Speaker
Description
Wind turbine rotor blades are continuously exposed to harsh environmental impacts such as erosion and contamination during operation. Even small-scale surface disturbances can induce premature laminar–turbulent transition of the aerodynamic boundary layer, leading to significant performance losses. Therefore, a measurement method is needed to detect emerging local disturbances on the rotor blade surface at their earliest stage, operating contactlessly without blade modification or downtime. The goal is to identify whether the disturbance type is additive or subtractive and to estimate its position and size.
Infrared thermography (IRT) is an established flow visualization method that has been successfully applied to detect the laminar-turbulent transition on wind turbine rotor blades. Previous studies demonstrated that the disturbance-induced wake flow can be visualized and that the wake patterns are sensitive to the disturbance size and position. However, existing investigations have mainly focused on qualitative evaluation or a small set of disturbance configurations. A quantitative assessment of how wake flow patterns relate to disturbance parameters in the field of wind turbine rotor blades is missing. In particular, the feasibility of inverse estimation of disturbance parameters from thermographic wake patterns has not yet been addressed.
Therefore, wind tunnel experiments are conducted on an airfoil with a height adjustable cylindrical disturbance on its surface. Cylindrical surface disturbances are considered as a generic disturbance model to investigate how thermographic wake patterns respond to the disturbance’s position, type, diameter and height. In the present study, the influence of disturbance height is investigated as a first step toward systematic inverse estimation. The airfoil chord length is L = 60 mm and the disturbance is located at 6 mm (0.1 x/L). The disturbance diameter remains fixed (d = 3 mm), while its height is varied between h = 0.25 mm and h = 4.5 mm. IRT measurements for each height are performed at a Reynolds number of of 10⁵ and at angles of attack of 0°, 5°, and 10°.
The experimental results show that, for additive surface disturbances, the thermographically detected wake area can be quantified and the wake area increases with increasing disturbance height at a Reynolds number of 10⁵. This trend is clearly observed at angles of attack of 0° and 5°, enabling a quantitative relationship between the thermographically detectable wake area and the sought-after disturbance height. At an angle of attack of 10°, however, no consistent trend can be identified. A validation using the leave-one-out cross-validation method shows increasing height estimation errors with increasing angle of attack.
Thus, the results show the feasibility of inverse estimation of disturbance parameters based on the evaluation of thermographic wake patterns. In particular, quantitative information on additive surface disturbances is obtainable under certain aerodynamic conditions. However, a cross-sensitivity w.r.t. the angle of attack exists, which must be considered in practical applications. As a result, the present study provides a foundation for the future development of inverse modelling approaches for IRT-based indirect disturbance characterization, including investigations at higher Reynolds numbers.
