Post-weld shift (PWS) significantly degrades the coupling efficiency and consistency of butterfly semiconductor lasers during automated packaging. This study proposes a novel pre-compensation method based on a Particle Swarm Optimization–Backpropagation (PSO-BP) neural network to mitigate these effects. Finite element simulations and experimental analyses reveal that PWS consistently presented negative Y-direction displacement, driven by localized thermal contraction and residual stress. Results indicate that PWS magnitude is highly sensitive to laser energy and welding joint layout. By integrating PSO, the neural network achieved global convergence in just 25 iterations—15 fewer than the conventional BP model—with superior prediction accuracy. Implementation of this pre-compensation framework into an automated packaging system reduced the welding offset range from 0.5–4 μm to 0.5–2 μm, yielding a 26% average increase in optical coupling efficiency. This research provided a high-performance predictive framework suitable for the high-precision mass production of optoelectronic devices.

Composite combined shell structures, represented by underwater equipment, are inevitably subjected to initial stress fields and complex external loads. To overcome the current limitations in vibration suppression research for such structures, this study establishes a vibration suppression framework incorporating dynamic vibration absorbers (DVAs). Based on the principles of model equivalence and model simplification, the framework integrates the first-order shear deformation theory (FSDT) with the spectral-Jacobi method (SJM) to develop a dynamic model of combined shells equipped with DVAs under initial stress fields. The dynamic behavior of the system is accurately predicted through the energy variational principle. The core of this framework lies in utilizing the energy coupling and resonance separation mechanisms of DVAs to effectively suppress the vibration energy of combined shells. By comparison with benchmark solutions such as literature data and finite element method (FEM) results, the maximum absolute error of the SJM is 3.3282%, which verifies the accuracy of the proposed method while addressing the issues of repetitive FEM modeling procedures and high computational cost. The influence of key DVA parameters on the dynamic behavior of combined shells is further analyzed from the perspective of vibration energy. Combined with the results of the convergence analysis and the numerical validation, it can be concluded that the proposed research framework provides an accurate and effective prediction for the vibration absorption behavior of combined shells under an initial stress field.

Uncertainty analysis is critical for the robust vibro-acoustic design of thin-walled plate structures. Nevertheless, conventional methods are limited by computational efficiency bottlenecks in tackling uncertainty in the vibro-acoustic behaviours of cracked laminated plates. To address this issue, an integrated spectral element-multiple response Gaussian process (MRGP) approach is proposed for the uncertainty-based vibro-acoustic analysis of cracked laminated plates. A reliable deterministic model is established using the spectral element method, and an MRGP surrogate model is developed to implement analytical uncertainty propagation and variance-based global sensitivity analysis, circumventing the exhaustive sampling overhead of standard Monte Carlo simulation (MCS). Numerical analysis demonstrates that, compared with MCS, the proposed method substantially improves computational efficiency by drastically reducing the number of required model evaluations, while maintaining high computational accuracy. Uncertainty parameters perturb vibro-acoustic responses, primarily manifested as resonant frequency drifts and amplitude fluctuations, which jointly induce variations in the bandwidth of uncertainty responses. Global sensitivity analysis covering material, geometric and loading variables further quantifies the effect of uncertainty levels on the relative importance of each parameter. Overall, this study provides a highly efficient and novel analytical framework for the robust vibro-acoustic design of composite structures with multi-source uncertainties.

The optical loss in diamond infrared windows arises from the high refractive index of diamond and Fresnel reflection at the air-diamond interface. Drawing inspiration from the anti-reflective properties of moth eyes, this study investigates the anti-reflective subwavelength structures (ASS) on the diamond surface through numerical simulations and experiments, aiming to achieve broadband high transmittance in the long-wave infrared range. The geometrical parameters, including period, depth, and filling factor, of the ASS are optimized using the finite difference time domain method, effective medium theory, and thin-film theory. A three-dimensional femtosecond laser back-processing technique, exploiting double-pulse sequence Bessel beam, is employed to examine the effects of various laser and process parameters on the morphologies, enabling the fabrication of ASS square frustum arrays. The resulting structures achieve a maximum transmittance of 96.4% at a wavelength of 11.1 mu m, demonstrating significant potential for application in extreme environments.

Few-shot defect detection holds significant importance for adapting to complex industrial environments and enhancing detection accuracy. Addressing the issue where existing few-shot defect detection methods are prone to compromised feature representation under varying defect scales, this study proposes a distance-guided prototype network (DGPN) with explicit meta learning. Building upon the prototype network, our method leverages distance information between normal and anomalous image features to guide feature transformation and fusion processes. An attention gating block (AGB) is introduced to convert distance representation vectors into feature maps while enhancing the capture capability for fine-grained targets. A multiscale feature fusion module (MSFF) is proposed to further strengthen the network’s ability to extract multiscale features and semantic information. Additionally, an upsample fusion module (UF) is designed to fully integrate and exchange multiscale contextual information between shallow and deep layers, decoding semantic information and spatial details to improve the precision of small defect detection. Extensive experiments on Industrial-5i, Visa, and a self-built chip dataset validate the superiority of the proposed method and the practical applicability.

Silver nanoparticle paste is a promising interconnect material for high-temperature power electronics, but its direct bonding with silicon is hindered by the high chemical stability of silicon surfaces. This paper proposes a femtosecond laser method to induce micro-nanostructures on silicon surfaces for low-temperature, high-strength bonding with nano-silver paste. By adjusting the laser fluence, multi-level micro-nano structures comprising micrometer cones, nanoparticles, and periodic patterns are fabricated on silicon, with the optimal average microstructure depth reaching 7.65 μm at 3.11 J/cm2. Shear strength tests demonstrate that untreated flat silicon achieves only 3.1 MPa, while structured silicon reaches a peak strength of 30.5 MPa at 3.11 J/cm2, representing a tenfold increase. The transition in failure mode from interfacial adhesion to substrate internal failure confirms mechanical anchoring, and cross-sectional analysis reveals complete filling of the microstructures by the silver paste, forming three-dimensional interlocking. Thermal cycling and high-temperature aging yield average shear strengths of 15.1 MPa and 10.5 MPa, respectively. Finite element simulations further reveal that 150 °C serves as a critical transition temperature for stress relaxation in the Ag layer during cooling. This work provides an effective approach for reliable power device packaging.

As a key low-expansion material for high-end equipment such as aerospace and precision instruments, the surface quality of Invar alloy directly determines the operational performance of devices. To fill the research gap in the multi-parameter synergy and mechanism of Invar alloy laser polishing, this study performs polishing experiments on Invar alloy using a burst-mode femtosecond laser, with a repetition rate of 1 MHz and four sub-pulses per burst. The results indicate that energy density plays a dominant role in the polishing effect: with the increase in energy density, the surface roughness first decreases and then increases. A stable molten pool is formed under medium energy density (0.47-0.64 J/cm2), and under the optimal parameter conditions, the surface roughness is reduced to 394 ± 50 nm, representing a 52% reduction compared to the original surface (821 nm). Scanning speed and scanning pitch affect the polishing effect by synergistically regulating energy input: increasing scanning speed under high energy density can inhibit the rise in roughness, while a small scanning pitch can lower the threshold of optimal energy density. Amplitude spectrum analysis reveals that the medium-scale surface undulations are significantly improved after polishing. A four-layer Fully Connected Neural Network (FCNN) model is established to achieve high-precision prediction of polishing effects with a coefficient of determination R2 = 0.92, which enables rapid prediction of unknown polishing parameter combinations and provides a new solution path for the optimization of polishing effects. This study clarifies the interaction mechanism between a burst-mode laser and Invar alloy, proposes an efficient ultra-precision polishing method for Invar alloy, and lays a theoretical foundation for its application in the field of high-end manufacturing.

The high output power and excellent heat dissipation of fiber lasers have always been the goals pursued in laser applications. In this paper, a novel structure for coupling a ten channel laser into a multimode fiber through beam shaping is proposed and named full-planar fiber coupled diode laser (FPFCDL). First, the size of the fiber coupled diode laser is reduced by optimizing the angle of the mirrors in the optical path to merge the spatial beams of the laser diodes to form an optical step. Then, a laser array arrangement with an all-planar structure is used to realize the heat dissipation uniformity of the laser diode on the heat sink surface. The experimental results show that the output power of the FPFCDL is greater than 172W and the fiber coupling efficiency reaches 86.44%. When the water flow rate and heat power are 3.09 L/min and 20 W}, the temperature difference of the full-planar base is only 1.4 degrees C. It is also found that increasing the thermal conductivity of the full-planar base and heat sink liquid can effectively improve the heat dissipation performance of FPFCDL. The FPFCDL structure we designed is easy to implement, shows excellent heat dissipation performance, and has a promising future in the application field of high-power fiber laser systems.

Dual-band antireflection (DBAR) windows based on surface microstructures offer a promising solution for mid-wave infrared (MWIR) and long-wave infrared (LWIR) co-aperture composite imaging. However, micro-nano manufacturing technology faces significant challenges in efficiently producing highly uniform microstructures with characteristic dimensions of ∼1 μm across hundreds of millimeters. Here, we report a laser optical field modulation (LOFM) technology for the rapid manufacture of ultra-large-scale arrays of antireflection microholes (ARMHs) on large-aperture and non-perfectly planar windows. LOFM technology, which modulates laser pulses in both temporal and spatial domains, enhances ARMH aspect ratios from 0.1 to 0.8 without reducing manufacturing time, and maintains processing accuracy even with laser focus shifts, thereby addressing inconsistencies in large-area processing. As a proof of concept, approximately 7 billion ARMHs are fabricated on a 100-mm-diameter zinc sulfide (ZnS) window at a rate of 20 000 holes per second using LOFM technology assisted by machine learning. The fabricated DBAR ZnS window exhibits ultra-broadband (3.5–14 μm), high transmittance (91.1%), wide-angle transmission, wear-resistant, and self-cleaning, making it suitable for environments with multiple interference factors. Dual-band imaging applications demonstrate the significant advantages of DBAR windows in target recognition, multi-scenario robustness, and information acquisition.

As a chemical widely used in industrial production and daily life, alcohol poses significant safety hazards due to its flammable and explosive properties, thus the development of reliable alcohol detection technologies is crucial. However, existing detection methods typically rely on external energy sources or involve complex manufacturing processes, which severely limit their popularization and application in on-site scenarios. Herein, inspired by the special wettability of Thalia dealbata fraser (TDF) leaf surfaces, we fabricated a stainless steel (SS) surface with both ultra-robust superhydrophobicity and superalcoholphilic via a strategy combining femtosecond laser ablation and vacuum annealing (VA-LT-SS), enabling efficient alcohol detection. Specifically, alcohol of different concentrations exhibits distinct spreading behaviors on this surface. Through structural design, the surface can efficiently identify alcohol concentrations and detect alcohol molecules in gas without external energy input. This surface maintains excellent hydrophobic performance even after being exposed to the atmosphere for 180 days, tape peeling 500 times, and water erosion for 12 h. The special superwettable material and its simple preparation method developed in this study hold broad application prospects in the field of energy-free alcohol detection.

Laser remote damage technology has garnered significant attention due to its advantages in precision strike and rapid response. However, high resistance protective materials such as alumina ceramics pose severe challenges to efficient laser damage due to their high ablation threshold and low laser absorptivity. In this study, a 500 mm long-focal-length lens was employed to investigate the long distance and large range damage of ceramic materials through the coupling of femtosecond laser filamentation and continuous wave (CW) laser. The research demonstrates that femtosecond filamentation assisted CW laser enables reliable ablation of transparent hard and brittle materials, achieving full penetration of a 3 mm ceramic sample within 15 ms, showcasing remarkable efficiency and reliability. Using a high-speed dynamic monitoring system, the enhancement mechanism of plasma shockwaves induced by femtosecond filaments on CW laser energy deposition was successfully captured. Furthermore, by equivalently incorporating plasma shockwaves into the ablation model, a numerical simulation model for femtosecond-CW combined laser interaction was established, revealing the variation patterns of shockwaves in melt pool ejection behavior. These shockwaves effectively overcome energy shielding effects and serve as the dominant factor in improving ablation rate, achieving an ablation rate of 1.37 × 107 μm3/J under 700 W CW laser irradiation. This study provides important theoretical and experimental support for the practical application of long distance, large range laser damage technology.

Objective The micro-hemispherical resonator (MHR) acts as the core sensing component in compact, high-precision inertial navigation systems. However, material and fabrication defects create a discrepancy between the two eigenfrequencies of the MHR operating mode, defined as frequency split. The frequency split of the MHR limits performance improvement, making tuning processes essential for reducing its adverse effects. Femtosecond laser machining serves as an effective mass-removal tuning method owing to its short pulse duration and ultrahigh peak power. Nevertheless, microcrack formation and high surface roughness induced during material removal degrade the quality factor (Q) of the resonator. Existing studies in this field mainly focus on optimizing individual performance indicators, while lacking systematic analysis of the correlation between surface defects and the Q, as well as the relationship between laser parameters and damage suppression. This study aims to clarify the correlation between femtosecond laser processing and fused silica surface defects and to optimize laser tuning parameters, so as to realize the synergistic suppression of frequency split while maintaining a high Q to the maximum extent. Methods First, the dissipation mechanisms governing the Q of the MHR are clarified. This investigation focuses on the influence of surface loss, thermoelastic damping, anchor loss, and air damping loss on the Q through theoretical analysis. A correlation model connecting the damaged layer and surface-related Q is established by combining formula derivation and parameter calculation, confirming that surface loss serves as the dominant factor affecting overall Q performance. Second, a tuning system is built. This system consists of a femtosecond laser (Pharos, Light Conversion Ltd., central wavelength of 1030 nm and pulse width of 216 fs), a laser Doppler vibrometer (Sunnyinnova, LV-S01), a vacuum chamber with a transparent window (below 10(-2) Pa), a motorized motion stage, and auxiliary optical components. Using fused silica as the experimental substrate, a series of elaborate experiments are designed to explore the effects of laser power (10?25 mW), scanning speed (10?500 mu m/s), and scanning interval on surface roughness and material removal efficiency. Surface topography is characterized via an optical profilometer (WYKO NT9100), and quantitative calculation of mass removal is implemented accordingly. Four optimized processing schemes are selected after preliminary screening. Third, comparative damage experiments are carried out on practical MHR devices. The Q is calculated using the formula Q=pi f tau, where f refers to the mode frequency and tau represents the vibration decay time constant; tau is extracted from measured decay curves. Appropriate tuning parameters are determined based on the decay time constant tau. Finally, frequency split tuning is implemented along four low-frequency axis orientations of the MHR. Gradual expansion and deepening of laser-ablated removal grooves effectively reduce the inherent frequency split. Results and Discussions Theoretical analysis indicates that the thickness of surface damage exerts exponential influence on surface loss. This finding verifies that the suppression of surface roughness is critical for reducing Q degradation. Processing experiments show that surface roughness decreases first and then increases with rising scanning speed. At low scanning speeds, cumulative laser energy deposition increases surface roughness, while insufficient overlapping coverage at high scanning speeds also leads to roughness deterioration. In addition, higher laser power corresponds to a faster scanning speed for minimum roughness, and the optimal scanning interval reaches approximately 60% of the single ablation line width. After two rounds of screening experiments focusing on removal volume and Q variation, the optimized tuning parameters are determined as a laser power of 20 mW and a scanning speed of 100 mu m/s. Subsequent tuning verification experiments confirm that the frequency split of the MHR decreases from 8.21 Hz to 0.096 Hz, while the Q slightly declines from 444850 to 441050 with an ultra-low attenuation of only 0.85%. The results demonstrate that the proposed tuning strategy realizes high-precision frequency trimming with excellent Q retention capability. Conclusions This study combines theoretical analysis and experimental verification to clarify the dominant effect of surface loss on the Q of MHRs. The findings reveal the inherent correlation between the Q and laser-induced damage. Experimental results indicate that maintaining laser power and scanning speed within appropriate ranges is essential to achieve low surface roughness and minimal damage during tuning. After multiple comparative evaluations, a laser power of 20 mW and a scanning speed of 100 mu m/s are selected to balance material removal efficiency and damage suppression. The experimental results demonstrate that the proposed tuning method effectively mitigates Q degradation while reducing the frequency split of the MHR. This work provides a critical theoretical foundation and experimental support for the high-precision frequency tuning of micro-hemispherical resonators.

As a key material for next-generation aero-engine components, the surface layer quality of a GH4169 superalloy after machining directly determines the wear performance of critical parts. This study focused on the tribological behaviour of GH4169 superalloy workpiece prepared by longitudinal torsional ultrasonic vibration-assisted milling (LTUVAM) over a wide temperature range. First, the surface topography and microstructural characteristics of milled specimens of conventional milling (CM) and LTUVAM were characterized. Subsequently, the effects of ultrasonic amplitude, sliding load and test temperature on the characterization parameters of wear resistance were summarized. The wear behaviours of the dual friction system, including a milling workpiece and friction ball, at room/high temperatures were investigated. Additionally, based on the microstructural characteristics of the milled specimen, the sliding wear mechanism of the dual friction system at room/high temperature was elucidated, along with the mechanism for enhancing surface layer wear resistance. Research findings indicate that high-temperature conditions promote thermal and osmotic transfer between the specimen and friction balls, increasing adhesive, oxidative wear and element transfer during sliding. The main forms of wear on CM and LTUVAM specimens at room/high temperatures are abrasive, adhesive, delamination, oxidative wear and element transfer. The LTUVAM specimens exhibited higher wear resistance compared to those prepared by CM, owing to the inhibition of inter-surface friction node growth (by the uniform ultrasonic vibration texture) and sub-surface crack nucleation and extension (by the refined surface layer microstructure). Results provide theoretical support for improving the wear resistance of nickel-based superalloy components under extreme conditions.

High-entropy alloy nanoparticles (HEA-NPs) have garnered significant interest across diverse fields. However, thus far, research on their applications has predominantly focused on electrocatalysis. Expanding the applications of HEA-NPs beyond current fields is timely and desirable but remains a challenge. Here, we demonstrate the successful femtosecond laser synthesis of HEA-NPs on the laser-induced graphene (LIG) for realizing high-performance Joule heating applications. This prepared composites (HEAs/LIG) exhibits exceptional electrothermal conversion ability with efficiency up to ~285.4 °C cm2 W-1. Furthermore, the HEAs/LIG also shows high broadband infrared emissivity of ~0.98 across the wavelength range from 2.5 to 20 μm. Finally, we present the applications of HEAs/LIG as an efficient Joule heater, which consumes ~49.1% less energy compared to conventional electrical heaters in winter. This work expands the application of HEA-NPs into the Joule heating field, and underlining the importance of further development in efficient energy utilization technology.

With the large-scale deployment of supercomputing and smart computing centers, high-speed optical devices have become key nodes governing high-performance communications in these systems. The packaging quality of micro-optical components in optical devices directly determines their communication performance. In this article, a new microlens packaging strategy for high-speed optical devices is proposed. First, a multilevel aggregation network (MANet) is constructed for efficient microlens detection, achieving a posture recognition accuracy of +/- 0.15 degrees and an average processing time of 1.2 s. Then, an optical-bus-controlled microlens grasping system and a ceramic gripper are developed, achieving a microlens grasping success rate of more than 95%. Finally, a packaging system is designed and implemented, and the packaged device is experimentally evaluated. Experimental results demonstrate a microlens packaging success rate of 92.64%, a power fluctuation of less than 1.8%, and an extinction ratio fluctuation of less than 0.25 dB.

Microlens recognition in industry automation is currently a challenge due to complex background, dim target, and blurred edge. In this study, a multiscale feature hierarchical convolutional attention network (MHC-Net) is proposed to realize the accurate recognition of microlens for the automatic optical detection. The MHC-Net is designed with encoder-decoder structure, which includes down-sampling, up-sampling, multiscale feature fusion block (MFB), spatial-attention neighborhood enhancement module (SNEM), depth-separable atrous asymmetric convolution module (DAAM). The MFB adaptively selects and fine fuses of both high-dimensional and low-dimensional features, which enhances the prominence of microlens. The SNEM fuses neighborhood information and spatial attention for enhancing the feature representation of semantic information. The DAAM integrates depth-separable convolution, atrous convolution, and asymmetric convolution, which can effectively extract multi-scale features to capture location information of small microlens. The MHC-Net achieves 94.94

Flexible pressure sensors with a porous architecture are highly desirable for wearable health monitoring and intelligent human-machine interaction, owing to their excellent comfort and conformability to human motion. However, conventional porous sensors often suffer from poor signal accuracy and unstable output, which limit their capability for precision sensing. To address these challenges, we designed and fabricated a flexible pressure sensor with exceptional linearity by mimicking the unique surface structure of Iron Cross Begonia (Begonia masoniana) leaves. The sensor is constructed using a readily available melamine foam as the backbone: a porous sensing scaffold is first obtained via a simple dip-coating process, and a film featuring bioinspired protrusions is fabricated by repeated replica molding. Lamination of these two components yields a stacked sensor device. Characterization demonstrates that the sensor achieves a broad pressure detection range of up to 350 kPa, with a minimum resolvable pressure of 250 Pa, and exhibits an excellent linearity of 0.999 over its entire working range (0-350 kPa). Moreover, the sensor shows stable responses under varying loading frequencies, is capable of detecting low-frequency signals, and retains its performance without notable degradation even after 5000 repeated loading-unloading cycles. In practical applications, the sensor accurately monitors flexion and extension movements of the wrist, finger, neck, and knee, capturing human motion signals with high fidelity. Furthermore, it enables information encoding and transmission through finger gestures. The proposed bioinspired structural design strategy effectively enhances the overall performance of porous pressure sensors, offering a new paradigm for the development of flexible sensing devices with promising applications in wearable health monitoring, human motion detection, and human-machine interaction.

ABSTRACT Flexible pressure sensors are pivotal for wearable electronics and intelligent robotics, yet conventional single‐layer uniform microstructures suffer from a critical tradeoff between sensitivity and sensing range. Inspired by the dome‐shaped pressure receptors of varying sizes found in crocodile skin, this paper proposes and fabricates an asymmetric, bilayer microstructure array flexible piezoresistive pressure sensor based on one‐step femtosecond laser fabrication technology. The sensor uses polydimethylsiloxane (PDMS) as a flexible substrate, with micro‐dome arrays of different sizes fabricated on both surfaces using a femtosecond laser. This ensures high fidelity and consistency in the microstructure morphology. Benefiting from the asymmetric bilayer structural design, the sensor achieves high sensitivity (64.1 kPa − 1 ) and excellent linearity (R 2 = 0.985) across an extremely wide operating range of 0.5–2000 kPa, while also demonstrating fast response and recovery times (104 ms/38 ms), excellent cyclic stability (>10 000 cycles), and resistance to vibration interference. The microstructure remained stable even after 10 000 fatigue load cycles. It enables reliable geological identification for detector vehicles and precise object‐size classification for mechanical claws, demonstrating great promise for high‐performance robotic perception applications.

The simultaneous improvement in sensitivity, response speed, and operating range of flexible pressure sensors is crucial for their applications in human-computer interaction, health monitoring, and robotic perception. Inspired by the microgroove structure of scorpion slit sensilla, a biomimetic pressure sensor based on a rigid-flexible hybrid design strategy is proposed in this study. This sensor is composed of a flexible layer of polydimethylsiloxane (PDMS) and a rigid layer of spring steel. A microgroove array template is fabricated on a zirconia ceramic substrate using a femtosecond laser, and a biomimetic microgroove array is formed on the PDMS surface via the template replication method. Silver nanoparticles are sputtered onto the surface as the conductive layer. Experimental results show that the sensor has a high sensitivity of 1.50 ± 0.04 kPa-1, a wide operating range of 0.2 kPa to 140 kPa, an average linearity (R2) as high as 0.995 within the 0.2 kPa to 20 kPa range, a response time of 45 ± 4 ms, and a recovery time of 40 ± 3 ms. By integrating the sensor into the sole of a quadruped robot and combining it with a Bluetooth wireless transmission module and machine learning algorithms, the system achieves high-precision identification of four types of geological environments.