Executive Summary : | Damage in composites can significantly affects the performance of structures, which when undetected can lead to catastrophic failures. These damages can be developed during manufacturing or while service and can manifest in failure mechanisms such as fiber failure, matrix cracking and delamination. This motivates to monitor damage by developing reliable damage detection systems and methods. Health monitoring of composites is traditionally based on vibration, guided waves or embedded sensors rely on external sensors that do not contribute to the performance of structures. Instead, it is preferred to have self-sensing composites that use their stimulus-responsive properties such as piezoresistivity to sense damage without external sensors. Though few composites such as carbon fiber reinforced polymer (CFRP) are inherently piezoresistive, commonly used structural composites such as glass fiber reinforced polymer (GFRP) need addition of fillers to improve their electrical conductivity. This work proposes to dope carbon nanofiber (CNF) to GFRP to make them a self-sensing material. The optimum concentration of CNF that improves both electrical and mechanical properties of GFRP will be studied. The developed self-sensing GFRP will be combined with electrical impedance tomography (EIT) to detect and estimate damage. EIT models the physics of electric passage (using finite element method) through structure to determine the spatial distribution of resistance and thereby defects. The use of EIT is well studied for medical applications, but relatively less studies focused on use of EIT to composites structures. This study will explore use of surface mounted electrodes that is more accessible for composite structures, instead of the more common use of edge mounted electrodes that is assessable for medical applications. Moreover, to better sense damage in thicker laminates, it is proposed to use electrodes on both top and bottom of the laminate. These new studies will involve reformulation of the forward and inverse EIT problems and improve the numerical methods for optimization techniques (genetic or machine learning algorithm) and regularization terms. Improvement in both damage detection and estimation capabilities are expected using the developed techniques. Finally, it is proposed to develop methodology to reliably predict the residual strength of the damaged composite based on the estimated damage from EIT. In summary, this project aims to improve the ability to detect and estimate damage in GFRP composites using EIT. This will be achieved by developing self-sensing GFRP and efficient health monitoring methodologies based on EIT. |