Bulletin of the American Physical Society
67th Annual Meeting of the APS Division of Fluid Dynamics
Volume 59, Number 20
Sunday–Tuesday, November 23–25, 2014; San Francisco, California
Session L28: Turbulence: Jets and Wakes |
Hide Abstracts |
Chair: Roberto Paoli, CERFACS Room: 2011 |
Monday, November 24, 2014 3:35PM - 3:48PM |
L28.00001: Large-eddy simulations of turbulent plane and radial wall-jets Rayhaneh Banyassady, Ugo Piomelli Large-eddy simulations of turbulent plane and radial wall-jets were conducted at different Reynolds numbers. The results were validated with the available experimental data. The radial wall-jets decay faster compared to the plane ones, due to the extra expansion in the azimuthal direction. This causes the pressure-gradient distributions to be different in radial and plane wall-jets (e.g. the inner layer in the plane case is under a favorable pressure-gradient, while in the radial case it subjected to an adverse pressure-gradient). However, these pressure gradients are not strong enough to cause any structural difference between plane and radial wall-jets. In both cases, the local Reynolds number (based on the local maximum velocity and local boundary-layer thickness) is an important determining factor in characterization of the flow. The joint probability-density function analysis shows that the local Reynolds number determines the level of intrusion of the outer layer into the inner layer: the lower the local Reynolds number the stronger is the interaction of inner and outer layers. These results were used to clarify some of the observations reported in literature; as an example the scatter of the reported log-law constants can be explained using the above-mentioned results. [Preview Abstract] |
Monday, November 24, 2014 3:48PM - 4:01PM |
L28.00002: Large eddy simulations of a Mach 0.9 jet with fully-turbulent nozzle-exit boundary layer Guillaume Bres, Frank Ham, Peter Jordan From past studies, it is well known that the state of the nozzle-exit boundary layer is a key parameter for the flow development and noise characteristics of a jet. However, because of the computational cost of simulating high Reynolds number wall-driven turbulence, the nozzle boundary layer is typically assumed to be laminar or weakly disturbed in most jet simulations. This approach often leads to enhanced laminar to turbulent shear-layer transition and increased noise due to vortex pairing. In the present work, large eddy simulations of an isothermal Mach 0.9 jet (Re $=$ 1E6) issued from a convergent-straight nozzle are performed using the compressible flow solver ``Charles'' developed at Cascade Technologies. Localized adaptive mesh refinement, synthetic turbulence and wall modeling are used inside the nozzle to ensure fully turbulent profiles at the nozzle exit. This resulted in significant improvements for the flowfield and sound predictions, compared to the typical approach based on laminar flow assumption in the nozzle. The far-field noise spectra now remarkably match the measurements from the companion experiment conducted at Pprime Institute, within 0.5 dB for most angles and relevant frequencies. As a next step toward better understanding of jet noise, the large transient database collected during the simulation is currently being mined using reduced order modeling and wavepacket analysis. [Preview Abstract] |
Monday, November 24, 2014 4:01PM - 4:14PM |
L28.00003: Numerical study of high-speed turbulent jets in crossflow Prahladh Iyer, Xiaochuan Chai, Krishnan Mahesh Large-eddy Simulation (LES) is used to study (i) a sonic jet injected into a supersonic crossflow, and (ii) a supersonic jet injected into a subsonic crossflow, whose conditions are based on experiments by Santiago {\it et al.} (1997) and Beresh {\it et al.} (2005) respectively. An unstructured, finite volume compressible solver (Park \& Mahesh 2007) along with the Dynamic Smagorinsky Model (DSM) (Moin {\it et al.} 1991) is used in the simulations. Qualitative and quantitative comparison with experiment show good agreement for both flows. Dynamic Mode Decomposition (DMD) of the three-dimensional flow field is performed to identify dominant frequencies and their corresponding flow features. [Preview Abstract] |
Monday, November 24, 2014 4:14PM - 4:27PM |
L28.00004: Assessing grid resolution effects in large-eddy simulations of a jet-in-cross-flow Anthony Ruiz, Guilhem Lacaze, Joseph Oefelein Calculations using the Large Eddy Simulation technique are conducted over a range of resolutions in a Jet In Cross Flow. The configuration corresponds to the experiment of Su and Mungal (2004). A turbulent jet at a Reynolds Number of 5,000 is injected into a laminar cross flow, with a jet to cross-flow velocity ratio of 5.7. Resolutions are varied from typical engineering spatial and temporal resolutions to near-DNS resolution. The near-DNS resolution has been extensively validated against experimental results in a previous study [Ruiz et al., Phys. Fluid (2014)], which also focused on the evolution of the major scales of turbulence. The grid resolution is coarsened to investigate the impact on the topology of coherent structures and mixing. Fourier analysis is conducted at all resolutions to observe the impact of filtering on the turbulent energy cascade, identify the main hydrodynamic frequencies, and determine the degree to which these are grid-dependent. Once hydrodynamics modes are known, phase-locked analysis of the flow field shows the spatial structure of these modes as a function of resolution. This enables a clear understanding of the impact of resolution on the flow. [Preview Abstract] |
Monday, November 24, 2014 4:27PM - 4:40PM |
L28.00005: The Far Field Structure of a Jet in Cross-Flow Nicolas Lanitis, James Dawson Stereoscopic PIV measurements were performed in the far field of a cross-flow jet. Measurements were taken in a water channel in the spanwise-wall normal plane (y-z) containing the Counter-Rotating vortex pair (CVP). The jet's Reynolds number was $Re_{jet} = 2 \times 10^{4}$ and had an exit diameter of $d_{j} = 4mm$. Measurements were taken for a jet to cross-flow velocity ratio of $V_{r} = 10$ at three downstream positions of $x/d_{j} = 30, 55, 85$ and for a $V_{r} = 15, 20$ at $x/d_{j} = 85$. Two point spatial correlations hint at the presence of arch shaped structures titled in the streamwise x-direction on the windward side of the CVP as well as straight vortex tubes extending into the wake. The arched shaped structure is compounded by PDFs of the location of streamwise vorticity peaks (vortex tubes) in the instantaneous field indicating the presence of a vortex structure aligned in the spanwise direction. This information together with the use of High Speed Stereoscopic PIV and Taylor's Hypothesis, which allowed for the extraction of 3D structures, led to the development of an eddy model comprised of hairpin, roller and wake structures to predict turbulence statistics of a jet in cross-flow. [Preview Abstract] |
Monday, November 24, 2014 4:40PM - 4:53PM |
L28.00006: Amplitude and frequency modulation in a turbulent jet at high Reynolds number Daniele Fiscaletti, Bharathram Ganapathisubramani, Gerrit Elsinga In this work, the amplitude and frequency modulation of the small scales of turbulence is investigated experimentally in a jet at high Reynolds number. Hot-wire anemometry (HWA) and long-range $\mu$PIV measurements are performed in the fully developed region of the jet. From HWA, time series are converted into space series by applying the Taylor hypothesis. Using spectral filters, two signals representative of the large, and the small scales are constructed. It was found that for positive large-scale fluctuations, the associated small-scale signal is stronger in amplitude (amplitude modulation), and presents locally a higher number of local maxima and minima (frequency modulation). Moreover, the local standard deviation of the small-scale signal (representative of the amplitude of the small-scale signal) increases with the local strength of the large-scale fluctuations. A further investigation with PIV allowed to resolve the small scales of turbulence, without the need for Taylor hypothesis. From this analysis, the amount of amplitude modulation was found to be only $25\%$ of the value obtained with HWA. This difference can be explained considering that the structures of intense vorticity travel, on average, at velocities higher than the mean velocity of the flow. [Preview Abstract] |
Monday, November 24, 2014 4:53PM - 5:06PM |
L28.00007: Pressure ratio effects on self-similar scalar mixing of high-pressure turbulent jets in a pressurized volume Adam Ruggles, Lyle Pickett, Jonathan Frank Many real world combustion devices model fuel scalar mixing by assuming the self-similar argument established in atmospheric free jets. This allows simple prediction of the mean and rms fuel scalar fields to describe the mixing. This approach has been adopted in super critical liquid injections found in diesel engines where the liquid behaves as a dense fluid. The effect of pressure ratio (injection to ambient) when the ambient is greater than atmospheric pressure, upon the self-similar collapse has not been well characterized, particularly the effect upon mixing constants, jet spreading rates, and virtual origins. Changes in these self-similar parameters control the reproduction of the scalar mixing statistics. This experiment investigates the steady state mixing of high pressure ethylene jets in a pressurized pure nitrogen environment for various pressure ratios and jet orifice diameters. Quantitative laser Rayleigh scattering imaging was performed utilizing a calibration procedure to account for the pressure effects upon scattering interference within the high-pressure vessel. [Preview Abstract] |
Monday, November 24, 2014 5:06PM - 5:19PM |
L28.00008: Numerical simulation of the flow field from a radially lobed nozzle and validation via HWA Noushin Amini, Aarthi Sekaran With a constant need for higher performance and efficiency in engineering (particularly aerospace) applications, lobed nozzles have experienced a regained interest in the recent past, owing to their superior mixing capabilities. Although previous experimental studies (Hu et al 1999, Hu et al 2008) have analyzed the flow field from lobed nozzles and made conjectures about the physics and flow mechanisms involved, the absence of a ``complete'' 3D dataset elicits unanswered questions. The present numerical study is intended as a complement to an existing experimental (single component hot wire anemometry) investigation, involving the analysis of the flow field downstream of a six lobed nozzle (N. Amini et al, 2012). A full 3D URANS simulation of the lobed nozzle is carried out, initially validated with experimental data, and then used to examine the stream-wise vortices and obtain a visual corroboration of the structure formation and breakup mechanism as described earlier (Hu et al, 2008). Further, the study takes a close look at the nature of the instabilities which trigger and enhance the mixing process in lobed nozzles in order to determine the precise role of the lobes and eventually obtain more effective mixing in industrial applications. [Preview Abstract] |
Monday, November 24, 2014 5:19PM - 5:32PM |
L28.00009: Large-eddy simulations of a solid-rocket booster jet Roberto Paoli, Adele Poubeau, Daniel Cariolle Emissions from solid-rocket boosters are responsible for a severe decrease in ozone concentration in the rocket plume during the first hours after a launch. The main source of ozone depletion is due to hydrogen chloride that is converted into chlorine in the high temperature regions of the jet (afterburning). The objective of this study is to evaluate the active chlorine concentration in the plume of a solid-rocket booster using large-eddy simulations. The gas is injected through the entire nozzle of the booster and a local time-stepping method based on coupling multi-instances of a fluid solver is used to extend the computational domain up to 600 nozzle exit diameters. The methodology is validated for a non-reactive case by analyzing the flow characteristics of supersonic co-flowing under expanded jets. Then, the chemistry of chlorine is studied offline using a complex chemistry solver and the LES data extracted from the mean trajectories of sample fluid particles. Finally, the online chemistry is analyzed by means of the multispecies version of the LES solver using a reduced chemistry scheme. The LES are able to capture the mixing of the exhaust with ambient air and the species concentrations, which is also useful to initialize atmospheric simulations on larger domains. [Preview Abstract] |
Monday, November 24, 2014 5:32PM - 5:45PM |
L28.00010: Direct numerical simulations of turbulent wakes with non-equilibrium similarity scalings Thibault Dairay, John Christos Vassilicos Recently, turbulent flow regions with dissipation scalings incompatible with equilibrium Richardson-Kolmogorov phenomenology have been discovered in the lee of regular and fractal grids. Considering the non-equilibrium dissipation law with a similarity analysis, new scaling laws have recently been obtained for the streamwise evolution of the centreline wake mean profiles (PRL 111, 144503 (2013)). In the present study, DNS of spatially evolving wakes generated by bluff plates with both simple square and irregular edge peripheries (the latter allowing the formation of jet-wake flows) have been carried out using the in-house code Incompact3d. The Reynolds number based on the plate length L, equal to the square-root of the plate area, and the freestream velocity is 5000. The self-similarity of the mean flow, Reynolds-stresses and the dissipation rate of turbulent kinetic energy have been analysed as well as the scaling laws. In the region where the flow is found to be axisymmetric and self-similar, the viscous term in the momentum equation is two orders of magnitude smaller than the other terms and the mean flow profile evolves in accordance with the non-equilibrium law up to 100L. Furthermore, the non-equilibrium dissipation law is observed for both regular and irregular plates. [Preview Abstract] |
Monday, November 24, 2014 5:45PM - 5:58PM |
L28.00011: Asymmetries in the high Reynolds number wake of a submarine model in pitch Anand Ashok, Tyler Van Buren, Alexander Smits Experiments are reported in the wake of a submarine model (DARPA SUBOFF) over a wide range of Reynolds numbers based on the length Re$_{\mathrm{L}}$ between 10$^{5}$ and 30 x 10$^{6}$ at a pitch angle of 8 degrees. Two-component velocity measurements were taken at five cross-stream planes, downstream of the stern of the model (2\textless x/D\textless 14), using hot wire anemometry. The wake is distinguished by two principal vortex structures, but the strength of the two vortices are not equal, leading to an asymmetric wake that slowly rotates. The asymmetry appears to be endemic, and is not affected by freestream turbulence, changes in tripping, surface roughness, and small angles of yaw. They persist across all the Reynolds numbers measured, and the effects of Reynolds number are only important at low Reynolds number in that the wake becomes independent of Reynolds number when Re$_{\mathrm{L}} \ge $ 4.8 x 10$^{6}$. This work was supported under ONR Grant N00014-13-1-0174 (Ron Joslin). [Preview Abstract] |
Monday, November 24, 2014 5:58PM - 6:11PM |
L28.00012: Steady imperfect bifurcation with generic 3D bluff bodies at large Reynolds numbers Olivier Cadot, Luc Pastur, Antoine Evrard, Guillaume Soyer The turbulent wake of parallelepiped bodies exhibits a strong bi-modal behavior. The wake randomly undergoes symmetry breaking reversals, between two mirror asymmetric steady modes (RSB modes). The characteristic time for reversals is about 2 or three orders of magnitudes larger than the natural time for vortex shedding. Such a dynamics has been recently observed on real car which points out its importance about industrial applications. Both the viscosity and the proximity of a wall in the vicinity of the parallelepiped body (similarly to the road with a car model), stabilize the RSB modes on a single symmetric mode. It is shown that these stabilizations occur through imperfect fork bifurcations at large Reynolds numbers. The extra drag due to the presence of the RSB modes is evidenced. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700