基于车身绕流的低雷诺数湍流模型改进研究

本文将汽车绕流模块化为各典型局部流动,通过常用湍流模型对各典型局部流动进行数值模拟,结果验证了湍流模型对转捩的捕捉能力是准确模拟汽车绕流的关键. 在分析汽车绕流分离及转捩机理的基础上,优化了稳态和瞬态求解方法,改进了湍流模型对转捩的预测能力,进而提高了湍流模型在汽车流场模拟上的精度. 针对汽车绕流的稳态问题,将流线曲率因子及 响应阈值引入 LRN $k$-$\varepsilon $ (low Reynolds number $k$-$\varepsilon $) 模型,获得了一种能够更准确预 测转捩的改进低雷诺数湍流模型 (modified LRN $k$-$\varepsilon $),改善了原模型对湍流耗散率的过强依赖性及全应力发展预测不足等问题;针对汽车绕流瞬态求解,通过分析 RANS/LES 混合湍流模型的构造思想及特点,引入约束大涡模拟方法,结合本文提出的改进的 LRN $k$-$\varepsilon $ 湍流模型,提出了一种能准确捕捉转捩现象 的转捩 LRN CLES 模型. 分别将改进的模型用于某实车外流场和风振噪声仿真中,通过 Ansys Fluent 求解器计算,并将计算结果与常用湍流模型的仿真结果、HD-2 风洞试验结果和实车道路实验结果进行对比,表明改进后的湍流模型能够更准确模拟复杂实车的稳态和瞬态特性,为汽车气动特性的研究提供了可靠理论依据及有效数值解决方法.      

RESEARCH ON IMPROVEMENTS OF LRN TURBULENCE MODEL BASED ON FLOW AROUND AUTOMOBILE BODY 1)

本文将汽车绕流模块化为各典型局部流动,通过常用湍流模型对各典型局部流动进行数值模拟,结果验证了湍流模型对转捩的捕捉能力是准确模拟汽车绕流的关键. 在分析汽车绕流分离及转捩机理的基础上,优化了稳态和瞬态求解方法,改进了湍流模型对转捩的预测能力,进而提高了湍流模型在汽车流场模拟上的精度. 针对汽车绕流的稳态问题,将流线曲率因子及 响应阈值引入 LRN $k$-$\varepsilon $ (low Reynolds number $k$-$\varepsilon $) 模型,获得了一种能够更准确预 测转捩的改进低雷诺数湍流模型 (modified LRN $k$-$\varepsilon $),改善了原模型对湍流耗散率的过强依赖性及全应力发展预测不足等问题;针对汽车绕流瞬态求解,通过分析 RANS/LES 混合湍流模型的构造思想及特点,引入约束大涡模拟方法,结合本文提出的改进的 LRN $k$-$\varepsilon $ 湍流模型,提出了一种能准确捕捉转捩现象 的转捩 LRN CLES 模型. 分别将改进的模型用于某实车外流场和风振噪声仿真中,通过 Ansys Fluent 求解器计算,并将计算结果与常用湍流模型的仿真结果、HD-2 风洞试验结果和实车道路实验结果进行对比,表明改进后的湍流模型能够更准确模拟复杂实车的稳态和瞬态特性,为汽车气动特性的研究提供了可靠理论依据及有效数值解决方法.     

Computation of Vortex Shedding and Radiated Sound for a Circular Cylinder: Subcritical to Transcritical Reynolds Numbers

The Lighthill acoustic analogy combined with Reynolds-averaged Navier–Stokes flow computations are used to investigate the ability of existing technology to predict the tonal noise generated by vortex shedding from a circular cylinder for a range of Reynolds numbers (100 < Re < 5 million). Computed mean drag, mean coefficient of pressure, Strouhal number, and fluctuating lift are compared with experiment. Two-dimensional calculations produce a Reynolds number trend similar to experiment but incorrectly predict many of the flow quantities. Different turbulence models give inconsistent results in the critical Reynolds number range (Re≈ 100000). The computed flow field is used as input for noise prediction. Two-dimensional inputs overpredict both noise amplitude and frequency; however, if an appropriate correlation length is used, predicted noise amplitudes agree with experiment. Noise levels and frequency content agree much better with experiment when three-dimensional flow computations are used as input data. Received 5 May 1998 and accepted 28 September 1998     

Unified turbulence models for LES and RANS,FDF and PDF simulations

A review of existing basic turbulence modeling approaches reveals the need for the development of unified turbulence models which can be used continuously as filter density function (FDF) or probability density function (PDF) methods, large eddy simulation (LES) or Reynolds-averaged Navier–Stokes (RANS) methods. It is then shown that such unified stochastic and deterministic turbulence models can be constructed by explaining the dependence of the characteristic time scale of velocity fluctuations on the scale considered. The unified stochastic model obtained generalizes usually applied FDF and PDF models. The unified deterministic turbulence model that is implied by the stochastic model recovers and extends well-known linear and nonlinear LES and RANS models for the subgrid-scale and Reynolds stress tensor.      

The compressible Coanda wall jet—an experimental study of jet structure and breakaway

An underexpanded jet issuing from a convergent slot and blowing over a surface of convex streamwise curvature was studied experimentally. The jet was confined between side walls, with the slot aspect ratio varying between 40 and 6, but tests showed that in the area of interest close to the slot the flow was effectively two-dimensional. The ratio of slot width to the radius of curvature of the downstream surface varied between 0.05 and 0.33. The main techniques used were Schlieren and shadowgraph to show the jet structure, and surface flow visualization which revealed areas of separation and reattachment. Surface static pressures were also measured on the curved surface. The curved jet proved to have a shock cell structure similar to that of a plane jet. However, the cell structure disappeared more rapidly as the outer shear layer grew more quickly due to the destabilizing effect of the curvature on the turbulence in the shear layer. Even at modest upstream jet pressures, a separated region on the Coanda surface became evident. This region was characterized by a stagnant constant pressure part followed by a region of strongly reversed flow before reattachment took place. The separation was caused by the compression at the end of the first shock cell, with reattachment taking place where expansion in the second cell started. The separated region grew rapidly as the upstream pressure was increased, until, finally, reattachment failed to occur and the jet suddenly broke away from the surface. This work is related to studies of the Coanda flare, where the jet is axisymmetric. The high level of turbulence causes rapid entrainment of air and so gives us clean combustion. However there should be more general application to devices that use the Coanda effect, varying from fluidic devices to blown jet flaps on wings.     

Large Eddy Simulations Using the Subgrid-Scale Estimation Model and Truncated Navier–Stokes Dynamics

We describe a procedure for large eddy simulations of turbulence which uses the subgrid-scale estimation model and truncated Navier–Stokes dynamics. In the procedure the large eddy simulation equations are advanced in time with the subgrid-scale stress tensor calculated from the parallel solution of the truncated Navier–Stokes equations on a mesh two times smaller in each Cartesian direction than the mesh employed for a discretization of the resolved quantities. The truncated Navier–Stokes equations are solved through a sequence of runs, each initialized using the subgrid-scale estimation model. The modeling procedure is evaluated by comparing results of large eddy simulations for isotropic turbulence and turbulent channel flow with the corresponding results of experiments, theory, direct numerical simulations, and other large eddy simulations. Subsequently, simplifications of the general procedure are discussed and evaluated. In particular, it is possible to formulate the procedure entirely in terms of the truncated Navier–Stokes equation and a periodic processing of the small-scale component of its solution. Received 27 April 2001 and accepted 16 December 2001     

Computation of a drastic flow pattern change in an annular swirling jet caused by a small decrease in inlet swirl

This paper investigates the flow pattern change in an annular jet caused by a sudden change in the level of inlet swirl. The jet geometry consists of an annular channel followed by a specially designed stepped‐conical nozzle, which allows the existence of four different flow patterns as a function of the inlet swirl number. This paper reports on the transition between two of them, called the ‘open jet flow high swirl’ and the ‘Coanda jet flow.’ It is shown that a small sudden decrease of 4% in inlet swirl results in a drastic and irreversible change in flow pattern. The objective of this paper is to reveal the underlying physical mechanisms in this transition by means of numerical simulations. The flow is simulated using the unsteady Reynolds‐averaged Navier–Stokes (URANS) approach for incompressible flow with a Reynolds stress turbulence model. The analysis of the numerical results is based on a study of different forces on a control volume, which consists of the jet boundaries. The analysis of these forces shows that the flow pattern change consists of three different regimes: an immediate response regime, a quasi‐static regime and a Coanda regime. The simulation reveals that the pressure–tangential velocity coupling during the quasi‐static regime and the Coanda effect at the nozzle outlet during the Coanda regime are the driving forces behind the flow pattern change. These physical mechanisms are validated with time‐resolved stereo‐PIV measurements, which confirm the numerical simulations. Copyright © 2008 John Wiley & Sons, Ltd.     

Pollution dispersion analysis using the puff model with numerical flow field data

The computations of the flowfield and pollutant dispersion over a flat plate and the Russian hills of various slopes are described. The Gaussian plume and the puff model have been used to calculate concentration of pollutant. The Reynolds-averaged unsteady incompressible Navier–Stokes equation with low Reynolds k– model has been used to calculate the flowfield. The flow data of a flat plate and the Russian hills from Navier–Stokes equation solutions has been used as the input data for the puff model. The computational results of flowfield agree well with experimental results of both a flat plate and Russian hills. The concentration prediction by the Gaussian plume model and the Gaussian puff model also agrees fairly well with experiments.     

Computational analysis of impinging shock-wave boundary layer interaction under conditions of incipient separation

The interaction of an oblique shock wave with a turbulent boundary layer under conditions of incipient separation is analyzed by means of large-eddy simulation (LES) and Reynolds-averaged Navier–Stokes (RANS) turbulence models, with the objective to explore their predictive capabilities, in particular with respect to the unsteady features of the interaction. Consistent with earlier direct numerical simulations, we have found that the flow dynamics in the interaction zone is characterized by strong intermittency associated with the formation of scattered spots of flow reversal near the nominal position of the reflected shock. Comparison with experimental results (at much larger Reynolds number) show that the qualitative features of the interaction are predicted reasonably well by both LES and RANS models. RANS models supplemented with a semi-empirical closure are also found to provide reasonable estimate of the fluctuating pressure loads at the wall.     

Studies of the unsteady supersonic base flows around three afterbodies

Unsteady supersonic base flows around three afterbodies, cylindrical （Cy）, boattailed （BT） and three-step （MS）, are investigated in this paper. Reynolds-averaged Navier-Stokes （RANS） and two RANS/LES （large-eddy simulation） hybrid methods, detached eddy simulation （DES） and delayed-DES （DDES）, are used to predict the base flow characteristics around the baseline Cy afterbody. All the RANS and hybrid methods are based on the two-equation SST （shear-stress transport） model with compressible corrections （CC）. According to the comparison of measurements, both DES and DDES can produce more satisfactory results than RANS. RANS can only present the ＂stable＂ flow pat- terns, while the hybrid methods can demonstrate unsteady flow structures. DDES and DES results are little different from one another although the latter exhibits better agreement with the experiment. DES is taken to investigate the 5° BT and three-step afterbodies. The mean flow data and the instantaneous turbulent coherent structures are compared against available measurements.     

Assessment of Subgrid-Scale Models for Decaying Compressible MHD Turbulence

Large eddy simulation method is formulated for study of compressible magnetohydrodynamic turbulence and assessment of different subgrid-scale models as applied for decaying case is performed. The filtered equations of compressible magnetopause using the mass-weighted filtering operation are obtained. Mass-weighted filtered equations for large-scale turbulent component comprise subgrid-scale terms and five models for closure of the subgrid-scale terms are suggested. In present paper the obtained results of numerical computations for large eddy simulation are compared with the results of direct numerical simulation of three-dimensional compressible magnetohydrodynamic turbulence. Assessment of five subgrid-scale models of large eddy simulation for MHD flows is fulfilled. The comparisons between large eddy simulation and direct numerical simulation are carried out regarding the temporal evolution of the global quantities kinetic and magnetic energy, cross helicity and the spectra of kinetic and magnetic energy.     

The development of an axisymmetric curved turbulent wall jet

An experimental study has been carried out of the low speed Coanda wall jet with both streamwise and axisymmetric curvature. A single component laser Doppler technique was used, and by taking several orientations at a given point, values of the three mean velocities and five of the six Reynolds stresses were obtained. The lateral divergence and convex streamwise curvature both enhanced the turbulence in the outer part of the jet compared with a plane two-dimensional wall jet. The inner layer exhibited a large separation of the positions of maximum velocity and zero shear stress. It was found that the streamwise mean velocity profile became established very rapidly downstream of the slot exit. The profile appeared fairly similar at later downstream positions, but the mean radial velocity and turbulence parameters showed the expected nonself preservation of the flow. Removal of the streamwise curvature resulted in a general return of the jet conditions toward those expected of a plane wall jet. The range and accuracy of the data may be used for developing turbulence models and computational techniques for this type of flow.     

Direct Numerical Simulation of Statistically Stationary One- and Two-Phase Turbulent Combustion: A Turbulent Injection Procedure

A new turbulent injection procedure dedicated to fully compressible direct numerical simulation (DNS) or large eddy simulation (LES) solvers is proposed. To avoid the appearance of spurious acoustic waves, this method is based on an accurate tracking of the turbulent structures crossing the boundary at the inlet of the domain. A finite difference DNS solver has been coupled with a spectral simulation in which a statistically stationary homogeneous turbulence evolves to provide fluctuating boundary conditions.A new turbulence forcing method, dedicated to spectral solvers, has been developed as well to control the major properties of the injected flow (turbulent kinetic energy, dissipation rate and integral length scale). One-dimensional Navier–Stokes characteristic boundary conditions extended to non-stationary flows are coupled with the injection procedure to evaluate is potential in four various configurations: spatially decaying turbulence, dispersion of vaporizing sprays, propagation of one- and two-phase V-shape turbulent flames.     

Towards a new partially integrated transport model for coarse grid and unsteady turbulent flow simulations

A new two-equation model is proposed for large eddy simulations (LESs) using coarse grids. The modeled transport equations are obtained from a direct transposition of well-known statistical models by using multiscale spectrum splitting given by the filtering operation applied to the Navier–Stokes equations. The model formulation is compatible with the two extreme limits that are on one hand a direct numerical simulation and on the other hand a full statistical modeling. The characteristic length scale of subgrid turbulence is no longer given by the spatial discretization step size, but by the use of a dissipation equation. The proposed method is applied to a transposition of the well-known k- statistical model, but the same method can be developed for more advanced closures. This approach is intended to contribute to non-zonal hybrid models that bridge Reynolds-averaged Navier–Stokes (RANS) and LES, by using a continuous change rather than matching zones. The main novelty in the model is the derivation of a new equation for LES that is formally consistent with RANS when the filter width is very large. This approach is dedicated to applications to non-equilibrium turbulence and coarse grid simulations. An illustration is made of large eddy simulations of turbulence submitted to periodic forcing. The model is also an alternative approach to hybrid models. PACS 47.27.Eq     

Optimization of an inclined elliptic impinging jet with cross flow for enhancing heat transfer

This work presents a parametric study and optimization of a single impinging jet with cross flow to enhance heat transfer with two design variables. The fluid flow and heat transfer have been analyzed using three-dimensional compressible Reynolds-averaged Navier–Stokes equations with a uniform heat flux condition being applied to the impingement plate. The aspect ratio of the elliptic jet hole and the angle of inclination of the jet nozzle are chosen as the two design variables, and the area-averaged Nusselt number on a limited target plate is set as the objective function. The effects of the design variables on the heat transfer performance have been evaluated, and the objective function has been found to be more sensitive to the angle of inclination of the jet nozzle than to the aspect ratio of the elliptic jet hole. The optimization has been performed by using the radial basis neural network model. Through the optimization, the area-averaged Nusselt number increased by 7.89% compared to that under the reference geometry.     

Modeling air entrainment and transport in a hydraulic jump using two-fluid RANS and DES turbulence models

Both RaNS (Reynolds-averaged Navier–Stokes) and DES (Detached Eddy Simulation) type turbulence models were used in conjunction with a two-fluid model of bubbly flow and a new subgrid air entrainment model to predict air entrainment and transport in a hydraulic jump. It was found that the void fraction profiles predicted by both methods are in agreement with the experimental data in the lower shear layer region, which contains the air bubbles entrained at the so-called toe of the hydraulic jump. In contrast, in the upper roller region behind the toe, the averaged results of the DES turbulence model gives accurate predictions while a RaNS turbulence model does not. This is because the DES turbulence model successfully captures the strong fluctuations on the free surface which allows it to entrain air near the top of the roller region. In contrast, RaNS type turbulence model results in a steady, smooth interface which fails to capture the wave-induced bubble sources in that region. To our knowledge, this study is the first successful quantitative numerical simulation of the overall void fraction profiles in a hydraulic jump.     

URANS Computations of Shock-Induced Oscillations Over 2D Rigid Airfoils: Influence of Test Section Geometry

The present article deals with recent numerical results from on-going research conducted at ONERA/DMAE regarding the validation of turbulence models for unsteady transonic flows, in which the mechanism of the shock-wave/boundary-layer interaction is important. The main goal is to predict the onset and extent of shock induced oscillations (SIO) that appear over the suction side of two-dimensional rigid airfoils and lead to the formation of unsteady separated areas. Computations are performed with the ONERA object-oriented software "elsA", using the URANS-type approach. In this approach, the unsteady mean turbulent flow is resolved using the standard Reynolds-averaged Navier–Stokes (RANS) equations and closure relationships involving standard transport equation-type models without any explicit modification due to unsteadiness. Applications are provided and discussed for two different test cases, one of which is rather well documented for CFD validation and described by mean-flow, phase-averaged and fluctuating data. Results demonstrate the importance of modelling the upper and lower walls of the test section when trying to capture SIO as precisely as possible with 2D computations, even though the adaptation of wind tunnel walls had been carefully considered. Finally, turbulence validation has been performed using one- and two-transport equation-type models, one of them resulting from in-house investigations for other turbulent flows applications.     

Numerical study of gas-cyclone airflow: an investigation of turbulence modelling approaches

A numerical study of unsteady single-phase vortical flow inside a cyclone is presented. Two different geometric configurations have been considered, with the goal of assessing several different turbulence modelling approaches for this class of problem. The models investigated include three Reynolds-averaged Navier–Stokes models: a commonly used two-equation eddy-viscosity model, a differential Reynolds stress model (DRSM) and an eddy-viscosity model sensitised to rotational and curvature (RC) effects which was recently developed and implemented into a commercial CFD (computational fluid dynamics) code by the authors. Results were also obtained using large eddy simulation (LES). The computational results are analysed and compared with available experimental data. The RC-sensitised eddy-viscosity model shows significant improvement over the standard eddy-viscosity model. The RC-sensitised model, DRSM and LES model predictions of the mean flowfield are in good agreement with the experimental data. The results suggest that curvature- and rotation-sensitive eddy-viscosity models may provide a practical alternative to more computationally intensive approaches.     

A Stochastic Backscatter Model for Grey-Area Mitigation in Detached Eddy Simulations

A new stochastic backscatter model is proposed for detached eddy simulations that accelerates the development of resolved turbulence in free shear layers. As a result, the model significantly reduces so-called grey areas in which resolved turbulence is lacking after the computation has switched from a Reynolds-averaged Navier–Stokes simulation to a large eddy simulation. The new stochastic model adds stochastic forcing to the momentum equations with a rate of backscatter from the subgrid to the resolved scales that is consistent with theory. The effectiveness of the stochastic model is enhanced by including spatial and temporal correlations of the stochastic forcing for scales smaller than the cut-off scale. The grey-area mitigation is demonstrated for two canonical test cases: the plane free shear layer and the round jet.