Premium
Generalized Gaussian smoothing for baseline‐free debonding assessment of sandwich panels
Author(s) -
Villalobos A.,
Ruiz R. O.,
Meruane V.
Publication year - 2021
Publication title -
structural control and health monitoring
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.587
H-Index - 62
eISSN - 1545-2263
pISSN - 1545-2255
DOI - 10.1002/stc.2727
Subject(s) - hyperparameter , structural engineering , vibration , gaussian , smoothing , mathematics , computer science , algorithm , engineering , statistics , acoustics , physics , quantum mechanics
The present work proposes a baseline‐free method to assess debonding on composite panels. The method is based on regressive Gaussian processes (GP) used to obtain vibration modes free of noise. A Bayesian scheme is implemented to allow the automatic determination of the model hyperparameters. The estimation of the undamaged vibration modes is carried out by modifying the optimal hyperparameters determined employing the damaged plate. The curvatures associated with the vibration modes are calculated as the second analytical derivative of the kernel function of the GP; thus, the use of numerical methods is avoided. A damage index is obtained comparing the differences between the curvatures of the damaged and undamaged modes. Finally, using k‐means clustering, the damage size and location could be identified. The effectiveness of the proposed method is verified by means of numerical simulations and experimental tests. The predicted damage is compared with the actual damage in terms of size, location, and intersection. Both numerical simulations and experimental tests offer favorable results.