To reduce the measurement uncertainties, first the anemometer has to be calibrated at normal ambient conditions in an accredited wind tunnel. Subsequently a measurement in the variable air density wind tunnel is performed at nearly the same temperature and pressure prevailing during the first measurement. The anemometer itself serves as a transfer standard. With both calibrations done, a transfer factor for measurements in the variable air density wind tunnel can be determined.
The advantage of this procedure is that blockage effects are negligible. Only the uncertainties due to repeatability (type ‘A’) and the uncertainty of the basic calibration remain and have to be taken into account. For this report, the following calibration procedure is used. However, this is not a standardized procedure.
During the calibration in the variable air density wind tunnel, the anemometer output is measured for flow speeds between 4 m/s and 15.5 m/s. For each measurement point adequate time is allowed to generate stable flow conditions. After this settling time, the flow speed is kept constant for a sampling interval of at least 30 s. The sampling frequency is at least 4 Hz. To cover rising and falling flow speeds the following sequence is chosen: 4, 8, 12, 15.5, 14, 10, 6 m/s.
(1) Calibration procedure for varying air temperatures
After determining the transfer factor for the anemometer under investigation, the air temperature is decreased down to -20 °C and a calibration run is done. The air temperature is then increased in 5 °C steps up to the maximum air temperature of 40 °C, while the air pressure is kept constant. At each temperature the calibration is done as described above. To decrease the type ‘A’ uncertainty, the minimum duration of the sampling interval is 30 s. For each calibration run a linear regression analysis is performed, resulting in a calibration function with slope and offset for one particular air temperature.
(2) Calibration procedure for varying air pressures
The calibration procedure for varying air pressures is done similar to varying temperatures. After the measurement for the determination of the transfer factor, the air pressure is increased in 50 hPa steps from 700 hPa up to 1100 hPa. To cover an air density of about 0.8 kg/m³ to 1.4 kg/m³, the measurements for different air pressures is done at a fixed air temperature of 10 °C. For each calibration run with variable ambient pressure a linear regression analysis is performed, resulting in a calibration function with slope and offset for each particular ambient air pressure.