Thermal Image Velocimetry

Abstract

A method for measuring the two-dimensional distribution of wind velocity vectors near a surface exposed to solar radiation, by tracking brightness temperature images instead of particle images, is proposed. It is based on time-sequential thermography with the algorithm used for particle image velocimetry. This thermal image velocimetry (TIV) was tested on a full-scale building wall covered by polystyrene boards attached side-by-side over a vertically elongated area measuring 22.2 m by 2.73 m. A thermal infrared camera was installed 8 m from the test wall to capture the wall-surface temperature at 30 Hz frequency. A sonic anemometer was also installed 35 mm from the surface used for validation of the TIV. The advection velocity estimated from thermal infrared imagery had a linear relationship with the wind velocity measured by the sonic anemometer, irrespective of the wind speed and direction. This linear slope was multiplied by the advection velocity of the thermal infrared image to rescale it to the wind velocity, and the term ‘TIV velocity’ was then used. A histogram and power spectra of the TIV velocity showed quantitatively good agreement with the velocity measured by the sonic anemometer, except for the high-frequency region of the spectra, where the TIV velocity was overestimated compared with that of the sonic anemometer. The method was also tested on ground covered by artificial turf to demonstrate its application to a horizontal plane with a wider area, extending for more than 80 m by 60 m.