Classification of turbulent flow patterns with fuzzy clustering

A pattern recognition technique for the identification of critical points in turbulent flows based on fuzzy C-means clustering is presented. The technique deals with, in particular, the problem that experimental data is normally obtained with coarsely spaced sensors making the identification of critical points uncertain. The technique was applied to hot-wire data from two different turbulent flows; one is the highly ordered wake of a mesh strip, while the second is the wake of a circular cylinder, which has a higher level of disorder. Foci and saddle points in the velocity fields were accurately detected and classified.     

Large-eddy simulation of turbulent flow in a stirred tank driven by a Rushton turbine

Large-eddy simulation (LES) of mixing process in a baffled tank was presented. The impeller rotation was modeled using the sliding mesh technique. In this study the CFD code was used for simulation of a standard vessel agitated by a 6-blade Rushton turbine and results were evaluated in terms of the predicted flow field, power number, mean velocity components, mixing time, turbulent kinetic energy and turbulent dissipation rate using published experimental data. Subsequently, the effects of varying injection position of the passive scalar have been investigated. The results show that LES is a reliable tool to investigate the unsteady behavior of the turbulent flow in stirred tank.     

Filter shape dependence and effective scale separation in large-eddy simulations based on relaxation filtering

The influence of the filter shape on the effective scale separation and the numerical accuracy of large-eddy simulations based on relaxation filtering (LES-RF) is investigated. The simulation of the turbulent flow development of a high-Reynolds number low-subsonic compressible mixing layer is performed using the LES-RF procedure, for discrete filters of order 2–10. A reference solution is first obtained using high-order numerical algorithms and shows a good agreement with experimental data found in the literature. Discrete filters of order 2, 4, 6, 8 and 10 are then considered to study the influence of the filter shape on numerical results. The 2nd-order scheme turns out to be too dissipative and prevents the emergence of unsteady motions within the mixing layer. For higher order schemes, from 4th- to 10th-order, the flow solutions are turbulent but exhibit mean flows and turbulent intensities depending on the filter. The investigation of the one-dimensional kinetic energy spectra then shows that the 4th-order filter may still be too dissipative whereas large scales remain unaffected using the 6th-, 8th- and 10th-order filters. A further study of the kinetic energy spectra nonetheless demonstrates that the effective spatial bandwidth of the LES increases with the order of the filtering scheme. Simulations using the 6th-, 8th- and 10th-order filters, with mesh sizes chosen to provide the same effective LES cut-off wavenumber, are performed and yield similar results. It is hence found that the value of the effective LES cut-off wavenumber, rather than to the filter shape itself, is mainly responsible for the discrepancies between the flow statistics obtained using different filters. One may conclude that filter shape independence is consequently achieved in the present LES of a mixing layer.     

A Parallel Overset Grid High-Order Flow Solver for Large Eddy Simulation

This work describes the development and validation of a parallel high-order compact finite difference Navier–Stokes solver for application to large-eddy simulation (LES) and direct numerical simulation. The implicit solver can employ up to sixth-order spatial formulations and tenth-order filtering. The parallelization of the solver is founded on the overset grid technique. LES were then performed for turbulent channel flow with Reynolds numbers ranging from Re _τ=180 to 590, and flow past a circular cylinder with a transitional wake at Re _D =3900. The channel flow solutions were obtained using both an implicit LES (ILES) approach and a dynamic sub-grid scale model. The ILES method obtained virtually identical solutions at half the computational cost. The original vector and new parallel solvers produce indistinguishable mean flow solutions for the circular cylinder. Repeating the cylinder simulation on a much finer mesh resulted in significantly better agreement with experimental data in the near wake than the coarse grid solution and other previous numerical studies.     

Implementation of synthetic turbulence inlet for turbomachinery LES

Large eddy simulation (LES) has the potential to model complex separated flows, where Reynolds Averaged Navier–Stokes (RANS) based methods often fail. An important aspect of LES is specifying correlated turbulent fluctuations at the inlet boundary. This is particularly important in turbomachines, where turbulence length scale and intensity play a key role in the correct prediction of component performance.In this work, a method is implemented into an unstructured Computational Fluid Dynamics (CFD) solver to impose correlated turbulent fluctuations in a compressible form. It is shown that compressibility effects are particularly important in turbomachinery and must be taken into account. The method uses a pre-processing method to generate a cube of isotropic, homogeneous turbulence. The velocity fluctuations so obtained are used to determine a fluctuating Mach number in order to evaluate the instantaneous total pressure and temperature fluctuations at domain inlet. In the authors knowledge this is one of the first attempts to define correlated fluctuations in a compressible form.The method is successfully applied to two turbomachinery related flows. Firstly, the jet flow from a propelling nozzle is investigated. Following this, the flow over a low pressure (LP) turbine blade is predicted. Results from the LES simulations show that modifications to the inlet conditions can significantly affect flow development. For the jet, changes in the shear layer and peak shear stress are shown, important in the context of high frequency sideline noise generated by the jet. Despite what is suggested in the literature the differences in shear stresses are important also in a non-swirling jet.For the LP turbine, incoming turbulent fluctuations modify the onset of transition and the extent of separation bubble. Without imposed turbulence fluctuations, loss is overpredicted by up to 50%. Moreover it is important to use a compressible solver. Despite the fact that the majority of the results proposed in literature on LP turbine is using incompressible solvers, the difference in terms of pressure coefficient, Cp, is comparable to turbulence contribution.     

A zonal RANS-LES method to determine the flow over a high-lift configuration

High Reynolds number flows are particularly challenging problems for large-eddy simulations (LES) since small-scale structures in thin and often transitional boundary layers are to be resolved. The range of the turbulent scales is enormous, especially when high-lift configuration flows are considered. For this reason, the prediction of high Reynolds number flow over the entire airfoil using LES requires huge computer resources. To remedy this problem a zonal RANS-LES method for the flow over an airfoil in high-lift configuration at Re_c=1.0×10⁶ is presented. In a first step, a 2D RANS solution is sought, from which boundary conditions are formulated for an embedded LES domain, which comprises the flap and a sub-part of the main airfoil. The turbulent fluctuations in the boundary layers at the inflow region of the LES domain are generated by controlled forcing terms, which use the turbulent shear stress profiles obtained from the RANS solution. The comparison with an LES solution for the full domain and with experimental data shows likewise results for the velocity profiles and wall pressure distributions. The zonal RANS-LES method reduces the computational effort of a full domain LES by approx. 50%.     

A large-eddy simulation method for low Mach number flows using preconditioning and multigrid

Large-eddy simulation (LES) has become one of the major tools to investigate the physics of turbulent compressible and incompressible flows. At low Mach numbers the performance of LES codes developed for the conservation equations of compressible fluids deteriorates due to the presence of two different time scales associated with acoustic and convective waves. In many subsonic turbulent flows low Mach number regions exist, which require large integration times until a fully developed flow is established. In such cases, the efficiency of algorithms for compressible flows can be improved considerably by low Mach number preconditioning methods. In this paper an efficient method of solution for low subsonic flows is developed based on an implicit dual time stepping scheme combined with low Mach number preconditioning and a multigrid acceleration technique. To validate the efficiency and the accuracy of the method, large-eddy simulations of turbulent channel flow at Re_τ = 590 and cylinder flow at Re = 3900 are performed for several Mach numbers and the data are compared with numerical and experimental findings from the literature. The speedup compared to a purely explicit approach is in the range of 6–40.     

Studies of shock/turbulent shear layer interaction using Large-Eddy Simulation

Shock/shear/turbulence interactions are simulated using Large-Eddy Simulation (LES) with a new localized subgrid closure approach. Both normal and oblique shocks interactions with turbulence are considered. The LES methodology adopted here combines a hybrid numerical scheme that switches automatically and locally between a shock-capturing scheme and a low-dissipation high-order central scheme.The fundamental role of the diffusion of turbulent kinetic energy by pressure fluctuations in the problem of normal shock/isotropic turbulence interaction is stressed in the DNS study, and accounted for in the closure model. The study of the interaction between two oblique shocks and a turbulent shear layer shows that the turbulence evolution is mostly affected by two competing phenomena. An amplification of the turbulent levels occurs downstream of the interaction, and the mixing layer growth rate is significantly increased. However, the integrated production of turbulent energy across the mixing layer is reduced, and the increase in mixing is found to be localized in space, the turbulent statistics quickly relaxing to their undisturbed levels. Furthermore, the increase in vorticity from the compression of the mixing layer remains small, unaffected by the presence of turbulent and coherent structures.     

Large-eddy simulation of multi-component compressible turbulent flows using high resolution methods

The ability of a finite volume Godunov and a semi-Lagrangian large-eddy simulation (LES) method to predict shock induced turbulent mixing has been examined through simulations of the half-height experiment [Holder and Barton. In: Proceedings of the international workshop on the physics of compressible turbulent mixing, 2004]. Very good agreement is gained in qualitative comparisons with experimental results for combined Richtmyer-Meshkov and Kelvin-Helmholtz instabilities in compressible turbulent multi-component flows. It is shown that both numerical methods can capture the size, location and temporal growth of the main flow features. In comparing the methods, there is variability in the amount of resolved turbulent kinetic energy. The semi-Lagrangian method has constant dissipation at low Mach number, thus allowing the initially small perturbations to develop into Kelvin-Helmholtz instabilities. These are suppressed at the low Mach stage in the Godunov method. However, there is an excellent agreement in the final amount of fluid mixing when comparing both numerical methods at different grid resolutions.     

Study of jet precession, recirculation and vortex breakdown in turbulent swirling jets using LES

Large eddy simulations (LES) are used to investigate turbulent isothermal swirling flows with a strong emphasis on vortex breakdown, recirculation and instability behaviour. The Sydney swirl burner configuration is used for all simulated test cases from low to high swirl and Reynolds numbers. The governing equations for continuity and momentum are solved on a structured Cartesian grid, and a Smagorinsky eddy viscosity model with the localised dynamic procedure is used as the sub-grid scale turbulence model. The LES successfully predicts both the upstream first recirculation zone generated by the bluff body and the downstream vortex breakdown bubble. The frequency spectrum indicates the presence of low frequency oscillations and the existence of a central jet precession as observed in experiments. The LES calculations well captured the distinct precession frequencies. The results also highlight the precession mode of instability in the center jet and the oscillations of the central jet precession, which forms a precessing vortex core. The study further highlights the predictive capabilities of LES on unsteady oscillations of turbulent swirling flow fields and provides a good framework for complex instability investigations.     

Numerical and Physical Instabilities in Massively Parallel LES of Reacting Flows

LES of reacting flows is rapidly becoming mature and providing levels of precision which can not be reached with any RANS (Reynolds Averaged) technique. In addition to the multiple subgrid scale models required for such LES and to the questions raised by the required numerical accuracy of LES solvers, various issues related to the reliability, mesh independence and repetitivity of LES must still be addressed, especially when LES is used on massively parallel machines. This talk discusses some of these issues: (1) the existence of non physical waves (known as ‘wiggles’ by most LES practitioners) in LES, (2) the effects of mesh size on LES of reacting flows, (3) the growth of rounding errors in LES on massively parallel machines and more generally (4) the ability to qualify a LES code as ‘bug free’ and ‘accurate’. Examples range from academic cases (minimum non-reacting turbulent channel) to applied configurations (a sector of an helicopter combustion chamber).     

Application of Compact Schemes to Large Eddy Simulation of Turbulent Jets

We present 3-D large eddy simulation (LES) results for a turbulent Mach 0.9 isothermal round jet at a Reynolds number of 100,000 (based on jet nozzle exit conditions and nozzle diameter). Our LES code is part of a Computational Aeroacoustics (CAA) methodology that couples surface integral acoustics techniques such as Kirchhoff's method and the Ffowcs Williams– Hawkings method with LES for the far field noise estimation of turbulent jets. The LES code employs high-order accurate compact differencing together with implicit spatial filtering and state-of-the-art non-reflecting boundary conditions. A localized dynamic Smagorinsky subgrid-scale (SGS) model is used for representing the effects of the unresolved scales on the resolved scales. A computational grid consisting of 12 million points was used in the present simulation. Mean flow results obtained in our simulation are found to be in very good agreement with the available experimental data of jets at similar flow conditions. Furthermore, the near field data provided by the LES is coupled with the Ffowcs Williams–Hawkings method to compute the far field noise. Far field aeroacoustics results are also presented and comparisons are made with experimental measurements of jets at similar flow conditions. The aeroacoustics results are encouraging and suggest further investigation of the effects of inflow conditions on the jet acoustic field.     

Initial conditions for Large Eddy Simulations of piston engine flows

The exact knowledge of the flow in a piston engine chamber is of vital interest in engine design. These flows feature 3D highly unsteady turbulent phenomena combined with combustion processes. Large Eddy Simulations (LES) appear to be a promising way to simulate them. However, computing several engine cycles results in excessive computational costs. Therefore, a different approach, namely the single-cycle strategy (SC), is to perform several simulations just of those parts of one engine cycle that are of interest. In this study, non-reacting LES is undertaken with a SC strategy for the injection of gas into a tumbling motion. Measured data are used for both the initialization and the validation of the computations. In addition, the initial field is varied using a proper orthogonal decomposition analysis on the experimental data to mimic realistic cycle-to-cycle variations of the tumble before the injection. Satisfactory results are obtained by using a simple procedure for creating initial conditions based on experimental data. By changing the initial field, it is demonstrated that initial conditions have a very significant influence on the LES results. This influence may restrict the use of SC strategies in favour of mutiple-cycle computations.     

Neural networks based subgrid scale modeling in large eddy simulations

In this paper a multilayer feed-forward neural network (NN) is used as subgrid scale (SGS) model in a large eddy simulation (LES). The NN was previously off-line trained using numerical data generated by a LES of a channel flow at Re_τ=180 with Bardina's scale similar (BFR) SGS model. Results show the ability of NNs to identify and reproduce the highly nonlinear behavior of the turbulent flows, and therefore the possibility of using NN techniques in numerical simulations of turbulent flows.     

LES of turbulent square jet flow using an MRT lattice Boltzmann model

In this paper we consider the application of multiple-relaxation-time (MRT) lattice Boltzmann equation (LBE) for large-eddy simulation (LES) of turbulent flows. The implementation is discussed in the context of 19-velocity (D3Q19) MRT-LBE model in conjunction with the Smagorinsky subgrid closure model. The MRT-LBE-LES is then tested in the turbulent square jet flow case. We compare MRT-LBE-LES results with (a) single-relaxation-time (SRT) or BGK LBE results and (b) experimental data. Computed results include the distribution of centerline mean streamwise velocity, jet spread, and spanwise profiles of mean streamwise velocity in the near-field region. The phenomenon of axis switching is investigated. The advantages of MRT over SRT are demonstrated. Reasonable agreement between our numerical results and experimental data demonstrate that the MRT-LBE is a potentially viable tool for LES of turbulence.     

Horizontal Large Eddy Simulation of Stratified Mixing in a Lock-Exchange System

This paper presents analytical and numerical results for two new anisotropic modifications of the Rational and Clark-α LES models. The main difference from their standard form is that in this study horizontal (as opposed to isotropic) spatial filtering is used, which is appropriate for turbulent mixing in stratified flows. We present several mathematical results regarding the horizontal Rational and Clark-α LES models. We also present numerical experiments that support the analytical developments and show that both horizontal LES models perform better than their standard, isotropic counterparts in approximating mixing in a 3D lock-exchange problem at Reynolds number Re=10,000.     

A numerical study of temporal shallow mixing layers using BGK-based schemes

A numerical study of the temporal shallow mixing layers is performed. The depth-averaged shallow water equations are solved by the finite volume method based on the Bhatnagar–Gross–Krook (BGK) equation. The filtering operation is applied to the governing equations and the well-known Smagorinsky model for the subgrid-scale (SGS) stress is employed in order to present a large eddy simulation (LES). The roll-up and pairing processes are clearly shown and the corresponding kinetic energy spectra are calculated. The effects of the Froude number and the bottom friction are numerically investigated. It is shown that the growth rate of the mixing layer decreases as the Froude number increases, which is very similar to the compressible mixing layers when considering the effects of the Mach number. The numerical results also indicate that the increase in bottom friction can enhance the stability of the flows, which is physically reasonable and consistent with the theoretical and experimental findings.     

An Irregularly Portioned Lagrangian Monte Carlo Method for Turbulent Flow Simulation

A novel computational methodology, termed “Irregularly Portioned Lagrangian Monte Carlo” (IPLMC) is developed for large eddy simulation (LES) of turbulent flows. This methodology is intended for use in the filtered density function (FDF) formulation and is particularly suitable for simulation of chemically reacting flows on massively parallel platforms. The IPLMC facilitates efficient simulations, and thus allows reliable prediction of complex turbulent flames. Sample results are presented of LES of both premixed and non-premixed flames via this method, and the results are assessed via comparison with laboratory data.     

LES of intermittency in a turbulent round jet with different inlet conditions

Large eddy simulation (LES) is a promising technique for accurate prediction of turbulent free shear flows in a wide range of applications. Here the LES technique has been applied to study the intermittency in a high Reynolds number turbulent jet with and without a bluff body. The objective of this work is to study the turbulence intermittency of velocity and scalar fields and its variation with respect to different inlet conditions. Probability density function distributions (pdf) of instantaneous mixture fraction and velocity have been created from which the intermittency has been calculated. The time averaged statistical results for a round jet are first discussed and comparisons of velocity and passive scalar fields between LES calculations and experimental measurements are seen to be good. The calculated probability density distributions show changes from a Gaussian to a delta function with increased radial distance from the jet centreline. The effect of introducing a bluff body into the core flow at the inlet changes the structure of pdfs, but the variation from Gaussian to delta distribution is similar to the jet case. However, the radial variation of the intermittency indicates differences between the results with and without a bluff body at axial locations due the recirculation zone created by the bluff body.     

Parallel Simulations of Swirling Turbulent Flames

The feasibility of using massively paralleled computations as an engineering design tool is evaluated. A parallel Large-Eddy Simulation (LES) algorithm which simulates turbulent reacting flows using a space and time-accurate method, is used to model the complex flow found inside a realistic gas-turbine combustor. The parallelization philosophy and its implementation as a platform-independent solver is discussed. A performance analysis is carried out to determine the communication and storage requirements, and the associated overhead. As a case study, the LES methodology is used for a parametric investigation of swirl effects on the turbulent reacting flow in the gas-turbine.