Moham Supp Dec 2018 WJ 201815layout.

qxp_Layout 1 11/8/18 2:49 PM Page 338

WELDING RESEARCH

Influence of Cold-Wire Submerged Arc Welding

on the Toughness of Microalloyed Steel
A detailed analysis of fracture toughness indicated that
the addition of an electrically cold wire to the TSAW process
improved the toughness in the HAZ of welded X70 microalloyed steel

BY M. MOHAMMADIJOO, L. COLLINS, R. LAZOR, H. HENEIN, AND D. G. IVEY

is generally believed to exhibit the poorest fracture toughness

ABSTRACT relative to the rest of the steel structure (Ref. 1). Several
methods have been commonly used to assess the fracture
The structural performance of microalloyed steel weld
toughness in the CGHAZ (Refs. 2–5); however, during the
joints is essentially dependent on the microstructural
characteristics of the heat-affected zone (HAZ). However, past decade, several endeavors have been devoted to develop-
the microstructure is adversely impacted by the heat ing a fracture toughness test method, which is less conserva-
input of the welding process. As such, an X70 microalloyed tive relative to the conventional methods according to engi-
steel was welded via a recently developed welding neering critical assessment (ECA) (Refs. 4, 5–8).
process, i.e., tandem submerged arc welding (TSAW) with An ECA, also known as a fracture toughness analysis, is a
an additional cold wire (cold-wire TSAW), and compared fitness-for-service procedure that uses fracture mechanics
with the conventional TSAW process in terms of the HAZ principles to determine the defect tolerance of critical safety
microstructure and toughness. The heat introduced to the items, such as pipelines and pressure vessels. An ECA is used
weldment was essentially reduced by the addition of a to decide whether a given flaw is safe from brittle fracture,
cold wire in the welding process, which changed the prop-
plastic collapse, creep, and fatigue crack growth under cer-
erties in the HAZ. The results of fracture toughness testing
using single-edge notched tension (SENT) testing, a tain loading conditions. According to the loading conditions
recently developed test method, and Charpy V-notch that steel pipelines are exposed to, the single-edge notched
(CVN) impact testing showed an improvement in the tension (SENT) test method has been developed to investi-
toughness of the HAZ of the cold-wire TSAW weld. This gate the toughness of the steel pipelines (Refs. 6–10). The
improvement was attributed to the formation of a lower fracture toughness in the heat-affected zone (HAZ) is con-
fraction of M-A constituents and smaller ferrite/bainite trolled by several microstructural factors, such as grain size,
grains with higher boundary misorientation angle, due to types of phases, and morphology of secondary phases,
the reduced heat input by cold wire addition. Furthermore, which are directly influenced by the heat input of the weld-
the microstructural analysis indicated the morphology and ing process (Refs. 11–13).
interspacing of the M-A constituents within the coarse-
During the past several decades, the relationship between
grained heat-affected zone (CGHAZ) played a significant
role on the toughness of the HAZ. microstructure and mechanical properties of high-strength
microalloyed steels has been extensively studied in an effort
to improve the fracture toughness and strength of this type
KEYWORDS of steel without substantially increasing the costs of steel
production and structure manufacturing, i.e., welding/join-
• Microalloyed Steel • Cold-Wire TSAW ing. However, although an excellent combination of
• Fracture Toughness • M-A Constituents • CGHAZ strength and toughness can be achieved by thermomechani-
cal-controlled processing (TMCP) and alloying design, the
fracture toughness deteriorates in the HAZ, particularly the
Introduction CGHAZ, due to the high welding heat input. This properties
deterioration is caused by the microstructural changes, i.e.,
In developing microalloyed steel structures, such as the formation of large prior austenite grains (PAGs) and the
pipelines, pressure vessels, and automotives, the steel is sub- martensite-austenite (M-A) constituents in the CGHAZ
jected to welding/joining, which results in the formation of a (Refs. 14–16), as a result of the thermal cycle and peak tem-
relatively narrow region adjacent to the fusion zone known as peratures (above ~ 1300˚C) that the steel experiences during
the coarse-grained heat-affected zone (CGHAZ). The CGHAZ welding.

https://doi.org/10.29391/2018.97.029

338-s WELDING JOURNAL / DECEMBER 2018, VOL. 97

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 3:12 PM Page 339

WELDING RESEARCH

A B
B A

Fig. 2 — A — Schematic of a circumferential flaw formed

through the wall thickness of a typical pipe; B — SENT speci-
Fig. 1 — A — X70 steel joint geometry indicating both ID and men geometry with a notch placed in the HAZ for 13.4-mm-
OD bevels; B — schematic view of cold-wire TSAW process thick pipe; and C — extraction of subsize CVN specimen from
with positioning of the electrodes and cold wire. the weld sample.

The studies done by Moeinifar et al. (Refs. 17, 18) showed on the penetration depth and macrostructure of the weld and,
that fraction and morphology of M-A particles, as well as the consequently, the properties of the weld. The penetration
PAG size, influence the Charpy impact toughness in the CG- depth and width of the weld decrease as the welding current
HAZ. Gharibshahiyan et al. (Ref. 19) have reported that the and voltage are reduced, which result in the formation of de-
coarser PAGs formed in the CGHAZ adversely affect the fects (due to the incomplete fusion) through the weld and the
toughness of the HAZ. Davis et al. (Refs. 20, 21) and Reichert formation of crack initiation sites. Accordingly, the weld speci-
et al. (Ref. 22) found that cleavage crack initiation and trans- mens were fabricated in a two-pass welding condition (i.e., one
granular fracture take place in the HAZ due to the formation inner-diameter [ID] and one outer-diameter [OD] welding
of a network of enlarged M-A constituents along and within run) using the recently developed cold-wire TSAW process
the PAGs in the CGHAZ. Li et al. (Ref. 23) showed that the for- (Ref. 35) and compared with the conventional TSAW process.
mation of a lower fraction of small M-A islands did not cause a The developed process improves the throughput of the weld-
detrimental effect on the HAZ toughness. However, the frac- ing process by addition of an electrode, which is electrically
tion and size of M-A constituents were essentially dependent cold, i.e., no current and voltage, while retaining appropriate
on the PAG size. The work done by Yu et al. (Ref. 24) conclud- weld appearance characteristics.
ed that more M-A constituents with a larger size were formed In the TSAW process, the welding deposition rate is in-
by coarsening the PAG in the CGHAZ. creased by increasing the current and voltage of the electrodes
It was also found that a microstructure comprised of a (i.e., heat input), which affects the properties in the HAZ.
mixture of the coarse PAGs and coarse M-A constituents re- However, in cold-wire TSAW, the additional cold wire fed into
sulted in the promotion of brittle fracture in the CGHAZ. the weld pool increases the deposition rate with no increase in
The temperature and kinetics of phase transformations tak- the welding heat input (Refs. 35–37). In addition, the heat in-
ing place during cooling are influenced by the PAG size put is moderated by the cold wire due to consumption of
(Refs. 25, 26). The martensite start temperature (Ms) in- some of the excess energy of the trail electrode, which low-
creases with an increase in the PAG size, which results in a ers the amount of the heat introduced to the weldment. As
higher volume fraction of martensite (Refs. 27, 28). Howev- such, a weld with smaller dilution and a narrower HAZ is
er, the PAG size and the fraction of the M-A constituents are achieved due to the lower welding heat input (Refs. 35, 38).
greatly influenced by the welding heat input. Shome (Ref. Moreover, since heat input directly influences the HAZ mi-
29) and Zhang et al. (Ref. 30) studied the dependency of the crostructure and properties, lowering the welding heat input
PAG size on the local thermal cycle in the HAZ. They report- results in a refinement of the microstructure with a corre-
ed the PAG size increased by increasing the welding heat in- sponding effect on the properties, which are discussed in de-
put. Refining the PAG size in the HAZ influenced the trans- tail in this paper. Microstructural alterations in the CGHAZ
formation products, particularly the M-A constituents, and are characterized using tint-etching optical microscopy
consequently the toughness (Refs. 23, 31, 32). However, (TEOM), scanning electron microscopy (SEM), and electron
there has only been limited work devoted to the correlation backscattered diffraction (EBSD). The Charpy V-notch
of the PAG size and M-A constituents in the HAZ with the (CVN) testing and SENT testing were conducted to evaluate
welding heat input (Refs. 31–34). the toughness changes in the HAZ by cold wire addition.
With the above in mind, due to the adverse effect of high
welding heat input on the properties of the HAZ, this study
evaluates the microstructure and the fracture toughness of the Experimental Procedure
HAZ when the welding heat input is reduced through the addi-
tion of a cold wire in the tandem submerged arc welding (cold- Materials and Welding Procedure
wire TSAW) process. It is possible to reduce the heat input by
reducing the current and voltage of the electrodes. However, Cold-wire TSAW and the conventional TSAW processes
this type of reduction can result in substantial adverse effects were conducted on beveled X70 microalloyed steel plates

DECEMBER 2018 / WELDING JOURNAL 339-s

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 340

WELDING RESEARCH

A B

C D

Fig. 3 — Macrographs of the following: A — Cold-wire TSAW and B — TSAW samples indicating the position of the initial crack for
SENT testing with the initial crack tip positioned in the CGHAZ; C and D — side and top views, respectively, of a SENT specimen for
the cold-wire TSAW sample.

with a thickness of 13.4 ± 0.3 mm, manufactured through ing crack susceptibility index for low-carbon microalloyed
thermomechanical-controlled processing (TMCP) (Ref. 39). steels given in Table 1, was calculated according to the Ito-
A typical X70 microalloyed steel has a yield strength (ys) Bessyo equation (Ref. 40).
and ultimate tensile strength (uts) of 480 and 610 MPa, re- Two 85 ± 5 deg V-shaped bevels, 4.0 ± 0.2 mm and 2.5 ±
spectively, with a CVN impact energy as high as 150 J. Two 0.2 mm in depth, were machined in the steel plates prior to
4-mm-diameter BA-S2Mo electrodes, and one cold wire with welding — Fig. 1A. One ID and one OD weld run were con-
the same diameter and composition as the electrodes, were ducted for each welding process to ensure complete joint
chosen according to AWS A5.23 for arc generation and fill- penetration and complete filling of the root opening be-
ing of the bevel. The BA-S2Mo electrodes are mainly used tween the two pieces of steel — Fig. 1B. Any opening be-
for welding of X70 microalloyed steels to provide a slightly tween the steel sheets would act as a crack during fracture
higher carbon content in the weld pool. The higher carbon toughness testing.
content in the weld retards the transformation of austenite In this study, the aim was to specifically evaluate the in-
to ferrite in the weld until the austenite-to-ferrite transfor- fluence of cold wire addition to the OD welding pass on
mation starts in the CGHAZ. The delay in the austenite-to- welding heat input and, consequently, the microstructure
ferrite transformation in the weld due to the slightly higher and toughness of the HAZ. Accordingly, the welding param-
carbon content results in the formation of low-temperature- eters for the ID weld run (performed via TSAW) were similar
transformation products, particularly acicular ferrite in the for all the weld samples; only the OD welding parameters
weld, which lead to higher fracture toughness of the weld. varied (Table 2). Direct current electrode positive (DCEP)
The BA-S2Mo electrode with the specified composition is and square wave alternating current (ACSQ) polarity were
also widely used by many pipe manufacturers. The granular employed, using constant current type power sources to op-
flux BF 6.4 was selected according to EN 760 (Bavaria, Ger- erate the lead and trail electrodes in the OD weld run, re-
many) with a basicity index of 1.7, and is typically used for spectively. Both the lead and trail electrodes in the ID weld
welding X70 and X80 microalloyed steels. run were operated in ACSQ polarity. As stated earlier, the
Table 1 indicates the chemical compositions of the mi- additional wire is electrically cold with no current and volt-
croalloyed steel and consumable electrodes. Pcm, the weld- age. The OD welding parameters and the welding process

Table 1 — Nominal Chemical Composition of X70 Microalloyed Steel and the Electrodes (wt-%)

X70 Steel C N V + Mo + Nb + Ti P S Mn Si Cu + Ni + Cr Pcm

0.046 0.0098 0.21 0.01 0.001 1.76 0.24 0.54 0.175
Electrode + C P S Mn Si Mo Ni Cr Cu
Cold Wire 0.10 0.007 0.01 1.04 0.1 0.56 0.02 0.03 0.03

340-s WELDING JOURNAL / DECEMBER 2018, VOL. 97

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 341

WELDING RESEARCH

Table 2 — ID and OD Welding Process Parameters

Process Parameters ID-Weld OD-Weld OD-Weld

(Polarity) Cold-Wire TSAW TSAW
(Polarity) (Polarity)
Current – Lead Electrode (A) 880 (ACSQ) 1040 (DCEP) 1040 (DCEP)
Current – Trail Electrode (A) 770 (ACSQ) 830 (ACSQ) 830 (ACSQ)
Voltage – Lead Electrode (V) 28 30 30
Voltage – Trail Electrode (V) 30 34 34
Welding Travel Speed (cm/min) 160 160 160
Cold Wire Position NA Lagging NA
Cold Wire Angle NA 63 deg NA
Cold Wire Feed Speed (cm/min) NA 25.4 NA

setup were developed based on the optimized welding pa- Figure 3A and B show macrographs of samples welded us-
rameters, which are discussed elsewhere (Refs. 35, 38). The ing cold-wire TSAW and TSAW processes, respectively, indi-
fabricated welds were visually inspected prior to further in- cating the position of the initial crack. Figure 3C and D de-
vestigation to ensure that no macro-level defects, such as pict a side and top macroview, respectively, of the final
undercutting, melt-through, surface porosity, and hot crack- notched SENT specimen including side grooves. The SENT
ing, were formed in the weld region and the HAZ. The nomi- tests were conducted at three different temperatures, i.e.,
nal heat input for both welding processes was constant and room temperature (RT), –5 and –30C. For the colder tem-
set at 22.2 kJ/cm; however, as stated earlier, cold wire addi- peratures, the specimens were cooled by circulating ethanol
tion results in a reduction in the welding heat input. All through a small chamber containing the specimen length
welding parameters were the same for both processes other between the grip faces. The flow rate of ethanol was varied
than the additional cold wire. within the chamber to control the testing temperature
at –5 and –30C. The instantaneous temperature at the
Fracture Toughness Testing vicinity of the crack tip was measured using a thermocouple.
For testing at low temperature, the temperature was kept
From the eight welding runs conducted by the TSAW and at –30 ± 3°C according to E1820 Standard through a dis-
cold-wire TSAW processes, eight weld coupons were cut in tance of at least ± 2 B from the notch. To restrict all rota-
the transverse direction relative to the weld direction to ex- tional degrees of freedom during SENT testing, the speci-
amine the HAZ fracture toughness via SENT testing. Prior mens were clamped using hydraulic grips. A SENT specimen
to machining, the SENT specimen blanks were straightened was then loaded/unloaded/reloaded under displacement
through a bend flattening procedure (described in Part 2 of control with a displacement rate of 0.01 mm/s, representing
the BS7448 Standard (Ref. 41)) to compensate for weld dis- quasistatic conditions. The methodology of the loading/
tortion and to preserve as much specimen thickness as pos- unloading cycles has been adopted from the recommended
sible. The specimens were then machined to a final thick- practice by CanmetMaterials (Ref. 7).
ness (B) and width (W) of 10.8 ± 0.05 mm as proposed by At the initial stage of the test, some loading/unloading
Shen et al. (Ref. 42). The specimen length (L) was ~ 108 mm cycles were carried out to calculate the initial crack size prior
(L = 10 W), plus additional length to accommodate gripping to the initial elastic part of the load-CMOD (crack mouth
of the ends with hydraulic wedges. The specimens were sur- opening displacement) test. Then, the calculated value was
face notched by electrical discharge machining (EDM) from compared with the measured initial crack size, a0. The load
the OD surface and then fatigue precracked (according to varied between 0.25PY and 0.5PY for the initial load/unload
ISO 12135:2002/ISO 15653:2010) to target the microstruc- cycles. Afterward, six unloading cycles were performed in
ture of the HAZ, particularly the CGHAZ. The desired initial the elastic regime with an applied force, PY, equal to
crack depth was a0 = 0.25 W (W = B).
Due to differences in the constraints at mid-thickness PY = ys + uts (W – a) Be (1)
(toward plane strain) and at the sides (plane stress) of the 2
specimen, crack tunneling was expected (Ref. 43). Tunneling where ys and uts are the yield strength and ultimate tensile
degrades the accuracy of the crack growth measurements strength of the specimen, respectively. W and Be are the
(Ref. 44). Accordingly, following fatigue precracking, the specimen width and the effective specimen width, respec-
SENT specimens for J-R curve development were side tively, and a is the crack length of the SENT specimen. The
grooved to a depth of approximately 0.8 mm on each side effective specimen width, Be, is calculated from Equation 2:
(for a total reduction of up to 15% of the thickness) with a
groove radius of approximately 0.25 ± 0.05 mm and an angle Be = BN (2B – BN)/B (2)
of 45 deg. As such, the net thickness, BN, was equal to 0.85
B. Side grooving was conducted to minimize final crack cur- where BN is the net thickness of the side grooved specimen
vature by generating relatively uniform constraints across (0.85 B). The ys and uts values of the CGHAZ were deter-
the crack tip. The geometry of the SENT specimen is shown mined using microhardness testing. According to Cheng
in Fig. 2B. (Ref. 45), the relationship between the hardness and the

DECEMBER 2018 / WELDING JOURNAL 341-s

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 342

WELDING RESEARCH

Fig. 4 — Charpy absorbed energy and J-value of the CGHAZ B

for steel samples welded by TSAW and cold-wire TSAW
processes. Dashed lines and solid lines represent the SENT
and CVN test results, respectively.

yield strength was empirically determined based on analysis

of the dimensions of the indentations. In addition, one em-
pirical equation is widely used in the literature, ys = HV/0.3,
which approximates the relationship between the hardness
HV and the yield strength. A relationship has been devel-
oped for welded C-Mn steels (ISO 15653 Standard (Ref. 46))
that can be used to estimate the ys (in the range of 160 HV
to 495 HV) and uts (in the range of 100 to 400 HV) of the
steel based on the Vickers hardness (HV) value, as follows:

ys (MPa) = 3.28HV – 221 (3)

UTS (MPa) = 3.3HV – 8 (4)

Fig. 5 — A typical load vs. CMOD plot for the following: A — A
According to Equations 3 and 4, the ys and uts values of ductile SENT specimen from a cold-wire TSAW weld; B — a
the CGHAZ of the TSAW and cold-wire TSAW specimens are brittle SENT specimen from a TSAW weld showing pop-in.
559 and 776 MPa, and 520 and 737 MPa, respectively, for the
corresponding microhardness values of 237.8 HV0.5 (TSAW
possible to extract full-size Charpy specimens from the
weld) and 225.9 HV0.5 (cold-wire TSAW weld). After the elas-
HAZ. As such, subsize specimens were extracted as close to
tic unloading cycles, plastic unloading cycles were performed
the specimen surface as possible to ensure positioning of
with CMOD intervals of 0.02 and 0.04 mm for the first five
half of the notch in the CGHAZ and the other half in the
unloading cycles and the subsequent unloading cycles, respec-
fine-grained heat-affected zone (FGHAZ). Figure 2C illus-
tively. In the unloading compliance method, a monotonically
trates the location of the notch for the Charpy samples.
increasing relation between crack depth and compliance is as-
sumed. Accordingly, the specimen is initially unloaded and re-
loaded at predefined intervals of CMOD, as described above, to Microstructural Analysis
evaluate the compliance during the test. The compliance is de-
fined as the slope of a linear regression line through the un- The TEOM technique was utilized to analyze the forma-
loading data set. tion of microstructural features, particularly PAG size and
In addition to SENT fracture toughness testing, CVN M-A constituents, in the HAZ of the weld samples (Ref. 48).
tests were performed on subsize bars extracted from the For this purpose, a chemical solution containing 4 g of picric
weld specimens, with dimensions of 5  10  55 mm, at acid in 96-mL ethanol, along with a few drops of HCl acid, was
RT, –30˚ and –45˚C. At least five Charpy specimens per test prepared to clearly reveal the PAG boundaries. The mean line-
condition were machined in the transverse orientation rela- al intercept method (ASTM E112 (Ref. 49)) was used to inves-
tive to the welding direction according to ASTM E23-12c tigate the PAG size in the CGHAZ. Freshly polished specimens
(Ref. 47). Because of the small size of the HAZ and the bul- were then tint etched through a separate process using modi-
bous shape of the weld metal (WM) and the HAZ, it was not fied LePera’s etchant for 30–50 s to reveal different mi-

342-s WELDING JOURNAL / DECEMBER 2018, VOL. 97

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 343

WELDING RESEARCH

A B G H

C D I J

E F K L

Fig. 6 — Optical macrographs of the fracture surfaces of SENT specimens after testing as follows: A, B — –30˚C; C, D — –5˚C; and
E, F — RT for cold-wire TSAW weld samples; G, H — –30˚C; I, J — –5˚C; and K, L — RT for TSAW weld samples. The left column repre-
sents the initial crack (fatigue precrack) size measurements and the right column represents the final crack length measure-
ments. The small region on top of each specimen represents the EDM notch, which was machined prior to precracking of the
specimens. Small grooves along both sides of each specimen are the side grooves machined on the specimens after precracking
and prior to SENT testing.

crostructural features. Quantitative analysis of the various Fracture Toughness Analysis Methodology
phase fractions formed during welding in the FGHAZ and the
CGHAZ was conducted using ImageJ commercial image According to fracture mechanics principles, the J-integral is
analysis software according to ASTM E562 (Ref. 50). To evalu- separated into elastic (Jel,i) and plastic (Jpl,i) components (Equa-
ate the ferrite/bainite (/B) grain size variation by cold wire tion 5) to generate J-R curves (ASTM E1152-87 (Ref. 51)).
addition, electron backscattered diffraction (EBSD) analysis
was conducted on the CGHAZ of the weld samples using a J = Jel,i + Jpl,i (5)
Zeiss Sigma field emission gun SEM, operating at 20 kV and
equipped with an HKL EBSD system. Jel,i is directly calculated from the stress intensity factor, Ki,

Table 3 — Initial and Final Crack Straightness Assessment*

Notch Welding Test Average Measurement Straightness Average Measurement Straightness

Position Process Temperature Initial Crack Difference, Validity Final Crack Difference, Validity
(°C) Length (mm) mm (% B) Length (mm) mm (% B)

Cold-Wire –30 2.74 0.14 (1.2) Valid 3.04 0.13 (1.1) Valid
TSAW –5 2.63 0.21 (1.8) Valid 5.65 0.11 (0.9) Valid
RT 2.72 0.21 (1.7) Valid 4.71 0.57 (4.7) Valid
CGHAZ
–30 2.67 0.11 (1.0) Valid 2.84 0.13 (1.1) Valid
TSAW –5 2.58 0.11 (0.9) Valid 3.18 0.13 (1.1) Valid
RT 2.42 0.36 (3.0) Valid 4.87 0.47 (3.9) Valid

*ASTM E1820-15 assesses crack straightness based on the maximum difference of all nine physical crack measurements from the average measurement and considers
the straightness valid when the maximum difference is less than or equal to 0.05 B.

DECEMBER 2018 / WELDING JOURNAL 343-s

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 344

WELDING RESEARCH

A B

C D

Fig. 7 — A — Typical J-R curve indicating the methodology for measuring the Jq value from the J-R curve. J-R curves for the CGHAZ
of cold-wire TSAW and TSAW specimens at the following: B — RT; C — –5˚C; and D — –30˚C.

for a plane strain crack, which is defined in Equation 6.

Equation 7 expresses the relationship for Jpl,i, which is a
( )
CMOD
V Vpl 
function of the plastic area under the load-CMOD curve. J pl ,i =  J pl ,i–1 + i
1 Ai pl  Ai1 

BN bi1 
2 LLD
(1 v 2 )
 Pi ai   a 
i1 
J el =   G  i   (6)  1 b (ai  ai1 ) (7)
E  BBN W  W
i1 


Table 4 — J-Integral Values for the CGHAZ of Weld Specimens Tested with the SENT Method

Notch Position Welding Process Test Initiation Toughness, Toughness Jm

Temperature (°C) J (kJ/m2) Qualification (kJ/m2)
Cold-Wire RT 966 Jq 459
TSAW –5 955 Jq 457
–30 320 Jqc 320
CGHAZ
TSAW RT 815 Jq 474
–5 576 Jqc 576
–30 228 Jqc 228

344-s WELDING JOURNAL / DECEMBER 2018, VOL. 97

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 345

WELDING RESEARCH

A B

C D

E F

Fig. 8 — Micrographs showing the microstructural features in the CGHAZ of the (A, C) TSAW weld and (B, D) cold-wire TSAW weld.
The PAG boundaries in the CGHAZ are highlighted in E and F for the TSAW and cold-wire TSAW weld, respectively. Images A, B, E,
and F, along with C and D, are TEOM and SEM secondary electron (SE) micrographs, respectively.

DECEMBER 2018 / WELDING JOURNAL 345-s

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 346

WELDING RESEARCH

A B

Fig. 9 — A — Fraction of microstructural constituents formed in the CGHAZ and FGHAZ; B — PAG and /B grain size analysis in the
CGHAZ of welded X70 microalloyed steels by TSAW and cold-wire TSAW processes.

where E is the Young’s modulus,  is Poisson’s ratio, ai is the ment of the center of the remaining ligament toward the
current crack size, and Pi is the load at the beginning of the load line plays a significant role in the crack propagation. Ac-
unload/load cycle. G(ai/W) is a function of the crack size to cordingly, the value of elastic compliance given in Equation
width ratio, bi is the ligament size (W – ai) and Apl,i is the 8 is corrected according to Equation 9 as follows:
plastic area under the load-CMOD curve. The parameters
iCMOD and iLLD have been defined by the finite element Ci
analyses conducted by Shen et al. (Ref. 52) under a two- Cc ,i = (9)
a  P 
dimensional plane strain assumption. These parameters are 1 – 0.165 0  i 
defined separately in the Appendix at the end of this paper. W  PY 
Equation 8 was developed to determine the crack size based
on CMOD elastic compliance (Ci) measurements via finite where Cc,i is the compliance after specimen rotation correc-
element analysis. The equation is valid for a/W ratios in the tion, which is established by means of two-dimensional,
range of 0.05 to 0.95. plane-strain finite element analysis for the SENT specimens
with length to width ratio of 10 and a0/W between 0.2 and
0.5. Pi and PY are the load at the beginning of the unload/
10
ai ri load cycle and the load limit, respectively.
= (8)
W i = 1 BeCi E + 1
Results
According to the ISO 12135 Standard, ri is the coefficient The average CVN impact energy (average of at least five
of a polynomial least-squares fitting function (Ref. 53), Charpy tests per test condition) at three different tempera-
which is defined in the Appendix. It is interesting to note tures for the weld samples is presented in Fig. 4. The average
that, according to the formulas mentioned in ASTM E1820- impact energy was increased by the addition of a cold wire
11 (Refs. 54, 55), the changes in CMOD compliance as a to the conventional TSAW process, with a corresponding re-
function of a/W for clamped SENT specimens (expressed as duction in the heat introduced to the weldment. Due to the
BeCiE in Equation 8) are significantly larger than for single proximity of the test temperature at –30˚C to the ductile-to-
edge notched bend (SENB) specimens. The work done by brittle transition temperature (DBTT) for this type of steel
Cravero et al. (Ref. 56) and Joyce et al. (Ref. 57) showed that (i.e., ~ –30˚C), large error bars for the CVN impact energy of
specimen rotation during SENT testing due to the move- the HAZ weld samples were observed. Shome et al. (Ref. 25)

Table 5 — Characteristics of M-A Constituents in the CGHAZ of Different Welds

Weld Sample Fraction (vol-%) Mean Size (m) Interspacing (m)

Cold-Wire TSAW 3.0 ± 0.1 0.3 ± 0.02 2.5 ± 0.1

TSAW 4.1 ± 0.1 0.48 ± 0.03 4.9 ± 0.2

346-s WELDING JOURNAL / DECEMBER 2018, VOL. 97

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 347

WELDING RESEARCH

A B

C D

Fig. 10 — EBSD maps showing /B grains based on boundaries with these angles: A, C — > 10 deg; B, D — > 15 deg in the CGHAZ of
(A, B) TSAW and (C, D) cold-wire TSAW samples.

reported that the Charpy energy for ferritic steels common- from the load-CMOD plots, it was essential to assess the
ly exhibits large scatter in the DBTT region. straightness of the initial fatigue precrack and the final
Figure 5A and B depict typical load-displacement curves crack during SENT testing for a sufficient estimation of the
for a fully ductile specimen (e.g., RT test specimen) and a toughness. For this reason, according to ASTM E1820-11
specimen with brittle fracture (e.g., –30°C test specimen). procedures, the assessment of the initial and final crack
The pop-in phenomenon (Fig. 5B) is typically observed for straightness was conducted, and the specimens were sec-
load-displacement curves of brittle materials (an interrup- tioned to ensure the crack propagated in the HAZ, particu-
tion in the load-displacement at relatively low displace- larly the CGHAZ.
ments) and is due to brittle instability of the specimen (Ref. According to the standard, the SENT specimens were sec-
58). The CMOD compliance relationships developed by Can- tioned at cryogenic temperatures and heat tinted and then
metMaterials for SENT specimens (Refs. 55, 56) were used nine physical crack measurements were performed. The
to generate J-R curves by regression of the load-displace- crack straightness was evaluated based on the maximum dif-
ment data. However, prior to calculation of the J-R curves ference of the nine crack length measurements from the av-

DECEMBER 2018 / WELDING JOURNAL 347-s

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 348

WELDING RESEARCH

erage value. As per Section 9.1 of the standard, the crack

straightness was considered valid if the maximum difference
was ≤ 0.05 B. Figure 6 shows macrographs of the fracture
surface of the SENT tested weld samples indicating the
EDM notch and the length of the initial and final crack.
Crack straightness assessment and average length of the ini-
tial and final cracks are summarized in Table 3.
The J-R curves for weld samples tested in the CGHAZ at
RT, –5° and –30°C are shown in Fig. 7. A typical J-R curve indi-
cating the methodology of measuring J-integral values from
the J-R curve (Refs. 59, 60) is shown in Fig. 7A. Initially, a con-
struction line, with a slope that was a function of the yield and
ultimate tensile strengths of the material and the crack exten-
sion was drawn. This was followed by drawing an offset line
parallel to the construction line at 0.2 mm offset.
Jq is the J-integral at the onset of ductile tearing (0.2 mm
offset) according to ASTM E1820-11 Annex A9 (Refs. 54,
55). Jq and Jm (J-integral at the maximum load) or Jqc (J-
integral at pop-in with less than 0.2 mm of prior ductile
tearing) values for SENT testing in the CGHAZ of the weld Fig. 11 — The /B boundary misorientation angle distribution
in the CGHAZ of microalloyed steel welded by the cold-wire
specimens are presented in Table 4. Jq was measured as the TSAW and conventional TSAW processes.
fracture toughness for the specimens with ductile fracture;
however, for specimens with a brittle fracture mode, Jqc
and/or Jm were considered as the fracture toughness values five grain size measurements conducted on the TEOM and
because there was not sufficient data in the J-R curves for EBSD micrographs is shown in Fig. 9B. These alterations in
evaluation of Jq. The large scatter in the J-R curves for weld grain size and microstructural features were attributed to a re-
specimens tested at –30°C was primarily due to the brittle duction in the heat introduced to the steel and, consequently,
behavior of the specimen at this temperature. According to a decrease in the retention time in the austenitic region, i.e.,
the J-integral results shown in Fig. 4, there is an increase in 1100˚–1400˚C, and an increase in the cooling rate, which are
the J value at the three different test temperatures through discussed in detail in the next section.
the addition of a cold wire to the conventional TSAW
process, with a corresponding reduction in the heat intro- Discussion
duced to the weldment. The trend in the J-integral values
for the CGHAZ of the cold-wire TSAW samples relative to The addition of a cold wire to the conventional TSAW
the TSAW samples correlates with the CVN test results. process resulted in an increase in the absorbed impact energy
As stated earlier, the fracture toughness in the CGHAZ and J-value of the CGHAZ obtained by CVN and SENT testing,
was influenced by the microstructural changes taking place respectively. The heat input and the thermal cycle that the
during welding as a result of the thermal cycle the steel ex- steel experiences during welding essentially governs the varia-
periences. The microstructure of the base X70 steel was pre- tions in fracture toughness values for the HAZ of welded mi-
dominantly polygonal ferrite (PF – 85.5%) and granular bai- croalloyed steel (Ref. 29). In addition, property changes in the
nite (GB – 9.8%), with a small fraction of bainitic ferrite CGHAZ were attributed to the alterations in the PAG size and
(BF – 3.2%) and M-A constituents (1.5%). However, the the fraction and morphology of the microstructural features,
TEOM (Fig. 8A and B) and SEM (Fig. 8C and D) images indi- particularly the M-A constituent, which are directly influenced
cate the microstructure was significantly altered in the CG- by the heat introduced to the steel. Considering the fact the
HAZ. The shape, size, and fraction of the M-A constituents characteristics of the M-A constituents (as the key microstruc-
in the CGHAZ were changed by cold wire addition and the tural feature formed in the CGHAZ of microalloyed steels)
consequent variation in the welding heat input. The phase govern both the crack initiation and propagation stages, a sim-
amounts formed in the CGHAZ and FGHAZ are shown in ilar trend was observed in the SENT and CVN toughness test
Fig. 9A. Because no significant changes in the microstruc- results.
ture of the FGHAZ were observed, the micrographs are not According to the TEOM and SEM micrographs of the
shown here. CGHAZ, shown in Fig. 8, the PAG size in the CGHAZ was de-
In addition to the phase transformations taking place in the creased by addition of the cold wire. The reduction in the PAG
CGHAZ, the PAG and ferrite/bainite grain size were reduced size in the CGHAZ from 68 m (TSAW weld) to 55 m (cold-
by the additional cold wire. Micrographs highlighting the PAG wire TSAW weld) was attributed to a reduction in the heat in-
boundaries using the TEOM technique and the ferrite/bainite troduced to the weldment and, consequently, a reduction in
grain boundaries via EBSD are shown in Fig. 8E and F and Fig. the retention time in the austenitization temperature range
10A–D, respectively. Individual ferrite/bainite grains were (1100˚–1400˚C) and an increase in the cooling rate by feeding
identified either as having boundary angles greater than 10 a cold wire to the welding process. An analysis of the heat in-
deg or greater than 15 deg. A 10–15 deg misorientation of two put, based on the thermodynamic principles, which is dis-
neighboring grains was considered as the transition from low- cussed in Ref. 61, indicated the heat input of the cold-wire
angle to high-angle grain boundaries. The average of at least TSAW process was reduced by 9% relative to the TSAW

348-s WELDING JOURNAL / DECEMBER 2018, VOL. 97

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 349

WELDING RESEARCH

process with a corresponding increase in the cooling rate by However, large slender shaped M-A constituents formed in
10%. According to Fig. 8, the CGHAZ microstructure of the the CGHAZ of the TSAW sample can contribute more as
TSAW sample (with higher heat input) is made up of large crack initiation sites and promote the propagation of micro-
PAGs, PF, GB, BF, and large M-A constituents. In contrast, the cracks due to the relatively higher stress concentration at
CGHAZ microstructure of the cold-wire TSAW sample (with the elongated M-A constituents. The fine M-A constituents
lower heat input) is composed of finer PAGs, PF, GB, and BF with small interspacing formed in the CGHAZ of the cold-
associated with fine, blocky shaped M-A constituents. Phase wire TSAW sample can slow down, if not arrest, any cracks
fraction analysis (Fig. 9) showed less M-A is formed within the initiated at one of the M-A constituents from further propa-
CGHAZ of the cold-wire TSAW weld sample compared with gation. The work done by Davis et al. (Ref. 21) and Reichert
the TSAW sample. However, in addition to the fraction and et al. (Ref. 69) also confirm that the slender M-A con-
shape, the size and interspacing of M-A constituents in the stituents are the most detrimental to fracture toughness.
CGHAZ of the weld samples play a significant role in the In addition to the formation of M-A constituents, the fer-
fracture toughness. rite/bainite grain size affects the properties of the CGHAZ
Table 5 indicates the characteristics of the M-A con- of welded microalloyed steels. Low angle boundaries (small-
stituents formed in the CGHAZ of the two weld specimens. er than 10–15 deg) are not effective obstacles for crack
The fraction of M-A constituents in the CGHAZ was reduced propagation and, as a result, do not significantly contribute
as a consequence of PAG size reduction (Refs. 23, 24) by the to the toughness. High angle grain boundaries (larger than
cold wire addition. Quantitative analysis of the M-A regions 10–15 deg), on the other hand, have a greater effect on
indicates the fraction of M-A constituents in the CGHAZ properties (Refs. 70–72). Accordingly, 10- and 15-deg
with sizes larger than 2 m for the TSAW and cold-wire thresholds were chosen as the criteria for the EBSD mapping
TSAW samples is 3.2 ± 0.2% and 1.0 ± 0.1%, respectively. of ferrite/bainite grains — Fig. 10. Although the ferrite/bai-
The size analysis indicates there is a reduction in the M-A nite grains in the CGHAZ were significantly coarser than
size with a corresponding reduction in M-A spacing upon those in the base X70 microalloyed steel (4.1 ± 0.5 m), the
the addition of a cold wire. The M-A interparticle spacing addition of a cold wire at a rate of 25.4 cm/min reduced the
was determined using the equation developed by Somekawa grain size (grains with boundary angles larger than 15 deg)
(Ref. 62). in the CGHAZ from 17.6 ± 5.0 m (TSAW weld) to 15.3 ±
The larger M-A constituents in the CGHAZ of the TSAW 4.5 m (cold-wire TSAW weld). The misorientation angle
sample were due to the higher martensite start temperature distribution of the ferrite/bainite boundary (≥ 15 deg) in
(Ms), which correlates with the larger PAGs (Refs. 28, 63). the CGHAZ of the two welds is shown in Fig. 11. Although
Bhadeshia (Ref. 64), Heinze et al. (Ref. 65), and Guimaraes more grains with boundary angles smaller than 31 deg
(Ref. 66) suggested that a decrease in Ms temperature corre- formed in the CGHAZ of the TSAW weld, a relatively higher
lates with a decrease in the PAG size, which results in a low- fraction of grains formed with boundary angles greater than
er volume fraction of martensite. According to Koistinen 31 deg in the CGHAZ of the cold-wire TSAW weld. Because a
and Marburger (Ref. 67) and Fisher et al. (Ref. 68), the frac- larger grain boundary angle generally has a greater effect on
tion of martensite is a function of the amount of undercool- the fracture toughness, the grains formed in the CGHAZ of
ing below the Ms temperature. As such, the volume fraction the cold-wire TSAW weld are likely to provide more barriers
of martensite, which was formed in the early stages of the to crack propagation in the CGHAZ relative to the CGHAZ
austenite transformation, was proportional to the PAG size of the TSAW weld. As such, in addition to the reduction in
cubed. As such, “for a larger PAG, the fraction of the trans- the fraction of M-A constituents and the refinement in the
formation needed to detect Ms is reached at a smaller un- morphology of M-A constituents, the addition of a cold wire
dercooling” (Ref. 63). Therefore, the fraction and size of (which lowers the welding heat input) resulted in grain size
formed M-A constituent increased for a coarser PAG size. In refinement due to the lower peak temperature and faster
this regard, the work done by Li et al. (Ref. 23) and Yu et al. cooling rate in the CGHAZ. These two metallurgical factors
(Ref. 24) showed that a microstructure comprised of coarse provided an improvement in the HAZ fracture toughness.
PAGs and coarse M-A constituents promotes brittle fracture
in the CGHAZ. As such, both grain size refinement and the Conclusions
M-A transformation concurrently play a significant role in
the HAZ toughness of welded microalloyed steels. The influence of a recently developed welding process, i.e.,
Due to the increased transformation of austenite to the tandem submerged arc welding with additional cold wire (cold-
M-A constituent, a higher proportion of localized brittle wire TSAW), on the microstructure and fracture toughness of
zones, i.e., the M-A constituents, was formed in the CGHAZ an X70 microalloyed steel was evaluated. Single-edge notched
of the TSAW sample relative to that of the cold-wire TSAW tension (SENT) and the Charpy V-notch (CVN) impact tests
sample, which resulted in higher microhardness values in indicated the fracture toughness of the coarse-grained heat-
the CGHAZ for the TSAW sample (237.8 ± 5.5 HV0.5) rela- affected zone (CGHAZ) was improved by a cold wire addition
tive to the cold-wire TSAW sample (225.9 ± 4.5 HV0.5). into the welding process. The cold wire addition primarily led
Moreover, a finer distribution of small M-A constituents to a reduction in the prior austenite grain (PAG) size in the
within the ferritic matrix of the CGHAZ of the cold-wire CGHAZ. However, microstructural evaluation indicated the
TSAW sample resulted in higher J-integral values and fraction, size, and interspacing of martensite-austenite (M-A)
Charpy energies at different test temperatures. The M-A constituents in the CGHAZ were reduced, and the shape was
constituent was significantly harder than the internal grain altered from slender to blocky when a cold wire was fed into
microstructure, so that cracks initiate easily at the M-As. the TSAW process. Furthermore, ferrite/bainite grains were

DECEMBER 2018 / WELDING JOURNAL 349-s

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 350

WELDING RESEARCH

refined in the CGHAZ by addition of the cold wire. A combina- and Koo, J. 2001. Ultra-high strength, weldable steels with excel-
tion of relatively large PAGs and ferrite/bainite grains, and lent ultra-low temperature toughness. ExxonMobil Upstream Re-
large elongated M-A constituents, which were mostly formed search Co. Houston, Tex., Nippon Steel Corp., Tokyo, Japan.
along the boundaries in the CGHAZ of the TSAW samples, was 13. Honeycombe, R. W. K., and Bhadeshia, H. K. D. H. 1996.
Steels: Microstructure and Properties, Metallurgy and Materials Sci-
responsible for the inferior toughness properties in the HAZ
ence Series. Edward Arnold, London, Great Britain.
relative to the cold-wire TSAW samples. The formation of larg- 14. Li, C., Wang, Y., and Chen, Y. 2011. Influence of peak tem-
er M-A constituents in the CGHAZ can stimulate the forma- perature during in-service welding of API X70 pipeline steels on
tion of microcracks, which lead to intergranular fracture. microstructure and fracture energy of the reheated coarse grain
heat-affected zones. J. Mater. Sci. 46: 6424–6431. DOI: 10.1007/
s10853-011-5592-7
Acknowledgments 15. Silwal, B., Li, L., Deceuster, A., and Griffiths, B. 2013. Effect
of postweld heat treatment on the toughness of heat-affected zone
The authors would like to acknowledge the Natural Sci- for grade 91 steel. Welding Journal 93: 80-s to 87-s.
ences and Engineering Research Council (NSERC) of Cana- 16. Lambert, A., Drillet, J., Gourgues, A. F., Sturel, T., and
Pineau, A. 2000. Microstructure of martensite-austenite con-
da, Evraz Inc. NA, TransCanada PipeLines Ltd., and En-
stituents in heat affected zones of high strength low alloy steel
bridge Pipelines Inc. for their financial support. The assis- welds in relation to toughness properties. Sci. Technol. Weld. Join. 5:
tance of Jonas Valloton with EBSD is appreciated. SGS Lud- 168–173. DOI: 10.1179/136217100101538164
wig Canada and the R&D Division of Evraz Inc. NA are also 17. Moeinifar, S., Kokabi, A. H., and Hosseini, H. R. M. 2011.
recognized for technical assistance in conducting SENT test- Effect of tandem submerged arc welding process and parameters of
ing and CVN testing, respectively. Gleeble simulator thermal cycles on properties of the intercritically
reheated heat affected zone. Mater. Des. 32: 869–876. DOI:
10.1016/j.matdes.2010.07.005
References 18. Moeinifar, S., Kokabi, A. H., and Hosseini, H. R. M. 2011.
Role of tandem submerged arc welding thermal cycles on proper-
ties of the heat affected zone in X80 microalloyed pipe line steel.
1. Moeinifar, S. 2011. Microstructure and toughness properties J. Mater. Process. Technol. 211: 368–375. DOI: 10.1016/
of subcritically, intercritically and supercritically heat affected j.jmatprotec.2010.10.011
zones in X80 microalloyed pipeline steel. Adv. Mater. Res.: 383–390, 19. Gharibshahiyan, E., Honarbakhsh, A., Parvin, N., and
5886–5893. DOI: 10.4028/www.scientific.net/AMR.383-390.5886 Rahimian, M. 2011. The effect of microstructure on hardness and
2. Kimura, H., Yokota, N., TomoyukiIshikawa, Kakihara, S., toughness of low carbon welded steel using inert gas welding.
Kondo, J., and Yokota, T. 2016. Development of grade X80 high Mater. Des. 32: 2042–2048. DOI: 10.1016/j.matdes.2010.11.056
charpy energy linepipe by MA formation control. 11th Int. Pipeline 20. Davis, C. L., and King, J. E. 1993. Effect of cooling rate on
Conf. American Society for Mechanical Engineers pp. 1–6. intercritically reheated microstructure and toughness in high
3. Ju, J.-B., Kim, W., and Jang, J. 2012. Variations in DBTT and strength low alloy steel. Mater. Sci. Technol. 9: 8–15.
CTOD within weld heat-affected zone of API X65 pipeline steel. 21. Davis, C. L., and King, J. E. 1994. Cleavage initiation in the
Mater. Sci. Eng. A 546: 258–262. DOI: http://dx.doi.org/10.1016/ intercritically reheated coarse-grained heat-affected zone: Part I.
j.msea.2012.03.062 Fractographic evidence. Metall. Mater. Trans. A 25: 563–573. DOI:
4. McCabe, D. E. 1972. Evaluation of the compact tension speci- 10.1007/BF02651598
men for determining plane-strain fracture toughness of high 22. Reichert, J. M., Garcin, T., Militzer, M., and Poole, W. J.
strength materials. J. Mater. Sci. 7: 449–542. 2014. A new approach using EBSD to quantitatively distinguish
5. Zhu, X.-K. 2016. Review of fracture toughness test methods complex transformation products along the HAZ in X80 linepipe
for ductile materials in low-constraint conditions. Int. J. Press. Ves- steel. 9th Int. Pipeline Conf. American Society of Mechanical Engi-
sel. Pip.: 139, 140, 173–183. neering, Calgary, AB.
6. Det-Norske-Veritas. 2006. Control for pipeline installation 23. Li, X., Ma, X., Subramanian, S. V., Shang, C., and Misra, R. D.
methods introducing cyclic plastic strain. K. 2014. Influence of prior austenite grain size on martensite-
7. Canmet Materials. 2010. Recommended practice: Fracture austenite constituent and toughness in the heat affected zone of
toughness testing using SE(T) samples with fixed-grip loading. 700MPa high strength linepipe steel. Mater. Sci. Eng. A 616: 141–
8. Park, D.-Y., Gravel, J.-P., Arafin, M., Liang, J., and Hari Manoj 147. DOI: 10.1016/j.msea.2014.07.100
Simha, C. 2014. Evaluation of two low-constraint toughness test 24. Yu, L., Wang, H. H., Hou, T. P., Wang, X. L., Wan, X. L., and
methods in a single specimen. J. Eng. Mater. Technol. 137: 11003. Wu, K. M. 2014. Characteristic of martensite-austenite con-
DOI: http://dx.doi.org/10.1115/1.4028728 stituents in coarse grained heat affected zone of HSLA steel with
9. ExxonMobil. 2010. Measurement of crack tip opening dis- varying Al contents. Sci. Technol. Weld. Join. 19: 708–714. DOI:
placement (CTOD) – Fracture resistance curves using single-edge 10.1179/1362171814Y.0000000246
notched tension (SENT) specimens. 25. Shome, M., Gupta, O. P., and Mohanty, O. N. 2004. Effect of
10. Park, D.-Y., Gravel, J. P., Arafin, M., Liang, J., and Hari simulated thermal cycles on the microstructure of the heat-affect-
Manoj Simha, C. 2015. Evaluation of two low-constraint toughness ed zone in HSLA-80 and HSLA-100 steel plates. Metall. Mater.
test methods in a single specimen. J. Eng. Mater. Technol. 137: Trans. A 35A: 985–996.
011003-1-011003-9. DOI: 10.1115/1.4028728 26. Spanos, G., Fonda, R. W., Vandermeer, R. A., and Matuszes-
11. Koo, J. Y., Luton, M. J., Bangaru, N. V., Petkovic, R. A., ki, A. 1995. Microstructural changes in HSLA-100 steel thermally
Fairchild, D. P., Petersen, C. W., Asahi, H., Hara, T., Terada, Y., cycled to simulate the heat-affected zone during welding. Metall.
Sugiyama, M., Tamehiro, H., Komizo, Y., Okaguchi, S., Hamada, Mater. Trans. A 26A: 3277–3293.
M., Yamamoto, A., and Takeuchi, I. 2004. Metallurgical design of 27. Yang, H. S., and Bhadeshia, H. K. D. H. 2009. Austenite
ultra high-strength steels for gas pipelines. Int. J. Offshore Polar grain size and the martensite-start temperature. Scr. Mater. 60:
Eng. 14: 2–10. 493–495. DOI: 10.1016/j.scriptamat.2008.11.043
12. Tamehiro, H., Asahi, H., Hara, T., Terada, Y., Luton, M. J., 28. Garcia-Junceda, A., Capdevila, C., Caballero, F. G., Garcia,

350-s WELDING JOURNAL / DECEMBER 2018, VOL. 97

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 351

WELDING RESEARCH

C., and Andre, D. 2008. Dependence of martensite start tempera- analysis, and indentation measurements. Mater. Sci. Eng. R Reports
ture on fine austenite grain size. Scr. Mater. 58: 134–137. DOI: 44: 91–149. DOI: http://dx.doi.org/10.1016/j.mser.2004.05.001
10.1016/j.scriptamat.2007.09.017 46. ISO, 15653:2010 Metallic Materials-Method of Test for the De-
29. Shome, M. 2007. Effect of heat-input on austenite grain size termination of Quasistatic Fracture Toughness of Welds. 2010.
in the heat-affected zone of HSLA-100 steel. Mater. Sci. Eng. A 445– 47. ASTM, E23-12C: Standard Test Methods for Notched Bar Im-
446: 454–460. DOI: 10.1016/j.msea.2006.09.085 pact Testing of Metallic Materials. 2012. ASTM International, Pa.
30. Zhang, Y. Q., Zhang, H. Q., Li, J. F., and Liu, W. M. 2009. Ef- 48. LePera, F. S. 1979. Improved etching technique for the de-
fect of heat input on microstructure and toughness of coarse grain termination of percent martensite in high-strength dual-phase
heat affected zone in Nb microalloyed HSLA steels. J. Iron Steel Res. steels. Metallography 12: 263–268. DOI: 10.1016/0026-0800(79)
Int. 16: 73–80. DOI: 10.1016/S1006-706X(10)60014-3 90041-7
31. Li, X., Fan, Y., Ma, X., Subramanian, S. V., and Shang, C. 49. ASTM, E112-12: Standard Test Methods for Determining Aver-
2015. Influence of martensite-austenite constituents formed at age Grain Size. 2012. ASTM International, Pa. DOI: 10.1520/
different intercritical temperatures on toughness. Mater. Des. 67: E0112-12.1.4
457–463. DOI: 10.1016/j.matdes.2014.10.028 50. ASTM, E562-11: Standard Test Method for Determining Vol-
32. Prawoto, Y., Jasmawati, N., and Sumeru, K. 2012. Effect of ume Fraction by Systematic Manual Point Count. 2011. ASTM Inter-
prior austenite grain size on the morphology and mechanical prop- national, Pa. DOI: 10.1520/E0562-11.2
erties of martensite in medium carbon steel. J. Mater. Sci. Technol. 51. Zhu, X. K., and Joyce, J. A. 2012. Review of fracture tough-
28: 461–466. DOI: 10.1016/S1005-0302(12)60083-8 ness (G, K, J, CTOD, CTOA) testing and standardization. Eng. Fract.
33. Cao, R., Li, J., Liu, D. S., Ma, J. Y., and Chen, J. H. 2015. Mi- Mech. 85: 1–46. DOI: http://dx.doi.org/10.1016/ j.engfracmech.
cromechanism of decrease of impact toughness in coarse-grain 2012.02.001
heat-affected zone of HSLA steel with increasing welding heat in- 52. Shen, G., and Tyson, W. R. 2009. Crack size evaluation using
put. Metall. Mater. Trans. A 46A: 2999–3014. DOI: 10.1007/ unloading compliance in single-specimen single-edge notched ten-
s11661-015-2916-2 sion fracture toughness testing. J. Test. Eval. 37(4): JTE102368.
34. Xia, Z. H., Wan, X. L., Tao, X. L., and Wu, K. M. 2012. Effect 53. ISO, 12135: Metallic Materials – Unified Method of Test for the
of heat input on toughness of coarse-grained heat-affected zone of Determination of Quasistatic Fracture Toughness. 2014.
an ultra low carbon acicular ferrite steel. Adv. Mater. Res. 538–541: 54. ASTM, E1820-13: Standard Test Method for Measurement of
2003–2008. DOI: 10.4028/www.scientific.net/AMR.538-541.2003 Fracture Toughness. 2013. ASTM International, Conshohocken, Pa.
35. Mohammadijoo, M., Kenny, S., Collins, L., Henein, H., and 55. Shen, G., Gianetto, J. A., and Tyson, W. R. 2008. Develop-
Ivey, D. G. 2017. Influence of cold-wire tandem submerged arc ment of procedure for low-constraint toughness testing using a
welding parameters on weld geometry and microhardness of mi- single-specimen technique. MTL Report No. 2008-18(TR), Canmet-
croalloyed pipeline steels. Int. J. Adv. Manuf. Technol. 88: 2249– Materials, Ottawa, Canada.
2263. DOI: 10.1007/s00170-016-8910-z 56. Cravero, S., and Ruggieri, C. 2007. Estimation procedure of
36. Mruczek, M. F., and Konkol, P. J. 2006. Cold wire feed sub- J-resistance curves for SE(T) fracture specimens using unloading
merged arc welding: Technical report. Advanced Technology Insti- compliance. Eng. Fract. Mech. 74: 2735–2757. DOI: 10.1016/
tute (ATI), Johnstown, Pa. j.engfracmech.2007.01.012
37. Ramakrishnan, M., and Muthupandi, V. 2013. Application 57. Joyce, J. A., and Link, R. E. 1995. Effect of constraint on up-
of submerged arc welding technology with cold wire addition for per shelf fracture toughness. Fract. Mech. ASTM STP 1256. ASTM
drum shell long seam butt welds of pressure vessel components. Int., Philadelphia, Pa.: 142–177.
Int. J. Adv. Manuf. Technol. 65: 945–956. DOI: 10.1007/ 58. Kirk, P. B. S., and Chum, V. 2014. Practical application of low
s00170-012-4230-0 constraint SENT fracture toughness testing for pipeline girth
38. Mohammadijoo, M., Kenny, S., Wiskel, J. B., Ivey, D. G., and welds. 10th Int. Pipeline Conf.: 1–11.
Henein, H. 2015. Cold-wire tandem submerged arc welding: A nov- 59. Park, D.-Y., Gravel, J.-P., Hari Manoj Simha, C., Liang, J., and
el technique for pipeline manufacturing. 54th Annual Conf. Metall. Duan, D.-M. 2016. Low-constraint toughness testing of two single-
Canadian Institute of Mining, Metallurgy and Petroleum, Toronto, edge notched tension methods in a single specimen. J. Press. Vessel
ON: 1–13. Technol. 138: 051401. DOI: 10.1115/1.4032998
39. Shikanai, N., Mitao, S., and Endo, S. 2007. Recent develop- 60. Lucon, E., Weeks, T. S., Gianetto, J. A., Tyson, W. R., and
ment in microstructural control technologies through the thermo- Park, D.-Y. 2015. Fracture toughness characterization of high-pres-
mechanical control process (TMCP) with JFE steel’s high-perfor- sure pipe girth welds using single-edge notched tension [SE(T)]
mance. JFE Technical Report No. 18 Plates, JFE Steel, Tokyo. specimens. Mater. Perform. Charact. 4: 20130098. DOI: 10.1520/
40. De Meester, B. 1997. The weldability modern structural MPC20130098
TMCP steels. ISIJ Int. 37: 537–551. 61. Mohammadijoo, M., Collins, L., Henein, H., and Ivey, D. G.
41. BS, 7448-1: Fracture Mechanics Toughness Tests – Part 1: 2017. Evaluation of cold wire addition effect on heat input and
Method for determination of KIc, Critical CTOD and Critical J Values productivity of tandem submerged arc welding for low-carbon mi-
of Metallic Materials. 1991. British Standard Institute (BSI). croalloyed steels. Int. J. Adv. Manuf. Technol. in press: 1–13. DOI:
42. Shen, G., Gianetto, J. A., and Tyson, W. R. 2009. Measure- 10.1007/s00170-017-0150-3
ment of J-R curves using single-specimen technique on clamped 62. Somekawa, H., and Mukai, T. 2006. Fracture toughness in
SE(T) specimens. 19th Int. Offshore Polar Eng. Conf., Osaka, Japan. an extruded ZK60 magnesium alloy. Mater. Trans. 47: 995–998.
43. Lan, W., Deng, X., and Sutton, M. A. 2010. Investigation of 63. Yen, H.-W., Huang, C.-Y., and Yang, J.-R. 2009. Characteriza-
crack tunneling in ductile materials. Eng. Fract. Mech. 77: 2800– tion of interphase-precipitated nanometer-sized carbides in a Ti–
2812. DOI: http://dx.doi.org/10.1016/j.engfracmech.2010.06.010 Mo-bearing steel. Scr. Mater. 61: 616–619. DOI: http://dx.doi.org/
44. Riemelmoser, F., Pippan, R., Weinhandl, H., and Kolednik, 10.1016/j.scriptamat.2009.05.036
O. 1999. The influence of irregularities in the crack shape on the 64. Bhadeshia, H. K. D. H. 2013. About calculating the charac-
crack extension measurement by means of the direct-current- teristics of the martensite-austenite constituent. Int. Semin. Weld.
potential-drop method BT – The influence of irregularities in the High Strength Pipeline Steels, CBMM, and The Minerals, Metals
crack shape on the crack extension measurement by means of the and Materials Society (TMS): 99–106.
direct-Cu. J. Test. Eval. 27: 42–46. 65. Heinze, C., Pittner, A., Rethmeier, M., and Babu, S. S. 2013.
45. Cheng, Y. T., and Cheng, C. M. 2004. Scaling, dimensional Dependency of martensite start temperature on prior austenite

DECEMBER 2018 / WELDING JOURNAL 351-s

Moham Supp Dec 2018 WJ 201815layout.qxp_Layout 1 11/8/18 2:49 PM Page 352

WELDING RESEARCH

grain size and its influence on welding-induced residual stresses. The coefficients of the polynomial regression (ti) based
Comput. Mater. Sci. 69: 251–260. DOI: 10.1016/j.commatsci. on finite element analyses conducted by Shen et al. (Ref. 55)
2012.11.058 are shown in Table A1.
66. Guimaraes, J. R. C., and Rios, P. R. 2010. Martensite start Parameters iCMOD and iLLD in Equation 7 are defined by
temperature and the austenite grain-size. J. Mater. Sci. 45: 1074–
the finite element analyses performed by Shen et al. (Ref.
1077. DOI: 10.1007/s10853-009-4044-0
67. Koistinen, D. P., and Marburger, R. E. 1959. A general equa- 52) under a two-dimensional plane strain assumption.
tion prescribing the extent of the austenite-martensite transfor-
mation in pure iron-carbon alloys and plain carbon steels. Acta
11 i
Metall. 7: 59–60. DOI: 10.1016/0001-6160(59)90170-1 a 
68. Fisher, J. C., Hollomon, J. H., and Turnbull, D. 1949. Kinet- C MOD = 
i  i

W
(A2)
ics of the austenite-martensite transformation. Trans. Am. Inst. i = 1
Min. Metall. Eng. 185: 691–700.
69. Reichert, J. M., Garcin, T., Militzer, M., and Poole, W. J.
2012. Formation of martensite/austenite (M/A) in X80 linepipe
steel. 9th Int. Pipeline Conf., Calgary, AB, Canada.  ai  
' LLD 
LLD =
LLD  1  1
(A3)
70. Flower, H. M., and Lindley, T. C. 2000. Electron backscatter- W  
LLD

ing diffraction study of acicular ferrite, bainite, and martensite
steel microstructures. Mater. Sci. Technol. 16: 26–40. DOI:
10.1179/026708300773002636
71. Rodriguez-Ibabe, J. M. 1998. The role of microstructure in
toughness behaviour of microalloyed steels. Mater. Sci. Forum. 284– 11 ai  k
286: 51–62. DOI: 10.4028/www.scientific.net/MSF.284-286.51 where LLD = i = 1 k
W 
 (A4)
72. Caballero, F. G., Capdevila, C., and Chao, J. 2010. The mi-
crostructure of continuously cooled tough bainitic steels. 2nd Int.
Conf. Super-High Strength Steels. Associazione italiana di metallur-
gia: 10–18.
Coefficients i and k are given in Table A2. The first de-
rivative of LLD, i.e., 'LLD, is also defined as
Appendix
The G(ai/W) parameter in Equation 6 is a function of the 11 k1
' a 
crack size to width ratio (a/W) as follows:
LLD =
k k  i

W
(A5)
12 i1 i = 1
 ai   ai 
G  = ti   (A1)
W W The ri in Equation 8 is the coefficient of a polynomial least-
i = 1
squares fitting function; values are provided in Table A2.

Table A1 — Coefficient ti in Equation A1 for 0.05  a/W  0.95 and L/W = 10

i 1 2 3 4 5 6 7 8 9 10 11 12

ti 1.20 –2.13 23.89 –69.05 100.46 –41.40 –36.14 51.21 –6.61 –52.32 18.57 19.47

Table A2 — Coefficients ϕi in Equation A2 and ψi in Equation A4 Used to Calculate η’CMOD and ηCMOD, Respectively; Coefficient ri in Equation 8 for
L/W = 10

i 1 2 3 4 5 6 7 8 9 10 11

i 1 –1.089 9.519 –48.572 109.225 –73.116 –77.984 38.487 101.401 43.306 –110.77
i –0.88 15.19 –35.44 18.64 18.40 –1.27 –12.76 –12.20 –4.45 5.40 14.19
ri 2.04 –15.73 73.24 –182.90 175.65 60.93 –113.99 –113.03 8.55 142.84

HANI HENEIN and DOUGLAS G. IVEY (divey@ualberta.ca) are with the Department of Chemical and Materials Engineering, University of
Alberta, Edmonton, AB, Canada. MOHSEN MOHAMMADIJOO and LAURIE COLLINS are with the R&D Division, Evraz Inc. NA, Regina, SK,
Canada. ROBERT LAZOR is with TransCanada PipeLines Ltd., Calgary, AB, Canada.

352-s WELDING JOURNAL / DECEMBER 2018, VOL. 97