Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Direct numerical simulation”

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.

At least 1,333 records · Page 74

Wave-current interaction study in the Gulf of Alaska for detection of eddies by synthetic aperture radar

High resolution Esa Remote Sensing Satellite-1 (ERS-1) Synthetic Aperture Radar (SAR) images are used to detect a mesoscale eddy. Such features limit dispersal of pollock larvae and therefore likely influence recruitment of fish in the Gulf of Alaska. During high sea states and high winds, the direct surface signature of the eddy was not clearly visible, but the wave refraction in the eddy area was observed. The rays of the wave field are traced out directly from the SAR image. The ray pattern gives information on the refraction pattern and on the relative variation of the wave energy along a ray through wave current interaction. These observations are simulated by a ray-tracing model which incorporates a surface current field associated with the eddy. The numerical results of the model show that the waves are refracted and diverge in the eddy field with energy density decreasing. The model-data comparison for each ray shows the model predictions are in good agreement with the SAR data.

Liu, Antony K.↗

Wake Vortex Transport in Proximity to the Ground

A sensitivity study for aircraft wake vortex transport has been conducted using a validated large eddy simulation (LES) model. The study assumes neutrally stratified and nonturbulent environments and includes the consequences of the ground. The numerical results show that the nondimensional lateral transport is primarily influenced by the magnitude of the ambient crosswind and is insensitive to aircraft type. In most of the simulations, the ground effect extends the lateral position of the downwind vortex about one initial vortex spacing (b(sub o)) in the downstream direction. Further extension by as much as one b(sub o) occurs when the downwind vortex remains 'in ground effect' (IGE) for relatively long periods of time. Results also show that a layer-averaged ambient wind velocity can be used to bound the time for lateral transport of wake vortices to insure safe operations on a parallel runway.

Hamilton, David W.↗

Multi-Scale Modeling of Magnetospheric Reconnection

One of the major challenges in modeling the magnetospheric magnetic reconnection is to quantify the interaction between large-scale global magnetospheric dynamics and microphysical processes in diffusion regions near reconnection sites. There is still considerable debate as to what degree microphysical processes on kinetic scales affect the global evolution and how important it is to substitute numerical dissipation and/or ad hoc anomalous resistivity by a physically motivated model of dissipation. Comparative studies of magnetic reconnection in small scale geometries demonstrated that MHD simulations that included non-ideal processes in terms of a resistive term $\eta J$ did not produce the fast reconnection rates observed in kinetic simulations. For a broad range of physical parameters in collisionless magnetospheric plasma, the primary mechanism controlling the dissipation in the vicinity of the reconnection site is non-gyrotropic effects with spatial scales comparable with the particle Larmor radius. We utilize the global MHD code BATSRUS and incorporate nongyrotropic effects in diffusion regions in terms of corrections to the induction equation. We developed an algorithm to search for magnetotail reconnection sites, specifically where the magnetic field components perpendicular to the local current direction approaches zero and form an X-type configuration. Spatial scales of the diffusion region and magnitude of the reconnection electric field are calculated selfconsistently using MHD plasma and field parameters in the vicinity of the reconnection site. The location of the reconnection sites is updated during the simulations. To clarify the role of nongyrotropic effects in diffusion region on the global magnetospheric dynamic we perform simulations with steady southward IMF driving of the magnetosphere. Ideal MHD simulations with magnetic reconnection supported by numerical resistivity produce steady configuration with almost stationary near-earth neutral line (NENL). Simulations with non-gyrotropic corrections demonstrate dynamic quasi-periodic response to the steady driving condition. The loading/unloading cycle in non-gyrotropic MHD results has a non-stationary reconnection site in the magnetotail, with the retreating during the stretching phase and then a new NENL forming in the resulting thin plasma sheet. We expect that this model will lead to improved representations of space weather event in the magnetosphere.

Kuznetsova, M. M.↗

Multigrid simulation of asymptotic curved-duct flows using a semi-implicit numerical technique

Asymptotic flows inside curved ducts of rectangular as well as polar cross section are analyzed using the Navier-Stokes equations in terms of the axial velocity and vorticity and the cross-flow stream function. Numerical solutions of the three second-order coupled elliptic partial differential equations governing this flow are obtained efficiently using the coupled alternating-direction implicit (ADI) method as well as the multigrid strongly-implicit (SI) scheme. For the flow configuration studied, the ADI method is found to be more sensitive to the time steps used than is the SI scheme. Use of the multigrid-coupled-strongly-implicit (MG-SI) scheme makes it possible to efficiently obtain fine-grid solutions for configurations having strong secondary flow. It is shown that, for this asymptotic curved-duct flow, the similarity parameter of significance is the Dean's number K rather than the Reynolds number Re. Results are obtained for curved ducts with square cross sections for K up to 900, which here corresponds to Re = 9,000 for this internal flow configuration.

Ghia, K. N.↗

Flight test and numerical simulation of transonic flow around YAV-8B Harrier II wing

A computational fluid dynamics (CFD) method is used to study the aerodynamics of the YAV-8B Harrier II wing in the transonic region. A numerical procedure is developed to compute the flow field around the complicated wing-pylon-fairing geometry. The surface definition of the wing and pylons were obtained from direct measurement using theodolite triangulation. A thin-layer Navier-Stokes code with the Chimera technique is used to compute flow solutions. The computed pressure distributions at several span stations are compared with flight test data and show good agreement. Computed results are correlated with flight test data that show the flow is severely separated in the vicinity of the wing-pylon junction. Analysis shows that shock waves are induced by pylon swaybrace fairings, that the flow separation is much stronger at the outboard pylon and that the separation is caused mainly by the crossflow passing the geometry of wing-pylon junction.

Gea, Lie-Mine↗

A Numerical Study of Directional Solidification and Melting in Microgravity

A computational model is presented for the study of the solidification and melting of a pure substance and of a binary alloy. The enthalpy method has been used, and incorporated into a commercial CFD code. Three examples of the use of the model are described: the three-dimensional solidification of a pure substance (succinonitrile), the results of which are compared with experiment; an example of the solidification of a bismuth-tin alloy; and a simulation of a solidification and melting experiment done in space known as the MEPHISTO program.

Chen, P. Y. P.↗

Unsteady Aerodynamic Effects on the Flight Characteristics of an F-16XL Configuration

Unsteady aerodynamic models based on windtunnel forced oscillation test data and analyzed with a fuzzy logic algorithm arc incorporated into an F-16XL flight simulation code. The reduced frequency needed in the unsteady models is numerically calculated by using a limited prior time history of state variables in a least-square sense. Numerical examples arc presented to show the accuracy of the calculated reduced frequency. Oscillatory control inputs are employed to demonstrate the differences in the flight characteristics based on unsteady and quasi-steady aerodynamic models. Application of the unsteady aerodynamic models is also presented and the results are compared with one set of F16XIL longitudinal maneuver flight data. It is shown that the main differences in dynamic response are in the lateral-directional characteristics, with the quasi-steady model being more stable than the flight vehicle, while the unsteady model being more unstable. Similar conclusions can also be made in a simulated rapid sideslipping roll. To improve unsteady aerodynamic modeling, it is recommended to acquire test data with coupled motions in pitch, roll and yaw.

Wang, Zhongjun↗

Ordering Unstructured Meshes for Sparse Matrix Computations on Leading Parallel Systems

The ability of computers to solve hitherto intractable problems and simulate complex processes using mathematical models makes them an indispensable part of modern science and engineering. Computer simulations of large-scale realistic applications usually require solving a set of non-linear partial differential equations (PDES) over a finite region. For example, one thrust area in the DOE Grand Challenge projects is to design future accelerators such as the SpaHation Neutron Source (SNS). Our colleagues at SLAC need to model complex RFQ cavities with large aspect ratios. Unstructured grids are currently used to resolve the small features in a large computational domain; dynamic mesh adaptation will be added in the future for additional efficiency. The PDEs for electromagnetics are discretized by the FEM method, which leads to a generalized eigenvalue problem Kx = AMx, where K and M are the stiffness and mass matrices, and are very sparse. In a typical cavity model, the number of degrees of freedom is about one million. For such large eigenproblems, direct solution techniques quickly reach the memory limits. Instead, the most widely-used methods are Krylov subspace methods, such as Lanczos or Jacobi-Davidson. In all the Krylov-based algorithms, sparse matrix-vector multiplication (SPMV) must be performed repeatedly. Therefore, the efficiency of SPMV usually determines the eigensolver speed. SPMV is also one of the most heavily used kernels in large-scale numerical simulations.

Oliker, Leonid↗

Growth characteristics of a plane crack subjected to three-dimensional loading

The closed form expressions for the stress intensity factors due to concentrated forces applied to the surfaces of a half plane crack in an infinite body are used to generate solutions for distributed loads in this geometry. The stress intensity factors for uniformly distributed loads applied over a rectangular portion of the crack surface are given in closed form. An example of non-uniformly distributed loads which can be treated numerically is also included. In particular, combinations of normal and shear stresses on the crack which simulate the case of loading at an angle to the crack front are considered. The resulting stress intensity factors are combined with the strain energy density fracture criterion for the purpose of predicting the most likely direction of crack propagation. The critical value of the energy density factor can then be used for determining the allowable load on a specimen with a crack front not perpendicular to the tensile axis.

Hartranft, R. J.↗

Numerical simulation in three space dimensions of time-dependent thermal convection in a rotating fluid

Three-dimensional time-dependent convection in a plane layer of fluid, uniformly heated from below and subject to vertical shear and to rotation about an axis tilted from the vertical, was simulated by the numerical solution of the Boussinesq equations, including all Coriolis terms. Rotation about a vertical axis produces smaller convection cells with diminished heat fluxes and considerable vorticity. When the rotation axis is tilted from the vertical to represent tropical latitudes, the convection cells become elongated in a N-S direction. Imposed flows with constant vertical shear produce convective rolls aligned with the mean flow. When the rotation vector is tilted from the vertical, the competing effects due to rotation and shear can stabilize the convective motions.

Hathaway, D. H.↗

Numerical modeling of the interstellar medium in galactic disks

We have been developing detailed hydrodynamic models of the global interstellar medium in the hope of understanding the mass and volume occupied by various phases, as well as their structure and kinematics. In our model, the gas is modeled by one fluid while representative Pop 1 stars are modeled by a second fluid. The two fluids are coupled in that the gas forms into stars at a rate given by a Schmidt law while stellar mass loss returns matter into the gas phase (on a time scale of 100 Myr). Also, the stars heat the gas through stellar winds and the gas cools through optically thin radiation. The time behavior of these two fluids is studied in two spatial dimensions with the Eulerian finite difference numerical hydrodynamic code Zen. The two spatial dimensions are along the plane of a disk (x, total length of 2 kpc) and perpendicular to the disk (z, total height of +/- 15 kpc) and a galactic gravitational field in the z direction, typical of that at the solar circle, is imposed upon the simulation; self-gravity and rotation are absent. For the boundary conditions, outflow is permitted at the top and bottom of the grid (z = +/- 15 kpc) while periodic boundary conditions are imposed upon left and right sides of the grid. As initial conditions, we assumed a gaseous distribution like that seen for the H1 by earlier researchers, although the results are insensitive to the initial conditions. We have run simulations in which the heating due to stars, parameterized as a stellar wind velocity, a, is varied from low (a = 150 km/s), to intermediate (a = 300 km/s), to high (a = 600 km/s). Since the intermediate case is roughly equivalent to the Galactic energy injection rate from supernovae, this summary will concentrate on results from this simulation.

Rosen, A.↗

Numerical simulation of the leading-edge separation vortex for a wing and strake-wing configuration

In the present investigation, the 'thin layer' Navier-Stokes equations are used to compute the flow about a delta wing and a strake-delta wing configuration. Both configurations possess blunt noses and rounded leading edges. The computational grid about these configurations is generated using a newly developed three-dimensional grid-generation code which solves a set of Poisson equations with spherical coordinate variables using an alternating direction implicit (ADI) scheme. Computational results are obtained for an isolated wing with 60 deg sweep and a strake-wing configuration with 80-60 deg sweep. Computations for angles of attack in the range from 6 to 30 deg are presented. Attention is given to the effect of the angle of attack, the Mach number, and the Reynolds number. Theoretical results are compared with experimental data.

Fujii, K.↗

Optimization as a Tool for Consistency Maintenance in Multi-Resolution Simulation

The need for new approaches to the consistent simulation of related phenomena at multiple levels of resolution is great. While many fields of application would benefit from a complete and approachable solution to this problem, such solutions have proven extremely difficult. We present a multi-resolution simulation methodology that uses numerical optimization as a tool for maintaining external consistency between models of the same phenomena operating at different levels of temporal and/or spatial resolution. Our approach follows from previous work in the disparate fields of inverse modeling and spacetime constraint-based animation. As a case study, our methodology is applied to two environmental models of forest canopy processes that make overlapping predictions under unique sets of operating assumptions, and which execute at different temporal resolutions. Experimental results are presented and future directions are addressed.

TEMPORAL RESOLUTION↗

Large Eddy Simulation of a Near Sonic Turbulent Jet and Its Radiated Noise

In this paper numerical simulations are used to calculate the turbulence dynamics simultaneously with the sound field for a high-speed near-sonic (Ma=0.9) compressible jet at two Reynolds numbers of 3,600 and 72,000. LES (Large Eddy Simulation) in conjunction with accurate numerical schemes is used to calculate the unsteady flow and sound in the near field of the jet. It is shown that the jet mean parameters, mean velocity fields and turbulence statistics are in good agreement with experimental data and results from other simulations. The sound in the near-field is calculated directly from the simulations. The calculations are shown to capture the peak in the dilatation and pressure spectra around a Strouhal number St=0.25-0.3, in agreement with typical jet-noise spectra measured in experiments. Dilatation contours in the near-field show the formation of acoustic waves with a dominant wavelength of 3.2-4 jet diameters, corresponding to the peak in the dilatation spectra. As expected, the non-compact noise sources are found to be most dominant in the region corresponding to the end of the potential core. The contribution of the LES model to the radiated noise appears to be weak and does not contaminate the sound field with spurious high-frequency noise. However, the frequency spectra of the sound show a rapid falloff away from the peak frequency. This is attributed to the quasi-laminar state of the shear-layers in the region prior to potential core closure, and a possible effect of insufficient azimuthal resolution at the observed location. Further analysis of the effect of the LES model, especially at high frequencies, is needed.

Constantinescu, G. S.↗

Assessment of the Partially Resolved Numerical Simulation (PRNS) Approach in the National Combustion Code (NCC) for Turbulent Nonreacting and Reacting Flows

This paper describes an approach which aims at bridging the gap between the traditional Reynolds-averaged Navier-Stokes (RANS) approach and the traditional large eddy simulation (LES) approach. It has the characteristics of the very large eddy simulation (VLES) and we call this approach the partially-resolved numerical simulation (PRNS). Systematic simulations using the National Combustion Code (NCC) have been carried out for fully developed turbulent pipe flows at different Reynolds numbers to evaluate the PRNS approach. Also presented are the sample results of two demonstration cases: nonreacting flow in a single injector flame tube and reacting flow in a Lean Direct Injection (LDI) hydrogen combustor.

Shih, Tsan-Hsing↗

Can Eccentric Debris Disks Be Long-lived? A First Numerical Investigation and Application to Zeta(exp 2) Reticuli

Context. Imaging of debris disks has found evidence for both eccentric and offset disks. One hypothesis is that they provide evidence for massive perturbers, for example, planets or binary companions, which sculpt the observed structures. One such disk was recently observed in the far-IR by the Herschel Space Observatory around Zeta2 Reticuli. In contrast with previously reported systems, the disk is significantly eccentric, and the system is several Gyr old. Aims. We aim to investigate the long-term evolution of eccentric structures in debris disks caused by a perturber on an eccentric orbit around the star. We hypothesise that the observed eccentric disk around Zeta2 Reticuli might be evidence of such a scenario. If so, we are able to constrain the mass and orbit of a potential perturber, either a giant planet or a binary companion. Methods. Analytical techniques were used to predict the effects of a perturber on a debris disk. Numerical N-body simulations were used to verify these results and further investigate the observable structures that may be produced by eccentric perturbers. The long-term evolution of the disk geometry was examined, with particular application to the Zeta2 Reticuli system. In addition, synthetic images of the disk were produced for direct comparison with Herschel observations. Results. We show that an eccentric companion can produce both the observed offsets and eccentric disks. These effects are not immediate, and we characterise the timescale required for the disk to develop to an eccentric state (and any spirals to vanish). For Zeta2 Reticuli, we derive limits on the mass and orbit of the companion required to produce the observations. Synthetic images show that the pattern observed around Zeta2 Reticuli can be produced by an eccentric disk seen close to edge-on, and allow us to bring additional constraints on the disk parameters of our model (disk flux and extent). Conclusions. We conclude that eccentric planets or stellar companions can induce long-lived eccentric structures in debris disks. Observations of such eccentric structures thus provide potential evidence of the presence of such a companion in a planetary system. We considered the specific example of Zeta2 Reticuli, whose observed eccentric disk can be explained by a distant companion (at tens of AU) on an eccentric orbit (ep greater than approx. 0.3).

planetary systems↗

Buoyancy Driven Shear Flows of Bubble Suspensions

In this work the gas volume fraction and the root-mean-squared fluid velocity are measured in buoyancy driven shear flows of bubble suspensions in a tall, inclined, rectangular channel. The experiments are performed under conditions where We << 1a nd Re >> 1, for which comparisons are made with kinetic theory and numerical simulations. Here Re = gamma(a(exp 2)/nu is the Reynolds number and We = rho(gamma(exp 2))a(exp 3)/sigma is the Weber number; gamma is the shear rate, a is the bubble radius, nu is the kinematic viscosity of the liquid, rho is the density of the liquid, and sigma is the surface tension of the gas/liquid interface. Kang et al. calculated the bubble phase pressure and velocity variance of sheared bubble suspensions under conditions where the bubbles are spherical and the liquid phase velocity field can be approximated using potential flow theory, i.e. We= 0 and Re >> 1. Such conditions can be achieved in an experiment using gas bubbles, with a radius of O(0.5mm), in water. The theory requires that there be no average relative motion of the gas and liquid phases, hence the motivation for an experimental program in microgravity. The necessity of performing preliminary, Earth based experiments, however, requires performing experiments where the gas phase rises in the liquid, which significantly complicates the comparison of experiments with theory. Rather than comparing experimental results with theory for a uniform, homogeneous shear flow, experiments can be compared directly with solutions of the averaged equations of motion for bubble suspensions. This requires accounting for the significant lift force acting on the gas phase when the bubbles rise parallel to the average velocity of the sheared suspension. Shear flows can be produced in which the bubble phase pressure gradient, arising from shear induced collisions amongst the bubbles, balances a body force (centrifugal or gravitational) on the gas phase. A steady, non-uniform gas volume fraction can be measured, from which the bubble phase pressure gradient can be obtained and compared to theory and numerical simulations. The presence of bounding walls further complicates the experiments, since the detailed interactions of the bubbles with bounding walls is not well understood, especially in the presence of gravity, where the momentum and energy exchange depends on the inclination of the wall.

Koch, D. L.↗

Direct Harmonic Linear Navier-Stokes Methods for Efficient Simulation of Wave Packets

Wave packets produced by localized disturbances play an important role in transition in three-dimensional boundary layers, such as that on a swept wing. Starting with the receptivity process, we show the effects of wave-space energy distribution on the development of packets and other three-dimensional disturbance patterns. Nonlinearity in the receptivity process is specifically addressed, including demonstration of an effect which can enhance receptivity of traveling crossflow disturbances. An efficient spatial numerical simulation method is allowing most of the simulations presented to be carried out on a workstation.

Streett, C. L.↗