大型复合材料结构快速无损检测与损伤识别方法研究

Aim at imminence requirement of rapidly and real time test for large-scale composite materials structure, mathematics model of thermoelastic ultrasound generation by laser in carbon fiber laminated composite materials is built. Diffuse and change law of ultrasound is studied in the composite materials. The develop thought of continuous distributed optical fiber F-P sensor for collect multi-ultrasound information is put forward. Continuous distributed optical fiber ultrasound sensing theory is studied. And each factor that influence collect precision of sensor is analyzed. Induction array distributing principle of sensor is studied based on the characteristic of large-scale composite materials structure. Distributing test point is optimized and designed. The multi-point array of sensor is constructed in order to ensuring the sensor's sensitivity for detecting multi-ultrasound information. The arithmetic of damage recognition on composite materials is studied. The rapidly and qualitative detection system of laser ultrasound is constructed. Through above studies, the theory of thermoelastic ultrasound generation by laser and diffuse characteristic of ultrasound in composite materials will be opened out. And continuous distributed optical fiber ultrasound sensor's practicality and commodity will be realized. New method of rapidly and real time detection for large-scale composite materials structure is built. The research of this item has important significance to fast development of our national composite materials technology. It will offer theoretic and experimental fundamental for our national NDT development.

针对大型复合材料结构微缺陷快速实时检测的迫切需求，以碳纤维铺层复合材料为主要研究对象，建立其激光热弹激发超声的数学模型，研究超声在复合材料内部的传播变化规律；提出一种适用于多个超声信息特征拾取的连续分布式光纤F-P传感器研制思想，研究连续分布式光纤F-P超声传感机理，分析影响传感器拾取精度的各种因素，结合大型复合材料结构特性，研究传感器的感应阵列分布原则，优化设计分布监测点，构建传感器多点阵列，以保证传感器探测多个超声信息的灵敏度；研究复合材料损伤识别算法，构建快速定性化激光超声检测系统。通过上述研究，可望揭示复合材料中激光热弹激发超声的机理和超声传播的特性，实现连续分布式光纤F-P超声传感器的实用化和商品化，建立大型复合材料结构快速实时在线检测新方法。项目的研究成果对于我国复合材料技术的快速发展具有重要意义，也可以为我国无损检测新技术的开拓提供理论和实验基础。

针对大型复合材料结构微缺陷快速实时检测的迫切需求，考虑材料组织参数和激光参数，建立了激光线源在单向纤维复合材料中激发超声的三维有限元模型，仿真研究了不同参数影响超声在复合材料内部的传播变化规律；设计了非本征型光纤F-P传感器的结构，确定传感器工作点稳定技术，建立了双波长光纤F-P传感系统的DE算法优化模型，研制了一正交高精度的光纤F-P传感系统，实验研究了该传感系统的方向性，结果显示其具有较强方向灵敏性；建立基于光纤F-P超声传感系统的碳纤维复合材料缺陷实时在线超声检测系统，进行实验研究，结果表明，该系统能有效实现碳纤维复合材料缺陷的实时在线检测，易于工程实用化；建立复合材料激光超声可视化检测系统，实验结果表明，该系统可实时动态展示构件内部缺陷信息，也可对超声传播信息按时间帧回放、选取，利于分析缺陷的位置和尺寸；通过EMD对检测的超声波形进行分析，提出了基于BP神经网络的复合材料缺陷分类和模式识别方法，在实测样本的基础上对缺陷分类进行了仿真试验研究，取得较好效果，说明将EMD与BP神经网络应用于超声检测复合材料缺陷模式识别与分类是切实可行。项目研究成果对于我国复合材料的快速实时在线检测具有重要意义，也为将激光超声检测技术推向工程应用奠定扎实的理论和技术基础。