研究目的
Investigating the application of time-frequency analysis (TFA) by wavelet transform for effective defect features extraction in laser ultrasonic signal processing, especially under low signal-to-noise rate (SNR) conditions.
研究成果
The TFA method, particularly wavelet transform, effectively extracts defect features from laser ultrasonic signals, even under low SNR conditions. It significantly improves the accuracy of defect depth prediction by 7.9dB, offering a promising tool for material evaluation and signal processing in various fields.
研究不足
The study focuses on aluminum plates and may not directly apply to other materials without further validation. The effectiveness of TFA under extremely low SNRs or for defects with complex geometries was not explored.
1：Experimental Design and Method Selection:
The study employs wavelet transform for TFA to analyze laser-generated surface acoustic wave (SAW) signals containing defect features. A simulation model using finite element method (FEM) is established in an aluminum plate with varying surface defect depths.
2：Sample Selection and Data Sources:
The model simulates an aluminum plate with defects of varying depths (0.1mm to 0.9mm) and a fixed width of 0.5mm. A receiving point is set to detect the vertical displacement of the surface.
3：1mm to 9mm) and a fixed width of 5mm. A receiving point is set to detect the vertical displacement of the surface.
List of Experimental Equipment and Materials:
3. List of Experimental Equipment and Materials: The simulation uses parameters of aluminum, including thermal conductive coefficient, thermal wave velocity, material density, and specific heat capacity.
4：Experimental Procedures and Operational Workflow:
The study involves simulating the propagation of SAW in the aluminum plate, analyzing the echo wave signal at the receiving point, and applying TFA to extract defect features under different SNRs.
5：Data Analysis Methods:
The main frequency of initial and echo SAW is analyzed using wavelet transform. The accuracy of defect depth prediction is evaluated by comparing the results with and without TFA.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
