Assessment Measures for LES Applications

Author(s):  
I. Celik ◽  
M. Klein ◽  
J. Janicka

Anticipating that Large Eddy Simulations will increasingly become the future engineering tool for research, development and design, it is deemed necessary to formulate some quality assessment measures that can be used to judge the resolution of turbulent scales and the accuracy of predictions. In this context some new and refined measures are proposed above and beyond those already published by the authors in the common literature. These new measures involve (a) fraction of total turbulent kinetic energy, (b) relative grid size with respect to Kolmogorov or Taylor scales, (c) relative effective sub-grid/numerical viscosity with respect to molecular viscosity, and (d) some property related to power spectra of turbulent kinetic energy. In addition, an attempt is made to segregate the contributions from numerical and modeling errors. Proposed measures are applied to various benchmark cases, and validated against fully resolved LES and/or DNS whenever possible. Along the same line of thinking, the authors present a perspective for verification of under-resolved direct numerical simulations.

2009 ◽  
Vol 131 (3) ◽  
Author(s):  
I. Celik ◽  
M. Klein ◽  
J. Janicka

Anticipating that large eddy simulations will increasingly become the future engineering tool for research, development, and design, it is deemed necessary to formulate some quality assessment measures that can be used to judge the resolution of turbulent scales and the accuracy of predictions. In this context some new and refined measures are proposed and compared with those already published by the authors in the common literature. These measures involve (a) fraction of the total turbulent kinetic energy, (b) relative grid size with respect to Kolmogorov or Taylor scales, and (c) relative effective subgrid/numerical viscosity with respect to molecular viscosity. In addition, an attempt is made to segregate the contributions from numerical and modeling errors. Proposed measures are applied to various test cases and validated against fully resolved large eddy simulation and/or direct numerical simulation whenever possible.


2020 ◽  
Author(s):  
Mohamed Sayed ◽  
Muhamed Hadziabdic ◽  
Abdelouahab Dehbi ◽  
Bojan Niceno ◽  
Konstantin Mikityuk

Author(s):  
Martin Söder ◽  
Lisa Prahl Wittberg ◽  
Björn Lindgren ◽  
Laszlo Fuchs

The effect of compression on a swirling/tumbling flow is studied using Large-Eddy Simulations (LES). In this study the geometry investigated is a cylinder with an artificially created swirling/tumbling motion. During compression the evolution of turbulence and vorticity are investigated. An increase of turbulence and vorticity is observed and linked to vorticity-dilatation interaction. It is shown that for swirling/tumbling flows turbulent kinetic energy available at Top Dead Center (TDC) is introduced by the piston through the vorticity-dilatation interaction and that turbulence increases independently of the presence of instability of the large scale flow structures.


Author(s):  
Mohamed Sayed ◽  
Muhamed Hadziabdic ◽  
Abdelouahab Dehbi ◽  
Bojan Niceno ◽  
Konstantin Mikityuk

2019 ◽  
Vol 867 ◽  
pp. 906-933 ◽  
Author(s):  
Riccardo Togni ◽  
Andrea Cimarelli ◽  
Elisabetta De Angelis

In this work we present and demonstrate the reliability of a theoretical framework for the study of thermally driven turbulence. It consists of scale-by-scale budget equations for the second-order velocity and temperature structure functions and their limiting cases, represented by the turbulent kinetic energy and temperature variance budgets. This framework represents an extension of the classical Kolmogorov and Yaglom equations to inhomogeneous and anisotropic flows, and allows for a novel assessment of the turbulent processes occurring at different scales and locations in the fluid domain. Two relevant characteristic scales, $\ell _{c}^{u}$ for the velocity field and $\ell _{c}^{\unicode[STIX]{x1D703}}$ for the temperature field, are identified. These variables separate the space of scales into a quasi-homogeneous range, characterized by turbulent kinetic energy and temperature variance cascades towards dissipation, and an inhomogeneity-dominated range, where the production and the transport in physical space are important. This theoretical framework is then extended to the context of large-eddy simulation to quantify the effect of a low-pass filtering operation on both resolved and subgrid dynamics of turbulent Rayleigh–Bénard convection. It consists of single-point and scale-by-scale budget equations for the filtered velocity and temperature fields. To evaluate the effect of the filter length $\ell _{F}$ on the resolved and subgrid dynamics, the velocity and temperature fields obtained from a direct numerical simulation are split into filtered and residual components using a spectral cutoff filter. It is found that when $\ell _{F}$ is smaller than the minimum values of the cross-over scales given by $\ell _{c,min}^{\unicode[STIX]{x1D703}\ast }=\ell _{c,min}^{\unicode[STIX]{x1D703}}Nu/H=0.8$, the resolved processes correspond to the exact ones, except for a depletion of viscous and thermal dissipations, and the only role of the subgrid scales is to drain turbulent kinetic energy and temperature variance to dissipate them. On the other hand, the resolved dynamics is much poorer in the near-wall region and the effects of the subgrid scales are more complex for filter lengths of the order of $\ell _{F}\approx 3\ell _{c,min}^{\unicode[STIX]{x1D703}}$ or larger. This study suggests that classic eddy-viscosity/diffusivity models employed in large-eddy simulation may suffer from some limitations for large filter lengths, and that alternative closures should be considered to account for the inhomogeneous processes at subgrid level. Moreover, the theoretical framework based on the filtered Kolmogorov and Yaglom equations may represent a valuable tool for future assessments of the subgrid-scale models.


2013 ◽  
Vol 444-445 ◽  
pp. 281-285 ◽  
Author(s):  
Tao Guo ◽  
Jun Zhou ◽  
Xiao Nan Liu

The vibration intensity is strong in Francis turbine occurred under the small opening conditions, such as Lijia Gorges and Three Gorges project. In paper we use large eddy simulation (LES) method base on Vreman SubGrid-Scale model to study the generation and evolution process of turbulence flow, capturing the details of the flow structures and the dissipation of the turbulent kinetic energy. The SIMPIEC algorithm is applied to solve the coupled equation of velocity and pressure. The result shows that the small guide vane opening conditions deviate the optimal conditions most. So some unstable flow characters been induced. Such as the turbulent kinetic energy of fluid in guide vanes zone, the blade passage and the draft tube are very strong. The unstable flow phenomenon including the swirl, flow separation, interruption and vortex strip. It can be deduced that the vibration of unit is induced by these flow characteristic.


2015 ◽  
Vol 93 (10) ◽  
pp. 1124-1130 ◽  
Author(s):  
T. Wang ◽  
P. Li ◽  
J.S. Bai ◽  
G. Tao ◽  
B. Wang ◽  
...  

The subgrid-scale (SGS) terms of turbulence transport are modelled by the stretched-vortex SGS stress model, and a large-eddy simulation code multi-viscous fluid and turbulence (MVFT) is developed to investigate the MVFT problems. Then one AWE shock tube experiment of interface instability is simulated numerically by MVFT code, which reproduces the development process of the interface. The obtained numerical images of interface evolution and wave structures in flow field are consistent with the experimental results. The evolution of perturbed interface and propagation of shock waves in flow field and their interactions are analyzed in detail. The statistics features of turbulence mixing in the form of finer quantities, such as the turbulent kinetic energy, enstrophy, density variance, and turbulent mass flux are investigated, which also proves that the SGS model has a key role in large-eddy simulation. The turbulent kinetic energy and enstrophy decay with time as a power law.


Sign in / Sign up

Export Citation Format

Share Document