Full Direct Numerical Simulation Research Articles

Large/small eddy simulations: A high-fidelity method for studying high-Reynolds number turbulent flows

Direct numerical simulations (DNS) are one of the main ab initio tools to study turbulent flows. However, due to their considerable computational cost, DNS are primarily restricted to canonical flows at moderate Reynolds numbers, in which turbulence is isolated from the realistic, large-scale flow dynamics. In contrast, lower fidelity techniques, such as large eddy simulations (LES), are employed for modeling real-life systems. Such approaches rely on closure models that make multiple assumptions, including turbulent equilibrium, small-scale universality, etc., which require prior knowledge of the flow and can be violated. We propose a method, which couples a lower-fidelity, unresolved, time-dependent calculation of an entire system (LES) with an embedded small eddy simulation (SES) that provides a high-fidelity, fully resolved solution in a sub-region of interest of the LES. Such coupling is achieved by continuous replacement of the large SES scales with a low-pass filtered LES velocity field. The method is formulated in physical space, with no assumptions of equilibrium, small-scale structure, and boundary conditions. A priori tests of both steady and unsteady homogeneous, isotropic turbulences are used to demonstrate the method's accuracy in recovering turbulence properties, including spectra, probability density functions of the intermittent quantities, and sub-grid dissipation. Finally, SES is compared with two alternative approaches: one embedding a high-resolution region through static mesh refinement and a generalization of the traditional volumetric spectral forcing. Unlike these methods, SES is shown to achieve DNS-level accuracy at a fraction of the cost of the full DNS, thus opening the possibility to study high-Re flows.

Acceleration of turbulent combustion DNS via principal component transport

We investigate the implementation of principal component (PC) transport to accelerate the direct numerical simulation (DNS) of turbulent combustion flows. The acceleration is achieved using the transport of PCs and the tabulation of the closure terms in the PC-transport equations using machine learning. Further acceleration is achieved by a treatment for bottlenecks associated with the acoustic time steps for low Mach number flows. The approach is implemented in 2D and 3D on a laboratory scale lean premixed methane-air flame stabilized on a slot burner. DNS based on the transport of thermochemical scalars (species and energy) is also carried out, first to develop a 2D DNS database for PC-transport equations’ closure terms and, second, to validate the approach against species DNS in 2D and 3D, a principal goal of the present effort. The results show that surrogate PC DNS can reproduce instantaneous profiles as well as statistics associated with turbulence, flame topology properties and measures of flame-turbulence interactions. The study also demonstrates that parametric simulations with surrogate PC DNS can be implemented at a fraction of the cost of a full 3D DNS with species and energy transport.

Dendrite-resolved, full-melt-pool phase-field simulations to reveal non-steady-state effects and to test an approximate model

We study the epitaxial, columnar growth of (multiply oriented) dendrites/cells for a spot melt in a polycrystalline Al–Cu substrate using two-dimensional, phase-field, direct numerical simulations (DNS) at the full-melt-pool scale. Our main objective is to compare the expensive DNS model to a much cheaper but approximate “line” model in which a single-crystal phase-field simulation is confined to a narrow rectangular geometry. To perform this comparison, we develop algorithms that automatically extract quantities of interest (QoIs) from both DNS and line models. These QoIs allow us to quantitatively assess the assumptions in the line model and help us analyze its discrepancy with the DNS model. We consider four sets of heat source parameters, mimicking welding and additive manufacturing conditions, that create a combination shallow and deep melt pools. Our largest DNS simulation used 16K × 14K grid points in space. Our main findings can be summarized as follows. Under AM conditions, the QoIs of line models are in excellent agreement with the full DNS results for both shallow and deep melt pools. Under welding conditions, the primary spacing of the DNS model is smaller than the prediction of line model. We identify a geometric crowding effect that accounts for the discrepancies between the DNS and line models. We propose two potential mechanisms that determine the response of the microstructure to geometric crowding.

On locally embedded two-scale solution for wall-bounded turbulent flows

Recent findings on wall-bounded turbulence have prompted a new impetus for modelling development to capture and resolve the Reynolds-number-dependent influence of outer flow on near-wall turbulence in terms of the ‘foot-printing’ of the large-scale coherent structures and the scale-interaction associated ‘modulation’. We develop a two-scale method to couple a locally embedded near-wall fine-mesh direct numerical simulation (DNS) block with a global coarser mesh domain. The influence of the large-scale structures on the local fine-mesh block is captured by a scale-dependent coarse–fine domain interface treatment. The coarse-mesh resolved disturbances are directly exchanged across the interface, while only the fine-mesh resolved fluctuations around the coarse-mesh resolved variables are subject to periodic conditions in the streamwise and spanwise directions. The global near-wall coarse-mesh region outside the local fine-mesh block is governed by the augmented flow governing equations with forcing source terms generated by upscaling the space–time-averaged fine-mesh solution. The validity and effectiveness of the method are examined for canonical incompressible channel flows at several Reynolds numbers. The mean statistics and energy spectra are in good agreement with the corresponding full DNS data. The results clearly illustrate the ‘foot-printing’ and ‘modulation’ in the local fine-mesh block. Noteworthy also is that neither spectral-gap nor scale-separation is assumed, and a smooth overlap between the global-domain and the local-domain energy spectra is observed. It is shown that the mesh-count scaling with Reynolds number is potentially reduced from $O(R{e^2})$ for the conventional fully wall-resolved large-eddy simulation (LES) to $O(Re)$ for the present locally embedded two-scale LES.

Investigation of Görtler vortices in high-speed boundary layers via an efficient numerical solution to the non-linear boundary region equations

Streamwise vortices and the associated streaks evolve in boundary layers over flat or concave surfaces due to disturbances initiated upstream or triggered by the wall surface. Following the transient growth phase, the fully-developed vortex structures become susceptible to inviscid secondary instabilities resulting in early transition to turbulence via `bursting' processes. In high-speed boundary layers, more complications arise due to compressibility and thermal effects, which become more significant for higher Mach numbers. In this paper, we study G\"{o}rtler vortices developing in high-speed boundary layers using the boundary region equations (BRE) formalism, which we solve using an efficient numerical algorithm. Streaks are excited using a small transpiration velocity at the wall. Our BRE-based algorithm is found to be superior to direct numerical simulation (DNS) and ad-hoc nonlinear parabolized stability equation (PSE) models. BRE solutions are less computationally costly than a full DNS and have a more rigorous theoretical foundation than PSE-based models. For example, the full development of a G\"{o}rtler vortex system in high-speed boundary layers can be predicted in a matter of minutes using a single processor via the BRE approach. This substantial reduction in calculation time is one of the major achievements of this work. We show, among other things, that it allows investigation into feedback control in reasonable total computational times. We investigate the development of the G\"{o}rtler vortex system via the BRE solution with feedback control parametrically at various freestream Mach numbers $M_\infty$ and spanwise separations $\lambda$ of the inflow disturbances.

Direct numerical simulation of incompressible turbulent boundary layers and planar jets at high Reynolds numbers initialized with implicit large eddy simulation

A direct numerical simulation (DNS) initialized with an implicit large eddy simulation (ILES) is performed for temporally evolving planar jets and turbulent boundary layers. In the ILES, an initial laminar flow develops into a fully developed state of the planar jet or the boundary layer. Subsequently, the DNS is started from the flow field obtained by the ILES. This hybrid ILES/DNS methodology is tested for the planar jet and boundary layer by comparing the results with full DNS started from the initial laminar flow. The ILES results used as the initial conditions of the DNS do not possess small-scale fluctuations. However, the small-scale fluctuations in the DNS grow with time and develop well within an interval of the integral time scale, where the influences of initial conditions taken from the ILES disappear for an energy spectrum of velocity fluctuations. The DNS initialized with the ILES well reproduces small-scale characteristics of turbulence, such as Reynolds number dependence of skewness and flatness of velocity derivative and energy spectrum of velocity fluctuations in the inertial subrange and viscous range. The DNS initialized with the ILES predicts well statistics dominated by large scales, such as 1st- and 2nd-order statistics and longitudinal auto-correlation function, in agreement with previous experimental and numerical studies. Reynolds number dependence of the mean velocity, root-mean-squared velocity fluctuations, Reynolds stress, shape factor, and skin friction in the turbulent boundary layers in the present DNS are consistent with previous experimental studies. These investigations confirm advantages of applying the ILES at the transitional flow region in the DNS of turbulent shear flows at high Reynolds numbers.

A local scattering approach for the effects of abrupt changes on boundary-layer instability and transition: a finite-Reynolds-number formulation for isolated distortions

We investigate the influence of abrupt changes on boundary-layer instability and transition. Such changes can take different forms including a local porous wall, suction/injection and surface roughness as well as junctions between rigid and porous walls. They may modify the boundary conditions and/or the mean flow, and their effects on transition have usually been assessed by performing stability analysis for the modified base flow and/or boundary conditions. However, such a conventional local linear stability theory (LST) becomes invalid if the change occurs over a relatively short scale comparable with, or even shorter than, the characteristic wavelength of the instability. In this case, the influence on transition is through scattering with the abrupt change acting as a local scatter, that is, an instability mode propagating through the region of abrupt change is scattered by the strong streamwise inhomogeneity to acquire a different amplitude. A local scattering approach (LSA) should be formulated instead, in which a transmission coefficient, defined as the ratio of the amplitude of the instability wave after the scatter to that before, is introduced to characterize the effect on instability and transition. In the present study, we present a finite-Reynolds-number formulation of LSA for isolated changes including a rigid plate interspersed by a local porous panel and a wall suction through a narrow slot. When the weak non-parallelism of the unperturbed base flow is ignored, the local scattering problem can be cast as an eigenvalue problem, in which the transmission coefficient appears as the eigenvalue. We also improved the method to take into account the non-parallelism of the unperturbed base flow, where it is found that the weak non-parallelism has a rather minor effect. The general formulation is specialized to two-dimensional Tollmien–Schlichting (T–S) waves. The resulting eigenvalue problem is solved, and full direct numerical simulations (DNS) are performed to verify some of the predictions by LSA. A parametric study indicates that conventional LST is valid only when the change is sufficiently gradual, and becomes either inaccurate or invalid when the scale of the local distortion is short. A local porous panel enhances T–S waves, while a local suction with a moderate mass flux significantly inhibits T–S waves. In the latter case, a comprehensive comparison is made between the theoretical predictions and experimental data, and a satisfactory quantitative agreement was observed.

Toward computationally efficient combustion DNS with complex fuels via principal component transport

We investigate the potential of accelerating chemistry integration during the direct numerical simulation (DNS) of complex fuels based on the transport equations of representative scalars that span the desired composition space using principal component analysis (PCA). The transport of principal components (PCs) can reduce the number of transported scalars and improve the spatial and temporal resolution requirements. The strategy is demonstrated using DNS of a premixed methane–air flame in a 2D vortical flow and is extended to the 3D geometry to demonstrate the resulting enhancement in the computational efficiency of PC transport. The PCs are derived from a priori PCA of the same composition space using DNS. This analysis is used to construct and tabulate the PCs’ chemical source terms in terms of the PCs using artificial neural networks (ANN). Comparison of DNS based on a full thermo-chemical state and DNS based on PC transport with six PCs shows excellent agreement even for terms that are not included in the PCA reduction. The transported PCs reproduce some of the salient features of strongly curved and strongly strained flames. The results also show a significant reduction of two orders of magnitude in the computational cost of the simulations, which enables an extension of the solution approach to 3D DNS under similar computational requirements.

Bubble generated turbulence and direct numerical simulations

Gas–liquid two phase flows are widely encountered in industry. The design parameters include two phase pressure drop, mixing and axial mixing in both the phases, effective interfacial area, heat and mass transfer coefficients. Currently, there is a high degree of empiricism in the design process of such reactors owing to the complexity of coupled flow and reaction mechanism. Hence, we focus on synthesizing recent advances in computational and experimental techniques that will enable future designs of such reactors in a more rational manner by exploring a large design space with high-fidelity models (computational fluid dynamics) that are validated with high-fidelity measurements (hot film anemometry (HFA), Laser Doppler anemometry (LDA), particle image velocimetry (PIV), etc.) to provide a high degree of rigor. Understanding the spatial distributions of dispersed phases and their interaction during scale up are key challenges that were traditionally addressed through pilot scale experiments, but now can be addressed through advanced modelling.For practically complete knowledge of the fluid mechanical parameters, it is desirable to implement direct numerical simulations (DNS). However, the current computational power does not permit full DNS for real bubble columns. Therefore, we have been using simplified turbulence models (such as large eddy simulation, Reynolds stress, k–ε, etc.) which need the knowledge of turbulence parameters. For the estimation of these parameters, currently semi-empirical procedures are being used pending the knowledge of turbulence. Further, the formulation of governing equations in all the CFD models (except DNS), the knowledge of interface forces (drag, lift, virtual mass, Basset, etc.) is needed and for their estimations empirical correlations are being employed, again pending the knowledge of fluid mechanics under turbulent conditions in bubble columns.In gas–liquid dispersions, the gas is sparged in the form of bubbles. During the bubble rise, the mechanism of wake detachment creates turbulence which can be called as wake generated turbulence. In addition, energy gets transferred from the gas phase to liquid phase. The quantitative amounts are negligible when bubble motion is not hindered and the gas–liquid dispersion is homogenous. The amounts increase with an increase in the extent of hindrance. However, in the homogenous regime, even under extreme conditions, the extent of energy transfer in the bulk gas–liquid dispersions (volume other than wake volume) is fairly limited. On contrast, in the heterogeneous regime, the rates of energy transfer become sizeable. The energy received by the liquid (in both the regimes) also creates turbulent motion and termed as bulk generated turbulence. In turbulent flows a compendium of eddies (flow structures) of different length and time scales contribute towards improved/enhanced mixing, momentum transfer, heat transfer, and mass transfer (transport phenomena). Hence, a proper understanding of the dynamics of these turbulent flow structures, and their role in the transport phenomena, can bring substantial improvement in the scale-up and design procedures. The present paper brings out the current status of knowledge on bubble generated turbulence. All the published literature in experimental measurements and DNS simulations has been critically analysed and coherently presented.

A numerical reduced model for thin liquid films sheared by a gas flow

The non-linear dynamics of thin liquid films sheared by a laminar gas flow in a channel is investigated. Such a two-layer flow is driven by pressure gradient and possibly by the gravity force. We describe the liquid phase with a long-wave integral model, with the aim to save computational cost with respect to the full Direct Numerical Simulation (DNS) of the Navier–Stokes equations. We derive this long-wave model by the integration of the Navier–Stokes equations over the film thickness, and by an asymptotic expansion up to the first order in terms of a long-wave parameter. These depth-integrated (or shallow water) equations are discretized by means of an augmented system, which holds an evolution equation for the surface tension in order to avoid numerical instabilities of classical upwind and centered schemes. On the other side, we study the gas phase with compressible Navier–Stokes equations, and we discretize them by means of a low-Mach scheme, accounting also for moving meshes (ALE). In order to analyze liquid–gas interactions, we introduce then a coupling methodology between depth-integrated equations and Navier–Stokes equations. This approach represents a compromise between the two existing methods: the full DNS, and the full long-wave model applied to both phases. In order to validate this approach, we present comparisons with DNS, showing a good agreement of spatio-temporal evolutions of the film thickness and the stress field. Furthermore, interfacial shear stress and pressure gradient evolutions are shown to be in accordance with those provided by two-layer second-order low-dimensional models.

Drag, lift and torque coefficients for ellipsoidal particles: From low to moderate particle Reynolds numbers

An accurate prediction of the translational and rotational motion of ellipsoidal particles can be only given if a complete set of correlations of the drag, lift and pitching torque coefficients is known. The present study is thus devoted to the assessment of the available correlations in the literature through a comparison with numerical results of the forces acting on a particle given by a full body-fitted direct numerical simulation (DNS) in the case of a uniform flow, for three different ellipsoidal particles, and for Reynolds number ranging from 0.1 to 290. The comparison between the computed force (or hydrodynamic coefficients) and the literature correlations shows clearly that certain precautions must be taken. The mean deviation of the most accurate correlation considered in the present study from our full DNS results can be of the order of 20% for the drag coefficient. For the lift and pitching torque coefficients, the deviations can increase up to roughly 25%. This comparison shows that further works are definitively necessary to develop a complete set of correlations for ellipsoidal particles outside Stokes regime.

Numerical simulations of separated flows at moderate Reynolds numbers appropriate for turbine blades and unmanned aero vehicles

Flows over airfoils and blades in rotating machinery, for unmanned and micro-aerial vehicles, wind turbines, and propellers consist of a laminar boundary layer near the leading edge that is often followed by a laminar separation bubble and transition to turbulence further downstream. Typical RANS turbulence models are inadequate for such flows. Direct numerical simulation (DNS) is the most reliable, but is also the most computationally expensive alternative. This work assesses the capability of Immersed Boundary (IB) methods and Large Eddy Simulations (LES) to reduce the computational requirements for such flows and still provide high quality results. Two-dimensional and three-dimensional simulations of a laminar separation bubble on a NACA-0012 airfoil at Rec=5×104 at 5° of incidence have been performed with an IB code and a commercial code using body fitted grids. Several Subgrid Scale (SGS) models have been implemented in both codes and their performance evaluated. For the two-dimensional simulations with the IB method the results show good agreement with DNS benchmark data for the pressure coefficient Cp and the friction coefficient Cf but only when using dissipative numerical schemes. There is evidence that this behavior can be attributed to the ability of dissipative schemes to damp numerical noise coming from the IB. For the three-dimensional simulations the results show a good prediction of the separation point, but inaccurate prediction of the reattachment point unless full DNS resolution is used. The commercial code shows good agreement with the DNS benchmark data in both two and three-dimensional simulations, but the presence of significant, unquantified numerical dissipation prevents a conclusive assessment of the actual prediction capabilities of very coarse LES with low order schemes in general case.

Predictive model for heat transfer performance of oblique and normally impinging jet arrays

A model is created to predict the heat transfer performance of oblique and normal jet impingement arrays of arbitrary extent. The simplicity of the model, being based on just two equations, allows for ease of implementation and relatively few computations. Comparing the performance of the predictive model with experiment yielded a mean absolute error of 2.9%. Comparison with two frequently referenced correlations in literature show mean absolute errors of 14% and 18% for considerably different test conditions, and comparisons with raw data from these studies have yielded even lower error. The proposed numerical model has a higher accuracy than current CFD implementations and is less computationally involved than full direct numerical simulation.

Development of Chemistry Coordinate Mapping Approach for Turbulent Partially Premixed Combustion

Direct numerical simulation (DNS) coupling with chemistry coordinate mapping (CCM) is presented to simulate flame propagation and auto-ignition in a partially premixed syngas/air mixture. In the CCM approach, the physical domain is mapped into a low dimensional phase space with a few thermodynamic variables as the independent variables. The integration of the chemical reaction rates and heat release rate are done on the grid in the phase space. Previously we showed that for premixed mixtures, two variables temperature and specific element mass fraction of H atom, can be sufficient to construct the phase space for a satisfactory mapping. However, for partially premixed combustion mode, a third phase space variable is required to map the physical cell into the phase space. It is shown that scalar dissipation rate of the element mass ratio of H atom can be used as the third dimension of the phase space. An investigation is carried out on the behavior of CCM and the choice over the element on which the local element mass ratio should be based. Mapping error in the CCM is investigated. It is shown that if the element mass ratio is based on the element involved in the most diffusive molecules, the error of the mapping can approach zero when the grid in the phase space is refined. To validate the CCM approach the results of DNS coupled with CCM (DNS-CCM) are compared with full DNS that integrates the chemical reaction rates and heat release rate directly in physical space. Good agreement between the results from DNS and DNS-CCM is obtained while the computational time is reduced at least by 70 %.

Toward application of conformal decomposition finite elements to non‐colloidal particle suspensions

SUMMARYParticle suspensions play an important role in many engineering applications, yet their behavior in a number of respects remains poorly understood. In conjunction with careful experiments, modeling and simulation of these systems can provide key insight into their complex behavior. However, these two‐phase systems pose the challenge of simultaneously, accurately, and efficiently capturing the complex geometric structure, kinematics, and dynamics of the particulate discrete phase and the discontinuities it introduces into the variables (e.g., velocity, pressure, density) of the continuous phase. To this end, a new conformal decomposition finite element method (CDFEM) is introduced for solid particles in a viscous fluid. The method is verified in several simple test problems that are representative of aspects of particle suspension behavior. In all cases, we find the CDFEM to perform accurately and efficiently leading to the conclusion that it forms a prime candidate for application to the full direct numerical simulation of particle suspensions. Copyright © 2012 John Wiley & Sons, Ltd.

A multi-zone chemistry mapping approach for direct numerical simulation of auto-ignition and flame propagation in a constant volume enclosure

A direct numerical simulation (DNS) coupling with multi-zone chemistry mapping (MZCM) is presented to simulate flame propagation and auto-ignition in premixed fuel/air mixtures. In the MZCM approach, the physical domain is mapped into a low-dimensional phase domain with a few thermodynamic variables as the independent variables. The approach is based on the fractional step method, in which the flow and transport are solved in the flow time steps whereas the integration of the chemical reaction rates and heat release rate is performed in much finer time steps to accommodate the small time scales in the chemical reactions. It is shown that for premixed mixtures, two independent variables can be sufficient to construct the phase space to achieve a satisfactory mapping. The two variables can be the temperature of the mixture and the specific element mass ratio of H atom for fuels containing hydrogen atoms. An aliasing error in the MZCM is investigated. It is shown that if the element mass ratio is based on the element involved in the most diffusive molecules, the aliasing error of the model can approach zero when the grid in the phase space is refined. The results of DNS coupled with MZCM (DNS-MZCM) are compared with full DNS that integrates the chemical reaction rates and heat release rate directly in physical space. Application of the MZCM to different mixtures of fuel and air is presented to demonstrate the performance of the method for combustion processes with different complexity in the chemical kinetics, transport and flame–turbulence interaction. Good agreement between the results from DNS and DNS-MZCM is obtained for different fuel/air mixtures, including H2/air, CO/H2/air and methane/air, while the computational time is reduced by nearly 70%. It is shown that the MZCM model can properly address important phenomena such as differential diffusion, local extinction and re-ignition in premixed combustion.

Optimally amplified large-scale streaks and drag reduction in turbulent pipe flow

The optimal amplifications of small coherent perturbations within turbulent pipe flow are computed for Reynolds numbers up to one million. Three standard frameworks are considered: the optimal growth of an initial condition, the response to harmonic forcing and the Karhunen-Loève (proper orthogonal decomposition) analysis of the response to stochastic forcing. Similar to analyses of the turbulent plane channel flow and boundary layer, it is found that streaks elongated in the streamwise direction can be greatly amplified from quasistreamwise vortices, despite linear stability of the mean flow profile. The most responsive perturbations are streamwise uniform and, for sufficiently large Reynolds number, the most responsive azimuthal mode is of wave number m=1 . The response of this mode increases with the Reynolds number. A secondary peak, where m corresponds to azimuthal wavelengths λ_{θ}^{+}≈70-90 in wall units, also exists in the amplification of initial conditions and in premultiplied response curves for the forced problems. Direct numerical simulations at Re=5300 confirm that the forcing of m=1,2 and m=4 optimal structures results in the large response of coherent large-scale streaks. For moderate amplitudes of the forcing, low-speed streaks become narrower and more energetic, whereas high-speed streaks become more spread. It is further shown that drag reduction can be achieved by forcing steady large-scale structures, as anticipated from earlier investigations. Here the energy balance is calculated. At Re=5300 it is shown that, due to the small power required by the forcing of optimal structures, a net power saving of the order of 10% can be achieved following this approach, which could be relevant for practical applications.

Velocity and conformation statistics based on reduced Karhunen–Loeve projection data from DNS of viscoelastic turbulent channel flow

To investigate the effectiveness of the Karhunen–Loeve (K–L) method as a data reduction approach, we study here its effect on the velocity and conformation statistics in a drag reducing turbulent polymer flow. The K–L method has been used to construct a set of basis velocity eigenfunctions from a large number of independent realizations of the velocity. Those were obtained from direct numerical simulation (DNS) of a viscoelastic turbulent channel flow using the Giesekus model. A subset of the K–L eigenfunctions, large enough to contain more than 90% of the fluctuating kinetic energy of the flow on the average, has then been subsequently used to obtain time series of projection coefficients of the velocity fields generated further from DNS. In a post-processing step, velocity fields were reconstructed using selected subsets of the projection coefficients. Those reconstructed velocity fields were then used to evaluate turbulent statistics as well as to integrate the constitutive equation. The turbulent statistics (r.m.s. velocities, Reynolds stress etc.) thus constructed showed good agreement with the full results from DNS. The Reynolds stress anisotropy was also calculated in this work for the first time. It was found to increase with viscoelasticity that was well reproduced in the reduced K–L data except near the channel centerline where the K–L data showed some loss of anisotropy. The biggest differences however between the K–L reduced data and the full DNS results were seen in the conformation statistics. The average polymer conformation extracted from the K–L reduced data was significantly less than that corresponding to the full DNS results anywhere except in the shear-dominated wall region. A further comparison of the energy and dissipation spectra between the full DNS and the K–L reconstructed data illustrated the impact of the K–L process in resulting to a significant damping of small turbulent scales even those contributing to the maximum in turbulent dissipation. This may also be the principal reason behind the poor quality of the K–L reconstructed conformation data.

Direct three-dimensional numerical simulation of nucleate boiling using the level contour reconstruction method

The nucleate boiling process is known to be a very efficient mode of heat transfer. It is desirable to operate many engineering applications in this mode since high heat transfer rates and convection coefficients are associated with small values of the excess temperature. Despite its importance, nucleate boiling has not been fully numerically simulated until very recently because of the complexity of dealing with the phase change problem in addition to the interface method itself. Here we perform a full direct numerical simulation of nucleate boiling in a fully three-dimensional geometry using the Level Contour Reconstruction Method (Shin and Juric, 2002). Our work is aimed at predicting nucleate boiling heat flux values more accurately on a real surface by including the effect of nucleation site density in the numerical model. This has been achieved by changing the surface area for a single nucleate bubble corresponding to the wall temperature. A large body of research in nucleate boiling deals with developing experimental correlations for heat transfer during the nucleate boiling process. Here we briefly review some of the heat transfer correlations developed for wall heat transfer in nucleate boiling. Tien (1962) obtained a

Equation-free/Galerkin-free POD-assisted computation of incompressible flows

We present a Galerkin-free, proper orthogonal decomposition (POD)-assisted computational methodology for numerical simulations of the long-term dynamics of the incompressible Navier–Stokes equations. The approach is based on the “equation-free” framework: we use short, appropriate initialized bursts of full direct numerical simulations (DNS) of the Navier–Stokes equations to observe, estimate, and accelerate, through “projective integration”, the evolution of the flow dynamics. The main assumption is that the long-term dynamics of the flow lie on a low-dimensional, attracting, and invariant manifold, which can be parametrized, not necessarily spanned, by a few POD basis functions. We start with a discussion of the consistency and accuracy of the approach, and then illustrate it through numerical examples: two-dimensional periodic and quasi-periodic flows past a circular cylinder. We demonstrate that the approach can successfully resolve complex flow dynamics at a reduced computational cost and that it can capture the long-term asymptotic state of the flow in cases where traditional Galerkin-POD models fail. The approach trades the overhead involved in developing POD-Galerkin and POD-nonlinear Galerkin codes, for the repeated (yet short, and on demand) use of an existing full DNS simulator. Moreover, since in this approach the POD modes are used to observe rather than span the true system dynamics, the computation is much less sensitive than POD-Galerkin to values of the system parameters (e.g., the Reynolds number) and the particular simulation data ensemble used to obtain the POD basis functions.