SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
The resolution requirements for direct numerical simulations of transition to turbulence are investigated. A reliable resolution criterion is determined from the results of several detailed simulations of channel and boundary-layer transition.
The resolution requirements for direct numerical simulations of transition to turbulence are investigated. A reliable resolution criterion is determined from the results of several detailed simulations of channel and boundary-layer transition.
Compressible wall-modeled large-eddy simulations of Mach 8 turbulent boundary-layer flows over a flat plate were carried out for the conditions of the hypersonic wind tunnel at Sandia National Laboratories. The simulations provide new insight into the effect of wall cooling on the aero-optical path distortions for hypersonic turbulent boundary-layer flows. Four different wall-to-recovery temperature ratios, 0.3, 0.48, 0.71, and 0.89, are considered. Despite the much lower grid resolution, the mean velocity, temperature, and resolved Reynolds stress profiles from the simulation for a temperature ratio of 0.48 are in good agreement with those from a reference direct numerical simulation. The normalized root-mean-square optical path difference obtained from the present simulations is compared with that from reference direct numerical simulations, Sandia experiments, as well as predictions obtained with a semi-analytical model by Notre Dame University. Here the present analysis focuses on the effect of wall cooling on the wall-normal density correlations, on key underlying assumptions of the aforementioned model such as the strong Reynolds analogy, and on the elevation angle effect on the optical path difference. Wall cooling is found to increase the velocity fluctuations and decrease the density fluctuations, resulting in an overall reduction of the normalized optical path distortion. Compared to the simulations, the basic strong Reynolds analogy overpredicts the temperature fluctuations for cooled walls. Also different from the strong Reynolds analogy, the velocity and temperature fluctuations are not perfectly anticorrelated. Finally, as the wall temperature is raised, the density correlation length, away from the wall but inside the boundary layer, increases significantly for beam paths tilted in the downstream direction.
While recent approaches, such as the macroscopic forcing method (MFM) or Green's function-based approaches, can be used to compute Reynolds-averaged Navier-Stokes closure operators using forced direct numerical simulations, MFM can also be used to directly compute moments of the effective nonlocal and anisotropic eddy diffusivities. The low-order spatial and temporal moments contain limited information about the eddy diffusivity but are often sufficient for quantification and modeling of nonlocal and anisotropic effects. However, when using MFM to compute eddy diffusivity moments, the statistical convergence can be slow for higher-order moments. In this work, we demonstrate that using the same direct numerical simulation (DNS) for all forced MFM simulations improves statistical convergence of the eddy diffusivity moments. We present its implementation in conjunction with a decomposition method that handles the MFM forcing semianalytically and allows for consistent boundary condition treatment, which we develop for both scalar and momentum transport. We demonstrate that for a two-dimensional Rayleigh-Taylor instability case study, using the same DNS for all forced MFM simulations results in convergence with 𝒪(100) simulations rather than 𝒪(1000) simulations. In conclusion, we then demonstrate the impacts of improved convergence on the quantification of the eddy diffusivity.
This paper presents a new deep learning data-driven model for predicting structure dependent pore-fluid velocity fields in rock. The model is based on a Convolutional Auto-Encoder (CAE) artificial neural network capable of learning from image data generated by direct numerical simulations of fluid flow through pore-structures, such as by Lattice Boltzmann or molecular dynamics methods. The main novelty of the model in comparison to previous CAE-based data-driven approaches consists of three parts. The first is a methodology for decomposing the full-domain of the porous media into sub-regions, or “sub-domains”, in order to reduce the overall size of the CAE, batch process the sub-domains in parallel, and enable the CAE to learn local and generalizable nonlinear mappings of pore-fluid velocities. The second consists of embedding the finite difference solutions of the incompressible Navier-Stokes and continuity equations into convolutional layers prior to the CAE in order to provide the CAE with knowledge of fluid dynamics physics (PhyFlow). The third main novelty is that the training of the CAE is regularized with a hierarchical loss function that encourages the learning of fluid flow patterns (in a way similar to ranked modes in principal component analysis), ranking from most to least important. This is shown to increase the stability in learning, reduce over-fitting, and promote interpretability of the CAE neural network layers (HierCAE). The comprehensive new data-driven model, which we call the PhyFlow-HierCAE model, is shown to exhibit improved accuracy and generalizability of flow field predictions over conventional CAE models, attributable to the embedded physical knowledge and the hierarchical regularization, as well as realize orders of magnitude speed-ups in computation times as a surrogate for the direct numerical simulations. Examples of training and forward predictions on unseen pore-structures are provided and evaluated for data from Lattice Boltzmann and molecular dynamics simulations of pore-fluid flow. The model is shown to be a fast and accurate emulator (or “surrogate”) for predicting effective permeability of unseen pore-structures based on learning from relatively small direct numerical simulation datasets.
The dynamic subgrid-scale (SGS) model of Germano et al. (1991) is generalized for the large eddy simulation (LES) of compressible flows and transport of a scalar. The model was applied to the LES of decaying isotropic turbulence, and the results are in excellent agreement with experimental data and direct numerical simulations. The expression for the SGS turbulent Prandtl number was evaluated using direct numerical simulation (DNS) data in isotropic turbulence, homogeneous shear flow, and turbulent channel flow. The qualitative behavior of the model for turbulent Prandtl number and its dependence on molecular Prandtl number, direction of scalar gradient, and distance from the wall are in accordance with the total turbulent Prandtl number from the DNS data.
Three direct numerical simulations of time-evolving turbulent plane wakes developing in the presence of various irrotational plane strains have been generated. A pseudospectral numerical method with up to 26 million modes is used to solve the equations in a reference frame moving with the irrotational strain. The initial condition for each simulation is taken from a previous turbulent self-similar plane wake direct numerical simulation at a velocity deficit Reynolds number, Re, of about 2000. Three different plane strains (of the same magnitude) are imposed. In the first two simulations the strain is in a plane normal to the streamwise wake direction (the two cases having strain of opposite sign); in the third the wake is compressed in the streamwise direction and stretched in the inhomogeneous cross-stream direction. The two flows that are stretched in the cross-stream direction experience an exponential increase of Re; flow visualization indicates many small-scale vortices with little or no organized large-scale structure. In the flow that is compressed in the cross-stream direction Re decays exponentially and the layer appears to be relaminarizing. The evolution of several turbulence statistics in each of these flows is examined.
A k-epsilon model that accounts for viscous and wall effects is presented. The proposed formulation does not contain the local wall distance thereby making very simple the application to complex geometries. The formulation is based on an existing k-epsilon model that proved to fit very well with the results of direct numerical simulation. The new form is compared with nine different two-equation models and with direct numerical simulation for a fully developed channel flow at Re = 3300. The simple flow configuration allows a comparison free from numerical inaccuracies. The computed results prove that few of the considered forms exhibit a satisfactory agreement with the channel flow data. The model shows an improvement with respect to the existing formulations.
As pointed out by Rodi standard integral solutions for jets and plumes developed for discharge into infinite, quiescent ambient are difficult to extend to complex situations, particularly in the presence of boundaries such as the sea floor or ocean surface. In such cases the assumption of similarity breaks down and it is impossible to find a suitable entrainment coefficient. The models are also incapable of describing any but the most slowly varying unsteady motions. There is therefore a need for full time-dependent modeling of the flow field for which there are three main approaches: (1) Reynolds averaged numerical simulation (RANS), (2) large eddy simulation (LES), and (3) direct numerical simulation (DNS). Rodi applied RANS modeling to both jets and plumes with considerable success, the test being a match with experimental data for time-averaged velocity and temperature profiles as well as turbulent kinetic energy and rms axial turbulent velocity fluctuations. This model still relies on empirical constants, some eleven in the case of the buoyant jet, and so would not be applicable to a partly laminar plume, may have limited use in the presence of boundaries, and would also be unsuitable if one is after details of the unsteady component of the flow (the turbulent eddies). At the other end of the scale DNS modeling includes all motions down to the viscous scales. Boersma et al. have built such a model for the non-buoyant case which also compares well with measured data for mean and turbulent velocity components. The model demonstrates its versatility by application to a laminar flow case. As its name implies, DNS directly models the Navier-Stokes equations without recourse to subgrid modeling so for flows with a broad spectrum of motions (high Re) the cost can be prohibitive - the number of required grid points scaling with Re(exp 9/4) and the number of time steps with Re(exp 3/4). The middle road is provided by LES whereby the Navier-Stokes equations are formally filtered with the filter chosen to only exclude the smallest turbulent motions. If successful, LES should provide much of the detail available to DNS but at more bearable cost. Fatica et al. in comparing LES with DNS for a low Reynolds number jet showed that the LES could simulate the temporally evolving behavior including growth of the jet thickness. It is the intention of this report to explore the application of an LES model to jets and plumes. As always, before tackling complex situations, the model must be tested for the simplest of cases and so we address only two, a non-buoyant axisymmetric jet issuing steadily from an orifice into a semi-infinite stationary environment and a buoyant jet in the same environment. The work is a continuation of Basu and Mansour.
Among the various types of advanced nuclear technologies that exist, high-temperature gas-cooled reactors (HTGRs) stand out for their co-generation capabilities and exceptional passive safety systems. HTGRs rely on a reactor cavity cooling system (RCCS) to remove decay heat from the reactor pressure vessel (RPV) during the course of long-term transients. Several studies have focused on experimental and numerical modeling of the RCCS; however, the literature reveals a notable gap when it comes to high-fidelity simulations such as direct numerical simulations (DNS) or large eddy simulation (LES) models of the RCCS. An LES model was developed for forced convection inside the RCCS so as to obtain a high-fidelity representation of the RCCS at a scale never before attempted. Furthermore, the LES model was compared against three different Reynolds-averaged Navier–Stokes (RANS) models (Realizable k − ϵ, k − ω SST, and Reynolds stress model [RSM]). This comparison verified the robustness and accuracy of the turbulence modeling approach of the RANS models for the RCCS. Finally, based on the comparison between the LES and the RANS models, the k − ω SST RANS model of the experimental facility was developed and compared against the available experimental data.
Direct numerical simulations of turbulent flows are predominantly carried out using structured, hexahedral meshes despite decades of development in unstructured mesh methods. Tetrahedral meshes offer ease of mesh generation around complex geometries and the potential of an orientation free grid that would provide un-biased small-scale dissipation and more accurate intermediate scale solutions. However, due to the lack of consistent multi-dimensional numerical formulations in conventional schemes for triangular and tetrahedral meshes at the cell interfaces, numerical issues exist when flow discontinuities or stagnation regions are present. The space-time conservative conservation element solution element (CESE) method - due to its Riemann-solver-free shock capturing capabilities, non-dissipative baseline schemes, and flux conservation in time as well as space - has the potential to more accurately simulate turbulent flows using unstructured tetrahedral meshes. To pave the way towards accurate simulation of shock/turbulent boundary-layer interaction, a series of wave and shock interaction benchmark problems that increase in complexity, are computed in this paper with triangular/tetrahedral meshes. Preliminary computations for the normal shock/turbulence interactions are carried out with a relatively coarse mesh, by direct numerical simulations standards, in order to assess other effects such as boundary conditions and the necessity of a buffer domain. The results indicate that qualitative agreement with previous studies can be obtained for flows where, strong shocks co-exist along with unsteady waves that display a broad range of scales, with a relatively compact computational domain and less stringent requirements for grid clustering near the shock. With the space-time conservation properties, stable solutions without any spurious wave reflections can be obtained without a need for buffer domains near the outflow/farfield boundaries. Computational results for the isotropic turbulent flow decay, at a relatively high turbulent Mach number, show a nicely behaved spectral decay rate for medium to high wave numbers. The high-order CESE schemes offer very robust solutions even with the presence of strong shocks or widespread shocklets. The explicit formulation in conjunction with a close to unity theoretical upper Courant number bound has the potential to offer an efficient numerical framework for general compressible turbulent flow simulations with unstructured meshes.
Many Reynolds averaged Navier-Stokes solvers use closure models in conjunction with 'the law of the wall', rather than deal with a thin, viscous sublayer near the wall. This work is motivated by the need for better models to compute near wall turbulent flow. The authors use direct numerical simulation of fully developed channel flow and one of three dimensional turbulent boundary layer flow to develop new models. These direct numerical simulations provide detailed data that experimentalists have not been able to measure directly. Another objective of the work is to examine analytically the effects of rotation on turbulence, using Rapid Distortion Theory (RDT). This work was motivated by the observation that the pressure strain models in all current second order closure models are unable to predict the effects of rotation on turbulence.
This paper investigates the flow behavior of involute-plate research reactors by performing Reynolds-Averaged Navier Stokes simulation (RANS), Large Eddy Simulation (LES) and Direct Numerical Simulation (DNS) of the channel flow between fuel plates. By modeling turbulence with different numerical approaches, this study provides data with three levels of fidelity. For the RANS simulation, three widely used turbulence models, i.e., k-ε, k-ω, Reynolds Stress Turbulence model (RST) are applied by using the commercial CFD code STAR-CCM +. For LES and DNS, the open-source CFD code, Nek5000, is used given its outstanding scalability on High Performance Computer (HPC) and high-order technique. The results from RANS simulations are compared with that from LES and DNS for benchmarking. Both macroscale parameters and turbulence statistics, such as velocity magnitude, lateral velocity and turbulence kinetic energy, are presented and analyzed. The results from RANS simulation achieve good agreement with LES and DNS on velocity and turbulence kinetic energy prediction. The RST turbulence model predicts the most similar flow pattern of lateral velocity as compared to LES and DNS. The Lambda-2 (λ2) criterion with a reasonable threshold is used to demonstrate the instantaneous vortices distribution in the involute channel from both LES and DNS calculation. The DNS simulation captures more detailed turbulence especially near the corner, which explains the discrepancy between LES and DNS results near the corner. The normalized RMS error are defined and calculated to assess the performance of those turbulence models. The RST model captures the anisotropic feature of turbulence, which enable it to outperform other turbulence models for predicting the flow behavior in an involute channel. Although some discrepancies are found between LES and DNS results in the corner, the overall deviations between LES and DNS are found to be small. In conclusion, given that the computational cost of DNS calculation is an order of magnitude higher, using LES data for benchmarking RANS model is a cost-effective approach.
The case of isotropic compressible turbulence subjected to rapid isotropic compression is studied using inviscid rapid distortion theory and direct numerical simulation. An exact solution to the rapid distortion problem is given, and results are compared to those of direct numerical simulation. Implications for modelling turbulent flows are discussed.
Direct numerical simulations of time-evolving turbulent plane wakes developing in the presence of various irrotational plane strains have been generated. A pseudospectral numerical method with up to 25 million modes is used to solve the equations in a reference frame moving with the irrotational strain. The initial condition for each simulation is taken from a previous turbulent self-similar plane wake direct numerical simulation at a velocity deficit Reynolds number, R(sub e), of about 2,000. All the terms in the equations governing the evolution of the Reynolds stresses have been calculated. The relative importance of the various terms is examined for the different strain geometries and the behavior of the individual terms is used to better assess whether the strained wakes are evolving self-similarly.
A weakly nonlinear theory is developed to study the interaction of Tollmien-Schlichting (TS) waves and Dean vortices in curved channel flow. The predictions obtained from the theory agree well with results obtained from direct numerical simulations of curved channel flow, especially for low amplitude disturbances. Some discrepancies in the results of a previous theory with direct numerical simulations are resolved.
Here, this paper presents the process, strategy, and results associated with porting a typical combustion physics flow solver to current state-of-the-art and future massively-parallel computer architectures. Major focus is placed on the distinct algorithmic structure of these types of codes and how it can be integrated with modern programming paradigms for heterogeneous platforms (i.e., distributed many-core systems with accelerators). An end-to-end case study is presented that exemplifies the process in a generic manner, which then serves as a clear guide with respect to the strategy and best practices leading to a robust and adaptable framework that performs well, is durable over time, is portable, and requires minimal human-effort. This end is accomplished beginning with the use of a mature, validated, structured, multiblock code framework optimized for application of both Large Eddy Simulation (LES) and Direct Numerical Simulation (DNS). This code has been ported to a variety of platforms over the past decade, including most recently the Oak Ridge Leadership Computing Facility’s “Summit” Platform. The experience gained on these multiple platforms provides general insights and thus the results presented are not specific to any one code or platform other than the overarching trend toward distributed many-core systems with accelerators in order to move toward exascale performance. The resultant performance and scalability of the ported code is demonstrated on a real-world application; a state-of-the-art rotating detonation rocket engine simulation that matches the complex geometry and boundary conditions imposed as part of a companion experimental campaign.
In recent years, isogeometric analysis (IGA) has attracted significant attention from the computational mechanics community due to its ability to integrate design and analysis. Besides, IGA is also a higher-order discretization technique for solving partial differential equations, showing high approximation capability per degree of freedom. In this paper, we extend the application realm of IGA to particle-laden flows based on Eulerian–Eulerian description that couples Navier–Stokes equations with a density transport equation through a Boussinesq approximation. The coupled systems are solved by using quadratic non-uniform rational B-spline (NURBS) functions and a recently developed residual-based variational multiscale (VMS) formulation, which introduces coupling between the fine velocity scales and density equation residuals. We deploy the proposed approach to perform large-eddy simulations (LES) of dilute particle-laden flows over a flat surface at Reynolds number = 10,000. We compare the simulation results against direct numerical simulation (DNS) results from the literature. We find that combining VMS and IGA, the proposed approach enables accurate prediction of a wide range of flow/particle statistics with a relatively lower mesh resolution.