Bulletin of the American Physical Society
71st Annual Meeting of the APS Division of Fluid Dynamics
Volume 63, Number 13
Sunday–Tuesday, November 18–20, 2018; Atlanta, Georgia
Session Q31: Boundary Layer Flows over Rough Surfaces III 
Hide Abstracts 
Chair: Michael Schultz, US Naval Academy Room: Georgia World Congress Center B403 
Tuesday, November 20, 2018 12:50PM  1:03PM 
Q31.00001: Receptivity analysis of flows over structured corrugated surfaces Mihailo R Jovanovic, Wei Ran, Armin Zare We develop a modelbased approach to quantify the receptivity of laminar and turbulent channel flows over corrugated surfaces. For smallamplitude roughness elements, we utilize perturbation analysis to determine the modification to the mean flow. The dynamics of velocity fluctuations around the resulting base velocity profile are studied using the linearized NavierStokes equations. In turbulent regime, the effect of background turbulence on the mean velocity is captured by a turbulent viscosity which can be determined from secondorder statistics of velocity fluctuations. We utilize the secondorder statistics resulting from the stochastically forced linearized model to compute a correction to the turbulent viscosity. This correction in turn influences the turbulent mean velocity and modifies skinfriction drag. Our simulationfree approach paves the way for a systematic analysis of energy amplification and skinfriction drag in the presence of roughness elements of various height and spacing. 
Tuesday, November 20, 2018 1:03PM  1:16PM 
Q31.00002: Bypass transition in a boundary layer past a localized strip of distributed roughness Aditya Anand, Sourabh S Diwan In this work, we study two cases of bypass transition induced by surface roughness – one in which transition proceeds through the appearance of turbulent spots (‘spotty’ transition) and the other in which no distinct turbulent spots are seen (‘nonspotty’ transition). Velocity measurements are carried out in a boundary layer developing on a flat plate downstream of a localized strip of distributed roughness. The spotty transition is obtained at relatively low speeds and at sufficiently downstream distances from the roughness, whereas the nonspotty transition is observed at higher speeds and much closer the roughness. We use the Wavelet transform to contrast the frequencytime behavior of the two transition scenarios. Furthermore, we compare broad features of the velocity profiles representative of the two cases at similar Reynolds numbers. While the mean velocity and turbulence intensity profiles show similar variations, distinct differences are found in the shape of the energy spectrum – for the nonspotty transition, there is a single broad spectral hump, whereas for the spotty transition two humps are observed over a range of wallnormal distances. The implications of the present findings will be discussed during the talk. 
Tuesday, November 20, 2018 1:16PM  1:29PM 
Q31.00003: Roughnessinduced laminarturbulent transition prediction over complex geometries Francis Lacombe, JeanPierre Hickey 
(Author Not Attending)

Q31.00004: Abstract Withdrawn Multiple studies have shown that drag coefficients are not spatially uniform in canopy flows such as forests, urban areas and coral reefs, and yet, most studies use one single drag coefficient to model flows in such canopies. Since drag force is the main momentum term that balances the flow driving forces, the knowledge of drag coefficient spatial distribution is crucial. In this presentation we show results of drag coefficient distributions that were obtained in a randomly distributed glass plates array using three independent methods. We used PIV measurements and momentum balance calculations to extract the drag force and calculate the drag coefficient based on the local velocity. We then verified the results using two independent force measurement techniques, one is based on a swinglike device and the other is based on specially designed drag force sensors. The results show that the drag coefficient around the canopy entrance increases in the flow direction, reaching peak values (~10) and then decreases. The results also show that the drag coefficient in the fully developed region increases with depth, from zero above the canopy to a peak value near the bottom, where the velocities are low. 
Tuesday, November 20, 2018 1:42PM  1:55PM 
Q31.00005: The pressure field within the canopy of urbanlike roughness Manuel Ferreira, Bharathram Ganapathisubramani Urban areas affect the development of atmospheric boundary layers. The tendency is to characterize their effect based on density parameters, which are closely related to the effective roughness height and the zeroplane displacement. Standard morphological models successfully predict the nonlinear behaviour of these quantities, but are insensitive to the variability of the roughness height. In this context, we investigated the flow over two surfaces: a staggered array of cubes with uniform height, and a staggered array of cuboids with random height distribution but identical packing densities. Extensive measurements of the flow field were acquired using 2DPIV from which pressure was reconstructed. This enabled estimating the magnitude and location of the streamwise forces acting on individual roughness obstacles. To assess the viability of this method, the pressure distribution and the estimated friction velocity U_{τ} were compared against reported wind tunnel data and numerical simulations. The centre of pressure and magnitude of the load applied on a an arbitrary obstacle appear to be primarily dictated by the relative height of its upstream neighbor. The existence of a functional form for the normalised pressure distribution within the canopy will be addressed in this talk. 
Tuesday, November 20, 2018 1:55PM  2:08PM 
Q31.00006: Volume Averaging for Urban Canopies Manuel F. Schmid, Marco G. Giometto, Gregory A. Lawrence, Marc B. Parlange Vertical profiles of urban canopy flows are often obtained by averaging flow quantities in horizontal direction. This averaging has to account for the fact that a significant part of the averaging volume may be occupied by solid obstacles, resulting in a reduced fluid volume fraction within the canopy layer. As a consequence, the averaging operation can be defined in two different ways depending on whether the average is normalized by the total volume or by the fluid volume only. 
Tuesday, November 20, 2018 2:08PM  2:21PM 
Q31.00007: 3D3C Mean Velocity Measurements Below the Building Height in an Urban Canopy Flow Michael Benson, Gawoon Shim, John Kelly Eaton, Christopher J. Elkins The flow structure below the building tops in an urban canopy is of interest because it influences momentum transport into the canopy, contaminant dispersion, and pedestrian safety. Measurements within the canopy also illuminate mechanisms by which wall roughness modifies boundary layers. We have acquired 3component magnetic resonance velocimetry measurements throughout an array of 48 cubical buildings with a single building 3X taller. The spatial resolution of 0.06 building lengths allows identification of separation bubbles and dominant vortex structures around each building. The meanflow structure is highly sensitive to the freestream flow orientation relative to the street canyons. When they are aligned, separation bubbles fill the gap between subsequent buildings and vertical axis vortices form behind the rear corners. Separation bubbles are much larger when the flow is skewed and the streamwise mass flux within the canopy is significantly reduced. The skewed building array has larger variations in the local vertical velocity due to the presence of streamwise vortices shed from the upper building corners. In the skewed case, streamlines released at half building height meander around separation bubbles resulting in rapid spanwise transport. 
Tuesday, November 20, 2018 2:21PM  2:34PM 
Q31.00008: Effect of moderate fouling on ship drag penalty Bharathram Ganapathisubramani, Takfarinas Medjnoun, Ralf Reinartz, Manuel Ferreira, Bagus Nugroho, Jason Monty, Nicholas Hutchins Biofouled ship hull surfaces have been subject of focus for decades as they give rise to higher frictional drag, which results in less efficiency, more fuel consumption thus more emissions. These surfaces are home to various marine species such as slimes, calcareous tubeworms and barnacles, which for hydrodynamicists represent rough walls with a range of scales that can interact with the growing turbulent boundary layer developing over the hull. In order to assess the drag penalty due to the moderate biofouling, a test coupon coated with calcareous tubeworms and barnacles was scanned, scaled and replicated for windtunnel testing to determine the equivalent sandgrain roughness ks. Direct drag measurements were performed using a floating element balance, while flow field measurements were obtained by PIV. Fully rough conditions were met in the drag measurements and this allowed the estimation of ks. The wake parameter of this rough wall was found to be different from that of classical smooth walls. This information was used in conjunction with the equivalent sandgrain roughness to predict the drag for a full scale cruising ship. 
Tuesday, November 20, 2018 2:34PM  2:47PM 
Q31.00009: Drag partition revisited at low packing densities Mingwei Ge, Xiang Yang It is a direct consequence of the conventional drag partition theory that the ground skin friction and the drag coefficient of a single roughness element are nonincreasing functions of the roughness packing density. This talk will present empirical evidence from DNS and LES that challenges the above conclusion. We perform direct numerical simulations (DNS) of flow over a cube at a moderate Reynolds number. Three different aspect ratios (H/W=0.5, 1, 2) are considered. An increase in the ground skin friction is found behind the cube for all three aspect ratios. Detailed analysis shows that the increase in the ground skin friction is due to a strong downwash motion behind the cube. Due to this downwash motion, the drag coefficient on a single cubic roughness element is an increasing function of the packing density if the roughness elements are sparsely packed. This is shown by largeeddy simulations of flow over aligned and staggered arranged cube arrays. In the light of the above findings, the conventional drag partition theory may need to be revised. 
Tuesday, November 20, 2018 2:47PM  3:00PM 
Q31.00010: Detailed velocity underwater measurements obtained inside five coral reefs Shai Asher, Uri Shavit The flow of water through a coral reef creates a turbulent flow field which controls the mass, momentum and energy transport between the reef and its surroundings. Biological functions such as feeding, respiration and reproduction depend on this transport phenomenon, and yet, most coral reef flow studies treat the flow using depth averaged variables, avoiding the need to present the spatial flow variations within the reef. To overcome this limitation we present a combination of detailed laboratory and field measurements of turbulent flow properties within reefs composed of two different types of branching corals in three different, naturally common, spatial configurations. Together with detailed geometrical mapping of coral skeletons using CT scanning, our laboratory measurements suggest the existence of an additional source of mixing deep within the reef that is unique to the coral canopy geometry. Using a custombuilt submersible Particle Image Velocimetry (PIV) system, our field measurements demonstrate the effect of colony type and spatial arrangement on mean flow properties, estimated particle retention time and distribution of large coherent structures within the reef under natural flow conditions. 
Tuesday, November 20, 2018 3:00PM  3:13PM 
Q31.00011: On Plume Dispersion over Roughness Elements after a GroundLevel Line Source in Crossflows ChunHo Liu, Ziwei Mo Pollutant plume transport over urban areas is commonly observed nowadays but its parameterization for urban settings is rather limited. An analytical solution to the vertical dispersion coefficient is developed in terms of downwind distance after the source, turbulent boundary layer thickness and drag coefficient (in response to building blocks). It is then validated by wind tunnel experiments for flows over arrays of roughness elements (ribs and cubes of different sizes/separations). Water vapor is atomized by ultrasonic to simulate tracer transport after a groundlevel line source in crossflows. The tracer concentrations over various rough surfaces consistently exhibit the conventional Gaussian distribution while the vertical dispersion coefficient agrees well with the newly derived analytical solution (coefficient of determination is 0.93). Comparing the fluxes in the streamwise and vertical direction helps elucidate the transport processes and the influence of surface roughness on turbulent pollutant dispersion. The analytical solution and wind tunnel result collectively suggest an improved parameterization of vertical dispersion coefficient for the air quality forecast in urban areas. 
Tuesday, November 20, 2018 3:13PM  3:26PM 
Q31.00012: Effect of spacing on the 3D flow around a pair of cubic roughness elements embedded in a turbulent channel flow resolved using tomographic holography Jian Gao, Karuna Agarwal, Joseph Katz The effect of spacing on the flow structure around a pair of roughness cubes embedded in the inner part of a turbulent channel flow at Re_{τ}=2500 is measured using tomographic holography. The cube height, a=1 mm, corresponds to 4% of the half channel height and 90 wall units. The spacings between cubes are 1.0a, 1.6a, and 2.5a. The boundary layer separates upstream of cube, causing formation of horseshoe vortex that rolls up in front of the cube, and then wraps around it, forming a pair of counterrotating “legs”. A vortical “canopy”, dominated by wallnormal vorticity along the cube sides and spanwise vorticity above it, covers the entire cube and part of the near wake behind it, appearing as an arch surrounding the recirculation region. The canopy is asymmetric with respect to the cube center due to influence of the neighboring canopy. The highspeed jetting between the cubes deforms the adjacent canopy, realigning the vertical vorticity in the axial direction. The resulting large streamwise structures engulf the adjacent horseshoe vortex legs. This process occurs earlier with decreasing cube spacing, presumably because of the corresponding faster jetting between cubes. Conversely, behind the cubes, the streamwise momentum deficit increases with decreasing spacing. 
Follow Us 
Engage
Become an APS Member 
My APS
Renew Membership 
Information for 
About APSThe American Physical Society (APS) is a nonprofit membership organization working to advance the knowledge of physics. 
© 2023 American Physical Society
 All rights reserved  Terms of Use
 Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 207403844
(301) 2093200
Editorial Office
1 Research Road, Ridge, NY 119612701
(631) 5914000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 200452001
(202) 6628700