Session Z08: Boundary Layers: Turbulent III

Thermal boundary condition and unsteady effects in high-speed turbulent channels.

We conduct direct numerical simulations of compressible turbulent channels in high-speed regimes, to study the near-wall behavior of turbulent statistics and the effects of Mach numbers as well as different thermal wall-boundary conditions, in particular, isothermal, adiabatic and pseudo-adiabatic. We have shown that the asymptotic behavior of Reynolds stresses change with the Mach number as well as thermal boundary conditions identifying the fluctuating dilatation at the wall as the responsible parameter. In this work, we will show that the asymptotic behavior of turbulent heat fluxes also depend on thermal boundary conditions, an effects that, in turns, depends on the level of compressibility. These effects are also studied in the budget equations which shed light on the relative importance of physical processes at different distances from the wall. We discuss how these results extend to unsteady effects which are analyzed with decaying channel simulations.   

Scaling surface drag in the atmospheric boundary layer using laboratory concepts

The work demonstrates how the drag scaling from laboratory Turbulent Boundary Layers (TBLs) can be extended to scale the drag or surface friction velocity in near-neutral and convective Atmospheric Boundary Layers (ABLs). Here, The estimation of surface drag in ABLs is enabled solely from the Radiosonde measured wind speed profiles. For this, The model presented in Dixit et al. (2020) for laboratory TLBs (i.e., the non-conductive NC model) is modified to compensate for convection effects using the convective velocity scale 

and the ABL bulk velocity scale. This modified convective model (i.e., the C model) is able to estimate the drag to within a 20% error band with a probability of 80%, which is remarkable given the inherent uncertainty of ±20% for the measured drag values. The benefit of this new model is that it requires only the Radiosonde profile data for the estimation of surface drag. These results are applicable to convective and near-neutral atmospheric conditions and could be of immense importance for data assimilation where Radiosonde profiles are available, but the surface fluxes are not known due to the unavailability of costly fast-response tower instrumentation.   

Transition from attached to detached eddies in stably stratified boundary layers

Research on canonical turbulent boundary layers (TBLs) in recent decades has provided growing support for A. A. Townsend's hypothesis regarding attached eddies. The height-dependent size of these eddies is realized in physical space in part by the behavior of uniform momentum zones. Yet, canonical conditions are rarely realized in practical applications such as the atmospheric boundary layer (ABL) where buoyancy effects are typically relevant. This presentation uses large-eddy simulations of the ABL to investigate the influence of stable stratification on the properties of the attached eddies. The organization of eddies in canonical TBLs is similarly observed for the stable ABL: the turbulent flow field preferentially organizes into zones of uniform velocity and temperature separated by thin layers with enhanced gradients. The "jump" in velocity and temperature across the gradient layers remains proportional to the surface shear stress and heat flux regardless of stratification. However, the zones become smaller and lose their dependence on the wall position as the stratification increases. These trends in the turbulence organization can be used to explain logarithmic mean profile deviations in stratified conditions that are traditionally modelled using empirical similarity relations.   

Boundary layer structure over standing waves

The boundary layer development and the onset and interaction of coherent structures over periodic standing waves were experimentally studied for various frequencies of wall oscillations. A customized Arduino-controlled mechanical system under a flexible wall allowed the formation of sinusoidal standing waves of 5 mm amplitude and 100 mm wavelength. The actuated wall has a section of 305 mm by 800 mm. It was placed and operated in a specially designed wind tunnel of 305 mm by 305 mm cross-section. We explored the near-wall turbulence development and large-scale coherent structure interaction for various Reynolds numbers and reduced frequency using phase-lock particle image velocimetry at different phase angles of the standing waves. We will discuss the modulation of distinct wall frequency on the turbulent energy of the flow and the coherence of energetic motions.   

Unsteady Surface Pressure Spectra with Small Levels of Wall Vibration

The spatial-temporal characteristics of the Unsteady Surface Pressure (USP) due to turbulent boundary layer flow is important for understanding forcing mechanisms in flow-induced vibration. Often, the USP is considered to be determined by the turbulent flow characteristics only, and is not influenced by wall vibration. This is sometimes referred to as the "blocked pressure assumption". An experiment was conducted in an anechoic wind tunnel in which the USP was measured directly downstream of a flexible membrane. The data indicated that changes in USP amplitude up to 5dB were observed over a broad range of frequencies when the wall was allowed to vibrate. This was found even when the wall motion was less than 20 viscous units in amplitude. Tonal acoustic forcing was considered, in which the wall vibration amplitude was in excess of 100 viscous units. These conditions led to increases in USP up to 8dB over a wide range of frequencies.   

Fluid-structure interaction of a viscoelastic compliant wall in turbulent channel flow

Fluid-structure interaction of a viscoelastic wall in an incompressible turbulent channel is performed using Direct Numerical Simulation (DNS). The simulations include boundary condition effects by clamping the wall on the four sides. For this fluid-solid coupling problem, we cannot arbitrarily set the average fluid pressure, as it would lead to a non-physical wall deformation. Certain modifications to the sequential fixed-point iteration necessary to compute the average fluid pressure are discussed. Using the DNS wall-pressure fluctuations computed on a rigid wall at friction Reynolds numbers 180 and 400, we propose modifications to wall-pressure models to account for the low-Reynolds number effect. Furthermore, model-based predictions of plate excitation and radiated sound are compared to one-way coupled DNS-based predictions for an elastic plate.   

Wall-resolved LES investigation of Reynolds number effects on a near-stall airfoil flowfield up to Rec=107

We have conducted a wall-resolved LES around the Aerospatiale A-airfoil at Re_c=10⁷ to understand the Reynolds number effects of the trailing-edge stall phenomena and to construct turbulence databases at a high Reynolds number (https://www.klab.mech.tohoku.ac.jp/database/). The computational grid contains approximately 38 billion grid points, and the simulation is conducted by the supercomputer Fugaku. Compared to the prior LES at Re_c=2.1x10⁶ [Asada & Kawai, Phys. Fluids (2018)], the present LES shows typical Reynolds number effects, such as the delay of the turbulent separation and the earlier turbulence transition. We analyze the development of the boundary layer over the upper airfoil surface through the integral relation as in the prior study [Tamaki et al., Phys. Rev. Fluids (2020)] to clarify the causes of the Reynolds number effects. The analysis shows that the primary difference in the boundary layer thickness is caused in the laminar and transitional regions, and then the difference is amplified in the downstream region.

    

Simultaneous flow-field measurements of velocity and passive-scalar mixture-fraction from a point-source in a turbulent boundary layer

High resolution simultaneous Particle Image Velocimetry (PIV) and Planar Laser-induced Fluorescence (PLIF) are conducted to study transport and mixing of a passive scalar within the boundary layer. This was performed by an acetone-air mixture injected isokinetically at h/δ~0.2 to compare with previous point-measurements. The field-of-view was approximately 2δ x 1.4δ which encompassed the entire boundary layer height and the injection point of the scalar. This setup captured the initial advection and mixing of the stream within the boundary layer. Post-processing of the data generated quantitative high spatial resolution velocity and scalar fields in the aforementioned plane. From these measurements, single- and two-point correlations are computed from the velocity and scalar fields, as well as cross-correlation between the two fields. These values are compared to data provided by single point measurements in the literature. The multipoint statistics conducted in this experiment provide a comprehensive picture of the coupled structure of velocity and scalar fields, and of the scalar transport by the turbulent scales.   

Direct Numerical Simulation of Turbulent Boundary Layer Flows Over a Permeable Bed Using Continuum and Pore-Resolved Approaches

A direct numerical simulation (DNS) is performed for a turbulent boundary layer over a porous sediment bed at permeability Reynolds number of Re_K= 2.56 (Re_τ=270) representative of aquatic systems. A continuum approach based on the volume-averaged Navier-Stokes (VANS) equations is used by defining smoothly varying porosity across the bed interface and modeling the drag force in the porous bed using a modified Ergun equation with Forchheimer corrections for inertial terms (Wood et al., Annual Review of Fluid Mechanics, 2020). The results from the continuum approach DNS are compared with a pore-resolved DNS in which turbulent flow over a randomly packed sediment bed is performed using a fictitious domain method to enforce the rigidity and no-slip condition on the monodispersed spherical particles representing the sediment bed. A spatially varying porosity profile generated from the pore-resolved DNS is used in the continuum approach. Mean flow and Reynolds stress statistics and net momentum exchange between the free-stream and the porous bed are compared between the two DNS studies, showing good agreement. The continuum VANS approach allows for a significant reduction in the computational costs thereby allowing to study hyporheic exchange of mass and momentum in large scale aquatic domains with combined influence of bedform and bed roughness.   

Reynolds stress budgets in Cartesian and orthogonal curvilinear coordinates for a turbulent flow over a curved ramp via DNS

We study the dynamics of turbulence in a spatially developing turbulent boundary layer (SDTBL) separating over a curved ramp via direct numerical simulation (DNS). We have performed DNS using FastRK3 which is an explicit, third-order Runge-Kutta based projection method which requires solving the Poisson equation for pressure only once per time step to solve the incompressible Navier-Stokes equations in orthogonal curvilinear coordinates (Aithal & Ferrante, J. Comput. Phys, 2020). We have derived, for the first time, the budget equations of the turbulence kinetic energy (TKE) and Reynolds stresses in orthogonal curvilinear coordinates, and compared them to those in Cartesian coordinates. The results of such comparison and physical mechanisms will be discussed.