Pub Date : 2026-02-10DOI: 10.1016/j.optlastec.2026.114910
Yanming Liu , Weiwei Liu , Yali Ma , Zhi Wang , Hongchao Zhang
Nickel-based superalloys are highly sensitive to thermal history during laser directed energy deposition (L-DED), which in turn directly affects the cladding layer geometry and heat flow characteristics. In this study, a multiphysics computational fluid dynamics (CFD) model is established based on the volume of fluid (VOF) method, and a combination of simulation and experimental approaches is employed to systematically investigate the effect of substrate temperature on the cladding layer geometry and melt pool flow behavior of the IN718 superalloy. The results show that as the substrate temperature increases, the width of the cladding layer increases, the cladding height exhibits a nonlinear variation, and the peak temperature within the melt pool also increases. Furthermore, the flow dynamics at the free surface of the melt pool influence the relationship between the substrate temperature and the geometry of the cladding layer. Finally, a dimensionless parameter analysis quantitatively reveals that increased substrate temperatures intensify Marangoni convection. This study provides a theoretical basis for precisely controlling the cladding layer geometry by modulating the substrate temperature.
{"title":"Role of substrate temperature in clad geometry and flow behavior during directed energy deposition of IN718 superalloy: Multiphysics modeling and experimental validation","authors":"Yanming Liu , Weiwei Liu , Yali Ma , Zhi Wang , Hongchao Zhang","doi":"10.1016/j.optlastec.2026.114910","DOIUrl":"10.1016/j.optlastec.2026.114910","url":null,"abstract":"<div><div>Nickel-based superalloys are highly sensitive to thermal history during laser directed energy deposition (L-DED), which in turn directly affects the cladding layer geometry and heat flow characteristics. In this study, a multiphysics computational fluid dynamics (CFD) model is established based on the volume of fluid (VOF) method, and a combination of simulation and experimental approaches is employed to systematically investigate the effect of substrate temperature on the cladding layer geometry and melt pool flow behavior of the IN718 superalloy. The results show that as the substrate temperature increases, the width of the cladding layer increases, the cladding height exhibits a nonlinear variation, and the peak temperature within the melt pool also increases. Furthermore, the flow dynamics at the free surface of the melt pool influence the relationship between the substrate temperature and the geometry of the cladding layer. Finally, a dimensionless parameter analysis quantitatively reveals that increased substrate temperatures intensify Marangoni convection. This study provides a theoretical basis for precisely controlling the cladding layer geometry by modulating the substrate temperature.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"198 ","pages":"Article 114910"},"PeriodicalIF":5.0,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192643","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-10DOI: 10.1016/j.optlastec.2026.114909
Benyue Li , Daryl Tan , Kok-Sing Lim , Shufeng Sun , Fengyi Chen
A novel three-reflector fibre Bragg grating Fabry–Perot interferometer (3R-FPI) based on femtosecond (fs) laser direct writing is proposed and experimentally demonstrated for inclination angle detection. The sensor structure consists of three short Bragg gratings, including one standard core grating and two edge-core gratings inscribed at precise positions along a single-mode fiber. The combination of edge-core and core gratings enables bending-induced asymmetrical electric field distribution to be effectively converted into optical intensity variations. Two 3R-FPI configurations are fabricated with carefully designed cavity lengths to generate distinct, non-overlapping spatial frequency components in the spatial frequency domain, thereby allowing efficient multiplexing through spatial frequency-division multiplexing (sFDM). The sensor’s output is demodulated using selective local Fourier transform, which significantly reduces computational complexity by focusing only on the resonant spatial frequency components of interest. Experimental results show that the proposed 3R-FPI exhibits clear and repeatable spectral responses to inclination changes within a range of − 50° to + 50°, with linear sensitivities of 6 × 10−7a.u./° and 3 × 10−7a.u./° for Configurations I and II, respectively (R2 > 0.98). The developed system further demonstrates biaxial inclination/tilt measurement capability when two 3R-FPI sensors are integrated orthogonally in a single array. The compact design, high sensitivity, and efficient frequency-domain multiplexing make the fs-laser-inscribed 3R-FPI an excellent candidate for distributed and vector inclination sensing in structural monitoring applications.
{"title":"Spatial-frequency-division-multiplexed edge-core Bragg grating Fabry-Perot interferometer for biaxial inclination sensing","authors":"Benyue Li , Daryl Tan , Kok-Sing Lim , Shufeng Sun , Fengyi Chen","doi":"10.1016/j.optlastec.2026.114909","DOIUrl":"10.1016/j.optlastec.2026.114909","url":null,"abstract":"<div><div>A novel three-reflector fibre Bragg grating Fabry–Perot interferometer (3R-FPI) based on femtosecond (fs) laser direct writing is proposed and experimentally demonstrated for inclination angle detection. The sensor structure consists of three short Bragg gratings, including one standard core grating and two edge-core gratings inscribed at precise positions along a single-mode fiber. The combination of edge-core and core gratings enables bending-induced asymmetrical electric field distribution to be effectively converted into optical intensity variations. Two 3R-FPI configurations are fabricated with carefully designed cavity lengths to generate distinct, non-overlapping spatial frequency components in the spatial frequency domain, thereby allowing efficient multiplexing through spatial frequency-division multiplexing (sFDM). The sensor’s output is demodulated using selective local Fourier transform, which significantly reduces computational complexity by focusing only on the resonant spatial frequency components of interest. Experimental results show that the proposed 3R-FPI exhibits clear and repeatable spectral responses to inclination changes within a range of − 50° to + 50°, with linear sensitivities of 6 × 10<sup>−7</sup>a.u./° and 3 × 10<sup>−7</sup>a.u./° for Configurations I and II, respectively (R<sup>2</sup> > 0.98). The developed system further demonstrates biaxial inclination/tilt measurement capability when two 3R-FPI sensors are integrated orthogonally in a single array. The compact design, high sensitivity, and efficient frequency-domain multiplexing make the fs-laser-inscribed 3R-FPI an excellent candidate for distributed and vector inclination sensing in structural monitoring applications.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"198 ","pages":"Article 114909"},"PeriodicalIF":5.0,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192333","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-10DOI: 10.1016/j.optlastec.2026.114866
Fazal Badshah , Zia Ullah , Haiyang Zhang , Muhammad Idrees , Misbah Qurban , Haibo Huang , Yuan Zhou
The phenomenon of light dragging refers to the alteration in the propagation path of light as it traverses through a moving or dynamic medium. This subtle effect is pivotal for detecting extremely low group velocities of light, a capability with significant implications in quantum technologies such as coherent state transfer, the implementation of quantum gates, and long-lived quantum memories. In this work, we report to the best of our knowledge the first theoretical demonstration of the light-dragging effect in the context of Landau-quantized graphene. The underlying mechanism responsible for this effect originates from nonlinear chirality induced during a Raman gain process in the Landau levels of graphene. Specifically, the interaction of a magnetic dipole transition with an electric dipole transition facilitates Raman gain-assisted chirality. This interplay gives rise to strong magnetoelectric cross-coupling, allowing two Raman pathways to interfere via a shared magnetic-dipole transition. We explore the resulting modifications in both the group index and refractive index spectra, revealing pronounced signatures of magnetoelectric-induced dispersion. Our proposed framework paves the way for the design of advanced photonic devices with enhanced optical performance, potentially offering a novel route toward sub-wavelength imaging and precise resolution of nanoscale structures.
{"title":"Light dragging in Landau-quantized graphene with pump-induced magnetoelectric chirality","authors":"Fazal Badshah , Zia Ullah , Haiyang Zhang , Muhammad Idrees , Misbah Qurban , Haibo Huang , Yuan Zhou","doi":"10.1016/j.optlastec.2026.114866","DOIUrl":"10.1016/j.optlastec.2026.114866","url":null,"abstract":"<div><div>The phenomenon of light dragging refers to the alteration in the propagation path of light as it traverses through a moving or dynamic medium. This subtle effect is pivotal for detecting extremely low group velocities of light, a capability with significant implications in quantum technologies such as coherent state transfer, the implementation of quantum gates, and long-lived quantum memories. In this work, we report to the best of our knowledge the first theoretical demonstration of the light-dragging effect in the context of Landau-quantized graphene. The underlying mechanism responsible for this effect originates from nonlinear chirality induced during a Raman gain process in the Landau levels of graphene. Specifically, the interaction of a magnetic dipole transition with an electric dipole transition facilitates Raman gain-assisted chirality. This interplay gives rise to strong magnetoelectric cross-coupling, allowing two Raman pathways to interfere via a shared magnetic-dipole transition. We explore the resulting modifications in both the group index and refractive index spectra, revealing pronounced signatures of magnetoelectric-induced dispersion. Our proposed framework paves the way for the design of advanced photonic devices with enhanced optical performance, potentially offering a novel route toward sub-wavelength imaging and precise resolution of nanoscale structures.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"198 ","pages":"Article 114866"},"PeriodicalIF":5.0,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192641","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-10DOI: 10.1016/j.optlastec.2026.114884
Jian Pu , Fang Hu , Yuan Li , Chuang Yue
In recent years, the increasing global demand for clean energy has spurred extensive research into hydrogen production via water electrolysis—a highly promising renewable technology. Electrocatalysts play a critical role in determining the efficiency and purity of this process, serving as a core component for improving overall performance. Laser ablation technology, boasting remarkable advantages like high efficiency, precision, and non-contact operation, has blazed a new route for the preparation of high-performance electrocatalysts. This technique enables the construction of micro/nano framework structures that offer abundant active sites, while allowing precise control over the composition, structure, and morphology of catalysts at the atomic level. As a result, it facilitates comprehensive enhancement of catalytic activity, stability, and electrical conductivity. This article provides a systematic review of recent advances in laser ablation for electrocatalyst preparation and offers a forward-looking perspective on future developments, aiming to serve as a valuable reference for further breakthroughs in the field.
{"title":"Advances in laser ablation-assisted water electrolysis for hydrogen production","authors":"Jian Pu , Fang Hu , Yuan Li , Chuang Yue","doi":"10.1016/j.optlastec.2026.114884","DOIUrl":"10.1016/j.optlastec.2026.114884","url":null,"abstract":"<div><div>In recent years, the increasing global demand for clean energy has spurred extensive research into hydrogen production via water electrolysis—a highly promising renewable technology. Electrocatalysts play a critical role in determining the efficiency and purity of this process, serving as a core component for improving overall performance. Laser ablation technology, boasting remarkable advantages like high efficiency, precision, and non-contact operation, has blazed a new route for the preparation of high-performance electrocatalysts. This technique enables the construction of micro/nano framework structures that offer abundant active sites, while allowing precise control over the composition, structure, and morphology of catalysts at the atomic level. As a result, it facilitates comprehensive enhancement of catalytic activity, stability, and electrical conductivity. This article provides a systematic review of recent advances in laser ablation for electrocatalyst preparation and offers a forward-looking perspective on future developments, aiming to serve as a valuable reference for further breakthroughs in the field.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"198 ","pages":"Article 114884"},"PeriodicalIF":5.0,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192644","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-10DOI: 10.1016/j.optlastec.2026.114892
Weimin Tang , Zhanqiang Liu , Wenjun Lyu , Bing Wang , Xiaoliang Liang , Jinfu Zhao , Liangliang Li
GH4169 alloy is extensively utilized in aerospace components owing to its superior mechanical properties and thermal stability. Ensuring high-quality surface integrity during milling is critical for the machined component performance. The study proposes a cost-effective 3D machined surface topography reconstruction approach. Optical image analysis and mathematical modeling are integrated through the co-optimization framework. In-situ monitoring and real-time evaluation of surface quality are enabled during the milling process. The proposed methodology leverages the U-Net pre-training on optical images and matched height point clouds acquired via laser scanning confocal microscopy (LSCM). Subsequent fine-tuning is performed using paired optical and point cloud datasets obtained from the digital microscope, enabling efficient surface reconstruction. The approach significantly enhances reconstruction speed and fidelity. In pre-training, the most suitable model has a structure similarity index measure (SSIM) of 0.9929, mean squared error (MSE), and mean absolute error (MAE) of 3.07 × 10-4 and 1.20 × 10-2 on the test set. In fine-tuning training, the best model has an SSIM of 0.8631, MSE, and MAE of 4.56 × 10-3 and 5.40 × 10-2 on the validation set when the smoothing coefficient is equal to 12. Then, the simulation roughness is calculated by mathematical modeling and compared with the reconstructed surface roughness to correct the reconstruction result. Finally, the interactive software is developed for engineering applications, which supports 2D/3D visualization, roughness evaluation, and simulation calculation, and systematically demonstrates the complete steps from data acquisition to result output. The study presents a method for rapid and in-situ detection of metal processing surface quality.
{"title":"Machined GH4169 surface topography reconstruction via optical imaging and mathematical modeling","authors":"Weimin Tang , Zhanqiang Liu , Wenjun Lyu , Bing Wang , Xiaoliang Liang , Jinfu Zhao , Liangliang Li","doi":"10.1016/j.optlastec.2026.114892","DOIUrl":"10.1016/j.optlastec.2026.114892","url":null,"abstract":"<div><div>GH4169 alloy is extensively utilized in aerospace components owing to its superior mechanical properties and thermal stability. Ensuring high-quality surface integrity during milling is critical for the machined component performance. The study proposes a cost-effective 3D machined surface topography reconstruction approach. Optical image analysis and mathematical modeling are integrated through the co-optimization framework. In-situ monitoring and real-time evaluation of surface quality are enabled during the milling process. The proposed methodology leverages the U-Net pre-training on optical images and matched height point clouds acquired via laser scanning confocal microscopy (LSCM). Subsequent fine-tuning is performed using paired optical and point cloud datasets obtained from the digital microscope, enabling efficient surface reconstruction. The approach significantly enhances reconstruction speed and fidelity. In pre-training, the most suitable model has a structure similarity index measure (SSIM) of 0.9929, mean squared error (MSE), and mean absolute error (MAE) of 3.07 × 10<sup>-4</sup> and 1.20 × 10<sup>-2</sup> on the test set. In fine-tuning training, the best model has an SSIM of 0.8631, MSE, and MAE of 4.56 × 10<sup>-3</sup> and 5.40 × 10<sup>-2</sup> on the validation set when the smoothing coefficient is equal to 12. Then, the simulation roughness is calculated by mathematical modeling and compared with the reconstructed surface roughness to correct the reconstruction result. Finally, the interactive software is developed for engineering applications, which supports 2D/3D visualization, roughness evaluation, and simulation calculation, and systematically demonstrates the complete steps from data acquisition to result output. The study presents a method for rapid and in-situ detection of metal processing surface quality.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"198 ","pages":"Article 114892"},"PeriodicalIF":5.0,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192329","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-10DOI: 10.1016/j.optlastec.2026.114870
Fan Qiu , Zhenhua Tang , Zhong-Jie Chen , Yu-Xiang Wu , Yan-Ping Jiang , Shui-Feng Li , Xin-Gui Tang , Xueqing Xu , Yi-Chun Zhou , Antonio Guerrero
Bias-tunable positive and negative photoconductivity is a crucial capability for modulating the photoelectric effect, providing substantial support for the development of high-performance optoelectronic devices and systems. Nonetheless, a limited number of two-terminal optoelectronic devices exist concerning the attainable positive and negative photoconductivity conversion under applied bias voltage. Herein, the Au/MXene/BFCO/FTO heterostructure is innovatively employed to effectively simulate artificial optoelectronic synapses, demonstrating exceptional analog resistive switching behavior and showcasing the diverse attributes of synaptic plasticity, encompassing short-term plasticity (STP, STD) and long-term plasticity (LTP, LTD). Interestingly, an intriguing bias-induced conversion between positive and negative photoconductivity was observed in Au/MXene/BFCO/FTO thin-film devices, and was attributed to the photogate effect (PGE). Furthermore, by implementing a convolutional neural network (CNN) architecture in conjunction with a stochastic adaptive optimization technique, we achieved enhanced recognition accuracies of 93% and 72% on the MNIST and Fashion MNIST datasets, respectively. These results may offer a feasible approach for combining BFCO materials with two-dimensional materials to construct optoelectronic synaptic devices for neuromorphic computing.
{"title":"Bias-tunable positive and negative photoconductivity in MXene/BFCO heterojunctions optoelectronic memristor for neuromorphic computing","authors":"Fan Qiu , Zhenhua Tang , Zhong-Jie Chen , Yu-Xiang Wu , Yan-Ping Jiang , Shui-Feng Li , Xin-Gui Tang , Xueqing Xu , Yi-Chun Zhou , Antonio Guerrero","doi":"10.1016/j.optlastec.2026.114870","DOIUrl":"10.1016/j.optlastec.2026.114870","url":null,"abstract":"<div><div>Bias-tunable positive and negative photoconductivity is a crucial capability for modulating the photoelectric effect, providing substantial support for the development of high-performance optoelectronic devices and systems. Nonetheless, a limited number of two-terminal optoelectronic devices exist concerning the attainable positive and negative photoconductivity conversion under applied bias voltage. Herein, the Au/MXene/BFCO/FTO heterostructure is innovatively employed to effectively simulate artificial optoelectronic synapses, demonstrating exceptional analog resistive switching behavior and showcasing the diverse attributes of synaptic plasticity, encompassing short-term plasticity (STP, STD) and long-term plasticity (LTP, LTD). Interestingly, an intriguing bias-induced conversion between positive and negative photoconductivity<!--> <!-->was observed<!--> <!-->in Au/MXene/BFCO/FTO thin-film devices,<!--> <!-->and was attributed to<!--> <!-->the photogate effect (PGE). Furthermore, by implementing a convolutional neural network (CNN) architecture in conjunction with a stochastic adaptive optimization technique, we achieved enhanced recognition accuracies of 93% and 72% on the MNIST and Fashion MNIST datasets, respectively. These results may offer a feasible approach for combining BFCO materials with two-dimensional materials to construct optoelectronic synaptic devices for neuromorphic computing.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"198 ","pages":"Article 114870"},"PeriodicalIF":5.0,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-10DOI: 10.1016/j.optlastec.2026.114862
Wei Wang, Wenyi Yang, Tong Zhang, Xia Yang
Linear servomotors (LSM) are extensively applied in machine tools, robotics, and precision automation, where accurate mover positioning is critical. Image matching has been widely explored in industrial measurement, with least-squares (LS) methods combined with gradient-based optimization serving as the dominant approach for subpixel accuracy. Nevertheless, the iterative nature of these methods not only increases computational burden but also makes convergence sensitive to the choice of initial values. To address these limitations, this paper proposes an iteration-free LS image matching framework, which directly estimates subpixel displacements without iterative refinement. Within this framework, a representative algorithm—the sum-table Gauss–Newton (ST-GN) method—is developed and applied to LSM mover positioning. Comprehensive simulations and experimental validations demonstrate that the proposed framework achieves high-precision matching, with a mean absolute error below 0.5 μm, thereby offering a reliable and efficient solution for high-accuracy image-based measurement in LSM applications.
{"title":"Iteration-free framework of least squares image matching for LSM mover positioning","authors":"Wei Wang, Wenyi Yang, Tong Zhang, Xia Yang","doi":"10.1016/j.optlastec.2026.114862","DOIUrl":"10.1016/j.optlastec.2026.114862","url":null,"abstract":"<div><div>Linear servomotors (LSM) are extensively applied in machine tools, robotics, and precision automation, where accurate mover positioning is critical. Image matching has been widely explored in industrial measurement, with least-squares (LS) methods combined with gradient-based optimization serving as the dominant approach for subpixel accuracy. Nevertheless, the iterative nature of these methods not only increases computational burden but also makes convergence sensitive to the choice of initial values. To address these limitations, this paper proposes an iteration-free LS image matching framework, which directly estimates subpixel displacements without iterative refinement. Within this framework, a representative algorithm—the sum-table Gauss–Newton (ST-GN) method—is developed and applied to LSM mover positioning. Comprehensive simulations and experimental validations demonstrate that the proposed framework achieves high-precision matching, with a mean absolute error below 0.5 μm, thereby offering a reliable and efficient solution for high-accuracy image-based measurement in LSM applications.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"198 ","pages":"Article 114862"},"PeriodicalIF":5.0,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192334","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
To address the limitations of CFRP laser machining process prediction in methodological benchmarking and mechanistic interpretability, this study proposes a morphology prediction framework that simultaneously integrates point-prediction accuracy, uncertainty quantification, and interpretability. Physically derived features are introduced to bridge external process parameters and morphological responses through a causally constrained energy-flow pathway, while a concise and reliable model is identified through systematic evaluation. Six baseline machine learning models are comparatively assessed along two principal dimensions—accuracy and uncertainty. By incorporating the coefficient of variation, maximal information coefficient, and recursive feature elimination, physical features exhibiting low dispersion, low collinearity, and high importance are selected to construct a three-layer causal chain of raw process parameters–physically derived features–morphological indicators. A dual-layer SHAP analysis is subsequently employed to hierarchically delineate the contribution pathways from process parameters to morphological responses.The results demonstrate that Gaussian Process Regression outperforms the other models in both predictive accuracy and uncertainty representation. Compared with models using only raw features, the inclusion of physically derived features enhances the reliability of uncertainty characterization and establishes physically constrained causal linkages between process parameters and morphological indicators. The three-layer causal chain, combined with the dual-layer SHAP analysis, jointly elucidates the distributional patterns and mechanistic contributions of morphological responses, thereby strengthening the causal consistency and interpretability of the predictive model. This work provides an efficient, robust, and interpretable technical paradigm for morphology prediction and process optimization in CFRP laser machining.
{"title":"Interpretable machine learning for laser machining morphology prediction of CFRP driven by physical-derived features","authors":"Ping Huang, Guanghui Zhang, Zhichuang Chen, Xinping He, Qingan Lu, Yuxing Huang, Hui Jiao, Jia Zhou, Yuhong Long","doi":"10.1016/j.optlastec.2026.114872","DOIUrl":"10.1016/j.optlastec.2026.114872","url":null,"abstract":"<div><div>To address the limitations of CFRP laser machining process prediction in methodological benchmarking and mechanistic interpretability, this study proposes a morphology prediction framework that simultaneously integrates point-prediction accuracy, uncertainty quantification, and interpretability. Physically derived features are introduced to bridge external process parameters and morphological responses through a causally constrained energy-flow pathway, while a concise and reliable model is identified through systematic evaluation. Six baseline machine learning models are comparatively assessed along two principal dimensions—accuracy and uncertainty. By incorporating the coefficient of variation, maximal information coefficient, and recursive feature elimination, physical features exhibiting low dispersion, low collinearity, and high importance are selected to construct a three-layer causal chain of raw process parameters–physically derived features–morphological indicators. A dual-layer SHAP analysis is subsequently employed to hierarchically delineate the contribution pathways from process parameters to morphological responses.The results demonstrate that Gaussian Process Regression outperforms the other models in both predictive accuracy and uncertainty representation. Compared with models using only raw features, the inclusion of physically derived features enhances the reliability of uncertainty characterization and establishes physically constrained causal linkages between process parameters and morphological indicators. The three-layer causal chain, combined with the dual-layer SHAP analysis, jointly elucidates the distributional patterns and mechanistic contributions of morphological responses, thereby strengthening the causal consistency and interpretability of the predictive model. This work provides an efficient, robust, and interpretable technical paradigm for morphology prediction and process optimization in CFRP laser machining.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"198 ","pages":"Article 114872"},"PeriodicalIF":5.0,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192645","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-10DOI: 10.1016/j.optlastec.2026.114903
Yu-Che Wu, Kuo-Chih Chang, Shu-Chun Chu
Digital lasers control the laser beam by dynamically updating the phase patterns of the spatial light modulator (SLM) within the laser cavity. Due to the presence of nonlinear effects, such as mode competition and gain saturation in digital laser systems, it is often necessary to rely on specifically manually tailored approach or iteration processes to find suitable loaded phases in Digital lasers. This study proposes a model based on Conditional Generative Adversarial Networks (cGAN) and a modified U-Net architecture, with designed loss functions to inverse design the loaded phases. This study employs deep neural networks to learn an effective nonlinear relation between light field intensity and the corresponding SLM-loaded phase in simulated L-shaped digital lasers, enabling the prediction of SLM-loaded phases for both analytical and non-analytical arbitrary structured light fields. The results demonstrate superior performance on non-analytical light fields compared to the current methods in L-shaped Digital lasers. Furthermore, a transfer learning strategy is introduced, allowing knowledge obtained from one class of structured beams to be effectively reused for another, as well as to cavity-length variations. Thereby enhances generalization and improves performance under limited training data. To the best of our knowledge, this is the first deep-learning-based inverse intracavity phase design framework specifically demonstrated for digital laser systems. Providing an efficient alternative for generating structured light in other digital laser systems.
{"title":"Inverse-designed phase prediction in digital lasers using deep learning and transfer learning","authors":"Yu-Che Wu, Kuo-Chih Chang, Shu-Chun Chu","doi":"10.1016/j.optlastec.2026.114903","DOIUrl":"10.1016/j.optlastec.2026.114903","url":null,"abstract":"<div><div>Digital lasers control the laser beam by dynamically updating the phase patterns of the spatial light modulator (SLM) within the laser cavity. Due to the presence of nonlinear effects, such as mode competition and gain saturation in digital laser systems, it is often necessary to rely on specifically manually tailored approach or iteration processes to find suitable loaded phases in Digital lasers. This study proposes a model based on Conditional Generative Adversarial Networks (cGAN) and a modified U-Net architecture, with designed loss functions to inverse design the loaded phases. This study employs deep neural networks to learn an effective nonlinear relation between light field intensity and the corresponding SLM-loaded phase in simulated L-shaped digital lasers, enabling the prediction of SLM-loaded phases for both analytical and non-analytical arbitrary structured light fields. The results demonstrate superior performance on non-analytical light fields compared to the current methods in L-shaped Digital lasers. Furthermore, a transfer learning strategy is introduced, allowing knowledge obtained from one class of structured beams to be effectively reused for another, as well as to cavity-length variations. Thereby enhances generalization and improves performance under limited training data. To the best of our knowledge, this is the first deep-learning-based inverse intracavity phase design framework specifically demonstrated for digital laser systems. Providing an efficient alternative for generating structured light in other digital laser systems.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"198 ","pages":"Article 114903"},"PeriodicalIF":5.0,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192649","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-10DOI: 10.1016/j.optlastec.2026.114896
Wenhui Wang , Haolong Jia , Guozhong Lei , Jiaming Xu , JingQi Liu , Wenchang Lai , Yan Wang , Kai Han
Single-pixel complex amplitude detection offers significant potential for applications in biomedical imaging, three-dimensional topography measurement, adaptive optics, and related fields. However, conventional interferometric methods require four phase-shifting steps to reconstruct intensity and phase, limiting the detection speed. This paper introduces a novel two-step phase-shifting technique that requires only two phase-shifted intensity measurements and one DC measurement for reconstruction. We develop a theoretical model and conduct numerical simulations. Then experimentally compare the four-step and two-step methods and validate the generality of the model by testing different illumination patterns. The proposed method not only achieves detection quality comparable to the conventional four-step approach but also improves the detection speed by approximately , demonstrating a significant advance in single-pixel imaging technology.
{"title":"Two-step phase-shifting single-pixel complex amplitude detection technique","authors":"Wenhui Wang , Haolong Jia , Guozhong Lei , Jiaming Xu , JingQi Liu , Wenchang Lai , Yan Wang , Kai Han","doi":"10.1016/j.optlastec.2026.114896","DOIUrl":"10.1016/j.optlastec.2026.114896","url":null,"abstract":"<div><div>Single-pixel complex amplitude detection offers significant potential for applications in biomedical imaging, three-dimensional topography measurement, adaptive optics, and related fields. However, conventional interferometric methods require four phase-shifting steps to reconstruct intensity and phase, limiting the detection speed. This paper introduces a novel two-step phase-shifting technique that requires only two phase-shifted intensity measurements and one DC measurement for reconstruction. We develop a theoretical model and conduct numerical simulations. Then experimentally compare the four-step and two-step methods and validate the generality of the model by testing different illumination patterns. The proposed method not only achieves detection quality comparable to the conventional four-step approach but also improves the detection speed by approximately <span><math><mn>33</mn><mi>%</mi></math></span>, demonstrating a significant advance in single-pixel imaging technology.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"198 ","pages":"Article 114896"},"PeriodicalIF":5.0,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192650","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号