Bulletin of the American Physical Society
70th Annual Meeting of the APS Division of Fluid Dynamics
Volume 62, Number 14
Sunday–Tuesday, November 19–21, 2017; Denver, Colorado
Session D28: Turbulence: Theory & Experiment IExperimental Turbulence
|
Hide Abstracts |
Chair: Michael Wilczek, Max Planck Institute for Dynamics and Self-Organization Room: 207 |
Sunday, November 19, 2017 2:15PM - 2:28PM |
D28.00001: Second-order structure function in high-resolution DNSs of turbulence $-$ Where is the inertial subrange? Takashi Ishihara, Yukio Kaneda, Koji Morishita, Mitsuo Yokokawa, Atsuya Uno We report some results of a series of high resolution direct numerical simulations (DNSs) of forced incompressible isotropic turbulence with up to $12288^3$ grid points and Taylor microscale Reynolds number $R_\lambda\sim 2300$. The DNSs show that there exists a scale range, approximately at $100 |
Sunday, November 19, 2017 2:28PM - 2:41PM |
D28.00002: First Results at ultra-high $R_{\lambda}$ in a wind tunnel Christian Kuechler, Eberhard Bodenschatz, Gregory P. Bewley With a new active grid installed, the Variable Density Turbulence Tunnel (VDTT) at the Max-Planck-Institute for Dynamics and Self-Organization produced homogeneous turbulence at Reynolds numbers up to $R_{\lambda} \approx 7500$. The active grid consisted of $111$ individually controllable flaps that produced more intense turbulence than classical fixed grids. We varied the Reynolds number by changing the pressure of sulfur hexafluoride gas in the tunnel between 0.5 and 15 bar, which changes the viscosity of the gas\footnote{E. Bodenschatz et al, \textit{Rev. Sci. Inst.} \textbf{82}(11)}. With hot wire probes called NSTAPs that were 30 microns long\footnote{S. Bailey et al, \textit{J. Flu. Mech.} \textbf{663}}, we measured velocity spectra and structure functions. While a bottleneck is present in the spectra at Reynolds numbers up to R$_{\lambda}<3000$, the bottleneck weakens and disappears at higher R$_{\lambda}$. We compare this observation with measurements made in the field and with computer simulations. [Preview Abstract] |
Sunday, November 19, 2017 2:41PM - 2:54PM |
D28.00003: K62SCAN: Experimental Assessment of the Refined Similarity Hypothesis John Lawson, Anna Knutsen, James Dawson, Eberhard Bodenschatz, Nicholas Worth Experimental and numerical studies of turbulence widely confirm departures from Kolmogorov's famed 1941 Similarity Hypothesis scaling. Kolmogorov anticipated this and later introduced his 1962 Refined Similarity Hypothesis (RSH), which posits that the local, volume averaged dissipation rate determines the scaling of velocity differences at that scale. Previous experimental tests have supported the RSH, but must be regarded as inconclusive, due to systematic errors introduced by the use of surrogates for the local dissipation rate. Recent advances in Scanning PIV now permit us to dispense with this technical limitation. During the K62SCAN EuHIT project, we conducted spatially resolved Stereo PIV and Scanning PIV measurements in the GTF3 von-K{\'a}rm{\'a}n mixing tank facility at MPIDS. We collected over $2\times10^5$ independent time-series Scanning PIV measurements of the turbulence at $R_\lambda=200$ in a cubic measurement volume approximately 42 Kolmogorov lengthscales to a side. The measurement is able to resolve the instantaneous dissipation and velocity fields, eliminating the need for surrogates. This enables us to present the first direct, experimental assessment of the RSH. Our results complement recent numerical investigations into the RSH scaling. [Preview Abstract] |
Sunday, November 19, 2017 2:54PM - 3:07PM |
D28.00004: Circulation in turbulent flows Katepalli Sreenivasan, Kartik Iyer, P.K Yeung, Xiaomeng Zhai Circulation around Eulerian contours has been a valuable conceptual tool in classical fluid dynamics and aerodynamics, but its properties have not been explored and exploited much in the turbulence literature, especially in comparison with multi-point objects such as velocity increments. The initial theoretical work of Migdal (Int.~J.~Mod.~Phys.~A {\bf 9} 1197-1238 (1994)) has been followed up only in a small number of empirical papers (e.~g.~Umeki, JPSJ {\bf 69}, 3788-3791 (1993), Cao et al.~PRL {\bf 76} 616-619 (1996) and Benzi et al.~PRE {\bf 55} 3739-3742 (1997)) and these latter papers use direct numerical simulations data on relatively small grids and low Reynolds numbers. Using our recent data base of simulations (Yeung et al.~PNAS {\bf 112} 12633-12638 (2015)) of isotropic and homogeneous turbulence on $8192^3$ grids (and others on smaller boxes down to $256^3$), we explore here the statistical properties of circulation, such as the probability density functions of circulation around contours of various sizes within the inertial range and its scaling properties. Among the results obtained, the one that stands out is that circulation statistics can be described very closely by a lognormal process, and, to within experimental accuracy by a uni-fractal of dimension 2.8. [Preview Abstract] |
Sunday, November 19, 2017 3:07PM - 3:20PM |
D28.00005: Multi-scale analysis of local flow topology in isotropic turbulence Mohammad Danish, Charles Meneveau Knowledge of local flow-topology, as described by the velocity gradients, is useful to develop insights of turbulence processes, such as energy cascade, material element deformation, etc. Much has been learned in recent past about flow-topology at the smallest (viscous) scales of turbulence. However, less is known at larger (or inertial) scales of turbulence. In this work, we present a detailed study on the scale-dependence of various quantities of our interest, like population fraction of different flow-topologies, joint probability distribution of second and third invariants of velocity gradient tensor, etc. We use a new filter proposed by Eyink \& Aluie to decompose the flow into small and large scales. We provide further insights for the observed behavior of scale-dependence by examining the probability fluxes appearing in the Fokker-Plank equation. Specifically, we aim to understand whether the differences observed between the viscous and inertial range are due to different effects caused by pressure, subgrid-scale or viscous stresses, or various combination thereof. For this purpose, we use the isotropic turbulence dataset at $Re_\lambda=433$ available at JHTDB and the analysis tools developed for SciServer, including FFT to evaluate filtering and gradients. [Preview Abstract] |
Sunday, November 19, 2017 3:20PM - 3:33PM |
D28.00006: A novel bridging relation connecting Eulerian and Lagrangian statistics Michael Wilczek, Cristian Lalescu The complexity of fully developed turbulent flows can be perceived either from an Eulerian or a Lagrangian point of view. The Eulerian frame is particularly suited for characterizing the spatial structure of turbulence. Following tracer particles, i.e. adopting the Lagrangian perspective, adds temporal information. Understanding the relationship between Eulerian and Lagrangian properties of turbulence is important for a range of problems, including turbulent mixing and transport as well as uncovering the origin of Lagrangian intermittency. The key challenge in relating Eulerian and Lagrangian statistics lies in the fact that Lagrangian tracers sample the flow spatio-temporally in a non-trivial manner: while tracer particles disperse, the flow field evolves in time. Here we present a novel bridging relation which captures both of these effects in an effective Lagrangian dispersion, relating the statistics of instantaneous Eulerian to temporal Lagrangian velocity fluctuations. We discuss the derivation of the bridging relation along with the properties of the effective Lagrangian dispersion. Furthermore, we benchmark our predictions against high-resolution direct numerical simulations of homogeneous and isotropic turbulence. [Preview Abstract] |
Sunday, November 19, 2017 3:33PM - 3:46PM |
D28.00007: On the small-scale structure of turbulence and its impact on the pressure field Dimitar Vlaykov, Michael Wilczek Understanding the non-locality of the pressure field in incompressible flows is one of the fundamental challenges in turbulence. The pressure field is encoded in the non-linear structure of turbulence by a Poisson equation. We present a quantitative investigation of the link between intense small-scale vortical and strain structures and the non-locality of the pressure field. Specifically, we show that the pressure in the neighborhood of intense vorticity regions is determined primarily by the local structures and with little influence from far-field contributions. Moreover, the degree of locality increases with the intensity of the reference regions. This is explained by the generation of a shielding shear region around a given vortex through the Biot-Savart law on the one hand, and the rapid decorrelation of the velocity gradient field on the other. Strong shear regions display similar but less pronounced insulating features. This behavior depends on the Reynolds number in two ways. Firstly, due to the intermittent nature of the velocity gradient, the intensity and frequency of occurrence of extreme events grow quickly with increasing Reynolds number. Secondly, the degree of locality of regions of given intensity tends to change slowly but monotonically with Reynolds number. [Preview Abstract] |
Sunday, November 19, 2017 3:46PM - 3:59PM |
D28.00008: Emergence of multi-scaling in fluid turbulence Diego Donzis, Victor Yakhot We present new theoretical and numerical results on the transition to strong turbulence in an infinite fluid stirred by a Gaussian random force. The transition is defined as a first appearance of anomalous scaling of normalized moments of velocity derivatives (or dissipation rates) emerging from the low-Reynolds-number Gaussian background. It is shown that due to multi-scaling, strongly intermittent rare events can be quantitatively described in terms of an infinite number of different ``Reynolds numbers'' reflecting a multitude of anomalous scaling exponents. We found that anomalous scaling for high order moments emerges at very low Reynolds numbers implying that intense dissipative-range fluctuations are established at even lower Reynolds number than that required for an inertial range. Thus, our results suggest that information about inertial range dynamics can be obtained from dissipative scales even when the former does not exit. We discuss our further prediction that transition to fully anomalous turbulence disappears at $R_\lambda < 3$. [Preview Abstract] |
Sunday, November 19, 2017 3:59PM - 4:12PM |
D28.00009: Structure and scale interaction in anisotropic homogeneous turbulence Douglas Carter, Filippo Coletti The structure and dynamics of anisotropic turbulence have mostly been investigated in shear flows. Here we use particle image velocimetry (PIV) to investigate anisotropic homogeneous turbulence with negligible mean shear, generated by two facing planar jet arrays. The homogeneous region is much larger than the integral scales, which allow for the natural development of the Richardson-Kolmogorov cascade. Moreover, with respect to advective flows with large mean velocity, the zero-mean-flow condition effectively enhances the dynamic range of the PIV turbulent fluctuation measurements. We obtain high-order statistics along directions parallel and normal to the jet axis, as well as in arbitrary directions within the plane of measurement. Up to the considered Reynolds numbers of 500 (based on the Taylor microscale), we find evidence of anisotropy extending throughout the inertial scales and down to the dissipative scales. In addition, we investigate the scale-to-scale energy transfer utilizing the generalized Karman-Howarth equation. We explore the correlation between large-scale and small-scale velocity fluctuations, which has been recently highlighted in shear flows, and the footprint of the multi-scale anisotropy on the topology of high-dissipation and high-enstrophy regions. [Preview Abstract] |
Sunday, November 19, 2017 4:12PM - 4:25PM |
D28.00010: What is turbulence and which way does it cascade? Carl Gibson Turbulence is defined as an eddy-like state of fluid motion, where the inertial vortex forces v x curl v of the eddies are larger than any other forces that tend to damp the eddies out. In the beginning at Planck conditions, it is assumed that the relevant dimensional parameters were the speed of light c, the Planck constant h, the Newton constant G, and the Boltzmann constant k. The first turbulence appeared at 10$^-^4^3$ s, 10$^-^3^5$ m, 10$^3^2$ K when the Kolmogorov scale first matched the horizon scale ct. Inertial vortex forces of adjacent fluid particles with the same spin cause them to merge, so the turbulence cascade is always from small scales to large, as observed, contrary to the standard Taylor-Lumley model which must be abandoned. Adjacent fluid particles with opposite spin repel each other and are repelled by walls, which explains boundary layer separation and turbulent diffusion. The second turbulence appeared at 10$^1^2$ seconds at density 10$^-^1^7$ kg m$^-^3$ with the fragmentation of protogalaxies along fossil big bang turbulence vortex lines. Life began at 10$^1^3$ seconds in Jeans mass clumps of a trillion planets. [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