Thin film sensor network for condition assessment of wind turbine blades

Existing sensing solutions facilitating continuous condition assessment of wind turbine blades are limited by a lack of scalability and clear link signal-to-prognosis. With recent advances in conducting polymers, it is now possible to deploy networks of thin film sensors over large areas, enabling low cost sensing of large-scale systems. Here, we propose to use a novel sensing skin consisting of a network of soft elastomeric capacitors (SECs). Each SEC acts as a surface strain gage transducing local strain into measurable changes in capacitance. Using surface strain data facilitates the extraction of physics-based features from the signals that can be used to conduct condition assessment. We investigate the performance of an SEC network at detecting damages. Diffusion maps are constructed from the time series data, and changes in point-wise diffusion distances evaluated to determine the presence of damage. Results are benchmarked against time-series data produced from off-the-shelf resistive strain gauges. This paper presents data from a preliminary study. Results show that the SECs are promising, but the capability to perform damage detection is currently reduced by the presence of parasitic noise in the signal.