基于Lamb波信号时频分析的铝合金结构损伤检测的研究

发布时间：2018-04-09 22:20

本文选题：结构健康监测　切入点：时频分析　出处：《电子科技大学》2014年硕士论文

【摘要】：作为一种有效的实时在线检测方法,基于Lamb波的结构健康监测技术成为最近十余年国内外的研究热点。该技术通过分析嵌入或附着在待测结构上的传感器所发送的数据,实时地监测结构的整体状况,及时对结构的损伤位置和程度进行分析和诊断,在航空航天、大型土木结构等领域具有广泛的应用前景。本文研究薄板中的Lamb波理论,结合HHT等时频分析技术对铝合金结构进行损伤检测。具体的研究工作如下:首先,从理论上研究Lamb波的传播特性,通过数值法求解频散方程,并绘制出Lamb波的频散曲线和振动幅度,为有效利用Lamb波进行结构损伤检测提供理论依据。在此基础上结合实验,从中心频率、波峰数以及激励形式等几个方面对Lamb波激励信号进行优化分析。其次,对整个Lamb波损伤检测系统进行了设计,包括系统的硬件组成以及部分软件操作。鉴于时域上分析Lamb波信号存在一定的局限性,本文引入HHT、波包分解等信号处理技术,结合频域分析,提取信号中的损伤特征信息。通过信号预处理以及小波多尺度分解对检测信号进行滤波去噪,结合实例分析进一步论证HHT以及波包分解技术处理检测信号的有效性。最后,搭建了基于Lamb波的结构损伤检测实验平台,对铝合金结构进行损伤识别分析以及初步的损伤定位研究。以1mm厚的7050铝合金板为检测对象,获取Lamb波在完好无损、存在裂纹损伤以及圆孔缺陷的板材结构中传播时的响应信号。实验中通过快速傅立叶变换提取出信号基波和二次谐波的能量以判定损伤是否存在;使用希尔伯特-黄变换提取出信号的最大能量幅值,将其作为表征损伤的特征参数,估计损伤程度;采用椭圆定位原理,通过希尔伯特-黄变换、波包分解技术获取信号能量峰值的到达时刻,实现铝合金结构损伤的初步定位。实验分析结果表明,所使用的损伤检测方法能够识别损伤以及确定损伤的大致位置,取得了明显的实验效果。
[Abstract]:As an effective real-time on-line detection method, structural health monitoring technology based on Lamb wave has become a hot research topic in recent ten years at home and abroad.Large-scale civil structures and other fields have a wide range of applications.In this paper, the theory of Lamb wave in thin plate is studied, and the damage detection of aluminum alloy structure is carried out with HHT isotime-frequency analysis technique.The specific research work is as follows: firstly, the propagation characteristics of Lamb wave are studied theoretically, the dispersion equation is solved by numerical method, and the dispersion curve and vibration amplitude of Lamb wave are drawn.It provides a theoretical basis for the effective use of Lamb wave in structural damage detection.On the basis of the experiments, the excitation signal of Lamb wave is optimized and analyzed from several aspects, such as center frequency, peak number and excitation form.Secondly, the whole Lamb wave damage detection system is designed, including the hardware composition of the system and part of the software operation.In view of the limitation of analyzing Lamb wave signal in time domain, this paper introduces signal processing techniques such as HHT, wave packet decomposition, and combines frequency domain analysis to extract damage characteristic information from the signal.Signal preprocessing and wavelet multiscale decomposition are used to filter and Denoise the detection signal, and the effectiveness of HHT and wave packet decomposition technology in processing the detected signal is further demonstrated by the analysis of practical examples.Finally, an experimental platform of structural damage detection based on Lamb wave is set up to identify and analyze the damage of aluminum alloy structure and to study the damage location of aluminum alloy structure.In the experiment, the energy of signal fundamental wave and second harmonic wave is extracted by fast Fourier transform to determine the existence of damage, and the maximum energy amplitude of signal is extracted by Hilbert-Huang transform, which is regarded as the characteristic parameter to characterize the damage.The damage degree is estimated and the arrival time of the peak energy of the signal is obtained by using the Hilbert-Huang transform and wave packet decomposition technique using the elliptical localization principle to realize the initial location of the damage of aluminum alloy structure.The experimental results show that the damage detection method can identify the damage and determine the general location of the damage, and achieve obvious experimental results.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TG115.28;TN911.6

【参考文献】