Session MA: Turbulent Boundary Layers: Experiments III

Scaling of near-wall turbulence in pipe flow

Experiments were conducted in the Princeton/ONR Superpipe. Profiles of the streamwise velocity component were measured in fully developed pipe flow at Reynolds numbers from $25\times10^3$ to $150\times10^3$. The turbulence intensity profiles non-dimensionalized with inner coordinates indicate that the magnitude of the near-wall peak is invariant with Reynolds number in both location and magnitude. The results agree with previous pipe flow data that have sufficient spatial resolution to avoid spatial filtering effects, but stand in contrast to similar results obtained in boundary layers, where the peak displays a strong Reynolds number dependence, although its position, in wall units, is fixed at the same location as in pipe. Although it is expected that the current experiments have sufficient spatial resolution at all Reynolds numbers, experiments are being conducted at constant $l^+$ to investigate the potential impact of probe spatial filtering on these results.   

Reynolds number effects on scale energy analysis of turbulent boundary layers

Scale energy analysis combines two approaches of studying wall- bounded turbulent flows - analysis in physical space and analysis in scale space. Previously, scale energy analysis has been performed on DNS channel flow data for a range of friction Reynolds numbers Re$_{\tau} = 180-934$ and dual plane PIV boundary layer data at Re$_{\tau} = 1100$. The dual plane technique allows determination of the full velocity gradient tensor in the measurement plane. Dual Plane PIV data were acquired in streamwise-spanwise planes in the logarithmic region of a water channel boundary layer at two higher Reynolds numbers - Re$_{\tau} = 2400$ and $3000$. The results of this study will be described and compared with the lower Re data. It is observed that in general, the production and scale transfer terms of the turbulent kinetic energy increase with increasing Reynolds number. The cross-over scale, which divides the range of scales into a transfer-dominated region and a production- dominated region, increases with increasing Reynolds numbers, resulting in a larger range of transfer-dominated scales at higher Reynolds numbers.   

Vortex organization in wall turbulence: inner or outer scaling?

Simultaneous hotwire measurements in the near-neutral atmospheric surface layer at $Re_\tau = \delta u_\tau /\nu \simeq 10^6$ are compared with time-resolved PIV measurements in a flat plate turbulent boundary layer at $Re_\tau = 5 \cdot 10^2$. We observe in both flows a similar strong signature in the two point correlation function of the streamwise velocity fluctuation near the wall, confirming the key statistical role of ramp-like vortex organization. However, at the lower Reynolds number, the organized structures were observed to extend up to the boundary layer thickness $\delta$, implying that in the atmospheric surface layer we should observe similar patterns up to heights of the order of 50 m, providing outer scaling holds. The effect of the Reynolds number on the scaling of ramp-like structures, hairpins and hairpin packets (Adrian, 2007) is investigated.   

PIV Characterization of Turbulent Channel Flow with Rib Patterned Superhydrophobic Walls

We report PIV measurements characterizing turbulent flow in a channel with superhydrophobic surfaces, structured and wetting surfaces, and smooth bottom surfaces. The superhydrophobic and structured surfaces were fabricated with alternating ribs and cavities. Both longitudinal and transverse rib/cavity orientations were considered. The widths of the ribs and cavities were 8 and 32 micrometers, respectively, and the cavity depths were 15 micrometers. PIV measurements were acquired over the Reynolds number range of 4800 -- 10000. The smooth bottom wall was used as a basis for comparison. The hydraulic diameter of the channel was nominally 8.2 mm with an aspect ratio of 8.9. A spanwise-averaged velocity profile was obtained at the channel centerline from the PIV data. The time-averaged velocity profiles reveal no discernible time-mean slip velocity at the superhydrophobic wall. However, the different surfaces affect the turbulence intensities, total and turbulent shear stress distributions, turbulence production in the channel, and local friction factors. Superhydrophobic surfaces with the ribs and cavities aligned with the flow show an 11{\%} decrease in the friction factor while the same surfaces aligned in the transverse direction show a modest increase in the friction factor.   

Behavior of Different Turbulent Length Scales in a Boundary Layer with Isotropic Freestream Turbulence

The effects of different length scales on interactions between nearly isotropic freestream turbulence (FST) and a favorable pressure gradient, turbulent boundary layer (TBL) over a rough surface are studied. Measurements are obtained using Laser Doppler and Hotwire Anemometry, at Re$_{\theta }\le $ 4,300. An active grid is used to generate FST levels of up to 7{\%}. Profiles of mean turbulent statistics show that the classical view of TBL flows is not able to collapse the data when FST is present. The different effects of FST on the streamwise and wall-normal variances result in that the addition of isotropic FST promotes anisotropy in the boundary layer. Second-order structure functions are examined to identify which turbulence length scales contribute mostly to creating the anisotropy. The analysis demonstrates that the effect of FST resides in a wide range of length scales, and is not limited to the largest scales of the flow as in the zero pressure gradient case. Spectral analysis is performed to more accurately identify the relevant scales.   

Experimental evidence of the Self Sustaining Process in a flat plate boundary layer

Streamwise velocity streaks and streamwise vorticity are both key features of the boundary layer and are involved into its transition to turbulence. It has been proposed that a Self Sustaining Process (SSP) exists between these structures from a given Reynolds number. We use bluff-body vortex generators to steadily force a flat plate boundary layer with counter-rotating streamwise vortices (CRSV) in a hydrodynamic tunnel at moderate Reynolds numbers. Using 2-Component 2-Dimensional Particle Image Velocimetry we reconstruct the 3-Dimensional time averaged flow field for both streamwise and spanwise components. By associating the modulation on the spanwise component to the CRSVs and the modulation on the streamwise component to the velocity streaks we witness the evolution of these structures. Using a parametric study on the Reynolds number we study the evolution of the streamwise streaks transient growth. Monitoring the evolution of the amplitude of the CRSV we show the transition from their decay at low Reynolds number to the existence of the SSP. Looking at the instantaneous vector fields we propose the destabilization of the streamwise streaks as the process responsible to the existence of a critical Reynolds number for the SSP.   

The Mechanism of Heat Transfer Augmentation in Stagnation Flow Subject to Freestream Turbulence

A physical model is presented which predicts the time-resolved heat transfer coefficient based on the properties of the coherent structures present. Water tunnel experiments have been performed to investigate the mechanism of heat transfer augmentation in stagnation flow subject to freestream turbulence. The experiment combined Time-Resolved Digital Particle Image Velocimetry with an array of simultaneous time-resolved heat transfer measurements. Passive grids produced freestream turbulence with an intensity of 5{\%} at length scales of 1cm, 2cm, and 3cm. The measurements reveal flow fields dominated by coherent structures whose number and strength strongly correlate with the length scale of the freestream turbulence. By examining the transient circulation and location of the identified structures, we observe that stretching and vorticity amplification significantly affects the near-wall flow. The transient heat transfer correlates well with the flow field induced by these structures. The time-resolved model developed represents a great advance over previous time-average predictors.   

Scaling of Polymer Degradation Rate within a High-Reynolds-Number Turbulent Boundary Layer

An experiment conducted at the U.S. Navy's Large Cavitation Channel on a 12.9 m long flat-plate test model produced the first quantitative measurements of polymer molecular weight within a turbulent boundary layer. Testing was conducted at speeds to 20 m/s and downstream distance based Reynolds numbers to 220 million. These results showed that the rate of polymer degradation by scission of the polymer chains increases with increased speed, downstream distance and surface roughness. With the surface fully rough at 20 m/s there was no measureable level of drag reduction at the first measurement location (0.56 m downstream of injection). These results are scaled with the assumption that the rate of degradation is dependent on the polymer residence time in the flow and the local shear rate. A successful collapse of the data within the measurement uncertainty was achieved over a range of flow speed (6.6 to 20 m/s), surface roughness (smooth and fully rough) and downstream distance from injection (0.56 to 9.28 m).   

Superhydrophobic Surfaces for Turbulent Drag Reduction: Effect of Geometry and Reynolds Number

Recently, it has been demonstrated that superhydrophobic surfaces are capable of reducing drag in turbulent flows. Superhydrophobic surfaces are chemically hydrophobic with micron or nanometer scale surface features which can support a shear-free air-water interface. In this talk, we will consider the effect of microfeature geometry and flow velocity on the observed drag reduction. Microridge geometries from 15$\mu $m to 60$\mu $m will be considered with shear free area ratios from 20{\%} to 80{\%}. Drag reductions are found to increase with increasing shear free area ratio up to 75{\%}. For any given shear free area ratio, drag reduction was seen to initiate at lower Reynolds numbers for larger microfeature spacings. Experiments were conducted over the range of Reynolds numbers 1000$<$\textit{Re}$<$100000, demonstrating that drag reduction in the turbulent regime is distinct from drag reduction observed over superhydrophobic surfaces in the laminar regime and persists well past transition to turbulence.   

Drag reduction in flows over superhydrophic surfaces

Recent research has suggested that large reductions in both laminar and turbulent skin friction might be realized in water flows over over superhydrophobic surfaces due to the modification of the no-slip boundary condition at the liquid-solid interface by a thin layer of trapped air. In our work, superhydrophobic surfaces have been fabricated by first laser-etching a textured pattern into an aluminum baseplate, and then coating the surface with a hydrophobic coating. The superhydrophbic characteristics and the presence of a trapped air layer have been characterized at a variety of operating pressures. The drag characteristics of the baseline and superhydrophobic surfaces have also been measured under both laminar and turbulent flow conditions using three different measurement apparatus - a narrow gap laminar flow channel, a large gap turbulent flow channel and a closed return water tunnel. The results from these measurements will be presented and discussed in detail.