High-resolution Direct Numerical Simulations Research Articles

Reynolds-number dependence of the dimensionless dissipation rate in homogeneous magnetohydrodynamic turbulence.

This paper examines the behavior of the dimensionless dissipation rate C_{ɛ} for stationary and nonstationary magnetohydrodynamic (MHD) turbulence in the presence of external forces. By combining with previous studies for freely decaying MHD turbulence, we obtain here both the most general model equation for C_{ɛ} applicable to homogeneous MHD turbulence and a comprehensive numerical study of the Reynolds number dependence of the dimensionless total energy dissipation rate at unity magnetic Prandtl number. We carry out a series of medium to high resolution direct numerical simulations of mechanically forced stationary MHD turbulence in order to verify the predictions of the model equation for the stationary case. Furthermore, questions of nonuniversality are discussed in terms of the effect of external forces as well as the level of cross- and magnetic helicity. The measured values of the asymptote C_{ɛ,∞} lie between 0.193≤C_{ɛ,∞}≤0.268 for free decay, where the value depends on the initial level of cross- and magnetic helicities. In the stationary case we measure C_{ɛ,∞}=0.223.

Energy spectrum in high-resolution direct numerical simulations of turbulence

Direct numerical simulations with 12,288${}^{3}$ grid points up to ${R}_{\ensuremath{\lambda}}=2300$ finds small systematic differences from Kolmogorov's ${k}^{-5/3}$ scaling of the inertial range of the energy spectrum.

Preferential Rotation of Chiral Dipoles in Isotropic Turbulence.

We introduce a new particle shape which shows preferential rotation in three dimensional homogeneous isotropic turbulence. We call these particles chiral dipoles because they consist of a rod with two helices of opposite handedness, one at each end. 3D printing is used to fabricate these particles with a length in the inertial range and their rotations are tracked in a turbulent flow between oscillating grids. High aspect ratio chiral dipoles preferentially align with their long axis along the extensional eigenvectors of the strain rate tensor, and the helical ends respond to the extensional strain rate with a mean spinning rate that is nonzero. We use Stokesian dynamics simulations of chiral dipoles in pure strain flow to quantify the dependence of spinning on particle shape. Based on the known response to pure strain, we build a model that gives the spinning rate of small chiral dipoles using velocity gradients along Lagrangian trajectories from high resolution direct numerical simulations. The statistics of chiral dipole spinning determined with this model show surprisingly good agreement with the measured spinning of much larger chiral dipoles in the experiments.

Effect of roll number on the statistics of turbulent Taylor-Couette flow

High-resolution direct numerical simulations of turbulent Taylor-Couette flow in large periodic domains reveal that the large-scale rollers that are formed are very nearly fixed in space and their axial wavelength may sensitively depend upon the initial conditions.

Numerical simulations of shoaling internal solitary waves of elevation

We present high-resolution, two- and three-dimensional direct numerical simulations of large amplitude internal solitary waves of elevation on the laboratory scale, shoaling onto and over a small-amplitude shelf. The three-dimensional, mapped coordinate, spectral collocation method used for the simulations allows for accurate modelling of both the shoaling waves and the bottom boundary layer. The shoaling of the waves is characterized by the formation of a quasi-trapped core which undergoes a spatially growing stratified shear instability at its edge and a lobe-cleft instability in its nose. Both of these instabilities develop and three-dimensionalize concurrently, leading to strong bottom shear stress. We explore significant regions of Schmidt and Reynolds number space and demonstrate that the formation of shear instabilities during shoaling is robust and should be readily observable in a number of standard laboratory setups. In the experiments with a corrugated bottom boundary, boundary layer separation is found inside each of the corrugations during shoaling. This more complex boundary layer phenomenology precludes the formation of the lobe-cleft instability almost completely and hence provides a different mechanism for fluid and material exchange across the bottom boundary layer. Our analyses suggest that all of these wave-induced instabilities can lead to enhanced turbulence in the water column and increased shear stress on the bottom boundary. Through the generation and evolution of these instabilities, the shoaling of internal solitary waves of elevation is likely to provide systematic mechanisms for material mixing, cross-boundary layer transport, and sediment resuspension.

Dynamics and structure of unstably stratified homogeneous turbulence

ABSTRACTWe characterise the properties of unstably stratified homogeneous turbulence by means of high-resolution direct numerical simulations and a two-point statistical spectral model based on a quasi-normal closure proposed by Burlot et al. Both approaches agree very well regarding the evolution of one- and two-point turbulent statistics, showing that the model is valid at even higher Reynolds numbers than previously considered. From a parametric study with different initial conditions, we confirm that the energy distribution at large scale influences strongly the late time dynamics of the flow. In particular, we assess the existence of backscatter transfer of energy, and evaluate its role in the growth rate of several turbulent quantities. Moreover, thanks to the statistical model, we analyse the scale-by-scale anisotropy of the flow through the decomposition of turbulent spectra in terms of directional anisotropy and polarisation anisotropy, for a refined characterisation of the structure of the flow which is strongly anisotropic in the large scales. This also allows us to study how isotropy is restored in the inertial scales.

The influence of surfactants on thermocapillary flow instabilities in low Prandtl melting pools

Flows in low Prandtl number liquid pools are relevant for various technical applications and have so far only been investigated for the case of pure fluids, i.e., with a constant, negative surface tension temperature coefficient ∂γ/∂T. Real-world fluids containing surfactants have a temperature dependent ∂γ/∂T > 0, which may change sign to ∂γ/∂T < 0 at a critical temperature Tc. Where thermocapillary forces are the main driving force, this can have a tremendous effect on the resulting flow patterns and the associated heat transfer. Here we investigate the stability of such flows for five Marangoni numbers in the range of 2.1 × 106 ≤ Ma ≤ 3.4 × 107 using dynamic large eddy simulations, which we validate against a high resolution direct numerical simulation. We find that the five cases span all flow regimes, i.e., stable laminar flow at Ma ≤ 2.1 × 106, transitional flow with rotational instabilities at Ma = 2.8 × 106 and Ma = 4.6 × 106, and turbulent flow at Ma = 1.8 × 107 and Ma = 3.4 × 107.

A unified large/small-scale dynamo in helical turbulence

We use high resolution direct numerical simulations (DNS) to show that helical turbulence can generate significant large-scale fields even in the presence of strong small-scale dynamo action. During the kinematic stage, the unified large/small-scale dynamo grows fields with a shape-invariant eigenfunction, with most power peaked at small scales or large $k$, as in \citet{SB14}. Nevertheless, the large-scale field can be clearly detected as an excess power at small $k$ in the negatively polarized component of the energy spectrum for a forcing with positively polarized waves. Its strength $bar{B}$, relative to the total rms field $B_{rms}$, decreases with increasing magnetic Reynolds number, $R_m$. However, as the Lorentz force becomes important, the field generated by the unified dynamo orders itself by saturating on successively larger scales. The magnetic integral scale for the positively polarized waves, characterizing the small-scale field, increases significantly from the kinematic stage to saturation. This implies that the small-scale field becomes as coherent as possible for a given forcing scale, which averts the $R_m$-dependent quenching of $\bar{B}/B_rms$. These results are obtained for $1024^3$ DNS with magnetic Prandtl numbers of $P_m=0.1$ and $10$. For $P_m=0.1$, $\bar{B}/B_{rms}$ grows from about $0.04$ to about $0.4$ at saturation, aided in the final stages by helicity dissipation. For $P_m=10$, $\bar{B}/B_{rms}$ grows from much less than 0.01 to values of the order the $0.2$. Our results confirm that there is a unified large/small-scale dynamo in helical turbulence.

Settling velocity of small inertial particles in homogeneous isotropic turbulence from high-resolution DNS

The gravitational settling velocity of small heavy particles in a three-dimensional turbulent flow remains a controversial topic. In a homogeneous turbulence of zero mean velocity, both enhanced settling velocity and reduced settling velocity have been reported relative to the still-fluid terminal velocity. Dominant mechanisms for enhanced settling include the preferential sweeping and particle-particle hydrodynamic interactions. The reduced settling could result from loitering (falling particles spend more time in the regions with upward flow), vortex trapping, and drag nonlinearity. Here high-resolution direct numerical simulations (DNS) are used to investigate the settling velocity of non-interacting small heavy particles, for an extended range of flow Taylor microscale Reynolds numbers (up to Rλ=500) with varying particle terminal velocity (relative to the Kolmogorov velocity) and particle inertia, by changing the particle-to-fluid density ratio and energy dissipation rate. For the parameter regimes considered here, the preferential sweeping has a dominant effect leading to an increase of the average settling velocity relative to the terminal velocity; and this increase is mainly governed by particle Froude number (the ratio between the particle inertial response time and the residence time of the particle in a Kolmogorov eddy) and its magnitude depends linearly on the square root of the energy dissipation rate. The reduction of settling due to loitering rarely occurs in a homogeneous turbulence without organized large-scale vortical structures, but is found to emerge only if the particle horizontal motions are blocked (thus removing the preferential sweeping effect), as shown in Good et al. (2014). The DNS results were used to develop a parameterization that relates the settling velocity to the particle inertia (St), Froude number, and Rλ. Finally, sensitivities of the DNS results to the large-scale forcing method and to the drag nonlinearity are also briefly discussed.

On the Dynamical Regimes of Pattern-Accelerated Electroconvection.

Recent research has established that electroconvection can enhance ion transport at polarized surfaces such as membranes and electrodes where it would otherwise be limited by diffusion. The onset of such overlimiting transport can be influenced by the surface topology of the ion selective membranes as well as inhomogeneities in their electrochemical properties. However, there is little knowledge regarding the mechanisms through which these surface variations promote transport. We use high-resolution direct numerical simulations to develop a comprehensive analysis of electroconvective flows generated by geometric patterns of impermeable stripes and investigate their potential to regularize electrokinetic instabilities. Counterintuitively, we find that reducing the permeable area of an ion exchange membrane, with appropriate patterning, increases the overall ion transport rate by up to 80%. In addition, we present analysis of nonpatterned membranes, and find a novel regime of electroconvection where a multivalued current is possible due to the coexistence of multiple convective states.

A third-order multistep time discretization for a Chebyshev tau spectral method

A time discretization scheme based on the third-order backward difference formula has been embedded into a Chebyshev tau spectral method for the Navier-Stokes equations. The time discretization is a variant of the second-order backward scheme proposed by Krasnov et al. (2008) 3. High-resolution direct numerical simulations of turbulent incompressible channel flow have been performed to compare the backward scheme to the Runge-Kutta scheme proposed by Spalart et al. (1991) 2. It is shown that the Runge-Kutta scheme leads to a poor convergence of some third-order spatial derivatives in the direct vicinity of the wall, derivatives that represent the diffusion of wall-tangential vorticity. The convergence at the wall is shown to be significantly improved if the backward scheme is applied.

Stably stratified turbulence in the presence of large-scale forcing.

We perform two high-resolution direct numerical simulations of stratified turbulence for Reynolds number equal to Re≈25000 and Froude number, respectively, of Fr≈0.1 and Fr≈0.03. The flows are forced at large scale and discretized on an isotropic grid of 2048(3) points. Stratification makes the flow anisotropic and introduces two extra characteristic scales with respect to homogeneous isotropic turbulence: the buoyancy scale, L(B), and the Ozmidov scale, ℓ(oz). The former is related to the number of layers that the flow develops in the direction of gravity, and the latter is regarded as the scale at which isotropy is recovered. The values of L(B) and ℓ(oz) depend on the Froude number, and their absolute and relative amplitudes affect the repartition of energy among Fourier modes in ways that are not easy to predict. By contrasting the behavior of the two simulated flows we identify some surprising similarities: After an initial transient the two flows evolve towards comparable values of the kinetic and potential enstrophy and energy dissipation rate. This is the result of the Reynolds number being large enough in both flows for the Ozmidov scale to be resolved. When properly dimensionalized, the energy dissipation rate is compatible with atmospheric observations. Further similarities emerge at large scales: The same ratio between potential and total energy (≈0.1) is spontaneously selected by the flows, and slow modes grow monotonically in both regimes, causing a slow increase of the total energy in time. The axisymmetric total energy spectrum shows a wide variety of spectral slopes as a function of the angle between the imposed stratification and the wave vector. One-dimensional energy spectra computed in the direction parallel to gravity are flat from the forcing up to buoyancy scale. At intermediate scales a ∼k(-3) parallel spectrum develops for the Fr≈0.03 run, whereas for weaker stratification, the saturation spectrum does not have enough scales to develop and instead one observes a power law compatible with Kolmogorov scaling. Finally, the spectrum of helicity is flat until L(B), as observed in the nocturnal planetary boundary layer.

Two-dimensional homogeneous isotropic fluid turbulence with polymer additives.

We carry out an extensive and high-resolution direct numerical simulation of homogeneous, isotropic turbulence in two-dimensional fluid films with air-drag-induced friction and with polymer additives. Our study reveals that the polymers (a) reduce the total fluid energy, enstrophy, and palinstrophy; (b) modify the fluid energy spectrum in both inverse- and forward-cascade régimes; (c) reduce small-scale intermittency; (d) suppress regions of high vorticity and strain rate; and (e) stretch in strain-dominated regions. We compare our results with earlier experimental studies and propose new experiments.

Resolving the paradox of oceanic large-scale balance and small-scale mixing.

A puzzle of oceanic dynamics is the contrast between the observed geostrophic balance, involving gravity, pressure gradient, and Coriolis forces, and the necessary turbulent transport: in the former case, energy flows to large scales, leading to spectral condensation, whereas in the latter, it is transferred to small scales, where dissipation prevails. The known bidirectional constant-flux energy cascade maintaining both geostrophic balance and mixing tends towards flux equilibration as turbulence strengthens, contradicting models and recent observations which find a dominant large-scale flux. Analyzing a large ensemble of high-resolution direct numerical simulations of the Boussinesq equations in the presence of rotation and no salinity, we show that the ratio of the dual energy flux to large and to small scales agrees with observations, and we predict that it scales with the inverse of the Froude and Rossby numbers when stratification is (realistically) stronger than rotation. Furthermore, we show that the kinetic and potential energies separately undergo a bidirectional transfer to larger and smaller scales. Altogether, this allows for small-scale mixing which drives the global oceanic circulation and will thus potentially lead to more accurate modeling of climate dynamics.

Nonhelical inverse transfer of a decaying turbulent magnetic field.

In the presence of magnetic helicity, inverse transfer from small to large scales is well known in magnetohydrodynamic (MHD) turbulence and has applications in astrophysics, cosmology, and fusion plasmas. Using high resolution direct numerical simulations of magnetically dominated self-similarly decaying MHD turbulence, we report a similar inverse transfer even in the absence of magnetic helicity. We compute for the first time spectral energy transfer rates to show that this inverse transfer is about half as strong as with helicity, but in both cases the magnetic gain at large scales results from velocity at similar scales interacting with smaller-scale magnetic fields. This suggests that both inverse transfers are a consequence of universal mechanisms for magnetically dominated turbulence. Possible explanations include inverse cascading of the mean squared vector potential associated with local near two dimensionality and the shallower k^{2} subinertial range spectrum of kinetic energy forcing the magnetic field with a k^{4} subinertial range to attain larger-scale coherence. The inertial range shows a clear k^{-2} spectrum and is the first example of fully isotropic magnetically dominated MHD turbulence exhibiting weak turbulence scaling.

Generalized Batchelor functions of isotropic turbulence

We generalize Batchelor’s parameterization of the autocorrelation functions of isotropic turbulence in a form involving a product expansion with multiple small scales. The richer small scale structure acquired this way, compared to the usual Batchelor function, is necessary so that the associated energy spectrum approximate well actual spectra in the universal equilibrium range. We propose that the generalized function provides an approximation of arbitrary accuracy for actual spectra of isotropic turbulence over the universal equilibrium range. The degree of accuracy depends on the number of higher moments which are determinable and it is reflected in the number of small scales involved. The energy spectrum of the generalized function is derived, and for the case of two small scales is compared with data from high-resolution direct numerical simulations. We show that the compensated spectra (which illustrate the bottleneck effect) and dissipation spectra are encapsulated excellently, in accordance with our proposal.

Magnetic helicity and the evolution of decaying magnetohydrodynamic turbulence.

Ensemble-averaged high resolution direct numerical simulations of reverse spectral transfer are presented, extending on the many single realization numerical studies done up to now. This identifies this type of spectral transfer as a statistical property of magnetohydrodynamic turbulence and thus permits reliable numerical exploration of its dynamics. The magnetic energy decay exponent from these ensemble runs has been determined to be nE=(0.47±0.03)+(13.9±0.8)/Rλ for initially helical magnetic fields. We show that even after removing the Lorentz force term in the momentum equation, thus decoupling it from the induction equation, reverse spectral transfer still persists. The induction equation is now linear with an externally imposed velocity field, thus amenable to numerous analysis techniques. A new door has opened for analyzing reverse spectral transfer, with various ideas discussed.

The statistics of a passive scalar in field-guided magnetohydrodynamic turbulence

A variety of studies of magnetised plasma turbulence invoke theories for the advection of a passive scalar by turbulent fluctuations. Examples include modelling the electron density fluctuations in the interstellar medium, understanding the chemical composition of galaxy clusters and the intergalactic medium, and testing the prevailing phenomenological theories of magnetohydrodynamic turbulence. While passive scalar turbulence has been extensively studied in the hydrodynamic case, its counterpart in MHD turbulence is significantly less well understood. Herein we conduct a series of high-resolution direct numerical simulations of incompressible, field-guided, MHD turbulence in order to establish the fundamental properties of passive scalar evolution. We study the scalar anisotropy, establish the scaling relation analogous to Yaglom’s law, and measure the intermittency of the passive scalar statistics. We also assess to what extent the pseudo Alfvén fluctuations in strong MHD turbulence can be modelled as a passive scalar. The results suggest that the dynamics of a passive scalar in MHD turbulence is considerably more complicated than in the hydrodynamic case.

Direct numerical simulations of particle-laden density currents with adaptive, discontinuous finite elements

Abstract. High-resolution direct numerical simulations (DNSs) are an important tool for the detailed analysis of turbidity current dynamics. Models that resolve the vertical structure and turbulence of the flow are typically based upon the Navier–Stokes equations. Two-dimensional simulations are known to produce unrealistic cohesive vortices that are not representative of the real three-dimensional physics. The effect of this phenomena is particularly apparent in the later stages of flow propagation. The ideal solution to this problem is to run the simulation in three dimensions but this is computationally expensive. This paper presents a novel finite-element (FE) DNS turbidity current model that has been built within Fluidity, an open source, general purpose, computational fluid dynamics code. The model is validated through re-creation of a lock release density current at a Grashof number of 5 × 106 in two and three dimensions. Validation of the model considers the flow energy budget, sedimentation rate, head speed, wall normal velocity profiles and the final deposit. Conservation of energy in particular is found to be a good metric for measuring model performance in capturing the range of dynamics on a range of meshes. FE models scale well over many thousands of processors and do not impose restrictions on domain shape, but they are computationally expensive. The use of adaptive mesh optimisation is shown to reduce the required element count by approximately two orders of magnitude in comparison with fixed, uniform mesh simulations. This leads to a substantial reduction in computational cost. The computational savings and flexibility afforded by adaptivity along with the flexibility of FE methods make this model well suited to simulating turbidity currents in complex domains.

Energy spectrum of buoyancy-driven turbulence.

Using high-resolution direct numerical simulation and arguments based on the kinetic energy flux Π(u), we demonstrate that, for stably stratified flows, the kinetic energy spectrum E(u)(k)∼k(-11/5), the potential energy spectrum E(θ)(k)∼k(-7/5), and Π(u)(k)∼k(-4/5) are consistent with the Bolgiano-Obukhov scaling. This scaling arises due to the conversion of kinetic energy to the potential energy by buoyancy. For weaker buoyancy, this conversion is weak, hence E(u)(k) follows Kolmogorov's spectrum with a constant energy flux. For Rayleigh-Bénard convection, we show that the energy supply rate by buoyancy is positive, which leads to an increasing Π(u)(k) with k, thus ruling out Bolgiano-Obukhov scaling for the convective turbulence. Our numerical results show that convective turbulence for unit Prandt number exhibits a constant Π(u)(k) and E(u)(k)∼k(-5/3) for a narrow band of wave numbers.