Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “parallel, scaling”

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 1,423 records · Page 79

Distributed control of reactive power flow in a radial distribution circuit with high photovoltaic penetration

We show how distributed control of reactive power can serve to regulate voltage and minimize resistive losses in a distribution circuit that includes a significant level of photovoltaic (PV) generation. To demonstrate the technique, we consider a radial distribution circuit with a single branch consisting of sequentially-arranged residential-scale loads that consume both real and reactive power. In parallel, some loads also have PV generation capability. We postulate that the inverters associated with each PV system are also capable of limited reactive power generation or consumption, and we seek to find the optimal dispatch of each inverter's reactive power to both maintain the voltage within an acceptable range and minimize the resistive losses over the entire circuit. We assume the complex impedance of the distribution circuit links and the instantaneous load and PV generation at each load are known. We compare the results of the optimal dispatch with a suboptimal local scheme that does not require any communication. On our model distribution circuit, we illustrate the feasibility of high levels of PV penetration and a significant (20% or higher) reduction in losses.

24 POWER TRANSMISSION AND DISTRIBUTION↗

On the Short-scale Spatial Variability of Electron Inflows in Electron-only Magnetic Reconnection in the Turbulent Magnetosheath Observed by MMS

We investigate the detailed properties of electron inflow in an electron-only reconnection event observed by the four Magnetospheric Multiscale (MMS) spacecraft in the Earth's turbulent magnetosheath downstream of the quasi-parallel bow shock. The lack of ion coupling was attributed to the small-scale sizes of the current sheets, and the observed bidirectional super-Alfvénic electron jets indicate that the MMS spacecraft crossed the reconnecting current sheet on both sides of an active X-line. Remarkably, the MMS spacecraft observed the presence of large asymmetries in the two electron inflows, with the inflows (normal to the current sheet) on the two sides of the reconnecting current layer differing by as much as a factor of four. Furthermore, even though the four MMS spacecraft were separated by less than seven electron skin depths, the degree of inflow asymmetry was significantly different at the different spacecraft. The asymmetry in the inflow speeds was larger with increasing distances downstream from the reconnection site, and the asymmetry was opposite on the two sides of the X-line. We compare the MMS observations with a 2D kinetic particle-in-cell (PIC) simulation and find that the asymmetry in the inflow speeds stems from in-plane currents generated via the combination of reconnection-mediated inflows and parallel flows along the magnetic separatrices in the presence of a large guide field.

79 ASTRONOMY AND ASTROPHYSICS↗

Effects of Anisotropic Thermal Conductivity in Magnetohydrodynamics Simulations of a Reversed-Field Pinch

A compressible magnetohydrodynamics simulation of the reversed-field pinch is performed including anisotropic thermal conductivity. When the thermal conductivity is much larger in the direction parallel to the magnetic field than in the perpendicular direction, magnetic field lines become isothermal. As a consequence, as long as magnetic surfaces exist, a temperature distribution is observed displaying a hotter confined region, while an almost uniform temperature is produced when the magnetic field lines become chaotic. To include this effect in the numerical simulation, we use a multiple-time-scale analysis, which allows us to reproduce the effect of a large parallel thermal conductivity. The resulting temperature distribution is related to the existence of closed magnetic surfaces, as observed in experiments. The magnetic field is also affected by the presence of an anisotropic thermal conductivity.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Development and Calibration of a System-Integrated Rotorcraft Finite Element Model for Impact Scenarios

Two full-scale crash tests of an MD-500 helicopter were conducted in 2009 and 2010 at NASA Langley's Landing and Impact Research Facility in support of NASA s Subsonic Rotary Wing Crashworthiness Project. The first crash test was conducted to evaluate the performance of an externally mounted composite deployable energy absorber (DEA) under combined impact conditions. In the second crash test, the energy absorber was removed to establish baseline loads that are regarded as severe but survivable. The presence of this energy absorbing device reduced the peak impact acceleration levels by a factor of three. Accelerations and kinematic data collected from the crash tests were compared to a system-integrated finite element model of the test article developed in parallel with the test program. In preparation for the full-scale crash test, a series of sub-scale and MD-500 mass simulator tests were conducted to evaluate the impact performances of various components and subsystems, including new crush tubes and the DEA blocks. Parameters defined for the system-integrated finite element model were determined from these tests. Results from 19 accelerometers placed throughout the airframe were compared to finite element model responses. The model developed for the purposes of predicting acceleration responses from the first crash test was inadequate when evaluating more severe conditions seen in the second crash test. A newly developed model calibration approach that includes uncertainty estimation, parameter sensitivity, impact shape orthogonality, and numerical optimization was used to calibrate model results for the full-scale crash test without the DEA. This combination of heuristic and quantitative methods identified modeling deficiencies, evaluated parameter importance, and proposed required model changes. The multidimensional calibration techniques presented here are particularly effective in identifying model adequacy. Acceleration results for the calibrated model were compared to test results and the original model results. There was a noticeable improvement in the pilot and copilot region, a slight improvement in the occupant model response, and an over-stiffening effect in the passenger region. One lesson learned was that this approach should be adopted early on, in combination with the building-block approaches that are customarily used, for model development and pretest predictions. Complete crash simulations with validated finite element models can be used to satisfy crash certification requirements, potentially reducing overall development costs.

Annett, Martin S.↗

Generation of large-scale, barrier-free diffuse plasmas in air at atmospheric pressure using array wire electrodes and nanosecond high-voltage pulses

This paper introduces a method to generate large-scale diffuse plasmas by using a repetition nanosecond pulse generator and a parallel array wire-electrode configuration. We investigated barrier-free diffuse plasmas produced in the open air in parallel and cross-parallel array line-line electrode configurations. We found that, when the distance between the wire-electrode pair is small, the discharges were almost extinguished. Also, glow-like diffuse plasmas with little discharge weakening were obtained in an appropriate range of line-line distances and with a cathode-grounding cross-electrode configuration. As an example, we produced a large-scale, stable diffuse plasma with volumes as large as 18 × 15 × 15 cm{sup 3}, and this discharge region can be further expanded. Additionally, using optical and electrical measurements, we showed that the electron temperature was higher than the gas temperature, which was almost the same as room temperature. Also, an array of electrode configuration with more wire electrodes had helped to prevent the transition from diffuse discharge to arc discharge. Comparing the current waveforms of configurations with 1 cell and 9 cells, we found that adding cells significantly increased the conduction current and the electrical energy delivered in the electrode gaps.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Future directions in computing and CFD

In recent years CFD on massively parallel machines has become a reality. This paper summarizes some recent trends both in high performance computing, and in CFD using parallel machines. The long term computational requirements for accomplishing some of the large scale problems in computational aerosciences, and current hardware and architecture trends are discussed. Performance results obtained from the implementation of some CFD applications on the Connection Machine CM-2 and the Intel iPSC/860 at NASA Ames Research Center are presented. It is argued that only massively parallel machines will be able to meet these grand challenge requirements.

Bailey, F. R.↗

Cu/Nb nano-composite wires processed by severe plastic deformation for applications in high pulsed magnets: effects of the multi-scale microstructure on the mechanical properties

Copper-based high strength and high electrical conductivity nano-composite wires reinforced by Nb nano-tubes are prepared by severe plastic deformation, applied with an Accumulative Drawing and Bundling process (ADB), for the windings of high pulsed magnets. The ADB process leads to a multi-scale Cu matrix containing up to N = 85{sub 4} (52.2.10{sup 6}) continuous parallel Nb tubes with diameter down to few tens nano-meters. After heavy strain, the Nb nano-tubes exhibit a homogeneous microstructure with grain size below 100 nm. The Cu matrix presents a multi-scale microstructure with multi-modal grain size distribution from the micrometer to the nano-meter range. The use of complementary characterization techniques at the microscopic and macroscopic level (in-situ tensile tests in the TEM, nano-indentation, in-situ tensile tests under high energy synchrotron beam) shed light on the interest of the multi-scale nature of the microstructure to achieve extreme mechanical properties, therefore allowing for design guidelines to further improve these properties. (authors)

36 MATERIALS SCIENCE↗

PLUM: Parallel Load Balancing for Unstructured Adaptive Meshes

Dynamic mesh adaption on unstructured grids is a powerful tool for computing large-scale problems that require grid modifications to efficiently resolve solution features. By locally refining and coarsening the mesh to capture physical phenomena of interest, such procedures make standard computational methods more cost effective. Unfortunately, an efficient parallel implementation of these adaptive methods is rather difficult to achieve, primarily due to the load imbalance created by the dynamically-changing nonuniform grid. This requires significant communication at runtime, leading to idle processors and adversely affecting the total execution time. Nonetheless, it is generally thought that unstructured adaptive- grid techniques will constitute a significant fraction of future high-performance supercomputing. Various dynamic load balancing methods have been reported to date; however, most of them either lack a global view of loads across processors or do not apply their techniques to realistic large-scale applications.

Oliker, Leonid↗

Fluctuations in electron-positron plasmas: Linear theory and implications for turbulence

Linear kinetic theory of electromagnetic fluctuations in a homogeneous, magnetized, collisionless electron-positron plasma predicts two lightly damped modes propagate at relatively long wavelengths: an Alfven-like mode with dispersion {omega}{sub r}=k{sub ||}v-tilde{sub A} and a magnetosonic-like mode with dispersion {omega}{sub r}{approx_equal}kv-tilde{sub A} if {beta}{sub e}<<1. Here v-tilde{sub A} is the Alfven speed in an electron-positron plasma and || refers to the direction relative to the background magnetic field B{sub o}. Both modes have phase speeds {omega}{sub r}/k which monotonically decrease with increasing wavenumber. The Alfven-like fluctuations are almost incompressible, but the magnetosonic-like fluctuations become strongly compressible at short wavelengths and propagation sufficiently oblique to B{sub o}. Using the linear dispersion properties of these modes, scaling relations are derived which predict that turbulence of both modes should be relatively anisotropic, with fluctuating magnetic energy preferentially cascading in directions perpendicular to B{sub o}. Turbulent spectra in the solar wind show two distinct power-law regimes separated by a distinct breakpoint in observed frequency; this characteristic should not be present in electron-positron turbulence because of the absence of whistler-like dispersion. Linear theory properties of the cyclotron and mirror instabilities driven by either electron or positron temperature anisotropies are generally analogous to those of the corresponding instabilities in electron-proton plasmas.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Status Report on the Water-Based NSTF RCCS Capability: Preparations and Design for the Transformation of the Natural Convection Shutdown Heat Removal Test Facility (NSTF) From Air to Water-based Cooling

The following report provides an overview of the current status of the water-based transformation, including details of the design, geometry, and fabrication of the test section and network piping, for the ½ scale water-based NSTF program at Argonne National Laboratory. In early FY2015, efforts towards a water-based design began in parallel with on-going air-based work, and began by identifying top level objectives of the transformation. Scaling studies and preparation tasks with computational modeling were carried out to guide the design decisions, minimize distortions, and ensure relevant data among scales. This entailed close collaboration with AREVA, whose water-based RCCS (included as part of their 625 MWt SC-HTGR) served as the primary design basis for incorporation into the NSTF. A literature review of previous works was made, with an emphasis placed on similar facilities, relevance to passive decay heat removal, and details pertaining to two-phase flow and measurement techniques. The study focused on input from key designers and a thermal analysis of existing cooling panels (e.g. at TAMU and UW-Madison) and two proposed AREVA cooling panels. A base case was used as a reference point for parametric studies of effects from varying tube spacing (pitch), tube diameter, fin thickness, and materials. Then, a structural analysis was performed to ensure safe material stresses during high temperature operation. Details of the pipe network geometry and guidelines for design of the water storage tank are finally presented, followed by engineering drawings in the appendix. The results of this study have identified a suitable configuration to support both the ANL/DOE project goals and the DOE vision to provide AREVA with data suitable for characterizing the RCCS of their full scale HTGR design. The dimensions were not selected for optimal RCCS performance, but instead to serve as a representative yet bounding configuration for future implementation into a full scale design. Scaling distortions are unavoidable; however, they can be accurately predicted based on earlier works of derived similarity solutions. Flexibility remained a primary design philosophy, and will allow the test facility to easily accommodate future alterations. The final design for the test section and network piping is as follows: riser tubes: 1.5” Schedule 160, 5.91” (150-mm) pitch, 304L stainless, heat transfer panels: 5/16” plates, 4.01” (102-mm) width, full penetration weld to risers, 304L stainless, split into 4 panels per riser separated by horizontal 1/8” gaps, test section: Eight (x8) riser tubes and nine (x9) heat transfer panels, fabricated into banks of two (x2) riser tubes and three (x3) fins, joined to form single section, and network geometry: 4.0” Schedule 40, 304L stainless. Details of the water storage tank will reflect competing effects such as thermal hydraulic phenomena during single and two-phase operation. These considerations include mixing during single-phase jet penetration, condensation of rising bubbles during two-phase discharge, bubble entrainment into the liquid outlet, and liquid carry-over into the steam outlet.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Weak scaling of the contact distance between two fluctuating interfaces with system size

A pair of flat parallel surfaces, each freely diffusing along the direction of their separation, will eventually come into contact. If the shapes of these surfaces also fluctuate, then contact will occur when their centers-of-mass remain separated by a nonzero distance ℓ. An example of such a situation is the motion of interfaces between two phases at conditions of thermodynamic coexistence, and in particular the annihilation of domain wall pairs under periodic boundary conditions. Here we present a general approach to calculate the probability distribution of the contact distance ℓ and determine how its most likely value ℓ* depends on the surfaces' lateral size L. Using the Edward-Wilkinson equation as a model for interfaces, we demonstrate that ℓ* scales weakly with system size, i.e., the dependence of ℓ* on L for both (1+1)- and (2+1)-dimensional interfaces is such that lim L→∞ (ℓ*/L) = 0. In particular, for (2+1)-dimensional interfaces ℓ* is an algebraic function of logL, a result that is confirmed by computer simulations of slab-shaped domains formed under periodic boundary conditions. Overall, this weak scaling implies that such domains remain topologically intact until ℓ becomes very small compared to the lateral size of the interface, contradicting expectations from equilibrium thermodynamics.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

symPACK: A GPU-Capable Fan-Out Sparse Cholesky Solver

Sparse symmetric positive definite systems of equations are ubiquitous in scientific workloads and applications. Parallel sparse Cholesky factorization is the method of choice for solving such linear systems. Therefore, the development of parallel sparse Cholesky codes that can efficiently run on today’s large-scale heterogeneous distributed-memory platforms is of vital importance. Modern supercomputers offer nodes that contain a mix of CPUs and GPUs. To fully utilize the computing power of these nodes, scientific codes must be adapted to offload expensive computations to GPUs. We present symPACK, a GPU-capable parallel sparse Cholesky solver that uses one-sided communication primitives and remote procedure calls provided by the UPC++ library. We also utilize the UPC++ "memory kinds" feature to enable efficient communication of GPU-resident data. We show that on a number of large problems, symPACK outperforms comparable state-of-the-art GPU-capable Cholesky factorization codes by up to 14x on the NERSC Perlmutter supercomputer.

Bellavita, Julian↗

Reinforcement Learning for Load-balanced Parallel Particle Tracing

We explore an online reinforcement learning (RL) paradigm to dynamically optimize parallel particle tracing performance in distributed-memory systems. Our method combines three novel components: (1) a work donation algorithm, (2) a high-order workload estimation model, and (3) a communication cost model. First, we design an RL-based work donation algorithm. Our algorithm monitors workloads of processes and creates RL agents to donate data blocks and particles from high-workload processes to low-workload processes to minimize program execution time. The agents learn the donation strategy on the fly based on reward and cost functions designed to consider processes' workload changes and data transfer costs of donation actions. Second, we propose a workload estimation model, helping RL agents estimate the workload distribution of processes in future computations. Third, we design a communication cost model that considers both block and particle data exchange costs, helping RL agents make effective decisions with minimized communication costs. We demonstrate that our algorithm adapts to different flow behaviors in large-scale fluid dynamics, ocean, and weather simulation data. Our algorithm improves parallel particle tracing performance in terms of parallel efficiency, load balance, and costs of I/O and communication for evaluations with up to 16,384 processors.

Distributed and parallel particle tracing↗

Distributed Parallel Processing and Dynamic Load Balancing Techniques for Multidisciplinary High Speed Aircraft Design

Multidisciplinary design optimization (MDO) for large-scale engineering problems poses many challenges (e.g., the design of an efficient concurrent paradigm for global optimization based on disciplinary analyses, expensive computations over vast data sets, etc.) This work focuses on the application of distributed schemes for massively parallel architectures to MDO problems, as a tool for reducing computation time and solving larger problems. The specific problem considered here is configuration optimization of a high speed civil transport (HSCT), and the efficient parallelization of the embedded paradigm for reasonable design space identification. Two distributed dynamic load balancing techniques (random polling and global round robin with message combining) and two necessary termination detection schemes (global task count and token passing) were implemented and evaluated in terms of effectiveness and scalability to large problem sizes and a thousand processors. The effect of certain parameters on execution time was also inspected. Empirical results demonstrated stable performance and effectiveness for all schemes, and the parametric study showed that the selected algorithmic parameters have a negligible effect on performance.

Krasteva, Denitza T.↗

Towards improved speed and accuracy of laser powder bed fusion simulations via multiscale spatial representations

Due to the growing popularity of laser powder bed fusion (LPBF) as a metal additive manufacturing technique, there is a strong need to be able to accurately predict build outcomes. Full fidelity simulations of this process are not feasible due to the vast range of length and time scales inherent to it. While part-scale codes for simulating residual stress and distortion have shown reasonable predictive capability, they often neglect many aspects of the process occurring over smaller length/time scales, and thus are unable to capture effects of process parameter adjustments or the behavior of fine features. One way of capturing aspects at more refined length scales is through the use of adaptive mesh refinement (AMR). AMR allows for the process to be simulated at scales approaching the physical spatial dimensions without drastically increasing the total degrees of freedom in the simulation. This manuscript describes the implementation of an AMR algorithm within a multiphysics, parallelized finite element code, and its application to the LPBF problem. In this work, part-scale examples are provided where the use of AMR has allowed for higher fidelity thermal and thermomechanical simulations, as compared to experimental measurements. Results from these higher resolution simulations show that while AMR is a necessary component for increased accuracy in a computationally efficient manner, other improvements are also necessary, including handling of the multiple time scales inherent to the problem and the need for improved AM-specific material models.

42 ENGINEERING↗

Parallel Computing Strategies for Irregular Algorithms

Parallel computing promises several orders of magnitude increase in our ability to solve realistic computationally-intensive problems, but relies on their efficient mapping and execution on large-scale multiprocessor architectures. Unfortunately, many important applications are irregular and dynamic in nature, making their effective parallel implementation a daunting task. Moreover, with the proliferation of parallel architectures and programming paradigms, the typical scientist is faced with a plethora of questions that must be answered in order to obtain an acceptable parallel implementation of the solution algorithm. In this paper, we consider three representative irregular applications: unstructured remeshing, sparse matrix computations, and N-body problems, and parallelize them using various popular programming paradigms on a wide spectrum of computer platforms ranging from state-of-the-art supercomputers to PC clusters. We present the underlying problems, the solution algorithms, and the parallel implementation strategies. Smart load-balancing, partitioning, and ordering techniques are used to enhance parallel performance. Overall results demonstrate the complexity of efficiently parallelizing irregular algorithms.

Biswas, Rupak↗

Electron magnetic reconnection without ion coupling in Earth’s turbulent magnetosheath

Magnetic reconnection in current sheets is a magnetic-to-particle energy conversion process that is fundamental to many space and laboratory plasma systems. In the standard model of reconnection, this process occurs in a minuscule electron-scale diffusion region. On larger scales, ions couple to the newly reconnected magnetic-field lines and are ejected away from the diffusion region in the form of bi-directional ion jets at the ion Alfvén speed. Much of the energy conversion occurs in spatially extended ion exhausts downstream of the diffusion region. In turbulent plasmas, which contain a large number of small-scale current sheets, reconnection has long been suggested to have a major role in the dissipation of turbulent energy at kinetic scales. However, evidence for reconnection plasma jetting in small-scale turbulent plasmas has so far been lacking. Here we report observations made in Earth’s turbulent magnetosheath region (downstream of the bow shock) of an electron-scale current sheet in which diverging bi-directional super-ion-Alfvénic electron jets, parallel electric fields and enhanced magnetic-to-particle energy conversion were detected. Contrary to the standard model of reconnection, the thin reconnecting current sheet was not embedded in a wider ion-scale current layer and no ion jets were detected. Observations of this and other similar, but unidirectional, electron jet events without signatures of ion reconnection reveal a form of reconnection that can drive turbulent energy transfer and dissipation in electron-scale current sheets without ion coupling.

79 ASTRONOMY AND ASTROPHYSICS↗

A time-parallel method for scalable heat transfer simulations of additive manufacturing

Here, a major challenge in simulating the thermal behavior in additive manufacturing processes is the disparate length and time scales between transport phenomena occurring in the melt pool and the component. A common simulation approach relies on spatial decomposition for parallel computing, but due to the nature of heat transfer in AM, where most of the computational expenditure is localized near the melt pool, the computational speedup from spatial parallelization saturates quickly. Therefore, additional parallelism by means of time-domain decomposition is needed to fully take advantage of high-performance computing (HPC) resources. This work introduces a time-parallel method to improve the computational scalability of additive manufacturing simulations on HPC systems, while maintaining high temporal resolution of heat transfer near the melt pool. The method, inspired by the nonlinear paraexp formalism, performs an iterative superposition of nonlinear solutions to the initial value problem, integrating the heat equation across overlapping time-parallel intervals. For a single layer of the NIST AMB2018–01 L7 benchmark problem, the method achieves a 38.51x speedup in wall-clock time with a maximum error in the global temperature solution of 0.99%. This reduces the total solution time from 196.72 min to 5.11 min on 128 nodes of the ORNL Frontier supercomputer. The tradeoff between accuracy and total wall-clock time is investigated and recommendations for time-parallel deployment for AM problems are made.

Additive manufacturing↗