| Περίληψη: | The scope of this thesis is to reveal, identify and investigate promising routes for developing multifunctional composite structures with damage sensing capabilities towards the development of integrated non-destructive inspection (NDI) and structural health monitoring (SHM) capabilities. Two routes were identified and selected for further investigation; the enhancement of multifunctionality of composite systems through the use of nanotechnology and the development of novel damage sensing techniques based on the electrical properties of the composites. Both were selected in the view of better integration of NDI/SHM functionality.
Initially, the use of nanotechnology to control the properties of polymer foam systems as part of multifunctional sandwich composite structures has been validated and proven feasible. Electrically conductive nano-composite foams were developed at a varying range of densities. The level of conductivity was controlled by the CNT concentration. The underlying mechanisms for the formation of the CNT network were analyzed closely and as a result a practical processing-structure-property map was proposed for describing the material capabilities.
Towards the development of self-sensing functionality, the established electrical conductivity was studied as an index of strain and damage in nano-composite foams. A 1D electrical resistance sensing approach was followed during mechanical compression testing (Electrical Resistance Change Method – ERCM). The variations of the recordings revealed the strain and damage formation within the material. The distinct regions in the response curve were correlated to micro-structural strain and damage mechanisms, effectively demonstrating the capability to develop multifunctional structural materials with self-sensing capabilities.
In the direction of novel sensing techniques, answering to the need for an electrical based approach that is transferable and scalable to 2D and even further to more complex 3D shell geometries, the concept of Electrical tomography and the ET inverse problem solution were proposed and studied as a tool for NDI and damage assessment of composite materials. The approach is based on the inherent electrical conductivity of the material and leads the step from conventional 1D electrical sensing to 2D imaging, offering a viable route for utilizing electrical sensing techniques in real applications. The technique delivers a conductivity change map which corresponds to the studied geometry and changes in conductivity are correlated with real damage. For each map, two features were extracted through automated algorithms; the Centre of Interest and the corresponding Region of Interest. It was found that the sensing principle was sensitive enough to extremely small variations of conductivity (less than 0.1% of the inspected area). The post-processing and feature extraction technique was effective in indicating to the location of the developed damage.
Taking a step further, the knowledge of the composite material microstructure and expected failure modes have been translated and formulated into an additional mathematical constraint. The formulation is applied to constrain the solution of the ERT inverse problem greatly enhancing the solution and the damage localization in ERT.
A concept for merging the two proposed routes for the development of multifunctional structures is then proposed and investigated. The establishment of a conductive 3D network of CNT is exploited using the previously formulated tomographic approaches. The development of a continuous artificial 3D CNT network within the matrix of a structural composite has been shown to provide electrical conductivity to previously non-conductive composites. This 3D network is used for the damage assessment of the composite as any structural damage introduced discontinuities in the 3D network which are located using tomographic approaches. ERT was applied providing 2D imaging for the NDI of composites based on electrical measurements taken from a CNT doped GFRP, effectively sensing variations in the electrical fields and identifying the location of the induced damage.
Having shown that ERT can provide useful information on the health/damage state of composite materials, a step further was taken to identify the required steps to apply ERT to larger composite components with more complex geometry. The studied cylindrical component provided a case study to demonstrate the procedure for applying ERT to existing structural components while formulating the ERT inverse problem to cover cases that could not be covered with the up-to-date formulations.
In parallel, an alternative approach for post-processing the electrical measurements taken using ET was proposed; the dipole technique. This observational technique was described, formulated and applied to the available experimental data. It was concluded that the dipole technique is effective in delivering a swarm of Damage Estimation Locations which formed convex Region of Interest, effectively locating the damage with small relative error and large inspection area suppression (reaching over 90%).
Finally, a practical electrical-based approach was formulated for monitoring a real case of aeronautical component. The goal for monitoring the integrity of composite patch repair on an aluminium component was achieved by proposing a mapping technique to translate distributed 1D electrical measurements to a 2D damage probability map. The proposed approach was formulated theoretically and verified on experimental level under simulated service conditions. It was concluded that the technique can effectively identify the location of damage which was verified by thermographic imaging techniques. This final approach essentially bridges the area between 1D ERCM techniques on specimen level and the ERT approach proposed in this thesis.
|
|---|
