Studies on the effects of cryogenic cooling on microstructure and mechanical properties of plasma arc welded SS 316

The stainless steel of grade 316 has significant use in nuclear engineering, aerospace industry and submarines and has become a material of choice due to its diversified properties. While welding this material, the heat input supplied weakens the fusion zone by grain growth, widening the heat-affected zone (HAZ), reducing yield and tensile strengths, and introducing distortion. Therefore, to address these issues, this study investigated the effect of cryogenic cooling during the welding process of thin sheets of SS 316 to improve the microstructure, mechanical properties, and reducing the distortion of the welded material. The keyhole mode plasma arc welding process was used to weld in a single pass without using filler with three different cooling methods. The microstructures, microhardness, and tensile properties of cryogenically and conventionally cooled weld samples were investigated at room temperature. The microstructural behavior of samples was characterized by metallurgical microscopy and scanning electron microscopy. The SEM Analysis reveals γ austenite and δ ferrite phases in conventionally welded test samples. M23C6 is formed in small amounts from δ ferrite, surrounding the δ ferrite on grain boundaries. In cryogenic cooled samples, delta ferrite is detected on grain boundaries of the austenitic matrix. Additionally, traces of (Cr, Fe)2 N are also revealed in specific cryogenic cooled samples due to liquid nitrogen impingement. In gel cooled samples, M23C6 is shown due to comparatively prolonged duration of cooling, and the results reveal that the liquid nitrogen and thermo gel improved average grain size up to 83.53% and 66.84%, respectively, as compared to an average grain size of conventional plasma weld. The reduction in HAZ is observed to be about 43.38% and 7.92% for liquid nitrogen cooled and thermo-gel samples, respectively, compared to conventional weld. Moreover, the tensile and yield strength of liquid nitrogen-cooled weldments increased up to 22.28% and 28.96%, respectively, while for gel-cooled welded sample, a 10.50% improvement in tensile strength and 3.10% in yield strength was observed. Furthermore, a reduction of 75% in distortion is achieved for welded samples with liquid nitrogen cooling.


Introduction
AISI stainless steel (SS316), a particular grade of Austenitic Stainless Steel(ASS), stands out for its superior corrosion, radiation resistance, and robust mechanical properties at elevated temperatures. These properties make it suitable for high-strength structures for high-pressure and temperature applications, such as pressurized water and nuclear reactors [1][2][3][4]. In addition to its standard industrial applications, SS316 is also preferred for medical, pharmaceutical, and food processing due to its ability to maintain weld integrity, rust-free qualities, and ease of sensitization [5]. It is also completely recyclable, making it a sustainable and environmentally friendly choice for various applications. As a result, SS316 steel alloy has become the material of choice for a wide range of industrial products subjected to high temperatures and pressures, such as jet engine parts, heat exchangers, evaporators, tanks, the oil industry, and submarines [6][7][8].
The robust properties of SS316 are due to its alloying elements including carbon(0.08%), silicon(0.75%), chromium(16%-18%), nickel(10%-14%), molybdenum(2%-3%), and nitrogen(0.1%) [9]. Carbon increases strength and reduces intergranular corrosion resistance at high temperatures due to carbide formations [10], while chromium imparts corrosion resistance by forming a thin layer of chromium oxide on the surface when it reacts with oxygen. However, chromium can react with carbon at high temperatures to form chromium carbides, reducing corrosion resistance [11]. Nickel helps stabilize the austenitic microstructure at room and low temperatures, while molybdenum increases resistance to pitting caused by certain acids and improves heat and wear resistance [12]. Manganese and nitrogen improve machinability, decrease brittleness, and enhance flow strength and corrosion resistance [13].
A passive oxide film that is formed on the surface of SS316, makes it stainless and protects it from rusting. This film being refractory material, has high melting point and during the welding process make it challenging to achieve a qualified weld joint. It requires a high level of focused heat input to remove it, which can lead to weld burn-through and prevent complete joint penetration. Under this scenario, a technique is required that ensures precise heat input control for high-quality welds and is cost-effective. Plasma arc welding (PAW) has emerged as a viable option for joining thin structures of SS316 due to its ability to produce a concentrated arc with a small heat-affected zone (HAZ). This reduces the risk of weld defects and minimizes the spread of heat input into the material due to its focused arc, reducing the likelihood of weld burn-through. Additionally, PAW offers comparable weld quality to laser and electron beam welding in the advanced manufacturing of thin structures, but at a relatively lower cost.
During welding SS316 alloy, PAW maintains the quality of the weldment by using a keyhole mode that generates a sharp, focused, high-velocity, and high-temperature plasma beam capable of removing the thin, invisible, corrosive film formed by the reaction of chromium and molybdenum with oxygen on the sample surface [14]. Figure 1 shows a schematic of the keyhole mode of PAW. In this process, a constricted and concentrated plasma arc is generated between a tungsten electrode and a narrow orifice in a constraint nozzle, surrounded by a shielding gas nozzle to protect the arc and fusion zone from the environment [15]. The resulting plasma-based heat input is used for welding the sample plates. This welding technique is preferred because it produces narrow bead welds with minimal thermal effects on the base material due to the concentrated plasma beam.
The high heat concentration can penetrate the joint in a single pass, removing the passive oxide film and other contaminants and producing a sound weld [16][17][18]. However, conventional welding techniques, including PAW, generate high thermal energy during welding, leading to undesirable residual stresses in the welded structures. These stresses can negatively impact the material properties of welded structures, such as corrosion resistance, fatigue strength, and distortion [18,19]. Keyhole plasma arc welding with dynamically controlled low stress no distortion (DCLSND) mode was introduced to address these issues. This mode minimizes the high heat input during welding, reducing associated residual stresses, and improving the microstructure and mechanical properties of the welded structure. One widely adopted technique in DCLSND welding is trailing cryogenic cooling medium during the welding process to reduce the weld-induced stresses. Cryogenic cooling is effective in reducing residual stresses in welding processes. The National Aeronautics and Space Administration (NASA), as part of their Saturn program, first demonstrated the significance of cryogenic cooling in welding aluminum airframes, showing that it can alleviate warpage and residual stresses, and initial experimental investigations found that the use of cryogenic liquids and auxiliary heat led to a 5% decrease in residual stresses [20]. However, the mechanism behind this reduction was not understood, and no mathematical model existed to explain the cooling process. Therefore, the first 2D model to analyze the effects of heat sink thermal capacity and distance on welding with a variable plasma arc process was introduced, but the model overpredicted the effects, indicating the need for a 3D model [21]. Thus, a 3D ellipsoidal model of welding heat input was developed, which led to the introduction of a dynamically controlled low stress no distortion (DC-LSND) welding technique and applied it to non-linear and non-regular welds, which eliminated distortion, but further investigation was needed to understand its mechanism and control [22]. Toward this end, a first simulation study was conducted, which used liquid nitrogen(LN2) to study the effects of cryogenic cooling on the hot cracking of welds, and found that an increase in standoff distance led to an increase in hot cracking [23]. Based on the results and limitations of previous studies, in the last two decades, a thorough investigation of the mechanism behind the reduction in welding residual and buckling stresses due to cryogenic cooling was conducted. The results found that the use of a trailing heat sink with increasing cooling strength and decreasing distance from the welding arc can eliminate buckling distortion in thin plates, but it was also revealed that no significant reduction occurred in longitudinal stresses in stainless steel(SS) materials when using LN2 as coolant [24,25]. Other similar studies investigated the effects of coolant flow rates, revealing that after a certain threshold of coolant flow rate is achieved, the cooling intensity does not increase further [26]. The combination of GTAW with LN2 trailing at a variable distance behind the welding arc (DW) was investigated, and the results revealed a reduction in longitudinal stresses and out-of-plane distortion [27].
In the past few years, numerous studies have examined the effects of trailing cryogenic coolants at a fixed distance behind the welding arc on various aspects of welding, including temperature distribution, residual stresses, and weld distortion. Techniques such as Friction Stir Welding (FSW) [28], Gas Metal Arc Welding (GMAW) [29][30][31] have been tested with various cryogenic cooling mediums, including liquid nitrogen (LN2) and liquid carbon dioxide (LCO2). These studies have found that cryogenic cooling can reduce distortion and residual stresses, increase tensile properties and microhardness, and improve grain size in welded joints. Inprocess cooling has also been found to enhance grain size and mechanical properties [32], while CO2 gas has been effective as a post-weld coolant in reducing residual stress [33]. In contrast to planar welds, an SS316 circumferential weld quality was evaluated to examine the effects of LN2 impingement with Gas Tungsten Arc Welding (GTAW), and results revealed a notable reduction in the weld distortion while enhancing the delta ferrite contents in SS 316 [34]. Additionally, cryogenic treatment has been found to reduce residual stresses, refine microstructure, and increase impact strength in Tig welding of AISI 304 and laser beam (LW) and electron beam (EBW) welds [35]. However, in the case of dissimilar welding of Al 6061-T6 and mild steel using FSW, the high-strength weld joint was produced using water as a trailing coolant, while trailing LN2 resulted in the lowest tensile shear strength of the lap joint [36].
A detailed summary of the recent studies on trailing cryogenic cooling in welding processes, as discussed above, is presented in table 1. It can be seen that although various studies have investigated the potential benefits of cryogenic cooling on the weld quality of thin structures with different welding techniques but the implementation of PAW in DCLSND mode is yet to be investigated. Therefore, the innovative aspect of current study is that the use of trailing LN2 in combination with PAW was explored on the weld quality of 2 mm thin SS316 sheets. As per author's knowledge, no other study has examined the PAW in conjunction with trailing LN2 as a heat sink. The effectiveness of this LN2 cryogenic cooling was validated based on analysis of microstructure and mechanical properties of weld samples. Additionally, a novel thermo-gel was developed, and its influence on microstructure and mechanical properties of the weld samples was studied. It is worth mentioning that thermo gel is a very innovative concept, and no published literature, according to our knowledge, has investigated its effects on the microstructure and quality of the weld. The main contributions of this study are as follows: • Design and modification of the welding torch of the plasma arc auto-linear welding machine for trailing liquid cooling medium behind the welding torch.
• Optimization of the welding process parameters using the response surface methodology to achieve highquality welds.
• Plasma arc welding of 2 mm thin sheets of SS316 with liquid nitrogen as cryogenic cooling medium. • Investigation and analysis of microstructure and mechanical properties of welded specimens employing different material characterization techniques, including radiography analysis, metallography study, SEM analysis, Tensile Testing, and Micro-Hardness Testing.
• Investigation and analysis of the effects of liquid nitrogen cooling on the width of the HAZ and weld-zone grain size.
• Investigation of the effects of trailing a novel thermo gel-based heat sink on the weldment's grain size and mechanical properties.

Material and methods
The methodologies adopted in this study are briefly discussed in the following sub-sections, including process parameter optimization, modification of welding setup, welding of samples, and material characterization techniques.

Optimization of plasma weld parameters
The different process parameters of PAW were optimized by conducting several trial runs designed using the design of experiment (DOE) method in Minitab. These experiments were designed with three main objectives: reduction in heat input supplied during single pass welding, achievement of complete joint penetration of weld bead without using the filler wire, and improving the microstructure and, ultimately, the mechanical properties of welded samples. A randomized two-level factorial matrix was built, and nineteen trail runs designed by DOE were conducted using arbitrary welding parameters based on previous welding experience to identify the significance and operational range limits of input parameters and output responses. After screening factors and identifying the main effects, interactions, and their effective operational ranges, the matrix of input process variables was designed for required experimental runs for optimization. After thirty designed experimental runs, the matrix was updated for achieved responses and analyzed to achieve optimized parameters. The optimized input parameters and resultant responses were further validated by welding four different samples on these parameters and achieving satisfactory results. Afterward, these experimentally found optimized process parameters were used to weld five test samples with Identification Numbers (ID) of 1M, 2M, 3M, 4M, and 5M. Sample 1M represents a naturally cooled sample (natural convection), 2M-4M samples represent liquid nitrogen cooling (forced convection), and sample 5M relates to the trailing thermo-gel cooling.

Modification of welding torch for cryogenic cooling
The welding torch of the conventional PAW machine was modified to facilitate the impingement of cryogenic cooling medium on the surface of the weld sample. One end of an SS tube is attached with bolted strip to the torch assembly, at a specific distance from the welding torch, for the impingement of cryogenic cooling medium (LN2) on the sample with three different jet nozzle diameters. On the other end, this tube was attached to a flexible SS pipe connected to a Dewalt (a double-walled flask of metal or silvered glass with a vacuum between the walls, used to hold liquids well below ambient temperature) LN2 storage container. Adjacent to this cooling nozzle, a separator of bake lite material was placed between the welding torch and liquid nitrogen cooling jet to avoid arc blow and disturbance of the weld pool during welding. The modified torch assembly on the PAW machine in the actual experimental setup is shown in figure 2 .Moreover, figure 3 shows the detailed schematic and corresponding experimental setup. Table 2 presents the specifications of the welding torch and nitrogen jet. Moreover, the cooling medium is trailed at a flow rate of 1.5 liter min −1 (l/m), and the distance of the liquid jet nozzle from the welding arc is optimized to a value of 70 mm after conducting various experiment runs conducted by varying this distance in a range of 40-100 mm. These experimental results revealed that the standoff distance below 70 mm extinguishes the welding arc and the value above that retards the cooling effect.
Furthermore, the elevation of the jet nozzle from the test sample base is also optimized to a value of 20 mm obtained as a result of varying this elevation in a range of 5-30 mm for different experimental runs. It is worth mentioning that the use of liquid nitrogen may cause a potential hazard in the form of frostbite due to its significantly lower boiling point (−195.8°C) and suffocating characteristics, and an extremely high expansion ratio of 1:696. Such characteristics of liquid nitrogen require strict adherence to safety measures, including air change every 10 min in the working area [37], which was achieved by blowing off four large exhaust fans during welding to control the hazard.

Welding of test samples
Two SS316 plates of dimensions 300 × 150 × 2 mm were used for each test in this study and are welded on an auto-control weld machine, as shown in figure 4(a). The LN2 delivery system for cryogenic cooling weld samples is shown in figure 4(b). These samples in a rolled and annealed condition are welded in a single pass (without filler) with a square butt joint and nil gap using the optimized process parameters as shown in figures 4(c) and (d). Moreover, the 2% thoriated tungsten electrode of 2.4 mm diameter with an electrode-to-work angle of 90°w as used on an auto linear welding machine with 4.5 lpm argon gas flow for plasma formation. The PAW was used for welding all samples, but a different cooling configuration was used for each sample. For instance, test sample 1M was left to naturally cool while LN2 was impinged on the subsequent three samples (2M, 3M and 4M) with different nozzle diameter (1 mm, 1.2 mm and 1.6 mm) respectively. Figure 5 shows the conventionally welded and cryogenically cooled welded samples after post-weld cleaning. Lastly, sample 5M was cooled with a newly developed thermo-gel, and the process is shown in figure 6. Precautions were taken during the DCLSND welding process to prevent frostbite or eye damage from the liquid nitrogen. Four pressure exhaust fans were  frequently operated to avoid suffocation and any other adverse effects in the working area because the vaporization of 1-liter liquid nitrogen expands to a 24.6 ft 3 of nitrogen gas volume.

Development and application of a novel thermo gel
The use of thermo gel in PAW of SS316 thin sheets is a very innovative concept. The gel is a semi-solid that can have properties ranging from soft and weak to hard and tough [38][39][40].This is a gel concentrate which when applied on a hot surface adheres to the surface and initially quenches the heat out of the surface and then acts as a thermal insulator when cured. Thermo gel properties can be easily altered for a specific application via substitution and modification of components in di block and tri block copolymer systems [41]. The composition of gel is as follows: 30%Poly vinyl acetate (PVA) glue (C4H6O2), 60% Corn starch (C6H10O5)n and 10% Baking Soda (NaHCO3) by weight. The constituents are blended with some drops of water to make its slurry for ease of application. Figure 6(a) demonstrate the application of thermo gel during the welding process while figures 6(b) and (c) show the post weld state of the thermo-gel cooled sample 5M.

Characterization of welded samples
Different material characterization techniques were implemented to thoroughly analyze the effects of cryogenic cooling on the microstructure and mechanical properties of welded samples. The microstructure analysis of welded samples was conducted using metallographic testing (MT), radiography testing, and scanning electron microscopy (SEM), while the mechanical properties were evaluated based on tensile and microhardness testing of these samples.  2.5.1. Surface preparation and microstructure analysis of welded samples The metallographic samples of 25 × 25 mm were extracted from the FZ and HAZ based on ISO 15614-11 sampling scheme using a water jet cutting machine, then cross-sectional parts were cut from these samples by hacksaw and cold mounted with Bakelite for further processing using cloth polishing machine. These samples were initially grinded using SiC 100-1200 μm thick emery papers followed by further polishing on an electric polishing machine with three different grades (6 μm, 3 μm, and 1 μm) of diamond paste. These polished samples were then cleaned with a mixture of detergent and ethanol, and the sample preparation was then carried out according to the ASTM E-3 standard. Moreover, the sample preparation was initiated with the etching of the sample using a Kaling reagent, followed by an electrical etching. The former reagent constitutes 2 g CuCl 2 , 33 ml HCl, 33 ml HNo 3, and 33 ml distilled water, while the latter includes 10 g oxalic acid 100 ml distilled water 6volt DC for 15 s for grain size. Furthermore, a Leica-DM6-Z metallurgical microscope with MC170 HD camera HD and LAS 4.12 software was used for the metallography of samples. The final samples Furthermore, the Kaling reagent reveals the internal structure of welded samples. The microstructure and elemental composition of the selected welded samples were examined using a thermionic emission SEM (TESCAN VEGA3) with a tungsten-heated cathodebased electron gun. In contrast to optical microscopic analysis, SEM analysis provides enhanced micrograph resolution and facilitates very large magnifications without any deterioration in the quality of micrographs. Moreover, the SEM system examined the samples in the secondary electron mode, which involves detecting secondary electrons emitted from atoms excited by an electron beam with an Everhart-Thornley detector.

Microhardness and tensile testing of welded samples
The cross-sectional test samples of 50 × 25 × 2 mm size for the microhardness test were extracted using Replica Kit, followed by further pre-processing, including cutting, grinding, and polishing of samples before rinsing in detergent and ethanol solutions. The prepared samples were then etched with fresh etchant aqua regia with a ratio of 1:3 HNO 3 and HCl. At this stage, the microhardness test was performed with an applied load of 980.7mN over a dwell time of 12 s. The plain tensile samples for the tensile test were extracted according to the ASTM E 8M standards, while the welded samples were prepared according to standard ISO 4136, as shown in figure 7, by laser cutting machine. These tensile samples were tested at 21°C (ambient temperature) under dynamic loading conditions on a universal testing machine which allows a maximum loading of 250KN. The test results were segregated based on ASTM A666 standard, and the selected tensile Samples were tagged with ID Numbers as 0T,1T,2T,3T,4T, and 5T. The sample 0T represents a plain tensile, 1T conventionally welded sample while the sample 5T refers to trailing gel cooled, and the remaining samples(2T,3T,4T) represent cryogenically cooled samples for LN2 jet nozzle diameters of 1 mm,1.2 mm,1.6 mm.

Optimization of welding parameters
In this study, the welding process parameters are optimized to achieve an optimal combination of process parameters that ensure high-quality welds. The main parameters include bottom shielding gas flow rate, welding current, welding voltage, and welding speed, and the optimized value was achieved by investigating the influence of these parameters on weld bead penetration. The process continued till qualified weld joints for a specific combination of process parameters was achieved, which were found to be 20 liters per minute (lpm) of shielding gas flowrate, 100 A current, 24V voltage, and 300 mms −1 welding speed. Five test samples were then welded using these parameters on a welding setup shown in figures 2-4, and all of the test samples showed complete joint penetration with an adequate crown profile of the weld bead without any undercut, underfill, or any other visual imperfections. Furthermore, a constant value of heat input during welding was achieved by using these optimized values of process parameters which also facilitates the comparison of results of welded samples for various welding configurations.

Radiographic analysis
Radiographic testing (RT) of selected welded samples was performed to investigate the internal structure of the weld bead to segregate the qualified samples from the rest of the samples. Radiography of welded samples was performed based on the standard test method ASME V with IQI ASTME 747 using a smart 160E x-ray tube. The generated x-ray images were captured by Carestream imaging plate using Carestream industrex Digital Viewing Software 4.1. Figure 8 shows the radiographs of the weld bead of welded samples for three unique welding configurations, including conventionally welded, cryogenically cooled welded, and gel-cooled samples. The visual analysis of the  radiographs revealed that no welding defects were observed for conventional and cryogenically cooled welds. However, for the thermo-gel cooled sample, white spots outside the weld bead were observed, representing the thermocouple failure due to the insulation effect of the thermo-gel after curing, which led to the overheating of the thermocouple. The overall results demonstrated that all selected samples passed the RT check.

Macrographic analysis
A macro examination is principally used for evaluating the quality of welds. The Leica-DM6-Z metallurgical microscope was used on 12× magnification, and results revealed complete joint penetration (CJP), Complete side wall fusion, and fusion zone free from other imperfections, as shown in figures 9-11.

Effect of conventional welding on microstructure
Metallographic testing (MT) is used to investigate the weldments for thermal analysis, buried imperfections, visual defects, contaminations, and cracks to find the weld failure's root cause. Figures 12(a) and 12(b) show 100× and 50× magnified microstructure of base metal which reveals the presence of two material phases: ferrite and austenite. Generally, the initial form of manufactured ASS after solidification and ambient cooling contains small traces of the ferrite phase, which is retained on grain boundaries and in the matrix of austenite, as shown by figure 12 [42]. The various fusion zones and microstructure of conventionally plasma arc welded samples are shown in 50× magnified figure 12(c). The analysis reveals that the fusion zone (FZ) consists of a coarse dendritic structure formed because of the slow cooling of the sample by the natural air convection, just like in casted metals. Adjacent to the FZ zone of the weldment, a narrow, partially melted fine grain region is observed at the  weld interface as it is in direct contact with cold base metal; a rapid heat transfer occurs due to conduction resulting in refined grain size.
Moreover, this region is enriched with alloying elements, which increase the brittleness of the welded sample, and because of the pull of adjacent weld metal when contracted on solidification, the stresses are also induced in the sample, which may cause potential weld failure during tensile testing [43]. Another development is that the grain shape in this region transitions from an intermixed cellular structure at the boundary of the weld metal region to a columnar structure at the HAZ boundary. Adjacent to the partially melted weld interface, the results reveal the heat-affected zone (HAZ), a portion of non-melted base metal whose microstructure is affected by the welding heat conducted across the width of this region. Moreover, two different structures are revealed in HAZ: a coarse-grained region and a fine-grained region formed due to continuous non-uniform heating and cooling during welding process. It is worth mentioning that the presence of two different grain structure weakens the zone increasing the probability of material failure when sample is loaded or undergoes tensile testing. Lastly, it is evident from the microstructure of conventionally plasma arc welded sample that it has four different grain sizes in it: dendritic or columnar in the weld bead area, intermixed cellular in partially melted weld interface, coarse and fine grains composite in HAZ, and the regular grain structure of the base metal. This simultaneous occurrence of various grain size and structure reduces the overall strength of welded sample when compared with base metal as demonstrated by tensile testing results of the samples.

Effect of cryogenic cooling during welding on microstructure
The effect of liquid nitrogen as a cooling medium on grain size and mechanical properties of welded samples was investigated. The microstructures of three cryogenically cooled welded samples, 2M, 3M, and 4M, for nozzle diameters of 1 mm, 1.2 mm, and 1.6 mm, respectively, are shown in figure 13. The microstructure analysis reveals that the increase in the cooling nozzle diameter results in grain size refinement of the welded sample because of reduction in heat input during welding. Moreover, two distinct developments were observed, one in a partially melted zone and the other in HAZ. It was observed that the partially melted zone was diminished under forced convection due to nitrogen cooling, and the grain structure in HAZ was revealed to be more refined in the center but with columnar grain growth on the upper and lower side of samples. This grain growth is more prominent in sample 2M, associated with a 1 mm cooling nozzle than in the remaining two samples, 3M and 4M samples, where the reduction in grain growth is observed with an increase in cooling nozzle diameter from 1.2 mm to 1.66 mm.
3.6. Effect of novel thermo-gel on microstructure A novel homemade thermo-gel based on starlight material was developed and analyzed for its influence on the grain size, structure, and mechanical properties of the welded sample. The liquified gel was trailed on the weld bead behind the welding arc across the whole length of the welding sample. Initially, the gel acted as a heat sink because of the trapped moisture content, which produced the quenching effect of liquid, thus extracting heat from the sample, but as the gel is cured with continuous heat extraction, it starts to act as a thermal insulator after a few minutes. Figure 14 shows the gel-cooled sample's microstructure, and the results' analysis reveals the presence of the partially melted zone and HAZ with coarse-grained and fine-grained grain structures. Moreover, the continued presence of a partially melted region is attributed to the rapid curing of the liquified gel such that the quenching of the liquefied gel did not last long enough for a partially melted zone to diminish completely, as in cryogenically cooled samples. Furthermore, a refinement of grain structure is noted in FZ on the lower side of weldment, where the grain size is observed to be comparatively smaller than that of conventional weld but is greater than that of cryogenically cooled samples.

HAZ width analysis
The HAZ is a crucial region in any welded sample as it is characterized by physical and mechanical properties different from weld metal and adjacent base metal because of microstructural changes due to the high heat input provided during welding. These variations in microstructural properties can be significant learning to weld failures [42] and thus introduces a need to thoroughly investigate the HAZ. In this study, the width of HAZ is measured on the Leica-DM6-Z metallurgical microscope, and the average values of HAZ width are calculated, shown in figure 15 shows the variation in averaged HAZ width for different samples. The results show that the lowest HAZ width is obtained for sample 4M, representing cryogenically cooled welded samples. This HAZ width is due to the intense cooling by impinging liquid nitrogen jet during the welding process such that heat is rapidly quenched from the weld sample, thus reducing the width of HAZ.

SEM analysis
The SEM analysis of selected welded samples was conducted to investigate the surface properties of these samples welded under three different PAW configurations. Figure 16 shows micrographs of four different welded samples, including one conventionally welded and three cryogenically cooled welded samples. Figure 16(a) reveals the presence of a few whiteish-grey spots of carbide near the grain boundaries formed    because of the carbide precipitation in the sample coupled with blackish ferrite content deposited on the grain boundaries. The micrograph of welded sample 2M cooled by a 1 mm cooling nozzle is shown in figure 16(b), which reveals the presence of (Cr, Fe) 2 N and δ-ferrite content in the welded sample like all other cryogenically cooled samples. The (Cr, Fe) 2 N is formed as a result of liquid nitrogen impingement and retained δ-ferrite is due to the rapid cooling effect that reduced the transformation of δ ferrite into austinite.
Further analysis found that the (Cr, Fe) 2 N is generally formed in the matrix while the retained δ-ferrite content is found to be deposited on the grain boundaries. As the diameter of the cooling nozzle increases, the percentage of these elements is also significantly reduced because of the more prolonged contact of liquid nitrogen coolant on the sample surface, as demonstrated in figures 16(c) and (d). The tensile test results also validate this phenomenon on welded samples 3M and 4M associated with 1.2 and 1.6 mm diameter cooling nozzle, respectively. Figures 16(e) and (f) show the microstructure of the gel-cooled sample, which only reveals the formation of carbides precipitated near the grain boundaries because of the slow cooling of the welded sample. This slow cooling is attributed to the short time till the complete curing of liquified thermo-gel, transforming the gel into a thermal insulator. Additionally, an Energy Dispersive Spectroscopy (EDS) analysis was conducted to analyze the elemental composition of welded samples and to support the microstructure results further. Analysis results revealed the elemental percentages found in the Fusion zone (FZ) of conventionally welded samples, cryogenically cooled samples, and thermo gel cooled samples, as shown in figure 17. In the cryogenically cooled sample, the presence of nitrogen was confirmed in the Fusion zone (FZ) which demonstrates that impingement of LN2 was effective.

Microhardness analysis
A microhardness test for five samples was conducted to investigate the strength of the welded sample's FZ, HAZ and base metal. Figure 19 shows the macrograph of the specific welded sample at 10× magnification, revealing the micro indentations of the microhardness test performed by the Vicker hardness tester. Further detailed analysis of all tested samples found that the highest microhardness value is associated with FZ followed by the partially melted zone and then the value decreases in HAZ due to relative coarse grain structure compared to the base metal. A numerical comparison of the microhardness for different welded samples shown in figure 18 reveals that the cryogenically cooled samples have higher microhardness values than gel-cooled conventionally welded samples. Moreover, the highest microhardness value is observed for sample 4M associated with the largest diameter of the cooling nozzle, which consequently cools the sample more effectively and refines the grain size.

Tensile strength analysis
The tensile strength is a common yet crucial factor in the performance evaluation of any welded sample as it directly estimates the maximum load the weld can bear, which is significant in the design and fabrication of welded structures. Table 3 presents the results of tensile testing of six different samples, including conventionally welded, cryogenically cooled, and thermo-gel cooled samples. The numerical results include ultimate tensile strength(UTS), yield strength(YS) and elongation at 30 mm gauge length. Based on the anlysis of these results, the cryogenically cooled sample 4T is ranked at the top with the highest value of UTS (675 MPa) and YS (374 MPa), while the plain sample 0T and conventional weld sample 1T is ranked at the bottom with the lowest  values of YS (272 MPa) and UTS (552 MPa) respectively. The relatively high strength of the nitrogen-cooled sample is due to the formation of nitrides on the sample surface, as revealed in the microstructure results. These nitrides formed due to the impingement of liquid nitrogen on the sample surface at elevated temperatures drastically increase the overall sample strength because even a minimal amount (up to 0.08% weight) of nitrogen can significantly improve the tensile and yield strength [43,44].
Moreover, the gel-cooled sample showed greater tensile strength than the conventionally welded sample compared to the plain tensile sample with the highest tensile strength. This behavior can be explained by the higher precipitation of carbides, as revealed in the microstructure study, which decreases the sample ductility but does not significantly impact the tensile and yield strength compared to the behavior of base metal [45,46]. These tensile testing results are validated by associated stress-strain curves and other mechanical properties, as shown in and figures 20 and 21. All the selected samples were tested on the UTS machine, and broken tensile samples are shown in figure 22.

Fractographic analysis
The selected broken tensile samples 0T, 1T, 2T, 3T, 4T, and 5T were further analyzed by fractography using SEM (TESCAN VEGA3). The results of the 0T sample revealed ductile dimples and micro voids, indicating a ductile fracture, as shown in figures 23(a) and (b). The conventional welded sample showed comparatively fewer dimples, more micro voids, and cleavage, indicating a brittle fracture, as shown in figures 23(c) and (d). All three cryogenic samples indicated a transition from ductile to brittle fracture as dimples are decreasing and cleavage and striations are increasing, as shown in figure 24. Higher ductility of these samples showed an increasing amount of δ ferrite due to rapid cooling of weld. Higher UTS and Y.S may be due to nitrides formed by LN2 impingement, as revealed by EDS.  Moreover, it can be observed that the increase in LN2 nozzle diameter tends to refine the grain structure of the samples. Figures 25(a) and (b) revealed the formation of comparatively larger dimples and a smaller number of micro voids in the thermo gel cooled weld sample. Carbides formed in this sample may have contributed to its UTS and Y.S.

Distortion analysis
Distortion analysis of weldment is significant because it directly relates to weld quality and associated repair costs incurred due to a post-weld treatment of straightening the distorted welded structures. The present study quantified the weld-induced distortion by a digital multi-function weld inspect gauge and feeler gauge in the longitudinal and transverse direction for different welded samples. Figure 26 shows the weld-induced distortion in conventionally welded and cryogenically cooled samples, and table 4 presents the same results for all five samples. The comparative analysis of the results ranks the cryogenically cooled sample 4M on top with a

Conclusions
The main objective of this research study was to improve the microstructural and mechanical properties of SS316 to produce lightweight and high-strength fabricated structures. In this study, the influence of trailing liquid nitrogen as a cooling medium during PAW on the microstructure and the mechanical properties of the thin SS316 welded plates were experimentally investigated. Furthermore, the effect of trailing a novel thermo gel during the welding on the overall weld quality was also analyzed. The main conclusions and analysis results are summarized as follows: • The application of cryogenic cooling results in a refinement of about 83.53% in the average grain size and a 43% reduction in the HAZ width of the welded structure. It also reduced the longitudinal and transverse distortion up to 75.51% and 74.68%, respectively. The comparison of the conventional and cryogenically cooled welded samples revealed an improvement of up to 22.28% UTS and 28.96% YS in cryogenically cooled welded samples. Furthermore, the metallographic analysis of the cryogenically cooled welded samples also revealed nitride formation, which has enhanced the tensile and yield strengths. Moreover, the rapid cooling of  liquid nitrogen has given less time to transform δ ferrite to γ austenite, resulting in the deposition of retained δ ferrite on the grain boundaries, which has contributed to increased ductility of the welded structure.
• The thermo gel, based on the properties of its constituents, has been observed to impart quenching and heat insolating effect on weldment as dictated by the resulting properties of the weldment. Furthermore, the promising effects of homemade thermo gel on grain structure and mechanical properties are demonstrated by  the improvement of up to 10.5% and 3.1% in UTS and YS, respectively, in thermo-gel based welded samples as compared to conventionally welded samples. Compared to the conventionally welded samples, the gelcooled welded sample exhibited an average grain size refinement of about 66.84% and a reduction of 54% and 65% in longitudinal and transverse distortion, respectively. The development and application of thermo gel is a fundamental study in its nature; therefore, the detailed thermal properties of the gel are not available so far.
• Future research direction includes investigating the effect of thermal properties of developed gel on weldment.
This research serves as guidelines for the plasma arc welding of SS316 with cryogenically cooling of weldment for lightweight and high-strength fabricated structures of diverse applications, including automobiles, aerospace, submarine cruise launch systems, and pressure vessels of nuclear plants.