Turbulence Sensor Dynamic Calibration Using Real-Time Spectral Computations

P. G. Mestayer Institut de Mécanique Statistique de la Turbulence, Marseille, France

Search for other papers by P. G. Mestayer in
Current site
Google Scholar
PubMed
Close
,
S. E. Larsen Department of Meteorology and Wind Energy, Risø National Laboratory, Roskilde, Denmark

Search for other papers by S. E. Larsen in
Current site
Google Scholar
PubMed
Close
,
C. W. Fairall Department of Meteorology, Pennsylvania State University, University Park, Pennsylvania

Search for other papers by C. W. Fairall in
Current site
Google Scholar
PubMed
Close
, and
J. B. Edson Department of Meteorology, Pennsylvania State University, University Park, Pennsylvania

Search for other papers by J. B. Edson in
Current site
Google Scholar
PubMed
Close
Restricted access

Abstract

The integration of plug-in Fast Fourier Transform (FFT) boards in data acquisition computers allows a considerable development in the dynamic calibration of turbulence sensors. The spectral transfer function of a fast and sensitive turbulence sensor can be obtained in situ from a slow sensor having an absolute calibration, by computing in real time either the power spectra of the two signals or their complex cross-spectrum. The real-time spectral method allows calibration of sensors with relatively complex responses and, in most cases, nonlinear transfer functions. When used in conjunction with appropriate control and correction algorithms, this method can take care of numerous sources of error such as electronic noise, line pickup, and sensor malfunctions. This study shows that it can be extended to sensor arrays, including X-wire dual-component anemometers.

Abstract

The integration of plug-in Fast Fourier Transform (FFT) boards in data acquisition computers allows a considerable development in the dynamic calibration of turbulence sensors. The spectral transfer function of a fast and sensitive turbulence sensor can be obtained in situ from a slow sensor having an absolute calibration, by computing in real time either the power spectra of the two signals or their complex cross-spectrum. The real-time spectral method allows calibration of sensors with relatively complex responses and, in most cases, nonlinear transfer functions. When used in conjunction with appropriate control and correction algorithms, this method can take care of numerous sources of error such as electronic noise, line pickup, and sensor malfunctions. This study shows that it can be extended to sensor arrays, including X-wire dual-component anemometers.

Save