SEARCH · Search NASA
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.
A detailed description of the data mapping and rotation algorithms for the Massively Parallel Processor is given.
Goodyear Aerospace delivered the Massively Parallel Processor (MPP) to NASA/Goddard in May 1983, over three years ago. Ever since then, Goodyear has tried to look in a forward direction. There is always some debate as to which way is forward when it comes to supercomputer architecture. Improvements to the MPP's massively parallel architecture are discussed in the areas of data I/O, memory capacity, connectivity, and indirect (or local) addressing. In I/O, transfer rates up to 640 megabytes per second can be achieved. There are devices that can supply the data and accept it at this rate. The memory capacity can be increased up to 128 megabytes in the ARU and over a gigabyte in the staging memory. For connectivity, there are several different kinds of multistage networks that should be considered.
An experimental and computational study has been performed on a close rotor-blade/vortex interaction. Surface pressure data was obtained from a rotor operating close to the tip-vortex from an upstream wing in a wind tunnel. Data was obtained for a wide range of blade-vortex proximities, orientations, and blade-tip Mach numbers (up to the transonic regime). A numerical model of these interactions was constructed using the unsteady, three-dimensional, full-potential rotor code called FPR. The model employed an undistorted full-field representation of the measured vortex. This simple model gave excellent comparisons with the data for a wide range of conditions, including parallel head-on interactions. Computational studies have also been performed on the manner of vortex representation and the influence of vortex-core size.
The subject of input/output (I/O) was often been neglected in the design of parallel computer systems, although for many problems I/O rates will limit the speedup attainable. The I/O problem is addressed by considering the role of files in parallel systems. The notion of parallel files is introduced. Parallel files provide for concurrent access by multiple processes, and utilize parallelism in the I/O system to improve performance. Parallel files can also be used conventionally by sequential programs. A set of standard parallel file organizations is proposed, based on common data partitioning techniques. Implementation strategies for the proposed organizations are suggested, using multiple storage devices. Problem areas are also identified and discussed.
Two-dimensional optical strain measurements on high temperature test specimens are presented. This two-dimensional capability is implemented through a rotatable sensitive strain axis. Three components of surface strain can be measured automatically, from which the first and second principal strains are calculated. One- and two-dimensional strain measurements at temperatures beyond 750 C with a resolution of 15 microstrain are demonstrated. The system is based on a one-dimensional speckle shift technique. The speckle shift technique makes use of the linear relationship between surface strain and the differential shift of laser speckle patterns in the diffraction plane. Laser speckle is a phase effect that occurs when spatially coherent light interacts with an optically rough surface. Since speckle is generated by any diffusely reflecting surface, no specimen preparation is needed to obtain a good signal. Testing was done at room temperature on a flat specimen of Inconel 600 mounted in a fatigue testing machine. A load cell measured the stress on the specimen before and after acquiring the speckle data. Strain components were measured at 0 C (parallel to the load axis) and at plus or minus 45 C, and plots indicate the calculated values of the first and second principal strains. The measured values of Young's modulus and Poisson's ratio are in good agreement with handbook values. Good linearity of the principal strain moduli at high temperatures indicate precision and stability of the system. However, a systematic error in the high-temperature test setup introduced a scale factor in the slopes of the two-dimensional stress-strain curves. No high temperature effects, however, have been observed to degrade speckle correlation.
Microwave and infrared measurements are reported for the van der Waals complex ArCH3CCH and its isotopomers. The structure is T-shaped with equilibrium center-of-mass separation of 3.73 A and an angle of 82 deg between the molecule symmetry axis and the van der Waals bond. The infrared and microwave spectra are complex due to the effects of a slightly hindered internal rotor. Analysis of the spectral data shows that the dipole moment is almost parallel to the dimer b axis. A modified vibration/rotation Hamiltonian that includes an internal rotor potential is used to show that the barrier to internal rotation is near 10.8/cm.
In order to simulate the effects of the impingement of hot exhaust jets of High Performance Aircraft on landing surfaces a multi-disciplinary computation coupling flow dynamics to heat conduction in the runway needs to be carried out. Such simulations, which are essentially unsteady, require very large computational power in order to be completed within a reasonable time frame of the order of an hour. Such power can be furnished by the latest generation of massively parallel computers. These remove the bottleneck of ever more congested data paths to one or a few highly specialized central processing units (CPU's) by having many off-the-shelf CPU's work independently on their own data, and exchange information only when needed. During the past year the first phase of this project was completed, in which the optimal strategy for mapping an ADI-algorithm for the three dimensional unsteady heat equation to a MIMD parallel computer was identified. This was done by implementing and comparing three different domain decomposition techniques that define the tasks for the CPU's in the parallel machine. These implementations were done for a Cartesian grid and Dirichlet boundary conditions. The most promising technique was then used to implement the heat equation solver on a general curvilinear grid with a suite of nontrivial boundary conditions. Finally, this technique was also used to implement the Scalar Penta-diagonal (SP) benchmark, which was taken from the NAS Parallel Benchmarks report. All implementations were done in the programming language C on the Intel iPSC/860 computer.
In the Concurrent Computing Laboratory in the Department of Physics and Astronomy at Louisiana State University we have constructed a heterogeneous computing environment that permits us to routinely simulate complicated three-dimensional fluid flows and to readily visualize the results of each simulation via three-dimensional animation sequences. An 8192-node MasPar MP-1 computer with 0.5 GBytes of RAM provides 250 MFlops of execution speed for our fluid flow simulations. Utilizing the parallel virtual machine (PVM) language, at periodic intervals data is automatically transferred from the MP-1 to a cluster of workstations where individual three-dimensional images are rendered for inclusion in a single animation sequence. Work is underway to replace executions on the MP-1 with simulations performed on the 512-node CM-5 at NCSA and to simultaneously gain access to more potent volume rendering workstations.
This viewgraph presentation concludes that a Monte Carlo probability density function (PDF) solution successfully couples with an existing finite volume code; PDF solution method applied to turbulent reacting flows shows good agreement with data; and PDF methods must be run on parallel machines for practical use.
This paper describes an expert system which is designed to perform automatic data analysis, identify anomalous events, and determine the characteristic features of these events. We have employed both artificial intelligence and neural net approaches in the design of this expert system. The artificial intelligence approach is useful because it provides (1) the use of human experts' knowledge of sensor behavior and faulty engine conditions in interpreting data; (2) the use of engine design knowledge and physical sensor locations in establishing relationships among the events of multiple sensors; (3) the use of stored analysis of past data of faulty engine conditions; and (4) the use of knowledge-based reasoning in distinguishing sensor failure from actual faults. The neural network approach appears promising because neural nets (1) can be trained on extremely noisy data and produce classifications which are more robust under noisy conditions than other classification techniques; (2) avoid the necessity of noise removal by digital filtering and therefore avoid the need to make assumptions about frequency bands or other signal characteristics of anomalous behavior; (3) can, in effect, generate their own feature detectors based on the characteristics of the sensor data used in training; and (4) are inherently parallel and therefore are potentially implementable in special-purpose parallel hardware.
In January 1992, the IML-1 FES experiment produced a set of classic experimental data and a 40 hour holographic "movie" of an ensemble of spheres in a fluid in microgravity. Because the data are in the form of holograms, we can study the three-dimensional distribution of particles with unprecedented detail by a variety of methods and for a wide variety of interests. The possession of the holographic movie is tantamount to having a complex experiment in space while working in an easily accessible laboratory on earth. The movie contains a vast amount of useful data, including residual g, g-jitter, convection and transport data, and particle fluid interaction data. The information content in the movie is so great that we have scarcely begun to tap into the data that is actually available in the more than 1000 holograms, each containing as much as 1000 megabytes of information. This ground-based project is exploiting this data and the concept of holographic storage of spaceflight data to provide an understanding of the effects of microgravity in materials processing. This paper provides the foundation, objectives, and status of the ground based project. The primary objective of this project is to advance the understanding of microgravity effects on crystal growth, convection in materials processing in the space environment, and complex transport phenomena at low Reynolds numbers. This objective is being achieved both experimentally and theoretically. Experiments are making use of existing holographic data recorded during the IML- I spaceflight. A parallel theoretical effort is providing the models for understanding the particle fields and their physics in the microgravity environment.
During this final period, BOOMERANG was deployed to McMurdo Mtn., Antarctica in late 1998 and successfully flew a 10.5 day long duration flight. The experiment returned excellent data, and produced the first resolved images of the early universe. These results, as well as those produced during a test flight over North America in August, 1997, are given in the references below. Analysis of the data from the 1998 flight is continuing. In parallel, we have begun to prepare the payload for a long-duration flight from McMurdo in December 2001. For this flight, the focal plane is being outfitted with polarization sensitive detectors, with the goal of detecting the polarization of the CMB that is predicted to exist at degree angular scales.
In order to be truly portable, a program must be tolerant of a wide range of development and execution environments, and a parallel program is just one which must be tolerant of a very wide range. This paper first defines the term "tolerant programming", then describes many layers of tools to accomplish it. The primary focus is on F-Nets, a formal model for expressing computation as a folded partial-ordering of operations, thereby providing an architecture-independent expression of tolerant parallel algorithms. For implementing F-Nets, Cooperative Data Sharing (CDS) is a subroutine package for implementing communication efficiently in a large number of environments (e.g. shared memory and message passing). Software Cabling (SC), a very-high-level graphical programming language for building large F-Nets, possesses many of the features normally expected from today's computer languages (e.g. data abstraction, array operations). Finally, L2(sup 3) is a CASE tool which facilitates the construction, compilation, execution, and debugging of SC programs.
In this work we show how automatic relative debugging can be used to find differences in computation between a serial program and an OpenMP parallel version of that program. Backtracking and re-execution are used to determine the first OpenMP parallel region that produces a difference in computation that may lead to an incorrect value the user has indicated. Tool-parallelized programs are addressed by utilizing static analysis and directive information from the parallelization tool. Manually-parallelized programs are addressed as well by performing data dependence and directive analysis.
A processing system has been developed to meet increasing demands for detailed noise measurement of individual model components. The Microphone Array Phased Processing System (MAPPS) uses graphical user interfaces to control all aspects of data processing and visualization. The system uses networked parallel computers to provide noise maps at selected frequencies in a near real-time testing environment. The system has been successfully used in the NASA Ames 7- by 10-Foot Wind Tunnel.
The Flexible Peripheral Component Interconnect (PCI) Input/Output (I/O) Card is an innovative circuit board that provides functionality to interface between a variety of devices. It supports user-defined interrupts for interface synchronization, tracks system faults and failures, and includes checksum and parity evaluation of interface data. The card supports up to 16 channels of high-speed, half-duplex, low-voltage digital signaling (LVDS) serial data, and can interface combinations of serial and parallel devices. Placement of a processor within the field programmable gate array (FPGA) controls an embedded application with links to host memory over its PCI bus. The FPGA also provides protocol stacking and quick digital signal processor (DSP) functions to improve host performance. Hardware timers, counters, state machines, and other glue logic support interface communications. The Flexible PCI I/O Card provides an interface for a variety of dissimilar computer systems, featuring direct memory access functionality. The card has the following attributes: 8/16/32-bit, 33-MHz PCI r2.2 compliance, Configurable for universal 3.3V/5V interface slots, PCI interface based on PLX Technology's PCI9056 ASIC, General-use 512K 16 SDRAM memory, General-use 1M 16 Flash memory, FPGA with 3K to 56K logical cells with embedded 27K to 198K bits RAM, I/O interface: 32-channel LVDS differential transceivers configured in eight, 4-bit banks; signaling rates to 200 MHz per channel, Common SCSI-3, 68-pin interface connector.
A gridded retarding potential analyzer (RPA) has been developed to characterize the magnetized plasma exhaust of the 10 kW Variable Specific Impulse Magnetoplasma Rocket (VX-10) experiment at NASA's Advanced Space Propulsion Laboratory. In this system, plasma is energized through coupling of radio frequency waves at the ion cyclotron resonance (ICR). The particles are subsequently accelerated in a magnetic nozzle to provide thrust. Downstream of the nozzle, the RPA's mounting assembly enables the detector to make complete axial and radial scans of the plasma. A multichannel collimator can be inserted into the RPA to remove ions with pitch angles greater than approximately 1 deg. A calculation of the general collimator transmission as a function over velocity space is presented, which shows the instrument's sensitivity in detecting changes in both the parallel and perpendicular components of the ion energy. Data from initial VX-10 ICRH experiments show evidence of ion heating.
A method and system for rendering the shape of a multi-core optical fiber or multi-fiber bundle in three-dimensional space in real time based on measured fiber strain data. Three optical fiber cores arc arranged in parallel at 120.degree. intervals about a central axis. A series of longitudinally co-located strain sensor triplets, typically fiber Bragg gratings, are positioned along the length of each fiber at known intervals. A tunable laser interrogates the sensors to detect strain on the fiber cores. Software determines the strain magnitude (.DELTA.L/L) for each fiber at a given triplet, but then applies beam theory to calculate curvature, beading angle and torsion of the fiber bundle, and from there it determines the shape of the fiber in s Cartesian coordinate system by solving a series of ordinary differential equations expanded from the Frenet-Serrat equations. This approach eliminates the need for computationally time-intensive curve-tilting and allows the three-dimensional shape of the optical fiber assembly to be displayed in real-time.