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.
Here, this work shows how it is possible to reconstruct SuperKEKB’s beam parameters using a Neural Network with beamstrahlung signal from the Large Angle Beamstrahlung Monitor (LABM) as input. We describe the device, the model, and discuss the results.
Methanol, an important liquid fuel and chemical feedstock, has yet to be produced using solar energy, H 2 O, and CO 2 as sole inputs in a standalone device. Here, this study directly addresses this longstanding challenge through presenting the first demonstration of unbiased solar methanol production from CO 2 and H 2 O with a monolithic artificial leaf design, surpassing the previous best energy efficiency in solar alcohol production by at least 1 order of magnitude. We first develop a new generation of photocathodes based on Si micropillar arrays and a cobalt tetraaminophthalocyanine molecular catalyst. By integrating a C 60 interlayer that facilitates unidirectional electron transfer through the semiconductor/catalyst interface, we realize a photovoltage of 500 mV, one of the highest recorded for single-junction Si-based photoelectrodes in aqueous CO 2 reduction, as well as unprecedented methanol formation with a Faradaic efficiency of 30% and a partial current density of 6.3 mA cm –2 . We further integrate the photocathode with a multijunction perovskite photovoltaic minimodule to afford a standalone solar fuel system, which demonstrates a light-to-methanol conversion efficiency of 0.8%, 32 times higher than the present record in light-to-alcohol conversion with an artificial leaf.
We demonstrate an analog O/E/O electronic link to allow integrated laser neurons to accept many distinguishable, high bandwidth input signals simultaneously. This device utilizes wavelength division multiplexing to achieve multi-channel fan-in, a photodetector to sum signals together, and a laser cavity to perform a nonlinear operation. Its speed outpaces accelerated-time neuromorphic electronics, and it represents a viable direction towards scalable networking approaches.
This article presents GaN two-dimensional electron gas (2DEG) Hall plates with low residual offset and noise at 3 V input bias. We studied devices made from three consecutive fabrication generations through current spinning offset measurements in a zero-field chamber. When operated above 1 V, the first-generation devices charted high residual offsets >1 mT. We reduced these residual offsets by three orders of magnitude in later device generations. Two experiments were performed to confirm the improvements. First, the GaN 2DEG Hall-effect plates were measured with infrared microscopy during current spinning to investigate current crowding and Joule heating. The low offset devices had minimal thermal gradients while the high offset devices had large thermal variations during current spinning. Second, zero-field current spinning was used to compare GaN 2DEG Hall-effect plates against typical commercial silicon Hall-effect plates. The Si devices had residual offsets >10μT biased at 3 V and AlGaN/GaN devices had residual offset <4 μT at 3 V, and an InAlN/GaN devices operated at 2 V had a residual offset <4 μT . Thus, for the first time, GaN 2DEG Hall-effect plates were shown to operate with high bias with low offsets and noise, which enables high fidelity sensing applications in extreme environments.
When the SNS site was built around 20 years ago, the Conventional Facilities (CF) control systems were de-signed using 2 communication protocols to allow pro-grammable logic controllers (PLCs) to interface with motors, variable frequency drives (VFDs), and remote input and output (I/O) devices. The protocol chosen to control motors and VFDs is DeviceNet, a CANbus-based protocol developed by Allen-Bradley, a subsidiary of Rockwell Automation. The protocol chosen to communi-cate with remote I/O is ControlNet, another protocol developed by Allen-Bradley. Both of these protocols are obsolete and present reliability and maintainability is-sues, particularly DeviceNet. As the Control Systems Section at SNS is working to modernize control systems throughout the machine, a major goal for PLC-based systems is to remove the obsolete communication proto-cols in favor of standard, ubiquitous Ethernet. To this end, any new VFDs installed use Ethernet communica-tion. Many VFDs are currently being replaced in the Central Utilities Building (CUB) and the Central Exhaust Facility (CEF) and are being removed from DeviceNet in favor of Ethernet communication. Planning is underway to retrofit Eaton Intelligent Technology motor control centers (MCCs) in the Target Building to remove particu-larly troublesome DeviceNet adaptors and replace them with Ethernet adaptors for each motor starter. The Con-trolNet network in the CUB has been demolished, with I/O drops integrated into a local Ethernet network, im-proving sustainability and maintainability.
Despite recent efforts to reduce standby power consumption in plug loads, new trends in the miniaturization and wide distribution of electronics necessitates devices with zero standby consumption. This work introduces two zero standby solutions that wake a device using an external input of energy harvested from a 5 mW laser pointer. These solutions are applicable to electronics that are remotely activated or have a fiber optic connection. The first utilizes a cascoded header switch to allow for simultaneous low-voltage harvesting and high-voltage blocking. The second involves the use of a charge pump to boost the harvested voltage to a level appropriate for the gate of a high-voltage switch. Prototypes for each method are developed in order to demonstrate functionality and identify the associated benefits and drawbacks. The results show that combining the two methods allows for optimal activation range (up to 25 m) and component count.
Electromagnetic device accurately indicates the responses of various sensors in high performance flight research aircraft to conditions encountered in flight. The device responds to sensor inputs to move a slideable armature along an indicator scale by the force of currents induced in the armature winding.
Addition of a secondary pneumatic modulation input improves the standard devices used on thermostatic expansion valves. The valves normally meter incoming refrigerant flow through an orifice valve controlled by a valve-actuating diaphragm.
Electronic amplifier contains matched pair of metal oxide semiconductor field effect transistor devices which have high input impedance and long-term stability. Thermistor in feedback loop provides temperature compensation for large drifts in the sensor.
Methods used to simulate dissipative media environments for antennas are reported, and special techniques are developed to determine the electrical properties of antennas in such media. A new device for measuring antenna input admittance is presented which permits a simple measurement of admittance over a broad frequency range where conventional methods are cumbersome in field use. An apparatus for measuring antenna current and charge distributions is described, and the associated probing errors are discussed. Measured admittances and current and charge distributions are compared with current theories for thin wire linear and loop antennas in the dissipative media and are found to be in good agreement.
The High Efficiency Solar Cell Task was assigned the objective of understanding and developing high efficiency solar cell devices that would meet the cost and performance goals of the Flat Plate Solar Array (FSA) Project. The need for research dealing with high efficiency devices was considered important because of the role efficiency plays in reducing price per watt of generated energy. The R&D efforts conducted during the 1982 to 1986 period are summarized to provide understanding and control of energy conversion losses associated with crystalline silicon solar cells. New levels of conversion efficiency were demonstrated. Major contributions were made both to the understanding and reduction of bulk and surface losses in solar cells. For example, oxides, nitrides, and polysilicon were all shown to be potentially useful surface passivants. Improvements in measurement techniques were made and Auger coefficients and spectral absorption data were obtained for unique types of silicon sheets. New modelling software was developed including a program to optimize a device design based on input characteristics of a cell.
As data communications rates climb toward 10 Gbits/s, clock recovery and synchronization become more difficult, if not impossible, using conventional electronic circuits. The high-speed photonic clock regenerator described in this article may be more suitable for such use. This photonic regenerator is based on a previously reported photonic oscillator capable of fast acquisition and synchronization. With both electrical and optical clock inputs and outputs, the device is easily interfaced with fiber-optic systems. The recovered electrical clock can be used locally and the optical clock can be used anywhere within a several kilometer radius of the clock/carrier regenerator.
Goal Directed Model Inversion (GDMI) is an algorithm designed to generalize supervised learning to the case where target outputs are not available to the learning system. The output of the learning system becomes the input to some external device or transformation, and only the output of this device or transformation can be compared to a desired target. The fundamental driving mechanism of GDMI is to learn from success. Given that a wrong outcome is achieved, one notes that the action that produced that outcome 0 "would have been right if the outcome had been the desired one." The algorithm then proceeds as follows: (1) store the action that produced the wrong outcome as a "target" (2) redefine the wrong outcome as a desired goal (3) submit the new desired goal to the system (4) compare the new action with the target action and modify the system by using a suitable algorithm for credit assignment (Back propagation in our example) (5) resubmit the original goal. Prior publications by our group in this area focused on demonstrating empirical results based on the inverse kinematic problem for a simulated robotic arm. In this paper we apply the inversion process to much simpler analytic functions in order to elucidate the dynamic behavior of the system and to determine the sensitivity of the learning process to various parameters. This understanding will be necessary for the acceptance of GDMI as a practical tool.
Goal Directed Model Inversion (GDMI) is an algorithm designed to generalize supervised learning to the case where target outputs are not available to the learning system. The output of the learning system becomes the input to some external device or transformation, and only the output of this device or transformation can be compared to a desired target. The fundamental driving mechanism of GDMI is to learn from success. Given that a wrong outcome is achieved, one notes that the action that produced that outcome "would have been right if the outcome had been the desired one." The algorithm makes use of these intermediate "successes" to achieve the final goal. A unique and potentially very important feature of this algorithm is the ability to modify the output of the learning module to force upon it a desired syntactic structure. This differs from ordinary supervised learning in the following way: in supervised learning the exact desired output pattern must be provided. In GDMI instead, it is possible to require simply that the output obey certain rules, i.e., that it "make sense" in some way determined by the knowledge domain. The exact pattern that will achieve the desired outcome is then found by the system. The ability to impose rules while allowing the system to search for its own answers in the context of neural networks is potentially a major breakthrough in two ways: 1) it may allow the construction of networks that can incorporate immediately some important knowledge, i.e. would not need to learn everything from scratch as normally required at present, and 2) learning and searching would be limited to the areas where it is necessary, thus facilitating and speeding up the process. These points are illustrated with examples from robotic path planning and parametric design.
Goal Directed Model Inversion (GDMI) is an algorithm designed to generalize supervised learning to the case where target outputs are not available to the learning system. The output of the learning system becomes the input to some external device or transformation, and only the output of this device or transformation can be compared to a desired target. The fundamental driving mechanism of GDMI is to learn from success. Given that a wrong outcome is achieved, one notes that the action that produced that outcome "would have been right if the outcome had been the desired one." The algorithm makes use of these intermediate "successes" to achieve the final goal. A unique and potentially very important feature of this algorithm is the ability to modify the output of the learning module to force upon it a desired syntactic structure. This differs from ordinary supervised learning in the following way: in supervised learning the exact desired output pattern must be provided. In GDMI instead, it is possible to require simply that the output obey certain rules, i.e., that it "make sense" in some way determined by the knowledge domain. The exact pattern that will achieve the desired outcome is then found by the system. The ability to impose rules while allowing the system to search for its own answers in the context of neural networks is potentially a major breakthrough in two ways: (1) it may allow the construction of networks that can incorporate immediately some important knowledge, i.e., would not need to learn everything from scratch as normally required at present; and (2) learning and searching would be limited to the areas where it is necessary, thus facilitating and speeding up the process. These points are illustrated with examples from robotic path planning and parametric design.
We have used the "MicroStripes" code (Flomerics, Inc.) to perform full-, near- and far-field diffraction modeling of metal mesh performance on substrates. Our Miles Code software, which approximates the full calculation in a quick, gui-based window, is useful as an iterative device by adjusting the input parameters (index of refraction, thickness, etc.) to provide agreement with the full calculation. However, despite the somewhat extravagant claims by the MicroStripes manufacturer, this code is also not perfect because numerous free parameters must be set. Key among these, as identified in our earlier papers and proposal documents, is the high frequency (i.e., far IR) character of the real and imaginary parts of the index of refraction of the metal mesh, the high frequency character of the real and imaginary parts of the index of refraction of the substrate, and the character of the interface between the mesh and the substrate material, and in particular the suppression (or possible enhancement) of surface effects at the interface.
A methodology for designing operational facilities for collaboration by multiple experts has begun to take shape as an outgrowth of a project to design such facilities for scientific operations of the planned 2003 Mars Exploration Rover (MER) mission. The methodology could also be applicable to the design of military "situation rooms" and other facilities for terrestrial missions. It was recognized in this project that modern mission operations depend heavily upon the collaborative use of computers. It was further recognized that tests have shown that layout of a facility exerts a dramatic effect on the efficiency and endurance of the operations staff. The facility designs (for example, see figure) and the methodology developed during the project reflect this recognition. One element of the methodology is a metric, called effective capacity, that was created for use in evaluating proposed MER operational facilities and may also be useful for evaluating other collaboration spaces, including meeting rooms and military situation rooms. The effective capacity of a facility is defined as the number of people in the facility who can be meaningfully engaged in its operations. A person is considered to be meaningfully engaged if the person can (1) see, hear, and communicate with everyone else present; (2) see the material under discussion (typically data on a piece of paper, computer monitor, or projection screen); and (3) provide input to the product under development by the group. The effective capacity of a facility is less than the number of people that can physically fit in the facility. For example, a typical office that contains a desktop computer has an effective capacity of .4, while a small conference room that contains a projection screen has an effective capacity of around 10. Little or no benefit would be derived from allowing the number of persons in an operational facility to exceed its effective capacity: At best, the operations staff would be underutilized; at worst, operational performance would deteriorate. Elements of this methodology were applied to the design of three operations facilities for a series of rover field tests. These tests were observed by human-factors researchers and their conclusions are being used to refine and extend the methodology to be used in the final design of the MER operations facility. Further work is underway to evaluate the use of personal digital assistant (PDA) units as portable input interfaces and communication devices in future mission operations facilities. A PDA equipped for wireless communication and Ethernet, Bluetooth, or another networking technology would cost less than a complete computer system, and would enable a collaborator to communicate electronically with computers and with other collaborators while moving freely within the virtual environment created by a shared immersive graphical display.
All efforts to develop electronic equipment reach a stage where they need a board test station for each board. The SMAP digital system consists of three board types that interact with each other using interfaces with critical timing. Each board needs to be tested individually before combining into the integrated digital electronics system. Each board needs critical timing signals from the others to be able to operate. A bench test system was developed to support test of each board. The test system produces all the outputs of the control and timing unit, and is delivered much earlier than the timing unit. Timing signals are treated as data. A large file is generated containing the state of every timing signal at any instant. This file is streamed out to an IO card, which is wired directly to the device-under-test (DUT) input pins. This provides a flexible test environment that can be adapted to any of the boards required to test in a standalone configuration. The problem of generating the critical timing signals is then transferred from a hardware problem to a software problem where it is more easily dealt with.