Spectral analysis of laser velocimeter data with the slotted correlation method
Abstract
This paper presents a theoretical and experimental investigation of the slotted correlation method for determining power spectra from laser velocimeter (LV) measurements. The estimate of the autocorrelation function is analyzed regarding aliasing, variance, and velocity biasing. The bias of the estimate is predictable and increases with increasing width of the lagged time slot Delta tau. When Delta tau is sufficiently large, aliasing occurs and can be detected and avoided by decreasing Delta tau until a stable estimate of the power spectral density is obtained. Expressions for the variance of the estimate of the autocorrelation function are derived and agree with results obtained by numerical experiment. The variance can be decreased by increasing the sampling rate, total sampling time, Delta tau, and the time between restarts of the estimate of the autocorrelation function. When velocity biasing occurs with LV data, the histogram of the number of points per lagged time slot shows a dependency on the autocorrelation function. This dependency can be used as an indication of the presence of velocity biasing.
 Publication:

4th Fluid Mechanics, Plasma Dynamics and Lasers Conference
 Pub Date:
 May 1986
 Bibcode:
 1986fmpd.confR....B
 Keywords:

 Autocorrelation;
 Laser Anemometers;
 Spectrum Analysis;
 Algorithms;
 Bias;
 Fast Fourier Transformations;
 Histograms;
 Instrumentation and Photography