Influence of filler wire on metallurgical, mechanical, and corrosion behaviour of 430 ferritic stainless steel using a fusion welding process

The advantage of ferritic stainless steels (FSS’s) over austenitic stainless steels is that they are less expensive alloys. This is due to low or negligible nickel in its alloying element which makes the steel affordable. This type of steel is highly recommended against chloride attack and is also machinable to produce various components for engineering applications. This study examines the effect of various fillers on 430 ferritic stainless steel (FSS). The austenitic (308) and ferritic (410) grades of filler were used to study the weldability, microstructure, mechanical properties, and corrosion resistance using tungsten inert gas welding. The findings showed the emergence of various complex phases in both the weld sample. The sample welded with 410 filler shows acicular ferrite, martensite and austenite. Whereas, austenite and vermicular ferrite are observed in the sample welded with 308 filler. Based on compositions and solidification modes, the mechanical properties of welded joints also vary. It was found that ferritic mode solidified welds dominated in terms of qualities, which was found in 410 filler. In the chloride solution, the behaviour of the pitting corrosion resistance of each weld varied. The sample welded with 410 was superior corrosion resistance. This is due to more δ-ferrite in the weld sample. Whereas, 308 showed poorer resistance against the simulated seawater solution. In 410 welds, a greater degree of sensitization was observed, as compared to 308 welds.


Introduction
Ferritic Stainless steels (FSS's) are inexpensive alloys and have the advantage of being more economical than austenitic stainless steels. This is due to the lack of nickel in its alloying element. Showing a lower coefficient of expansion compared to non-austenitic stainless steel, which is a great advantage when resisting cyclic loading [1]. Fuel filters, protection tubes, storage containers, electrical appliances, solar water heaters and household appliances are some of the applications of ferritic stainless steel. Also, FSS has material advantages over austenitic such as, improved machinability and stress protection against chloride corrosion cracking [2]. FSS is generally an alloy of iron and chromium having a body-centered cubic (BCC) crystal structure. Chromium content in the FSS range from 11 wt% to 30 wt%. However, the FSS has poor strength at high temperatures as compared to austenitic stainless steel [2].
The FSS has either a ferritic microstructure or a combination of ferritic and martensitic microstructure during welding [3]. The suppression of austenitic formation during solidification at elevated temperatures results in the formation of a fully ferritic structure during welding. However, the formation of martensite is achieved by the solid-state transformation or during the last stage of transformation during solidification [3]. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. microstructure development of FSS. As shown in figure 1 the phase to solidify from the liquid pool is δ-ferrite. Further cooling transforms δ-ferrite to austenite. This austenite phase again transforms to α-ferrite and carbides during equilibrium cooling conditions. However, the development of the martensite phase is due to the nonequilibrium cooling during solidification [4].
In fabrication industries, the joining of steel is generally done by the welding process. Various welding methods used for joining stainless steels are shield metal arc welding (SMAW), gas tungsten arc welding (GTAW), tungsten inert gas welding (TIG), laser welding, and friction welding. Within this welding process, SMAW and TIG are commonly used. Due to easy-to-use and economical welding processes. Ankur et al investigated on effect of welding processes on microstructure, mechanical properties, and corrosion behavior of low-nickel austenitic stainless steels [6]. It was found the TIG welding process shows better mechanical and corrosion properties as compared to the SMAW welding process. The major problem associated with the welding of FSS's is grain coarsening. The excessive grain growth can be controlled by the lowering of heat input in fusion welding or by introducing the nitride and carbide formers such as Al, B, V and Zr in FSS [7]. Thus, during welding, it suppresses the formation of grain growth. Also, the reduction of ductility has been reported by the author, due to the formation of secondary phases (i.e. σ phase) in the heat-affected zone [8]. Two forms of martensite phase have been reported during the applied heat input to the ferritic stainless steel. The formation of these phases is mainly due to the diffusion-less phase transformation, which is formed above the Ms temperature. The closed packed (111) planes in the austenite, ε-phase having hexagonal (hcp) crystal structure and secondly the ά-phase which has BCC or BCT structures [9]. Khorammi et al investigated the study on microstructure and mechanical characteristics of low-carbon steel and ferritic stainless-steel joints and observed that mechanical properties drastically increased by the addition of filler as compared to autogenous weld. It was due to the formation of martensite in the ferrite matrix in the fusion zone [10]. Subodh et al studied the effect of heat dissipation on the microstructure and mechanical properties of gas tungsten arc welded (GTAW) 304 SS joints. He reports that heat input affects the dendritic size and interdendritic distance [11].
Due to fusion welding, several brittle phases (such as the sigma phase and chromium carbides) can occur, impacting the corrosion resistance and mechanical characteristics of the welded joints [12]. Ambade et al investigated on corrosion behaviour of 409 M FSS. They observed that the increase in heat input decreases the corrosion rate. Also, they have investigated the welding process. They studied three welding processes namely GTAW, GMAW, and SMAW. They found lower corrosion resistance in GTAW than in other welding processes [13]. Cui et al reported that the austenite site is the preferential site for pit formation in a chloride environment [14]. The role of δ-ferrite has a significant impact on the behavior of pit in ASS [15]. The extra delta ferrite in the welding area results in better resistance to pitting [15]. However, it is reported that the role of δ-ferrite has a have detrimental effects on the toughness. The increase in δ-ferrite in the weld structure increases the ductilebrittle transition temperature. This results in a decrease in notch toughness and creep ductility at higher temperatures also there is a high chance to form precipitates at the weld [14].
From the above studies, it was observed that microstructure during solidification in welding plays important role in mechanical and corrosion behaviour. The change in microstructure can be obtained by varying the filler or by changing the heat input during welding. Therefore, in this study, the 430 ferritic stainless steel was welded using the TIG welding process and two different fillers (308 and 410 filler) were used to weld. Microstructural, mechanical and corrosion tests were performed to compare the welding process using different fillers.

Materials and method
The ferritic stainless steel was procured with 3 mm thickness in as-rolled and mill-annealed condition. The samples for welding were cut into desired dimensions of 250 mm (length) × 75 mm (width) using an electric discharge machine (EDM). The chemical composition of the base metal and fillers are given in table 1. Butt weld joints were prepared for welding with a gap of 1.2 mm. The welding was carried out using TIG welding with a constant current of 100 A. Before welding, the samples were cleaned properly with ethanol. This reduces the risk of producing welding defects during the welding operation, which is caused due to oil, grease and dust particle. Figure 2 shows the schematic representation of welding. It was reported that above 120 A weld defects occur. However, below 70 A there was a lack of penetration in the weld [16]. The pitting resistance equivalent number is calculated by equation (1).
The heat input was evaluated by using equation (2).
Where η is the efficiency of TIG welding which is taken as 0.6, I is the current in ampere, V is the voltage in volts and S is the welding speed in mm/s. The gas flow rate was 7 lit/min. Table 2 shows the heat input of both welding samples. After the welding is performed the weld samples were again cleaned. Figure 3 shows the schematic representation of the weld zone used for the calculation of dilution in the welding. Table 3 shows the dilution for different samples. Where A WD (A WD = A TR + A RR + A RG + A BF ) is the total area of the weld deposit, A TR is the top area of the weld, A RR is the root area, A BF is the area of base metal fusion and A RG is the area of the root gap.  Metallurgical analysis the samples were cut in the dimension of 4 cm (length) × 1cm (width). The samples were prepared as per ASTM E3-95 standard for microstructure in the transverse direction. Various series of emery paper were used, ending with 1200 grit size emery paper followed by cloth polish using diamond paste (0.25 μm). The sample was cleaned using an ultrasonic bath and ethanol. The sample was etched using the marbles reagent with a reaction time of 12s. The prepared sample was then placed under the metallurgical microscope (Zeiss AxioLab A1). The image was processed using image analyzing software. The ferrite content in the sample was measured using Ferrotiscope ((Fischer Ferritoscope FMP30) at the weld zone. Vicker's microhardness (Simadzu Micro-hardness Tester) was used to measure the hardness with a load of 500 gm for 10s holding time. The tensile test was carried out as per ASTM E-08M-04 by Instron universal testing machine (model-4467). Fractured ends of the tensile tested after failure were analyzed using a scanning electron microscope (SEM-JOEL 6380A Japan).
The corrosion test of the sample was performed using the potentiodynamic polarization test. Three cell electrode was used in which platinum acted as a counter electrode, a saturated calomel electrode (SCE) was used as the reference electrode and the working sample was used as the working electrode. The corrosion testing samples were cloth polished followed by cleaning with an ultrasonic bath using ethanol. Before the start of the experiment, the sample solution was purged for 1 h (nitrogen purging) followed by immersion of the sample for 30 min for stabilization at open circuit potential. For reproducibility, the sample data were reproduced three times. the electrolyte used for corrosion testing was synthetic seawater solution (3.5% NaCl solution). The potential was set from −300 mV (E ocp ) to 1200 mV at a scanning speed of 1.5 V/s.

Results and discussions
3.1. Predictive diagram Espy diagram was used as a productive diagram to measure the amount of ferrite present in the weld metal. Also, it was used to predict the microstructure in the weldment. The Cr eq and Ni eq were calculated according to equations (3) and (4) [17].  Figure 4 shows the distinct morphology of weld structures and different ferritic content in the fusion zone. The Cr eq and Ni eq were calculated for the base material also. Which resembles the solidification mode in the weld metal as shown in figure 5. From figure 4 it was observed that both the filler material has shown different microstructure. Weld sample joined resembles austenite and ferrite in its microstructure. Also, the ferrite content was found to be nearly 10 wt% for the sample welded with 308L filler whereas, the sample welded with 410 filler would have higher ferrite content in it. Figure 4 shows the solidification mode of both the filler material. The sample welded with 308 electrode shows austenite with lathy ferrite, due to a lower Cr eq /Ni eq   Figure 4 shows the solidification mode for various modes. The Cr eq /Ni eq value below 1.3 shows A mode, the Cr eq /Ni eq value between 1.3 -1.4 shows AF solidification mode, the value lies between 1.4 -1.8 shows FA solidification mode, and a value more than 1.8 shows F mode of solidification. In the present study sample  welded with 308L was found to be FA mode, whereas the sample welded with 410 filler shows F mode of solidification. It is well-known fact that the presence of δ-ferrite in the weld zone reduces the chances of solidification cracking. Also, δ-ferrite has the capability to absorb impurities during solidification [6]. Figure 6 shows the optical micrograph of base materials in as-rolled and mill-annealed conditions. The figure shows equiaxed ferrite grains in its microstructure with a small amount of carbides. Similar microstructure was reported by Zhou et al [18]. Figure 7 shows the optical micrograph of weld samples. Figures 6(a)-(c) shows the optical micrograph of the sample welded with 308 filler whereas, figures 6(d)-(f) shows the optical micrograph of the sample welded with 410 filler. The figure shows the HAZ region having coarse grains whereas the base metal has fine grains in its microstructure. The width of the HAZ region and length of inter-dendritic spacing was listed in table 4. It was observed that the nearly same width of HAZ was observed in both samples.  This may be due to the fact that almost the same heat input was applied to both samples. Also, the base material was the same therefore the coefficient of heat transfer was the same for both the welding samples. The average inter-dendritic spacing was also measured for both welding samples. The inter-dendritic spacing was observed more in the sample welded with 308 filler as compared to 410 filler. The sample welded with 308 filler shows an austenitic structure and vermicular ferrite. The formation of vermicular ferrite in the steel increases with the increase in the Cr eq /Ni eq ratio [19]. The vermicular ferrite in weld metal has a similar crystallographic orientation. It is reported that the molten metal when solidifies transforms to δ-ferrite at ∼1530°C and solidification completes at the temperature of ∼1455°C. Below this temperature the δ-ferrite transforms to γ austenite. The transformation of δ to γ continues till the metal cools down to 877°C. The solidification mode of the 308 filler was found to be FA mode. In which the liquid metal transforms to L + δ. This δ-ferrite again transforms into γ austenite L + δ + γ and finally to δ + γ. It has been reported that the fusion boundary for all the FA mode weld sample exhibit planer austenite [19]. The amount of δ-ferrite content in the fusion zone depends on the cooling rate of the sample. Slower the colling rate, the more chances of transforming the delta ferrite into austenite [20]. In the present study, the cooling rate was not sufficient to transform the δ-ferrite into full austenite. The sample welded with 410 filler shows austenite, ferrite, and some martensite in the microstructure. The formation of acicular ferrite is due to the faster cooling rate and the formation of fine ferrite in the microstructure. Intergranular nucleation is the major characteristic of acicular ferrite in the weld microstructure [21]. Similar results were obtained by Teker et al [22]. In the present study, there is some martensite formation in the fusion zone. This occurs due to the transformation of austenite which is retained between ferritic crystals during cooling [21]. In the present study, the solidification mode was found to be F mode for 410 filler, which was found using a Pseudo binary phase diagram. In this mode, the molten liquid transforms into L + δ. This L + δ transforms to a fully δ ferrite. This δ ferrite on further cooling transforms itself into δ + γ. It was reported that a higher cooling rate, lowers the formation of martensite and austenite in the F solidification mode [23]. The ferrite content for both weld zone was obtained using a ferritoscope as shown in table 4. It was found that more ferrite content in the weld metal of 410 filler as compared to the 308 filler. The findings of the above results validate the findings of the Espy diagram. Figures 8(a), (b) shows the XRD pattern of 430 SS welded with 308 and 410 filler. The figures show various sharp peaks of various phases. Figure 8(a) shows the sample having majorly two peaks. Namely austenite ( and ferrite after solidification of the sample welded with 308 filler. However, figure 8(b) in which the sample was welded with 410 filler shows the presence of austenite, ferrite and martensite in its structure. It was reported a small amount of δ-ferrite is necessary to avoid the problem of hot cracking during weld solidification. Fully austenitic stainless steel weld deposits are susceptible to microfissuring during cooling upon solidification [24]. The result of the XRD analysis justifies the finding of Espy diagram and microstructural analysis.

Mechanical properties 3.4.1. Tensile test
The result of the tensile test was recorded in table 5. It was observed that the ultimate tensile strength of the base material was found to be 510.8 MPa. The sample welded with 430 base metal with 308 electrodes was observed to be 640.5 MPa and the sample welded with 410 filler was observed to be 730.4 MPa. The increase in tensile strength in the sample welded with 410 filler is due to the more ferrite content in the fusion zone. It is reported that martensite formation in the fusion zone enhances the tensile strength in the weld sample [25]. Moreover, the inter-dendritic spacing is less when compared to the sample welded with 308 filler which favours the increase in strength. Also, the heat input was slightly higher in the 410 samples. An increase in heat input helps to dissolve the carbide present in the sample, which homogenizes the sample and makes the grain finer [26]. This resulted in improved strength of the tested sample. The elongation of the weld metal was more than compared to the weld samples. This is attributed to the change in chemical composition for both the weld and base metal and also the change in microstructural changes involved in the fusion zone resulting in a decrease in elongation percentage [25].  Figures 9(a), (b) shows the microhardness of both weld samples. The microhardness value of weld metal and base metal was found to be approximately 195 HV (308 weld sample) and 252 HV (410 weld sample) for weld zone and, 220 for the base metal. In figure 9(a) there is a sudden increase in the hardness near the fusion boundary region followed by a decrease in the hardness value. Figure 7(b) shows the sample welded with 410 filler. It was observed that there is an increase in the hardness value (252 HV) in the fusion zone as compared to the sample welded with 308 fillers. This is attributed to the formation of martensite in the ferrite grains and the precipitation of intergranular carbides. It is reported that a decrease in grain size increases the hardness of the material. Followed by a decrease in the hardness value in the HAZ region. Which is due to the grain coarsening in the microstructure. There may be two reasons for the increase in hardness value near the fusion boundary. First is the formation of martensite formation at the ferrite grain boundary and the second one is the carbide precipitate on the grains. The formation of carbide precipitate is due to the formation of M 23 C 6 . when the carbide precipitates form there is a depletion of chromium near the grains boundary which concentrates at the    [27]. Also, there is a reduction in the hardness value near the fusion boundary (HAZ region). This reduction is mainly due to the grain coarsening. As compared to austenitic stainless steel FSS is more prone to grain coarsening. When the temperature (sensitization temperature) is maintained at a particular duration causes the grain to enlarge and cause grain coarsening [28]. The grain coarsening results in a decrease in the toughness of the material. Figures 10(a), (b) shows the fracture surface of both samples. It was observed that the sample welded with 308 filler shows a fine dimple as shown in figure 10(a). This is attributed to the large plastic deformation of the sample before failure. Similar observations were observed by Karthick et al [29]. However, figure 10(b) shows the quasi-cleavage pattern followed by flat facets over the surface when welded with 410 fillers.

Fractography analysis
3.6. Corrosion behaviour 3.6.1. Potentiodynamic polarization test PDP test at 3.5 weight percent NaCl solution is shown in figure 11. It has been noted that 308 filler has a higher corrosion current density (i corr ) than weld. The discrepancy between the i corr for 410 (4.521 × 10 −7 ) and 308 filler (5.589 × 10 −5 ) was caused by more ferrite in the microstructure of 410 filler. The value of E corr is active for 308 fillers (−0.290 V) as compared to 410 fillers (−0.195 V). All of the welded samples exhibit some passive behaviour. It was observed that 410 filler welding samples had slightly lower pitting potential (−0.030 V) as compared to 308 filler samples (−0.026 V). Exceptional levels of delta ferrite in the weld are primarily responsible for high corrosion resistance in 410 sample [30]. Figure 12 shows the pitting micrograph after the potentiodynamic polarization test. Ankur et al investigated two different welding processes namely SMAW and  TIG welding. They found out that pit initiation and propagation start from the austenitic phase [6]. In the present investigation, we can attribute that majority of the pits initiates from the austenitic phase and cause pitting at the weld sample. As the 308 weld sample shows more austenitic phase than the ferritic hence there are more pits on the sample as compared to other weld samples. It is reported that discontinuity in the phases restricts the growth of pits and improves the corrosion rate. Hence more ferrite phase in the 410 filler helps improve the pitting resistance than the 308 weld sample [31].

Double loop electrochemical reactivation test
DLEPR curve of the welded sample with filler 308 and 410 has been shown in figure 13. It is reported that DLEPR curve is used to calculate the degree of sensitization (DOS) [32]. During experimentation, two types of scan take place, i.e., forward scan during which maximum current indicates activation current (I a ) and reverse scan where maximum current shows reactivation current (I r ) [33]. The degree of sensitization (DOS) is calculated by taking the ratio of I r / I a [34]. From figure 13 it is observed that I a and I r of 308 filler are 0.034 and 0.011, whereas for 410 filler I a and I r are 0.031 and 0.018, respectively. So DOS of 308 is 32.35% and 410 filler gives nearly 46.87%. From the calculated result, it is evident that DOS of 410 filler welded material is more compared to 308 filler weld. This is due to the more δ-ferrite content in 410 filler welded material. In 410 filler, more chromium diffuses to grain boundaries and formed chromium carbide which increases sensitization [35]. However, in the case of 310 filler, it is known that chromium is a ferrite stabilizer so the solubility of chromium content is very high in ferrite and hence content of chromium remains very less for diffusion to grain boundaries which gives results in less formation of chromium carbide compare to 410 filler which results in lesser DOS.