Self-piercing riveting (SPR) is a rapidly developing, green and efficient mechanical joining technology with significant potential in aerospace applications, particularly for joining aircraft skin and skeleton structures. It effectively joins similar, dissimilar, and multi-layer sheet materials. This study investigates SPR joints formed between TA1 and 5A06 aviation alloy sheets. To ensure joint quality, a novel lightweight deep convolutional neural network (LDCNN) is proposed for real-time detection and supervision of SPR joint appearance defects. Concurrently, the static mechanical performance of joints was systematically evaluated. Results demonstrate that LDCNN achieves a real-time detection speed of up to 166 FPS which is 19.4
In the overhead welding of oil and gas pipelines, defects such as sidewall lack of fusion and interpass lack of fusion are prone to occur. To address the issue of fusion in narrow-gap overhead welding, this study employs an oscillating pulsed gas metal arc welding (P-GMAW) process to investigate the influence of oscillation amplitude on arc behavior and weld bead formation, thereby providing a theoretical basis for the design of oscillation amplitude. Experiments revealed that when the oscillation amplitude was increased to 3.2 mm, the sidewall penetration depth decreased, adversely affecting weld bead formation. Through high-speed cinematography and three-dimensional numerical simulations, the arc behavior and droplet transfer characteristics during narrow-gap P-GMAW were elucidated. Excessive oscillation amplitude causes the arc to “climb” rapidly along the sidewalls, directing arc heat to ineffective regions of the sidewalls. Meanwhile, as the oscillation amplitude changes, the current density and electromagnetic force at the extreme positions exhibit asymmetric distributions, influencing the distribution of arc heat input and subsequently affecting the weld bead morphology. This study identifies key factors influencing the design of oscillation amplitude in narrow-gap welding of pipelines, providing a basis for optimizing the process parameters of narrow-gap overhead position P-GMAW and enhancing welding quality.
This study employs in situ laser remelting technology to investigate its effects on the forming quality, microstructure, and properties of LPBF-produced CuCrZr alloy. Research demonstrates that the density of alloy increases significantly while porosity decreases after laser remelting. Surface blackening is essentially eliminated, with no significant surface protrusions observed. Grain size decreases slightly (73.9 μm in the XOY plane, 137.1 μm in the XOZ plane) and its distribution becomes more uniform. However, the short thermal cycle induced by laser remelting provides limited contribution to the precipitation of Cr-rich nanoscale phases. The laser remelting leads to a redistribution of crystallographic texture, and the proportion of low-angle grain boundaries (LAGBs) decreases markedly. Some subgrain boundaries transform from dislocation-entangled networks into relatively ordered linear dislocation structures after laser remelting, and local lattice misorientation and GND density also decrease markedly. The hardness in the XOZ plane increases notably (by approximately 25
Significant challenges exist in the dissimilar metal welding of aluminum and steel due to substantial differences in their physical properties and poor metallurgical compatibility. In this study, dissimilar metal welding between 1060 commercial pure aluminum and 304 stainless steel was achieved using two distinct filler metals, including Cu and NiMnCu2, via laser welding. The effects of these filler metals on the microstructure and mechanical properties of the welded joints were investigated. It was found that both filler metals produced welded joints free of macroscopic defects. Notably, Al-Cu phases, without Fe-Al intermetallic phases, were detected in the weld seam of Joint-Cu. The weld seam of Joint-NiMnCu2 consisted of a multicomponent face-centered cubic (FCC) solid solution, while the transition zone on the aluminum side comprised α-Al and Fe4Al13 phases. The Cu filler metal was more favorable for ductility and fracture resistance of the dissimilar metal joints, whereas the NiMnCu₂ filler metal produced a harder but more brittle weld region due to Fe₄Al₁₃ formation .
Aluminium alloys have gained widespread applications due to their excellent corrosion resistance, thermal, electrical conductivity and workability. However, achieving high-quality composite aluminium structure remains challenging. Brazing, as one of the predominant methods, exhibits minimal impact on the aluminium structure and mechanical properties. It produces smooth and uniform surfaces, while offering high production efficiency and exceptional adaptability to automated processes. Al-Si filler metals are particularly suitable for aluminium structure brazing due to their good wettability, high strength, and excellent corrosion resistance. Limitations include the poor plasticity of certain Al-Si-based filler metals (e.g., Al-Si-Cu brazing alloys) as well as the presence of coarse Si phases. Additionally, the melting temperatures of some aluminium alloys (such as 6xxx series) are close to those of conventional Al-based filler metals, potentially compromising joint performance. Standard solutions involved advanced filler metal preparation techniques and alloying strategies. This review summarises the effects of preparation methods and alloying on Al-Si-based filler metals, emphasizing the need for more multicomponent (quinary or higher) designs. Achieving such complex compositions required reliance on computational modelling and machine learning approaches.
AA2060 aluminium–lithium alloys are attractive for aerospace structures because of their high specific strength and damage tolerance; however, their friction stir welded (FSW) joints may experience mechanical loading, sliding wear, and corrosive exposure during service. Previous studies have largely examined these responses separately, leaving the integrated durability of AA2060 joints insufficiently understood. This study evaluates the mechanical, tribological, and corrosion performance of AA2060-T8E30 joints produced using a Taguchi L9 FSW matrix. Tensile testing, wear analysis, electrochemical assessment, microscopy, and CRITIC–TOPSIS with GRA validation were used to identify a balanced processing condition. The optimum conventional condition, 1200 rpm, 3 mm s−1, and 6 kN, achieved 383.9 MPa tensile strength, 7.2
To address the limitations of existing twin-wire welding technologies, a single-power twin-wire single-arc MAG welding system was independently developed. Utilizing high-speed imaging and synchronous electrical signal acquisition, droplet transfer behaviors and weld formation characteristics were systematically investigated across varying parameters. Results indicate that the electromagnetic attractive force between the parallel, co-directional current-carrying wires is the primary driver of droplet coalescence and transfer behavior. As welding current increases and control modes vary, droplet transfer transitions through four distinct modes: independent short-circuit, common-droplet short-circuit, common-droplet pulsed, and common-droplet spray transfer. Under constant voltage (CV) control, increasing the current from 150 to 350 A raised deposition rate from 1.98 to 5.2 kg/h; however, at high currents, the explosive rupture of common droplets exacerbated spatter and induced defects like humping and undercut. Conversely, the pulsed mode allows precise regulation at high currents. Under 450 A pulsed conditions, a stable one-droplet-one-pulse transfer was achieved, yielding an optimal weld-form factor of 1.7, a deposition rate of 9.1 kg/h, and a defect-free surface, making it the ideal process for balancing efficiency and quality. At 550 A (pulsed), the transfer frequency peaked at 296 Hz, with a deposition rate of 11.7 kg/h and a penetration depth of 9.2 mm, suitable for deep-penetration welding of medium-thick plates. This study provides crucial theoretical insights and process data for the industrial implementation of this highly efficient technology.
Magnetic pulse crimping (MPC) is an efficient solid-state technique for force- and form-fit joining of similar and dissimilar materials. Aluminum alloy (AA) is lightweight and corrosion-resistant, whereas Inconel offers high-temperature durability. The aluminum-Inconel joint is used for heat transfer in rocket engine components, where the inner Inconel tube withstands high temperatures, and the outer AA layer conducts heat for fuel reheat. In this manuscript, a detailed experimental and numerical study has been performed to identify the optimum joining conditions. An AA 1050 flyer tube has been joined to an Inconel 625 superalloy target rod at variable stand-off distances (SOD) of 1 mm to 3 mm and discharge energies ranging from 4 to 8 kJ using MPC. A central composite design (CCD) based on response surface methodology (RSM) was employed to quantify the effects of SOD and discharge energy on the joint strength. Characterization of the samples was performed using optical microscopy, SEM, and EDS to understand the joint interface behaviour. The numerical simulation was performed using the LSDYNA EM module to determine the magnetic field density and impact velocity of the flyer tube. The RSM analysis identified a favorable processing region, while the additional experimental investigation established 1.5 mm SOD and 6 kJ discharge energy as the optimum condition, producing a maximum pull-out load of 2904 N. The combined statistical, experimental, microstructural, and numerical results were used to establish the experimentally observed joining window for AA1050–Inconel 625 samples.
This study integrates laser wire additive manufacturing technology with numerical simulations and high-speed camera-assisted techniques to systematically investigate the influence mechanisms of process parameters on single-pass single-layer and multi-layer depositions. Numerical simulations indicate that a concentrated heat source induces a sharp rise in molten pool temperature, forming a significant thermal gradient, while residual stresses concentrate in the deposited layer. The thermal accumulation effect during multi-layer deposition leads to stress accumulation and reduced interlayer cooling rates. The deposited layers dimensions are directly proportional to laser power and wire feed speed, but inversely proportional to welding speed. Intense beam oscillation promotes molten pool flow and uniform energy distribution, facilitating grain refinement and transforming columnar crystals into equiaxed grains. Through comprehensive parameter optimization, circular oscillation with a frequency of 50 Hz and an amplitude of 1.5 mm was identified as the optimal condition for multi-layer deposition. A regression model established for molten pool dimensions based on oscillation parameters shows good agreement between experimental and predicted values. High-speed cameras of single-pass ten-layer deposition demonstrate stable interlayer and continuous liquid bridge transitions, with dense and uniform interlayer microstructures and distinct fusion lines. Hardness testing indicates that the non-oscillation specimen exhibits a maximum average hardness of 64.1 HV, while inverted triangular oscillation yields the most uniform hardness distribution with a variance of 7.4. Circular oscillation shows the highest hardness fluctuation with a variance of 15. For multi-layer specimens, hardness decreases sharply from the base material to the deposited layer, fluctuating between 65.2 HV and 119.2 HV.
This paper investigates the quality control during the wire arc additive manufacturing (WAAM) process of nickel-based alloy structural components with overhang features. Firstly, a finite element model was developed in Fluent to study the temperature field evolution and melt pool flow of the nickel alloy components with overhang features in WAAM. A periodic coefficient φ was incorporated into the double-ellipsoidal heat source considering the periodic short-circuiting characteristics of CMT process. The effects of gravity, Marangoni force, electromagnetic force, and other forces on the flow behavior of the molten pool under different torch inclination angles are studied during WAAM of nickel alloy overhang structural components. With simulation results as references, WAAM experiments were performed on 60° inclined thin-walled overhang structural components. The results demonstrate that the established numerical model can effectively predict the forming quality of nickel-based alloy overhanging structures, providing a theoretical basis for improving forming accuracy and workpiece performance.
The study analyzed how variations in welding groove geometric parameters (groove angle, root gap, root face) affect joint flatness deviations, maintaining a cross-sectional area tolerance within 2.3
This paper investigates the primary factors influencing the fatigue behavior of butt-welded joints connecting normal- and high-strength steels of varying thicknesses. The research focused on dissimilar joints composed of 20 mm S690QL and 15 mm AH36, with similar joints (20 mm S690QL and 15 mm S690QL) included for comparison. Comprehensive experimental measurements of fatigue strength, misalignment, local weld geometry, and local material hardness were carried out for both joint types featuring thickness steps. Fatigue tests and fracture surface analysis were conducted to identify failure locations for crack initiation. Fatigue strength assessments using the structural stress and effective notch stress approaches were carried out to evaluate the influence of geometry-dependent stress concentration on fatigue failure behavior. The experimental results indicate that the failure locations varied depending on the specimen type. For the dissimilar joints, the locations were identified at the weld toe, specifically at the boundary between the heat-affected zone and weld metal in S690 side, which exhibited lower microhardness values. A comparison of local stress ranges, accounting for the effects of misalignments and local weld geometries, showed higher stress ranges at the failure locations. The scatter of fatigue test results was reduced significantly when the local stress ranges were considered. Consequently, the study confirms that the interaction of local material properties and stress concentrations from global and local weld geometries influences the fatigue failure behavior of high-strength dissimilar welded joints with thickness steps.
Welding X-ray inspection is essential for ensuring joint integrity and process reliability in manufacturing, yet conventional vision-only detectors may struggle with ambiguous indications, overlapping defect candidates, and limited interpretability. This paper presents Weld-LLaVA, a visual-prompt-guided vision-language framework for welding X-ray defect recognition and decision support in visual question answering tasks. The proposed workflow integrates radiographic image enhancement, YOLOv8-assisted automatic candidate localization, colored bounding-box visual prompting, chain-of-thought (CoT)-style dialogue construction, and domain-specific fine-tuning of LLaVA-1.5-7B. On the human-annotated validation set containing 1798 images and 3465 defect instances, Weld-LLaVA achieves 87.13
Basing resistance spot weld (RSW) joint quality on nugget penetration, although conventional, may result in high strength welds being determined as non-conforming. This is especially true when welding dissimilar RSW in advanced high strength steels (AHSS) with high stacking ratios. This study examines the relationship between nugget penetration and mechanical strength in three-sheet lap joints made from AHSS materials with thicknesses ranging from 0.65 to 2.0 mm and tensile strengths varying from 280 to 2100 MPa. A hybrid design of experiments framework combining Box Behnken Design (BBD) and Latin Hypercube Sampling (LHS) was implemented across 80 experimental runs to capture both interactions driven process trends and broad process window coverage in a complex HSR dissimilar RSW system. Joints were evaluated using lap shear and cross tension testing and detailed microstructural characterization. Results demonstrated that joints without visible nugget penetration into the thin top sheet could achieve high mechanical strengths. Interrupted welding experiments confirmed that bonding between sheets with high joint strength and no nugget penetration was due to either diffusion bonding or localized brazing. SEM and EDS analysis distinguished two distinct fusion interfaces: complete fusion zones with full nugget penetration and brazed interfaces, each exhibiting unique diffusion mechanisms. The results showed that joints without visible nugget penetration into the top thin sheet could still satisfy the required tensile shear strength (TSS) and cross tension strength (CTS) levels for the investigated stack up. Interrupted welding and interfacial characterization indicated that this retained performance was associated with alternative bonding mechanisms at the thin, thick interface, including solid state diffusion bonding and localized brazing. These findings indicate that thin sheet penetration should not be treated as a standalone indicator of weld quality in the present HSR dissimilar AHSS joint and that mechanical performance together with failure mode and interfacial bonding state should be considered in weld assessment.
This study investigates the influence of manual shielded metal arc welding (SMAW, referred to as M1) and robotic flux-cored arc welding (FCAW, referred to as R1) on the fracture toughness of welded joints in API 5L X80 steel. Single-edge notched bend (SE(B)) specimens were prepared according to ASTM E1820 and tested using the crack tip opening displacement (CTOD) method to quantify fracture resistance. The coarse-grained heat-affected zone (CGHAZ) exhibited significant reductions in fracture toughness. The CTOD of the R1 process is approximately 22
Al2O3 ceramic and Fe–Mn–Al–C low–density steel were successfully joined by air brazing using an Ag–10 mol
The bonding of carbon-fiber-reinforced carbon (C/C) composites to copper (Cu) is essential for thermal management applications but hindered by thermal expansion mismatch and weak interfaces. This study demonstrates oscillatory pressure as a novel method to significantly enhance C/C–Cu joint performance. Compared with static pressure bonding, oscillatory pressure increased the shear strength nearly threefold to 30 MPa. Nanoindentation indicated softening of the Ti interlayer and hardening of the TiC reaction layer. Microstructural analysis revealed that the oscillatory pressure formed TiC layer was denser, more uniform (∼0.75 ± 0.06 μm thick, ∼1.53 × thicker than static pressure), and exhibited a strong (111) preferred orientation with higher dislocation density. The process also refined the Ti interlayer grain structure, improving its deformability. The synergistic enhancement in hardness and ductility underpinned the strength improvement. The oscillatory pressure bonding is thus an effective strategy for achieving high-performance C/C–metal joints.
This study aims to clarify the challenges associated with applying conventional fracture evaluation methods to wire arc additive manufactured (WAAM) carbon steel components. Charpy impact tests were conducted on wall components fabricated using four types of carbon steel welding wires (SG2, SG3, ER110, and ELCOR M70) over a wide temperature range. The results revealed that significant scatter in absorbed energy and fracture morphology can occur even at identical test temperatures, particularly for SG2 and ELCOR M70. Fracture surface observations and quantitative analysis of cleavage facet size distributions demonstrated that this variability is significantly influenced by inherent microstructural heterogeneity in WAAM-fabricated materials. In addition, local microstructural variation at the notch root, which is intrinsically included in the Charpy testing method, is reflected in the results. In contrast, SG3 and ER110 exhibited relatively stable fracture behavior with smaller scatter, corresponding to their more homogeneous microstructures. These findings indicate that the scatter observed in Charpy impact test results reflects underlying microstructural heterogeneity, suggesting that careful interpretation of test results is required. Furthermore, detailed fractographic analysis enables the detection of toughness heterogeneity in these materials. This provides a practical basis for interpreting fracture test results in heterogeneous additively manufactured materials.
Liquid metal embrittlement (LME) remains a significant challenge in resistance spot welding (RSW) of zinc-coated advanced high-strength steels (AHSS), particularly in dissimilar sheet combinations where thermal and metallurgical interactions are not yet well understood. Although several studies have reported that high current pre-pulses intensify LME in similar AHSS joints, the present work shows this behavior does not necessarily extend to dissimilar configurations. This study examines the influence of a short 16 kA, 20 ms pre-pulse on LME severity in third generation 980 MPa AHSS (3G-980) and Interstitial-Free (IF) welds by comparing a standard two-pulse schedule with an otherwise identical schedule incorporating the pre-pulse. The pre-pulse reduced the cracking index (CI) from 0.56 to 0.13 and shifted the weld nugget approximately 0.08 mm toward the IF sheet, increasing the distance between the 3G-980 bottom sheet surface and the fusion boundary by 64
The fatigue life of welded components can be assessed using finite element simulations and fatigue damage parameters. The underlying models are often based on idealized weld geometries, although the real-world geometry of hand-welded seams can deviate substantially from these idealized geometries, particularly in notch regions that are critical for fatigue crack initiation. Although 3D-scanning has already been used to implement the real weld geometry in numerical fatigue assessment, the influence of scan post-processing, meshing, and material modeling is often not documented in sufficient detail. As a result, it remains scientifically unclear how strongly these parameters affect the calculated simulation results. This study provides detailed and reproducible guidance for fatigue life assessment based on 3D-scanned weld geometries using commercial and widely available tools and highlights the main modeling decisions. The results show that the crack path on the weld surface can be predicted well and that damage parameters accounting for the stress gradient improve fatigue life prediction by a factor higher than 10. Furthermore, the influence of post-processing and smoothing of the specimen surface is shown to be significant as it has the potential to nearly halve the deviation when using the FDPSWT. The prediction quality of the processed 3D-scanned geometries is also compared with that of idealized geometries. With the currently available damage parameters, fatigue life prediction based on the idealized geometry is still more precise and shows less scatter than those based on 3D-scanned geometries.
