Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “parallel coordinates”

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 91 records · Page 5

HPC formulations of optimization algorithms for tensor completion

Tensor completion is a powerful tool used to estimate or recover missing values in multi-way data. It has seen great success in domains such as product recommendation and healthcare. Tensor completion is most often accomplished via low-rank sparse tensor factorization, a computationally expensive non-convex optimization problem which has only recently been studied in the context of parallel computing. In this work, we study three optimization algorithms that have been successfully applied to tensor completion: alternating least squares (ALS), stochastic gradient descent (SGD), and coordinate descent (CCD++). We explore opportunities for parallelism on shared- and distributed-memory systems and address challenges such as memory- and operation-efficiency, load balance, cache locality, and communication. Among our advancements are a communication-efficient CCD++ algorithm, an ALS algorithm rich in level-3 BLAS routines, and an SGD algorithm which combines stratification with asynchronous communication. Furthermore, we show that introducing randomization during ALS and CCD++ can accelerate convergence. We evaluate our parallel formulations on a variety of real datasets on a modern supercomputer and demonstrate speedups through 16384 cores. Further, these improvements reduce time-to-solution from hours to seconds on real-world datasets. We show that after our optimizations, ALS is advantageous on parallel systems of small-to-moderate scale, while both ALS and CCD++ provide the lowest time-to-solution on large-scale distributed systems.

97 MATHEMATICS AND COMPUTING↗

Breast Patient Setup Error Assessment: Comparison of Electronic Portal Image Devices and Cone-Beam Computed Tomography Matching Results

Purpose: To quantify the differences in setup errors measured with the cone-beam computed tomography (CBCT) and electronic portal image devices (EPID) in breast cancer patients. Methods and Materials: Repeat CBCT scan were acquired for routine offline setup verification in 20 breast cancer patients. During the CBCT imaging fractions, EPID images of the treatment beams were recorded. Registrations of the bony anatomy for CBCT to planning CT and EPID to digitally reconstructed-radiographs (DRRs) were compared. In addition, similar measurements of an anthropomorphic thorax phantom were acquired. Bland-Altman and linear regression analysis were performed for clinical and phantom registrations. Systematic and random setup errors were quantified for CBCT and EPID-driven correction protocols in the EPID coordinate system (U, V), with V parallel to the cranial-caudal axis and U perpendicular to V and the central beam axis. Results: Bland-Altman analysis of clinical EPID and CBCT registrations yielded 4 to 6-mm limits of agreement, indicating that both methods were not compatible. The EPID-based setup errors were smaller than the CBCT-based setup errors. Phantom measurements showed that CBCT accurately measures setup error whereas EPID underestimates setup errors in the cranial-caudal direction. In the clinical measurements, the residual bony anatomy setup errors after offline CBCT-based corrections were {Sigma}{sub U} = 1.4 mm, {Sigma}{sub V} = 1.7 mm, and {sigma}{sub U} = 2.6 mm, {sigma}{sub V} = 3.1 mm. Residual setup errors of EPID driven corrections corrected for underestimation were estimated at {Sigma}{sub U} = 2.2mm, {Sigma}{sub V} = 3.3 mm, and {sigma}{sub U} = 2.9 mm, {sigma}{sub V} = 2.9 mm. Conclusion: EPID registration underestimated the actual bony anatomy setup error in breast cancer patients by 20% to 50%. Using CBCT decreased setup uncertainties significantly.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

2018 Benefits Study of Emergency IEA Oil Stocks: Final CRADA Report

This CRADA work was performed with the International Energy Agency (IEA) by Oak Ridge National Laboratory (ORNL) to support ongoing IEA, and U.S. DOE, discussions on future oil market challenges by IEA member countries. ORNL completed an update of a 2012 study for the IEA that accounts for changes in oil market and global economic conditions since 2013. The ORNL study evaluates the collective benefits of IEA emergency oil stocks, counting benefits to all countries, not just IEA, that are net oil importers. The IEA provided oil market data and projections, provided recommendations on parameter values and key case assumptions, in coordination with IEA member countries and coordinated the ORNL study with a parallel study on stockholding costs by an IEA subcontractor. IEA also combined the results of the ORNL benefits study and the IEA-funded cost study in a summary report by IEA, Insights Series 2018 - Costs and Benefits of Emergency Stockholding.

02 PETROLEUM↗

AGEING PROCEDURES ON LITHIUM BATTERIES IN AN INTERNATIONAL COLLABORATION CONTEXT

The widespread introduction of electrically-propelled vehicles is currently part of many political strategies and introduction plans. These new vehicles, ranging from limited (mild) hybrid to plug-in hybrid to fully-battery powered, will rely on a new class of advanced storage batteries, such as those based on lithium, to meet different technical and economical targets. The testing of these batteries to determine the performance and life in the various applications is a time-consuming and costly process that is not yet well developed. There are many examples of parallel testing activities that are poorly coordinated, for example, those in Europe, Japan and the US. These costs and efforts may be better leveraged through international collaboration, such as that possible within the framework of the International Energy Agency. Here, a new effort is under development that will establish standardized, accelerated testing procedures and will allow battery testing organizations to cooperate in the analysis of the resulting data. This paper reviews the present state-of-the-art in accelerated life testing in Europe, Japan and the US. The existing test procedures will be collected, compared and analyzed with the goal of international collaboration.

25 ENERGY STORAGE↗

Grand challenges of wind energy science – meeting the needs and services of the power system

The share of wind power in power systems is increasing dramatically, and this is happening in parallel with increased penetration of solar photovoltaics, storage, other inverter-based technologies, and electrification of other sectors. Recognising the fundamental objective of power systems, maintaining supply–demand balance reliably at the lowest cost, and integrating all these technologies are significant research challenges that are driving radical changes to planning and operations of power systems globally. In this changing environment, wind power can maximise its long-term value to the power system by balancing the needs it imposes on the power system with its contribution to addressing these needs with services. A needs and services paradigm is adopted here to highlight these research challenges, which should also be guided by a balanced approach, concentrating on its advantages over competitors. The research challenges within the wind technology itself are many and varied, with control and coordination internally being a focal point in parallel with a strong recommendation for a holistic approach targeted at where wind has an advantage over its competitors and in coordination with research into other technologies such as storage, power electronics, and power systems.

17 WIND ENERGY↗

Coordinate Projection-based Solver for ODE with Invariants

CPODES is a general purpose (serial and parallel) solver for systems of ordinary differential equation (ODE) with invariants. It implements a coordinate projection approach using different types of projection (orthogonal or oblique) and one of several methods for the decomposition of the Jacobian of the invariant equations.

Serban, Radu↗

Full-field drift Hamiltonian particle orbits in axisymmetric tokamak geometry

A Hamiltonian/Lagrangian theory to describe guiding center orbit drift motion that is canonical in Boozer magnetic coordinates is developed to include full electrostatic and electromagnetic perturbed fields in axisymmetric tokamak geometry. Furthermore, the radial component of the equilibrium magnetic field in the covariant representation is retained and the background equilibrium state extends to anisotropic plasma pressure conditions. A gauge transformation on the perturbed vector potential is imposed to guarantee canonical structure in the Boozer frame. Perturbed field nonlinear wave-wave interactions affect only the evolution of the guiding center particle parallel gyroradius. The evolution of the particle coordinate positions retains only linear wave-particle interactions. For particle motion in magnetohydrodynamic (MHD) instability structures, the electrostatic potential is linked mainly to the binormal component of the perturbed displacement vector when finite {delta}A{sub perpendicular} components are included.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Accelerating Collective Communication in Data Parallel Training across Deep Learning Frameworks

This work develops new techniques within Horovod, a generic communication library supporting data parallel training across deep learning frameworks. In particular, we improve the Horovod control plane by implementing a new coordination scheme that takes advantage of the characteristics of the typical data parallel training paradigm, namely the repeated execution of collectives on the gradients of a fixed set of tensors. Using a caching strategy, we execute Horovod’s existing coordinator-worker logic only once during a typical training run, replacing it with a more efficient decentralized orchestration strategy using the cached data and a global intersection of a bitvector for the remaining training duration. Next, we introduce a feature for end users to explicitly group collective operations, enabling finer grained control over the communication buffer sizes. To evaluate our proposed strategies, we conduct experiments on a world-class supercomputer — Summit. We compare our proposals to Horovod’s original design and observe 2x performance improvement at a scale of 6000 GPUs; we also compare them against tf.distribute and torch.DDP and achieve 12% better and comparable performance, respectively, using up to 1536 GPUs; we compare our solution against BytePS in typical HPC settings and achieve about 20% better performance on a scale of 768 GPUs. Finally, we test our strategies on a scientific application (STEMDL) using up to 27,600 GPUs (the entire Summit) and show that we achieve a near-linear scaling of 0.93 with a sustained performance of 1.54 exaflops (with standard error +- 0.02) in FP16 precision.

Romero, Joshua↗

Abacavir and warfarin modulate allosterically kinetics of NO dissociation from ferrous nitrosylated human serum heme-albumin

Human serum albumin (HSA) participates to heme scavenging, in turn HSA-heme binds gaseous diatomic ligands at the heme-Fe-atom. Here, the effect of abacavir and warfarin on denitrosylation kinetics of HSA-heme-Fe(II)-NO (i.e., k{sub off}) is reported. In the absence of drugs, the value of k{sub off} is (1.3 {+-} 0.2) x 10{sup -4} s{sup -1}. Abacavir and warfarin facilitate NO dissociation from HSA-heme-Fe(II)-NO, the k{sub off} value increases to (8.6 {+-} 0.9) x 10{sup -4} s{sup -1}. From the dependence of k{sub off} on the drug concentration, values of the dissociation equilibrium constant for the abacavir and warfarin binding to HSA-heme-Fe(II)-NO (i.e., K = (1.2 {+-} 0.2) x 10{sup -3} M and (6.2 {+-} 0.7) x 10{sup -5} M, respectively) were determined. The increase of k{sub off} values reflects the stabilization of the basic form of HSA-heme-Fe by ligands (e.g., abacavir and warfarin) that bind to Sudlow's site I. This event parallels the stabilization of the six-coordinate derivative of the HSA-heme-Fe(II)-NO atom. Present data highlight the allosteric modulation of HSA-heme-Fe(II) reactivity by heterotropic effectors.

60 APPLIED LIFE SCIENCES↗

An accurate conservative level set/ghost fluid method for simulating turbulent atomization

This paper presents a novel methodology for simulating incompressible two-phase flows by combining an improved version of the conservative level set technique introduced in [E. Olsson, G. Kreiss, A conservative level set method for two phase flow, J. Comput. Phys. 210 (2005) 225-246] with a ghost fluid approach. By employing a hyperbolic tangent level set function that is transported and re-initialized using fully conservative numerical schemes, mass conservation issues that are known to affect level set methods are greatly reduced. In order to improve the accuracy of the conservative level set method, high order numerical schemes are used. The overall robustness of the numerical approach is increased by computing the interface normals from a signed distance function reconstructed from the hyperbolic tangent level set by a fast marching method. The convergence of the curvature calculation is ensured by using a least squares reconstruction. The ghost fluid technique provides a way of handling the interfacial forces and large density jumps associated with two-phase flows with good accuracy, while avoiding artificial spreading of the interface. Since the proposed approach relies on partial differential equations, its implementation is straightforward in all coordinate systems, and it benefits from high parallel efficiency. The robustness and efficiency of the approach is further improved by using implicit schemes for the interface transport and re-initialization equations, as well as for the momentum solver. The performance of the method is assessed through both classical level set transport tests and simple two-phase flow examples including topology changes. It is then applied to simulate turbulent atomization of a liquid Diesel jet at Re=3000. The conservation errors associated with the accurate conservative level set technique are shown to remain small even for this complex case.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Coupling of in-situ X-ray Microtomography Observations with Discrete Element Simulations-Application to Powder Sintering

The sintering of both a powder with a wide particle size distribution (0-63 {mu}m) and of a powder with artificially created pores is investigated by coupling in situ X-ray microtomography observations with Discrete Element simulations. The micro structure evolution of the copper particles is observed by microtomography all along a typical sintering cycle at 1050 deg. C at the European Synchrotron Research Facilities (ESRF, Grenoble, France). A quantitative analysis of the 3D images provides original data on interparticle indentation, coordination and particle displacements throughout sintering. In parallel, the sintering of similar powder systems has been simulated with a discrete element code which incorporates appropriate sintering contact laws from the literature. The initial numerical packing is generated directly from the 3D microtomography images or alternatively from a random set of particles with the same size distribution. The comparison between the information drawn from the simulations and the one obtained by tomography leads to the conclusion that the first method is not satisfactory because real particles are not perfectly spherical as the numerical ones. On the opposite the packings built with the second method show sintering behaviors close to the behaviors of real materials, although particle rearrangement is underestimated by DEM simulations.

36 MATERIALS SCIENCE↗

Multiomics resolution of molecular events during a day in the life of Chlamydomonas

The unicellular green alga Chlamydomonas reinhardtii displays metabolic flexibility in response to a changing environment. We analyzed expression patterns of its three genomes in cells grown under light–dark cycles. Nearly 85% of transcribed genes show differential expression, with different sets of transcripts being up-regulated over the course of the day to coordinate cellular growth before undergoing cell division. Parallel measurements of select metabolites and pigments, physiological parameters, and a subset of proteins allow us to infer metabolic events and to evaluate the impact of the transcriptome on the proteome. Among the findings are the observations that Chlamydomonas exhibits lower respiratory activity at night compared with the day; multiple fermentation pathways, some oxygen-sensitive, are expressed at night in aerated cultures; we propose that the ferredoxin, FDX9, is potentially the electron donor to hydrogenases. The light stress-responsive genes PSBS , LHCSR1 , and LHCSR3 show an acute response to lights-on at dawn under abrupt dark-to-light transitions, while LHCSR3 genes also exhibit a later, second burst in expression in the middle of the day dependent on light intensity. Each response to light (acute and sustained) can be selectively activated under specific conditions. Our expression dataset, complemented with coexpression networks and metabolite profiling, should constitute an excellent resource for the algal and plant communities.

59 BASIC BIOLOGICAL SCIENCES↗

Particle-in-cell simulation of multipactor discharge on a dielectric in a parallel-plate waveguide

An original 2D3V (two-dimensional in coordinate space and three-dimensional in velocity space) particle-in-cell code has been developed for simulation of multipactor discharge on a dielectric in a parallelplate metal waveguide with allowance for secondary electron emission (SEE) from the dielectric surface and waveguide walls, finite temperature of secondary electrons, electron space charge, and elastic and inelastic scattering of electrons from the dielectric and metal surfaces. The code allows one to simulate all stages of the multipactor discharge, from the onset of the electron avalanche to saturation. It is shown that the threshold for the excitation of a single-surface multipactor on a dielectric placed in a low-profile waveguide with absorbing walls increases as compared to that in the case of an unbounded dielectric surface due to escape of electrons onto the waveguide walls. It is found that, depending on the microwave field amplitude and the SEE characteristics of the waveguide walls, the multipactor may operate in two modes. In the first mode, which takes place at relatively low microwave amplitudes, a single-surface multipactor develops only on the dielectric, the surface of which acquires a positively potential with respect to the waveguide walls. In the second mode, which occurs at sufficiently high microwave intensities, a single-surface multipactor on the dielectric and a two-surface multipactor between the waveguide walls operate simultaneously. In this case, both the dielectric surface and the interwall space acquire a negative potential. It is shown that electron scattering from the dielectric surface and waveguide walls results in the appearance of high-energy tails in the electron distribution function.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

NSTX Report on FES Joint Facilities Research Milestone 2010

Annual Target: Conduct experiments on major fusion facilities to improve understanding of the heat transport in the tokamak scrape-off layer (SOL) plasma, strengthening the basis for projecting divertor conditions in ITER. The divertor heat flux profiles and plasma characteristics in the tokamak scrape-off layer will be measured in multiple devices to investigate the underlying thermal transport processes. The unique characteristics of C-Mod, DIII-D, and NSTX will enable collection of data over a broad range of SOL and divertor parameters (e.g., collisionality ν*, beta β, parallel heat flux q||, and divertor geometry). Coordinated experiments using common analysis methods will generate a data set that will be compared with theory and simulation.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Creating a Parallel Version of VisIt for Microsoft Windows

VisIt is a popular, free interactive parallel visualization and analysis tool for scientific data. Users can quickly generate visualizations from their data, animate them through time, manipulate them, and save the resulting images or movies for presentations. VisIt was designed from the ground up to work on many scales of computers from modest desktops up to massively parallel clusters. VisIt is comprised of a set of cooperating programs. All programs can be run locally or in client/server mode in which some run locally and some run remotely on compute clusters. The VisIt program most able to harness today's computing power is the VisIt compute engine. The compute engine is responsible for reading simulation data from disk, processing it, and sending results or images back to the VisIt viewer program. In a parallel environment, the compute engine runs several processes, coordinating using the Message Passing Interface (MPI) library. Each MPI process reads some subset of the scientific data and filters the data in various ways to create useful visualizations. By using MPI, VisIt has been able to scale well into the thousands of processors on large computers such as dawn and graph at LLNL. The advent of multicore CPU's has made parallelism the 'new' way to achieve increasing performance. With today's computers having at least 2 cores and in many cases up to 8 and beyond, it is more important than ever to deploy parallel software that can use that computing power not only on clusters but also on the desktop. We have created a parallel version of VisIt for Windows that uses Microsoft's MPI implementation (MSMPI) to process data in parallel on the Windows desktop as well as on a Windows HPC cluster running Microsoft Windows Server 2008. Initial desktop parallel support for Windows was deployed in VisIt 2.4.0. Windows HPC cluster support has been completed and will appear in the VisIt 2.5.0 release. We plan to continue supporting parallel VisIt on Windows so our users will be able to take full advantage of their multicore resources.

97 MATHEMATICS AND COMPUTING↗

Simulation of neoclassical transport with the continuum gyrokinetic code COGENT

The development of the continuum gyrokinetic code COGENT for edge plasma simulations is reported. The present version of the code models a nonlinear axisymmetric 4D (R, v∥, μ) gyrokinetic equation coupled to the long-wavelength limit of the gyro-Poisson equation. Here, R is the particle gyrocenter coordinate in the poloidal plane, and v∥ and μ are the guiding center velocity parallel to the magnetic field and the magnetic moment, respectively. The COGENT code utilizes a fourth-order finite-volume (conservative) discretization combined with arbitrary mapped multiblock grid technology (nearly field-aligned on blocks) to handle the complexity of tokamak divertor geometry with high accuracy. Furthermore, topics presented are the implementation of increasingly detailed model collision operators, and the results of neoclassical transport simulations including the effects of a strong radial electric field characteristic of a tokamak pedestal under H-mode conditions.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Magnetic structure of Sr{sub 2}Fe{sub 2}O{sub 5} brownmillerite by single-crystal Mössbauer spectroscopy

In order to determine orientation of the Fe{sup 3+} magnetic moments and electric field gradient (efg) axes in the brownmillerite-type strontium ferrite structure for both iron sublattices where the efg tensor is not axially symmetric, the Mössbauer spectra of powdered and oriented single-crystal Sr{sub 2}Fe{sub 2}O{sub 5} were analyzed by solving the complete Hamiltonian for hyperfine interactions in the excited and ground states of the {sup 57}Fe nuclei. The magnetic moments of both octahedrally and tetrahedrally coordinated iron cations lie on the ac-plane of the orthorhombic unit cell and are parallel to the shortest c-axis, whilst the main efg axes are parallel to the longest crystallographic axis, b. This orientation is similar to that in Ca{sub 2}Fe{sub 2}O{sub 5}, in spite of the structural differences of strontium and calcium ferrite brownmillerites at low temperatures. - Graphical abstract: Mössbauer spectra of powdered and oriented single-crystal Sr{sub 2}Fe{sub 2}O{sub 5}, analyzed by solving the complete Hamiltonian for hyperfine interactions in the excited and ground states of the {sup 57}Fe nuclei, show that the magnetic moments of both octahedrally and tetrahedrally coordinated iron cations are parallel to the shortest c-axis of the orthorhombic brownmillerite unit cell . Display Omitted - Highlights: • Single-crystal Mössbauer spectroscopy is used to study magnetic structure of Sr{sub 2}Fe{sub 2}O{sub 5}. • Complete Hamiltonian for hyperfine interactions in excited and ground states was solved. • Fe{sup 3+} magnetic moments are parallel to the shortest crystallographic axis c. • The orientation of nuclear electric field gradient is similar to that in Ca{sub 2}Fe{sub 2}O{sub 5}.

36 MATERIALS SCIENCE↗

Formation of Gd coordination polymer with 1D chains mediated by Bronsted acidic ionic liquids

One dimensional coordination polymer Gd[(SO{sub 4})(NO{sub 3})(C{sub 2}H{sub 6}SO){sub 2}] (1) is prepared through the mediation of Bronsted acid ionic liquid, which crystallized in the monoclinic space of C2/c. In this polymer, adjacent Gd atoms are linked by two SO{sub 4}{sup 2-} ions to generate a 1-D chain, and all oxygen atoms in SO{sub 4}{sup 2-} groups are connected to three nearest Gd atoms in µ{sup 3}:η{sup 1}:η{sup 1}:η{sup 2} fashion. Gd, S and N from SO{sub 4}{sup 2-} and NO{sub 3}{sup -} are precisely coplanar. The planar is coordinated by a pair of DMSO molecules, which is parallel and linked by hydrogen bonding to form a three-dimensional supramolecular network. Magnetic susceptibility measurement of 1 reveals weak antiferromagnetic interactions between the Gd (III) ions. It exhibits relatively large magneto-caloric effect with –ΔS{sub m}=28.8 J Kg{sup −1} K{sup −1} for ΔH=7 T. - Graphical abstract: Coordination polymer Gd[(SO{sub 4})(NO{sub 3})(C{sub 2}H{sub 6}SO){sub 2}] was obtained mediated by Bronsted acid Ionic Liquid, which presents a 1-D chains collected by SO{sub 4}{sup 2-} groups. Magnetic susceptibility of the polymer reveals weak antiferromagnetic interactions between the Gd(III) ions with the relatively large magneto-caloric effect of –ΔS{sub m}=28.8 J Kg{sup −1} K{sup −1} for ΔH= 7T.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗