Turbulent channel flow simulations are performed using second- and fourth-order finite difference codes. A systematic comparison of the large-eddy simulation (LES) results for different grid resolutions, finite difference schemes, and several turbulence closure models is performed. The use of explicit filtering to reduce numerical errors is compared to results from the traditional LES approach. Filter functions that are smooth in spectral space are used, as the findings of this investigation are intended for application of LES to complex domains. Explicit filtering introduces resolved subfilter-scale (RSFS) as well as subgrid-scale (SGS) turbulence terms. The former can be theoretically reconstructed; the latter must be modelled. The dynamic Smagorinsky model, the dynamic mixed model, and the new dynamic reconstruction model are all studied. It is found that for explicit filtering, increasing the reconstruction levels for the RSFS stress improves the mean velocity as well as the turbulence intensities. When compared to LES without explicit filtering, the difference in the mean velocity profiles is not large; however the turbulence intensities are improved for the explicit filtering case.

SUMMARYA high-order wall treatment is proposed and implemented into a Cartesian grid method and the wall treatment is evaluated for incompressible turbulent flows. The Cartesian grid method employs a sequence of locally refined, uniformly spaced, Cartesian grids. In order to achieve a high-order accuracy, a wall treatment procedure has been developed for arbitrarily shaped geometries. The procedure consists of high-order Lagrangian polynomial interpolations and extrapolations for determining the dependent variables around the wall boundaries. The wall treatment procedure and the Cartesian grid method are used together with a highly efficient multi-grid acceleration method and a local grid refinement strategy for optimal distribution of the grid points. The high-order Cartesian grid method is evaluated using test functions as well as for laminar and turbulent flows. The proposed approach maintains the high-order discretization and yields high-order accuracy of the numerical results. Large eddy simulation of a turbulent swirling flow indicates that the high-order wall treatment leads to significantly different results from those calculated using a low-order piecewise constant wall description. The differences in the results are smaller at a low level of turbulence near the inlet region, but become significant in the region far away from the inlet where the turbulence is more intense. In the latter situation the effect of the wall treatment is as important as the choice of the subgrid scale stress model.

Numerical simulation is performed to evaluate flow field and thermal characteristics in heat exchange matrices. The heat exchanger geometry evaluated consists of an isotropic open lattice structure containing group-interconnected, millimeter-scale, thermally conductive ligaments. The Reynolds number range of interest is Re = 3-1000. The flow fields investigated in this paper are both laminar and turbulent. The laminar flow fields are solved using direct numerical simulation (DNS), while turbulent flow fields are solved by large-eddy simulation (LES). The mean velocity profiles and velocity fluctuations are compared to each other and to experimental data for a packed bed of spheres. Friction factor and Colburn j-factor are calculated from the simulated flow fields and averaged in time. They are compared to the algebraic correlations reported in the literature for a packed bed of spheres and to experimental data for a screen laminate.

No abstract

The tonal noise of a notebook radial blower in a free-field environment is investigated using an hybrid method based on unsteady Reynolds Averaged Navier-Stokes simulations and an Ffowcs Williams and Hawking's analogy to compute the acoustic far-field from the wall-pressure fluctuations recorded in the simulations. Incompressible and compressible simulations have been performed to demonstrate the effect of the compressibility on the noise sources in the very constrained environment due to the casing specific design. The complex flow in the blower yields distributed noise sources. The tongue interaction with the blade wakes and the inlet flow distortions are major contributors at the blade passing frequency, while smaller structures in the clearance are important noise sources at higher harmonics. The influence of a daisy obstruction on one of the blower inlet is also investigated. The additional inlet distortion strongly affects the noise levels at the blade passing frequency. The acoustic predictions are compared to experimental measurements and provide good agreement within the variability of the acoustic level measured between blowers of the same conception batch. The compressible simulations provide the best agreement at higher frequency when the compactness limit is overpassed.