Search NASA⌕ Search

SEARCH · Search NASA

Results for “variational methods”

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 55 records · Page 3

Improved Variable-Transformation Algorithm for the Hexagonal Variational Nodal Method

Hexagonal Variational Nodal Method (VNM) has been widely employed in fast reactor core calculation codes. It provides the weak form solution of the second order transport equation. It employs a direct three-dimensional expansion instead of the transverse integration technique, with the singular term caused by transverse integration over the hexagonal nodal avoided. However, as the increasing Pn expansion order in angle, the number of moments increases dramatically, which significantly increases the requirements in both memory and computational time. As a consequence, Yang proposed the Variable-Transformation Algorithm. It firstly used the reverse Cuthill-McKee (RCMK) reordering method to transform the response matrices into block-diagonal. Then, for the unreducible matrices, orthogonal transformation was applied, after which the matrices were sparse enough to be reduced by using the RCMK reordering once more. In this summary, based on a newly developed hexagonal VNM code NECP-VioletHex, the Variable Transformation Algorithm was investigated and improved by adopting more constraints to the basis functions to further take advantage of the symmetric properties of the hexagonal-z nodes. (authors)

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Uncertainty Quantification in Wind Plant Energy Estimation: Preprint

Data characterizing a wind power plant's production is important for the validation, calibration, and refinement of the wake loss models used to simulate performance for pre-construction energy estimates. In this process, there is the potential for significant variation in the methods used by an analyst for, first, the preparation of the operational production data for comparison to model data and, second, the generation of analogous production data from simulations. In this work, the uncertainty introduced by these method variations is quantified by considering ensembles of analysis methods. Specifically, 1920 methods variations were developed for preparing the operational data as a predictable production benchmark and simulations taking as input sample sets of the observed wind resource ranged in size from covering 4% to 125% of a full year were run using 3 engineering wake loss models in FLORIS simulations. The benchmark and simulation input uncertainties place lower bounds on the precision of the calibration between the simulation model estimates and operational benchmark data. Even within subsets of method variations (representative of what might be seen within a single company), the uncertainty in the predictable expected production is estimated to be between 1% and 3% the median predictable expected production while the uncertainty in the simulated expected production remains at best on the order of 2%. To put this into context: the simulated expected production is very closely related to the annual energy production (AEP) metric commonly used to quantify a plant's predicted performance and a 3% error in the pre-construction AEP estimate can result in a $17M loss due to additional financing costs for a typical 100MW Texas plant.

17 WIND ENERGY↗

Observations on variational and projector Monte Carlo methods

Variational Monte Carlo and various projector Monte Carlo (PMC) methods are presented in a unified manner. Similarities and differences between the methods and choices made in designing the methods are discussed. Both methods where the Monte Carlo walk is performed in a discrete space and methods where it is performed in a continuous space are considered. It is pointed out that the usual prescription for importance sampling may not be advantageous depending on the particular quantum Monte Carlo method used and the observables of interest, so alternate prescriptions are presented. The nature of the sign problem is discussed for various versions of PMC methods. A prescription for an exact PMC method in real space, i.e., a method that does not make a fixed-node or similar approximation and does not have a finite basis error, is presented. This method is likely to be practical for systems with a small number of electrons. Approximate PMC methods that are applicable to larger systems and go beyond the fixed-node approximation are also discussed.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Analytical and variational numerical methods for unstable miscible displacement flows in porous media

The miscible displacement of one fluid by another in a porous medium has received considerable attention in subsurface, environmental and petroleum engineering applications. When a fluid of higher mobility displaces another of lower mobility, unstable patterns – referred to as viscous fingering – may arise. Their physical and mathematical study has been the object of numerous investigations over the past century. In writing this paper the objective is to present a review of these contributions with particular emphasis on variational methods. These algorithms are tailored to real field applications thanks to their advanced features: handling of general complex geometries, robustness in the presence of rough tensor coefficients, low sensitivity to mesh orientation in advection dominated scenarios, and provable convergence with fully unstructured grids.

42 ENGINEERING↗

Ultracold atoms at unitarity within quantum Monte Carlo methods

Variational and diffusion quantum Monte Carlo (VMC and DMC) calculations of the properties of the zero-temperature fermionic gas at unitarity are reported. Our study differs from earlier ones mainly in that we have constructed more accurate trial wave functions and used a larger system size, we have studied the dependence of the energy on the particle density and well width, and we have achieved much smaller statistical error bars. The correct value of the universal ratio of the energy of the interacting to that of the noninteracting gas, {xi}, is still a matter of debate. We find DMC values of {xi} of 0.4244(1) with 66 particles and 0.4339(1) with 128 particles. The spherically averaged pair-correlation functions, momentum densities, and one-body density matrices are very similar in VMC and DMC, which suggests that our results for these quantities are very accurate. We find, however, some differences between the VMC and DMC results for the two-body density matrices and condensate fractions, which indicates that these quantities are more sensitive to the quality of the trial wave function. Our best estimate of the condensate fraction of 0.51 is smaller than the values from earlier quantum Monte Carlo calculations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Solitary Waves of the MRLW Equation by Variational Iteration Method

In a recent publication, Soliman solved numerically the modified regularized long wave (MRLW) equation by using the variational iteration method (VIM). In this paper, corrected numerical results have been computed, plotted, tabulated, and compared with not only the exact analytical solutions but also the Adomian decomposition method results. Solitary wave solutions of the MRLW equation are exactly obtained as a convergent series with easily computable components. Propagation of single solitary wave, interaction of two and three waves, and also birth of solitons have been discussed. Three invariants of motion have been evaluated to determine the conservation properties of the problem.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Hylleraas-configuration-interaction nonrelativistic energies for the {sup 1}S ground states of the beryllium isoelectronic sequence

In a previous work, Sims and Hagstrom [“Hylleraas-configuration-interaction study of the 1 {sup 1}S ground state of neutral beryllium,” Phys. Rev. A 83, 032518 (2011)] reported Hylleraas-configuration-interaction (Hy-CI) method variational calculations for the {sup 1}S ground state of neutral beryllium with an estimated accuracy of a tenth of a microhartree. In this work, the calculations have been extended to higher accuracy and, by simple scaling of the orbital exponents, to the entire Be 2 {sup 1}S isoelectronic sequence. The best nonrelativistic energies for Be, B{sup +}, and C{sup ++} obtained are −14.6673 5649 269, −24.3488 8446 36, and −36.5348 5236 25 hartree, respectively. Except for Be, all computed nonrelativistic energies are superior to the known reference energies for these states.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Helium-4 gravitational form factors: Exchange currents

We evaluate the leading exchange corrections to the helium-4 gravitational form factors (GFFs) to momenta of the order of the nucleon mass. We use both the K-harmonic method with simple pair nucleon potential, and a Jastrow trial function using the Argonne 𝑣 14 potential, to evaluate the helium-4 GFFs. The exchange current contributions include the pair interaction, plus the seagull and the pion exchange interactions, modulo the recoil corrections. To estimate the off-shellness of the pion nucleon coupling in this momenta range, we discuss the results using either the pseudoscalar (PS) or pseudovector (PV) pion-nucleon couplings. When the PV coupling is used, the pair diagram contribution is higher order in the nonrelativistic expansion. The results for the helium-4 A-GFF are comparable to those given by the impulse approximation, especially for the PS coupling using both the K-harmonic method and variational method. The exchange current contributions with the PS coupling for the charge form factor of helium-4, yield better agreement with the existing data over a broad range of momenta, especially when the Argonne 𝑣 14 potential including the D-wave admixture is used.

A ≤ 5↗

A parallel in energy Krylov solution algorithm for the variational nodal method

To reduce the computational expense caused by the large number of energy groups in thermal and fast reactor calculations, we researched using parallelization in energy (i.e. group) of the multi-group response matrix equations derived from the Variational Nodal Method. This study focuses on solving steady-state eigenvalue problems in diffusion theory that are typically needed for depletion and reactivity coefficient calculations. To examine serial and parallel performance, GMRES, a Krylov algorithm, is applied to the multi-group system of equations at each fission source iteration. We also study the impact of using Wielandt acceleration on the eigenvalue/eigenvector problem. Numerical results based on a 72 group thermal and a 216 group fast reactor core problems are obtained for both serial and parallel calculations. The results show reduced CPU times when using GMRES and Wielandt acceleration in both serial and parallel calculations. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Effect of the duration of the growth process on the properties of GaN grown by the sublimation method

Variation in the structural and morphological features and luminescent characteristics of thick epitaxial GaN layers grown by the sublimation sandwich method with the duration of the crystallization process has been studied. This was, in particular, done by means of scanning electron microscopy in the secondary-electron and color-cathodoluminescence modes. It was found that rather high-quality GaN layers with a thickness of up to 0.5 mm can be grown in a time of about 1.5 h, with their surface hardly exhibiting any luminescence in the visible spectral range. However, making the growth process longer in order to obtain thicker layers impairs the quality of a crystal being grown, which is accompanied by an increase in the intensity of cathodoluminescence from its surface layer in the visible (predominantly yellow) region of the spectrum. Reasons for the poorer quality of GaN layers in this case are discussed. It is suggested that, as the evaporation rate from the source decreases, the amount of active nitrogen near the growth surface becomes lower.

36 MATERIALS SCIENCE↗

Implementation and validation of the variational nodal expansion method core simulator

The variational nodal expansion method (VNEM) has been developed by the OECD Halden Reactor Project for the purpose of implementation into a full-core simulator. This version of the VNEM does not make use of the diffusion approximation, but instead provides a fast converged solution to the even-parity neutron transport equation in several minutes using nodal methods and variational techniques. This 'method to simulator' implementation has been completed and the VNEM-simulator is currently undergoing continued validation efforts to ensure that it can track more nuclear reactor cycles effectively. This paper presents the results of the most recent development on the 3D VNEM core simulator implementation and validation efforts. The method is briefly described and data is presented and discussed for three cases for the initial criticality tests at the beginning of life, hot stand-by state for the Ringhals-3 PWR. Also presented here are the results for one year of operation of the maiden cycle of this unit. The effective neutron multiplication factor, k{sub eff}, and relative neutron detector readings are compared showing excellent agreement. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗