Monitoring of Damage in Composite Structures Using an Optimized Sensor Network: A Data-Driven Experimental Approach

Due to the complexity of the fracture mechanisms in composites, monitoring damage using a vibration-based structural response remains a challenging task. This is also complex when considering the physical implementation of a health monitoring system with its numerous uncertainties and constraints, including the presence of measurement noise, changes in boundary and environmental conditions of a tested object, etc. Finally, to balance such a system in terms of efficiency and cost, the sensor network needs to be optimized. The main aim of this study is to develop a cost- and performance-effective data-driven approach to monitor damage in composite structures and validate this approach through tests performed on a physically implemented structural health monitoring (SHM) system. In this study, we combined the mentioned research problems to develop and implement an SHM system to monitor delamination in composite plates using data combined from finite element models and laboratory experiments to ensure robustness to measurement noise with a simultaneous lack of necessity to perform multiple physical experiments. The developed approach allows the implementation of a cost-effective SHM system with validated predictive performance.

Automated summary

Monitoring damage in composites using vibration-based structural response is challenging due to the complexity of fracture mechanisms. Implementing a health monitoring system is also complex, considering uncertainties and constraints like measurement noise and changing environmental conditions. This study aims to develop a cost-effective data-driven approach to monitor damage in composites and validate it with tests on a physically implemented SHM system. By combining finite element models and laboratory experiments, the developed approach enables the implementation of a cost-effective SHM system with validated predictive performance.