Abstract

This study performs a substantial statistical analysis on the effects of damage characteristics, such as position and severity, on a specific damage metric composed of coefficients obtained from the discrete wavelet transform. The results of the analysis serve as the basis for developing more sophisticated and optimized damage-identifying methods in structural health monitoring. For the methodological procedure, a carbon fiber reinforced polymer beam was numerically modeled using the finite element method where the damage was induced considering a local reduction of stiffness in different positions and levels of severity. The numerical methodology was validated with modal experimental tests. Results obtained through analysis of variance and design of experiments indicated the sensitivity of each factor and the response surface provided the best combination of factors for each response. Knowing the sensitivity and interaction of these parameters is essential for the correct formulation of damage indexes or other techniques in damage identification. The results obtained in this paper can be used as a guideline for engineers when developing new wavelet transform-based strategies for structural health monitoring.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call