSS and in its weld ... 19
2.2.1. Aluminium and Silicon ... 22
2.2.2. Niobium ... 23
2.2.3. Vanadium ... 24
2.2.4. Titanium ... 25
2.2.5. Zirconium ... 27
2.2.6. Boron and Copper... 27
2.2.7. Manganese and Nickel ... 28
2.2.8. Rare-earth elements ... 28
2.3. Microstructure of welded HSS structure ... 29
2.3.1. Microstructure and physical features of the HAZ ... 31
2.3.2. Microstructure of weld ... 34
2.4. Undermatched, matched and overmatched filler metal ... 37
2.5. Heat input and cooling time ... 42
3. SCOPE OF THE RESEARCH ... 47
4. AIM OF THE RESEARCH... 50
5. RESEARCH METHODS ... 52
6. EXPERIMENTAL INVESTIGATIONS ... 53
6.1. Experimental arrangement... 53
6.1. Joint geometries and preparation ... 55
6.3. Test set up ... 58
6.4. Material properties ... 61
6.5. Standard tests ... 69
6.6. Additional material test ... 71
6.6.1. CTOD test ... 72
6.6.2. Compared microstructure examination ... 81
7. RESULTS AND DISCUSSION ... 83
7.1. Visual test ... 83
7.2. Macro photography ... 83
7.3. Micro photography ... 92
7.4. Radiographic tests ... 103
7.5. Surface crack detection ... 103
7.6. Transverse tensile test ... 104
7.7. Transverse bend test ... 112
7.8. Impact test ... 115
7.9. Hardness test ... 123
7.9. CTOD tests ... 129
vi
7.11. Additional microstructure tests ... 135
7.11.1. Microstructure of the base material ... 136
7.11.2. Microstructure of weld metal ... 137
7.11.3. Microstructure of HAZ of QT and TMCP HSS ... 138
7.11.4. Comparison of HAZ microstructure of steels QT and TMCP ... 147
7.11.5. Microstructure study of CTOD samples after simulated welding thermal cycle ... 150
8. DIVERGENCE IN MANUFACTURERS’ HSS’s WITH DIFFERENT HEAT INPUTS ... 152
9. CONCLUSIONS ... 154
10. FUTURE WORK ... 157
11. SUMMARY ... 158
References... 160
vii LIST OF ABBREVIATIONS AND SYMBOLS
Abbreviations Explanation
9R Cu A copper particle type
A Ampere
A Austenitising
A5 Elongation at break %
AC1 The temperature at which austenite starts to form when heated.
Ac3 In hypoeutectoid steel, the temperature at which the transformation of ferrite into austenite is completed.
AcC Accelerated-Cooled
AF Acicular Ferrite
AHSS Advanced High Strength Steel
Al Aluminium
APFIM Atom Probe Field Ion Microscopy
ASTM American Society for Testing and Materials a/W Overall crack depth/ specimen width
B Boron
BH Bake Hardenable
Bs Temperature where bainite starts to form
C Carbon
CCT Continue-Cooling-Temperature (diagram)
CEV Carbon Equivalent Value (IIW)
CET Carbon Equivalent Value (SEW 088) CGHAZ Coarse-Grain Heat Affected Zone
CJP Complete Joint Penetration
CMn Carbon Manganese
Cr Chromium
CTOD Crack-Tip Opening Displacement
Cu Copper
DP-CP Dual Phase or Complex Phase
DQ Direct Quenching
viii
DQ&T Direct Quenching and Tempering
E Welding Energy
EN European Standard
exp Exponent
Fe Iron
FCAW Flux-Cored Arc Welding
FGHAZ Fine Grane Heat Affected Zone
FSP Ferrite Site Plate
GBF Grain Boundary Ferrite
GMA Gas Metal Arc
GMAW Gas Metal Arc Welding
HAZ Heat Affected Zone
HB Brinell Hardness
HBW Brinell Hardness specifies the use of a tungsten car- bide ball indenter
HIZ Heat Impact Zone
HSLA High Strength Low Alloy
HSS High Strength Steel
HV Vickers Hardness
HY High Yield Strength
ICCGHAZ Intercritically reheated Coarse-grain Heat Affected Zone
ICHAZ Inter-Critical Heat Affected Zone IF-HS High Strength Interstitial Free IIW International Institute of Welding
IS Isotropic
ISO International Standard Organization
J Joule
lHAZ/e HAZ width to sample thickness
K Kelvin
kg Kilogram
kJ/mm Kilo Joule/ millimeter
M Thermomechanically rolled
ix M-A, M/A Martensite-Austenite
MAG Metal Active Gas (welding)
Mg Magnesium
MIG Metal Inert Gas
min Minute
ML Lath Martensite
mm Millimeter
Mn Manganese
MnS Manganese Sulphate
Mo Molybdenum
MPa MegaPascal
MS Martensitic
Ms Temperature where martensites start to form
N Nitrogen
N Normalized
N Newton
Nb Niobium, Columbium
NDT Non-destructive Testing
Ni Nickel
Nital HNO3 + ethanol
O Oxygen
P Phosphorus
P180, P400 Degree of coarseness
PCM Carbon equivalent formula according to Ito-Bessyo
pf polygonal ferrite
PF Pearlite and Ferritic
PJP Partial Joint Penetration
ppm Parts per million
pWPS Preliminary Welding Procedure Specification
Q Quenched
Q Heat amount
QL Quenched and Tempered+ Low notch toughness
temperature
x
QT Quenched and Tempered
S Sulphur
s Second
s Plate Thickness
SA-Weld Submerged Arc Weld
SE(B) Three point bend specimen
SFS Finnish Standard Association
Si Silicon
SiC Silicon carbide
SMA Submerged Arc (Welding)
Sn Tin
StPSPU St. Petersburg State Polytechnic University
T Tempered
t8/5 Cooling time from 800 °C to 500 °C
∆t8/5 Cooling time from 800 °C to 500 °C
Ta Tantalum
TEM Transmission Electron Microscopy
Ti Titanium
TiN Titanium Nitride
TiO Titanium Oxide
TM Thermomechanical
TMCP Thermomechanical Controlled Process
Tp Peak Temperature
TRIP Transformation-Induced Plasticity
TTT Time-Temperature-Transformation (diagram)
U.S.Navy United State Navy
V Vanadium
V Voltage
W Watt
W Tungsten
Wf Windmanstatten ferrite
WM Weld Metal
WPS Welding Procedure Specification
xi
wt% Mass fraction
X-ray Röntgen radiation
Zr Zirconium
YAG Yttrium-Aluminium-Garnet–laser
ε-Cu epsilon copper
μm micrometer
α alpha
α ferrite
ɣ gamma
ɣ austenite
π pi
μ mu
δ delta
λ lambda
σ sigma
∞ Infinite
°C degrees Celsius, degrees centigrade
% percent
∆ delta
η eta
η arc heat efficiency
xii
Standards
ASTM E 1290-2 Standard test method for crack-tip opening displacement (CTOD) fracture toughness measurement
ASTM E 112-10 Standard Test Methods for Determining Average Grain Size SEW 088:1993 German standard, weldable fine grained steels; guidelines
for processing, particular for fusion welding SFS-EN 10204 Metallic products. Types of inspection documents SFS-EN 1321 Destructive tests on welds in metallic materials.
Macroscopic and microscopic examination of welds SFS-EN 1435 Non-destructive examination of welds. Radiographic
examination of welded joints
SFS-EN 571-1 Non destructive testing. Penetrant testing. Part 1: General principles
SFS-EN ISO 148-1 Metallic materials. Charpy pendulum impact test. Part 1:
Test method (ISO 148-1:2009)
SFS-EN ISO 15164-1Specification and qualification of welding procedures for metallic materials. Welding procedure test. Part 1: Arc and gas welding of steels and welding of nickel and nickel alloys.
SFS-EN ISO17637 Non-destructive testing of welds. Visual testing of fusion- welded joints (ISO 17637:2003)
SFS-EN ISO 23277 Non-destructive testing of welds. Penetrant testing of welds.
Acceptance levels (ISO 23277:2006)
SFS-EN ISO 5173 Destructive tests on welds in metallic materials. Bend tests (ISO 5173:2009)
SFS-EN ISO 4063 Welding and allied processes. Nomenclature of processes and reference numbers (ISO 4063:2009, Corrected version 2010-03-01)
SFS-EN ISO 4136 Destructive tests on welds in metallic materials. Transverse tensile test (ISO 4136:2001)
xiii
SFS-EN ISO 6057-1Metallic materials. Vickers hardness test. Part 1: Test method (ISO 6507-1:2005)
SFS-EN ISO 6892-1Metallic materials. Tensile testing. Part 1: Method of test at room temperature (ISO 6892-1:2009)
14
1. INTRODUCTION
Welding is the most commonly used method to join different types of structures.
In many respects, joints are the most critical components of a load-bearing steel structure. In order for the final product to be properly developed, a number of factors must be considered when manufacturing individual components, including design, processes, inspection and quality control of structure. At low service temperatures, questions about the ductility of the welded joint can arise, as the welded structure tends to low transition temperatures. This is especially the case if the joint is produced from high strength steels (HSSs).
HSS has been in use in workshops since the 1980’s. At the time, the significance of the term HSS differed from the modern conception because maximum yield strength of HSSs has increased nearly every year. During the 1980’s, the maximum yield strength of weldable HSS was 500 MPa, whereas today it is at least 1000 MPa or more. In the beginning, only a few manufacturers had HSS, which was represented through a limited assortment of products. Today, HSS is constructed worldwide with most of the modern global production consisting of structural steel which is measured in tons with an approximate yield strength 355 MPa.
1.1. Background
The need of utilization of HSS grows continuously. Currently, HSSs are used more frequently and in a diverse number of industries. Primarily, HSS was just used in the car industry, but today the material is used in a more diverse assortment of industries and locations including the arms of cranes and the frames of lumber carriers, although this list is by no means extensive.
15
To date, HSS has not been formally standardized. At the lower end, structure steels have a yield strength in the range of 235-355 MPa. Recent literature has stated that strong steels should have yield strength of at least 460 MPa, while steels with a yield strength of more than 550 MPa should be categorized as ultra HSSs. Today, the yield strength of some steel has increased to 1100 MPa, while in the commercial sector, steel with a rating of up to 1300 MPa (1500 MPa) is sold.
There are three different ways to make HSS. First, the oldest method is the QT method (quenched and tempered method), followed by the TMCP (thermomechanical controlled process) and finally, the last method is direct quenching (DQ). The common goal of all of these above mentioned production methods is to create a steel of high yield strength and good ductility. All the steels that are created using one of these three different methods (QT, TMCP or DQ) have a bainite and/or martensite small microstructure in the main structure. TMCP steel can also have a ferrite-bainite main structure. This small microstructure is created through the alloying of various microelements such as niobium, titanium, vanadium, and boron, which in turn make inclusions like carbides and nitrides. Together with fast cooling and tempering, the resulting microstructure is small and the hardness of structure is high despite the small content of carbon. Some manufacturers have developed DQ steel to replace QT steel using this new method (Porter 2006).
Additionally, chromium, nickel, molybdenum, aluminium, carbon, magnesium, silicon, phosphorus and other alloying elements are added (or are not taken away during the manufacturing process) to iron to make HSS. It is typical of HSSs to have a low carbon content which gives the steel a lower CEV (Carbon Equivalent Value) and good weldability.
Before starting to use HSS in old structures, the entire structure must be redesigned. Simply thinning the structures is not enough as buckling, springing, or bending can easily occur. In their publication from GMA-welded AHSS structure, Kaputska et al. (2008) explained that it is important for designers and
16
manufacturing engineers to understand the factors that may be affected in these performances. As there are a large variety of manufacturers that make HSSs using different methods, it is important to clarify differences between these steels. Sampath (2006) explained that manufacturers must exercise extreme caution when transferring allowable limits of certified secondary construction practices from one type of HSS plate steel to another, even for same plate thickness.
2. STATE OF THE ART
A large number of scientific reports and design guidelines have been published regarding the welding of HSSs (Zeman 2009, Shi & Han 2008, Liu et al. 2007, Pacyna & Dabrowski 2007, Yayla et al. 2006, Juan et al. 2003, Keehan et al.
2003, Miki et al. 2002, Zaczek & Cwiek 1993). Special attention has been devoted to welding HSSs with matching filler material, however, only a limited number of publications consider welding HSSs with undermatching filler material (Rodriques et al 2004a). In the 1980s HSS was pioneered in Japan and organized so that individual manufacturers had their own research projects on specific steels. As a result of this rigorous research, today’s steels are of much better caliber and quality.
There are three different popular and widely available HSSs on the market including those manufactured through the QT, TMCP and DQ processes. QT has been available the longest and DQ HSS has only recently been developed and acquirable on the market. Consequently, most of the research has focused on QT steels, however DQ steel research has emerged in the 2000s and recently, comparing all three HSSs has been an emerging field of investigation.
17 2.1. What is HSS?
The term HSS is variable concept. Today, HSSs are steels with a yield strength greater than 550 MPa. Classifying steels according to their yield strength allows for the correct comparison between different types of steels. Fig. 1 (World Auto Steels 2009) depicts the classifications of different HSS types.
Conventional HSSs (HSS) have a yield strength lower than 550 MPa. Included in this group of steels are IF-HS (High Strength Interstitial Free) steels, BH (Bake Hardenable) steels, IS (Isotropic steels), CM (Carbon Magnanese) steels, and HSLA (High Strength Low Alloy) steels (World Auto Steel 2009).
Advanced HSSs (AHSS) have yield strengths greater than 550 MPa. Some steels that fit into this category are TRIP (Transformation-Induced Plasticity) steels, DP-CP (Dual Phase or Complex Phase) steels, and MS (Martensitic) steels. MS steels are used in many different industries and can be found in cranes, earth-movers, harvesters, and more.
Traditional HSSs, such as high-strength low-alloy (HSLA), have more than three decades of shop experience upon which to build a technology base. In contrast, users of AHSS demanded a fast track accumulation of knowledge and dissemination as they implemented these new steels. A considerable challenge arises along the total elongation and yield strength axes, as the trend shows that higher strengths steels have decreasing total elongation percentages.
Manufacturers are currently looking for ways to maintain the total elongation percentages with steels of increased yield strength.
18
Figure 1. Relationship between yield strength and total elongation for various types of steels (World Auto Steel 2009).
Fig. 2 depicts the developmental history of HSS for commercial use. The first HSS, S355, was developed in the 1940s with a yield strength of 355 MPa. By the 1970s, HSSs with a yield strength of up to 690 MPa had been created. By 1990, the maximum MPa had been increased to 960 MPa, and currently, HSSs with a yield strength of up to 1300 MPa can be found (Kömi 2009).
History of Ultra High Strength steels
Yield Strength, MPa Hardness, HBW
Figure 2. The history of ultra HSS (modified from Jukka Kömi figure 2009, Rautaruukki Ltd).
19
HSSs have been used in the war industry since 1946. The U.S. Navy has used high yield (HY) strength steel, including HY-80, HY-100, and HY-130 steels (Moon et al. 2000 according to Holsberg, P.W. et al. 1989). However, these steels were originally quite expensive to make and additionally, the knowledge of this new generation of steel was kept within the government and therefore the private sector was, for a time, excluded from this new industry. The HY- strength steel corresponds with the ISO system, where the tensile strength of HY-70 (70 ksi) corresponds to 490 MPa, HY-80 (80 ksi) corresponds to 700 MPa, HY-100 (100 ksi) corresponds to 780 MPa, HY-120 (120 ksi) corresponds to 840 MPa and HY-130 (130 ksi) corresponds to 910 MPa.
2.2. Effects of alloying elements in HSS and in its weld
Alloying elements are used in HSSs to reduce the phase microstructure. There are many appropriate alloying elements that can be used when making HSSs, including Cr, W, Mo, V, B, Ti, Nb, Ta, Zr, Ni, Mn and Al. Every alloy or blend of alloys has a different effect on the steel. These elements compose inclusions and precipitations such as nitrides, carbides, carbonitrides and composites in the HSS and inhibit grain growth. In order to create a HSS with a small grain size an alloy or combination of alloys should be used, and additionally planned rolling can contribute to the creation of a steel with the above mentioned desired characteristics.
To prevent the growth of austenite grains, a maximum temperature, which is dependent on the alloying element, where carbides and nitrides will dissolve to austenite, must not be exceeded. Fig. 3 shows how carbide and nitride inclusions quickly dissolve into austenite once these temperatures have been exceeded.
20
Figure 3. The effects of microalloying on Al, Zr and Ti to austenite grain growth starting temperature (modified from Harri Nevalainen figure 1984).
Titanium, niobium, zirconium, and vanadium are also effective grain growth in- hibitors during reheating. However, for steels that are heat treated (QT, TMCP and DQ steels) these four elements may have adverse effects on hardenability because their carbides are quite stable and difficult to dissolve in austenite prior to quenching (Metal Handbook 1990).
In many research projects alloying elements of HSSs and its welds have been under examination. For example, Kou (2003) reported that increasing the alloying content of weld metal increases its hardenability by pushing the nose of continuous cooling curves to longer times. Moon et al. (2000) noticed that the microhardness variations in the weld and HAZ areas can be examined to correspond with the microstructure of the weldment. At the same time they concluded that the HAZ of the base metal was the hardest region in each weldment examined, regardless of filler metal type, base metal, or heat input.
Maximum hardness was reached about midway through the HAZ of each
21
weldment studied. Fig. 4 describes hardness areas with different heat inputs (4.33 kJ/mm, 2.17 kJ/mm and 1.18 kJ/mm) using HSSs HSLA100 and HY80.
Figure 4. Microhardness maps of welds made with three different filler metals and different welding parameters. The corresponding microhardness scale is included at the bottom of this figure (Moon et al. 2000).
Hamada (2003) reported that it is necessary to combine the values of the constituents in the steel material and the welding conditions after taking into account the necessary joint properties. In their research, they used five different HSSs, HT50, HT60, HT80 and two HT100. They concluded that the properties of the weld HAZ, especially those of the coarse grain HAZ and fine grain HAZ
22
heated to more than the AC3 transformation point, are determined by the composition of the steel along welding conditions, as seen in fig 5.
Figure 5. Structural distribution within multi-layer welded joint HAZ (Hamada 2003 according to Shishida et al. 1987).
Toughness deterioration is one of worst things that can happen when welding HSSs. Caballero et al. (2009) investigated HS bainite steel and concluded that a high degree of microstructural banding, as a result of an intense segregation of manganese during dendritic solification, leads to a dramatic deterioration in toughness in these advanced bainitic steels. They concluded that the stress concentration associated with heterogeneous hardness distribution in the microstructure can be considered a possible factor contributing to premature crack nucleation.
2.2.1. Aluminium and Silicon
Aluminium (Al) is widely used as a deoxidizer and it was the first element used to control austenite grain growth during reheating. When Al or silicon (Si) reacts with oxygen, soft oxides are formed. These soft oxides do not create crack initiations of growth similar to what is seen in TiO precipitations (Vähäkainu 2003). However, in HSSs it has been noticed that niobium (Nb) and titanium (Ti) are more effective grain refiners than Al (Metal Handbook 1990). High Al
23
content weakens the toughness of steel, as it promotes the formation of preferred orientation of ferrite and upper bainite. Free Al promotes forming local areas which contain high contents of carbon, which are known as M-A islands.
This mechanism prevents carbon diffusion and the formation of carbides (Matsuda et al. 1995).
With regard to Al, Kaputska et al. (2008) have also observed that while Al has many effects in steel making, the CEV does not consider Al in its calculation.
Si is one of the principal deoxidizers used in steel making. Killed steels may contain moderate amounts of Si, from 0 to a 0.6 % maximum (Metal Handbook 1990). Low-alloy steels are reinforced by Si, but Si does not affect the features of low carbon steels (Harrison & Wall 1996).
2.2.2. Niobium
As an alloying element, Nb has an important role in HSS. The effects of niobium on steel and HAZ are not solely derived from niobium. Niobium affects steel and HAZ when it is combined with other alloying elements, such as Ti and V, and precipitations. In the welded joints of HS steels, the effects of niobium depend upon the heat input. If welding and using a low heat input, this will increase impact toughness, while if a high heat input is used it will decrease the impact toughness in the HAZ. In these HSSs, as carbon content increases, there in an inverse relationship as the impact toughness decreases (Tian 1998; Hatting &
Pienaar 1998).
In certain amounts, Nb (0.02-0.05 wt.%) increases austenite recrystallization temperature, provides strengthening by forming thermally stable, Nb(C,N) and Nb,Ti(C,N) precipitates. During fusion welding, the precipitates limit austenite grain growth in the weld HAZ, and thereby limit hardenability or improve weldability. Excessive amounts of Nb (>0.05 wt.%) can potentially impair HAZ toughness in high heat input weldments (Sampath 2005). Small additions of Nb
24
increase the yield strength of carbon steel. The addition of 0.02 % Nb can increase the yield strength of medium-carbon steel from 490 MPa to 700 MPa.
This increased strength may be accompanied by considerably impaired notch toughness unless special measures are used to refine grain size during rolling.
Grain refinement during rolling involves special thermomechanical processing techniques such as controlled rolling practices, low finishing temperatures for final reduction passes, and accelerated cooling after rolling is completed (Metal Handbook 1990).
In HSLA steel with niobium, granular bainite is dominant within a wider cooling rate range. In addition, martensite is observed at high cooling rates with Nb 0.026 %, but is not produced in the same steel without Nb (Zhang et al. 2009).
Zhang also reports that at lower cooling rates, under 32 °C/s, Nb addition suppresses grain boundary ferrite transformation and promotes the formation of granular bainite. Li et al. (2001) have reported that the addition of 0.031 % Nb to low carbon micro alloyed steel produced the largest size and greatest area of M-A phase.
2.2.3. Vanadium
Vanadium (V) increases the austenite recrystallization temperature in HS steels.
It provides room temperature strengthening by forming VN, V(C,N) and (V,Ti)N precipitates in ferrite (Sampath 2005). V also strengthens HSLA steels in two ways. First, the precipitation hardens the ferrite and secondly, the precipitation refines the ferrite grain size. The precipitation of V carbonitride in ferrite can develop a significant increase in strength that depends not only on the rolling process used, but also on the base composition. Carbon content above 0.13 to 0.15 % and Mn content of 1 % or more enhances the precipitation hardening, particularly when nitrogen content is at least 0.01 %. Grain size refinement depends on thermal processing (hot rolling) variables, as well as V content (Metal Handbook 1990).
25
Chen et al. (2006) have reported that there is a correlation between V content and the size of M-A particles. This is a direct correlation as the size of M-A particles increase with increased V content from 0 % to 0.151 %. When increasing V content, there is a decrease in the impact toughness in HSS. The coarse austenite and ferrite grain and M-A constituent were thought to be the main factors resulting in impact toughness deterioration.
Both Chen et al. (2006) and Zhang et al. (2009) reported after their experiments on that the concentration of V should be limited to a low level, near 0.05 %. If the V content is 0.1 % or more, this results in a greater area fraction of the M-A phase, larger average and maximum sizes of M-A particles, and deterioration in toughness.
2.2.4. Titanium
When considering the welding of steel, Ti is most important micro alloying element. Stable Ti nitrides that form in high temperatures inhibit grain growth in the HAZ. Consequently, because of this grain size CGHAZ cannot grow destructively (Liu & Liao 1998).
Ti is unique among common alloying elements, because it provides both precipitation strengthening and sulfide shape control. Small amounts of Ti (<0.025 %) are also useful in limiting austenite grain growth in HSSs. However, it is only useful in fully killed (aluminium deoxidized) steels because of its strong deoxidizing effect. The versatility of Ti is limited because variations in O, N, and S affect the contribution of Ti as a carbide strengthener (Metal Handbook 1990).
In controlled amounts (0.01-0.02 wt.%) Ti acts as a grain refiner, increases rerystallization temperature, fixes solute nitrogen as TiN, and provides strengthening by forming thermally stable, complex Ti(C,N) precipitates. During fusion welding, TiN precipitates limit austenite grain growth in the weld HAZ, thereby limiting hardenability and improving the HAZ strength and toughness.
26
Precipitation of TiN invariably reduces the HAZ toughness, especially at low temperatures (Sampath 2005).
Ti can react with nitrogen in liquid condition. Large TiN precipitates will grow in steel and their formation is easier when the Ti/N ratio is large. These kinds of precipitates cannot prevent grain growth as the precipitates which form in lower temperature. Precipitates which are big and angular can nucleate cracks and decrease fatigue durability (Lee & Pan 1995). The size of some inclusions are explained in fig. 6.
Figure 6. The nucleation ability of various inclusions (Lee & Pan 1995).
Ti improves HAZ microstructure and toughness of welded structure with three inter-related mechanism. Those mechanism are refining of ferrite grains by the pinning effect of thermally stable Ti-nitride and Ti-oxide particles which are distributed in austenite, by formation of pure Ti-nitride and Ti-oxide particles which disperse in austenite at high temperature and then this particles can be as nucleation sites for acicular ferrite during the ɣ-α transformation. Third mechanism is formation of fine nitrides which decrease the detrimental effect of soluble nitrogen in ferrite (Rak et al. 1997).
27 2.2.5. Zirconium
Zirconium can also be added to killed high-strength low-alloy steels to improve inclusion characteristics. This occurs with sulfide inclusions, where the changes in inclusion shape improve ductility in transverse bending (Metal Handbook 1990).
2.2.6. Boron and Copper
Boron (B) is added to fully killed steel to improve hardenability. The average B content in steels ranges from 0.0005 to 0.003 %. When B is substituted in part for other alloys, it should be done only to alter the hardenability. The lowered alloy content may be harmful for some applications; however B is most effective in lower carbon steels (Metal Handbook 1990).
According to Moon et al. (2008), the addition of B to high strength low alloy plate steel makes a fine martensite microstructure, which increases hardenability by making the prior austenite grain boundary more stable.
Vickers hardness of base steels and CGHAZ increasing Cu and B content, solid-solution hardening as uncovered by Moon et al. (2008) investigation. In the same investigation, it was also noticed that Charpy V-notch toughness showed an opposite tendency. This is mainly due to the formation of the hard phase by increasing hardenability with Cu and B addition and where toughness in the CGHAZ is decreased as compared to base steels.
The results published by Moon et al. (2008) indicate that Cu addition is not useful to improve the toughness of the HAZ in high strength low alloy plate steel. Hwang at al. (1998) studied that the structure of low-carbon (C 0.04 %) copper-bearing (Cu 1.8 %) alloy steel plate manufactured by the DQ&T process has been transformed into a fine structure with high dislocation density. During tempering, fine NbC and ɛ-Cu particles are precipitated in large amounts, which
28
do not get coarsened even when the tempering temperatures rise, resulting in excellent mechanical properties. The results of Hwang et al. (1998) indicate that the addition of alloying elements and the application of the DQ&T process to low-carbon alloy steel plates contribute to the production of plates with excellent strength and toughness.
2.2.7. Manganese and Nickel
Manganese (Mn) improves the strength of steel without decreasing its impact toughness and is commonly used in steel making. Mn reacts with oxygen and sulphur quite easily and makes precipitations and is important because all non hopeless effects are outclosed. The use of Mn needs to limited to under 1.5 % as steel with over 1.5 % Mn content can be brittle (Vähäkainu 2003, Lindroos at el. 1986). Excessive amounts Mn increase hardenability and reduce weldability (Sampath 2005).
In his study, Keehan (2004) investigates the effects of Ni and Mn in weld metal.
TEM investigations in conjunction with APFIM (Atom Probe Field Ion Microscopy) concluded a mixed microstructure of martensite, bainitic and retained austenite at an alloying level where a fully martensite microstructure would normally be expected. For increased levels of Mn, a harder and more brittle mainly martensite microstructure formed. At lower levels of Mn a softer, tougher and more easily tempered microstructure with greater amount of bainite is formed. Ni reduction with Mn levels at 2 wt% lead to an increase in toughness. Hardness results showed that lower Mn and Ni levels lead to a softer weld metal (Keehan 2004).
2.2.8. Rare-earth elements
Rare-earth elements, principally cerium, lanthanum, and praseodymium, can be used to provide shale control of sulphide inclusions. Sulphide inclusions, which are plastic at rolling temperatures and thus elongate and flatten during rolling,
29
adversely affect ductility in the short transverse (through-thickness) direction.
The chief role of rare-earth additives is to produce rare-earth sulfide and oxysulphide inclusions, which have negligible plasticity at even the highest rolling temperatures. Excessive amount of cerium (>0.02 %) and other rare- earth elements lead to oxide of oxysulphide stringers that may affect directionally. Treatment with rare-earth elements is seldom used because they produce relatively dirty steels. Treatment with calcium is preferred, because it helps with sulphide inclusions shape control (Metal Handbook 1990).
2.3. Microstructure of welded HSS structure
The structure of the base metal of HSS is homogenous and the grain size is small and regular, fig. 7. When the steel is heated during welding, the homogenous microstructure changes immediately. The heat input in the HAZ is different depending on how far the area is from the fusion zone. Many features, such as hardness, ductility and impact toughness change radically, and in many cases, to defective direction. The main structure in the base of HSS is tempered martensite and/or bainite. In addition, there are other phases such as ferrite and M-A constituent. Other important parts of structure are segregations of inclusions and precipitations such as nitrides, carbides, carbonitrides and composites.
Figure 7. Microstructure of TMCP HSS (own image 2010). Aspect ratio is 1:500.
30
Fig. 8 shows the schematic description from the HAZ area temperature during steel welding. The width of the HAZ depends on heat input and cooling time. A large proportion of inclusions and precipitations dissolve when the temperature is high. When this happens, there are no nether inclusions or precipitations in or near the fusion line.
HAZ area
Temperature
Weld
HAZ area
Liquid Liquid + γ Austenite
Bainite
A1-boundary 723 °C Martensite
20 1. Weld metal, 2. Fusion
line, 3. CGHAZ, 4. FGHAZ, 5. Partly austenite zone 6. ICHAZ. T curve describe maximum temperature of base material during welding.
Figure 8. Maximum temperature of base material during welding and HAZ microstructure after welding in steel (modified from Hitsaajan opas 2003).
Inclusions and precipitations are important in HSS making, as they are processes which constrain the grain growth. The same texture, inclusions and precipitations, occur in HSS weld metal. Inclusions of different shapes and textures, including spherical and faceted, and agglomerations of particles were observed in the weld metals when welding HSSs with matching filler metal. The inclusions core mainly consists of a mixture of oxides of Ti, Mn, Si, and Al in different proportions, reflecting a very complex deoxidation product.
Additionally, phases rich in either Mn and S, Si or Zr, C, and N, which indicates the presence of Mn sulphides, Si, or Zr carbonitrides, were also observed (Ramirez 2008).
31
2.3.1. Microstructure and physical features of the HAZ
Near the fusion zone, the phase structure of base metal is coarse as a result of the high temperature of the base metal during welding. In multi-run welding, ICCGHAZ (intercritically reheated coarse grained heat affected zone) is the worst area in the base metal (Li et al. 2001; Kim et al. 1991; Davis & King 1993).
Both heat input and t8/5 (cooling time from 800 °C to 500 °C) time change the microstructure of the welded base metal and these two factors must be under control while welding. There are numerous recommendations from manufacturers regarding heat input and t8/5 time. The main differences between recommendations relate to preheating and post-heating. In specifications, however, there are also differences in spotheating temperature. Using recommended values, it is possible to successfully weld HSS.
In the study done by Kaputska et al. (2008), it was concluded that the fusion zone microstructure and hardness were found to be affected by the base metal chemistry, the cooling rate conditions, and the filler metal composition.
The elongation of the welded structure decreases as the yield strength of HSS grows. Yasuyama et al (2007) compares steels with yield strengths ranging from 270 MPa to 980 MPa. In the study, steels were welded by the YAG laser, mash seam, and plasma arc methods. It was confirmed that the elongation of the weldment declined compared to that of the base metal, regardless of the base metal strength. This was determined by conducting a tensile test both parallel and perpendicular to the weld line. It was therefore concluded, that the elongation is very low in high strength welded structures (Yasuyama et al.
2007).
32
Lambert et al. (2000) studied the microstructure of the martensite-austenite constituent in HAZ of HSLA steel welds in relation to toughness properties. The material used in the research was HSLA steel, with a yield strength of 433 MPa.
Charpy impact test results indicated that the correlation between the toughness and microstructure of low carbon steel simulated HAZs is rather complex. The amout of M-A constituents and coarseness of bainite are major metallurgical factors affecting the impact properties (Lampert et al. 2000). In the same study, Lampert et al. (2000) also noticed that retained austenite and low carbon transformed martensite have significantly different influences on cleavage fracture and impact properties of simulated HAZ microstructure, where freshly transformed high carbon martensite is much more deleterious than retained austenite.
Metallographic investigations demonstrated the existence of different M-A constituents. In the most brittle zones (the ICCGHAZ), retained austenite was mostly located between bainitic packets, whereas blocky martensite and mixed M-A constituents were located at prior austenite grain boundaries. In mixed M-A constituents, austenite was distributed at the periphery, while martensite was located at the centre. This distribution of retained austenite could be a result of chemicals and/or the mechanical stabilization mechanism (Lambert et al. 2000).
Furthermore, through TEM, Lambert et al. (2000) found a constituent retained austenite at room temperature. The presence of constituent may influence the thermal stability of retained austenite, as they propagate before transformation.
These observations constitute preliminary investigations of the transformation mechanism of retained austenite islands.
Moon et al. (2000) compared two new ultra-low-carbon matching filler metals, with HY steel (High yield, quenched and tempered, steel) of HSLA steel.
Despite the low heat input, 1.2 kJ/mm, the fusion zone hardness of two of the new ultra-low-carbon matching filler metals are comparable to the base metal hardness. The results were achieved through researching the microhardness variations in the weld and HAZ areas and corresponding this with the
33
microstructure of the weldment. In addition, the heat affected zone of the base metal was the hardest region in each weldment examined, regardless of filler metal type, base metal, or heat input. The maximum hardness occurs about midway through the HAZ of each weldment studied, rather than adjacent to the fusion boundary (Moon et al. 2000).
Additionally, Moon et al. (2000) studied that the fusion zone consists predominantly of lath ferrite with varying amounts (depending on location) of untempered fine lath martensite, as well as small amounts of interlath retained austenite and oxide inclusions. No polygonal ferrite or solid-state precipitates such as carbides or carbonitrides were observed in the fusion zone. The local variations in microhardness correlate well with the local variations in the microstructure.
Research carried out to study the research done by Mohandas et al. (1999) has displayed that the high Ms and Bs temperatures of steel are also responsible for low softening tendency. Steel, which has longer critical cooling time for full martensite transformation, exhibited greater resistance for softening with high heat inputs.
In the investigation of heat input it was realized that the number and morphology of the ML (lath martensite) in the HAZ had some variations under different weld heat inputs (E= 0.92 ~ 1.86 kJ/mm). The carbon gathers near the grain boundary and then becomes a carbide with Fe, Mn, Mo etc. so that the impact toughness decreases. The carbide has strong direction bonds with the lath microstructure which provides the low energy passage for the impact fracture and increases brittle crack sensitivity. The fine precipitate distributed inside the grain or at the boundary is favorable to improve toughness. By controlling weld heat input (E ≤ 2.0 kJ/mm), the presence of carbides in the HAZ can be removed, and therefore the impact toughness in this zone can be assured. It was also indicated from the test results of Juan at al. (2003) that the cooling time (t8/5) should be controlled (t8/5 10-20 s) to improve toughness in the HAZ. This is so, because the cooling time increases with larger weld heat
34
inputs, which increases the potential for the deterioration of impact toughness in the HAZ (Juan et al. 2003).
When welding ultra HSS, with a yield strength of more than 900 MPa, with MAG welding, it is important to precisely and accurately control heat input to the lowest possible temperatures. Zeman (2009b) examined ultra HSS, with a yield strength of 1100 MPa. In the case of the joint made by the MAG method, the weld is characterized by its bainitic structure. In the HIZ (Heat Impact Zone), Zeman observed a purely martensite structure or mixture of bainite and martensite structures (Zeman 2009b). In the same study, Zeman (2009b) noticed that ultra HSS requires the linear energy of welding to be precisely adjusted. If the linear energy of welding is too low, there could be excessive hardening of the HIZ, which increases the risk of cold cracking, whereas if the linear energy of welding is too high, the strength properties can decrease.
2.3.2. Microstructure of weld
The microstructure of the weld in welded HSSs should be small and homogeneous. Alloying elements are used to make inclusions in the weld and these inclusions are the beginnings of solidifications. The inclusion density tends to be quite high but the volume fraction is comparatively small. Ramirez (2008) found in his research that in the HSS filler metal the volume fraction of nonmetallic inclusions in most deposit HSS weld metals ranged from 0.2 to 0.6
%. In a few welds, the volume fraction was from 0.8 to 1.1 %. The inclusion density observed in the welds ranged from 1.2 x 108 to 5.4 x 108 particles per mm3, while the average inclusion diameter ranged from 0.3 to 0.6 μm and the maximum inclusion diameter from 0.9 to 1.7 μm.
O and S levels correlate with the inclusion size and higher levels of O and S increase the inclusion size. The average inclusion size does not drastically change with combined O and S levels up to about 400 ppm. However, above
35
400 ppm, the average inclusion size increases with an increase of both O and S levels in the weld metal (Ramirez 2008).
Ramirez (2008) has stated that there are dozens of different inclusions in HSS filler metal. Table 1 describes these inclusions, while fig. 9 a and b show the acceptable shape of spherical and angular inclusions, respectively. Finally, Fig.
10 shows the phase structure of one inclusion. The chemical composition of the inclusion in region a (fig.10) is 32.2O-0.5Al-1.3Si-0.9S-51.4Ti-13.7Mn (TiO2), in region b (fig.10) MnS, and in region c (fig.10) Ti-oxide.
a) b)
Figure 9. Inclusion of the weld in HSS (a) Spherical, (b) Angular (Ramirez 2008).
Figure 10. Composites of inclusion (Ramirez 2008).
The weld microstructure can be formed from many starting values. Fig. 11 illustrates those elements which must be taken into consideration when estimating the microstructure of a weld. Additionally, Mistra et al. (2005) researched different types of inclusions as seen in Table 1.
36
Figure 11. The various factors that play a role in deciding weld microstructure (Modified from Basu & Raman 2002).
CHEMISTRY
WELD METAL MICROSTRUCTURE HARDENABILITY
ELEMENTS
INCLUSION CONTENT
JOINT DESIGN PARAMETERS
HEAT INPUT
• Plate thickness
• Heat input
• Thermal diffusivi- ty
• Joint geometry
• Current
• Speed
• Voltage
• Grain Boundary Ferrite
• Pearlite Ferrite
• Ferrite Site Plate
• Acicular Ferrite
37
Table 1. Characteristic of nonmetallic inclusions (modified from Ramirez 2008).
INCLUSION CHARACTERISTIC
INCLUSION CHEMICAL COMPOSITION DESCRIPTION
1
Region a — 50.1O-0.7Mg-1.6Al-3.9Si-2.8S-
19.6Ti-21.4Mn O, Al, Si, S, Ti, Mn rich Region b — 48.2O-0.9Mg-1.6Al-3.4Si-2.3S-
22.2Ti-21.4Mn
2 51.4O-1.4Al-4.5Si-1.7S-18.1Ti-22.8Mn
O, Al, Si, S, Ti, Mn rich 3
Region a — 32.2O-0.5Al-1.3Si-0.9S-51.4Ti-
13.7Mn (Ti-O2) Composite inclu- sion Region b MnS, Region c Ti-Oxide
4
Region a — 32.3O-1.5Al-0.7Si-50.4Ti-15.1Mn
Ti-Mn oxide Region b — 35.4O-3.2Al-6.1Si-0.8S-26.5Ti-
28.0Mn
Region c — 35.3O-4.4Al-9.6Si-1.4S-3.6Ti- 45.8Mn
Table 2. Classification of precipitates of HSS with a yield strength 770 MPa into type I – IV based on size and morphology (Misra et al. 2005).
2.4. Undermatched, matched and overmatched filler metal
Filler metal also has quite a considerable effect on the welded structure of HSS depending of the yield strength of filler metal corresponding with the yield strength of base metal on the filler wire used. The filler metal can be classified as either undermatched, matched or overmatched. The filler metal is undermatched when the yield strength of the filler metal is below the yield strength of the base metal. Matched filler metals have the same yield strength
38
as base metals, and overmatched filler metals have yield strength greater than the base metals. Generally, HSSs are welded by undermatched or matched filler metal, and overmatched filler metal is infrequently used as confirmed by Porter (2006). Welding HSS requires a high quality welding process, however, it is not economical to use overmatched filler metal for HSS as it does not garnish any additional benefits.
Structural steels, whose yield strength is between 235 MPa and 460 MPa, are usually welded with overmatched or matched filler material. The yield strength of structural steels is lower compared to HSSs, and there are more possibilities when welding these steels. The flexibility has allowed for a greater variety of filler material research to be carried out with regards to structural steels.
Only a few research projects have used undermatching filler metal when welding HSSs. A maximum undermatching valve of 10 % can be accepted for class 690 MPa yield strength HSS (Toyota 1986, Satoh & al. 1975). Pisarski and Dolby (2003) found out that in assessing the toughness of softened HAZs, the test specimen must match the practical situation in terms of yield strength, mismatch between weld deposit, and parent metal. They explained that the fracture toughness of softened HAZ regions depended on the mismatch in strength between the weld deposit and parent plate. Their research confirmed that the worst case fracture toughness of softened HAZs occurred when the HAZ undermatched in strength both the weld deposit and parent metal. Higher toughnesses were measured when either the weld metal or parent steel undermatched the HAZ in strength.
Their conclusions also elaborated that the tolerance to flaws in softened HAZs critically depends on the fracture toughness of the HAZ region where tolerance reduces rapidly in a situation where the cleavage is the dominant failure mechanism (Pisarski and Dolby 2003).
In a study carried out by Umekuni and Masubuchi (1997), the tensile strength test showed that the tensile strength of the undermatched weld increases due to
39
restraint by surrounding matched welds and the base metal. Results of fatigue testing showed that both undermatched and matched welds exhibited a similar relationship between crack growth rates and the stress intensity factor.
Undermatched welds have proven to be effective with HSSs, reducing the need for preheating. Undermatched welds lead to lower residual stresses than matched welds, which has the potential to reduce crack initiation. The properties of the weld metal are also a factor in the effectiveness of undermatched welds on HSSs (Umekuni & Masubuchi 1997).
The results of restraint cracking tests indicated that the application of undermatched welds to HSSs leads to the reduction of minimum preheating temperatures and thus preventing cold cracking on the weld metal. It is necessary to consider not only the strength of weld metal, but also its ductility, fracture toughness, and hydrogen content when selecting weld metals for undermatching (Umekuni & Masubuchi 1997).
Undermatched welds have similar fatigue characteristics to matched welds, where both undermatched and matched welds have similar crack propagation rates (Umekuni & Masubuchi 1997).
Additionally, with a WM undermatched yield strength level 12 %, the concentration of plastic flow in the weakest zone increased, while the strength and ductility of the weld loaded in tension decrease. This experiment was conducted with two different heat inputs (2.0 kJ/mm and 5.0 kJ/mm) on a 25 mm thick piece of 700 MPa HSS, yield strength 700 MPa. Mismatching yield strength grade between WM / BM was 0.815, when heat input was 2.0 kJ/mm and 0.765 when heat input was 5.0 kJ/mm (Loureiro 2002).
Welding high strength and high hardness QT steel involves HAZ softening and is a characteristic feature of fusion welding processes and consumables used (Rodrigues et al. 2004b).
40
Initiating a simulation is one possible way to evaluate the features of a welded structure in HSS. Rodrigues et al. (2004b) used this method and concluded that the tensile strength of the soft zone determines the overall strength of the joint.
In fact, independent to the level of the undermatched yield stress, the joints achieved the base plate strength in all overmatched tensile strength situations.
For matched and undermatched cases, the strength of joint was strongly dependent on the HAZ dimensions. For the cases in which the ratio width of the HAZ to sample thickness was less than 1/3, the loss of strength never exceeded 10 %, even in cases of extreme strength undermatch. However, the joint strength decreased linearly with increased HAZ widths. In almost all the cases, mismatch lead to a decrease in joint ductility, which varied depending on HAZ dimensions and hardening values (Rodrigues et al. 2004b).
Rodrigues et al (2004b) also wrote that the mechanical behaviour of the overall joint depends on the plastic distribution inside the HAZ. They noticed that the large undermatched tensile strength promotes strain localization in the HAZ from the start of deformation. When the HAZ dimension is very small (width <
1/3 of the thickness), it was found that the soft material can achieve the base plate strength. They further stated that if the undermatched level of yield stress is large and the HAZ width is equal to the sample thickness, then the constraint promotes premature failure in the soft zone and the overall strength of the joint decrease even more. In the matched situation of tensile strength, the HAZ constraint induces deformation to spread to the adjacent material, whereas the soft HAZ material avoids deformation. There is an apparent increase in the material strength in almost all the undermatched cases and for lHAZ/e (HAZ width to sample thickness) ratios lower than unity, which is due to constraint (Rodrigues et al. 2004b).
In Complete Joint Penetration (CJP), where matching filler metal is required, one recommendation stipulates that there should be groove welds in the tension application. Duane (1997) wrote that using undermatched filler metal is useful in welds such as Partial Joint Penetration (PJP) groove welds and filler welds. In these situations, using undermatched filler material is a cost-effective and
41
desirable alternative to matched welds. Duane (1997) also explained that when welding higher strength steels with undermatched weld metal, it is important that the level of diffusible hydrogen in the deposit weld metal is appropriate for the higher strength steel that is being welded.
An analysis of the microstructure and the resulting fusion zone hardness indicated that dilution of the filler metal by the base metal does play a role in weld metal microstructure evolution. Hardness traverses indicated that the weld has regions of significant hardening and softening depending on the base metal grade, filler metal type, and cooling rate conditions. The location of greatest hardening in the near HAZ (adjacent to the fusion boundary), is where the far HAZ experienced softening. The potential implications of the hardness increased in the near HAZ region are not well understood (Kapustka et al.
2008).
In dynamic tensile test results of the 780 MPa butt joint and of the DP780 steels, all of these specimens failed in the softened region of the HAZ (Kapustka et al. 2008).
It is clear from a large amount of research that the lower the weld strength mismatching, the higher the fracture toughness of the HAZ (Shi et al. 1998).
When undermatched filler metal is used in welding HSS, a number of items must be taken into consideration. First of all, heat input and t8/5 time are two of the most important aspects to consider. These two elements depend on a number of factors, including thickness of steel, preheating, current, voltage, and the speed of welding. Some of these factors can be altered while others cannot.
For example, metallurgic and chemical effects depend on base and filler material and predescrible the effects in the weld.
42 2.5. Heat input and cooling time
Welding HSS is considerably more complex than welding lower yield strength structural steels. When welding HSSs, a number of quantity modifications are made during the heating process. The HAZ area has many different phase zones, and the CGHAZ is quite often the worst zone in HSS after welding. The phase structure depends on the thermal cycle, which in turn depends on heat input, work piece geometry, material properties, etc.
In earlier research (Vilpas et al. 1985) low heat input was under 2.0 kJ/mm, but today low heat input correspond to values 0.5 kJ/mm or lower. When welding ultra HSSs heat input must be very low according to the recommendations of manufacturers.
HSS has been studied in a number of research using different consumables and welding processes. Nevasmaa et al. (1992b) researched Accelerated- Cooled (AcC) high strength TMCP steel X80 and noticed that those steel do not need to be preheated in the arc energy range from 1.5 to 5.0 kJ/mm. They also concluded that in SA-weld metals, the toughness requirement of 40 J at -40 °C was exceed throughout the arc energy range from 2.0 to 5.0 kJ/mm.
Magudeeswaran et al. (2008) researched QT steel of two different types; (1) consumable made from austenitic stainless steel, and (2) low hydrogen ferritic steel. Welding with different heat inputs and two different methods (GMAW and FCAW), they concluded that the alloying content of manganese and nickel are important in the solidification process of HSS weld metals. They also noticed that the SMAW process is more useful for welding HSSs than the FCAW process. The joints produced by using the SMAW process exhibited superior tensile and impact properties and lesser degree of CGHAZ softening compared to their FCAW counterparts.
43
Wang et al. (2003) and Juan et al. (2003) researched heat input of HSS and the test results indicated that implementing a cooling time (t8/5 =10 - 20 s) improves toughness in the HAZ (when corresponding weld heat input is 1.31 - 1.86 kJ/mm). This is true, because the larger the weld heat input, the longer the cooling time and the easier it is for the deterioration of impact toughness in the HAZ.
In another study carried out of Shi and Han (2008) on 800 MPa yield strength HSLA steel it was reported that the presence of allotriomorphic ferrite, bainitic ferrite and martensite exists for simulated HAZ of the test steel. This happens, because at a temperature range of 800-1300 °C, the austenite decomposes to various ferrite morphologies. In the subsequent cooling process from 800 °C to 300°C, the austenite decomposes to various ferrite morphologies. The austenite to ferrite decomposition starts with the formation of allotriomorphic ferrite at prior austenite boundaries and eventual coverage of these boundaries. With the continued cooling, the side plate ferrite may nucleate at the ferrite/ austenite boundaries and extend into the untransformed austenite grain interiors. Further cooling to even lower temperatures increases the possibility of bainitic ferrite or acicular ferrite formation. When carbide-free bainitic ferrite is formed, the remaining austenite is enriched into carbon and becomes stable. The carbon content of remaining austenite may reach 0.5 – 0.8 wt%. With further cooling as the temperature settles to room temperature, the remaining austenite may completely or partially transform to martensite (Shi & Han 2008).
As the M/A constituent forms in the HAZ during bainite transformation, the carbon-enriched, untransformed regions will partially transform into martensite at low temperatures. The carbon-enriched austenite regions are formed by the rejection of carbon from ferrite to austenite following the transformation of bainite ferrite. The transformation of M/A constituent leads to the deterioration of toughness in the HAZ (Shi & Han 2008).
Shi & Han (2008) also noticed that when the cooling time in simulated 800 MPa yield strength HSS is 18 s, the fracture toughness in the simulated HAZ is
44
highest. Additionally, when the value of t8/5 is 45 s or longer, the toughness of the weld deteriorates. A remarkable decrease in toughness is observed with the increased size of austenite grain and the volume fraction of the M/A constituent.
The fact that the fracture toughness deteriorated drastically for the partially phase transformed HAZ may be related to the formation of a mixed microstructure, in which the M/A constituent is a distributed shape of networks (Shi & Han 2008).
Liu et al. (2007) noticed in double thermal experiments that the impact toughness decreases dramatically and obvious brittlement happens in the intercritical region of CGHAZ. They investigated copper-bearing steel with a tensile strength of no less than 685 MPa. The decreased toughness and brittlement occurred, because pearlite is formed on the interface of original austenite and coarse granular bainite, which can reduce the impact toughness.
The higher heat input, the more serious brittlement becomes. Thus, during multilayer welding, it is proposed to strictly control heat input. Single thermal cycle experiments show that the copper-bearing steel has a narrow range of heat-input and brittlement can easily occur in the region of CGHAZ with higher heat-input. Granular bainite transformed from austenite leads to brittlement, and the softening starts when t8/5 time is more than 7 s. The dissolution of ε-Cu and coarse lath bainite and more ferrite can cause softening of the CGHAZ.
Many HSSs, particularly copper-bearing steels, have a narrow range of heat input when welding. The effective measure to avoid or reduce the softening phenomenon of CGHAZ is to limit or control the heat input during welding.
During the welding thermal cycle, with increasing heat input, lath bainite becomes coarser and the amount of ferrite increases. Coarse lath bainite decreases dislocation density and ferrite is in a soft phase. Therefore, coarse lath bainite and more ferrite can cause the softening of CGHAZ (Liu, W-Y.
2007).
The features of steel can vary with the cooling rate. Pacyna and Dabrovski (2007) investigated CEV 0.39 low-C, Mn-Mo, Al killed steel using different
45
cooling time in the manufacturing process. They noticed that depending on the rate of cooling, and within the air to water cooling temperature range, the new steel can attain a tensile strength between 504 MPa and 1122 MPa. The corresponding proof stress range is from 286 MPa to 478 MPa and the structure of the air cooled steel consists of ferrite, pearlite, and bainite. This research concluded that a low carbon equivalent allows for good weldability under any conditions.
Depending on the welding current and travel speed combination used, significantly different dependencies on all the influencing parameters were observed even though the heat input was same. This can be attributed to differences in the weld bead morphologies. Different weld bead morphologies are likely to lead to different weld cooling rates that will affect the microstructure by itself and also different microstructural features, such as austenite grain size, inclusion parameters, which in turn, will further contribute to the final AF content (Basu & Roman 2002).
The increase to the heat input increases the yield and undermatched tensile strength of the WM, and also produces an undermatched HAZ (Loureiro 2002).
When the heat input is greater (4.5 kJ/mm), the weld metal can undermatch, despite the use of matching filler material (Nevasmaa & al. 1992a). If undermatching is 10 % or less, then a maximum heat input 2.0 kJ/mm can be accepted according to Nevasmaa et al. (1992a).
An example of the microstructure of HSS is in fig. 12, which illustrates QT steel with a yield strength of 690 MPa or more and the CCT-diagram shows cooling curves from 1000 °C to room temperature, and together with table 3, it shows the main microstructure and hardness for this steel after different cooling times.
This type of CCT-diagram can be used to describe the microstructure of high strength QT steels with a standard yield strength 690 MPa. That microstructure will form in different zones of QT steels HAZ (yield strength 690 MPa) after cooling.
46 CCT diagram QT steel 690 MPa
Figure 12. CCT-diagram of QT steel which yield strength is 690 MPa or more (Modified from Dillinger Hüttenwerke AG 2008).
Table 3. Example of microstructure, austenite grain size and hardness for QT steel (yield strength 690 MPa) after different maximum heating temperatures when t8/5 is 20 s (Modified from Dillinger Hüttenwerke AG 2008).
PHASE STRUC- TURE, HARD-
NESS
800
°C
900
°C
1000
°C
1100
°C
1200
°C
1350
°C
Martensite % 5 10 35 50 60 70
Bainite % 55 80 60 50 40 30
Ferrite % 40 10 5 - - -
HV10 227 223 275 313 328 319
Austenite grain
size (ASTM) 11 12 10-11 6 6 2-3
