j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2

Available online at www.sciencedirect.com

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

Original Article

Microstructural characterisation and

micromechanical investigation of the heat-affected
zone in multi-pass welding of 9Ni steel pipes


Pedro P.N. Maia a,**, Emerson M. Mina
a, Giovani Dalpiaz b,
Ricardo R. Marinho b, Marcelo T.P. Paes b, Marcelo F. Motta a,
H
elio C. de Miranda a, Cleiton C. Silva a,*
a

, Building 729, Fortaleza,
Department of Metallurgical and Materials Engineering, Federal University of Ceara
60440-554, Brazil
b
Leopoldo Am
erico Miguez de Mello Research and Development Center, Petrobras, Rio de Janeiro, 21941-915, Brazil

article info abstract

Article history: 9Ni steels have been recently adopted in supercritical CO2 injection systems in deepwater
Received 22 November 2022 oil fields. The manufacture of these reinjection systems involves multi-pass welding pro-
Accepted 12 March 2023 cedures, which produce an Heat-Affected Zone (HAZ) with a high heterogeneity level
Available online 17 March 2023 regarding the microstructural features and the local mechanical properties. An extensive
microstructural and micromechanical characterisation was performed over the HAZ of
Keywords: three welded joints with different heat-input conditions to evaluate the effects of the
9Ni steels reheating cycles and the welding parameters on the microconstituents. Light Optical Mi-
Welding croscopy (LOM), Scanning Electron Microscopy (SEM) and Electron Backscatter Diffraction
Heat-affected zone (EBSD) analyses were performed to identify microstructural features that correlate to the
Microstructural features local mechanical responses evaluated through an extensive microhardness mapping.
Micromechanical characterisation Regarding the Coarse Grained HAZ (CGHAZ), the highest microhardness values for all
welding conditions are found at the Supercritically Reheated CGHAZ (SCR-CGHAZ), char-
acterised by its refined microstructure and a quite low area fraction of coarse martensite
laths. The Subcritically Reheated CGHAZ (SC-CGHAZ) and the Intercritically Reheated
CGHAZ (IC-CGHAZ) e regions where wider martensite blocks and higher coarse martensite
lath area fractions were observed e composed the softer zones of the microhardness map.
It was also found that reheating at intercritical temperatures induces the formation of
supersaturated fresh martensite and may contribute to retained/reversed austenite parti-
cles’ C-enrichment, which may degrade the mechanical properties at the IC-CGHAZ.
© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC
BY license (http://creativecommons.org/licenses/by/4.0/).

* Corresponding author.
** Corresponding author.
E-mail addresses: [email protected] (P.P.N. Maia), [email protected] (C.C. Silva).
https://doi.org/10.1016/j.jmrt.2023.03.092
2238-7854/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2 1717

(TRIP) effect is activated. Many authors classify the 9Ni steel as

1. Introduction a TRIP steel [19e21], proposing that the martensite trans-
formation induced by plastic strain would enhance plasticity
The challenges of the offshore oil and gas production have by the introduction of deformation mechanisms [22]. How-
been, through decades, promoting many technological ad- ever, Barrick and DuPont [23], in a study on simulated HAZ in a
vancements to the petroleum industry, like the prospecting of 10 wt% Ni steel, demonstrated that the toughness is not solely
oil in deep waters [1], the deep offshore drilling technologies derived from the austenite content, which is unexpected for
[2], and the manufacturing of complex subsea systems [3]. In TRIP steels. The authors verified that, in some cases, austenite
Brazil, the oil production in the pre-salt region of the Santos content does not even follow a positive correlation with
basin faces a challenge related to the high concentration of toughness [23]. Moreover, recent papers point out martensite
carbon dioxide (CO2), which may reach values up to 12%. As morphology heterogeneities as one major factor in deter-
soon as modules process the CO2 removal on the Floating, mining deformability in low-carbon Ni-alloyed steels [24].
Production, Storage and Offloading (FPSO) unit, the separated Morsdorf et al. [25] developed extensive work towards the
CO2 has been pressurised and then transported through characterisation of microstructural heterogeneities in 5Ni
pipelines to injection wells [4]. This process mitigates the steel. The authors identified coarse martensite laths as soft
emission of CO2 and plays an essential role in oil recovery, zones within a much harder matrix of thin laths, proposing
improving wells’ efficiency [5]. that these constituents form a nano-composite-like micro-
Favouring the efficiency of the injection processes, the CO2 structure, in which the coarse martensite lath could signifi-
must be transported in a liquid or supercritical state, cantly enhance the steel's toughness.
demanding the application of high pressures, scaling up to Drawing the relationship between microstructure and
550 bar [6,7]. In this scenario, depressurisation, caused by mechanical properties in 9Ni steels has proven to be a com-
cracking or by a malfunction in flange sealing rings, may lead plex task, which can be hampered when evaluating the
to an abruptly localised decrease in temperature, which may reheated HAZ resulting from multi-pass welding. Although
reach
90  C, eventually promoting localised embrittlement few studies were presented regarding the effects of multi-pass
of the pipeline [8]. welding on microstructure and toughness of the HAZ in 9Ni
In this context, 9Ni steels stand out as a great option for steels [12e14,26,27], there is a lack of information on how the
manufacturing the top side portion of these injection systems. numerous microstructural features may impact the mechan-
Presenting a DuctileeBrittle Transition Temperature (DBTT) ical properties of this HAZ. In this context, the present study
as low as
196  C [9], those steels are commonly adopted in aims to conduct a detailed microstructural investigation
the transport and storage of Liquefied Natural Gas (LNG), and through the different HAZ regions, exposing extensive ana-
its applications are recently expanding to the supercritical CO2 lyses and quantification of the microstructural features and
injection in deepwater oil fields [10]. The manufacture of the micromechanical responses and bringing scientific and engi-
injection system involves welding procedures, which, must neering insights on the microstructure-strengthening rela-
follow a multi-pass approach due to the large wall thickness of tionship. Samples of three heat-input conditions were
the pipes. The resulting HAZ microstructure is essentially submitted to broad microhardness mapping and microstruc-
martensitic due to the high hardenability of 9Ni steels [11]. tural characterisation by LOM and SEM with EBSD analysis.
However, the superposition of welding heat cycles will dras-
tically impact the HAZ microstructure by introducing hetero-
geneity upon microstructural features, such as the size and 2. Experimental procedure
morphology of constituents and the fraction of second phases.
It is reported that the multi-pass welding can introduce Local This work analysed three distinct welded joints of quenched
Brittle Zones (LBZ) in 9Ni steels HAZ [12,13], which would be and tempered ASTM A333 Gr.8 (9Ni) steel pipes, which
related to the presence of Martensite-Austenite (M-A) con- chemical composition (Table 1) was evaluated through optical
stituents [14]. emission spectroscopy with a Shimadzu PDA-7000. The pipes
Along with the features associated with the M-A constit- presented a diameter of 219.1 mm and thickness of 31.7 mm
uents, other microstructural factors may also impact the 9Ni and were girth welded through Gas Metal Arc Welding
steel toughness. The retained austenite, for example, is clas- (GMAW) using a Markle HighPULSE 550 RS power source in a
sically attributed as the most significant agent for toughness robotic workbench equipped with a KUKA KR16 robot and a
improvement in low-carbon Ni-alloyed steels [15,16]. The role KUKA DKP-400 positioner. The adopted filler metal was the Ni-
of retained austenite on toughness of high strength based superalloy Inconel 625 (AWS ER NiCrMo-3). For the root
martensitic steels has been the subject of several studies. pass, a heat input of 0.5 kJ/mm was aimed in all joints, using a
Besides low-carbon Ni-alloyed steels, medium Mn steels are continuous direct current of 119 A, a voltage of 16,4 V, and a
an example of another high strength steel class in which its 250 mm/min welding speed, achieving a short-circuiting
toughness is commonly related to the metastable austenite
within its microstructure [17]. According to Kuzmina et al. [18],
the nucleation and growth of austenite at the grain boundary
Table 1 e Chemical composition (wt.%) of the ASTM A333
play an essential role in improving the impact toughness, first
Gr.8 (9Ni) steel pipes.
due to a cleaning effect, removing Mn from these interfaces,
and second, by a crack blunting mechanism, via martensitic C Mn Si Ni Cr S P Fe
transformation when the Transformation Induced Plasticity 0.032 0.496 0.278 9.460 0.032 <0.005 <0.005 Bal.
1718 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2

metal transfer. For the fill and cap passes, pulsed GMAW was evaluate the size distribution of martensite constituents along
applied, imposing different heat-input conditions for each of the HAZ of each sample.
the three welded joints. The aimed parameters for each con-
dition are presented in Table 2 [28], which also presents the
heat-input values calculated by considering a GMAW arc ef- 3. Results
ficiency of 0.85 [29]. The total number of passes for the low,
medium and high heat-input conditions was 36, 18 and 11, 3.1. Microhardness testing
respectively.
Samples of the welded joints’ complete cross-section were Fig. 2 shows the microhardness profiles for the cap passes
removed and submitted to conventional metallographic HAZ in each sample, presenting insets of LOM micrograph
preparation, being grinded with sandpaper and mechanically montages which supported the HAZ segmentation into
polished up to 1mm diamond paste. CGHAZ, Fine Grained HAZ (FGHAZ) and Intercritical HAZ
With surfaces in as-polished condition, microhardness (ICHAZ). Regarding the CGHAZ and the FGHAZ, complete
Vickers mapping was carried out with a LECO AMH55 LM- austenitization (T > Ac3) is provided in both cases. However,
110AT tester on the HAZ of each sample. A load of 100 gf in CGHAZ, the peak temperature is sufficiently high to pro-
and a dwell time of 13 s were applied, while the samples were mote significant grain growth, achieving a coarse micro-
orthogonally displaced, guaranteeing a 100mm distance be- structure, while, in FGHAZ, a lower austenitization
tween each indentation. The mapped areas covered the temperature will result in a refined microstructure. The
reheated HAZ of at least three fill passes located at the half- ICHAZ, in turn, is submitted to intercritical temperatures
height of the samples. Also, microhardness profiles were (Ac3 > T > Ac1) in which partial austenitization occurs, also
evaluated on the cap passes HAZ of each sample, aiming to allowing carbon diffusion and precipitation, mainly along
compare a non-reheated HAZ with the reheated mapped HAZ. with grains and lath boundaries [30]. Some authors may also
The samples’ HAZ and Base Metal (BM) were then chemi- discriminate a fourth HAZ region in carbon and low alloy
cally etched by immersion in 4% Picral solution for 45 s fol- steels welding, the so-called Subcritical HAZ (SCHAZ), sub-
lowed by immersion in 2% Nital solution for 20 s. LOM analysis mitted to temperatures below Ac1, providing a tempering ef-
was then performed as a first evaluation step with a Carl Zeiss fect [31]. The SCHAZ wasn't identified in any of the profiles of
Axiovision Z1M. A deeper analysis was carried out by SEM Fig. 2, as the BM microstructure already consists of tempered
using a FEI Quanta 450 FEG operated at 20 kV in secondary lath martensite, which is the expected microstructure for the
electron mode. Fig. 1 shows the etched low heat-input sample, SCHAZ.
highlighting the HAZ mapped area. From the microhardness profiles of Fig. 2, it is clear that the
For macrohardness measurements Hardness Vickers CGHAZ consists of a softer HAZ region at the cap passes of all
testing was conducted in a ZwickRoell Indentec ZHU250CL samples, while the higher microhardness values are observed
universal hardness tester. A load of 10 kgf and a dwell time of at the FGHAZ, and moderate values are observed at the ICHAZ.
10 s were applied for 6 indents scattered along the fill passes’ These results oppose Fernandes et al. [32] observations, which
HAZ of each welded joint. report the CGHAZ as the hardest and most brittle HAZ region.
Also, size-reduced samples of the welding interfaces were Nevertheless, having the CGHAZ as softer HAZ regions cor-
extracted for microstructural investigation through EBSD ana- roborates the recent proposals by Morsdorf et al. [25], which
lyses. Those samples were submitted to a similar metallo- points out the coarse martensite lath as a significant agent for
graphic preparation with the addition of a final electropolishing mechanical properties determination in low-carbon Ni-
step, which was performed with a mixture of 70% perchloric alloyed steels, being identified as soft zones inside a thin lath
acid in 99.8% ethanol in a volume ratio of 1:9. The polishing was martensite matrix, and capable of significantly enhance the
conducted at 10  C, applying 25 V during 50 s. EBSD measure- steel's toughness. The author verified that a larger austenite
ments with 0.5 mm step size were carried out in a FEI Quanta 250 grain size would increase the coarse lath area fraction [25].
SEM coupled with Oxford Nordlys Max EBSD detector operated Therefore, it is expected that the CGHAZ might present higher
at 20 kV. Inverse Pole Figure (IPF) maps and microstructural coarse lath area fraction in comparison to the other HAZ
interfaces were then constructed using Channel 5 package to regions.
Regarding the reheating imposed by the multi-pass weld-
ing, Fig. 3 shows the microhardness maps over the fill passes
HAZ in each sample, revealing a heterogeneous but stand-
Table 2 e Parameters for fill and cap passes in welding of
ASTM A333 Gr.8 (9Ni) steel pipes [28]. ardised HAZ, where it is possible to discriminate regions ac-
cording to their microhardness values. In Fig. 3aec, the upper
Heat-input Arc Mean Welding Heat-
condition voltage welding speed (mm/ input (kJ/ area of light blue tones consists of the FZ, while the lower area,
(V) current (A) min) mm) composed of dark blue tones (lower microhardness), consists
of the BM.
Low heat- 470 0.6
input To evaluate the reheated HAZ, we may introduce new la-
Medium 30 184 220 1.3 bels to describe the impact of the reheating thermal cycles
heat- over the previous passes' HAZ. Therefore, the previously
input introduced HAZ segmentation may be sub-categorised ac-
High heat- 132 2.1 cording to the reheating temperature. The CGHAZ, for
input
example, can be segmented into: Subcritically reheated
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2 1719

Fig. 1 e Low heat-input sample macrography showing the multi-pass weld cross-section. The dark zone corresponds to the
Inconel 625 fusion zone, and the light grey along with the interface is the HAZ. Sample etched by immersion in 4% Picral and
2% Nital solutions (a). Macrography of the enlarged mapped region as indicated in Fig. 1a (b). LOM micrograph montage of
the microhardness mapped region (c).

CGHAZ (SC-CGHAZ); Intercritically reheated CGHAZ (IC- in a close to radial manner. This configuration indicates that
CGHAZ); Supercritically reheated CGHAZ (SCR-CGHAZ), con- the reheating cycles greatly impact the mechanical properties
dition achieved when low super-critical reheating tempera- of the HAZ, surpassing the influence of the initial HAZ
tures are imposed, resulting in a microstructure similar to that microstructure defined by the previous welding pass. The ef-
of the FGHAZ; Unaltered CGHAZ (UA-CGHAZ), condition ach- fects of the reheating cycles are also highlighted by the
ieved when high super-critical reheating temperatures are microhardness values periodicity observed when evaluating
imposed or when isn't observed any relevant effect from profiles parallel to the weld interface. This behaviour is
reheating [13,33]. demonstrated by the microhardness profile presented in
This segmentation relates to the high heterogeneity of Fig. 5, which follows the vector displayed in Fig. 3a. The profile
mechanical properties that is clearly observed in the micro- exhibit two valleys e at X ¼ 1.0 mm and X ¼ 4.9 mm e and two
hardness maps, where a wide range of microhardness values peaks e at X ¼ 1.7 mm and X ¼ 5.7 mm. The valleys and peaks
is obtained at the HAZ of each sample. Fig. 4a shows statistics correspond to specific HAZ regions with particular micro-
about the macrohardness measurements taken on the HAZ of structural features responsible for the different mechanical
each sample, pointing out a drop in microhardness values responses. Also, the profile exhibited in Fig. 5 was used as a
with the increase of heat-input, which relates to the increase reference during the microstructural characterisation of the
in softer zones area fraction observed when comparing the CGHAZ distinct regions, as the highlighted indents mark the
maps of Fig. 3. The wide range of microhardness values results areas where microscopy analyses were initially conducted.
in the large error bars presented in Fig. 4, hindering the proper The micrographs presented in Fig. 7 refer to the highlighted
evaluation of the heat-input effect on the micromechanical section of the microhardness profile in Fig. 5, with the corre-
properties. However, when evaluating a specific CGHAZ spondent indents being marked with red borders on the
segment (i.e., SC-CGHAZ, IC-CGHAZ, or SCR-CGHAZ), it is micrographs.
possible to verify a clear trend between the heat-input and the
microhardness values, where higher heat inputs generate 3.2. Microstructural characterisation
softer microstructures. Nevertheless, at the IC-CGHAZ
considerably large error bars are obtained due to microstruc- To understand the impact of the heating cycles over the HAZ
tural features exposed in the following subsection. microhardness distribution it is necessary to carry out an
The microhardness maps revealed that high hardness extensive microstructural investigation over this complex
zones, as well as low hardness zones, may reach out the whole zone. By an initial HAZ macro analysis, it is possible to tell that
HAZ extension, from the weld interface to the BM, expanding the radial regions of the microhardness maps correspond to
1720 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2

Fig. 2 e Microhardness profiles for the cap passes HAZ in samples of low heat-input (a), medium heat-input (b) and high
heat-input (c). The profiles coverd from the Fusion Zone (FZ) to the BM.

regions submitted to similar reheating cycles. Fig. 6a and b segmentation. Fig. 6ceg shows micrographs of the regions
reveal that the soft HAZ zones of the map initially presented in highlighted in Fig. 6a. In those images, the microhardness
Fig. 3a are invariably located to the left of intensely etched indents were superimposed by geometries coloured according
regions, which exhibit darker tones in LOM analyses. As one to the microhardness scale introduced in Fig. 3. Also, higher
may see by the HAZ pattern shown in Fig. 6a and b, those soft magnification LOM images are presented in Fig. 7 following
zones relate to the HAZ regions that are the farthest from the the microhardness profile of Fig. 5. The highlighted indents in
heat source of the respective welding pass and, therefore, are Fig. 5 are marked with red borders at the micrographs of Fig. 7.
submitted to a low temperature reheating, which will promote Fig. 6c exhibit a micrograph over the softer CGHAZ region.
a tempering effect on those regions. Through the same im- The microstructure is highly etched on the right side, exhib-
ages, it's possible to verify that this reheating cycle isn't pro- iting a higher contrast in LOM analysis, while the micro-
nounced at the HAZ/BM interface, as the BM already exhibits a structure on the left side doesn't share the same aspect. A
microstructure composed of tempered martensite. Those ob- higher magnification LOM micrograph of the later region is
servations validate the fact that the soft HAZ regions of the presented in Fig. 7a. As discussed above, the HAZ micro-
microhardness maps from Fig. 3 consist of subcritically hardness pattern indicates that this region consists of the SC-
reheated areas. CGHAZ, presenting microhardness values ranging from 250
However, considering the high microstructural complexity HV to 271 HV for the medium heat-input condition (Fig. 4b),
of the evaluated reheated HAZ, the proper identification and while the non-reheated CGHAZ of the cap passes presented a
characterisation of the different HAZ regions demand a higher mean value of 357.4 HV. At the SC-CGHAZ registered in Fig. 7a,
magnification analysis. Be that as it may, to evaluate the effect it is possible to identify dark etched features, which resemble
of reheating on the microstructure, efforts were concentrated precipitates nucleated at the microstructure interfaces, such
on investigations close to the weld interface, seeking to obtain as lath, block and packet boundaries that may act as nucle-
information regarding the CGHAZ and its previously described ating agents.
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2 1721

Fig. 3 e Microhardness mapping over reheated HAZ of fill passes on samples of medium heat-input (a), low heat-input (b)
and high heat-input (c).

This precipitation is evident in the SEM image shown in intense decoration by coarse particles is observed at the
Fig. 8a, where a clear decoration can be observed at the packet boundaries and the PAG boundaries. This decoration
boundaries of the highlighted coarse martensite lath, as was then evaluated in SEM analysis, as shown in Fig. 8c,
pointed out by the red arrows, and Fig. 8b, which shows the identifying the presence of fine structures of fresh martensite
distribution of the precipitated particles throughout the pre- e indicated by the yellow arrows e and faceted particles of
viously mentioned microstructure interfaces, with the yellow retained/reversed austenite e indicated by blue and red ar-
arrows indicating the precipitation throughout the Prior rows. Those observations provide solid information for clas-
Austenite Grain (PAG) boundaries, while the red and blue ones sifying this region as an IC-CGHAZ, which experienced an
point to precipitates at the martensite packets and blocks in- intercritical reheating capable of partially austenitizing the
terfaces, respectively. Those particles at the SC-CGHAZ microstructure. The austenite mainly nucleates at the packet
mainly consist of carbides that may experience growth due and PAG boundaries, being in good agreement with works on
to the tempering effect resulting from the subcritical reheat intercritical heat treatments on lath martensite steels, which
[25]. Furthermore, intense carbide precipitation is also propose that the interface segregation of austenite stabilising
observed in the interior of the coarse laths on SC-CGHAZ, as elements (here C and Ni) along with the release of elastic
can be seen in Fig. 8a. Those nanometric carbides typically stress from martensite matrix can promote the austenite
extend in three orientation variants, agreeing with results reversion at martensite boundaries [36e38]. The austenite is
from Fonda and Spanos [11], which suggest that the coarse later submitted to high cooling rates, resulting in supersatu-
lath is initially formed as a supersaturated component and the rated fresh martensite and some amount of retained
precipitation of carbides takes place during subsequent ther- austenite. During holding at intercritical temperature, Ni can
mal activation, discarding the hypothesis of bainite coales- be rapidly segregated into austenite due to its higher diffu-
cence proposed by other authors [34,35]. sivity [39]. The Ni-enrichment is observed through Energy
The particle precipitation along the microstructure in- Dispersive X-ray Spectroscopy (EDS) analyses on points 1 and
terfaces is even more evident in the region registered in Fig. 6d 2 (coarse lath interior and retained/reversed austenite,
and 7b, which consists of the previously mentioned highly respectively) in Fig. 8d. The EDS spectrum of each point is
etched area. As pointed out in Fig. 7b, in LOM analysis, an presented in Fig. 9 and the semi-quantitative result of
1722 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2

Fig. 4 e Mean values and error bars from macrohardness measurements taken on the HAZ (a). Mean values and error bars
from microhardness measurements taken on the SC-CGHAZ (b), the IC-CGHAZ (c) and the SCR-CGHAZ (d) within the mapped
areas.

chemical composition in weight percentage is presented in values may be a consequence of the fresh martensite
Table 3, which reveals a significant increase in Ni content at nucleation.
the austenite particle located at the coarse martensite lath Continuing the microstructure scanning through the
interface. welding interface, Fig. 6e and 7c show LOM images over the
Also, the diffusive activity may be noted by analysing the hardest CGHAZ region, exhibiting an average microhardness
coarse laths interior. As one may see, in Fig. 8d, a lack of value of 335 HV e for the medium heat-input condition
precipitates is observed within the highlighted coarse laths, (Fig. 4d). Coming from the IC-CGHAZ, this region is strongly
which exhibit a featureless surface, contrasting with the characterised by an abrupt increase in microhardness values
previously presented coarse laths of the SC-CGHAZ. During which follows an evident change in microstructural features,
holding at intercritical temperature, the carbides precipitated mainly related to the refinement of martensite structures.
within the coarse martensite laths may experience dissolu- Fig. 8e shows a SEM image of this region, revealing a micro-
tion, and the C atoms will segregate to the coarse lath in- structure mainly composed of thin laths and finely dispersed
terfaces, acting on the stabilisation of austenite while precipitates located at the lath boundaries. Although the area
“cleaning” the interior of coarse laths. Thereby, the condition fraction of coarse lath is drastically decreased, Fig. 8f exposes
presented in Fig. 8d will be achieved, where the coarse laths a coarse martensite lath detected in this region, revealing
exhibit a featureless surface, and their boundaries are deco- nanometric carbides precipitation in the interior of this
rated by austenite particles. Compared to the SC-CGHAZ, a microconstituent, in a similar condition to the one observed
wide range of microhardness values is observed on IC-CGHAZ, on SC-CGHAZ. Following the previous classifications, it is ex-
presenting values from 248 HV to 289 HV for the medium heat- pected that this region would consist of a SCR-CGHAZ. It
input condition (Fig. 4c), where the higher microhardness makes sense that a low-supercritical reheat could achieve a

Fig. 5 e Microhardness profile obtained by following the vector displayed in Fig. 3a.
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2 1723

Fig. 6 e Overview of the mapped area in the medium heat-input sample (a) with an overlay of the microhardness map
introduced in Fig. 3a (b). Micrographs of the regions highlighted in Fig. 3a, revealing the reheated CGHAZ segmentation
(ceg).

refinement of the microstructure. Morsdorf et al. [25] reported microhardness values presented in Fig. 5. The same micro-
that low austenitization temperatures can substantially structure pattern is observed for all samples, not presenting
decrease the area fraction of coarse laths. The resulting regions that would be characteristic of an UA-CGHAZ. In fact,
refinement is reflected in the high microhardness values the only reheated CGHAZ region with microhardness values
observed in this area. Also, these results agree with the initial similar to the CGHAZ of cap passes is the SCR-CGHAZ, con-
analysis at the cap passes HAZ, which shows the FGHAZ e firming that the deposition pattern of the welding fill passes
region submitted to low-supercritical temperatures e as the was capable of avoiding the presence of UA-CGHAZ regions,
hardest HAZ region (Fig. 2). which are reported as LBZ in 9Ni reheated HAZ [12e14].
Fig. 6f is representative of a gradual transition zone, where As exhibited in Fig. 10, each CGHAZ region presents similar
higher supercritical temperatures will result in coarser microstructural features in all welding conditions. However,
martensite structures, decreasing the microhardness values. differences arise when evaluating the morphological aspects
Fig. 7def exhibit the microstructure evolution through this of the lath martensite microstructure, such as martensite
transition zone, where the registered regions correspond to blocks width and coarse laths area fraction.
the respective markers over the microhardness profile in Aiming to evaluate the variation in martensite block width,
Fig. 5. The progressive temperature increase is expected to each sample's CGHAZ microstructure was analysed via EBSD.
lead to an UA-CGHAZ microstructure. Instead, the effects of Fig. 11 exhibit, for example, IPF maps used to measure
the reheating promoted by the next welding pass were martensite block width along the CGHAZ of the low heat-
observed, resulting in a microstructure characteristic of the input sample. According to reports by lei Long et al. [40],
SC-CGHAZ, as shown in Fig. 6g and 7g. Continuing the high angle interface boundaries were reconstructed over the
microstructural analysis over the microhardness profile, maps: boundary orientation angles from 15 to 45 for PAGs,
Fig. 7f reveals the beginning of another IC-CGHAZ, exposing a from 45 to 55 for martensite packets, and from 55 to 65 for
periodic microstructural pattern related to the periodicity of martensite blocks, those boundaries are also highlighted in
1724 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2

Fig. 7 e CGHAZ LOM images. Micrographs were taken over the highlighted indents of the microhardness profile presented in
Fig. 5(aeh). The red-bordered geometries identify indents contained in the profile.

Fig. 11def. At least 40 martensite blocks were measured for equivalent widths for both zones. Also, we may notice that the
each region in each welding condition. application of higher heat-inputs can promote the obtainment
Fig. 12 shows the trend regarding the martensite block size of a coarser microstructure, represented by the increment in
on the evaluated samples through boxplots, which indicate width of the martensite blocks. An inverse correlation be-
the maximum and minimum values, in addition to the me- tween the coarser microstructure and the microhardness
dian and the first and the third quartiles, while the diamond values was found. As coarser as the microstructure is, it would
marker inside each box represents the mean values. It is exhibit lower microhardness, a relation observed while eval-
observed that the SCR-CGHAZ has the most refined micro- uating the increase in heat-input, which promotes the
structure among the CGHAZ regions while the SC-CGHAZ and decrease in microhardness values of all CGHAZ regions. Also,
IC-CGHAZ tend to exhibit larger martensite blocks, with when evaluating the different CGHAZ regions on the same
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2 1725

Fig. 8 e Low heat-input sample CGHAZ SEM images. Micrographs highlighting the microstructural features of the SC-CGHAZ
(aeb), the IC-CGHAZ (ced) and the SCR-CGHAZ (eef). The yellow dashed lines highlight coarse lath martensites. The red
arrows in Fig. 8a indicate precipitates at the coarse laths boundaries. The yellow, red and blue arrows in Fig. 8b indicate
precipitates throughout PAG boundaries, martensite packets and martensite blocks, respectively. The yellow, red and blue
arrows in Fig. 8c indicate fresh martensite and various retained austenite clusters, respectively.

sample, it is observed that the variation of microhardness outline the coarse martensite laths. The results obtained
values follows a strong relation with the martensite blocks through the quantification are summarised in Table 4, which
width: presenting considerably lower block width values, the reveals a significant decrease in the area fraction of coarse
SCR-CGHAZ exhibit much higher microhardness values, while laths for the SCR-CGHAZ. This behaviour certainly contributes
the coarser regions, composed by the SC-CGHAZ and the IC- to the higher microhardness values of this region in compar-
CGHAZ, are softer, with the SC-CGHAZ presenting slightly ison to the SC-CGHAZ and IC-CGHAZ, which present similar
wider martensite blocks, and also slightly lower microhard- values of coarse lath area fraction on each of the evaluated
ness values. conditions. Furthermore, it can also be noticed that increasing
Regarding the area fraction of coarse martensite laths, the heat-input enlargers the fractions of coarse lath in all of
multiple high-resolution SEM micrographs in secondary the CGHAZ zones. This outcome agrees with the results of
electron mode were evaluated, resulting in the cover of at Morsdorf et al. [25], which suggested that higher austenitiza-
least 22,000mm2, in each CGHAZ region, for the quantification tion temperatures would guarantee a softer austenite matrix
of the area fractions. As an example of the evaluated images, that imposes less resistance to the coarse laths growth,
Fig. 13 exhibits micrographs of the three CGHAZ regions for resulting in microstructures with larger coarse lath area
the high heat-input welding condition, where the dashed lines fraction.
1726 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2

Fig. 9 e EDS spectra of analyses on coarse lath interior (a) and retained/reversed austenite particle (b). Points 1 and 2 in
Fig. 8d, respectively.

This carbon redistribution will reduce the C supersaturation in

4. Discussion the vicinity of dislocation cores which act as carbon sinks. To
relate the chemical composition and the solid solution
By relating the microhardness values to the yield strength, we strengthening in lath martensite steels is a complex task. In
may discuss the observed microstructural features’ contri- fact, recent reports on strengthening mechanisms of these
bution to the local mechanical properties of the CGHAZ re- alloys tends to concentrate efforts in the discussion of grain
gions. In steels, the yield strength (sy) can be evaluated as a size and dislocation strengthening [46].
sum of strengthening mechanism which contributes to hin- The HallePetch equation has been commonly used to ex-
dering dislocation slip: press the effect of grain size on steel strengthening:
sy ¼ s0 þ sss þ sgs þ sdis þ sppt ; (1)
sgs ¼ Khp d
1=2 ; (2)
where s0 is the intrinsic lattice friction of pure iron, and sss, sgs,
where Khp is the HallePetch coefficient and d is the mean value
sdis and sppt are the strengthening from solid solution, grain
of the effective grain size. The effective grain size in lath
size, dislocation and precipitation, respectively [41,42].
martensite steels is usually attributed to the block size since
Solid solution strengthening may be closely related to the
the block boundaries are considered the smallest crystallo-
type and the concentration of the alloying elements [43].
graphic discontinuity able to pin or significantly impair the
However, the complexity of the microstructure and the
dislocation mobility, following a consistent HallePetch rela-
composition of lath martensitic steels make it challenging to
tionship [40,47]. Also, it is important to notice that the block
clarify this hardening mechanism. It is known that the me-
size does not follow a simple dependence on the PAG size,
chanical properties of martensitic steels are significantly
being also affected by alloying elements, such as Mn, that may
impacted by the distribution of C atoms in the microstructure
impact the subdivision of martensite packets into blocks [48].
and the eventual precipitation of carbides during the
Thus, although the PAG size is an important feature in the
tempering process [44]. Exhibiting high martensite start tem-
general analysis of HAZ microstructure, it may blur some as-
perature e approximately 330  C [11] e, 9Ni steels undergo an
pects of the microstructure-strengthening relationship.
unavoidable autotempering during quenching, which is
As shown in Fig. 12, martensite block widths along the
especially observed at the coarse laths, microconstituents that
CGHAZ influence the local mechanical properties. The block
form close to the martensite start temperature [25]. According
refinement promoted by the superposition by FGHAZ or by
to Morsdorf et al. [45], this phenomenon induces the decom-
applying a lower heat-input may increase the microhardness
position of the Zener-ordered supersaturated lattice, leading
values of the HAZ. Although the HallePetch equation de-
to agglomeration of C atoms in Body-Centered Tetragonal
scribes the connection between the effective grain size and
(BCT) clusters within the Body-Centered Cubic (BCC) matrix.
the yield strength, it is recognised that the hardness measures
These clusters may further evolve into transition carbides and
a material's plastic flow resistance and is, therefore, a function
precipitated cementite, with the C atoms diffusing and
of yield strength and strain hardening. Considering the pro-
agglomerating along dislocations to form cementite nuclei.
portionality between microhardness and yield strength, it is
possible to evaluate that the results initially presented in
Table 3 e EDS semi-quantitative results of analyses taken Fig. 12 follow a strong HallePetch relationship, as exhibited in
over the points indicated in Fig. 8d. Fig. 14a. The variation in martensite block width promoted by
differences in welding pass superpositions and heat-input
Element Weight percentage (%)
conditions draws a consistent HallePetch relationship which
Point 1 Point 2
corroborates with the results of studies which has performed
Si 0.30 0.26 actual uniaxial tensile experiments to verify the effects of the
Fe 89.08 82.72 lath martensite hierarchical microstructure on the mechani-
Ni 10.62 17.03
cal properties of lath martensite steels [40,47,49].
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2 1727

Fig. 10 e SEM micrographs registering, respectively, the SC-CGHAZ, the IC-CGHAZ and the SCR-CGHAZ in the samples of low
(aec), medium (def) and high (gei) welding heat-input.

Fig. 11 e IPF maps of the SC-CGHAZ (a), the IC-CGHAZ (b), and the SCR-CGHAZ (c) for the low heat-input welding condition.
The grain boundaries of each map are shown in Fig. 11def.
1728 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2

Fig. 12 e Boxplots representing the martensite blocks width distribution on the different CGHAZ regions, and
microhardness values tendency between those regions. The blue cross markers represent outliers. Measurements taken
over the samples of low (a), medium (b) and high (c) heat-inputs.

Regarding dislocation strengthening, this mechanism may lath growth during the subsequent cooling, resulting in the
be partially related to the morphological aspects of the lath low coarse lath area fraction values observed in Table 4. The
martensite microstructure. Morsdorf et al. [25] developed welding heat-input effect is also observed in the values pre-
extensive characterisation work towards the lath martensite sented in Table 4. As expected, a higher heat-input will pro-
constituents, identifying the coarse laths as a low dislocation mote the formation of larger PAGs at the CGHAZ, increasing
density zones, which presents itself as “soft” micro- the coarse lath area fraction on the final microstructure. This
constituents dispersed inside a much harder matrix effect seemingly extends to the SCR-CGHAZ, which exhibited
composed of thin martensite laths. The authors formulate a significant increase in coarse lath area fraction for the high
that this condition is a result of the martensite transformation heat-input condition.
sequence, in which the coarse laths are the first martensite Fig. 14b shows the relationship between the local micro-
constituents to grow into the initially “soft” austenite grain. As hardness and the coarse lath area fraction, exhibiting values
they grow and the temperature decreases, the significant in- for all of the evaluated CGHAZ regions in each welding con-
crease in dislocation density of the highly constrained un- dition. Considering the same CGHAZ region at the different
transformed austenite will inhibit further coarse lath growth welding conditions, it is verified that the increase in heat-
e which will undergo an autotempering process e and catal- input promotes the increase in coarse lath area fraction,
yse the nucleation of thin martensite laths, that will exhibit which is followed by the decrease in dislocation density and
significantly higher dislocation density in comparison to the microhardness values. Also, when comparing the different
coarse martensite laths [25,50]. Morsdorf et al. [25] also CGHAZ regions in the same heat-input condition, it is
observed that the coarse lath area fraction increases with the observed that the lower microhardness values are obtained in
PAG size, a factor related to the increase in martensite start the regions with a larger area fraction of coarse lath. As
temperature, which contributes to the lower resistance of the exhibited in Fig. 14b, this relationship follows a strong linear
initial austenite structure. These observations suggest that tendency; however, it is difficult to measure the impact of the
the CGHAZ would present the highest coarse lath area fraction coarse lath at the local mechanical properties, as its area
along the HAZ, and this is, in fact, observed, with the SC- fraction increment is usually followed by an increase in
CGHAZ and the IC-CGHAZ exhibiting higher values of coarse martensite block size. Both observations seem to promote a
lath area fraction e thus exhibiting a larger area with low decrease in microhardness values, indicating that grain size
dislocation density and lower microhardness values e for all and dislocation strengthening significantly contribute to the
welding conditions (Table 4). However, those high fractions local mechanical properties.
are not observed at the harder SCR-CGHAZ, which corre- Besides the morphological aspects, the precipitation
sponds to a CGHAZ region submitted to reheat at low super- pattern also presents itself as a microstructural feature
critical temperatures, promoting the complete austenitization capable of promoting the observed heterogeneity in HAZ
and the refinement of this zone. The refined austenite grains microhardness values. When comparing Fig. 8c and d to
obtained during reheating will hinder the coarse martensite Fig. 8a and b and Fig. 8e and f, the effects of an intercritical
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2 1729

Fig. 13 e Large field of view SEM images in secondary electron mode highlighting the coarse laths in SC-CGHAZ (a), IC-
CGHAZ (b) and SCR-CGHAZ (c) for the high heat-input condition.

reheating on the precipitation interlath and intralath steels that have characterised the observed constituents as
martensite structure are clearly observed. In the martensite retained/reverse austenite particles [16,19,21,51e53]. Nonethe-
transformation sequence, the coarse laths are the first con- less, those austenite structures are also quite similar to the
stituents to transform from the prior austenite grain, being microconstituents characterised by Ramachandran et al. [54] as
subjected to an auto-tempering process during the subse- necklace-type M-A (red arrows in Fig. 8c) and slender-type M-A
quent cooling. As a result of this low range diffusion process, constituents (blue arrows in Fig. 8c). These M-A constituents are
nanocarbides precipitate in the interior of the initially super- considered deleterious structures capable of significantly
saturated coarse martensite lath, as exhibited in Fig. 8a and f. reducing the toughness due to the significant hardness difference
These microstructural features agree with previous studies between the M-A and the surrounding matrix microstructure,
performed by Morsdorf et al. [25]. When this constituent is acting on crack initiation mechanisms [54,55]. This assumption
submitted to reheating at intercritical temperature, the energy agrees with previous reports that have characterised the IC-
provided is high enough to promote the dissolution of these CGHAZ as LBZ within the HAZ of 9Ni steel welded joints,
carbides and the C atoms diffusion to the coarse laths relating this observation to the chemical composition of M-A
boundary. The accumulation of carbon along with these constituents [12,13].
boundaries leads to the carbon clustering, which subse- Although higher resolution microscopy is needed to
quently can nucleate reverted austenite. Additionally, the adequately characterise these constituents, the presence of
segregation of C atoms to dislocations and boundaries can deleterious M-A constituents in 9Ni steel welded joints is
fastly diffuse it towards the retained austenite, previously already documented [12,13]. It is believed that the austenite C-
formed under the quench. This carbon enrichment will act on enrichment promoted by the intercritical reheat is responsible
stabilising reverted and retained austenite, resulting in a for forming these brittle constituents. However, a detailed APT
seemingly featureless coarse lath surrounded by a micro- analysis is needed to verify the previously described C-
structure containing C-rich retained/reverted austenite par- enrichment phenomenon.
ticles, as registered in Fig. 8c and d. Also, regarding the intercritical reheating effects over the
This microstructural landscape described here for the 9Ni steel lath martensite microstructure, it is clear that the partial
IC-CGHAZ has some similarities with studies on lath martensitic austenitization is responsible for the precipitation of fresh
martensite along the PAG and packet boundaries, as indicated
by the yellow arrows in Fig. 8c. The presence of these hard
Table 4 e Mean values of coarse lath area fraction.
supersaturated martensite constituents may explain the high
SC-CGHAZ IC-CGHAZ SCR-CGHAZ deviation of microhardness values observed in the IC-CGHAZ
Low heat-input 5.6% 5.3% 1.4% for all the heat-input conditions, resulting in the slightly
Medium heat-input 6.1% 6.5% 1.9% higher mean microhardness value in comparison to the SC-
High heat-input 7.4% 7.2% 3.6% CGHAZ.
1730 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2

Fig. 14 e Plot of microhardness vs reciprocal square root block width (a) and plot of microhardness vs coarse lath are fraction
(b) at the CGHAZ regions in each evaluated condition.

Concerning the SCR-CGHAZ, the refined microstructure microconstituents are characterised as “soft” zones of low
impacts the dispersion of precipitates, as the precipitation dislocation density.
mainly occurs at the microstructural interfaces [44,50]. Be that  Although it is expected that the precipitates would act in
as may, the precipitation hardening may be blurred by the the alloy strengthening, the evaluation of this hardening
grain size hardening. Also, the proper evaluation of the mechanism is hindered by the precipitation pattern at the
hardening promoted by the nanocarbides located within the microstructure interfaces, which may blur the line be-
coarse laths demands high spatial resolution mechanical tween grain size strengthening and precipitation
testing to probe the microconstituents. Nanoindentation has strengthening. Nonetheless, the effect of the precipitates
been the favored technique to evaluate the carbon may be observed at the IC-CGHAZ, where the intercritical
redistribution-induced softening [25,56]. reheating promotes the dissolution of the nanocarbides
within coarse laths and may contribute to the of C-
enrichment of retained/reversed austenite particles and
5. Conclusion the formation of M-A constituents.

The present work conducted a microstructural and micro-

mechanical investigation along the HAZ of 9Ni steel pipes
multi-pass girth welds, revealing features related to reheating Data availability
cycles and heat-input conditions. This study yielded the
following conclusions. The data that support the findings of this study are available
from the corresponding author, upon reasonable request.
 The reheating cycles imposed by subsequent welding
passes are the most impacting factors on the local micro-
hardness, standing out in comparison to the impact of the Declaration of Competing Interest
prior microstructure, resulting in the obtaining of hard and
soft layers that reach out the entire radial extension of the The authors declare that they have no known competing
HAZ. financial interests or personal relationships that could have
 The highest microhardness values, for all welding condi- appeared to influence the work reported in this paper.
tions, are founded at the SCR-CGHAZ e agreeing with the
analysis at the cap passes, which exhibit the FGHAZ as the
hardest HAZ region e, while the SC-CGHAZ and the IC-
Acknowledgements
CGHAZ exhibit lower microhardness values, although the
later presents a high deviation that is probably related to
The authors would like to thank the Laborato
rio de Pesquisa e
the presence of hard fresh martensite and possible M-A

Tecnologia em Soldagem (LPTS) of the Universidade Federal do Ceara
constituents resulted from C-enrichment of austenite
(UFC); the Central Analı́tica-UFC (funded by FINEP-CT-INFRA,
particles at the intercritically reheated zone.
CAPES-Pr
o-Equipamentos, and MCTI-CNPq-SisNano2.0); the
 The variation in the mechanical response of the CGHAZ
Ag^encia Nacional do Petr
oleo, Ga

s Natural e Biocombustı́veis (ANP),
regions is mainly related to the grain size strengthening, as
the Financiadora de Estudos e Projetos (FINEP) and the Minist
erio da
the martensite block refinement promoted by reheating
Ci^encia, Tecnologia e Inovaça
~o (MCTI) for the financial support
and by heat-input condition follows a strong HallePetch
through the Programa de Recursos Humanos da ANP para o Setor
relation, and to the coarse lath area fraction, as these
Petr
oleo e Ga

s (PRH-ANP/MCTI). The authors would also like to
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2 1731

thank the Petr

oleo Brasileiro S. A. (Petrobras) for the donation of Mater Sci Eng, A 2003;340:68e79. https://doi.org/10.1016/
the 9Ni steel pipes. This study was financed in part by the S0013-7944(02)00111-X.
~o de Aperfeiçoamento de Pessoal de Nı́vel Superior e
Coordenaça [15] Marshall CW, Hehemann RF, Troiano AR. The characteristics
of 9% nickel low carbon steels. Transactions of American
Brasil (CAPES) e Finance Code 001.
Society for Metals 1962;55:135.
[16] Zhang JM, Li H, Yang F, Chi Q, Ji LK, Feng YR. Effect of heat
treatment process on mechanical properties and
references
microstructure of a 9% Ni steel for large LNG storage tanks. J
Mater Eng Perform 2013;22:3867e71. https://doi.org/10.1007/
s11665-013-0701-1.
[1] Hedberg HD, Moody JD, Hedberg RM. Petroleum prospects of [17] Hu J, Du L-X, Xu W, Zhai J-H, Dong Y, Liu Y-J, et al. Ensuring
the deep offshore. AAPG (Am Assoc Pet Geol) Bull combination of strength, ductility and toughness in
1979;63:286e300. https://doi.org/10.1306/C1EA55F8-16C9- medium-manganese steel through optimization of nano-
11D7-8645000102C1865D. scale metastable austenite. Mater Char 2018;136:20e8.
[2] Duhen T, Sagot A, Kerbart Y. Deep offshore slim hole drilling. https://doi.org/10.1016/j.matchar.2017.11.058.
In: 1997 offshore technology conference. Houston, TX: [18] Kuzmina M, Ponge D, Raabe D. Grain boundary segregation
Offshore Technology Conference; 1997. https://doi.org/ engineering and austenite reversion turn embrittlement into
10.4043/8557-MS. toughness: example of a 9wt.% medium Mn steel. Acta Mater
[3] Cezar AP, Pereira AR, Daniel LL, Oazen E, Trovoado L, de 2015;86:182e92. https://doi.org/10.1016/
Almeida MC. Subsea solutions in the pre-salt development j.actamat.2014.12.021.
projects. In: 2015 offshore technology conference. Houston, [19] Zhang XJ. Microhardness characterisation in developing high
TX: Offshore Technology Conference; 2015. https://doi.org/ strength, high toughness and superior ballistic resistance
10.4043/25773-MS. low carbon Ni steel. Mater Sci Technol 2012;28:818e22.
[4] Beltra~ o RLC, Sombra CL, Lage ACVM, Netto JRF, https://doi.org/10.1179/1743284712Y.0000000015.
Henriques CCD. Challenges and new technologies for the [20] Isheim D, Hunter AH, Zhang XJ, Seidman DN. Nanoscale
development of the presalt cluster. In: 2009 offshore analyses of high-nickel concentration martensitic high-
technology conference. Houston, TX: Offshore Technology strength steels. Metall Mater Trans 2013;44:3046e59. https://
Conference; 2009. https://doi.org/10.4043/19880-MS. doi.org/10.1007/s11661-013-1670-6.
[5] Ding M, Yuan F, Wang Y, Xia X, Chen W, Liu D. Oil recovery [21] Jain D, Seidman DN, Barrick EJ, DuPont JN. Atom-probe
from a CO2 injection in heterogeneous reservoirs: the tomographic investigation of austenite stability and carbide
influence of permeability heterogeneity, CO2-oil miscibility precipitation in a TRIP-assisted 10 wt pct Ni steel and its
and injection pattern. J Nat Gas Sci Eng 2017;44:140e9. weld heat-affected zones. Metall Mater Trans
https://doi.org/10.1016/j.jngse.2017.04.015. 2018;49:1031e43. https://doi.org/10.1007/s11661-018-4470-1.
[6] Eldevik F, Graver B, Torbergsen LE, Saugerud OT. [22] Olson GB, Azrin M. Transformation behavior of TRIP steels.
Development of a guideline for safe, reliable and cost Metall Trans A 1978;9:713e21. https://doi.org/10.1007/
efficient transmission of CO2 in pipelines. Energy Proc BF02659928.
2009;1:1579e85. https://doi.org/10.1016/j.egypro.2009.01.207. [23] Barrick EJ, DuPont JN. Mechanical properties and
[7] Knoope MMJ, Ramı́rez A, Faaij APC. A state-of-the-art review microstructural characterization of simulated heat-affected
of techno-economic models predicting the costs of CO2 zones in 10 wt pct Ni steel. Mater Sci Eng, A
pipeline transport. Int J Greenh Gas Control 2013;16:241e70. 2019;748:189e204. https://doi.org/10.1016/j.msea.2019.01.085.
https://doi.org/10.1016/j.ijggc.2013.01.005. [24] Morsdorf L, Jeannin O, Barbier D, Mitsuhara M, Raabe D,
[8] Scheid A, Fe
lix LM, Martinazzi D, Renck T, Kwietniewski CEF. Tasan CC. Multiple mechanisms of lath martensite plasticity.
The microstructure effect on the fracture toughness of Acta Mater 2016;121:202e14. https://doi.org/10.1016/
ferritic Ni-alloyed steels. Mater Sci Eng, A 2016;661:96e104. j.actamat.2016.09.006.
[9] Shin H-S, Lee H-M, Kim M-S. Impact tensile behaviors of 9% [25] Morsdorf L, Tasan CC, Ponge D, Raabe D. 3D structural and
nickel steel at low temperature. Int J Impact Eng atomic-scale analysis of lath martensite: effect of the
2000;24:571e81. https://doi.org/10.1016/S0734-743X(99) transformation sequence. Acta Mater 2015;95:366e77.
00181-5. https://doi.org/10.1016/j.actamat.2015.05.023.
[10] Cipriano Farias FW, Payao Filho JdC, Barreto de Azevedo LM. [26] Jang J-I, Yang Y-C, Kim W-S, Kwon D. Evaluation of cryogenic
Microstructural and mechanical characterization of the fracture toughness in SMA-welded 9% Ni steels through
transition zone of 9% Ni steel cladded with Ni-based modified CTOD test. Met Mater 1997;3:230e8. https://doi.org/
superalloy 625 by GTAW-HW. Metals 2018;8:1007. https:// 10.1007/BF03025929.
doi.org/10.3390/met8121007. [27] Jang J-I, Yang Y-C, Kim W-S, Kwon D. A study of fracture
[11] Fonda RW, Spanos G. Effects of cooling rate on toughness and microstructures in the weld heat-affected
transformations in a Fe-9 pct Ni steel. Metall Mater Trans zone of QLT-processed 9% Ni steel. In: Balachandran UB,
2014;45:5982e9. https://doi.org/10.1007/s11661-014-2588-3. Gubser DG, Hartwig KT, Reed RP, Warnes WH, Bardos VA,
[12] Jang J-I, Lee B-W, Ju J-B, Kwon D, Kim W-S. Crack-initiation editors. Advances in cryogenic engineering materials.
toughness and crack-arrest toughness in advanced 9 pct Ni Boston, MA: Springer; 1998. p. 41e8.
steel welds containing local brittle zones. Metall Mater Trans [28] de Miranda HC, Silva CC. Estudo da Soldabilidade de Aços
2002;33:2615e22. https://doi.org/10.1007/s11661-002-0383-z. Inoxida
veis Duplex e Superduplex: relato
rio Te

cnico LPTS-
[13] Jang J-I, Lee B-W, Ju J-B, Kwon D, Kim W-S. Experimental CNPEM-9%Ni. Fortaleza, CE: LPTS/DEMM/UFC; 2016.
analysis of the practical LBZ effects on the brittle fracture Technical Report 3.
performance of cryogenic steel HAZs with respect to crack [29] Kou S. Welding metallurgy. 2 ed. Hoboken, NJ: John Wiley &
arrest toughness near fusion line. Eng Fract Mech Sons; 2003.
2003;70:1245e57. [30] Lippold JC. Welding metallurgy and weldability. Hoboken, NJ:
[14] Jang J-I, Ju J-B, Lee B-W, Kwon D, Kim W-S. Effects of John Wiley & Sons; 2015.
microstructural change on fracture characteristics in coarse- [31] Bhadeshia HKDH, Honeycombe R. Steels: microstructure and
grained heat-affected zones of QLT-processed 9% Ni steel. properties. 3 ed. Oxford: Butterworth-Heinemann; 2017.
1732 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 3 ; 2 4 : 1 7 1 6 e1 7 3 2

[32] Fernandes VS, Farias FWC, da Cruz Paya ~ o Filho J. Fracture quenched martensites (0.1e0.5%C). Acta Mater
toughness of a 9% Ni steel pipe girth welded with Ni-based 2011;59:5845e58. https://doi.org/10.1016/
superalloy 625 filler metal operating in a sour environment. J j.actamat.2011.05.061.
Mater Res Technol 2020;9:6305e21. https://doi.org/10.1016/ [45] Morsdorf L, Emelina E, Gault B, Herbig M, Tasan C. Carbon
j.surfcoat.2019.06.084. redistribution in quenched and tempered lath martensite.
[33] Kim BC, Lee S, Kim NJ, Lee DY. Microstructure and local Acta Mater 2021;205:116521. https://doi.org/10.1016/
brittle zone phenomena in high-strength low-alloy steel j.actamat.2020.116521.
welds. Metall Trans A 1991;22:139e49. https://doi.org/ [46] Liu H, Shang X, He B, Liang Z. Strain rate dependence of
10.1007/BF03350956. strengthening mechanisms in ultrahigh strength lath
[34] Bhadeshia HKDH, Keehan E, Karlsson L, Andre
n H-O. martensite. Int J Plast 2023;161:103495. https://doi.org/
Coalesced bainite. Trans Indian Inst Met 2006;59:689e94. 10.1016/j.ijplas.2022.103495.
[35] Keehan E, Karlsson L, Bhadeshia HKDH, Thuvander M. [47] Morris JW. On the ductile-brittle transition in lath
Three-dimensional analysis of coalesced bainite using martensitic steel. ISIJ Int 2011;51:1569e75. https://doi.org/
focused ion beam tomography. Mater Char 2008;59:877e82. 10.2355/isijinternational.51.1569.
https://doi.org/10.1016/j.matchar.2007.07.011. [48] Morito S, Yoshida H, Maki T, Huang X. Effect of block size on
[36] Hu J, Zhang J-M, Sun G-S, Du L-X, Liu Y, Dong Y, et al. High the strength of lath martensite in low carbon steels. Mater
strength and ductility combination in nano-/ultrafine-grained Sci Eng, A 2006;438e440:237e40. https://doi.org/10.1016/
medium-Mn steel by tuning the stability of reverted austenite j.msea.2005.12.048.
involving intercritical annealing. J Mater Sci 2019;54:6565e78. [49] Luo H, Wang X, Liu Z, Yang Z. Influence of refined
https://doi.org/10.1007/s10853-018-03291-w. hierarchical martensitic microstructures on yield strength
[37] Raabe D, Sandlo € bes S, Milla
n J, Ponge D, Assadi H, Herbig M, and impact toughness of ultra-high strength stainless steel. J
et al. Segregation engineering enables nanoscale martensite Mater Sci Technol 2020;51:130e6. https://doi.org/10.1016/
to austenite phase transformation at grain boundaries: a j.jmst.2020.04.001.
pathway to ductile martensite. Acta Mater 2013;61:6132e52. [50] Shamsujjoha M. Evolution of microstructures, dislocation
https://doi.org/10.1016/j.actamat.2013.06.055. density and arrangement during deformation of low carbon
[38] Raabe D, Herbig M, Sandlo € bes S, Li Y, Tytko D, Kuzmina M, lath martensitic steels. Mater Sci Eng, A 2020;776:139039.
et al. Grain boundary segregation engineering in metallic https://doi.org/10.1016/j.msea.2020.139039.
alloys: a pathway to the design of interfaces. Curr Opin Solid [51] Fultz B, Kim JI, Kim YH, Kim HJ, Fior GO, Morris JW. The
State Mater Sci 2014;18:253e61. https://doi.org/10.1016/ stability of precipitated austenite and the toughness of 9Ni
j.cossms.2014.06.002. steel. Metall Trans A 1985;16:2237e49. https://doi.org/
[39] Yang Y-H, Cai Q-W, Tang D, Wu H-B. Precipitation and 10.1007/BF02670423.
stability of reversed austenite in 9Ni steel. Int J Miner Metall [52] Fultz B, Kim JI, Kim YH, Morris JW. The chemical
Mater 2010;17:587e95. https://doi.org/10.1007/s12613-010- composition of precipitated austenite in 9Ni steel. Metall
0361-1. Trans A 1986;17:967e72. https://doi.org/10.1007/BF02661262.
[40] lei Long S, long Liang Y, Jiang Y, Liang Y, Yang M, liang Yi Y. [53] Jain D, Isheim D, Zhang XJ, Ghosh G, Seidman DN. Thermally
Effect of quenching temperature on martensite multi-level stable Ni-rich austenite formed utilizing multistep
microstructures and properties of strength and toughness in intercritical heat treatment in a low-carbon 10 wt pct Ni
20CrNi2Mo steel. Mater Sci Eng, A 2016;676:38e47. https:// martensitic steel. Metall Mater Trans 2017;48:3642e54.
doi.org/10.1016/j.msea.2016.08.065. https://doi.org/10.1007/s11661-017-4146-2.
[41] Yu L, Xiao X, Chen L, Cheng Y, Duan H. A hierarchical [54] Ramachandran DC, Kim S-D, Moon J, Lee C-H, Chung J-H,
theoretical model for mechanical properties of lath Biro E, et al. Classification of martensite-austenite
martensitic steels. Int J Plast 2018;111:135e51. https:// constituents according to its internal morphology in high-
doi.org/10.1016/j.ijplas.2018.07.012. strength low alloy steel. Mater Lett 2020;278:128422. https://
[42] Hou W, Liu Q, Wen H, Gu J. Effect of cyclic intercritical doi.org/10.1016/j.matlet.2020.128422.
tempering on the microstructure and mechanical properties [55] Lee SG, Sohn SS, Kim B, Kim WG, Um K-K, Lee S. Effects of
of a low-carbon Cu-bearing 7Ni steel. Metall Mater Trans martensite-austenite constituent on crack initiation and
2020;51:3981e95. https://doi.org/10.1007/s11661-020-05811-3. propagation in inter-critical heat-affected zone of high-
[43] Xiao X, Terentyev D, Yu L, Bakaev A, Jin Z, Duan H. strength low-alloy (HSLA) steel. Mater Sci Eng, A
Investigation of the thermo-mechanical behavior of 2018;715:332e9. https://doi.org/10.1016/j.msea.2018.01.021.
neutron-irradiated Fe-Cr alloys by self-consistent plasticity [56] Morsdorf L, Kashiwar A, Kübel C, Tasan C. Carbon
theory. J Nucl Mater 2016;477:123e33. https://doi.org/10.1016/ segregation and cementite precipitation at grain
j.jnucmat.2016.05.012. boundaries in quenched and tempered lath martensite.
[44] Hutchinson B, Hagstro € m J, Karlsson O, Lindell D, Tornberg M, Mater Sci Eng, A 2023;862:144369. https://doi.org/10.1016/
Lindberg F, et al. Microstructures and hardness of as- j.msea.2022.144369.