How accurate are ultrasonic anemometers, calibrated in a laminar wind tunnel, under turbulent conditions?

Abstract

Two ultrasonic anemometers, namely the Gill HS-100 and the Anemoment TriSonica Mini, were placed in the VKI L-1A wind tunnel to assess their accuracy and precision. A comparison was made between these instruments and a Prandtl probe, using a range of wind speeds up to 20 m/s, eighteen azimuth angles (wind directions) at 20° intervals, five tilt angles ranging from -10 to 10°, and two levels of turbulence. As the normal level of turbulence in the wind tunnel, determined by hot wire anemometry, is less than 1%, a grid was installed to allow a second level of turbulence of 6.1% to be evaluated. The analysis shows that the Gill HS-100 accurately measures wind speed in low turbulence conditions in accordance with the manufacturer’s specifications. However, in a turbulent atmosphere, the measured wind speeds are not only overestimated by 2-8% compared to the Prandtl probe but also show a systematic shift compared to the low turbulence case. As expected from the manufacturer’s specifications, the TriSonica Mini is less accurate and precise than the HS-100. In a low-turbulence environment, it fails to meet the manufacturer’s specifications. In a more turbulent atmosphere, however, it matches and even outperforms the HS-100, making it suitable for real-world applications. However, it did show reduced accuracy at non-horizontal wind angles, requiring careful consideration for drone-based measurements. More complex calibrations involving multiple turbulence levels could be considered for both anemometers. However, the feasibility of this approach remains uncertain in light of the results obtained.