1,721,015 research outputs found
Assessing the sensitivity of turbine cascade flow to inflow disturbances using direct numerical simulation
Noise due to unsteady flow past trailing edges
This paper presents two-dimensional direct numerical simulations (DNS) of noise generated at trailing edges (TE) with zero thickness. The simulations are conducted specifying either no-slip or slip walls in order to investigate viscous effects. In both cases, small amplitude disturbances are introduced close to the inflow boundary that serve as pressure disturbances at the TE. DNS data reveals that the unsteady Kutta condition is not satisfied, irrespective of the wall boundary condition. However, it appears that the validity of the unsteady Kutta condition is not essential for making an accurate prediction of the far field noise. The far field pressure is predicted as a function of the surface pressure difference using a 2-D modification of Amiet's classical theory, and compared with the far field pressure computed directly. Directivity plots provide evidence that the presence of boundary layers and noise generated by an unsteady wake in the no-slip cases lead to smearing of individual lobes, and that the downstream pointing lobes in no-slip wall cases are probably due to nonlinear noise generation in the wake. The simulations are conducted using a high-order accurate numerical method which is free of upwinding, artificial dissipation or any form of explicit filtering, and employs a novel boundary treatment
Numerical investigation of turbulent supersonic axisymmetric wakes
Numerical experiments are conducted of turbulent supersonic axisymmetric wakes at Mach number M= 2.46 and Reynolds number, based on free-stream velocity and base diameter, ReD = 1 x 10(5). Direct numerical simulations (DNS) are used to study the effect of approach flow conditions, and of specific azimuthal modes, on the near-wake behaviour. To that end, DNS are performed with laminar and turbulent approach boundary layers, and additional turbulent approach flow DNS with reduced circumferential size are conducted to deliberately eliminate certain azimuthal/helical modes. DNS with turbulent approach flow show an increased turning angle and increased growth of the separating shear layer, leading to a shorter recirculation region, a stronger recompression shock system, and ultimately good agreement with experimental data at considerably higher Reynolds number. A similar wake structure is found for laminar and turbulent inflow conditions, giving further evidence of the wake structure being a consequence of the global near-wake instabilities and not a result of upstream conditions. Stability analyses of two-dimensional basic states are carried out by computing the temporal pulse response using forced Navier–Stokes simulations to investigate which azimuthal modes are dominant for fully turbulent wakes and how the stability behaviour is influenced by the choice of basic state. Using the time- and azimuthally averaged data from three-dimensional DNS with turbulent inflow as basic state, an absolute instability of the axisymmetric mode was found and helical modes m= 4, 5, 6 were found to be linearly most unstable, in contrast to results obtained earlier using an axisymmetric flow solution as the basic state. The addition of a turbulence viscosity in the forced DNS retains most of the stability characteristics but reduces the wavenumber of the linearly most-amplified modes
Flow and noise predictions for a tandem cylinder configuration using novel hybrid RANS/LES approaches
The performance of several novel hybrid RANS/LES methodologies for accurate flow and noise predictions of the NASA Tandem Cylinder Experiment are investigated. Simulations are performed using three different hybrid RANS/LES methodologies which employ different techniques to transform the baseline RANS model into a turbulence resolving subgrid scale model. The approaches investigated are the Scale-Adaptive-Simulation which computes the subgrid dissipation-rate from a transport equation that is sensitive to the v. Karman length scale, the IDDES approach and a modified Flow-Simulation-Methodology (FSM) which both rely on replacing the modeled turbulent length scale with a measure of the local grid spacing. Since we expect that the predictions in the RANS and LES region depend on the capabilities of the baseline RANS model, the hybrid RANS/LES approaches used in this study are based on an Explicit-Algebraic-Stress Model. These models are assessed and compared with formulations based on a traditional two-equation model. The simulations are performed with the open-source finite volume code OpenFOAM and acoustic predictions are obtained using an acoustic analogy based on Curle's theory. The IDDES and FSM approaches were found to predict the hydrodynamic field in very good agreement with reference data but showed some deficiencies in capturing the higher harmonics in the acoustic spectra and broadband noise levels at high Strouhal numbers. The SAS approaches lack in accuracy for predicting the hydrodynamic field but resolve the higher harmonics in the acoustic spectra
Acoustic and hydrodynamic analysis of the flow around an aerofoil with trailing-edge serrations
Direct numerical simulations of the flow around a NACA-0012 aerofoil are conducted, employing an immersed boundary method to represent flat-plate trailing-edge extensions both with and without serrations. Properties of the turbulent boundary layer convecting over the trailing edge are similar for both cases. For cases with serrations, the trailing-edge noise produced by the flow over the aerofoil is observed to decrease in amplitude, and the frequency interval over which the noise reduction occurs differs depending on the serration length. The directivity and spanwise coherence of the trailing-edge noise appears largely unaffected by the serrations. The hydrodynamic behaviour in the vicinity of the trailing-edge extensions is investigated. The streamwise discontinuity imparted upon the turbulent flow by the straight trailing edge can clearly be observed in statistical quantities, whereas for the serrated case no spanwise homogeneous discontinuities are observed. The trailing-edge serrations appear to break up the larger turbulent structures convecting into the wake, and to promote the development of horseshoe vortices originating at the serrations themselves
Efficient parallel computing with a compact finite difference scheme
This paper proposes an efficient parallel computing approach based on a high-order accurate compact finite difference scheme in conjunction with a conventional domain decomposition method and MPI libraries. The proposed parallel computing approach consists of two major features: (a) a newly developed compact finite difference scheme with extended stencils containing halo points around subdomain boundaries, and (b) a predictor–corrector type implementation of a compact filter that effectively suppresses spurious errors from the subdomain boundaries. The current work employs three halo cells for the inter-node communication, based on which the coefficients of the new compact scheme at the subdomain boundaries are optimized to achieve as high level of resolution and accuracy as the interior compact scheme provides. Also, an optimal set of cut-off wavenumbers of the compact filter that minimizes spurious errors is suggested. It is shown that the level of errors from the proposed parallel calculations lies within the same order of magnitude of that from the single-domain serial calculations. The overall accuracy and linear stability of the new parallel compact differencing-filtering system are confirmed by grid convergence tests and eigenvalue analyses. The proposed approach shows a substantial improvement with respect to existing methods available
Direct numerical simulations of trailing-edge noise generated by turbulent boundary-layers
Direct numerical simulations (DNS) are conducted of turbulent flow passing an infinitely thin trailing edge (TE).The objective is to investigate the turbulent flow field in the vicinity of theTE and the associated broadband noise generation.To generate a turbulent boundary layer a short distance from the inflow boundary, a technique is employed which exploits the dynamical features of the outer and inner part of the boundary layer.The DNS data are Fourier transformed in time and point spectra at various locations are used to identify representative frequencies. Several flow quantities are visualized for these frequencies and compared to predictions made with an acoustic analogy and a modified version of Amiet's classical trailing edge noise theory. A rapid increase in skin friction and a decrease in boundary layer thickness is observed at the trailing edge, consistent with analytical predictions using triple deck theory, leading to suppression of very low frequencies as found from point spectra
Influence of free stream effects on jet noise generation and propagation within the Goldstein acoustic analogy for fully turbulent jet inflow boundary conditions
Large-scale compressible-flow direct numerical simulations
Direct numerical simulation (DNS) of turbulent flows began in the early 1970s with the study of incompressible isotropic turbulence by Orszag and Patterson (Phys Rev Lett 28(2), 76–79, 1972) [8], and compressible DNS were not conducted until the 1980s.</p
- …
