Detailed schlieren and laser lightsheet visualizations of the near wake of micro vortex
generator (MVG) revealed large structures that were different from those of the undisturbed
turbulent boundary layer. These structures were attributed to the rapid breakdown
of the primary trailing vortex pair. The breakdown was thought to arise from a cylindrical
Kelvin–Helmholtz-like instability surface. The structures appear to be hairpin or ring-like
in nature that showed eruptions into the freestream flow, entraining it.
AIAA 2010-462340th Fluid Dynamics Conference and Exhibit1im28 June -1 July 2010, Chicago, Illinois Experimental Study of Near Wake of Micro VortexGenerators in Supersonic Flow Frank K. Lu.* Adam J. Piercef and Yusi Shiht University of Texas at Arlington, Arlington, Texas, 76019 Detailed schlieren and laser lightsheet visualizations of the near wake of micro vortexgenerator (MVG) revealed large structures that were different from those of the undis-turbed turbulent boundary layer. These structures were attributed to the rapid breakdownof the primary trailing vortex pair. The breakdown was thought to arise from a cylindricalKelvin-Helmholtz-like instability surface. The structures appear to be hairpin or ring-likein nature that showed eruptions into the freestream flow, entraining it. I. Introduction .ICRO vortex generators (MVGs), whose height is less than the boundary layer thickness, have been.lately proposed for mitigating the adverse effects of shock/boundary-layer interactions (SBLIs).1-17There is belief that MVGs energize the boundary-layer by vortex entrainment of the freestream althoughthe precise physical mechanisms are still not well understood. This belief is based on surface flow visualiza-tions6,11,17 which indicate two bright lines downstream of the MVG that apparently disappears after abouttwo MVG lengths downstream (Fig.1). The disappearance of these lines has been suggested to be due tothe vortices lifting off the surface due to upwash from their mutual interaction.6,11However, Lu et al.17through a detailed study of the topology around an MVG suggest that symmetry breaking at the rear ofthe MVG leads to a unsteady wake that cannot be captured by surface flow visualization, this techniquebeing regarded as a steady-state one. Moreover, digitally enhanced images18 show that these two topologicallines actually trail further downstream than previously thought. Further, it is not clear from topologicalconsiderations that the pair of (primary) vortices are specifically associated with the pair of bright lines.Additionally, based on observations of experimental and numerical surface flow visualizations, topologicalrules suggest a number of secondary vortex pairs and vortex filament pairs being shed in addition to theprimary vortex pair. There is actually a distinct possibility that the bright lines seen in the surface flowvisualization are due to a pair of corner (secondary) vortices. A different downstream flowfield was suggested by Blinde et al. via detailed particle image velocimetry(PIV) mapping. TThhee pposotsutluatlaedt efdlowfield, shown in Fig. 2, includes hairpin vortices streaming from thetrailing edges of an MVG array, separated by high-speed channels. There are no trailing vortices in thismodel and this can be interpreted to be due to the breakdown of the trailing vortices but interpreted byothers as a lifting from the mutual upwash of the trailing vortex pair. PIV unfortunately was unable toresolve the extremely small vortex filaments. The mechanism that affects the subsequent shock/boundary-layer interaction, especially separated inter-actions, also remains the subject of investigation. Thus, the aforementioned model of entrainment of the ( *Professor and D irector, A erodynamics Research Ce n ter, De p artmen t of Mechanical and Aero s pace Engin e ering. Associate Fellow AIAA. ) ( TGraduate Research Assistant, Aerodynamics R e search C e nter, D e partment of Me c hanical and Aerospace Eng i neering.Student Member AIAA. ) freestream by the counter-rotating trailing vortices to energize the boundary layer may be insufficient. Itappears that Blinde et al.’s’ model may point to new physical mechanisms involving impinging hairpin orring vortices on the shock, thereby disrupting it. The present paper reports flow visualizations which suggestthat the trailing vortices downstream of the MVG breaks down. Symmetry breaking19,20 at the rear of theMVG destabilizes the trailing vortices in what can be considered as a Kelvin-Helmholtz-like instability.This instability is seen to cause the trailing vortices to breakdown rapidly into large vortical structures,identified previously as vortex rings in Li and Liu15,16 and as hairpin vortices by Blinde et al.7 Here, theemphasis is on identifying the vortex breakdown mechanism experimentally. This breakdown mechanismcan happen rapidly unlike the upwash of a trailing vortex pair in supersonic flow. This latter point is basedon observations of the persistence of the signature of trailing vortices shed from protuberances.22 Figure 1. Surface flow visualization showingdisappearance of wake signature.17 Figure 2. Conceptual sketch of flow downstream of an MVGarray, showing the presence of hairpin vortices.' Before examining the supersonic wake of an MVG, it will be useful to briefly review the effect of three-dimensionality on a vortex line perpendicular to the flow direction convecting past a three-dimensionalsurface mounted obstacle. The vorticity equation, ignoring body forces, can be written as The middle two terms on the RHS of the above equation are absent for incompressible flow while the lastterm is the viscous dissipation of vorticity, being the stress tensor. For a two-dimensional flow, the firstterm on the RHS also vanishes since the velocity vector lies in a plane that is perpendicular to the vorticityvector. Therefore, vorticity is not created and, instead, for a steady, two-dimensional flow, is dissipated. However, for an obstacle with a finite span mounted on a surface, an incoming, two-dimensional boundarylayer that the obstacle is submerged in develops three-dimensionality.a Therefore, the first term on the RHS,which is known as the vortex stretching term, is now nonzero due to crossflow terms so that vorticityis created. As has been observed, the vorticity is distributed in such a way that one or more so-calledhorseshoe vortices wrap around the object, regardless ofwhether the protuberance is sub-boundary layer inheight or taller.26-28 Such a horseshoe vortex system is extremely persistent since lengthening it increasesits vorticity and its well-organized structure even in a turbulent flow. Studies of the downstream evolutionof the horseshoe vortex have shown that it eventually breaks down, depending on the Reynolds number ( Flows past tall obstacles a re k nown a s junction flows. Both protuberance an d junction flows ar e common occu r rences, ranging f rom flows past buildings and other structures, v e hicles and circuit boards. M o reover, pr o tuberances of various shapesare used f or enhancing heat t r ansfer,23,24 i mproving f l ows when occurring as vo r tex generators and for tripping boundary layers.25 ) of the flow and the geometry of the obstacle.29 Compressibility introduces extra terms into the vorticityequation, namely, the second term on the RHS is vortex stretching due to dilatation and the third termis known as the baroclinic torque. Depending on the velocity, density and pressure gradients, these termscan enhance or diminish vorticity. Specifically, for protrusions, such as the MVG in the present study,these gradients increase vorticity due to shocks forming around the obstacle, amongst other causes. Thevorticity is mostly confined within the boundary layer, with some arising from the shock envelope aroundthe MVG. The confined vorticity is forced into trailing vortices which stretch, strengthening them. Thesecounter-rotating trailing vortices are regions with high momentum. Sidewash caused by the MVG results ina low momentum, or momentum deficit, region along the centerline of the MVG wake4,6,7,11,14-16,30 and thehigh-momentum vortex flows on either side. The shape of the MVG is such that the trailing vortices convergewhich destabilizes them. Finally, visualizations of a small negative value of X2 [31] and an instantaneousnumerical schlieren image are shown in Fig. 3.15 Figure 3(a) shows what are thought to be vortex rings.Analogous, large density structures are visible in the instantaneous schlieren image of Fig. 3(b). Thesenumerical visualizations reveal large structures using different diagnostics than the stereo PIV mapping ofBlinde et al. (a) Iso-surface of a small negative value of X2. (b) Instantaneous numerical schlieren. Figure 3. Visualizations of [15. Lightsheet visualizations are reported in this paper, including a rarely used local lightsheet technique.32,33These visualizations are combined with schlieren visualizations to reveal a highly unsteady wake with largeroller structures. There appears to be a distinct difference between these structures and those in the turbulentboundary layer ahead. The present visualizations provide further evidence of a potential vortex breakdownmechanism. The train of either hairpin vortices or vortex rings impinging on the downstream shock isthought to be the key mechanism for disrupting any separated SBLIs downstream which are the subject offuture papers. II. Experiment The experimental approach is summarized here with details available in [34]. A blowdown wind tunneloperating at a Mach number of 2.47±0.005 was used. The tunnel reached steady-state pressure conditionsin 2-3 s.35 The total pressure was generally kept at 200±6 kPa (29±0.9 psia). The run times of less than 10 s resulted in a total temperature drop of only 1-2 K. The temperature can be considered to be nearlysteady and so can the unit Reynolds number be considered at 43 million per m. The test section was 15.2 cm² × 81.28 cm long(6 in.2× 2.67 ft). It was outfitted with extensiveoptical access from both sides and from the top.A flat plate, 73 cm long(28.75 in.), with a sharpleading edge of 15 deg was mounted in the test sec-tion36 to develop a boundary layer naturally, withtransition occurring within 3 cm (1.2 in.) from theleading edge (Fig. 4). The flat plate was supportedin the test section by a sharp-tipped rail on eachside. The plate was made in layers. The top layerwas made of a number of small, thin plates buttedtightly together to form a continuous, flat surface.This modular design allowed for quick configurationchanges. A cavity below allowed pressure tubing,transducer wiring and other elements to be placed. Figure 4.Test section showing flat plate.Flow fromright to left. The wiring and tubing were channeled to the rear from the side of the test section. Finally, a bottom coverencases the cavity. Six MVGs in the form of skewed tetrahedra were distributed evenly across the flat plate such that theirleading edges were located 272 mm (10.7 in.) downstream of the leading edge of the flat plate, Fig. 5. TheMVGs were 12.95 mm (0.51 in.) long and 1.57 mm (0.062 in.) high. The front of each MVG was 11.7 mm(0.46 in.) wide. The center-to-center spacing between the MVGs was 30.5 mm (1.2 in.). Two styles of MVGswere fabricated based on the designs from [16], with the trailing edge angle of either 45 or 70 deg, labeled asMVG45 and MVG70 respectively. Flow visualizations of either MVG45 and MVG70 are qualitatively verysimilar. The “nvg431”style based on [37] was also available. Figure 5. Micro-vortex generator array. Diagnostics that were applied included schlieren still and video imaging,38 mean surface pressure mea-surements downstream of an MVG and at the bisection of two MVGs, surface flow visualization18 and laserlightsheet visualization using PIV hardware and software. The same setup was used to obtain quantitativevelocity data. The schlieren images were obtained by a Nikon D300S camera with a VR 18-200 mm f/3.5-5.6lens. In still mode, the shutter speed is 1/8000 s and the image consisted of 12.3 megapixels. High-definitionvideo from the camera was at 24 fps and 720 p. Due to the uniqueness of the lightsheet visualization, some details are provided here. A schematic ofthe laser lightsheet system is given in Fig. 6. The imaging hardware is a LaVision FlowMaster 3D SPIVsystem that was installed on a custom-built mounting system. A cylindrical lens with a focal length of-10 mm spreaded a beam of light from a New Wave Research Solo 120 Nd:YAG double-pulsed laser into a 2.5 mm thick sheet. The figure shows the lightsheet incident on the centerline of the flat plate, which also bisected one of the MVGs. An alternate arrangement had the lightsheet in a crosswise plane, 20.6 (0.81 in.)downstream of the MVG trailing edge. The lasers were pulsed at 5 Hz with a pulse duration of 0.7 us.The images were captured by a FlowMaster 3S ImagerIntense/ImagePro ICCD camera with 1376 × 1040 pixel resolution and a 12-bit dynamicrange. Other major components of the imaging system included a frame grabber, control electronics and ahost computer running the DaVis 7.1 PIV data acquisition and processing software. A high-pressure seeding system was installed in the plenum chamber of the wind tunnel to spray calciumcarbonate particles (Specialty Minerals CalEssence 70) with an average diameter of 0.7 um. It is thoughtthat this is the first time that calcium carbonate has been used for such a purpose.39TThe suitability ofcalcium carbonate for seeding high-speed aerodynamics flows and a novel surface treatment to reduce glareare discussed in 34. In addition to seeding the flow from the plenum chamber, local seeding was accomplished by naturallyaspirating acetone from a 2.8 mm (0.11 in.) diameter surface pressure tap 138 mm (5.44 in.) upstreamalong the centerline of an MVG. A flexible tubing connected the pressure tap to a vial of acetone. Sincethe pressure in the test section was subatmospheric during a run, the acetone was drawn into the boundarylayer via the tap. This local lightsheet visualization technique had been used previously in three-dimensionalshock/boundary layer interaction studies.33,40 Figure 6. Schematic of laser lightsheet visualization hardware. III. Results and Discussion A. Schlieren Visualization Figure 7 shows schlieren images of the boundary layer past the flat plate and with MVG45. The flow isfrom right to left as indicated by the large arrow. Due to the high resolution of the cameras, the imagesshow a large number of weak wave reflections which can be traced to joints and surface roughness. Notingthat the schlieren technique is sensitive to density gradients, the boundary layer is visible as a bright bandwith the horizontally mounted knife edge. Figure7(a) shows a slight irregularity on the boundary-layer edgeattributable to a wave from the side wall. Since a schlieren image is integrated along the line of sight, thisirregularity appears exaggerated by the dark wave reflection off the boundary layer. Figure 7(b) shows a curved oblique wave system impinging from the leading edge of the MVG. Thecurvature is due to the lower Mach number of the flow closer to the surface. The wave system appears tohave a 入-foot shape but with the leading and trailing legs inclined close to one another, indicating a weak Figure 7. Schlieren images of Mach 2.47 flow. separation. Downstream is a light uniform zone due to the flow past the top of the MVG. This is followed by a darkzone as the flow expands past the trailing edge of the MVG. A more complicated wave system is visibleimpinging the top of the MVG rear. These waves coalesce just at the edge of the boundary layer to form asingle wave. These features are visible in Babinsky et al. as well as in the instantaneous schlieren imagesof Herges et al. An important feature in Fig.7(b) is a broad wispy line downstream of the MVG, this being pointed to bythe thin white arrow. This feature is apparently not visible in the instantaneous schlieren images of Hergeset al.11but is visible and highlighted in Babinsky et al. Babinsky et al. identified this faint line as theouter boundary of a low-momentum region. B. Lightsheet Visualization Figure 8 presents global laser lightsheet visualization along the centerline of MVG70..The flat plate isoutlined in the image for clarity. Due to the multiple MVGs and the camera perspective, blurred imagesof the other MVGs outside of the camera’s depth of focus are also visible in Fig. 8. The figure shows largestructures in the undisturbed boundary layer as faint dark objects inclined downstream. In the near wake ofthe MVG, however, the visualization shows a distinct train of large billowing structures that appears to bedifferent from those upstream. The increased prominence of the large structures downstream of the MVGcan be compared with the numerical schlieren results of Fig. 3(b). Note that since the boundary layer is devoid of particles in global lightsheet visualization, boundary layerstructures tend to appear as dark images. Further revelation of the influence of the MVG on the boundarylayer is obtained using local laser lightsheet visualization where the boundary layer is seeded by a portlocated directly upstream of the MVG (see Section II). Figure 9 shows local lightsheet visualizations of theflat plate and when an MVG is mounted on it. The change in the large structures due to the presence ofthe MVG is obvious. Figure 9(b) is able to show smaller structures that are not visible in Fig. 8. It is also worthwhile to compare these images with those of Fig. 7. It appears that the faint line identifiedin Fig. 7(b) may be associated with the intermittent outer boundary of the large downstream structures,abseent in the boundary layer over a flat plate. This intermittent outer boundary appears as a faint featurein the schlieren image since it does not spread over the entire span of the flat plate, noting that schlierenimaging integrates all density gradients along its field of view. Babinsky et al.6remarked that the faintline as indicated in Fig. 7(b) is the upper boundary ofa low-momentum region. The present visualizationsshown in Fig. 7(b) indicates a broad band instead of a thin line. Thus, the present visualizations suggestmore specifically that this faint broad band marks the intermittent interface between large structures in theMVG wake. The visualizations also suggest that time-resolved schlieren may not be able to reveal such aband whereas an averaging procedure can. Figure 8. Global lightsheet visualization of MVG70 with lightsheet aligned along MVG axis. (b) With MVG70 Figure 9. Comparison between local lightsheet visualization of flat plate and MVG70. Kleine42 remarked that videographic imaging allows small or obscure features to be revealed that mayotherwise be thought to be“noise" or other flaws. Following this argument, even if the images are not timeresolved, video images are expected to help in understanding the MVG wake. Thus, Figs. 10 and 11 showconsecutive frames from local lightsheet visualizations of the flat plate and with MVG70 respectively. Despitethe low framing rate of the camera, two important observations can be made. First, as in the previous globalvisualization, there is a distinct difference in the shape of the large turbulent structures over the bare flatplate and downstream of an MVG. Secondly, there are features around the MVG that appear relativelysteady and others that show a highly intermittent edge. The visualizations show that the flowfield nearestthe MVG, namely, the primary vortex is turbulent but does not show evidence of large-scale intermittency.Further downstream, it appears that the primary vortex breaks down to form the large roller structures which exhibit severe intermittency, identified by Li and Liu15 as ring vortices and Blinde et al.’ as hairpinvortices. For further clarification, portions of Fig. 11(j) are enlarged and shown in Fig. 12. The difference betweenthe large structures of the incoming boundary layer as shown in Fig. 12(a) compared to downstream of theMVG as shown in Fig. 12(c) is clearly evident. In the latter, large swirls of boundary layer fluid are seen toerupt into the free stream. It is fair to conclude, at least along the centerplane of the MVG wake, that thereis significant entrainment of the freestream fluid and a high level of intermittency. Further, Fig. 12(b) shows a sharp inclined demarcation between the incoming boundary layer and theMVG wake. The demarcation appears to be relatively steady with some flutter between frames.Thisdemarcation is thought to indicate the primary vortex shed off the MVG side.17 Note that the lightsheetis located at the centerplane while the primary vortex is slightly off the centerplane. Therefore, less light isscattered off the primary vortex and thus it appears a little darker in all the frames of Fig. 11. Figure 12(b)shows the rapid breakdown of the primary vortex at about two lengths behind the MVG trailing edge. Thisis consistent with surface flow visualizations which shows the disappearance of two bright lines.6,11,17(Notethat digitally enhanced images show that these two lines extend much further downstream.8) c..Further Discussion Babinsky et al.6aand Herges et al.11 tthought that the disappearance of the bright lines is due to thetwo primary vortices being lifted from the surface due to mutual upwash. However, based on topologicalconsiderations, the two bright lines in surface flow visualization are not associated with the primary vortexpair but may be attributed to two secondary vortices along the corner of the MVG with the flat plate thatoriginate from the leading-edge tips of the MVG. These vortices,as they clash behind the MVG, are likelyto be unsteady due to symmetry breaking and break down. However, it is the break down of the primary vortex pair that is important here due to their relativelylarge size and the outcome on downstream interactions. The breakdown of the primary vortex pair may bedue to the presence of an inflection surface of the velocity profile in the MVG wake, the so-called momentumdeficit. This inflection surface is somewhat circular15 and is analogous to the inflection point in plane shearflow. Thus, this type of instability is labeled Kelvin-Helmholtz-like21 and observed in the wakes of wall-mounted obstacles at low speeds by various authors.29,43-45Mason and Morton43 suggested that innervortices, which may be the primary vortices of MVGs, are dominant and, even though they become unstableand break up, the presence of these vortices is noticeable far downstream. Velte et al.46 observed the presence of embedded longitudinal vortices shed from sub-boundary layervanes and discussed how vorticity rearranges the axial momentum distribution to the crosswise direction.Logdberg et al.’s47 visualizations showed a pair of turbulent vortical structures being convected downstream.Their hotwire measurements indicated the presence of strong counter-rotating vortices emanating from theMVGs but no evidence of unsteadiness was offered. These authors suggested the presence of a “hooklike”vortex core motion persisting as far as 200 vortex heights downstream. Other studies have provided evidence that the flow downstream of an MVG array is unsteady.48 Angeleand Grewe48 suggested that the unsteadiness contributes to maximum Reynolds stresses around the meanvortex centers which should not be the case if the vortices were steady. Duriez et al.49 suggested that steadyspatial forcing at different wavelengths produces a self-sustaining process between the streamwise velocitystreaks in a turbulent boundary layer and the streamwise vortices generated by an MVG array. The circular upwash with momentum deficit has been observed in the wakes of surface-mounted obstaclesat low and high speeds. Present evidence of this feature is provided indirectly by visualizations where thelightsheet is located in a crosswise plane 20.6 mm (0.81 in.) downstream of the MVG trailing edge. Afew sequential frames at 15 Hz are shown in Fig. 13(a)-13(d). These individual frames show an unsteadyroundish blob of light scattered by the acetone droplets. This local lightsheet visualization technique wasdeliberately used to highlight the flow within the boundary layer, as contrasted against the spanwise globallightsheet visualization of Bur et al.8In the latter technique, the lack of seeding in the boundary layer Figure 11. Sequence of frames at 15 Hz from local laser lightsheet visualization of MVG. 25 Figure 12. Enlargements of Fig. 11(j). resulted in these regions being dimly lit. The rms of the light intensity of 118 images is shown in Fig. 13(g) to highlight the roundish structure.The figure reveals that there is a structure with an unsteady rim. This unsteadiness can be interpreted tobe due to the convection of the large structures arising from the trailing vortex breakdown. It is not entirelyclear in the visualization whether these structures are haipin or ring shaped, although this distinction maynot be a critical one on how these vortical structures interact with a downstream separated zone. IV. Conclusions Detailed lightsheet visualizations of the near wake of an MVG reveal the presence of large roller structuresthat were previously identified as hairpin or ring vortices. These structures show a high level of intermittency.These structures are considered to arise from the rapid break down of the primary vortex pair due to acylindrical Kelvin-Helmholtz-like instability surface. Acknowledgments The authors gratefully acknowledge funding for this work via AFOSR Grant No. FA9550-08-1-0201monitored by Dr. John Schmisseur. The authors thank the assistance of Rod Duke and David Whaley withthe experiments. David Whaley was supported by a High School Research Internship under a Texas Youthin Technology grant from the Texas Workforce Commission, administered by Dr. J. Carter M. Tiernan. Theauthors also acknowledge Richard R. Mitchell for designing the high-pressure seed injection system. Finally,the authors appreciate discussions with Prof. C. Liu and Dr. Q. Li. (g) RMS of 118 frames. Figure 13. SSpanwise, local laser light sheet visualization of MVG45. References Babinsky, H., Makinson, N. J.,and Morgan,C. E.,“Micro-Vortex Generator Flow Control for Supersonic Engine Inlets,”AIAA Paper 2007-2007,2007. Holden, H. and Babinsky,H.,“Effect of Microvortex Generators on Separate Normal Shock/BoundaryLayer Interactions,”Journal of Aircraft, Vol. 44, No. 1, 2007, pp. 89-96. Shinn, A. F., Vanka, S. P., Mani, M., Dorgan, A., and Michal,T.,“Application of BCFD Unstructured Grid Solver toSimulation of Micro-Ramp Control of Shock/Boundary Layer Interactions,” AIAA Paper 2007-3914, 2007. Babinsky, H. and Ogawa, H., “SBLI Control for Wings and Inlets," Shock Waves, Vol. 18, No. 2, 2008, pp. 89-96. 5Anderson, B. H., Mace, J. L., and Mani, M.,,“Active ‘Fail Safe’Micro-Array Flow Control For Advanced EmbeddedPropulsion Systems,"AIAA Paper 2009-0920,2009. Babinsky, H., Li, Y., and Pitt Ford, C. W.,“Microramp Control of Supersonic Oblique Shock-Wave/Boundary-LayerInteractions,”AIAA Journal, Vol. 47, No. 3, 2009, pp. 668-675. Blinde, P. L., Humble, R. A., van Oudheusden, B. W., and Scarano, F., “Effects of Micro-Ramps on a ShockWave/Turbulent Boundary Layer Interaction,” Shock Waves, Vol.19, No. 6, 2009, pp. 507-520. ǒBur, R., Coponet, D., and Carpels, Y.,“Separation Control by Vortex Generator Devices in a Transonic Channel Flow,”Shock Waves, Vol. 19, No. 6, 2009, pp.521-530. Domel, N. D., Baruzzini, D., and Miller, D. N.,“CFD Results for Shock-Boundary Layer Flow Control with Micro-Rampsat Various Grid Densities,” AIAA Paper 2009-4016,2009. 10Galbraith,M. C., Orkwis, P. D., and Benek, J. A.,“Multi-Row Micro-Ramp Actuators for Shock Wave Boundary-LayerIn-teraction Control,”AIAA Paper 2009-0321,2009. 11 Anderson, T., Kroeker, E., Elliott, G., and Dutton, J.,“Micro-Ramp Flow Control of Normal Shock/Boundary LayerInteractions,"AIAA Paper 2009-0920,2009. 12Lee, S., Loth, E., Georgiadis,N. J., and DeBonis, J. R.,“Effect of Mach Number on Flow Past Micro-Ramps,”AIAAPaper 2009-4181,2009. 13Rybalko,M., Loth, E., Chima, R. V., Hirt,S. M., and DeBonis,J. R.,“Micro-Ramps for External Compression Low-BoomInlets," AIAA Paper 2009-4206,2009. 14Lee, S., Goettke, M. K., Loth, E., Tinapple, J., and Benek, J.,“Microramps Upstream of an Oblique-Shock/Boundary-Layer Interaction,”AIAA Journal, Vol. 48, No. 1, 2010,pp. 104-118. 15Li, Q. and Liu, C., “LES for Supersonic Ramp Control Flow Using MVG at M = 2.5 and Reg= 1440,"AIAA Paper2010-0592,2010. 16Li, Q. and Liu, C., “Numerical Investigation on the Effects of the Declining Angle of the Trailing Edge of MVG,”AIAAPaper 2010-0714,2010. 17Lu, F. K., Pierce, A. J., Shih, Y.,Liu, C., and Li, Q.,“Experimental and Numerical Study of Flow Topology Past MicroVortex Generators,” AIAA Paper 2008-4463,2010. 18Pierce, A. J., Lu, F. K., Bryant, D. S., and Shih, Y.,“New Developments in Surface Oil Flow Visualization,”AIAAPaper 2010-4353, 2010. 19Crawford, J. D. and Knobloch, E., “Symmetry and Symmetry-Breaking Bifurcations in Fluid Dynamics,” Annual Reviewof Fluid Mechanics, Vol. 23, 1991, pp. 341-387. 20Bridges,D. H., “Toward a Theoretical Description of Vortex Wake Asymmetry,” Progress in Aerospace Sciences, Vol. 46,No. 2-3,2010, pp. 62-80. 21Lim, T. T., “On the Role of Kelvin-Helmholtz-Like Instability in the Formation of Turbulent Vortex Rings,”FluidDynamics Research, Vol. 21, No. 1, 1997, pp. 47-56. 22Sedney, R., “A Survey of the Effects of Small Protuberances on Boundary-Layer Flows,”AIAA Journal, Vol. 11, No. 6,1973, pp.782-792. 23Jacobi, A. M. and Shah, R. K., “Heat Transfer Surface Enhancement Through the Use of Longitudinal Vortices: AReview of Recent Progress,”Experimental Thermal and Fluid Science, Vol. 11, 1995, pp. 295-309. 24 Chyu, M. K. and Natarajan, V.,“Heat Transfer on the Base Surface of Three-Dimensional Protruding Elements,”International Journal of Heat and Mass Transfer, Vol. 39, 1996, pp. 2925-2935. 25 Klebanoff, P. S., Cleveland, W. G., and Tidstrom, K. D., “On the Evolution of a Turbulent Boundary-Layer Induced bya Three-Dimensional Roughness Element,” Journal of Fluid Mechanics, Vol.237, 1992, pp. 101-187. 26Baker, C. J., “The Laminar Horseshoe Vortex,”Journal of Fluid Mechanics, Vol. 95,1979, pp.347-367. 27Baker, C. J., “The Turbulent Horseshoe Vortex,” Journal of Wind Engineering and Industrial Aerodynamics, Vol. 6,1980, pp. 9-23. 28Simpson, R. L.,“Junction Flows,” Annual Review of Fluid Mechanics, Vol. 33,2001, pp. 415-443. 29Sau, A., Hwang, R. R., Sheu, T. W., and Wang, W. C., “Interaction of Trailing Vortices in the Wake of a Wall-MountedRectangular Cylinder,” Physical Review E, Vol. 68, No. 5, 2003, Article No. 056303. 30Ghosh, S., Choi, J.-I., and Edwards, J. R.,“RANS and Hybrid LES/RANS Simulation of the Effects of Micro VortexGenerators Using Immersed Boundary Methods," AIAA Paper 2008-3728,2008. Jiang, M., Machiraju, R., and Thompson, D. S.,“Detection and Visualization of Vortices,”Visualization Handbook,edited by C. R. Johnson and C. D. Hansen, Elsevier, 2005, pp. 295-309. Settles, G. S. and Teng, H. Y., “Flow Visualization Methods for Separated 3-Dimensional Shock-Wave TurbulentBoundary-Layer Interactions,"AIAA Journal, Vol. 21, No.3, 1983, pp. 390-397. 33Settles, G. S.,“Modern Developments in Flow Visualization,”AIAA Journal, Vol. 24, No. 8,1986, pp. 1313-1323.34Pierce, A. J., Experimental Study of Micro-Vorter Generators at Mach 2.5, MSAE thesis, University of Texas at Arling-ton, 2010. 35Braun, E. M., Lu, F. K., Mitchell, R. R., Wilson, D. R., and Dutton,J. C., “Supersonic Blowdown Wind Tunnel ControlUsing LabVIEW,” AIAA Paper 2008-0852, 2008. 36Mitchell, R. R. and Lu, F. K.,“Development of a Supersonic Aerodynamic Test Section using Computational Modeling,”AIAA Paper 2009-3573,2009. 37Anderson, B. H., Tinapple, J., and Sorber, L., “Optimal Control of Shock Wave Turbulent Boundary Layer InteractionsUsing Micro-Array Actuation,” AIAA Paper 2006-3197, 2006. 38Pierce, A. J. and Lu, F. K., “Laser Alignment Method for Portable Schlieren System,AIAA Paper 2009-3574, 2009. 39Raffel, M., Willert, C., Wereley, S., and Kompenhans, J., Particle Image Velocimetry: A Practical Guide, Springer,Berlin, 2nd ed., 2007. 40Settles, G. S. and Lu, F. K.,“Conical Similarity of Shock/Boundary-Layer Interactions Generated by Swept and UnsweptFins,” AIAA Journal, Vol. 23, No. 7,1985, pp. 1021-1027. 41Settles, G.S., Bogdonoff,S. M., and Vas, I. E., “Incipient Separation of a Supersonic Turbulent Boundary Layer at HighReynolds Numbers,”AIAA Journal, Vol. 14,1976, pp. 50-56. 42Kleine, H.,“Filming the Invisible-Time-Resolved Visualization of Compressible Flows,” European Physical Journal-Special Topics, Vol. 182,2010, (to appear). 43Mason, P. J. and Morton, B. R.,.,“Trailing Vortices in the Wakes of Surface-Mounted Obstacles,” Journal of FluidMechanics, Vol. 175,1987, pp. 247-293. 44Hwang, J. Y. and Yang, K. S., “Numerical Study of Vortical Structures Around a Wall-Mounted Cubic Obstacle inChannel Flow," Physics of Fluids, Vol. 16,2004, pp. 2382-2394. 45Yakhot, A., Anor, A., Liu, H. P., and Nikitin, N., “Direct Numerical Simulation of Turbulent Flow Around a Wall-Mounted Cube: Spatio-Temporal Evolution of Large-Scale Vortices,” Journal of Fluid Mechanics, Vol. 566,2006, pp. 1-9. 46Velte, C. M., Hansen, M. O. L., and Okulov, V. L., "Helical Structure of Longitudinal Vortices Embedded in TurbulentWall-Bounded Flow,” Journal of Fluid Mechanics, Vol. 619, 2009, pp. 167-177. 47Logdberg, O., Fransson, J. H., and Alfredsson, P. H., “Streamwise Evolution of Longitudinal Vortices in a TurbulentBoundary Layer,” Journal of Fluid Mechanics, Vol.623,2009, pp. 27-58. 48 Angele, K. P. and Grewe, F., “Instantaneous Behavior of Streamwise Vortices for Turbulent Boundary Layer SeparationControl,” Journal of Fluids Engineering, Vol. 129, 20006,226-235. 49Duriez,T., Aider, J.-L., and Wesfreid, J. E.,“Self-Sustaining Process through Streak Generation in a Flat-Plate BoundaryLayer," Physical Review Letters, Vol. 103, 2009, Article No. 144502. of merican Institute of Aeronautics and AstronauticsCopyright C by Frank K. Lu. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission. ofmerican Institute of Aeronautics and Astronautics
还剩12页未读,是否继续阅读?
北京欧兰科技发展有限公司为您提供《发生器中超音速流场中微型漩涡发生器近尾迹区的实验研究检测方案(粒子图像测速)》,该方案主要用于其他中超音速流场中微型漩涡发生器近尾迹区的实验研究检测,参考标准--,《发生器中超音速流场中微型漩涡发生器近尾迹区的实验研究检测方案(粒子图像测速)》用到的仪器有德国LaVision PIV/PLIF粒子成像测速场仪、Imager sCMOS PIV相机、PLIF平面激光诱导荧光火焰燃烧检测系统