Search NASA⌕ Search

SEARCH · Search NASA

Results for “Functional 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 1,495 records · Page 83

Direct Calculation of the Scattering Amplitude Without Partial Wave Decomposition: Inclusion of Correlation Effects - III

In the first two papers in this series, we developed a method for studying electron-hydrogen scattering that does not use partial wave analysis. We constructed an ansatz for the wave function in both the static and static exchange approximations and calculated the full scattering amplitude. Here we go beyond the static exchange approximation, and include correlation in the wave function via a modified polarized orbital. This correlation function provides a significant improvement over the static exchange approximation: the resultant elastic scattering amplitudes are in very good agreement with fully converged partial wave calculations for electron-hydrogen scattering. A fully variational modification of this approach is discussed in the conclusion of the article Popular summary of Direct calculation of the scattering amplitude without partial wave expansion. III ....." by J. Shertzer and A. Temkin. In this paper we continue the development of In this paper we continue the development of a new approach to the way in which researchers have traditionally used to calculate the scattering cross section of (low-energy) electrons from atoms. The basic mathematical problem is to solve the Schroedinger Equation (SE) corresponding the above physical process. Traditionally it was always the case that the SE was reduced to a sequence of one-dimensional (ordinary) differential equations - called partial waves which were solved and from the solutions "phase shifts" were extracted, from which the scattering cross section was calculated.

Shertzer, Janine↗

JOVIAN STRATOSPHERE AS A CHEMICAL TRANSPORT SYSTEM: BENCHMARK ANALYTICAL SOLUTIONS

We systematically investigated the solvable analytical benchmark cases in both one- and two-dimensional (1D and 2D) chemical-advective-diffusive systems. We use the stratosphere of Jupiter as an example but the results can be applied to other planetary atmospheres and exoplanetary atmospheres. In the 1D system, we show that CH{sub 4} and C{sub 2}H{sub 6} are mainly in diffusive equilibrium, and the C{sub 2}H{sub 2} profile can be approximated by modified Bessel functions. In the 2D system in the meridional plane, analytical solutions for two typical circulation patterns are derived. Simple tracer transport modeling demonstrates that the distribution of a short-lived species (such as C{sub 2}H{sub 2}) is dominated by the local chemical sources and sinks, while that of a long-lived species (such as C{sub 2}H{sub 6}) is significantly influenced by the circulation pattern. We find that an equator-to-pole circulation could qualitatively explain the Cassini observations, but a pure diffusive transport process could not. For slowly rotating planets like the close-in extrasolar planets, the interaction between the advection by the zonal wind and chemistry might cause a phase lag between the final tracer distribution and the original source distribution. The numerical simulation results from the 2D Caltech/JPL chemistry-transport model agree well with the analytical solutions for various cases.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Dynamics of polymers: A mean-field theory

We derive a general mean-field theory of inhomogeneous polymer dynamics; a theory whose form has been speculated and widely applied, but not heretofore derived. Our approach involves a functional integral representation of a Martin-Siggia-Rose (MSR) type description of the exact many-chain dynamics. A saddle point approximation to the generating functional, involving conditions where the MSR action is stationary with respect to a collective density field ρ and a conjugate MSR response field ϕ, produces the desired dynamical mean-field theory. Besides clarifying the proper structure of mean-field theory out of equilibrium, our results have implications for numerical studies of polymer dynamics involving hybrid particle-field simulation techniques such as the single-chain in mean-field method.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A recipe for free-energy functionals of polarizable molecular fluids

Classical density-functional theory is the most direct approach to equilibrium structures and free energies of inhomogeneous liquids, but requires the construction of an approximate free-energy functional for each liquid of interest. We present a general recipe for constructing functionals for small-molecular liquids based only on bulk experimental properties and ab initio calculations of a single solvent molecule. This recipe combines the exact free energy of the non-interacting system with fundamental measure theory for the repulsive contribution and a weighted density functional for the short-ranged attractive interactions. We add to these ingredients a weighted polarization functional for the long-range correlations in both the rotational and molecular-polarizability contributions to the dielectric response. We also perform molecular dynamics calculations for the free energy of cavity formation and the high-field dielectric response, and show that our free-energy functional adequately describes these properties (which are key for accurate solvation calculations) for all three solvents in our study: water, chloroform, and carbon tetrachloride.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

The Physics and Chemistry of Cluster-based Catalyst Systems

In this project we used first principles calculations to investigate the physical and chemical properties of atomic clusters, small clumps of matter containing up to a few hundred atoms. The calculations are based on density functional theory, the most efficient and accurate method available for computing and thereby predicting the properties of molecules, clusters, and solids. An initial focus of our work involved studying the reactivity of small Pd clusters. We demonstrated that H2 molecules readily dissociate on Pd clusters and that multiple dissociations can occur on clusters larger than a few Pd atoms. For example, up to eight H2 molecules will dissociate on Pd13 at room temperature. A second focus of the project was to study the evolution of the electrical behavior of clusters, in particular, the evolution of metallic behavior in clusters of metal atoms. We developed a technique for using the electric dipole polarizability of the clusters – the response of the cluster to an applied electric field -- as a probe of this behavior. By decomposing the polarizability into so-called charge transfer (or metallic) and local (or insulating) contributions, we obtained the remarkable result that the per atom charge transfer contributions in sodium clusters attain the same value as occurs in bulk sodium metal for clusters containing as few as 10 atoms. Our calculated total polarizabilities matched experimentally measured trends essentially perfectly, validating the reliability of the calculations and supporting our conclusion that tiny clusters already display bulk-like metallic character. Similar results were obtained for other metal atom clusters, including potassium, aluminum, and zinc. We discovered that the evolution toward bulk metallic behavior for each of these cluster species is essentially identical, when judged using these polarizability-based descriptors. This suggests a universality in the approach of metal clusters to bulk-like metallicity. This discovery is a major outcome of our project. We also used our polarizability-based metrics to investigate the evolution of the properties of Si and NaCl clusters. These systems are non-metallic in the bulk limit. For Si, we find the behavior of small clusters to be metal-like and we attribute this to bonding arrangements in the clusters that are very different from the open structure of crystalline silicon. As the clusters get larger, the bonding in the cluster interior becomes more and more like that of the bulk material. We expect a transition from metal-like to insulator-like behavior to accompany this change is structure. Our results point to a transition at sizes beyond 146 atoms, the largest size treated in our study. We hope to study this transition further in the future. A third thrust of our project involved work on a new approach to eliminating unphysical electron self-interaction from density functional theory calculations. Two electrons exert repulsive forces on each other, but a single electron should not interact with itself. Because of approximations made in density functional theory, individual electrons effectively interact with themselves. The effect of this self-interaction is often small, especially when atoms are near their equilibrium positions in molecules. But the effects can be pronounced when bonds are stretched and also in some situations involving transition metals even when the atoms are near equilibrium. Our research involved the further development of a method for removing self-interaction called the Fermi-Löwdin orbital self-interaction correction or FLO-SIC. We conducted the first systematic FLO-SIC calculations for the individual atoms ranging from Li – Kr. We also carried out a critical comparison of the results of FLO-SIC with an older method known as the Perdew-Zunger self-interaction correction (PZ-SIC). FLO-SIC can be thought of as a modified approach to PZ-SIC that is expected to be computationally more efficient. We demonstrated that the two methods are distinct, but give essentially the same results for quantities such as molecular binding energies or electron removal energies. We are continuing to develop the FLO-SIC methodology with a goal of creating a computationally efficient method that is free of self-interaction effects.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Sensitivity analysis and approximation methods for general eigenvalue problems

Optimization of dynamic systems involving complex non-hermitian matrices is often computationally expensive. Major contributors to the computational expense are the sensitivity analysis and reanalysis of a modified design. The present work seeks to alleviate this computational burden by identifying efficient sensitivity analysis and approximate reanalysis methods. For the algebraic eigenvalue problem involving non-hermitian matrices, algorithms for sensitivity analysis and approximate reanalysis are classified, compared and evaluated for efficiency and accuracy. Proper eigenvector normalization is discussed. An improved method for calculating derivatives of eigenvectors is proposed based on a more rational normalization condition and taking advantage of matrix sparsity. Important numerical aspects of this method are also discussed. To alleviate the problem of reanalysis, various approximation methods for eigenvalues are proposed and evaluated. Linear and quadratic approximations are based directly on the Taylor series. Several approximation methods are developed based on the generalized Rayleigh quotient for the eigenvalue problem. Approximation methods based on trace theorem give high accuracy without needing any derivatives. Operation counts for the computation of the approximations are given. General recommendations are made for the selection of appropriate approximation technique as a function of the matrix size, number of design variables, number of eigenvalues of interest and the number of design points at which approximation is sought.

Murthy, D. V.↗

Deducing spectroscopic factors from wave-function asymptotics

In a coupled-channel model, we explore the effects of coupling between configurations on the radial behavior of the wave function and, in particular, on the spectroscopic factor (SF) and the asymptotic normalization coefficient (ANC). We evaluate the extraction of a SF from the ratio of the ANC of the coupled-channel model to that of a single-particle approximation of the wave function. We perform this study within a core+n collective model, which includes two states of the core that connect by a rotational coupling. To get additional insights, we also use a simplified model that takes a {delta} function for the coupling potential. Calculations are performed for {sup 11}Be. Fair agreement is obtained between the SF inferred from the single-particle approximation and the one obtained within the coupled-channel models. Significant discrepancies are observed only for large coupling strength and/or large admixture, that is, a small SF. This suggests that reliable SFs can be deduced from the wave-function asymptotics when the structure is dominated by one configuration, that is, for a large SF.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Parallelized POD-based suboptimal economic model predictive control of a state-constrained Boussinesq approximation

Motivated by an energy efficient building application, we want to optimize a quadratic cost functional subject to the Boussinesq approximation of the Navier-Stokes equations and to bilateral state and control constraints. Since the computation of such an optimal solution is numerically costly, we design an efficient strategy to compute a sub-optimal (but applicationally acceptable) solution with significantly reduced computational effort. We employ an economic Model Predictive Control (MPC) strategy to obtain a feedback control. The MPC sub-problems are based on a linear-quadratic optimal control problem subjected to mixed control and state constraints and a convection-diffusion equation, reduced with proper orthogonal decomposition. Finally, to solve each sub-problem, we apply a primal-dual active set strategy. The method can be fully parallelized, which enables the solution of large problems with real-world parameters.

97 MATHEMATICS AND COMPUTING↗

Accurate atomic quantum defects from particle–particle random phase approximation

The accuracy of calculations of atomic Rydberg excitations cannot be judged by the usual measures, such as mean unsigned errors of many transitions. We show how to use quantum defect (QD) theory to (a) separate errors due to approximate ionisation potentials, (b) extract smooth QDs to compare with experiment, and (c) quantify those defects with a few characteristic parameters. The particle–particle random phase approximation (pp-RPA) produces excellent Rydberg transitions that are an order of magnitude more accurate than those of time-dependent density functional theory with standard approximations. Here, we even extract reasonably accurate defects from the lithium Rydberg series, despite the reference being open-shell. Our methodology can be applied to any Rydberg series of excitations with four transitions or more to extract the underlying threshold energy and characteristic QD parameters. Our pp-RPA results set a demanding challenge for other excitation methods to match.

74 ATOMIC AND MOLECULAR PHYSICS↗

Bosonic field digitization for quantum computers

Quantum simulation of quantum field theory is a flagship application of quantum computers that promises to deliver capabilities beyond classical computing. The realization of quantum advantage will require methods that can accurately predict error scaling as a function of the resolution and parameters of the model and that can be implemented efficiently on quantum hardware. In this paper, we address the representation of lattice bosonic fields in a discretized field amplitude basis, develop methods to predict error scaling, and present efficient qubit implementation strategies. A low-energy subspace of the bosonic Hilbert space, defined by a boson occupation number cutoff, can be represented with exponentially good accuracy by a low-energy subspace of a finite-size Hilbert space. The finite representation construction and the associated errors are directly related to the accuracy of the Nyquist-Shannon sampling and the finite Fourier transforms of the boson number states in the field and the conjugate-field bases. We analyze the relation between the boson mass, the discretization parameters used for wave function sampling, and the finite representation size. Numerical simulations of small size Φ 4 problems demonstrate that the boson mass optimizing the sampling of the ground state wave function is a good approximation to the optimal boson mass yielding the minimum low-energy subspace size. However, we find that accurate sampling of general wave functions does not necessarily result in accurate representation. Finally, we develop methods for validating and adjusting the discretization parameters to achieve more accurate simulations.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

TU-F-CAMPUS-I-05: Parameterization of the Noise Power Spectrum in X-Ray Computed Tomography

Purpose: The purpose of this work was to develop a method so that the noise power spectrum (NPS) can be approximated for arbitrary levels of mAs, from a single determination in CT. Methods: The NPS is factorized into 2 components, 1) a parameterized function representing the 1D normalized spatial frequency distribution and 2) a function to scale the magnitude of 1) for arbitrary values of mAs. The 1D NPS, normalized by image variance (NNPS), was determined for 2 FBP reconstruction kernels (smoothing and edge enhancing) for 400 mAs. The NNPS were fit to the parameterized function and a scaling function was established to approximate the image variance at arbitrary values of mAs. Using the root mean square error normalized by the maximum value (NRMSE), the NPS approximated with the factorization method was compared to the NPS determined at 5 different mAs levels. Results: The factorization resulted in a set of 7 coefficients that can be used to approximate the 1D NPS, for arbitrary levels of mAs, for the convolution kernels studied in this work. The approximated NPS (factorization) agreed well with the determined NPS for all mAs levels. The greatest NRMSE was 0.02 and was observed for the edge enhancing kernel. Conclusion: The proposed factorization method has been demonstrated as applicable for FBP reconstruction. It can be used to approximate the 1D NPS for arbitrary levels of mAs, from a single NPS determination. Furthermore, approximations of the 1D NPS can conveniently be distributed since the factorization method only used 7 coefficients in the approximation.

60 APPLIED LIFE SCIENCES↗

Exchange-correlation magnetic fields in spin-density-functional theory

In spin-density-functional theory for noncollinear magnetic materials, the Kohn-Sham system features exchange-correlation (xc) scalar potentials and magnetic fields. The significance of the xc magnetic fields is not very well explored; in particular, they can give rise to local torques on the magnetization, which are absent in standard local and semilocal approximations. We obtain exact benchmark solutions for two electrons on four-site extended Hubbard lattices over a wide range of interaction strengths, and compare exact xc potentials and magnetic fields with approximations obtained from orbital-dependent xc functionals. Here, the xc magnetic fields turn out to play an increasingly important role as systems becomes more and more correlated and the electrons begin to localize; the effects of the xc torques, however, remain relatively minor. The approximate xc functionals perform overall quite well, but tend to favor symmetry-broken solutions for strong interactions.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

A Multi-Resolution Nonlinear Mapping Technique for Design and Analysis Application

This report describes a nonlinear mapping technique where the unknown static or dynamic system is approximated by a sum of dimensionally increasing functions (one-dimensional curves, two-dimensional surfaces, etc.). These lower dimensional functions are synthesized from a set of multi-resolution basis functions, where the resolutions specify the level of details at which the nonlinear system is approximated. The basis functions also cause the parameter estimation step to become linear. This feature is taken advantage of to derive a systematic procedure to determine and eliminate basis functions that are less significant for the particular system under identification. The number of unknown parameters that must be estimated is thus reduced and compact models obtained. The lower dimensional functions (identified curves and surfaces) permit a kind of "visualization" into the complexity of the nonlinearity itself.

Phan, Minh Q.↗

A Multi-Resolution Nonlinear Mapping Technique for Design and Analysis Applications

This report describes a nonlinear mapping technique where the unknown static or dynamic system is approximated by a sum of dimensionally increasing functions (one-dimensional curves, two-dimensional surfaces, etc.). These lower dimensional functions are synthesized from a set of multi-resolution basis functions, where the resolutions specify the level of details at which the nonlinear system is approximated. The basis functions also cause the parameter estimation step to become linear. This feature is taken advantage of to derive a systematic procedure to determine and eliminate basis functions that are less significant for the particular system under identification. The number of unknown parameters that must be estimated is thus reduced and compact models obtained. The lower dimensional functions (identified curves and surfaces) permit a kind of "visualization" into the complexity of the nonlinearity itself.

Phan, Minh Q.↗

Kernel learning backward SDE filter for data assimilation

In this paper, we develop a kernel learning backward SDE filter method to estimate the state of a stochastic dynamical system based on its partial noisy observations. A system of forward backward stochastic differential equations is used to propagate the state of the target dynamical model, and Bayesian inference is applied to incorporate the observational information. Further, to characterize the dynamical model in the entire state space, we introduce a kernel learning method to learn a continuous global approximation for the conditional probability density function of the target state by using discrete approximated density values as training data. Numerical experiments demonstrate that the kernel learning backward SDE is highly effective.

97 MATHEMATICS AND COMPUTING↗

Time-dependent potential-functional embedding theory

We introduce a time-dependent potential-functional embedding theory (TD-PFET), in which atoms are grouped into subsystems. In TD-PFET, subsystems can be propagated by different suitable time-dependent quantum mechanical methods and their interactions can be treated in a seamless, first-principles manner. TD-PFET is formulated based on the time-dependent quantum mechanics variational principle. The action of the total quantum system is written as a functional of the time-dependent embedding potential, i.e., a potential-functional formulation. By exploiting the Runge-Gross theorem, we prove the uniqueness of the time-dependent embedding potential under the constraint that all subsystems share a common embedding potential. We derive the integral equation that such an embedding potential needs to satisfy. As proof-of-principle, we demonstrate TD-PFET for a Na{sub 4} cluster, in which each Na atom is treated as one subsystem and propagated by time-dependent Kohn-Sham density functional theory (TDDFT) using the adiabatic local density approximation (ALDA). Our results agree well with a direct TDDFT calculation on the whole Na{sub 4} cluster using ALDA. We envision that TD-PFET will ultimately be useful for studying ultrafast quantum dynamics in condensed matter, where key regions are solved by highly accurate time-dependent quantum mechanics methods, and unimportant regions are solved by faster, less accurate methods.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Understanding ferromagnetism and optical absorption in 3d transition metal-doped cubic ZrO{sub 2} with the modified Becke-Johnson exchange-correlation functional

The electronic structure, magnetic, and optical properties in cubic crystalline phase of Zr{sub 1−x}TM{sub x}O{sub 2} (TM = V, Mn, Fe, and Co) at x = 6.25% are studied using density functional theory with the Generalized Gradient Approximation and the modified Becke-Johnson of the exchange-correlation energy and potential. In our calculations, the zirconia is a p-type semiconductor and has a large band gap. We evaluated the possibility of long-range magnetic order for transition metal ions substituting Zr. Our results show that ferromagnetism is the ground state in V, Mn, and Fe-doped ZrO{sub 2} and have a high value of energy in Mn-doped ZrO{sub 2}. However, in Co-doped ZrO{sub 2}, antiferromagnetic ordering is more stable than the ferromagnetic one. The exchange interaction mechanism has been discussed to explain the responsible of this stability. Moreover, it has been found that the V, Mn, and Fe transition metals provide half-metallic properties considered to be the leading cause, responsible for ferromagnetism. Furthermore, the optical absorption spectra in the TM -doped cubic ZrO{sub 2} are investigated.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Should muffin tin radius vary in different structures of a material?: A case study

Quantum mechanical calculations based on density functional theory and a generalized gradient approximation (GGA) have been used to study the structural properties of YbN. Its predicted unit cell lattice parameter in NaCl (B1) structure is 4.7810Å and in CsCl (B2) structure it is 2.8685Å. In the determination of lattice parameter the muffin tin radius (R{sub MT}) of constituent atoms play important role. In both the structures the muffin tin radius for Yb and N converges to 2.3 and 1.4 a.u., respectively.

36 MATERIALS SCIENCE↗