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.
Results of a performance analysis of an airborne Doppler radar wind shear detection system are given in vugraph form. It was concluded that both CO sub 2 and Ho:YAG lasers are feasible for dry microburst applications, but with limited performance in wet microbursts. The Ho:YAG performs better than the CO sub 2 for a set of identical lidar parameters.
Performance analysis of rendezvous charts for Apollo 9 mission
The performance of conventional thermal propulsion systems is fundamentally constrained by the specific energy limitations associated with chemical fuels and the thermal limits of available materials. Electromagnetic thrust augmentation represents one intriguing possibility for improving the fuel composition of thermal propulsion systems, thereby increasing overall specific energy characteristics; however, realization of such a system requires an extremely high-energy-density electrical power source as well as an efficient plasma acceleration device. This Technical Publication describes the development of an experimental research facility for investigating the use of cross-field magnetohydrodynamic (MHD) accelerators as a possible thrust augmentation device for thermal propulsion systems. In this experiment,a 1.5-MW(sub e) Aerotherm arc heater is used to drive a 2-MW(sub e) MHD accelerator. The heatsink MHD accelerator is configured as an externally diagonalized, segmented channel, which is inserted into a large-bore, 2-T electromagnet. The performance analysis and engineering design of the flow path are described as well as the parameter measurements and flow diagnostics planned for the initial series of test runs.
The methodology and standardization layer provided by the Performance Application Programming Interface (PAPI) has played a vital role in application profiling for almost two decades. It has enabled sophisticated performance analysis tool designers and performance-conscious scientists to gain insights into their applications by simply instrumenting their code using a handful of PAPI functions that “just work” across different hardware components. In the past, PAPI development had focused primarily on hardware-specific performance metrics. However, the rapidly increasing complexity of software infrastructure poses new measurement and analysis challenges for the developers of large-scale applications. In particular, acquiring information regarding the behavior of libraries and runtimes—used by scientific applications—requires low-level binary instrumentation, or APIs specific to each library and runtime. No uniform API for monitoring events that originate from inside the software stack has emerged. In this article, we present our efforts to extend PAPI’s role so that it becomes the de facto standard for exposing performance-critical events, which we refer to as software-defined events (SDEs), from different software layers. Upgrading PAPI with SDEs enables monitoring of both types of performance events—hardware- and software-related events—in a uniform way, through the same consistent PAPI. The goal of this article is threefold. First, we motivate the need for SDEs and describe our design decisions regarding the functionality we offer through PAPI’s new SDE interface. Second, we illustrate how SDEs can be utilized by different software packages, specifically, by showcasing their use in the numerical linear algebra library MAGMA-Sparse, the tensor algebra library TAMM that is part of the NWChem suite, and the compiler-based performance analysis tool Byfl. Third, we provide a performance analysis of the overhead that results from monitoring SDEs and discuss the trade-offs between overhead and functionality.
A performance analysis of a Discovery Class Venus Exploration Mission in which aerocapture is used to capture a spacecraft into a 300km polar orbit for a two year science mission has been conducted to quantify its performance. A preliminary performance assessment determined that a high heritage 70 sphere-cone rigid aeroshell with a 0.25 lift to drag ratio has adequate control authority to provide an entry flight path angle corridor large enough for the mission s aerocapture maneuver. A 114 kilograms per square meter ballistic coefficient reference vehicle was developed from the science requirements and the preliminary assessment s heating indicators and deceleration loads. Performance analyses were conducted for the reference vehicle and for sensitivity studies on vehicle ballistic coefficient and maximum bank rate. The performance analyses used a high fidelity flight simulation within a Monte Carlo executive to define the aerocapture heating environment and deceleration loads and to determine mission success statistics. The simulation utilized the Program to Optimize Simulated Trajectories (POST) that was modified to include Venus specific atmospheric and planet models, aerodynamic characteristics, and interplanetary trajectory models. In addition to Venus specific models, an autonomous guidance system, HYPAS, and a pseudo flight controller were incorporated in the simulation. The Monte Carlo analyses incorporated a reference set of approach trajectory delivery errors, aerodynamic uncertainties, and atmospheric density variations. The reference performance analysis determined the reference vehicle achieves 100% successful capture and has a 99.87% probability of attaining the science orbit with a 90 meters per second delta V budget for post aerocapture orbital adjustments. A ballistic coefficient trade study conducted with reference uncertainties determined that the 0.25 L/D vehicle can achieve 100% successful capture with a ballistic coefficient of 228 kilograms per square meter and that the increased ballistic coefficient increases post aerocapture V budget to 134 meters per second for a 99.87% probability of attaining the science orbit. A trade study on vehicle bank rate determined that the 0.25 L/D vehicle can achieve 100% successful capture when the maximum bank rate is decreased from 30 deg/s to 20 deg/s. The decreased bank rate increases post aerocapture delta V budget to 102 meters per second for a 99.87% probability of attaining the science orbit.
As part of the Control Center Operations, the Performance Analysis Group has responsibility for validating all committed Deep Space Network support to the various user projects. The group provides the Network with the analysis support necessary to assure that actual Network performance meets or exceeds committed levels throughout the mission. Validation is carried out as the signals are received in real time by monitoring the data streams and comparing the data quality against predicted values and established performance standards and tolerance limits. An important function is to generate the predictions used to point the antennas, acquire the radio frequency, and validate the monitored Network performance. Another function is anomaly investigation. An integral part of validation is the support task of generating and maintaining performance standards and tolerance limits for each flight project. The Performance Analysis Group is broadly divided into two subgroups: real-time personnel and nonreal-time personnel.
A recently developed algorithm for nonlinear system performance analysis has been applied to an F16 aircraft to begin evaluating the suitability of the method for aerospace problems. The algorithm has a potential to be much more efficient than the current methods in performance analysis for aircraft. This paper is the initial step in evaluating this potential.
This work examines the performance of leading-edge systems designed for machine learning computing, including the NVIDIA DGX-2, Amazon Web Services (AWS) P3, IBM Power System Accelerated Compute Server AC922, and a consumer-grade Exxact TensorEX TS4 GPU server. Representative deep learning workloads from the fields of computer vision and natural language processing are the focus of the analysis. Performance analysis is performed along with a number of important dimensions. Performance of the communication interconnects and large and high-throughput deep learning models are considered. Different potential use models for the systems as standalone and in the cloud also are examined. The effect of various optimization of the deep learning models and system configurations is included in the analysis.
Understanding the behavior of software in execution is a key step in identifying and fixing performance issues. This is especially important in high performance computing contexts where even minor performance tweaks can translate into large savings in terms of computational resource use. To aid performance analysis, developers may collect an execution trace —a chronological log of program activity during execution. As traces represent the full history, developers can discover a wide array of possibly previously unknown performance issues, making them an important artifact for exploratory performance analysis. However, interactive trace visualization is difficult due to issues of data size and complexity of meaning. Traces represent nanosecond-level events across many parallel processes, meaning the collected data is often large and difficult to explore. The rise of asynchronous task parallel programming paradigms complicates the relation between events and their probable cause. Here, to address these challenges, we conduct a continuing design study in collaboration with high performance computing researchers. We develop diverse and hierarchical ways to navigate and represent execution trace data in support of their trace analysis tasks. Through an iterative design process, we developed Traveler , an integrated visualization platform for task parallel traces. Traveler provides multiple linked interfaces to help navigate trace data from multiple contexts. We evaluate the utility of Traveler through feedback from users and a case study, finding that integrating multiple modes of navigation in our design supported performance analysis tasks and led to the discovery of previously unknown behavior in a distributed array library.
The paper deals with the performance of two parallel variants of the simplex algorithm on a message-passing system. First, the simplex algorithm is reviewed, two possible parallelizations of the algorithm are discussed, and results of benchmark speedups of the alternatives are presented. Between column and row partitionings, the row partitioning method is found to be generally superior, while the column partitioning method is more efficient when the number of rows is small, and the number of columns is much greater that the number of rows. Various performance analysis tools are then applied to examine the reasons for relative performance differences, and communication idle time due to global minimization and load imbalances is noted as the main factor in execution slowdown.
A communications service, called Path Service, is currently being developed by the Consultative Committee for Space Data Systems (CCSDS) to provide a mechanism for the efficient transmission of telemetry data from space to ground for complex space missions of the future. This is an important service, due to the large volumes of telemetry data that will be generated during these missions. A preliminary analysis of performance of Path Service is presented with respect to protocol-processing requirements and channel utilization.
Robot coordination and control systems for remote teleoperation applications are by necessity implemented on distributed computers. Modeling and performance analysis of these distributed robotic systems is difficult, but important for economic system design. Performance analysis methods originally developed for conventional distributed computer systems are often unsatisfactory for evaluating real-time systems. The paper introduces a formal model of distributed robotic control systems; and a performance analysis method, based on scheduling theory, which can handle concurrent hard-real-time response specifications. Use of the method is illustrated by a case of remote teleoperation which assesses the effect of communication delays and the allocation of robot control functions on control system hardware requirements.
Methodology used in guidance system performance analysis /assessment/
The past performance analysis conducted on three High Pressure Oxygen Turbopump (HPOTP) bearings from the Space Shuttle Main Engine is presented. Metallurgical analysis of failed bearing balls and races, and wear track and crack configuration analyses were carried out. In addition, one bearing was tested in laboratory at very high axial loads. The results showed that the cracks were surface initiated and propagated into subsurface locations at relatively small angles. Subsurface cracks were much more extensive than was appeared on the surface. The location of major cracks in the races corresponded to high radial loads rather than high axial loads. There was evidence to suggest that the inner races were heated to elevated temperatures. A failure scenario was developed based on the above findings. According to this scenario the HPOTP bearings are heated by a combination of high loads and high coefficient of friction (poor lubrication). Different methods of extending the HPOTP bearing life are also discussed. These include reduction of axial loads, improvements in bearing design, lubrication and cooling, and use of improved bearing materials.
Internet performance is highly correlated with key economic development metrics of a region. According to World Bank, the economic growth of a country increases 1.3% with a 10% increase in the speed of the Internet. Therefore, it is necessary to monitor and understand the performance of the Internet links in the region. It helps to figure out the infrastructural inefficiencies, poor resource allocation, and routing issues in the region. Moreover, it provides healthy suggestions for future upgrades. Therefore, the objective of this paper is to understand the Internet performance and routing infrastructure of South Asian countries in comparison to the developed world and neighboring countries using end-to-end Internet performance measurements. The South Asian countries comprise nearly 32% of the Internet users in Asia and nearly 16% of the world. The Internet performance metrics in the region are collected through the PingER framework. The framework is developed by the SLAC National Accelerator Laboratory, USA and is running for the last 20 years. PingER has 16 monitoring nodes in the region, and in the last year PingER monitors about 40 sites in South Asia using the ubiquitous ping facility. The collected data is used to estimate the key Internet performance metrics of South Asian countries. The performance metrics are compared with the neighboring countries and the developed world. Particularly, the TCP throughput of the countries is also correlated with different development indices. Further, worldwide Internet connectivity and routing patterns of the countries are investigated to figure out the inconsistencies in the region. Furthermore, the performance analysis revealed that the South Asia region is 7-10 years behind the developed regions of North America (USA and Canada), Europe, and East Asia.
Calculations were performed to verify that the Outer Shield Assembly will support the weight of the Water Extended Polyester (WEP) shielding and the lead shielding.
Parallel programming paradigms include process level parallelism, thread level parallelization, and multilevel parallelism. This viewgraph presentation describes a detailed performance analysis of these paradigms for Shared Memory Architecture (SMA). This analysis uses the Paraver Performance Analysis System. The presentation includes diagrams of a flow of useful computations.
A research study on the topology design and performance analysis for the Space Station Information System (SSIS) network is conducted. It is begun with a survey of existing research efforts in network topology design. Then a new approach for topology design is presented. It uses an efficient algorithm to generate candidate network designs (consisting of subsets of the set of all network components) in increasing order of their total costs, and checks each design to see if it forms an acceptable network. This technique gives the true cost-optimal network, and is particularly useful when the network has many constraints and not too many components. The algorithm for generating subsets is described in detail, and various aspects of the overall design procedure are discussed. Two more efficient versions of this algorithm (applicable in specific situations) are also given. Next, two important aspects of network performance analysis: network reliability and message delays are discussed. A new model is introduced to study the reliability of a network with dependent failures. For message delays, a collection of formulas from existing research results is given to compute or estimate the delays of messages in a communication network without making the independence assumption. The design algorithm coded in PASCAL is included as an appendix.