直径为数十到数百微米的微孔阵列越来越广泛用于航空航天、能源、化纤等行业。超快激光微孔加工具有对材料无选择性、灵活度高和效率高的优势，但存在孔内激光能量传输及孔型形成机理不明确、高通量加工热损伤和加工质量难以检测等难题。本文采用900 fs超快激光，研究了金属材料上旋切加工中微孔内部激光能量的传输及孔型形成机理，通过多光束和旋光钻孔融合建立了微孔阵列高通量控型加工工艺，提出了高通量微孔加工热损伤抑制策略和基于深度学习的质量检测方法。 通过原位成像直接观测结合仿真研究了激光束在孔内的多重反射和反射光束对材料去除过程的影响，阐明了平行和垂直于线偏振两个方向上的加工孔壁角度由能量吸收率和材料烧蚀阈值共同决定，发现了孔壁对反射光束的汇聚效应，揭示了线偏振激光微孔加工时“收缩-发散”孔型的形成机理。提出了反射激光束开始主导孔型演变的深度阈值，验证了该深度阈值由脉冲能量、旋切半径、材料特性和激光偏振共同确定，圆偏振激光加工时可得到比线偏振激光更大的深度阈值。 开发并建立了旋光系统-激光参量-平台运动协调控制的软硬件平台，量化研究了旋光加工微孔的直径和锥度调控原理，在铜、不锈钢、钛等材料上加工出了深径比5:1的正、负及零锥度直孔。提出了基于动态参数调控的异型微孔旋光加工方法，实现了收缩-扩张性、锥直型等异型孔加工；通过旋光系统与多光束加工融合，实现了异型微孔阵列的高通量控型加工，平均加工效率提升10倍以上。 为解决大面积微孔阵列高通量加工热损伤，提出了基于热输入时-空分离的加工路径和参数优化模型，实现了多光束激光微孔加工时更均匀的温度分布，建立并验证了有限差分模型求解温度场；自主开发软件实现了加工路径的快速优化，工件最高温度和导致材料氧化的过热持续时间分别降低了10%、50%，工件的热变形、氧化变色和重铸层等缺陷明显下降。另外，为解决大面积微孔阵列加工质量检测问题，提出了基于深度学习的加工质量快速检测方法，可同时识别和提取微孔轮廓、重铸层、碎屑颗粒、微裂纹等多个特征，进而从几何质量和表面质量两个维度实现了微孔阵列加工质量的快速、精准评价。
Micro-hole arrays with diameters ranging from tens to hundreds of micrometers are increasingly being uesd in industries such as aerospace, energy, and synthetic fibers. Ultrafast laser drilling provides several advantages, including non-selective materials, high flexibility, and high efficiency. However, it faces challenges such as the unclear mechanism of laser energy transmission and hole shape formation, thermal damage during high-throughput processing, and difficulty in evaluating processing quality. To address these challenges, this study has investigated the transmission of laser energy inside micro-holes and the mechanism for hole-shape formation in metallic materials using a 900-fs laser. Furthermore, a new high-throughput drilling technique using multi-beam helical drilling has been proposed. Finally, this paper proposed strategies to reduce thermal damage in the high-throughput processing of micro-hole arrays with large areas. Additionally, a novel quality evaluation method, based on deep learning, has been presented.Multi-reflections of laser beams inside micro-holes and the influence of the reflected beams on the material ablation process were studied through in-situ imaging and simulation. It was revealed that the difference in hole wall angles between the directions parallel and perpendicular to linear polarization is dependent on variances in energy absorption and material ablation thresholds. The early formed hole wall causes a converging effect on the reflected beams. Besides, the mechanism of the "contraction-divergence" hole-shape formation during linearly polarized laser drilling was also revealed. A depth threshold was proposed, from which the hole shape evolution is dominated by the reflected laser beam. It was verified that this depth threshold is jointly determined by the pulse energy, rotation radius of the laser spot, material properties, and laser polarization.A coordinated platform consisting of the helical drilling system, laser source, and motion stage was established. The principles of diameter and taper control in helical drilling were quantitatively investigated , and straight holes with positive, negative, and zero tapers with a depth-to-diameter ratio of 5:1 were processed in copper, stainless steel, and titanium. In addition, a method for drilling shaped micro-holes using dynamic control of drilling parameters was proposed, which has realized the processing of “shrinkage-expansion” shaped micro-holes and straight holes with different tapers at both ends. By integrating the helical drilling system with multi-beam processing, high-throughput drilling of irregular-shaped micro-hole arrays was achieved , and the average processing efficiency was increased by more than 10 times.To address thermal damage during high-throughput processing of large-area micro-hole arrays, a heat input regulation strategy based on temporal and spatial separation was proposed. This strategy resulted in a more uniform temperature distribution during multi-beam laser drilling of micro-hole arrays. A finite difference model for solving the temperature field was established and verified, and an independently developed software was used to achieve rapid optimization of the processing path.The maximum workpiece temperature and the duration of overheating causing material oxidation were reduced by 10% and 50%, respectively. Improvements in thermal deformation, oxidation discoloration and recast layer were observed. In addition, a new deep learning-based quality evaluation method for large-area micro-hole arrays was introduced. It can simultaneously identify and extract multiple features such as hole profiles, recast layers, debris, and microcracks, allowing the fast and accurate evaluation of processing quality of micro-hole arrays from both geometric and surface quality perspectives.
