Damage localization using electromechanical impedance technique based on inverse implementation

Structural Health Monitoring (SHM) systems allow maintenance planning which reduces maintenance costs. These systems may also allow lifetime extension of structures which allows reduction in lifecycle costs of structures. As a result, SHM systems have been of interest to researchers in civil and mechanical engineering sectors. The SHM system should be ideally low cost, suitable for continuous monitoring, and able to detect small levels of damage. The sensitivity to small levels of damage is even more important for composite structures as composites have low damage tolerance. Electromechanical Impedance (EMI)-based techniques have been proposed due to their sensitivity to low levels of damage. Furthermore, due to their high frequency of operation they are immune to ambient noise. Most of the studies using the EMI technique were limited to damage detection only. Recently some studies have used a network of sensors for EMI measurements and used them for damage localization. This paper follows on this trend of using a network of sensors for damage localization using EMI technique. A novel inverse algorithm is proposed for the damage localization. For the inverse technique, an analytical model for the correlation of distance and angle with the damage index (DI) is developed. This model is then used to generate a database of the DIs for a large number of damage scenarios. The measured DIs from the structure are then compared with this database and the scenario with the least mean error in the DIs is the predicted damage scenario. The algorithm takes into consideration the modeling errors and the measurement errors through regularization and in turn allows more accurate damage localization. The method is applied on different damage scenarios on an unidirectional composite sample. Some sensitivity studies such as the performance of the method using different fits have been carried out. In addition, the accuracy of damage localization of the proposed technique has been compared to a localization technique found in literature. The results show that indeed the proposed algorithm works and improves the damage localization using the EMI technique.