Search NASA⌕ Search

SEARCH · Search NASA

Results for “Accelerators, Computing”

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 1,729 records · Page 96

ArborX 2.0

ArborX library tackles a problem of efficiently finding geometric objects that are close in space. Variations of this problem, such as finding the nearest neighbors of a point, or finding all objects within a certain distance, are inherent components of applications in many fields. The data may be large so that solving the problem efficiently may require significant computational resources, such as multiple processors or accelerators such as general purpose GPUs. ArborX' main advantage in its ability to solve large problems efficiently utilizing a combination of distributed and on-node parallelism. ArborX can be run efficiently on a wide variety of hardware, including GPUs from different vendors, which distinguishes it from other available libraries which typically choose only few of these. The other advantage is that it supports both types of user problems: spatial problems (useful for intersections and finding objects within certain distance), and nearest neighbor problems. ArborX also supports flexible interface in its interaction with a user. Particularly, it allows a user to call user's own function on a positive match, a functionality not rarely available in other libraries. ArborX implements construction and traversal algorithms using efficient tree structures, such as bounding volume hierarchy (BVH). At its core, ArborX uses linear BVH for its low construction cost and sufficient quality. ArborX implements both spatial and nearest-neighbor traversal algorithms. ArborX also provides several clustering algorithms (minimum spanning tree, DBSCAN, HDBSCAN*), interpolation using minimum least squares and ray tracing. ArborX is written using C++, and is parallelized using the message passing interface (MPI) for the distributed communication, and the Kokkos library for on-node parallelism. This approach allows ArborX to be run on a wide variety of hardware, from common laptops and desktops to supercomputers while using the same codebase.

Prokopenko, Andrey [Oak Ridge National Laboratory ↗

Active Learning Framework

Machine learning (ML) of interatomic potentials show great promise to accelerate scientific simulation, e.g., by emulating expensive computations at a high accuracy but much reduced computational cost. Training datasets are calculated from computationally expensive ab initio quantum mechanics methods, density functional theory (DFT). Trained on this data, an ML model can be very successful in predicting energy and forces for new atomic configurations. A critical factor is the quality and diversity of the training dataset. Thus, a highly automated approach to dataset construction based on active learning framework is designed suitable for material physics. The active learning scheme begins with fully randomized atomic configurations. Then, many Molecular Dynamics (MD) trajectories are simulated using current ML potentials, where each MD trajectory is initialized to a random disordered configuration. The temperature is varied in order to diversify the sampled configuration during these simulations. The variance of predictions for eight neural networks within an ensemble is analyzed to determine whether the model is operating as expected. This helps in determining whether collecting more data would be helpful to the model by checking the ensemble variance is greater than the threshold. In this case, the MD trajectory is terminated and the final atomic configuration is placed on a queue (SQL database) for DFT calculations and added to training dataset. Periodically, ML model is retrained to the updated training model. This Active Learning loop is iterated until the cost of MD simulations becomes prohibitively expensive. The MD simulations will hopefully be sufficiently robust to support nucleation after many active learning iterations. In this sense, active learning scheme must automatically discover the important low energy and nonequilibrium physics.

Nebgen, Benjamin↗

Theoretical and Computational Modeling of a Plasma Wakefield BBU Instability

Plasma wakefield accelerators achieve accelerating gradients on the order of the wave-breaking limit, m c² k_{p}/e, so that higher accelerating gradients correspond to shorter plasma wavelengths. Small-scale accelerating structures, such as plasma and dielectric wakefields, are susceptible to the beam break-up instability (BBU), which can be understood from the Panofsky-Wenzel theorem: if the fundamental accelerating mode scales as b⁻¹ for a structure radius b, then the dipole mode must scale as b⁻³, meaning that high accelerating gradients necessarily come with strong dipole wake fields. Because of this relationship, any plasma-accelerator-based future collider will require detailed study of the trade-offs between extracting the maximum energy from the driver and mitigating the beam break-up instability. Recent theoretical work* predicts the tradeoff between the witness bunch stability and the amount of energy that can be extracted from the drive bunch, a so-called efficiency-instability relation . We will discuss the beam break-up instability and the efficiency-instability relation and the theoretical assumptions made in reaching this conclusion. We will also present preliminary particle-in-cell simulations of a beam-driven plasma wakefield accelerator used to test the domain of validity for the assumptions made in this model.

43 PARTICLE ACCELERATORS↗

Searching for Strongly Coupled Dark Sectors with Unsupervised and Generative Learning

Recipient of the URA Early Career Award for groundbreaking searches for dark matter arising from strongly coupled dark sectors with the CMS detector, pioneering work in ML-based model-independent anomaly detection for collider and astrophysics experiments, and leadership in the development of new AI/ML techniques to improve event reconstruction and detector simulation in particle physics, as well as novel strategies to accelerate AI inference and throughput with heterogeneous computing using coprocessors as a service.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Reduction of Simulation Times for High-Q Structures using the Resonance Equation

Simulating steady state performance of high quality factor (Q) resonant RF structures is computationally difficult for structures with sizes on the order of more than a few wavelengths because of the long times (on the order of ~ 0.1 ms) required to achieve steady state in comparison with maximum time step that can be used in the simulation (typically, on the order of ~ 1 ps). This paper presents analytical and computational approaches that can be used to accelerate the simulation of the steady state performance of such structures. The basis of the proposed approach is the utilization of a larger amplitude signal at the beginning to achieve steady state earlier relative to the nominal input signal. Finally, the methodology for finding the necessary input signal is then discussed in detail, and the validity of the approach is evaluated.

43 PARTICLE ACCELERATORS↗

Stressing the importance of plant specialized metabolites: omics-based approaches for discovering specialized metabolism in plant stress responses

Plants produce a diverse range of specialized metabolites that play pivotal roles in mediating environmental interactions and stress adaptation. These unique chemical compounds also hold significant agricultural, medicinal, and industrial values. Despite the expanding knowledge of their functions in plant stress interactions, understanding the intricate biosynthetic pathways of these natural products remains challenging due to gene and pathway redundancy, multifunctionality of proteins, and the activity of enzymes with broad substrate specificity. In the past decade, substantial progress in genomics, transcriptomics, metabolomics, and proteomics has made the exploration of plant specialized metabolism more feasible than ever before. Notably, recent advances in integrative multi-omics and computational approaches, along with other technologies, are accelerating the discovery of plant specialized metabolism. In this review, we present a summary of the recent progress in the discovery of plant stress-related specialized metabolites. Emphasis is placed on the application of advanced omics-based approaches and other techniques in studying plant stress-related specialized metabolism. Additionally, we discuss the high-throughput methods for gene functional characterization. These advances hold great promise for harnessing the potential of specialized metabolites to enhance plant stress resilience in the future.

59 BASIC BIOLOGICAL SCIENCES↗

deeprob/ThioesteraseEnzymeSpecificity: EnZymClass-first-release

Characterizing the functional properties of plant acyl-ACP thioesterases (TEs), a key enzyme class used in the production of renewable oleochemicals in microbial hosts, experimentally, can be an expensive and time consuming process since it requires manual screening of thousands of candidates in a database. Using amino acid sequence to computationally predict an enzyme’s function might accelerate this process; however obtaining the necessary amount of information on previously characterized enzymes and their respective sequences required by standard Machine Learning (ML) based approaches to accurately infer sequence-function relationships can be prohibitive, especially with a low-throughput testing cycle. Experimental noise, unbalanced dataset where high sequence similarity does not always imply identical functional properties will further prevent robust prediction performance. Herein we present a ML method, Ensemble method for enZyme Classification (EnZymClass), that is specifically designed to address these issues. We used EnZymClass to classify TEs into short, long and mixed free fatty acid substrate specificity categories. While general guidelines for inferring substrate specificity have been proposed before, prediction of chain-length preference from primary sequence has remained elusive for plant acyl-ACP TEs. By applying EnZymClass to a subset of TEs in the ThYme database, we identified two medium chain TEs, ClFatB3 and CwFatB2, with previously uncharacterized activity in E. coli fatty acid production hosts.

Banerjee, Deepro↗

Analytical study of laser-supported combustion waves in hydrogen

Laser supported combustion (LSC) waves are an important ingredient in the fluid mechanics of CW laser propulsion using a hydrogen propellant and 10.6 micron lasers. Therefore, a computer model has been constructed to solve the one-dimensional energy equation with constant pressure and area. Physical processes considered include convection, conduction, absorption of laser energy, radiation energy loss, and accurate properties of equilibrium hydrogen. Calculations for 1, 3, 10 and 30 atm were made for intensities of 10 to the 4th to 10 to the 6th W/sq cm, which gave temperature profiles, wave speed, etc. To pursue the propulsion application, a second computer model was developed to describe the acceleration of the gas emerging from the LSC wave into a variable-pressure, converging streamtube, still including all the above-mentioned physical processes. The results show very high temperatures in LSC waves which absorb all the laser energy, and high radiative losses.

Kemp, N. H.↗

Low cost solar array project production process and equipment task: A Module Experimental Process System Development Unit (MEPSDU)

Several major modifications were made to the design presented at the PDR. The frame was deleted in favor of a "frameless" design which will provide a substantially improved cell packing factor. Potential shaded cell damage resulting from operation into a short circuit can be eliminated by a change in the cell series/parallel electrical interconnect configuration. The baseline process sequence defined for the MEPSON was refined and equipment design and specification work was completed. SAMICS cost analysis work accelerated, format A's were prepared and computer simulations completed. Design work on the automated cell interconnect station was focused on bond technique selection experiments.

Source record↗

Four-Cylinder Stirling Engine Control Simulation

Four-cylinder, Stirling-engine, transient-engine-simulation computer program developed. Program intended for control analysis. Associated engine model simplified to shorten computer calculation time. Model includes engine mechanical-drive dynamics and vehicle-load effects. Computer program also includes subroutines that allow acceleration of engine by addition of hydrogen to system and braking of engine by short circuiting of working spaces.

Daniele, C. J.↗

An unconditionally stable Runge-Kutta method for unsteady flows

A quasi-three dimensional analysis was developed for unsteady rotor-stator interaction in turbomachinery. The analysis solves the unsteady Euler or thin-layer Navier-Stokes equations in a body fitted coordinate system. It accounts for the effects of rotation, radius change, and stream surface thickness. The Baldwin-Lomax eddy viscosity model is used for turbulent flows. The equations are integrated in time using a four stage Runge-Kutta scheme with a constant time step. Implicit residual smoothing was employed to accelerate the solution of the time accurate computations. The scheme is described and accuracy analyses are given. Results are shown for a supersonic through-flow fan designed for NASA Lewis. The rotor:stator blade ratio was taken as 1:1. Results are also shown for the first stage of the Space Shuttle Main Engine high pressure fuel turbopump. Here the blade ratio is 2:3. Implicit residual smoothing was used to increase the time step limit of the unsmoothed scheme by a factor of six with negligible differences in the unsteady results. It is felt that the implicitly smoothed Runge-Kutta scheme is easily competitive with implicit schemes for unsteady flows while retaining the simplicity of an explicit scheme.

Jorgenson, Philip C. E.↗

Development of a Navier-Stokes algorithm for parallel-processing supercomputers

An explicit flow solver, applicable to the hierarchy of model equations ranging from Euler to full Navier-Stokes, is combined with several techniques designed to reduce computational expense. The computational domain consists of local grid refinements embedded in a global coarse mesh, where the locations of these refinements are defined by the physics of the flow. Flow characteristics are also used to determine which set of model equations is appropriate for solution in each region, thereby reducing not only the number of grid points at which the solution must be obtained, but also the computational effort required to get that solution. Acceleration to steady-state is achieved by applying multigrid on each of the subgrids, regardless of the particular model equations being solved. Since each of these components is explicit, advantage can readily be taken of the vector- and parallel-processing capabilities of machines such as the Cray X-MP and Cray-2.

Swisshelm, Julie M.↗

An unconditionally stable Runge-Kutta method for unsteady flows

A quasi-three-dimensional analysis was developed for unsteady rotor-stator interaction in turbomachinery. The analysis solves the unsteady Euler or thin-layer Navier-Stokes equations in a body-fitted coordinate system. It accounts for the effects of rotation, radius change, and stream surface thickness. The Baldwin-Lomax eddy viscosity model is used for turbulent flows. The equations are integrated in time using a four-stage Runge-Kutta scheme with a constant time step. Implicit residual smoothing was employed to accelerate the solution of the time accurate computations. The scheme is described and accuracy analyses are given. Results are shown for a supersonic through-flow fan designed for NASA Lewis. The rotor:stator blade ratio was taken as 1:1. Results are also shown for the first stage of the Space Shuttle Main Engine high pressure fuel turbopump. Here the blade ratio is 2:3. Implicit residual smoothing was used to increase the time step limit of the unsmoothed scheme by a factor of six with negligible differences in the unsteady results. It is felt that the implicitly smoothed Runge-Kutta scheme is easily competitive with implicit schemes for unsteady flows while retaining the simplicity of an explicit scheme.

Jorgenson, Philip C. E.↗

Development of an aerodynamics algorithm for parallel-processing supercomputers

An explicit flow solver, applicable to the hierarchy of model equations ranging from Euler to full Navier-Stokes, is combined with several techniques designed to reduce computational expense. The computational domain consists of local grid refinements embedded in a global coarse mesh, where the locations of these refinements are defined by the physics of the flow. Flow characteristics are also used to determine which set of model equations is appropriate for solution in each region, thereby reducing not only the number of grid points at which the solution must be obtained, but also the computational effort required to get that solution. Acceleration to steady-state is achieved by applying multigrid on each of the subgrids, regardless of the particular model equations being solved. Since each of these components is explicit, advantage can readily be taken of the vector- and parallel-processing capabilities of machines such as the Cray X-MP and Cray-2.

Swisshelm, Julie M.↗

Control of robot dynamics using acceleration control

Acceleration control of robotic devices can provide improvements to many space-based operations using flexible manipulators and to ground-based operations requiring better precision and efficiency than current industrial robots can provide. This paper reports on a preliminary study of acceleration measurement on robotic motion during parabolic flights on the NASA KC-135 and a parallel study of accelerations with and without gravity arising from computer simulated motions using TREETOPS software.

Workman, G. L.↗

Green's function methods in heavy ion shielding

An analytic solution to the heavy ion transport in terms of Green's function is used to generate a highly efficient computer code for space applications. The efficiency of the computer code is accomplished by a nonperturbative technique extending Green's function over the solution domain. The computer code can also be applied to accelerator boundary conditions to allow code validation in laboratory experiments.

Wilson, John W.↗

On some recent advances in multidisciplinary analysis of hypersonic vehicles

This paper presents pertinent details of the development and application of an integrated, multidisciplinary finite element analysis tool for the effective modeling and simulation of aerospace vehicles including hypersonic ones. Recent advances in this connection include more efficient CFD solution techniques and also aeroservoelastic stability analysis methods that employ unsteady aerodynamic forces for computation of flutter and divergence speeds. An accelerated Euler solution procedure based on the Aitken acceleration technique has recently been implemented that effects considerable improvement in solution efficiency for the same level of solution accuracy, as evidenced by example problems presented in the paper. A summary of finite element numerical formulations for the various individual disciplines as well as for the unified aeroservoelastic analysis and some relevant numerical examples are also presented in the paper.

Gupta, K. K.↗

Ponderomotive effects on distributions of O(+) ions in the auroral zone

Test particle calculations are used to compute the effects of gravity and ponderomotive acceleration by shear Alfven wave oscillations on the distribution function of O(+) ions along auroral field lines, assuming an ionospheric Maxwellian source of the ions at 2000 km altitude with approximately 0.5 eV of thermal energy in the parallel component of velocity. The electric field model corresponds to a standing wave oscillation with a frequency approximately 1 Hz in the azimuthal direction superimposed on the background dipole field, in which the wave amplitude is either increasing or decreasing in time. The electric field is taken to be primarily in the perpendicular direction. The time varying wave produces broad distributions with widths of 2 to 10 times the initial 0.5-eV thermal energy of the Maxwellian source, and the density and flux of upward going O(+) ions at one Earth radius are both enhanced in this model. The oxygen ion distribution functions at 1 R(sub E) altitude resulting from interaction with waves whose amplitudes are increasing in time have a more gradual lower energy cutoff than do the distribution functions resulting from decaying waves. The high-energy part of the distribution functions in growing waves reflects the temperature of the Maxwellian source, while the high-energy part of the distributions resulting from decaying waves steepens with time, independent of the source temperature.

Witt, E.↗