Effect of the Shielding Gas and Heat Treatment in Inconel 625 Coatings Deposited by GMAW Process

Abstract

Friction, wear, and corrosion of engineering components operating in harsh environments can be substantially improved by applying hard, corrosion-resistant coatings to prolong their useful lives. Nickel superalloys are particularly relevant due to their excellent mechanical properties and corrosion resistance at elevated temperatures. Among the various coating techniques, arc welding processes are suitable due to their good deposition rate and reliability. This work aimed to evaluate the effect of the shielding gas and after-deposition heat treatment on the microstructure and mechanical properties of Inconel 625 coatings deposited by the GMAW process. The coatings were deposited onto carbon steel plates using two mixtures of shielding gases (Ar+25%CO₂ and Ar+25%He) without interpass temperature control. The specimens were analyzed both as welded and after heat treatment (heating for 1 h at 1000 °C and air cooling) using Vickers hardness tests, scanning electron microscopy, energy-dispersive X-ray spectroscopy (EDS), and wavelength dispersion spectrometry (WDS). The coatings that used Ar+25%He-shielding gas were harder and showed more precipitate formation, which was associated with the higher cooling rates involved. As for the heat treatment, it led to a reduction in the segregation of the alloying elements in the interdendritic region via diffusion and a reduction in surface hardness.

1. Introduction

In coastal and offshore installations, material deterioration due to corrosion and wear can decrease reliability and increase maintenance costs of engineering components. To mitigate this, deposition techniques using wear and corrosion-resistant coatings can increase the useful life of engineering components in harsh environments [1]. This strategy enables us to reduce costs and optimize performance by combining cheaper and easier-to-manufacture substrates with more expensive and noble coating materials [2].

Inconel 625 is a nickel-based superalloy predominantly used as an overlay coating in critical industries such as aerospace, chemical industry, electrochemical, power generation plants, turbines, nuclear plants, and oil and gas drilling [3,4] due to its excellent corrosion resistance and high mechanical strength, particularly at elevated temperatures [5,6]. When produced under equilibrium conditions, the mechanical strength at high temperatures of Inconel 625 derives from the solid solution hardening effect of the refractory metals (niobium and molybdenum) in a nickel–chromium matrix [7]. However, welding Inconel 625 to produce overlays can lead to the formation of a complex microstructure where niobium and molybdenum may form precipitates varying from γ’ and γ” precipitates to carbides, Laves, and δ phases [8].

Different welding techniques have been used to produce Inconel overlays. Low-dilution coating processes such as cold metal transfer [9], laser [10,11], plasma transferred arc (PTA) [2,4], and gas tungsten arc welding (GTAW) [12] minimize the alteration of the chemical composition of the coating with elements from the substrate and reduce segregation, ensuring coatings with superior properties [13]. Alternatively, there is also the possibility of using gas metal arc welding (GMAW), which, on the one hand, leads to substantially greater dilution but, on the other hand, promotes greater production capacity [6,8]. Among the processes cited, GMAW presents the highest production capacity, the lowest equipment cost, and is the easiest to operate. Moreover, it can be easily automated [14]. As an overlay method, the use of multilayers is normally required to reduce the dilution of the substrate into the GMAW coating [15].

Most works in the literature are limited to using inert shielding gases (particularly pure Ar) to deposit Inconel 625 coatings by GMAW due to the inherently high dilution of the process [16,17]. However, the arc is more unstable in a pure Ar atmosphere. Moreover, the surface tension of the molten pool is higher with pure Ar, producing more convex beads. This may lead to discontinuities (welding defects) in the coatings produced by GMAW [18]. Alternatively, mixtures of Ar with active gases (e.g., CO₂) can potentially lead to a more stable arc and reduce the convexity index of the beads, thus reducing defects in the coatings at the expense of even higher dilutions. In previous works, our groups have deposited Inconel 625 coatings using GMAW under Ar+25%He and Ar+25%CO₂ shielding gases with interpass temperature control [6,14,18]. The microstructures of the coatings, verified via scanning electron microscopy (SEM) coupled to energy-dispersive X-ray spectroscopy (EDS), were composed of γ (Ni,Cr-CFC) dendrites surrounded by an interdendritic phase rich in Nb-Mo precipitates and carbides (Nb,Mo MC type and Cr₂₃C₆). The use of Ar+25%CO₂ shielding gas led to greater dilution when compared to Ar+25%He [18].

The precipitates formed during welding can affect the corrosion resistance, mechanical properties, and tribological performance of Inconel overlays produced by arc welding [19], in particular when the dilution of the substrate into the coating is high [20]. A very influential factor in the microstructure of the Inconel deposits is the realization of the post-weld heat treatment (PWHT). Some authors have investigated the influence of PWHT on different characteristics of Inconel 625 overlays [13,15,21]. The temperature and time used in PWHT can modify the microstructure and properties of the cladding. Heat treatments at around 1000 °C have been shown to decrease internal stresses by reducing the density of dislocations and, consequently, the coating hardness [10,13,15].

This work aimed to evaluate the effect of the shielding gas and heat treatment on the microstructure, hardness, and chemical composition of Inconel 625 GMAW weld overlays onto carbon steel surfaces to select the parameters that provide better mechanical properties for applications in aggressive environments, aiming to reduce maintenance and manufacturing costs.

The main scientific contributions of this work were to evaluate the effects of using Ar+25%CO₂-shielding gas on coatings applied by the GMAW process, carried out without stopping (without interpass temperature control). This can potentially reduce the cost of producing the coatings, but it is not commonly used, in addition to analyzing the effects of heat treatment on the Inconel coatings produced under the conditions studied.

2. Materials and Methods

2.1. Materials

ASTM A-36 carbon steel sheets with a thickness of 10 mm were used as the base material (substrates), whereas ER NiCrMo-3 wire (Inconel 625) with a diameter of 1.2 mm was used as filler metal. The chemical composition of the substrate and filler metal wire can be observed in

Table 1. Chemical composition of the base material (ASTM A-36) and filler metal wire (Inconel 625), in weight percentage (%).

Material	Ni %	Cr %	Mn %	Si%	Al %	Ti %	Fe %	C %	Mo %	Nb %
ASTM A-36	0.02	0.02	0.67	0.09	0.03	-	Bal.	0.23	-	-
Inconel 625	Bal.	22.46	-	-	0.26	0.26	0.02	0.02	8.84	3.46

. Two different shielding gas compositions (Ar+25%He and Ar+25%CO₂) were used to analyze the influence of He and CO₂ on the microstructure and microhardness of the coatings. Ar+25%He was selected because it is an inert gas widely used in applications with Inconel 625. Ar+25%CO₂ was selected because it leads to good wettability and has low cost, seeking to evaluate the applicability of this shielding gas, despite not being commonly recommended for applications with Inconel 625 since it tends to increase dilution.

2.2. Surface Coating Methods

The weld overlays were applied via the GMAW process, according to

Table 2. Welding parameters.

Welding voltage	25 V
Feeding speed	10 m/min
Welding speed	25 cm/min
Contact tip distance	17 mm
Shielding gas flow	16 L/min
Number of layers	2

. The equipment used was a Digiplus A7 source and a Yaskawa^® HP20 robot (Yaskawa, Kyoto, Japan). A coating technique was used with an overlap of 50% of the previous bead to create each layer. No stopping was carried out between the deposition of the individual beads, i.e., the layers were produced without interpass temperature control. This approach aimed to investigate if it was possible to reduce the coating deposition time without compromising the microstructure and hardness of the overlays. Two layers were made with each shielding gas to reduce the dilution effect, meeting the minimum requirements proposed by the NACE MR0175 standard [22].

2.3. Sample Preparation

After the deposition of the coatings, the samples were cut with dimensions of 35 mm × 25 mm × 9 mm and then ground to obtain flat surfaces.

Table 3. Nomenclature used for each sample.

Nomenclature	Shielding Gas	Condition
HE	Ar+25%He	As welded
CO2	Ar+25%CO2	As welded
HET	Ar+25%He	After heat treatment
CO2T	Ar+25%CO2	After heat treatment

summarizes the conditions and nomenclature used for each sample. The heat treatment was carried out at a temperature of 1000 °C for one hour, followed by cooling in open air without agitation. Such treatment is intended to reduce internal stresses (reducing the density of dislocations) and, consequently, a hardness reduction in the Inconel 625 coating, as carried out by Li and Wei [10], meeting the requirements proposed by the NACE MR0175 standard [22]. It must be pointed out that the heat treatment should also change the microstructure of the substrate, but the substrate has very little effect on the surface properties of thick coatings, such as friction, wear, and corrosion. For that reason, microstructural changes in the substrate were not in the scope of this work, but in the case of a thin coating, the microstructural changes in both the coating and the substrate should be addressed.

2.4. Metallographic Preparation

The samples were mounted with a semiconductor resin, polished with water sandpaper (grit sizes 600, 800, 1200, and 2000), and then polished with diamond paste (1 μ). To observe the microstructures under SEM, the samples were not etched, in order to avoid that some very small precipitates could be removed during etching due to loss of mechanical support when the matrix around them is etched.

2.5. Hardness Tests

Hardness tests were carried out using Mitutoyo model HV100 equipment with a Vickers indenter at a load of 10 kgf applied for 30 s, following the ANSI/NACE MR0175/ISO 15156-2 standard [22]. Ten hardness measurements were performed directly on the surface of each cladding sample.

2.6. Scanning Electron Microscopy and X-ray Spectroscopy Analyses

Scanning electron microscopy (SEM) analysis was performed on Jeol JSM IT300LV equipment under an electron acceleration voltage of 15 kV using backscattered electron (BSE) detection mode, which provides chemical contrast. Since the specimens were not etched, secondary electron (SE) images were not used. Compositional analyses were carried out by energy-dispersive X-ray spectroscopy (EDS) using an Oxford Instruments XMax-n 80 X-ray detector coupled to the microscope.

Wavelength dispersion spectrometry (WDS) tests were performed in a JEOL JXA-8230 electron microprobe using five WDS detectors with artificial layered dispersive elements (LDE1 and LDE2), thallium acid phthalate (TAP), pentaerythritol (PET/L-H), and lithium fluoride (LIF-L/H) crystals. A JEOL energy dispersive spectrometer (EDS) coupled to the microscope was also used for control. Analyses were performed under a current condition of 15 kV and 50 nA with a beam aperture of 70 µm. Si was detected with TAP, Cr, Ti, and Mo with PET/L, Ni and Fe with LIF-H, and Nb with PET/H.

X-ray diffraction (XRD) tests were carried out in a Bruker D8 Discover diffractometer, using monochromatic radiation from a tube with a copper anode coupled to a Johansson monochromator for Ka_1 operating at 40 kV and 40 mA, Bragg Brentano θ–2θ configuration, Lynxeye one-dimensional detector, 2θ range from 3° to 100°, with a step of 0.01°. The samples were kept rotating at 15 rpm during the measurements.

3. Results

3.1. Electronic Microprobe Analysis

Figure 1 shows images generated by a scanning electron microscope in backscattered electron (BSE) detection modes. The samples showed dendritic microstructure. The effect of the 50% overlap between the weld beads can be seen in Figure 1a,c. It can also be observed that there was a slight decrease in the overlapping effect between the beads due to the heat treatment, as seen in Figure 1b,d. The microstructures have a dendritic morphology, characteristic of Inconel 625 in the as-welded condition [4,6,16]. According to Ahn and Yoon [20], in austenitic alloys such as Inconel 625, which contain Nb and Mo, the non-equilibrium solidification that occurs during welding can promote the segregation of these elements to the interdendritic regions, which are the last to solidify. This segregation increases the solidification range of the alloy and promotes the final solidification of a low-melting-point, Nb-rich eutectic.

The chemical composition of the coated surfaces in each condition, quantified by WDS, can be observed in

Table 4. Chemical composition of coated surfaces in weight percentage (%).

Sample	Ni %	Cr %	Mo %	Nb %	Fe %	Ti %	Si %
CO2T	62.90	22.75	8.74	3.44	2.29	0.20	0.08
CO2	62.98	22.80	8.76	3.51	1.84	0.18	0.08
HET	64.33	23.12	8.95	3.60	0.14	0.26	0.08
HE	64.03	23.13	8.95	3.69	0.10	0.26	0.08

, which agrees with the nominal composition of Inconel 625 [23,24]. However, there was a variation in the mass percentage of the element Fe in the samples welded with CO₂, and the HE samples were compared. This greater amount of Fe results from the greater dilution induced by CO₂ compared with He, as also observed by Macedo and Celente [18]. CO₂ presents a higher ionization potential and thermal conductivity than He; thus, the energy transferred to the melting pool is higher [23], increasing the diffusion of iron from the substrate to the top of the coating. It is important to point out that this Fe dilution should be higher when approaching the interface with the substrate [9].

3.2. Scanning Electron Microscopy Analyses

Figure 2 shows SEM images in backscattered electron (BSE) and secondary (SE) detection modes for the coatings deposited with CO₂ without subsequent heat treatment. The images in (A) and (B) show a general view of the coating, where the BSE image in (a) evidences the dendritic structure and the SE image in (E) evidences small pores in the coatings since SE images only provide topographical contrast. The porosity found in the samples was already expected, given that the CO₂ shielding gas during the welding process adds the element oxygen to the welding puddle, which promotes an increase in porosity in the weld bead. The magnification BSE image in (C) shows a lighter interdendritic region, richer in elements with higher atomic size. Further magnification (D) evidences the presence of small white precipitates. The microstructure observed in Figure 2 agrees with the literature for welded Inconel 625 [3,14,16,18]. The dendrites appear darker in color, while the interdendritic regions appear lighter, suggesting that the interdendritic regions are rich in elements with a higher atomic number. This results from the segregation of elements such as Nb and Mo into the interdendritic regions. Due to the high content of these elements, precipitates form at some points, which can be of the MC type ((Mo, Nb, Ti) C), Laves phase (Ni, Fe, Cr)₂(Nb, Mo), phase γ” Ni₃(Nb, Mo), and δ phase (Ni₃Nb), as found by other authors in the literature [2,4,12].

According to Durandcharre [13], for welded coatings, the behavior of the solid/liquid interface is governed by a high-temperature gradient that quickly solidifies the molten metal. Dendritic solidification promotes a variation in the chemical composition between the center of the dendrites and its immediately adjacent region because the first regions to solidify reject the solute into the remaining liquid as solidification occurs (solute segregation). According to Cieslak and Headley [24], the solidification of Inconel 625 begins with the formation of the austenitic γ phase in the form of dendrites and continues with the formation of second phases and carbides. The segregation of elements such as Nb, Si, and C into the interdendritic region makes solidification prone to forming secondary phases at the end of solidification through an eutectic reaction leading to γ + various Nb-rich phases such as Laves and Nb carbides. Higher carbon contents normally promote the formation of Nb carbides, while Si promotes the formation of the Laves phase. According to DuPont [25], weldability reduces with increasing Nb, Si, and C content, which is attributed to the decrease in the solidification temperature range and the tendency to form second phases at the end of solidification. The equilibrium process of alloy solidification in welding occurs through the reactions: L → L + γ → L + γ + NbC → L+ γ + NbC + Laves. The solidification of the pool begins with the L → γ reaction, which causes the accumulation of elements such as Nb, Ti, and C in the intergranular regions and the liquid part of the grain contours. Subsequently, the MC-type carbide (M can be Nb or Ti) precipitates in these regions through the eutectic reaction L → γ + NbC, which consumes a large part of the C present. Elements such as Nb, Cr, and Fe also accumulate in the intergranular regions and in the liquid films present at the grain boundaries, which promotes the formation of the Laves phase (Fe₂(Nb,Ti,Mo)) through the reaction: L → γ + Laves. Finally, the simultaneous presence of Ni and elements such as Al and Ti causes the dispersed phase γ’ to precipitate, causing hardening and increasing the mechanical properties of the weld metal [24,25,26].

Post-welding heat treatment of the coatings deposited using CO₂ increases the size of the interdendritic regions and reduces their chemical contrast in relation to the dendrites, as shown in Figure 3. Comparing the images of the samples in the as-welded state (Figure 2D) and after heat treatment (Figure 3C), heat treatment leads to diffusion of the elements segregated in the interdendritic region. This region becomes wider but shows a darker color in Figure 3 when compared with the interdendritic region before heat treatment shown in Figure 2. This difference in color is caused by the diffusion of the heavier elements, which appear brighter in the BSE contrast.

Figure 4 shows that the use of He gas led to slightly different microstructures, where a high number of precipitates are observed in the interdendritic regions. Moreover, the precipitates are larger and the dendritic arms are thinner than for the deposits using CO₂ (compare Figure 2 and Figure 4). Macedo and Celente [18] measured higher cooling rates when welding Inconel 625 with He than with CO₂, probably justifying the thinner dendrites. The cause of more intense precipitation when using He must be associated with the intense segregation of Nb and Mo in the interdendritic regions [25,27] caused by the higher cooling rates. Analyzing the images of the samples in the as-welded state and after heat treatment, it is noted that the chemical contrast between the dendrites and the interdendritic region almost disappeared after heat treatment (Figure 5). To verify which elements are present in the interdendritic regions, EDS maps were obtained, as shown in Figure 6, Figure 7, Figure 8 and Figure 9.

EDS maps for all the conditions are presented in Figure 6, Figure 7, Figure 8 and Figure 9. The maps for Mo and Nb confirm that the precipitates observed in the interdendritic region are rich in those elements before heat treatment, both for CO₂ (Figure 6) and He (Figure 8) shielding gases. Some Ti precipitates can also be observed, particularly for the He shielding gas (Figure 8). In the solidification process, the liquid transforms into the γ phase, leading to an accumulation of Nb, Mo, and Ti in the spacing between the dendrites and at the grain boundaries. Due to this, the Laves phases, MC-type carbides (including MoC, NbC, and TiC), M₆C, and the intermetallic δ-Ni₃Nb can precipitate in these regions. O maps showed that this element was uniformly distributed on the surfaces for all the samples, as exemplified in Figure 6, Figure 7 and Figure 8, except for some oxygen that coincided with Ti, suggesting the formation of a small amount of TiO₂ for the specimens welded with CO₂, but not observed for weldings with He. The Fe maps showed that the element was uniformly distributed for all the samples, as exemplified in Figure 9 for the HET sample. Comparing the state as welded and after heat treatment for overlays using CO2 (Figure 6 and Figure 7) and He (Figure 8 and Figure 9), the maps suggest the dissolution of the precipitates in the interdendritic region caused by the heat treatment.

Comparing the welded state of the samples that used He gas and CO₂, in agreement with the studies by Macedo and Celente [18], the use of He compared to CO₂ gas provided greater plasma conductivity. The higher conductivity of the plasma caused an increase in the cooling speed of the weld pool and, consequently, an increase in the amount of segregation during the solidification of the alloy. According to Floreen and Fuchs [28], the formation of the Laves phase, δ phase, and/or Nb carbides plays an important role in the solidification of Inconel 625, strongly affecting its mechanical and corrosion properties.

3.3. X-ray Diffraction Analysis

X-ray diffraction (XRD) measurements were carried out to determine the crystalline phases present in the coatings using the Xpert High Score Plus Software. It is observed in Figure 10 that in all samples, the phases can be considered similar; that is, a γ Ni,Cr phase was identified with face-centered cubic symmetry, a Laves phase (Fe₂(Nb,Ti,Mo)), and carbides as secondary phases. The formation of carbides in the precipitation hardening process is not common in nickel alloys solidified under equilibrium conditions. However, in the welded condition, carbon may react with elements added to these alloys to form carbides, the most common of which are MC, M₆C, M₇C₃, and M₂₃C₆ [29]. Carbides of the M₂₃C₆ type, in addition to the decomposition of the primary carbide, can also be formed by residual carbon in the alloy matrix. They are often present at grain boundaries but can also be found along the boundaries of twins and stacking faults. The formation of this carbide is favored in alloys with high contents of Cr, Ti, and Al, while high contents of Mo and W tend to more easily form the M₆C type carbide, while high contents of Nb and Ta promote the formation of the carbide primary MC [30]. With heat treatment, the relative intensity between precipitates and the γ Ni,Cr phase decreased, again indicating dissolution of the precipitates. The carbide peaks were less intense with He than with CO₂ shielding gas.

According to the micrographs and the identification of the phases via X-ray analysis, the formation of the γ(FeNi) phases constituting the matrix and interdendritic region with dispersed carbides was verified, in addition to the presence of the Laves phase.

3.4. Hardness of the Coated Surfaces

The results obtained during the Vickers hardness test are shown in Figure 11. The as-welded samples that used He shielding gas had a higher average hardness compared to the samples that used CO₂ gas. The average hardness of the as-welded samples that used He gas converges with the results of Li and Wei [10], who attributed the increased hardness of welded samples to the rapid cooling involved (approximately 380 °C/s). Therefore, hardening elements such as molybdenum and niobium precipitate in the nickel matrix, inducing an intense formation of dislocations in the precipitated regions and the formation of secondary phases such as Laves. Since He increases the conductivity of the plasma, there is an increase in segregation, resulting in the observed increase in hardness. Moreover, lower iron dilution from the substrate into the coating was observed for He shielding gas compared to CO₂, further contributing to the increased hardness of the as-welded HE specimen. It is relevant to point out that a higher iron content via dilution of the substrate is detrimental to the corrosion resistance of the coatings. The approach used to minimize the problem in this work was the use of a two-layer coating.

As expected, comparing the samples as welded and after heat treatment, a tendency to reduce hardness is observed. These results agree with those presented by Li and Wei [10]. The authors, by raising the heat treatment temperature to 1000 °C for 1 h, achieved a reduction in the hardness of the Inconel 625 coating via the dissolution of elements such as Nb and Mo and secondary phases in the matrix.

The heat treatment showed greater hardness reduction for the samples that used He shielding gas. This occurred because the hardness of the CO₂ samples before heat treatment was lower. The lower cooling rate that occurs when welding Inconel 625 with CO₂ [18] leads to more iron in the coating due to the dilution of the substrate, as well as reducing the number of precipitates, reducing the coating’s hardness.

The hardness limit recommended by NACE MR0175 is 250 HV (22 HRC) for welded joints subjected to aggressive environments. The results often obtained in the literature [10,31] suggest the need for subsequent thermal treatments to improve and adapt the coating properties. With GMAW deposition, this work showed that weldings with CO₂ comply with the NACE MR0175 Standard, while the HE samples exceeded the limit pre-established by the standard.

4. Conclusions

This work investigated the microstructure of Inconel 625 deposited by GMAW using different shielding gases in the as-welded and after-heat treatment conditions. Regardless of the type of shielding gas used, an austenitic microstructure was observed with a dendritic shape (cellular and equiaxed dendritic) and segregation of the elements Nb and Mo into the interdendritic regions, which are the last to solidify.

The coating carried out using He shielding gas generated more precipitates in the interdendritic regions associated with intense segregation of Nb and Mo than with CO₂ shielding gas, as well as higher Fe content and coarser dendrites.

With both shielding gases, heat treatment caused diffusion in the interdendritic region, but this was more intense in samples that used He gas. Heat treatment also reduced hardness via the dissolution of secondary phases in the matrix.

The as-welded samples that used He shielding gas had a higher average hardness compared to the samples that used CO₂ gas, which was related to the rapid cooling to which the sample was subjected after the welding process. Heat treatment showed a greater reduction in the hardness of the coated surface in the samples that used He gas.

In the conditions studied (without waiting for interpass temperature), the samples welded with Ar+25%CO₂ complied with the NACE MR0175 Standard, while the He sample exceeded the established limit.