materials

Article Investigation on Microstructure and Properties of Duplex Welds by Underwater with Different

Kai Wang 1, Changlei Shao 2, Xiangdong Jiao 1,3, Jialei Zhu 3,*, Zhihai Cai 4 and Congwei Li 3

1 School of Mechanical Engineering, Beijing University of Chemical Technology, Beijing 100029, China; [email protected] (K.W.); [email protected] (X.J.) 2 Shanghai Nuclear Engineering Research & Design Institute, Shanghai 200233, China; [email protected] 3 Beijing Institute of Petrochemical Technology, Beijing 102617, China; [email protected] 4 Army Academy of Armored , Beijing 100072, China; [email protected] * Correspondence: [email protected]

Abstract: Taking S32101 duplex stainless steel as the research object, underwater laser wire filling welding technology was used for U-groove filling welding. The influence of different shielding gas compositions on the ferrite content, microstructure, mechanical properties and pitting corrosion resistance was studied by simulating a water depth of 15 m in the hyperbaric chamber. The results show that, under the same process parameters, the size and proportion of austenite in the weld when using pure as the shielding gas are larger than those protected by other shielding gases. In a mixed shielding gas, the increase in nitrogen content has little effect on the strength and toughness  of the weld. Regardless of the shielding gas used, the base metal was the weakest part of the weld.  At the same time, intermetallic inclusions have an adverse effect on the impact toughness of the weld. Citation: Wang, K.; Shao, C.; Jiao, X.; The pitting corrosion resistance of the welds depends on the Cr2N content in the heat-affected zone. Zhu, J.; Cai, Z.; Li, C. Investigation on The precipitation and enrichment of Cr2N causes local deficiency, which is the main factor Microstructure and Properties of for the weak pitting corrosion ability of the heat-affected zone. Pure nitrogen protection has a better Duplex Stainless Steel Welds by corrosion resistance than other gas protection. Underwater Laser Welding with Different Shielding Gas. Materials Keywords: duplex stainless steel; underwater laser wire filling welding; hyperbaric chamber; 2021, 14, 4774. https://doi.org/ intermetallic inclusions; pitting corrosion resistance 10.3390/ma14174774

Academic Editor: Filippo Berto

Received: 1 August 2021 1. Introduction Accepted: 18 August 2021 Duplex stainless steel (DSS) has been widely developed in recent years because of Published: 24 August 2021 its good corrosion resistance and economic benefits. Due to both its dual-phase equilib- rium microstructure of ferrite and austenite and high content of alloy elements, DSS has Publisher’s Note: MDPI stays neutral a high strength and toughness and excellent corrosion resistance [1,2]. Therefore, DSS is with regard to jurisdictional claims in increasingly used as structural materials in nuclear power plants, offshore engineering published maps and institutional affil- construction and the petrochemical industry [3]. In order to reduce the cost of operation iations. and maintenance, underwater welding is usually used to repair cracks and surface corro- sion defects in nuclear power plants. Underwater welding is mainly divided into three methods: underwater dry welding, underwater wet welding and underwater local dry welding [4–7] Underwater dry welding has good welding quality, but the welding process Copyright: © 2021 by the authors. is complex, resulting in a significant increase in cost; the underwater wet welding device Licensee MDPI, Basel, Switzerland. is simple, but it cannot guarantee the welding quality; local dry welding combines the This article is an open access article advantages of the first two methods, which not only ensures the welding quality, but also distributed under the terms and reduces the equipment cost. It is very suitable for the construction [8] and maintenance of conditions of the Creative Commons underwater nuclear power plants [4]. Attribution (CC BY) license (https:// As a new welding technology, underwater laser welding (ULW) has gradually become creativecommons.org/licenses/by/ the most promising maintenance technology for nuclear power plant underwater welding. 4.0/).

Materials 2021, 14, 4774. https://doi.org/10.3390/ma14174774 https://www.mdpi.com/journal/materials Materials 2021, 14, 4774 2 of 16

Feng et al. studied the effects of protective materials and water on the repair quality of Ni-Al bronze by underwater laser welding [9]. Ning G. et al. proved the effect of water depth on the weld quality and welding process in underwater laser welding [10]. At the same time, Ning G. carried out welding experiments on a Ti-6Al-4V alloy with filler wire underwater laser beam welding (ULBW) technology, and obtained joints close to the welding quality in air [11]. Tomków et al. completed the butt welding experiment of high- strength low-alloy s460n steel by using a local cavity method, and tested the mechanical properties of the welded joints in order to obtain a high quality [12]. Based on the existing research results, the microstructure and properties of underwater laser welding of duplex stainless steel were studied using the local cavity method. In DSS laser welding, the proportion of the equilibrium phase is very important; the equal fraction of ferrite and austenite is the best combination of mechanical properties and corrosion resistance [13]. ULW technology has a high applicability in the field of nuclear power plant repair. However, some welding metallurgical challenges limit its applications in underwater maintenance. The metallurgical solidification of DSS is fully ferritic, followed by the transformation from solid-state ferrite to austenite controlled by thermal cycle . When laser welding is carried out in an , the rapid cooling rate largely suppresses the formation of austenite and restricts the optimum phase equilibrium in DSS. Nitrogen is a strong austenite- element because of its high diffusion rate [14,15]. When using ULW, the rapid heating and cooling cycle restricts the austenite formation and disturbs the optimum phase equilibrium in DSS, resulting in nitrogen atom loss and a high ferrite content [16], which is not conducive to the pitting corrosion resistance of welded joints. The decrease in nitrogen content is not limited to weld metal (WM): the heat-affected zone (HAZ) is also affected by the nitrogen diffusion from HAZ to WM [17–19]. According to Sakai et al., the austenite content of DSS should be higher than 50% in order to obtain high impact toughness [20]. Miura and Ogawa also pointed out that the lowest pitting corrosion rate is also associated with an austenite content of 50% or more [21]. In this paper, the effects of pure argon, pure nitrogen and different proportions of the nitrogen–argon mixture as a shielding gas on the mechanical properties and pitting corrosion resistance of DSS welded joints were studied, and the influence mechanism was analyzed.

2. Experimental Procedure 2.1. Equipment and Materials The ULW was performed using RCL-C6000 fiber laser of Ruike (specifications given in Table1), with a maximum output power of 6 kW and wavelength of 1070 nm. As shown in Figure1, the experimental system is mainly composed of two parts: the hyperbaric experimental chamber and the ULW experimental platform, which can meet the require- ments of rapid and accurate positioning of underwater welding and simulation of environment. During the welding process, the pressure of the cabin was kept at 0.15 MPa by filling compressed gas to simulate the water depth of 15 m. The wet underwater en- vironment was simulated by injecting water into the experimental tank, and the water depth in the tank was 30 mm. A sealed underwater laser head and local drainage device were developed. The laser beam was transmitted to the local dry space inside the local drainage device through optical fiber, and the ULW repair was realized by laser wire filling technology. Figure2a,b was underwater laser welding test platform and hyperbaric test chamber respectively. S32101 DSS was used as the base metal, and the dimensions of test plates were prepared with 600 mm × 300 mm × 25.8 mm. The design of U-groove filling joint is shown in Figure3a,b, as part of the weld section. The chemical compositions of the BM and filler metal are summarized in Table2. The mechanical properties of the base metal are shown Materials 2021, 14, 4774 3 of 16

in Table3. The chemical composition and mechanical properties are provided by Anshan Iron and Steel Group Co., Ltd., located in Liaoning, China.

Table 1. Specifications of RCL-C6000 fiber laser machine.

Parameter (Unit) Value Rated power (W) 6000 Wavelength (nm) 1075–1085 Materials 2021, 14, x FOR PEER REVIEWDivergence angle (Rad) <0.1 3 of 16 Materials 2021 , 14, x FOR PEER REVIEW 3 of 16 Operation mode Fiber core diameter (µm) 100

FigureFigure 1. SchematicFigure 1.diagram diagram Schematic of of underwater underwater diagram of laser laserunderwater weldin weldingg lasertest test system weldin system ing in hyperbarictest hyperbaric system environment. in environment. hyperbaric environment.

Materials 2021, 14, x FOR PEER REVIEW 4 of 16

Table 3. Mechanical properties of S32101 duplex stainless steel.

Tensile Strength/MPa Yield Strength/MPa Hardness Elongation% Impact/J Ferrite Content% Mpa Mpa HB Figure 2. Hyperbaric25 °C underwater 130 °C laser welding 25 °C test system: 130 °C (a ) underwater laser welding test platform; (b) hyperbaric test Figure 2. HyperbaricFigure underwater 2. Hyperbaric laser underwater welding test laser system: welding (a) underwater test system: laser (a welding) underwater test platform; laser welding (b) hyperbaric test test chamber. 703chamber. 602platform; (b)453 hyperbaric test371 chamber. 49 207 98 56.5

Table 1. Specifications of RCL-C6000 machine. Table 1. Specifications of RCL-C6000 fiber laser machine. Parameter (Unit) Value Parameter (Unit) Value Rated power (W) 6000 Rated power (W) 6000 Wavelength (nm) 1075–1085 Wavelength (nm) 1075–1085 Divergence angle (Rad) <0.1 Divergence angle (Rad) <0.1 Operation mode Continuous wave Operation mode Continuous wave Fiber core diameter (μm) 100 Fiber core diameter (μm) 100 S32101 DSS was used as the base metal, and the dimensions of test plates were pre-

pared with 600 mmS32101 × 300 DSS mm was × 25.8 used mm. as Thethe basedesign metal, of U-groove and the filling dimensions joint is ofshown test plates in were pre- Figure 3. U-grooveFigureFigure design 3.3a,b, andU-groovepared as cross-sectional part with designof the600 andweld mmappearance cross-sectional ×section. 300 mm of The(a ×) the25.8 chemical appearance welded mm. The compositionsjoint of design and (a) the(b )of the welded ofU-groove schematic the jointBM filling anddiagram. and filler (b joint) the is shown in schematicmetal are diagram.summarizedFigure 3a,b, in as Table part 2.of The the mechanicalweld section. properties The chemical of the basecompositions metal are shownof the BM and filler in Table 3.2.2. Themetal Welding chemical are summarized Process composition in andTable mechanical 2. The mechanical properties properties are provided of the by baseAnshan metal are shown in Table 3. The chemical composition and mechanical properties are provided by Anshan Iron and Steel GroupIn 0.15 Co., MPa Ltd., pressure, located the in Liaoning, welding processChina. of U-type groove with laser wire filling Iron and Steel Group Co., Ltd., located in Liaoning, China. was carried out. The microstructure and mechanical properties of underwater welds with Table 2. Chemical composition of base metal and filler metals (wt.%). different shielding gas were studied. The process parameters are shown in Table 4. Table 2. Chemical composition of base metal and filler metals (wt.%). Materials C Si Mn P S Cr Ni Mo N Cu Co Nb Base metal Materials0.023 0.59C 4.9 Si Table 0.0199Mn 4. Welding 0.001 P parameters 21.5 S and 1.62 composition Cr 0.26 Ni of0.21 shielding Mo0.24 gas. N0.025 Cu - Co Nb Filler metalBase 0.012 metal 0.350.023 1.590.59 0.015 4.9 Focal 0.001 0.0199 Spot 22.56 0.001 8.62 21.5 3.05 1.620.15 0.260.06 0.210.049 0.240.002 0.025 - Wire Speed Laser Power Shielding Gas and SpecimenFiller metal No. 0.012 Layers 0.35 1.59 Diameter 0.015 0.001 22.56 8.62 3.05Speed 0.15 (m/min) 0.06 0.049 0.002 (m/min) (kW) Flow Rate (L/min) (mm) Root 5 3.2 5000 0.6 A Pure Ar, 25 Filling/finishing 5 2.6 5000 0.6 Root 5 3.2 5000 0.6 B 90%Ar + 10%N2 25 Filling/finishing 5 2.6 5000 0.6 Root 5 3.2 5000 0.6 C 50%N2 + 50%Ar, 25 Filling/finishing 5 2.6 5000 0.6 Root 5 3.2 5000 0.6 D Pure N2, 25 Filling/finishing 5 2.6 5000 0.6

The microstructure of weld specimens was characterized by optical microscopy (OM) and scanning electron microscopy (SEM) combined with energy dispersive (EDS). MTS E45 testing machine was used for tensile test, and the dimensions of the tensile specimens were prepared in accordance with ASTM E-8 standard. Charpy V-notch impact tests of the WM and HAZ were performed at −40 °C, and the impact specimens were pre- pared in accordance with ASTM A370 (2017). The microhardness was measured from HAZ to WM on a Vickers hardness tester. The test load was 500 g and the dwell time was 10 s. The percentage elongation, tensile and hardness were determined at ambient tem- perature.

2.3. Electrochemical Corrosion Tests The evaluation of weld pitting sensitivity by potentiodynamic tests has been con- firmed as a time-saving and non-destructive method [22]. A potentiostat with a three- cell was used, with saturated calomel electrode (SCE) and platinum electrode as

Materials 2021, 14, 4774 4 of 16

Table 2. Chemical composition of base metal and filler metals (wt.%).

Materials C Si Mn P S Cr Ni Mo N Cu Co Nb Base metal 0.023 0.59 4.9 0.0199 0.001 21.5 1.62 0.26 0.21 0.24 0.025 - 0.012 0.35 1.59 0.015 0.001 22.56 8.62 3.05 0.15 0.06 0.049 0.002

Table 3. Mechanical properties of S32101 duplex stainless steel.

Tensile Strength Yield Strength Elongation Hardness Impact Ferrite Content (MPa) (MPa) (%) HB (J) (%) 25 ◦C 130 ◦C 25 ◦C 130 ◦C 703 602 453 371 49 207 98 56.5

2.2. Welding Process In 0.15 MPa pressure, the welding process of U-type groove with laser wire filling was carried out. The microstructure and mechanical properties of underwater welds with different shielding gas were studied. The process parameters are shown in Table4.

Table 4. Welding parameters and composition of shielding gas.

Focal Spot Wire Speed Laser Power Shielding Gas and Specimen No. Layers Diameter Speed (m/min) (m/min) (kW) Flow Rate (L/min) (mm) Root 5 3.2 5000 0.6 A Pure Ar, 25 Filling/finishing 5 2.6 5000 0.6 Root 5 3.2 5000 0.6 B 90%Ar + 10%N , 25 Filling/finishing 5 2.6 5000 0.6 2 Root 5 3.2 5000 0.6 C 50%N + 50%Ar, 25 Filling/finishing 5 2.6 5000 0.6 2 Root 5 3.2 5000 0.6 D Pure N , 25 Filling/finishing 5 2.6 5000 0.6 2

The microstructure of weld specimens was characterized by optical microscopy (OM) and scanning electron microscopy (SEM) combined with energy dispersive spectroscopy (EDS). MTS E45 testing machine was used for tensile test, and the dimensions of the tensile specimens were prepared in accordance with ASTM E-8 standard. Charpy V-notch impact tests of the WM and HAZ were performed at −40 ◦C, and the impact specimens were prepared in accordance with ASTM A370 (2017). The microhardness was measured from HAZ to WM on a Vickers hardness tester. The test load was 500 g and the dwell time was 10 s. The percentage elongation, tensile and hardness were determined at ambient .

2.3. Electrochemical Corrosion Tests The evaluation of weld pitting sensitivity by potentiodynamic tests has been confirmed as a time-saving and non-destructive method [22]. A potentiostat with a three-electrode cell was used, with saturated calomel electrode (SCE) and platinum electrode as reference electrode and counter electrode, respectively, and the test piece as working electrode. The ◦ potentiodynamic test was carried out in 1 m NaCl at 60 C. The N2 purging was carried out for 30 min for deoxidization before each test. The specimens were composed of WM and HAZ with an exposed area of 25 mm2. In order to reduce the effect of the working electrode surface condition on the test result, the specimens were grounded to 2000 grit and polished with diamond paste. Materials 2021, 14, x FOR PEER REVIEW 5 of 16

reference electrode and counter electrode, respectively, and the test piece as working elec- trode. The potentiodynamic test was carried out in 1 m NaCl solution at 60 °C. The N2 purging was carried out for 30 min for deoxidization before each test. Materials 2021, 14, 4774 The specimens were composed of WM and HAZ with an exposed area5 of of 16 25 mm2. In order to reduce the effect of the working electrode surface condition on the test result, the specimens were grounded to 2000 grit and polished with diamond paste. In order to improve the accuracy of the experiment, the working electrode was ca- In order to improve the accuracy of the experiment, the working electrode was ca- thodically polarized at room temperature for 10 min. When the steady-state open circuit thodically polarized at room temperature for 10 min. When the steady-state open circuit potential (Eocp) was reached (approximately 10 min), the experiment was started and the potential (Eocp) was reached (approximately 10 min), the experiment was started and the specimen was anodically polarized up to the potential of +300 mVSCE. specimen was anodically polarized up to the potential of +300 mVSCE.

3. Results and3. Results Discussion and Discussion 3.1. Microstructural3.1. Microstructural Characterization Characterization The microstructureThe microstructure of BM is presented of BM is in presented Figure4a. in Figure Figure4b 4a. displays Figure the4b displays two phases the two phases volume fractions,volume confirmed fractions, asconfirmed almost equal as almost (46 ± equal1 vol.% (46 Ferrite), ± 1 vol.% in Ferrite), the microstructure in the microstructure measured bymeasured the software by the Axio software vert.a1 Axio configured vert.a1 configured by an optical by microscopean optical microscope (OM, Zeiss). (OM, Zeiss). The austeniteThe (light austenite phase) (light is distributed phase) is as distributed laths in the as ferrite laths matrix. in the However,ferrite matrix. in the However, laser in the welding process,laser welding the nitrogen process, in thethe moltennitrogen pool in the can molt been lost pool and can the be austenite lost and contentthe austenite of content the weld isof reduced the weld [23 is]. reduced [23].

FigureFigure 4. (a) Microstructure 4. (a) Microstructure of the BM;of the (b BM;) phases (b) phases volume volume fraction fraction of the BM.of the BM.

On the basisOn of the the basis DeLong of the diagram, DeLong thediagram, American the American Welding Society Welding (AWS) Society recom- (AWS) recom- mends themends WRC~92 the structureWRC~92 structure diagram, diagram, where nitrogen where ni istrogen a strong is a austenite-formingstrong austenite-forming ele- element [24ment]. It can [24]. not It onlycan not improve only improve the strength the ofstreng DSS,th but of alsoDSS, increase but also the increase toughness, the toughness, reduce thereduce formation the offormation the harmful of the intermetallic harmful intermet phaseallic and phase reduce and the reduce tendency the of tendency the of the precipitatedprecipitated phase, due phase, to a high due chromium to a high andchromium molybdenum and molybdenum content. content. Therefore, this paper investigates the effect of nitrogen on the microstructure and Therefore, this paper investigates the effect of nitrogen on the microstructure and mechanical properties of DSS-welded joints by underwater laser wire filling welding. mechanical properties of DSS-welded joints by underwater laser wire filling welding. Rapid cooling will restrict the formation of austenite and disturb the optimal phase Rapid cooling will restrict the formation of austenite and disturb the optimal phase balance in DSS. In Amir Baghdadhi’s experiments, laser reheating is used to avoid low balance in DSS. In Amir Baghdadhi’s experiments, laser reheating is used to avoid low austenite fraction and nitride formation [25]. Compared with air, the cooling rate of austenite fraction and nitride formation [25]. Compared with air, the cooling rate of the the molten pool of underwater laser welding is faster and the temperature gradient is molten pool of underwater laser welding is faster and the temperature gradient is larger. larger. Therefore, in the experiment, the method of reducing the welding speed was Therefore, in the experiment, the method of reducing the welding speed was used to im- used to improve the welding heat input, and a microstructure similar to that in air was prove the welding heat input, and a microstructure similar to that in air was obtained. obtained. Figure5 presents the microstructure of S32101 stainless steel welds under Figure 5 presents the microstructure of S32101 stainless steel welds under different shield- different shielding gas compositions. Figure6a–d are the weld microstructure of samples A, ing gas compositions. Figure 6a–d are the weld microstructure of samples A, B, C and D, B, C and D, respectively. It is observed that the microstructure is mainly composed of two phases: therespectively. light phase is It austenite is observed and that the the dark microstr phase isucture ferrite. is mainly However, composed compared of withtwo phases: the the base metal,light therephase are is significantaustenite and differences the dark in phase the morphology is ferrite. However, of the γ-phases compared formed with the base in the grainmetal, boundaries there are of thesignificantα-phase differences in the weld in zone. the Moreover,morphology the of recrystallized the γ-phases size formed in the of the γ-phasegrain in boundaries the A and B of specimens the α-phase is coarser in the thanweld that zone. in Moreover, the C and D the specimens. recrystallized This size of the is due to theγ-phase austenite in the transformation A and B specimens of weld metalis coarser under th thean that action in ofthe a C welding and D thermalspecimens. This is cycle during underwater welding. The previous-transformation austenite (PTA) and grain boundary austenite (GBA) began precipitating along the ferrite grain boundaries. A large amount of austenite precipitated as Widmanstätten austenite (WA) and grew into austenite grains in the ferrite. At the same time, a large number of intragranular austenite (IGA) particles with an acicular morphology were observed within the ferrite grains. Compared with the welds prepared with argon as a shielding gas, the volume of WA, IGA and GBA Materials 2021, 14, x FOR PEER REVIEW 6 of 16

due to the austenite transformation of weld metal under the action of a welding thermal cycle during underwater welding. The previous-transformation austenite (PTA) and grain boundary austenite (GBA) began precipitating along the ferrite grain boundaries. A large amount of austenite precipitated as Widmanstätten austenite (WA) and grew into austen- Materials 2021, 14, 4774 6 of 16 ite grains in the ferrite. At the same time, a large number of intragranular austenite (IGA) particles with an acicular morphology were observed within the ferrite grains. Compared with the welds prepared with argon as a shielding gas, the volume of WA, IGA and GBA of those preparedof those with prepared the nitrogen with the mixture nitrogen gas ormixture pure nitrogen gas or pure were nitrogen increased, were and increased, the and the three types ofthree austenite types wereof austenite coarsened were (Figure coarsened5d). However, (Figure 5d). the However, type of austenite the type in of the austenite in the weld did notweld change did obviouslynot change under obviously the different under the shielding different gas shielding conditions. gas Therefore,conditions. Therefore, increasing theincreasing proportion the ofproportion nitrogen inof thenitrogen shielding in the gas shielding can significantly gas can significantly increase the increase the austenite content,austenite but content, the transformation but the transformation type of the austenite type of the was austenite not affected. was not affected.

Figure 5. WeldFigure microstructure: 5. Weld microstructure: (a) sample A in (a pure) sample Ar shielding A in pure gas; Ar ( shieldingb) sample gas;B in (90%b) sample Ar + 10%N B in 90%2 hybrid Ar + shielding Materialsgas; 2021 (c), 14sample, x FOR10% C PEER in N 50% hybridREVIEW Ar + shielding50% N2 hybrid gas; (c )shielding sample C gas; in 50% (d) sample Ar + 50% D Nin purehybrid N2 shielding shielding gas; gas. (d ) sample D in 7 of 16 2 2 pure N2 shielding gas. Figure 6 shows the effects of different shielding gas compositions on the microstruc- ture in the HAZs. The microstructure near the fusion line was composed of WM, a transi- tion zone (TZ) and HAZ. Due to fact that the temperature of HAZ near the fusion line can increase to approximately 1000 °C, all of the γ-phases of the original duplex microstruc- tures will further evolve. In the subsequent cooling process, some ferrite transforms into austenite, but the growth of the austenite grains was inhibited due to the decrease in the high temperature residence time in the underwater environment. Therefore, the HAZ con- sists of a large number of coarse polygonal ferrite and irregular austenite. It is also ob- served that the width of the polygonal ferrite band was approximately 50 μm. However, the microstructure of TZ was quite different under different shielding gas conditions. It can be seen that the more nitrogen content in the shielding gas, the more austenite content and the more uniform the distribution. In addition, due to the composition gradient and diffusion effect of the weld and HAZ, the of Cr and Ni in HAZ near the fusion line were decreased, thus the γ-phase began to nucleate in the ferrite or dislocation line. However, due to the increase in undercooling, the γ-phase formed is very small, which is different from the weld γ-phase, and has little difference with ferrite in compo- sition, and so belongs to non-diffusion transformation.

Figure 6. HAZFigure microstructure: 6. HAZ microstructure:(a) sample A in (purea) sample Ar shielding A in pure gas; Ar (b shielding) sample B gas; in 90% (b) sample Ar + 10% B in N 90%2 hybrid Ar gas; (c)

sample C in 50%+ 10%Ar + N50%2 hybrid N2 hybrid gas; gas; (c) sample(d) sample C in D 50% in pure Ar +N2 50% shielding N2 hybrid gas. gas; (d) sample D in pure N2 shielding gas. Figure 7 shows the austenite content of four kinds of welds in different shielding gas compositions and BM. As can be seen from Figure 7, with the increase in the nitrogen content in the shielding gas, the content of austenite in WM and HAZ gradually increases. In addition, when the nitrogen content in the shielding gas increases to 50%, the austenite content increases greatly, and when the shielding gas is pure nitrogen, the increase range of the austenite content decreases. The reason for this phenomenon may be because the nitrogen element in the WM reached the saturation state at the austenite transformation temperature, and then reached the maximum value of the α/γ transformation rate. How- ever, there was an obvious temperature gradient in HAZ near the fusion line, and the of nitrogen in ferrite decreases rapidly with the decrease in temperature, so lim- ited austenization occurred in the HAZs.

Figure 7. The effect of the shielding gas composition on the austenite fractions of the DSS welds.

Materials 2021, 14, x FOR PEER REVIEW 7 of 16

Materials 2021, 14, 4774 7 of 16

Figure6 shows the effects of different shielding gas compositions on the microstructure in the HAZs. The microstructure near the fusion line was composed of WM, a transition zone (TZ) and HAZ. Due to fact that the temperature of HAZ near the fusion line can increase to approximately 1000 ◦C, all of the γ-phases of the original duplex microstruc- tures will further evolve. In the subsequent cooling process, some ferrite transforms into austenite, but the growth of the austenite grains was inhibited due to the decrease in the high temperature residence time in the underwater environment. Therefore, the HAZ consists of a large number of coarse polygonal ferrite and irregular austenite. It is also observed that the width of the polygonal ferrite band was approximately 50 µm. However, the microstructure of TZ was quite different under different shielding gas conditions. It Figure 6. HAZ microstructure: (a) sample A in pure Ar shielding gas; (b) sample B in 90% Ar + 10% N2 hybrid gas; (c) can be seen that the more nitrogen content in the shielding gas, the more austenite content sample C in 50% Ar + 50% N2 hybridand the gas; more (d) uniformsample D the in distribution. pure N2 shielding In addition, gas. due to the composition gradient and diffusion effect of the weld and HAZ, the concentrations of Cr and Ni in HAZ near the fusionFigure line 7 shows were decreased, the austenite thus the contentγ-phase beganof four to kinds nucleate of in welds the ferrite in different or dislocation shielding gas compositionsline. However, and due BM. to the As increase can be in undercooling,seen from Fi thegureγ-phase 7, with formed the is veryincrease small, whichin the nitrogen contentis different in the fromshielding the weld gas,γ-phase, the content and has of little austenite difference in with WM ferrite and in HAZ composition, gradually and increases. so belongs to non-diffusion transformation. In addition,Figure when7 shows the the nitrogen austenite content content ofin four the kindsshielding of welds gas in increases different shielding to 50%, gas the austenite contentcompositions increases and greatly, BM. As and can when be seen the from shielding Figure7, withgas is the pure increase nitrogen, in the nitrogenthe increase range of thecontent austenite in the shieldingcontent gas,decreases. the content The of austenitereason for in WM this and phenomenon HAZ gradually may increases. be because the nitrogenIn addition, element when in the the nitrogen WM reached content in the the satura shieldingtion gas state increases at the to 50%, aust theenite austenite transformation content increases greatly, and when the shielding gas is pure nitrogen, the increase range temperature, and then reached the maximum value of the α/γ transformation rate. How- of the austenite content decreases. The reason for this phenomenon may be because the ever,nitrogen there was element an inobvious the WM temperature reached the saturation gradie statent in at HAZ the austenite near the transformation fusion line, and the solubilitytemperature, of nitrogen and then in reachedferrite thedecreases maximum rapi valuedly of with the α the/γ transformation decrease in temperature, rate. How- so lim- ited ever,austenization there was anoccurred obvious temperaturein the HAZs. gradient in HAZ near the fusion line, and the solubility of nitrogen in ferrite decreases rapidly with the decrease in temperature, so limited austenization occurred in the HAZs.

Figure 7. The effect of the shielding gas composition on the austenite fractions of the DSS welds. Figure3.2. 7. Mechanical The effect Performance of the shielding Testing gas composition on the austenite fractions of the DSS welds. The stress–strain curves and tensile test results of the BM and welded joints are presented in Figure8 and Table5. Compared to the BM, the joints had a higher tensile strength and yield strength but their elongation was slightly lower than that of BM. The experimental results also present that the increase in nitrogen content in the shielding gas did not affect the strength and toughness of the joint. The four kinds of welded joints were all strengthened, and the BM was the weakest part of the joint.

Materials 2021, 14, x FOR PEER REVIEW 8 of 16

3.2. Mechanical Performance Testing The stress–strain curves and tensile test results of the BM and welded joints are pre- sented in Figure 8 and Table 5. Compared to the BM, the joints had a higher tensile strength and yield strength but their elongation was slightly lower than that of BM. The experimental results also present that the increase in nitrogen content in the shielding gas did not affect the strength and toughness of the joint. The four kinds of welded joints were all strengthened, and the BM was the weakest part of the joint. There are three main reasons, which are as follows: • The ferrite and austenite in the BM are banded along the rolling direction, whereas the ferrite and austenite in the WM and HAZ are interlaced with different directions, and the grain boundaries are increased, which can lock the dislocation and strengthen the joint; • The atoms of Cr, Mo and Ni in the weld metal can be remelted at a high temperature, which can replace Fe atoms in the lattice and disturb the original lattice arrangement, and can also make dislocation movement difficult and strengthen the joint. As in HSLA steel, the addition of Mn and other alloying elements, such as copper (Cu), (TI) and vanadium (V), both provide strengthening and an obtain ideal mi- crostructure [26,27]; • The nitrogen atoms in the shielding gas are intercalated into the lattice in the form of an interstitial solid solution. The strengthening effect of the interstitial solid solution is more obvious than that of the replacement solid solution. The nitrogen atoms are mainly concentrated in the austenite phase and directly strengthen the austenite. Materials 2021, 14, 4774 8 of 16 Therefore, the strength performance of WM is better than that of BM.

Figure 8. FigureStress–strain 8. Stress–strain curve. curve. Table 5. Results of tensile tests.

Tensile Strength Yield Strength No. Elongation/% Fracture Location Table 5. ResultsRm/MPa of tensileRp0.2/MPa tests. A 749 589 28.5 BM B 721Tensile 581 Strength 30.5 Yield BM Strength Fracture Loca- C 747 592 26 BM DNo. 748 586 26.5 BM Elongation/% BM 766Rm/MPa 608 32.5Rp0.2/MPa BM tion A 749 589 28.5 BM There are three main reasons, which are as follows: • The ferriteB and austenite in the BM are721 banded along the rolling direction,581 whereas 30.5 BM the ferrite and austenite in the WM and HAZ are interlaced with different directions, and the grainC boundaries are increased, which747 can lock the dislocation and592 strengthen 26 BM the joint; • The atomsD of Cr, Mo and Ni in the weld748 metal can be remelted at a high temperature,586 26.5 BM which can replace Fe atoms in the lattice and disturb the original lattice arrangement, and canBM also make dislocation movement766 difficult and strengthen the joint. 608 As in 32.5 BM HSLA steel, the addition of Mn and other alloying elements, such as copper (Cu), titanium (TI) and vanadium (V), both provide strengthening and an obtain ideal 3.3.microstructure Charpy [ 26V-Notch,27]; Impact Tests • The nitrogen atoms in the shielding gas are intercalated into the lattice in the form of an interstitial solid solution. The strengthening effect of the interstitial solid solution is moreCharpy obvious than V-notch that of the replacementimpact solidtests solution. were The performed nitrogen atoms on the weld zone and HAZ at −40 °C are mainly concentrated in the austenite phase and directly strengthen the austenite. andTherefore, the results the strength performanceare shown of WM in is better Figure than that 9. of In BM. order to explain the relationship between impact

3.3.energy Charpy V-Notch and Impact ferrite Tests content, the ferrite content in different areas of weld is expressed by a polylineCharpy V-notch diagram. impact tests wereIn the performed figure, on the weld the zone average and HAZ at absorbing−40 ◦C energy of the WM of the A and B and the results are shown in Figure9. In order to explain the relationship between impact energysamples and ferrite at content, −40 the°C ferrite were content 24.74 in different J and areas 25.25 of weld isJ, expressed respectively, by which is merely 62.5% of the BM, a polyline diagram. In the figure, the average absorbing energy of the WM of the A and

Materials 2021, 14, x FOR PEER REVIEW 9 of 16

Materials 2021, 14, 4774 9 of 16 with an average value of 40 J. The average absorbed energy of the WM of the C and D samples increased to 32.62 J and 33.75 J, which is approximately 82.5% of the BM. Com- B samples at −40 ◦C were 24.74 J and 25.25 J, respectively, which is merely 62.5% of the paredBM, with with an average the BM, value the of 40average J. The average absorbing absorbed energy energy ofof thethe WM underwater of the C and Dlaser welds de- creased,samples increased but the to A 32.62 and J andB samples 33.75 J, which exhibited is approximately a more distinct 82.5% of thefall BM. than Compared that of C and D. The mainwith the reason BM, the is averagedue to absorbingthe change energy in austen of the underwaterite content laser in the welds weld. decreased, According but to Karlsson etthe al., A andthe B preferential samples exhibited growth a more orientation distinct fall [100 than] thatof ferrite of C and is D.in The conjunction main reason with the easiest is due to the change in austenite content in the weld. According to Karlsson et al., the cleavage planes, which would seriously damage the resulting toughness below the duc- preferential growth orientation [100] of ferrite is in conjunction with the easiest cleavage tile-to-brittleplanes, which would transition seriously temperature damage the resulting[28]. Th toughnesse change below in HAZ the ductile-to-brittleimpact energy under differ- enttransition shielding temperature gas conditions [28]. The change is similar in HAZ to impactthat of energy the weld under zone, different which shielding is also the result of thegas conditionsinfluence is of similar the austenite to that of the content. weld zone, which is also the result of the influence of the austenite content.

55 WM WM 70 50 HAZ HAZ BM 60 45 50 40 35.75 35.75 33.7534.5 40 35 32.5 32.625 30 IMPACT (J) 30 25.25

24.74 content (%) Austenite 25 20

20 10

15 0 weld A weld B weld C weld D BM

Figure 9. 9.Austenite Austenite fractions fractions and absorbing and absorbing energy of energy the WMs of in the different WMs shielding in different gas. shielding gas. Figure 10 shows the fractured surface morphologies after the impact tests at −40 ◦C. It canFigure be seen that10 shows the fractured the fractured surface of surface welds A andmorphologies B was relatively after flat the compared impact to tests at −40 °C. Itthat can of be welds seen C andthat D. the The fractured fractured surfacesurface is of characterized welds A and by a B quasi-cleavage was relatively plane, flat compared to thattearing of ridge welds and C fluvial and pattern,D. The which fractured is consistent surface with is the characterized occurrence of a quasi-cleavageby a quasi-cleavage plane, tearingfracture. ridge Some larger and fluvial holes were pattern, also revealed which in is welds cons Aistent and B, with which the may occurrence have resulted of a quasi-cleav- from the peeling of inclusions in the impact testing. At a higher magnification (1000×), agea large fracture. number ofSome high-density, larger holes short and were curved also tearing revealed ridges in were welds observed A and in weld B, which A. may have resultedThe crack sourcefrom radiatedthe peeling a fluvial of patterninclusions from thein the middle impact to the testing. surrounding, At a and higher there magnification (1000×),were many a micro-dimpleslarge number around of high the-density, crack source. short Therefore, and curved the quasi-cleavage tearing ridges fracture were observed in welddominated A. The the crack weld A source fracture radiated mode. The a fluvial size of thepattern deformation from the and middle the amount to the of surrounding, dimples in weld B were larger than those in weld A, and some micro-holes caused by andplastic there deformation were many were micro-dimples also revealed around around the quasi-cleavage the crack source. small section,Therefore, which the quasi-cleav- ageindicates fracture that the dominated impact toughness the weld of weld A Bfracture is improved. mode. Figure The 10 csize depicts of obviousthe deformation tear and the amountdimples on of the dimples fracture surfacein weld of weldB were C, whichlarger indicates than those that the in impact weld fractureA, and at some the micro-holes causedweld joint by appears plastic as deformation dimple ductile were fracture also characteristics. revealed around At the the same quasi-cleavage time, there are small section, obvious dimples and a fibrous network on the impact fracture surface of weld D, which is which indicates that the impact toughness of weld B is improved. Figure 10c depicts ob- characterized by a ductile fracture. vious tear dimples on the fracture surface of weld C, which indicates that the impact frac- ture at the weld joint appears as dimple ductile fracture characteristics. At the same time, there are obvious dimples and a fibrous network on the impact fracture surface of weld D, which is characterized by a ductile fracture.

Materials 2021,, 14,, 4774x FOR PEER REVIEW 10 of 16

Figure 10. SEMSEM micrographs micrographs showing showing the the fracture fracture surface surface morphology morphology after after Charpy Charpy impact impact testing testing at − at40 −°C:40 (a◦)C: the (a )frac- the fracturedtured surface surface of weld of weld A; A;(b) (bthe) the fractured fractured surface surface of ofweld weld B; B; (c ()c the) the fractured fractured surface surface of of weld weld C; C; ( (dd)) the the fractured fractured surface of weld D.

Therefore, it can be concluded that the addition of nitrogen in the shielding gas can reduce the degree of the quasi-cleavage fracture, and that the size and quantity of nitride in the impact fracture were reduced. When the shielding gas changes from pure argon to pure nitrogen, the quasi-cleavage fracture transformedtransformed intointo ductileductile fracture.fracture. The second-phase particles in the quasi-cleavage section were observed by high- magnificationThe second-phase SEM, and particles the chemical in the composition quasi-cleavage and locationsection were of inclusions observed were by high-mag- measured nificationby EDS (Figure SEM, 11 and). The the resultschemical show composition that the inclusion and location was a Fe-Crof inclusions intermetallic were compound. measured Whenby EDS the (Figure chromium 11). The content results was show high, that the brittlethe inclusion and hard wasσ-phase a Fe-Cr begins intermetallic to precipitate com- pound.from δferrite When at the 820 chromium◦C. The σ -phasecontent is was a Fe-Cr high, intermetallic the brittle and compound hard σ-phase of WCR begins 45%, to which pre- cipitatewill make from the δ metalferrite more at 820 brittle. °C. The On theσ-phase other hand,is a Fe-Cr the formation intermetallic of inclusions compound will of reduce WCR 45%,the stress which area, will blockmake thethe propagationmetal more brittle. path of On plastic the other deformation hand, the and formation also reduce of inclu- the sionsimpact will toughness. reduce the Due stress to the area,σ-phase block precipitating the propagation at the path grain of boundary, plastic deformation a large amount and alsoof chromium reduce the will impact be consumed, toughness. reducing Due to the the σ corrosion-phase precipitating resistance of at the the material. grain boundary,

Materials 2021, 14, x FOR PEER REVIEW 11 of 16

Materials 2021, 14, 4774 11 of 16 a large amount of chromium will be consumed, reducing the corrosion resistance of the material.

FigureFigure 11. 11. TheThe results results of of EDS EDS analysis analysis of the incl inclusionsusions in impact fracture surface.

3.4. Hardness Tests 3.4. Hardness Tests Figure 12 shows the change in microhardness in different weld cross sections. It can Figure 12 shows the change in microhardness in different weld cross sections. It can be seen that the average hardness of WM (average: ~280 HV) is higher than that of HAZ be seen that the average hardness of WM (average: ~280 HV) is higher than that of HAZ andand BM BM (average (average micro-hardness: micro-hardness: ~240 HV). HV). The The reason reason is that that the the higher higher Ni and and Cr Cr atoms atoms inin the the weld weld can can produce produce an an obvious obvious solid solid solution solution strengthening strengthening effect effect by by remelting and replacingreplacing the the iron iron atoms atoms in in the lattice. In In addition, addition, the the ferrite ferrite and and austenite in in the weld metalmetal interlace with eacheach other,other, the the number number of of grain grain boundaries boundaries increases increases and and the the direction direc- tionis different, is different, which which leads leads to the to increase the increase in hardness. in hardness. Weld BWeld and weldB and C weld obtained C obtained similar similarmicro-hardness micro-hardness values values (275 HV), (275 which HV), which were slightlywere slightly lower lower than than weld weld D (280 D (280 HV) HV) and andhigher higher than than weld weld A (270 A HV).(270 HV). This mayThis bemay due be to due the higherto the higher content content of austenite of austenite in weld in D, weldwhich D, leads which to leads the increase to the increase in dislocation in dislocation density density and the and increase the increase in micro-hardness in micro-hard- in nessthe weld. in the However, weld. However, the content the ofcontent austenite of austenite in the argon-gas-protected in the argon-gas-protected weld is low, weld which is low,leads which to the leads decrease to the in decrease the Cr element in the andCr element the effect and of the solid effect solution of solid strengthening, solution strength- which, ening,in turn, which, results in in turn, a decrease results inin hardness.a decrease in hardness.

MaterialsMaterials2021 2021, 14, 14, 4774, x FOR PEER REVIEW 12 12of of16 16 Materials 2021, 14, x FOR PEER REVIEW 12 of 16

Figure 12. Micro-hardness measurements on the welded joints. FigureFigure 12.12. Micro-hardness measurements measurements on on the the welded welded joints. joints. In addition, the dilution ratio of alloy elements also has an important influence on InIn addition,addition, thethe dilutiondilution ratioratio ofof alloyalloy elementselements alsoalso hashas anan importantimportant influenceinfluence onon the the microstructure and hardness. microstructurethe microstructure and and hardness. hardness. In order to further determine the change in Ni and Cr elements from WM to HAZ InIn orderorder toto furtherfurther determinedetermine thethe change change in in Ni Ni and and Cr Cr elements elements from from WM WM to to HAZ HAZ near near the fusion line, spot and line scanning by EDS were conducted from the HAZ to the thenear fusion the fusion line, line, spot spot and lineand line scanning scanning by EDSby EDS were were conducted conducte fromd from the the HAZ HAZ to to the the side side of the weld in Figure 13. The line scanning result is shown in Figure 13a, and the ofside the of weld the weld in Figure in Figure 13. The 13. line The scanning line scanning result result is shown is shown in Figure in Figure 13a, and 13a, the and element the element compositions of points A and B are listed in Figure 13b. As can be seen in Figure compositionselement compositions of points of A points and B A are and listed B are in listed Figure in Figure13b. As 13b. can As be can seen be in seen Figure in Figure 13a, the 13a, the microhardness of the transition zone is significantly different from that of the microhardness13a, the microhardness of the transition of the transition zone is significantlyzone is significantly different different from that from of thethat weld of the zone weld zone and substrate. Table 6 shows the experimental results of EDS point scanning andweld substrate. zone and Table substrate.6 shows Table the 6 experimental shows the experimental results of EDS results point of scanning EDS point for scanning four welds for four welds under different shielding gas conditions. Due to the fact that the change underfor four different welds under shielding different gas conditions.shielding gas Due conditions. to the fact Due that to the the change fact that trend the change is similar, trend is similar, the EDS line scanning test is not repeated for the other three samples. thetrend EDS is similar, line scanning the EDS test line is scanning not repeated test is for not the repeated other three for the samples. other three samples.

(a) ((bb)) Figure 13. EDS result of line scanning for (a), and points scanning for (b); A, B points are HAZ and WM scanning zone by FigureFigure 13. 13.EDS EDS result result ofof line scanning for for (a (a),), and and points points scanning scanning for for (b); ( bA,); B A, points B points are HAZ are HAZ and WM and WMscanning scanning zone by zone EDS. by EDS.

TableTable 6.6. EDSEDS resultsresults ofof points points shown shown inin FigureFigure 14a 14a13a (at.%). (at.%). (at.%). Weld Points Fe Cr Ni Mn Mo Si Weld Points Points Fe Cr Cr Ni Ni Mn Mn Mo Mo Si Si 1 65.494 65.494 22.793 22.793 7.3327.332 2.4482.448 1.0561.056 0.8770.877 A 1 65.494 22.793 7.332 2.448 1.056 0.877 A 2 67.624 21.552 4.156 3.889 0.828 1.951 22 67.624 21.552 21.552 4.156 3.889 3.889 0.828 0.828 1.951 1.951 1 66.004 66.004 22.154 22.154 7.2967.296 3.0583.058 0.5130.513 0.9750.975 B 1 66.004 22.154 7.296 3.058 0.513 0.975 B 2 67.786 21.364 4.514 3.891 1.447 0.998 22 67.786 21.364 21.364 4.514 3.891 3.891 1.447 1.447 0.998 0.998 1 64.691 64.691 23.269 23.269 6.8036.803 3.0433.043 0.2930.293 1.9011.901 C 1 64.691 23.269 6.803 3.043 0.293 1.901 C 2 66.826 22.416 3.115 4.168 1.096 2.379 22 66.826 22.416 22.416 3.115 4.168 4.168 1.096 1.096 2.379 2.379 D 1 64.574 64.574 23.763 23.763 6.7716.771 2.4482.448 0.7850.785 1.6921.692 1 64.574 23.763 6.771 2.448 0.785 1.692 D 2 67.872 22.310 3.350 4.196 1.048 1.224

Materials 2021, 14, x FOR PEER REVIEW 13 of 16

Materials 2021, 14, 4774 13 of 16 2 67.872 22.310 3.350 4.196 1.048 1.224

Due to the local composition gradient and the diffusion effect of alloy elements, there was Duean obvious to the local composition gradient gradient between and the WM diffusion and HAZ. effect Since of alloy the elements,transition therezone washad ana considerable obvious concentration amount of gradientCr with betweena high affinity WM and for HAZ.carbon, Since the interstitial the transition atoms zone of had a considerable amount of Cr with a high affinity for carbon, the interstitial atoms of carbon migrated from the partially melted zone in HAZ during remelting. At this time, a carbon migrated from the partially melted zone in HAZ during remelting. At this time, carbon-poor ferritic band with a polygonal morphology was formed adjacent to the fusion a carbon-poor ferritic band with a polygonal morphology was formed adjacent to the boundary in the HAZ. Polygonal ferrite and irregular austenite were interlaced in order fusion boundary in the HAZ. Polygonal ferrite and irregular austenite were interlaced to increase dislocation energy, which is considered to be the major factor for the increase in order to increase dislocation energy, which is considered to be the major factor for the in hardness. The existence of the ferrite-banded region has been verified in many litera- increase in hardness. The existence of the ferrite-banded region has been verified in many tures. Some scholars call it the ferrite-stabilized region [29], carbon-poor region [30,31] literatures. Some scholars call it the ferrite-stabilized region [29], carbon-poor region [30,31] and decarburized region [32]. It is reported [30] that this carbon-poor zone has a jagged and decarburized region [32]. It is reported [30] that this carbon-poor zone has a jagged morphology and its width depends upon both the temperature gradient and the carbon morphology and its width depends upon both the temperature gradient and the carbon content of the substrate. content of the substrate.

3.5.3.5. PotentiodynamicPotentiodynamic PolarizationPolarization 2 InIn orderorder toto studystudy thethe effecteffect ofof NN2 contentcontent inin thethe protectiveprotective gasgas onon thethe pittingpitting behaviorbehavior ofof S32101 welds by by ULW, ULW, the the polarization polarization test test was was conducted conducted at at60 60°C ◦atC the at thescanning scanning rate rateof 1.67 of 1.67mV/s mV/s in 1 m in NaCl 1 m NaCl solution. solution. Figure Figure 14 shows 14 shows the effects the effects of different of different shielding shielding gases gaseson the on anodic the anodic polarization polarization responses responses of the DSS of the WM DSS and WM BM. and From BM. the From slope the of slopethe pas- of thesivation passivation currentdensities, densities, it can be it seen can bethat seen, with that, the withaustenite the austenite phase increased, phase increased, the spec- theimen specimen with pure with N2 pureprotection N2 protection exhibited exhibited a lower corrosion a lower corrosioncurrent density current range density compared range comparedwith that of with the pure that ofAr theprotection. pure Ar The protection. difference The in the difference responses in to the the responses localized tocorro- the localizedsion was the corrosion result of was the the change result in of nitrogen the change content, in nitrogen because content,nitrogen becausepromotes nitrogen the for- promotesmation of thethe formationaustenite stability of the austenite phase. stability phase.

FigureFigure 14.14. AnodicAnodic polarizationpolarization curvescurves ofof thethe DSSDSS jointsjoints weldedwelded inin differentdifferent shieldingshielding gases.gases.

TableTable7 7shows shows the the electrochemical electrochemical parameter parameter values values in in Figure Figure 14 14.. The The transpassive transpassive potentialpotential (E(Et)t) determinesdetermines the the pitting pitting corrosion corrosion susceptibility, susceptibility, and and the the passive passive current current density den- (sityip) determines (ip) determines the corrosionthe corrosion rates rates of the of welded the welded zone zone in the in passivation the passivation range. range.

TableTable 7.7.Electrochemical Electrochemical corrosioncorrosion parametersparameters after after anodic anodic polarization polarization tests. tests.

Icorr Icorr EcorrE corr ip ip EEt t EtE−t-EEff ShieldingShielding Gas Gas 2 2 μA/cmµA/cm2 mVmVSCE SCE μA/cmµA/cm2 mVmVSCESCE mV A A151 1513297 3297 52 52357 357 528 B B145 1453056 3056 50 50348 348 521 C 74 2756 36 799 1002 D 68 2731 33 847 1010 BM 67 2689 34 824 919

Materials 2021, 14, x FOR PEER REVIEW 14 of 16

C 74 2756 36 799 1002 Materials 2021, 14, 4774 D 68 2731 33 84714 of 16 1010 BM 67 2689 34 824 919

As shown in TableAs shown7, the valuein Table of 7, Et the was value increased of Et was significantly increased withsignificantly the addition with the addition of of nitrogen gas,nitrogen and the gas, value and of ipthe was value decreased. of ip was This decrea meanssed. that This the means pitting that corrosion the pitting corrosion resistance of theresistance passive film of the is superiorpassive film to that is superior of pure Arto that as a of shielding pure Ar gas. as a The shielding value ofgas. The value of Et with a hybrid shielding gas in 90% Ar + 10% N2 is slightly greater than pure argon, but Et with a hybrid shielding gas in 90% Ar + 10% N2 is slightly greater than pure argon, but the ip value is almost unchanged. Furthermore, the value of (Et − Ef), which shows the the ip value is almost unchanged. Furthermore, the value of (Et − Ef), which shows the passive region,passive is increased region, with is increased the addition with of the nitrogen addition in the of nitrogen shielding in gas. the Theshielding above gas. The above results show that,results with show the addition that, with of nitrogen,the addition the valueof nitrog of ipen, decreases, the value the of resistanceip decreases, of the resistance the passive filmof to the pitting passive corrosion film to is pitting improved corrosion and the is stability improved increases. and the Among stability them, increases. Among pure nitrogen asthem, a shielding pure nitrogen gas has anas excellenta shielding pitting gas resistancehas an excellent property. pitting Kyröläinen resistance property. and Lukkari [33Kyröläinen] and the Outokumpu and Lukkari Welding [33] and Handbook the Outokumpu (Outokumpu, Welding 2010) Handbook have noted (Outokumpu, 2010) that a low austenitehave contentnoted that can a lead low toaustenite nitride precipitation,content can lead which to nitride has a negative precipitation, effect on which has a neg- weld corrosionative properties effect andon weld toughness corrosion [34]. properties and toughness [34]. Figure 15 demonstrateFigure 15 thedemonstrate SEM images the SEM of the images pitting of morphology the pitting morphology formed on formed the on the sur- surface of the DSSface weldsof the weldedDSS welds in fourwelded different in four shielding different gas shielding conditions gas conditions after anodic after anodic po- polarization tests.larization It is obvious tests. It thatis obvious the pitting that the corrosion pitting degreecorrosion in thedegree HAZ in wasthe HAZ much was much more more severe thansevere in the than WM in andthe WM BM, soand the BM, pitting so the resistance pitting resistance of the whole of the welded whole joint welded joint was was predominatedpredominated by the HAZ. by the HAZ.

(a) (b)

(c) (d)

Figure 15. SEMFigure images 15. SEM of selected images of corroded selected corrodedregions in regions the DSS in joint the DSS after joint potentiostatic after potentiostatic test: (a,c test:) HAZ (a,c near) HAZ the WM and (b,d) HAZ near the WMBM. and (b,d) HAZ near the BM.

In HAZ, pittingIn tends HAZ, to bepitting on the tends polygonal to be on ferrite the bandpolygona nearl the ferrite fusion band line near (Figure the 15fusiona) line (Figure and on the boundaries15a) and on of austenitethe boundaries and ferrite. of austenite The reasonand ferrite. is that The the reason lean carbonis that the and lean carbon and high a-phase contenthigh a-phase of the polygonalcontent of ferritethe polygonal zone leads ferrite to the zone precipitation leads to the of precipitation Cr2N. As of Cr2N. As the nucleationthe point, nucleation Cr2N both point, causes Cr2N the both concentration causes the ofconcentration corrosion stress of corrosion and diffuses stress and diffuses around it, resultingaround in corrosionit, resulting pits in (Figurecorrosion 15c,d). pits (Figure 15c,d). Therefore, it is concluded that the increase in nitrogen content in the shielding gas can significantly promote the transformation of austenite, increase the solubility of Cr2N, reduce the width of the chromium-poor zone and increase the overall corrosion resistance of the weld. Materials 2021, 14, 4774 15 of 16

4. Conclusions 1. The addition of nitrogen in the shielding gas can increase the austenite content in the weld zone to approximately 51.6% higher than that of pure argon (the nitrogen shielded is 51.6%, the pure argon shielded is 32.2%), but the transformation type of the ferrite–austenite is not affected significantly; 2. The increase in nitrogen content in the shielding gas does not affect the strength of the joint, and the base metal is still the weakest part of the joint; 3. The evaporation loss of nitrogen in the weld pool means that it is easy to form harmful phases, such as an Fe-Cr intermetallic compound in the weld, which is not conducive to the impact toughness of the weld; 4. The addition of nitrogen in the shielding gas is beneficial to austenite regeneration during solidification. The pitting corrosion resistance of the four kinds of welds is not as good as that of base metal; The heat affected zone has poor corrosion resistance. The weld with pure nitrogen protection has the highest Et value and the best corrosion resistance; 5. There is a carbon-poor ferrite band near the fusion line in the HAZ. Due to the fact that the solubility of nitrogen in ferrite decreases rapidly with the decrease in temperature, the supersaturated nitrogen combines with chromium to form chromium nitride precipitation. The precipitation of Cr2N results in partial Cr depletion, which is the main reason for the weak pitting resistance of HAZ.

Author Contributions: Conceptualization, C.S. and X.J.; methodology, J.Z.; software, K.W.; val- idation, K.W.; formal analysis, K.W.; investigation, K.W. and C.L.; resources, C.S., X.J. and Z.C.; data curation, K.W.; writing—original preparation, K.W.; writing—review and editing, K.W.; visualization, K.W.; supervision, J.Z.; project administration, C.S.; funding acquisition, J.Z. and X.J. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by National Science and Technology Major Project, grant num- ber 2018ZX06002006 and Award Cultivation Foundation from Beijing Institute of Petrochemical Technology, grant number Project No.BIPTACF-010. Conflicts of Interest: The authors declare no conflict of interest.

References 1. Westin, E.M.; Johansson, M.M.; Bylund, L.-Å.; Pettersson, R.F.A. Effect on microstructure and properties of super duplex stainless steel welds when using backing gas containing nitrogen and hydrogen. Weld. World 2014, 58, 347–354. [CrossRef] 2. Chen, W. The study on microstructure and corrosion resistance properties of duplex stainless steel. Mater. Sci. 2019, 9, 632–638. 3. Fan, Y.; Liu, T.G.; Xin, L.; Han, Y.M.; Lu, Y.H.; Shoji, T. Thermal aging behaviors of duplex stainless steels used in nuclear power plant: A review—Sciencedirect. J. Nucl. Mater. 2020, 544, 152693. [CrossRef] 4. Zhu, J.; Jiao, X.; Zhou, C.; Gao, H. Applications of Underwater in Nuclear Power Plant Maintenance. Energy Procedia 2012, 16, 153–158. [CrossRef] 5. Zhu, J.; Jiao, X.; Chen, M.X.; Zhou, C.; Hui, G. Research of chamber local dry underwater welding system and drainage properties. China Weld. 2013, 16, 27–29. 6. Zhou, C.; Jiao, X.; Xue, L.; Chen, J. Study on automatic hyperbaric welding applied in sub-sea pipelines repair. In Proceedings of the International Offshore and Polar Engineering Conference 2010, Beijing, China, 20–25 June 2010. 7. Zhou, C.F.; Jiao, X.D.; Zhu, J.L.; Gao, H.; Shen, Q.P.; Yu, Y.; Zhang, J.B. Study on Local Dry Welding of 304 Stainless Steel in Nuclear Power Stations Repair. Adv. Mater. Res. 2012, 460, 415–419. [CrossRef] 8. Zhang, X.; Ashida, E.; Shono, S.; Matsuda, F. Effect of shielding conditions of local dry cavity on weld quality in underwater Nd:YAG laser welding. J. Mater. Process. Tech. 2006, 174, 34–41. [CrossRef] 9. Feng, X.; Cui, X.; Zheng, W.; Lu, B.; Dong, M.; Wen, X.; Zhao, Y.; Jin, G. Effect of the protective materials and water on the repairing quality of nickel aluminum bronze during underwater wet laser repairing. Opt. Laser Technol. 2019, 114, 140–145. [CrossRef] 10. Guo, N.; Xing, X.; Zhao, H.; Tan, C.; Feng, J.; Deng, Z. Effect of water depth on weld quality and welding process in underwater fiber laser welding. Mater. Des. 2017, 115, 112–120. [CrossRef] 11. Guo, N.; Cheng, Q.; Zhang, X.; Fu, Y.; Huang, L. Microstructure and Mechanical Properties of Underwater Laser Welding of Titanium Alloy. Materials 2019, 12, 2703. [CrossRef] Materials 2021, 14, 4774 16 of 16

12. Tomków, J.; Janeczek, A.; Rogalski, G.; Wolski, A. Underwater Local Cavity Welding of S460N Steel. Materials 2020, 13, 5535. [CrossRef] 13. Gunn, R. Duplex Stainless Steels: Microstructure, Properties and Applications; Anti-Corrosion Methods and Materials; Woodhead Publishing: Cambridge, UK, 2003; p. 45. 14. Kiser, S.D. Welding and of Stainless Steels; John Wiley and Sons Ltd.: Hoboken, NJ, USA, 2009. 15. Park, Y.H.; Lee, Z.H. Effect of nitrogen and heat treatment on the microstructure and tensile properties of 25Cr-7Ni-1.5Mo-3WxN duplex stainless steel castings. Mater. Sci. Eng. A 2001, 297, 78–84. [CrossRef] 16. Hosseini, V.A.; Wessman, S.; Hurtig, K.; Karlsson, L. Nitrogen loss and effects on microstructure in multipass TIG welding of a super duplex stainless steel. Mater. Des. 2016, 98, 88–97. [CrossRef] 17. Hosseini, V.A.; Karlsson, L. Physical and kinetic simulation of nitrogen loss in high temperature heat affected zone of duplex stainless steels. Materials 2019, 6, 100325. [CrossRef] 18. Nowacki, J.; Łukojc, A. Microstructural transformations of heat affected zones in duplex steel welded joints. Mater. Charact. 2006, 56, 436–441. [CrossRef] 19. Chen, T.; Yang, J.-R. Microstructural characterization of simulated heat affected zone in a nitrogen-containing 2205 duplex stainless steel. Mater. Sci. Eng. A 2002, 338, 166–181. [CrossRef] 20. Sakai, Y.; Aida, G.; Suga, T.; Nakano, T. Development of various flux-cored wires and their application in Japan. In IIW/IISS Doc. XII-1131-89; Welding Company: Tokyo, Japan, 1989. 21. Miura, M.; Ogawa, K. Hydrogen embrittlement cracking in duplex stainless steel weld metal. In IIW Doc., IX-1461-87; Welding Company: Tokyo, Japan, 2000. 22. Zhang, Z.; Jing, H.; Xu, L.; Han, Y.; Zhao, L. Investigation on microstructure evolution and properties of duplex stainless steel joint multi-pass welded by using different methods. Mater. Des. 2016, 109, 670–685. [CrossRef] 23. Keskitalo, M.; Mäntyjärvi, K.; Sundqvist, J.; Powell, J.; Kaplan, A. Laser welding of duplex stainless steel with nitrogen as shielding gas. J. Mater. Process. Technol. 2015, 216, 381–384. [CrossRef] 24. Ornig, H.; Richter, M. WRC-1992-Diagramm löst DeLong-Diagramm ab. Schweiss. Schneid. 1997, 49, 467–470. 25. Baghdadchi, A.; Hosseini, V.A.; Hurtig, K.; Karlsson, L. Promoting austenite formation in laser welding of duplex stainless steel-impact of shielding gas and laser reheating. Weld. World 2020, 65, 499–511. [CrossRef] 26. Pouraliakbara, H.; Khalajb, G.; Jandaghia, M.R.; Khalajb, M.J. Study on the correlation of toughness with chemical composition and tensile test results in microalloyed API pipeline steels. J. Min. Metall. 2015, 51, 173–178. [CrossRef] 27. Khalaj, G.; Azimzadegan, T.; Khoeini, M.; Etaat, M. Artificial neural networks application to predict the ultimate tensile strength of X70 pipeline steels. Neural Comput. Appl. 2012, 23, 2301–2308. [CrossRef] 28. Karlsson, L.; Arcini, H.; Bergquist, E.-L.; Weidow, J.; Börjesson, J. Effects of Alloying Concepts on Ferrite Morphology and Toughness of Lean Duplex Stainless Steel Weld Metals. Weld. World 2010, 54, R350–R359. [CrossRef] 29. Kurt, B.; Orhan, N.; Ozel, S. Interface structure of diffusion bonded duplex stainless steel and medium couple. Sci. Technol. Weld. Join. 2009, 12, 1035–1040. [CrossRef] 30. Wang, S.; Ma, Q.; Li, Y. Characterization of microstructure, mechanical properties and corrosion resistance of dissimilar welded joint between 2205 duplex stainless steel and 16MnR. Mater. Des. 2011, 32, 831–837. [CrossRef] 31. Pan, C.; Zhang, Z. Morphologies of the transition region in dissimilar austenitic-ferritic welds. Mater. Charact. 1996, 36, 5–10. [CrossRef] 32. Celik, A.; Alsaran, A. Mechanical and Structural Properties of Similar and Dissimilar Steel Joints—Influence of weld metal on corrosion. Mater. Charact. 1999, 43, 311–318. [CrossRef] 33. Korhonen, S. Ruostumattomat Teräkset ja Niiden Soveltuvuus Konepajan Ohutlevytuotantoon; Lutpub: Lahti, Finland, 2010. 34. Outokumpu. Outokumpu Welding Handbook; Outokumpu: Helsinki, Finland, 2010.