Local Damage Detection with the Global Fitting Method Using Operating Deflection Shape Data

The present study extends the previous vibration based damage detection method (Yoon et al. in J. Nondestruct. Eval., 2008) by using Operating Deflection Shape (ODS) derived from experimental Frequency Response Function (FRF) data to detect the locations and extent of damage in steel beams, composite beams and plate-like structures. The present method requires to use FRF data obtained only from the damaged structure based on an assumption that the undamaged structure is homogeneous and smooth. The procedure uses the Global Fitting Method (GFM) that fits a smooth and analytic ODS to the measured ODS. Both the analytic and measured ODS are differentiated twice to yield the Curvature Operating Shapes (COS). The Structural Irregularity Index (SII) at each grid point on the structures is the difference between the analytic and measured COS. The procedure is repeated for each frequency in the FRF. Then, SII are averaged over the selected frequencies to obtain the frequency-averaged SII, which are statistically treated to locate damage. The present method is originally developed for one-dimensional (1D) beam models but can be extended further to any two-dimensional (2D) plate-like models by adding the results in the horizontal and/or vertical directions. The present method using experimental FRF successfully identified the locations and extents of the notches in steel beams, delaminations in the composite beams and plates, and dry spots in a composite hull structure. The present method is compared with different damage detection methods and showed improved performance in terms of detecting damage and computation cost. The present method also showed improved performances when ODS rather than mode shapes (MS) data are used.