Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “nonlinear model reduction”

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 541 records · Page 30

HerMES: COSMIC INFRARED BACKGROUND ANISOTROPIES AND THE CLUSTERING OF DUSTY STAR-FORMING GALAXIES

We present measurements of the auto- and cross-frequency power spectra of the cosmic infrared background (CIB) at 250, 350, and 500 {mu}m (1200, 860, and 600 GHz) from observations totaling {approx}70 deg{sup 2} made with the SPIRE instrument aboard the Herschel Space Observatory. We measure a fractional anisotropy {delta}I/I = 14% {+-} 4%, detecting signatures arising from the clustering of dusty star-forming galaxies in both the linear (2-halo) and nonlinear (1-halo) regimes; and that the transition from the 2- to 1-halo terms, below which power originates predominantly from multiple galaxies within dark matter halos, occurs at k{sub {theta}} {approx} 0.10-0.12 arcmin{sup -1} (l {approx} 2160-2380), from 250 to 500 {mu}m. New to this paper is clear evidence of a dependence of the Poisson and 1-halo power on the flux-cut level of masked sources-suggesting that some fraction of the more luminous sources occupy more massive halos as satellites, or are possibly close pairs. We measure the cross-correlation power spectra between bands, finding that bands which are farthest apart are the least correlated, as well as hints of a reduction in the correlation between bands when resolved sources are more aggressively masked. In the second part of the paper, we attempt to interpret the measurements in the framework of the halo model. With the aim of fitting simultaneously with one model the power spectra, number counts, and absolute CIB level in all bands, we find that this is achievable by invoking a luminosity-mass relationship, such that the luminosity-to-mass ratio peaks at a particular halo mass scale and declines toward lower and higher mass halos. Our best-fit model finds that the halo mass which is most efficient at hosting star formation in the redshift range of peak star-forming activity, z {approx} 1-3, is log(M{sub peak}/M{sub Sun }) {approx} 12.1 {+-} 0.5, and that the minimum halo mass to host infrared galaxies is log(M{sub min}/M{sub Sun }) {approx} 10.1 {+-} 0.6.

79 ASTRONOMY AND ASTROPHYSICS↗

Constitutive modeling for isotropic materials

The first year of progress on a NASA-Lewis contract with the General Electric Co is documented. The purpose of this contract (NAS3-23927) is to develop and evaluate unified constitutive equations for applications to hot-path components of aircraft gas turbine engines such as high pressure turbine blades and vanes. To accomplish this goal, uniaxial, notched, and multiaxial specimens made of conventionally cast Rene 80 are being tested under conditions that simulate engine operating conditions. To reduce the raw data, automated data reduction techniques are being developed that produce computer files containing the information needed to analyze proposed constitutive theories. Described are the analytical methods being developed to determine the parameters for these nonlinear unified theories by using the reduced data files. In another activity, a dedicated finite-element computer code is being developed to use unified theories in the structural analysis of hot-section components. This code was extensively verified for one such theory by successfully predicting the strain histories measured experimentally at the notch root of complex specimens taken from complex laboratory specimens.

Ramaswamy, V. G.↗

Electroelastic investigation of drying rate in the direct contact ultrasonic fabric dewatering process

Ultrasonic vibrations, used to atomize liquids into a fine mist, are a promising solution for the future of efficient clothes drying technology. The world’s first ultrasonic dryer—demonstrated by researchers at Oak Ridge National Laboratory—successfully applies the scientific principles behind ultrasonic drying, and several working prototypes have been demonstrated. This technology is based on direct mechanical coupling between mesh piezoelectric transducers and wet fabric. During the atomization process, vertical oscillations of a contained liquid, called Faraday excitations, result in the formation of standing waves on the liquid surface. At increasing amplitudes and frequencies of oscillation, wave peaks become extended and form “necks” connecting small secondary droplets to the bulk liquid. When the oscillation reaches an acceleration threshold, the droplet momentum is sufficient to break the surface tension of the neck and enable the droplets to travel away from the liquid. In this work, we investigate the atomization process using an ultrasonic transducer as it pertains to moisture retained within a fabric. An experimentally validated electromechanical analytical-numerical model is proposed. This model bridges the vibrations of a piezoelectric mesh transducer to the critical acceleration needed for fabric drying to occur. Then, the drying rate model is developed, consisting of an initial nonlinear region due to atomization, followed by a linear thermal evaporation region. The models developed identify the influence of key parameters on ultrasonic drying and will aid in improving atomizer design for efficient, timely fabric drying. This study is the first proposed model for the ultrasonic atomization of fabrics saturated with water, applicable to any type of transducer. The results present a non-dimensional equation for the ultrasonic dewatering of fabrics, dependent only on transducer acceleration and the surface area of the cloth. Furthermore, the development of this technology using the proposed physical models will allow for global reductions in electrical demand related to clothes drying.

42 ENGINEERING↗

Accurate and Efficient Parametric Model-Order Reduction for Turbulent Thermal Transport

This project produced new algorithms and software for low-cost exploration of turbulent thermal-fluids behavior under parametric variation by using reduced-order models (ROMs). The ROMs numerically solve the governing equations for fluid motion by using a small set of basis functions (typically, N=20-200 modes) to represent the solution. The base modes are computed as optimal combinations of solutions from expensive high-fidelity "anchor-point" solutions involving millions of unknowns, which are typically generated by solving the full Navier-Stokes equations on a supercomputer. The ROM solution is itself a combination of the base modes, where the basis coefficients are determined by evolving an NxN system of nonlinear equations. The overarching idea is to use the inexpensive ROM to predict solutions under conditions where the parameters differ from the anchor-point conditions. Several ingredients are required to make ROMs useful for thermal hydraulics analysis. These include: a stable and accurate ROM that is capable of reproducing the large-scale dynamics of turbulent flow, error indicators than can guide the choice of anchor points, and low-cost mechanisms for evaluating nonlinear terms in the reduced equations.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Reduced order models for thermal analysis : final report : LDRD Project No. 137807.

This LDRD Senior's Council Project is focused on the development, implementation and evaluation of Reduced Order Models (ROM) for application in the thermal analysis of complex engineering problems. Two basic approaches to developing a ROM for combined thermal conduction and enclosure radiation problems are considered. As a prerequisite to a ROM a fully coupled solution method for conduction/radiation models is required; a parallel implementation is explored for this class of problems. High-fidelity models of large, complex systems are now used routinely to verify design and performance. However, there are applications where the high-fidelity model is too large to be used repetitively in a design mode. One such application is the design of a control system that oversees the functioning of the complex, high-fidelity model. Examples include control systems for manufacturing processes such as brazing and annealing furnaces as well as control systems for the thermal management of optical systems. A reduced order model (ROM) seeks to reduce the number of degrees of freedom needed to represent the overall behavior of the large system without a significant loss in accuracy. The reduction in the number of degrees of freedom of the ROM leads to immediate increases in computational efficiency and allows many design parameters and perturbations to be quickly and effectively evaluated. Reduced order models are routinely used in solid mechanics where techniques such as modal analysis have reached a high state of refinement. Similar techniques have recently been applied in standard thermal conduction problems e.g. though the general use of ROM for heat transfer is not yet widespread. One major difficulty with the development of ROM for general thermal analysis is the need to include the very nonlinear effects of enclosure radiation in many applications. Many ROM methods have considered only linear or mildly nonlinear problems. In the present study a reduced order model is considered for application to the combined problem of thermal conduction and enclosure radiation. The main objective is to develop a procedure that can be implemented in an existing thermal analysis code. The main analysis objective is to allow thermal controller software to be used in the design of a control system for a large optical system that resides with a complex radiation dominated enclosure. In the remainder of this section a brief outline of ROM methods is provided. The following chapter describes the fully coupled conduction/radiation method that is required prior to considering a ROM approach. Considerable effort was expended to implement and test the combined solution method; the ROM project ended shortly after the completion of this milestone and thus the ROM results are incomplete. The report concludes with some observations and recommendations.

42 ENGINEERING↗

Analytical and numerical investigation of structural response of compliant wall materials

Surface motion of compliant walls in drag reduction experiments was analyzed. The spectrum of surface motion indicates that membranes over deep cavities respond at low frequencies and large wavelengths. The membrane over a deep cavity is therefore found not to yield the desired reponse predicted by the postulated mechanism. The membrane over a thin air gap is found to act as a wavelength chopper, and analysis of the nonlinear response of the compliant surface indicates its possible suitability for compliant wall experiments. Periodic structures are found to lock in the desired wavelengths of motion. Laminated structures are found to be very ineffective as compliant models, except when there is no bonding between the membrane and the backing. Computer programs developed for these analyses are documented.

Goglia, G. L.↗

Sparsified time-dependent Fourier neural operators for fusion simulations

This paper presents a sparsified Fourier neural operator for coupled time-dependent partial differential equations (ST-FNO) as an efficient machine learning surrogate for fluid and particle-based fusion codes such as NIMROD (Non-Ideal Magnetohydrodynamics with Rotation - Open Discussion) and GTC (Gyrokinetic Toroidal Code). ST-FNO leverages the structures in the governing equations and utilizes neural operators to represent Green's function-like numerical operators in the corresponding numerical solvers. Once trained, ST-FNO can rapidly and accurately predict dynamics in fusion devices compared with first-principle numerical algorithms. In general, ST-FNO represents an efficient and accurate machine learning surrogate for numerical simulators for multi-variable nonlinear time-dependent partial differential equations, with the proposed architectures and loss functions. The efficacy of ST-FNO has been demonstrated using quiescent H-mode simulation data from NIMROD and kink-mode simulation data from GTC. The ST-FNO H-mode results show orders of magnitude reduction in memory and central processing unit usage in comparison with the numerical solvers in NIMROD when computing fields over a selected poloidal plane. The ST-FNO kink-mode results achieve a factor of 2 reduction in the number of parameters compared to baseline FNO models without accuracy loss.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Discrete models of fluids: spatial averaging, closure and model reduction

We consider semidiscrete ODE models of single-phase fluids and two-fluid mixtures. In the presence of multiple fine-scale heterogeneities, the size of these ODE systems can be very large. Spatial averaging is then a useful tool for reducing computational complexity of the problem. The averages satisfy exact balance equations of mass, momentum, and energy. These equations do not form a satisfactory continuum model because evaluation of stress and heat flux requires solving the underlying ODEs. To produce continuum equations that can be simulated without resolving microscale dynamics, we recently proposed a closure method based on the use of regularized deconvolution. Here we continue the investigation of deconvolution closure with the long term objective of developing consistent computational upscaling for multiphase particle methods. The structure of the fine-scale particle solvers is reminiscent of molecular dynamics. For this reason we use nonlinear averaging introduced for atomistic systems by Noll, Hardy, and Murdoch-Bedeaux. We also consider a simpler linear averaging originally developed in large eddy simulation of turbulence. We present several simple but representative examples of spatially averaged ODEs, where the closure error can be analyzed. Based on this analysis we suggest a general strategy for reducing the relative error of approximate closure. For problems with periodic highly oscillatory material parameters we propose a spectral boosting technique that augments the standard deconvolution and helps to correctly account for dispersion effects. We also conduct several numerical experiments, one of which is a complete mesoscale simulation of a stratified two-fluid flow in a channel. In this simulation, the operation count per coarse time step scales sublinearly with the number of particles.

Multiscale modeling, discrete models, fluid flow↗

MHD simulation of RF current drive in MST

Auxiliary heating and current drive using RF waves such as the electron Bernstein wave (EBW) promises to advance the performance of the reversed field pinch (RFP). In previous computational work [1], a hypothetical edge-localized current drive is shown to suppress the tearing activity which governs the macroscopic transport properties of the RFP. The ideal conditions for tearing stabilization include a reduced toroidal induction, and precise width and radial position of the Gaussian-shaped external current drive. In support of the EBW experiment on the Madison Symmetric Torus, an integrated modeling scheme now incorporates ray tracing and Fokker-Plank predictions of auxiliary current into single fluid MHD. Simulations at low Lundquist number (S ∼ 10{sup 4}) generally agree with the previous work; significantly more burdensome simulations at MST-like Lundquist number (S ∼ 3×10{sup 6}) show unexpected results. The effect on nonlinearly saturated current profile by a particular RF-driven external force decreases in magnitude and widens considerably as the Lundquist number increases toward experimental values. Simulations reproduce the periodic current profile relaxation events observed in experiment (sawteeth) in the absence of current profile control. Reduction of the tearing mode amplitudes is still observable; however, reduction is limited to periods between the large bursts of magnetic activity at each sawtooth. The sawtoothing pattern persists with up to 10 MW of externally applied RF power. Periods with prolonged low tearing amplitude are predicted with a combination of external current drive and a reduced toroidal loop voltage, consistent with previous conclusions. Finally, the resistivity profile is observed to have a strong effect on the optimal externally driven current profile for mode stabilization.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Multi-dimensional instability of obliquely propagating ion acoustic solitary waves in electron-positron-ion superthermal magnetoplasmas

The solitary structures of multi–dimensional ion-acoustic solitary waves (IASWs) have been considered in magnetoplasmas consisting of electron-positron-ion with high-energy (superthermal) electrons and positrons are investigated. Using a reductive perturbation method, a nonlinear Zakharov-Kuznetsov equation is derived. The multi-dimensional instability of obliquely propagating (with respect to the external magnetic field) IASWs has been studied by the small-k (long wavelength plane wave) expansion perturbation method. The instability condition and the growth rate of the instability have been derived. It is shown that the instability criterion and their growth rate depend on the parameter measuring the superthermality, the ion gyrofrequency, the unperturbed positrons-to-ions density ratio, the direction cosine, and the ion-to-electron temperature ratio. Clearly, the study of our model under consideration is helpful for explaining the propagation and the instability of IASWs in space observations of magnetoplasmas with superthermal electrons and positrons.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Effects of q-profile structure on turbulence spreading: A fluctuation intensity transport analysis

This paper studies effects of q-profile structure on turbulence spreading. It reports results of numerical experiments using global gyrokinetic simulations. We examine propagation of turbulence, triggered by an identical linear instability in a source region, into an adjacent, linearly stable region with variable q-profile. The numerical experiments are designed so as to separate the physics of turbulence spreading from that of linear stability. The strength of turbulence spreading is measured by the penetration depth of turbulence. Dynamics of spreading are elucidated by fluctuation intensity balance analysis, using a model intensity evolution equation which retains nonlinear diffusion and damping, and linear growth. It is found that turbulence spreading is strongly affected by magnetic shear s, but is hardly altered by the safety factor q itself. There is an optimal range of modest magnetic shear which maximizes turbulence spreading. For high to modest shear values, the spreading is enhanced by the increase of the mode correlation length with decreasing magnetic shear. However, the efficiency of spreading drops for sufficiently low magnetic shear even though the mode correlation length is comparable to that for the case of optimal magnetic shear. The reduction of spreading is attributed to the increase in time required for the requisite nonlinear mode-mode interactions. The effect of increased interaction time dominates that of increased mode correlation length. Our findings of the reduction of spreading and the increase in interaction time at weak magnetic shear are consistent with the well-known benefit of weak or reversed magnetic shear for core confinement enhancement. Weak shear is shown to promote locality, as well as stability.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Zonal structure effect on the nonlinear saturation of reverse shear Alfven eigenmodes

A gyrokinetic ion/fluid electron hybrid model is used to study the nonlinear evolution of reverse shear Alfven eigenmodes (RSAE) driven by energetic particles. When only the energetic particle nonlinear effects are included, the saturation amplitude of a single-n RSAE follows the trapping scaling, δB/B~(γ/ω) 2 . When the nonlinear effects of thermal ions and electrons are included, zonal structures are force generated but do not affect the saturation amplitude for γ/ω≤0.03. No spontaneous generation of zonal structures is observed, in contrast to ion-temperature-gradient-driven modes. At larger energetic particle drive, the effects of zonal structures cause a significant reduction in the RSAE saturation amplitude. The reduction is not caused by the zonal flow shearing of the RSAE, but by the force-generated n = 0 component in the thermal ion distribution function and the electron density. Finally, these n = 0 perturbations lead to nonlinear evolution of the RSAE mode structure and enhance damping.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Kernel principal component analysis for stochastic input model generation

Highlights: {yields} KPCA is used to construct a reduced order stochastic model of permeability. {yields} A new approach is proposed to solve the pre-image problem in KPCA. {yields} Polynomial chaos is used to provide a parametric stochastic input model. {yields} Flow in porous media with channelized permeability is considered. - Abstract: Stochastic analysis of random heterogeneous media provides useful information only if realistic input models of the material property variations are used. These input models are often constructed from a set of experimental samples of the underlying random field. To this end, the Karhunen-Loeve (K-L) expansion, also known as principal component analysis (PCA), is the most popular model reduction method due to its uniform mean-square convergence. However, it only projects the samples onto an optimal linear subspace, which results in an unreasonable representation of the original data if they are non-linearly related to each other. In other words, it only preserves the first-order (mean) and second-order statistics (covariance) of a random field, which is insufficient for reproducing complex structures. This paper applies kernel principal component analysis (KPCA) to construct a reduced-order stochastic input model for the material property variation in heterogeneous media. KPCA can be considered as a nonlinear version of PCA. Through use of kernel functions, KPCA further enables the preservation of higher-order statistics of the random field, instead of just two-point statistics as in the standard Karhunen-Loeve (K-L) expansion. Thus, this method can model non-Gaussian, non-stationary random fields. In this work, we also propose a new approach to solve the pre-image problem involved in KPCA. In addition, polynomial chaos (PC) expansion is used to represent the random coefficients in KPCA which provides a parametric stochastic input model. Thus, realizations, which are statistically consistent with the experimental data, can be generated in an efficient way. We showcase the methodology by constructing a low-dimensional stochastic input model to represent channelized permeability in porous media.

97 MATHEMATICS AND COMPUTING↗

On the multistream approach of relativistic Weibel instability. III. Comparison with full-kinetic Vlasov simulations

The saturation of the Weibel instability in the relativistic regime is investigated within the Hamiltonian reduction technique based on the multistream approach developed in paper I in the linear case and in paper II for the nonlinear saturation. In this work, the study is compared with results obtained by full kinetic 1D2V Vlasov-Maxwell simulations based on a semi-Lagrangian technique. For a temperature anisotropy, qualitatively different regimes are realized depending on the excitation of the longitudinal (plasma) electric field, in contrast with the existing theories of the Weibel instability based on their purely transverse characters. The emphasis here is on gaining a better understanding of the nonlinear aspects of the Weibel instability. The multistream model offers an alternate way to make calculations or numerical experiments more tractable, when only a few moments of the velocity distribution of the plasma are considered.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

ELECTRON HEATING IN MAGNETOROTATIONAL INSTABILITY: IMPLICATIONS FOR TURBULENCE STRENGTH IN THE OUTER REGIONS OF PROTOPLANETARY DISKS

The magnetorotational instability (MRI) drives vigorous turbulence in a region of protoplanetary disks where the ionization fraction is sufficiently high. It has recently been shown that the electric field induced by the MRI can heat up electrons and thereby affect the ionization balance in the gas. In particular, in a disk where abundant dust grains are present, the electron heating causes a reduction of the electron abundance, thereby preventing further growth of the MRI. By using the nonlinear Ohm's law that takes into account electron heating, we investigate where in protoplanetary disks this negative feedback between the MRI and ionization chemistry becomes important. We find that the “e-heating zone,” the region where the electron heating limits the saturation of the MRI, extends out up to 80 AU in the minimum-mass solar nebula with abundant submicron-sized grains. This region is considerably larger than the conventional dead zone whose radial extent is ∼20 AU in the same disk model. Scaling arguments show that the MRI turbulence in the e-heating zone should have a significantly lower saturation level. Submicron-sized grains in the e-heating zone are so negatively charged that their collisional growth is unlikely to occur. Our present model neglects ambipolar and Hall diffusion, but our estimate shows that ambipolar diffusion would also affect the MRI in the e-heating zone.

79 ASTRONOMY AND ASTROPHYSICS↗

ARCADE IMPLOSION CAUSED BY A FILAMENT ERUPTION IN A FLARE

Coronal implosions—the convergence motion of plasmas and entrained magnetic field in the corona due to a reduction in magnetic pressure—can help to locate and track sites of magnetic energy release or redistribution during solar flares and eruptions. We report here on the analysis of a well-observed implosion in the form of an arcade contraction associated with a filament eruption, during the C3.5 flare SOL2013-06-19T07:29. A sequence of events including the magnetic flux-rope instability and distortion, followed by a filament eruption and arcade implosion, lead us to conclude that the implosion arises from the transfer of magnetic energy from beneath the arcade as part of the global magnetic instability, rather than due to local magnetic energy dissipation in the flare. The observed net contraction of the imploding loops, which is found also in nonlinear force-free field extrapolations, reflects a permanent reduction of magnetic energy underneath the arcade. This event shows that, in addition to resulting in the expansion or eruption of an overlying field, flux-rope instability can also simultaneously implode an unopened field due to magnetic energy transfer. It demonstrates the “partial opening of the field” scenario, which is one of the ways in 3D to produce a magnetic eruption without violating the Aly–Sturrock hypothesis. In the framework of this observation, we also propose a unification of three main concepts for active region magnetic evolution, namely the metastable eruption model, the implosion conjecture, and the standard “CSHKP” flare model.

79 ASTRONOMY AND ASTROPHYSICS↗

On the marginally stable saturation spectrum of unstable type I equatorial electrojet irregularities

Formulation of a self-consistent convective nonlinear theory of type I irregularities in the equatorial electrojet. It is found that a combination of three mechanisms - convective amplification, quasi-linear polarization electric field reduction, and nonlinear particle orbit diffusion damping - accounts for radar backscatter observations of a ubiquitous marginally stable (or 'constant ion-acoustic Doppler shift') saturation spectrum better than any of the three mechanisms treated separately. In particular, no spatially homogeneous theory without wave refraction can account for the observations. Wave refraction alone or with quasi-linear polarization electric field reduction is also inadequate. Wave refraction, quasi-linear polarization reduction, and particle orbit diffusion theory appear to account for type I observations at radar elevation angles less than 60 deg. Vertical type I backscatter cannot be explained without modifying the present laminar electrojet model.

Lee, K.↗

One-dimensional particle simulations of Knudsen-layer effects on D-T fusion

Particle simulations are used to solve the fully nonlinear, collisional kinetic equation describing the interaction of a high-temperature, high-density, deuterium-tritium plasma with absorbing boundaries, a plasma source, and the influence of kinetic effects on fusion reaction rates. Both hydrodynamic and kinetic effects influence the end losses, and the simulations show departures of the ion velocity distributions from Maxwellian due to the reduction of the population of the highest energy ions (Knudsen-layer effects). The particle simulations show that the interplay between sources, plasma dynamics, and end losses results in temperature anisotropy, plasma cooling, and concomitant reductions in the fusion reaction rates. However, for the model problems and parameters considered, particle simulations show that Knudsen-layer modifications do not significantly affect the velocity distribution function for velocities most important in determining the fusion reaction rates, i.e., the thermal fusion reaction rates using the local densities and bulk temperatures give good estimates of the kinetic fusion reaction rates.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗