Abstract

A parametric approach to estimating the acoustic entropy detected over the course of fatigue damage is presented. Information entropy and relative entropy is estimated through a parametric approach where trial probability density functions (PDFs) are fitted to each individual received acoustic signal as the material degrades over the cycles of loading. The PDF that produces the maximum cumulative entropy is selected to model the signals. This selection criterion is due to the fact that the PDF with higher cumulative entropy results in less bias during the selection process. The evolution trends of both information entropy and relative entropy show the stages of fatigue damage observed in the fatigue indicators such as change in hardness. The acoustic entropy has an advantage over the conventional indices of damage as it can be employed directly in the online sensor based structural health monitoring schemes as a diagnosis feature.

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