结构梁损伤的频响函数幅值变化特征分析

为了识别振动筛下横梁的损伤特征,探寻结构梁损伤在线监测的有效途径,通过理论计算与实验模态分析,从两方面系统地研究了结构损伤对其局部刚度的影响,得到了弯曲振动梁的频响函数特征变化与其弯曲刚度下降及梁的损伤程度之间的关系,并分别给出了梁在自由状态与约束状态下的结构损伤程度与频响函数幅值变化之间的定量关系,这种定量关系可以用来进行结构梁损伤的在线监测。     

一种基于高频频响函数的无基准疲劳裂纹识别方法

传统的线性频响函数损伤识别方法依赖于与健康基准对比来识别损伤,限制了其应用范围。提出了一种基于非线性高频频响函数的无基准疲劳裂纹损伤识别方法。通过分析出现裂纹时损伤界面的非线性特征,提出利用不同幅值激励下高频频响函数不同的特性,在无基准情况下提取非线性频响成分,构造损伤指数表征非线性损伤,同时分析了不同频率段对疲劳裂纹的敏感程度。实验表明该方法可以在无基准情况下有效识别疲劳裂纹。     

基于RCSA的深孔内圆磨床主轴端点频响函数预测

基于响应耦合子结构分析法预测了深孔内圆磨床主轴端点的频响函数。首先对磨床主轴进行子结构划分,计算各子结构自由状态下的频响函数矩阵,然后顺序刚性耦合各子结构的频响函数矩阵,对轴承支撑点使用结构修改法修改轴承约束下的已耦合子结构频响函数矩阵,直至耦合到最后一个子结构,得到主轴端点的频响函数。以某深孔内圆磨床为研究对象,分别基于该方法和有限元法,对其主轴端点的频响函数进行预测,并对其进行实验测试。实验及分析结果表明：该方法预测精度高于有限元分析方法预测精度、计算速度快,便于深孔内圆磨床主轴系统的结构优化。     

用实测频响函数的相对灵敏度对机床状态进行监测

通过实测机床频响函数，利用频响函数的相对灵敏度和机床模态参数间的关系，提取特征信息，通过对信息的分析，对机床运行状态进行监测。     

频响函数在结构损伤检测中应用的试验研究

结构动力测试中,频响函数可直接测得,不仅避免了在模态提取时的误差,而且在相同频段上可提供更多结构损伤信息,因此可方便地应用于结构损伤的检测中。对于结构基础的随机激励,试验结果表明频响函数法可有效地检测结构的损伤。同时说明传感器布置在结构的不同位置都可检测结构的损伤。并且随机振动的大小不影响结构损伤检测的效果。为频响函数法在现场实测结构的损伤提供了依据。     

用半理论法预测主轴系统刀尖点频响函数

采用半理论法,即理论与试验相结合的方法预测主轴系统刀尖点频响函数。首先,介绍半理论法的预测原理;然后,应用半理论法预测主轴系统刀尖点频响函数的流程,包括利用梁理论计算自由-自由状态刀具两端的阻抗矩阵、搭建主轴-刀柄频响函数测试系统、测试装卡短光滑圆柱的主轴-刀柄系统的频响函数、根据半理论法计算刀尖点频响函数和试验验证;最后,以某立式加工中心主轴系统为研究对象,应用该方法对刀尖点频响函数进行预测,并与试验进行对比以证明该方法的有效性。     

损伤识别在车用储氢容器模态试验中的应用

针对车用高压储氢容器的复合纤维缠绕结构在其使用过程中的安全问题,开展了体积为0.074m3车用纤维缠绕储氢容器的模态试验研究。根据完好容器、反复充装引起的桶身轻微损伤和疲劳破裂两种疲劳损伤状态容器的模态分析,探讨了基于模态参数进行损伤识别的可行性。研究表明,无论是轻微损伤还是疲劳破裂状态,局部的疲劳裂纹对损伤前后的频响函数有一定影响但并不敏感,无法基于频率响应函数判定损伤是否发生并确定损伤位置,但基于损伤前后振型比较可以实现损伤判定并预测损伤位置。该研究为确定车用储氢结构的在线损伤识别奠定了实践基础。     

激励点优化和加速度信号的结构损伤识别

通过对激励点优化布置获取结构的响应信息,提出了一种以加速度信号差曲率函数作为损伤指标,直接利用输出信号快速判断结构损伤位置的方法.首先计算模态振型,以模态保证准则(MAC)矩阵非对角元素最小值作为适应度函数,采用改进粒子群算法(MPSO)优化激励点数量和位置,再运用平均加速度幅值和均方根评价准则选择较优的激励点布置方案;然后试验激励对应的位置,获取加速度信号后计算测点处损伤前后加速度差的平方的积分值,运用曲率指标函数确定损伤位置,并对加速度信号通过巴特沃斯滤波后作为改进多尺度样本熵(MMSE)的输入样本;最后根据MMSE均值的变化,判定各工况相对损伤程度变化.结果表明:利用结构响应的加速度信号差曲率函数适合作为损伤识别的判别指标,通过三维桁架振动台中螺栓连接的状态模拟损伤,可以对不同损伤工况进行损伤诊断.     

频响函数作为神经网络输入参数的应用研究

给出一个BP神经网络,应用频响函数作为神经网络的输入参数,用来识别结构的状态信息,由于频响函数数据量大,直接作为神经网络输入参量容易造成网络训练收敛慢或不收敛。将频响函数在(0,1)范围进行归一化处理,通过用BP网络试验,新方法收敛速度提高了25倍,表明这是一种简单有效的方法。     

频响函数作为神经网络输入参数的应用研究

给出一个BP神经网络,应用频响函数作为神经网络的输入参数,用来识别结构的状态信息.由于频响函数数据量大,直接作为神经网络输入参量容易造成网络训练收敛慢或不收敛.将频响函数在(0,1)范围进行归一化处理,通过用BP网络试验,新方法收敛速度提高了25倍,表明这是一种简单有效的方法.     

Frequency response function shape-based methods for structural damage localisation

The ultilisation of structural shape signals for damage localisation has shown some promise, especially in the applications where an accurate finite element model of the structure is not available. For this purpose, traditional shape signals, like mode shapes, flexibility matrices, uniform load surface (ULS) and operational deflection shapes (ODS) have been widely used. Using frequency response function (FRF) shapes for structural damage localisation is however, a relatively new but promising technique. Unlike mode shapes, ULS and ODS, FRF shapes are defined on broadband data and so have potential to reveal damage location more clearly. Another advantage of using FRF shapes is that the test data can be directly used without the necessity of conducting modal identification. Nevertheless, some problems associated with this approach still remain to be solved. No solid foundation or deduction about the use of FRF shapes for damage localisation has been given in any literature so far. In addition, it has been observed that this method only works for a low-frequency range. This limitation of FRF shapes has not been explained or well treated so far. In this study, a scheme of using FRF shapes for structural damage localisation is proposed. Methods within this scheme include some important modifications like using the imaginary parts of FRF shapes and normalising FRF shapes before comparison. The theoretical explanation of using FRF shapes for damage localisation is presented and the limitations of the previous FRF shape methods have been overcome. The proposed methods have shown great potential in structural damage localisation.     

应变模态振型获取的一种简便方法

根据应变模态分析原理,定义了一个新的应变模态振型系数,提出了基于应变响应获取结构应变模态振型的一种简便方法,并通过简支梁实验进行了验证。研究结果表明,采用该方法无需测量位移模态,仅需采用单点激励,用电阻应变计测量结构上各测点的应变响应信息,即可获得被测结构应变模态振型,大大简化了应变模态在工程结构损伤识别中的实验检测分析过程。     

Time-domain analysis of piezoelectric impedance-based structural health monitoring using multilevel wavelet decomposition

A new approach to analyze the response from a piezoelectric wafer in an impedance-based structural health monitoring (SHM) method is proposed. It is shown that the time-domain response of a piezoceramic wafer provides information on the electromechanical impedance (EMI) variation when a monitored structure is damaged. Practical analysis was carried out using wavelet transform in two different levels. This approach simplifies EMI based SHM and the results show that it is more sensitive to damage than methods based on impedance measurements in the frequency domain. The efficiency of this new approach is demonstrated through experiments using an aluminum plate. The piezoelectric wafer was excited using a chirp signal and its response was analyzed using both frequency response functions (FRF) and the proposed method. The results confirm that this new approach is more sensitive to detect damage than FRF based methods.     

Locating damage in waveguides from the phase of point frequency response measurements

A method for detecting damage in uniform waveguide structures from two or more point frequency response functions (FRFs) is described. Attention is focussed on bending waves in beams although the method can in principle be applied to any waveguide structure. The input FRF is the superposition of directly injected waves and waves reflected from the damage and from other scattering regions in the structure. The phase of this FRF modulates with wavenumber, with the period of modulation in wavenumber space being related to the distances between the excitation point and scattering locations. The phase spectrum of the input FRF is found: the phase is determined, dispersive effects are removed by transforming from the frequency domain to the wavenumber domain, and the inverse Fourier transform from the wavenumber to the space domain found. Peaks in this phase spectrum indicate the distance to the scatterer. Two (or more) input FRFs can be used to determine the location of the scatterer unambiguously. Signal processing issues are discussed. Numerical results for a uniform beam with a breathing crack are given and experimental results for beams with a slot cut into them are presented. The approach lies in the middle ground between low frequency, modal methods and high frequency, ultrasonic methods. It allows one to interrogate a region of a structure rather than the complete structure and there is no requirement for a validated model of the structure in order to locate damage, apart from an estimate of the dispersion relation.     

On-line fan blade damage detection using neural networks

This paper presents a methodology for monitoring the on-line condition of axial-flow fan blades with the use of neural networks. In developing this methodology, the first stage was to utilise neural networks trained on features extracted from on-line blade vibration signals measured on an experimental test structure. Results from a stationary experimental modal analysis of the structure were used for identifying global blade mode shapes and their corresponding frequencies. These in turn were used to assist in identifying vibration-related features suitable for neural network training. The features were extracted from on-line blade vibration and strain signals which were measured using a number of sensors.The second stage in the development of the methodology entails utilising neural networks trained on numerical Frequency Response Function (FRF) features obtained from a Finite Element Model (FEM) of the test structure. Frequency domain features obtained from on-line experimental measurements were used to normalise the numerical FRF features prior to neural network training. Following training, the networks were tested using experimental frequency domain features. This approach makes it unnecessary to damage the structure in order to train the neural networks.The paper shows that it is possible to classify damage for several fan blades by using neural networks with on-line vibration measurements from sensors not necessarily installed on the damaged blades themselves. The significance of this is that it proves the possibility to perform on-line fan blade damage classification using less than one sensor per blade. Even more significant is the demonstration that an on-line damage detection system for a fan can be developed without having to damage the actual structure.     

Transient suppression in FRF measurement: Comparison and analysis of four state-of-the-art methods

This paper compares four well selected methods for computing the non-parametric Frequency Response Function (FRF) of a periodically excited linear time invariant system. The suppression of the transient is mandatory when its influence in the data is large. Better suppression of the transient leads to a better non-parametric FRF estimate. A good non-parametric FRF estimate can be used to validate the parametric transfer function model in a second step. The suppression of the transient will be highlighted using the mean squared error of the non-parametric FRF estimate. Temperature transients caused by heat diffusion are used as example. The selected methods consist of two standard windowing methods and two methods based on the Local Polynomial Method (LPM). LPM was designed to find a non-parametric FRF estimate in the presence of nonlinearities. This paper will modify LPM to find a non-parametric FRF estimate for linear systems using a single experiment. The mean squared error of the four non-parametric FRF estimates will be compared and analyzed, based on a simulation and a measurement example.     

Vibration based damage detection using large array sensors and spatial filters

In this paper, a novel approach for vibration based damage detection is proposed. The approach relies on the use of a large network of sensors (possibly hundreds of them) to which a programmable linear combiner is attached. The linear combiner is programmed to work as a modal filter. The frequency content of the output of the modal filter is proposed as feature for damage detection. It is shown that if a local damage is present, spurious peaks appear in the FRF of the modal filter whereas if temperature changes are considered, the FRF of the modal filter is shifted but its shape remains unchanged. The approach is interesting because of the ability to differentiate between local damage and global environmental changes to a structure. Issues about the practical implementation of the method are discussed.     

基于压缩频响函数的结构损伤识别

结构的损伤识别可作为一个优化问题来处理。本文直接应用频响函数(FRF)进行结构的损伤识别。通过对FRF的主成分分析(PCA)实现数据压缩和特征提取,建立基于压缩FRF的优化目标函数。为了提高算法的收敛速度,以结合局部搜索算法(LS)的遗传算法(GA)为优化工具,并进一步结合子结构识别法来求解。基于桁架的计算结果表明,这种方法具有很好的鲁棒性和识别效果。