Degree Type


Date of Award


Degree Name

Doctor of Philosophy


Aerospace Engineering


Wind Energy Science, Engineering, and Policy; Engineering Mechanics

First Advisor

Chao Hu

Second Advisor

Stephen Holland


Condition evaluation of large-scale (or mesoscale) structures, including civil, aerospace, and energy structures, is difficult due to their large sizes, complex geometries and lack of economic and scalable sensing technologies capable of detecting, localizing, and quantifying faults over a structure's global area. A key challenge in the monitoring of a mesoscale structure is the need to distinguish between faults in the structure's global (e.g. changing load paths, loss in global stiffness) and local (e.g. crack propagation, composite delamination) conditions. This work presents a flexible sensing skin for the cost-effective monitoring of large-scale structures, with a special focus on the monitoring of wind turbine blades. The use of sensing skins, also termed electronic artificial skins (e-skins) or dense sensor networks (DSNs), for the condition assessment of structures is an emerging technology enabling a broad range of sensors and their associated electronics to be integrated onto a single sheet. Sensing skins offer a logical solution to the local/global detection problem as it allows for the localized and discrete monitoring of a structure over the structure's entire global area, and as such, they mimic the ability of biological skin to detect and localize damage.

This work proposes, develops specialized algorithms for, and experimentally validates a sensing skin based on a soft elastomeric capacitor (SEC). The SEC is a large-area electronic that transduces a structure's strain into a measurable change in capacitance. The SEC is highly scalable due to its low cost and ease of fabrication, and can, therefore, be used for the cost-effective monitoring of large-scale components. When arranged in a network configuration, SECs deployed onto the surface of a structure can be used to reconstruct strain maps. These strain maps can then be used to detect, localize, and quantify damage on a structure. One particularly useful attribute of the SEC is its capability to measure the additive strain of a structure (ɛx+ɛy).

In order to effectively utilize a network of SECs, an algorithm that fuses sensor geometry, along with the sensor locations and measured strain values is proposed and validated. This algorithm allows for the reconstruction of more accurate additive strain maps of a structure without having to increase the number of sensors deployed within a sensing skin. In situations where the structure's unidirectional strain maps are needed, the main challenge is to decompose the SEC's additive strain map (i.e. ɛx+ɛy) into its linear strain components along two orthogonal directions. To address this challenge, two algorithms were developed that leverage a hybrid dense sensor network (HDSN) of SECs and traditional unidirectional strain sensors (e.g. resistive strain gauges and fiber Bragg grated optical sensors) to decompose the additive strain measurements made by the SECs. In cases where quantifying and predicting the health of a structure is the key consideration, algorithms with trackable damage sensitives features must be developed. To this end, this work presents two data fusion techniques that were specially developed for the high-channel-count sensing skin discussed in this work. To experimentally validate the use of an SEC-based sensing skin for the real-time structural health monitoring of wind turbine blades, an experimental validation was conducted by deploying an HDSN consisting of 12 SECs (measuring 38 x 38 mm2) and eight RGSs on the interior of a scaled model wind turbine blade, mounted inside a wind tunnel to simulate an operational environment. The real-time strain maps were then reconstructed and damage sensitive features were then used to track the degrading health of the wind turbine blade under a series of controlled damage cases. Results demonstrate that the SEC-based sensing skin, working in collaboration with the newly proposed algorithms, is able to successfully detect, localize, and quantify damage in a model wind turbine blade. Additionally, these experimental results demonstrate the capability of the SECs to operate in the electromagnetically noisy environment of a wind tunnel, showing that the SEC would be capable of operating inside the similarly noisy environment of a wind turbine blade.

Copyright Owner

Austin Robert Johnson Downey



File Format


File Size

307 pages