With the increasing demand for integrated sensing in wearable devices and intelligent systems, developing flexible sensors that can simultaneously decouple multiple physical quantities just using one single signal has become a key to simplifying system architecture and reducing costs. Addressing the aforementioned pain point, this study proposes a single-signal dual-modal flexible sensor that successfully achieves high-performance sensing and precise separation of pressure and temperature using only a single resistance signal output. Through the synergistic effect of high-performance sensitive materials and a microstructured dielectric layer, this sensor has a sensitivity of 7.130 kPa-1 within the pressure range of 0-10 kPa, and a response time of 200 ms; and it can achieve monitoring accuracy of 1 degrees C within the temperature range of 30-50 degrees C, with a response time of 13 s. To address the signal crosstalk problem, based on the difference in material response time to pressure (milliseconds) and temperature (seconds), combined machine learning classification algorithms with improved ensemble empirical mode decomposition (EEMD), the sensor can decouple dual-mode information from one single signal. This single-signal decoupling strategy does not require the design of complex multi-channel acquisition circuits or additional hardware discrimination units, which greatly simplifies the sensor structure and lower the difficulty of system integration. This study provides a new approach for developing simple and efficient multi-modal flexible sensing systems, and has important application potential in fields such as electronic skin, soft robotics, and health monitoring.

Laser Directed Energy Deposition (LDED) is widely applied in producing complex structural components and repairing parts. As the fundamental unit of processing, the LDED melt pool undergoes repeated melting, cooling, and solidification, which determine the final quality of the formed part. Consequently, monitoring the temperature field of the melt pool is a significant research focus. However, existing studies are limited to 2D imaging of the temperature field and fail to achieve 3D imaging. This lack of dimensional results in incomplete monitoring. Therefore, this paper innovatively proposes a 3D monitoring system for the LDED melt pool temperature field. First, this study designs an in-situ monitoring experimental platform based on visible and thermal imaging cameras. Subsequently, a 3D imaging algorithm is developed, comprising the following steps: 1) Utilizing a biprism and visible camera with Ball-Pivoting and mesh subdivision to achieve 3D morphology reconstruction and point cloud densification; 2) Applying a super-resolution model to enhance the resolution of the thermal imaging camera results; 3) Establishing pixel-level matching and fusion based on perspective transformation principles; 4) Assigning temperature information to the 3D morphology using a proposed image processing algorithm to realize 3D thermal field imaging. Experimental results show width error of 3.46%, height error of 3.53%, and temperature error of 0.78%. Finally, the research reveals the relationship between the melt pool temperature field and its morphology, as well as the causes of melt pool depression defects. This study provides new perspectives for defect control, dynamic monitoring of the melt pool, and process parameter optimization.

Flexible electronics advances smart wearable devices and human motion monitoring systems. with Digital Light Processing (DLP) 3D printing favored for high precision. However, existing recyclable DLP printing materials generally suffer from several limitations, including stringent recycling conditions, performance degradation after multiple recycling cycles, and insufficient functional integration. These challenges restrict their processability and practical application, preventing them from fully meeting the requirements of green and sustainable flexible electronics. In this study, a castor oil-based recyclable elastomer (CIPMT) incorporating dynamic hindered urea bonds (HUBs) is developed. Taking advantage of the reversible dissociation and recombination of HUBs in the original resin solvent at 50 degrees C, the elastomer enables a mild and efficient closed-loop recycling process. Notably, the recycled resin solution can be directly reused for subsequent DLP 3D printing without requiring additional reagents. After six recycling and reprinting cycles, the elastomer maintains excellent mechanical performance (elongation at break > 349%), good thermal stability (T-5%= 251 degrees C), and high elongation retention at break (up to 90%). Based on this recyclable CIPMT material combined with DLP 3D printing, a capacitive sensor was fabricated. After recycling and reprinting, the electrical properties of this sensor remained virtually unchanged compared to the original sensor. Furthermore, by doping with the conductive ionic liquid [DEIM][TFSI], the functionality of the CIPMT elastomer was extended to fabricate resistive sensors. The sensor exhibits segmented response characteristics, and outstanding cyclic stability (>500 cycles), enabling precise detection of diverse human movement states. This study offers novel insights into the eco-friendly, multi-cycle fabrication of DLP 3D-printed flexible sensors, providing significant reference value for advancing their applications in smart wearable devices and personalized medicine.

In laser welding, the plume and molten pool represent critical surface and internal phenomena related to gas flow dynamics and penetration states. Their significant temporal correlation and complementary information make their joint analysis indispensable for evaluating weld stability and tracking defect evolution. Accurate extraction of plume and molten pool features requires high-performance image segmentation models. However, plumes exhibit low contrast and highly diffusive dynamics, while molten pools exhibit light saturation with sharply defined edges-resulting in a significant appearance gap. Coupled with the millisecond-level inference demands in industrial settings, these factors cause traditional morphological algorithms to falter in complex scenarios, while fully supervised semantic-segmentation networks rely on expensive pixel-level annotations and still face challenges in meeting real-time industrial requirements. To address these challenges, this study presents a comprehensive paradigm, Minimize Diligent Student-Strict Teacher (MDS-MT), for real-time feature extraction of plume and molten pool. The fundamental semi-supervised semantic segmentation module Diligent Student-Strict Teacher (DS-MT) combines illumination enhancement and multi-scale feature learning: it employs adaptive illumination enhancement and a confidence-guided mechanism to improve the edge representation of the plume and molten pool, while utilizing multi-scale feature extraction and an attention module to enhance the perception of structures at disparate scales. With only 10% labeled data, the model still achieves 96.88% PA, 91.37% mPrec, 85.81% mIoU, and 92.12% mDice, effectively mitigating the segmentation performance bottleneck caused by the scarcity of labeled data. The proposed method demonstrates outstanding performance in cross-material experiments utilizing only 200 labeled images, as well as in cross-scenario steel defect monitoring with merely 15% annotation coverage. To address industrial real-time requirements, the proposed method integrates knowledge distillation and TensorRT inference acceleration, achieving 0.63 ms per frame while maintaining 96.56% PA, 91.02% mPrec, 84.88% mIoU, and 91.57% mDice. This study presents a high-accuracy, low-annotation-cost, real-time cooperative perception solution for plume-molten pool features, offering an efficient and industrially viable approach for laser welding visual monitoring and advancing theoretical research and engineering implementation of intelligent welding process perception.

Laser preheating can address the challenge of depositing soft particles onto hard substrates by cold spraying, and optimizing the laser linear energy density is critical for coating quality. This study uses a 10 kW high-power laser to preheat Q235 steel substrate and investigates the effect of linear energy density on the coating's microstructure, mechanical (bonding strength, microhardness, residual stress) and functional (electrical conductivity, electrochemical performance after 168 h immersion) properties. The results show that increasing linear energy density initially optimizes the microstructure and properties of copper coatings. When the linear energy density is 7.96 J/mm, the coating exhibits optimal comprehensive performance with a bonding strength >82 MPa, a microhardness of 129.2 HV0.1, a residual stress of -95.1 MPa, an electrical conductivity of 53.59 MS/m, a corrosion potential of -0.032 V (168 h), a corrosion current density of 4.810 & times;10(-8) A/cm(-2) (168 h), a porosity of 0.05%, and a fine average grain size of 0.89 mu m. However, exceeding this linear energy density threshold results in decreased electrical conductivity and corrosion resistance. SEM images of the corroded specimen reveal that a dense passivation film formed on the surface of the coating prepared under laser preheating after 168 h of prolonged immersion, which significantly enhances the coating's corrosion resistance. Overall, this study exhibits good reliability of high-performance pure copper coatings prepared by cold spray technology under laser preheating.

Femtosecond laser ablation (FLA) offers distinct advantages for high-performance machining, such as controllable energy input and extremely high peak power. To further enhance its capabilities for processing ultra-hard materials, burst-mode FLA has emerged as a prominent research focus, which improves machining efficiency and limits thermal defects. This paper comprehensively investigates the machining of SiC via ultraviolet FLA in MHz burst mode. Temperature fields based on the two-temperature model were simulated, which investigated the electron-lattice thermal coupling disparities under varying sub-pulse numbers. The mechanism of how thermal accumulation synergy mechanism in burst-mode enhances was revealed. It was found that low-energy and multisub-pulse output strategy of burst-mode FLA induces sustained residual thermal accumulation in lattice, thereby enhancing energy utilization efficiency. Material removal mechanisms exhibit diverse pathways with dynamic sub-pulse energies. Furthermore, this study establishes the correlation between the decline in MRR under high laser repetition rates and the shielding effect through discrete wavelet transform time-frequency analysis of acoustic emission signals. These results demonstrate that precise energy modulation of sub-pulses in burst-mode FLA is critical for the improvement of surface finish and machining efficiency, providing significant guidance for future applications achieving high-quality and high-efficiency micromachining of SiC.

In recent years, with the in-depth research on the coolants of the fourth-generation nuclear reactors, the liquid lead-bismuth eutectic (LBE) has gradually come into people’s view as a potential coolant. However, at high temperatures, liquid LBE poses a serious threat to the integrity of structural components such as fuel cladding tubes and coolant pipes in lead-cooled fast reactors (LFR). The compatibility issue between liquid LBE and structural materials, especially liquid metal corrosion (LMC) and liquid metal embrittlement (LME), has become the main factor restricting the development of LFRs. High-entropy alloys (HEA), due to their unique microstructure and excellent properties, have the potential to be used as candidate structural materials for higher-temperature LFRs. Based on this, this article first briefly analyzes the potential of HEA corrosion resistance. At present, studies on the corrosion behavior of HEAs in liquid LBE are primarily categorized into two types: HEA coatings and HEA bulk materials. Differences in their fabrication techniques and morphological characteristics result in distinct microstructural features and corrosion mechanisms in liquid LBE environments. Accordingly, this article offers a comprehensive review of compatibility studies between HEA coating and bulk materials in liquid LBE, examining the research from three key perspectives: preparation methods, HEA systems, and element influence, thereby contributing valuable insights to the field. Finally, this article outlines recommendations for future research directions in three aspects: corrosion testing, preparation, and design of HEAs, providing theoretical support and technical guidance for the application of HEAs in LFRs.

Addressing complex shaped and tumor-associated bone defects remains a significant challenge in the field of bone tissue engineering. In this study, we developed a rare earth metal compound lanthanum hexaboride (LaB6) enabled multi-functionalized 3D-printed polylactic acid (PLA) based composite bone scaffolds, aiming to combine photothermal-mediated shape memory and thermotherapy functionalities. The 3D-printed PLA/LaB6 composite scaffolds exhibit an ordered, interconnected porous structure with excellent mechanical properties. These scaffolds demonstrate efficient and controllable photothermal responses to target set-point temperatures through modulation of material composition and near-infrared (NIR) light exposure conditions. This controllable photothermal behavior not only enables effective tumor cell suppression and ablation, showing high efficacy against osteosarcoma cells, but also remotely activates the shape memory capability of PLA, providing shape adaptation. The composite scaffolds display component-dependent, tunable degradation profiles and sustained release of bioactive ions of lanthanum and boron, which enhance biomineralization due to accelerated ion exchange and deposition processes. Furthermore, these scaffolds stimulate cellular proliferation and enhance osteogenic differentiation. This work presents a promising approach for the development of multifunctional, 3D-printed polymer composite bone scaffolds based on rare earth metal particles, offering potential solutions for complex irregular and tumor-related bone defects.

Resolving the dichotomy between wide detection ranges and low mechanical hysteresis remains a critical challenge in flexible electronics, largely governed by the intrinsic viscoelastic creep of polymeric dielectrics. Drawing inspiration from the distinctive load-bearing mechanisms of traditional Chinese Sparrow Brace architecture, we report a mechanically optimized tilted micro-architecture designed to enhance structural resilience. Unlike conventional soft elastomeric pillars that easily succumb to mechanical failure, this BOPS-based tilted geometry provides excellent load-bearing capacity, effectively preventing premature failure. Finite element analysis (FEA) confirms that this tilted geometry forces a fundamental shift from conventional bulk compression to structural bending. Because this bending-dominated architecture drives rapid elastic recovery, it significantly mitigates the severe effects of the polymer’s viscoelastic creep under the tested loading conditions, achieving reliable signal reversibility with low hysteresis. We fabricated this specific architecture via programmable femtosecond laser direct writing (FsLDW) on biaxially oriented polystyrene (BOPS) films, harnessing the material’s entropy-driven self-growth kinetics. By merging this localized growth mechanism with the architectural design, we effectively bypassed the complexities of traditional molding, achieving mask-free, in situ growth of large-scale, highly uniform dielectric micro-arrays. The resulting sensor delivers a remarkably broad working range (up to ~2.28 MPa) coupled with a negligible recovery error (~1.3%), an agile dynamic response (~70/80 ms), and consistent operational durability. Ultimately, this work combines architecture-inspired structural design with advanced femtosecond laser surface microengineering, providing a conceptually novel and scalable pathway for next-generation flexible sensing.

To address the challenges posed by irregular and tumor-associated bone defects in the field of bone tissue engineering, this study develops a multi-element CuFeSe2 ceramic nanoparticle multifunctionalized 3D-printed polylactic acid (PLA) bone scaffold with photothermal-triggered shape memory and thermal therapy capabilities. The results indicate that the 3D-printed PLA/CuFeSe2 composite bone scaffold exhibits an ordered and interconnected three-dimensional porous structure with enhanced mechanical properties. Simultaneously, the composite scaffold demonstrates excellent and controllable photothermal responsiveness, enabling the regulation of steady-state temperature ranges by adjusting CuFeSe2 content and near-infrared light (NIR) irradiation conditions. This controllable photothermal response not only makes the scaffold with high-efficiency inhibition capability against HOS osteosarcoma cells (inhibition rate about 90 %) but also remotely activates the shape memory property of PLA, granting the scaffold morphological adaptability for minimally invasive implantation and defect matching.

Addressing complex bone defects associated with infection and tumors requires advanced surface and biointerface engineering of biomaterials. This study presents a surface functionalization strategy for constructing a multifunctional coating on a 3D-printed shape-memory polymer composite (PLA/PEG/MoS2) substrate. Utilizing a polydopamine (PD)-mediated deposition process, cerium oxide nanoparticles (CeO2 NPs) were uniformly immobilized onto the scaffold surface to create a tailored, functionally active coating. This coating markedly altered the surface physicochemical properties, reducing the water contact angle from approximately 81° to 48°, and enhanced interfacial biomineralization and the osteoblastic response of MC3T3-E1 cells. Moreover, the designed substrate-coating architecture integrated bulk photothermal responsiveness with surface-mediated nanoparticle activity, enabling a synergistic therapeutic mode at the biointerface. Under near-infrared (NIR) irradiation, the photothermal heat generated by the MoS2-incorporated scaffold triggered shape-memory recovery for adaptive defect fitting while simultaneously amplifying the antibacterial and antitumor efficacy of the PD/CeO2 coating. As a result, the multifunctional scaffold achieved antibacterial rates of 92% against E. coli and 94% against S. aureus, together with a 92% ablation rate of human osteosarcoma (HOS) cells. Overall, this work demonstrates a potential functional-coating strategy for engineering a multifunctional biointerface on 3D-printed bone scaffolds, integrating surface bioactivity, NIR-triggered shape-memory behavior, and synergistic photothermal/nanoparticle therapy. These findings provide a promising design for functional coatings and surface-engineered biomaterials in the treatment of complex bone defects.

The inherent layered structure of stereolithography (SLA)-fabricated alumina ceramics results in inadequate mechanical performance and notable anisotropy, severely limiting their practical application. Inspired by the intricate brick-and-mortar microstructure of nacre, we propose an intermittent laser scanning (ILS) strategy designed to control anisotropy by tailoring the interlayer microstructure. Guided by the optimal parameters identified from orthogonal experiments, the ILS strategy successfully produces a precisely controlled stepped-layer structure. A systematic investigation is conducted into the effects of the ILS strategy on printing accuracy, surface quality, pyrolysis behavior, microstructure, mechanical properties, and anisotropy. The results demonstrate that the ILS strategy not only preserves excellent surface quality but also significantly reduces curing shrinkage along the printing direction and alleviates anisotropy in sintering shrinkage. Crucially, compared with the conventional layered structure, the tailored stair-stepped structure improves the sectionmodulus and promotes crack deflection, yielding a three-point flexural strength of 338.44 MPa and a fracture toughness of 4.5 MPa & sdot;m1/2 for the sintered ceramics; the corresponding anisotropy factors are reduced to 2.1% and 1.5%, respectively. This work clarifies the crucial role of the interlayer microstructure in controlling the anisotropy of SLA ceramics, thereby providing a pathway for the fabrication of high-performance and lowanisotropic SLA ceramic parts.

The demand for modern lightweight design has increased the need for variable cross-section holes in functional components. However, conventional mechanical drilling and two-dimensional or 3D dynamic-focusing galvanometer scanning drilling, are difficult to process such holes in carbon fiber-reinforced plastic (CFRP) laminates due to tool-geometry constraints and the fixed laser angle of incidence (AOI). Herein, this study introduced a novel femtosecond laser-based method for fabricating stepped holes in CFRP laminates by utilizing a five-axis galvanometer system capable of dynamic incident-angle adjustment. A critical laser fluence distribution mechanism was demonstrated both theoretically and experimentally, showing that increasing the laser angle of incidence (AOI) from 1 degrees to 9 degrees enhances the deep-hole laser fluence by 157.84%. This significant increase directly counteracts the occlusion effect of the sidewall taper, fundamentally explaining factor that a five-axis galvanometer system is able to reduce the sidewall taper from 38.3 degrees to 1 degrees. Furthermore, the transition from thermal ablation to thermo-mechanical removal is also governed by laser fluence. In the thermo-mechanical removal process, the ejection speed of fibers and particles reaches 34.83 m/s, which subsequently affects the uniform distribution of spot energy. As the laser spot diffuses along the hole depth, the fluence decreases to 93.03% of its initial surface value at a depth of 1.5 mm, resulting in taper accumulation. These lead to a gradient spot energy distribution, which in turn induces the formation of a gradient sidewall microstructure, and this structure is further influenced by the fiber orientation. It can be addressed by increasing the AOI to improve the effective energy distribution and maintain uniform removal. These findings demonstrate that dynamic AOI control can adjust the laser fluence distribution to overcome geometric constraints and material heterogeneity, offering a promising approach for the high-quality laser machining of stepped hole in CFRP laminates.

Digital image correlation (DIC) with speckle patterns becomes unreliable under heterogeneous large-amplitude deformation regimes because correspondence estimation is highly amplitude-dependent: identical local textures may imply different displacement magnitudes, leading to scale ambiguity and heavy-tailed errors. We pro pose an amplitude-decoupled RAFT-DIC network (ADR-DICnet) that explicitly decouples deformation amplitude from correspondence estimation and uses amplitude conditioning to stabilize iterative refinement. The model predicts a per-image amplitude indicator and supports three inference modes (pred/snap/oracle), enabling an oracle-gap diagnosis that attributes residual errors to amplitude estimation versus matching difficulty. To further account for robustness across amplitude regimes, we also investigate GroupDRO training on amplitude groups as a worst-group-oriented training strategy. On Speckle Dataset 3.0, our model achieves 0.073 Mean-EPE, 0.230 P95, and 0.088 Worst-A (pixel units), outperforming representative learning-based baselines such as ECI-Net (0.159/0.329/0.220) and R3-DICnet (0.174/0.329/0.195), while substantially improving tail behavior under large deformation. Additional ablations analyze the roles of amplitude conditioning, robustness-oriented training strategies, and inference modes, and the iteration-budget study provides a practical accuracy-efficiency trade-off. Overall, the results demonstrate that explicitly modeling deformation amplitude is an effective and deployment-friendly strategy for robust speckle matching under heterogeneous cross-scale deformations within the tested 1-10 px benchmark.

Femtosecond laser ablation (FLA) has huge potential in precise machining of ultra-hard materials such as tungsten carbide (WC) and diamond-related materials due to its exceptionally high peak power and ultrashort pulse duration. However, the material removal mechanisms in FLA are complex because of the rapid laser/matter interaction, leading to the instability of machining efficiency. In this study, the transition of machining mechanisms and its effects on dynamic material removal rate (MRR) in the FLA of WC is fundamentally investigated using the acoustic emission (AE) sensing technique. Influences of laser parameters on machining efficiency is theoretically analyzed via time-domain and frequency-domain of AE signals. According to the metrics of measured area of rectified signal envelop (MARSE), root mean square (RMS), and short-time Fourier transform (STFT), the negative trend between amplitudes of MARSE/RMS and energy efficiency rate is revealed. Excessive laser fluence leads to severe plasma shielding effect, which reduces the MRR. Increasing the scanning speed enhances energy utilization efficiency. During the first four ablation cycles, MARSE/RMS amplitudes and the characteristic frequency power in the STFT spectrum remain consistent, with minimal defocusing effects. Amplitudes of MARSE/RMS and the power of 0.2 MHz component show decreasing trends, whereas the harmonics exhibit the opposite trend. AE provides an effective in-situ monitoring method for the identification of transitions in material removal mechanisms and the dynamic changes in MRR, offering valuable insights for the optimization of machining efficiency and precision in femtosecond laser machining.

Spatter is a detrimental by-product in Laser Powder Bed Fusion (LPBF) that adversely affects the quality of fabricated components. In particular, the redeposition of large spatter particles onto the powder bed can significantly impair surface finish and mechanical performance. Understanding the transport and deposition behavior of such particles is therefore essential for improving LPBF build quality. In this study, a CFD-DPM model was established and, combined with experimental analysis, employed to investigate the effects of scanning strategies and initial particle velocity, size, and ejection angle on spatter trajectories and deposition patterns. Different scanning strategies were represented by varying the relative angle between the spatter ejection direction and the shielding gas flow. Simulation and experimental results reveal that large spatters exhibit highly concentrated deposition, primarily on the part surface, particularly near the downstream side of the shielding gas flow, as well as in the powder bed regions adjacent to the part edges. Among the tested scan strategies, the S-0 scanning strategy, where the spatter ejection direction is aligned with the shielding gas flow, large spatter particles exhibit the lowest deposition amount in the vicinity of the build area. These findings provide quantitative insight into how scanning strategy and initial spatter dynamics jointly determine deposition patterns and defect formation. The results offer practical guidance for optimizing scan paths and mitigating lack-of-fusion (LOF) defects, thereby improving the quality and reliability of LPBF components.

Femtosecond laser shock peening (fs-LSP) stands out as a precision surface enhancement technique, offering superior advantages over conventional nanosecond laser shock peening (ns-LSP) in microscale processing and thermal damage mitigation. However, traditional fs-LSP treatment of titanium encounters a core trade-off: insufficient shockwave intensity at low energies versus excessive surface roughness (Ra >1.2 mu m) at high energies. Here, we propose an innovative dual-pulse coupled defocusing strategy to address the power limitations in conventional fs-LSP processing. Defocusing effectively curbs excessive heat flux, minimizing ablative damage on titanium surfaces. Experiments show optimal performance at defocus distance of 0.8-1.0 mm, yielding surface roughness below Ra 0.8 mu m and microhardness enhancement over 210 HV, comparable to conventional LSP outcome. This work systematically study the micro-morphology, microstructure, shockwave dynamics, and residual stress of the hardened layer. The fs-LSP hardening mechanism is mainly oxidation-driven, with an affected depth of similar to 50 mu m featuring a distinct martensitic transformation at the interface. Notably, the process introduce unique coloration, which closely relate with surface oxidation behavior and microstructure but minimally with laser-induced periodic surface structures. The coloration patterns mirror hardening efficacy, with green-toned surfaces exhibiting superior surface quality and hardness. Our defocusing approach enables fs-LSP to achieve both superior surface quality/hardness and anti-counterfeiting functionality, establishing a new way for titanium surface engineering.

Liquid tin (Sn) possesses excellent thermophysical properties and demonstrates significant application potential in high-temperature energy systems, nuclear engineering, and advanced manufacturing. In this study, the corrosion behaviors and microstructural mechanisms of Ni-based superalloys (IN625, IN718), austenitic stainless steels (316L, 316Ti, 310S), and titanium materials (CP Ti, Ti-6Al-4V) were systematically evaluated after static exposure to liquid Sn at 500 °C for 500 h. The results indicate that the overall corrosion resistance follows the decreasing order of Ti-6Al-4V ≈ CP Ti > 316Ti > 316L > IN625 > 310S > IN718. Titanium materials exhibit the most outstanding corrosion resistance, with the interaction strictly confined to a highly sluggish and diffusion-controlled interfacial mutual dissolution stage due to the extremely low mutual solubility between Ti and Sn. Austenitic stainless steels generally suffer from the selective leaching of the highly active Ni, accompanied by the formation of porous FeSn2. Ni-based superalloys display divergent corrosion mechanisms. IN625 shifts the corrosion process to diffusion control by forming a continuous Cr- and Mo-rich intermediate layer, whereas IN718 experiences severe bulk dissolution and morphological collapse. This study reveals the critical roles of elemental solubility, diffusion capability, and corrosion layer stability in the evolution of liquid metal corrosion, providing an important theoretical basis for structural material selection and compositional design for high-temperature liquid Sn environments.

By employing a cost-effective cold spray process in nitrogen atmosphere combined with in-situ laser assistance, Al6061 deposits are successfully fabricated on ZL114A substrates. The results indicate that with increasing laser power, the deposit’s microstructure undergoes a pronounced evolution from a severely plastically deformed structure with a high dislocation density to an equiaxed recrystallized structure with low dislocation density, and eventually to a coarsened grain structure. An optimal laser power window of approximately 5 kW was identified, within which the synergistic effects of plastic rheology and dynamic recrystallization are most prominent, resulting in a uniform, stable microstructure with low dislocation density. Consequently, the deposit exhibits the best overall wear resistance, with a wear rate only 50% that of the substrate. Moreover, the bonding strength between the deposit and the substrate shows a non-monotonic variation with increasing laser power, reaching a peak value of 79 MPa at 3 kW. The evolution of bonding strength is initially dominated by high-energy impact and thermal softening, whereas at higher powers (>4 kW), it becomes primarily governed by the complex coupling between adiabatic rheological behavior and the pre-deposition crater effect.

The helical drilling process exhibits significant potential for applications in the drilling of high-temperature alloys and carbon fiber reinforced plastic (CFRP) laminates. However, due to the limitations of the optical modulation, the current helical drilling technique is exclusively applicable for small-hole machining with diameters less than 1 mm. Given this, an eccentric helical drilling process for machining large holes in CFRP laminate has been proposed. High-quality 10 mm diameter holes with an average hole taper of less than 0.6 degrees have been successfully drilled in 6.2 mm-thick CFRP laminates. The integrated theoretical and experimental analysis revealed the formation mechanisms of hole taper and heat-affected zone (HAZ), and their correlations with microscopic morphology and fiber-orientation dependent quality under different laser deflection angles and helical rotary speeds. Results show that increasing the deflection angle to 3 degrees enhances laser fluence by 78.15 % compared to 0.5 degrees, effectively mitigating taper and HAZ caused by hole sidewall obstruction. At 7000 rpm and 3 degrees deflection, rotary speed has limited effect on taper reduction but enables rapid material removal and suppresses heat accumulation, reducing HAZ by 49.33 % at the entrance and 82.14 % at the exit. Larger deflection angles reduce sidewall defects (e.g., adhesive resin, interface cracking, fiber fractures), while higher rotary speeds improve kerf uniformity through enhanced polarization consistency and geometric independence of the laser beam. These advantages highlight the potential of the eccentric helical process for drilling large holes in CFRP laminates.