Eigensensitivity analysis of subgrid-scale stresses in large-eddy simulation of a turbulent axisymmetric jet

The study of complex turbulent flows by means of large-eddy simulation approaches has become increasingly popular in many scientific and engineering applications. The underlying filtering operation of the approach enables to significantly reduce the spatial and temporal resolution requirements by means of representing only large-scale motions. However, the small-scale stresses and their effects on the resolved flow field are not negligible, and therefore require additional modeling. As a consequence, the assumptions made in the closure formulations become potential sources of model-form uncertainty that can impact the quantities of interest. The objective of this work, thus, is to perform a model-form sensitivity analysis in large-eddy simulations of an axisymmetric turbulent jet following an eigenspace-based strategy recently proposed. The approach relies on introducing perturbations to the decomposed subgrid-scale stress tensor within a range of physically plausible values. These correspond to discrepancy in magnitude (trace), anisotropy (eigenvalues) and orientation (eigenvectors) of the normalized, small-scale stresses with respect to a given tensor state, such that propagation of their effects can be assessed. The generality of the framework with respect to the six degrees of freedom of the small-scale stress tensor makes it also suitable for its application within data-driven techniques for improved subgrid-scale modeling.     

Subgrid continuum modeling of particle motion in a turbulent flow

An Eulerian continuum approach to modeling the motion of dispersed particles within the framework of the large-eddy simulation method is developed. The approach is based on a kinetic equation for the filtered probability density function for the particle velocity in a turbulent flow. Models for the subgrid turbulent stresses of the dispersed phase are presented.     

Features of a Turbulent Jet at High Supersonic Velocities

The results of modeling a turbulent supersonic jet at M = 5 using large-eddy simulation (LES) are presented. The structural features of turbulence formed in this flow are analyzed. The possibilities of the large-eddy simulation method and the complexities of simulation of the compressibility effects in jet flows at high Mach numbers are considered. Such features of the supersonic jet as the local turbulent shocklets and Mach waves are reproduced numerically. It is shown that in the neighborhood of the jet the trajectories of ejection flow are located along the front of Mach waves. Anisotropic turbulent structures whose longitudinal scale is greater than the transverse scale by an order of magnitude are revealed in the jet. An estimate of the baroclinic effects shows their weak influence on the vorticity generation in the jet flow considered.     

Implementation and Evaluation of an Embedded LES-RANS Solver

In the current work, we present the development and application of an embedded large-eddy simulation (LES) - Reynolds-averaged Navier Stokes (RANS) solver. The novelty of the present work lies in fully embedding the LES region inside a global RANS region through an explicit coupling at the arbitrary mesh interfaces, exchanging flow and turbulence quantities. In particular, a digital filter method (DFM) extracting mean flow, turbulent kinetic energy and Reynolds stress profiles from the RANS region is used to provide meaningful turbulent fluctuations to the LES region. The framework is developed in the open-source computational fluid dynamics software OpenFOAM. The embedding approach is developed and validated by simulating a spatially developing turbulent channel flow. Thereafter, flow over a surface mounted spanwise-periodic vertical fence is simulated to demonstrate the importance of the DFM and the effect of the location of the RANS-LES interface. Mean and second-order statistics are compared with direct numerical simulation (DNS) data from the literature. Results indicate that feeding synthetic turbulence at the LES interface is essential to achieve good agreement for the mean flow quantities. However, in order to obtain a good match for the Reynolds stresses, the LES interface needs to be placed sufficiently far upstream, which in the present case was six spoiler heights before the fence. Further, a realistic spoiler configuration with finite-width in the spanwise direction and inclined at 30 degrees was simulated using the embedding approach. As opposed to the vertical fence case this is a genuinely (statistically) three-dimensional case and a very good match with mean and second-order statistics was obtained with the experimental data. Finally, in order to test the present solver for high sub-sonic speed flows the flow over an open cavity was simulated. A good match with reference data is obtained for mean and turbulence profile comparisons. Tones in the pressure spectra were predicted reasonably well and an overall sound pressure level with a maximum deviation of 2.6 d B was obtained with the present solver when compared with the experimental data.     

Temporal large-eddy simulation of unstratified and stably stratified turbulent channel flows

Recently, Pruett et al. [Pruett, C.D., Gatski, T.B., Grosch, C.E., Thacker, W.D., 2003. The temporally filtered Navier–Stokes equations: properties of the residual stress. Phys. Fluids 15, 2127–2140] proposed an approach to large-eddy simulation (LES) based on time-domain filtering; their approach was termed temporal large-eddy simulation or TLES. In a continuation of their work, Pruett and collaborators tested their methodology by successfully performing TLES of unstratified turbulent channel flow up to Reynolds number of 590 (based on channel half-height and friction velocity) [Pruett, C.D., Thomas, B.C., Grosch, C.E., Gatski, T.B., 2006. A temporal approximate deconvolution model for LES. Phys. Fluids 18, 028104, 4p]. Here, we carefully analyze the TLES methodology in order to understand the role of its key components and in the process compare TLES to more traditional approaches of spatial LES. Furthermore, we extend the methodology to stably stratified turbulent channel flow.     

Adaptive large eddy simulation with moving grids

The paper presents adaptive mesh moving methods for large eddy simulation (LES) of turbulent flows. With this approach, a given number of grid points is redistributed with respect to an appropriately selected criterion. The Arbitrary Lagrangian–Eulerian formulation is applied to solve the governing equation on moving grids employing a collocated finite volume formulation. A dynamic moving mesh partial differential equation based on a variational principle is solved for the corner points of the grid by means of a dedicated solver. Adaptation is performed in a statistical sense so that statistical quantities of interest are employed. Various LES-specific design criteria and combination of them are proposed, such as the time-averaged gradient of streamwise velocity, turbulent kinetic energy and production rate. These are investigated in the framework of elementary and balanced monitor functions. These are tested for the three-dimensional flow in a channel with periodic constrictions. The numerical results are compared to a highly resolved LES reference solution. The independence of the moving mesh method from the initial LES is shown, and its potential to improve the efficient resolution of turbulent flow features is demonstrated.     

Active control of flow separation over an airfoil using synthetic jets

We perform large-eddy simulation of turbulent flow separation over an airfoil and evaluate the effectiveness of synthetic jets as a separation control technique. The flow configuration consists of flow over an NACA 0015 airfoil at Reynolds number of 896,000 based on the airfoil chord length and freestream velocity. A small slot across the entire span connected to a cavity inside the airfoil is employed to produce oscillatory synthetic jets. Detailed flow structures inside the synthetic-jet actuator and the synthetic-jet/cross-flow interaction are simulated using an unstructured-grid finite-volume large-eddy simulation solver. Simulation results are compared with the 2005 experimental data of Gilarranz et al., and qualitative and quantitative agreements are obtained for both uncontrolled and controlled cases. As in the experiment, the present large-eddy simulation confirms that synthetic-jet actuation effectively delays the onset of flow separation and causes a significant increase in the lift coefficient. Modification of the blade boundary layer due to oscillatory blowing and suction and its role in separation control is discussed.     

Large-eddy simulation study of an air flow past a heated square cylinder

A large-eddy simulation (LES) technique has been used for the calculation of an air flow past a heated square cylinder. The LES method is a conventional one with the Smagorinsky eddy-viscosity model, and the computational grid is small enough to be handled by workstations. The computed turbulent flow field quantities agree well with the experimental results reported by Lyn et al. (J Fluid Mech 304:285–319, 1995). Particular attention has been spent to the convective heat transfer and the prediction of the thermal fluctuations in case of a fluid with Pr = 0.71. The LES results agree well with the empirical correlations proposed by Hilpert and Sparrow et al. for the average Nusselt number.     

Deconvolutional artificial-neural-network framework for subfilter-scale models of compressible turbulence

Acta Mechanica Sinica - We establish a deconvolutional artificial-neural-network (D-ANN) approach in large-eddy simulation (LES) of compressible turbulent flow. Filtered variables in the...     

Effect of solid mass flux on anisotropic gas–solid flow in risers determined with an LES-SOM model

Within the framework of the two-fluid approach, gas was treated with a large-eddy simulation and a sub-grid-scale (SGS) turbulent kinetic energy model while particles were treated with a second-order-moment method to describe the anisotropy of the fluctuating velocity. A modified Simonin model was derived for the gas–solid interphase fluctuating energy transfer. The anisotropic gas–solid flow in a circulating fluidized bed was investigated. Predictions were in good agreement with experimental data. The distributions of the second-order moment of particles and SGS-turbulent kinetic energy of gas were simulated at different solid mass fluxes. The effects of the solid mass flux on the particle second-order moment, particle anisotropic behavior, gas SGS-turbulent kinetic energy and gas SGS energy dissipation were analyzed for the circulating fluidized bed.     

Numerical modeling of flow past a volumeless and thin rigid body using direct forcing immersed boundary method

The new capability has been added as the numerical method for modeling volumeless and thin rigid bodies to the direct forcing immersed boundary (DFIB) method. The DFIB approach is based on adding a virtual force to the Navier–Stokes equations of incompressible flow to account for the interaction between the fluid and structures. The volume of a solid function (VOS) identifies the stationary or moving solid structures in a given fluid domain. A new VOS-based algorithm was developed to identify thin, rigid structure boundary points in fluid flow and ensure that the fluid cannot cross through the boundary of a thin rigid structure while moving or stationary. The DFIB method was first validated in a three-dimensional (3D) turbulent flow over a circular cylinder. The large-eddy simulation simulated the turbulent flow scales. The proposed algorithm was tested using a 3D turbulent flow past a stationary and rotating Savonius wind turbine that functions as a thin, rigid body. The validation results showed that the selected DFIB approach, combined with the novel algorithm, could simulate a thin, volumeless, rigid structure that is stationary and rotating in incompressible turbulent flows. The current method is also applicable for two-way fluid-structure interaction problems.     

Turbulence structure in a diabatically heated forest canopy composed of fractal Pythagoras trees

We investigate the turbulent flow through a heterogeneous forest canopy by high-resolution numerical modeling. For this purpose, a novel approach to model individual trees is implemented in our large-eddy simulation (LES). A group of sixteen fractal Pythagoras trees is placed in the computational domain and the tree elements are numerically treated as immersed boundaries. Our objective is to resolve the multiscale flow response starting at the diameter of individual tree elements up to the depth of the atmospheric surface layer. A reference run, conducted for the forest flow under neutral thermal stratification, produces physically meaningful turbulence statistics. Our numerical results agree quantitatively with data obtained from former field-scale LESs and wind tunnel experiments. Furthermore, the numerical simulations resolve vortex shedding behind individual branches and trunks as well as the integral response of the turbulent flow through the heterogeneous forest canopy. A focus is the investigation of the turbulence structure of the flow under stable thermal stratification and in response to the heating of the fractal tree crowns. For the stratified flows, statistical quantities, e.g. turbulent kinetic energy and vorticity, are presented and the turbulent exchange processes of momentum and heat are considered in detail. The onset and formation of coherent structures such as elevated shear layers above the diabatically heated forest canopy are analyzed. For the stably stratified flow, temperature ramps above the forest canopy were simulated in agreement with previous observations. Thermally driven vortices with a typical diameter of the canopy height were simulated when the tree crowns were diabatically heated. The impact of the coherent flow structures on the heat flux is investigated.     

Large-eddy simulation of low-frequency unsteadiness in a turbulent shock-induced separation bubble

The need for better understanding of the low-frequency unsteadiness observed in shock wave/turbulent boundary layer interactions has been driving research in this area for several decades. We present here a large-eddy simulation investigation of the interaction between an impinging oblique shock and a Mach 2.3 turbulent boundary layer. Contrary to past large-eddy simulation investigations on shock/turbulent boundary layer interactions, we have used an inflow technique which does not introduce any energetically significant low frequencies into the domain, hence avoiding possible interference with the shock/boundary layer interaction system. The large-eddy simulation has been run for much longer times than previous computational studies making a Fourier analysis of the low frequency possible. The broadband and energetic low-frequency component found in the interaction is in excellent agreement with the experimental findings. Furthermore, a linear stability analysis of the mean flow was performed and a stationary unstable global mode was found. The long-run large-eddy simulation data were analyzed and a phase change in the wall pressure fluctuations was related to the global-mode structure, leading to a possible driving mechanism for the observed low-frequency motions.      

Computational modeling of turbulent mixing in a jet in crossflow

The jet in crossflow is a configuration of highest theoretical and practical importance, in which the turbulent mixing plays a major role. High-resolution measurements using Particle Image Velocimetry combined with Laser Induced Fluorescence have been conducted and used to validate simulations ranging from simple steady-state Reynolds-averaged Navier Stokes to sophisticated large-eddy simulation. The reasons for the erratic behavior of steady-state simulations in the given case, in which large-scale structures dominate the turbulent mixing, have been discussed. The analysis of intermittency proved to be an appropriate framework to account for the influence of these flow structures on the jet in crossflow, contributing to the explanation of the poor performance of the steady-state simulations.     

A von Neumann–Smagorinsky turbulent transport model for stratified shear flows

A simple subgrid turbulent diffusion model based on an analogy to the von Neumann–Richtmyer artificial viscosity is explored for use in modelling mixing in turbulent stratified shear flow. The model may be more generally applicable to multicomponent turbulent hydrodynamics and to subgrid turbulent transport of momentum, composition and energy. As in the case of the von Neumann artificial viscosity and many subgrid-scale models for large-eddy simulation, the turbulent diffusivity explicitly depends on the grid size and is not based on a quantitative model of the unresolved turbulence. In order to address the issue that it is often not known a priori when and where a flow will become turbulent, the turbulent diffusivity is set to zero when the flow is expected to be stable on the basis of a Richardson/Rayleigh–Taylor stability criterion, in analogy to setting the von Neumann artificial viscosity to zero in expanding flows. One-dimensional predictions of this model applied to a simple shear flow configuration are compared to those obtained using a K–ε model. The density and velocity profiles predicted by both models are shown to be very similar.     

A ghost-cell immersed boundary method for large-eddy simulations of compressible turbulent flows

A methodology to perform a ghost-cell-based immersed boundary method (GCIBM) is presented for simulating compressible turbulent flows around complex geometries. In this method, the boundary condition on the immersed boundary is enforced through the use of ‘ghost cells’ that are located inside the solid body. The computations of variables on these ghost cells are achieved using linear interpolation schemes. The validity and applicability of the proposed method is verified using a three-dimensional (3D) flow over a circular cylinder, and a large-eddy simulation of fully developed 3D turbulent flow in a channel with a wavy surface. The results agree well with the previous numerical and experimental results, given that the grid resolution is reasonably fine. To demonstrate the capability of the method for higher Mach numbers, supersonic turbulent flow over a circular cylinder is presented. While more work still needs to be done to demonstrate higher robustness and accuracy, the present work provides interesting insights using the GCIBM for the compressible flows.     

Spectral-based simulations of particle-laden turbulent flows

In this paper, we discuss the application of spectral-based methods to simulation of particle-laden turbulent flows. The primary focus of the article is on the past and ongoing works by the authors. The particles are tracked in Lagrangian framework, while direct numerical simulation (DNS) or large-eddy simulation (LES) is used to describe the carrier-phase flow field. Two different spectral methods are considered, namely Fourier pseudo-spectral method and Chebyshev multidomain spectral method. The pseudo-spectral method is used for the simulation of homogeneous turbulence. DNS of both incompressible and compressible flows with one- and two-way couplings are reported. For LES of particle-laden flows, two new models, developed by the authors, account for the effect of sub-grid fluctuations on the dispersed phase. The Chebyshev multidomain method is employed for the works on inhomogeneous flows. A number of canonical flows are discussed, including flow past a square cylinder, channel flow and flow over backward-facing step. Ongoing research on particle-laden LES of inhomogeneous flows is briefly reported.     

Towards high-fidelity erosion prediction: On time-accurate particle tracking in turbomachinery

Erosion and fouling caused by ingested particles causes performance degradation and safety issues in turbomachinery components. Simulating these processes is a complex multiphysics and multiscale problem which has not reached a satisfactory level of maturity yet. The current state of the art approach is based on RANS solutions, which provide an averaged carrier phase on which the particles are advanced in an a posteriori manner. Upon wall impact, the particle quantities are then fed to the erosion and rebound models. In this work, we present as an alternative to this approach an Eulerian/Lagrangian simulation framework of high-order accuracy in space and time for the time-resolved prediction of particle motion in complex flows. We apply it to the LES of a particle-laden flow over a T106C low pressure turbine linear cascade. We then contrast time-averaged and time-accurate flow fields as carrier phases for the particles and highlight how the associated modeling assumptions influence the solution. Based on the particle Stokes number, we identify characteristic regimes and their interaction with the flow phase. By a detailed comparison of the particle statistics, we highlight the effects of turbulent small scale behavior and define the modeling challenges associated with finding accurate particle closure models for time-averaged simulations. This framework constitutes a first step towards high-fidelity erosion prediction for turbomachinery applications.     

Mesh Node Distribution in Terms of Wall Distance for Large-eddy Simulation of Wall-bounded Flows

In this note, basic turbulent statistics in a pipe flow are computed accurately by large-eddy simulation using a mesh resolution coarser than the viscous sublayer. These results are obtained when a regular Cartesian mesh is used for the spatial discretization of the circular pipe thanks to an immersed boundary method combined with high-order schemes. In this particular computational configuration, the near-wall features of mean velocity and Reynolds stress profiles are found to be correctly captured at a scale significantly smaller than the mesh size. Comparisons between channel and pipe flow configurations suggest that an irregular mesh distribution in terms of wall distance may be a favourable condition to explicitly compute by large-eddy simulation reliable wall turbulence without any extra-modelling in the near-wall region.