Prognostics and health management of composite structures under multiple impacts through electromechanical behavior and a particle filter

Self-sensing techniques are restricted to monitoring the various types of damage caused during repeated impact testing, and only a few studies have investigated the prognostics of carbon fiber reinforced plastics (CFRPs); in these studies, the electrical resistance of CFRPs was gauged in real time during multiple-impact testing. Therefore, real-time prognostics and health management using electromechanical behavior data obtained from CFRP structures under repeated impact testing are proposed herein. The health condition of the CFRP is observed in real time during impact testing using mechanical and electromechanical behavior data. Further, the types of failure observed during impact testing are investigated using real-time self-sensing data. Moreover, a particle filter is used for predicting the electromechanical behavior and the remaining number of useful impacts during repeated impact testing conducted using a physics-based prog-nostics tool. The applicability of the proposed methodology was confirmed by monitoring and predicting impact damage growth on the wind-turbine blade within a 5% prediction error. An advanced-condition- based monitoring technique with the diagnostics and prognostics of the current health state was designed successfully, and an application of the introduced method was demonstrated for industrial use.(c) 2022 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/).

Automated summary

This study proposes a method for real-time prognosis and health management using electromechanical behavior data obtained from CFRP structures under repeated impact testing. The types of failure observed during impact testing are investigated using real-time self-sensing data. A particle filter is used for predicting the electromechanical behavior and the remaining number of useful impacts during repeated impact testing. The proposed methodology was successfully applied to monitor and predict impact damage growth on a wind-turbine blade with a 5% prediction error.