NDT Approaches to Optimize Acoustics Based SHM Systems for Anisotropic Composite Structures

Appropriate and reliable monitoring with SHM systems providing a high probability of detection is a challenge in itself. Compared to traditional nondestructive testing (NDT), where a near to endless sampling series can be performed, the sensor pattern for a SHM system is fixed and as such also sampling is limited with regard to the fixed transducer positions. Finding the optimum position of a SHM system’s transducer network is therefore crucial and can only be reasonably achieved through simulation. This situation of appropriateness of the SHM system is aggravated when it comes to structures made of anisotropic materials such as fibre-reinforced materials where the elastic constants may not be known per se. An approach to alleviate this problem is therefore presented. It is based on using novel techniques developed in NDT which have to be used prior to configuring an SHM system optimized to reliably detect a tolerable damage in the composite structure considered. This starts by applying conventional ultrasonic testing, possibly on the basis of phased array transducers and to sample data in a way that the materials’ effective stiffness constants describing anisotropy are based on the full matrix capture (or sampling phased array) and reverse phase matching methods, which allow those constants to be determined on an iterative basis. Once these constants are determined a guided wave based simulation can start which allows wave patterns in an anisotropic materials environment to be generated for undamaged and damaged conditions. Looking then either at non-linear dynamic properties within the material or the difference in guided wave signals between an undamaged and a damaged condition of a structure for a given actuation pattern will allow sensor locations to be determined at the locations where the difference in signals between the undamaged and the damaged condition becomes maximum for a tolerable damage defined. Based on these simulation results the sensor pattern for the SHM system considered to be optimum is determined. The approach is illustrated on different examples performed.