Pulsed Eddy Current Testing Research Based On The Analysis Of Characteristics

Defects detection and quantitative analysis of aircraft multi-layered structure is an actual challenge for the aerospace nondestructive testing(NDT) currently. Pulsed eddy current(PEC) testing is a new kind of eddy current testing technique developed in recent years. Owing to some technical advantages such as rich spectrum, simple operation, low cost, it has become the research hotspot in nondestructive testing field. Based on the analysis of state-of-art pulsed eddy current testing technology, new feature analysis methods are developed in this paper by extracting the different characteristics and processing those using different methods.Signal classification based on principal component analysis(PCA) is studied. The first three principal components of pure lift-off signals, pure corrosion signals and corrosion plus lift-off signals are extracted and analyzed. Results show that with PCA method pure lift-off signals can be identified from other signals,and different depth of pure corrosion signals can be recognized. Pure lift-off signals can be recognized completely from other signals combined with K-means algorithm.Lift-off effect reduction is studied based on high-pass filter. A new feature, the peak value of differential signal after high-pass filtering, is proposed to indicate the height of lift-off, which has a clear physical background. Based on this new feature, a method to reduce lift-off effect is proposed and put to experiments. The experimental results show that this method can effectively suppress lift-off effect in pulsed eddy current testing.Dynamic trajectory of feature vector is studied and applied in pulsed eddy current testing. Dynamic trajectories of fast Fourier transform(FFT) results are used in flaw detection and life-off reduction. Dynamic trajectories of amplitudes of two odd harmonic frequencies are applied to detect defects, identify existence of lift-off and measure the depth of the defect. Amplitude and phase of FFT are combined into a feature vector and used to lift-off reduction and significant effects are obtained.Finally, the conclusions of this paper are given and the suggestions for further research are provided.