A Novel Structural Health Monitoring Method for Full-Scale CFRP Structures

Abstract

Polymer-based composite materials are used to manufacture ultra-light panels, massively introduced in the last decades in the aeronautical industry, as a potential solution to reduce the energy consumption and combustion emissions. Safety being the main concern in aeronautical applications, structural monitoring is essential. However, the currently used survey techniques are not usually adapted to these new materials, and potential improvements are constantly being investigated. New Structural Health Monitoring (SHM) systems have emerged during the last years as an interesting option. A SHM system is based on a precise damage detection strategy, and yields information about structural integrity. Moreover, in spite of their exceptionally high mechanical performances, polymer-based composite materials are more vulnerable than metallic alloys when confronted to aggressive environmental ageing factors. The main motivation of this thesis is then to set the basic foundations of a novel, complete, reliable, robust, fast and low-cost SHM system to be applied to full-scale composite structures. In order to address the ageing characterization problem, a standard method for artificial weathering has been developed. This thesis has focused on the influence of three factors: temperature, relative humidity and UV radiation, and their synergetic interaction. The goal of artificial ageing is to subject the specimens to controlled ageing protocols, in order to observe changes in the elastic properties. Depending on the protocol, the extent of the ageing can reach a degradation of 15 % of the elastic moduli. Concerning data acquisition, the PVDF technology offers an alternative. Proving the survivability of PVDF sensors was essential to justify their integration in a composite structure. The validation has been carried out by comparing the ability of PVDF sensors to provide quality signals, which can prove to be as accurate as those from non-aged sensors (accelerometers). In order to measure the elastic properties, vibrational behaviour could provide information about the internal properties of composite structure. Indeed, structural stiffness is related to the eigenfrequencies, with the mass as an intermediary. However, the orthotropic nature of composites makes this problem more complex, and requires an identification algorithm. The method adopted in this thesis minimizes the difference between experimental and FE simulation results in order to identify the best estimation of the elastic properties. The eigenfrequencies and the mode shapes vehicle information about the internal health of the structure. Concerning the acquisition of structural data, operational modal analysis (OMA) methods were privileged, because of their attributes of robustness and rapidity. The use of OMA has been justified by the difficulty to classic structural perform tests. The frequency domain decomposition (FDD) variant technique uses the notion of singular value decomposition (SVD) to extract the necessary information exclusively from the output sensors. As part of the feature extraction solution, the optimal placement problem (OSP) has been extensively addressed in the current state-of-the-art. Among the different approaches, the effective independence (EI) method was successfully adapted to this specific case. Taking into account the results and observations in the development of the different tools, a global SHM method has been proposed.