Journal of Materials Processing Technology 238 (2016) 373382

Contents lists available at ScienceDirect

Journal of Materials Processing Technology

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

Heat input and metal transfer inuences on the weld geometry and
microstructure during underwater wet FCAW
Yong Zhang, Chuanbao Jia , Bo Zhao, Jiakun Hu, Chuansong Wu
Key Laboratory for Liquid-Solid Structural Evolution and Processing of Materials Ministry of Education, Shandong University, Jinan 250061, PR China

a r t i c l e

i n f o

Article history:
Received 14 March 2016
Received in revised form 27 June 2016
Accepted 13 July 2016
Available online 15 July 2016
Keywords:
Underwater wet welding
FCAW
Heat input
Weld geometry
Microstructure
Metal transfer

a b s t r a c t
FCAW experiments were conducted under water and in the air. Inuences of surrounding water on the
weld geometry and weld metal microstructure were investigated through analyzing the heat input and
metal transfer. The underwater welds have a lower proportion of acicular ferrite and pro-eutectoid ferrite.
Two metal transfer modes in wet FCAW were observed, i.e. repelled globular transfer and surface tension
based quasi-short-circuit transfer. The former one is the main cause of asymmetric and less uniform weld
appearance.
2016 Elsevier B.V. All rights reserved.

1. Introduction
Characterized by low cost, good adaptability, simple equipment,
the underwater wet welding has been widely applied. Particularly,
the underwater wet ux-cored arc welding (FCAW) has been developed for automatic and semi-automatic welding processes in deep
water. It has a great potential to be applied to repair or even to
construct oceanographic structures.
Liu et al. (1993) mentioned that the harsh underwater environment has signicant effect on the welding process, which
dramatically alters the cooling rate during welding and further
affects the nature of the weld-metal phase transformation. The
behavior of the welding arc, metal transfer and weld pool is
seriously inuenced by the special environment which results in
completely different metallurgical process with weld geometry and
micro-structure changes. Consequently the weld joint quality and
mechanical properties might be lowered.
Tsai and Masubuchi (1979) investigated the bubbles dynamics,
arc heat distribution, and heat input mechanism. They concluded
that a fast cooling of the weld zone during underwater welding
was the result of surface heat losses in weld area behind the arc.
Liu et al. (1993) found that the much lower cooling time t8/5 (from
800 C to 500 C) during underwater wet shielded metal arc welding

Corresponding author.
E-mail address: jiachuanbao@sdu.edu.cn (C. Jia).
http://dx.doi.org/10.1016/j.jmatprotec.2016.07.024
0924-0136/ 2016 Elsevier B.V. All rights reserved.

(SMAW) resulted in signicant amount of heat affected zone (HAZ)

martensite in nearly all low carbon steels. Wang and Yang (1997)
studied the underwater welding arc plasma by means of spectral
diagnostics. They concluded that the temperature of underwater
arc plasma was lower than in air and the difference was bigger
for deeper water. Shi and Zheng (2013) studied the inuence of
welding parameters on weld bead geometry in wet FCAW. They
concluded that comparing with travel speed and arc voltage, welding current was less sensitive to the geometries of the bead. Jia et al.
(2013) observed that the width of the underwater weld was about
two-thirds in air, presuming that the arc plasma was compressed
by the water environment even in shallow water.
Gao et al. (2015b) observed more pro-eutectoid ferrite in the
weld metal during wet shielded metal arc welding. They assumed
that this phenomenon is due to the loss of alloying elements under
water deeper than 11 m. Perez et al. (2003) presented evidence of
toughness improvements on wet welds by adding nickel to the
coating of rutile type electrodes. Al-Abbas et al. (2011) revealed
that the wet weld columnar grains were ner than those found in
the dry welds at low welding current but with different columnar
grain morphology, whereas were coarser at high welding current
with similar columnar grain morphology. Guo et al. (2015b) concluded that the addition of Ni can be helpful for suppressing the
formation of the coarse strip pro-eutectoid ferrite (PF) in the columnar grain zone of the weld metal. Di et al. (2015) simulated the
local dry underwater welding process. They concluded that as the
cooling rate increased, the volume fraction of proeutectoid ferrite

374

Y. Zhang et al. / Journal of Materials Processing Technology 238 (2016) 373382

and ferrite side plate decreased, acicular ferrite increased accompanied with rened grain. Their ndings suggested that rapid cooling
rate can improve the impact toughness and tensile strength of
weld metal in local dry underwater welding. Gao et al. (2015a)
obtained martensitic dominated coarse grain heat affected zone in
wet shielded metal-arc welding. They improved the fatigue property by the combination of grinding and underwater ultrasonic
impact treatment. Garasic et al. (2009) carried out research to
determine optimum parameters that minimize the generation of
cold cracks in wet SMAW and self-shielding FCAW at three levels of
heat input. They concluded that FCAW provided better conditions
for reducing the risk of cold cracks occurrence.
Zhang et al. (2004) observed that direct wet underwater laser
welding produced a kind of plasma with strong ultraviolet emission and obtained smaller penetration when the water depth is
less than 3 millimeters. Cui et al. (2014) conducted friction taper
plug welding experiments in wet conditions. They found a large
volume of lath bainite, small amounts of acicular ferrite, polygonal ferrite, martensite in HAZ and the lower region of the weld
metal. Meanwhile, they observed mainly lath martensite in the
upper weld metal region.Yin et al. (2015) investigated the material
ow inuence on the weld formation and mechanical performance
in underwater friction taper plug welds. They indicated that the
microstructure of both weld zone and HAZ mainly consists of lath
bainite which results in an overmatching condition (i.e. specimen
breaks in the base material far away from the plug area and HAZ
during tensile testing) through the weld.
However, only few of the studies mentioned above analyzed
the physics of FCAW welding process in underwater environment,
especially metal transfer and arc behaviors. This partly owing to the
difculty to acquire visual images as evidence. Guo et al. (2015a)
acquired clear images of droplets transfer via X-ray photographing. They (Guo et al., 2016) also analyzed the effect of boric acid
concentration on the arc stability but the bubbles and welding
arc could hardly be recognized from the images they provided.
Actually, the generation and burst of bubbles are signicant factors to determine the welding arc stability and metal transfer. Jia
et al. (2014) obtained clear images of droplets and welding arc
during underwater FCAW. They found a typical repelled globular
droplet transfer mode and a wandering compressed welding arc
behavior.
Thus, an imaging system was developed to acquire the images
of bubbles, arc and droplets at the same time (Fig. 1). This paper is

Table 1
Chemical composition of Q235 carbon steel mass%.
C

Mn

Si

Fe

0.17

0.350.65

0.28

0.038

0.038

Balance

Parameters

Air 1

Water 1

Air 2

Water 2

Arc voltage (V)

Wire feed speed (cm/s)
Travel speed (mm/min)
Surrounding

30
6.32
240
Air

30
6.32
240
Water

31
6.32
240
Air

31
6.32
240
Water

Table 2
Experimental parameters.

aimed at exploring the inuence of surrounding water on the weld

geometry and weld metal microstructure of wet FCAW through
analyzing the heat input and metal transfer.
2. Experimental method
Experiments were conducted under water (fresh water, 0.4 m
depth) and in the air with the same parameters. The welding voltage
was set constant and the current varied according to the arc length.
Four sets of bead-on-plate experiments were conducted in direct
current electrode positive (DCEP) mode. The travel speed and the
wire feed speed were maintained xed. The base material is Q235
mild steel with the dimensions of 300 mm 100 mm 8 mm. Its
composition is listed in Table 1.
The imaging system (Fig. 1) was developed based on a high
speed camera (Optronis CamRecord 5000 2) with frequency
2000 fps and resolution 512 512. Visual and electrical signals
were collected simultaneously. During welding, the torch was kept
stationary, and the test piece was controlled to move uniformly
and straightly. Through the transparent glasses a high speed camera could sense the welding process perpendicularly to the welding
direction. The welding power source was LET 500. A rutile type uxcored wire made by E. O. Paton Electric Welding Institute specially
for underwater wet welding was used. Other experimental parameters were shown in Table 2. Air 1 and Water 1 were set as group
A, while Air 2 and Water 2 as group B.
The cross-sections of weld beads were submitted to standard
mechanical grinding, polishing and corrosion procedures. Bead

Fig. 1. Schematic of underwater wet ux-cored arc welding imaging system.

Y. Zhang et al. / Journal of Materials Processing Technology 238 (2016) 373382

375

Fig. 2. Appearance of the weld bead.

Fig. 3. The uctuation of bead penetration.

cross-section microstructure examination was carried out using

an optical microscope XJP-6A. Micro-hardness values were measured using a Vickers microhardness tester (DHV-1000) with a load

of 2 N and loading time of 10 s. Microhardness tests were conducted in longitudinal direction i.e. parallel to the base material
surface.

376

Y. Zhang et al. / Journal of Materials Processing Technology 238 (2016) 373382

Table 3
Bead geometry parameters.

A
B

Air 1
Water 1
Air 2
Water 2

Table 4
The effective power of the arc and heat input.
W (mm)

P (mm)

R (mm)

13.8
11.8
15.8
10.2

2.1
2.7
1.8
2.6

3.5
3.9
3.2
4.2

3. Results and discussion

3.1. Weld geometry characteristics
The weld beads deposited in air have serious slag defects and the
bead widths (W) seem larger (Fig. 2). The bead widths of the weld
deposited under water are smaller and less slag defects are found
on the surface. However, the bead reinforcement (R) of underwater
welds is not uniform. For Water 1, there are many irregular humps
and the middle part is almost broken.
Jia et al. (2013) observed a remarkable uctuation of the current
during underwater FCAW. It is assumed that the bead penetration
(P) might uctuate accordingly so the specimens were cut longitudinally along the seam midline to evaluate the uctuation. The
level of base metal is marked by dotted lines (Fig. 3). The penetration prole as well as the locations of the maximum and minimum
penetration are marked by solid lines.
Comparing the images, the minimum penetration of Air 1 and
Water 1 are about 22% and 18% smaller than its maximum value,
while for Air 2 and water 2 the gap are 17% and 15%. It can be
concluded that the uctuation of penetration is unrelated to the
uctuation of the current (i.e. the uctuation of heat input). Averaged data are used in Table 3.
As one can see on Table 3, the underwater weld widths in group
A are about 14% smaller than those in air. While the penetration
and reinforcement of underwater welds are 20% and 14% bigger
than air welds. For group B the weld widths of underwater welds

A
B

Air 1
Water 1
Air 2
Water 2

Effective power (w)

Heat input (J/mm)

5265
4907
5407
4915

1052
980
1080
982

are 35% smaller than air welds. The penetration and reinforcement
differences are 44% and 31%.
The compression and asymmetry of underwater weld beads can
be obviously observed (Fig. 4). Pore defects are discovered both at
the bottom and the top part of the weld metal. The shapes of two
underwater welds HAZ are like a crescent, the widths of which
increase with penetration. This phenomenon can well reect the
much different cooling process in the two directions (i.e. through
the substrate thickness and across the weld seam width).
First, it implies that the arc burning atmosphere (bubbles) and
the water environment cause rapid cooling on the substrate surface.
Second, although the high temperature gradient exist between the
weld pool and the backside surface of the base metal, the thick plate
provides bigger volume for the thermal conduction in the direction
of thickness.
3.2. Microstructure analysis
Grains appear with bigger sizes from the base metal to the
heat affected zone (Fig. 5a). The microstructure of weld zone
(Fig. 5b) is uniform. The pro-eutectoid ferrite (PF) distributes on
the boundary of the columnar grain and ingrown carbon-free bainite grows in parallel to the prior austenite grain boundary. The
other microstructures within the grain are acicular ferrite (AF) plus
a little granular bainite.
Grains of overheated zone are coarse with block or strip PF scattered on the boundary of the coarse prior austenite grain (Fig. 5).
Plenty of side plate ferrite (SPF), a little granular bainite plus mas-

Fig. 4. The cross-sections of the weld bead.

Y. Zhang et al. / Journal of Materials Processing Technology 238 (2016) 373382

377

Fig. 5. Microstructure of air welds (a shows the whole weld bead; b shows the weld zone; c and d show the HAZ).

Fig. 6. Microstructure of underwater welds (a shows the whole weld bead; b and c show the weld zone and d shows the HAZ).

sive pearlite constitute typical widmannstatten structure. Outside

the overheated zone, the structures change little except that the
grain size is much smaller and the proportion of PF increases a
lot. For air specimens, lower cooling speed provides the grains
sufcient growing time to go through the high temperature transformation region of austenite. Hence, the diffusion of carbon will

become favorable and eventually facilitates the formation of coarse

proeutectoid ferrite (Zhang et al., 2011).
Lath martensite, granular bainite plus a little PF constitute the
microstructure of underwater specimens HAZ (Fig. 6d). The grain
size and the width of underwater welds HAZ are smaller. AF, granular bainite, side plate ferrite and a large amount of PF constitute the
microstructure of underwater weld metal. The AF intends to clus-

378

Y. Zhang et al. / Journal of Materials Processing Technology 238 (2016) 373382

higher carbon content, the underwater weld metal has lower hardness than air welds. Similarly a lower proportion of PF which has
lower carbon content leads to higher microhardness in HAZ.
Different welding thermal processes caused by different surroundings might be responsible for the differences of weld
geometry and microstructure between air and underwater specimens. The effective power and the heat input are calculated to
analyze the effect of thermal processes. The equations are given
below.
1
P=
T

T
u(t) i (t) dt

(1),

PT 
P
Q =
=
S
V
Fig. 7. Microhardness test result.

ter together and its proportion is much lower than in air specimens
while the PF in HAZ is also lower.
Different cooling rate could explain the microstructure differences. For underwater specimens HAZ, after the nucleation of PF
at the boundary of the austenite, the rapid cooling rate caused by
the water environment does not provide sufcient time for PF to
grow up. The amount of PF decreases a lot. As for lath martensite, lower alloying content of base metal (i.e. Q235) plus the rapid
cooling rate makes it difcult for undercooled austenite to retain
stable, therefore most of them will transform to lath martensite.
The microhardness value of base material is about 150 HV (Fig. 7).
At the 3 mm test line, the hardness have a sharp increase for both
specimens. It is concluded that due to a lower proportion of AF with

(2),

where, P means the effective power, T means time, Q means heat

input, S means distance, V means travel speed,  = 0.8.
The real heat input listed in Table 4 during underwater welding
is about 8% less than in air. Wang and Yang (1997) concluded that
the arc temperature of underwater welding was about 4% lower
than the welding arc in air and the difference was bigger for deeper
water. The wire feeding speeds were set constant for both welding
in the air and under water, and the energy required to melt the wire
is assumed to be the same. Although Tsai and Masubuchi (1979)
concluded that the energy lost to the water environment from the
plasma column only accounted for 4% of the total arc power. Liu
et al. (2007) thought that thermoconductivity was very small, the
heat conduction of water could be ignored. The energy lost to the
environment from the plasma column still can be concluded bigger
than in air. Due to its lower total heat input and bigger thermoconductivity, the temperature of molten pool during underwater
welding will be lower.

Fig. 8. The waveforms of electrical signals.

Y. Zhang et al. / Journal of Materials Processing Technology 238 (2016) 373382

379

Fig. 9. Metal transfer process in air FCAW.

Fig. 10. A repelled globular transfer process in underwater FCAW.

Much fast cooling rate during underwater welding will lead to

a compressed welding arc and a high temperature gradient in the
molten pool. A compressed arc will have a stronger stirring effect
on the molten pool and a high temperature gradient accelerates the
motion of the molten pool. These can explain less slag on the bead
deposited under water although the temperature of its molten pool
is lower. Besides, much fast cooling rate interrupt the grain growth
and lead much smaller grain size than those in the air. Due to the
less uniform cooling rate and lower temperature, the microstruc-

ture of the weld zone is less uniform and the HAZ contains less
PF(Fig. 8).
3.3. Metal transfer process
As being well-know, the metal transfer modes have signicant
effect on the formation of weld bead. Droplets transfer modes were
investigated to reveal the water surroundings effect on droplets
transfer process. Though, two sets of experiments (Groups A and

380

Y. Zhang et al. / Journal of Materials Processing Technology 238 (2016) 373382

Fig. 11. Electrical signals under a repelled globular transfer mode.

Fig. 12. A quasi-short-circuit transfer process in underwater FCAW.

Fig. 13. Electrical signals under a quasi-short-circuit transfer mode.

B) were conducted with different parameters, there are no signicant changes in metal transfer modes for both underwater and air
welding conditions. The group A was elected to typically discuss
the metal transfer differences between air and underwater FCAW.

Fig. 9 shows a typical cycle of a droplets formation and detachment from the wire tip during FCAW in air. A newly formed droplet
repelled from the arc center by the upward forces and some spatters are clearly observed ying off the welding area at 1.104 s. The
droplet is not of a regular spherical form due to a complex action

Y. Zhang et al. / Journal of Materials Processing Technology 238 (2016) 373382

of forces the droplet is subjected to. Its volume is equivalent to a

globe with 0.85 times diameter of the wire (i.e. about 1.36 mm).
After 21 ms, the droplet grows to its peak volume with a diameter
of 2.19 mm (i.e. about 1.37 times of the wire diameter). The droplet
detaches from the wire tip (Fig. 9c) and transfers to the weld pool
along the axis of the wire. Meanwhile, the welding arc immediately
ignites between the wire and the detached droplet which already
become a part of the weld pool.
For this typical small droplet transfer in air, the frequency uctuates between 20HZ and 30HZ (i.e. each period of metal transfer
lasts about 3340 ms).
Two metal transfer modes of underwater welding
(Figs. 10 and 11) are identied with the frequency from 2.6 Hz to
3 Hz. The droplets diameters of typical repelled globular transfer
process (Fig. 10) range from 4 mm to 7 mm. This mode accounts
for about 70% of the whole metal transfer process. The trail of
droplets towards the molten pool are random what might worsen
the stability of the underwater wet welding, and intend to form an
asymmetric and less uniform weld seam appearance. However, the
bubbles generation and bursting during this process is relatively
stable which is advantageous to the stability of the welding.
The second transfer mode (Fig. 12) is dened as surface tension based quasi-short-circuit transfer mode. From 2.0 s to 2.31 s,
the droplet is repelled by the repulsive forces and grows at the
wire tip. By now, this process is similar with the globular repelled
mode and the bubbles generation and burst are uniform. After
0.03 s, the droplet contacts the molten pool and is merged with
it (Fig. 12f) which seem like short circuit transfer mode. However,
the waveform of voltage during this process (Fig. 13) doesnt show
the characteristics of short circuit transfer (the voltage is almost
constant).
Wang and Song (2012) mentioned that droplets transferred during FCAW in air were constituted by slag and molten metal. So
it is assumed that the droplets of underwater wet FCAW are also
mixed. Before such a hybrid droplet contacts the molten pool, the
arc has extinguished and the current drops which will decrease the
decomposition of core ingredients and the evaporation of water.
As a result, the bubbles shrink to some extent and keep at a rather
stable state during the quasi-short-circuit period (Fig. 12f). As soon
as the droplet contacts the molten pool, the circuit is shorted. This
immediately causes a current increment and an accelerated decomposition of core ingredients. The bubbles generate explosively with
the maximum diameter up to 16 mm and the droplet nishes its
transfer at the same time by the effect of surface tension and short
circuit. After that the current decreases and the bubbles last for
a short time and then break into a succession of small bubbles
(Fig. 12i).
The droplets transferred during underwater welding are much
bigger than in the air. It is believed that different forces the droplets
subjected to are the main cause of such big droplets. Guo et al.
(2015a) mentioned that the periodically upward motion of bubbles
can inevitably apply a kind of gas ow drag force (FL ) on the droplets
in the direction of bubbles oating. According to (Liao (1999)), FL is
represented
FL = 3DU

(3),

where,  is the viscosity coefcient of gas in bubble, D, the diameter

of droplet, U, the ow velocity.
The resistance for metal transfer is increased obviously because
of the effect of FL and leads to big droplets whose max diameter is
up to 6 mm. Comparing the electrical signals of two transfer modes,
the current of repelled globular transfer mode and its bubbles
behavior (generation and burst) are more stable which is advantageous to the stability of underwater welding process. However,
unlike the quasi-short-circuit transfer mode, the trail of globular
repelled modes hybrid droplet towards the molten pool is more

381

random and it occupies about 70% of the total transfer process. It is

believed to be the main cause of the asymmetric and less uniform
weld seam appearance. Generally speaking, each transfer mode has
advantages and disadvantages. How to achieve the optimum proportion of two transfer modes to improve the stability of welding
and the formation of bead deserve further investigation.
4. Conclusions
FCAW experiments were conducted under water and in the air
with the same parameters. The effect of surrounding water on
the weld geometry and weld metal microstructure of wet FCAW
has been investigated through analyzing the heat input and metal
transfer. Main conclusions are summarized as follows:
1 An asymmetric and less uniform weld appearance with varied
penetration and reinforcement is found for underwater beads.
2 The compressed arc and high temperature gradient accelerate
the ow of the molten pool, resulting in less slag on the bead
deposited under water.
3 Compared with the air welds, the underwater welds have less AF
in the weld zone and less PF in the HAZ due to the rapid cooling
effect and lower temperature.
4 The underwater wet FCAW has much bigger molten droplet size
and thus lower metal transfer frequency than the air FCAW.
5 Two metal transfer modes are observed, i.e. repelled globular
transfer (occupied 70%) and surface tension based quasi-shortcircuit transfer. The former one is the main cause of asymmetric
and less uniform weld appearance.
Acknowledgements
The authors acknowledge the nancial support from the Natural
Science Foundation of China (Grant No. 51105237) and the Fundamental Research Funds of Shandong University (No. 2015TB002).
References
Al-Abbas, F.M., Al-Ghamdi, T.A., Liu, S., 2011. Comparison of solidication behavior
between underwater wet welding and dry welding. Proceedings of the
International Conference on Offshore Mechanics and Arctic Engineering
OMAE, 285296.
Cui, L., Yang, X., Wang, D., Hou, X., Cao, J., Xu, W., 2014. Friction taper plug welding
for S355 steel in underwater wet conditions: welding performance,
microstructures and mechanical properties. Mater. Sci. Eng. A 611, 1528.
Di, X., Ji, S., Cheng, F., Wang, D., Cao, J., 2015. Effect of cooling rate on
microstructure, inclusions and mechanical properties of weld metal in
simulated local dry underwater welding. Mater. Des. 88, 505513.
Gao, W., Wang, D., Cheng, F., Deng, C., Liu, Y., Xu, W., 2015a. Enhancement of the
fatigue strength of underwater wet welds by grinding and ultrasonic impact
treatment. J. Mater. Process. Technol. 223, 305312.
Gao, W.B., Wang, D.P., Cheng, F.J., Deng, C.Y., Xu, W., 2015b. Underwater wet
welding for HSLA steels: chemical composition, defects, microstructures, and
mechanical properties. Acta Metall. Sinica 28, 112.
I., Kralj, S., Kozuh, Z., 2009. Investigation into cold cracking in underwater
Garasic,
wet welding of API 5L X70 steel. Trans. Famena, 2534.
Guo, N., Guo, W., Du, Y., Fu, Y.L., Feng, J., 2015a. Effect of boric acid on metal
transfer mode of underwater ux-cored wire wet welding. J. Mater. Process.
Technol. 223, 124128.
Guo, N., Liu, D., Guo, W., Li, H., Feng, J., 2015b. Effect of Ni on microstructure and
mechanical properties of underwater wet welding joint. Mater. Des. 77, 2531.
Guo, N., Xu, C., Du, Y., Wang, M., Feng, J., Deng, Z., Tang, D., 2016. Effect of boric acid
concentration on the arc stability in underwater wet welding. J. Mater. Process.
Technol. 229, 244252.
Jia, C., Zhang, T., Maksimov, S.Y., Yuan, X., 2013. Spectroscopic analysis of the arc
plasma of underwater wet ux-cored arc welding. J. Mater. Process. Technol.
213, 13701377.
Jia, C., Hu, J., Du, Y., Zhang, G., Zhao, B., 2014. Visual sensing of welding arc and
metal transfer during underwater FCAW. In: Proceedings of the 19th National
Welding Conference of China Welding Society, Tianjin, China.
Liao, S., 1999. On the viscous ow past a sphere: a SimpliedDescription. Commun.
Nonlinear Sci. Numer. Simul. 4, 104109.

382

Y. Zhang et al. / Journal of Materials Processing Technology 238 (2016) 373382

Liu, S., Olson, D.L., Ibarra, S., 1993. Underwater Welding of Steels. Welding, Brazing,
and SolderingASM Materials Handbook, V.6, 10th ed., pp. 10101015,
December 1993.
Liu, X., Wang, G., Shi, Y., Zhong, J., 2007. Finite element simulation of
three-dimensional temperature eld in underwater welding. China Welding
16, 5965.
Perez, F., Liu, S., Smith, C., Rodriguez, E., 2003. Effect of nickel on toughness of
underwater wet welds. ASME 2003 International Conference on Offshore
Mechanics and Arctic Engineering.
Shi, Y.H., Zheng, Z.P., 2013. Sensitivity model for prediction of bead geometry in
underwater wet ux cored arc welding. Trans. Nonferrous Met. Soc. China 23,
19771984.
Tsai, C.L., Masubuchi, K., 1979. Mechanisms of rapid cooling in underwater
welding. Appl. Ocean Res. 1, 99110.

Wang, B., Song, Y.L., 2012. The Phenomena of Arc During the Welding and the
Weldability of Welding Materials, 1st ed. Cina Machine Press, Beijing.
Wang, G.R., Yang, Q.M., 1997. Spectrum diagnosis of underwater welding arc. J.
South China Univ., 1116.
Yin, Y., Yang, X., Cui, L., Wang, F., Li, S., 2015. Material ow inuence on the weld
formation and mechanical performance in underwater friction taper plug
welds for pipeline steel. Mat. Des. 88, 990998.
Zhang, X., Chen, W., Ashida, E., Matsuda, F., 2004. Relationship between weld
quality and optical emissions in underwater Nd: YAG laser welding. Opt. Lasers
Eng. 41, 717730.
Zhang, L., Li, Y., Wang, J., Jiang, Q., 2011. Effect of acicular ferrite on cracking
sensibility in the weld metal of Q690+ Q550 high strength steels. ISIJ Int. 51,
09151559, 11321136%@.