Search NASA⌕ Search

SEARCH · Search NASA

Results for “Computing continuum”

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 37 records · Page 2

QoS-aware edge AI placement and scheduling with multiple implementations in FaaS-based edge computing

Resource constraints on the computing continuum require that we make smart decisions for serving AI-based services at the network edge. AI-based services typically have multiple implementations (e.g., image classification implementations include SqueezeNet, DenseNet, and others) with varying trade-offs (e.g., latency and accuracy). The question then is how should AI-based services be placed across Function-as-a-Service (FaaS) based edge computing systems in order to maximize total Quality-of-Service (QoS). To address this question, we propose a problem that jointly aims to solve (i) edge AI service placement and (ii) request scheduling. These are done across two time-scales (one for placement and one for scheduling). Here we first cast the problem as an integer linear program. We then decompose the problem into separate placement and scheduling subproblems and prove that both are NP-hard. We then propose a novel placement algorithm that places services while considering device-to-device communication across edge clouds to offload requests to one another. Our results show that the proposed placement algorithm is able to outperform a state-of-the-art placement algorithm for AI-based services, and other baseline heuristics, with regard to maximizing total QoS. Additionally, we present a federated learning-based framework, FLIES, to predict the future incoming service requests and their QoS requirements. Our results also show that our FLIES algorithm is able to outperform a standard decentralized learning baseline for predicting incoming requests and show comparable predictive performance when compared to centralized training.

97 MATHEMATICS AND COMPUTING↗

Workflows Community Summit 2022: A Roadmap Revolution

Scientific workflows have become integral tools in broad scientific computing use cases. Science discovery is increasingly dependent on workflows to orchestrate large and complex scientific experiments that range from the execution of a cloud-based data preprocessing pipeline to multi-facility instrument-to-edge-to-HPC computational workflows. Given the changing landscape of scientific computing (often referred to as a computing continuum) and the evolving needs of emerging scientific applications, it is paramount that the development of novel scientific workflows and system functionalities seek to increase the efficiency, resilience, and pervasiveness of existing systems and applications. Specifically, the proliferation of machine learning/artificial intelligence (ML/AI) workflows, need for processing large-scale datasets produced by instruments at the edge, intensification of near real-time data processing, support for long-term experiment campaigns, and emergence of quantum computing as an adjunct to HPC, have significantly changed the functional and operational requirements of workflow systems. Workflow systems now need to, for example, support data streams from the edge-to-cloud-to-HPC, enable the management of many small-sized files, allow data reduction while ensuring high accuracy, orchestrate distributed services (workflows, instruments, data movement, provenance, publication, etc.) across computing and user facilities, among others. Further, to accelerate science, it is also necessary that these systems implement specifications/standards and APIs for seamless (horizontal and vertical) integration between systems and applications, as well as enable the publication of workflows and their associated products according to the FAIR principles.

97 MATHEMATICS AND COMPUTING↗

Characterization of the Finite Element Computational Fluid Dynamics Capabilities in the Multiphysics Object Oriented Simulation Environment

We report the multiphysics object-oriented simulation environment (moose) is a code package that couples a variety of physics modules, allowing for highly accessible multiphysics simulations. The physics modules include a finite element Navier–Stokes (N–S) module that is designed to solve laminar fluid dynamics problems. The usage of this module in multiple recent studies coupled with the growing interest in moose for usage in nonlight water reactor safety studies by the Nuclear Regulatory Commission (NRC) prompted the authors to investigate the computational fluid dynamics capabilities of moose. A two-dimensional laminar flow past a circular cylinder scenario is simulated in the moose framework to investigate the effectiveness of the N–S module. Simulations assumed an unsteady laminar flow with a Reynolds number of 200. To verify the results from moose, similar simulations were conducted using the well-utilized simulation of turbulent flow in arbitrary regions—computational continuum mechanics C++ (star-ccm + ) finite volume code. Results from both codes are also compared to some results from literature. Velocity and pressure profiles of both transient simulations were compared. The numerical and input errors in moose are also visualized with contour plots to qualitatively understand the evolution of the errors across time and space. The comparisons between moose and star-ccm + showed nearly perfect agreement between the codes for velocity and pressure, especially after the development of the vortex street in later time-steps. The force coefficients showed excellent agreement after the development of the vortex street, but demonstrated notable discrepancies prior to the vortex street development, which is likely due to how each code simulated the approach to the vortex street in earlier time-steps.

97 MATHEMATICS AND COMPUTING↗

Conservative velocity mappings for discontinuous Galerkin kinetics

Continuum computational kinetic plasma models evolve the distribution function of a plasma species f s on a phase-space grid over time. In many problems of interest the distribution function has limited extent in velocity space; hence, using a uniform, highly refined mesh would be costly and slow. Nonuniform velocity grids can reduce the computational cost by placing more degrees of freedom where f s is appreciable and fewer where it is not. In this work we introduce a first-of-its kind discontinuous Galerkin approach to nonuniform velocity-space discretization using mapped velocity coordinates. This new method is presented in the context of a gyrokinetic model used to study magnetized plasmas. We create discretizations of collisionless and collisional terms using mappings in a way that exactly conserves particles and energy. Numerical tests of such properties are presented, and we show that this new discretization can reproduce earlier gyrokinetic simulations using grids with up to 6–60 times fewer cells and 22X-60X speed-ups depending on dimensionality, geometry and plasma parameters.

Discontinuous Galerkin↗

ON HIGHLY CLUMPED MAGNETIC WIND MODELS FOR COOL EVOLVED STARS

Recently, it has been proposed that the winds of non-pulsating and non-dusty K and M giants and supergiants may be driven by some form of magnetic pressure acting on highly clumped wind material. While many researchers believe that magnetic processes are responsible for cool evolved stellar winds, existing MHD and Alfven wave-driven wind models have magnetic fields that are essentially radial and tied to the photosphere. The clumped magnetic wind scenario is quite different in that the magnetic flux is also being carried away from the star with the wind. We test this clumped wind hypothesis by computing continuum radio fluxes from the {zeta} Aur semiempirical model of Baade et al., which is based on wind-scattered line profiles. The radio continuum opacity is proportional to the electron density squared, while the line scattering opacity is proportional to the gas density. This difference in proportionality provides a test for the presence of large clumping factors. We derive the radial distribution of clump factors (CFs) for {zeta} Aur by comparing the nonthermal pressures required to produce the semiempirical velocity distribution with the expected thermal pressures. The CFs are {approx}5 throughout the sub-sonic inner wind region and then decline outward. These implied clumping factors lead to excess radio emission at 2.0 cm, while at 6.2 cm it improves agreement with the smooth unclumped model. Smaller clumping factors of {approx}2 lead to better overall agreement but also increase the discrepancy at 2 cm. These results do not support the magnetic clumped wind hypothesis and instead suggest that inherent uncertainties in the underlying semiempirical model probably dominate uncertainties in predicted radio fluxes. However, new ultraviolet line and radio continuum observations are needed to test the new generations of inhomogeneous magnetohydrodynamic wind models.

79 ASTRONOMY AND ASTROPHYSICS↗

The 2D Rydberg series in Al I

High quality ab initio electonic structure calculations were performed on the 2D Rydberg series in Al I. The configuration 3s3p2(2D) is shown to contribute substantially to the lowest four 2D Rydberg states. The same configuration also contributes substantially to a 2D state embedded in the ionization continuum. Computed oscillator strengths for the first six members of the 2D Rydberg transitions are given: these should be of substantially high accuracy than currently available values.

Taylor, Peter R.↗

The 2D Rydberg series in Al I

High-quality ab initio electronic structure calculations have been performed on the 2D Rydberg series in Al I. The configuration 3s3p2(2D) is shown to contribute substantially to the lowest four 2D Rydberg states. The same configuration also contributes substantially to a 2D state embedded in the ionization continuum. Computed oscillator strengths for the first six members of the 2D Rydberg transitions are given: these should be of substantially higher accuracy than currently available values.

Taylor, Peter R.↗

A computational approach to continuum damping of Alfven waves in two and three-dimensional geometry

While the usual way of calculating continuum damping of global Alfven modes is the introduction of a small artificial resistivity, we present a computational approach to the problem based on a suitable path of integration in the complex plane. This approach is implemented by the Riccati shooting method and it is shown that it can be transferred to the Galerkin method used in three-dimensional ideal magneto-hydrodynamics (MHD) codes. The new approach turns out to be less expensive with respect to resolution and computation time than the usual one. We present an application to large aspect ratio tokamak and stellarator equilibria retaining a few Fourier harmonics only and calculate eigenfunctions and continuum damping rates. These may serve as an input for kinetic MHD hybrid models making it possible to bypass the problem of having singularities on the path of integration on one hand and considering continuum damping on the other.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

The role of Rydberg and continuum levels in computing high harmonic generation spectra of the hydrogen atom using time-dependent configuration interaction

We study the role of Rydberg bound-states and continuum levels in the field-induced electronic dynamics associated with the High-Harmonic Generation (HHG) spectroscopy of the hydrogen atom. Time-dependent configuration-interaction (TD-CI) is used with very large atomic orbital (AO) expansions (up to L= 4 with sextuple augmentation and off-center functions) to describe the bound Rydberg levels, and some continuum levels. To address the lack of ionization losses in TD-CI with finite AO basis sets, we employed a heuristic lifetime for energy levels above the ionization potential. The heuristic lifetime model is compared against the conventional atomic orbital treatment (infinite lifetimes), and a third approximation which is TD-CI using only the bound levels (continuum lifetimes go to zero). The results suggest that spectra calculated using conventional TD-CI do not converge with increasing AO basis set size, while the zero lifetime and heuristic lifetime models converge to qualitatively similar spectra, with implications for how best to apply bound state electronic structure methods to simulate HHG. The origin of HHG spectral features including the cutoff and extent of interference between peaks is uncovered by separating field-induced coupling between different types of levels (ground state, bound Rydberg levels, and continuum) in the simulated electronic dynamics. Thus the origin of deviations between the predictions of the semi-classical three step model and the full simulation can be associated with particular physical contributions, which helps to explain both the successes and the limitations of the three step model.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Computational modeling of continuum scale constitutive equations to improve biomass feedstock material handling and conveying systems

Existing design practices for common feedstock handling equipment such as hoppers and screw feeders have not been successful in many situations. The lack of an appropriate toolset to characterize and model biomass flow behavior is deemed the root cause of plugging, stalls, and flow-related issues in biomass handling systems design and operations. New standardized characterization protocols in conjunction with appropriate constitutive models will lead to successful biomass handling system design and operations. This study investigates constitutive models for bulk solid flow based on the first principles to overcome issues in biomass handling. The focus is on two major issues in biomass handling, i.e., initiating and sustaining of flow in the hopper that is widely used in biomass conveying operation. Accordingly, this article focuses on analytical models of biomass hopper flow. Mohr-Coulomb, Drucker-Prager, and modified Cam-clay models are investigated as constitutive bulk solid flow models. Parameters of those models are determined using a set of triaxial tests. Triaxial tests are conducted using the cubical triaxial tester (CTT) in the Food and Biomaterials Lab of the Pennsylvania State University. Milled 1mm corn stover and 2mm wood chip biomass materials having different flow characteristics are examined, and their respective flow behavior out of a hopper are simulated using the Finite Element method implementing identified constitutive models and determined fundamental mechanical properties. It is found that, with the same hopper geometry, ground corn stover 1mm tends to exhibit funnel flow whereas ground Douglas fir 2 mm tends to exhibit mass flow. These simulation results are validated with hopper flow experiments. This study demonstrates how biomass flow problems can be predicted and mitigated by numerically modeled fundamental constitutive equations.

09 BIOMASS FUELS↗

Continuum Mechanical and Computational Aspects of Material Behavior

Fluid flows are typically classified as laminar or turbulent. While the glassy, regular flow of water from a slightly opened tap is laminar, the sinuous, irregular flow of water from a fully opened tap is turbulent. In a laminar flow, the velocity and other relevant fields are deterministic functions of position and time. Photos taken at different times, no matter how far removed, of steady laminar flow from a tap will be identical. In a turbulent flow, the velocity and other relevant fields manifest complex spatial and temporal fluctuations. A video of steady turbulent flow from a tap will exhibit a constantly changing pattern and many length and time scales. In nature and technology, laminar flows are more the exception than the rule. Fluvial, oceanic, pyroclastic, atmospheric, and interstellar flows are generally turbulent, as are the flows of blood through the left ventricle and air in the lungs. Flows around land, sea, and air vehicles and through pipelines, heating, cooling, and ventilation systems are generally turbulent, as are most flows involved in industrial processing, combustion, chemical reactions, and crystal growth. Over the past year, a significant portion of our research activity has focused on numerical studies of Navier-Stokes-αβ model and extensions thereof. Our results regarding these and other approaches to turbulence modeling are described below.

42 ENGINEERING↗

The dynamics and control of large flexible space structures. Part B: Development of continuum model and computer simulation

The equations of motion of an arbitrary flexible body in orbit were derived. The model includes the effects of gravity with all its higher harmonics. As a specific example, the motion of a long, slender, uniform beam in circular orbit was modelled. The example considers both the inplane and three dimensional motion of the beam in orbit. In the case of planar motion with only flexible vibrations, the pitch motion is not influenced by the elastic motion of the beam. For large values of the square of the ratio of the structural modal frequency to the orbital angular rate the elastic motion was decoupled from the pitch motion. However, for small values of the ratio and small amplitude pitch motion, the elastic motion was governed by a Hill's 3 term equation. Numerical simulation of the equation indicates the possibilities of instability for very low values of the square of the ratio of the modal frequency to the orbit angular rate. Also numerical simulations of the first order nonlinear equations of motion for a long flexible beam in orbit were performed. The effect of varying the initial conditions and the number of modes was demonstrated.

Bainum, P. M.↗

Analysis of the anisotropic viscoplastic-damage response of composite laminates - Continuum basis and computational algorithms

The mathematical structure underlying the rate equations of a recently-developed constitutive model for the coupled viscoplastic-damage response of anisotropic composites is critically examined. In this regard, a number of tensor projection operators have been identified, and their properties were exploited to enable the development of a general computational framework for their numerical implementation using the Euler fully-implicit integration method. In particular, this facilitated (i) the derivation of explicit expressions of the (consistent) material tangent stiffnesses that are valid for both three-dimensional as well as subspace (e.g. plane stress) formulations, (ii) the implications of the symmetry or unsymmetry properties of these tangent operators from a thermodynamic standpoint, and (iii) the development of an effective time-step control strategy to ensure accuracy and convergence of the solution. In addition, the special limiting case of inviscid elastoplasticity is treated. The results of several numerical simulations are given to demonstrate the effectiveness of the schemes developed.

Saleeb, A. F.↗

Direct simulation Monte Carlo on petaflop supercomputers and beyond

The gold-standard definition of the Direct Simulation Monte Carlo (DSMC) method is given in the 1994 book by Bird [Molecular Gas Dynamics and the Direct Simulation of Gas Flows (Clarendon Press, Oxford, UK, 1994)], which refined his pioneering earlier papers in which he first formulated the method. In the intervening 25 years, DSMC has become the method of choice for modeling rarefied gas dynamics in a variety of scenarios. The chief concern to applying DSMC to more dense or even continuum flows is its computational expense compared to continuum computational fluid dynamics methods. The dramatic (nearly billion-fold) increase in speed of the largest supercomputers over the last 30 years has thus been a key enabling factor in using DSMC to model a richer variety of flows, due to the method’s inherent parallelism. We have developed the open-source SPARTA DSMC code with the goal of running DSMC efficiently on the largest machines, both current and future. It is largely an implementation of Bird’s 1994 formulation. In this work, we describe algorithms used in SPARTA to enable DSMC to operate in parallel at the scale of many billions of particles or grid cells, or with billions of surface elements. We highlight a few examples of the kinds of fundamental physics questions and engineering applications that DSMC can address at these scales.

97 MATHEMATICS AND COMPUTING↗

Simulation and Emulation of X-Ray Diffraction from Dynamic Compression Experiments

Many important aspects of the dynamic thermo-mechanical response of materials occur at the mesoscale, i.e. a physical scale of interactions smaller than what can be adequately described by homogenous behaviors, yet larger than the scale of the atomic lattice. Concurrent advancements in computational power, continuum theory, and experimental diagnostics are enabling unprecedented understanding of such interactions. However, we cannot develop a sufficient level of confidence in such mesoscale capability until the constitutive description of the underlying constituents is reliably representative of their actual physical behavior. Therefore, there is a strong need to combine experimental, modeling, and data-science techniques to validate models of the thermomechanical response of individual single crystals. One experimental diagnostic with high potential impact to shock physics and materials science is in-situ x-ray diffraction. This paper is primarily focused on simulation of x-ray diffraction in shock physics, but with an aim toward quantifying parametric uncertainty of simulation models. Here, we develop and demonstrate a data-science and model-driven approach to constrain the parameterization of continuum models of crystal lattice deformation associated with the shock response of crystalline materials. The framework is built around the connection between continuum hydrodynamic simulations of lattice deformation and a new Bragg diffraction simulation code, BarberShop. The dynamic deformation of a crystal lattice is modeled using the DiscoFlux model within an arbitrary Lagrangian-Eulerian hydrodynamic code, FLAG. These detailed continuum simulations of lattice deformation can be computationally slow, thus a statistical model is used to emulate the evolution of lattice deformation fields in time and across the considered model parameter space. Emulated lattice deformation fields can then be generated rapidly for any combination of physics model parameters. In turn, these fields can be fed into BarberShop to realize a rapid prediction of Bragg diffraction patterns associated with particular values of physics model parameters. The framework enables parameterization of the single crystal model to obtain Bragg diffraction patterns that most closely resemble a corresponding measurement. Furthermore, the framework naturally provides sensitivities of the lattice deformation to the physics parameters. We highlight the utility of this framework through the application to a synthetic closed-loop inverse problem leading to the parameterization of a single crystal material model. As a model problem, we consider the dynamic response of the energetic molecular crystal, cyclotrimethylenetrinitramine (or RDX), under dynamic compression induced by simulated flyer plate impact experiments.

36 MATERIALS SCIENCE↗

Computational aspects of the continuum quaternionic wave functions for hydrogen

Over the past few years considerable attention has been given to the role played by the Hydrogen Continuum Wave Functions (HCWFs) in quantum theory. The HCWFs arise via the method of separation of variables for the time-independent Schrödinger equation in spherical coordinates. The HCWFs are composed of products of a radial part involving associated Laguerre polynomials multiplied by exponential factors and an angular part that is the spherical harmonics. In the present paper we introduce the continuum wave functions for hydrogen within quaternionic analysis ((R)QHCWFs), a result which is not available in the existing literature. In particular, the underlying functions are of three real variables and take on either values in the reduced and full quaternions (identified, respectively, with R{sup 3} and R{sup 4}). We prove that the (R)QHCWFs are orthonormal to one another. The representation of these functions in terms of the HCWFs are explicitly given, from which several recurrence formulae for fast computer implementations can be derived. A summary of fundamental properties and further computation of the hydrogen-like atom transforms of the (R)QHCWFs are also discussed. We address all the above and explore some basic facts of the arising quaternionic function theory. As an application, we provide the reader with plot simulations that demonstrate the effectiveness of our approach. (R)QHCWFs are new in the literature and have some consequences that are now under investigation.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Creating a Simple Single Computational Approach to Modeling Rarefied and Continuum Flow About Aerospace Vehicles

We proposed to create a single computational code incorporating methods that can model both rarefied and continuum flow to enable the efficient simulation of flow about space craft and high altitude hypersonic aerospace vehicles. The code was to use a single grid structure that permits a smooth transition between the continuum and rarefied portions of the flow. Developing an appropriate computational boundary between the two regions represented a major challenge. The primary approach chosen involves coupling a four-speed Lattice Boltzmann model for the continuum flow with the DSMC method in the rarefied regime. We also explored the possibility of using a standard finite difference Navier Stokes solver for the continuum flow. With the resulting code we will ultimately investigate three-dimensional plume impingement effects, a subject of critical importance to NASA and related to the work of Drs. Forrest Lumpkin, Steve Fitzgerald and Jay Le Beau at Johnson Space Center. Below is a brief background on the project and a summary of the results as of the end of the grant.

Goldstein, David B.↗

First-principles equation of state and electronic properties of warm dense oxygen

We perform all-electron path integral Monte Carlo (PIMC) and density functional theory molecular dynamics (DFT-MD) calculations to explore warm dense matter states of oxygen. Our simulations cover a wide density-temperature range of 1–100 g cm{sup −3} and 10{sup 4}–10{sup 9} K. By combining results from PIMC and DFT-MD, we are able to compute pressures and internal energies from first-principles at all temperatures and provide a coherent equation of state. We compare our first-principles calculations with analytic equations of state, which tend to agree for temperatures above 8 × 10{sup 6} K. Pair-correlation functions and the electronic density of states reveal an evolving plasma structure and ionization process that is driven by temperature and density. As we increase the density at constant temperature, we find that the ionization fraction of the 1s state decreases while the other electronic states move towards the continuum. Finally, the computed shock Hugoniot curves show an increase in compression as the first and second shells are ionized.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗