Engineering Failure Analysis 149 (2023) 107252

Contents lists available at ScienceDirect

Engineering Failure Analysis

journal homepage: www.elsevier.com/locate/engfailanal

Investigation of welding of dissimilar high thickness SA106 Gr.C

carbon steel and SA335P12 alloy steel to eradicate the failure in
boiler components fabricated through SAW and SMAW processes
N. Prasanna, Anand Ramanathan *, N. Siva Shanmugam
Department of Mechanical Engineering, National Institute of Technology, Tiruchirappalli, India

A R T I C L E I N F O A B S T R A C T

Keywords: This study reports the effect of Shielded metal arc welding (SMAW) and Submerged arc welding
Submerged arc welding (SAW) processes used to fabricate high-thickness end cover plates by welding of dissimilar ma­
As-welded terials SA 335 P12 steel pipes and ASTM A 106 grade C carbon steel. The study emphasizes the
Hardness
structure-property relation of the welded joints made by two different techniques using macro­
Microstructure
Welding of dissimilar material
structure and microstructural examination with optical and electron microscopy and hardness
Scanning electron microscopy evaluation. The joints were subjected to magnetic particle non-destructive testing and the SAW
Energy dispersive x-ray spectroscopy joint was found to be defect-free. Microstructural examination revealed acicular ferrite micro­
structure on the SAW joint but bainite and martensite microstructure on the SMAW joint. After
heat treatment hardness values disclosed the dissolution of the hard phases. The joints fabricated
with SAW demonstrated high ductility with an increase in elongation of 23% and 37.5%, during
the bend test and the tensile test respectively. Higher thickness welding of dissimilar joints by
SAW is prefereed over SMAW joints, as it requires fewer weld passes and produce superior weld
quality without any defects.

1. Introduction

Components of thermal boilers, such as headers and steam piping, are made of creep-resistant Cr-Mo steels, such as T91/P91 and
P22/T22 steel, joined by dissimilar welding [1]. Specific concerns encountered in dissimilar steel welding include coefficient of
thermal expansion mismatch, carbon migration, asymmetric weld beads, and unmixed zone formation [2,3]. Temperature and
pressure parameters at various junctions in these boilers function are different. In power plant circuit, temperature and pressure varies
depending upon the design considerations. To with stand these differences different materials of varying thickness were used. Hence
dissimilar material joints between various thicknesses are inevitable. Welding procedures are required to manufacture these con­
nections [4], and welding is inevitably used to join various components in power plants. Among various types of weld joints, the
circumferential butt joint is predominantly used for pipeline systems. Generally, welding of these pipes is accomplished by several
weld passes, as the pipe’s wall thicknesses are high [5]. Arc welding of high thickness components can only be achieved by welding
multipass layers one over the other in the welding groove [6]. In multi-pass groove welds, subsequently deposited weld passes have a
considerable impact on the overall generation of through-thickness residual stresses [7]. In addition, hydro end covers are made up of
an ASTM A 106 seamless pressure pipe plate, also known as SA 106Gr. C, welded to an ASTM A335 P12 pipe, known as SA 335 P12.

* Corresponding author.
E-mail address: [email protected] (A. Ramanathan).

https://doi.org/10.1016/j.engfailanal.2023.107252
Received 22 November 2022; Received in revised form 27 March 2023; Accepted 9 April 2023
Available online 10 April 2023
1350-6307/© 2023 Elsevier Ltd. All rights reserved.
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Nomenclature

SMAW Shielded Metal Arc Welding

SAW Submerged Arc Welding
AW As-Welded
PWHT Post-Weld Heat Treatment
Cr-Mo Chromium-Molybdenum
ASTM American Society for Testing and Materials
HAZ Heat Affected Zone
ASME American Society of Mechanical Engineers
IBR Indian Boiler Regulations
HV Hardness Value
SEM Scanning Electron Microscopy
EDS Energy Dispersive x-ray Spectroscopy
MPI Magnetic Particle Inspection

SMAW is used to produce such welding of dissimilar materials. Cracks occur in SMAW due to uneven shrinkage during solidification
caused by inadequate preheating and interpass temperature [8]. The magnitude of residual stress and its distribution is greatly
influenced by preheating. Average temperature reached during welding and cooling and peak tensile residual stress and distortion are
significantly reduced by preheating. Thus, cold cracking in joints can be prevented by preheating before welding [9]. In previous
studies, a significant number of tests were undertaken on the impact of welding technique, welding factors, number of welding passes
on shrinkage, and shrinkage stress in pipes [10-12]. High residual stresses due to excessive martensite with very high preheat tem­
perature resulted in cracks. Preheating reduces cracking in metals, and the recommended range for preheat and interpass temperature
is 200–300 ◦ C [13,14]. Maintaining correct preheating and interpass temperatures in multi-pass welds has practical difficulties. Hence
failure to maintain a preheat temperature of 150 ◦ C during welding of SA 335 P12 and SA 106 Gr. C materials by SMAW often results in
crack formation. Interpass temperature or the temperature of parent metal before welding can be reduced by stopping the machine
after internal welding or changing the weld point [15]. Conventional welding processes with consumable electrodes could improve the
deposition rate, but such welds usually suffered from a loss of alloying elements, low toughness, and high level of oxygen content and
non-metallic inclusions [16,17]. The main concern in welding dissimilar Cr-Mo steels was carbon migration. Laha, K et al. [18] re­
ported that failure occurred in the soft zone of the P22 weld metal interface. Other researchers reported similar failures in multi-pass
welding [19]. Various methods like the use of alternative filler wire and functionally graded materials to overcome carbon migration
were suggested. But these methods were not feasible in industrial compounds due to production-related failures [3]. During repair
welding a large amount of heat transfer occurs when arc temperature exceeds 1500 ◦ C [20]. A large amount of heat flow in a short
duration during repair welding causes unsteady thermal stress on the weldment and leads to crack formation in the welding seam.
Further, they interact with existing cracks resulting in crack propagation which reduces the compound’s life. The major reason for
crack development in HAZ was that material in the HAZ area changed from a compressed to tensioned condition during repair welding
with crack growth being manifest in the early stages of cooling [21]. To avoid such difficulties in manual welding, this study undertook
welding of the component by a SAW machine.
Before commissioning, thermal boilers are erected and subjected to a hydro test in which the pressure of the components is raised
1.5 times the actual rating of the boiler. The pressure in the steam pipe lines are blocked by a component called hydro end covers,
which are constructed using pipes with outer diameter ranging from 219 to 762 mm and thicknesses ranging from 25 mm to 105 mm.
In addition, hydro end cover plates function as drain points or circuit closing points. These plates range in thickness from 60 to 110 mm
and are welded at the ends of the circuits inside the pipe’s inner bore diameter. These hydro end cover plates are welded in various
locations in a thermal power plant construction site. Hydro end cover plates with thicknesses ranging from 60 mm to 110 mm are
welded inside the pipes to withstand high pressure and impact loads. After the hydro test, the collected metallic waste is removed by
cutting the plates; the circuits are then connected to the respective water line and the turbine circuits. The pipes are bored inside to seat
the hydro end cover plate and welded manually by SMAW. Preheat for welding is 200 ◦ C and interpass temperature during welding is
150 ◦ C. After welding, post-heating is carried out at 200 ◦ C for 2 h. Then these pipes are heated by Liquefied Petroleum Gas (LPG)
burners for about 15–20 min to ensure preheat for welding. After attaining the required temperature, the groove is welded by manual
SMAW. Manual welding and continuous heating is not possible simultaneously due to the heat experienced by the welder. Hence
welding is carried out intermittently. Therefore heating is required every 15–20 min. Higher thickness of the plate and the pipe makes
it difficult to maintain heat in the weld region. Hence the job and weldment cool faster leading to cracks in the weld. The purpose of
this study is to eradicate the failure of weld joints during hydro testing of newly erected thermal boilers thereby reducing accidents
during commissioning and to enhance the safety.
SAW is extensively preferred in heavy industries such as piping due to the superior weld bead, to deep penetrated welds as indicated
by Kiran et al. [22], and to better joining and deposition rate as reported by Fang et al. [23]. Further, the study of welding of dissimilar
joints between SA 335 P12 and carbon steel has not been reported in the literature. Hence, the present investigation was carried out to
explore the welding of hydro end covers by the SAW method. In this work, samples were welded both by the SMAW and SAW processes
and the effect of the SAW process on macro and microstructure was analyzed. Further, hardness test, SEM, and EDS study were carried

2
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

out on the welded joints between SA 335 P12 and SA 106 Gr. C materials, and the results analyzed.

2. Experimental procedure

2.1. Material

The plate of a hydro end cover is made of SA 106Gr. C carbon steel and the pipe of SA335P12 alloy steel. Chemical composition of
the materials is tested as per ASTM E415:2021 and given in Table 1 and the chemical compositions of filler materials are given in
Table 2. The dimensions of the pipe and plate material welded for sample preparation are given in Table 3.
Before welding, the materials were ground to remove their protective coating. Two sets of samples were prepared. In the first set,
the plate was inserted into the pipe and welded manually by the SMAW method. Raw material pieces of the plate and pipe and their fit-
up for welding are shown in Fig. 1 and Fig. 2 shows the design details for welding the plate inside the pipe.
Fig. 3 shows the welding of these samples by the SMAW and SAW processes. In the second set, the plate and pipe were welded by the
SAW method. The samples were then parted by a power saw. The selected samples were subjected to post-weld heat treatment, macro
examination, micro examination, hardness test, and nondestructive testing, and the results discussed.

2.2. Welding of samples by the SMAW and SAW methods

Both SMAW and SAW processes were carried out by maintaining the preheating temperature of 150 ◦ C and the interpass tem­
perature of 200 ◦ C. Parameters used for welding by the SMAW and SAW processes are shown in Table 4.
The welded samples were then cut into two halves; one half being used for testing in the as-welded (AW) condition while the other
was tested after post-weld heat treatment (PWHT). In the PWHT cycle, the samples were heated to 650 ◦ C, the rate of heating and the
rate of cooling being maintained at 110 ◦ C per hour with a soaking time of 250 min. Welding parameters and guidelines for PWHT were
maintained in accordance with ASME Sec IX and IBR /ASME Sec IX.

2.3. Metallurgical characterization

For macro examination, a slice of about 25 mm thickness in the transverse direction of the weld joint with base metals on both sides
was removed from the mock-up coupons. Both sides of the slices were milled to ensure parallel seating. The surfaces of the samples
were polished using a polishing machine, BAINPOL-VT, Chennai Metco, India, and water was used as a coolant to avoid overheating.
The macro samples were then etched with a nital solution (1.5% nitric acid in methanol). Macrostructure was obtained through a
microscope OLYMPUS – SZ61, Japan, with magnification up to 45X. For the microstructure examination, the specimens were finely
ground using different emery grits (120, 220, 400, and 600 grit sizes) and cloth polished using alumina to ensure a mirror-like surface
without fine scratches. After polishing, the samples were etched with a nital solution (3% nitric acid in methanol). Etching action on
the polished surface revealed the metallurgical structure of the materials. The samples were examined through a metallurgical mi­
croscope (OLYMPUS –BX51M) with a magnification of 500 X. Field-emission scanning electron microscopy (FESEM, ZEISS) charac­
terized the weld metal’s microstructure. Hardness measurements were conducted on a Vickers scale with a 500 g load with dwell time
of 30 s in the hardness testing machine (AFFRI). Hardness test was carried out across the weld joints, covering plate material, HAZ on
both sides, weld metal, and pipe materials in accordance with ASTM E384.

3. Results and discussion

3.1. Non destructive testing

Welded samples were checked for surface and subsurface flaws. A magnifying glass was used for visual inspection to locate gross
surface flaws in the samples. Further, samples were subjected to Magnetic Particle Inspection (MPI) to locate surface and subsurface
flaws. MPI tests were conducted on the specimens in accordance with the ASTM Volume 3.03 Non-Destructive Testing procedure. The
results obtained from visual inspection and the MPI tests are tabulated in Table 5. From the below contents it can be seen that the
weldment produced by the SAW process was free from surface and subsurface flaws.

3.2. Macro examination

Fig. 4 shows the macrographs of the specimens welded by the SMAW and SAW processes at preheat and interpass temperatures of
150 ◦ C and 200 ◦ C, respectively, in the AW and PWHT conditions. The chemical composition and mechanical qualities of the weldment

Table 1
Chemical composition of materials in weight percentage.
Component Material C Mn P S Si Cr Mo

Pipe A 335P 12 0.05–0.15 0.30–0.6 0.025 0.025 0.5 max 0.8–1.25 0.44–0.65
Plate ASTM A 106 Gr C 0.35 max 0.29–1.06 0.035 0.035 0.1 min 0.4 max 0.15 max

3
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Table 2
Chemical composition of filler materials in weight percentage.
Element C Si Mn Mo Ni Cr P Ti S N Cu

E 8018 B3 0.05 0.8 0.9 0.9–1.2 NIL 2–2.5 0.03 NIL 0.03 NIL NIL
AWS A5.23 EG 0.07 0.25 1.65 0.003 0.02 0.03 0.009 0.115 0.005 0.003 0.01

Table 3
Dimensions of sample pipe and plate in mm.
Component Material Outer Dia (mm) Thickness (mm)

Pipe A 335P 12 323 57

Plate ASTM A 106 Gr C 207 67

Fig. 1. Photographic view of (a) plate ASTM A 106 Gr. C material, (b) pipe A 335P 12 material, and (c) fit up of plate inside a pipe for welding.

are determined by the weld geometry of the weld bead.

The weld’s superiority is dependent on a number of process variables. Welding current and voltage, which control welding heat
input, weld bead shape, and microstructural characteristics, are the most important process variables. An improperly formed weld
bead leads to several welding defects, like undercuts, lack of fusion, and cracks [24].
Macro specimens of the SMAW process revealed lack of fusion, non-uniform welding, and lack of penetration at the welded joint of
the two materials as seen in the Fig. 4a, Fig. 4b and in Fig. 4e. Lack of fusion at the root observed in the samples welded by the SMAW
process could be due to insufficient voltage. Non-uniform welding was due to manual welding which led to distortion. Further Fig. 5
revealed that 30 beads were needed from the macro specimens welded by the SMAW process, whereas only 22 beads were required
from the SAW macro specimens. This was attributed to the automated welding process as the flux and electrodes were fed auto­
matically. The joint between two metal components is filled with a filler material to form a weld bead. The width and height of the bead
varies, depending on the welding process and filler wire used. In general, the wider and higher/larger reinforcement of weld bead,
produces stronger joint. When the bead has unform width and height, without cracks or holes and without any distortion it can be
considered as quality bead. The weld bead shape and penetration are affected by the orientation angle of the welding electrode and
uneven weld bead is caused by the high arc pressure on the weld pool during SMAW process. In general, it is desired that weld beads
expected to have equal dimension as much as depth. Recommendations vary from a 1:1 to 1.4:1 width to depth ratio to avoid defects.
Accordingly weld to depth ratio of samples welded by SMAW and by SAW were calculated and tabulated below in Table 6.
From the above it can found that the average width to depth ratio of samples welded by SMAW is 1.55 whereas the average width to
depth ratio of samples welded by SAW is 1.14. Hence it can be observed that samples welded by SMAW are prone for cracking. The
higher productivity can also be attributed, in addition to the factors mentioned, by the higher heat input produced by SAW, increasing
the deposition rate of the process when compared to SMAW. Also, uniform welding insignificant distortion was observed, as shown in
Fig. 4c and Fig. 4d.
Improper fusion involes lack of complete melting and fusion of the weld metal in a joint. It is manily caused by incorrect position of
electrodes, low welding current. Undercut represents a depression into the base metal adjacent to the toe of the weld. Undercut can be
avoided by correct choice of current range and good manipulation of electrode during welding. The most often, poor welding tech­
niques used before, during, and after the welding process cause a weld to crack. Cracking is frequently brought on by improper work
piece preparation, poor part fitment, incorrect welder settings, and contaminated filler materials. It occurs at elevated temperature
during cooling and solidifying. Preheating and proper cooling rate will reduce the probabilities of cracking.

4
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 2. (a) Isometric view of the pipe and flat end cover (b) Sectional view of the pipe and flat end cover (c) Sectional view of the pipe and flat end
cover joint with weld bead (d) Dimensions of the plate and bevel provision for welding.

3.3. Micro examination

Fig. 6a and Fig. 6b show the parent metal microstructure of the SA-106 Grade C and SA 335 P12 samples. Parent metal SA-106
Grade C consists of polygonal ferrite (F) and pearlite (P) while SA 335 P12 consists of polygonal ferrite and tempered bainite (B).
The microstructure of the welded samples of the SMAW and SAW processes at different locations prior to heat treatment are shown
in Fig. 7.
The microstructure of HAZ shows a combination of coarse and fine grains of pearlite and bainite. The microstructure of the HAZ
inboth processes does not show any hard phases like martensite. The interface regime of both welding processes reveals columnar
grains of ferrite and carbides with inter-bead refinement. The interface region in the SMAW process showed segregation of more
carbon than in SAW. All welds had an acicular ferrite (AF) structure. Microstructure of the welded samples of the SMAW and SAW
processes at different locations is shown in Fig. 8. The HAZs in the welded samples of PWHT of both welding processes showed coarse
and fine grains of tempered bainite (B) and pearlite.
The weldment shows columnar grains of ferrite and carbides with inter-bead refinement. Equiaxial/polygonal fine grains of ferrite
were observed in the refined zone. Flux shielding processes like SMAW are more susceptible to hydrogen pick-up, as moisture in the
flux is the major source for hydrogen generation. Diffusion of hydrogen below 150 ◦ C affects the bainite and martensite micro­
structures. Hard or brittle microstructures like bainite and martensite (M) are highly vulnerable to hydrogen cracking. In general, these
microstructures were formed due to rapid cooling. Improper or low preheat and interpass temperatures, thick sections, low heat input

5
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 3. Photographic view of samples being welded by (a) Manual welding SMAW method and (b) SAW method.

Table 4
Parameters used for welding by SMAW and by SAW processes.
Process Filler Metal Dia (mm) Current Range (A) Range (V) Process Efficiency (%)

SMAW E 8018 B3 4 150–170 22–26 80

SAW AWS A5.23 EG 4.8 375–415 27–31 98

Table 5
Results of visual, LPI and MPI carried out on the welded samples.
Samples welded by

SMAW SAW

Visual Inspection Uneven rough surface Smooth finished surface

Cracks No cracks were found
MPI Cracks, discontinuities and lack of fusion were detected No aberrations detected

are reasons for fast cooling. The most susceptible microstructure is untempered martensite whereas acicular ferrite, austenite, ferrite,
and pearlite resist cracking [25].

3.4. Hardness test

Esfahani, M.N et al. [26] reported that microstructural homogeneity can be estimated by microhardness. Hence, hardness test was
carried out at 3 places in the pipe’s parent metal, 2 places in the pipe’s HAZ, 10 places in the weld metal, 2 places in plate HAZ, and at 2
places in plate’s parent metal as shown in Fig. 9 and Table 7. From the values in Table 7, the hardness values of all samples in the AW
condition were higher than those in the PWHT condition, indicating the effectiveness of PWHT on the samples.
However, hardness measurements on the weld joints indicated hardness values of the weld metal were higher than that of the plate
and pipe materials, revealing the strength of the weld metal [27]. Hardness measurements were made as per Test Method - ASTM E92:
2017 in AW and PWHT conditions for all the samples. The hardness values of all samples in the AW condition were higher than those in
the PWHT condition which was in line with the results of Sultan et al. [28]. Hardness values of the welded metal were uniform before
and after heat treatment in the SAW process with the standard deviation of 6.38 and 6.15 in AW and PWHT conditions, respectively as
indicated in Fig. 10. In both the AW and PWHT circumstances, the hardness values of the welded metal were lower in the SAW process
than in the SMAW process.
Similarly, Fig. 10 depicts the average hardness value comparison of welded samples by the SMAW and SAW procedures at different

6
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 4. Macrographs of the specimens welded by (a) SMAW in AW condition; (b) SMAW after PWHT; (c) SAW in AW condition; (d) SAW after PWHT
and (e) SMAW indicating defects.

locations before and after heat treatment. The SAW method yielded a hardness value of 193 HV at HAZ, while the SMAW approach
yielded a hardness value of 213.5 HV under the AW condition. The hardness value of the samples at HAZ was found to be lower when
welded using the SAW method against the SMAW method. In the PWHT condition, the hardness value at HAZ by the SAW technique

7
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 5. Representation of weld beads in samples welded by (a) SMAW and (b) SAW processes.

Table 6
Weld to Depth ratio of welded samples.
Sample Welded by SMAW Welded by SAW

Width (mm) Depth (mm) W/D ratio Width (mm) Depth (mm) W/D ratio

1 7.10 4.80 1.48 8.40 7.40 1.14

2 7.00 4.70 1.49 8.50 7.60 1.12
3 7.10 5.20 1.37 8.20 8.00 1.03
4 7.40 5.10 1.45 8.30 7.70 1.08
5 7.50 5.00 1.50 8.40 7.90 1.06
6 7.60 4.90 1.55 8.70 7.80 1.12
7 7.70 4.60 1.67 8.90 7.80 1.14
8 8.00 5.20 1.54 9.10 8.10 1.12
9 7.50 5.10 1.47 9.20 7.40 1.24
10 7.70 4.60 1.67 9.00 7.70 1.17
11 7.90 4.80 1.65 9.10 7.80 1.17
12 7.40 4.90 1.51 9.30 7.50 1.24
13 7.80 5.10 1.53 8.90 7.60 1.17
14 7.70 4.90 1.57 8.60 7.60 1.13
15 7.90 4.70 1.68 8.70 7.80 1.12
16 7.90 4.60 1.72 8.80 8.10 1.09
17 7.80 5.10 1.53 8.70 7.40 1.18
18 7.70 5.30 1.45 8.90 7.70 1.16
Average 1.55 Average 1.14

was 146.5 HV and 156.5 HV by the SMAW method.

The average hardness value of the weld metal by the SMAW process was 213 HV in AW, which was reduced to 191.5 HV after
PWHT. Similarly, the average hardness value of the weld metal by the SMAW process was 200.2 HV in AW and 161.6 HV after PWHT.
Similarly, the average hardness value of the weld metal by the SAW process was 200.2 HV in AW and 161.6 HV after PWHT. When the
hardness value of 213 HV obtained by the SMAW process was compared to the hardness value of 200.2 HV obtained by the SAW
process in the AW condition, a reduction of 12.8 HV (6%) in hardness was obtained. Similarly, when the hardness value of 191.5 HV
obtained by the SMAW process was compared to the hardness value of 161.6 HV obtained by the SAW process in the PWHT condition,
a reduction of 29.9 HV (15%) hardness was obtained. The SAW process resulted in a decrease in martensite volume fraction and a
decrease in hardness values at the weld. This conclusion is qualitatively similar to that of Zhou et al. [29]. Zhou et al. investigated the
hardness and residual stresses of high tensile strength, high-thickness steel welded using a multi-pass SAW method using the Finite
Element Method (FEM).
The findings revealed that as the martensite volume fraction increases, so do the hardness values, and as the number of beads
increases, so do the residual stresses. Xue Wang [30] used 3D-FEM to investigate the residual welding stresses of plates and pipes. The
findings revealed that weld shape constraints influence martensite transformation. As a result, in butt welding pipe, using fewer beads
is preferable to reducing residual stresses. Due to the increased peak temperature during welding, the hardness value usually increases
from parent metal to weldment. As a result of the microstructure and hardness studies, it was clear that the SAW welded joint

8
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 6. Microstructure of (a) Parent metal SA-106 Grade C and (b) SA 335 P12.

Fig. 7. Microstructure of HAZ in (a) as-welded condition by SMAW (b) as-welded condition by SAW (c) PWHT condition by SMAW and (d) PWHT
condition by SAW processes.

demonstrated improved stability.

3.5. Tensile test

Tensile testing of weld is carried out to evaluate the ductility and strength of the weld. To assess the strength of the weld joint, a

9
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 8. Microstructure of weldment in (a) as-welded condition by SMAW (b) as-welded condition by SAW (c) PWHT condition by SMAW and (d)
PWHT condition by SAW processes.

transverse weld test was conducted. Specimens welded by both SMAW and SAW process were tested and the results are tabulated in
Table 8.
Fig. 11 (a) and (b) shows the specimens on which tensile testing and bend test were conducted. Both the tensile and bend tests were
conducted on Universal Testing Machine (UTM) of make TINIUS OLSEN, with a load capacity of 50 kN and strain rate of 1 mm / min.
HORIZON software was used to evaluate the results.
The ultimate tensile strength and elongation of welded specimens of quantity 03, were tested using a universal testing machine. The
average ultimate tensile stress with the SMAW-welded specimen was 460 MPa with the standard deiviation of 2.26 as opposed to 547
MPa with the standard deiviation of 2.19 in the SMAW-welded specimen. The average elongation achieved by the SAW specimen was
19.8% with the standard deiviation of 2.43 as compared to 14.4% with the standard deiviation of 2.52 for the SMAW specimen. As a
result, it can be noted that the specimen welded using the SAW method had greater strength values.

3.6. Bend test

The bend test measures the ductility and the soundness of material. Weld quality is assessed in relation to ductility, or the ability to
withstand cracking during bending. Bending specimens made in accordance with ASTM E190-92. The test is conducted on a fixed
support, and the specimen’s midpoint is loaded (three-point bend test). Bend tests were conducted on specimens of quantity 03 welded
by both SMAW and SAW process and the results are tabulated in Table 9.
During the bend test, a SMAW-welded specimen resisted an ultimate stress of 906 MPa with the standard deviation of 2.31 as
compared to the 965 MPa with standard diviation of 2.38 sustained by a SAW-welded specimen. Similar to this, the elongation
achieved by the SAW specimen was 11% with standard diviation of 2.45 as compared to the SMAW specimen’s 8.94% with standard
diviation of 2.23. As a result, it can be shown that the specimen that was welded using the SAW process has greater ductility than the
specimen that was welded using the SMAW technique.

3.7. Scanning electron microscopy (SEM) and energy-dispersive x-ray spectroscopy (EDS) analysis

SEM images help in the quantifiable assessment of size, fraction, and distribution while EDS analyzes the key elements involved in
inclusions and their effectiveness.

10
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 9. Photographic view of the weld joint and schematic representation of the location of hardness measurements.

Table 7
Hardness values comparison of welded samples by SMAW and SAW process at different locations before and after heat treatment.
Weld location Hardness value (HV) AW Hardness value (HV)
PWHT

SMAW SAW SMAW SAW

Pipe PM 193 175 149 133

Pipe PM 205 186 151 136
Pipe PM 206 184 153 134
Pipe HAZ 208 182 145 145
Pipe HAZ 219 204 168 148
Weld Metal 202 204 189 167
Weld Metal 224 210 184 157
Weld Metal 225 200 187 157
Weld Metal 205 208 175 158
Weld Metal 209 197 193 157
Weld Metal 206 189 189 167
Weld Metal 223 198 191 172
Weld Metal 220 191 203 169
Weld Metal 213 202 205 159
Weld Metal 203 203 199 153
Plate HAZ 218 203 222 215
Plate HAZ 225 248 214 178
Plate PM 178 195 160 138
Plate PM 172 172 145 139

Fig. 12 (a), Fig. 12 (b) and Fig. 12 (c) show SEM images of the weldment obtained by the SMAW process and its corresponding
element overlay map shown in Fig. 12 (d). Correspondingly, Fig. 13 (a) and Fig. 13 (b) are SEM images of the weldment obtained by the
SAW process. Its corresponding element overlay map is shown in Fig. 13 (c). Figs. 12 and 13 reveal that cracking was noticed in Fig. 12
(c) but not found in Fig. 13. Fig. 12 (d) shows a silicon presence of 7% in the SMAW weldment without any trace of silicon in Fig. 13 (c).
Performance at the weld joint was affected by the heterogeneous microstructure formed by the reheating process [31]. Wang, Shang,
and Wang (2015) [32] noted that the correlation between toughness and the microstructure of weldments was limited and rare.
Further constituents detected in the heat-affected zone of multi-pass welding affected toughness. Fig. 12 (c) shows a crack on the

11
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 10. Average hardness values comparison of welded samples by SMAW and SAW process at different locations before and after heat treatment.

Table 8
Tensile properties of welded specimens.
Process Ultimate Stress MPa Standard deviation (%) Total Elongation (%) Standard deviation (%)

SMAW 455 14.0

460 2.26 14.4 2.43
475 14.7
SAW 535 19.3
547 2.19 19.8 2.52
559 20.3

sample welded by the SMAW process.

As acicular ferrite structure was formed by the interlocking structure, it was the most appropriate constituent for Cr Mn steel weld
metals. Crack formation was effectively controlled by small grain size as it had more boundaries and differing crystallographic ori­
entations. Microstructure refinement due to acicular ferrite effectively controlled parameters, such as inclusions and alloying elements.
In addition acicular ferrite acts as a barrier to crack initiation. Appropriate parameters, such as initial temperature and heat input for
steel beams, were required to avoid cold cracks during welding. The most probable microstructure which could influence cold cracking
was non-tempered martensite. Further, inclusions like sulfur and silicon also assist cold cracking, which occurs at grain boundaries
forming martensite [33]. An increase in Si content promotes martensite and raises hardness values. Further, Si alters the micro­
structure of the weld metal and induces cracking when Si content increased beyond 2% [34].
Fig. 12 (d) shows the presence of 7% Silicon in the EDS image, which was responsible for cracking in the SMAW method. Similarly,
Fig. 13 (c) shows the EDS image and its micro constituents where silicon was absent, and a crack-free weld was obtained by the SAW
process.

3.8. Heat input and electrical energy consumption

During the manual SMAW process, to maintain the preheat temperature of 150 ◦ C and interpass temperature of 200 ◦ C, the joint
was heated by an LPG burner on both sides. The burners were removed after the preheat temperature was maintained and manual
welding started. But rapid cooling occurred due to heavy thickness and hence welding was stopped with heating being carried out
again. This process of heating-welding-heating was repeated till the required fillet dimension was obtained. Due to the differential
expansion and shrinkage and intermittent heating, several times cracks occur [35] in the element during the SMAW process. Cracks
occur deep in the groove and these cracks are to be removed by grinding. Hence, welding of hydro end cover was carried out as per the
parameters mentioned in Table 4, and the time taken to weld hydro end cover was measured, and found that the SAW process con­
sumes 19 h less than the SMAW process. Using IBR-certified, experienced welders, numerous experiments were conducted to establish

12
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 11. Specimens tested for tensile and bend tests (a) SMAW process and (b) SAW process.

Table 9
Results of Bend Test.
Process Ultimate Stress (MPa) Standard Total Elongation (%) Standard
deviation (%) Deviation (%)

SMAW 886 8.74

906 2.31 8.94 2.23
928 9.14
SAW 942 10.72
965 2.38 11 2.45
988 11.26

the welding parameters for both processes.

Heat input and total electrical energy required to complete the welding process was computed. Table 10 indicates the current, volt
input, time taken and weld speed to complete welding in both welding processes. Heat input was calculated using the following
equation (1) and energy consumption was calculated using the equation (2)
η × V × I × 60
Heat input = kJ/mm (1)
Weld speed × 1000
Where ɳ indicates process efficiency, V indicates voltage in V, I the current in A, and weld speed in mm/min.
V × I × T × 3600
Energy Consumption = kJ (2)
1000
Where V indicates voltage in V, I the current in A, and T the time in hours. From the above table it can be seen that heat input given
during SAW welding is 0.072 kJ/mm lesser when compared with heat input of SMAW process. Similarly, above calculations also reveal
that electrical energy required to complete welding by the SAW process was 206190 kJ whereas the SMAW process required 331776
kJ, demonstrating that the SMAW process required 125586 kJ more electrical energy than the SAW process.

4. Conclusions

In this study, welding of dissimilar material joint between high thickness SA335P12 pipe and SA 106Gr. C plate using the SMAW
and SAW processes was analyzed. Visual inspection, macrograph analysis of the weld bead, metallurgical investigation, and other
characterization techniques like SEM and EDS were used to assess the weld joint’s weld integrity. Major conclusions drawn from the
work are listed below.

13
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 12. SEM analysis on the weldments welded by SMAW showing (a) voids (b) Shrinkage cavity (c) Crack and (d) Element overlay and EDS
sum spectrum.

• High thickness SA335P12 pipe and SA 106Gr. C plate successfully welded using SAW process was free from cracks, porosity,
shrinkage cavities, and voids.
• In the SAW-welded specimen, an acicular ferrite microstructure was observed. Hard phases such as bainite and martensite were
absent, resulting in enhanced toughness. The presence of more than 2% silicon in the SMAW process, on the other hand, induced
cracking.
• Compared to the SMAW process, the SAW approach yielded an overall hardness reduction of 6% in AW and 15% in PWHT
conditions.
• The number of beads required to finish a weld in the SAW process was decreased from 30 to 22. A lower number of beads minimizes
residual stresses in joints.

14
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Fig. 13. (a) and (b) SEM analysis on the weldments welded by SAW and (c) Element overlay EDS sum spectrum.

Table 10
Welding parameters for both SMAW and SAW process.
Process Current Volt (V) Welding Time Weld Speed Heat input (kJ/mm) Electrical Energy Consumed (kJ)
(A) (Hours) (mm / min)

SMAW 160 24 24 125 1.475 331,776

SAW 395 29 5 480 1.403 206,190

• The joints produced with SAW demonstrated an increase in elongation of 23% and 37.5%, respectively, during the bend test and the
tensile test.
• Compared to the SMAW method, the SAW approach used 37.8% less electrical energy, finished in 20.8% less time, and required
reduced heat input.

When the two methods were compared, the SAW method was superior to the SMAW method for welding of dissimilar higher
thickness component, resulting in reduced costs, higher output, and improved safety.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

15
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

Data availability

Data will be made available on request.

Acknowledgments

The authors thank BHEL, Tiruchirappalli, and the Director, National Institute of Technology, Tiruchirappalli, India for extending
the facilities to carry out the research work.

References

[1] D. Jandová, J. Kasi, V. Kanta, Creep resistance of similar and dissimilar weld joints of P91 steel, in: Mater. High Temp., Science Reviews Ltd, 2006, pp. 165–170.
10.1179/mht.2006.010.
[2] R.S. Vidyarthy, A. Kulkarni, D.K. Dwivedi, Study of microstructure and mechanical property relationships of A-TIG welded P91–316L dissimilar steel joint,
Mater. Sci. Eng. A. 695 (2017) 249–257, https://doi.org/10.1016/j.msea.2017.04.038.
[3] A. Kulkarni, D.K. Dwivedi, M. Vasudevan, Study of mechanism, microstructure and mechanical properties of activated flux TIG welded P91 Steel-P22 steel
dissimilar metal joint, Mater. Sci. Eng. A. 731 (2018) 309–323, https://doi.org/10.1016/j.msea.2018.06.054.
[4] S. Mahajan, R. Chhibber, Experimental investigations on P22/P91 dissimilar shielded metal arc welds for power plant applications, Proc. Inst. Mech. Eng. Part L
J. Mater. Des. Appl. 234 (2020) 1313–1324, https://doi.org/10.1177/1464420720939115.
[5] J. Ni, X. Wang, J. Gong, M. Abdel Wahab, Thermal, metallurgical and mechanical analysis of circumferentially multi-pass welded P92 steel pipes, Int. J. Press.
Vessel. Pip. 165 (2018) 164–175, https://doi.org/10.1016/j.ijpvp.2018.06.009.
[6] Y.L. Sun, G. Obasi, C.J. Hamelin, A.N. Vasileiou, T.F. Flint, J.A. Francis, M.C. Smith, Characterisation and modelling of tempering during multi-pass welding,
J. Mater. Process. Technol. 270 (2019) 118–131, https://doi.org/10.1016/j.jmatprotec.2019.02.015.
[7] P.K. Taraphdar, M.M. Mahapatra, A.K. Pradhan, P.K. Singh, K. Sharma, S. Kumar, Evaluation of through-thickness residual stresses in conventional and narrow
grooved stainless steel welds, Proc. Inst. Mech. Eng. Part L J. Mater. Des. Appl. 234 (2020) 1165–1179, https://doi.org/10.1177/1464420720930355.
[8] J.R. K.S, D.B. Nagesh, I.N.N. Kumar, Application of flux cored arc welding (fcaw) without pre-heating for heavy structures in ship building fabrication &
construction of heavy offshore structural steels, J. Constr. Eng. Technol. Manag. 10 (2020) 20–28. http://engineeringjournals.stmjournals.in/index.php/
JoCETM/article/view/3683 (accessed June 26, 2021).
[9] M. Zubairuddin, S.K. Albert, M. Vasudevan, S. Mahadevan, V. Chaudhri, V.K. Suri, Thermomechanical analysis of preheat effect on grade P91 steel during GTA
welding, Mater. Manuf. Process. 31 (2016) 366–371, https://doi.org/10.1080/10426914.2015.1025964.
[10] M. Taghipour, et al., Root cause analysis of a failure in a flange-pipe welded joint in a steam line in an ammonia plant: Experimental investigation and
simulation assessment, Eng. Fail. Anal. 129 (2021), 105730.
[11] D. Deng, H. Murakawa, W. Liang, Numerical and experimental investigations on welding residual stress in multi-pass butt-welded austenitic stainless steel pipe,
Comput. Mater. Sci. 42 (2008) 234–244, https://doi.org/10.1016/j.commatsci.2007.07.009.
[12] C. Pandey, H.K. Narang, N. Saini, M.M. Mahapatra, P. Kumar, Microstructure and transverse shrinkage stress analysis in GTA welds of P91 steel pipe, Int. J. Steel
Struct. 17 (2017) 763–774, https://doi.org/10.1007/s13296-017-6030-8.
[13] M. Sireesha, S.K. Albert, S. Sundaresan, Microstructure and mechanical properties of weld fusion zones in modified 9Cr-1Mo steel, J. Mater. Eng. Perform. 10
(2001) 320–330, https://doi.org/10.1361/105994901770345033.
[14] C. Pandey, M.M. Mahapatra, P. Kumar, N. Saini, Some studies on P91 steel and their weldments, J. Alloys Compd. 743 (2018) 332–364, https://doi.org/
10.1016/j.jallcom.2018.01.120.
[15] K. Sirin, S.Y. Sirin, E. Kaluc, Influence of the interpass temperature on t8/5 and the mechanical properties of submerged arc welded pipe, J. Mater. Process.
Technol. 238 (2016) 152–159, https://doi.org/10.1016/j.jmatprotec.2016.07.008.
[16] M. Vasudevan, V. Arunkumar, N. Chandrasekhar, V. Maduraimuthu, Genetic algorithm for optimisation of A-TIG welding process for modified 9Cr-1Mo steel,
Sci. Technol. Weld. Join. 15 (2010) 117–123, https://doi.org/10.1179/136217109X12577814486773.
[17] S. Krishnan, D.V. Kulkarni, A. De, Multipass pulsed current gas metal arc welding of P91 steel, Sci. Technol. Weld. Join. 21 (2016) 171–177, https://doi.org/
10.1179/1362171815Y.0000000080.
[18] K. Laha, S. Latha, K. Bhanu Sankara Rao, S.L. Mannan, D.H. Sastry, Comparison of creep behaviour of 2⋅25Cr - 1Mo/ 9Cr - 1Mo dissimilar weld joint with its base
and weld metals, Mater. Sci. Technol. 17 (2001) 1265–1272. 10.1179/026708301101509188.
[19] R.K. Shiue, The study of carbon migration in dissimilar welding of the modified 9Cr-1Mo steel, Artic. J. Mater. Sci. Lett. (2001), https://doi.org/10.1023/A:
1011616232396.
[20] M. Weglowski, Determination of GTA and GMA welding arc temperatures, Weld. Int. 19 (2005) 186–192, https://doi.org/10.1533/wint.2005.3398.
[21] K. Yang, Y. Zhang, J. Zhao, Elastoplastic fracture analysis of the P91 steel welded joint under repair welding thermal shock based on XFEM, Metals (Basel). 10
(2020) 1–26, https://doi.org/10.3390/met10101285.
[22] D.V. Kiran, B. Basu, A. De, Influence of process variables on weld bead quality in two wire tandem submerged arc welding of HSLA steel, J. Mater. Process.
Technol. 212 (2012) 2041–2050, https://doi.org/10.1016/j.jmatprotec.2012.05.008.
[23] C. Fang, Y. Chen, Z. Yang, J. Wang, M. Wu, K. Qi, Cable-type welding wire submerged arc surfacing, J. Mater. Process. Technol. 249 (2017) 25–31, https://doi.
org/10.1016/j.jmatprotec.2017.05.020.
[24] D. Pathak, R.P. Singh, S. Gaur, V. Balu, Influence of input process parameters on weld bead width of shielded metal arc welded joints for AISI 1010 plates, in:
Mater. Today Proc., Elsevier Ltd, 2020: pp. 24–28. 10.1016/j.matpr.2020.05.516.
[25] K. Wu, Y. Yan, R. Cao, X. Li, Y. Jiang, F. Yang, X. Jia, J. Chen, Microstructure and charpy impact toughness of a 2.25Cr-1Mo-0.25V steelweld metal, Materials
(Basel) 13 (2020) 3013, https://doi.org/10.3390/ma13133013.
[26] M.N. Esfahani, J. Coupland, S. Marimuthu, Microstructure and mechanical properties of a laser welded low carbon-stainless steel joint, J. Mater. Process.
Technol. 214 (2014) 2941–2948, https://doi.org/10.1016/j.jmatprotec.2014.07.001.
[27] M.L. Zhu, F.Z. Xuan, Correlation between microstructure, hardness and strength in HAZ of dissimilar welds of rotor steels, Mater. Sci. Eng. A. 527 (2010)
4035–4042, https://doi.org/10.1016/j.msea.2010.03.066.
[28] A.R. Sultan, R. Ravibharath, R. Narayanasamy, Study of Dissimilar Header Welding Between 2.25Cr–1Mo Steel and 9Cr–1Mo Steel with 9018 B9 Electrode
Under Various Conditions of Post Weld Heat Treatment, Trans. Indian Inst. Met. 70 (2017) 2079–2092, https://doi.org/10.1007/s12666-016-1029-y.
[29] H. Zhou, Q. Zhang, B. Yi, J. Wang, Hardness prediction based on microstructure evolution and residual stress evaluation during high tensile thick plate butt
welding, Int. J. Nav. Archit. Ocean Eng. 12 (2020) 146–156, https://doi.org/10.1016/J.IJNAOE.2019.09.004.
[30] X. Wang, L. Hu, Q. Xu, D. Chen, S. Sun, Influence of martensitic transformation on welding residual stress in plates and pipes, Http://Dx.Doi.Org/10.1080/
13621718.2016.1263711. 22 (2016) 505–511, https://doi.org/10.1080/13621718.2016.1263711.
[31] J.C.F. Jorge, L.F.G. De Souza, M.C. Mendes, I.S. Bott, L.S. Arau, V.R. Santos, J.M.A. Rebello, G.M. Evans, Microstructure characterization and its relationship
with impact toughness of C e Mn and high strength low alloy steel weld metals e a review, (2020).
[32] X. Wang, C. Shang, X. Wang, Characterization of the multi-pass weld metal and the effect of post-weld heat treatment on its microstructure and toughness, in:
7th Int. Conf. High Strength Low Alloy Steels, HSLA Steels 2015, Int. Conf. Microalloying 2015, Microalloying 2015 Int. Conf. Offshore Eng. Steels 2015, OES
2015, John Wiley and Sons Inc., 2015: pp. 481–488. 10.1007/978-3-319-48767-0_57.

16
N. Prasanna et al. Engineering Failure Analysis 149 (2023) 107252

[33] A.R.V. Nunes, A. Zeemann, L.H. De Almeida, The contribution of impurities to unexpected cold cracks in a thick C-Mn steel plate, J. Mater. Res. Technol. 8
(2019) 4364–4373, https://doi.org/10.1016/j.jmrt.2019.07.047.
[34] R. Cao, Z.S. Chan, J.J. Yuan, C.Y. Han, Z.G. Xiao, X.B. Zhang, Y.J. Yan, J.H. Chen, The effects of Silicon and Copper on microstructures, tensile and Charpy
properties of weld metals by refined X120 wire, Mater. Sci. Eng. A. 718 (2018) 350–362, https://doi.org/10.1016/j.msea.2018.01.080.
[35] S. Yi, et al., Analysis of crack causes and effects of the A333 low carbon pipeline steel after thermite welding, Eng. Fail. Anal. 130 (2021), 105774.

17