Bayesian-Based Approach for the Thermographic Measurement of Flow Transition on Wind Turbine Rotor Blades

Abstract

The position of the laminar–turbulent flow transition affects the aerodynamic efficiency of wind turbine rotor blades. An established diagnostic tool is infrared thermography, which enables flow visualization on in-service wind turbines, including the detection of the flow transition position. For the first time, the capabilities of a Bayesian-based image evaluation on the basis of previous knowledge are investigated for maximizing the measurement quality in particular for those weather conditions with a low contrast-to-noise ratio. The Bayesian framework is assessed using simulated and measured thermographic images, incorporating a probability distribution of the transition position. Results indicate that utilizing previous knowledge, especially when normally distributed around the true transition position with a standard deviation of 3 px, significantly reduces uncertainty for thermographic images with a contrast-to-noise ratio <7. Additionally, the Bayesian framework enhances the visualization of transition progression along the radial blade axis, yielding a less noisy result. Previous experimental data can be used to reduce uncertainty for erroneous transition position detections. In conclusion, the integration of high-quality previous knowledge through Bayesian inference proves to be effective in lowering the uncertainty of the position measurement of the laminar–turbulent transition on wind turbine rotor blades, with no compromise of the spatiotemporal resolution.