Health Monitoring of Flexible Structures Via Surface-mounted Microsensors: Network Optimization and Damage Detection

Abstract

Flexible and composite structures often develop hidden damages, e.g. cracks or delamination between laminae or phases. Such events could be sensed through embedded structural health monitoring (SHM) systems, but past experimental studies in the literature proved that embedding sensors for SHM purposes may decrease the reliability of the structure, as the sensor acts as an inclusion. In former studies, the authors proposed to adopt a surface-mounted SHM approach based on inertial MEMS (micro electro-mechanical systems) sensors, which has two advantages: it is low cost and also it avoids the afore-mentioned detrimental effects on the endurance limit state of the structure. However, the low accuracy of the MEMS sensors and the type of response that they can measure may hinder an effective monitoring of the structural state; this can be overcome through redundancy and an efficient sensor placement. In this article, an automated approach is discussed for obtaining the optimal topology of a sparse MEMS sensor network. In this regard, the scenarios are assumed unknown in terms of extent and location of stiffness degradation due to damage. The optimal sensor locations are obtained via maximization of the global sensitivity of the measurements to damage. The method can be also implemented in a multi-scale framework, to efficiently handle (micro) sensors, (meso) damaged regions and (macro) structural components of (by far) different sizes. Data related to composite specimens and panels are discussed, with the aim of assessing the identifiability of damage through self-adapting theoretical/numerical (digital) twins of the monitored structures.