Search

This study investigates the particle dispersion characteristics in the turbulent wake of a simplified vehicle model (Ahmed body) for two values of the rear slant angle ϕ (25 ° and 40 °) to study the effect of flow separation. In the experiments (Reynolds number Rel=1.90 × 10 5 ), smoke particles were released from a source and visualized with a laser sheet. Concentration fields were analyzed to calculate the vertical (Sy) and lateral (Sz) smoke spread. The findings indicate that the flow topology and concentration fields in the wake are highly dependent on ϕ. In the ϕ = 40 ° case, separation on the rear slant disrupts the trailing vortices originating on the rear slant edges and significantly alters the wake topology. The growth of vertical smoke spread saturates after the recirculation region. The uniform mixing and absence of trailing vortices concentrate the smoke particles in the model midplane. In the ϕ = 25 ° case, the signatures of trailing vortices were observed in the concentration fields behind the model. The vertical smoke spread is less, and the lateral smoke spread is more compared to the ϕ = 40 ° case. The growth of the smoke spread ( S y , S z ) driven by the trailing vortices persists for a long distance, even after the recirculation region. The disruption of trailing vortices brought about by the flow separation appears to be an important effect driving the vertical smoke spread in the wake. The connection between the turbulent velocity structures and concentration structures will need to be explored with combined velocity and concentration measurements in the wake. [ABSTRACT FROM AUTHOR]

Moving vehicles pick up the particulates/dust on the road surface, and the wake structures behind the vehicle determine the dispersion of these particles. This study aims at understanding the characteristics of particle dispersion in the turbulent wake of a moving vehicle. The flow over a simplified vehicle model known as the Ahmed body is analysed. Experimental measurements and computational simulations of particle dispersion in the near wake of Ahmed body are presented. Unsteady simulations are performed for the prediction of the flow topology and particle dispersion around the Ahmed body, at R e l = 0.7 × 10 5 (Reynolds no. based on vehicle model length 'l'), the rear slant angle of Ahmed body considered is ϕ = 25 o . Governing equations solved are unsteady Reynolds Averaged Navier–Stokes (URANS) equations, and the turbulence model employed was SST k - ω . Experiments involved the release of fog particles (Ethylene-Glycol, d p ≈ 1 μ m ) from a source behind the model. Dispersion of the particles was measured in the wind tunnel using a laser sheet for fog particle visualisation at three downstream locations, X=1, 2, and 3, at R e l = 1.4 × 10 5 . It was observed that the flow structure in the near wake of the model consisted of two counter-rotating spanwise vortices and a pair of trailing vortices. URANS simulations capture both the turbulent flow features in the near wake. Experimental results were compared with the URANS simulation results, and it was found that the dominant turbulent structures (recirculation vortices and trailing vortices) in the near wake contribute to particle dispersion. The dispersion patterns obtained are dependent on the size and strength of turbulent structures. The URANS simulations perform reasonably well in capturing the vertical dispersion of particles. However, they do not accurately estimate the lateral dispersion of smoke particles. The lateral dispersion is underestimated by the URANS simulations. [ABSTRACT FROM AUTHOR]

We present measurements of Eulerian longitudinal velocity autocorrelations in homogeneous, isotropic, high-intensity (∼9%) free-stream turbulence behind an active grid. Spatial correlations are measured using particle image velocimetry as well as with two-point hot-wire anemometry (HWA), while temporal correlations are measured using HWA. The temporal correlations are transformed into spatial correlations by using Taylor's ‘frozen’ hypothesis with both the mean as well as instantaneous velocities. A model relating Eulerian spatial and temporal autocorrelations is also used for this purpose. The differences from the measured spatial correlation resulting from the use of Taylor's hypothesis on the temporal correlation is quantified; even at this moderately high level of turbulent intensity, the result from the use of the instantaneous velocity as convection velocity is practically indistinguishable from that obtained using the mean velocity. Use of the model produces a good agreement between the estimates of the spatial correlation function. A relation between Eulerian spatial and temporal integral scales is also derived. [ABSTRACT FROM PUBLISHER]

Abstract: In this paper we present experimental measurements of buoyancy driven turbulent exchange flow in a vertical pipe (L/d ratios of 9–12). The flow is driven by an unstable density difference across the ends of the pipe, created using brine and distilled water. Away from either end, a fully developed region of turbulence exists with a linear density gradient. Using a mixing length model that accounts for the end effects, we obtain the turbulent scales and flux. The Nusselt number scales like the square root of the Rayleigh number (Nu ∼ Ra 1/2). We give an empirical relation to quantify the end effects and hence calculate the flux of the salt (NaCl) given the aspect ratio of the pipe and the overall density difference across it. [Copyright &y& Elsevier]

We study the stability of surface waves on the radial film flow created by a vertical cylindrical water jet striking a horizontal plate. In such flows, surface waves have been found to be unstable and can cause transition to turbulence. This surface-wave-induced transition is different from the well-known Tollmien–Schlichting wave-induced transition. The present study aims at understanding the instability and the transition process. We do a temporal stability analysis by assuming the flow to be locally two-dimensional but including spatial variations to first order in the basic flow. The waves are found to be dispersive, mostly unstable, and faster than the mean flow. Spatial variation is the major destabilizing factor. Experiments are done to test the results of the linear stability analysis and to document the wave breakup and transition. Comparison between theory and experiments is fairly good and indicates the adequacy of the model. [ABSTRACT FROM AUTHOR]