In femtosecond laser precision manufacturing, unfavorable factors such as nonperpendicular laser incidence, noncircular focused laser spots, and asymmetric energy distribution in the focal spot severely affect the accuracy and morphology of laser-induced microstructures. Typically, significant efforts are made to avoid these adverse effects. However, by leveraging these unfavorable factors, asymmetric microgrooves can be easily fabricated on solid surfaces through femtosecond laser direct writing. On the overheated grooved surface, Leidenfrost droplets can be unidirectionally propelled due to the rectification effect of the vapor flow beneath the Leidenfrost droplets by the laser-induced asymmetric microgrooves. The maximum driving velocity of the droplets can reach ~400 mm·s ^−1 , and the droplets can even climb an 18° inclined slope against gravity. This Leidenfrost droplet propulsion method is effective for a wide range of volatile liquids (e.g . , water, ethanediol, ethanol, cyclohexane, n-hexane, and acetone) and is applicable to various substrate materials (e.g . , aluminum, stainless steel, Ti alloy, Mg alloy, silicon, and ceramics). In particular, Leidenfrost droplets moving along curved or arced trajectories have been realized on laser-designed double-layered crisscrossing microgroove arrays. This achievement provides a new approach for designing and flexibly controlling the motion paths of Leidenfrost droplets, whereas previously reported strategies are typically limited to driving droplets in straight lines. The current simple and effective approach for Leidenfrost droplet transport has significant application potential in fundamental physics, energy conversion, mechanical engineering, and chemical engineering.
Basin insulators are critical components in power systems, and internal bubble defects can degrade insulation performance and threaten operational safety. Accurate detection is therefore essential. However, traditional contact-based methods are costly and complex. Although infrared thermal imaging provides a non-contact solution, it still faces challenges in segmenting bubble defects, including blurred boundaries, missed small targets, and noise interference. To address these issues, this paper proposes a Residual-Attention U-Net (URA-Net) for precise segmentation of bubble defects. Specifically, a Res-SE module is designed to suppress noise, enhance key features, and alleviate gradient degradation while preserving structural information. In addition, a convolutional block attention module is introduced to further enhance defect-related features while retaining weak signals. An infrared thermal imaging system has been developed to construct a self-collected dataset for training and evaluation. Furthermore, quantitative defect size estimation is achieved by extracting contours and calculating diameters from segmentation results. Experimental results show that the proposed method achieves an MIoU of 92.53%, a recall of 96.20%, a precision of 96.45%, and an F1-score of 96.28%. The method can effectively detect defects with diameters from 0.7 mm to 5.5 mm and depths from 0.5 mm to 2.0 mm, demonstrating accurate and robust performance.
The photocatalytic removal of antibiotic pollutants has been widely investigated for water purification, where continued pollutant decay after light irradiation (“dark-stage removal”) represents an intriguing post-illumination phenomenon in photocatalytic systems. Although charge storage provides an important explanation for post-illumination activity, adsorption relaxation after light-off may also contribute to dark-stage removal. In this work, a Ni(OH)2-modified Bi2WO6 system was employed as a representative model to investigate non-equilibrium adsorption in photocatalytic dark-stage tetracycline removal. The photocatalytic performance under visible light was first evaluated to establish a high-activity reaction system. Subsequently, a combination of acid extraction, solution–surface distribution analysis, and charge probing experiments was used to clarify the contributions of adsorption redistribution and charge storage. The results demonstrate that charge storage is negligible and unlikely to dominate the dark-stage behavior in the Bi2WO6/Ni(OH)2 system. Instead, the apparent dark-stage removal is primarily governed by a non-equilibrium adsorption process, arising from the imbalance between rapid photocatalytic consumption and adsorption–desorption dynamics during illumination. After light-off, the system gradually relaxes toward equilibrium, leading to continued redistribution of tetracycline from the solution to the catalyst surface. This work reveals a photocatalysis-induced non-equilibrium adsorption pathway for dark-stage removal, providing a complementary perspective for understanding post-illumination behavior in photocatalytic systems.
High-temperature co-electrolysis of H2O and CO2 in solid oxide electrolysis cells (SOECs) offers a promising route for syngas production that integrates carbon utilization with renewable energy storage. However, the interplay between the reverse water-gas shift (RWGS) reaction and the electrochemical reductions of H2O and CO2 leads to complex competitive dynamics that remain insufficiently understood. In this study, we develop multi-physics models at two scales: a two-dimensional button-cell model and a three-dimensional commercial-scale planar SOEC model with crossflow channels to investigate the reaction mechanisms and competitive behavior under different operating conditions. The models are validated against experimental polarization curves and show good agreement. The results reveal that the RWGS reaction occurs primarily in the thick support layer and is largely unaffected by polarization due to its spatial separation from the electrochemical reaction zone. By comparing gas composition variations under open-circuit and polarized conditions, the respective contributions of the RWGS reaction and CO2 electrolysis to CO production are estimated, and their competitive relationship under different inlet gas compositions is clarified. The two-dimensional model with an asymmetric electrode configuration reveal that in regions not covered by the oxygen electrode, the local potential distribution leads to simultaneous CO2 electrolysis and H2 oxidation. Under polarization, H2O electrolysis dominates in the active layer, with its competitive advantage varying with the inlet gas composition. The three-dimensional model further reveals a sawtooth-shaped distribution of reaction rates resulting from channel overlap in the crossflow configuration. These findings provide theoretical insights into the reaction competition mechanisms in SOECs and offer practical guidance for tuning syngas composition by adjusting operating conditions.
In dynamic multi-objective optimization, environmental changes can rapidly invalidate converged solution sets. Therefore, achieving fast convergence recovery while maintaining solution-set distribution stability remains a key challenge. Existing methods often lack effective coordination between change response and subsequent optimization, which limits their recovery speed and long-term tracking stability. To address this issue, this paper proposes a dynamic multi-objective evolutionary algorithm with direction-guided hybrid response and sensitive-dimension-based two-stage evolution. First, a direction-guided hybrid response strategy is developed to improve post-change recovery. It estimates the population migration direction from the elite centroid and the population centroid, and then generates a response population through directional translation, change-intensity-related perturbation, and small-scale random reinitialization. The response population then provides the decision-space basis for sensitive-dimension identification. The identified dimensions further provide search guidance for the subsequent two-stage evolution. Second, a sensitive-dimension identification mechanism is constructed to locate decision variables that are more sensitive to environmental changes. Based on the dispersion characteristics of elite solutions in the decision space, the identified sensitive dimensions are used to guide a two-stage evolutionary process. Predictive search is performed on sensitive dimensions to accelerate Pareto-front tracking, while local refinement is conducted on relatively stable dimensions to improve solution-set distribution. Finally, a multi-source fusion elite recombination strategy is introduced to enhance the robustness of population updating. It removes duplicates and performs controlled replacement among the original population, two-stage candidate solutions, and elite information. Experimental results on FDA, DF, and JY benchmark problems under different change-frequency settings show that the proposed algorithm outperforms the competing algorithms on most test problems and remains competitive on the others. Further studies on aluminum electrolysis operating-parameter optimization and the dynamic multi-objective knapsack problem demonstrate its practical applicability in both continuous- and discrete-variable decision scenarios.
