Monitoring of thermo-cycles in fibre laser welding of duplex stainless steel 2205 sheets and its correlation with microstructures and mechanical properties

The study reports the influence of change in the heat supplied (43 J mm−1 to 18.5 J mm−1) on the microstructures as well as mechanical properties of weld joints obtained by welding of Duplex stainless steel 2205 using fibre laser. In-process thermal monitoring of the molten weld pool was carried out using IR pyrometer. Cooling rates (i.e. solidification and solid) were calculated from the thermo-profiles of weld pool, and it increases with the decrease in heat input. From the optical images, it is observed that columnar grains originated from the fusion zone walls and merged at the center. Since, solidification front velocity is comparable on both sides’ leads to a central edge. Ferrite phase content observed in fusion zone microstructure, increases with the increase in solid cooling rate. The result suggests that the joints fabricated at lowest heat input displayed highest tensile strength. The maximum tensile strength been reported to be 872.5 ± 10.8 MPa, and failure occurred at parent metal. Tensile strength of weld joints of DSS 2205 was found to have improved with increasing cooling rate. Higher cooling rate results in the formation of fine dendritic grains as well as higher ferrite content in the weld metal.


Introduction
The Duplex stainless steel 2205 (DSS 2205) have nearly equal proportion of austenite and ferrite phase and it is cheap in comparison of austenitic stainless steel due to addition of lower amounts of Ni.The equal phase combinations in DSS 2205 facilitate excellent mechanical properties and good corrosion resistance.It is commonly used in fabrication of heat exchangers, offshore structural components, water treatment plant devices [1,2].Despite being readily weldable, heat input is the key factor in welding of DSS 2205 to achieve desired microstructure and avoid formation of deleterious phases.Various welding methodologies such as shielded metal arc (SMA), Gas tungsten arc (GTA), Gas metal arc (GMA), and Friction Stir welding (FSW) have been used to study the joining of DSS 2205 alloy [3].McPherson et al carried out welding of 2205 DSS using submerged arc welding.The weld zone was observed to have ferrite structures and acicular austenite.Precipitation in the weld metal of DSS was scarce and occasional amount of chi-phase was witnessed in the HAZ [4].Production of intermetallic phases and embrittlement, constitute a significant risk, markedly reducing both corrosion resistance and toughness of the weld joint [5].For example, the sigma phase (Cr-rich phase) can cause considerable reduction in plasticity and impairment in corrosion resistance [6,7].Taban carried out welding of DSS using plasma arc welding in the key-hole mode by using direct current and without any filler.Heat input was varied from 1.82 kJ mm −1 to 1.96 kJ mm −1 and single pass welding was executed while being shielded by pure Argon gas.Grains structure similar to what is obtained during rolling process were formed in the weld metal when heat supplied was lower.In the fusion pool the main solidified part is primarily ferrite.This is due to the quick chilling taking place and therefore diffusional transformation to austenite is suppressed [8].Sathiya et al carried out GTA welding of DSS 2205 sheets and examined the effect on the micro-structure in weld due to changing of shielding gas.They reported that helium shielding results in higher aspect (width/penetration) ratio than argon shielded weld metal.Also, welds achieved by shielding with helium demonstrated better toughness due to the presence of high Mn along with incidence of high austenite phase and lesser amount of ferrite phase [9].Heat input should be supplied in controlled way to obtain favorable microstructures.Köse and Kaçar, reported that low heat input and fast cooling rate of LBW results in formation of delta ferrite phase in weld metal of duplex stainless steel.Delta ferrite phase results in rise of tensile strength and hardness of weld joints, and failure takes place in weld metal [10].If the heat input was too low, precipitation of chromium nitride and higher concentration of ferrite content were reported [11,12].On the contrary, welding at high heat input or when the weld is subjected to high temperatures in the range of 700 °C-1000 °C for long duration, inter-metallic phase such as χ (Chi) or σ (Sigma) precipitates formed which are very brittle thereby deteriorating the properties of the joint [13].Mourad et al carried out comparative study between GTA and CO2 laser beam welding of 2205 DSS.Weld produced using LBW has smaller FZ and has a more satisfactory weld profile when compared to that of GTAW.This smaller FZ obtained by LBW has better resistance to corrosion than that of GTA welding [14].It has been reported by many authors that σ phase has a drastic negative effect on the mechanical properties of the weld joint along with causing reduced capacity to corrosion.Therefore, it is considered as the most dangerous phase.Chi (χ) phase is formed after σ phase and is found to be rich in Molybdenum.Kordatos et al studied the effect of cooling rate on microstructure, strength and corrosion resistance of 2205 duplex stainless steel weld joints.The authors reported that cooling of weld metal by water is faster when compared to air-cooling and therefore, it caused a decrease in width of the ferritization zone to 500 μm, which is 600 μm when cooled by air.Cooling by water caused harder weld metal when compared to air-cooling because of enhanced formation of harder ferrite in terms of volume fraction [15].From the literature review, it has been found that there is lack of literature available on thermal monitoring of laser welding of DSS 2205.Thus, the present study focuses primarily on monitoring of thermal history that includes surface peak temperature and rate of cooling of the weld pool during fiber laser welding of DSS 2205 sheets at different heat input.Further, the rate of cooling is correlated with the evolution of the microstructure of fusion zone, and their influence on the mechanical properties of weld joints was investigated.

Materials and methods
DSS 2205 sheets of dimension 30 mm × 60 mm × 1 mm was used for experimental work.Microstructure comprising a dual phase i.e., ferrite and austenite grain, was observed for DSS 2205, as is shown in figure 1. Microstructure of the parent material had a volume fraction of ferrite and austenite in approximately equivalent proportion with grains of roughly the same size.The chemical composition obtained from spectrochemical analysis is given in table 1. Edge preparation of respective abutting edges was carried out on milling machine to ensure flatness of the edges and then cleaned with acetone bath to remove surface contaminants.Before experiment specimens were clamped properly using rigid fix-tures to avoid gap between edges of two sheets and to minimize angular distortion of weld specimen.Welding experiments were performed using Yb-fibre laser (make: IPG, Germany, model: YLR 2000) with nominal power supply of 2.15 KW and having mul-timode beam intensity profile.Laser head was affixed on a 5-axis CNC workstation that could move at a speed up to 20 m min −1 effectively.Raw beam of spot diameter 0.45 mm with Gaussian profile was focused on the surface with the help of 200 mm plano-convex lens equipped with laser head.Technical specifications of Yb-fibre laser are presented in table 2. The figure 2 illustrates the diagram of the experimental setup utilized in the present study.
For weld samples B1 to B5, process parameters for full penetration were obtained by performing sets of trial run experiments.Table 2 presents process parameters used to carry out LBW experiments of DSS 2205 sheets.As illustrated in table 2, for samples B1 to B5 laser output power is kept constant at 2150 W, and speed of welding varied from 3000 mm min −1 to 7000 mm min −1 and correspondingly heat input will change.Heat input used for different welding runs is calculated from equation (1).
The focused laser beam was kept at the surface of abutting edges of the specimen to be welded.The argon gas at the flow rate of about 10 l min −1 was purged on the sample to avoid oxidation of the weld samples.Thermal history of the molten weld-pool was observed and noted down using a single-spot monochromatic non-contact type Infrared pyrometer (IR) (Model: CTLM-2HCF3-C3H by Micro-Epsilon).It operates at a wavelength of 1.6 μm with sampling rate of 1 kHz and 0.7 mm vision zone.Though the IR pyrometer was designed to operate  at 1.6 μm wavelength, the reflected laser radiation was found to affect the recorded signal.In order to block the reflected laser radiation from entering into the pyrometer, a 1064 ± 25 nm notch filter with an optical density of 3 was mounted in front of the optical components.A proper calibration procedure was followed by melting a variety of metals, including Al, SS, and Cu, and comparing the temperature signal with the metals' known melting points.The pyrometer operated within a temperature range of 600 °C-2100 °C.The accuracy of the pyrometer as per the specification is ± (0.3% of reading + 2 °C) [16,17].A pair of diode laser was used to guide and focus the pyrometer at the intended zone.The substrate and pyrometer were kept stationary while the laser beam traversed linear motion while the thermal monitoring of weld pool was being carried out.
For metallography analysis, specimens obtained from laser welding were cut on a plane that was transverse to welding direction of the weld track.They were further mirror polished with sand papers of grit size P-600 to P-2000, followed by 1 μm and 1/2 μm diamond pastes.Kalling's No. 2 (5 g CuCl2, 100 ml HCl and 100 ml C2H5OH) etchant was used for etching of the polished weld samples to expose the microstructures and geometry of the FZ welds joints.The microstructures of the fusion zone were assessed using an optical microscope (Model-OLYMPUS BX51M).A micro hardness tester with a 200 g load cell and a 10 s dwell period was used to conduct the micro hardness test.Micro-hardness was performed along the cross-section at a depth of 0.5 mm, measuring from the top surface.Tensile tests were performed on a UTM machine (Make: Zwick Roell) at a ramp rate of 1 mm/min.Tensile specimens were fabricated as per ASTM E8/E8M standard as has been shown in figure 3. Fractographical analysis of tensile tested joints was performed using FE-SEM to determine the failure mode.

Thermal history molten weld pool
The temperature signal is an important tool for the process control because it effectively conveys the data regarding thermal history of the laser material interaction.The top surface temperature was measured using a single spot monochromatic non-contact type IR pyrometer.In order to obtain the precise measurements, the pyrometer was focused at the top surface along middle section of weld pool using a pair of guide diode lasers provided on it.Figure 4(a) represents thermo-cycle plot illustrating different stages in the molten weld pool thermo-cycle.As depicted in figure 4(a), a typical thermo-cycle of molten weld pool comprises the heating cycle (OA) and the cooling cycle (AE).The gradient of the curve is used to calculate the rate of cooling.During the heating cycle OA, surface of the fusion zone (FZ) is irradiated by the focused laser beam, which raises its temperature beyond the melting point of the base alloy.As laser beam proceeds, during cycle AD, the molten metal behind the laser beam begins to cool to room temperature.The section BC denoted cooling during nucleation termed as solidification cooling rate, whereas section DE characterizes solid cooling.Solidification begins at location B and ends at position C, which are the liquidus and solidus temperatures of the material under investigation, respectively.The slope of the curve BC, i.e., (ΔT/Δt) is used to determine the solidification cooling rate as shown in figure 4(a).Here, ΔT is the change in temperature and Δt is the corresponding cooling time interval determined between point B (liquidus temperature) and C (solidus temperature) [16][17][18].It significantly impacts the grain size and grain structure of the fusion zone.Similarly, the gradient of curve during DE was used to calculate the solid cooling rate.During period DE solid-state phase transformation or the transformation of ferrite to austenite and different phases takes place [12].
The variation of thermo profile obtained using an infrared pyrometer, is shown in figure 4(b), revealed that the highest temperature was over the melting point of DSS 2205 for weld samples B1 to B5. Results of measured solidification and solid cooling rates are presented in figures 4(c) & (d) for the weld samples B1 to B5 with change in heat input from 43 J mm −1 to 18.5 J mm −1 .In all melt-runs, fast solidification cooling rates in the range of 1.78 × 103 °C s −1 to 4.7 × 103 °C s −1 were measured, and it was observed that higher heat input causes cooling rate to slow down.When the heat input decreases, an increment in the rate of heat transfer between the molten weld pool to the surrounding base material results in increasing thermal gradient, and subsequently cooling rate of molten weld pool increases [19].Also, solid cooling rate increases from 0.86 × 103 °C s −1 to 1.85 × 103 °C s −1 with the decrease in heat input from weld sample B1 to B5. Due to fast solid cooling rate all weld samples returned to a temperature around 700 °C in less than 1 min, giving no chance for formation of brittle phases such as χ (Chi) or σ (Sigma), which are assumed to deteriorate the mechanical properties of weld joints.b-d) depicts the weld bead profiles taken along the cross-section of FZ at different heat input.The weld bead show full penetration for all the weld samples; however the variation in weld bead shapes is observed due to change in the heat input.The weld joints interfaces are free of pores and many such defects.The weld bead geometry for a maximum heat input (B1-43 J mm −1 ) has an interface shaped as a 'barrel' with an equal top (1.10 mm) and bottom (1.06 mm) width as is shown in figure 5(b).However, a further decrease in heat input (B3-25.8J/mm) leads to the formation of an hourglass-shaped weld bead (figure 5(c)).When the heat input is further reduced to minimum value (B5-18.5 J mm −1 ), weld profile divulges a concave shape (figure 5(d)), with wider top width (0.72 mm) and shorter root (0.31 mm).At lower heat input specimen interaction volume with laser beam reduces, and this results in the reduction of the weld bead size, which is comparable to the results obtained by Kumar et al and Kumar and Shahi [20,21].

Microstructure
Figure 6 demonstrates SEM micrograph of the weld joint of weld sample B2 (32.2 J mm −1 ).A sharp transition of crystal structure (size and shape) between the base material (DSS 2205) and the weld metal was perceived.Figures 7(a)-(d) depicts the optical micrographs of the welded weld samples at varying heat inputs.From the optical images, it is observed that columnar grains originated from the fusion zone walls and merged at the center.Since solidification front velocity is comparable on both sides, leads to a central edge.Similar microstructure for laser welding of steel is reported in the literature [22].It is evident from these figures 7(a-d) that microstructure of the fusion zone changes with the change in heat input.Presence of both ferrite and austenite phases are observed from the microstructures depicted in figures 7(a-d).The dark region in the microstructures indicates the ferrite phase and the brighter region represents the austenite phase [23][24][25].At a temperature just below the melting point, ferrite (F) is directly produced from the liquid as solidification begins.Ferrite (F) will change into austenite (A) phase as the temperature slowly drops below 900 °C; as a result, both ferrite and austenite phases are present [10].The transformation cycle in duplex stainless steel during the welding process can be described in summary as follows: L → L + F → F → F + A. With the decrease in heat input from weld sample B1 (43 J mm −1 ) to B5 (18.5 J mm −1 ), the dark region looks more apparent in comparison of brighter region in optical micrograph, which indicates high ferrite content in the fusion zone at lowest heat input.The reason attributed to be due to increase in solid cooling rate from 0.86 × 103 °C s −1 to 1.85 × 103 °C s −1 , corresponding to weld sample B1 to B5.At highest cooling rate, time available for the transformation of ferrite to austenite phase is less and due to this the resultant microstructure comprises of a large amount of ferrite [12,26].

Microhardness
The change in microhardness along the weld cross-section with varying heat input (43-18.5 J/mm) for weld samples B1to B5 is shown in figure 8.In this study, the micro-hardness of the fusion zone is higher than parent metal (285.6 ± 7.4 HV), irrespective of the change in heat input, due to higher fraction of ferrite grains in FZ of weld samples.The average hardness value of the fusion zone for the weld sample (B1-B5) increases from 314.5 ± 10.3 HV to 351.3 ± 9.7 HV as the heat input is reduced from 43 J mm −1 to 18.5 J/mm.The reason attributed due to higher cooling rate at lower heat input (18.5 J mm −1 ) results in more ferrite formation along with uniform and finer grain weld microstructure in comparison to higher heat input (43 J mm −1 ).The enhanced  ferrite formation and fine dendrites cause higher hardness which is in consonance with the studies published by Mourad et al and Saravanan et al [14,27].

Tensile test
The transverse tensile strength has been evaluated for all the weld joints with varying heat input.Figure 9 depicts the stress-strain curves for all the weld samples and parent metal (DSS 2205).It is observed from the stressstrain curve that weld joints are ductile in nature, having large zone of plastic deformation.The average value of tensile strength and their corresponding percentage elongations for different weld samples, thus obtained are shown in figures 10(a) and (b).The tensile strength and percentage elongation of the welded joints increases from weld sample B1 to B5.For weld sample B1, the average tensile strength has been calculated to be 805.4±11.2 MPa with elongation percentage of 28.7 ± 1.4%, and the failure occurred in the FZ.Similarly, for weld sample B5, it was reported that the average tensile strength is 872.5 ± 10.8 MPa with elongation percentage of 33.1 ± 1.65%, and failure occurred at parent metal.It is observed that at lowest heat input (B5-18.5 J mm −1 ), the average tensile and elongation percentage of the joint of the weld are highest due to smaller width of the weld, high ferrite content and finer grain formation [14].
On the contrary, condition of the obtained by supplying maximum heat input has a comparatively larger weld width, lower ferrite content, and coarser grain size.As we move from weld sample B1 to B5, smaller FZ size and finer homogenized grains facilitate more pinning and dislocation thereby providing higher strength to weld samples B4 and B5.

Conclusions
• Weld bonded joints with narrow fusion zone and no convexity or undercut at the surface was observed.
• The solid and solidification cooling rate of weld samples increases with the decreases in heat input from samples B1 to B5.
• Fusion zone microstructures in the DSS 2205 joints revealed the occurrence of both austenite and ferrite phases.Microstructures of weld sample B5 comprises of higher ferrite phase than B1.The transformation of ferrite to austenite phase depends on solid cooling rate and fraction of ferrite is more at higher cooling rate.At   high cooling rate less time is available for this transformation and consequently, the resultant microstructure contains large fraction of ferrite phase.
• The lowest hardness value (314.5 ± 10.3 HV) is higher than the hardness of parent metal DSS 2205 (285.6 ± 7.4 HV), and it increases with an increase in rate of cooling of weld metal from B1 to B5.The reason attributed that higher cooling result in more ferrite formation along with uniform and finer grain weld microstructure in comparison of samples solidified at lower cooling rate.
• Tensile strength of weld joints of DSS 2205 was found to have improved with increasing cooling rate.Higher cooling rate results in the formation of fine dendritic grains as well as higher ferrite content in the weld metal.
• The maximum failure strength was reported to be 872.5 ± 10.8 MPa with percentage elongation of 33.1 ± 1.65% for weld sample B5, and failure occurred in the parent metal.

Figure 1 .
Figure 1.SEM image of microstructure of base metal DSS 2205.

Figure 2 .
Figure 2. Experimental setups used for laser beam welding.

Figure 5 (
a) demonstrates the macro-morphology of the top surface of weld sample, and figures 5(

Figure 4 .
Figure 4. (a) Thermo-cycle plot depicting the various stages of molten pool thermal history (b) variation of thermo-cycle, (c) solidification cooling rate, and (d) solid cooling rate of weld samples at different heat input.

Figure 8 .
Figure 8. Variation of microhardness in the fusion zone of weld samples at different heat inputs.
) and (b) show the SEM image of fracture surfaces of the tensile tested weld samples B1 and B5.Dimples of different sizes and shapes were found in the fractured surfaces, indicating that the fracture occurred in ductile mode.From figure 11(a), it is witnessed that the fractured surface image of weld sample B1 (43 J mm −1 ) comprises coarse and elongated dimples along with coarse voids.The fractured surface image of weld sample B5 (18.5 J mm −1 ), as shown in figure 11(b), contains a large population of smaller dimples within the larger dimples along with microvoids coalescence, which signifies relatively higher tensile strength as well as ductility.

Figure 9 .
Figure 9. Stress-strain curves for weld samples at different heat inputs and the base metal DSS 2205.

Figure 10 .
Figure 10.Variation of (a) tensile strength, (b) % elongation of weld samples at different heat inputs.

Table 2 .
Process parameters used for laser beam welding of DSS 2205 sheets.