Abstract
The partially integrated transport modeling (PITM) method first introduced in Refs. Schiestel and Dejoan (Theor Comput Fluid Dyn 18:443, 2005) and in Chaouat and Schiestel (Phys Fluids 17:065106, 2005) provides a continuous approach for hybrid RANS-LES simulations. Inspired from the multiple scale approach, the basis of the development of the method is the spectral space in quasi-homogeneous turbulence. The PITM method embodies a partitioning control function that monitors the ratio of subfilter energy to total turbulent energy by reference to the cutoff wavenumber location. How this procedure behaves in inhomogeneous flows is an important question. The present paper demonstrates that the same control function can be used both in homogeneous and in non-homogeneous flows as well. Further on, an analysis of the effect of anisotropic filters generally used for wall flows is conducted for computing the equivalent cutoff wavenumber suitable for determining the subfilter energy of the spectrum and see how it interferes with control function. Illustrations in the turbulent plane channel flow are given that confirm the efficiency of the procedure and DNS data have been used to support and supplement the discussion.
Similar content being viewed by others
References
Bardina, J., Ferziger, J.H., Reynolds, W.C.: Improved subgrid scale models for large eddy simulations. AIAA-paper 1357, 1–10 (1980)
Befeno, I., Schiestel, R.: Non-equilibrium mixing of turbulence scales using a continuous hybrid RANS/LES approach: application to the shearless mixing layer. Flow Turbul. Combust 78, 129–151 (2007)
Cambon, C., Jeandel, D., Mathieu, J.: Spectral modelling of homogeneous non-isotropic turbulence. J. Fluid Mech. 104, 247–262 (1981)
Chaouat, B., Peyret, C.: Investigation of the wall scalar fluctuations effect on passive scalar turbulent fields at several Prandtl numbers by means of direct numerical simulations. J. Heat Transfer ASME 141, 1–9 (2019)
Chaouat, B., Schiestel, R.: Extension of the partially integrated transport modeling method to the simulation of passive scalar turbulent fluctuations at various Prandtl numbers. Submitted to Int. J. Heat Fluid Flow (2020)
Chaouat, B., Schiestel, R.: Simulations of turbulent flows out of spectral equilibrium using the PITM method, 23ème Congrès Français de Mécanique, 2491-715X, 1–16 (2017)
Chaouat, B.: Subfilter scale transport model for hybrid RANS/LES simulations applied to a complex bounded flow. J. Turbul. 11(51), 1–30 (2010)
Chaouat, B.: An efficient numerical method for RANS/LES turbulent simulations using subfilter scale stress transport equations. Int. J. Numer. Methods Fluids 67, 1207–1233 (2011)
Chaouat, B.: Simulation of turbulent rotating flows using a subfilter scale stress model derived from the partially integrated transport modeling method. Phys. Fluids 24(045108), 1–35 (2012)
Chaouat, B.: The state of the art of hybrid RANS/LES modeling for the simulation of turbulent flows. Flow Turbul. Combust. 99, 279–327 (2017)
Chaouat, B.: Commutation errors in PITM simulations. Int. J. Heat Fluid Flow 67, 138–154 (2017)
Chaouat, B.: Application of the PITM method using inlet synthetic turbulence generation for the simulation of the turbulent flow in a small axisymmetric contraction. Flow Turbul. Combust. 98, 987–1024 (2017)
Chaouat, B., Schiestel, R.: A new partially integrated transport model for subgrid-scale stresses and dissipation rate for turbulent developing flows. Phys. Fluids 17(065106), 1–19 (2005)
Chaouat, B., Schiestel, R.: From single-scale turbulence models to multiple-scale and subgrid-scale models by Fourier transform. Theoret. Comput. Fluid Dyn. 21, 201–229 (2007)
Chaouat, B., Schiestel, R.: Progress in subgrid-scale transport modelling for continuous hybrid non-zonal RANS/LES simulations. Int. J. Heat Fluid Flow 30, 602–616 (2009)
Chaouat, B., Schiestel, R.: Analytical insights into the partially integrated transport modeling method for hybrid Reynolds averaged Navier–Stokes equations-large eddy simulations of turbulent flows. Phys. Fluids 24(085106), 1–34 (2012)
Chaouat, B., Schiestel, R.: Hybrid RANS-LES simulations of the turbulent flow over periodic hills at high Reynolds number using the PITM method. Comput. Fluids 84, 279–300 (2013)
Chaouat, B., Schiestel, R.: Partially integrated transport modeling method for turbulence simulation with variable filters. Phys. Fluids 25(125102), 1–39 (2013)
Davidson, L., Friess, C.: A new formulation of \(f_k\) for the PANS model. J. Turbul. 20(5), 322–336 (2019)
Deardorff, J.W.: A numerical study of three-dimensional turbulent channel flow at large Reynolds numbers. J. Fluid Mech. 41, 453–480 (1970)
Durbin, P.: Some recent developments in turbulence closure modeling. Annu. Rev. Fluid Mech. 50, 77–103 (2017)
Egorov, Y., Menter, F.T., Lechner, R., Cokljat, D.: The scale-adaptive simulation method for unsteady turbulent flow prediction? Part 2? Application to complex flows. Flow Turbul. Combust 85, 139–165 (2020)
Fasel, H.F., Seidel, J., Wernz, S.: A methodology for simulations of complex turbulent flows. J. Fluids Eng. ASME 124(4), 933–942 (2002)
Foroutan, H., Yavuzkurt, S.: A partially averaged Navier Stokes model for the simulation of turbulent swirling flow with vortex breakdown. Int. J. Heat Fluid Flow 50, 402–416 (2014)
Friess, C., Davidson, L.: A formulation of PANS capable of mimicking IDDES. Int. J. Heat Fluid Flow 86(108666), 1–25 (2020)
Fröhlich, J., Von Terzi, D.: Hybrid LES/RANS methods for the simulation of turbulent flows. Prog. Aerosp. Sci. 44, 349–377 (2008)
Germano, M.: Turbulence? The filtering approach. J. Fluid Mech. 41(238), 325–336 (1992)
Girimaji, S.S., Abdol-Hamid, K.S.: Partially-averaged Navier Stokes model for turbulence: implementation and validation. AIAA-paper 0502, 1–14 (2005)
Girimaji, S.S.: Partially-averaged Navier–Stokes method for turbulence: a Reynolds averaged Navier–Stokes to direct numerical simulation bridging method. ASME J. Appl. Mech. 73(3), 413–421 (2006)
Girimaji, S.S., Jeong, E., Srinivasan, R.: Partially averaged Navier–Stokes method for turbulence? Fixed point analysis and comparisons with unsteady partially averaged Navier–Stokes. ASME J. Appl. Mech. 73(3), 422–429 (2006)
Hamba, F.: Log-layer mismatch and commutation error in hybrid RANS/LES simulation of channel flow. Int. J. Heat Fluid Flow 30, 20–31 (2009)
Heinz, S., Mokhtarpoor, R., Stoellinger, M.: Physics-based control of hybrid RANS-LES methods. AIAA-paper 0327, 1–12 (2019)
Heinz, S.: The large eddy simulation capability of Reynolds-averaged Navier–Stokes equations: analytical results. Phys. Fluids 31(021702), 1–6 (2019)
Heinz, S.: A review of hybrid RANS-LES methods for turbulent flows: concepts and applications. Prog. Aerosp. Sci. 114(100595), 1–25 (2020)
Heinz, S., Mokhtarpoor, R., Stoellinger, M.: Theory-based Reynolds-averaged Navier–Stokes equations with large eddy simulation capability for separated turbulent flow simulations. Phys. Fluids 32(065102), 1–20 (2020)
Hinze, J.O.: Turbulence. Mc Graw-Hill, New York (1975)
Jakirlic, S., Maduta, R.: Extending the bounds of steady RANS closures: toward an instability-sensitive Reynolds stress model. Int. J. Heat Fluid Flow 51, 175–194 (2015)
Jeandel, D., Brison, J.F., Mathieu, J.: Modeling methods in physical and spectral spaces. Phys. Fluids 100(21), 169–181 (1978)
Kamble, C., Girimaji, S.S., Chen, H.: Partially averaged Navier–Stokes formulation of a two-layer turbulence model. AIAA J. 58(1), 174–183 (2020)
Klapwijk, M., Lloyd, T., Vaz, G.: On the accuracy of partially averaged Navier–Stokes resolution estimates. Int. J. Heat Fluid Flow 80(108484), 1–10 (2019)
Menter, F.R., Egorov, Y.: The scale-adaptive simulation method for unsteady turbulent flow prediction? Part 1? Theory and model description. Flow Turbul. Combust. 85, 113–138 (2020)
Mons, V., Cambon Sagaut, P.: A spectral model for homogeneous shear-driven anisotropic turbulence in terms of spherically averaged descriptors. J. Fluid Mech. 788, 147–182 (2016)
Press, W.H., Teukolsky, S.A., Vetterling, W.T., Flannery, B.P.: Numerical Recipes. Cambridge University Press, Cambridge (1992)
Razi, P., Tazraei, P., Girimaj, S.S.: Partially-averaged Navier–Stokes (PANS) simulations of flow separation over smooth curved surfaces. Int. J. Heat Fluid Flow 21, 252–263 (2017)
Schiestel, R.: Modeling and Simulation of Turbulent Flows. Wiley, ISTE Ltd and J (2008)
Schiestel, R., Dejoan, A.: Towards a new partially integrated transport model for coarse grid and unsteady turbulent flow simulations. Theoret. Comput. Fluid Dyn. 18, 443–468 (2005)
Scotti, A., Meneveau, Ch.: Generalized Smagorinsky model for anisotropic grids. Phys. Fluids A 5(9), 2306–2308 (1993)
Scotti, A., Meneveau, Ch., Fatica, M.: Dynamic Smagorinsky model on anisotropic grids. Phys. Fluids 9(6), 1856–1858 (1997)
Shur, M.L., Spalart, P.R., Strelets, M.K., Travin, K.: A hybrid RANS-LES approach with delayed-DES and wall-modelled LES capabilities. Int. J. Heat Fluid Flow 29, 1638–1649 (2008)
Spalart, P.R.: Strategies for turbulence modelling and simulations. Int. J. Heat Fluid Flow 21, 252–263 (2000)
Spalart, P.R.: Detached-eddy simulation. Annu. Rev. Fluid Mech. 41, 181–202 (2009)
Speziale, C.G.: Turbulence modeling for time-dependent RANS and VLES? A review. AIAA J. 36, 173–184 (1998)
Tazraei, P., Girimaji, S.S.: Scale-resolving simulations of turbulence: equilibrium boundary layer analysis leading to near-wall closure modeling. Phys. Rev. Fluids 4(104607), 1–17 (2019)
Vichnevetsky, R.: Fourier Analysis of Numerical Approximation of Hyperbolic Equations. SIAM, Philadelphia (1982)
Yassour, Y., Wolfshtein, M.: A spectral model for the calculation of homogeneous turbulent flows. Symposium on Turbulent Shear Flows IV, Karlsruhe Univ. 9.23-9.29 (1983)
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Appendices
Appendix A: Mathematical Properties
Let us consider
and
Eq. (129) can be developed in the following form as
It is then possible to get the averaging in statistical sense of (131) as
Strictly speaking, the filtering process leads to mathematical non-equality \(\overline{ \left\langle \phi \right\rangle } =\left\langle {\bar{\phi }}\right\rangle \ne \left\langle \phi \right\rangle \) for any variable \(\phi \). To recover \(\left\langle {\bar{\phi }}\right\rangle = \left\langle \phi \right\rangle \), the key concept is to consider the tangent homogeneous anisotropic turbulence field at the physical space location implying that the variation of the mean variable \(\phi \) is accounted for by the use of Taylor series expansion in space limited to the linear terms (Chaouat and Schiestel 2007, 2013; Chaouat 2017). In practice, \(\left\langle {\bar{\phi }}\right\rangle \approx \left\langle \phi \right\rangle \) only if the variation of the flow velocities over the filter width is not too large. In the framework of the tangent homogeneous space and the spectral cutoff filter, \(\overline{ \left\langle \phi \right\rangle } =\left\langle {\bar{\phi }}\right\rangle = \left\langle \phi \right\rangle \) and \(\left\langle \phi ^> \phi ^< \right\rangle = 0\), so that Eq. (132) reduces to
and we get also
In particular, the subfilter scale stress tensor defined in large eddy simulation is then
while the resolved scale stress tensor is
so that
In RANS modeling, the Reynolds stress tensor accounting for the total fluctuating velocities is defined as
Using the decomposition \(\phi =\left\langle \phi \right\rangle +\phi ^<+ \phi ^>\), the Reynolds stress tensor \(R_{ij}\) can be rewritten as
So that, assuming that the correlation between the small scale and large scale \(\left\langle u^<_i u^>_j\right\rangle \) are small compared to the other correlations, \(R_{ij}\) can be computed in a first approximation as the sum of the statistical average of subfilter and resolved stresses
and the contraction of the tensors appearing in Eq. (140) leads to
where \(k^{(s)}\) and \(k^{(r)}\) are the filtered subfilter and resolved turbulence energy, respectively. This relation is strictly true in the case of the spectral cutoff filter, and only approximate in the case of a sharp filter.
Appendix B: Formal Derivative Operators
For sake of clarity, we consider in this section the use of a uniform spectral cutoff filter (Chaouat and Schiestel 2007, 2013; Chaouat 2017). As recalled in “Appendix A”, each variable \(\phi \) can be decomposed into a statistical mean \(\left\langle \phi \right\rangle \), a macro-scale fluctuation \(\phi ^<\) and a micro-scale fluctuation \(\phi ^>\). Using this decomposition for the velocity, it is a simple matter to see that the mass conservation implies
But the spectral cutoff filter also yields
The particle derivative in the instantaneous flow is given by
The material derivative accounting for the statistical velocity is
and in particular
The material derivative accounting for the filtered velocity is
or equivalently
so that the averaging in the statistical sense of Eq. (147) is
On the other hand, applying a filtering process on Eq. (147) gives
Using (143), Eq. (149) can be rewritten as
leading to
More generally, from the definition
and one can also easily derive some useful relations such as
Finally, note that
because \(\left\langle u^<_i \phi ^>\right\rangle \) reduces to zero.
Appendix C: Remark on Turbulent Flows Near Energetic Equilibrium
In the framework of first-moment closure referring to eddy-viscosity turbulence models, the production term, without prejudice to the reasoning, can be expressed as
with
and
so that
where the last term is the mean spectral flux \( F^{(1)}\). As it is clear from Eq. (160), the averaging in a statistical sense of the subfilter production (including the mean flow production and the production from the large resolved scales giving rise to \(F^{(1)}\) approaches the RANS production, i.e., \( \left\langle P^{(s)} \right\rangle \approx P\) and the same result prevails for the subfilter dissipation-rate \( \left\langle \varepsilon ^{(s)} \right\rangle \approx \epsilon \) for flows close to energetic equilibrium so that
where \(\nu _{t}\) denotes the turbulent eddy viscosity in RANS modeling given by
involving the total kinetic energy. All these relations can be also rewritten
These relations are recalling the Heisenberg hypothesis is spectral space (Hinze 1975) assuming that the effect of the small eddies with wavenumbers larger than \(\kappa \) by which energy is withdrawn from larger eddies with wavenumbers smaller than \(\kappa \) is described through a small scales eddy viscosity. Then the spectral flux is represented as the product of two integrals in the domains \( \left[ {0,\kappa } \right] \) and \( \left[ {\kappa ,\infty } \right] \) as in Eq. (163).
Appendix D: Order of Magnitude of the Resolved Part of the Turbulent Diffusion
We consider the resolved part of the turbulent advection denoted \(J^{(r)}\) of any arbitrary variable \(\psi \) due to the macro-scale fluctuation velocity \(u^<_i\) as
and which becomes a turbulent diffusion term once statistically averaged.
The large scale fluctuation variable \(\psi ^< = {\bar{\psi }} -\left\langle \psi \right\rangle \) is fully resolved in the numerical simulation but for convenience an analytical approximation can be easily obtained considering the convection process by means of the large scale fluctuating velocity \(u^<_i\) and using the time scale of the resolved eddies \(\tau ^{(r)}\) as
where
corresponds to the displacement time of an eddy before loosing its individuality. The approximation of the fluctuation corresponding to a wavelength range \([\kappa _1,\kappa _2]\) is approached by characteristic the time scale of this range of eddies multiplied by the gradient of the filtered field with a cutoff at \(\kappa _1\). If \(\kappa _1=0\), then the use of the gradient of the mean value is natural. Then, Taylor series expansion of Eq. (166) limited to the first order yields
So that
or in a first approximation by neglecting the cross-correlations
where \(C_\psi \) is a numerical coefficient introduced to describe the diffusion process, leading to
Moreover, for the instantaneous value \({\bar{\psi }}= \left\langle \psi \right\rangle + \psi ^<\), a fluctuating part remains
Similarly, the diffusion process caused by the micro-scale fluctuation \(\psi ^>\) is then can be computed as
with
leading to
so that
The total mean diffusion involving the micro-scale and the macro-scale is therefore
or in a more compact form
that is to say
Note that the mean diffusion originating from the resolved scales is non zero only in the case of triple correlations. Indeed, as already remarked, double correlations between large scale and small scale fluctuations are always zero \(\left\langle u_i^{<}u_i^{>}\right\rangle =0 \). In contrast, for triple correlations like \(\left\langle u_j^{<}k^{(s)'}\right\rangle \) with \(k^{(s)'}=k^{(s)<}+k^{(s)>}\), \(k^{(s)<}=u_i^{<}u_i^{<}-\left\langle u_i^{<}u_i^{<}\right\rangle \) and \(k^{(s)>}=u_i^{>}u_i^{>}-\left\langle u_i^{>}u_i^{>}\right\rangle \), it appears \(\left\langle u_j^{<} k^{(s)'}\right\rangle =\left\langle u_j^{<} k^{(s)<}\right\rangle \). In practice, an approximation of \(\left\langle J^{(s)}_\psi \right\rangle \) and \(\left<J^{(r)}_\psi \right>\) for practical use are thus
and
We have thus approximated both the large scale fluctuating diffusion term and the small scale averaged one in a consistent way.
Appendix E: Variational Calculus of the Diffusion Terms \(J^{(s+r)}_{\psi} (\psi )\)
The general expressions of the diffusion terms \(J^{(r)}_\psi (\psi )\) and \(J^{(s)}_\psi (\psi )\) are given by Eqs. (172) and (176) in which \(k^{(r)}\), \(k^{(s)}\) and \(\epsilon ^{(s)}\) are all functions of \((x_i,t)\). In a first approximation, the functional variations of \(\psi (x_i,t)\) can be computed as \(\delta \psi (x_i,t) = \zeta (x_i , t)\delta \alpha \), where \(\zeta (x_i ,t)\) is a function in space and time while \(\delta \alpha \) is a small scalar increment, it clearly emphasizes the fact that \(\delta \) is a functional variation of \(\psi \) whose magnitude is controlled by \(\delta \alpha \) while keeping the same shape \( \zeta (x_i ,t) \). If \(\partial /\partial x_i\) is the usual partial differential operator in \(x_i\) (or t ) corresponding to the usual first partial derivative, then one can write
regardless the particular shape of the \(\psi (x_i , t)\) curve. Then, assuming that the function \( \zeta (x_i ,t) \) roughly keeps the same shape as the original function \(\psi (x_i , t)\), that is to say
we get the following estimate
Applied to Eq. (178), one gets
and thus
for \({\psi } = k^{(s)}(x_i ,t)\) and \({\psi }=\epsilon ^{(s)}(x_i ,t)\), respectively, considering that \(k^{(s+r)}=k^{(s)}+k^{(r)}\) remains approximatively constant when the cutoff is varied, Eq. (186) reduces to
Due to the implicit j summation in Eq. (178), one may reasonably wonder if Eq. (185) is valid. This equation is still verified because for each value of j, (j=1,2 and 3), the same approximation given by Eq. (186) is obtained so that
Note that the relations (182) and (184) are also valid for instantaneous quantities but in this case \(\delta \alpha \) becomes fluctuating.
Rights and permissions
About this article
Cite this article
Chaouat, B., Schiestel, R. Energy Partitioning Control in the PITM Hybrid RANS/LES Method for the Simulation of Turbulent Flows. Flow Turbulence Combust 107, 937–978 (2021). https://doi.org/10.1007/s10494-021-00259-w
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10494-021-00259-w