Distributed Roughness Elements Research Articles

Boundary layer transition due to distributed roughness: effect of roughness spacing

The influence of roughness spacing on boundary layer transition over distributed roughness elements is studied using direct numerical simulation and global stability analysis, and compared with isolated roughness elements at the same Reynolds number $Re_h=U_eh/\nu$ ( $U_e$ is the boundary layer edge velocity, h is roughness height and $\nu$ is the kinetic viscosity of the fluid). Small spanwise spacing ( $\lambda _z=2.5h$ ) inhibits the formation of counter-rotating vortex pairs (CVP) and, as a result, hairpin vortices are not generated and the downstream shear layer is steady. For $\lambda _z=5h$ , the CVP and hairpin vortices are induced by the first row of roughness, perturbing the downstream shear layer and causing transition. The temporal periodicity of the primary hairpin vortices appears to be independent of the streamwise spacing. Distributed roughness leads to a lower critical roughness Reynolds number for instability to occur and a more significant breakdown of the boundary layer compared with isolated roughness. When the streamwise spacing is $\lambda _x=5h$ , the high-momentum fluid barely moves downward into the cavities and the wake flow has little impact on the following roughness elements. The leading unstable varicose mode is associated with the central low-speed streaks along the aligned roughness elements, and its frequency is close to the shedding frequency of hairpin vortices in the isolated case. For larger streamwise spacing ( $\lambda _x=10h$ ), two distinct modes are obtained from global stability analysis. The first mode shows varicose symmetry, corresponding to the primary hairpin vortex shedding induced by the first-row roughness. The high-speed streaks formed in the longitudinal grooves are also found to be unstable and interact with the varicose mode. The second mode is a sinuous mode with lower frequency, induced as the wake flow of the first-row roughness runs into the second row. It extracts most energy from the spanwise shear between the high- and low-speed streaks.

Hypersonic Boundary-Layer Transition on Cold-Wall Canonical Geometries with Quantified Distributed Roughness Elements

Given the challenges of obtaining a natural turbulent boundary layer on common test model geometries in a shock tunnel, this work aims to investigate the influence of roughness elements on the boundary layer with respect to transition to turbulence. The experiments were conducted in the High Enthalpy Shock Tunnel Göttingen at the German Aerospace Center on a 1100-mm-long, 7°-half-angle cone and a 602-mm-long flat plate at Mach 7.4. Roughness elements were applied on the nosetip of the cone and near the leading edge of the flat plate. The roughness elements were scanned with a laser profilometer, allowing their specification in terms of a roughness Reynolds number based on the 70th-percentile element height and an exceedance probability distribution. Transition was examined for the cone geometry using streamwise-aligned coaxial thermocouples on the 0° meridian. This assisted sizing the roughness elements required for transition to occur as far upstream as detectable. Breakdown of roughness-induced vortical structures generated by the roughness elements with a similar roughness Reynolds number was then examined using the flat plate geometry with temperature-sensitive paint applied downstream of the roughness elements. It was found that roughness-induced vortices required a finite distance (persistence length) to break down into turbulent structures. The persistence length was successfully reduced by interspersing roughness elements with smaller ones.

Impact of high size distributed roughness elements on wind turbine performance

NOTITLE.

Spatial resolution issues in rough wall turbulence

The question whether spatial resolution effects should be accounted for when performing hot-wire (or particle image velocimetry) measurements in turbulent wall-bounded flows over rough surfaces in the fully rough regime is addressed and answered by exploiting an existing direct numerical simulation database of open-channel flows at Re_tau =500 over a rough surface consisting of randomly distributed roughness elements. The results show sizeable attenuation effects of the streamwise velocity fluctuation intensity, similar to smooth-wall flows with up to 70% for a wire width of around 100 viscous units. A suitable correction scheme is applied to successfully compensate for the spatial resolution effects. Both results indicate that spatial resolution effects are relevant also in fully rough flows and can be corrected a posteriori when the state-of-the-art criteria on sensor size cannot be achieved in experiments.Graphical

Numerical investigation of distributed roughness effects on separated flow transition over a highly loaded compressor blade

Large eddy simulations were conducted to investigate the effects of distributed roughness on separated flow transitions over a highly loaded compressor blade at a Reynolds number (Re) of 1.5 × 105. The distributed roughness elements were located downstream of the velocity peak on the suction surface, and four numerical cases with increasing peak amplitude of the roughness elements (k+ = 0, 23, 50, and 112) were considered. The results showed that low- and high-speed streamwise streaks appeared alternately along the spanwise direction over the distributed roughness elements. The streaks remained steady earlier; however, as the streamwise counter-rotating vortices were induced by a significant spanwise velocity component, the low-momentum fluid in the near-wall region was transported away from the blade surface and interacted with the outer separated shear layers, which caused unsteady merging of streaks and promoted the destabilization of separated shear layers. Compared with the baseline case (k+ = 0), the strong shear effect between the low- and high-speed streamwise streaks near the roughened blade surface accelerated the distortion of spanwise vortices, and three-dimensional hairpin vortex structures broke down into small-scale turbulent eddies at a shorter streamwise distance. With increase in the roughness magnitude, the level of the production term of turbulent kinetic energy was reduced due to weakened vortex dynamics, and the viscous dissipation in turbulent boundary layers also became weaker. Therefore, the profile losses of the three roughness cases, k+ = 23, 50, and 112, were decreased by 7.2%, 10.1%, and 15.5%, respectively.

Effect of Surface Roughness Geometry on Boundary-Layer Transition and Far-Field Noise

Surface roughness elements are often used to force laminar to turbulent transition in aerodynamic and aeroacoustic wind-tunnel experiments. The statistical features and spectral content of the pressure fluctuations in the resulting turbulent boundary layer at the trailing edge can affect far-field noise. To elucidate this dependence, boundary-layer transition induced by randomly distributed roughness elements and a zigzag strip of the same height over a NACA 0012 airfoil is investigated experimentally. The effects of roughness geometry on the near-field flow topology, transition location, and far-field noise are addressed in the common experimental setting for the first time. For a fixed roughness height, distributed roughness elements are less effective in forcing transition than the zigzag strip at low freestream velocity (). As increases, the transition front for the distributed roughness elements moves closer to the roughness location, reaching the same or even further upstream locations compared to the transition onset in the presence of the zigzag strip. The far-field noise depends on the transition location. For , a higher noise level is measured for the distributed roughness elements with respect to the zigzag strip. In contrast, for , the earlier onset of transition with the distributed surface roughness leads to a lower noise level than that with the zigzag strip. The data confirm that an adequate characterization of the boundary-layer transition is necessary when measuring the far-field noise during wind-tunnel experiments.

Direct numerical simulation of flow over dissimilar, randomly distributed roughness elements: A systematic study on the effect of surface morphology on turbulence

Direct numerical simulations of flow over fourteen dense rough walls with varying geometry are conducted. The effect of a departure from same-size element roughness on turbulence statistics is studied, and the correlation of the roughness function with wall-normal fluctuations at the roughness crest is examined.

Turbulent thermal convection over rough plates with varying roughness geometries

We present a systematic investigation of the effects of roughness geometry on turbulent Rayleigh–Bénard convection (RBC) over rough plates with pyramid-shaped and periodically distributed roughness elements. Using a parameter $\unicode[STIX]{x1D706}$ defined as the height of a roughness element over its base width, the heat transport, the flow dynamics and the local temperatures are measured for the Rayleigh number range $7.50\times 10^{7}\leqslant Ra\leqslant 1.31\times 10^{11}$ and Prandtl numbers $Pr$ from 3.57 to 23.34 at four values of $\unicode[STIX]{x1D706}$ (0.5, 1.0, 1.9 and 4.0). It is found that the heat transport scaling, i.e. $Nu\sim Ra^{\unicode[STIX]{x1D6FC}}$ where $Nu$ is the Nusselt number, may be classified into three regimes in turbulent RBC over rough plates. In Regime I, the system is in a dynamically smooth state. The heat transport scaling is the same as that in a smooth cell. In Regimes II and III, the heat transport is enhanced. When $\unicode[STIX]{x1D706}$ is increased from 0.5 to 4.0, $\unicode[STIX]{x1D6FC}$ increases from 0.36 to 0.59 in Regime II and it increases from 0.30 to 0.50 in Regime III. The experiment thus clearly demonstrates that the heat transport scaling in turbulent RBC can be manipulated using $\unicode[STIX]{x1D706}$ in the heat transport enhanced regime. Previous studies suggest that the transition to heat transport enhanced regime, i.e. from Regime I to Regime II, occurs when the thermal boundary layer (BL) thickness becomes smaller than the roughness height. Direct measurements of the viscous BL in the present study suggest that the transition from Regime II to Regime III is likely a result of the viscous BL thickness becoming smaller than the roughness height. The scaling exponent of the Reynolds number $Re$ with respect to $Ra$ changes from 0.471 to 0.551 when $\unicode[STIX]{x1D706}$ is increased from 0.5 to 4.0, suggesting a change of the dynamics of the large-scale circulation. Interestingly, the transition from Regime II to Regime III in terms of the heat transport scaling is not reflected in the $Re$ scaling with $Ra$. It is also found that increasing $\unicode[STIX]{x1D706}$ increases the clustering of thermal plumes which effectively increases the plume lifetime. This leads to a great increase in the probability of observing large temperature fluctuations in the bulk flow, which corresponds to the formation of more coherent plumes or plume clusters that are ultimately responsible for the enhanced heat transport.

Law-of-the-wall in a boundary-layer over regularly distributed roughness elements

ABSTRACTAs opposed to the log-region, the roughness sublayer present above rough surfaces is still poorly understood due to the complex interaction between wakes developing behind roughness elements. To investigate the spatially averaged flow velocity in this region, a data-set has been collected from several direct numerical simulations and wind-tunnel experiments available in the literature. A generalised law-of-the-wall has been derived, applicable to a roughness sublayer present over regularly distributed roughness elements. The key roughness parameter of this new law is the effective height ϵ, which characterises the interaction between the roughness and the outer flow in a temporally and spatially averaged sense. A morphometric study reveals that ϵ is closely related to a new roughness density parameter, λ2, that accounts for the roughness element shape and the inter-element spacing. This allows ϵ to be a universal parameter on roughness characterisation. The derived values of the classical roughness length z0 of the log-law compare well with previous experimental data and geometrical model predictions. Finally, the main properties of the roughness sublayer such as its height are discussed using the geometrical and the roughness parameters proposed in the study.

On the excitation of Görtler vortices by distributed roughness elements

Gortler vortices evolve in boundary layers over concave surfaces as a result of the imbalance between centrifugal effects and radial pressure gradients. Depending on various geometrical and flow conditions, these vortices can significantly distort the baseline flow and lead to secondary instabilities and ultimately to transition. In this study, the growth of Gortler vortices excited by distributed roughness elements is analyzed using the solution to the nonlinear boundary region equations with upstream boundary conditions derived previously via an asymptotic analysis applied in the vicinity of the roughness elements. Generalized Rayleigh pressure equation derived based on the assumption that the baseline flow is a function of the transverse coordinates only is used to determine the growth rates associated with the secondary instabilities. Within the analysis, the roughness shape, height and diameter as well as the spanwise separation between the roughness elements are varied in the linear regime, while keeping the same Gortler number. It is found that bell-shaped distributed roughness elements are more likely to excite the Gortler instabilities than sharp-edge-type (e.g., cylindrical) roughness elements, and by increasing the roughness diameter, the strength of Gortler vortices associated with the bell-shaped roughness elements increases as expected, but the strength of Gortler vortices associated with cylindrical-shaped roughness elements decreases.

Effect of Freestream Turbulence on Roughness-induced Crossflow Instability

The effect of freestream turbulence on generation of crossflow disturbances over swept wings is investigated through direct nu- merical simulations. The set up follows the experiments performed by Downs et al. (2012). In these experiments the authors use ASU(67)-0315 wing geometry which promotes growth of crossflow disturbances. Distributed roughness elements are locally placed near the leading edge with a given spanwise wavenumber to excite the corresponding stationary crossflow vortices. In present study, we partially reproduce the isotropic homogenous freestream turbulence through direct numerical simulations using freestream spectrum data from the experiments. The generated freestream fields are then applied as the inflow boundary condition for direct numerical simulation of the wing. The distributed roughness elements are modelled through wing surface deformation and placed near the leading edge to trigger the stationary crossflow disturbances. The effects of the generated freestream turbulence on the initial amplitudes and growth of the boundary layer perturbations are then studied.

Mars Science Laboratory Heat Shield Aerothermodynamics: Design and Reconstruction

The Mars Science Laboratory heat shield was designed to withstand a fully turbulent heat pulse using information from ground testing and computational analysis on a preflight design trajectory. Instrumentation on the flight heat shield measured in-depth temperatures to permit reconstruction of the surface heating. The data indicate that boundary-layer transition occurred at five of seven measurement locations before peak heating. Data oscillations at three pressure measurement locations may also indicate transition. This paper presents the heat shield temperature and pressure data, possible explanations for the timing of boundary-layer transition, and a comparison of reconstructed and computational heating on the actual trajectory. A smooth-wall boundary-layer Reynolds number that was used to predict transition is compared with observed transition at various heat shield locations. A single transition Reynolds number criterion does not uniformly explain the timing of boundary-layer transition observed during flight. A roughness-based Reynolds number suggests that transition due to discrete or distributed roughness elements occurred. However, the distributed roughness height from acreage heat shield material would have needed to be larger than expected. The instrumentation confirmed the predicted location of maximum turbulent heat flux near the lee-side shoulder. The reconstructed heat flux at that location is bounded by smooth-wall turbulent convective heating, indicating that a significant augmentation due to surface roughness did not occur. Turbulent heating on the downstream side of the heat shield nose exceeded smooth-wall convective levels, assuming a supercatalytic surface, indicating that roughness may have augmented heating. The stagnation region heating also exceeded calculated convective heating; the cause of elevated heating may be attributed to a combination of shock-layer radiation and a heating augmentation of unknown origin that was also evident in ground test data.

Numerical modelling of aeolian erosion over a surface with non‐uniformly distributed roughness elements

ABSTRACTThe present study is focused on the analysis of the mean wall friction velocity on a surface including roughness elements exposed to a turbulent boundary layer. These roughness elements represent non‐erodible particles over an erodible surface of an agglomeration of granular material on industrial sites. A first study has proposed a formulation that describes the evolution of the friction velocity as a function of geometrical parameters and cover rate with different uniform roughness distributions. The present simulations deal with non‐uniform distributions of particles with a random sampling of diameters, heights, positions and arrangements. The evolution (relative to geometrical parameters of the roughness elements) of the friction velocity for several non‐uniform distributions of roughness elements was analysed by the equation proposed in the literature and compared to the results obtained with the numerical simulations. This comparison showed very good agreement. Thus, the formulation developed for uniform particles was found also to be valid for a larger spectrum of particles noted on industrial sites. The present work aims also to investigate in detail the fluid mechanics over several roughness particles. Copyright © 2013 John Wiley & Sons, Ltd.

Numerical Investigation of Passive Separation Control for an Airfoil at Low-Reynolds-Number Conditions

Two different passive flow control strategies were investigated for a modified NACA airfoil at a chord-based Reynolds number of and an angle of attack of . For these conditions, the laminar boundary layer separates from the suction side, resulting in a loss of lift and a drag increase. Distributed roughness elements with roughness Reynolds numbers of and 446 that were mounted near the leading edge and scalloped leading edges with serration amplitudes of 5 and 0.5% of the chord were considered. The large roughness elements and the scalloped leading edge reduce the flow separation and enhance performance. The flow physics are, however, different. For , the roughness elements result in high-frequency shedding. The shedding results in an accelerated transition of the separated boundary layer. For the scalloped leading edge with 5% serration amplitude, laminar separation bubbles are situated in the leading-edge troughs. The turbulent wedges that originate from these bubbles coalesce near midchord. For a serration amplitude of 0.5%, the separation line is deformed in the spanwise direction in a manner that is reminiscent of stall cells.

Low-wavenumber turbulent boundary layer wall-pressure measurements from vibration data over smooth and rough surfaces in pipe flow

The vibration response of a thin cylindrical shell excited by fully developed turbulent pipe flow is measured and used to extract the fluctuating pressure levels generated by the boundary layer. Parameters used to extract the turbulent fluctuating pressure levels are determined via experimental modal analyses of the water-filled pipe and measured vibration levels from flow through the pipe at 5.8m/s. Measurements are reported for hydraulically smooth and fully rough surface conditions. Smooth wall-pressure levels are compared to the turbulent boundary layer pressure model of Chase [The character of the turbulent wall pressure at subconvective wavenumbers and a suggested comprehensive model. Journal of Sound and Vibration112 (1) (1987) 125–147] and the measurements of Bonness et al. [Low-wavenumber turbulent boundary layer wall-pressure measurements from vibration data on a cylinder in pipe flow. Journal of Sound and Vibration329 (2010) 4166–4180]. Results for the smooth pipe match the predicted smooth wall-pressure spectrum and correspond to a normalized low wavenumber-white level which is −41dB below the maximum level at the convective peak. Pressure levels from the fully rough condition display a low-wavenumber-white level which is 28dB below the convective peak level. This suggests an increase of 13dB in low-wavenumber wall pressure for the uniformly distributed roughness elements in this study over a hydraulically smooth surface.

Stabilization of a swept-wing boundary layer by distributed roughness elements

AbstractThe stabilization of a swept-wing boundary layer by distributed surface roughness elements is studied by performing direct numerical simulations. The configuration resembles experiments studied by Saric and coworkers at Arizona State University, who employed this control method in order to delay transition. An array of cylindrical roughness elements are placed near the leading edge to excite subcritical cross-flow modes. Subcritical refers to the modes that are not critical with respect to transition. Their amplification to nonlinear amplitudes modifies the base flow such that the most unstable cross-flow mode and secondary instabilities are damped, resulting in downstream shift of the transition location. The experiments by Saric and coworkers were performed at low levels of free stream turbulence, and the boundary layer was therefore dominated by stationary cross-flow disturbances. Here, we consider a more complex disturbance field, which comprises both steady and unsteady instabilities of similar amplitudes. It is demonstrated that the control is robust with respect to complex disturbance fields as transition is shifted from 45 to 65 % chord.

Transition Experiments on Slightly Blunted Cones with Distributed Roughness in Hypersonic Flight

‡Slightly blunted cones with smooth tips and roughened frusta were flown in the NASA Ames hypersonic ballistic range through quiescent air environments. Global surface intensity (temperature) distributions were optically measured and analyzed to determine transition onset and progression over the roughened surface. Real-gas Navier-Stokes calculations of model flowfields, including laminar boundary layer development in these flowfields, were conducted to predict key dimensionless parameters used to correlate transition on such configurations in hypersonic flow. For distributed roughness elements totally immersed within the laminar boundary layer, the critical roughness Reynolds number correlating approach was found to well model transition onset and progression over the roughened conic frusta. The critical value of the roughness Reynolds number for transition was found to be 269  20%, in agreement with the critical value of 250  20% determined in earlier experiments for transition on rough, blunt bodies in hypersonic free flight.

Anisotropic friction of rugby ball surfaces

In irregular, isotropic surface patterns with randomly distributed roughness elements, the kinetic coefficient of friction is expected to be independent of sliding direction. In regular, anisotropic surface patterns, at least two directions are expected to have different friction coefficients. Five different tackified rugby ball surfaces with different pimple heights and densities were tested; three with irregular (two directions at right angles tested) and two with regular patterns (four directions at 45° angles tested). The friction coefficient and the surface height profile were obtained in a standardised way with the Kawabata friction testing equipment. The average friction coefficients found ranged from 0.94 to 1.25. In the irregular patterns, only one exhibited a significant difference between the two perpendicular directions. In one regular surface, the difference between friction coefficients in two main directions was unexpectedly significant, in the other regular surface, an expected difference in friction coefficient turned out to be non-significant. The coefficient of friction correlated highly with roughness parameters, positively with the arithmetic mean roughness and the average slope of the roughness elements, and negatively with the kurtosis. About 18.3% of the coefficient of friction is accounted for uniquely by the arithmetic mean roughness, 46.1% by a combined contribution of the arithmetic mean roughness and the kurtosis and 0% by unique contribution of the kurtosis.

Modeling flows over gravel beds by a drag force method and a modified S–A turbulence closure

A double-averaged Navier–Stokes equations (DANS) model has been developed for depth-limited open channel flows over gravels. Three test cases are used to validate the model: an open-channel flow over a densely packed gravel bed with small-scale uniform roughness (D/d50∼13, d50=median diameter of roughness elements, D=water depth), open-channel flows over large-scale sparsely distributed roughness elements (D/Δ∼2.3–8.7, Δ=roughness height) and steep slope gravel-bed river flows with D/d50∼7–25. Various methods of treatment of the gravel-induced resistance effect have been investigated. The results show that the wall function approach (WFA) is successful in simulating flows over small gravels but is not appropriate for large gravels since the vertical profile of the longitudinal velocity does not follow the logarithmic–linear relationship. The drag force method (DFM) performs better but the non-logarithmic velocity distribution generated by sparsely distributed gravels cannot be simulated accurately. Noting that the turbulence length scale within the gravel layer is governed by the gravel size, the DANS model incorporating the DFM and a modified Spalart–Allmaras (S–A) turbulence closure is proposed. The turbulence length scale parameter in the S–A model is modified to address the change in the turbulence structure within the gravel layer. The computed velocity profiles agree well with the corresponding measured profiles in all cases. Particularly, the model reproduces the S-shape velocity profile for sparsely distributed large size roughness elements. The modeling methodology is robust and can be easily integrated into the existing numerical models.

Transition Experiments on Blunt Bodies with Isolated Roughness Elements in Hypersonic Flight

Smooth titanium hemispheres with isolated three-dimensional (3D) surface roughness elements were flown in the NASA Ames hypersonic ballistic range through quiescent CO2 and air environments. Global surface intensity (temperature) distributions were optically measured and thermal wakes behind individual roughness elements were analyzed to define tripping effectiveness. Real-gas Navier-Stokes calculations of model flowfields, including laminar boundary layer development in these flowfields, were conducted predict key dimensionless parameters used to correlate transition on blunt bodies in hypersonic flow. For isolated roughness elements totally immersed within the laminar boundary layer, critical roughness Reynolds numbers for flights in air were found to be higher than those measured for flights in CO2, i.e., it was easier to trip the CO2 boundary layer to turbulence. Tripping effectiveness was found to be dependent on trip location within the subsonic region of the blunt body flowfield, with effective tripping being most difficult to achieve for elements positioned closest to the stagnation point. Direct comparisons of critical roughness Reynolds numbers for 3D isolated versus 3D distributed roughness elements for flights in air showed that distributed roughness patterns were significantly more effective at tripping the blunt body laminar boundary layer to turbulence.