This study investigated the effect of heat input on the microstructure and tensile properties of electron beam welded joints between 1J50 soft magnetic alloy and 06Cr19Ni10 stainless steel. The weld zone (WZ) was primarily composed of coarse columnar grains of FeNi solid solution with 6-8 wt% Cr. Segregation of Si and P was detected at grain boundaries. As the heat input decreased from 60.5 J/mm to 38.5 J/mm, the average grain size in the WZ reduced from 42.43 mu m to 20.06 mu m and the local area fraction of segregated phases decreased from 5.73 % to 1.94 %. These microstructural changes contributed to an increase in joint tensile strength from 396.7 MPa to 426.7 MPa, reaching 98 % of that of the 1J50 base metal (BM). However, the joint produced with the lowest heat input failed prematurely due to incomplete fusion. The reduction in heat input also resulted in the disappearance of fine equiaxed grains in weld center, leading to decreased joint elongation. The microhardness of the WZ was lower than that of both BMs because of the coarse columnar grains.

Severe crack defects significantly limit the application of NiTi SMA/Ti6Al4V fusion welded joints in aerospace engineering. Currently, there is a lack of effective electron beam welding techniques to control the formation of such cracks. This study developed an electron beam offset barrier welding technology based on beam offset to eliminate crack defects. Results demonstrated that merely adjusting the electron beam offset distance toward the Ti6Al4V side could not suppress crack formation. The weld zone produced abundant brittle Ti2Ni intermetallic compounds and extremely high transverse residual tensile stresses, leading to longitudinal cracks. Crack-free joints were achieved by introducing a Mo barrier layer and optimizing its thickness, height, and beam offset distance. The 300 mu m thick Mo layer effectively prevented base metals mixing, substantially reducing Ti2Ni formation. The diffusion layers, primarily formed through Mo-Ti interdiffusion, established robust joining between the Mo layer and both base materials. The joint strength increased from 76 MPa to 308 MPa, with the fracture location shifting from the Ti2Ni reaction layer to the Mo barrier layer. This study provides a novel strategy for controlling crack defects in dissimilar material welded joints with poor metallurgical compatibility, and promotes the engineering application of NiTi shape memory alloy/Ti6Al4V alloy electron beam welded joints.

Microstructure and properties of joints produced by electron beam direct welding and beam offset welding of Zr-4 and TC4 are investigated. The microstructure of the fusion zone in both types of joints mainly consists of (3 phase and alpha' martensite. A 30-mu m reaction layer composed of alpha phase is detected near the Zr-4 fusion line in direct-welded specimens. The reaction layer exhibits relatively low hardness and tensile strength. Beam offset welding toward Zr-4 eliminates the reaction layer, increasing the tensile strength of the joint from 437 MPa to 563 MPa, which is close to that of Zr-4 base metal.

This study successfully achieved the dissimilar material joining of Ta/GH4169 through electron beam welding and systematically investigated the effect of different welding speeds on the microstructure and mechanical properties of the joints. Microstructural analysis revealed that as the welding speed increased, the weld morphology gradually transitioned from a bowl shape to a nail-like shape, and the proportion of eutectic content in the weld decreased. The weld hardness decreased from 478 HV to 332 HV. However, regardless of the welding speed, the welds were composed of a dual-phase structure of FCC and FCC + Laves eutectic phases. The FCC phase exhibited good plasticity, and deformation mainly occurred through dislocation slip mechanisms. The uniformly distributed hard Laves phase significantly hindered dislocation slip, forming various dislocation configurations, which helped improve the joint's strength and toughness. Nevertheless, due to the weak bond between the weld and the Ta base material, the joint typically failed at the fusion line on the Ta side. Ultimately, the tensile strength of the joint at room temperature and 400 degrees C were 339 MPa (at moderate welding speed, approximately 56% of the Ta base material strength) and 332 MPa (at high welding speed), respectively. These findings provide valuable guidance for achieving high-quality joining of Ta with nickel-based high-temperature alloys.

This study joins CoCrFeMnNi HEA/SA203 steel dissimilar materials using vacuum electron beam welding (VEBW), achieving a joint strength of 87% of the base material. However, despite this high mechanical strength, the joint is compromised by pronounced morphological undulations along the HEA-side fusion line, posing potential risks. Through a validated multi-physics simulation, we demonstrate that this defect is not due to compositional variations but is a direct consequence of the intrinsic thermophysical property mismatch between the materials. The HEA's poor thermal conductivity and high molten-state viscosity create severe asymmetric thermal-flow fields, establishing a dynamic instability within the molten pool. This instability is characterized by the convergence of stratified vortices and a periodic, high-velocity mass exchange, which is identified as the direct mechanism responsible for sculpting the irregular fusion line. The simulation results demonstrate excellent agreement with experimental microstructures and morphologies, validating the proposed mechanism. This work provides a fundamental understanding of the molten pool dynamics governing defect formation in dissimilar metal welding involving HEAs, offering a theoretical basis for mitigating such defects and advancing the reliable application of HEA-based cryogenic components.

This study investigated the effects of different beam offset distances on the microstructure and mechanical properties of Ta/GH4169 welded joints. The results showed that all joints fractured at the fusion line on the Ta side. The strength of the joint first increased and then decreased with the increase of the offset distance to the Ta side. When the beam offset distance was 0.5 mm, the strength of the joint was the highest, reaching 348 MPa at room temperature and 298 MPa at 400 degrees C, which were 268% and 481% higher than those of the joint without beam offset. The microstructure analysis indicated that the weld was mainly composed of FCC phase, FCC + Laves eutectic phase and BCC particles. Among them, the FCC phase formed a stable coherent interface between the Laves phases. In addition, a high density of nanoscale coherent L12 precipitates was uniformly distributed within the FCC matrix. These precipitates contributed to the strengthening of the FCC matrix and participated in local dislocation-mediated deformation, as evidenced by the formation of dislocation channels and deformation microbands. However, this beneficial effect was mainly confined to the microscopic scale within the weld metal. The overall tensile response of the joint was still predominantly governed by premature fracture near the brittle reaction layer on the Ta side. These results provide important references for the high-quality bonding of Ta with Ni-Fe-based materials, as well as for the development of high-strength eutectic high entropy alloys.

Electron beam welding was performed on the Mo-10Re alloy tube-rod structure. This study systematically analyzed the microstructure and properties of the joints welded under different beam currents. The results demonstrated that increasing the beam current coarsened the equiaxed grains within the weld, yet effectively promoted pore escape. The joint fabricated at 11 mA exhibited the maximum tensile strength at both room temperature and elevated temperature, with fracture occurring intergranularly in the weld. Although higher beam currents exacerbated weld softening, they markedly enhanced the corrosion resistance of the joints.

The Electron beam welding of AlCoCrFeNi₂.₁ eutectic high-entropy alloy to 304 stainless steel disrupts the original lamellar structure in the fusion zone, causing severe degradation of joint performance. To address this, we propose a combined strategy of Al alloying and post-weld heat treatment (PWHT). The optimized PWHT-3h joint exhibits an ultimate tensile strength of 765 MPa and a uniform elongation of 36.8%, with fracture shifting from the weld to the 304SS base metal, indicating a strength-overmatched joint. Microstructural analysis reveals that this enhancement arises from the formation of a hierarchical heterogeneous structure. The as-welded joint consists of an FCC matrix and a disordered BCC secondary phase. After PWHT, L12 precipitates form coherently within the FCC matrix, while the BCC phase transforms into an ordered B2 structure containing uniformly dispersed Cr-rich nanoprecipitates. The B2 ordering transformation and Cr-rich precipitation enhance interfacial stability and deformation compatibility, whereas the coherent L12 precipitates suppress dislocation pile-ups and strain localization through a dislocation-shearing mechanism. This synergistic multiscale strengthening leads to the remarkable improvement in both strength and ductility. Our findings provide a novel theoretical framework and technical pathway for tailoring microstructures and enabling engineering applications of high-performance HEA/stainless steel dissimilar joints.

Aluminum-lithium alloy was welded by electron beam in vacuum with three different scanning paths, non-scanning, sinusoid and cochleoid path, indicating the identical heat input, welding speed, and focusing state. Numerical simulation is conducted to compare evolution of keyhole, fluid flow in molten pool, and weld formation. The recoil pressure of metal vapor on keyhole wall is theoretically modified in view of the vapor behavior in vacuum. The non-scanning mode facilitates rapid keyhole initiation and favorable early stability, indicating excessive local heat, disordered fluid flow, frequent keyhole collapse, and weld defects such as collapse and porosity. The sinusoid scanning path redistributes heat through periodic oscillation, corresponding to the keyhole instability. The velocity along keyhole wall ranges from 1.280 m/s to 3.028 m/s, the average and maximum depth of keyhole are 2.24 mm and 2.94 mm, respectively, which still induces strong flow turbulence and keyhole collapse. The cochleoid scanning path provides the most uniform heat distribution and the most ordered flow field, giving a more concentrated molten pool velocity ranging from 1.328 m/s to 2.760 m/s and smaller average and maximum keyhole depths of 2.09 mm and 2.85 mm, respectively. This suppresses excessive bottom melting, effectively inhibits keyhole and weld collapse, and markedly improves process stability and welding quality.

The application of superconducting radio frequency cavities involved the joining of Nb and stainless steel. However, severe crack defects were observed in the fusion welded joints. Limited reports existed on cracks control of Nb/304 stainless steel joints fabricated by electron beam welding. To eliminate crack defects and improve joint performance, Cu interlayers with varying thicknesses were introduced. Compared to conventional electron beam offset welding methods, this approach significantly suppressed the formation of intermetallic compounds. The 0.9 mm Cu interlayer yielded the optimal joint morphology. At 0.3 mm Cu interlayer thickness, the weld microstructure primarily consisted of a-Fe matrix. Increasing the Cu interlayer thickness from 0.5 to 0.9 mm promoted the gradual dominance of Cu-rich solid solution, ultimately replacing a-Fe as the matrix phase. It progressively eliminated the c(Fe2Nb) Laves phase and retained only the mu (Fe7Nb6) reaction layer. The results demonstrated that the introduction of Cu interlayer significantly improved the mechanical properties of the joints. With increasing Cu interlayer thickness, the microhardness of the weld decreased while the joint strength exhibited progressive enhancement. The ultimate tensile strength of joint with 0.9 mm Cu interlayer reached 155 MPa, which was 60 % that of the Nb base metal of 258 MPa.

No continuous brittle phase precipitates at the fracture location of the molybdenum/Kovar alloy electron beam welded joint, indicating that unstable phase interfaces resulting from solid-state phase transitions constitute the fundamental cause of joint brittleness. To inhibit solid-state phase transitions and optimize interfacial stability, Permalloy was added to the weld. The original solid-state eutectoid reaction (R(Fe5Mo3)-* alpha-Fe + mu(Fe3Mo2)) was successfully replaced by a liquid-phase eutectic reaction (L-* gamma-Ni + MoNi). Since the eutectic gamma-Ni + MoNi in contact with alpha-Mo precipitates directly from the liquid phase, no habitus growth occurs between the products and the parent phase during solid-state phase transitions. The eutectic nuclei grow along lattice planes exhibiting minimal distortion with alpha-Mo to minimize interface energy with adjacent phases. After adding Permalloy, the lattice misfit at the alpha-Mo/MoNi and alpha-Mo/gamma-Ni interfaces is reduced to 11.2 % and 6.3 %, respectively, significantly lower than the excessive misfits of 22.1 % and 30.6 % observed at the original alpha-Mo/eutectoid alpha-Fe + mu (Fe3Mo2) interfaces without Permalloy addition. The transformation of the primary phase interface within the weak zone from incoherent to nearly coherent significantly enhances the mechanical properties of the welded joint, resulting in a tensile strength of 312 MPa, a 23 % increase compared to joints welded without Permalloy addition.

Electron beam welding (EBW) of Ti2AlNb and TA15 was experimented with and without an Al interlayer to analyze the effect of Al on the evolution of microstructure in the fusion zone (FZ) and its impact on joint performance. When welded without Al layer, significant softening occurred in the FZ with a microhardness of 250 HV, which was much lower than that of two base metals (BMs). The FZ was mainly composed of B2 phase which was softer than that of the O and alpha 2 phase in Ti2AlNb BM as well as the alpha+(3 phase in TA15 BM. Furthermore, the weld was dominated by coarse columnar grains, which further resulted in the softening of the weld. The tensile strength of the joint at room temperature reached 914 MPa. Al element was introduced in subsequent experiments to improve the softening problem. Adding 50 mu m thick Al layer could not completely improve the softening phenomenon, but the softening problem was solved with the 100 mu m thick Al layer. When the thickness of Al layer was 50 mu m, the main component phase of FZ changed to alpha ' martensite which was harder than B2 phase, and the microhardness increased by 30 HV. The tensile strength of the joint reached 946 MPa. The formation of the alpha 2 (Ti3Al) phase in FZ during welding with 100 mu m Al layer significantly enhanced its microhardness to 380 HV. And there was no obvious change in tensile strength when the softening problem was solved. All joints were fractured in the FZ. As the thickness of Al layer increased, the fracture mechanism gradually changed from quasi-dissociative fracture to brittle fracture.

During electron beam welding (EBW) of Ti2AlNb/TA15, the fusion zone (FZ) was mainly composed of the softer B2 phase, which led to the softening of the joint. In this study, beam offset TA15 alloy was utilized to analyze its influence on the evolution of microstructure in FZ and joint performance. Results showed that the main component phase of FZ was transformed into martensitic phase by the beam offset process, the microhardness of the FZ was increased about by 110 HV and the softening problem was solved. When the beam offset distance reached 0.4 mm, the tensile strength of the joint at room-emperature was also increased by about 55 MPa, and the fracture position changed from FZ to heat-affected zone (HAZ) on Ti2AlNb side. In addition, the plasticity of joint was slightly improved.

In this study, the microstructure and mechanical properties of electron beam welded dissimilar joint of Nb to 304 austenitic stainless steel were analyzed. The results showed that the joint obtained were asymmetric due to the large difference in physical properties between the Nb and the 304 austenitic stainless steel. The weld mainly consisted of eutectic structure and brittle intermetallic compounds reaction layer. The eutectic phase was composed of a-Fe ferrite and 8(Fe2Nb) intermetallic compound. The brittle phases in the reaction layer were 8(Fe2Nb) and mu (Fe7Nb6). The microhardness of the reaction layer of the joint was up to 1065 HV, and the presence of 8(Fe2Nb) in the eutectic region contributed to its high microhardness as well. The tensile strength of the welded joint was 137 MPa, which was 53 % of that of the Nb base metal. The joint failed at the reaction layer by brittle cleavage fracture.

Strong inference from light, smoke and spatters often increases errors of seam tracking based on structured light vision in robotic welding. Deep learning methods including the semantic segmentation networks make progress in extracting laser strips from the interference accurately. However, previous studies did not consider the match between the semantic complexity and network structure. Besides, the tracking accuracy was generally focused while the computational costs were neglected. A novel local-add U-net (LAU-net) is developed in this work based on the original U-net to achieve both low tracking errors and low computational costs. The backbone of the U-net was first adjusted to match the semantic complexity of laser stripe images. Secondly, the weighted-loss function was adopted to solve the imbalance between positive and negative pixels. The weight factor was optimized to avoid degression of network performance on structured light datasets. Finally, a new local-add connect was proposed to make full use of local information in feature maps. The segmentation accuracy was significantly improved by the local-add connect at the price of only a 1.1 ms increase in computational latency. The seam tracking system based on the LAU-net possessed good adaptability and robustness. The average tracking error was controlled within 1.48 pixels and the tracking speed reached 45 fps, which was competitive compared with other traditional or deep-learning seam tracking methods.

A validated numerical simulation model coupling multiple phases and physics is developed. The role of horizontal gaps in the laser-arc hybrid welded bottom-locking joints is focused. A new adjustment method for the surface tension momentum source is proposed to suppress the spontaneous closure of internal gaps. The weld profile, metal flows, keyhole dynamics and heat flux distribution under different gap sizes are compared. The results indicated that horizontal gaps deteriorated the weld formation. The weld depth was lower than 3 mm when the gap size reached 0.8 mm, which failed to meet the design requirements. In addition, horizontal gaps induced a split-flow once the molten metal penetrated the upper butt part. The split-flow reduced the density of upward back flows on the gap side while hardly affected that on the other side. Consequently, more keyhole wall humps formed under the unbalanced upward flows compared with gap-free conditions. Moreover, more severe keyhole fluctuations hindered the laser beam from reaching the keyhole bottom. The average keyhole depth was thus reduced with the increase of gap sizes. Correspondingly, the heat transfer was also affected. Higher heat flux was obtained when no gap existed while more heat energy was located at the keyhole wall under gap conditions, which was responsible for the reduction of the weld depth. This study clarifies weld pool dynamics of bottomlocking joints and will provide more modeling guidance on welding CFD analysis involving gap conditions.

Electron beam welding with varying beam offsets was employed to investigate the evolution of microstructure and mechanical properties in Nb/304 stainless steel joints. The weld zone microstructure primarily consisted of a eutectic matrix of 8(Fe2Nb) and alpha-Fe. As the electron beam was deflected from the Nb side to the 304 stainless steel side, the amount of molten Nb decreased, reducing the formation of brittle intermetallic compounds. The proeutectic phase in the weld zone transitioned from 8(Fe2Nb) to alpha-Fe, and the thickness of the 8(Fe2Nb) and mu(Fe7Nb6) intermetallic compounds reaction layer on the Nb side decreased by 48 %. With the beam offset toward the 304 stainless steel side, the microhardness of the weld zone decreased by 22 %, while the tensile strength of the joint increased by 51 %. The reaction layer exhibited the highest hardness, reaching 1147 HV. The joint with +0.6 mm beam offset achieved the maximum tensile strength of 156 MPa, 61 % of Nb base metal. All joints failed in a brittle manner at the reaction layer, characterized by cleavage fracture. The mechanism underlying joint weakening was revealed.

This paper proposes a new method: the pre-melting electron beam freeform fabrication method. This new method aims to mitigate the excessive heat input associated with directly forming a molten pool on the deposit and the poor formation caused by severe agitation of the molten pool. Consequently, it seeks to refine the internal microstructure of the part and enhance its mechanical properties. Using TC4 alloy as the research subject, this study performs a comparative analysis between the EBF and PEBF methods in terms of structural composition, deposit formation, cross-sectional morphology, microstructure, tensile strength, fracture path, and grain size. The research findings indicate that the PEBF method, due to its significantly lower heat input, results in the maximum alpha phase width being reduced to 1/4 to 1/5 of that observed with the EBF method. This reduction in alpha phase width leads to an increase in the tensile strength of the deposit from 784 to 840 MPa and an enhancement in the fracture strain from 0.22 to 0.34, which is close to the strength difference (68 MPa) calculated for the two methods using the Hall-Petch equation. The comparative results underscore the superiority of the PEBF method, demonstrating its potential to improve mechanical properties.