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 1,189 records · Page 66

MICADO: Parallel implementation of a 2D-1D iterative algorithm for the 3D neutron transport problem in prismatic geometries

The large increase in computing power over the past few years now makes it possible to consider developing 3D full-core heterogeneous deterministic neutron transport solvers for reference calculations. Among all approaches presented in the literature, the method first introduced in [1] seems very promising. It consists in iterating over resolutions of 2D and ID MOC problems by taking advantage of prismatic geometries without introducing approximations of a low order operator such as diffusion. However, before developing a solver with all industrial options at EDF, several points needed to be clarified. In this work, we first prove the convergence of this iterative process, under some assumptions. We then present our high-performance, parallel implementation of this algorithm in the MICADO solver. Benchmarking the solver against the Takeda case shows that the 2D-1D coupling algorithm does not seem to affect the spatial convergence order of the MOC solver. As for performance issues, our study shows that even though the data distribution is suited to the 2D solver part, the efficiency of the ID part is sufficient to ensure a good parallel efficiency of the global algorithm. After this study, the main remaining difficulty implementation-wise is about the memory requirement of a vector used for initialization. An efficient acceleration operator will also need to be developed. (authors)

97 MATHEMATICS AND COMPUTING↗

Electrified vehicle DC power conversion with distributed control

An electrified vehicle has a battery pack comprising series-connected battery units providing a main voltage. To supply a lower voltage bus equally from all battery units, a plurality of DC/DC converters are powered by respective units and have their outputs in parallel. A central module has an outer loop controller generating a target current to regulate the bus voltage to a predetermined voltage and has an allocator distributing the target current into a plurality of allocated current commands according to respective states of charge of the battery units. A plurality of local controllers each adjusts a current of a respective DC/DC converter. Each local controller receives a respective allocated current command as a respective feedforward control variable. Each local controller uses an error between the bus voltage and the predetermined voltage to be integrated as a respective feedback control variable only when the error is above a threshold.

33 ADVANCED PROPULSION SYSTEMS↗

Towards Lightweight Data Integration Using Multi-Workflow Provenance and Data Observability

Modern large-scale scientific discovery requires multidisciplinary collaboration across diverse computing facilities, including High Performance Computing (HPC) machines and the Edge-to-Cloud continuum. Integrated data analysis plays a crucial role in scientific discovery, especially in the current AI era, by enabling Responsible AI development, FAIR, Reproducibility, and User Steering. However, the heterogeneous nature of science poses challenges such as dealing with multiple supporting tools, cross-facility environments, and efficient HPC execution. Building on data observability, adapter system design, and provenance, we propose MIDA: an approach for lightweight runtime Multi-workflow Integrated Data Analysis. MIDA defines data observability strategies and adaptability methods for various parallel systems and machine learning tools. With observability, it intercepts the dataflows in the background without requiring instrumentation while integrating domain, provenance, and telemetry data at runtime into a unified database ready for user steering queries. We conduct experiments showing end-to-end multi-workflow analysis integrating data from Dask and MLFlow in a real distributed deep learning use case for materials science that runs on multiple environments with up to 276 GPUs in parallel. We show near-zero overhead running up to 100,000 tasks on 1,680 CPU cores on the Summit supercomputer.

Santos Souza, Renan↗

Stress-induced α″ phase in a beta Ti–19Nb–1.5Mo–4Zr–8Sn alloy

Highlights: • Morphologies/distributions of α″ produced by stretching/rolling are similar. • Under small deformation, α″ martensite consists of parallel lamellar structures. • Under larger deformations, martensite has ladder-like pattern and parallel ellipsoids. • With the increase of deformation hardness decreases firstly and then increases. - Abstract: The morphological evolution of the stress-induced α″ martensite with increases in the stretching and rolling deformation degrees as well as the related orientation relationship were investigated in a β titanium alloy (Ti–19Nb–1.5Mo–4Zr–8Sn)·For a stretching deformation of 1%, the α″ martensite exhibited a packet morphology consisting of parallel lamellar structures. On the other hand, in the 2%-rolled sample, the α″ martensite exhibited a long lath-like distribution, which was accompanied by a few small pieces. With an increase in the deformation degree, the α″ martensite gradually exhibited a ladder-like structure consisting of long and short α″ laths or ellipsoids. The existence of these self-accommodating martensitic variants decreased the elastic energy. The parallel α″ laths present within a single packet were found to have the same orientation. The interface between the α″ and β phases was coherent, and their lattice constant relationships were identified as √2a{sub β}≈b{sub α″}≈c{sub α″} and a{sub β} ≈ a{sub α″}. Further, with the increase in the deformation degree, the hardness of the alloy first decreased and then increased; this was the case both the stretched as well as the rolled samples.

36 MATERIALS SCIENCE↗

White x-ray microdiffraction analysis of defects, strain and tilts in a free standing GaN film.

A novel white-beam microdiffraction analysis of defects, strains and tilts in a free standing m-plane GaN film grown via hydride vapor phase epitaxy is presented. It is shown that misfit dislocations are grouped within cell boundaries creating local lattice rotations (tilts) between the growing cells. Distribution of lattice rotations in the film is not homogeneous. Regions of large rotations are separated by low rotations regions. The dominating rotation axis is parallel [11{bar 2}0] direction. High in plane shear stress component is observed along [0001].

36 MATERIALS SCIENCE↗

Kinetic ion effect on geodesic acoustic Alfven modes in tokamaks

Using a quasitoroidal set of coordinates with coaxial circular magnetic surfaces, the Vlasov equation is solved for collisionless plasmas, and the dielectric tensor is found for large aspect ratio tokamaks in a low frequency band. Taking into account q-profile and charge separation parallel electric field, it is found that the Alfven wave continuum is deformed by ion geodesic effects producing continuum minimum at the rational magnetic surfaces. Low frequency geodesic ion induced Alfven waves are found below the continuum minimum where collisionless damping has a gap for Maxwell distribution. In kinetic approach, the ion thermal motion defines the geodesic effect but the mode frequency is strongly corrected due to parallel motion of electrons.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Development of flow network analysis code for block type VHTR core by linear theory method

VHTR (Very High Temperature Reactor) is high-efficiency nuclear reactor which is capable of generating hydrogen with high temperature of coolant. PMR (Prismatic Modular Reactor) type reactor consists of hexagonal prismatic fuel blocks and reflector blocks. The flow paths in the prismatic VHTR core consist of coolant holes, bypass gaps and cross gaps. Complicated flow paths are formed in the core since the coolant holes and bypass gap are connected by the cross gap. Distributed coolant was mixed in the core through the cross gap so that the flow characteristics could not be modeled as a simple parallel pipe system. It requires lot of effort and takes very long time to analyze the core flow with CFD analysis. Hence, it is important to develop the code for VHTR core flow which can predict the core flow distribution fast and accurate. In this study, steady state flow network analysis code is developed using flow network algorithm. Developed flow network analysis code was named as FLASH code and it was validated with the experimental data and CFD simulation results. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Transport of cosmic-ray protons in intermittent heliospheric turbulence: Model and simulations

The transport of charged energetic particles in the presence of strong intermittent heliospheric turbulence is computationally analyzed based on known properties of the interplanetary magnetic field and solar wind plasma at 1 astronomical unit. The turbulence is assumed to be static, composite, and quasi-three-dimensional with a varying energy distribution between a one-dimensional Alfvénic (slab) and a structured two-dimensional component. The spatial fluctuations of the turbulent magnetic field are modeled either as homogeneous with a Gaussian probability distribution function (PDF), or as intermittent on large and small scales with a q-Gaussian PDF. Simulations showed that energetic particle diffusion coefficients both parallel and perpendicular to the background magnetic field are significantly affected by intermittency in the turbulence. This effect is especially strong for parallel transport where for large-scale intermittency results show an extended phase of subdiffusive parallel transport during which cross-field transport diffusion dominates. The effects of intermittency are found to depend on particle rigidity and the fraction of slab energy in the turbulence, yielding a perpendicular to parallel mean free path ratio close to 1 for large-scale intermittency. Investigation of higher order transport moments (kurtosis) indicates that non-Gaussian statistical properties of the intermittent turbulent magnetic field are present in the parallel transport, especially for low rigidity particles at all times.

79 ASTRONOMY AND ASTROPHYSICS↗

Experimental evaluation of a spatial resampling technique to improve the accuracy of pencil-beam dose calculation in proton therapy

Purpose: In proton therapy, pencil-beam algorithms (PBAs) are the most widely used dose calculation methods. However, the PB calculations that employ one-dimensional density scaling neglect the effects of lateral density heterogeneity on the dose distributions, whereas some particles included in such pencil beams could overextend beyond the interface of the density heterogeneity. We have simplified a pencil-beam redefinition algorithm (PBRA), which was proposed for electron therapy, by a spatial resampling technique toward an application for proton therapy. The purpose of this study is to evaluate the calculation results of the spatial resampling technique in terms of lateral density heterogeneity by comparison with the dose distributions that were measured in heterogeneous slab phantoms. Methods: The pencil beams are characterized for multiple residual-range (i.e., proton energy) bins. To simplify the PBRA, the given pencil beams are resampled on one or two transport planes, in which smaller sub-beams that are parallel to each other are generated. We addressed the problem of lateral density heterogeneity comparing the calculation results to the dose distributions measured at different depths in heterogeneous slab phantoms using a two-dimensional detector. Two heterogeneity slab phantoms, namely, phantoms A and B, were designed for the measurements and calculations. In phantom A, the heterogeneity slab was placed close to the surface. On the other hand, in phantom B, it was placed close to the Bragg peak in the mono-energetic proton beam. Results: In measurements, lateral dose profiles showed a dose reduction and increment in the vicinity of x= 0 mm in both phantoms at depths z= 142 and 161 mm due to lateral particle disequilibrium. In phantom B, these dose reduction/increment effects were higher/lower, respectively, than those in phantom A. This is because a longer distance from the surface to the heterogeneous slab increases the strength of proton scattering. Sub-beams, which were generated from the resampling plane, formed a detouring/overextending path that was different from that of elemental pencil beams. Therefore, when the spatial resampling was implemented at the surface and immediately upstream of the lateral heterogeneity, the calculation could predict these dose reduction/increment effects. Without the resampling procedure, these dose reduction/increment effects could not be predicted in both phantoms owing to the blurring of the pencil beam. We found that the PBA with the spatial resampling technique predicted the dose reduction/increment at the dose profiles in both phantoms when the sampling plane was defined immediately upstream of the heterogeneous slab. Conclusions: We have demonstrated the implementation of a spatial resampling technique for pencil-beam calculation to address the problem of lateral density heterogeneity. While further validation is required for clinical use, this study suggests that the spatial resampling technique can make a significant contribution to proton therapy.

61 RADIATION PROTECTION AND DOSIMETRY↗

A simple backprojection algorithm for 3D in vivo EPID dosimetry of IMRT treatments

Treatment plans are usually designed, optimized, and evaluated based on the total 3D dose distribution, motivating a total 3D dose verification. The purpose of this study was to develop a 2D transmission-dosimetry method using an electronic portal imaging device (EPID) into a simple 3D method that provides 3D dose information. In the new method, the dose is reconstructed within the patient volume in multiple planes parallel to the EPID for each gantry angle. By summing the 3D dose grids of all beams, the 3D dose distribution for the total treatment fraction is obtained. The algorithm uses patient contours from the planning CT scan but does not include tissue inhomogeneity corrections. The 3D EPID dosimetry method was tested for IMRT fractions of a prostate, a rectum, and a head-and-neck cancer patient. Planned and in vivo-measured dose distributions were within 2% at the dose prescription point. Within the 50% isodose surface of the prescribed dose, at least 97% of points were in agreement, evaluated with a 3D {gamma} method with criteria of 3% of the prescribed dose and 0.3 cm. Full 3D dose reconstruction on a 0.1x0.1x0.1 cm{sup 3} grid and 3D {gamma} evaluation took less than 15 min for one fraction on a standard PC. The method allows in vivo determination of 3D dose-volume parameters that are common in clinical practice. The authors conclude that their EPID dosimetry method is an accurate and fast tool for in vivo dose verification of IMRT plans in 3D. Their approach is independent of the treatment planning system and provides a practical safety net for radiotherapy.

61 RADIATION PROTECTION AND DOSIMETRY↗

CUDA Computation of the Feynman Distribution

In 2006, NVIDIA Corporation introduced the Computer Unified Device Architecture (CUDA) for C, C++, and Fortran programming languages. The CUDA parallel computing platform uses the Graphics Processing Unit (GPU) of NVIDIA graphics card, and it is designed for parallel computations. In these computations, the same algorithms are executed on the data in parallel. The GPU contains many Streaming Multiprocessors (SM) that can independently execute blocks of threads. A thread is a sequence of serial programmed instructions. The SM executes threads in groups of 32 and a group of 32 threads is called warp. The threads within a block can be synchronized and have access to a shared memory that is visible only to them. Conversely, blocks of threads cannot be synchronized and have access only to the global memory of the graphics card. This memory is called device memory and is independent of the host memory controlled by the Central Processing Unit (CPU). A single thread has access to a local memory (register) that is not visible to any other thread. The part of the program that is executed on the graphics card is called kernel. Each thread of the kernel has a unique identification index that is accessible by the in-built variable threadIdx; the latter is a 3-component vector. A threads block can contain up to 1024 threads. Blocks of threads have a unique identification index that is accessible by the in-built variable blockIdx; the latter is a 3-component vector. The in-built variable blockDim contains the number of threads per block; this variable is a 3-component vector. Threads and blocks represent different levels of parallelism; a kernel is executed as a grid of blocks of threads. Figure 1 illustrates a grid of thread blocks for a GPU with 2 SM. A SM can execute multiple blocks of threads in parallel; a block of threads is executed on a single SM. A GPU can execute multiple kernels concurrently. This work has used the Quadro M5000 graphics card with the following technical specifications: - 2048 cores with 861 MHz frequency, - 16 streaming multiprocessors, - 5.2 trillion transistors, - 8.2 Gb device memory, - 64 kb register memory (per thread), - 48 kb shared memory (per block) - 211.6 GB/s memory bandwidth, and - 3.5 GFLOPS floating point performance.

97 MATHEMATICS AND COMPUTING↗

Future Grid State Modeling for Transactive Systems

The simulation design and evaluation of a transactive energy system is contingent on the setting in which it would exist, which could vary significantly with the many drivers that impact the potential future states of the power system. This report summarizes the development of the Transactive Future Grid State (TFGS) model, which is a lightweight capacity expansion model that will support transactive energy research by allowing users to explore alternative future states. The TFGS model determines the long-term generation capacity needs of a given power system that satisfy specified constraints and support a least cost hourly dispatch. This is accomplished by solving a single optimization problem that simultaneously finds the generation fleet with capacity sufficient to meet peak load and the flexibility to meet hourly operational dispatch, all at lowest cost. The TFGS model was applied to a larger ongoing study taking place that will be simulating how distribution system operators (DSOs) could use a transactive energy system to manage distributed energy resources. The larger study, commonly referred to as the DSO+T study, involves two parallel analyses; one that represents the system today and one that represents a future system with a high degree of renewable energy penetration. The TFGS model provided the generation mix of the high renewable base case. As the TFGS model was being used to provide inputs to the DSO+T study, two issues became apparent. First, additional functionality would have been useful, and second, a larger-than-anticipated amount of user inputs and data were required to successfully use the TFGS model. These discoveries have lead to the recommendation of specific functionality that can be added in future development of the tool.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Comparison of Distributive Consensus Algorithms for AnomalyDetection on the Power-Grid

As an increasing number of devices in the power-grid rely on high-precision clocks to measure electrical properties and time stamp data, GPS receivers have become a vulnerability on the power-grid. GPS spoofing attacks, in which a fake signal is transmitted, can cause sensor measurement errors that can misinform protective equipment and grid operators for both the transmission and distribution grid. The technology being developed to secure precision clocks in the transmission grid are not appropriate for the distribution grid. Thus, there is a need to developed anomaly detection algorithms that are suitable for the hardware available on the distribution grid. Traditionally, a distributed detection framework is comprised of spatially distributed nodes which acquire observations about a quantity of interest and sends the data to the fusion center, where a global decision is made. Instead the distribution grid has low-power devices, such as smart inverters and microPMUs, that could be used to perform highly parallelized calculations. Having on node calculations is beneficial since it is more network and computation efficient, as the more calculations can be performed where the data is being captured instead of communicating across a slow network to a centralized fusion center. Instead, we could employ an alternate peer-to-peer type of information exchange on the distribution grid that can still reach a global decision. A decentralized approach to anomaly detection makes efficient use of the computational and network bandwidth that exists, and it also lacks the single point of failure of a centralized approach. Thus, a decentralized local information exchange and/or inference is preferred. Lastly, the unreliability of the distribution grid network connections and potential noise or error in the measurements must be considered. In order to produce an effective detection tool, a Byzantine-fault tolerant and asynchronous algorithm must be developed. Thus, the first phase of our project is to develop a distributed, decentralized, asynchronous, Byzantine-fault tolerant, collaborative autonomy-based anomaly detection algorithm to collectively analyze clusters of data for signs of a malicious attack.

97 MATHEMATICS AND COMPUTING↗

Theoretical Plasma Physics

Lattice Boltzmann algorithms are a mesoscopic method to solve problems in nonlinear physics which are highly parallelized – unlike the direction solution of the original problem. These methods are applied to both fluid and magnetohydrodynamic turbulence. By introducing entropic constraints one can enforce the positive definiteness of the distribution functions and so be able to simulate fluids at high Reynolds numbers without numerical instabilities. By introducing a vector distribution function for the magnetic field one can enforce the divergence free condition on the magnetic field automatically, without the need of divergence cleaning as needed in most direct numerical solutions of the resistive magnetohydrodynamic equations. The principal reason for the high parallelization of lattice Boltzmann codes is that they consist of a kinetic collisional relaxation step (which is purely local) followed by a simple shift of the relaxed data to neighboring lattice sites. In large eddy simulations, the closure schemes are highly nonlocal – the most famous of these schemes is that due to Smagorinsky. Under a lattice Boltzmann representation the Smagorinsky closure is purely local – being simply a particular moment on the perturbed distribution fucntions. After nonlocal fluid moment models were discovered to represent Landau damping, it was found possible to model these fluid models using an appropriate lattice Boltzmann algorithm. The close to ideal parallelization of the lattice Boltzmann codes permitted us to be Gordon Bell finalists on using the Earth Simulation in Japan. We have also been involved in the radio frequency propagation of waves into a tokamak and into a spherical overdense tokamak plasma. Initially we investigated the use of a quasi-optical grill for the launching of lower hybrid waves into a tokamak. It was found that the conducting walls do not prevent the rods from being properly irradiated, the overloading of the quasi-optical grill is not severe with the electric field only being about three times higher than in the ideal case. Moreover, the quasi-optical grill was significantly fewer structural elements that the multijunction grill. Nevertheless there has not been much interest from experimental fusion groups to implementing these structures. Hence we have returned to optimizing the multijunction grill so that the large number of coupling matrix elements can be efficiently evaluated using symmetry arguments. In overdense plasmas, the standard electromagnetic waves cannot propagate into the plasma center, but are reflected at the plasma edge. By optimizing mode conversion processes (in particular, the O-X-B wave propagation of Ordinary Mode converting to an Extraordinary mode which then converts into an electrostatic Bernstein wave) one can excite within the plasma an electrostatic Bernstein wave that does not suffer density cutoffs and is absorbed on the electron cyclotron harmonics. Finally we have started looking at other mesoscopic lattice algorithms that involve unitary collision and streaming steps. Because these algorithms are unitary they can be run on quantum computers when they become available – unlike their computational cousin of lattice Boltzmann which is a purely classical code. These quantum lattice gas algorithms have been tested successfully on exact analytic soliton collision solution. These calculations are hoped to be able to study Bose Einstein condensed atomic gases and their ground states in an optical lattice.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

OpenMP and MPI implementations of an elasto-viscoplastic fast Fourier transform-based micromechanical solver for fast crystal plasticity modeling

Here, we explore several parallel implementations of an elasto-viscoplastic fast Fourier transform (EVPFFT) model using Message Passing Interface (MPI), OpenMP, and a hybrid of MPI and OpenMP to efficiently predict micromechanical response of polycrystals. Performance studies using EVPFFT are performed based on domain decomposition over voxels of a periodic cell, which is a representative volume element (RVE) of polycrystalline copper. We begin by parallelizing the computationally intensive Newton–Raphson (NR) single crystal solver within EVPFFT. Next, we compare the performance of the serial and parallel FFTW (Fastest Fourier Transform in the West) using OpenMP (OpenMP-FFTW) and MPI (MPI-FFTW) with the original Numerical Recipes-based FOURN routine within EVPFFT. In the parallel environment, we find that the FFT calculations are best performed using the MPI version of FFTW. Finally, the remainder of the code, except read/write subroutines, is parallelized. Significant speedups of the original EVPFFT model are achieved using MPI on shared memory multicore workstations. Furthermore, results achieved on a distributed memory Cray supercomputer show promising strong and weak scalability and in some cases even super scalability for the single crystal NR solver in EVPFFT. MPI-FFTW also scales perfectly for microstructure RVEs larger than 64 3 FFT voxels. For example, the MPI-EVPFFT parallel version of the code accelerates the simulations for approximately two orders of magnitude using 64 cores over the old serial code for an RVE size of 128 3 . Finally, the parallel EVPFFT code developed in this work can run massive voxel-based microstructural RVEs taking the advantages of thousands of logical cores provided by more advanced clusters.

42 ENGINEERING↗

Regulation of Solar Wind Electron Temperature Anisotropy by Collisions and Instabilities

Abstract Typical solar wind electrons are modeled as being composed of a dense but less energetic thermal “core” population plus a tenuous but energetic “halo” population with varying degrees of temperature anisotropies for both species. In this paper, we seek a fundamental explanation of how these solar wind core and halo electron temperature anisotropies are regulated by combined effects of collisions and instability excitations. The observed solar wind core/halo electron data in ( β ∥ , T ⊥ / T ∥ ) phase space show that their respective occurrence distributions are confined within an area enclosed by outer boundaries. Here, T ⊥ / T ∥ is the ratio of perpendicular and parallel temperatures and β ∥ is the ratio of parallel thermal energy to background magnetic field energy. While it is known that the boundary on the high- β ∥ side is constrained by the temperature anisotropy-driven plasma instability threshold conditions, the low- β ∥ boundary remains largely unexplained. The present paper provides a baseline explanation for the low- β ∥ boundary based upon the collisional relaxation process. By combining the instability and collisional dynamics it is shown that the observed distribution of the solar wind electrons in the ( β ∥ , T ⊥ / T ∥ ) phase space is adequately explained, both for the “core” and “halo” components.

Yoon, Peter H. (ORCID:0000000181343790)↗

Electron tomography of (In,Ga)N insertions in GaN nanocolumns grown on semi-polar (112{sup -}2) GaN templates

We present results of scanning transmission electron tomography on GaN/(In,Ga)N/GaN nanocolumns (NCs) that grew uniformly inclined towards the patterned, semi-polar GaN(112{sup -}2) substrate surface by molecular beam epitaxy. For the practical realization of the tomographic experiment, the nanocolumn axis has been aligned parallel to the rotation axis of the electron microscope goniometer. The tomographic reconstruction allows for the determination of the three-dimensional indium distribution inside the nanocolumns. This distribution is strongly interrelated with the nanocolumn morphology and faceting. The (In,Ga)N layer thickness and the indium concentration differ between crystallographically equivalent and non-equivalent facets. The largest thickness and the highest indium concentration are found at the nanocolumn apex parallel to the basal planes.

36 MATERIALS SCIENCE↗