Fiber laser welding of HSLA steel by autogenous laser welding and autogenous laser welding with cold wire methods

Fiber laser welding of HSLA steel by autogenous laser welding and autogenous laser welding with cold wire methods

Journal of Materials Processing Tech. 275 (2020) 116353 Contents lists available at ScienceDirect Journal of Materials Processing Tech. journal home...

18MB Sizes 0 Downloads 36 Views

Journal of Materials Processing Tech. 275 (2020) 116353

Contents lists available at ScienceDirect

Journal of Materials Processing Tech. journal homepage:

Fiber laser welding of HSLA steel by autogenous laser welding and autogenous laser welding with cold wire methods⋆


Shiwei Zhanga,b, Junhao Suna,b, Minhao Zhuc, Lin Zhangc, Pulin Niea,b, Zhuguo Lia,b,


Shanghai Key laboratory of Materials Laser Processing and Modification, School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China b Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration, Shanghai, 200240, China c Jiangnan Shipyard (Group) Co., Ltd., Shanghai, 201913, China



Associate Editor: C.H. Caceres

6 mm thickness high strength low alloy (HSLA) steel butt joints were fabricated by fiber laser welding. The influence of welding parameters on morphologies, weld process behavior, microstructure and mechanical properties were investigated. Defects (undercut, spatters and humping) were observed with the increase of heat input (HI) during autogenous laser welding (ALW), and weld morphologies changed from partial penetration to root humping first, then fine result, and finally over penetration (cutting) via high-speed images (HSI). Three typical transition modes were observed during autogenous laser welding with cold wire (LWACW), including spreading-transition, liquid-bridge transition and explosion transition. Sound process stability was achieved in liquid-bridge transition with the optimized parameters of 6 kW laser power, 0.8 m/min welding speed and 3.5 m/min feeding wire speed. The microstructure in the fusion zone was composed of acicular ferrite with some equiaxed grains, while that of autogenous laser welding was mainly made up of the acicular ferrite. The microhardness of base metal was relatively lower than that of fusion zone in both types, and both joints fractured at the side of base metal during tensile test.

Keywords: HSLA steel Autogenous laser welding Process stability Laser welding with wire filler High speed images

1. Introduction Laser welding processes has been divided into laser welding with a cold wire (LWACW) and autogenous laser welding (ALW) based on whether a filler metal was used. Despite of many potentials for jointing HSLA steels, ALW still existed some limitations and problems such as poor bridge ability and high accurate edge preparation requirement of welded joint, reported by Schultz et al. (2014). By contrast, LWACW not only made up for the above limitations of ALW, but also offset the loss of elements and introduced the nucleation to promote grain refinement owing to the filler wire, noted by Lei et al. (2018). Besides, LWACW improved process stability and got sound weld shape to avoid collapse, blowhole and other defects. The morphology and properties of weld joints were significantly determined by the process stability. According to Li et al. (2014), most of studies focused on the process stability and weld quality of HSLA steel by ALW and LWACW. Frostevarg (2018a) and Haug et al. (2013) investigated root bead morphology of full penetration laser welding,

and root spatter criteria is classified into formation regimes. Salminen and Kujanpää (2003) reported that defects by ALW could be compensated via adjusting the heat input and decreasing the wire feeding tolerances by LWACW. Wire feed rate had a significant influence on the interactions between cold wire, laser beam, the welding stability and welding quality. Longfield et al. (2007) utilized LWACW and ALW to weld 2.5 mm butt joints, and observed that the gap-bridging capability respectively increased 0.2 mm and 0.4 mm. Yu et al. (2013b) obtained sound weld seam with full penetration was obtained on 10 mm steel plate by ALW and LWACW, and concave surface at the bottom of weld could be inhibited by the specific shielding gas nozzle. Although some efforts for ALW and LWACW of thick plate were made in the decades, the research of welding process behavior and defects formation mechanism was limited. In this work, the weldability of 6 mm thick HSLA steel plates was systematically analyzed by ALW and LWACW processes, respectively. The process stability of ALW and LWACW processes were evaluated. The relationship between the defect formation and the process stability

Author acknowledge both Dr. Caceres and Dr. Na as the AEs. Corresponding author at: Shanghai Key laboratory of Materials Laser Processing and Modification, School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China. E-mail address: [email protected] (Z. Li). ⋆ ⁎ Received 11 May 2019; Received in revised form 25 July 2019; Accepted 28 July 2019 Available online 29 July 2019 0924-0136/ © 2019 Elsevier B.V. All rights reserved.

Journal of Materials Processing Tech. 275 (2020) 116353

S. Zhang, et al.

Table 1 Nominal composition of HSLA steel and ER50-6 cold wire. Name

Steel ER50-6

Elements/wt.% Fe









Bal. Bal.

0.07 0.06˜0.15

1.42 1.40˜1.85

0.20 0.80˜1.15

0.007 ≤0.035

0.012 ≤0.025

0.15 –

0.02 –

0.02 ≤0.5

was studied. To evaluate the welding integrity and quality, the metallurgical characterization and mechanical properties of weld joints were also analyzed.

Table 2 Welding parameters used in the ALW process.

2. Experimental procedure The base materials (BM) used for the study were a kind of ship-used HSLA steel. The length, width and thickness of samples were 300 mm, 150 mm and 6 mm, respectively. The ER50-6 cold wire with a diameter of 1.2 mm was the filler metal. Table 1 demonstrated the chemical composition of BM and filler wire. The metallurgical microstructure of HSLA steel was composed of polygonal ferrite and fine-grained pearlite, as indicated in Fig. 1. The ALW and LWACW experiments were conducted with an integrated welding system, consisting of IPG fiber laser system and a KUKA® robot. The maximum output power and continuous wavelength of fiber laser was 10 kW and 1.06 μm, respectively. The laser head was tilted 6° from the horizontal axis for the sake of protection of the optics from laser reflections. The wire feeding system of Fronius was utilized during the LWACW process. And the laser-electrode separation was held constant at 0 mm. A mixed shielding gas of 80% Ar + 20% CO2 to protect the weld bead. The flow rate of shielding gas was held constant at 20 L/min. Prior to welding, the steel plates were polished, removed dirt and degreased with acetone, then welded in butt configuration. The fluid flow behavior of welded pool and weld root during ALW and LWACW processes were observed by a Photron Fast cam SA4 highspeed camera. An assistant laser light was used to highlight the welding zone for capturing the clear high-speed videos and images. The ALW and LWACW processes were observed at 2000 frames per second. Tables 2 and 3 showed the welding parameters of ALW and LWACW processes, respectively. After welding process, the weld specimens were cut from the stable state of the welding beams and ground with silicone papers, and then polished, following by etching with a 2% natal solution for 50 s. Optical microscopy (OM) and scanning electron microscopy (SEM, TESCAN-




Laser power Spot diameter Welding speed Defocus distance

3.0, 4.0, 4.6, 5.0, 5.5 0.72 0.8 0

kW mm m/min mm

Table 3 Welding parameters used in the LWACW process. Parameter



Laser power Spot diameter Welding speed Defocus distance Laser-electrode separation Wire feeding rate

5.0, 5.6, 6.0, 6.5 0.72 0.8 0 0 3.5

kW mm m/min mm mm m/min

MIRA3) were utilized to observe the weld bead morphology and microstructure of weld joints. The Vickers micro-hardness cross-section measurements were conducted on the welds with 200 gf under 15 s dwell time. To analyze hardness discrepancy at different locations, three lines of measurements were located at about 1.5–2 mm from the surface (top and bottom), and middle region of the joints. Based on ASTM: E8/E8M-13a standard, tensile test specimens were conducted. The tensile tests were performed via a Zwick T1-FR020TN-A50 testing machine at room temperature. The loading rate was a constant of 1 mm/min. Ultimate tensile strength (UTS) and elongation (A%) were recorded to evaluate the joint property of the welds. 3. Results and discussion 3.1. Weld morphologies of ALW and LWACW welded joints Figs. 2 and 3 showed the surface morphologies and cross-sections of the welded joints obtained by different laser powers during ALW and LWACW, respectively. As shown in Fig. 2, the penetration depth of ALW joints increased with the increase of laser power. Keyhole porosity formed at 3 kW laser power during ALW, indicating the incomplete penetration of the welded joint. It was because that low laser power caused the keyhole to be unstable and easily collapsed, leading to voids consisted of entrapped vapor, similar to the study by Elmer et al. (2015). Full penetration of the welded joints was produced as laser power exceeded 4 kW. The top surface showed concavity with undercut and underfill, and the humping was generated at the bottom surface of welded seam. At 4 kW the width of the welded seam was periodically wide-narrow-wide with deep undercut due to the unstable keyhole and burning loss of metals. The bottom surface formed many continuous root humps, which was attributed to the interactions of the keyhole stability, droplet impinging momentum and surface tension, as reported by Ohnishi et al. (2013). A relatively sound and continuous weld bead was achieved with undercut and small spatters at 5 kW. As the laser power increased to 5.2 kW, excessive penetration was obtained with deeper concavity, undercut and root humps. It was due to

Fig. 1. The metallurgical microstructure of HSLA steel. Yellow square box indicates higher multiples of pearlite. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article). 2

Journal of Materials Processing Tech. 275 (2020) 116353

S. Zhang, et al.

Fig. 2. The appearances and cross-sections of welded joints under the ALW conditions at v = 0.8 m/min, and P ranging from 3.0 kW to 5.2 kW.

two reasons: (1) the excessive heat input produced a great amount of molten metal; (2) elongation of the molten metal increased intensive downward melt flow, according to Nothdurft et al. (2018). Compared with the Fig. 2, The appearance and welded joint crosssections were different produced during the LWACW process, as shown in Fig. 3. As laser power increased from 5.0 kW to 5.6 kW, a partial penetration with narrow convex ridges was obtained at the top surface. According to Atabaki et al. (2016), the nominal heat input (HI) from the laser power was calculated as the following equation:


P × 60 v


attain the full penetration and the sound geometry of joints, the laser power should be increased to over 5.6 kW. At 6.0 kW, the top side of weld bead was continuous and uniform without apparent defects, such as spatters, cracks and undercut. And the convexity appeared on the surface of weld cross-section. At 6.5 kW, trapped pores, root humping and the undercut occurred due to excessive heat input. the laser plasma ejecting force sharply increased and disturbed the liquid fluid of molten pool into whirlpool. It’s easy for gas to melt into molten pool during unstable welding process. Escape velocity of pores (ve) was less than solidification rate (R) of molten pool, leading to pore formation. According to Stokes equation, escape velocity of pores was expressed as follows:


( Cp (Tm

Ta ))coldwire


ve =

Where P denotes laser power (kW), and v denotes welding speed (m/ min). ρ and Cp are density and heat capacity of the wire, respectively. And Tm and Ta are the melting point of cold wire and the room temperature. ( Cp (Tm Ta))coldwire denotes the heat input used to melt the cold wire. Part of laser power was consumed by the cold wire during LWACW process so that HI in the molten pool decreased, and the volume of the pool also decreased. Besides, insufficient HI resulted in the increase of surface tension and viscosity, which led to the poor wettability and spreading between solid metal and molten pool. The combination of surface tension, viscosity and the capillary motion caused the convex surface, as reported by Gusarov and Smurov (2010). To


G ) gr




Where ve is escape velocity of pores. ρG and ρ are density of gas and liquid metal and, respectively. g and r are gravity and the radius of trapped gases. η is liquid metal viscosity. By comparison and analysis, it has been demonstrated that as the laser power increased to 5 kW, the appearance of welded joint was well formed without obvious defects in ALW process, as shown in Fig. 4. LWACW made up for the shortcomings of autogenous laser welding, such as spatters and undercut, since the addition of welding wire not only increased the stability of welding process, but also eliminated 3

Journal of Materials Processing Tech. 275 (2020) 116353

S. Zhang, et al.

Fig. 3. The appearances and cross-sections of welded joints under the LWACW conditions at v = 0.8 m/min, and P ranging from 5.0 kW to 6.5 kW.

undercut on the top side of welded seam. Sound and continuous welded joint was attained and mechanical properties of welded joints was guaranteed, as shown in Fig. 5.

escaped from the bottom in full penetration and deteriorated the mass loss, resulting in the formation of the undercuts. Similar to the results by Zhang et al. (2018), with the increase of laser power, root humping was formed due to lots of molten metal accumulated at the weld bottom surface. Based on the high-speed images and the morphologies of samples above, Fig. 5 illustrated the various melt flow behaviors under different conditions. HI had a significant influence on melt flow behaviors and morphologies of welded joints, depending on kinds of combinations of laser power and welding speed. It was noteworthy that the penetration increased with the increase of HI during ALW process. According to Frostevarg (2018b), the melt flow behaviors could be divided into four cases: partial penetration, root humping, fine result and over penetration. Fig. 5(a) showed partial penetration as HI was small. When the laser light irradiated onto the metal surface, most of the energy was absorbed, while the rest lost due to the metallic plume and spatters escaped from the top surface. Fig. 5(b) showed root humping as the heat input increased. The keyhole of the laser was periodic open and close, which resulted in hydrodynamic instability of laser molten pool. Thus Plateau-Raleigh capillary instability was responsible for the formation of humping, according to Eq.(4) proposed by Debroy et al. (2017).

3.2. Process behavior The high-speed images with different laser powers during ALW process were demonstrated in Fig. 4. Fig. 4(a) demonstrated the top and bottom surface of the molten pools at 3 kW. The base metal was welded in partial penetration, and the metallic plume was very unstable in height and width. As illustrated in Fig. 4(b), at t0, the laser plume at the bottom was generated and the keyhole occurred, and then the liquid metal flowed via the channel to the humping zone (t0 +10ms). Based on the theory of surface tension, the higher the temperature, the lower the surface tension. The surface tension of keyhole inlet was much lower than that of the humping, leading to the formation of humping, according to Pan et al. (2016). The laser went in the welding direction away, solidification occurred at the bottom, impeding the molten metal flow (t0 + 15 ms). The metallic plume was escaped periodically and then the periodic humps were formed on the bottom side at t0 + 20 ms, as illustrated in Fig. 4(b). For full penetration welding process, a part of the laser energy, metallic plume and small spatters escaped from the bottom at 5 kW and 5.2 kW, and the keyhole penetrated through the base metal. Fig. 4(c) showed that the dynamics of the laser plume, demonstrating the keyhole and weld pool were stable. The oscillations of the plume were accompanied with the fluctuation of the keyhole on the bottom surface, and the keyhole was kept open for the plume to erupt. The metallic metal flowed into the keyhole and then solidified at the bottom surface, in concordance with results of Zhao et al. (2017). A relatively sound weld bead with smooth bottom surface could be obtained. As shown in Fig. 4(d), many spatters and metallic plume

L /D


Where L denotes the length of the molten metal. D denotes the width of molten pool. As the welding continued, D fluctuated with periodic open – close keyhole, leading to L/D instability. Therefore, the bottom melt pool divided into small spherical balls, commonly known as humping, to keep the uniform capillary pressure, as reported by Cho and Farson (2007). Fig. 5(c) illustrated relatively fine result. The appearance of the top surface was slight weld concavity and undercut, but the appearance of the bottom surface was smooth and sound without unfilled and 4

Journal of Materials Processing Tech. 275 (2020) 116353

S. Zhang, et al.

Fig. 4. High speed camera images observing the top and bottom surface of the ALW process at v = 0.8 m/min (right to left welding direction): (a) P = 3 kW; (b) P = 4 kW; (c) P = 5 kW; (d) P = 5.2 kW.

undercut defects. With the increase of HI, over penetration occurred, companied with metal plume and spatters escaped from the opening keyhole of the bottom surface, as demonstrated in Fig. 5(d). Once the laser keyhole penetrated, the undercut at the top surface occurred. The formation of the undercut was due to the downward melt flow and the pronounced vaporization of metal elements. Surface tension of molten metal had a great influence on the flow of fluid melt. According to Zhao et al. (2010), surface tension decreased as temperature increased, and the higher temperature of molten metal in the pool center was pulled outward solid material at the pool edge, thus possibly causing undercut. Compared with ALW, the addition of filler wire in LWACW improved the weld formation and affected the flow of the molten pool. Stable and continuous transition was the prerequisite for good weld

formation and weld quality. As the laser power increased from 5.0 kW to 6.5 kW, three typical transition modes occurred, including spreadingtransition, liquid-bridge transition and explosion transition. Fig. 6 showed the typical spreading transition mode during LWACW process at 5.0 kW. At the initial stage of t0, the filler wire was melted by high energy of laser beam and metal vapor radiation energy. The molten metal was transferred in the spreading transition mode, as shown in Fig. 6(a). Due to the periodicity of the melting and feeding of the wire, the flow of molten pool was unstable, leading to the generation of pores, as illustrated in Fig. 6(b) and (c). Under combination of laser recoil force and surface tension, molten metal flowed toward the molten pool rear, and heat input was so low that the poor wetting of molten pool resulted in poor spreading, as demonstrated in Fig. 6(d) 5

Journal of Materials Processing Tech. 275 (2020) 116353

S. Zhang, et al.

Fig. 5. Schematic illustration of melt flow behaviors: (a) partial penetration; (b) root humping; (c) fine result; (d) over penetration.

from the edge of the keyhole, according to Yu et al. (2013a). Spatters led to the loss of melt, resulting in undercut on the surface and thus poor joining, as depicted in Fig. 8(b) and (c). Molten pool produced liquid expulsion under the recoil pressure and then formed a whirlpool, as illustrated in Fig. 8(d) and (e). The liquid expulsion and whirlpool broke the dynamic balance of the molten pool and laser keyhole, which had a significant impact on the LWACW process. The schematic illustration of typical explosion transition mode was demonstrated in Fig. 8(f). By observing the process behaviors of ALW and LWACW, it was noticed that LWACW possessed a semi-hemispherical geometry on the top side and reduced the formation of undercut and spatters in liquid bridge mode, compared to that of ALW. LWACW improved the stability of weld process and enhanced the appearance of the weld geometry by the addition of filler wire. That was attributed to two reasons. On the one hand, the liquid bridge transition was stable and the melt metal made up for the loss of the metal which flowed to the bottom side; on the other hand, laser beam irradiated at the tip of the filler wire, instead of directly impacting the weld pool, keeping the keyhole stable, which decreased the fluctuation of laser beam and boosted the flow stability of the molten pool.

and (e), which was similar to the study of Luo et al. (2016). Fig. 6(f) showed the schematic illustrations of typical spreading transition. Once the edge of cold wire contacted with the laser beam and molten pool, it would be melted and enter the pool. Cold wire of the solid state continued entering the molten pool and was subsequently melted by the heat conduction in the molten pool. Fig. 7 showed the liquid-bridge transition mode during LWACW process at 6.0 kW. As shown in Fig. 7(a), the filler wire was melted and formed liquid flow under the combination of laser, metal vapor, and heat irradiation energy. And then the liquid metal flowed into the welded pool under the combined action of the gravity and surface tension. Fig. 7(b) showed the liquid-bridge transition mode, in which feeding rate and melting rate of the wire reached a dynamic equilibrium. And a stable liquid metal flow was generated between the surface of solid wire tip and the molten pool, which resembled a liquid bridge to make the metal liquid transfer stably, as depicted in Fig. 7(c)–(e). Fig. 7(f) demonstrated the schematic illustrations of typical liquid-bridge transition. Liquid metal flowed into the welded pool in liquid-bridge transition mode, and it guaranteed the stability of the keyhole and liquid flow. Thus liquid-bridge transition was an optimal transition mode during the LWACW process. Similar behavior has been observed by Ma et al. (2017). Fig. 8 showed the explosion transition mode at 6.5 kW during the LWACW process. As the laser power increased to 6.5 kW, the metal vapor/plasma ejecting force increased and the dynamic balance of the molten pool and keyhole was deteriorated, resulting in the drastic oscillation near the keyhole. The plasma ejecting force impacted fluid flow of the liquid metal ejected from weld pool and then the fluid flow disintegrated into droplets and formed spatters, as depicted in Fig. 8(a). The laser beam disturbed the liquid metal flow and tiny spatters ejected

3.3. Microstructure of welds Optical and SEM micrographs of cross-section welds by the ALW and LWACW processes under optimum parameters were depicted in Figs. 9 and 10 , respectively. The FZ of both specimens displayed the dendritic microstructures consisting predominately of acicular ferrite (AF) determining the mechanical property of welded joints. In addition, a handful of allotriomorphic ferrite called grain boundary ferrite (GBF) 6

Journal of Materials Processing Tech. 275 (2020) 116353

S. Zhang, et al.

Fig. 6. The wire melting and spreading-transition behavior at 5.0 kW for the LWACW process.

and side-plate ferrite (SPF) occurred under the ALW and LWACW processes, as reported by Wei et al. (2015). The HAZ of the joints consisted of sub-regions, i.e. coarse-grained HAZ (CGHAZ), fine-grained HAZ (FGHAZ), and inter-critical HAZ (ICHAZ), and the microstructure of HAZ was a mixture of ferrite (F), pearlite (P) and martensite with some bainite, similar to the report of Munro et al. (2012), as illustrated in Figs. 9 and 10, respectively. The microstructures of the ICHAZ containing ferrite and pearlite were very heterogeneous for each condition, as demonstrated n in Figs. 9(b) and 10 (b). In FGHAZ, the microstructure consisted of uniform and fine ferrite and pearlite. Grain refinement was ascribed to a great quantity of nucleation particles existed at the ferrite grain boundaries in the heating cycle, as illustrated in Figs. 9(c) and 10 (c). Figs. 9(d) and 10 (d) showed the microstructure of CGHAZ. The higher heat input by ALW process was expected to result in the coarse martensite with some bainite than that one by LWACW process. Since the ALW and LWACW processes were characterized with high energy-density and fast cooling rate, microstructures in the FZ mainly consisted of columnar grains under the optimum conditions of ALW and LWACW. Compared to the FZ micrographs of ALW, the FZ micrographs of LWACW demonstrated that the interwoven, needle-like acicular ferrite was tinier and finer, as demonstrated in Figs. 10(f) and (g). Besides, fine equiaxed grains formed in FZ, which was contributed to two aspects: on the one hand, laser energy of molten wire entered molten pool through liquid metal, thus resulting in the decrease of heat input reaching the base metal and the reduction of cooling rate, as stated by

Liu et al. (2015); on the other hand, foreign particles of filler wire were presented in the weld pool and acted as heterogeneous nucleus, the same result was confirmed by Shao et al. (2018). When the liquid metal crystallized, the crystal nucleus expected to attach to the exterior of these refractory particles and grow continually, as shown in Fig. 10(e). The grain sizes of GBF and SPF prepared by LWACW were smaller than those attained by ALW, due to the higher heat input of LWACW and the addition of refractory particles, as shown in Figs. 9(f) and 10 (f). 3.4. Mechanical properties Fig. 11 showed the Vickers’ microhardness distributions obtained by ALW and LWACW processes under optimum parameters, respectively. It was observed that the hardness curves were similar to saddle shaped distributions. The average hardness of HSLA steel plate was approximately 185 HV, but the hardness was higher in both FZ and HAZ during the ALW and LWACW processes due to high cooling rate, which indicated microstructure evolution occurred under the action of thermal cycles. The higher microhardness of the weld joints was contributed to the formation of martensite with some bainite, as demonstrated in Figs. 9 and 10. The highest hardness was due to the formation of martensite and/or some bainite in the fusion line as depicted in Figs. 9(d) and 10 (d) The acicular ferrite in FZ and grain refinement in HAZ were responsible for the increase of hardness. Fig. 11(a) displayed an average value of 265 ± 10 HV in the FZ and a maximum value of 293 HV in HAZ and then decreased from 290 to


Journal of Materials Processing Tech. 275 (2020) 116353

S. Zhang, et al.

Fig. 7. The wire melting and liquid-bridge transition behavior at 6.0 kW for the LWACW process.

185 ± 5 HV. It didn’t exist a softened region in the HAZ, which was attributed to the low heat input and fast cooling rate of ALW, as observed by Cao et al. (2011). Besides, at the mid-thickness, the width of FZ was narrower than the top and bottom regions and the highest hardness value was slightly less than that of the other regions. The weld metal hardness was nearly identical and the highest value in the HAZ was also similar, as depicted in Fig. 11(b). The difference between ALW and LWACW processes was that the total heat input was higher during LWACW than during ALW, which slowed weld cooling rate. The volume of martensite and/or some bainite of LWACW process was less than that of ALW process. Fig. 12 indicated the stress–strain behavior for the ALW and LWACW specimens under optimum parameters. Both welded joints produced the necking and subsequently continuous deformation was confined at the necking, as shown by the specimen insets in Fig. 12. Facture ultimately occurred at the base metal in ductile fracture mode. The typical behaviors were contributed to microstructure evolution in FZ and grain refinement in HAZ. Besides, the microstructures of FZ and HAZ were harder than that of the base metal. The transverse ultimate yield strength, tensile strength and elongation of LWACW joints were 487.35 MPa, 560.76 MPa and 15.87%, respectively, but those of ALW joints were 494.87 MPa, 572.36 MPa and 14.41%, respectively. By comparison and evaluation, mechanical properties of the joints of the two welding methods were similar. Ultimate yield strength and tensile strength of LWACW joints were slightly lower than those of ALW joints; elongation of LWACW joints were slightly higher than that of ALW

joints. This indicated fiber laser welding of HSLA steel was feasible and LWACW facilitated higher toughness of the joints. 4. Conclusions (1) Consistent, sound welded joints were successfully obtained with the convex reinforcement and full-penetration at a broad condition for LWACW process. The slight concave weld-bead morphology with the fine bottom was only formed in a narrow operating window for ALW process. (2) Continuous and sound transition was the prerequisite for good weld morphology and weld quality during the LWACW process. As laser power ranged from 5.0 kW to 6.5 kW, three typical transition modes occurred. The optimal mode of liquid transfer was liquid-bridge transition. (3) For both ALW and LWACW joints, microstructure of FZ was predominately composed of interwoven, needle-like acicular ferrite, and phases of HAZ were mainly a mixture of pearlite and ferrite. For LWACW process, filler wire slowed weld cooling rate and promoted heterogeneous nucleation. The microstructures were finer and more uniform, and mechanical properties were better. (4) Microhardness distributions in ALW and LWACW joints were similar. The hardness in FZ of LWACW joints was slight softener than that in ALW joints. Measurements verified that mechanical properties of those joints were nearly equal and all the joints raptured with the necking at the base metal.


Journal of Materials Processing Tech. 275 (2020) 116353

S. Zhang, et al.

Fig. 8. The wire melting and explosion transition behavior at 6.5 kW for the LWACW process.

Fig. 9. Microstructure of the different zones obtained by ALW under optimum parameters: at v = 0.8 m/min, P = 5.0 kW.


Journal of Materials Processing Tech. 275 (2020) 116353

S. Zhang, et al.

Fig. 10. Microstructure of the different zones obtained by LWACW under optimum parameters: at v = 0.8 m/min, P =6.0 kW, and vs = 3.5 m/min (vs denotes the feeding rate).

Fig. 11. Microhardness profiles of the welded joint obtained by ALW and LWACW under optimum parameters.

Acknowledgment This work was provided and supported by the National Nature Science Foundation of China (No. 51775338). References Atabaki, M.M., Yazdian, N., Ma, J., Kovacevic, R., 2016. High power laser welding of thick steel plates in a horizontal butt joint configuration. Opt. Laser Technol. 83, 1–12. Cao, X., Wanjara, P., Huang, J., Munro, C., Nolting, A., 2011. Hybrid fiber laser – arc welding of thick section high strength low alloy steel. Mater. Des. 32, 3399–3413. Cho, M.H., Farson, D.F., 2007. Simulation study of a hybrid process for the prevention of weld bead hump formation. Weld. J. 86, 253-s–262-s. Debroy, T., Wei, H.L., Zuback, J.S., Mukherjee, T., Elmer, J.W., Milewski, J.O., Beese, A.M., Wilson-Heid, A., De, A., Zhang, W., 2017. Additive manufacturing of metallic components – process, structure and properties. Prog. Mater. Sci. 92. Elmer, J.W., Vaja, J., Carlton, H.D., Pong, R., 2015. The effect of Ar and N-2 shielding gas on laser weld porosity in steel, stainless steels, and nickel. Weld. J. 94, 313s–325s. Frostevarg, J., 2018a. Factors affecting weld root morphology in laser keyhole welding. Opt. Laser Eng. 101, 89–98. Frostevarg, J., 2018b. Factors affecting weld root morphology in laser keyhole welding. Opt. Lasers Eng. 101, 89–98.

Fig. 12. The stress–strain behavior for ALW and LWACW specimens under optimum parameters. 10

Journal of Materials Processing Tech. 275 (2020) 116353

S. Zhang, et al. Gusarov, A.V., Smurov, I., 2010. Modeling the interaction of laser radiation with powder bed at selective laser melting. Phys. Procedia 5, 381–394. Haug, P., Rominger, V., Speker, N., Weber, R., Graf, T., Weigl, M., Schmidt, M., 2013. Influence of laser wavelength on melt bath dynamics and resulting seam quality at welding of thick plates. Phys. Procedia 41, 49–58. Lei, Z., Zhang, K., Zhou, H., Ni, L., Chen, Y., 2018. A comparative study of microstructure and tensile properties of Ti2AlNb joints prepared by laser welding and laser-additive welding with the addition of filler powder. J. Mater. Process. Technol. 255, 477–487. Li, S., Chen, G., Zhang, M., Zhou, Y., Zhang, Y., 2014. Dynamic keyhole profile during high-power deep-penetration laser welding. J. Mater. Proc. Technol. 214, 565–570. Liu, W., Ma, J., Liu, S., Kovacevic, R., 2015. Experimental and numerical investigation of laser hot wire welding. Int. J. Adv. Manuf. Technol. 78, 1485–1499. Longfield, N., Lieshout, T., Wit, I.D., Veldt, T.V.D., Stam, W., 2007. Improving laser welding efficiency. Weld. J. 86, 52–54. Luo, Y., Tang, X., Deng, S., Lu, F., Chen, Q., Cui, H., 2016. Dynamic coupling between molten pool and metallic vapor ejection for fiber laser welding under subatmospheric pressure. J. Mater. Process Technol. 229, 431–438. Ma, G., Li, L., Chen, Y., 2017. Effects of beam configurations on wire melting and transfer behaviors in dual beam laser welding with filler wire. Opt. Laser Technol. 91, 138–148. Munro, C., Nolting, A.E., Cao, X.J., Wanjara, P., 2012. Hybrid laser-arc welding of HSLA65 steel plate: microstructural and mechanical property evaluation of butt welds. Mater. Sci. Forum 706–709, 2992–2997. Nothdurft, S., Springer, A., Kaierle, S., 2018. Influencing the Weld Pool during Laser Welding. Ohnishi, T., Kawahito, Y., Mizutani, M., Katayama, S., 2013. Butt welding of thick, high strength steel plate with a high power laser and hot wire to improve tolerance to gap variance and control weld metal oxygen content. Sci. Technol. Weld. Join. 18,

314–322. Pan, Q., Mizutani, M., Kawahito, Y., Katayama, S., 2016. Effect of shielding gas on laser–MAG arc hybrid welding results of thick high-tensile-strength steel plates. Weld. World 60, 653–664. Salminen, A.S., Kujanpää, V.P., 2003. Effect of wire feed position on laser welding with filler wire. J. Laser Appl. 15, 2–10. Schultz, V., Seefeld, T., Vollertsen, F., 2014. Gap bridging ability in laser beam welding of thin aluminum sheets. Phys. Procedia 56, 545–553. Shao, Y., Liu, C., Yan, Z., Li, H., Liu, Y., 2018. Formation mechanism and control methods of acicular ferrite in HSLA steels: a review. J. Mater. Sci. Technol. 34. Wei, H.L., Blecher, J.J., Palmer, T.A., Debroy, T., 2015. Fusion zone microstructure and geometry in complete-joint-penetration laser-arc hybrid welding of low-alloy steel. Weld. J. 94, 135s–144s. Yu, Y., Huang, W., Wang, G., Wang, J., Meng, X., Wang, C., Yan, F., Hu, X., Yu, S., 2013a. Investigation of melting dynamics of filler wire during wire feed laser welding. J. Mech. Sci. Technol. 27, 1097–1108. Yu, Y.C., Yang, S.L., Yin, Y., Wang, C.M., Hu, X.Y., Meng, X.X., Yu, S.F., 2013b. Multi-pass laser welding of thick plate with filler wire by using a narrow gap joint configuration. J. Mech. Sci. Technol. 27, 2125–2131. Zhang, M., Zhang, Z., Tang, K., Mao, C., Hu, Y., Chen, G., 2018. Analysis of mechanisms of underfill in full penetration laser welding of thick stainless steel with a 10 kW fiber laser. Opt. Laser Technol. 98, 97–105. Zhao, C.X., Kwakernaak, C., Pan, Y., Richardson, I.M., Saldi, Z., Kenjeres, S., Kleijn, C.R., 2010. The effect of oxygen on transitional Marangoni flow in laser spot welding. Acta Mater. 58, 6345–6357. Zhao, S., Yang, L., Liu, T., Yang, R., Pan, J., 2017. Analysis of plasma oscillations by electrical detection in Nd: YAG laser welding. J. Mater Process Technol. 249, 479–489.