Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “function approximation”

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 1,477 records · Page 82
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

CO 2 storage site characterization using ensemble-based approaches with deep generative models

Estimating spatially distributed properties such as permeability from available sparse measurements is a great challenge in efficient subsurface CO 2 storage operations. In this paper, a deep generative model that can accurately capture complex subsurface structure is tested with an ensemble-based inversion method for accurate and accelerated characterization of CO 2 storage sites. We chose Wasserstein Generative Adversarial Network with Gradient Penalty (WGAN-GP) for its realistic reservoir property representation and Ensemble Smoother with Multiple Data Assimilation (ES-MDA) for its robust data fitting and uncertainty quantification capability. WGAN-GP are trained to generate high-dimensional permeability fields from a low-dimensional latent space and ES-MDA then updates the latent variables by assimilating available measurements. Several subsurface site characterization examples including Gaussian, channelized, and fractured reservoirs are used to evaluate the accuracy and computational efficiency of the proposed method and the main features of the unknown permeability fields are characterized accurately with reliable uncertainty quantification. Furthermore, the estimation performance is compared with a widely-used variational, i.e., optimization-based, inversion approach, and the proposed approach outperforms the variational inversion method in several benchmark cases. We explain such superior performance by visualizing the objective function in the latent space: because of nonlinear and aggressive dimension reduction via generative modeling, the objective function surface becomes extremely complex while the ensemble approximation can smooth out the multi-modal surface during the minimization. This suggests that the ensemble-based approach works well over the variational approach when combined with deep generative models at the cost of forward model runs unless convergence-ensuring modifications are implemented in the variational inversion.

42 ENGINEERING↗
DOE OSTI2015

An adaptive importance sampling algorithm for Bayesian inversion with multimodal distributions

Parametric uncertainties are encountered in the simulations of many physical systems, and may be reduced by an inverse modeling procedure that calibrates the simulation results to observations on the real system being simulated. Following Bayes' rule, a general approach for inverse modeling problems is to sample from the posterior distribution of the uncertain model parameters given the observations. However, the large number of repetitive forward simulations required in the sampling process could pose a prohibitive computational burden. This difficulty is particularly challenging when the posterior is multimodal. We present in this paper an adaptive importance sampling algorithm to tackle these challenges. Two essential ingredients of the algorithm are: 1) a Gaussian mixture (GM) model adaptively constructed as the proposal distribution to approximate the possibly multimodal target posterior, and 2) a mixture of polynomial chaos (PC) expansions, built according to the GM proposal, as a surrogate model to alleviate the computational burden caused by computational-demanding forward model evaluations. In three illustrative examples, the proposed adaptive importance sampling algorithm demonstrates its capabilities of automatically finding a GM proposal with an appropriate number of modes for the specific problem under study, and obtaining a sample accurately and efficiently representing the posterior with limited number of forward simulations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2015

A frozen Gaussian approximation-based multi-level particle swarm optimization for seismic inversion

In this paper, we propose a frozen Gaussian approximation (FGA)-based multi-level particle swarm optimization (MLPSO) method for seismic inversion of high-frequency wave data. The method addresses two challenges in it: First, the optimization problem is highly non-convex, which makes hard for gradient-based methods to reach global minima. This is tackled by MLPSO which can escape from undesired local minima. Second, the character of high-frequency of seismic waves requires a large number of grid points in direct computational methods, and thus renders an extremely high computational demand on the simulation of each sample in MLPSO. We overcome this difficulty by three steps: First, we use FGA to compute high-frequency wave propagation based on asymptotic analysis on phase plane; Then we design a constrained full waveform inversion problem to prevent the optimization search getting into regions of velocity where FGA is not accurate; Last, we solve the constrained optimization problem by MLPSO that employs FGA solvers with different fidelity. The performance of the proposed method is demonstrated by a two-dimensional full-waveform inversion example of the smoothed Marmousi model.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2020

DPM: A deep learning PDE augmentation method with application to large-eddy simulation

A framework is introduced that leverages known physics to reduce overfitting in machine learning for scientific applications. The partial differential equation (PDE) that expresses the physics is augmented with a neural network that uses available data to learn a description of the corresponding unknown or unrepresented physics. Training within this combined system corrects for missing, unknown, or erroneously represented physics, including discretization errors associated with the PDE's numerical solution. For optimization of the network within the PDE, an adjoint PDE is solved to provide high-dimensional gradients, and a stochastic adjoint method (SAM) further accelerates training. Additionally, the approach is demonstrated for large-eddy simulation (LES) of turbulence. High-fidelity direct numerical simulations (DNS) of decaying isotropic turbulence provide the training data used to learn sub-filter-scale closures for the filtered Navier–Stokes equations. Out-of-sample comparisons show that the deep learning PDE method outperforms widely-used models, even for filter sizes so large that they become qualitatively incorrect. It also significantly outperforms the same neural network when a priori trained based on simple data mismatch, not accounting for the full PDE. Measures of discretization errors, which are well-known to be consequential in LES, point to the importance of the unified training formulation's design, which without modification corrects for them. For comparable accuracy, simulation runtime is significantly reduced. A relaxation of the typical discrete enforcement of the divergence-free constraint in the solver is also successful, instead allowing the DPM to approximately enforce incompressibility physics. Since the training loss function is not restricted to correspond directly to the closure to be learned, training can incorporate diverse data, including experimental data.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2018

Analysis of the structural, electronic, elastic and thermodynamic properties of CuAl2X4 (X = O, S) spinel structure

Highlights: • Fundamental properties of CuAl{sub 2}X{sub 4} (X = O, S) aluminate spinels have been investigated. • The pressure dependence of the elastic constants C{sub ij} can be successfully fit by a straight line. • CuAl{sub 2}X{sub 4} (X = O, S) compounds are mechanically stable at ambient conditions. • Temperature and pressure dependences of some macroscopic parameters are achieved. - Abstract: The structural, electronic, elastic and thermodynamic properties of oxo- and thio-spinels have been predicted based on the ultrasoft pseudo-potential scheme as implemented in the CASTEP code. The exchange-correlation potential was treated within the generalized gradient approximation (GGA-PBE). The calculated structural parameters such as lattice constants and internal parameters are in good agreement with the available experimental data. The ground state electronic band structure revealed that CuAl{sub 2}X{sub 4} (XO, S) spinels are direct band gap semiconductors with a transition along Γ-Γ symmetry points. The calculated values of elastic constants satisfy the Born criteria at ambient conditions and confirm that the spinel structures of CuAl{sub 2}O{sub 4} and CuAl{sub 2}S{sub 4} are mechanically stable. The thiospinel structure may have a phase transition under pressure, due to the negative value of C{sub 44}. The influence of temperature and pressure on macroscopic characteristics of CuAl{sub 2}O{sub 4} and CuAl{sub 2}S{sub 4} compounds was estimated using the quasi-harmonic Debye model.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2020

SympNets: Intrinsic structure-preserving symplectic networks for identifying Hamiltonian systems

Here, we propose new symplectic networks (SympNets) for identifying Hamiltonian systems from data based on a composition of linear, activation and gradient modules. In particular, we define two classes of SympNets: the LA-SympNets composed of linear and activation modules, and the G-SympNets composed of gradient modules. Correspondingly, we prove two new universal approximation theorems that demonstrate that SympNets can approximate arbitrary symplectic maps based on appropriate activation functions. We then perform several experiments including the pendulum, double pendulum and three-body problems to investigate the expressivity and the generalization ability of SympNets. The simulation results show that even very small size SympNets can generalize well, and are able to handle both separable and non-separable Hamiltonian systems with data points resulting from short or long time steps. In all the test cases, SympNets outperform the baseline models, and are much faster in training and prediction. We also develop an extended version of SympNets to learn the dynamics from irregularly sampled data. This extended version of SympNets can be thought of as a universal model representing the solution to an arbitrary Hamiltonian system.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2018

Area-preserving scheme for efficiency enhancement in single-pass tapered free electron lasers

Efficiency enhancement of a single-pass short-wavelength high-gain tapered free electron laser (FEL) has recently been intensively studied. The goal is to sustain the growth of radiation power in the post-saturation regime. Among the various schemes, the undulator tapering is considered an effective route to achieve higher power extraction efficiency. The tapering strategy can be of constant or varying resonant phase along the undulator axis. In this paper we propose an efficiency-enhancement scheme based on preservation of the longitudinal phase space area which ensures trapping of resonant particles in the ponderomotive bucket as long as possible along the undulator axis before significant particle depletion occurs. In the meanwhile such a scheme takes advantage of the increase of the radiation field amplitude to precipitate the particle deceleration process at the middle stage of undulator tapering. We analyze such an area-preserving scheme of undulator tapering by formulating the post-saturation FEL interaction in a one-dimensional (1-D) model via introduction of the particle trapping fraction. The output performance is evaluated through numerical iteration and confirmed with 1-D particle tracking simulations. The results show that the optimal power extraction efficiency based on the proposed scheme, together with a prebunched beam, can be greatly improved within relatively short taper length compared with other schemes before radiation diffraction effect becomes significant. Besides, the undesired sideband effects are found to be effectively suppressed. For the proposed area-preserving taper scheme, we also derive an analytical approximate formula for the resonant phase as a function of undulator axis. In conclusion, we expect that the analysis can shed light on the aim to further enhance the power extraction efficiency in single-pass tapered FELs.

47 OTHER INSTRUMENTATION↗
DOE OSTI2016

Long-lived K isomer and enhanced γ vibration in the neutron-rich nucleus 172 Dy: Collectivity beyond double midshell

The level structure of 172Dy has been investigated for the first time by means of decay spectroscopy following in-flight fission of a 238U beam. A long-lived isomeric state with T1/2 = 0.71(5) s and Kπ = 8- has been identified at 1278 keV, which decays to the ground-state and γ -vibrational bands through hindered electromagnetic transitions, as well as to the daughter nucleus 172Ho via allowed β decays. The robust nature of the Kπ = 8- isomer and the ground-state rotational band reveals an axially-symmetric structure for this nucleus. Meanwhile, the γ -vibrational levels have been identified at unusually low excitation energy compared to the neighboring well-deformed nuclei, indicating the significance of the microscopic effect on the non-axial collectivity in this doubly mid-shell region. The underlying mechanism of enhanced γ vibration is discussed in comparison with the deformed Quasiparticle Random-Phase Approximation based on a Skyrme energy-density functional.

172Dy↗
DOE OSTI2024

Optimization of nonlinear turbulence in stellarators

We present new stellarator equilibria that have been optimized for reduced turbulent transport using nonlinear gyrokinetic simulations within the optimization loop. The optimization routine involves coupling the pseudo-spectral GPU-native gyrokinetic code GX with the stellarator equilibrium and optimization code DESC . Since using GX allows for fast nonlinear simulations, we directly optimize for reduced nonlinear heat fluxes. To handle the noisy heat flux traces returned by these simulations, we employ the simultaneous perturbation stochastic approximation (SPSA) method that only uses two objective function evaluations for a simple estimate of the gradient. We show several examples that optimize for both reduced heat fluxes and good quasi-symmetry as a proxy for low neoclassical transport. Finally, we run full transport simulations using the T3D stellarator transport code to evaluate the changes in the macroscopic profiles.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Extremely Slow Diffusion of Gold Nanoparticles under Confinement in Mesoporous Silica

Here, we report an X-ray photon correlation spectroscopy (XPCS) study of the mobility of colloidal gold nanoparticles with a diameter of approximately 4 nm in glycerol under confinement within the channels of SBA-15 mesoporous silica with pore diameters of approximately 6 and 12 nm. The XPCS correlation functions result from an effective heterodyne signal due to the mixing of the coherent scattering from the nanoparticles and the mesoporous silica. Over the range of wave vectors and hence length scales probed, the nanoparticle dynamics are well described by one-dimensional diffusion. The nanoparticle diffusivity varies with temperature in a manner expected based on the temperature-dependent viscosity of glycerol; however, the magnitudes of the diffusion coefficients are several orders of magnitude smaller than those of the nanoparticles in bulk glycerol. We consider mechanisms that might contribute to this reduction in diffusivity including enhancement in hydrodynamic drag under confinement, effects of nanoparticle adsorption to the pore walls, and slowing structural dynamics in glycerol due to the combination of the nanometer-scale confinement and the presence of the nanoparticles.

36 MATERIALS SCIENCE↗
DOE OSTI2019

Coverage-Dependent Adsorption of Phenol on Pt(111) from First Principles

In this work, we quantify the coverage dependence of the adsorbate-adsorbate and metal-adsorbate interactions for phenol on Pt(111) using density functional theory. For the four most favorable adsorption sites, we find that the adsorption energy of phenol decreases linearly as a function of coverage. As such, the repulsive phenol-phenol lateral interactions are strongest near saturation when the inter-molecular distances are less than ~4.5 Å, manifesting in a decrease of the C-O dihedral bond angle of ~4°. The linear dependence of the adsorption energy on coverage allows for the construction of a mean-field model. We validate our mean-field model for phenol adsorption by comparing the theoretically predicted differential heat of adsorption to the experimental data, where a maximum deviation of 0.11 eV is found. This compares well to earlier reported results that used a similar mean-field approach for the adsorption of benzene on Pt(111), suggesting that lateral interactions between functionalized aromatics can be parameterized using a simpler approximation. Overall, this work demonstrates a simple method toward the improvement in current state-of-the-art hydrodeoxygenation modeling at surfaces by allowing for the incorporation of coverage effects in systems with large, aromatic compounds without resorting to overly complex models.

09 BIOMASS FUELS↗
DOE OSTI2025

Improving the Quasi‐Biennial Oscillation via a Surrogate‐Accelerated Multi‐Objective Optimization

Accurate simulation of the quasi-biennial oscillation (QBO) is challenging due to uncertainties in representing convectively generated gravity waves. We develop an end-to-end uncertainty quantification workflow that calibrates these gravity wave processes in E3SM for a realistic QBO. Central to our approach is a domain knowledge-informed, compressed representation of high-dimensional spatio-temporal wind fields. By employing a parsimonious statistical model that learns the fundamental frequency from complex observations, we extract interpretable and physically meaningful quantities capturing key attributes. Building on this, we train a probabilistic surrogate model that approximates the fundamental characteristics of the QBO as functions of critical physics parameters governing gravity wave generation. Leveraging the Karhunen–Loève decomposition, our surrogate efficiently represents these characteristics as a set of orthogonal features, capturing cross-correlations among multiple physics quantities evaluated at different pressure levels and enabling rapid surrogate-based inference at a fraction of the computational cost of full-scale simulations. Finally, we analyze the inverse problem using a multi-objective approach. Our study reveals a tension between amplitude and period that constrains the QBO representation, precluding a single optimal solution. To navigate this, we quantify the bi-criteria trade-off and generate a set of Pareto optimal parameter values that balance the conflicting objectives. This integrated workflow improves the fidelity of QBO simulations and offers a versatile template for uncertainty quantification in complex geophysical models.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Magic in twisted transition metal dichalcogenide bilayers

The long-wavelength moiré superlattices in twisted 2D structures have emerged as a highly tunable platform for strongly correlated electron physics. We study the moiré bands in twisted transition metal dichalcogenide homobilayers, focusing on WSe 2 , at small twist angles using a combination of first principles density functional theory, continuum modeling, and Hartree-Fock approximation. We reveal the rich physics at small twist angles θ < 4°, and identify a particular magic angle at which the top valence moiré band achieves almost perfect flatness. In the vicinity of this magic angle, we predict the realization of a generalized Kane-Mele model with a topological flat band, interaction-driven Haldane insulator, and Mott insulators at the filling of one hole per moiré unit cell. The combination of flat dispersion and uniformity of Berry curvature near the magic angle holds promise for realizing fractional quantum anomalous Hall effect at fractional filling. We also identify twist angles favorable for quantum spin Hall insulators and interaction-induced quantum anomalous Hall insulators at other integer fillings.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2016

Full-scale computation for all the thermoelectric property parameters of half-Heusler compounds

The thermoelectric performance of materials relies substantially on the band structures that determine the electronic and phononic transports, while the transport behaviors compete and counter-act for the power factor PF and figure-of-merit ZT. These issues make a full-scale computation of the whole set of thermoelectric parameters particularly attractive, while a calculation scheme of the electronic and phononic contributions to thermal conductivity remains yet challenging. In this work, we present a full-scale computation scheme based on the first-principles calculations by choosing a set of doped half- Heusler compounds as examples for illustration. The electronic structure is computed using the WIEN2k code and the carrier relaxation times for electrons and holes are calculated using the Bardeen and Shockley’s deformation potential (DP) theory. The finite-temperature electronic transport is evaluated within the framework of Boltzmann transport theory. In sequence, the density functional perturbation combined with the quasi-harmonic approximation and the Klemens’ equation is implemented for calculating the lattice thermal conductivity of carrier-doped thermoelectric materials such as Tidoped NbFeSb compounds without losing a generality. The calculated results show good agreement with experimental data. Lastly, the present methodology represents an effective and powerful approach to calculate the whole set of thermoelectric properties for thermoelectric materials.

36 MATERIALS SCIENCE↗
DOE OSTI2015

Two-Yukawa fluid at a hard wall: Field theory treatment

We apply a field-theoretical approach to study the structure and thermodynamics of a two-Yukawa fluid confined by a hard wall. We derive mean field equations allowing for numerical evaluation of the density profile which is compared to analytical estimations. Beyond the mean field approximation, analytical expressions for the free energy, the pressure, and the correlation function are derived. Subsequently, contributions to the density profile and the adsorption coefficient due to Gaussian fluctuations are found. Both the mean field and the fluctuation terms of the density profile are shown to satisfy the contact theorem. We further use the contact theorem to improve the Gaussian approximation for the density profile based on a better approximation for the bulk pressure. The results obtained are compared to computer simulation data.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2016

A gas-dynamical approach to radiation pressure acceleration

The study of high intensity ion beams driven by high power pulsed lasers is an active field of research. Of particular interest is the radiation pressure acceleration, for which simulations predict narrow band ion energies up to GeV. We derive a laser-piston model by applying techniques for non-relativistic gas-dynamics. The model reveals a laser intensity limit, below which sufficient laser-piston acceleration is impossible. The relation between target thickness and piston velocity as a function of the laser pulse length yields an approximation for the permissible target thickness. We performed one-dimensional Particle-In-Cell simulations to confirm the predictions of the analytical model. These simulations also reveal the importance of electromagnetic energy transport. We find that this energy transport limits the achievable compression and rarefies the plasma.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2017

Magnetohydrodynamic motion of a two-fluid plasma

Here, the two-fluid Maxwell system couples frictionless electron and ion fluids via Maxwell’s equations. When the frequencies of light waves, Langmuir waves, and single-particle cyclotron motion are scaled to be asymptotically large, the two-fluid Maxwell system becomes a fast-slow dynamical system. This fast-slow system admits a formally-exact single-fluid closure that may be computed systematically with any desired order of accuracy through the use of a functional partial differential equation. In the leading order approximation, the closure reproduces magnetohydrodynamics (MHD). Higher order truncations of the closure give an infinite hierarchy of extended MHD models that allow for arbitrary mass ratio, as well as perturbative deviations from charge neutrality. The closure is interpreted geometrically as an invariant slow manifold in the infinite-dimensional two-fluid phase space, on which two-fluid motions are free of high-frequency oscillations. This perspective shows that the full closure inherits a Hamiltonian structure from two-fluid theory. By employing infinite-dimensional Lie transforms, the Poisson bracket for the all-orders closure may be obtained in closed form. Thus, conservative truncations of the single-fluid closure may be obtained by simply truncating the single-fluid Hamiltonian. Moreover, the closed-form expression for the all-orders bracket gives explicit expressions for a number of the full closure’s conservation laws. Notably, the full closure, as well as any of its Hamiltonian truncations, admits a pair of independent circulation invariants.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2018

Coupled electron-ion Monte Carlo simulation of hydrogen molecular crystals

We performed simulations for solid molecular hydrogen at high pressures (250 GPa ≤ P ≤ 500 GPa) along two isotherms at T = 200 K (phase III) and at T = 414 K (phase IV). At T = 200 K, we considered likely candidates for phase III, the C2c and Cmca12 structures, while at T = 414 K in phase IV, we studied the Pc48 structure. We employed both Coupled Electron-Ion Monte Carlo (CEIMC) and Path Integral Molecular Dynamics (PIMD). The latter is based on Density Functional Theory (DFT) with the van der Waals approximation (vdW-DF). The comparison between the two methods allows us to address the question of the accuracy of the exchange-correlation approximation of DFT for thermal and quantum protons without recurring to perturbation theories. In general, we find that atomic and molecular fluctuations in PIMD are larger than in CEIMC which suggests that the potential energy surface from vdW-DF is less structured than the one from quantum Monte Carlo. We find qualitatively different behaviors for systems prepared in the C2c structure for increasing pressure. Within PIMD, the C2c structure is dynamically partially stable for P ≤ 250 GPa only: it retains the symmetry of the molecular centers but not the molecular orientation; at intermediate pressures, it develops layered structures like Pbcn or Ibam and transforms to the metallic Cmca-4 structure at P ≥ 450 GPa. Instead, within CEIMC, the C2c structure is found to be dynamically stable at least up to 450 GPa; at increasing pressure, the molecular bond length increases and the nuclear correlation decreases. For the other two structures, the two methods are in qualitative agreement although quantitative differences remain. Finally, we discuss various structural properties and the electrical conductivity. Here, we find that these structures become conducting around 350 GPa but the metallic Drude-like behavior is reached only at around 500 GPa, consistent with recent experimental claims.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗