Rough Configurations Research Articles

Application of a Distributed Element Roughness Model to Additively Manufactured Internal Cooling Channels

Abstract Design for cooling effectiveness in turbine blades relies on accurate models for dynamic losses and heat transfer of internal cooling passages. Metal additive manufacturing (AM) has expanded the design space for these configurations, but can give rise to large-scale roughness features. The range of roughness length scales in these systems makes morphology resolved computational fluid dynamics (CFD) impractical. However, volumetric roughness models can be leveraged, as they have computational costs orders of magnitude lower. In this work, a discrete element roughness model (DERM), based on the double-averaged Navier–Stokes equations, is presented and applied to additively manufactured rough channels, representative of gas turbine blade cooling passages. Unique to this formulation of DERM is a generalized sheltering-based treatment of drag, a two-layer model for spatially averaged Reynolds stresses, and explicit treatment of dispersion. Six different AM rough surface channel configurations are studied, with roughness trough to peak sizes ranging from 15% to 60% nominal channel passage half-width, and the roughness Reynolds number ranges from Rek = 60 to 300. DERM predictions for spatially and temporally averaged mean flow quantities are compared to previously reported direct numerical simulation results. Good agreement in the mean velocity profiles, stress balances, and drag partitions are observed. While DERM models are typically calibrated to specific deterministic roughness morphologies at comparatively small roughness Reynolds numbers, the present more generalized DERM formulation has wider applicability. Here, it is demonstrated that the model can accommodate random roughness of large scale, typical of AM.

Modeling rough walls from surface topography to double averaged Navier-Stokes computation

The discrete element method was recently revisited using a double averaged Navier-Stokes formulation [Chedevergne F. A double-averaged navier-stokes turbulence model for wall flows over rough surfaces with heat transfer. J Turbul. 2021 Sep;22(11):713–734. doi:10.1080/14685248.2021.1973014] and a new closure relation for the drag coefficient [Chedevergne F, Forooghi P. On the importance of the drag coefficient modelling in the double averaged navier-stokes equations for prediction of the roughness effects. J Turbul. 2020 Aug;21(8):463–482. doi:10.1080/14685248.2020.1817465]. The developed model lies on the notion of representative elementary roughness whose characterisation needs to be generalised to provide a rigorous definition for randomly distributed rough configurations. From 3D scans of rough surfaces and simple image processing, a procedure was proposed to compute the blockage factor and the elementary diameter, the two main parameters of the representative elementary roughness. The procedure was successfully applied to two experimental configurations [Squire D, Morrill-Winter C, Hutchins N, et al. Comparison of turbulent boundary layers over smooth and rough surfaces up to high reynolds numbers. J Fluid Mech. 2016;795:210–240; Croner E, Léon O, Chedevergne F. Industrial use of equivalent sand grain height models for roughness modelling in turbomachinery. In: 55th 3AF International Conference on Applied Conference; Poitiers, France; Apr 2021. https://hal.archives-ouvertes.fr/hal-03228846]. Computed velocity profiles match experimental ones when the Reynolds number is varied, showing at the same time the relevance of the procedure and the validity of the double averaged Navier-Stokes model across the different rough regimes.

Flow patterns and free-surface dynamics in hydraulic jump on pebbled rough bed

Some basic characteristics of a classic hydraulic jump flow over a pebbled rough bed, as well as on a smooth bed as a reference, are presented in this experimental study. For the experiments, an inflow Froude number Fr1from 1.54 to 4.94 and inflow Reynolds number Re1from 42 000 to 230 000 were considered. Visual observations and measurements suggested some differences between the formation of a hydraulic jump on rough and smooth bed configurations, including different air entrainment processes, larger vortical structures in the roller length and stronger backward flow in the upper layer. Furthermore, the jump roller and aerated flow lengths were shorter on a pebbled rough bed than on a smooth bed, while the dimensionless advection velocity of large vortices was the same for both bed types. The instantaneous jump toe perimeter showed the largest variation at the largest Fr1and was generally larger on rough bed than on smooth bed. Larger oscillations of the free-surface profile were observed on smooth bed, highlighting that roughness resulted in smaller free-surface oscillations, suggesting the higher rate of energy dissipation.

A Computational Fluid Dynamics Investigation of using Large-Scale Geometric Roughness Elements in Open Channels

The hydraulic behavior of the flow can be changed by using large-scale geometric roughness elements in open channels. This change can help in controlling erosions and sedimentations along the mainstream of the channel. Roughness elements can be large stone or concrete blocks placed at the channel's bed to impose more resistance in the bed. The geometry of the roughness elements, numbers used, and configuration are parameters that can affect the flow's hydraulic characteristics. In this paper, velocity distribution along the flume was theoretically investigated using a series of tests of T-shape roughness elements, fixed height, arranged in three different configurations, differ in the number of lines of roughness element. These elements were used to find the best configuration of roughness elements that can be applied to change the flow's hydraulic characteristics. ANSYS Parametric Design Language, APDL, and Computational Fluid Dynamics, CFD, was used to simulate the flow in an open channel with roughness elements. CFD can be used to study the hydrodynamics of open channels under different conditions with inclusive details rather than relying on the costly field and time-consuming. Runs were implemented with different conditions, the discharge, and water depth in upstream and downstream of the flume. T-shape roughness elements with height equal to 3cm placed in three different configurations, two lines, four lines, and fully rough configurations were tested. The results show that the effect of roughness elements increasing with increasing the number of lines of roughness elements. Cases of four lines and fully rough configurations have almost the same hydraulic performance by having the same results of the velocity decrease percentage, which is decreased by approximately about 66% and 61% of the control case's velocity in the zone near the roughness elements consequently. But the difference is that four lines configuration is affected in a part of the test section. This behavior increases the velocity values by about 11% in the other side and by about 10% near the free surface in the case of four lines configuration and increased by about 32% above the roughness elements in a fully rough configuration.

A Computational Fluid Dynamics Investigation of using Large-Scale Geometric Roughness Elements in Open Channels

The hydraulic behavior of the flow can be changed by using large-scale geometric roughness elements in open channels. This change can help in controlling erosions and sedimentations along the mainstream of the channel. Roughness elements can be large stone or concrete blocks placed at the channel's bed to impose more resistance in the bed. The geometry of the roughness elements, numbers used, and configuration are parameters that can affect the flow's hydraulic characteristics. In this paper, velocity distribution along the flume was theoretically investigated using a series of tests of T-shape roughness elements, fixed height, arranged in three different configurations, differ in the number of lines of roughness element. These elements were used to find the best configuration of roughness elements that can be applied to change the flow's hydraulic characteristics. ANSYS Parametric Design Language, APDL, and Computational Fluid Dynamics, CFD, was used to simulate the flow in an open channel with roughness elements. CFD can be used to study the hydrodynamics of open channels under different conditions with inclusive details rather than relying on the costly field and time-consuming. Runs were implemented with different conditions, the discharge, and water depth in upstream and downstream of the flume. T-shape roughness elements with height equal to 3cm placed in three different configurations, two lines, four lines, and fully rough configurations were tested. The results show that the effect of roughness elements increasing with increasing the number of lines of roughness elements. Cases of four lines and fully rough configurations have almost the same hydraulic performance by having the same results of the velocity decrease percentage, which is decreased by approximately about 66% and 61% of the control case's velocity in the zone near the roughness elements consequently. But the difference is that four lines configuration is affected in a part of the test section. This behavior increases the velocity values by about 11% in the other side and by about 10% near the free surface in the case of four lines configuration and increased by about 32% above the roughness elements in a fully rough configuration.

The effect of surface roughness on laser-induced stress wave propagation.

We investigate laser-induced acoustic wave propagation through smooth and roughened titanium-coated glass substrates. Acoustic waves are generated in a controlled manner via the laser spallation technique. Surface displacements are measured during stress wave loading by the alignment of a Michelson-type interferometer. A reflective coverslip panel facilitates capture of surface displacements during loading of as-received smooth and roughened specimens. Through interferometric experiments, we extract the substrate stress profile at each laser fluence (energy per area). The shape and amplitude of the substrate stress profile are analyzed at each laser fluence. Peak substrate stress is averaged and compared between smooth specimens with the reflective panel and rough specimens with the reflective panel. The reflective panel is necessary because the surface roughness of the rough specimens precludes in situ interferometry. Through these experiments, we determine that the surface roughness employed has no significant effect on substrate stress propagation and smooth substrates are an appropriate surrogate to determine stress wave loading amplitude of roughened surfaces less than 1.2 μm average roughness (Ra). No significant difference was observed when comparing the average peak amplitude and loading slope in the stress wave profile for the smooth and rough configurations at each fluence.

Effect of realistic ballasted track in the underbody flow of a high-speed train via CFD simulations

The underbody flow of high-speed trains is a matter of concern in the railway industry because of its relation to the ballast flight phenomenon, and several experimental campaigns are focused on evaluating its characteristics. Computational simulations are considered as a flexible and economical alternative. However, the aforementioned simulations include only simplified track models with a flat ground that do not include the sleepers or the profiles of stones. The present study uses a previously developed methodology to reproduce the passing of a high-speed train over a scanned real profile of stones and sleepers and determines the effect on the surrounding flow and compares the results relative to those of a flat ballast bed. The simulations are validated relative to experimental data published in the bibliography with satisfactory results. The results confirm expected differences between flat and rough configurations, thereby revealing a more chaotic velocity field near the rough ballast and the presence of a larger number and smaller size of turbulence structures. The rough configuration also allows for the measurement of the pressure on the surface of the stones and the application of a predictive model for ballast flight. The obtained results are similar to those in experimental tests.

CFD modelling of wave damping over a fringing reef in the Colombian Caribbean

The understanding of physical processes over submerged reefs represents an important ongoing research topic when considering wave energy dissipation and coastal protection that these environments provide. Detailed analyses are required to assess wave damping based on the contribution of reef roughness and wave breaking. For this purpose, the CFD (computational fluid dynamics) toolbox OpenFOAM® is applied to simulate the wave energy dissipation process over reefs with explicit accounting for the complexities of coral shape instead of commonly applied parameterized approaches for bottom roughness and wave breaking. Model validation was performed through comparison with field measurements over a reef profile of Tesoro Island in the Colombian Caribbean. Quantitative analysis of wave damping caused by wave breaking and reef roughness was conducted for (1) moderate and extreme wave conditions, (2) smooth and rough seabed configurations and (3) changes in the water depth over the reef crest. Wave height attenuation is found to vary along the reef profile reaching differences of up to 55% between smooth and rough reef surface scenarios, particularly for moderate wave conditions. Wave breaking, high turbulent flows and detachment of undertow currents are among the reef roughness effects on hydrodynamics. The fore-reef terrace and the reef crest are identified as the most critical zones where dissipation takes place. Wave breaking from rough seabeds provides a global wave attenuation of 75.4–94.8%, with the reef roughness alone accounting for ~ 4–14%. Under extreme wave height scenarios, the wave damping from reef roughness is not significant. Further predictions regarding roughness effects on the reef hydrodynamics, wave set-up and undertow currents for moderate and extreme wave climate conditions are also shown. Directions for future research using CFD are presented to address limitations that arise from the limited span-wise domain in our approach that prevents development of large lateral coherent structures.

Analytical wall function including roughness corrections

Inspired by the work of Aupoix (2015a,b) and relying on the analytical wall function of Suga et al. (2006), this paper proposes a modified version of the wall model capable of accounting for roughness effects. The thermal correction was enhanced to capture roughness effects due to the increase of the wetted surface of the walls. A derivation of the model adapted to configurations with very large roughness is also proposed. The new model is compared to the former analytical wall function formulation using several rough configurations for which experimental data are available. The validation test cases have been chosen to highlight the improvements brought by the present work.

Dissolution of minerals with rough surfaces

We study dissolution of minerals with initial rough surfaces using kinetic Monte Carlo simulations and a scaling approach. We consider a simple cubic lattice structure, a thermally activated rate of detachment of a molecule (site), and rough surface configurations produced by fractional Brownian motion algorithm. First we revisit the problem of dissolution of initial flat surfaces, in which the dissolution rate rF reaches an approximately constant value at short times and is controlled by detachment of step edge sites. For initial rough surfaces, the dissolution rate r at short times is much larger than rF; after dissolution of some hundreds of molecular layers, r decreases by some orders of magnitude across several time decades. Meanwhile, the surface evolves through configurations of decreasing energy, beginning with dissolution of isolated sites, then formation of terraces with disordered boundaries, their growth, and final smoothing. A crossover time to a smooth configuration is defined when r=1.5rF; the surface retreat at the crossover is approximately 3 times the initial roughness and is temperature-independent, while the crossover time is proportional to the initial roughness and is controlled by step-edge site detachment. The initial dissolution process is described by the so-called rough rates, which are measured for fixed ratios between the surface retreat and the initial roughness. The temperature dependence of the rough rates indicates control by kink site detachment; in general, it suggests that rough rates are controlled by the weakest microscopic bonds during the nucleation and formation of the lowest energy configurations of the crystalline surface. Our results are related to recent laboratory studies which show enhanced dissolution in polished calcite surfaces. In the application to calcite dissolution in alkaline environment, the minimal values of recently measured dissolution rate spectra give rF∼10-9 mol/(m2 s), and the calculated rate laws of our model give rough rates in the range 10-6–10-5 mol/(m2 s). This estimate is consistent with the range of calcite dissolution rates obtained in a recent work after treatment of literature data, which suggests the universal control of kink site dissolution in short term laboratory works. The weak effects of lattice size on our results also suggest that smoothing of mineral grain surfaces across geological times may be a microscopic explanation for the difference of chemical weathering rate of silicate minerals in laboratory and in the environment.

Reducing Darcy coefficient by using drag reduction methods in open-channel flows: Effect on discharge capacity and potential application to mitigate river flooding impact

Drag reduction by polymer addition is a common strategy used to minimize friction losses in pipe flows but has not been tested in river flows. Present paper then aims at measuring backwater curves and velocity profiles within smooth and rough bed flume configurations to assess the capabilities of such polymer addition to decrease the water depth with regards to the use of plain water and thus increase the channel conveyance. The inclusion of a limited amount of polymers proves to be able to reduce the typical Darcy-Weisbach friction coefficient with regards to plain water by a factor 2 in smooth bed conditions and a factor 1.5 in rough bed conditions. Moreover, the vertical profiles of streamwise velocity appear to be hardly affected by the addition of such polymers. Whether such drag reduction would still be effective in real watercourses remains unknown and would now require field experiments at larger scale.

Selective enhancement of Selényi rings induced by the cross-correlation between the interfaces of a two-dimensional randomly rough dielectric film

By the use of both perturbative and non-perturbative solutions of the reduced Rayleigh equation, we present a detailed study of the scattering of light from two-dimensional weakly rough dielectric films. It is shown that for several rough film configurations, Selényi interference rings exist in the diffusely scattered light. For film systems supported by dielectric substrates where only one of the two interfaces of the film is weakly rough and the other planar, Selényi interference rings are observed at angular positions that can be determined from simple phase arguments. For such single-rough-interface films, we find and explain by a single scattering model that the contrast in the interference patterns is better when the top interface of the film (the interface facing the incident light) is rough than when the bottom interface is rough. When both film interfaces are rough, Selényi interference rings exist but a potential cross-correlation of the two rough interfaces of the film can be used to selectively enhance some of the interference rings while others are attenuated and might even disappear. This feature may in principle be used in determining the correlation properties of interfaces of films that otherwise would be difficult to access.

Semantic Object Segmentation in Tagged Videos via Detection.

Semantic object segmentation (SOS) is a challenging task in computer vision that aims to detect and segment all pixels of the objects within predefined semantic categories. In image-based SOS, many supervised models have been proposed and achieved impressive performances due to the rapid advances of well-annotated training images and machine learning theories. However, in video-based SOS it is often difficult to directly train a supervised model since most videos are weakly annotated by tags. To handle such tagged videos, this paper proposes a novel approach that adopts a segmentation-by-detection framework. In this framework, object detection and segment proposals are first generated using the models pre-trained on still images, which provide useful cues to roughly localize the semantic objects. Based on these proposals, we propose an efficient algorithm to initialize object tracks by solving a joint assignment problem. As such tracks provide rough spatiotemporal configurations of the semantic objects, a voting-based refinement algorithm is further proposed to improve their spatiotemporal consistency. Extensive experiments demonstrate that the proposed framework can robustly and effectively segment semantic objects in tagged videos, even when the image-based object detectors provide inaccurate proposals. On various public benchmarks, the proposed approach obtains substantial improvements over the state-of-the-arts.

Turbulence modulation by micro-particles in smooth and rough channels

Inertial micro-particles, dispersed in turbulent flows, are affected by the local dynamics of the carrier flow field. The presence of a relative dilute loading of particles entails some relevant changes in the mean velocity field. Moreover, the roughness of the solid boundary affects both the carrier and the carried phase dynamics and their mutual interaction. The underlying physics has been exhaustively tackled in the so-called one-way coupling regime, i.e., negligible action of particles onto the fluid, whereas many aspects of the particle back-reaction have been much less investigated and are still poorly understood. Thus the research in this area remains very active. In order to understand the mutual interaction between carrier and carried phases, direct numerical simulations are particularly suitable, even at low Reynolds number. Here, particle laden flow over a complex domain is investigated at friction Reynolds number Reτ = 180. The numerical analysis is based on the Euler-Lagrange approach, taking into account the fluid-particle interaction (two-way coupling). Point forces are used to represent the back-effect of particles on the turbulence and the effect of the wall’s roughness is taken into account modeling the elastic rebound of particles onto it, instead of using a virtual rebound model. The interest is focused on the effect of micro-particles of different inertia on fluid and particle statistics in the near-wall region. In particular, turbulent and solid phase statistics are compared with those obtained for a one-way coupled flow, for the same Reynolds number in smooth and rough channel flow configurations.

2D numerical flow modeling in a macro‐rough channel

AbstractA 2D numerical flow model, developed at the Research unit of Hydrology, Applied Hydrodynamics and Hydraulic Constructions at ULg, has been applied to flows in a macro‐rough channel. The model solves the shallow water equations (SWE) with a two length scale, depth‐integrated k‐type approach for turbulence modeling. Data for the comparison have been provided by experiments conducted at the Laboratory of Hydraulic Constructions at EPFL. In the experiments with different non‐prismatic channel configurations, namely large‐scale cavities at the side walls, three different 2D flow characteristics could be observed in cavities. With the used numerical model features, especially regarding turbulence and friction modeling, a single set of bottom and side wall roughness could be found for a large range of discharges investigated in a prismatic channel. For the macro rough configurations, the numerical model gives an excellent agreement between experimental and numerical results regarding backwater curves and flow patterns if the side wall cavities have low aspect ratios. For configurations with high aspect ratios, the head loss generated by the preservation of important recirculation gyres in the cavities is slightly underestimated. The results of the computations reveal clearly that the separation of turbulence sources in the mathematical model is of great importance. Indeed, the turbulence related to 2D transverse shear effects and the 3D turbulence, generated by bed friction, can have very different amplitude. When separating these two effects in the numerical models, most of the flow features observed experimentally can be reproduced accurately. Copyright © 2009 John Wiley & Sons, Ltd.

The effect of the slope of irregularly distributed roughness elements on turbulent wall-bounded flows

Wall roughness produces a downward shift of the mean streamwise velocity profile in the log region, known as the roughness function. The dependence of the roughness function on the height and arrangement of roughness elements has been confirmed in several studies where regular rough walls were analysed; less attention has been paid to non-regular rough walls. Here, a numerical analysis of turbulent flows over irregularly shaped rough walls is performed, clearly identifying the importance of a parameter, called the effective slope (ES) of the wall corrugations, in characterizing the geometry of non-smooth irregular walls. The effective slope proves to be one of the fundamental geometric parameters for scaling the roughness function. Specifically, for a moderate range of roughness heights, both in the transitionally and in the fully rough regime, ES appears to scale the roughness function for a wide range of irregular rough geometric configurations. The effective slope determines the relative importance of friction drag and pressure drag. For ES ~ 0.15 we find that the friction contribution to the total wall stress is nearly in balance with the pressure-drag contribution. This value separates the region where the roughness function ΔU+ = f(ES) is linear from that where a smooth nonlinear behaviour is observed. In the cases investigated, value ES ~ 0.15 also separates the transitionally rough regime from the fully rough regime.

Rough Gauge Fields, Smearing and Domain Wall Fermions

At a fixed lattice spacing, as determined by say m_\rho, adding additional fermion flavors to a dynamical simulation produces rougher gauge field configurations at the lattice scale. For domain wall fermions, these rough configurations lead to larger residual chiral symmetry breaking and larger values for the residual masses, m_{res}. We discuss ongoing attempts to reduce chiral symmetry breaking for N_f = 3 dynamical domain wall fermion simulations by different smoothing choices for the gauge fields.

Reynolds Number and Roughness Effects on Thick Airfoils for Wind Turbines

The effects of increased Reynolds number (Re) on clean and rough airfoils were investigated by wind tunnel experiments and by numerical simulation with an improved Navier-Stokes solver. By cooling the cryogenic wind tunnel in Cologne (KKK) of the German-Dutch Wind-Tunnel (DNW) down to 100 K, values of Re up to 10 × 106 were reached, based on a model chord of 500 mm. Whereas the clean surface configuration shows no drastic loss in maximum lift, the lift-to-drag ratio decreases from 95 to approximately 85, mainly due to an increase of drag. In contrast to that, rough surface configurations, including zig-zag tape for transition fixing and carborundum based distributed roughness around the nose, show increased performance for increased Re. In parallel, numerical simulations were carried out with a boundary layer stability code coupled to a Navier-Stokes solver. The trends in behaviour of the boundary-layer based properties were predicted well, in contrast to the separation pattern responsible for maximum lift.

Anderson localization of 3He with variable disorder provided by a 4He solid/liquid interface

When liquid helium is in contact with a graphite surface, and the pressure of the liquid is increased toward the bulk solidification pressure, the solid helium will pass through smooth and rough configurations during layering transitions. When 3He is added to the system, it preferentially resides at the solid/liquid interface. When the interface is atomically sharp, the 3He will diffuse easily on the surface, but when the interface is rough, the 3He, with a deBroglie wavelength of ∼1.5 nm, may experience Anderson localization. Currently we are using an NMR technique, which monitors the diffusion of the 3He, as a probe of this effect.

Effect of Rough Operating Surface on Rotary Spreader Distribution Pattern

Fertilizer spreader pattern tests are generally conducted on smooth surfaces, but in actual use, the spreaders are frequently operated on rough surfaces. A turf spreader, for example, must operate over tree and shrub roots, rocks, etc. in the turf. A laboratory test was conducted in which rough surfaces were simulated with tracks made of steel rods. Four materials were tested on a smooth surface and on four different rough tracks. The results showed significant skewing and pattern deterioration with all of the materials and with all rough track configurations.