Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “COMPUTING”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 775 records · Page 43
All sourcesNASA NTRSDOE OSTI
NASA NTRS1993

Efficiently modeling neural networks on massively parallel computers

Neural networks are a very useful tool for analyzing and modeling complex real world systems. Applying neural network simulations to real world problems generally involves large amounts of data and massive amounts of computation. To efficiently handle the computational requirements of large problems, we have implemented at Los Alamos a highly efficient neural network compiler for serial computers, vector computers, vector parallel computers, and fine grain SIMD computers such as the CM-2 connection machine. This paper describes the mapping used by the compiler to implement feed-forward backpropagation neural networks for a SIMD (Single Instruction Multiple Data) architecture parallel computer. Thinking Machines Corporation has benchmarked our code at 1.3 billion interconnects per second (approximately 3 gigaflops) on a 64,000 processor CM-2 connection machine (Singer 1990). This mapping is applicable to other SIMD computers and can be implemented on MIMD computers such as the CM-5 connection machine. Our mapping has virtually no communications overhead with the exception of the communications required for a global summation across the processors (which has a sub-linear runtime growth on the order of O(log(number of processors)). We can efficiently model very large neural networks which have many neurons and interconnects and our mapping can extend to arbitrarily large networks (within memory limitations) by merging the memory space of separate processors with fast adjacent processor interprocessor communications. This paper will consider the simulation of only feed forward neural network although this method is extendable to recurrent networks.

Farber, Robert M.↗
NASA NTRS1993

A programming environment for distributed complex computing. An overview of the Framework for Interdisciplinary Design Optimization (FIDO) project. NASA Langley TOPS exhibit H120b

The Framework for Interdisciplinary Design Optimization (FIDO) is a general programming environment for automating the distribution of complex computing tasks over a networked system of heterogeneous computers. For example, instead of manually passing a complex design problem between its diverse specialty disciplines, the FIDO system provides for automatic interactions between the discipline tasks and facilitates their communications. The FIDO system networks all the computers involved into a distributed heterogeneous computing system, so they have access to centralized data and can work on their parts of the total computation simultaneously in parallel whenever possible. Thus, each computational task can be done by the most appropriate computer. Results can be viewed as they are produced and variables changed manually for steering the process. The software is modular in order to ease migration to new problems: different codes can be substituted for each of the current code modules with little or no effect on the others. The potential for commercial use of FIDO rests in the capability it provides for automatically coordinating diverse computations on a networked system of workstations and computers. For example, FIDO could provide the coordination required for the design of vehicles or electronics or for modeling complex systems.

Townsend, James C.↗
NASA NTRS1995

Parallel processing for scientific computations

The scope of this project dealt with the investigation of the requirements to support distributed computing of scientific computations over a cluster of cooperative workstations. Various experiments on computations for the solution of simultaneous linear equations were performed in the early phase of the project to gain experience in the general nature and requirements of scientific applications. A specification of a distributed integrated computing environment, DICE, based on a distributed shared memory communication paradigm has been developed and evaluated. The distributed shared memory model facilitates porting existing parallel algorithms that have been designed for shared memory multiprocessor systems to the new environment. The potential of this new environment is to provide supercomputing capability through the utilization of the aggregate power of workstations cooperating in a cluster interconnected via a local area network. Workstations, generally, do not have the computing power to tackle complex scientific applications, making them primarily useful for visualization, data reduction, and filtering as far as complex scientific applications are concerned. There is a tremendous amount of computing power that is left unused in a network of workstations. Very often a workstation is simply sitting idle on a desk. A set of tools can be developed to take advantage of this potential computing power to create a platform suitable for large scientific computations. The integration of several workstations into a logical cluster of distributed, cooperative, computing stations presents an alternative to shared memory multiprocessor systems. In this project we designed and evaluated such a system.

Alkhatib, Hasan S.↗
NASA NTRS2000

An Overview of the NCC Spray/Monte-Carlo-PDF Computations

This paper advances the state-of-the-art in spray computations with some of our recent contributions involving scalar Monte Carlo PDF (Probability Density Function), unstructured grids and parallel computing. It provides a complete overview of the scalar Monte Carlo PDF and Lagrangian spray computer codes developed for application with unstructured grids and parallel computing. Detailed comparisons for the case of a reacting non-swirling spray clearly highlight the important role that chemistry/turbulence interactions play in the modeling of reacting sprays. The results from the PDF and non-PDF methods were found to be markedly different and the PDF solution is closer to the reported experimental data. The PDF computations predict that some of the combustion occurs in a predominantly premixed-flame environment and the rest in a predominantly diffusion-flame environment. However, the non-PDF solution predicts wrongly for the combustion to occur in a vaporization-controlled regime. Near the premixed flame, the Monte Carlo particle temperature distribution shows two distinct peaks: one centered around the flame temperature and the other around the surrounding-gas temperature. Near the diffusion flame, the Monte Carlo particle temperature distribution shows a single peak. In both cases, the computed PDF's shape and strength are found to vary substantially depending upon the proximity to the flame surface. The results bring to the fore some of the deficiencies associated with the use of assumed-shape PDF methods in spray computations. Finally, we end the paper by demonstrating the computational viability of the present solution procedure for its use in 3D combustor calculations by summarizing the results of a 3D test case with periodic boundary conditions. For the 3D case, the parallel performance of all the three solvers (CFD, PDF, and spray) has been found to be good when the computations were performed on a 24-processor SGI Origin work-station.

Raju, M. S.↗
NASA NTRS2001

Generic Divide and Conquer Internet-Based Computing

The rapid growth of internet-based applications and the proliferation of networking technologies have been transforming traditional commercial application areas as well as computer and computational sciences and engineering. This growth stimulates the exploration of new, internet-oriented software technologies that can open new research and application opportunities not only for the commercial world, but also for the scientific and high -performance computing applications community. The general goal of this research project is to contribute to better understanding of the transition to internet-based high -performance computing and to develop solutions for some of the difficulties of this transition. More specifically, our goal is to design an architecture for generic divide and conquer internet-based computing, to develop a portable implementation of this architecture, to create an example library of high-performance divide-and-conquer computing agents that run on top of this architecture, and to evaluate the performance of these agents. We have been designing an architecture that incorporates a master task-pool server and utilizes satellite computational servers that operate on the Internet in a dynamically changing large configuration of lower-end nodes provided by volunteer contributors. Our designed architecture is intended to be complementary to and accessible from computational grids such as Globus, Legion, and Condor. Grids provide remote access to existing high-end computing resources; in contrast, our goal is to utilize idle processor time of lower-end internet nodes. Our project is focused on a generic divide-and-conquer paradigm and its applications that operate on a loose and ever changing pool of lower-end internet nodes.

Radenski, Atanas↗
NASA NTRS2001

Inviscid Flow Computations of the Shuttle Orbiter for Mach 10 and 15 and Angle of Attack 40 to 60 Degrees

This report documents the results of a computational study done to compute the inviscid longitudinal aerodynamic characteristics of the Space Shuttle Orbiter for Mach numbers 10 and 15 at angles of attack of 40, 50, 55, and 60 degrees. These computations were done to provide limited aerodynamic data in support of the Orbiter contingency abort task. The Orbiter had all the control surfaces in the undeflected position. The unstructured grid software FELISA was used for these computations with the equilibrium air option. Normal and axial force coefficients and pitching moment coefficients were computed. The hinge moment coefficients of the body flap and the inboard and outboard elevons were also computed. These results were compared with Orbiter Air Data Book (OADB) data and those computed using GASP. The comparison with the GASP results showed very good agreement in Cm and Ca at all the points. The computed axial force coefficients were smaller than those computed by GASP. There were noticeable differences between the present results and those in the OADB at angles of attack greater than 50 degrees.

Prabhu, Ramadas K.↗
NASA NTRS2006

Computational Aeroelastic Modeling of Airframes and TurboMachinery: Progress and Challenges

Computational analyses such as computational fluid dynamics and computational structural dynamics have made major advances toward maturity as engineering tools. Computational aeroelasticity is the integration of these disciplines. As computational aeroelasticity matures it too finds an increasing role in the design and analysis of aerospace vehicles. This paper presents a survey of the current state of computational aeroelasticity with a discussion of recent research, success and continuing challenges in its progressive integration into multidisciplinary aerospace design. This paper approaches computational aeroelasticity from the perspective of the two main areas of application: airframe and turbomachinery design. An overview will be presented of the different prediction methods used for each field of application. Differing levels of nonlinear modeling will be discussed with insight into accuracy versus complexity and computational requirements. Subjects will include current advanced methods (linear and nonlinear), nonlinear flow models, use of order reduction techniques and future trends in incorporating structural nonlinearity. Examples in which computational aeroelasticity is currently being integrated into the design of airframes and turbomachinery will be presented.

Bartels, R. E.↗
NASA NTRS2011

Performance Analysis of Cloud Computing Architectures Using Discrete Event Simulation

Cloud computing offers the economic benefit of on-demand resource allocation to meet changing enterprise computing needs. However, the flexibility of cloud computing is disadvantaged when compared to traditional hosting in providing predictable application and service performance. Cloud computing relies on resource scheduling in a virtualized network-centric server environment, which makes static performance analysis infeasible. We developed a discrete event simulation model to evaluate the overall effectiveness of organizations in executing their workflow in traditional and cloud computing architectures. The two part model framework characterizes both the demand using a probability distribution for each type of service request as well as enterprise computing resource constraints. Our simulations provide quantitative analysis to design and provision computing architectures that maximize overall mission effectiveness. We share our analysis of key resource constraints in cloud computing architectures and findings on the appropriateness of cloud computing in various applications.

Stocker, John C.↗
NASA NTRS2014

Computational Aerodynamic Simulations of an 840 ft/sec Tip Speed Advanced Ducted Propulsor Fan System Model for Acoustic Methods Assessment and Development

Computational Aerodynamic simulations of an 840 ft/sec tip speed, Advanced Ducted Propulsor fan system were performed at five different operating points on the fan operating line, in order to provide detailed internal flow field information for use with fan acoustic prediction methods presently being developed, assessed and validated. The fan system is a sub-scale, lownoise research fan/nacelle model that has undergone extensive experimental testing in the 9- by 15- foot Low Speed Wind Tunnel at the NASA Glenn Research Center, resulting in quality, detailed aerodynamic and acoustic measurement data. Details of the fan geometry, the computational fluid dynamics methods, the computational grids, and various computational parameters relevant to the numerical simulations are discussed. Flow field results for three of the five operating conditions simulated are presented in order to provide a representative look at the computed solutions. Each of the five fan aerodynamic simulations involved the entire fan system, excluding a long core duct section downstream of the core inlet guide vane. As a result, only fan rotational speed and system bypass ratio, set by specifying static pressure downstream of the core inlet guide vane row, were adjusted in order to set the fan operating point, leading to operating points that lie on a fan operating line and making mass flow rate a fully dependent parameter. The resulting mass flow rates are in good agreement with measurement values. The computed blade row flow fields for all five fan operating points are, in general, aerodynamically healthy. Rotor blade and fan exit guide vane flow characteristics are good, including incidence and deviation angles, chordwise static pressure distributions, blade surface boundary layers, secondary flow structures, and blade wakes. Examination of the computed flow fields reveals no excessive boundary layer separations or related secondary-flow problems. A few spanwise comparisons between computational and measurement data in the bypass duct show that they are in good agreement, thus providing a partial validation of the computational results.

Tweedt, Daniel L.↗
NASA NTRS2015

Re-Computation of Numerical Results Contained in NACA Report No. 496

An extensive examination of NACA Report No. 496 (NACA 496), "General Theory of Aerodynamic Instability and the Mechanism of Flutter," by Theodore Theodorsen, is described. The examination included checking equations and solution methods and re‐computing interim quantities and all numerical examples in NACA 496. The checks revealed that NACA 496 contains computational shortcuts (time‐ and effort‐saving devices for engineers of the time) and clever artifices (employed in its solution methods), but, unfortunately, also contains numerous tripping points (aspects of NACA 496 that have the potential to cause confusion) and some errors. The re‐computations were performed employing the methods and procedures described in NACA 496, but using modern computational tools. With some exceptions, the magnitudes and trends of the original results were in fair‐to‐very‐good agreement with the re‐computed results. The exceptions included what are speculated to be computational errors in the original in some instances and transcription errors in the original in others. Independent flutter calculations were performed and, in all cases, including those where the original and re‐computed results differed significantly, were in excellent agreement with the re‐computed results. Appendix A contains NACA 496; Appendix B contains a Matlab(Reistered) program that performs the re‐computation of results; Appendix C presents three alternate solution methods, with examples, for the two‐degree‐of-freedom solution method of NACA 496; Appendix D contains the three‐degree‐of‐freedom solution method (outlined in NACA 496 but never implemented), with examples.

Perry, Boyd, III↗
NASA NTRS2022

Computational Analysis of the X-57 Maxwell Airplane at Unpowered Conditions (Preliminary Fuselage)

The X-57 Maxwell is an all-electric airplane that implements a distributed electric propulsion system to demonstrate that high-efficiency electric propulsion can be integrated with aerodynamics to increase the performance of an airplane. To this end, distributed electric fans were installed on the wing to provide increased flow over the wing at the low takeoff and landing speeds of the X-57. The low-speed lift augmentation allows for a reduction in wing area for cruise optimization. The X-57 wing area was reduced to 42 percent of the wing area of the baseline aircraft, a Tecnam P2006T. With this reduced wing area and the electric propulsion system, it is estimated that the X-57 will cruise on less than one-third the total energy compared to the baseline aircraft. To meet the cruise performance goal at a Mach number of 0.233 at an altitude of 8000 feet, the X-57 has a cruise lift coefficient of 0.7516 and needs to have a cruise drag coefficient of 0.05423 or less. The USM3D computational solver was used to investigate the X-57 performance, without the distributed electric propulsion high-lift system operating. The unpowered X-57 performance is of interest to quantify if the X-57 can meet the cruise drag performance goal, and to document the lift performance of the very small wing at takeoff and landing conditions. The primary configurations investigated in this paper include the cruise configuration with no flap deflection, a takeoff configuration with a 10◦ flap deflection, and a landing configuration with a 30◦ flap deflection. The conditions for the cruise configuration were a flight unit Reynolds number of 1.32E+06 per foot, an altitude of 8000 feet, a Mach number of 0.233, and angles of attack from −2° to 24° . At the cruise lift coefficient of 0.7516, the computed drag coefficient is 0.05275. This computed drag is less than the drag coefficient of 0.05423 that is required to meet the X-57 airplane performance goal. However, the computational airplane is a completely smooth geometry and does not account for protuberance drag, nor the drag from steps and gaps in the actual X-57 airplane. Therefore, based upon the CFD drag calculation there is a 10-percent margin to account for some of the differences between the as-built metal fuselage and empennage construction, and the smooth computational geometry. The computed cruise drag also does not account for an induced drag reduction due to the wing-tip propellers and a drag reduction due to laminar flow achieved on the wing. The computed lift to drag ratio is 14.14 at the cruise lift coefficient of 0.7516, and the maximum computed lift to drag ratio is 15.8. The maximum lift coefficient for the cruise configuration was 2.13 at an angle of attack of 15°. The conditions for the takeoff configuration with a 10° flap deflection were a flight unit Reynolds number of 0.986E+06 per foot, an altitude of 2500 feet, a Mach number of 0.149, and angles of attack from −2° to 22°. The maximum lift coefficient for the takeoff configuration was 2.21 at an angle of attack of 16°. The conditions for the landing configuration with a 30° flap deflection were a flight unit Reynolds number of 0.922E+06 per foot, an altitude of 2500 feet, a Mach number of 0.139, and angles of attack from −2° to 24°. The maximum lift coefficient for the landing configuration was 2.58 and occurred at two angles of attack, 10° and 14°. Based on the unpowered maximum lift coefficient of 2.58 for the 30° flap deflection, along with computations of the distributed electric propulsion lift augmentation (not shown in this paper), the X-57 Maxwell is estimated to meet its powered landing goal of a maximum lift coefficient of 4.0.

X-57 Maxwell Airplane↗
DOE OSTI2024

High-data throughput reconfigurable computing platform

A reconfigurable computing platform includes a reconfigurable computing device, electro-optical transceiver, and first voltage converter disposed on a multilayer board. The electro-optical transceiver converts an optical signal at least one of to and from an electrical signal, and the electrical signal is operatively coupled to the reconfigurable computing device. The electro-optical transceiver is disposed in proximity to the reconfigurable computing device, and the first voltage converter is operatively coupled to a common voltage distributed around a periphery of the multilayer board. The first voltage converter converts the common voltage to a first operating voltage, and the first voltage converter is disposed in proximity to the reconfigurable computing device. The first operating voltage is provided to the reconfigurable computing device as a first power source. A reconfigurable computing system includes a plurality of reconfigurable computing platforms operatively coupled together using an optical signal. A corresponding method of providing a reconfigurable computing platform is also disclosed.

Tang, Shaochun↗
DOE OSTI2017

On the Bayesian calibration of computer model mixtures through experimental data, and the design of predictive models

For many real systems, several computer models may exist with different physics and predictive abilities. To achieve more accurate simulations/predictions, it is desirable for these models to be properly combined and calibrated. Here, we propose the Bayesian calibration of computer model mixture method which relies on the idea of representing the real system output as a mixture of the available computer model outputs with unknown input dependent weight functions. The method builds a fully Bayesian predictive model as an emulator for the real system output by combining, weighting, and calibrating the available models in the Bayesian framework. Moreover, it fits a mixture of calibrated computer models that can be used by the domain scientist as a mean to combine the available computer models, in a flexible and principled manner, and perform reliable simulations. It can address realistic cases where one model may be more accurate than the others at different input values because the mixture weights, indicating the contribution of each model, are functions of the input. Inference on the calibration parameters can consider multiple computer models associated with different physics. The method does not require knowledge of the fidelity order of the models. We provide a technique able to mitigate the computational overhead due to the consideration of multiple computer models that is suitable to the mixture model framework. We implement the proposed method in a real-world application involving the Weather Research and Forecasting large-scale climate model.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

A Roadmap for Edge Computing Enabled Automated Multidimensional Transmission Electron Microscopy

The advent of modern, high-speed electron detectors has made the collection of multidimensional hyperspectral transmission electron microscopy datasets, such as 4D-STEM, a routine. However, many microscopists find such experiments daunting since analysis, collection, long-term storage, and networking of such datasets remain challenging. Some common issues are their large and unwieldy size that often are several gigabytes, non-standardized data analysis routines, and a lack of clarity about the computing and network resources needed to utilize the electron microscope. The existing computing and networking bottlenecks introduce significant penalties in each step of these experiments, and thus, real-time analysis-driven automated experimentation for multidimensional TEM is challenging. One solution is to integrate microscopy with edge computing, where moderately powerful computational hardware performs the preliminary analysis before handing off the heavier computation to high-performance computing (HPC) systems. In this work, we trace the roots of computation in modern electron microscopy, demonstrate deep learning experiments running on an edge system, and discuss the networking requirements for tying together microscopes, edge computers, and HPC systems.

47 OTHER INSTRUMENTATION↗
DOE OSTI2019

Enhancing materials property prediction by leveraging computational and experimental data using deep transfer learning

Abstract The current predictive modeling techniques applied to Density Functional Theory (DFT) computations have helped accelerate the process of materials discovery by providing significantly faster methods to scan materials candidates, thereby reducing the search space for future DFT computations and experiments. However, in addition to prediction error against DFT-computed properties, such predictive models also inherit the DFT-computation discrepancies against experimentally measured properties. To address this challenge, we demonstrate that using deep transfer learning, existing large DFT-computational data sets (such as the Open Quantum Materials Database (OQMD)) can be leveraged together with other smaller DFT-computed data sets as well as available experimental observations to build robust prediction models. We build a highly accurate model for predicting formation energy of materials from their compositions; using an experimental data set of $$1,643$$ 1 , 643 observations, the proposed approach yields a mean absolute error (MAE) of $$0.07$$ 0.07 eV/atom, which is significantly better than existing machine learning (ML) prediction modeling based on DFT computations and is comparable to the MAE of DFT-computation itself.

36 MATERIALS SCIENCE↗
DOE OSTI2018

An approach to quantum-computational hydrologic inverse analysis

Making predictions about flow and transport in an aquifer requires knowledge of the heterogeneous properties of the aquifer such as permeability. Computational methods for inverse analysis are commonly used to infer these properties from quantities that are more readily observable such as hydraulic head. We present a method for computational inverse analysis that utilizes a type of quantum computer called a quantum annealer. While quantum computing is in an early stage compared to classical computing, we demonstrate that it is sufficiently developed that it can be used to solve certain subsurface flow problems. We utilize a D-Wave 2X quantum annealer to solve 1D and 2D hydrologic inverse problems that, while small by modern standards, are similar in size and sometimes larger than hydrologic inverse problems that were solved with early classical computers. Our results and the rapid progress being made with quantum computing hardware indicate that the era of quantum-computational hydrology may not be too far in the future.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Software Systems for High-performance Quantum Computing

Quantum computing promises new opportunities for solving hard computational problems, but harnessing this novelty requires breakthrough concepts in the design, operation, and application of computing systems. We define some of the challenges facing the development of quantum computing systems as well as software-based approaches that can be used to overcome these challenges. Following a brief overview of the state of the art, we present models for the quantum programming and execution models, the development of architectures for hybrid high-performance computing systems, and the realization of software stacks for quantum networking. This leads to a discussion of the role that conventional computing plays in the quantum paradigm and how some of the current challenges for exascale computing overlap with those facing quantum computing.

high-performance computing↗
DOE OSTI2016

The Improved Simple Corner Balance Method and Efforts to Enhance its Computational Performance

The diamond difference (DD) spatial discretization has long been a work horse of the neutron transport community. For nearly as long as DD has been used in neutron transport, it has been observed that DD produces undamped, non-physical angular fluxes in spatially under resolved problems and can fail in the thick diffusion limit. A long list of alternative spatial discretizations, all with better physics properties: accuracy, robustness, and/or preservation of the thick diffusion limit, have been proposed as alternatives to DD. Unfortunately, these alternative spatial discretizations also require significantly more floating point operations (FLOPs) per unknown compared to DD. Historically, neutron transport problems of interest have been able to be resolved sufficiently that DD is accurate enough and FLOPs have been the computational performance bottleneck controlling time to solution. Thus, in practice DD has been and remains a preferred spatial discretization for neutron transport simulations. Trends in computing architecture are leading to a new paradigm where data movement, not FLOPs will control time to solution. This presents an opportunity to improve neutron transport physics by reconsidering more FLOP intensive spatial discretization than DD without incurring a time to solution penalty. In this work, we consider the improved simple corner balance (ISCB) spatial discretization, a unique alternative to DD that has the same memory footprint as DD, and detail efforts to improve its computational performance. The ISCB spatial discretization is a more FLOP intensive spatial discretization than DD with the same memory footprint as DD. Though designed to maximize FLOP intensity, third party libraries such as LAPACK are not tuned for the unique, frequent, small linear algebra needs of neutron transport problems. While hand coding a sparse Gauss elimination is likely to be faster than a dense matrix solve (CM vs. HGE), were LAPACK suitable for small linear algebra problems, we would not have seen a performance increase with decreased numbers of LAPACK calls for the Schur complement solve (SC4 vs. SC1). Significant effort has reduced the relative compute time of ISCB to DD, but on current computer architectures DD is still faster than ISCB. Future work comparing the overall efficiency of ISCB to DD may reveal that while ISCB is slower than DD on current computers, ISCB is more accurate than DD by a factor greater than the compute time slowdown. As new computer architectures become available, we will test our hypothesis that neutron transport simulations will be bandwidth limited, ideally rendering the additional FLOPs required by ISCB to be 'free' [in terms of time to solution]. (authors)

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗