Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “backing layers”

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 757 records · Page 42

AdjointBackMapV2: Precise reconstruction of arbitrary CNN unit’s activation via adjoint operators

Adjoint operators have been found to be effective in the exploration of CNN’s inner workings (Wan and Choe, 2022). However, the previous no-bias assumption restricted its generalization. We overcome the restriction via embedding input images into an extended normed space that includes bias in all CNN layers as part of the extended space and propose an adjoint-operator-based algorithm that maps high-level weights back to the extended input space for reconstructing an effective hypersurface. Such hypersurface can be computed for an arbitrary unit in the CNN, and here we prove that this reconstructed hypersurface, when multiplied by the original input (through an inner product), will precisely replicate the output value of each unit. We show experimental results based on the CIFAR-10 and CIFAR-100 data sets where the proposed approach achieves near 0 activation value reconstruction error.

97 MATHEMATICS AND COMPUTING↗

Influence of hydrogen effusion from hydrogenated silicon nitride layers on the regeneration of boron-oxygen related defects in crystalline silicon

The degradation effect boron doped and oxygen-rich crystalline silicon materials suffer from under illumination can be neutralized in hydrogenated silicon by the application of a regeneration process consisting of a combination of slightly elevated temperature and carrier injection. In this paper, the influence of variations in short high temperature steps on the kinetics of the regeneration process is investigated. It is found that hotter and longer firing steps allowing an effective hydrogenation from a hydrogen-rich silicon nitride passivation layer result in an acceleration of the regeneration process. Additionally, a fast cool down from high temperature to around 550 °C seems to be crucial for a fast regeneration process. It is suggested that high cooling rates suppress hydrogen effusion from the silicon bulk in a temperature range where the hydrogenated passivation layer cannot release hydrogen in considerable amounts. Thus, the hydrogen content of the silicon bulk after the complete high temperature step can be increased resulting in a faster regeneration process. Hence, the data presented here back up the theory that the regeneration process might be a hydrogen passivation of boron-oxygen related defects.

36 MATERIALS SCIENCE↗

Method Accelerates Training Of Some Neural Networks

Three-layer networks trained faster provided two conditions are satisfied: numbers of neurons in layers are such that majority of work done in synaptic connections between input and hidden layers, and number of neurons in input layer at least as great as number of training pairs of input and output vectors. Based on modified version of back-propagation method.

Shelton, Robert O.↗

Pulsed Electromagnetic Acceleration of Plasma: A Review

Much have been learned in the acceleration mechanisms involved in accelerating a plasma electromagnetically in the laboratory over the last 40 years since the early review by Winston Bostik of 1963, but the accumulated understanding is very much scattered throughout the literature. This literature extends back at least to the early sixties and includes Rosenbluth's snowplow model, discussions by Ralph Lovberg, Colgate's boundary-layer model of a current sheet, many papers from the activity at Columbia by Robert Gross and his colleagues, and the relevant, 1-D unsteady descriptions developed from the U. of Maryland theta-pinch studies. Recent progress on the understanding of the pulsed penetration of magnetic fields into collisionless or nearly collisionless plasmas are also be reviewed. Somewhat more recently, we have the two-dimensional, unsteady results in the collisional regime associated with so-called wall-instability in large radius pinch discharges and also in coaxial plasma guns (e.g., Plasma Flow Switch). Among other things, for example, we have the phenomenon of a high- density plasma discharge propagating in a cooaxial gun as an apparently straight sheet (vs paraboloid) because mass re-distribution (on a microsecond timescale) compensates for the 1/r- squared variation of magnetic pressure. We will attempt to collate some of this vast material and bring some coherence tc the development of the subject.

Thio, Y. C. Francis↗

Investigation of Ruthenium Dissolution in Advanced Membrane Electrode Assemblies for Direct Methanol Based Fuel Cell Stacks

Dissolution of ruthenium was observed in the 80-cell stack. Duration testing was performed in single cell MEAs to determine the pathway of cell degradation. EDAX analysis on each of the single cell MEAs has shown that the Johnson Matthey commercial catalyst is stable in DMFC operation for 250 hours, no ruthenium dissolution was observed. Changes in the hydrophobicity of the cathode backing papers was minimum. Electrode polarization analysis revealed that the MEA performance loss is attributed to changes in the cathode catalyst layer. Ruthenium migration does not seem to occur during cell operation but can occur when methanol is absent from the anode compartment, the cathode compartment has access to air, and the cells in the stack are electrically connected to a load (Shunt Currents). The open-to-air cathode stack design allowed for: a) The MEAs to have continual access to oxygen; and b) The stack to sustain shunt currents. Ruthenium dissolution in a DMFC stack can be prevented by: a) Developing an internally manifolded stacks that seal reactant compartments when not in operation; b) Bringing the cell voltages to zero quickly when not in operation; and c) Limiting the total number of cells to 25 in an effort to limit shunt currents.

alternative power sources↗

Possible Detection of Nitrates on Mars by the Sample Analysis at Mars (SAM) Instrument

Planetary models suggest that nitrogen was abundant in the early Martian atmosphere as dinitrogen (N2). However, it has been lost by sputtering and photochemical loss to space [1, 2], impact erosion [3], and chemical oxidation to nitrates [4]. Nitrates, produced early in Mars history, are later decomposed back into N2 by the current impact flux [5], making possible a nitrogen cycle on Mars. It is estimated that a layer of about 3 m of pure NaNO3 should be distributed globally on Mars [5]. Nitrates are a fundamental source for nitrogen to terrestrial microorganisms. Therefore, the detection of soil nitrates is important to assess habitability in the Martian environment. The only previous mission that was designed to search for soil nitrates was the Phoenix mission but was unable to detect evolved N-containing species by TEGA and the MECA WCL [6]. Nitrates have been tentatively identified in the Nakhla meteorite [7]. The purpose of this work is to determine if nitrates were detected in first solid sample (Rocknest) in Gale Crater examined by the SAM instrument.

Navarro-Gonzalez, R.↗

Quantum coherent devices with reduced energy dissipation

The disclosed technology can be used to increase lifetimes of quantum coherent devices such as qubits including transmons and xmons and to reduce the losses in resonators. Energy stored in superconducting devices can be lost by emission of phonons that couple to the environment. Defects in one or more materials in the quantum coherent device can cause the coupling to phonons. Patterning one or more of the layers creates a phononic bandgap that reduces or eliminates the emission of the phonons to the environment. The energy then couples back to the quantum coherent device.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

MoS{sub 2} functionalization for ultra-thin atomic layer deposited dielectrics

The effect of room temperature ultraviolet-ozone (UV-O{sub 3}) exposure of MoS{sub 2} on the uniformity of subsequent atomic layer deposition of Al{sub 2}O{sub 3} is investigated. It is found that a UV-O{sub 3} pre-treatment removes adsorbed carbon contamination from the MoS{sub 2} surface and also functionalizes the MoS{sub 2} surface through the formation of a weak sulfur-oxygen bond without any evidence of molybdenum-sulfur bond disruption. This is supported by first principles density functional theory calculations which show that oxygen bonded to a surface sulfur atom while the sulfur is simultaneously back-bonded to three molybdenum atoms is a thermodynamically favorable configuration. The adsorbed oxygen increases the reactivity of MoS{sub 2} surface and provides nucleation sites for atomic layer deposition of Al{sub 2}O{sub 3}. The enhanced nucleation is found to be dependent on the thin film deposition temperature.

36 MATERIALS SCIENCE↗

Design and application of ion-implanted polySi passivating contacts for interdigitated back contact c-Si solar cells

Ion-implanted passivating contacts based on poly-crystalline silicon (polySi) are enabled by tunneling oxide, optimized, and used to fabricate interdigitated back contact (IBC) solar cells. Both n-type (phosphorous doped) and p-type (boron doped) passivating contacts are fabricated by ion-implantation of intrinsic polySi layers deposited via low-pressure chemical vapor deposition and subsequently annealed. The impact of doping profile on the passivation quality of the polySi doped contacts is studied for both polarities. It was found that an excellent surface passivation could be obtained by confining as much as possible the implanted-and-activated dopants within the polySi layers. The doping profile in the polySi was controlled by modifying the polySi thickness, the energy and dose of ion-implantation, and the temperature and time of annealing. An implied open-circuit voltage of 721 mV for n-type and 692 mV for p-type passivating contacts was achieved. Besides the high passivating quality, the developed passivating contacts exhibit reasonable high conductivity (R{sub sh n-type} = 95 Ω/□ and R{sub sh p-type} = 120 Ω/□). An efficiency of 19.2% (V{sub oc} = 673 mV, J{sub sc} = 38.0 mA/cm{sup 2}, FF = 75.2%, and pseudo-FF = 83.2%) was achieved on a front-textured IBC solar cell with polySi passivating contacts as both back surface field and emitter. By improving the front-side passivation, a V{sub OC} of 696 mV was also measured.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Global change - Geoengineering and space exploration

Geoengineering options and alternatives are proposed for mitigating the effects of global climate change and depletion of the ozone layer. Geoengineering options were discussed by the National Academy of Science Panel on the Policy Implications of Greenhouse Warming. Several of the ideas conveyed in their published report are space-based or depend on space systems for implementation. Among the geoengineering options using space that are discussed include the use of space power systems as an alternative to fossil fuels for generating electricity, the use of lunar He-3 to aid in the development of fusion energy, and the establishment of a lunar power system for solar energy conversion and electric power beaming back to earth. Other geoengineering options are discussed. They include the space-based modulation of hurricane forces and two space-based approaches in dealing with ozone layer depletion. The engineering challenges and policy implementation issues are discussed for these geongineering options.

Jenkins, Lyle M.↗

Scale interactions and anisotropy in Rayleigh–Taylor turbulence

We study energy scale transfer in Rayleigh–Taylor (RT) flows by coarse graining in physical space without Fourier transforms, allowing scale analysis along the vertical direction. Two processes are responsible for kinetic energy flux across scales: baropycnal work Λ, due to large-scale pressure gradients acting on small scales of density and velocity; and deformation work Π, due to multiscale velocity. Our coarse-graining analysis shows how these fluxes exhibit self-similar evolution that is quadratic-in-time, similar to the RT mixing layer. We find that Λ is a conduit for potential energy, transferring energy non-locally from the largest scales to smaller scales in the inertial range where Π takes over. In three dimensions, Π continues a persistent cascade to smaller scales, whereas in two dimensions Π rechannels the energy back to larger scales despite the lack of vorticity conservation in two-dimensional (2-D) variable density flows. This gives rise to a positive feedback loop in 2-D RT (absent in three dimensions) in which mixing layer growth and the associated potential energy release are enhanced relative to 3-D RT, explaining the oft-observed larger α values in 2-D simulations. Despite higher bulk kinetic energy levels in two dimensions, small inertial scales are weaker than in three dimensions. Moreover, the net upscale cascade in two dimensions tends to isotropize the large-scale flow, in stark contrast to three dimensions. Furthermore, our findings indicate the absence of net upscale energy transfer in three-dimensional RT as is often claimed; growth of large-scale bubbles and spikes is not due to ‘mergers’ but solely due to baropycnal work Λ.

42 ENGINEERING↗

Specific features of the nonradiative relaxation of Er{sup 3+} ions in epitaxial Si structures

The specific features of the nonradiative relaxation of Er{sup 3+} ions in Si:Er layers grown by sublimation molecular-beam epitaxy (SMBE) are studied. In Si:Er/Si diode structures containing precipitation-type emitting centers, a resonance photoresponse at the wavelength λ ≈ 1.5 μm is observed, which is indicative of the nonradiative relaxation of Er3+ ions via the energy back-transfer mechanism. Saturation of the erbium-related photocurrent is for the first time observed at high temperatures. This allows estimation of the concentration of Er centers that undergo relaxation via the above-mentioned back-transfer mechanism (N{sub 0} ≈ 5 × 10{sup 16} cm{sup −3}). In terms of order of magnitude, the estimated concentration N{sub 0} corresponds to the concentration of optically active Er ions upon excitation of the Si:Er layers by means of the recombination mechanism. The features of the nonradiative relaxation of Er{sup 3+} ions in Si:Er/Si structures with different types of emitting centers are analyzed.

36 MATERIALS SCIENCE↗

Advanced antireflection for back-illuminated silicon photomultipliers to detect faint light

Silicon photomultipliers have attracted increasing attention for detecting low-density light in both scientific research and practical applications in recent years; yet the photon losses due to reflection on the light-sensitive planar silicon surface considerably limit its photon detection efficiency. Here we demonstrate an advanced light trapping feature by developing the multi-layer antireflection coatings and the textured silicon surface with upright random nano-micro pyramids, which significantly reduces the reflection of faint light in a wide spectrum, from ultraviolet to infrared. Integrating this advanced photon confinement feature into next-generation back-illuminated silicon photomultiplier would increase the photon detection efficiency with significantly lower reflection and much more active areas. This advanced design feature offers the back-illuminated silicon photomultiplier broader application opportunities exemplified in the emerging scenarios such as nuclear medical imaging, light detection and ranging for autonomous driving, detection of scintillation light in ionizing radiation, as well as high energy physics.

42 ENGINEERING↗

...And Eat it Too: High Read Performance in Write-Optimized HPC I/O Middleware File Formats

As HPC applications run on increasingly high process counts on larger and larger machines, both the frequency of checkpoints needed for fault tolerance and the resolution and size of Data Analysis Dumps are expected to increase proportionally. In order to maintain an acceptable ratio of time spent performing useful computation work to time spent performing I/O, write bandwidth to the underlying storage system must increase proportionally to this increase in the checkpoint and computation size. Unfortunately, popular scientific self-describing file formats such as netCDF and HDF5 are designed with a focus on portability and flexibility. Extra care and careful crafting of the output structure and API calls is required to optimize for write performance using these APIs. To provide sufficient write bandwidth to continue to support the demands of scientific applications, the HPC community has developed a number of I/O middleware layers, that structure output into write-optimized file formats. However, the obvious concern with any write optimized file format would be a corresponding penalty on reads. In the log-structured filesystem, for example, a file generated by random writes could be written efficiently, but reading the file back sequentially later would result in very poor performance. Simulation results require efficient read-back for visualization and analytics, and though most checkpoint files are never used, the efficiency of a restart is very important in the face of inevitable failures. The utility of write speed improving middleware would be greatly diminished if it sacrificed acceptable read performance. In this paper we examine the read performance of two write-optimized middleware layers on large parallel machines and compare it to reading data natively in popular file formats. The two I/O middleware layer examined in this paper are the Adaptable IO System (ADIOS), a library-based approach developed at Oak Ridge National Laboratory to provide a high-level IO API that can be used in place of netCDF or HDF5 to do much more aggressive write-behind and efficient reordering of data locations within the file; and the Parallel Log-structured Filesystem (PLFS), a stackable FUSE filesystem approach developed at Los Alamos National Laboratory that decouples concurrent writes to improve the speed of checkpoints. Since ADIOS is an I/O componentization that affords selection of different I/O methods at or during runtime, through a single API, users can have access to MPI-IO, Posix-IO, HDF5, netCDF, and staging methods. The ADIOS BP file format is a new log-file format that has a superset of the features of both HDF5 and netCDF, but is designed to be portable and flexible while being optimized for writing. PLFS, on the other hand, is mounted as a stackable filesystem on top of an existing parallel filesystem. Reads or writes to the PLFS filesystem are transparently translated into operations on per-process log files stored in the underlying parallel filesystem. Since PLFS performs this translation without application modification, users can write in HDF5, netCDF, or app-specific file formats and PLFS will store the writes in a set of efficiently written log-formatted files, while presenting the user with a logical 'flat' file on reads. Despite their different approaches, the commonality behind both of these middleware systems is that they both write to a log file format. As shown in previous publications, writes are fully optimized in both systems, sometimes resulting in 100x improvements over writing data in popular file formats. But as mentioned above, writing is only one part of the story. In this paper we examine the read performance of our middleware layers on large parallel machines and compare these to reading data either natively or from other popular file formats. We compare the reading performance in two different scenarios: (1) Reading back restarts from the same number of processors as wrote the data and (2) Reading back restart data from a different number of processors. We observe that not only can write-optimized I/O middleware be built to not penalize read speeds, but for important workloads, techniques that improve write performance can, perhaps counterintuitively, improve read speeds over reading from a contiguously organized file format. In the remainder of this paper, we investigate this further through case studies of PLFS and ADIOS on simulation checkpoint restarts.

99 GENERAL AND MISCELLANEOUS↗

First stage of reaction of molten Al with MgO substrate

The Al/MgO couple was produced in vacuum (~ 5 × 10{sup −} {sup 4} Pa) by contact heating from RT up to 1000 °C and holding at that temperature for 1 h of a small 4 × 4 × 4 mm aluminium (5 N) sample placed on the [100] MgO single crystal substrate. TEM observations backed with electron diffraction analysis indicated that the interaction between liquid aluminium and MgO starts from a redox reaction producing a continuous layer of MgAl{sub 2}O{sub 4} spinel on the substrate surface. Its growth is controlled by solid state out-diffusion of magnesium and oxygen towards the surface being in contact with liquid metal. The thickening of spinel layer is accompanied by its cracking and infiltration with aluminium. The above process enables local dissolution of the MgO substrate and formation in it of a thin region of interpenetrating metallic channels walled with spinel. The removal of dissolved magnesium through open aluminium channels towards the drop and to vacuum locally produces areas of aluminium enriched with dissolved oxygen, which results in the nucleation of α-Al{sub 2}O{sub 3} at spinel clad walls. The growth of α-Al{sub 2}O{sub 3} is controlled only by the dissolution rate of MgO by aluminium, liquid state diffusion of Mg to drop/vacuum and oxygen to the front of the of α-Al{sub 2}O{sub 3} crystallites growing into MgO substrate. - Highlights: • New unique evidence of first stages of interaction of liquid Al with MgO substrates • Interaction of liquid Al with MgO starts with the formation of a layer MgAl{sub 2}O{sub 4}. • Growth of MgAl{sub 2}O{sub 4} is slow as controlled by solid state out-diffusion of Mg and O. • MgAl{sub 2}O{sub 4} serves as a nucleation site for Al{sub 2}O{sub 3} and consumed by it soon after. • Growth of Al{sub 2}O{sub 3} is fast as controlled by diffusion in liquid state.

36 MATERIALS SCIENCE↗