SEARCH · Engineering Papers
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 computing system was designed with the capability of providing an effective throughput of one billion floating point operations per second for three dimensional Navier-Stokes codes. The methodology used in defining the baseline design, and the major elements of the numerical aerodynamic simulation facility are described.
The Burroughs Corporation solution to the problem of numeric aerodynamic simulation consists of a computing system designed to meet an effective throughput of one billion floating point operations per second for three-dimensional Navier-Stokes codes. In order to fully appreciate the design, its features, and subtleties, the methodology of the study which evolved this solution and the impact on the processor architecture evolution are described as well as details of the baseline design.
The basic module consists of a minicomputer, low cost peripheral storage device (disk) and a modest number (8-12) of microcomputer modules. A simple arrangement, where the microcomputers are connected to a single time multiplexed bus, only communicating to the host minicomputer, will be efficient. By running the machine in a dedicated mode for long periods of time, it will be possible to obtain a large number of solutions. As such, the device should be useful as a research tool. A scheme is outlined to assemble a number of these computing modules in parallel to decrease computing time. The advantages and disadvantages are discussed of using a number of these systems assembled in a loosely coupled configuration, each independently computing a separate flow, to give a very high throughput.
The results of a system definition study (theoretical) for an Advanced Meteorological Temperature Sounder (AMTS) is described. From the data the atmospheric temperature and humidity profiles can be determined over the entire earth's surface with a spatial resolution of 45 km. x 45 km; amounts and type of cloud cover as well as surface temperatures of the earth are also determined. The major purpose of the study was to determine the feasibility of cooling twenty-eight detectors to the 80-90 Kelvin region by means of a radiative cooler. Other related considerations were achieving high signal-to-noise ratios, maximizing optical throughput through the grating spectrometer, and reducing preamplifier noise. A detailed optical design of an f/5 Ebert-Fastie spectrometer was carried out to verify that image quality is adequate; field lenses near the spectrometer focal plane were designed to image the grating onto the smallest size detectors for each channel.
This paper proposes a process for inexpensive high performance solar cell fabrication that can be automated for further cost reduction and higher throughputs. The unique feature of the process is the use of oxides as doping sources for simultaneous n(+) junction formation and back p(+) layer, as a mask for metallization and as an in situ AR coating for spectrum matching. Cost analysis is performed to show that significant cost reductions over the conventional process is possible using the proposed scheme and the cost intensive steps are identified which can be further reduced to make the process compatible with the needed price goals of 50 cents/watt. The process was demonstrated by fabricating n(+)-p cells using Arsenic doped oxides. Simple n(+)-p structure cells showed corrected efficiencies of 14.5% (AMO) and 12% with doped oxide as an in situ antireflection coating.
The structure of unconfined flames stabilized on a NASA contra-swirl can has been examined. The structure depends on reference velocity and fuel-air ratio, and photographic illustrations of representative examples of each flame type are shown. A highly compact flame is obtained over a narrow band of fuel-air ratios, within the total range over which flames may be stabilized, at high reference velocity (27.8 m/s). The compact structure can only be obtained with careful fuel nozzle location. Chemical concentration profiles, temperature profiles and sodium chloride seeding experiments all confirm the visual impression of compactness. A flame with 50% greater fuel throughput has a much less compact structure, which is reflected in lower maximum temperature, much more complex chemical concentration distributions and a more diffuse recirculation boundary as observed by sodium chloride seeding than in the compact flame. Each of these flames has not only an evident axial recirculation core but also a toroidal recirculation region over the bluff region between the inner and outer swirlers. In contrast, laser Doppler velocimetry measurements carried out on a flame with lower reference velocity (5.65 m/s) do not indicate the presence of a secondary recirculation region in this flame.
The extrusion on a single screw extruder is examined. The process is divided into several steps: the dosage of the materials to be conveyed; the modification of the shape of the feeding opening which influences the feeding process and consequently the throughput of the extruder; optimizing the shape of the feeding zone to meet the specific material requirements; and plasticizing and homogenizing.
A comparison of MSS radiometric processing techniques identified as a preferred radiometric processing technique a procedure which equalizes the mean and standard deviation of detector-specific histograms of uncalibrated scene data. Evaluation of MSS calibration data demonstrated that the relationship between detector responses is essentially linear over the range of intensities typically observed in MSS data, and that the calibration wedge data possess a high degree of temporal stability. An analysis of the preferred radiometric processing technique showed that it could be incorporated into the MDP-MSS system without a major redesign of the system, and with minimal impact on system throughput.
A miniplant, consisting of a 5 cm-diameter fluidized-bed reactor and associated equipment was used to study the deposition parameters, temperature, reactant composition, seed particle size, bed depth, reactant throughput, and methods of reactant introduction. It was confirmed that the permissible range of fluidized-bed temperature was limited at the lower end by zinc condensation (918 C) and at higher temperatures by rapidly decreasing conversion efficiency. Use of a graded bed temperature was shown to increase the conversion efficiency over that obtained in an isothermal bed. Other aspects of the process such as the condensation and fused-salt electrolysis of the ZnCl2 by-product for recycle of zinc and chlorine were studied to provide information required for design of a 50 MT/year experimental facility. In view of the favorable technical and economic indications obtained, it was recommended that construction and operation of the 50 MT/year experimental facility be implemented.
Egg-white solution was chosen as the reference solution in order to assess the effects of operational parameters (voltage, flow rate, ampholine pH range and concentration, and protein concentration) of the RIEF apparatus on protein resolution. Topics of discussion include: (1) comparison of RIEF apparatus to conventional IEF techniques (column and PAG) with respect to resolution and throughput; (2) peptide and protein separation (AHF, Thymosin - Fraction 5, vasoactive peptide, L-asparaginase and ACP); and (3) detection of peptides - dansyl derivatives of amino acids and peptides, post-focusing fluorescent labeling of amino acids, peptides and proteins, and ampholine extraction from focused gels.
The various thermal diodes available for use in cryogenic systems are described. Two diode types, liquid-trap and liquid-blockage diodes, were considered to be the most attractive, and thermal models were constructed to predict their behavior in the reverse mode. The diodes, which are of similar size and throughput, were also examined experimentally in a parallel test setup under nominally identical conditions. Their characteristics were ascertained in terms of forward-mode and reverse-mode conductances, shutdown times and energies, and recovery to forward-mode operation with ethane as the working fluid in the temperature range 170 K to 220 K. Results show that the liquid-blockage diode is the quicker of the two diodes to shut down from the forward mode (8 min as opposed to 10 min). However, the liquid-blockage diode has a larger reverse-mode conductance which results in a greater overall evaporator temperature rise. The importance of the relative size and heat inputs to the condenser/reservoir configuration of the liquid-blockage diode and the evaporator trap configuration for the liquid-trap diode are demonstrated. Also included are data which show the susceptibility of the diodes to recovery to forward-mode operation. Guidelines for the choice of a particular diode for an actual application are given.
The pulse-height analyzer basically resembles a time-sharing multiplexing data-acquisition system which acquires analog data (from energetic particle spectrometers) and converts them into digital code. The PHA simultaneously acquires pulse-height information from the analog signals of the four input channels and sequentially multiplexes the digitized data to a microprocessor. The PHA together with the microprocessor form an on-board real-time data-manipulation system. The system processes data obtained during the rocket flight and reduces the amount of data to be sent back to the ground station. Consequently the data-reduction process for the rocket experiments is speeded up. By using a time-sharing technique, the throughput rate of the microprocessor is increased. Moreover, data from several particle spectrometers are manipulated to share one information channel; consequently, the TM capacity is increased.
There were three major issues examined in the feasibility study. First, the ability of the proposed system architecture to support the anticipated workload was evaluated. Second, the throughput of the computational engine (the flow model processor) was studied using real application programs. Third, the availability, reliability, and maintainability of the system were modeled. The evaluations were based on the baseline systems. The results show that the implementation of the Numerical Aerodynamic Simulation Facility, in the form considered, would indeed be a feasible project with an acceptable level of risk. The technology required (both hardware and software) either already exists or, in the case of a few parts, is expected to be announced this year.
There were three major issues examined in the feasibility study. First, the ability of the proposed system architecture to support the anticipated workload was evaluated. Second, the throughput of the computational engine (the flow model processor) was studied using real application programs. Third, the availability reliability, and maintainability of the system were modeled. The evaluations were based on the baseline systems. The results show that the implementation of the Numerical Aerodynamic Simulation Facility, in the form considered, would indeed be a feasible project with an acceptable level of risk. The technology required (both hardware and software) either already exists or, in the case of a few parts, is expected to be announced this year. Facets of the work described include the hardware configuration, software, user language, and fault tolerance.
The specific requirements for spacecraft computing systems are considered. These requirements are partly related to the constraints of limited resources of power, weight, and volume. Another important factor is the requirement of extremely high reliability. These reliability requirements have led to introduction of automated redundancy techniques on board the spacecraft. The various redundant computers check each other and provide recovery procedures when a computer is found to have failed. Past and future capabilities are considered along with distributed processing requirements. System considerations are discussed, taking into account suboptimum computer throughput, sensitivity to software modifications, hierarchic timing, I/O granularity, restricted communications, synchronous functions, hierarchic control, and concurrent error detection. A description is presented of the Unified Data System (UDS), which consists of a set of standard microcomputers connected by several buses. Attention is also given to synchronization and timing, the executive control structure, the programming language, and the executive program.
The paper presents an experimentally developed surface macro-structuring process suitable for high volume production of silicon solar cells. The process lends itself easily to automation for high throughput to meet low-cost solar array goals. The tetrahedron structure observed is 0.5 - 12 micron high. The surface has minimal pitting with virtually no or very few undeveloped areas across the surface. This process has been developed for (100) oriented as cut silicon. Chemi-etched, hydrophobic and lapped surfaces were successfully texturized. A cost analysis as per Samics is presented.
High area throughput rate will be required for the economical production of silicon dendritic web for solar cells. Web width depends largely on the temperature distribution on the melt surface while growth speed is controlled by the dissipation of the latent heat of fusion. Thermal models were developed to investigate each of these aspects, and were used to engineer the design of laboratory equipment capable of producing crystals over 4 cm wide; growth speeds up to 10 cm/min were achieved. The web crystals were characterized by resistivity, lifetime and etch pit density data as well as by detailed solar cell I-V data. Solar cells ranged in efficiency from about 10 to 14.5% (AM-1) depending on growth conditions. Cells with lower efficiency displayed lowered bulk lifetime believed to be due to surface contamination.
As LACIE Procedure 1 evolved from the Classification and Mensuration Subsystem smallfields procedures, it became evident that two computational systems would have merit-the LACIE/Earth Resources Interactive Processing System based on a large IBM-360 computer oriented for operational use with high computational throughput, and a smaller, highly interactive system based on a PDP 11-45 minicomputer and its display system, the IMAGE-100. The latter had advantages for certain phases; notably, interactive spectral aids could be implemented quite rapidly. This would allow testing and development of Procedure 1 before its implementation on the LACIE/Earth Resources Interactive Processing System. The resulting minicomputer system, called the Classification and Mensuration Subsystem IMAGE-100 Hybrid System, allowed Procedure-1 operations to be performed interactively, except for clustering, classification, and automatic selection of best acquisitions, which were offloaded to the LACIE/Earth Resources Interactive Processing System.