Effect of fuzzy C means technique in failure mode discrimination of glass/epoxy laminates using acoustic emission monitoring

Abstract

Acoustic emission is one of the powerful techniques that can be used for in situ structural health monitoring of composite laminates. One of the main issues of AE is to characterize the different damage mechanisms from the detected AE signals. Unsupervised Pattern recognition has been one of the techniques used for the identification of a specific failure mode in composites from Acoustic emission data. Cross ply composite laminate of size 300 × 300 mm is fabricated using Vacuum bag molding. ASTM D3039 Standard tensile specimens are cut from the laminate and these specimens are subjected to uni axial tensile test under Acoustic Emission monitoring. Fast Fourier transform analysis (FFT) and Short Time Fast Fourier Transform (STFFT) analysis are performed on the Wave forms of the AE hit data obtained during the conduct of tensile test to characterize the failure modes in crossply specimens. Fast Fourier Transform enabled calculating the frequency content of each damage mechanism. In this paper Fuzzy C Mean clustering is performed for the AE parameters obtained from the test and the efficiency of this technique is being investigated using FFT AND STFFT analysis.