<i>A posteriori</i> analysis of numerical errors in subfilter scalar variance modeling for large eddy simulation

Kaul, Colleen M.; Raman, Venkat

doi:10.1063/1.3556097

Cited by 32 publications

(30 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nearest the inlet [ Fig. 4(a) Z , the resolved dissipation depends on the square of the magnitude of the filtered scalar gradient and is prone to under prediction in LES computations using grid-based filtering [21,34]. It was verified that increasing the order of the finite di↵erence scheme used to compute Z,res increased that quantity's magnitude.…”

Section: Scalar Dissipation Rate Modelingmentioning

confidence: 88%

“…Prior studies have shown that the equilibrium approach is invalid even in the simplest of flows [21,22], leading to large underprediction of scalar variance. Consequently, the use of the VTE (Eq.…”

Section: Subfilter Models For Scalar Mixingmentioning

confidence: 99%

“…2). It has been shown that the STE provides a numerically accurate formulation [21]. Here, both VTE and STE formulations will considered in order to determine the role of numerical errors in the model predictions.…”

Section: Subfilter Models For Scalar Mixingmentioning

confidence: 99%

See 2 more Smart Citations

Large eddy simulation of a lifted ethylene flame using a dynamic nonequilibrium model for subfilter scalar variance and dissipation rate

Kaul

Raman

Knudsen

et al. 2013

Proceedings of the Combustion Institute

View full text Add to dashboard Cite

Accurate prediction of nonpremixed turbulent combustion using large eddy simulation (LES) requires detailed modeling of the mixing between fuel and oxidizer at scales finer than the LES filter resolution. In conserved scalar combustion models, the small scale mixing process is quantified by two parameters, the subfilter scalar variance and the subfilter scalar dissipation rate. The most commonly used models for these quantities assume a local equilibrium exists between production and dissipation of variance. Such an assumption has limited validity in realistic, technically relevant flow configurations. However, nonequilibrium models for variance and dissipation rate typically contain a model coe cient whose optimal value is unknown a priori for a given simulation. Furthermore, conventional dynamic procedures are not useful for estimating the value of this coe cient. In this work, an alternative dynamic procedure based on the transport equation for subfilter scalar variance is presented, along with a robust conditional averaging approach for evaluation of the

show abstract

Section: Scalar Dissipation Rate Modelingmentioning

confidence: 88%

Section: Subfilter Models For Scalar Mixingmentioning

confidence: 99%

See 1 more Smart Citation

Large eddy simulation of a lifted ethylene flame using a dynamic nonequilibrium model for subfilter scalar variance and dissipation rate

Kaul

Raman

Knudsen

et al. 2013

Proceedings of the Combustion Institute

View full text Add to dashboard Cite

show abstract

“…Algebraic models for the variance [8,9,10,11,12] and the dissipation rate [8,13,14] directly describe the modeled parameter using information about the local scalar field and filter width, and have the advantage of being conceptually simple and computationally inexpensive. Transport equation models for the variance [12,15,16] and for the dissipation rate [17,18,19] offer different advantages: they do not forcibly assume that production and dissipation processes are in equilibrium, they produce fields that are less noisy than algebraic models, and they incorporate a wider range of physics. Nonetheless, transport equation approaches are sensitive to the descriptions used to model unclosed subfilter terms, and the performance of these equations is not always superior to the performance of algebraic models [12].…”

Section: Introductionmentioning

confidence: 99%

Modeling scalar dissipation and scalar variance in large eddy simulation: Algebraic and transport equation closures

et al. 2012

View full text Add to dashboard Cite

The scalar dissipation rate and the scalar variance are important modeling parameters in large eddy simulations (LES) of reacting flows. Approaches ranging from constant coefficient algebraic equations to transport equations with complex closure models have been proposed to describe these parameters. While currently available approaches capture many of the physics behind the dissipation and variance, their modeling remains challenging. Here, two direct numerical simulations (DNS) are used to analyze LES dissipation rate and variance models, and to propose a new model for the dissipation rate that is based on a transport equation. The first DNS that is considered is a non-premixed auto-igniting C 2 H 4 jet flame simulation performed by Yoo et al. (Proc. Comb. Inst., 2010). An LES of this case is run using algebraic models for the dissipation rate and variance. It is shown that the algebraic models do not accurately reproduce the subfilter DNS fields. This motivates the introduction of a transport equation model for the LES dissipation rate. Closure of the equation is addressed by formulating a new adapted dynamic approach. This approach borrows dynamically computed information from LES quantities that, unlike the dissipation rate, do not reside on the smallest flow length scales. The adapted dynamic closure is analyzed by considering a second DNS of scalar mixing in homogeneous isotropic turbulence. Data from this second DNS is used to confirm that the adapted dynamic procedure successfully closes the dissipation rate equation over a wide range of LES filter widths. The first reacting jet case is then returned to and used to test the LES transport equation models. These models are shown to predict the dissipation rate and variance fields more accurately than the the algebraic LES models.

show abstract

“…However, numerical errors scale with wavenumber [34,35], and are highest close to the mesh-scale. Consequently, the near-filter scale features in LES are contaminated by numerical errors [36,37]. Practically, the evaluation of gradient-based quantities is poorly approximated [36,37].…”

Section: Numerical Errors In Dqmommentioning

confidence: 99%

A multivariate quadrature based moment method for LES based modeling of supersonic combustion

Donde

Koo

Raman

2012

Journal of Computational Physics

View full text Add to dashboard Cite

A posteriori analysis of numerical errors in subfilter scalar variance modeling for large eddy simulation

Cited by 32 publications

References 35 publications

Large eddy simulation of a lifted ethylene flame using a dynamic nonequilibrium model for subfilter scalar variance and dissipation rate

Large eddy simulation of a lifted ethylene flame using a dynamic nonequilibrium model for subfilter scalar variance and dissipation rate

Modeling scalar dissipation and scalar variance in large eddy simulation: Algebraic and transport equation closures

A multivariate quadrature based moment method for LES based modeling of supersonic combustion

Contact Info

Product

Resources

About