Performance Comparison of Different Wavelet Families Based on Vibration Damage Detection of a Timoshenko Beam

Abstract

The objective of this work focuses on performance comparison of different wavelet families based on vibration damage detection in a Timoshenko beam structure. Module 1 considers only rotational damaged data only. Module 2 considers the difference data between undamaged and damaged beam structure. Wavelet properties are important factor for wavelet selection process to applicable area. The different types of continuous wavelet transform like daubechies, symlets, Coiflets, Gaussian, dmeyer, morlet, biorsplines, reverse bior are used. The rectangular beam is modeled numerically. The modal analysis is performed in a Timoshenko beam with two ends clamped boundary condition. The natural frequency and mode shapes of different damaged conditions are obtained. Damage is created by diminishing the young's modulus value percentage of single or multiple element, Rotational mode shape data's are effective in damage detection up to damage case 2, and Difference data of rotational mode shape data's are effective all damage cases. Gaussian white noise SNR (Sound Noise ratio) 24 added with difference data, noise affected the damage detection algorithm performances in least damage case 6 and triple damage case 8. Damage localize by sym3, dmey, bior6.8 wavelets are effective comparing to all other wavelet families. The absolute wavelet coefficients are used for damage localization and the maximum absolute wavelet coefficient are intended for identifying the damage severity.