Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Transformers”

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 325 records · Page 18
All sourcesNASA NTRSDOE OSTI
DOE OSTI2015

Challenges at the Frontiers of Matter and Energy: Transformative Opportunities for Discovery Science

FIVE TRANSFORMATIVE OPPORTUNITIES FOR DISCOVERY SCIENCE As a result of this effort, it has become clear that the progress made to date on the five Grand Challenges has created a springboard for seizing five new Transformative Opportunities that have the potential to further transform key technologies involving matter and energy. These five new Transformative Opportunities and the evidence supporting them are discussed in this new report, “Challenges at the Frontiers of Matter and Energy: Transformative Opportunities for Discovery Science.” Mastering Hierarchical Architectures and Beyond-Equilibrium Matter Complex materials and chemical processes transmute matter and energy, for example from CO2 and water to chemical fuel in photosynthesis, from visible light to electricity in solar cells and from electricity to light in light emitting diodes (LEDs) Such functionality requires complex assemblies of heterogeneous materials in hierarchical architectures that display time-dependent away-from-equilibrium behaviors. Much of the foundation of our understanding of such transformations however, is based on monolithic single- phase materials operating at or near thermodynamic equilibrium. The emergent functionalities enabling next-generation disruptive energy technologies require mastering the design, synthesis, and control of complex hierarchical materials employing dynamic far-from-equilibrium behavior. A key guide in this pursuit is nature, for biological systems prove the power of hierarchical assembly and far- from-equilibrium behavior. The challenges here are many: a description of the functionality of hierarchical assemblies in terms of their constituent parts, a blueprint of atomic and molecular positions for each constituent part, and a synthesis strategy for (a) placing the atoms and molecules in the proper positions for the component parts and (b) arranging the component parts into the required hierarchical structure. Targeted functionality will open the door to significant advances in the harvesting, transforming (e.g., reducing CO2, splitting water, and fixing nitrogen), storing, and use of energy to create new materials, manufacturing processes, and technologies—the lifeblood of human societies and economic growth. Beyond Ideal Materials and Systems: Understanding the Critical Roles of Heterogeneity, Interfaces, and Disorder Real materials, both natural ones and those we engineer, are usually a complex mixture of compositional and structural heterogeneities, interfaces, and disorder across all spatial and temporal scales. It is the fluctuations and disorderly states of these heterogeneities and interfaces that often determine the system’s properties and functionality. Much of our fundamental scientific knowledge is based on “ideal” systems, meaning materials that are observed in “frozen” states or represented by spatially or temporally averaged states. Too often, this approach has yielded overly simplistic models that hide important nuances and do not capture the complex behaviors of materials under realistic conditions. These behaviors drive vital chemical transformations such as catalysis, which initiates most industrial manufacturing processes, and friction and corrosion, the parasitic effects of which cost the U.S. economy billions of dollars annually. Expanding our scientific knowledge from the relative simplicity of ideal, perfectly ordered, or structurally averaged materials to the true complexity of real-world heterogeneities, interfaces, and disorder should enable us to realize enormous benefits in the materials and chemical sciences, which translates to the energy sciences, including solar and nuclear power, hydraulic fracturing, power conversion, airframes, and batteries. Harnessing Coherence in Light and Matter Quantum coherence in light and matter is a measure of the extent to which a wave field vibrates in unison with itself at neighboring points in space and time. Although this phenomenon is expressed at the atomic and electronic scales, it can dominate the macroscopic properties of materials and chemical reactions such as superconductivity and efficient photosynthesis. In recent years, enormous progress has been made in recognizing, manipulating, and exploiting quantum coherence. This progress has already elucidated the role that symmetry plays in protecting coherence in key materials, taught us how to use light to manipulate atoms and molecules, and provided us with increasingly sophisticated techniques for controlling and probing the charges and spins of quantum coherent systems. With the arrival of new sources of coherent light and electron beams, thanks in large part to investments by the U.S. Department of Energy’s Office of Basic Energy Sciences (BES), there is now an opportunity to engineer coherence in heterostructures that incorporate multiple types of materials and to control complex, multistep chemical transformations. This approach will pave the way for quantum information processing and next-generation photovoltaic cells and sensors. Revolutionary Advances in Models, Mathematics, Algorithms, Data, and Computing Science today is benefiting from a convergence of theoretical, mathematical, computational, and experimental capabilities that put us on the brink of greatly accelerating our ability to predict, synthesize, and control new materials and chemical processes, and to understand the complexities of matter across a range of scales. Imagine being able to chart a path through a vast sea of possible new materials to find a select few with desired properties. Instead of the time-honored forward approach, in which materials with desired properties are found through either trial-and-error experiments or lucky accidents, we have the opportunity to inversely design and create new materials that possess the properties we desire. The traditional approach has allowed us to make only a tiny fraction of all the materials that are theoretically possible. The inverse design approach, through the harmonious convergence of theoretical, mathematical, computational, and experimental capabilities, could usher in a virtual cornucopia of new materials with functionalities far beyond what nature can provide. Similarly, enhanced mathematical and computational capabilities significantly enhance our ability to extract physical and chemical insights from vastly larger data streams gathered during multimodal and multidimensional experiments using advanced characterization facilities. Exploiting Transformative Advances in Imaging Capabilities across Multiple Scales Historically, improvements in imaging capabilities have always resulted in improved understanding of scientific phenomena. A prime challenge today is finding ways to reconstruct raw data, obtained by probing and mapping matter across multiple scales, into analyzable images. BES investments in new and improved imaging facilities, most notably synchrotron x-ray sources, free-electron lasers, electron microscopes, and neutron sources, have greatly advanced our powers of observation, as have substantial improvements in laboratory- scale technologies. Furthermore, BES is now planning or actively discussing exciting new capabilities. Taken together, these advances in imaging capabilities provide an opportunity to expand our ability to observe and study matter from the 3D spatial perspectives of today to true “4D” spatially and temporally resolved maps of dynamics that allow quantitative predictions of time-dependent material properties and chemical processes. The knowledge gained will impact data storage, catalyst design, drug delivery, structural materials, and medical implants, to name just a few key technologies. ENABLING SUCCESS Seizing each of these five Transformative Opportunities, as well as accelerating further progress on Grand Challenge research, will require specific, targeted investments from BES in the areas of synthesis, meaning the ability to make the materials and architectures that are envisioned; instrumentation and tools, a category that includes theory and computation; and human capital, the most important asset for advancing the Grand Challenges and Transformative Opportunities. While “Challenges at the Frontiers of Matter and Energy: Transformative Opportunities for Discovery Science” could be viewed as a sequel to the original Grand Challenges report, it breaks much new ground in its assessment of the scientific landscape today versus the scientific landscape just a few years ago. In the original Grand Challenges report, it was noted that if the five Grand Challenges were met, our ability to direct matter and energy would be measured only by the limits of human imagination. This new report shows that, prodded by those challenges, the scientific community is positioned today to seize new opportunities whose impacts promise to be transformative for science and society, as well as dramatically accelerate progress in the pursuit of the original Grand Challenges.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
NASA NTRS2013

Stacked Transformer for Driver Gain and Receive Signal Splitting

In a high-speed signal transmission system that uses transformer coupling, there is a need to provide increased transmitted signal strength without adding active components. This invention uses additional transformers to achieve the needed gain. The prior art uses stronger drivers (which require an IC redesign and a higher power supply voltage), or the addition of another active component (which can decrease reliability, increase power consumption, reduce the beneficial effect of serializer/deserializer preemphasis or deemphasis, and/or interfere with fault containment mechanisms), or uses a different transformer winding ratio (which requires redesign of the transformer and may not be feasible with high-speed signals that require a 1:1 winding ratio). This invention achieves the required gain by connecting the secondaries of multiple transformers in series. The primaries of these transformers are currently either connected in parallel or are connected to multiple drivers. There is also a need to split a receive signal to multiple destinations with minimal signal loss. Additional transformers can achieve the split. The prior art uses impedance-matching series resistors that cause a loss of signal. Instead of causing a loss, most instantiations of this invention would actually provide gain. Multiple transformers are used instead of multiple windings on a single transformer because multiple windings on the same transformer would require a redesign of the transformer, and may not be feasible with high-speed transformers that usually require a bifilar winding with a 1:1 ratio. This invention creates the split by connecting the primaries of multiple transformers in series. The secondary of each transformer is connected to one of the intended destinations without the use of impedance-matching series resistors.

Driscoll, Kevin R.↗
DOE OSTI2018

Trivalent manganese on vacancies triggers rapid transformation of layered to tunneled manganese oxides (TMOs): Implications for occurrence of TMOs in low-temperature environment

Tunneled Mn oxides (TMOs) are common minerals in natural environment, particularly in ferromanganese nodules of oceanic and lake sediments. Their structures host a considerable amount of transition and rare earth metals, thus mediating metal cycling and bearing potential economic interest for exploiting these metals. TMOs form through topotactic transformation of layered Mn oxides (LMOs), such as vernadite, in natural environment. Trivalent Mn (Mn(III)) in the LMO structure is a critical player in the transformation, and the transformation is believed to be extremely slow at room temperature. However, the specific role of Mn(III) and its impacts on the transformation kinetics remains unknown. In the present study, we show that the formation of Mn(III) on vacancies of an LMO is the initial transformation step leading to TMOs, and that the transformation can be rapid at room temperature and circumneutral pH. Specifically, after pre–adsorbed with Mn(II) on vacancies at pH 4, δ–MnO 2 , a hexagonal birnessite analogous to vernadite, starts to transform to a 4 × 4 TMO at 1 h upon incubation at pH 7 and 21 °C under anoxic conditions. The rapid transformation is triggered by the comproportionation reaction between the vacancy-adsorbed Mn(II) and Mn(IV) in δ-MnO 2 that produces Mn(III) on the vacancies. Such intermediate Mn(III)-rich product acts as a precursor for subsequent rapid structural rearrangement to form tunnels. An incubation at lower or higher pH retards the transformation due to an insufficient amount of Mn(III) (pH 6) or the formation of triclinic birnessite (pH 8) as an intermediate product. The presence of O 2 favors the formation of triclinic birnessite at pH 8 and thus retards the transformation whereas O 2 enhances production of Mn(III)–rich hexagonal birnessite at pH 6 and 7 and promotes the transformation. We propose a novel transformation mechanism of LMOs to TMOs, highlighting the role of vacancy–adsorbed Mn(III) in the transformation. Furthermore, this work changes our understanding of TMO formation kinetics and suggests TMOs can readily form in low-temperature redox-fluctuating environment, such as lake and oceanic sediments where Mn(II) often coexists with LMOs.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2016

Detecting the Extent of Eutectoid Transformation in U-10Mo

During eutectoid transformation of U-10Mo alloy, uniform metastable γ UMo phase is expected to transform to a mixture of α-U and γ’-U 2 Mo phase. The presence of transformation products in final U-10Mo fuel, especially the α phase is considered detrimental for fuel irradiation performance, so it is critical to accurately evaluate the extent of transformation in the final U-10Mo alloy. This phase transformation can cause a volume change that induces a density change in final alloy. To understand this density and volume change, we developed a theoretical model to calculate the volume expansion and resultant density change of U-10Mo alloy as a function of the extent of eutectoid transformation. Based on the theoretically calculated density change for 0 to 100% transformation, we conclude that an experimental density measurement system will be challenging to employ to reliably detect and quantify the extent of transformation. Subsequently, to assess the ability of various methods to detect the transformation in U-10Mo, we annealed U-10Mo alloy samples at 500°C for various times to achieve in low, medium, and high extent of transformation. After the heat treatment at 500°C, the samples were metallographically polished and subjected to optical microscopy and x-ray diffraction (XRD) methods. Based on our assessment, optical microscopy and image processing can be used to determine the transformed area fraction, which can then be correlated with the α phase volume fraction measured by XRD analysis. XRD analysis of U-10Mo aged at 500°C detected only α phase and no γ’ was detected. To further validate the XRD results, atom probe tomography (APT) was used to understand the composition of transformed regions in U-10Mo alloys aged at 500°C for 10 hours. Based on the APT results, the lamellar transformation product was found to comprise α phase with close to 0 at% Mo and γ phase with 28–32 at% Mo, and the Mo concentration was highest at the α/γ interface.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2022

Organic matter transformations are disconnected between surface water and the hyporheic zone

Abstract. Biochemical transformations of organic matter (OM) are a primary driver of river corridor biogeochemistry, thereby modulating ecosystem processes at local to global scales. OM transformations are driven by diverse biotic and abiotic processes, but we lack knowledge of how the diversity of those processes varies across river corridors and across surface and subsurface components of river corridors. To fill this gap we quantified the number of putative biotic and abiotic transformations of organic molecules across diverse river corridors using ultra-high-resolution mass spectrometry. The number of unique transformations is used here as a proxy for the diversity of biochemical processes underlying observed profiles of organic molecules. For this, we use public data spanning the contiguous United States (ConUS) from the Worldwide Hydrobiogeochemical Observation Network for Dynamic River Systems (WHONDRS) consortium. Our results show that surface water OM had more biotic and abiotic transformations than OM from shallow hyporheic zone sediments (1–3 cm depth). We observed substantially more biotic than abiotic transformations, and the numbers of biotic and abiotic transformations were highly correlated with each other. We found no relationship between the number of transformations in surface water and sediments and no meaningful relationships with latitude, longitude, or climate. We also found that the composition of transformations in sediments was not linked with transformation composition in adjacent surface waters. We infer that OM transformations represented in surface water are an integrated signal of diverse processes occurring throughout the upstream catchment. In contrast, OM transformations in sediments likely reflect a narrower range of processes within the sampled volume. This indicates decoupling between the processes influencing surface water and sediment OM, despite the potential for hydrologic exchange to homogenize OM. We infer that the processes influencing OM transformations and the scales at which they operate diverge between surface water and sediments.

58 GEOSCIENCES↗
NASA NTRS1993

High temperature co-axial winding transformers

The analysis and design of co-axial winding transformers is presented. The design equations are derived and the different design approaches are discussed. One of the most important features of co-axial winding transformers is the fact that the leakage inductance is well controlled and can be made low. This is not the case in conventional winding transformers. In addition, the power density of co-axial winding transformers is higher than conventional ones. Hence, using co-axial winding transformers in a certain converter topology improves the power density of the converter. The design methodology used in meeting the proposed specifications of the co-axial winding transformer specifications are presented and discussed. The final transformer design was constructed in the lab. Co-axial winding transformers proved to be a good choice for high power density and high frequency applications. They have a more predictable performance compared with conventional transformers. In addition, the leakage inductance of the transformer can be controlled easily to suit a specific application. For space applications, one major concern is the extraction of heat from power apparatus to prevent excessive heating and hence damaging of these units. Because of the vacuum environment, the only way to extract heat is by using a cold plate. One advantage of co-axial winding transformers is that the surface area available to extract heat from is very large compared to conventional transformers. This stems from the unique structure of the co-axial transformer where the whole core surface area is exposed and can be utilized for cooling effectively. This is a crucial issue here since most of the losses are core losses.

Divan, Deepakraj M.↗
DOE OSTI2020

Broadband Characterization and Circuit Model Development of Transmission-Scale Transformers

This report describes broadband measurements of transmission-scale transformers typical in the electric power grid. This work was performed as part of the EMP Resilient Grid LDRD project at Sandia National Laboratories to generate circuit models that can be used for high-altitude electromagnetic pulse (HEMP) coupling simulations and response predictions. The objective of the work was to obtain characterization data of substation yard equipment across a frequency range relevant to HEMP. Vector network analyzer measurements up to 100 MHz were performed on two power transformers at ABB-Hitachi and a single ITEC potential transformer. Custom cable breakouts were designed to interface with the transformer terminals and provide ground connections to the chassis at the base of the transformer bushings. The three-phase terminals of the power transformers were measured as a common mode impedance using a parallel resistive splitter, and the single-phase terminals of the potential transformer were measured directly. A vector fitting algorithm was used to empirically fit circuit models to the resulting two-port networks and input impedances of the measured objects. Simplified circuit representations of the input impedances were also generated to assess the degree of precision needed for high-altitude electromagnetic pulse response predictions, which were performed in Sandia's XYCE circuit simulator platform. HEMP coupling simulations using the transformer models showed significant reduction in the voltage peak and broadening in the pulse width seen at the power transformer compared to the traveling wave voltage. This indicated the importance of the load condition when defining the coupled insult in an electric power substation. Simplified circuit models showed a similar voltage at the transformer with a smoothed waveform. The presence of potential transformers in the simulation did not significantly change the simulated voltage at the power transformer. Single-port input impedance models were also developed to define load conditions when transfer characteristics were not necessary.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRS1994

Effect of radiance-to-reflectance transformation and atmosphere removal on maximum likelihood classification accuracy of high-dimensional remote sensing data

Many analysis algorithms for high-dimensional remote sensing data require that the remotely sensed radiance spectra be transformed to approximate reflectance to allow comparison with a library of laboratory reflectance spectra. In maximum likelihood classification, however, the remotely sensed spectra are compared to training samples, thus a transformation to reflectance may or may not be helpful. The effect of several radiance-to-reflectance transformations on maximum likelihood classification accuracy is investigated in this paper. We show that the empirical line approach, LOWTRAN7, flat-field correction, single spectrum method, and internal average reflectance are all non-singular affine transformations, and that non-singular affine transformations have no effect on discriminant analysis feature extraction and maximum likelihood classification accuracy. (An affine transformation is a linear transformation with an optional offset.) Since the Atmosphere Removal Program (ATREM) and the log residue method are not affine transformations, experiments with Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) data were conducted to determine the effect of these transformations on maximum likelihood classification accuracy. The average classification accuracy of the data transformed by ATREM and the log residue method was slightly less than the accuracy of the original radiance data. Since the radiance-to-reflectance transformations allow direct comparison of remotely sensed spectra with laboratory reflectance spectra, they can be quite useful in labeling the training samples required by maximum likelihood classification, but these transformations have only a slight effect or no effect at all on discriminant analysis and maximum likelihood classification accuracy.

Hoffbeck, Joseph P.↗
DOE OSTI2025

Mn(II)-induced phase transformation of Mn(IV) oxide in seawater

Manganese (Mn) oxides are key components of oceanic and lacustrine Mn nodules and influence metal cycling through oxidation and adsorption processes. Layered Mn oxides (LMOs) are the most common minerals in these nodules and the immediate products of microbially mediated Mn(II) oxidation by O 2 . LMOs can transform into tunneled Mn oxides (TMOs), Mn oxyhydroxides (MnOOH), or Mn(II,III) phase (Mn 3 O 4 ). LMOs often concur with Mn(II) in the environment and the adsorption and oxidation of Mn(II) by LMOs can greatly promote the transformation of LMOs to those phases. However, the Mn(II)-promoted transformation of LMOs in seawater—rich in various cations (300 mM Na + , 10 mM K + , 50 mM Ca 2+ , and 10 mM Mg 2+ ) remains poorly understood. We examined the transformation of δ-MnO 2 in artificial seawater (pH 8.2) under anoxic conditions with the Mn(II)/MnO 2 ratio (r) ranging from 0.08 to 3.83. To assess the effect of ionic strength (IS), parallel experiments were conducted in a mixed 530 mM NaCl and 10 mM KCl solution (having seawater ionic strength but without Ca 2+ and Mg 2+ ) and in 100 mM NaCl solution as a control. At low r (0.08), δ-MnO 2 transformed into triclinic birnessite and a 4 × 4 TMO in 100 mM NaCl solution, which, however, was suppressed in seawater due to strong interactions of Ca 2+ /Mg 2+ with δ-MnO 2 . In the mixed 530 mM NaCl and 10 mM KCl solution (the same ionic strength as of seawater), the transformation occurred extensively but the products had lower crystallinity compared to in 100 mM NaCl solution. At the high Mn(II)/MnO 2 ratios (0.5 ≤ r ≤ 3.83), δ-MnO 2 transformed extensively into MnOOH phases and hausmannite (Mn 3 O 4 ) in 100 mM NaCl solution. The seawater suppressed the transformation, but the suppression became weaker with increasing Mn(II)/MnO 2 ratio. For example, the transformation was completely suppressed at r = 0.5 but essentially negligible at r = 3.83. The suppression at these high Mn(II)/MnO 2 ratios was mainly ascribed to the influence of Ca 2+ and Mg 2+ rather than of the high IS, and the weaker suppression at the higher Mn(II)/MnO 2 ratio suggests stronger competition of Mn(II) with Ca 2+ /Mg 2+ for interacting with δ-MnO 2 . Moreover, the composition and crystallinity of the transformation products (i.e., the relative abundance of MnOOH (α, β, and γ) and Mn 3 O 4 ) were influenced by both the high ionic strength and the presence of Ca 2+ and Mg 2+ . Therefore, even though Ca 2+ and Mg 2+ concentrations are much lower than Na + in seawater, their impacts on the transformation are dominant. Our study explains why MnOOH phases, hausmannite, and TMOs are less common than LMOs in oceanic environments, partially because seawater chemistry suppresses their formation. LMOs are the most reactive for metal adsorption and oxidation among all Mn oxides. Thus, the high stability of LMOs in an oceanic environment confers the high impacts of Mn oxides on metal cycling in the ocean.

Divalent manganese↗
DOE OSTI2019

Sawtooth oscillation behavior with varying amounts of applied stellarator rotational transform.

In this study, tokamak-like sawtooth oscillations are observed in the Compact Toroidal Hybrid (CTH), a current-carrying stellarator. CTH has the unique ability to change the amount of the applied vacuum rotational transform from external stellarator coils relative to the rotational transform generated by the internal plasma current to investigate the effects of strong three-dimensional magnetic shaping on sawtooth behavior. The observed sawteeth in CTH, for plasmas with monotonically decreasing rotational transform profiles dominated by the plasma current, have characteristics of those observed on tokamaks including (1) a central emissivity rise and then a sudden crash with a well-defined inversion radius, (2) the presence of an m = 1 emissivity fluctuation, and (3) the normalized inversion surface radius scales with the total edge rotational transform. We explore the properties of an ensemble of discharges in CTH in which the fractional rotational transform, defined as the vacuum rotational transform divided by the total rotational transform, is systematically varied from 0.04 to 0.42 to observe changes in sawtooth oscillation dynamics. Over this range of the fractional rotational transform, the measured sawtooth period decreased by a factor of two. At a high fractional rotational transform, the sawtooth amplitude is observed to consist of only low-amplitude oscillations while the measured crash time of the sawtooth oscillation does not appear to have a strong dependence on the amount of the fractional transform applied. Lastly, experimental results indicate that the low-amplitude sawteeth are accompanied by a decrease in the sawtooth period and predominantly correlated with the mean elongation (due to the increasing fractional rotational transform) of the non-axisymmetric plasmas within CTH rather than other global equilibrium parameters.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS1974

An extension of the Laplace transform to Schwartz distributions

A characterization of the Laplace transform is developed which extends the transform to the Schwartz distributions. The class of distributions includes the impulse functions and other singular functions which occur as solutions to ordinary and partial differential equations. The standard theorems on analyticity, uniqueness, and invertibility of the transform are proved by using the characterization as the definition of the Laplace transform. The definition uses sequences of linear transformations on the space of distributions which extends the Laplace transform to another class of generalized functions, the Mikusinski operators. It is shown that the sequential definition of the transform is equivalent to Schwartz' extension of the ordinary Laplace transform to distributions but, in contrast to Schwartz' definition, does not use the distributional Fourier transform. Several theorems concerning the particular linear transformations used to define the Laplace transforms are proved. All the results proved in one dimension are extended to the n-dimensional case, but proofs are presented only for those situations that require methods different from their one-dimensional analogs.

Price, D. R.↗
DOE OSTI2023

A survey of techniques for optimizing transformer inference

Recent years have seen a phenomenal rise in the performance and applications of transformer neural networks. The family of transformer networks, including Bidirectional Encoder Representations from Transformer (BERT), Generative Pretrained Transformer (GPT) and Vision Transformer (ViT), have shown their effectiveness across Natural Language Processing (NLP) and Computer Vision (CV) domains. Transformer-based networks such as ChatGPT have impacted the lives of common men. However, the quest for high predictive performance has led to an exponential increase in transformers' memory and compute footprint. Researchers have proposed techniques to optimize transformer inference at all levels of abstraction. Further, this paper presents a comprehensive survey of techniques for optimizing the inference phase of transformer networks. We survey techniques such as knowledge distillation, pruning, quantization, neural architecture search and lightweight network design at the algorithmic level. We further review hardware-level optimization techniques and the design of novel hardware accelerators for transformers. We summarize the quantitative results on the number of parameters/FLOPs and the accuracy of several models/techniques to showcase the tradeoff exercised by them. We also outline future directions in this rapidly evolving field of research. We believe that this survey will educate both novice and seasoned researchers and also spark a plethora of research efforts in this field.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

In situ investigation of stress-induced martensitic transformation in granular shape memory ceramic packings

Stress-induced martensitic transformation can occur in granular shape memory materials when individual particles experience high stresses and transform from a high-symmetry austenite phase to a low-symmetry martensite phase. This involves a highly heterogeneous distribution of the driving force and very low mechanical constraint for martensite nucleation, so the transformation behavior can be dramatically different from the well-documented case of monolithic solids. In this work, we investigate the stress-induced martensitic transformation in granular shape memory ceramic packings, which consist of single-crystal micro-particles of ZrO 2 -12 at%CeO 2 and ZrO 2 -15 at%CeO 2 . Using in situ neutron diffraction, we study how the phase fraction, lattice strain, and integral peak broadness evolve during external loading, unloading, and subsequent heating. Several peculiar features are discovered, including (i) a continuous mode of transformation with a wide range of transformation loads, (ii) co-evolution of the packing structure, contact deformation, and martensitic transformation, and (iii) a strong correlation between the peak broadening and the transformed phase fraction. In addition, we show the first direct evidence of reversible stress-induced martensitic transformation in granular materials. We additionally discuss the mechanism for martensite nucleation and growth in granular packings and show how that leads to the observed transformation characteristics.

36 MATERIALS SCIENCE↗
DOE OSTI2014

Microstructure of warm rolling and pearlitic transformation of ultrafine-grained GCr15 steel

Pearlitic transformation mechanisms have been investigated in ultra-fine grained GCr15 steel. The ultrafine-grained steel, whose grain size was less than 1 μm, was prepared by thermo-mechanical treatment at 873 K and then annealing at 923 K for 2 h. Pearlitic transformation was conducted by reheating the ultra-fine grained samples at 1073 K and 1123 K for different periods of time and then cooling in air. Scanning electron microscope observation shows that normal lamellar pearlite, instead of granular cementite and ferrite, cannot be formed when the grain size is approximately less than 4(± 0.6) μm, which yields a critical grain size for normal lamellar pearlitic transformations in this chromium alloyed steel. The result confirms that grain size has a great influence on pearlitic transformation by increasing the diffusion rate of carbon atoms in the ultra-fine grained steel, and the addition of chromium element doesn't change this pearlitic phase transformation rule. Meanwhile, the grain growth rate is reduced by chromium alloying, which is beneficial to form fine grains during austenitizing, thus it facilitating pearlitic transformation by divorced eutectoid transformation. Moreover, chromium element can form a relatively high gradient in the frontier of the undissolved carbide, which promotes carbide formation in the frontier of the undissolved carbide, i.e., chromium promotes divorced eutectoid transformation. - Highlights: • Ultrafine-grained GCr15 steel was obtained by warm rolling and annealing technology. • Reduction of grain size makes pearlite morphology from lamellar to granular. • Adding Cr does not change normal pearlitic phase transformation rule in UFG steel. • Cr carbide resists grain growth and facilitates pearlitic transformation by DET.

36 MATERIALS SCIENCE↗
DOE OSTI2018

In-situ TEM observation of hcp-Ti to fcc-Ti phase transformation in Nb-Ti-Si based alloys

Highlights: • Hcp-Ti to fcc-Ti phase transformation has been investigated by direct TEM observation. • Two periodic structures nucleate and act as modulated structures during the hcp-Ti to fcc-Ti phase transformation. • The transformation proceeds through formation of growth ledges at interfaces. • Stability of Ti might be decreased in the following order: hcp, periodic structures, fcc. - Abstract: Phase transformation process of interfacial Ti precipitates in Nb-Ti-Si based alloys is systematically investigated by in-situ transmission electron microscopy (TEM) observation. The Ti precipitates exhibit an uncommon face-centered cubic (fcc) structure, which can be identified by analyzing its chemical composition and phase transformation characteristics. Fcc-Ti precipitates precipitate from Nbss and partially transform into hexagonal close-packed Ti (hcp-Ti) lamellae during heat-treatment, which will reversely transform into fcc-Ti lamellae during heating in TEM. Two periodic structures nucleate and act as modulated structures in the process of hcp-Ti to fcc-Ti reverse phase transformation. The transformation proceeds through formation of growth ledges at interfaces. Origins of the two periodic structures and ledge model for the phase transformation have been discussed.

36 MATERIALS SCIENCE↗
DOE OSTI2017

The expression of keratin 6 is regulated by the activation of the ERK1/2 pathway in arsenite transformed human urothelial cells

Urothelial cancers have an environmental etiological component, and previous studies from our laboratory have shown that arsenite (As{sup +3}) can cause the malignant transformation of the immortalized urothelial cells (UROtsa), leading to the expression of keratin 6 (KRT6). The expression of KRT6 in the parent UROtsa cells can be induced by the addition of epidermal growth factor (EGF). Tumors formed by these transformed cells have focal areas of squamous differentiation that express KRT6. The goal of this study was to investigate the mechanism involved in the upregulation of KRT6 in urothelial cancers and to validate that the As{sup +3}-transformed UROtsa cells are a model of urothelial cancer. The results obtained showed that the parent and the As{sup +3}-transformed UROtsa cells express EGFR which is phosphorylated with the addition of epidermal growth factor (EGF) resulting in an increased expression of KRT6. Inhibition of the extracellular-signal regulated kinases (ERK1/2) pathway by the addition of the mitogen-activated protein kinase kinase 1 (MEK1) and MEK2 kinase inhibitor U0126 resulted in a decrease in the phosphorylation of ERK1/2 and a reduced expression of KRT6. Immuno-histochemical analysis of the tumors generated by the As{sup +3}-transformed isolates expressed EGFR and tumors formed by two of the transformed isolates expressed the phosphorylated form of EGFR. These results show that the expression of KRT6 is regulated at least in part by the ERK1/2 pathway and that the As{sup +3}-transformed human urothelial cells have the potential to serve as a valid model to study urothelial carcinomas. - Highlights: • Keratin 6 is induced in As{sup +3}-transformed transformed UROtsa cells by EGF. • The ERK1/2 signaling pathway regulates the expression of KRT6. • As{sup +3}-transformed human urothelial cells can serve as a model to study urothelial carcinomas.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2025

A Novel Gene Stacking Method in Plant Transformation Utilizing Split Selectable Markers

Gene stacking, the process of introducing multiple genes into a single plant to enhance desired traits, is essential for plant genetic improvement through both conventional breeding and genetic transformation. In general, transformation-based gene stacking can be achieved through either co-transformation to simultaneously introduce multiple genes or sequential multi-round transformation. While co-transformation is generally faster and more efficient than sequential multi-round transformation, it often requires two selectable marker genes, which confer resistance to antibiotics, for selecting transgenic events. However, in most cases, there is only one best selectable marker gene for a specific plant species or genotype. Also, it is harder to optimize the concentrations of two antibiotics for co-transformation than using one antibiotic for selecting transgenic events. To overcome this challenge, we recently developed an innovative split selectable marker system for plant co-transformation, allowing the use of one selectable marker gene to select transgenic events. This method involves constructing two binary vectors, each carrying a subset of genes of interest and a partial fragment of the selectable marker gene, which is connected to a partial intein fragment. Following Agrobacterium -mediated co-transformation, plants harboring both binary vectors are selected using a single antibiotic, such as kanamycin. This split-marker system can be used to co-transform multiple genes into both herbaceous and woody plants, accelerating genetic improvement of polygenic traits or integrative improvement of multiple traits to simultaneously increase crop yield and quality.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2019

Embedded Optical Sensing Systems For Distribution Transformer Monitoring

Traditional utility monitoring systems do not provide real-time visibility into the condition of transformers or accurate performance measurements. This has resulted in the use of lagging indicators, such as oil sample analysis, typically requiring expensive field visits, grid outages, and laboratory sample testing. This presentation reviews the use of optical sensing systems for managing transformers and improving grid resiliency and safety. The sensing systems allow direct internal monitoring of critical transformer components. Versatility covering different monitoring parameters of interest in embedded configurations for new transformers and a retrofittable configuration for older transformers is achieved by applying hair-thin multiplexed fiber-optic and other low-profile optical sensors in combination with a high-resolution optical readout and advanced analytics. Under an initiative funded by the DOE Office of Electricity, PARC, GE and Con Edison built, tested, and qualified commercial network transformers with embedded optical sensing systems. The parameters monitored were validated across a range of scenarios with lab instruments typically used for qualification testing. An ongoing field trial to demonstrate the robustness of the technology for remote monitoring through the industrial internet under demanding grid deployment conditions in New York City is showing promise toward scale-up of the technology. This presentation will provide an overview of the typical pain points in managing transformers and parameters for monitoring. It will also cover the requirements for highly constrained installations that need to have a larger capacity and flexible rating but that must fit into a tight space. The unique capabilities of optical sensing systems and a low-cost approach to interpreting the signals from sensors in transformers for predictive maintenance will be covered. The presentation will discuss the compatibility of the sensors in the transformer environment. It will discuss the validation testing per IEEE standards that enabled the solution to be implemented in field-deployable commercial transformers. The initial results from the field deployment and the potential for the technology to provide systematic alerts for unsafe/unexpected events will be discussed. The technology could enable utility professionals to gain information they can use to better plan for maintenance/management, and improve grid performance. This would in turn accommodate DERs and other evolving grid market dynamics.

24 POWER TRANSMISSION AND DISTRIBUTION↗