| Composite laminated structures are widely used in solid rocket motor,aircraft skin,generator blade,pressure vessel and other industrial fields.Quality inspection is the technical basis to ensure their structural performance and safe service.For the quality inspection of these plate-like structures,non-destuctive testing based on Lamb waves is an effective method,because of the advantages of low attenuation,large detection area and high detection efficiency.However,the propagation characteristics of Lamb waves are very complex,especially when Lamb wave propagates in composite laminated structures,due to the influence of impedance discontinuity.The common problems of Lamb wave detection will occur,such as serious signal dispersion and aliasing,difficulty in multi-modal identification,low imaging resolution of defects.In view of the above problems,detection theory and method based on Lamb wave for composite laminated structures are studied in this paper.For the problem of complex dispersion property,a dispersion analysis method based on signal domain transformation is proposed without any material parameters.The signal is transformed from the time-frequency domain to the distance-wavenumber domain,the relationship between frequency and wavenumber is constructed,and the wavenumber dispersion curve within the bandwidth of the excitation signal is quickly and accurately reconstructed.Besides,the time of flight(ToF)of Lamb waves along different propagating angles is extracted by using the wavelet transform(WT),and the directionality of group velocity is calculated.For AAA plate(Aluminum-Acrylic-Aluminum,the thickness of each layer is 2mm),the wavenumber dispersion curves of A0,S0,A1,S1 Lamb modes are obtained,and the group velocity of A0,S0 Lamb modes are obtained.For the problem of multi-mode identification,a Lamb modes identification method based on dictionary learning is developed.The dispersion property is regarded as a linear-time invariant system,and Lamb waves can be stimulated based on the dispersion knowledge and the propagating distance.The signal features(ToF and maximum energy of WT)are extracted from the above stimulated waves to create a dictionary,and Lamb modes are identified by searching the dictionary.In addition,the estimated parameter is used to reconstruct the corresponding waves.The experimental verification of direct Lamb waves and reflected Lamb waves with propagation distance of 0.3m and 0.5m in AAA plate shows that A0 and S0 modes can be effectively identified,and each wavepacket can be reconstructed.For the problem of signal aliasing caused by the dispersion,a dispersion compensation method based on phase superposition is presented.The linear wavenumber is constructed from the nonlinear wavenumber,and the wavenumber difference is calculated.Using the mode and propagation distance provided by the Lamb modes identification,the phase superposition and waveform correction are performed on the dispersive signal to achieve the purpose of dispersion compensation.The experimental results show that the above four wavepackets can be compensated to the waveform of excitation.This experiment proves the effectiveness of the proposed dispersion compensation method.The probability-based diagnostic imaging is applied to damage visualization by combining the damage trajectory and Gaussian probability distribution.Firstly,the interaction mechanism between Lamb waves and damage is studied.Secondly,the features extracted from the scattered waves are used to determine the ellipse or hyperbola trajectory.Then,the damage probability of each pixel is calculated.Lastly,the probability map of damage existence is obtained by data fusion of multiple transmitter and receiver paths(T-R paths).The typical damage,through-thickness hole and debonding,is set in AAA plate respectively,and the position of damage can be accurately located by using the probability-based diagnostic imaging.This method provides a high-resolution damage imaging method determined by a small number of T-R paths. |
PDF Full Text Request