基于碳纳米线的三维编织复合材料损伤源定位研究

发布时间：2018-01-23 20:05

本文关键词： 三维编织复合材料碳纳米线传感器小波阈值降噪内部缺陷损检测四分矩阵奇异值分解　出处：《天津工业大学》2017年博士论文　论文类型：学位论文

【摘要】：复合材料因为其优异的比刚度、比强度,以及良好的耐久性和抗疲劳性能,已被广泛应用在航空航天、建筑、高铁和汽车材料等领域。实时高效的无损检测方法对复合材料结构的内部损伤检查和持续在线监测意义重大。本文将碳纳米传感测量和奇异值分解算法相结合应用于三维编织复合材料试件的损伤检测,对树脂基复合材料进行了损伤存在性与损伤位置确定的检测实验,构建了基于碳纳米线的三维编织复合材料损伤检测系统。本文首先介绍了复合材料的定义,基于其在航空航天等诸多领域的广泛应用分析了复合材料无损检测的重要性,并对复合材料健康监测的国内外研究现状进行了阐述。其次,阐述了碳纳米线的传感原理及制备方法,并对碳纳米线的传感特性进行分析,给出了碳纳米传感器用于三维编织复合材料的损伤检测的理论依据。随后,文章介绍了碳纳米线传感器嵌入三维编织复合材料试件的编织方法。实验中所用试件均采用三维四步六向编织工艺将碳纳米线传感器以轴向纱和六向纱形式嵌入复合材料。针对三维六向编织复合材料,基于四步法编织工艺携纱器的运动规律,构建了结构的三单胞几何模型,研究了材料各区域纱线的面内和空间运动轨迹,预测了材料内部碳纳米线传感器的走向。在此基础上,对碳纳米传感器阵列采集的数据矩阵用二维小波阈值法进行预处理。考虑到小波硬阈值函数不连续而软阈值函数与真实信号间存在恒定偏差的缺点,提出了一种动态阈值算法,使得改进后的阈值随分解层数的增加而动态改变,且该阈值函数连续可导。通过Matlab软件仿真证明了该阈值函数的降噪效果优于传统的阈值函数,为后续的数据处理打好了基础。最后,考虑复合材料健康监测中的数据处理问题,为保证系统良好的实时性,提出四分矩阵奇异值分解算法以简化数据,提高算法效率;同时提取数据矩阵的关键特征,通过与健康试件的特征值比对,确定损伤的存在并计算损伤位置。实验结果表明该算法应用于三维编织复合材料的损伤定位切实可行。
[Abstract]:Composite materials have been widely used in aerospace and construction because of their excellent specific stiffness, specific strength, good durability and fatigue resistance. In the field of high iron and automobile materials, real time and high efficient nondestructive testing method is of great significance for internal damage detection and continuous on-line monitoring of composite structures. In this paper, carbon nanosensor measurement and singular value decomposition algorithm are combined. It is used for damage detection of 3D braided composite material. In this paper, the damage existence and damage location of resin matrix composites are tested, and a three-dimensional braided composite damage detection system based on carbon nanowires is constructed. Firstly, the definition of composite material is introduced in this paper. Based on its extensive application in aerospace and other fields, the importance of nondestructive testing of composite materials is analyzed, and the research status of health monitoring of composite materials at home and abroad is described. Secondly. The sensing principle and preparation method of carbon nanowires are described, and the sensing characteristics of carbon nanowires are analyzed, and the theoretical basis for damage detection of 3D braided composites by carbon nanowires is given. In this paper, the braiding method of carbon nanowire sensor embedded in 3D braided composite specimen is introduced. In the experiment, the carbon nanowire sensor is embedded in the form of axial yarn and hexagonal yarn by using three-dimensional four-step six-direction braiding process to embed the carbon nanowire sensor in the form of axial yarn and six-direction yarn. Composites. Three dimensional hexagonal braided composites. Based on the motion law of the yarn carrier in the four-step braiding process, the three-cell geometric model of the structure is constructed, and the in-plane and spatial motion trajectories of the yarns in each region of the material are studied. The trend of carbon nanowire sensor is predicted. Two-dimensional wavelet threshold method is used to preprocess the data matrix collected by carbon nanosensor array. Considering the disadvantage of wavelet hard threshold function discontinuity and constant deviation between soft threshold function and real signal. In this paper, a dynamic threshold algorithm is proposed, which makes the improved threshold change dynamically with the increase of the number of decomposition layers. And the threshold function can be continuously differentiable. Matlab software simulation shows that the threshold function noise reduction effect is better than the traditional threshold function, which provides a good basis for the subsequent data processing. Finally. Considering the problem of data processing in composite material health monitoring, a quadrilateral matrix singular value decomposition (QSVD) algorithm is proposed to simplify the data and improve the efficiency of the algorithm in order to ensure the good real-time performance of the system. At the same time, the key features of the data matrix are extracted and compared with the eigenvalues of the healthy samples. The experimental results show that the proposed algorithm is feasible for 3D braided composite material damage location.
【学位授予单位】：天津工业大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TB33

【相似文献】