Damage study of structures using Hotteling's t2 curve in electromechanical impedance method

Abstract

The recent advances in sensor technologies have led to the daunting task of combining the information and robust decision making for damage detection in structural health monitoring (SHM) due to its capabilities of extracting multiple information. The Electromechanical Impedance (EMI) method employs high frequencies range in assessing the local structural response based on structural health monitoring (SHM). This work describes the quantification of the frequency domain on the Al plate using principal component analysis (PCA) based hotteling’s T² damage curve in describing the behaviour of signal. PCA used to reduce multivariable complex data set to lower dimension in order to reveal simplified statistical patterns. The EMI method used damage metrics as a tool to separate quantitatively or qualitative pre-process data of EMI spectra into classes depending on the damage presence, level and location. The information of sensor’s resistance (R) and conductance (G) is studied in the frequency domain and data fusion is realized at the variable level using fused variable F. The proposed methodology is tested and validated for Al material by creating drilled holes.