Spectral analysis and cross-correlation techniques for photon counting measurements on fluid flows

Photon counting techniques for the measurement of turbulent fluid flows are analyzed, and it is shown that considerable errors can result if conventional Fourier methods are used to transform count correlation records from LDV systems onto the frequency domain. Two alternative schemes are presented that overcome this difficulty. The first involves the use of high resolution spectral techniques to transform count autocorrelation records, and the second makes use of the count cross-correlation between signals from two detectors. A theoretical analysis is presented for the count cross-correlation process, and experiments in air flow show turbulence levels predicted by the two methods to be in close agreement.