Search NASA⌕ Search

SEARCH · Search NASA

Results for “Mathematics and Computing, Physics”

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 451 records · Page 25

Stochastic Eulerian Lagrangian methods for fluid-structure interactions with thermal fluctuations

We present approaches for the study of fluid-structure interactions subject to thermal fluctuations. A mixed mechanical description is utilized combining Eulerian and Lagrangian reference frames. We establish general conditions for operators coupling these descriptions. Stochastic driving fields for the formalism are derived using principles from statistical mechanics. The stochastic differential equations of the formalism are found to exhibit significant stiffness in some physical regimes. To cope with this issue, we derive reduced stochastic differential equations for several physical regimes. We also present stochastic numerical methods for each regime to approximate the fluid-structure dynamics and to generate efficiently the required stochastic driving fields. To validate the methodology in each regime, we perform analysis of the invariant probability distribution of the stochastic dynamics of the fluid-structure formalism. We compare this analysis with results from statistical mechanics. To further demonstrate the applicability of the methodology, we perform computational studies for spherical particles having translational and rotational degrees of freedom. We compare these studies with results from fluid mechanics. The presented approach provides for fluid-structure systems a set of rather general computational methods for treating consistently structure mechanics, hydrodynamic coupling, and thermal fluctuations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Cross-scale Efficient Tensor Contractions for Coupled Cluster Computations Through Multiple Programming Model Backends

Coupled-cluster methods provide highly accurate models of molecular structure by explicit numerical calculation of tensors representing the correlation between electrons. These calculations are dominated by a sequence of tensor contractions, motivating the development of numerical libraries for such operations. While based on matrix-matrix multiplication, these libraries are specialized to exploit symmetries in the molecular structure and in electronic interactions, and thus reduce the size of the tensor representation and the complexity of contractions. The resulting algorithms are irregular and their parallelization has been previously achieved via the use of dynamic scheduling or specialized data decompositions. We introduce our efforts to extend the Libtensor framework to work in the distributed memory environment in a scalable and energy efficient manner. We achieve up to 240 speedup compared with the best optimized shared memory implementation. We attain scalability to hundreds of thousands of compute cores on three distributed-memory architectures, (Cray XC30&XC40, BlueGene/Q), and on a heterogeneous GPU-CPU system (Cray XK7). As the bottlenecks shift from being compute-bound DGEMM's to communication-bound collectives as the size of the molecular system scales, we adopt two radically different parallelization approaches for handling load-imbalance. Nevertheless, we preserve a uni ed interface to both programming models to maintain the productivity of computational quantum chemists.

74 ATOMIC AND MOLECULAR PHYSICS↗

Separability of Mesh Bias and Parametric Uncertainty for a Full System Thermal Analysis

When making computational simulation predictions of multi-physics engineering systems, sources of uncertainty in the prediction need to be acknowledged and included in the analysis within the current paradigm of striving for simulation credibility. A thermal analysis of an aerospace geometry was performed at Sandia National Laboratories. For this analysis a verification, validation and uncertainty quantification workflow provided structure for the analysis, resulting in the quantification of significant uncertainty sources including spatial numerical error and material property parametric uncertainty. It was hypothesized that the parametric uncertainty and numerical errors were independent and separable for this application. This hypothesis was supported by performing uncertainty quantification simulations at multiple mesh resolutions, while being limited by resources to minimize the number of medium and high resolution simulations. Based on this supported hypothesis, a prediction including parametric uncertainty and a systematic mesh bias are used to make a margin assessment that avoids unnecessary uncertainty obscuring the results and optimizes computing resources.

42 ENGINEERING↗

Interface- and discontinuity-aware numerical schemes for plasma 3-T radiation diffusion in two and three dimensions

A set of numerical schemes is developed for two- and three-dimensional time-dependent 3-T radiation diffusion equations in systems involving multi-materials. To resolve sub-cell structure, interface reconstruction is implemented within any cell that has more than one material. Therefore, the system of 3-T radiation diffusion equations is solved on two- and three-dimensional polyhedral meshes. The focus of the development is on the fully coupling between radiation and material, the treatment of nonlinearity in the equations, i.e., in the diffusion terms and source terms, treatment of the discontinuity across cell interfaces in material properties, the formulations for both transient and steady states, the property for large time steps, and second order accuracy in both space and time. The discontinuity of material properties between different materials is correctly treated based on the governing physics principle for general polyhedral meshes and full nonlinearity. The treatment is exact for arbitrarily strong discontinuity. The scheme is fully nonlinear for the full nonlinearity in the 3-T diffusion equations. Three temperatures are fully coupled and are updated simultaneously. The scheme is general in two and three dimensions on general polyhedral meshes. The features of the scheme are demonstrated through numerical examples for transient problems and steady states. The effects of some simplifications of numerical schemes are also shown through numerical examples, such as linearization, simple average of diffusion coefficient, and approximate treatment for the coupling between radiation and material.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Learning functional priors and posteriors from data and physics

In this work, we develop a new Bayesian framework based on deep neural networks to be able to extrapolate in space-time using historical data and to quantify uncertainties arising from both noisy and gappy data in physical problems. Specifically, the proposed approach has two stages: (1) prior learning and (2) posterior estimation. At the first stage, we employ the physics-informed Generative Adversarial Networks (PI-GAN) to learn a functional prior either from a prescribed function distribution, e.g., Gaussian process, or from historical data and physics. At the second stage, we employ the Hamiltonian Monte Carlo (HMC) method to estimate the posterior in the latent space of PI-GANs. In addition, we use two different approaches to encode the physics: (1) automatic differentiation, used in the physicsinformed neural networks (PINNs) for scenarios with explicitly known partial differential equations (PDEs), and (2) operator regression using the deep operator network (DeepONet) for PDE-agnostic scenarios. We then test the proposed method for (1) meta-learning for one-dimensional regression, and forward/inverse PDE problems (combined with PINNs); (2) PDE-agnostic physical problems (combined with DeepONet), e.g., fractional diffusion as well as saturated stochastic (100-dimensional) flows in heterogeneous porous media; and (3) spatial-temporal regression problems, i.e., inference of a marine riser displacement field using experimental data from the Norwegian Deepwater Programme (NDP). The results demonstrate that the proposed approach can provide accurate predictions as well as uncertainty quantification given very limited scattered and noisy data, since historical data could be available to provide informative priors. In summary, the proposed method is capable of learning flexible functional priors, e.g., both Gaussian and non-Gaussian process, and can be readily extended to big data problems by enabling mini-batch training using stochastic HMC or normalizing flows since the latent space is generally characterized as low dimensional.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Two-dimensional Imaging Velocity Interferometry: Technique and Data Analysis

We describe the data analysis procedures for an emerging interferometric technique for measuring motion across a two-dimensional image at a moment in time, i.e. a snapshot 2d-VISAR. Velocity interferometers (VISAR) measuring target motion to high precision have been an important diagnostic in shockwave physics for many years Until recently, this diagnostic has been limited to measuring motion at points or lines across a target. We introduce an emerging interferometric technique for measuring motion across a two-dimensional image, which could be called a snapshot 2d-VISAR. If a sufficiently fast movie camera technology existed, it could be placed behind a traditional VISAR optical system and record a 2d image vs time. But since that technology is not yet available, we use a CCD detector to record a single 2d image, with the pulsed nature of the illumination providing the time resolution. Consequently, since we are using pulsed illumination having a coherence length shorter than the VISAR interferometer delay ({approx}0.1 ns), we must use the white light velocimetry configuration to produce fringes with significant visibility. In this scheme, two interferometers (illuminating, detecting) having nearly identical delays are used in series, with one before the target and one after. This produces fringes with at most 50% visibility, but otherwise has the same fringe shift per target motion of a traditional VISAR. The 2d-VISAR observes a new world of information about shock behavior not readily accessible by traditional point or 1d-VISARS, simultaneously providing both a velocity map and an 'ordinary' snapshot photograph of the target. The 2d-VISAR has been used to observe nonuniformities in NIF related targets (polycrystalline diamond, Be), and in Si and Al.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Criticality safety enhancements for SCALE 6.2 and beyond

SCALE is a widely used suite of tools for nuclear systems modeling and simulation that provides comprehensive, verified and validated, user-friendly capabilities for criticality safety, reactor physics, radiation shielding, and sensitivity and uncertainty analysis. Since 1980, regulators, industry, and research institutions around the world have relied on SCALE for nuclear safety analysis and design. SCALE 6.2 provides several new capabilities and significant improvements in many existing features for criticality safety analysis. Enhancements are realized for nuclear data; multigroup resonance self-shielding; continuous-energy Monte Carlo analysis for sensitivity/uncertainty analysis, radiation shielding, and depletion; and graphical user interfaces. An overview of these capabilities is provided in this paper, and additional details are provided in several companion papers.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Comparing Measured (a,n) Thick Target Yields to MCNP6 using TENDL 2012 Libraries - Paper 31

MCNP6, due to the heritage of MCNPX, has always possessed the capability to transport the light ions, such as alpha particles. Two types of competing physics take place during the particle transport: atomic and nuclear physics. Charged particles lose energy to media due to Coulombic interactions. MCNP6 models Coulombic interactions through the use of the continuous slowing down approximation. Stopping powers are calculated at a variety of energies using the Bethe-Bloch formula with Sternheimer and Peierls density effect correction. For energies less than 1.31 MeV/AMU, a low-energy SPAR model is used. When the energy of the particle is high enough, there is a transition from nuclear interactions to nucleon interactions. MCNP6 possess various physics models to account for these types of intra-nuclear cascades. These models are less accurate at low energies. Even though MCNP6 will transport these light ions to low energies (i.e. less than 20 MeV), most of the available model physics in MCNP6, for nuclear interactions, has a suggested minimum energy of ∼20-200 MeV. The result of the poorer prediction at low energies causes important light ion events, such as F-19(a,n)Na-22 from U-234 decay in UF{sub 6}, to produce orders of magnitude fewer neutrons in MCNP6 than should be emitted. As with neutrons, ACE (A Compact ENDF) formatted light ion libraries have been developed from several different evaluations. These libraries possess better physics at low energies where the physics models break down. The TENDL 2012 evaluations possess alpha transport libraries, in ACE format, such that they can easily be used with codes like MCNP6 that read ACE formatted cross sections. As a result, a capability was developed for MCNP6 to read and implement these library cross sections when the light ion reached energies below the library upper energy cutoff. Perfectly modelling an experiment for benchmark comparisons to a code calculation can be tricky. For complex setups, systematic error is always possible; and usually a hefty uncertainty analysis is required to quantify just the geometry setup and material combination choices. For an (a,n) interaction, a thick target yield is defined as the amount of neutrons emitted given that the alpha ranges out in the material. For alpha transport at low energy <10 MeV, in reasonably dense materials (>1-2 g/cc), with a material dimension length larger than a few hundred microns, alpha particles will tend to completely stop and range out in the material. For Li, Be, B, C, N, O, F, Na, Mg and Al thick target yields exist in 200 keV intervals from 1 to 9.8 MeV (most of the elements do have (a,n) thresholds above 1 MeV). Because modelling a thick target yield in MCNP6 is much easier than attempting to account for the complexity of systematic error from a full out experiment, we present here comparisons of (a,n) thick target yields to those calculated in MCNP6 using the TENDL 2012 alpha library evaluations.

61 RADIATION PROTECTION AND DOSIMETRY↗

Review of Hybrid (Deterministic/Monte Carlo) Radiation Transport Methods, Codes, and Applications at Oak Ridge National Laboratory

This paper provides a review of the hybrid (Monte Carlo/deterministic) radiation transport methods and codes used at the Oak Ridge National Laboratory and examples of their application for increasing the efficiency of real-world, fixed-source Monte Carlo analyses. The two principal hybrid methods are (1) Consistent Adjoint Driven Importance Sampling (CADIS) for optimization of a localized detector (tally) region (e.g., flux, dose, or reaction rate at a particular location) and (2) Forward Weighted CADIS (FW-CADIS) for optimizing distributions (e.g., mesh tallies over all or part of the problem space) or multiple localized detector regions (e.g., simultaneous optimization of two or more localized tally regions). The two methods have been implemented and automated in both the MAVRIC sequence of SCALE 6 and ADVANTG, a code that works with the MCNP code. As implemented, the methods utilize the results of approximate, fast-running 3-D discrete ordinates transport calculations (with the Denovo code) to generate consistent space- and energy-dependent source and transport (weight windows) biasing parameters. These methods and codes have been applied to many relevant and challenging problems, including calculations of PWR ex-core thermal detector response, dose rates throughout an entire PWR facility, site boundary dose from arrays of commercial spent fuel storage casks, radiation fields for criticality accident alarm system placement, and detector response for special nuclear material detection scenarios and nuclear well-logging tools. Substantial computational speed-ups, generally O(102-4), have been realized for all applications to date. This paper provides a brief review of the methods, their implementation, results of their application, and current development activities, as well as a considerable list of references for readers seeking more information about the methods and/or their applications.

97 MATHEMATICS AND COMPUTING↗

Subcritical Multiplication with a Fixed Source

In a subcritical, multiplying medium, the system multiplication describes the expected total number of neutrons created by a single source neutron. Subcriticality plays a large role in criticality safety and thus it is vital for the subcritical multiplication factor be accurate, especially as a system approaches criticality. This work examines the accuracy of calculating the system multiplication using the MCNP6.2 ® k-eigenvalue power iteration (KCODE) method when a fixed-point source is present in a multiplying medium, for near critical systems. This work compares the standard approach for calculating system multiplication, using the fixed-source calculational approach, to a new, single k-eigenvalue power iteration approach that incorporates a fixed-source component and a fission-source component into a single calculation. For the remainder of this paper, some theoretical background and numerical results for an approximate k eigenvalue approach, an accurate fixed-source approach and a new and more accurate k-eigenvalue approach to computing system multiplication are provided.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Adjoint-Based Constant-Mass Partial Derivatives

It is well known that the adjoint equation in transport theory provides an economical means of computing derivatives of a system response to system parameters. It is also well known, though not often remarked on, that adjoint-based partial derivatives with respect to mass density are constant-volume derivatives. Likewise, adjoint-based partial derivatives with respect to dimensions (i.e., internal interface locations and the outer system boundary) are constant-density derivatives. In this paper, we derive a formula for the constant-mass partial derivative of a response with respect to internal interface locations and the outer system boundary. The formula uses the partial derivatives for density and dimensions, which may be adjoint-based. This paper includes numerical comparisons for a two-dimensional (r-z) cylinder that also demonstrate the correctness of the regular adjoint-based derivatives for surface locations. Adjoint-based derivatives for the neutron problems were computed using the new SENSMG code at Los Alamos National Laboratory.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Entropy distance: New quantum phenomena

We study a curve of Gibbsian families of complex 3 Multiplication-Sign 3-matrices and point out new features, absent in commutative finite-dimensional algebras: a discontinuous maximum-entropy inference, a discontinuous entropy distance, and non-exposed faces of the mean value set. We analyze these problems from various aspects including convex geometry, topology, and information geometry. This research is motivated by a theory of infomax principles, where we contribute by computing first order optimality conditions of the entropy distance.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Using heteroclinic orbits to quantify topological entropy in fluid flows

Topological approaches to mixing are important tools to understand chaotic fluid flows, ranging from oceanic transport to the design of micro-mixers. Typically, topological entropy, the exponential growth rate of material lines, is used to quantify topological mixing. Computing topological entropy from the direct stretching rate is computationally expensive and sheds little light on the source of the mixing. Earlier approaches emphasized that topological entropy could be viewed as generated by the braiding of virtual, or “ghost,” rods stirring the fluid in a periodic manner. Here, we demonstrate that topological entropy can also be viewed as generated by the braiding of ghost rods following heteroclinic orbits instead. We use the machinery of homotopic lobe dynamics, which extracts symbolic dynamics from finite-length pieces of stable and unstable manifolds attached to fixed points of the fluid flow. As an example, we focus on the topological entropy of a bounded, chaotic, two-dimensional, double-vortex cavity flow. Over a certain parameter range, the topological entropy is primarily due to the braiding of a period-three orbit. However, this orbit does not explain the topological entropy for parameter values where it does not exist, nor does it explain the excess of topological entropy for the entire range of its existence. We show that braiding by heteroclinic orbits provides an accurate computation of topological entropy when the period-three orbit does not exist, and that it provides an explanation for some of the excess topological entropy when the period-three orbit does exist. Furthermore, the computation of symbolic dynamics using heteroclinic orbits has been automated and can be used to compute topological entropy for a general 2D fluid flow.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Computing the optimal path in stochastic dynamical systems

In stochastic systems, one is often interested in finding the optimal path that maximizes the probability of escape from a metastable state or of switching between metastable states. Even for simple systems, it may be impossible to find an analytic form of the optimal path, and in high-dimensional systems, this is almost always the case. In this article, we formulate a constructive methodology that is used to compute the optimal path numerically. The method utilizes finite-time Lyapunov exponents, statistical selection criteria, and a Newton-based iterative minimizing scheme. The method is applied to four examples. The first example is a two-dimensional system that describes a single population with internal noise. This model has an analytical solution for the optimal path. The numerical solution found using our computational method agrees well with the analytical result. The second example is a more complicated four-dimensional system where our numerical method must be used to find the optimal path. The third example, although a seemingly simple two-dimensional system, demonstrates the success of our method in finding the optimal path where other numerical methods are known to fail. In the fourth example, the optimal path lies in six-dimensional space and demonstrates the power of our method in computing paths in higher-dimensional spaces.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Entanglement-fidelity relations for inaccurate ancilla-driven quantum computation

It was shown by T. Morimae [Phys. Rev. A 81, 060307(R) (2010)] that the gate fidelity of an inaccurate one-way quantum computation is upper bounded by a decreasing function of the amount of entanglement in the register. This means that a strong entanglement causes the low gate fidelity in the one-way quantum computation with inaccurate measurements. In this paper, we derive similar entanglement-fidelity relations for the inaccurate ancilla-driven quantum computation. These relations again imply that a strong entanglement in the register causes the low gate fidelity in the ancilla-driven quantum computation if the measurements on the ancilla are inaccurate.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

The high explosives & affected targets (HEAT) dataset

Artificial Intelligence (AI) surrogate models offer a computationally efficient alternative to full-physics simulations, yet no existing datasets are publicly available for training, testing, and validation of machine learning models of the dynamics of high-explosive driven shocks through multiple materials. Shock propagation through materials is a computationally challenging problem because simulations must include material-specific equations of state (EOS) along with descriptions of other physical processes such as plastic deformation, phase change, damage processes, fluid instabilities, and multi-material interactions. Shocks are typically initiated by high-velocity impacts or explosive loading. The latter case necessitates the addition of models of reactive materials to represent high-explosive (HE) detonation. Here, to address the lack of an expansive dataset for multi-material shock propagation in the AI/ML community, we present the High-Explosives and Affected Targets (HEAT) Dataset. HEAT is a physics-rich collection of two-dimensional, cylindrically symmetric, simulations generated using an Eulerian, multi-material, shock-propagation code developed at Los Alamos National Laboratory. The dataset includes two partitions: (1) the expanding shock-cylinder (CYL) simulations, Figs. 1, and (2) the Perturbed Layered Interface (PLI) simulations, Fig. 2. Entries in both partitions consist of time series of arrays of thermodynamic fields (pressure, density, and temperature), kinematic fields (position and velocity), and additional fields that depend on thermodynamic and/or kinematic fields (e.g., material stress). Materials in the CYL partition include solids (aluminium, copper, depleted uranium, stainless steel, tantalum, and a generic polymer), a liquid (water), gases (air, nitrogen), and a generic detonating material (high explosive, HE). The PLI partition spans a highly varying geometry but consists of fixed materials across entries: Copper, aluminium, stainless steel, generic polymer, and generic HE. HEAT captures critical phenomena such as momentum transfer, shock propagation, plastic deformation, and thermal effects, making HEAT a valuable benchmark for development of AI/ML emulation of multi-material shock propagation.

36 MATERIALS SCIENCE↗