Spectrum Estimation of Spatial Velocity Component Pulsations

One of the objectives in the study of turbulent flows is to determine the spectral features of flow parameters. Typically, the longitudinal and transverse components of the velocity vector are parameters for which spectral analysis is carried out. In the present study, the spectrum of pulsations of a turbulent velocity sampled in the spatial domain is considered. An application with the Reynolds number of 3900 is being examined for a turbulent flow past a 3D cylindrical surface. Numerical simulation is performed using ANSYS Fluent commercial code and based on the Navier-Stokes equations. The spatial signal is sampled on the centerline downstream of the flow. The energy of the transversal component of the velocity vector is analyzed using Fourier transforms of a signal sampled in the spatial domain. Two approaches are considered to study the spectral properties of the signal. The traditional Energy Signal Density (ESD) estimation is analyzed using Parseval’s theorem. The alternative Energy Signal Spectra (ESS) estimation approach is based on the local signal energy definition. The energy spectra performances are compared to the “–5/3” law of A.N. Kolmogorov.