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 A29: Turbulence: CompressibilityCompressible Turbulence
|
Hide Abstracts |
Chair: Sharath Girimaji, Texas A&M University Room: 205 |
Sunday, November 19, 2017 8:00AM - 8:13AM |
A29.00001: Exact Theory of Compressible Fluid Turbulence Theodore Drivas, Gregory Eyink We obtain exact results for compressible turbulence with any equation of state, using coarse-graining/filtering. We find two mechanisms of turbulent kinetic energy dissipation: scale-local energy cascade and ``pressure-work defect'', or pressure-work at viscous scales exceeding that in the inertial-range. Planar shocks in an ideal gas dissipate all kinetic energy by pressure-work defect, but the effect is omitted by standard LES modeling of pressure-dilatation. We also obtain a novel inverse cascade of thermodynamic entropy, injected by microscopic entropy production, cascaded upscale, and removed by large-scale cooling. This nonlinear process is missed by the Kovasznay linear mode decomposition, treating entropy as a passive scalar. For small Mach number we recover the incompressible ``negentropy cascade'' predicted by Obukhov. We derive exact Kolmogorov 4/5th-type laws for energy and entropy cascades, constraining scaling exponents of velocity, density, and internal energy to sub-Kolmogorov values. Although precise exponents and detailed physics are Mach-dependent, our exact results hold at all Mach numbers. Flow realizations at infinite Reynolds are ``dissipative weak solutions'' of compressible Euler equations, similarly as Onsager proposed for incompressible turbulence. [Preview Abstract] |
Sunday, November 19, 2017 8:13AM - 8:26AM |
A29.00002: The influence of compressibility on nonlinear spectral energy transfer – Part 1: Fundamental mechanisms Divya Sri Praturi, Sharath Girimaji Nonlinear spectral energy transfer by triadic interactions is one of the foundational processes in fluid turbulence. Much of our current knowledge of this process is contingent upon pressure being a Lagrange multiplier with the only function of re-orienting the velocity wave vector. In this study, we examine how the nonlinear spectral transfer is affected in compressible turbulence when pressure is a true thermodynamic variable with a wave character. We perform direct numerical simulations of multi-mode evolution at different turbulent Mach numbers of $M_t = 0.03, 0.6$. Simulations are performed with initial modes that are fully solenoidal, fully dilatational and mixed solenoidal-dilatational. It is shown that solenoidal-solenoidal interactions behave in canonical manner at all Mach numbers. However, dilatational and mixed mode interactions are profoundly different. This is due to the fact that wave-pressure leads to kinetic-internal energy exchange via the pressure-dilatation mechanism. An important consequence of this exchange is that the triple correlation term, responsible for spectral transfer, experiences non-monotonic behavior resulting in inefficient energy transfer to other modes. [Preview Abstract] |
Sunday, November 19, 2017 8:26AM - 8:39AM |
A29.00003: The influence of compressibility on nonlinear spectral energy transfer - Part 2: Effect on hypersonic boundary layer transition Ankita Mittal, Sharath Girimaji We examine the effect of compressible spectral energy transfer in the nonlinear regime of transition to turbulence of hypersonic boundary layers. The nature of spectral energy transfer between perturbation modes is profoundly influenced by two compressibility mechanisms. First and foremost, the emergence of nonlinear pressure-dilatation mechanism leads to kinetic-internal energy exchange within the perturbation field. Such interchange is absent in incompressible flow as pressure merely reorients the perturbation amplitude vector while conserving kinetic energy. Secondly, the nature of triadic interactions also changes due to variability in density. In this work, we demonstrate that the efficiency of nonlinear spectral energy transfer is diminished in compressible boundary layers. Emergence of new perturbation modes or `broad-banding' of the perturbation field is significantly delayed in comparison to incompressible boundary layer undergoing transition. A significant amount of perturbation energy is transformed to internal energy and thus unavailable for `tripping' the flow into turbulent state. These factors profoundly change the nature of the nonlinear stage of transition in compressible boundary layer leading to delayed onset of full-fledged turbulence. [Preview Abstract] |
Sunday, November 19, 2017 8:39AM - 8:52AM |
A29.00004: Velocity-vorticity correlation structures (VVCS) in spatially developing compressible turbulent boundary layer Shi-Yao Li, Zhen-Su She, Jun Chen A velocity-vorticity correlation structure (VVCS) analysis is applied to the direct numerical simulation (DNS) of compressible turbulent boundary layer (CTBL) at Mach numbers, $Ma=2.25$, $4.50$ and $6.0$. It is shown that the VVCS analysis captures the geometry variation in the streamwise direction during the transition and in the wall-normal direction in the fully developed regime. Specifically, before transition, the VVCS captures the instability wave number, while in the transition region it displays a distinct scaling change of the dimensions. The fully developed turbulence regime is characterized by a nearly constant spatial extension of the VVCS. Particularly, after turbulence is well developed, a multi-layer structure in the wall normal direction is observed in the maximum correlation coefficient and in the length scales of the VVCS, as expected from a recent symmetry-based theory, the ensemble structure dynamics (SED). The most interesting outcome is an observed linear dependence of the length scale of the VVCS from $y^+\approx50$ to $200$, which is a direct support to Townsend’s attached-eddy theory. In conclusion, the VVCS analysis quantifies the geometrical characteristics of the coherent structures in turbulent compressible shear flows throughout the whole domain. [Preview Abstract] |
Sunday, November 19, 2017 8:52AM - 9:05AM |
A29.00005: Self-consistent viscous heating of rapidly compressed turbulence Alejandro Campos, Brandon Morgan Given turbulence subjected to infinitely rapid deformations, linear terms representing interactions between the mean flow and the turbulence dictate the evolution of the flow, whereas non-linear terms corresponding to turbulence-turbulence interactions are safely ignored. For rapidly deformed flows where the turbulence Reynolds number is not sufficiently large, viscous effects can't be neglected and tend to play a prominent role, as shown in the study of Davidovits \& Fisch (2016). For such a case, the rapid increase of viscosity in a plasma--as compared to the weaker scaling of viscosity in a fluid---leads to the sudden viscous dissipation of turbulent kinetic energy. As shown in Davidovits \& Fisch, increases in temperature caused by the direct compression of the plasma drive sufficiently large values of viscosity. We report on numerical simulations of turbulence where the increase in temperature is the result of both the direct compression (an inviscid mechanism) and the self-consistent viscous transfer of energy from the turbulent scales towards the thermal energy. A comparison between implicit large-eddy simulations against well-resolved direct numerical simulations is included to asses the effect of the numerical and subgrid-scale dissipation on the self-consistent viscous [Preview Abstract] |
Sunday, November 19, 2017 9:05AM - 9:18AM |
A29.00006: Simulation of Compressible Flows and Shock Turbulence Interaction Using Observable Euler and Navier-Stokes Equations Majid Allahyari, Kamran Mohseni We present the results of numerical simulation of several canonical problems involving shock and turbulence based on an inviscid regularization technique, termed the observable method. This technique applies the observable divergence theorem to the conservation laws to obtain the observable version of Euler and Navier-Stokes equations. The method is completely inviscid, does not introduce artificial dissipation, and eliminates the need for complex numerical schemes. Simulations of shock-vorticity/entropy wave interaction and decaying three-dimensional homogeneous isotropic turbulence demonstrate the good performance of the observable method. Moreover, the presented method shows consistent results compared to the DNS results for a canonical shock-turbulence interaction problem. [Preview Abstract] |
Sunday, November 19, 2017 9:18AM - 9:31AM |
A29.00007: Shock jumps in shock-turbulence interactions in the presence of very strong turbulence fluctuations Chang-Hsin Chen, Diego Donzis Shock-turbulence interactions are ubiquitous in nature and engineering and have, thus, been studied extensively with simulations and experiments. With the continuous increase of computational power available, investigations have considered an increasingly wide range of conditions in terms of the strength of turbulence and shock waves. One consistent observation in the literature is the departure from Rankine-Hugoniot relations due to turbulent fluctuations upstream of the shock. Based on the so-called quasi-equilibrium assumption, analytical solutions for thermodynamic jumps that take into account the observed "holes" in the shock are presented and compared to new DNS data using shock-resolving simulations. By considering the effect of subsonic regions upstream of the shock, the good agreement obtained in previous work is extended to the broken regime at $M_t/\Delta M > 2$. Based on these results we also present refined analytical solutions on the dilatation at the shock. This is discussed in the context of shock structure and the conditions for shock destruction by turbulence. An extended study of how changes experienced by turbulence can be characterized by the flow and thermodynamic conditions will also be discussed. [Preview Abstract] |
Sunday, November 19, 2017 9:31AM - 9:44AM |
A29.00008: New insights on compressible turbulent mixing in spectral space John Panickacheril John, Diego Donzis, Katepalli Sreenivasan Previous studies have shown that dilatational forcing has an effect in the dynamics of the velocity field in compressible turbulence. However, there has virtually been no studies of these effects on scalar mixing, the specific mechanisms responsible for compressibility effects and the scaling with governing parameters. Using a large DNS database, generated with different ratios of solenoidal to dilatational forcing, we find that the commonly used turbulent Mach number $(M_{t})$ fails to parametrize mixing efficiency. Instead, the dilatational Mach number $(M_{td}$) is a better scaling parameter to observe non-monotonic trends. We observe an accumulation of energy at large scales when compressibility is high; this has an effect on the energy and scalar cascade. We analyze both budgets to assess changes in global and inter-scale statistics for each mode and their interactions. For moderate compressibility levels, the normalized spectra of both modes do not collapse even when their own dissipation rates are used. Furthermore, a dilatational cascade is formed at high compressibilty levels with advection terms scaling with $\chi$, the ratio of dilatational to total kinetic energy. Results on scalar dissipation and their relation to thermodynamic variables are also presented. [Preview Abstract] |
Sunday, November 19, 2017 9:44AM - 9:57AM |
A29.00009: Theory of Relativistic Fluid Turbulence Gregory Eyink, Theodore Drivas Relativistic turbulence is expected in high-energy astrophysical flows, e.g. AGN outflow jets. We obtain exact theory by space-time coarse-graining the fluid stress-energy tensor, giving the analogue of Reynolds stress. Kinetic energy cascade is not natural in relativity, but cascade of internal energy is found with scale-transfer due to contraction of the Reynolds stress-energy tensor with the 4-gradient of the coarse-grained 4-velocity. Unlike non-relativistic turbulence, where energy flux is Galilei-invariant, Lorentz invariance of relativistic cascade is broken at finite Reynolds number but restored in the infinite-Reynolds limit. Otherwise, our results closely parallel those on non-relativistic compressible turbulence, with (i) a new mechanism of turbulent energy dissipation due to ``pressure-dilatation defect'' exemplified by relativistic shocks and (ii) an inverse cascade of entropy with microscopic entropy production as source and large-scale cooling as sink. We obtain Kolmogorov 4/5th-type laws that give estimates on turbulent scaling exponents. When speed of light goes to infinity, our theory recovers non-relativistic results. The analysis provides the framework for relativistic LES modeling and extends Onsager's ``dissipative Euler'' theory to relativistic turbulence. [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