Autogenous Fiber Laser Welding of 316L Austenitic and 2304 Lean Duplex Stainless Steels

This study presents results of experimental tests on quality of dissimilar welded joints between 316L austenitic and 2304 lean duplex stainless steels, welded without ceramic backing. Fiber laser welded butt joints at a thickness of 8 mm were subjected to non-destructive testing (visual and penetrant), destructive testing (static tensile test, bending test, and microhardness measurements) and structure observations (macro- and microscopic examinations, SEM, element distribution characteristics, and ferrite content measurements). Non-destructive tests and metallographic examinations showed that the welded joints meet the acceptance criteria for B level in accordance with EN ISO 13919–1 standard. Also the results of the destructive tests confirmed the high quality of the joints: specimens were fractured in base material with lower strength—316L austenitic stainless steel and a 180° bending angle was obtained confirming the high plasticity of the joints. Microscopic examination, SEM and EDS analysis showed the distribution of alloying elements in joints. The microhardness of the autogenous weld metal was higher by about 20 HV0.2 than that of the lean duplex steel. Ferrite content in the root was about 37% higher than in the face of the weld. The Schaeffler phase diagram was used to predict the phase composition of the welded joints and sufficient compliance with the magnetic method was found. The presented procedure can be used for welding of 316L–2304 stainless steels dissimilar welded joints of 8 mm thickness without ceramic backing.


Introduction
Lasers as a heat source have been widely applied in industrial processes [1,2]. Currently, they are used for welding, cutting, remelting, and various processes with surface modification of materials [3][4][5][6]. Laser sources are characterized by: high energy density, high efficiency, large temperature gradients, high repeatability of the process, and formation of a narrow heat affected zone (HAZ). Many authors report beneficial changes in the properties of various materials subjected to laser treatment in such sectors as: medical, energy, ocean engineering, etc. [7][8][9]. The area of application of various technological laser solutions includes processing of many types of metals and their alloys, ceramics, polymers, and composites [1,5,10,11], including underwater local dry welding and wet welding [12,13]. Limiting the disadvantageous effects of technological use of lasers in some applications, e.g., the small width of weld, can be obtained by combining it with other heat sources to create hybrid processes. The most popular solutions include laser-metal active gas arc welding (MAG) and laser-tungsten inert gas arc welding (TIG) processes [14][15][16].
Welding of dissimilar stainless steel joints can be done between two main types of filler metals: austenitic (e.g., 309L) or duplex (e.g., 2209). Vincente et al. stated that the higher corrosion resistance of joints made by the MAG process is favored by the use of duplex steel consumable [45]. Similarly, Rahmani et al. recommend using duplex instead of austenitic consumable when using the TIG process [46]. However, among the various austenitic consumables, the most beneficial, both in terms of strength properties, as well as morphology and corrosion resistance, is the use of 309L [47,48] or super austenitic 904L filler metal [49]. The literature also describes attempts to weld such a material combination without filler material: using different variants of the TIG process [50,51] or concentrated welding sources [33,52,53]. Ridha Mohammed et al. stated that the mechanical properties of the duplex-austenitic steel fiber laser joints were better compared with the base materials (BM) because of the small HAZ [52]. In the same work, the authors found the presence in welded joints the regions with a completely austenitic solidification mode which were susceptible to solidification cracking. An important observation is that when laser welding is used for super duplex and austenitic steels, the ferrite/austenite phase balance is not significantly changed by different heat input values [39]. Existence of an unmixed zone that originated from each base material was stated by Chun et al. [53]. The approximately 50:50 ferrite/austenite microstructural balance can be reached with solution annealing in the range of 1050-1100 • C, and it is especially important in autogenous welds [33]. Saravanan et al. stated that the higher microhardness of the weld zone is attributed by the formation of finer and uniform grains following high cooling rate [54].
Although researches on laser welding of dissimilar stainless steels have been reported, the study on structure and properties of laser beam welded austenitic/lean duplex steel joints is still relatively rarely mentioned. Therefore, the present work aims to show the ability of making 316L austenitic stainless steel-2304 lean duplex stainless steel dissimilar joints using the autogenous fiber laser welding process.

Materials and Methods
The test pieces were a flat plates with thickness of 8 mm made of 316L austenitic stainless steel (1.4404, UNS S31603) and 2304 lean duplex stainless steel (1.4362, UNS S32304) in delivery condition. Both materials were after heat treatment: 316L steel after solution annealing from 1050 • C and 2304 steel was hot rolled and solution annealed. The chemical composition of the used materials in accordance with the inspection certificate (spectral analysis) and the requirements of the EN 10088-2:2014 standard (minimum and maximum wt. %) are given in Table 1. The mechanical properties of the materials in accordance with the requirements of the EN 10088-2-2014 standard are presented in Table 2. Chromium and nickel equivalents (Creq and Nieq) were calculated according to [55].  A high power continuous wave ytterbium fiber laser IPG Photonics YLS-6000 (IPG Photonic, Oxford, MS, USA) with a maximum power of 6 kW and a wavelength of 1070 nm was used ( Figure 1). The joints were made at 6 kW laser output power. The laser beam was delivered through the optical feeding fiber (transmitting fiber) of 300 µm core diameter. A welding head having a focal length of 250 mm and a focus collimator lens 150 mm was used. Focusing position was set on the surface of the plate. Welding speed was set to 25 mm/s (1.5 m/min). Welding process was autogenous-without filler metal. Shielding gas was supplied by a gas nozzle mounted at laser head to avoid the oxidation of weld beads. The flow rate of shielding gas-argon 5.0 (I1 in accordance with ISO 14175)-was set at 16 L/min. Test pieces were welded without using a ceramic backing or forming gas.   Figure  1). The joints were made at 6 kW laser output power. The laser beam was delivered through the optical feeding fiber (transmitting fiber) of 300 µm core diameter. A welding head having a focal length of 250 mm and a focus collimator lens 150 mm was used. Focusing position was set on the surface of the plate. Welding speed was set to 25 mm/s (1.5 m/min). Welding process was autogenous-without filler metal. Shielding gas was supplied by a gas nozzle mounted at laser head to avoid the oxidation of weld beads. The flow rate of shielding gas-argon 5.0 (I1 in accordance with ISO 14175)-was set at 16 L/min. Test pieces were welded without using a ceramic backing or forming gas. In order to identify and then eliminate any troubles with welding process the first stage of research consisted of making preliminary test joints. Initial joints were made to determine the values of significant variables, such as focal length, laser power and welding speed. The parameters of laser welding for austenitic stainless steel and lean duplex stainless steel differ, therefore, after initial experiments with similar materials, parameters matching both grades were selected. Before welding, the elements were cleaned with abrasive paper and degreased with acetone. Preliminary tests on stainless steels showed the presence of spatter if the surface was not cleaned properly just before welding.
To assess the quality of the welded joints, non-destructive tests (NDT) were performed: visual testing (VT)-according to the EN ISO 17637 standard and penetrant testing (PT)-according to the EN ISO 571-1 standard. Tensile and bending tests were conducted at ambient temperature of 20 °C using Instron 1195 (INSTRON, Norwood, MA, USA) universal testing machine in accordance with EN ISO 6892-1 and EN ISO 4136 standards. For bending test according to EN ISO 15614-11 bend former with a diameter of ɸ = 27 mm was used (the bend former diameter for materials with elongation above 25% is four times the specimen thickness). The acceptance criterion was defined as a bend angle α = 180°. Specimens in the face and root areas were slightly grinded to remove notches before performing tensile and three-point bending tests.
In order to prevent changes in the material structure under the influence of temperature during cutting, the specimens for metallographic examinations (according to EN ISO 17639) were cut mechanically in a direction transverse to the welding axis at the cutting machine with intensive In order to identify and then eliminate any troubles with welding process the first stage of research consisted of making preliminary test joints. Initial joints were made to determine the values of significant variables, such as focal length, laser power and welding speed. The parameters of laser welding for austenitic stainless steel and lean duplex stainless steel differ, therefore, after initial experiments with similar materials, parameters matching both grades were selected. Before welding, the elements were cleaned with abrasive paper and degreased with acetone. Preliminary tests on stainless steels showed the presence of spatter if the surface was not cleaned properly just before welding.
To assess the quality of the welded joints, non-destructive tests (NDT) were performed: visual testing (VT)-according to the EN ISO 17637 standard and penetrant testing (PT)-according to the EN ISO 571-1 standard. Tensile and bending tests were conducted at ambient temperature of 20 • C using Instron 1195 (INSTRON, Norwood, MA, USA) universal testing machine in accordance with EN ISO 6892-1 and EN ISO 4136 standards. For bending test according to EN ISO 15614-11 bend former with a diameter of φ = 27 mm was used (the bend former diameter for materials with elongation above 25% is four times the specimen thickness). The acceptance criterion was defined as a bend angle α = 180 • . Specimens in the face and root areas were slightly grinded to remove notches before performing tensile and three-point bending tests.
In order to prevent changes in the material structure under the influence of temperature during cutting, the specimens for metallographic examinations (according to EN ISO 17639) were cut mechanically in a direction transverse to the welding axis at the cutting machine with intensive cooling.
Then it was grinded on abrasive papers with gradation 600-2400 and polishing on a polishing cloth using an aqueous suspension of 3 µm diamonds. This preparation of the specimen was followed by two-stage etching. The first stage was etching to reveal the austenite grain boundaries for which a mixture of 50 mL of boiling water, 3 mL of HNO 3 , and 1 mL of HF was used. The etching was performed by a two-minute immersion of specimens in an 80 • C temperature solution. Then the specimen was thoroughly rinsed and cooled under water. The next stage was etching in Beraha-type reagent (85 mL of water, 15 mL of HCl, and 1 g K 2 S 2 O 5 ). Etching was carried out by about 1-3 min immersion in 20 • C temperature solution (until the corresponding colored structure was obtained on the specimen). At the end the specimen was thoroughly rinsed and dried with a stream of compressed air. Macroscopic metallographic observations were performed using DSLR Nikon d7000 with Tamron 90 mm f/2.8 macro lens (Nikon Corporation, Tokyo, Japan), while metallographic microscopy tests were done on a light microscope (LM) Olympus BX51 (Olympus, Tokyo, Japan). It offers imaging in a bright field, dark field and polarized light. The tests were also carried out using the scanning electron microscope JOEL JSM-7800F (SEM) with the EDAX adapter (Japan Electronics Corporation, Tokyo, Japan) enabling EDS analysis.
The ferrite number was determined using a Fischer Feritscope FMP30 for both base materials and weld (Helmut Fischer GmbH Institut für Elektronik und Messtechnik 71069 Sindelfingen, Germany). The Feritscope was calibrated on calibration standards prior to measurements. The ferrite content measurements were carried out in accordance with ISO 17655, at 6 measuring points for each place of measure.

Non-Destructive Tests
To detect surface imperfections VT and PT were conducted. Positive results of those tests were found for all specimens. Figure 2 shows face and root side of one of the welded joints. cooling. Then it was grinded on abrasive papers with gradation 600-2400 and polishing on a polishing cloth using an aqueous suspension of 3 µm diamonds. This preparation of the specimen was followed by two-stage etching. The first stage was etching to reveal the austenite grain boundaries for which a mixture of 50 mL of boiling water, 3 mL of HNO3, and 1 mL of HF was used. The etching was performed by a two-minute immersion of specimens in an 80 °C temperature solution. Then the specimen was thoroughly rinsed and cooled under water. The next stage was etching in Beraha-type reagent (85 mL of water, 15 mL of HCl, and 1 g K2S2O5). Etching was carried out by about 1-3 min immersion in 20 °C temperature solution (until the corresponding colored structure was obtained on the specimen). At the end the specimen was thoroughly rinsed and dried with a stream of compressed air. Macroscopic metallographic observations were performed using DSLR Nikon d7000 with Tamron 90 mm f/2.8 macro lens (Nikon Corporation, Tokyo, Japan), while metallographic microscopy tests were done on a light microscope (LM) Olympus BX51 (Olympus, Tokyo, Japan). It offers imaging in a bright field, dark field and polarized light. The tests were also carried out using the scanning electron microscope JOEL JSM-7800F (SEM) with the EDAX adapter (Japan Electronics Corporation, Tokyo, Japan) enabling EDS analysis. Microhardness measurements-HV0.2 were carried out using FM-800 tester with a load of F = 1.9614 N (Future-Tech, Tokyo, Japan).
The ferrite number was determined using a Fischer Feritscope FMP30 for both base materials and weld (Helmut Fischer GmbH Institut für Elektronik und Messtechnik 71069 Sindelfingen, Germany). The Feritscope was calibrated on calibration standards prior to measurements. The ferrite content measurements were carried out in accordance with ISO 17655, at 6 measuring points for each place of measure.

Non-Destructive Tests
To detect surface imperfections VT and PT were conducted. Positive results of those tests were found for all specimens. Figure 2 shows face and root side of one of the welded joints.    Figure 2a shows the view of the face of welded joint. Underfilling of the weld face can be seen, which in the case of laser welding of large thicknesses without consumable is expected. The welded joints met the assumed acceptance criterion of quality level B in accordance with the EN ISO 13919-1 standard. Therefore, on a basis of NDT results it was determined that the next part of the research, consisting of destructive tests, can be carried out. Figure 3 shows the view of two specimens (signed A and B) which were subjected to transverse tensile tests. The results of the test are presented in Table 3. Minimum tensile strength required for the 316L austenitic stainless steel should be 530 MPa, and minimum tensile strength required for the 2304 duplex stainless steel should be 630 MPa ( Table 2). For this test, the acceptance criterion was to exceed the minimum tensile strength of 316L austenitic stainless steel.

Static Tensile Test
Materials 2020, 13, x FOR PEER REVIEW 6 of 16 1 standard. Therefore, on a basis of NDT results it was determined that the next part of the research, consisting of destructive tests, can be carried out. Figure 3 shows the view of two specimens (signed A and B) which were subjected to transverse tensile tests. The results of the test are presented in Table 3. Minimum tensile strength required for the 316L austenitic stainless steel should be 530 MPa, and minimum tensile strength required for the 2304 duplex stainless steel should be 630 MPa ( Table 2). For this test, the acceptance criterion was to exceed the minimum tensile strength of 316L austenitic stainless steel.  Base material 316L The obtained results meet the acceptance criterion and the obtained values are 10-15% higher compared to the requirements. The fracture was ductile and occurred in the 316L base material. It proves the correctness of welding parameters in terms of laser beam power, focus position and welding speed. The identified welding imperfections in the VT and PT tests did not affect the tensile strength. Tests showed that no structural changes reducing mechanical properties occurred under the influence of the laser welding thermal cycle. The absence of structural changes will be verified by metallographic examinations (Section 3.2.3).

Bending Test
The bending tests were carried out on four specimens-in two the tensiled side was the face of the weld and in two the tensiled side was the root of the weld. The aim of this test was to investigate the plastic properties of the welded joint. The bending tests are also carried out to reveal welding imperfections, e.g., incomplete fusion, lack of penetration, pores, and others. Specimens after the bending test are presented in Figure 4.  The obtained results meet the acceptance criterion and the obtained values are 10-15% higher compared to the requirements. The fracture was ductile and occurred in the 316L base material. It proves the correctness of welding parameters in terms of laser beam power, focus position and welding speed. The identified welding imperfections in the VT and PT tests did not affect the tensile strength. Tests showed that no structural changes reducing mechanical properties occurred under the influence of the laser welding thermal cycle. The absence of structural changes will be verified by metallographic examinations (Section 3.2.3).

Bending Test
The bending tests were carried out on four specimens-in two the tensiled side was the face of the weld and in two the tensiled side was the root of the weld. The aim of this test was to investigate the plastic properties of the welded joint. The bending tests are also carried out to reveal welding imperfections, e.g., incomplete fusion, lack of penetration, pores, and others. Specimens after the bending test are presented in Figure 4. The bending test was executed until an angle of 180° was reached. The welds were subjected to significant plastic deformation and no surface cracks were visible on the root or face side, which indicates that the laser welded dissimilar joints exhibited good ductility and adequate bending strength. Such a test result proves very good plastic properties, but also a lack of welding imperfections in the tested specimens. As in the tensile test, underfill did not lead to cracks on the tensiled surfaces during the bending test. Figure 5 presents the macrostructure of the welded joint observed on the cross-section of the weld axis. The joint has a regular symmetrical shape with visible underfilling of the weld face, without visible pores or excess penetration from the root side. Typical geometry of laser welded joint was observed. Keyhole laser welding forms a 'chalice' shaped weld bead profile. Examined welded joints were made without welding imperfections such as porosity and humping beads. Full penetration was achieved for investigated welded joints without changing beam focusing position.  The bending test was executed until an angle of 180 • was reached. The welds were subjected to significant plastic deformation and no surface cracks were visible on the root or face side, which indicates that the laser welded dissimilar joints exhibited good ductility and adequate bending strength. Such a test result proves very good plastic properties, but also a lack of welding imperfections in the tested specimens. As in the tensile test, underfill did not lead to cracks on the tensiled surfaces during the bending test. Figure 5 presents the macrostructure of the welded joint observed on the cross-section of the weld axis. The joint has a regular symmetrical shape with visible underfilling of the weld face, without visible pores or excess penetration from the root side. Typical geometry of laser welded joint was observed. Keyhole laser welding forms a 'chalice' shaped weld bead profile. Examined welded joints were made without welding imperfections such as porosity and humping beads. Full penetration was achieved for investigated welded joints without changing beam focusing position. The bending test was executed until an angle of 180° was reached. The welds were subjected to significant plastic deformation and no surface cracks were visible on the root or face side, which indicates that the laser welded dissimilar joints exhibited good ductility and adequate bending strength. Such a test result proves very good plastic properties, but also a lack of welding imperfections in the tested specimens. As in the tensile test, underfill did not lead to cracks on the tensiled surfaces during the bending test. Figure 5 presents the macrostructure of the welded joint observed on the cross-section of the weld axis. The joint has a regular symmetrical shape with visible underfilling of the weld face, without visible pores or excess penetration from the root side. Typical geometry of laser welded joint was observed. Keyhole laser welding forms a 'chalice' shaped weld bead profile. Examined welded joints were made without welding imperfections such as porosity and humping beads. Full penetration was achieved for investigated welded joints without changing beam focusing position.   Figure 6 shows the structure of base materials: 316L austenitic stainless steel and 2304 lean duplex stainless steel. Figure 6a presents a microstructure of 316L including twins, slip bands, and δ ferrite (darker due to Beraha reagent etching). Austenite grains are quite fine, elongated in the direction of forming and between them big amount of darker δ ferrite precipitates. As demonstrated in Figure 6b base material of 2304 lean duplex steel is characterized by a dark continuous matrix of the ferrite phase (δ) and white island of austenitic (γ) phase with characteristic twins made visible through two-stage etching. Visible directionality of the structure is related to the rolling process.

Macro-and Microscopic Examinations
Materials 2020, 13, x FOR PEER REVIEW 8 of 16 Figure 6 shows the structure of base materials: 316L austenitic stainless steel and 2304 lean duplex stainless steel. Figure 6a presents a microstructure of 316L including twins, slip bands, and δ ferrite (darker due to Beraha reagent etching). Austenite grains are quite fine, elongated in the direction of forming and between them big amount of darker δ ferrite precipitates. As demonstrated in Figure 6b base material of 2304 lean duplex steel is characterized by a dark continuous matrix of the ferrite phase (δ) and white island of austenitic (γ) phase with characteristic twins made visible through two-stage etching. Visible directionality of the structure is related to the rolling process.  Figures 7 and 8 show the transition areas from base material through the HAZ to the weld metal (WM). The austenitic structure of the weld consists of equiaxed coarse and fine dendrites. In addition, no eutectics or microcracks were found in the weld structure. During cooling the liquid metal in welding pool firstly solidify as ferrite. Further cooling cause partial transformation of ferrite to austenite. Austenite initially form as grain boundary allotriomorphic austenite (GBA), then as Widmanstätten side plates of austenite (WA) and finally as intragranular precipitates of austenite (IGA). In general, the first two types of austenite (GBA and WA) require less driving force than intragranular needle austenite, which means that the grain boundary and Widmanstätten austenite formed earlier at higher temperatures, whereas intragranular austenite particles precipitated on further cooling at a lower temperature [55,56]. GBA morphology resembles a coherent island arranged along the border of ferrite. WA morphology can be described as small needles in a ferrite matrix. While the IGA morphology is like square islands in a ferrite matrix.  Figures 7 and 8 show the transition areas from base material through the HAZ to the weld metal (WM). The austenitic structure of the weld consists of equiaxed coarse and fine dendrites. In addition, no eutectics or microcracks were found in the weld structure. During cooling the liquid metal in welding pool firstly solidify as ferrite. Further cooling cause partial transformation of ferrite to austenite. Austenite initially form as grain boundary allotriomorphic austenite (GBA), then as Widmanstätten side plates of austenite (WA) and finally as intragranular precipitates of austenite (IGA). In general, the first two types of austenite (GBA and WA) require less driving force than intragranular needle austenite, which means that the grain boundary and Widmanstätten austenite formed earlier at higher temperatures, whereas intragranular austenite particles precipitated on further cooling at a lower temperature [55,56]. GBA morphology resembles a coherent island arranged along the border of ferrite. WA morphology can be described as small needles in a ferrite matrix. While the IGA morphology is like square islands in a ferrite matrix.
Materials 2020, 13, x FOR PEER REVIEW 8 of 16 Figure 6 shows the structure of base materials: 316L austenitic stainless steel and 2304 lean duplex stainless steel. Figure 6a presents a microstructure of 316L including twins, slip bands, and δ ferrite (darker due to Beraha reagent etching). Austenite grains are quite fine, elongated in the direction of forming and between them big amount of darker δ ferrite precipitates. As demonstrated in Figure 6b base material of 2304 lean duplex steel is characterized by a dark continuous matrix of the ferrite phase (δ) and white island of austenitic (γ) phase with characteristic twins made visible through two-stage etching. Visible directionality of the structure is related to the rolling process.  Figures 7 and 8 show the transition areas from base material through the HAZ to the weld metal (WM). The austenitic structure of the weld consists of equiaxed coarse and fine dendrites. In addition, no eutectics or microcracks were found in the weld structure. During cooling the liquid metal in welding pool firstly solidify as ferrite. Further cooling cause partial transformation of ferrite to austenite. Austenite initially form as grain boundary allotriomorphic austenite (GBA), then as Widmanstätten side plates of austenite (WA) and finally as intragranular precipitates of austenite (IGA). In general, the first two types of austenite (GBA and WA) require less driving force than intragranular needle austenite, which means that the grain boundary and Widmanstätten austenite formed earlier at higher temperatures, whereas intragranular austenite particles precipitated on further cooling at a lower temperature [55,56]. GBA morphology resembles a coherent island arranged along the border of ferrite. WA morphology can be described as small needles in a ferrite matrix. While the IGA morphology is like square islands in a ferrite matrix. As presented in Figure 7 heat affected zones were barely visible. HAZ of laser welded stainless steel joints is very narrow, in this case it was about 20 µm for 316L steel and 50 µm for 2304 steel as can be seen on Figure 8a and b, respectively. Both for lower (LM- Figure 7) and higher magnifications (SEM- Figure 8), a difference in the austenite-ferrite ratio in the weld close to HAZ is noticeable. As expected, from the 316L steel side there is a larger amount of austenite in the weld metal than from the 2304 steel side. Linear EDS analysis was carried out through all welded joint: from 316L steel, first HAZ, weld metal, second HAZ ending at 2304 steel ( Figure 9). The distribution of the analyzed elements along the measuring line is visible on individual curves. Level of the line does not show the content of a given element in the analyzed alloy, but only show the variability of its content along the measuring line. As could be predicted, the weld metal has a composition resulting from mixing of both materials-due to autogenous laser welding. The content of nickel which is stabilizing austenite element increases towards 316L steel. The heterogeneous composition in the weld metal can be the reason for the different austenite-ferrite ratio on both sides of the weld axis. Due to diffusion, differences in elements content occur in HAZ. The nickel content in HAZ next to 2304 steel decreases compared to the weld. The chromium content increases in the HAZ on the 2304 steel side. This increase may be due to a visible larger ferrite grains in this area and chromium is ferrite stabilizer. This can be seen in the photo below the graph (Figure 9), and is the effect of heat affecting the structure. As presented in Figure 7 heat affected zones were barely visible. HAZ of laser welded stainless steel joints is very narrow, in this case it was about 20 µm for 316L steel and 50 µm for 2304 steel as can be seen on Figure 8a and b, respectively. Both for lower (LM- Figure 7) and higher magnifications (SEM- Figure 8), a difference in the austenite-ferrite ratio in the weld close to HAZ is noticeable. As expected, from the 316L steel side there is a larger amount of austenite in the weld metal than from the 2304 steel side.
Linear EDS analysis was carried out through all welded joint: from 316L steel, first HAZ, weld metal, second HAZ ending at 2304 steel ( Figure 9). The distribution of the analyzed elements along the measuring line is visible on individual curves. Level of the line does not show the content of a given element in the analyzed alloy, but only show the variability of its content along the measuring line. As could be predicted, the weld metal has a composition resulting from mixing of both materials-due to autogenous laser welding. The content of nickel which is stabilizing austenite element increases towards 316L steel. The heterogeneous composition in the weld metal can be the reason for the different austenite-ferrite ratio on both sides of the weld axis. Due to diffusion, differences in elements content occur in HAZ. The nickel content in HAZ next to 2304 steel decreases compared to the weld. The chromium content increases in the HAZ on the 2304 steel side. This increase may be due to a visible larger ferrite grains in this area and chromium is ferrite stabilizer. This can be seen in the photo below the graph (Figure 9), and is the effect of heat affecting the structure. As presented in Figure 7 heat affected zones were barely visible. HAZ of laser welded stainless steel joints is very narrow, in this case it was about 20 µm for 316L steel and 50 µm for 2304 steel as can be seen on Figure 8a and b, respectively. Both for lower (LM- Figure 7) and higher magnifications (SEM- Figure 8), a difference in the austenite-ferrite ratio in the weld close to HAZ is noticeable. As expected, from the 316L steel side there is a larger amount of austenite in the weld metal than from the 2304 steel side. Linear EDS analysis was carried out through all welded joint: from 316L steel, first HAZ, weld metal, second HAZ ending at 2304 steel ( Figure 9). The distribution of the analyzed elements along the measuring line is visible on individual curves. Level of the line does not show the content of a given element in the analyzed alloy, but only show the variability of its content along the measuring line. As could be predicted, the weld metal has a composition resulting from mixing of both materials-due to autogenous laser welding. The content of nickel which is stabilizing austenite element increases towards 316L steel. The heterogeneous composition in the weld metal can be the reason for the different austenite-ferrite ratio on both sides of the weld axis. Due to diffusion, differences in elements content occur in HAZ. The nickel content in HAZ next to 2304 steel decreases compared to the weld. The chromium content increases in the HAZ on the 2304 steel side. This increase may be due to a visible larger ferrite grains in this area and chromium is ferrite stabilizer. This can be seen in the photo below the graph (Figure 9), and is the effect of heat affecting the structure. The low chromium content in the welding pool caused the formation of a structure with a higher content of austenite than is observed in duplex laser-welded joints with the same parameters and geometry [56]. As a result, the austenite to ferrite ratio is closer to 50:50. The obtained volume fraction of austenite may also be induced by the difference in the thermal conductivity coefficient of the austenitic steel in relation to the duplex steel, which resulted in an increase in the cooling time and allowed the transformation of ferrite into austenite [31].
Metallographic microscopic examination-SEM and EDS analysis-did not show the segregation of alloying elements between the ferritic and austenitic phases. This indicates that the areas depleted in Cr and Ni, which can cause degradation phenomena, e.g., corrosion, were not formed.

Microhardness Measurements
Microhardness measurements were made in base materials, weld metal and HAZs in two lines as can be seen in Figure 5. Figure 10 shows the results of the measurements. The low chromium content in the welding pool caused the formation of a structure with a higher content of austenite than is observed in duplex laser-welded joints with the same parameters and geometry [56]. As a result, the austenite to ferrite ratio is closer to 50:50. The obtained volume fraction of austenite may also be induced by the difference in the thermal conductivity coefficient of the austenitic steel in relation to the duplex steel, which resulted in an increase in the cooling time and allowed the transformation of ferrite into austenite [31].
Metallographic microscopic examination-SEM and EDS analysis-did not show the segregation of alloying elements between the ferritic and austenitic phases. This indicates that the areas depleted in Cr and Ni, which can cause degradation phenomena, e.g., corrosion, were not formed.

Microhardness Measurements
Microhardness measurements were made in base materials, weld metal and HAZs in two lines as can be seen in Figure 5. Figure 10 shows the results of the measurements. For 316L austenitic stainless steel microhardness value was in a range of 186-209 HV0.2, which corresponds well to the literature [38,50]. The microhardness fluctuations in the 316L steel are due to the presence of a large number of delta ferrite precipitates in the austenitic structure. In the HAZ (on 316L steel side) microhardness values are lower (average 196 HV0.2), which confirms that both heat input and cooling rate were appropriate for this process. The average HV0.2 value for weld metal is 254 HV0.2, with higher values on weld face than in the weld root. Differences in hardness values between the line passing through the face of the weld in relation to the line close to the root can be explained by differences in the amount of heat accumulated in each area. The average value of hardness for 2304 steel was 233 HV0.2-typical for this grade. However, its decrease in the HAZ on 2304 steel side (average value 197 HV0.2) was observed, which is a result of the structural changes shown in the microscopic studies. Due to the different morphology and austenite arrangement between weld metal and lean duplex, the microhardness of the autogenous weld metal was higher (about 20 HV0.2) than microhardness of the lean duplex steel [33]. The results are distinctive for dissimilar austenitic-duplex stainless steel welded joints-which was also demonstrated by other authors [33,52,54]. For 316L austenitic stainless steel microhardness value was in a range of 186-209 HV0.2, which corresponds well to the literature [38,50]. The microhardness fluctuations in the 316L steel are due to the presence of a large number of delta ferrite precipitates in the austenitic structure. In the HAZ (on 316L steel side) microhardness values are lower (average 196 HV0.2), which confirms that both heat input and cooling rate were appropriate for this process. The average HV0.2 value for weld metal is 254 HV0.2, with higher values on weld face than in the weld root. Differences in hardness values between the line passing through the face of the weld in relation to the line close to the root can be explained by differences in the amount of heat accumulated in each area. The average value of hardness for 2304 steel was 233 HV0.2-typical for this grade. However, its decrease in the HAZ on 2304 steel side (average value 197 HV0.2) was observed, which is a result of the structural changes shown in the microscopic studies. Due to the different morphology and austenite arrangement between weld metal and lean duplex, the microhardness of the autogenous weld metal was higher (about 20 HV0.2) than microhardness of the lean duplex steel [33]. The results are distinctive for dissimilar austenitic-duplex stainless steel welded joints-which was also demonstrated by other authors [33,52,54].

Ferrite Content Measurements and Calculations
Points for ferrite measurement were chosen within the areas of 316L austenitic stainless steel and 2304 lean duplex stainless steel on the welded joint surface. Similarly, points were selected along the weld from the face and root of the weld. Table 4 shows the results of ferrite content measurements. The results of the delta ferrite measurements showed its correct, expected content in base materials. However, differences in the ferrite content between the face and the root of the weld were observed. Higher (by about 37%) ferrite content in the weld root is a consequence of different heat distribution between the border surfaces during and after the welding process. This distribution is also associated with various values of the thermal conductivity coefficient of base materials and the intensity of the laser heat source. Laser welding is characterized by high power density, which means that the energy distribution is constant over the entire depth (that's why a cylindrical model of a heat source is usually used for numerical simulations).
Geometry of the weld is not constant throughout the thickness of the specimen, but it changes significantly in the area of the weld face ( Figure 5). Forming a 'chalice' shaped weld bead is characteristic for the keyhole laser welding of thick materials. The change in the fusion line shape also indicates different heat distribution conditions in the welding area. This causes a change in the cooling rate of the joint on the upper and lower boundary surface (face/root). This phenomenon and different thermal properties can be explained by the increase in ferrite content in the weld root, as this leads to an increase in the cooling rate of the joint.
The results of ferrite content measurements were compared with the results obtained from the Schaeffler diagram-a graphic method for assessing the weldability of high alloy steels and dissimilar joints. Schaeffler diagram is a good preliminary prediction method for weldability of steels subjected to fusion welding [54]. It was assumed, according to the welding conditions presented in Chapter 2, that a dilution rate was 50%, so the values of Creq and Nieq for the weld metal are 21.9% and 8.9% respectively, and this corresponds to about 30% of the amount of ferrite ( Figure 11). Comparative analysis of the results of measurements by the magnetic method and the Schaeffler diagram confirm that Creq and Nieq in this form does not fully describe the ferrite forming tendency during laser welding. Obtained results of prediction of ferrite content were also confirmed using WRC-92 diagram.
Solidification mode of stainless steels can be predicted by the Fe-Ni-Cr pseudo-binary phase diagram shown in Figure 12 based on Creq/Nieq ratio [50]. As a result of welding of 316L steel without filler metal (Creq/Nieq = 1.71) the weld solidification mode is ferrite-austenite (FA). According to the Figure 12, Creq/Nieq ratio of 2304 steel equal to 3.80 during rapid cooling is enough to cause δ-ferrite solidification and then ferrite and austenite transformation (ferrite mode-F). Chemical composition of the weld metal, which is an alloy of 50% dilution rate, described by the value of Creq/Nieq = 2.46 also causes solidification in the F mode. The results of these analyzes are generally consistent with the results of metallographic examinations and ferrite content measurements. Materials 2020, 13, x FOR PEER REVIEW 12 of 16 Figure 11. Schaeffler constitutional diagram.
Solidification mode of stainless steels can be predicted by the Fe-Ni-Cr pseudo-binary phase diagram shown in Figure 12 based on Creq/Nieq ratio [50]. As a result of welding of 316L steel without filler metal (Creq/Nieq = 1.71) the weld solidification mode is ferrite-austenite (FA). According to the Figure 12, Creq/Nieq ratio of 2304 steel equal to 3.80 during rapid cooling is enough to cause δ-ferrite solidification and then ferrite and austenite transformation (ferrite mode-F). Chemical composition of the weld metal, which is an alloy of 50% dilution rate, described by the value of Creq/Nieq = 2.46 also causes solidification in the F mode. The results of these analyzes are generally consistent with the results of metallographic examinations and ferrite content measurements.

Conclusions
Tests regarding dissimilar laser welding of butt joints made of 316L austenitic stainless steel and 2304 lean duplex stainless steel showed that the use of IPG YLS-6000 fiber laser with a maximum power of 6 kW allowed to obtain sound butt joints meeting the requirements of quality level B in accordance with EN ISO 13919-1 standard. During welding, the only imperfection detected by VT and PT was the underfilling of the face, while microscopic examinations did not show the presence of any other welding imperfections. Due to the very low (as for laser welding) welding speed, it was possible to achieve full penetration of 8 mm thick plates with one-sided course of the laser beam.
Obtained joints have good strength properties (higher than the minimum values required by the standard), which were confirmed in the static tensile tests (average Rm = 600 MPa). All specimens

Conclusions
Tests regarding dissimilar laser welding of butt joints made of 316L austenitic stainless steel and 2304 lean duplex stainless steel showed that the use of IPG YLS-6000 fiber laser with a maximum power of 6 kW allowed to obtain sound butt joints meeting the requirements of quality level B in accordance with EN ISO 13919-1 standard. During welding, the only imperfection detected by VT and PT was the underfilling of the face, while microscopic examinations did not show the presence of any other welding imperfections. Due to the very low (as for laser welding) welding speed, it was possible to achieve full penetration of 8 mm thick plates with one-sided course of the laser beam.
Obtained joints have good strength properties (higher than the minimum values required by the standard), which were confirmed in the static tensile tests (average R m = 600 MPa). All specimens fractured in 316L austenitic stainless steel BM. The bending tests showed that plastic properties of base materials were not deteriorated and confirmed the absence of welding imperfections.
The microstructure of the dissimilar welded joint has a better austenite to ferrite ratio (closer to 50:50), compared to the microstructure observed on similar duplex steel welded joints [56]. This is caused by mixing of 316L steel and 2304 steel as well as by lower thermal conductivity of austenitic steel, which extends the cooling process and the time for austenite formation. Within the weld face, the weld width is greater, which affects a different amount of heat to be dissipated by the base material than in the case of weld root. The differences in thermal cycles, in these areas, cause changes in weld microhardness in the face and in the root of the weld. The microhardness of the autogenous weld metal (average 254 HV0.2) is higher than microhardness of the 2304 lean duplex stainless steel (average 233 HV0.2). The solidification microstructures of laser welds are generally consistent with the prediction from the Schaeffler diagram.
Based on the results of present study, autogenous fiber laser welding of 316L austenitic and 2304 lean duplex stainless steels without using a ceramic backing can be recommended as a suitable procedure for sound welded joints.