i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 3 ( 2 0 1 8 ) 9 3 9 9 e9 4 0 7

Available online at www.sciencedirect.com

ScienceDirect

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

The effect of hydrogen content and welding

conditions on the hydrogen induced cracking of the
API X70 steel weld

Ali Talebi Hanzaei, Seyed Pirooz Hoveida Marashi*, Eslam Ranjbarnodeh

Department of Mining and Metallurgical Engineering, Amirkabir University of Technology, Tehran, Iran

article info abstract

Article history: The self-restraint testing was used to investigate the influence of hydrogen content, pre-
Received 22 October 2017 heating, and post-heating on the sensitivity of welding of API X70 pipeline steel to
Received in revised form hydrogen induced cracking (HIC). The variation of hydrogen content was applied using a
20 December 2017 low hydrogen electrode E8018-G and a high hydrogen (cellulosic) electrode E8010-P1.
Accepted 28 March 2018 Diffusible hydrogen of these electrodes was measured by mercury displacement method.
Available online 22 April 2018 The average diffusible hydrogen content of cellulosic electrode E8010-P1 and low hydrogen
electrode E8018-G were 43.6 and 1.1 ml/100 g of weld metal, respectively. The results of
Keywords: visual inspection, penetrant test, and macroscopic examination showed that welding with
Hydrogen induced cracking cellulosic electrode leads to cracking unless both preheating and post-heating are applied.
Preheating However, in the case of low hydrogen electrode, cracking occurs only if no preheating or
Post-heating post-heating is applied. The microstructure of the welded specimens in different condi-
API 5L X70 steel tions by optical and scanning electron microscopy (SEM) showed that the dominant phase
in the weld zone of all specimens is bainite. The microhardness profile displayed that
hardness limitation (350 HV) cannot predict the sensitivity to cold cracking; therefore,
other parameters such as hydrogen content should also be considered.
© 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

The manufacturing of steel pipelines for the transfer of oil,

Introduction
Hydrogen embrittlement in steels (and other metallic alloys)
has historically been a major technological challenge in the oil
and gas industry [1e3]. With the recent focus on hydrogen
economy and its applicability across several other industrial
sectors, hydrogen embrittlement has recently gained greater
attention and become an active area of research [2,4]. It is
generally accepted that atomic hydrogen can diffuse through
steel and migrate to regions of high stress concentration, such
as an advancing crack tip, to cause failure [5e7].
gas, and most recently, hydrogen is done in accordance with
API 5L standard [8,9]. The need for high mechanical resistance,
and good fracture toughness at low temperatures and also
good weldability, requires the use of high strength low alloy
steels (HSLA) that are obtained by thermomechanical pro-
cessing [10].
The welds in ferritic steels have always been accompanied
by hydrogen induced cracking (HIC); however, the improve-
ment of steel making processes has led to high strength base
steels containing low alloy levels that are less susceptible to
HIC. In contrast, the as-cast nature of the weld metal does not

* Corresponding author.
E-mail address: pmarashi@aut.ac.ir (S.P.H. Marashi).
https://doi.org/10.1016/j.ijhydene.2018.03.216
0360-3199/© 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
9400 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 3 ( 2 0 1 8 ) 9 3 9 9 e9 4 0 7
provide the same opportunity to control the transformation
processes; therefore, the microstructures as precise as HSLA
steels cannot be achieved [11].
Previous studies have shown that the necessary conditions
for the formation of cold cracks in weld joints involve the
presence of diffusible hydrogen, tensile residual stress, sus-
ceptible microstructure, and a temperature below 200
C [12].
The possibility of having a susceptible microstructure in the
heat affected zone (HAZ) or weld metal depends on the
chemical composition of the base metal and weld metal, the
heat input, and preheating (which reduces the cooling rate)
[13,14]. Low hydrogen welding consumables, proper baking of
the consumables to remove their moisture content, and
appropriate preheating and/or post-heating conditions, which
would provide hydrogen with more time to diffuse out at high
temperature, are chosen to reduce hydrogen level [4,14].
Recently, a number of studies have been performed on the
HIC in the X70 steel weld [15e19]. Dickinson and Ries [20],
Magudeeswaran et al. [21], Chakraborty et al. [14], Pandey et al.
[22], and Saini et al. [4] studied the hydrogen cracking sus-
ceptibility by implant test. Dickinson and Ries [20] tested
various high strength steels under different welding condi-
tions and provided a model for predicting the influence of
various parameters, such as preheating, hydrogen content,
and heat input on the HIC. They expressed the results as a
function of carbon equivalent, martensite start temperature,
hardness of the HAZ, and hardenability. Magudeeswaran et al.
[21] compared the effect of welding by ferritic steel consum-
ables and austenitic stainless steel consumables on the HIC of
AISI 4340 (a quenched and tempered) steel. They concluded
that welds made by austenitic stainless steel consumables
offered a greater resistance to HIC. Chakraborty et al. [14]
assessed the susceptibility of DMR-249A (a HSLA) steel welds
made by the E8018-C1 electrode to HIC under different con-
ditions of electrode baking. They observed that neither the
steel nor the weld was susceptible to HIC even with a high
diffusible hydrogen content of 9 ml/100 g of weld metal. This
insensitivity was attributed to the presence of a mostly ferrite
structure and the absence of susceptible microstructure con-
stituents such as bainite and martensite in both weld metal
and HAZ. Pandey et al. [22] evaluated the effect of different
levels of diffusible hydrogen content in the deposited metal in
the HIC susceptibility of cast and forged P91 steel welds. They
concluded that P91 steel welded by the electrode having high
hydrogen level was more susceptible to HIC. Saini et al. [4] also
investigated the HIC susceptibility of P92 steel welds by
varying the electrode conditions. They concluded that the P92
steel plate welded by the contaminated electrode with a high
level of diffusible hydrogen had more susceptibility to HIC.
Law et al. [23] used four point bending method to develop a
test that isolated and quantitatively assessed the effects of
diffusible hydrogen on HIC in the weld metal. They believed
that this method was suitable for two reasons; first, it placed
the weld bead at the position of maximum stress, so that the
test would not be prematurely concluded due to the parent
metal or HAZ cracking, and second, it exposed the maximum
volume of weld metal to the imposed maximum stress.
A standard and close to actual conditions method in the
gas pipeline weld is the self-restraint test [24] which has not
been previously reported for the X70 steel weld. By self-
restraint test using a low hydrogen electrode and a high
hydrogen (cellulosic) electrode, the present study investigates
the effect of preheating and post-heating on the probability of
HIC in the X70 steel weld.
Experimental
Determination of hydrogen content in weld metal
Displacement of mercury standard method based on ISO 3690
[25] was used to measure diffusible hydrogen in weld metal.
For this purpose, a set-up was prepared according to the
standard to measure diffusible hydrogen, which is shown in
Fig. 1. Moreover, the welding fixture was made based on this
standard (Fig. 2). The fixture was supposed to rapidly transfer
heat to the copper blocks so that the specimen is cooled
immediately after welding to trap all the hydrogen.
The test piece assembly must be made of plain carbon non-
rimming steel with a carbon content of less than 0.18 wt% and
a sulfur content of less than 0.02 wt%. Test pieces were pre-
pared with dimensions of 30  15  10 mm3. The test assembly
was degassed at 650
C for 1 h to remove any bulk hydrogen
and was cooled in the furnace. Each test piece assembly was
finished with one operation on a surface grinder so as to
ensure a uniform width to obtain proper clamping.
The test piece, which was previously welded, cleaned and
immersed in liquid nitrogen, was removed from the liquid
nitrogen, rinsed with acetone at temperatures close to 0
C,
dried in a jet of air and transferred to the wide limb of the Y-
tube. Then, by evacuating the air, the probable residual of
acetone or air was removed from the Y-tube. Using a magnet,
the test piece was maneuvered into the position under the
capillary tube. After sufficient time, diffusible hydrogen
evolved from the test piece and collected in the capillary tube.
The length of the hydrogen gas column and the differential
level of mercury between the two limbs of the Y-tube were
measured. The ambient temperature and atmospheric pres-
sure were measured and recorded, too. Using these values and
Fig. 1 e The set-up prepared for measuring diffusible
hydrogen based on ISO 3690 standard.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 3 ( 2 0 1 8 ) 9 3 9 9 e9 4 0 7 9401

Fig. 2 e Welding fixture prepared based on ISO 3690 standard.

Table 1 e Chemical composition of the employed X70 steel (wt%).

C Si Mn P S Nb V Ti Al Cr Mo Ni Cu As
0.059 0.26 1.63 0.011 0.006 0.053 0.04 0.019 0.022 0.22 0.09 0.08 0.04 0.01

equation (1) [25], the volume of the hydrogen collected under Cold cracking test by self-restraint method
standard conditions (STP) can be calculated.
The chemical analysis of the used API 5L X70 steel is shown in
273  ðp  hÞpr2 C Table 1. The results of the tensile test in rolling direction and
VSTP ¼ (1)
760  ð273 þ TÞ  1000
perpendicular to the rolling direction are shown in Table 2. For
where welding the test pieces, the AWS A5.5 E8010-P1 cellulosic
electrode and the AWS A5.5 E8018-G low-hydrogen electrode
p is the atmospheric pressure, in millimeters of mercury. with a diameter of 3.2 mm were used. The results of their
h is the differential head, in millimeters, of mercury be- chemical analysis are shown in Table 3.
tween the two limbs of the Y-tube. The ISO17642-2 standard [24] was utilized to examine the
r is the inside radius, in millimeters, of the capillary tube. effect of the electrode type and the welding conditions that led
C is the length, in millimeters, of the gas column above the to cold cracking. The test consisted of examining transverse cut
mercury. faces of the weld by depositing a weld bead on the test specimen
T is the room temperature, in degrees Celsius, at the time of with pre-defined conditions in order to detect probable cracks
hydrogen measurement. either in the weld metal or in the HAZ. According to this stan-
dard, a number of specimens with a U-groove with dimensions
The volume at STP of diffusible hydrogen in deposited of 200  150  18 mm3 were made from X70 steel (Fig. 3). Since
metal, HD, in milliliters per 100 g, can be calculated from SMAW1 process is used for connecting pipes in the gas trans-
equation (2) [25]. mission lines, the same process of welding was used in this
study. Different conditions of preheating and post-heating for
100 both electrodes were examined on the specimens according to
HD ¼ VSTP  (2)
m2  m1
Table 4. Based on the chemical composition, specimen thick-
where ness, heat input, and hydrogen content of the cellulosic elec-
trode, according to EN 1011-2, the preheating temperature was
VSTP is the volume, in milliliters, of hydrogen gas at STP. set to 100
C and the post-heating temperature was considered
m2 is the final mass, in grams, of the test piece with to be 200
C for 2 h immediately after welding [26].
deposited metal. After depositing each weld bead in the standard specimen
m1 is the initial mass, in grams, of the test piece. groove, and after 48 h, the visual inspection and penetrant test
were done on the surface of the weld. Then, five slices with
equal distances were made on the specimen according to
Fig. 4 in order to obtain 10 faces for macro and micro exami-
nations. The cutting was carried out by water jet so that the
cutting process could not affect the crack growth and micro-
Table 2 e The results of the tensile test of the employed
structure. Specimens were polished and etched by 2% Nital
X70 steel.
solution and were examined by optical microscope. In order to
0.2% yield Ultimate tensile Elongation better detect the weld microstructure, the specimens were
strength (MPa) strength (MPa) (%)
also examined by scanning electron microscope (SEM). A
Longitudinal 460 657 30
Transverse 487 659 27 1
Shielded Metal Arc Welding.
9402 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 3 ( 2 0 1 8 ) 9 3 9 9 e9 4 0 7

Table 3 e Chemical compositions of the applied electrodes based on producer certificate (wt%).
C Mn Si P S Ni Cr Mo V
E8010-P1 0.14 0.76 0.24 0.014 0.005 0.88 0.03 0.02 <0.01
E8018-G 0.06 1.39 0.40 0.016 0.011 0.75 0.043 <0.01 <0.01

150

60

20°

A A
15

200 R6 18

9 2±0.2
A-A
15
60

Fig. 3 e Dimensions of prepared specimens from X70 steel for cold cracking test based on ISO 17642-2 standard (dimensions
in millimeters) [24].

results are shown in Table 5. The weld hydrogen contents of

Table 4 e Different welding conditions for the examined
the electrodes were slightly different, and could be expressed
specimens.
precisely. As expected, and as shown in Fig. 5, the low
Specimen no. Type of electrode Preheating Post-heating
hydrogen electrode E8018-G had much lower hydrogen con-
1 E8010-P1 No No tent than the cellulosic electrode E8010-P1 (1.1 versus 43.6 ml/
2 E8010-P1 No Yes 100 g).
3 E8010-P1 Yes No
4 E8010-P1 Yes Yes
Cold cracking
5 E8018-G No No
6 E8018-G No Yes
7 E8018-G Yes No The results of visual inspection and penetrant test on the
8 E8018-G Yes Yes surface of the specimens welded according to Table 4 showed
that the specimens 1, 2, 3, and 5 had longitudinal overall
microhardness test was also performed on the specimens cracks in the weld surface. To determine the depth of the
using the Vickers method applying 50 g force as a profile from crack and further scrutiny, as described in the experimental
the middle of the weld to the base metal at intervals of 500 mm. section of the present study, the macroscopic examination
was done on the specimen sections. Fig. 6 shows the macro-
graphic images along with the numbers associated with each
Results and discussion image related to Table 4. These images confirm the observed
surface cracks and also show that the cracks are deep and
Diffusible hydrogen content extended from the weld root to the surface. Having these
observations, it can be concluded that if the cellulosic elec-
Three series of specimens were prepared and welded by each trode is used for welding, no cold cracking is expected to occur
electrode, and their hydrogen content was measured; the in case of simultaneous application of both preheating and

Fig. 4 e Cutting location for checking cracks on the standard welded specimen based on ISO 17642-2 [24].
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 3 ( 2 0 1 8 ) 9 3 9 9 e9 4 0 7 9403

Table 5 e Diffusible hydrogen content of weld metal made by two types of electrodes.
Electrode Specimen VSTP (ml) m1 (g) m2 (g) HD (ml/100 g) Mean HD Standard deviation
(ml/100 g) for HD
Cellulosic electrode E8010-P1 a 1.47 35.07 38.30 45.43 43.62 1.94
b 1.56 35.06 38.82 41.58
c 1.32 35.08 38.10 43.85
Low hydrogen electrode E8018-G d 0.04 35.15 38.66 1.17 1.10 0.14
e 0.03 35.18 38.23 0.94
f 0.04 35.18 38.65 1.19

64 electrode E8010-P1 (Table 3), the diffusion coefficient of

hydrogen in the weld zone made by the E8018-G electrode is
32
higher than that of the weld zone made by the E8010-P1
16 electrode [9]. Therefore, although the initial concentration of
Diffusible 8 hydrogen in the E8010-P1 electrode is much higher than that
hydrogen of the E8018-G electrode, since the principle source of
(ml/100g) 4
hydrogen is the electrode, the relative concentration of
2
hydrogen in the interface of the base metal and weld,
1 compared to the other regions, is higher in the specimens
0.5 welded by the E8018-G electrode than in the specimens wel-
1 2 3 ded by the E8010-P1electrode, due to the higher diffusion co-
E8010-P1 45.4 41.6 43.8 efficient mentioned above. This means that one of the factors
E8018-G 1.2 0.9 1.2 affecting cold cracking e i.e. hydrogen content e has been
strengthened in this region. If there are conditions that lead to
Fig. 5 e Comparison of diffusible hydrogen content in the HIC, this region will be more sensitive than other regions. This
welds of two types of electrodes. can be the reason for the presence of cracks in the HAZ of
specimen 5, welded by the low hydrogen electrode, E8018-G.

post-heating. On the other hand, if no preheating and post- Microstructure

heating is applied, cold cracking is likely to occur in case of
using a low hydrogen electrode.
As seen in Fig. 6, cracks occurred in the weld metal in
specimens 1, 2, and 3, which were welded by cellulosic elec-
trodes. While in specimen 5 that was welded by low hydrogen
electrode, the crack was located in HAZ. Ayesha et al. [9]
revealed that the diffusion coefficient of hydrogen increases
by increasing the manganese in X70 steel. Since the manga-
nese content in the low hydrogen electrode E8018-G is
approximately twice the manganese content in the cellulosic
Fig. 7 shows the microstructure of the base metal. The ferritic
structure with the equiaxed grains, along with fine perlite
colonies, is evident in this figure. Formation of this structure is
due to the thermomechanical operations in the process of
producing HSLA [27]. Figs 8 and 9, respectively, show the op-
tical microscopic and SEM images of the weld metal of the
specimens that are welded with the conditions expressed in
Table 4. Comparing the optical microscopic and SEM images
and using ImageJ software, the formed phases in the

Fig. 6 e Macrographic images of cold cracking test (the numbers associated with each image is related to Table 4).
9404 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 3 ( 2 0 1 8 ) 9 3 9 9 e9 4 0 7
Fig. 7 e Optical microscopic image of the base metal
microstructure (etched with 2% Nital solution).
microstructure of the specimens and their approximate per-
centage were determined, the results of which are briefly
summarized in Table 6. Since the acicular ferrite structure is
an interlocked structure consisted of the bainite or Widman-
statten ferrite that has nucleated on inclusions in the prior
austenite grains, here the acicular ferrite is not considered an
independent microstructure [28]. In the preheated specimens
(specimens 3, 4, 7 and 8 in Table 4), 10%e20% of the primary
polygonal ferrite is observed. The reduction of the cooling rate
caused by preheating created the conditions for the formation
of this phase at the prior austenite grain boundary. Further-
more, in specimens welded without preheating (specimens 1,
2, 5, and 6), Widmanstatten ferrite and martensite are
observed due to the higher cooling rate after welding. The
results of the microhardness test, presented later in this
paper, prove the presence of these brittle phases. The per-
centage of brittle phases is reduced in specimen 2 as
compared to specimen 1 and also in specimen 6 as compared
to specimen 5, which can be due to the effect of post-heating.
Fig. 8 e Optical microscopic images of the microstructure of the weld zone. The numbers below each image are related to the
conditions applied according to Table 4. (Etched with 2% Nital solution).
Hardness
Fig. 10 shows the results of the microhardness test using the
Vickers method as a profile from the middle of the weld to the
base metal. The hardness variations in all specimens are
almost the same. It decreases from the middle of the weld to
the base metal. The maximum hardness achieved after
welding by cellulosic electrode is 366 HV (Fig. 10) that is higher
than the maximum hardness achieved after welding by low
hydrogen electrode (307 HV). The reason is that there is higher
carbon equivalent in cellulosic electrode (0.24%) than in low
hydrogen electrode (0.22%) [26]. Moreover, there are harder
microstructure constituents in the weld zone of specimens
welded by cellulosic electrode than in the weld zone of spec-
imens welded by low hydrogen electrode (Table 6).
Specimens 1 and 2, welded by cellulosic electrodes (Fig. 10
(a)) without preheating, had the highest hardness, and were
cracked after welding. The crack occurred because of the
presence of high amount of hydrogen due to the cellulosic
electrode (Table 5) along with the brittleness of these speci-
mens due to the presence of martensite and bainite in the
weld microstructure (Table 6). The hardness of specimen 2 in
HAZ was much lower than that of specimen 1, which can be
due to the effect of post-heating that reduced the cooling rate.
Welds of specimens 3 and 4, which were preheated before
welding, had a lower hardness than those of specimens 1 and
2. This was because of the absence of martensite in the weld
structure of specimens 3 and 4 while the weld structure of
specimens 1 and 2 consisted of 20 and 10% martensite,
respectively (Table 6). The hardness of specimen 3 was lower
than that of specimen 4 in the weld zone while it was higher in
HAZ because of the post-heating of specimen 4. The presence
of crack in the weld zone of specimen 3, despite its lower
hardness in the weld zone compared to that of specimen 4,
can be due to the higher concentration of hydrogen in the
weld zone of specimen 3. This shows the effect of post-
heating on specimen 4, which is the diffusion and removal
of hydrogen from the weld zone, and the higher resistance to
hydrogen cracking. As expected, specimen 5 had the highest
hardness and HAZ among the specimens welded by a low
hydrogen electrode (Fig. 10 (b)) because no preheating and
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 3 ( 2 0 1 8 ) 9 3 9 9 e9 4 0 7
Fig. 9 e SEM images of the weld zone microstructure. The numbers below each image are related to the conditions applied
according to Table 4. (Etched with 2% Nital solution).
Table 6 e Approximate percentage of phases in the weld zone of specimens.
Specimen no. Widmanstatten ferrite
1 10%
2 e
3 e
4 e
5 e
6 10%
7 e
8 e
post-heating were applied and it had the highest cooling rate,
which led to the formation of martensite in its microstructure.
The presence of crack in the HAZ of specimen 5 was because
of more brittle structure of this specimen than the structures
of specimens 6, 7, and 8, and also higher hydrogen concen-
tration due to the lack of preheating and post-heating.
Comparing the hardness results and the microstructure of
the weld metal, it can be observed that depending on the
preheating and post-heating conditions, the hardness is
reduced by decreasing the brittle phases in the structure.
Furthermore, preheating is the most effective factor in
reducing the hardness as it reduces the cooling rate and pre-
vents the formation of brittle phases such as martensite.
The maximum hardness is often limited to 350 HV in
welding fabrication to avoid cold cracking [29]. However, in
this study the maximum hardness of the cracked specimens
was 366 HV (specimen 1), 330 HV (specimen 2), 284 HV (spec-
imen 3), and 307 HV (specimen 5). All the specimens, except
specimen 1, had hardnesses lower than 350 HV and were
cracked. Therefore, this limitation (350 HV) cannot predict the
sensitivity to cold cracking, and other parameters such as
hydrogen content and residual stresses should also be
considered.
The microstructure and hardness of the welded specimens
in this study had similar behavior to the microstructure and
9405
Ferrite veins Polygonal ferrite Bainite Martensite
e e 70% 20%
10% e 80% 10%
e 20% 80% e
e 10% 90% e
10% e 80% 10%
e e 90% e
20% 20% 60% e
e 10% 90% e
hardness of a steel shaft after heat treatment [30] because in
both cases, modifying a cooling rate changes the microstruc-
ture, and the hardness as a result. Thus, similar relationship
between microhardness and microstructure is observed in a
welded specimen and heat treated steel shaft.
As depicted in Fig. 10, the weld hardness of specimen 4 is
higher than that of specimen 3 and the weld hardness of
specimen 8 is higher than that of specimen 7. It indicates that,
compared to welding with preheating and without post-
heating, the weld hardness increases by simultaneous appli-
cation of preheating and post-heating. This increase in the
weld hardness occurs in the welds made by both types of
electrodes, which could be due to some kind of secondary
hardness mechanisms caused by post-heating. However, its
definitive cause cannot be explained by the tests conducted in
this study, but further studies are needed.
Therefore, it can be concluded that preheating and post-
heating have two effects. First, they increase the diffusion
rate of hydrogen and increase the hydrogen removal from the
specimen, which reduces the local concentration of hydrogen
in the critical regions of microstructure and residual stress.
Second, they lead to a less brittle microstructure that is more
resistance to HIC. In a study done by Dickinson and Ries [20],
which examined the effect of different parameters on
hydrogen cracking susceptibility by implant testing,
9406 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 3 ( 2 0 1 8 ) 9 3 9 9 e9 4 0 7

a
400
Weld HAZ
350

HV (50 gf) 300

250

200

150
0 1 2 3 4 5 6
Distance from weld center line (mm)

1 (without preheating and post-heating) 2 (without preheating and with post-heating)

3 (with preheating and without post-heating) 4 (with preheating and post-heating)

b
400
Weld HAZ
350
HV (50 gf)

300

250

200

150
0 1 2 3 4 5 6
Distance from weld center line (mm)

5 (without preheating and post-heating) 6 (without preheating and with post-heating)

7 (with preheating and without post-heating) 8 (with preheating and post-heating)

Fig. 10 e The results of the microhardness test using the Vickers method with 50 g force as a profile from the middle of the
weld to the base metal; a) Specimens welded by cellulosic electrode E8010-P1, b) Specimens welded by low hydrogen
electrode E8018-G.

preheating had the greatest effect on reducing the hydrogen
cracking sensitivity. In this study, as shown in Table 6 and
Fig. 10, preheating has the greatest effect on reducing the
brittle phases and the hardness of the weld. Because pre-
heating reduces the cooling rate from 800 to 500
Celsius per
second, and according to the continuous cooling trans-
formation (CCT) diagrams, phases with less hardness, such as
ferrite, are formed [20]. The greatest effect of post-heating is
the surface removal of hydrogen from the specimen and
decreasing its concentration in critical areas.
Conclusions
The present study investigated the effect of the type of elec-
trode, preheating and post-heating on the cold cracking
sensitivity in X70 steel welded by the self-restraint method.
For this purpose, the amount of diffusible hydrogen in two
types of cellulosic and low hydrogen electrode was measured
via displacement of mercury. The sensitivity to cold cracking
was studied in self-restraint condition and using visual in-
spection, penetrant test, metallography and microhardness
test. The following results were obtained:
 Welding by the cellulosic electrode E8010-P1 led to
cracking, except in case of applying both preheating and
post-heating simultaneously. Therefore, despite the high
amount of hydrogen in the weld made by the cellulosic
electrode (43.6 ml/100 g), it was possible to prevent the
formation of HIC by applying proper preheating and post-
heating.
 The weld made by the low hydrogen electrode E8018-G was
cracked in the absence of preheating or post-heating.
Therefore, welding by a low hydrogen electrode (1.1 ml/
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 3 ( 2 0 1 8 ) 9 3 9 9 e9 4 0 7
100 g) could not prevent cracking, thus preheating or post-
heating was essential to prevent HIC.
 The dominant phase in the weld zone of all specimens was
bainite. In the specimens welded with preheating, there
was 10e20% of the polygonal ferrite, and brittle phases
such as Widmanstatten ferrite and martensite were pre-
sent in the specimens without preheating.
 The hardness limitation (350 HV) cannot predict the
sensitivity to cold cracking, and other parameters such as
hydrogen content should also be considered.
Acknowledgement
The authors are thankful to the support of the Iranian Gas
Engineering and Development Company. We also thank the
company of Yazd Ghadir Industrial Turbines Company and
Yazd Electrode Company for providing welding equipment.
references
[1] Perez TE. Corrosion in the oil and gas industry: an increasing
challenge for materials. JOM 2013;65:1033e42.
[2] Srinivasan R, Neeraj T. Hydrogen embrittlement of ferritic
steels: deformation and failure mechanisms and challenges
in the oil and gas industry. JOM 2014;66:1377e82.
[3] Thodla R, Piza Paes MT, Gerst B. Hydrogen assisted cracking
of AISI 4137M steel in O&G environments. Int J Hydrogen
Energy 2015;40:17051e64.
[4] Saini N, Pandey C, Mahapatra MM. Effect of diffusible
hydrogen content on embrittlement of P92 steel. Int J
Hydrogen Energy 2017;42:17328e38.
[5] Hirth J. Effects of hydrogen on the properties of iron and
steel. Metall Trans A 1980;11:861e90.
[6] da Silva BRS, Salvio F, Santos DSD. Hydrogen induced stress
cracking in UNS S32750 super duplex stainless steel tube
weld joint. Int J Hydrogen Energy 2015;40:17091e101.
[7] Yan Y, Yan Y, He Y, Li J, Su Y, Qiao L. Hydrogen-induced
cracking mechanism of precipitation strengthened
austenitic stainless steel weldment. Int J Hydrogen Energy
2015;40:2404e14.
[8] Institute AP. Specification for line pipe. ANSI/API specification
5L. Washington, DC: API Publishing Services; 2008.
[9] Ayesha J, Haq KM, Dunne DP, Calka A, Pereloma EV. Effect of
microstructure and composition on hydrogen permeation in
X70 pipeline steels. Int J Hydrogen Energy 2013;38:2544e56.
[10] Godefroid LB, Ca ^ ndido LC, Toffolo RVB, Barbosa LHS.
Microstructure and mechanical properties of two API steels
for iron ore pipelines. Mater Res 2014;17:114e20.
[11] Holdstock R. Inducing hydrogen assisted cold cracking in high
strength steel weld metal. University of Wollongong; 2009.
9407
[12] Viyanit E. Numerical simulation of hydrogen assisted
cracking in supermartensitic stainless steel welds. Hamburg:
Helmut-Schmidt; 2005.
[13] Kumar PG, Yu-ichi K. Diffusible hydrogen in steel
weldments. Trans JWRI 2013;42:39e62.
[14] Chakraborty G, Rejeesh R, Albert SK. Study on hydrogen
assisted cracking susceptibility of HSLA steel by implant test.
Def Technol 2016;12:490e5.
[15] Alvaro A, Olden V, Macadre A, Akselsen OM. Hydrogen
embrittlement susceptibility of a weld simulated X70 heat
affected zone under H2 pressure. Mat Sci Eng A Struct
2014;597:29e36.
[16] Olden V, Alvaro A, Akselsen OM. Hydrogen diffusion and
hydrogen influenced critical stress intensity in an API X70
pipeline steel welded joint e experiments and FE
simulations. Int J Hydrogen Energy 2012;37:11474e86.
[17] Alvaro A, Olden V, Akselsen OM. 3D cohesive modelling of
hydrogen embrittlement in the heat affected zone of an X70
pipeline steel. Int J Hydrogen Energy 2013;38:7539e49.
[18] Alvaro A, Olden V, Akselsen OM. 3D cohesive modelling of
hydrogen embrittlement in the heat affected zone of an X70
pipeline steel e part II. Int J Hydrogen Energy
2014;39:3528e41.
[19] Fiore S, Boring M. Evaluation of hydrogen cracking in weld
metal deposited using cellulosic-coated electrodes. EWI:
Materials Joining Technology; 2009.
[20] Dickinson D, Ries G. Implant testing of medium to high
strength steel-a model for predicting delayed cracking
susceptibility. Weld J 1979;58:205e11.
[21] Magudeeswaran G, Balasubramanian V,
Madhusudhanreddy G. Hydrogen induced cold cracking
studies on armour grade high strength, quenched and
tempered steel weldments. Int J Hydrogen Energy
2008;33:1897e908.
[22] Pandey C, Saini N, Mahapatra MM, Kumar P. Hydrogen
induced cold cracking of creep resistant ferritic P91 steel for
different diffusible hydrogen levels in deposited metal. Int J
Hydrogen Energy 2016;41:17695e712.
[23] Law M, Nolan D, Holdstock R. Method for the quantitative
assessment of transverse weld metal hydrogen cracking.
Mater Char 2008;59:991e7.
[24] ISO17642-2. Destructive tests on welds in metallic materials -
cold cracking tests for weldments - Arc welding processes.
Self-restraint tests. 2005.
[25] ISO3690. Welding and allied processes. Determination of
hydrogen content in arc weld metal. 2012.
[26] EN1011-2. Welding - recommendations for welding of
metallic materials. Arc Welding of Ferritic Steels; 2001.
[27] Hillenbrand H-G, Graf M, Kalwa C. Development and
production of high strength pipeline steels. Florida, USA:
Niobium 2001 Orlando; 2001.
[28] Thewlis G. Classification and quantification of
microstructures in steels. Mater Sci Tech Lond
2004;20:143e60.
[29] Bang I, Son Y, Oh K, Kim Y, Kim W. Numerical simulation of
sleeve repair welding of in-service gas pipelines. Weld J
2002;81. 273eS - 82-S.
[30] ASM Handbook. Heat treatingvol. 4; 1991.