SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
MemLeap is a GPU Memory Profiler built atop Nvbit (NVidia Binary Instrumentation Tool) framework. It analyzes memory accesses to memory objects in GPU memory, including Unified Memory, to optimize memory usage and reveal performance bottlenecks.
There is growing interest in combining microphysical models and polarimetric radar observations to improve our understanding of storms and precipitation. Mapping model-predicted variables into the radar observational space necessitates a forward operator, which requires assumptions that introduce uncertainties into model–observation comparisons. These include uncertainties arising from the microphysics scheme a priori assumptions of a fixed drop size distribution (DSD) functional form, whereas natural DSDs display far greater variability. To address this concern, this study presents a moment-based polarimetric radar forward operator with no fundamental restrictions on the DSD form by linking radar observables to integrated DSD moments. The forward operator is built upon a dataset of >200 million realistic DSDs from one-dimensional bin microphysical rain-shaft simulations, and surface disdrometer measurements from around the world. This allows for a robust statistical assessment of forward operator uncertainty and quantification of the relationship between polarimetric radar observables and DSD moments. Comparison of “truth” and forward-simulated vertical profiles of the polarimetric radar variables are shown for bin simulations using a variety of moment combinations. Higher-order moments (especially those optimized for use with the polarimetric radar variables: the sixth and ninth) perform better than the lower-order moments (zeroth and third) typically predicted by many bulk microphysics schemes.
This work presents techniques for non-invasive transverse profile measurements of high-intensity proton beams using an Electron Beam Profile Scanner (EBPS). The EBPS utilizes low-energy electrons as a probe to analyze the transverse size of proton beams, allowing for potential analysis on a single-bunch basis. Recent upgrades to the Fermilab Main Injector have enhanced beam power on target to 1 MW, with future developments targeting 2 MW. The higher beam power has increased the demand for non-invasive diagnostics, as invasive methods can disrupt operations.The techniques presented include 1) the slow scan technique, which serves as a proof of concept for the probe beam, 2) the one-shot scan technique for measuring horizontal beam profiles, and 3) the raster scan technique for analyzing horizontal beam profiles as a function of the longitudinal distribution of the beam. The profiles obtained will be crucial for studying and understanding instabilities in high-power, high-intensity proton beams. This will contribute to optimizing the operation of high-power proton accelerators by minimizing beam loss, activation, and damage to both the diagnostics and the accelerator components.
Accurate measurement of electron beam emittance is essential for optimizing high-brightness electron sources. The Pinhole Scan Technique measures the 4D phase space and hence the emittance by measuring the beam profile after clipping the beam using a pinhole followed by a drift section and then scanning the beam over the pinhole. This technique has been implemented in low energy (< 200 keV) beamlines at both Cornell University and Arizona State University. However, the technique poses several practical challenges. In this work, we analyze and address key issues affecting the 4D phase space and emittance measurements using this technique. We identify and investigate sources of inaccuracies like the pinhole aspect ratio, beam divergence, position-momentum correlations in the phase space, and the point-spread-function of the detector and suggest techniques to minimize them. Our findings offer a pathway to more accurate 4D phase space characterization in advanced electron beam systems.
Combined missions to Mars and Phobos/Deimos were considered in relation to large scientific returns at a low total cost. A matrix of all feasible major combined mission concepts was compiled. The most promising approach was selected from which a nominal mission profile was developed. A systems analysis was performed to trade off performance, configuration, and cost characteristics of candidate mission concepts. Configuration and subsystem optimization analysis were then conducted from which a baseline and an alternate program were selected. Program schedules and cost estimates were prepared for the recommended and alternate program.
A compact dual frequency feed horn is designed and optimized to operate at W-band (94.92 GHz) and G-band (238 GHz). The feed horn consists of a single spline profile horn, a diplexer having W- and G-band ports and a band pass filter at the W-band port. The filter provides isolation between the W- and G-band ports. Simulations show a directivity of 21.9 dB and -12 dB beam width of 28 degrees at 94.92 GHz and 26.27 dB and 17 degrees respectively at 238 GHz. The dual frequency horn will be used as part of a W/G-band breadboard atmospheric radar.
We report on a novel combined laser pulse shaping and dynamic wavelength encoding capability based on a simple architecture implementing direct space to time mapping. There are several potential applications that can be enabled by the ability to control the instantaneous intensity or wavelength of an optical waveform on a picosecond-to-nanosecond timescale. To our knowledge, no known methods can access this temporal regime with a practical architecture. Here, we demonstrate an extension of the Space–Time Induced Linearly Encoded Transcription for Temporal Optimization (STILETTO) technique that can generate optical waveforms with a programmable instantaneous wavelength vs. time. We experimentally demonstrate the technique by generating self-gated spectrograms and show that it can encode dynamic wavelength vs time profiles at timescales not achievable by any other known method.
Independent component analysis is an unsupervised machine learning algorithm that separates a set of mixed signals into a set of statistically independent source signals. Applied to high-quality gene expression datasets, independent component analysis effectively reveals both the source signals of the transcriptome as co-regulated gene sets, and the activity levels of the underlying regulators across diverse experimental conditions. Two major variables that affect the final gene sets are the diversity of the expression profiles contained in the underlying data, and the user-defined number of independent components, or dimensionality, to compute. Availability of high-quality transcriptomic datasets has grown exponentially as high-throughput technologies have advanced; however, optimal dimensionality selection remains an open question. We computed independent components across a range of dimensionalities for four gene expression datasets with varying dimensions (both in terms of number of genes and number of samples). We computed the correlation between independent components across different dimensionalities to understand how the overall structure evolves as the number of user-defined components increases. We then measured how well the resulting gene clusters reflected known regulatory mechanisms, and developed a set of metrics to assess the accuracy of the decomposition at a given dimension. We found that over-decomposition results in many independent components dominated by a single gene, whereas under-decomposition results in independent components that poorly capture the known regulatory structure. From these results, we developed a new method, called OptICA, for finding the optimal dimensionality that controls for both over- and under-decomposition. Specifically, OptICA selects the highest dimension that produces a low number of components that are dominated by a single gene. We show that OptICA outperforms two previously proposed methods for selecting the number of independent components across four transcriptomic databases of varying sizes. OptICA avoids both over-decomposition and under-decomposition of transcriptomic datasets resulting in the best representation of the organism’s underlying transcriptional regulatory network.
The large flexibility of the proposed QUASAR facility is leveraged in order to explore the effect of magnetic shear on adiabatic Ion Temperature Gradient (ITG) turbulence. The QUASAR facility is a reimagining of the National Compact Stellarator Experiment utilizing and expanding upon the already constructed coil set. Recent work using fixed boundary optimization of the LI383 equilibrium (upon which QUASAR is based) has suggested possible improvements to ITG turbulence. In this work, a different approach is taken, wherein a series of self-consistent free boundary VMEC equilibria are developed for QUASAR. These equilibria assume temperature and density profiles consistent with 2% beta and ohmic current drive. In each configuration, the toroidal field coils are energized to different values and the STELLOPT code is used to vary the modular coil current and net toroidal current. The edge value of rotational transform is targeted in the optimization, producing a magnetic shear scan. All these configurations share similar neoclassical transport levels, while nonlinear GENE flux tube simulations show up to a factor of four change in adiabatic ITG turbulence at various radii. Comparisons of proxy functions and linear flux tube runs are also made. This work highlights the capability of the QUASAR experiment as a tool to explore transport in 3D magnetic fields and the possibility of the further improvements to stellarators through optimization.
Reaction rates at spatially heterogeneous, unstable interfaces are notoriously difficult to quantify, yet are essential in engineering many chemical systems, such as batteries and electrocatalysts. Experimental characterizations of such materials by operando microscopy produce rich image datasets, but data-driven methods to learn physics from these images are still lacking because of the complex coupling of reaction kinetics, surface chemistry and phase separation. Here we show that heterogeneous reaction kinetics can be learned from in situ scanning transmission X-ray microscopy (STXM) images of carbon-coated lithium iron phosphate (LFP) nanoparticles. Combining a large dataset of STXM images with a thermodynamically consistent electrochemical phase-field model, partial differential equation (PDE)-constrained optimization and uncertainty quantification, we extract the free-energy landscape and reaction kinetics and verify their consistency with theoretical models. We also simultaneously learn the spatial heterogeneity of the reaction rate, which closely matches the carbon-coating thickness profiles obtained through Auger electron microscopy (AEM). Across 180,000 image pixels, the mean discrepancy with the learned model is remarkably small (<7%) and comparable with experimental noise. Our results open the possibility of learning nonequilibrium material properties beyond the reach of traditional experimental methods and offer a new non-destructive technique for characterizing and optimizing heterogeneous reactive surfaces.
Reaction rates at spatially heterogeneous, unstable interfaces are notoriously difficult to quantify, yet are essential in engineering many chemical systems, such as batteries 1 and electrocatalysts 2. Experimental characterizations of such materials by operando microscopy produce rich image datasets 3–6, but data-driven methods to learn physics from these images are still lacking because of the complex coupling of reaction kinetics, surface chemistry and phase separation 7. Here we show that heterogeneous reaction kinetics can be learned from in situ scanning transmission X-ray microscopy (STXM) images of carbon-coated lithium iron phosphate (LFP) nanoparticles. Combining a large dataset of STXM images with a thermodynamically consistent electrochemical phase-field model, partial differential equation (PDE)-constrained optimization and uncertainty quantification, we extract the free-energy landscape and reaction kinetics and verify their consistency with theoretical models. We also simultaneously learn the spatial heterogeneity of the reaction rate, which closely matches the carbon-coating thickness profiles obtained through Auger electron microscopy (AEM). Across 180,000 image pixels, the mean discrepancy with the learned model is remarkably small (<7%) and comparable with experimental noise. Our results open the possibility of learning nonequilibrium material properties beyond the reach of traditional experimental methods and offer a new non-destructive technique for characterizing and optimizing heterogeneous reactive surfaces.
4H-Silicon carbide (4H-SiC) is a wide bandgap semiconductor that offers superior material properties over silicon, including higher critical electric field, thermal conductivity, and electron saturation velocity. These advantages make 4H-SiC highly attractive for high-voltage, high-efficiency power electronics. However, realizing the full potential of SiC requires device technologies that are not only high-performing but also manufacturable and reliable under real-world operating conditions. This report summarizes the outcomes of a five-year R&D effort funded by the U.S. Department of Energy (DOE) under the Electric Drive Technologies Consortium (EDTC), focused on developing cost-competitive, high-performance, and highly reliable (CPR) power devices on 4H-SiC substrates. The program targeted scalable and manufacturable 1.2 kV-class SiC MOSFETs optimized for next-generation electric vehicles, renewable energy systems, and industrial power conversion. The project delivered transformative advancements in SiC power device performance and ruggedness. Particularly, Specific on-resistance (R on,sp ) was reduced by up to 37%, from ~4.0 m$\Omega \cdot$cm 2 in earlier designs to an industry-leading 2.40 m$\Omega \cdot$cm 2 , driven by optimized doping, refined JFET widths, and layout engineering. Breakdown voltages (BV) exceeded 1600 V, marking improvement over legacy baselines, and demonstrating the robustness of newly implemented junction profiles and edge terminations. Short-circuit withstand time (SCWT) saw a remarkable 4$\times$ increase, from ~2 $\mu$s to over 8 $\mu$s, achieved through the successful deployment of deep P-well structures (~1.8–2.0 $\mu$m) via channeling implantation. This innovative process breakthrough enabled precise junction formation without MeV-class implantation tools, reduced leakage under high field stress, and allowed even the shortest-channel devices (down to 0.3 $\mu$m) to achieve both high BV and excellent ruggedness—breaking the traditional trade-off between conduction efficiency and blocking capability. Several novel architectures pushed the performance envelope further. JBSFETs—featuring embedded Schottky portions—eliminated bipolar degradation and drastically reduced third-quadrant leakage, while Ladder MOSFETs introduced a clever orthogonal conduction path that achieved a 15.4% reduction in R on,sp over standard linear designs. Switching performance reached new benchmarks: short-channel devices showed a 31% reduction in total switching energy compared to 0.5 $\mu$m counterparts, while maintaining manageable gate drive requirements. Layout-optimized structures not only improved transconductance but also accelerated switching transitions, pointing to real-world benefits in converter-level efficiency. The devices also passed rigorous reliability validation. Stress-tested across TDDB, HTGB, HTRB, HVP, and burn-in, the devices screened under 30 V/10 hr and 43 V/1 s protocols consistently exhibited tighter lifetime distributions and long-term oxide robustness. These screening techniques proved effective in identifying latent defects and ensuring deployment-grade reliability. Meanwhile, advanced 3D TCAD simulations revealed and resolved electric field hotspots—particularly in HEXFET corners—where fields exceeding 4.8 MV/cm were mitigated through geometry-aware layout corrections. Overall, the results of this project demonstrate a manufacturable and scalable SiC power device platform that addresses key DOE performance targets for efficient, robust, and reliable 1.2kV 4H-SiC Power Devices. The developed technologies represent a meaningful step forward in the commercial readiness of high-voltage SiC solutions and provide a strong foundation for continued advancement in wide bandgap power electronics.
Residential energy consumers are charged based on a utility rate structure, such as net metering or feed-in tariff. To lower consumers' electricity bills, expensive batteries are deployed to reduce the electricity fed from the grid during peak hours. However, strategic photovoltaic (PV) panel placement enables the reduction of operational energy cost while considering the spatial feasibility and efficiency for hosting rooftop PV. In this paper, we present a framework to automatically identify the optimal location of rooftop PV panels on residential buildings. Our framework integrates multiple workflows, including energy and environmental simulation, parametric modeling, and optimization to identify the ideal location of PV panels to balance the demand and supply at a building and community scale. These workflows are linked using the Grasshopper plug-in for Rhinoceros CAD software. The framework includes two different workflows, each satisfying a target for optimal PV placement: (a) maximizing PV panel efficiency, where users aim to maximize energy generation, and (b) minimizing operational energy cost, where "best'' panels are selected considering utility rates for operational energy cost. Our framework is demonstrated in a residential community in Fort Collins, Colorado, to generate the optimal PV placement for each of the two aforementioned targets. Results from the two workflows are compared to illustrate the effect of PV location and orientation on solar energy production efficiency and operational energy cost. The developed workflows are introduced as tools within the Grasshopper plug-in to investigate the solar potential of rooftop PV panels while taking into account factors such as contextual shading, utility rate structures, and buildings' energy demand profiles.
We use the new simulation capabilities of the extended-magnetohydrodynamic (MHD) code, M3D-C1, to investigate the nonlinear MHD properties of a reactor-scale quasisymmetric stellarator equilibrium. Our model captures the self-consistent evolution of the magnetic field, temperature, density, and flow profiles without imposing restrictions on the structure of the first. We include the effects of resistivity using a realistic temperature-dependent Spitzer model, along with a model for heat transport that captures the key physical characteristic, namely, strongly anisotropic diffusion in directions perpendicular and parallel to the magnetic field. We consider a quasi-axisymmetric, finite-pressure equilibrium that was optimized for self-consistent bootstrap current, quasi-symmetry, and energetic particle confinement. Our assessment finds that the equilibrium is highly unstable to interchange-like pressure-driven instabilities near the plasma edge. The initially unstable modes rapidly destabilize other modes in the direction of the N-fold rotational symmetry (toroidal, in this case). For this equilibrium, N = 2, meaning destabilization of a large number of even-numbered toroidal Fourier modes. Thus, field-periodicity is likely to be an important factor in the nonlinear MHD stability characteristics of optimized stellarators.
SF-22-136 Deep learning has been shown as a successful method for various tasks, and its popularity results in numerous open-source deep learning software tools. Deep learning has been applied to a broad spectrum of scientific domains such as cosmology, particle physics, computer vision, fusion, and astrophysics. Scientists have performed a great deal of work to optimize the computational performance of deep learning frameworks. However, the same cannot be said for I/O performance. As deep learning algorithms rely on big-data volume and variety to effectively train neural networks accurately, I/O is a significant bottleneck on large-scale distributed deep learning training. DLIO, is a novel representative benchmark suite built based on the I/O profiling of the selected workloads. DLIO can be utilized to accurately emulate the I/O behavior of modern deep learning applications. Using DLIO, application developers and system software solution architects can identify potential I/O bottlenecks in their applications and guide optimizations to boost the I/O performance leading to lower training times. The storage vendor can also use DLIO as a guide for designing and optimize the storage and filesystem targeting at deep learning application.
In order for microreactors to be economically competitive, operation costs will need to be minimized through some degree of autonomous control. Previous work has demonstrated the effectiveness of reinforcement learning (RL) for load-following control in a drum-controlled microreactor. This study extends that work by exploring the potential of RL to independently control each of the reactor’s drums. We compare a single-agent RL approach with a multi-agent RL (MARL) framework, testing them for generalization across different load-following power profiles and control timescales, and for robustness in cases of randomly disabled control drums. Since the point kinetics simulation environment used in this study cannot resolve spatial effects, we assume that in the absence of spatially localized disturbances, optimal drum movements should be symmetrical. We demonstrate that single-agent RL is able to achieve accurate performance only when symmetric actions are ignored; otherwise, it fails to train a useful controller. Meanwhile, the MARL framework performs symmetric actions by design and trains a robust, accurate agent, as evidenced by mean absolute errors in power matching of 0.41% for the training power profile, 0.68% for a profile with half the drums disabled, and 0.21% for a profile on a realistic load-following time horizon.
Probabilistic CFD design is needed because we are asked to do more with less. To cost effectively accomplish the design task, we need to formally quantify the effect of uncertainties (variables) in the design. Probabilistic design is one effective method to formally quantify the effect of uncertainties. Our objective is to establish a revolutionary new early design process, by developing non-deterministic physics-based probabilistic design tools, which will include all the life cycle processes. Breakthroughs will be sought in speed, accuracy, intelligence, and usability of the system. This paper is concerned with the usefulness of parametric optimization method coupled with a Navier-Stokes analysis code for the aero-thermodynamic design of turbomachinery combustor liner. The interconnection between the CFD code and NESSUS codes facilitated the coupling between the thermal profiles and structural design. We have developed new concepts for reducing the computational cost of unsteady, three-dimensional, compressible aerodynamic analyses for multistage turbomachinery flows. The flow was modeled by the three-dimensional Favre-Reynolds-averaged Navier-Stokes equations using the k-epsilon turbulence closure, which was integrated using an implicit third-order upwind solver. The methodology developed in this paper is expected to lead to the design optimization of turbomachinery blades.
Power grids face challenges in their infrastructure related to the integration of electric vehicles (EV). In particular, EV charging stations may induce instability in key system parameters such as substantial voltage drops, active power losses, and transformer overload due to high demand during charging periods. This article presents a seasonal reconfiguration strategy based on the differential evolution (DE) algorithm, aimed at enhancing system performance under highly variable and stochastic load profiles, particularly those driven by EV charging. The DEA algorithm is hybridized with the find-union (FU) algorithm to efficiently ensure network radiality throughout the optimization process. The proposed methodology is validated on a hybrid distribution system composed of the IEEE 33-bus network, a modified IEEE 13-bus system, and a specific 13-bus microgrid. Results have demonstrated that seasonal reconfiguration significantly reduces active power losses and mitigates transformer loading during critical demand hours, thereby quantifiably increasing the system’s performance. As an integral component of the proposed approach, an analysis of CO2 emissions associated with energy losses is included, allowing a contextualized assessment of the environmental benefits of seasonal reconfiguration in various geographical areas.