Session AC: Turbulence Simulations I

Large-eddy simulation of the flow over two-dimensional dunes

When a fluid flows over a mobile sand bed the sediment transport generated by the interaction of the flow field with the bed often results in the periodic deformation of the bed in the form of dunes. Dunes generally reach an equilibrium shape, and slowly propagate in the direction of the flow, as sand is lifted in the high-shear regions, and redeposited in the separated-flow area. Our aim is to attempt to connect the flow-field characteristics with the bed deformation. As a first step, we perform large eddy simulation of the flow over a typical dune geometry at laboratory scale (the Reynolds number based on the average channel height and mean velocity is 18,900). We consider three dunes, with different heights (relative to the channel depth) but equal wavelengths, using approximately ten million grid points. The mean flow shows a recirculation region downstream of the dune crest, whose extent increases with dune height. After reattachment the shear stress becomes high, confirming that sediment is lifted up in this region. The Reynolds stresses are higher in the shear layer, where the high spanwise vorticity gives rise to coherent vortices. The budgets of turbulent kinetic energy show that, in addition to production and dissipation, the diffusion terms play an important role. In the reattachment region, diffusion and dissipation are more significant. The mean flow advection is important at the beginning of the shear layer.   

Large Eddy Simulation study of the fully developed wind-turbine array boundary layer

When wind turbines are deployed in large arrays, their efficiency decreases due to complex interactions among themselves and with the atmospheric boundary layer (ABL). For wind farms whose length far exceeds the height of the ABL, a fully developed flow regime can be established. Such a fully developed wind-turbine array boundary layer may be studied computationally using periodic boundary conditions in the horizontal direction. A suite of Large Eddy Simulations in which wind turbines are modeled using the classic ``drag disk'' concept are performed, in order to quantify the vertical transport of momentum and kinetic energy across the boundary layer. LES for various wind turbine arrangements, loading factors, and surface roughness are performed. Horizontally averaged statistics are documented. Results are compared with models for effective roughness length scales experienced by the ABL. This scale is often used to parameterize wind turbine arrays in models of atmospheric flow at regional or global scales. Based on the observed trends, a modified model is proposed showing improvements in the predicted effective roughness heights.   

Coherent structures in a turbulent flow over an urban canopy made of cubical obstacles

The study of turbulent heat or mass transport is of special interest in engineering, especially for heat exchangers. For instance, roughness elements (turbulators) are usually placed on the walls of the internal channels of a turbine blade to enhance the heat transfer. In the present paper, DNSs are carried out for passive heat transport in a turbulent channel flow with $\Lambda$ shape square ribs for w/k = 1, 3, 7, 15 (w being the pitch, k the height of the ribs turbulators. The angle of inclination of the lambda shape turbulators is 45 degrees. Numerical results show that $\Lambda$ shape square ribs are more efficient than square ribs in maximizing the heat transfer. The configuration with w/k=3 presents the largest heat flux. The increase in the heat transfer is due to a secondary motion which is generated by the $\Lambda$ shape turbulators. Two counter rotating vortices above the square ribs transport the heat out of the wall into the center of the channel. The distribution of the heat flux coefficient is not uniform in the channel and leads to temperature gradients at the wall. The total drag of the $\Lambda$ shape turbulators is larger than that over a smooth wall due to an increase of form drag.   

Direct numerical simulation of turbulent flow over a surface mounted obstacle

The direct numerical simulation of turbulent flow over a surface mounted obstacle is studied as a numerical experiment that takes place in a wind tunnel. For this reason, the incompressible, three dimensional, transient Navier-Stokes equations for Newtonian fluids are solved directly using Galerkin finite elements. The Reynolds number defined with respect to the height of the obstacle is in the range of $10^5$. The results include instantaneous streamline patterns that show the vortex shedding phenomenon and the flapping of the recirculation bubble downstream the obstacle. Energy spectra are studied along with Eulerian autocorrelation coefficients, longitudinal and lateral coefficients that yield the chaotic behavior of turbulence. The computer code developed for this work is a parallel program written in Fortran 90 that uses the MPI-paradigm and runs in distributed memory systems. Movies are shown where both streaklines and instantaneous streamlines are depicted that clearly demonstrate the transient characteristics of this prototype separated flow.   

Large-eddy simulation of flow over a multi-element airfoil

An accurate prediction of turbulent flow over a multiple element high-lift airfoil configuration remains a challenge to computational fluid dynamics. Maximum lift, drag, and pitching moment are difficult to accurately predict especially in the presence of flow separation on one or more of the airfoil elements. In this study, we investigate turbulent flow over a MD30P30N high-lift configuration using large-eddy simulation. The MD30P30N configuration consists of three elements: a slat, a main airfoil, and a flap. Four different attack angles, 16$^\circ$, 19$^\circ$, 21$^\circ$, and 24$^\circ$, are considered while deflection angles of the slat and flap are fixed to 30$^\circ$. The Reynolds number is $9\times 10^6$ based on the mounted-wing chord-length and freestream velocity. Simulation results obtained on a 54 million-element mesh agree well with experimental data in terms of pressure distribution, velocity profiles, and transition location. A grid sensitivity study is performed to identify the resolution effects on the prediction of flow transition, wakes, and turbulent boundary layers. Accurate prediction of laminar-to-turbulence transition on the slat surface and downstream evolution of the slat wake is found to be crucial for the global accuracy of the simulation.   

Numerical Simulation of Turbulence-Induced Bedform Initiation

Bedform initiation induced by near-wall turbulence structures on a sand bed is studied using large-eddy simulation. Due to the dilute sediment concentration during the simulation, transport of sediment is modeled with the Eulerian method. A second-order accurate arbitrary Lagrangian-Eulerian scheme is implemented that allows flow simulation over evolving bedforms. With a bed elevation model based on conservation of sediment mass to calculate changes in bed elevation, the present numerical model enables detailed observation of bedform instability caused by near-wall turbulence. It is found that the streak structure on the bed surface appears as the initial bed perturbation due to sediment erosion by turbulent sweeps, which in turn induce small pile-up at the downstream end of the inrush zone where the sweep diminishes. The continuous growth of the small sediment pile-up leads to the formation of bed defects, which alter the flow condition and the spatial distribution of near-bed sediment erosion. As a consequence, merging of multiple bed defects leads to the formation of ripple marks, which are a common bedform pattern in the subaqueous environment. The simulation results reveal interactions between turbulent structures and the sand bed and demonstrate the importance of the resolved turbulence in the simulation of bedform initiation.   

Direct numerical simulation of surface ablation by turbulent convection

Rapid erosion by a turbulent flow creates complex flow/surface phenomena arising from the evolving surface topography and its interaction with a turbulent flow that transports the erosive agent onto the surface. The non-equilibrium nature of the problem poses major challenges to current turbulent models and boundary conditions used in direct numerical simulation (DNS) algorithms. A generalized algorithm for turbulent erosion processes based on level-set and immersed boundary methods has been developed in a DNS flow solver to investigate the action of various erosive agents (heat, particles, chemical species) on erodible surfaces. This algorithm is applied to the ablation of a slab of ice by natural and forced convection of water. The study focuses on the characterization of the surface topography in relation to the evolution of coherent structures in the flow, as ablation proceeds.   

Comparing DNS and Experiments of Subcritical Flow Past an Isolated Surface Roughness Element

Results are presented from computational and experimental studies of subcritical roughness within a Blasius boundary layer. This work stems from discrepancies presented by Stephani and Goldstein (AIAA Paper 2009-585) where DNS results did not agree with hot-wire measurements. The near wake regions of cylindrical surface roughness elements corresponding to roughness-based Reynolds numbers \textit{Re}$_{k}$ of about 202 are of specific concern. Laser-Doppler anemometry and flow visualization in water, as well as the same spectral DNS code used by Stephani and Goldstein are used to obtain both quantitative and qualitative comparisons with previous results. Conclusions regarding previous studies will be presented alongside discussion of current work including grid resolution studies and an examination of vorticity dynamics.