Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “performance improvement”

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 937 records · Page 52

Collaborative Clustering for Sensor Networks

Traditionally, nodes in a sensor network simply collect data and then pass it on to a centralized node that archives, distributes, and possibly analyzes the data. However, analysis at the individual nodes could enable faster detection of anomalies or other interesting events, as well as faster responses such as sending out alerts or increasing the data collection rate. There is an additional opportunity for increased performance if individual nodes can communicate directly with their neighbors. Previously, a method was developed by which machine learning classification algorithms could collaborate to achieve high performance autonomously (without requiring human intervention). This method worked for supervised learning algorithms, in which labeled data is used to train models. The learners collaborated by exchanging labels describing the data. The new advance enables clustering algorithms, which do not use labeled data, to also collaborate. This is achieved by defining a new language for collaboration that uses pair-wise constraints to encode useful information for other learners. These constraints specify that two items must, or cannot, be placed into the same cluster. Previous work has shown that clustering with these constraints (in isolation) already improves performance. In the problem formulation, each learner resides at a different node in the sensor network and makes observations (collects data) independently of the other learners. Each learner clusters its data and then selects a pair of items about which it is uncertain and uses them to query its neighbors. The resulting feedback (a must and cannot constraint from each neighbor) is combined by the learner into a consensus constraint, and it then reclusters its data while incorporating the new constraint. A strategy was also proposed for cleaning the resulting constraint sets, which may contain conflicting constraints; this improves performance significantly. This approach has been applied to collaborative clustering of seismic and infrasonic data collected by the Mount Erebus Volcano Observatory in Antarctica. Previous approaches to distributed clustering cannot readily be applied in a sensor network setting, because they assume that each node has the same view of the data set. A view is the set of features used to represent each object. When a single data set is partitioned across several computational nodes, distributed clustering works; all objects have the same view. But when the data is collected from different locations, using different sensors, a more flexible approach is needed. This approach instead operates in situations where the data collected at each node has a different view (e.g., seismic vs. infrasonic sensors), but they observe the same events. This enables them to exchange information about the likely cluster membership relations between objects, even if they do not use the same features to represent the objects.

Wagstaff. Loro :/↗

Improved cache performance in Monte Carlo transport calculations using energy banding

We present an energy banding algorithm for Monte Carlo (MC) neutral particle transport simulations which depend on large cross section lookup tables. In MC codes, read-only cross section data tables are accessed frequently, exhibit poor locality, and are typically too much large to fit in fast memory. Thus, performance is often limited by long latencies to RAM, or by off-node communication latencies when the data footprint is very large and must be decomposed on a distributed memory machine. The proposed energy banding algorithm allows maximal temporal reuse of data in band sizes that can flexibly accommodate different architectural features. The energy banding algorithm is general and has a number of benefits compared to the traditional approach. In the present analysis we explore its potential to achieve improvements in time-to-solution on modern cache-based architectures.

97 MATHEMATICS AND COMPUTING↗

Streamline Integration Using MPI-Hybrid Parallelism on a Large Multicore Architecture

Streamline computation in a very large vector field data set represents a significant challenge due to the nonlocal and data-dependent nature of streamline integration. In this paper, we conduct a study of the performance characteristics of hybrid parallel programming and execution as applied to streamline integration on a large, multicore platform. With multicore processors now prevalent in clusters and supercomputers, there is a need to understand the impact of these hybrid systems in order to make the best implementation choice. We use two MPI-based distribution approaches based on established parallelization paradigms, parallelize over seeds and parallelize over blocks, and present a novel MPI-hybrid algorithm for each approach to compute streamlines. Our findings indicate that the work sharing between cores in the proposed MPI-hybrid parallel implementation results in much improved performance and consumes less communication and I/O bandwidth than a traditional, nonhybrid distributed implementation.

97 MATHEMATICS AND COMPUTING↗

Improved electrical performances of plasma-enhanced atomic layer deposited TaC{sub x}N{sub y} films by adopting Ar/H{sub 2} plasma

TaC{sub x}N{sub y} films were grown by a plasma-enhanced atomic layer deposition using Ta(N-t-C{sub 5}H{sub 11})[N(CH{sub 3}){sub 2}]{sub 3} as the precursor and H{sub 2} or Ar/H{sub 2} plasma as the reducing agent. The Ar/H{sub 2} plasma appeared to efficiently break the Ta-N bonds in the Ta precursor and formed more TaC{sub x}, which significantly decreased the resistivity of the films ({approx}255 {mu}{omega} cm) compared with the case of the H{sub 2} plasma ({approx}1570 {mu}{omega} cm). The Ar/H{sub 2} plasma also made the films denser and efficiently eliminated the oxygen from the films. This improved the resistance against the elemental diffusion as well as the aging characteristics of the films after exposure to air.

36 MATERIALS SCIENCE↗

Substitution of virgin carbon fiber with low-cost recycled fiber in automotive grade injection molding polyamide 66 for equivalent composite mechanical performance with improved sustainability

Recycled carbon fiber has historically proven challenging to integrate into composite manufacturing due in no small part to the low-density, randomly oriented, discontinuous fiber format that results from typical recycling. Discontinuous fiber requires the use of alternative technologies than those traditionally applied to continuous fiber (e.g., hand lay-up, winding). Extrusion compounding is one such applicable technology, but material transfer into the system requires alternative feeding equipment or the use of an altered procedure as trialed in this research. In this study, an injection molding compound for automotive applications was prepared with recycled carbon fiber and compared against an existing commercial compound. Input fibers and molded compound were evaluated for mechanical performance, while relevant variables such as compounded fiber alignment and aspect ratio were compared to the existing baseline material to confirm a like-for-like composite material structure. Overall, analysis indicates that recycled fiber performs similarly to virgin fiber reinforcement, demonstrating that recycled fiber may be a viable drop-in replacement for short-fiber discontinuous applications.

42 ENGINEERING↗

A Flexible-blocking Based Approach for Performance Tuning of Matrix Multiplication Routines for Large Matrices with Edge Cases

Efficient and scalable matrix operations are being highly demanding in the recent era of Machine Learning, Deep Learning, and Big Data Analytics. The two commonly used matrix-matrix operations in the Basic Linear Algebra Subprograms (BLAS) specification are General Matrix-Matrix multiplication (GEMM) and Symmetric Rank-k update (SYRK). The SYRK routine is a specialization of the GEMM routine, where half of the multiplications are skipped as the resultant matrix is known to be symmetric. Fortunately, several linear algebra libraries implement these BLAS routines quite efficiently. The libraries usually partition the input matrices into blocks and place them in processor caches, thus improving performance by leveraging the caches. However, the contemporary libraries are highly optimized for squarish matrices, but the performance degrades significantly for the matrices with edge case (strictly thin or strictly fat shapes) in the multicore machine. The primary reason is that the current state-of-the-art libraries make fixed block shapes based on a processor architecture, and do not consider the shape of the input matrices. In this paper, we propose a new blocking approach, we name it Flexible-blocking, to mitigate the scalability issues. In contrast to the contemporary libraries, our approach formulates the blocks of the input matrices based on the shapes of the matrices as well as the number of threads used in the implementation. Our proposed technique shows noticeable performance improvement on multicore shared-memory machines for the edge case matrices.

Hossain, Md Mosharaf↗

Using data tagging to improve the performance of Kanerva's sparse distributed memory

The standard formulation of Kanerva's sparse distributed memory (SDM) involves the selection of a large number of data storage locations, followed by averaging the data contained in those locations to reconstruct the stored data. A variant of this model is discussed, in which the predominant pattern is the focus of reconstruction. First, one architecture is proposed which returns the predominant pattern rather than the average pattern. However, this model will require too much storage for most uses. Next, a hybrid model is proposed, called tagged SDM, which approximates the results of the predominant pattern machine, but is nearly as efficient as Kanerva's original formulation. Finally, some experimental results are shown which confirm that significant improvements in the recall capability of SDM can be achieved using the tagged architecture.

Rogers, David↗

A Passivity-Shortage Based Control Design for Teleoperation With Time-Varying Delays

This letter investigates the effect of time-varying delays in bilateral teleoperation, with respect to stability and performance, using the properties of passivity-shortage. Until recently, the desired stability and performance characteristics were achieved using the concept of passivity. However, passivity is limited to systems of relative degree zero or one, while passivity-shortage includes systems of higher relative degrees and possibly of non-minimum phase. Passivity-shortage also arises naturally when data transmission is subject to delays, either constant or time-varying. In this letter, the properties of passivity-shortage are employed to design a simple negative feedback controller. We show that the proposed method provides a faster responding solution and improved performance compared to the existing approaches. The performance improvements include improved steady-state error convergence, and robustness against environmental disturbances, even in the presence of varying delays.

42 ENGINEERING↗

Fine-Grained Exploitation of Mixed Precision for Faster CNN Training

As deep convolutional neural networks (CNNs) have become increasingly popular and successful at an ever-widening number of machine learning tasks specialized hardware has become increasingly available for training and deploying them. NVIDIA's recent Volta architecture includes tensor cores which perform a fused operation reduced and mixed precision (16-bit multiply, 32-bit accumulate). Recent research indicates that, typically, very little is lost (in terms of training accuracy) when half precision is used in place of single precision, and performance gains can be made by doing arithmetic in reduced precision. In this work we demonstrate that making layer-by-layer choices as to the arithmetic/data precision can lead to further performance improvement. In our study of 25,200 CNNs we demonstrate an average speedup (over purely half precision) of 1.27x and speedups as high as 3.64x by appropriately combining single and half precision arithmetic and data types on a layer-by-layer basis.

Johnston, Travis↗

Improving the Comprehensive Performance of Na 0.7 MnO 2 for Sodium Ion Batteries by ZrO 2 Atomic Layer Deposition

Sodium ion batteries with Na-Mn-O compounds as cathode have been widely studied as substitutes for lithium ion batteries due to their abundant resources. Still, the relatively poor cycling stability and low capacity of Na-Mn-O compounds significantly limit their applications. Different approaches, including element substitution and surface modification, have been applied to improve the electrochemical performance of those cathode materials. Herein, element doping and coating of ZrO 2 on Na 0.7 MnO 2 particles have been achieved by atomic layer deposition (ALD) followed by post-annealing. The rate capability and cycling stability of the modified material were significantly improved, and the mechanism of performance enhancement was revealed. The ZrO 2 coatings acted as a stable interfacial layer to enhance the cycling stability of Na 0.7 MnO 2 by suppressing side reactions between the electrode and electrolyte. The doping of transition metal ions reduced energy barriers for sodium ion insertion and deintercalation during cycling, further improving the charge and discharge capacity and rate performance of Na 0.7 MnO 2 .

25 ENERGY STORAGE↗

Effects of fuel-capsule shimming and drive asymmetry on inertial-confinement-fusion symmetry and yield

Three orthogonal proton emission imaging cameras were used to study the 3D effects of low-mode drive asymmetries and target asymmetries on nuclear burn symmetry and yield in direct-drive, inertial-confinement-fusion experiments. The fusion yield decreased quickly as the burn region became asymmetric due to either drive or capsule asymmetry. Here, measurements and analytic scaling are used to predict how intentionally asymmetric capsule shells could improve performance by compensating for drive asymmetry when it cannot be avoided (such as with indirect drive or with polar direct drive).

01 COAL, LIGNITE, AND PEAT↗

Porous Diblock Copolymer Thin Films in High-Performance Semiconductor Microelectronics

The engine fueling more than 40 years of performance improvements in semiconductor integrated circuits (ICs) has been industry's ability to pattern circuit elements at ever-higher resolution and with ever-greater precision. Steady advances in photolithography - the process wherein ultraviolet light chemically changes a photosensitive polymer resist material in order to create a latent image - have resulted in scaling of minimum printed feature sizes from tens of microns during the 1980s to sub-50 nanometer transistor gate lengths in today's state-of-the-art ICs. The history of semiconductor technology scaling as well as future technology requirements is documented in the International Technology Roadmap for Semiconductors (ITRS). The progression of the semiconductor industry to the realm of nanometer-scale sizes has brought enormous challenges to device and circuit fabrication, rendering performance improvements by conventional scaling alone increasingly difficult. Most often this discussion is couched in terms of field effect transistor (FET) feature sizes such as the gate length or gate oxide thickness, however these challenges extend to many other aspects of the IC, including interconnect dimensions and pitch, device packing density, power consumption, and heat dissipation. The ITRS Technology Roadmap forecasts a difficult set of scientific and engineering challenges with no presently-known solutions. The primary focus of this chapter is the research performed at IBM on diblock copolymer films composed of polystyrene (PS) and poly(methyl-methacrylate) (PMMA) (PS-b-PMMA) with total molecular weights M{sub n} in the range of {approx}60K (g/mol) and polydispersities (PD) of {approx}1.1. These materials self assemble to form patterns having feature sizes in the range of 15-20nm. PS-b-PMMA was selected as a self-assembling patterning material due to its compatibility with the semiconductor microelectronics manufacturing infrastructure, as well as the significant body of existing research on understanding its material properties.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Improving the performance for direct electrolysis of CO 2 in solid oxide electrolysis cells with a Sr 1.9 Fe 1.5 Mo 0.5 O 6– δ electrode via infiltration of Pr 6 O 11 nanoparticles

Direct CO 2 electrolysis using solid oxide electrolysis cells (CO 2 -SOECs) holds promise to efficiently convert carbon dioxide to carbon monoxide and oxygen. Cathodes with desirable catalytic activity and chemical stability play a critical role in the development of direct CO 2 -SOECs. Although Sr 2 Fe 1.5 Mo 0.5 O 6–δ (SFM) has exhibited promise for direct CO 2 -SOECs due to its redox stability, it suffers from insufficient activity for the CO 2 reduction reaction (CO 2 RR). Here we report interface engineering of nanosized Pr 6 O 11 on the SFM cathode obtained through infiltration to promote the CO 2 RR performance for direct CO 2 -SOECs. The effect of Pr 6 O 11 loading on the performance of the CO 2 RR is systematically investigated. At 800 °C, the current density of the Pr 6 O 11 infiltrated SFM cathode with an optimum Pr 6 O 11 loading of 14.8 wt% reaches 1.61 A cm –2 at 1.5 V, more than double that of the SFM cathode (0.76 A cm –2 ) under the same operating conditions. X-ray photoelectron spectroscopy (XPS) characterization and in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) analysis indicate that the adsorption ability of CO 2 on the SFM cathode has been significantly improved by the formation of Pr 6 O 11 . Temperature-programmed desorption (TPD) of CO 2 measurements further manifest that a 14.8 wt% Pr 6 O 11 -SFM cathode has better CO desorption capacity. In addition, polarization resistance of the SFM cathode has significantly decreased with the addition of Pr 6 O 11 . Three-electrode measurement was used to analyze the improved electrode kinetics. Finally, these results demonstrate that the formation of Pr 6 O 11 in the SFM cathode through infiltration is a promising approach for increasing CO 2 RR activity for CO 2 -SOECs.

03 NATURAL GAS↗

Adaptive Evolution of Synthetic Cooperating Communities Improves Growth Performance

Symbiotic interactions between organisms are important for human health and biotechnological applications. Microbial mutualism is a widespread phenomenon and is important in maintaining natural microbial communities. Although cooperative interactions are prevalent in nature, little is known about the processes that allow their initial establishment, govern population dynamics and affect evolutionary processes. To investigate cooperative interactions between bacteria, we constructed, characterized, and adaptively evolved a synthetic community comprised of leucine and lysine Escherichia coli auxotrophs. The co-culture can grow in glucose minimal medium only if the two auxotrophs exchange essential metabolites — lysine and leucine (or its precursors). Our experiments showed that a viable co-culture using these two auxotrophs could be established and adaptively evolved to increase growth rates (by ~3 fold) and optical densities. While independently evolved co-cultures achieved similar improvements in growth, they took different evolutionary trajectories leading to different community compositions. Experiments with individual isolates from these evolved co-cultures showed that changes in both the leucine and lysine auxotrophs improved growth of the co-culture. Interestingly, while evolved isolates increased growth of co-cultures, they exhibited decreased growth in mono-culture (in the presence of leucine or lysine). A genome-scale metabolic model of the co-culture was also constructed and used to investigate the effects of amino acid (leucine or lysine) release and uptake rates on growth and composition of the co-culture. When the metabolic model was constrained by the estimated leucine and lysine release rates, the model predictions agreed well with experimental growth rates and composition measurements. While this study and others have focused on cooperative interactions amongst community members, the adaptive evolution of communities with other types of interactions (e.g., commensalism, ammensalism or parasitism) would also be of interest.

59 BASIC BIOLOGICAL SCIENCES↗

Application of boundary layer control to HSCT low speed configuration

The feasibility of using boundary layer control (BLC) on a high speed civil transport (HSCT) high lift configuration for low speed performance improvement is studied. The possibility of using a part of the suction system previously designed for cruise laminar flow control (LFC) for low speed BLC mode was explored. A suction scheme for BLC was devised for the HSCT high lift configuration. The extent of the suction regions and levels were determined via panel method based inviscid flow analysis coupled with three-dimensional boundary layer analyses. The compatibility of the BLC suction requirements was assessed, and the modifications necessary to operate the system in both modes of operation, were determined. An assessment of the potential aerodynamic performance improvement resulting from an implementation of the BLC concept compared to an optimized simple flap configuration without BLC was made by an Euler code analysis of the simple flap configuration.

Parikh, P. G.↗

Eu(III) Complexes of Octadentate 1-Hydroxy-2-pyridinones: Stability and Improved Photophysical Performance

The luminescence properties of lanthanoid ions can be dramatically enhanced by coupling them to antenna ligands that absorb light in the UV-visible and then efficiently transfer the energy to the lanthanoid centre. The synthesis and the complexation of Ln{sup III} cations (Ln = Eu, Gd) for a ligand based on four 1-hydroxy-2-pyridinone (1,2-HOPO) chelators appended to a ligand backbone derived by linking two L-lysine units (3LI-bis-LYS) is described. This octadentate Eu{sup III} complex ([Eu(3LI-bis-LYS-1,2-HOPO)]{sup -}) has been evaluated in terms of its thermodynamic stability, UV-visible absorption and luminescence properties. For this complex, the conditional stability constant (pM) is 19.9, which is an order of magnitude higher than diethylenetriaminepentacetic acid at pH = 7.4. This Eu{sup III} complex also shows an almost two-fold increase in its luminescence quantum yield in aqueous solution (pH = 7.4) when compared with other octadentate ligands. Hence, despite a slight decrease of the molar absorption coefficient, a much higher brightness is obtained for [Eu(3LI-bis-LYS-1,2-HOPO)]{sup -}. This overall improvement was achieved by saturating the coordination sphere of the Eu{sup III} cation, yielding an increased metal-centred efficiency by excluding solvent water molecules from the metal's inner sphere.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Improving the performance of stainless-steel DC high voltage photoelectron gun cathode electrodes via gas conditioning with helium or krypton

Gas conditioning was shown to eliminate field emission from cathode electrodes used inside DC high voltage photoelectron guns, thus providing a reliable means to operate photoguns at higher voltages and field strengths. Measurements and simulation results indicate that gas conditioning eliminates field emission from cathode electrodes via two mechanisms: sputtering and implantation, with the benefits of implantation reversed by heating the electrode. We have studied five stainless steel electrodes (304L and 316LN) that were polished to approximately 20 nm surface roughness using diamond grit, and evaluated inside a high voltage apparatus to determine the onset of field emission as a function of voltage and field strength. The field emission characteristics of each electrode varied significantly upon the initial application of voltage but improved to nearly the same level after gas conditioning using either helium or krypton, exhibiting less than 10 pA field emission at -225 kV bias voltage with a 50 mm cathode/anode gap, corresponding to a field strength of ~13 MV/m. Finally, field emission could be reduced with either gas, but there were conditions related to gas choice, voltage and field strength that were more favorable than others.

BastaniNejad, M.↗

Representing DC-Coupled PV+Battery Hybrids in a Capacity Expansion Model

Recent technology cost and performance improvements and the federal investment tax credit (ITC) have driven growing interest in coupling solar photovoltaic (PV) and battery systems. Combining these technologies into co-located or hybridized PV+battery systems has the potential to lower costs and increase energy output relative to multiple independent systems. In this work, we provide an overview of PV+battery systems and demonstrate methods for incorporating them into NREL’s Regional Energy Deployment System (ReEDS) capacity expansion model. Although the methods are applied to a specific model, we anticipate that the approaches used here can be useful for informing PV+battery method development for other capacity expansion models. The implemented method relies heavily on the native representations of PV and battery technologies; therefore, the focus of this work is on capturing and parameterizing the interactions between them for a configuration in which the PV and battery technologies share a single bi-directional inverter. This work also demonstrates the impacts of including PV+battery systems in the ReEDS optimization for the conterminous United States through 2050. In particular, we perform parametric sensitivities for input assumptions that are uncertain and expected to influence PV+battery deployment levels, including (a) the cost of PV+battery systems relative to independent PV and battery systems, (b) the battery component’s qualification for the ITC, and (c) future cost trajectories for PV and battery systems. We find that PV+battery deployment could occur throughout the conterminous United States if there are cost savings associated with DC coupling PV and battery technologies. If even modest (5%) cost savings can be achieved (through a shared inverter and balance-of-system costs), then approximately one-third of utility-scale PV deployment through 2050 adopts the DC-coupled hybrid configuration, resulting in total PV+battery deployment that exceeds the magnitude of PV+battery projects in U.S. interconnection queues in 2020. If greater cost savings can be achieved through DC coupling (e.g., due to a growing amount of shared balance-of-system costs, reduced financial risk, or modularity) or more rapid cost and performance improvements are realized for PV and battery technologies, then total PV+battery deployment and the share of PV and battery deployment that adopts the hybrid configuration grows (to >50%). In all cases, growing PV+battery deployment primarily displaces independent PV and battery technologies, indicating the strong competition between the hybrid and independent configurations comprising technologies with similar performance characteristics.

14 SOLAR ENERGY↗