Autocorrelation Techniques

2 Technique Determining a velocity ACF from experimental data is not a straightforward task, especially for laser Doppler velocimetry (LDV) data. Unlike measurements obtained from a hot-wire probe, the LDV signal is inherently intermittent by nature due to the random arrival of particles at the measurement volume. An advantage of the Poisson distributed arrival times is that the Nyquist criterion need not be satisfied since aliasing is not an issue when randomly sampling (Adrian and Yao, 1987). However, the randomly varying time interval between data points complicates the approach for determining the discretized ACF and introduces a velocity bias (see x2.2). As a result, the “slotting technique” was developed (e.g., Mayo et al., 1974; Gaster and Roberts, 1975) where data pairs are grouped into slots corresponding