Impact Damage Prediction for Wave Energy Converters

Abstract

Marine and hydrokinetic energy is of growing interest across the globe because it has the potential to provide a large source of renewable energy from the world’s oceans and rivers. These marine and hydrokinetic devices, such as wave energy converters, must operate remotely in all weather conditions, including severe storms. Thus, these devices can suffer from structural damage affecting their performance and lifespan. Therefore, there is interest in developing structural health monitoring systems that can identify new damage, estimate its severity, and then make a decision to provide crews with a maintenance or control recommendation. In this study, we investigate using the electromechanical impedance response of piezoelectric transducers to actively monitor the structural health of composite materials similar to those used in several marine and hydrokinetic devices. Recurring impact damage experiments were completed on five plates using a test drop stand, consisting of five consecutive impacts at the same location for each plate. Classification and regression methods were evaluated in an attempt to predict impact damage on a new plate. Machine learning algorithms were used on data collected over a frequency range of 10 kHz to 100 kHz for two types of piezoelectric transducers.