Machine Learning-Assisted Optimization of Femtosecond Laser-Induced Superhydrophobic Microstructure Processing DOI Creative Commons
Lifei Wang,

Yucheng Gu,

Xiaoqing Tian

и другие.

Photonics, Год журнала: 2025, Номер 12(6), С. 530 - 530

Опубликована: Май 23, 2025

Superhydrophobic surfaces have garnered significant attention due to their pivotal roles in various fields. Femtosecond laser technology provides a feasible means for inducing superhydrophobic microstructures on material surfaces. However, the unclear influence mechanisms of process parameters, as well high cost and time-consuming nature experiments, identifying optimal femtosecond processing parameters within space remains challenge. To address this issue, optimization framework that couples machine learning genetic algorithms was proposed successfully applied laser-induced groove structures TC4 alloy Firstly, based 64 sets experimental data, effects power, scanning speed, interval micro-groove wetting properties were discussed detail. Furthermore, by utilizing small sample dataset, employed establish prediction model contact angle, among which support vector regression demonstrated predictive accuracy. Three additional dimensional variables, i.e., number effective pulses, energy deposition rate, roughness, also added original dataset vectors extra dimensions participate guide training process. The further coupled into algorithm achieve quantitative design processing. Compared best hydrophobicity angle designed improved 5.5%. method an ideal solution accurately predicting processes, thereby accelerating development application microstructures.

Язык: Английский

Fabrication of Mechanically Robust Hydrophobic Surfaces Using Femtosecond Laser Shock Peening DOI Open Access
Chao Xu, Mengyu Jia,

Yucheng Gu

и другие.

Materials, Год журнала: 2025, Номер 18(9), С. 2154 - 2154

Опубликована: Май 7, 2025

The harsh service environment has increased the demand for hydrophobic surfaces with excellent mechanical properties; however, how to manufacture such remains a significant challenge. In this study, method fabricating properties using femtosecond laser shock peening (fs-LSP) is proposed, without need any additional processing steps. Taking CH1900A martensitic steel as an example, systematic analysis of microstructure was conducted after fs-LSP, revealing mechanisms by which fs-LSP affects surface morphology, grain structure, dislocation density, and boundary characteristics. high-density dislocations refinement induced significantly enhanced hardness introduced residual compressive stresses. Additionally, laser-induced periodic micro/nanostructures on ensured properties. effect single pulse energy number impacts also been discussed in detail. As were increased, material progressively optimized, evidenced refinement, increase geometrically necessary (GND) higher proportion high-angle boundaries (HAGBs). Such optimization not monotonous or unlimited; 75 μJ six achieved optimal effect, reaching up 8.2 GPa contact angle 135 degrees. proposed provides new strategy manufacturing properties, detailed discussion provide theoretical guidance process optimization.

Язык: Английский

Процитировано

0
Machine Learning-Assisted Optimization of Femtosecond Laser-Induced Superhydrophobic Microstructure Processing DOI Creative Commons
Lifei Wang,

Yucheng Gu,

Xiaoqing Tian

и другие.

Photonics, Год журнала: 2025, Номер 12(6), С. 530 - 530

Опубликована: Май 23, 2025

Superhydrophobic surfaces have garnered significant attention due to their pivotal roles in various fields. Femtosecond laser technology provides a feasible means for inducing superhydrophobic microstructures on material surfaces. However, the unclear influence mechanisms of process parameters, as well high cost and time-consuming nature experiments, identifying optimal femtosecond processing parameters within space remains challenge. To address this issue, optimization framework that couples machine learning genetic algorithms was proposed successfully applied laser-induced groove structures TC4 alloy Firstly, based 64 sets experimental data, effects power, scanning speed, interval micro-groove wetting properties were discussed detail. Furthermore, by utilizing small sample dataset, employed establish prediction model contact angle, among which support vector regression demonstrated predictive accuracy. Three additional dimensional variables, i.e., number effective pulses, energy deposition rate, roughness, also added original dataset vectors extra dimensions participate guide training process. The further coupled into algorithm achieve quantitative design processing. Compared best hydrophobicity angle designed improved 5.5%. method an ideal solution accurately predicting processes, thereby accelerating development application microstructures.

Язык: Английский

Процитировано

0