Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed parallelization”

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 1,207 records · Page 67

Flux Pinning-Induced Stress in a Long Hollow Cylindrical Superconductor Restricted by a Metal Tube

The tensile stress in a hollow cylindrical superconductor completely restricted by a metal tube in a parallel magnetic field is investigated. Using the plane strain approach, analytical expressions for the stress distribution of the superconductor and the contact stress between the metal tube and the superconductor are obtained and show that both the ratio of Young’s modulus and thickness ratio of the high-temperature superconductors and the metal tube significantly affect the stress distribution in the superconductor explicitly. In addition, the flux pinning-induced stress in the above-mentioned superconductor is systematically studied by the exponential critical state model during the field cooling process. The results indicate that the metal tube can reduce effectively the tensile stress in the hollow cylindrical superconductor during the magnetization process. The findings suggest that the appropriate selection of material parameters and size of the metal tube can improve superconductor performance efficiently.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Microstructure effects on fatigue crack growth in additively manufactured Ti–6Al–4V

In this article we investigate the relationship between the different microstructural features in additively-manufactured (AM) Direct Metal Laser Deposited Ti–6Al–4V alloys and the observed fatigue and fracture behaviors, using high resolution Digital Image Correlation experiments and microstructural imaging. Fatigue cracks in the AM specimens were observed to periodically propagate parallel to the α' laths and deflect at prior β grain boundaries. Corresponding plastic strain distributions show needle-like regions of high strain that correlated to the α' laths, suggesting significant influence of the microstructure on the plastic zone. In many cases, the fatigue cracks also propagated towards voids from unsintered powder or gas entrapment in the vicinity. In several experiments, premature failure of the AM specimens occurred under monotonic fracture or cyclic fatigue loading, which we attribute to the presence of void clusters located near the crack-tip. These fatigue and fracture characteristics were similar for AM specimens of different build-orientations.

36 MATERIALS SCIENCE↗

Finite Larmor radius effects on nondiffusive tracer transport in a zonal flow

Finite Larmor radius (FLR) effects on nondiffusive transport in a prototypical zonal flow with drift waves are studied in the context of a simplified chaotic transport model. The model consists of a superposition of drift waves from the linearized Hasegawa-Mima equation and a zonal shear flow perpendicular to the density gradient. High frequency FLR effects are incorporated by gyroaveraging the ExB velocity. Transport in the direction of the density gradient is negligible and we therefore focus on transport parallel to the zonal flows. A prescribed asymmetry produces strongly asymmetric non-Gaussian probability distribution functions (PDFs) of particle displacements, with Levy flights in one direction only. For k{sub perpendicular}{rho}{sub th}=0, where k{sub perpendicular} is the characteristic wavelength of the flow and {rho}{sub th} is the thermal Larmor radius, a transition is observed in the scaling of the second moment of particle displacements: {sigma}{sup 2}{approx}t{sup {gamma}}. The transition separates ballistic motion ({gamma}{approx_equal}2) at intermediate times from superdiffusion ({gamma}=1.6) at larger times. This change of scaling is accompanied by the transition of the PDF of particle displacements from algebraic decay to exponential decay. However, FLR effects seem to eliminate this transition. In all cases, the Lagrangian velocity autocorrelation function exhibits nondiffusive algebraic decay, C{approx}{tau}{sup -{kappa}}, with {kappa}=2-{gamma} to a good approximation. The PDFs of trapping and flight events show clear evidence of algebraic scaling with decay exponents depending on the value of k{sub perpendicular}{rho}{sub th}. The shape and spatiotemporal self-similar anomalous scaling of the PDFs of particle displacements are reproduced accurately with a neutral ({alpha}={beta}), asymmetric, effective fractional diffusion model, where {alpha} and {beta} are the orders of the spatial and temporal fractional derivatives, respectively.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Large amplitude relativistic plasma waves

Relativistic, longitudinal plasma oscillations are studied for the case of a simple water bag distribution of electrons having cylindrical symmetry in momentum space with the axis of the cylinder parallel to the velocity of wave propagation. The plasma is required to obey the relativistic Vlasov-Poisson equations, and solutions are sought in the wave frame. An exact solution for the plasma density as a function of the electrostatic field is derived. The maximum electric field is presented in terms of an integral over the known density. It is shown that when the perpendicular momentum is neglected, the maximum electric field approaches infinity as the wave phase velocity approaches the speed of light. It is also shown that for any nonzero perpendicular momentum, the maximum electric field will remain finite as the wave phase velocity approaches the speed of light. The relationship to previously published solutions is discussed as is some recent controversy regarding the proper modeling of large amplitude relativistic plasma waves.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Sheath and boundary conditions in a collisional magnetized warm electronegative plasma

The characteristics of a weakly collisional sheath in a warm electronegative plasma in the presence of an oblique magnetic field are investigated using a fluid model including the effects of ionization and ion-neutral collisions. The general sheath criterion imposed on the entrance velocity component of the positive ions perpendicular to the wall at the sheath-presheath edge is derived and discussed. It is shown that the boundary conditions are crucial to the sheath structure. Without including the entrance velocity components parallel to the wall, a pulse-like structure in the positive-ion density distribution near the sheath-presheath edge appears if the magnetic field is strong. With inclusion of all velocity components at the edge, the pulse-like structure disappears, resulting in a smooth sheath profile. It is also found that increasing the temperature and decreasing the concentration of the negative ions will increase the sheath thickness, and increasing the magnetic field will decrease the sheath thickness.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Statistical modelling of discharge behavior of atmospheric pressure dielectric barrier discharge

In this work, stochastic behavior of atmospheric pressure dielectric barrier discharge (DBD) has been investigated. The experiment is performed in a DBD reactor consisting of a pair of stainless steel parallel plate electrodes powered by a 50 Hz ac high voltage source. Current pulse amplitude distributions for different space gaps and the time separation between consecutive current pulses are studied. A probability distribution function is proposed to predict the experimental distribution function for the current pulse amplitudes and the occurrence of the transition regime of the pulse distribution. Breakdown voltage at different positions on the dielectric surface is suggested to be stochastic in nature. The simulated results based on the proposed distribution function agreed well with the experimental results and able to predict the regime of transition voltage. This model would be useful for the understanding of stochastic behaviors of DBD and the design of DBD device for effective operation and applications.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Note: Control of liquid helium supply to cryopanels of Kolkata superconducting cyclotron

The Kolkata superconducting cyclotron utilises liquid helium to cool the main magnet niobium-titanium (NbTi) coil and the cryopanels. Three liquid helium cooled cryopanels, placed inside the dees of the radio-frequency system, maintain the high vacuum in the acceleration region of the superconducting cyclotron. The small cryostat placed inside the cryogenic distribution manifold located at the basement of the superconducting cyclotron building supplies liquid helium in parallel branches to three cold heads, used for cooling their associated cryopanels. The level in the cryostat has to be maintained at an optimum value to ensure uninterrupted flow of liquid helium to these three cold heads. This paper describes the transfer function of the overall system, its tuning parameters, and discusses the actual control of cryostat level by using these parameters.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

3D numerical simulation of the long range propagation of acoustical shock waves through a heterogeneous and moving medium

Many situations involve the propagation of acoustical shock waves through flows. Natural sources such as lightning, volcano explosions, or meteoroid atmospheric entries, emit loud, low frequency, and impulsive sound that is influenced by atmospheric wind and turbulence. The sonic boom produced by a supersonic aircraft and explosion noises are examples of intense anthropogenic sources in the atmosphere. The Buzz-Saw-Noise produced by turbo-engine fan blades rotating at supersonic speed also propagates in a fast flow within the engine nacelle. Simulating these situations is challenging, given the 3D nature of the problem, the long range propagation distances relative to the central wavelength, the strongly nonlinear behavior of shocks associated to a wide-band spectrum, and finally the key role of the flow motion. With this in view, the so-called FLHOWARD (acronym for FLow and Heterogeneous One-Way Approximation for Resolution of Diffraction) method is presented with three-dimensional applications. A scalar nonlinear wave equation is established in the framework of atmospheric applications, assuming weak heterogeneities and a slow wind. It takes into account diffraction, absorption and relaxation properties of the atmosphere, quadratic nonlinearities including weak shock waves, heterogeneities of the medium in sound speed and density, and presence of a flow (assuming a mean stratified wind and 3D turbulent ? flow fluctuations of smaller amplitude). This equation is solved in the framework of the one-way method. A split-step technique allows the splitting of the non-linear wave equation into simpler equations, each corresponding to a physical effect. Each sub-equation is solved using an analytical method if possible, and finite-differences otherwise. Nonlinear effects are solved in the time domain, and others in the frequency domain. Homogeneous diffraction is handled by means of the angular spectrum method. Ground is assumed perfectly flat and rigid. Due to the 3D aspect, the code was massively parallelized using the single program, multiple data paradigm with the Message Passing Interfaces (MPI) for distributed memory architectures. This allows us to handle problems in the order of a thousand billion mesh points in the four dimensions (3 dimensions of space plus time). The validity of the method has been thoroughly evaluated on many cases with known solutions: linear piston, scattering of plane wave by a heterogeneous sphere, propagation in a waveguide with a shear flow, scattering by a finite amplitude vortex and nonlinear propagation in a thermoviscous medium. This validation process allows for a detailed assessment of the advantages and limitations of the method. Finally, applications to atmospheric propagation of shock waves will be presented.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DIRECT INTEGRATION OF THE COLLISIONLESS BOLTZMANN EQUATION IN SIX-DIMENSIONAL PHASE SPACE: SELF-GRAVITATING SYSTEMS

We present a scheme for numerical simulations of collisionless self-gravitating systems which directly integrates the Vlasov-Poisson equations in six-dimensional phase space. Using the results from a suite of large-scale numerical simulations, we demonstrate that the present scheme can simulate collisionless self-gravitating systems properly. The integration scheme is based on the positive flux conservation method recently developed in plasma physics. We test the accuracy of our code by performing several test calculations, including the stability of King spheres, the gravitational instability, and the Landau damping. We show that the mass and the energy are accurately conserved for all the test cases we study. The results are in good agreement with linear theory predictions and/or analytic solutions. The distribution function keeps the property of positivity and remains non-oscillatory. The largest simulations are run on 64{sup 6} grids. The computation speed scales well with the number of processors, and thus our code performs efficiently on massively parallel supercomputers.

79 ASTRONOMY AND ASTROPHYSICS↗

Radiation transport in kinetic simulations and the influence of photoemission on electron current in self-sustaining discharges

A kinetic description for electronic excitation of helium for principal quantum number n $\leqslant $ 4 has been included into a particle-in-cell (PIC) simulation utilizing direct simulation Monte Carlo (DSMC) for electron-neutral interactions. The excited electronic levels radiate state-dependent photons with wavelengths from the extreme ultraviolet (EUV) to visible regimes. Photon wavelengths are chosen according to a Voigt distribution accounting for the natural, pressure, and Doppler broadened linewidths. This method allows for reconstruction of the emission spectrum for a non-thermalized electron energy distribution function (EEDF) and investigation of high energy photon effects on surfaces, specifically photoemission. A parallel plate discharge with a fixed field (i.e. space charge neglected) is used to investigate the effects of including photoemission for a Townsend discharge. When operating at a voltage near the self-sustaining discharge threshold, it is observed that the electron current into the anode is higher when including photoemission from the cathode than without even when accounting for self-absorption from ground state atoms. As a result, the photocurrent has been observed to account for as much as 20% of the total current from the cathode under steady-state conditions.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Parallel transport modeling of linear divertor simulators with fundamental ion cyclotron heating *

Abstract The Material Plasma Exposure eXperiment (MPEX) is a steady state linear device with the goal to perform plasma material interaction studies at future fusion reactor relevant conditions. A prototype of MPEX referred as ‘Proto-MPEX’ is designed to carry out research and development related to source, heating and transport concepts on the planned full MPEX device. The auxiliary heating schemes in MPEX are based on cyclotron resonance heating with radio frequency (RF) waves. Ion cyclotron heating (ICH) and electron cyclotron heating in MPEX are used to independently heat the ions and electrons and provide fusion divertor conditions ranging from sheath-limited to fully detached divertor regimes at a material target. A hybrid particle-in-cell code- PICOS++ is developed and applied to understand the plasma parallel transport during ICH in MPEX/Proto-MPEX to the target. With this tool, evolution of the distribution function of MPEX/Proto-MPEX ions is modeled in the presence of (a) Coulomb collisions, (b) volumetric particle sources and (c) quasi-linear RF-based ICH. The code is benchmarked against experimental data from Proto-MPEX and simulation data from B2.5 EIRENE. The experimental observation of ‘density-drop’ near the target in Proto-MPEX and MPEX during ICH is demonstrated and explained via physics-based arguments using PICOS++ modeling. In fact, the density drops at the target during ICH in Proto-MPEX/MPEX to conserve the flux and to compensate for the increased flow during ICH. Furthermore, sensitivity scans of various plasma parameters with respect to ICH power are performed for MPEX to investigate its role on plasma transport and particle and energy fluxes at the target.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Efficient Probabilistic Visualization of Local Divergence of 2D Vector Fields with Independent Gaussian Uncertainty

This work focuses on visualizing uncertainty of local divergence of two-dimensional vector fields. Divergence is one of the fundamental attributes of fluid flows, as it can help domain scientists analyze potential positions of sources (positive divergence) and sinks (negative divergence) in the flow. However, uncertainty inherent in vector field data can lead to erroneous divergence computations, adversely impacting downstream analysis. While Monte Carlo (MC) sampling is a classical approach for estimating divergence uncertainty, it suffers from slow convergence and poor scalability with increasing data size and sample counts. Thus, we present a two-fold contribution that tackles the challenges of slow convergence and limited scalability of the MC approach. (1) We derive a closed-form approach for highly efficient and accurate uncertainty visualization of local divergence, assuming independently Gaussian-distributed vector uncertainties. (2) We further integrate our approach into Viskores, a platform-portable parallel library, to accelerate uncertainty visualization. In our results, we demonstrate significantly enhanced efficiency and accuracy of our serial analytical (speed-up up to 1946×) and parallel Viskores (speed-up up to 19698×) algorithms over the classical serial MC approach. We also demonstrate qualitative improvements of our probabilistic divergence visualizations over traditional mean-field visualization, which disregards uncertainty. We validate the accuracy and efficiency of our methods on wind forecast and ocean simulation datasets.

Ouermi, Timbwaoga [University of Utah]↗

Relative entropy and optimization-driven coarse-graining methods in VOTCA

We discuss recent advances of the VOTCA package for systematic coarse-graining. Two methods have been implemented, namely the downhill simplex optimization and the relative entropy minimization. We illustrate the new methods by coarse-graining SPC/E bulk water and more complex water-methanol mixture systems. The CG potentials obtained from both methods are then evaluated by comparing the pair distributions from the coarse-grained to the reference atomistic simulations.We have also added a parallel analysis framework to improve the computational efficiency of the coarse-graining process.

74 ATOMIC AND MOLECULAR PHYSICS↗

Efficient Machine Learning Approach for Optimizing Scientific Computing Applications on Emerging HPC Architectures

Efficient parallel implementations of scientific applications on multi-core CPUs with accelerators such as GPUs and Xeon Phis is challenging. This requires - exploiting the data parallel architecture of the accelerator along with the vector pipelines of modern x86 CPU architectures, load balancing, and efficient memory transfer between different devices. It is relatively easy to meet these requirements for highly structured scientific applications. In contrast, a number of scientific and engineering applications are unstructured. Getting performance on accelerators for these applications is extremely challenging because many of these applications employ irregular algorithms which exhibit data-dependent control-ow and irregular memory accesses. Furthermore, these applications are often iterative with dependency between steps, and thus making it hard to parallelize across steps. As a result, parallelism in these applications is often limited to a single step. Numerical simulation of charged particles beam dynamics is one such application where the distribution of work and memory access pattern at each time step is irregular. Applications with these properties tend to present significant branch and memory divergence, load imbalance between different processor cores, and poor compute and memory utilization. Prior research on parallelizing such irregular applications have been focused around optimizing the irregular, data-dependent memory accesses and control-ow during a single step of the application independent of the other steps, with the assumption that these patterns are completely unpredictable. We observed that the structure of computation leading to control-ow divergence and irregular memory accesses in one step is similar to that in the next step. It is possible to predict this structure in the current step by observing the computation structure of previous steps. In this dissertation, we present novel machine learning based optimization techniques to address the parallel implementation challenges of such irregular applications on different HPC architectures. In particular, we use supervised learning to predict the computation structure and use it to address the control-ow and memory access irregularities in the parallel implementation of such applications on GPUs, Xeon Phis, and heterogeneous architectures composed of multi-core CPUs with GPUs or Xeon Phis. We use numerical simulation of charged particles beam dynamics simulation as a motivating example throughout the dissertation to present our new approach, though they should be equally applicable to a wide range of irregular applications. The machine learning approach presented here use predictive analytics and forecasting techniques to adaptively model and track the irregular memory access pattern at each time step of the simulation to anticipate the future memory access pattern. Access pattern forecasts can then be used to formulate optimization decisions during application execution which improves the performance of the application at a future time step based on the observations from earlier time steps. In heterogeneous architectures, forecasts can also be used to improve the memory performance and resource utilization of all the processing units to deliver a good aggregate performance. We used these optimization techniques and anticipation strategy to design a cache-aware, memory efficient parallel algorithm to address the irregularities in the parallel implementation of charged particles beam dynamics simulation on different HPC architectures. Experimental result using a diverse mix of HPC architectures shows that our approach in using anticipation strategy is effective in maximizing data reuse, ensuring workload balance, minimizing branch and memory divergence, and in improving resource utilization.

Arumugam, Kamesh↗

ARPA E GO Competition

A Nonlinear Programming SC-ACOPF Framework with Parallel Computing Capabilities

24 POWER TRANSMISSION AND DISTRIBUTION↗

Multi-fidelity is the new annealing: Gradient-free learning of posterior densities via transport maps

To tackle concentrated, multi-modal Bayesian inference problems, we propose using an annealed importance sampling procedure. To do this, we form a sequence of annealed distributions and employ transport maps to act as a surrogate of each distribution. This process is demonstrated on a few examples, favorably showing its potential for efficiently parallelizing the process of PDE evaluations and allowing for surrogates that we can sample from exactly.

van Bloemen Waanders, Bart G [Sandia National Labo↗

RELATIVISTIC CYCLOTRON INSTABILITY IN ANISOTROPIC PLASMAS

A sufficiently large temperature anisotropy can sometimes drive various types of electromagnetic plasma micro-instabilities, which can play an important role in the dynamics of relativistic pair plasmas in space, astrophysics, and laboratory environments. Here, we provide a detailed description of the cyclotron instability of parallel propagating electromagnetic waves in relativistic pair plasmas on the basis of a relativistic anisotropic distribution function. Using plasma kinetic theory and particle-in-cell simulations, we study the influence of the relativistic temperature and the temperature anisotropy on the collective and noncollective modes of these plasmas. Growth rates and dispersion curves from the linear theory show a good agreement with simulations results.

79 ASTRONOMY AND ASTROPHYSICS↗

Eliminating Yield Anisotropy and Enhancing Ductility in Mg Alloys by Shear Assisted Processing and Extrusion

Solid phase processing techniques such as friction stir welding, Shear assisted processing and extrusion (ShAPE)/ friction extrusion and cold spray have been successfully demonstrated as promising thermomechanical methods to produce metallic materials with enhanced performance. In this study, AZ series with and without silicon, ZK60 Mg alloys in as-received forms (as-cast or as-extruded) were processed using Shear Assisted Processing and Extrusion (ShAPE). Microstructural characterization was performed using EBSD and TEM and revealed that as compared to the feedstock materials/ billets, friction extruded Mg alloys had more uniform microstructure, equiaxed grains, finer and homogeneously distributed precipitates and chemical homogeneity. It was also observed that basal planes were not oriented parallel to extrusion axis. As a result, rod products exhibited significantly reduced (in some cases eliminated) yield asymmetry and achieved enhanced ductility, which were uncommon or difficult to attain using conventional processing techniques. In addition, modified texture likely suppressed deformation twinning under compressive deformation.

Shear Assisted Processing and Extrusion, SHAPE, ma↗