Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “gradient sampling”

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 1,297 records · Page 72
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Learning Constrained Parametric Differentiable Predictive Control Policies With Guarantees

We present differentiable predictive control (DPC), a method for offline learning of constrained neural control policies for nonlinear dynamical systems with performance guarantees. We show that the sensitivities of the parametric optimal control problem can be used to obtain direct policy gradients. Specifically, we employ automatic differentiation (AD) to efficiently compute the sensitivities of the model predictive control (MPC) objective function and constraints penalties. To guarantee safety upon deployment, we derive probabilistic guarantees on closed-loop stability and constraint satisfaction based on indicator functions and Hoeffding’s inequality. We empirically demonstrate that the proposed method can learn neural control policies for various parametric optimal control tasks. In particular, we show that the proposed DPC method can stabilize systems with unstable dynamics, track time-varying references, and satisfy nonlinear state and input constraints. Our DPC method has practical time savings compared to alternative approaches for fast and memory-efficient controller design. Specifically, DPC does not depend on a supervisory controller as opposed to approximate MPC based on imitation learning. We demonstrate that, without losing performance, DPC is scalable with greatly reduced demands on memory and computation compared to implicit and explicit MPC while being more sample efficient than model-free reinforcement learning (RL) algorithms.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Stellar Abundance Maps of the Milky Way Disk

To understand the formation of the Milky Way's prominent bar it is important to know whether stars in the bar differ in the chemical element composition of their birth material as compared to disk stars. This requires stellar abundance measurements for large samples across the Milky Way's body. Such samples, e.g., luminous red giant stars observed by the Sloan Digital Sky Survey's APOGEE survey, will inevitably span a range of stellar parameters; as a consequence, both modeling imperfections and stellar evolution may preclude consistent and precise estimates of their chemical composition at a level of purported bar signatures, which has left current analyses of a chemically distinct bar inconclusive. Here, we develop a new self-calibration approach to eliminate both modeling and astrophysical abundance systematics among red giant branch (RGB) stars of different luminosities (and hence surface gravity $\mathrm{log} g$). We apply our method to 48,853 luminous APOGEE Data Release 16 RGB stars to construct spatial abundance maps of 20 chemical elements near the Milky Way's mid-plane, covering galactocentric radii of 0 kpc < R GC < 20 kpc. Our results indicate that there are no abundance variations whose geometry matches that of the bar, and that the mean abundance gradients vary smoothly and monotonically with galactocentric radius. We confirm that the high-α disk is chemically homogeneous, without spatial gradients. Furthermore, we present the most precise [Fe/H] versus R GC gradient to date with a slope of – 0.057 ±0.001 dex kpc –1 out to approximately 15 kpc.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS1994

On the temperature and velocity through the photosphere of a sunspot penumbra

We investigate the structure in depth of a sunspot penumbra by means of the inversion code of the radiative transfer equation proposed by Ruiz Cobo & del Toro Iniesta (1992), applied to a set of filtergrams of a sunspot, scanning the Fe I line at 5576.1 A, with a sampling interval of 30 mA, from -120 to 120 mA from line center (data previously analyzed by Title et al. 1993). The temperature structure of this penumbra is obtained for each of the 801 pixels selected (0.32 sec x 0.32 sec). On the average, the temperatures seem to decrease as we move inward, but the differences are of the order of the rms values (approximately equal 100-200 K) at a given distance to sunspot center. The outer parts of the penumbra have also a bigger curvature in the T versus log tau(sub 5) relation than the inner parts. We realize, however, that these differences might be influenced by possible stray light effects. Compared to the quiet Sun, penumbral temperatures are cooler at deep layers and hotter at high layers. A mean penumbral model atmosphere is presented. The asymmetries observed in the intensity profile (the line is magnetically insensitive) are deduced to be produced by strong gradients of the line-of-sight velocity that sharply vary spatially along slices of almost constant distance to sunspot center. These variations suggest that such gradients are not only needed to explain the broadband circular polarization observed in sunspots (see Sanchez Almeida & Lites 1992) but are a main characteristic of the fine-scale penumbra. The results are compatible with an Evershed flow present everywhere, but its gradient with depth turns out to vary so that the flow seems to be mainly concentrated in some penumbral fibrils when studied through Dopplergrams. Finally, as by-products of this study, we put constraints to the practical usefulness of the Eddington-Barbier relation, and we explain the values of the Fourier Dopplergrams to be carrying information of layers around the centroid of the generalized response function of Dopplergrams to velocity fluctuations.

Del Toro Iniesta, J. C.↗
NASA NTRS2003

Variability in Ocean Color Associated with Phytoplankton and Terrigenous Matter: Time Series Measurements and Algorithm Development at the FRONT Site on the New England Continental Shelf

Fronts in the coastal ocean describe areas of strong horizontal gradients in both physical and biological properties associated with tidal mixing and freshwater estuarine output (e.g. Simpson, 1981 and O Donnell, 1993). Related gradients in optically important constituents mean that fronts can be observed from space as changes in ocean color as well as sea surface temperature (e.g., Dupouy et al., 1986). This research program is designed to determine which processes and optically important constituents must be considered to explain ocean color variations associated with coastal fronts on the New England continental shelf, in particular the National Ocean Partnership Program (NOPP) Front Resolving Observational Network with Telemetry (FRONT) site. This site is located at the mouth of Long Island sound and was selected after the analysis of 12 years of AVHRR data showed the region to be an area of strong frontal activity (Ullman and Cornillon, 1999). FRONT consists of a network of modem nodes that link bottom mounted Acoustic Doppler Current Profilers (ADCPs) and profiling arrays. At the center of the network is the Autonomous Vertically Profiling Plankton Observatory (AVPPO) (Thwaites et al. 1998). The AVPPO consists of buoyant sampling vehicle and a trawl-resistant bottom-mounted enclosure, which holds a winch, the vehicle (when not sampling), batteries, and controller. Three sampling systems are present on the vehicle, a video plankton recorder, a CTD with accessory sensors, and a suite of bio-optical sensors including Satlantic OCI-200 and OCR-200 spectral radiometers and a WetLabs ac-9 dual path absorption and attenuation meter. At preprogrammed times the vehicle is released, floats to the surface, and is then winched back into the enclosure with power and data connection maintained through the winch cable. Communication to shore is possible through a bottom cable and nearby surface telemetry buoy, equipped with a mobile modem, giving the capability for near-real time data transmission and interactive sampling control.

Morrison, John R.↗
NASA NTRS2003

The Galileo Probe: How it Has Changed Our Understanding of Jupiter

The Galileo Mission to Jupiter, which arrived in December of 1995, provided the first study by an orbiter, and the first in-situ sampling via an entry probe, of an outer planet atmosphere. The rationale for an entry probe is that, even from an orbiter, remote sensing of the jovian atmosphere could not adequately retrieve the information desired. This paper provides a current summary of the most significant aspects of the data returned from the Galileo entry probe. As a result of the probe measurements, there has been a reassessment of our understanding of outer planet formation and evolution of the solar system. The primary scientific objective of the Galileo probe was to determine the composition of the jovian atmosphere, which from remote sensing remained either very uncertain, or completely unknown, with respect to several key elements. The probe found that the global He mass fraction is. significantly above the value reported from the Voyager Jupiter flybys but is slightly below the protosolar value, implying that there has been some settling of He to the deep jovian interior. The probe He measurements have also led to a reevaluation of the Voyager He mass fraction for Saturn, which is now determined to be much closer to that of Jupiter. The elements C, N, S, Ar, Kr, Xe were all found to have global abundances approximately 3 times their respective solar abundances. This result has raised a number of fundamental issues with regard to properties of planetesimals and the solar nebula at the time of giant planet formation. Ne, on the other hand, was found to be highly depleted, probably as the result of it being carried along with helium as helium settles towards the deep interior. The global abundance of O was not obtained by the probe because of the influence of local processes at the probe entry site (PES), processes which depleted condensible species, in this case H2O, well below condensation levels. Other condensible species, namely NH3 and H2S, were similarly affected but attained their deep equilibrium mixing ratios before the maximum depth sampled by the probe. Processes that might be capable of producing such effects on the condensibles are still under investigation. Measured isotopic ratios of noble gases and other heavy elements are solar, and (D + (Sup 3)He)/H is the same to within measurement uncertainties as in the local interstellar medium. No thick clouds were detected, and in particular no significant water cloud, but the PES location clearly affected the probe measurements of clouds. In fact, the probe data must be understood in the context of the location of the PES, which was within what is termed a 5 micron hot spot, a local clearing in the clouds that is bright near the 5 microns spectral region. The thermal structure at the PES was determined from approximately 1000 km above the 1 bar pressure level (10(exp -9 bars)) to 132 km 1 bar (22bars). The probe showed the atmosphere to have a generally sub-adiabatic temperature gradient (static stability) of = 0.1 K/km to as deep as the probe made measurements. In the upper atmosphere the probe derived a maximum positive vertical temperature gradient of approximately 5 K/km, and maximum temperature of = 900 K. The energy sources producing the warm upper atmosphere have yet to be completely identified. At first glance, Doppler tracking of the probe indicates that the long observed cloud level zonal winds extend to levels at least as deep as the probe made measurements. Zonal wind increases from = 80 m/s at pressures less than a bar to about 180 m/s near 5 bars, and remains approximately constant with depth thereafter. However, there is a question as to whether the winds measured from probe tracking are representative of the general wind field, or are considerably influenced by localized winds associated with the PES.

Young, Richard E.↗
DOE OSTI2024

Resolving crystallographic geometrically necessary dislocations in three dimensions in a hexagonal close packed titanium alloy

Abstract Geometrically necessary dislocation (GND) content is measured from mm 3 -scaled Ti7Al three-dimensional (3D) microstructural data using a theory extended for hexagonal close packed crystals, which accounts for basal, prismatic and pyramidal ⟨ c + a ⟩ type dislocation content. The Ti7Al samples have been mechanically pre-strained to two different strain levels, and will then be strained along the same axis in uniaxial tension during simulation. Both inter- and intragranular GNDs across the microstructures have been characterized, with a large contribution of pyramidal ⟨ c + a ⟩ GNDs, consistent with the relative slip activity involved in pre-straining. The spatially resolved crystallographic GND distributions within the 3D microstructures are used to instantiate a microstructure model for forward modeling deformation simulations by a dislocation density hardening elasto-viscoplastic fast Fourier transform framework. Coarsening the voxel resolution during the initial microstructure construction procedure is shown to strongly impact both the magnitude and spatial distribution of the GNDs and in turn the forward deformation response of the pre-strained material. This study indicates that the voxel resolution desired when transferring from measured to model microstructures need not only be proportionally scaled with the microstructure but also sufficiently fine to capture the subgranular orientation gradients that may already be present in the material.

Materials Science↗
DOE OSTI2021

Recovery of Forest Structure Following Large-Scale Windthrows in the Northwestern Amazon

The dynamics of forest recovery after windthrows (i.e., broken or uprooted trees by wind) are poorly understood in tropical forests. The Northwestern Amazon (NWA) is characterized by a higher occurrence of windthrows, greater rainfall, and higher annual tree mortality rates (~2%) than the Central Amazon (CA). We combined forest inventory data from three sites in the Iquitos region of Peru, with recovery periods spanning 2, 12, and 22 years following windthrow events. Study sites and sampling areas were selected by assessing the windthrow severity using remote sensing. At each site, we recorded all trees with a diameter at breast height (DBH) ≥ 10 cm along transects, capturing the range of windthrow severity from old-growth to highly disturbed (mortality > 60%) forest. Across all damage classes, tree density and basal area recovered to >90% of the old-growth values after 20 years. Aboveground biomass (AGB) in old-growth forest was 380 (±156) Mg ha -1 . In extremely disturbed areas, AGB was still reduced to 163 (±68) Mg ha -1 after 2 years and 323 (± 139) Mg ha -1 after 12 years. This recovery rate is ~50% faster than that reported for Central Amazon forests. The faster recovery of forest structure in our study region may be a function of its higher productivity and adaptability to more frequent and severe windthrows. These varying rates of recovery highlight the importance of extreme wind and rainfall on shaping gradients of forest structure in the Amazon, and the different vulnerabilities of these forests to natural disturbances whose severity and frequency are being altered by climate change.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

TaV_BCC_SolidSolution_128atoms_VASP6

We performed density functional theory (DFT) calculations for body-centered-cubic (BCC) structures with 128 lattices sites of solid solution binary alloys tantalum-vanadium (Tanu-V). The electronic structures of alloys have been calculated using Vienna Ab initio Simulation Package (VASP). Within this package the DFT approach is used to reduce many-body Schrodinger equation to set of single particle Kohn-Sham (KS) equations. The generalized electronic exchange-correlation functional is described by generalized gradient approximation with the Perdew-Burke-Ernzerhof parametrization. The electron-ion interactions is described by pseudopotentials developed within the plane-wave basis projector augmented-wave (PAW) approach. These pseudopotentials are available at the VASP portal (http://cms.mpi.univie.ac.at/vasp/). Our calculations have been run with the pseudopotentials treating s and p semi-core states as valence in case for the element V. For Ta, p semi-core states as valence were chosen. The electronic densities and potentials are expanded over plane-waves with energy cutoff of 350 eV. 2x2x2 k-mesh and normal precision were used. The alloys were modeled by supercell containing 128 randomly distributed atoms. At initial step the atoms occupy perfect bcc lattice cites. This initial structure was optimized until energy changes less than 1e-6 eV, while forces acting on atoms don't exceed 1e-2 eV/angstrom. The electron-ion interaction is described by PAW pseudopotentials. The calculations have been collected by sampling chemical compositions across the entire compositional range. The chemical compositions have been sampled by progressively changing the number of atoms per constituent by 4. For each chemical composition of binaries and ternaries, the first-principle calculations have been run for 100 randomized arrangements of the constituents on the BCC lattice sites. We collected data for a total of 3,100 randomized atomic structures over 31 chemical compositions. The calculations have been collected on NERSC-Perlmutter and OLCF-Summit using the VASP 6.3.2. The VASP calculations for every atomic structure have been performed in 2 main steps: 1. Starting from an ideal body-centered-cubic (BCC) structure, geometry optimization with low precision has been executed to perform a preliminary optimization of the atomic structure. The output for this calculations is available in the files 0.CONTCAR, 0.OUTCAR, rlx1.out. 2. Using the atomic structure resulting from the preliminary geometry optimization, a second geometry optimization has been performed using normal precision. The output for this calculations is available in the files CONTCAR, OUTCAR, rlx2.out, vaspout.h5, and vasprun.xml. Cases 1-10 have been run without generating the file 'vaspout.h5'. Every chemical composition sampled across the composition range in the dataset has its own directory. The convention used to name the directories for binary alloys is AXBY, where A and B refer to the constituents, whereas X and Y are positive integers that represent the number of atoms for each constituent and their values still sum up to 128. Each atomic structure associated with a specific chemical composition has its own sub-directory within the directory of the corresponding chemical composition. The sub-directories for each atomic structure for each chemical composition are named 'case-*', where * is a positive integer that spans all the values from 1 through 100, extremes included. The files contained in each sub-directory 'case-*' for each atomic structure are as follows: FILES contained in each subdirectory with name case-N where N ranges between 11 and 100, extremes included: 1. INCAR: input file that contains various parameters and settings for controlling the behavior of the electronic structure calculations 2. KPOINTS: input file that specifies the Bloch vectors (k points) used to sample the Brillouin zone 3. 0.POSCAR: input file that defines the atomic structure of a system 4. 0.CONTCAR: output file that provides the atomic positions and cell parameters after the first geometry optimization has been run with the precision variable set to PREC=Low in the INCAR file 5. 0.OUTCAR: output file that contains detailed information about the progress of a calculation after the first geometry optimization has been run with the precision variable set to PREC=Low in the INCAR file 6. rlx1.out: file with diagnostic information about the execution of the first geometry optimization with precision variable set to PREC=Low in the INCAR file 7. POSCAR: input file that defines the atomic structure of a system after the first geometry optimization has been run at low precision. This represents the input for the second geometry optimization run with the precision variable set to PREC=Normal in the INCAR file 8. CONTCAR: output file that provides the atomic positions and cell parameters after the second geometry optimization has been run with the precision variable set to PREC=Normal in the INCAR file 9. OUTCAR: output file that contains detailed information about the progress of a calculation after the second geometry optimization has been run with the precision variable set to PREC=Normal in the INCAR file 10. rlx2.out: file with diagnostic information about the execution of the second geometry optimization with precision variable set to PREC=Normal in the INCAR file 11. vaspout.h5: hierarchical HDF5 file containing the inputs and outputs of a VASP calculation. To analyze the data in this file we recommend using py4vasp. This file is only produced if the VASP version used is compiled with HDF5 support 12. vasprun.xml: contains similar information to OUTCAR, but in an xml format. Subdirectories with name case-N, where N ranges between 1 and 10 (extremes included) contain all the files listed above except 'vaspout.h5'. Subdirectories with name case-N, where N ranges between 41 and 60 (extremes included), contain a duplicate copy of the files listed above except for KPOINTS. The names of the duplicate files end with -bis, and correspond to a second VASP calculation that has converged to a different optimized geometry. This research is sponsored by the Artificial Intelligence Initiative as part of the Laboratory Directed Research and Development (LDRD) Program of Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the US Department of Energy under contract DE-AC05-00OR22725. This work used resources of the Oak Ridge Leadership Computing Facility, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC05-00OR22725, under Directorate Discretionary awards MAT025 (Materials Science) and LRN026 (Machine Learning), and INCITE award MAT201. This work also used resources of the National Energy Research Scientific Computing Center, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231, under award ERCAP0025216. REFERENCES (1) Kresse, G. and Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. review B 47, 558 (1993). (2) Kresse, G. and Hafner, J. Ab initio molecular-dynamics simulation of the liquid-metal-amorphous-semiconductor transition in germanium. Phys. Rev. B 49, 14251 (1994) (3) Kresse, G. and Furthmuller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. materials science 6, 15-50 (1996) (4) Kresse, G. and Furthmuller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. review B 54, 11169 (1996) (5) Kresse, G. and Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. review b 59, 1758 (1999)

36 MATERIALS SCIENCE↗
DOE OSTI2023

The Origin of Star Formation in Early-type Galaxies Inferred from Spatially Resolved Spectroscopy

Abstract We investigate the origin of star formation activity in early-type galaxies with current star formation using spatially resolved spectroscopic data from the Mapping Nearby Galaxies at Apache Point Observatory in the Sloan Digital Sky Survey (SDSS). We first identify star-forming early-type galaxies from the SDSS sample, which are morphologically early-type but show current star formation activity in their optical spectra. We then construct comparison samples with different combinations of star formation activity and morphology, which include star-forming late-type galaxies, quiescent early-type galaxies, and quiescent late-type galaxies. Our analysis of the optical spectra reveals that the star-forming early-type galaxies have two distinctive episodes of star formation, which is similar to late-type galaxies but different from quiescent early-type galaxies with a single star formation episode. Star-forming early-type galaxies have properties in common with star-forming late-type galaxies, which include stellar population, gas and dust content, mass, and environment. However, the physical properties of star-forming early-type galaxies derived from spatially resolved spectroscopy differ from those of star-forming late-type galaxies in the sense that the gas in star-forming early-type galaxies is more concentrated than their stars, and is often kinematically misaligned with stars. The age gradient of star-forming early-type galaxies also differs from those of star-forming late-type galaxies. Our findings suggest that the current star formation in star-forming early-type galaxies has an external origin including galaxy mergers or accretion gas from the cosmic web.

Astronomy & Astrophysics↗
DOE OSTI2024

Accelerating defect predictions in semiconductors using graph neural networks

First-principles computations reliably predict the energetics of point defects in semiconductors but are constrained by the expense of using large supercells and advanced levels of theory. Machine learning models trained on computational data, especially ones that sufficiently encode defect coordination environments, can be used to accelerate defect predictions. Here, we develop a framework for the prediction and screening of native defects and functional impurities in a chemical space of group IV, III–V, and II–VI zinc blende semiconductors, powered by crystal Graph-based Neural Networks (GNNs) trained on high-throughput density functional theory (DFT) data. Using an innovative approach of sampling partially optimized defect configurations from DFT calculations, we generate one of the largest computational defect datasets to date, containing many types of vacancies, self-interstitials, anti-site substitutions, impurity interstitials and substitutions, as well as some defect complexes. We applied three types of established GNN techniques, namely crystal graph convolutional neural network, materials graph network, and Atomistic Line Graph Neural Network (ALIGNN), to rigorously train models for predicting defect formation energy (DFE) in multiple charge states and chemical potential conditions. We find that ALIGNN yields the best DFE predictions with root mean square errors around 0.3 eV, which represents a prediction accuracy of 98% given the range of values within the dataset, improving significantly on the state-of-the-art. We further show that GNN-based defective structure optimization can take us close to DFT-optimized geometries at a fraction of the cost of full DFT. The current models are based on the semi-local generalized gradient approximation-Perdew–Burke–Ernzerhof (PBE) functional but are highly promising because of the correlation of computed energetics and defect levels with higher levels of theory and experimental data, the accuracy and necessity of discovering novel metastable and low energy defect structures at the PBE level of theory before advanced methods could be applied, and the ability to train multi-fidelity models in the future with new data from non-local functionals. The DFT-GNN models enable prediction and screening across thousands of hypothetical defects based on both unoptimized and partially optimized defective structures, helping identify electronically active defects in technologically important semiconductors.

Rahman, Md Habibur (ORCID:000000027705984X)↗
DOE OSTI2022

The impact of habitat fragmentation on domatia-dwelling mites and a mite-plant-fungus tritrophic interaction

Abstract Context Habitat fragmentation is a leading threat to biodiversity, yet the impacts of fragmentation on most taxa, let alone interactions among those taxa, remain largely unknown. Objectives We studied how three consequences of fragmentation—reduced patch connectivity, altered patch shape, and edge proximity—impact plant-dwelling mite communities and mite-plant-fungus interactions within a large-scale habitat fragmentation experiment. Methods We sampled mite communities from the leaves of Quercus nigra (a plant species that has foliar domatia which harbor fungivorous and predacious mites) near and far from edge within fragments of varying edge-to-area ratio (shape) and connectivity via corridors. We also performed a mite-exclusion experiment across these fragmentation treatments to test the effects of mite presence and fungal hyphal abundance on leaf surfaces. Results Habitat edges influenced the abundance and richness of leaf-dwelling mites; plants closer to the edge had higher mite abundance and species richness. Likewise, hyphal counts were higher on leaves near patch edges. Despite both mite and fungal abundance being higher at patch edges, leaf hyphal counts were not impacted by mite abundance on those leaves. Neither patch shape nor connectivity influenced mite abundance, mite species richness, or the influence of mites on leaf surface fungal abundance. Conclusion Our results suggest that mites and foliar fungi may be independently affected by edge-structured environmental gradients, like temperature, rather than trophic effects. We demonstrate that large-scale habitat fragmentation and particularly edge effects can have impacts on multiple levels of microscopic communities, even in the absence of cascading trophic effects.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

NbTa_BCC_SolidSolution_128atoms_VASP6

We performed density functional theory (DFT) calculations for body-centered-cubic (BCC) structures with 128 lattices sites of solid solution binary alloys niobium-tantalum (Nb-Ta). The electronic structures of alloys have been calculated using Vienna Ab initio Simulation Package (VASP). Within this package the DFT approach is used to reduce many-body Schrodinger equation to set of single particle Kohn-Sham (KS) equations. The generalized electronic exchange-correlation functional is described by generalized gradient approximation with the Perdew-Burke-Ernzerhof parametrization. The electron-ion interactions is described by pseudopotentials developed within the plane-wave basis projector augmented-wave (PAW) approach \cite{PAW}. These pseudopotentials are available at the VASP portal (http://cms.mpi.univie.ac.at/vasp/). Our calculations have been run with the pseudopotentials treating s and p semi-core states as valence in case for the elements Nb. For Ta, p semi-core states as valence were chosen. The electronic densities and potentials are expanded over plane-waves with energy cutoff of 350 eV. 2x2x2 k-mesh and normal precision were used. The alloys were modeled by supercell containing 128 randomly distributed atoms. At initial step the atoms occupy perfect bcc lattice cites. This initial structure was optimized until energy changes less than 1e-6 eV, while forces acting on atoms don't exceed 1e-2 eV/angstrom. The electron-ion interaction is described by PAW pseudopotentials. The calculations have been collected by sampling chemical compositions across the entire compositional range. The chemical compositions have been sampled by progressively changing the number of atoms per constituent by 4. For each chemical composition of binaries and ternaries, the first-principle calculations have been run for 100 randomized arrangements of the constituents on the BCC lattice sites. We collected data for a total of 3,100 randomized atomic structures over 31 chemical compositions. The calculations have been collected on NERSC-Perlmutter using the VASP 6.3.2. The VASP calculations for every atomic structure have been performed in 2 main steps: 1. Starting from an ideal body-centered-cubic (BCC) structure, geometry optimization with low precision has been executed to perform a preliminary optimization of the atomic structure. The output for this calculations is available in the files 0.CONTCAR, 0.OUTCAR, rlx1.out. 2. Using the atomic structure resulting from the preliminary geometry optimization, a second geometry optimization has been performed using normal precision. The output for this calculations is available in the files CONTCAR, OUTCAR, rlx2.out, vaspout.h5, and vasprun.xml. Cases 1-10 have been run without generating the file 'vaspout.h5'. Every chemical composition sampled across the composition range in the dataset has its own directory. The convention used to name the directories for ternary alloys is AXBYCZ, where A, B, and C refer to the constituents, and X, Y, and Z are positive integers that represent the number of atoms for each constituent and their values still sum up to 128. Each atomic structure associated with a specific chemical composition has its own sub-directory within the directory of the corresponding chemical composition. The sub-directories for each atomic structure for each chemical composition are named 'case-*', where * is a positive integer that spans all the values from 1 through 100, extremes included. The files contained in each sub-directory 'case-*' for each atomic structure are as follows: FILES contained in each subdirectory with name case-N where N ranges between 11 and 100, extremes included: 1. INCAR: input file that contains various parameters and settings for controlling the behavior of the electronic structure calculations 2. KPOINTS: input file that specifies the Bloch vectors (k points) used to sample the Brillouin zone 3. 0.POSCAR: input file that defines the atomic structure of a system 4. 0.CONTCAR: output file that provides the atomic positions and cell parameters after the first geometry optimization has been run with the precision variable set to PREC=Low in the INCAR file 5. 0.OUTCAR: output file that contains detailed information about the progress of a calculation after the first geometry optimization has been run with the precision variable set to PREC=Low in the INCAR file 6. rlx1.out: file with diagnostic information about the execution of the first geometry optimization with precision variable set to PREC=Low in the INCAR file 7. POSCAR: input file that defines the atomic structure of a system after the first geometry optimization has been run at low precision. This represents the input for the second geometry optimization run with the precision variable set to PREC=Normal in the INCAR file 8. CONTCAR: output file that provides the atomic positions and cell parameters after the second geometry optimization has been run with the precision variable set to PREC=Normal in the INCAR file 9. OUTCAR: output file that contains detailed information about the progress of a calculation after the second geometry optimization has been run with the precision variable set to PREC=Normal in the INCAR file 10. rlx2.out: file with diagnostic information about the execution of the second geometry optimization with precision variable set to PREC=Normal in the INCAR file 11. vaspout.h5: hierarchical HDF5 file containing the inputs and outputs of a VASP calculation. To analyze the data in this file we recommend using py4vasp. This file is only produced if the VASP version used is compiled with HDF5 support 12. vasprun.xml: contains similar information to OUTCAR, but in an xml format. Subdirectories with name case-N, where N ranges between 1 and 10 (extremes included) contain all the files listed above except 'vaspout.h5'. Subdirectories with name case-N, where N ranges between 41 and 60 (extremes included), contain a duplicate copy of the files listed above except for KPOINTS. The names of the duplicate files end with -bis, and correspond to a second VASP calculation that has converged to a different optimized geometry. This research is sponsored by the Artificial Intelligence Initiative as part of the Laboratory Directed Research and Development (LDRD) Program of Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the US Department of Energy under contract DE-AC05-00OR22725. This work used resources of the Oak Ridge Leadership Computing Facility, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC05-00OR22725, under Directorate Discretionary awards MAT025 (Materials Science) and LRN026 (Machine Learning), and INCITE award MAT201. This work also used resources of the National Energy Research Scientific Computing Center, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231, under award ERCAP0025216. REFERENCES (1) Kresse, G. and Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. review B 47, 558 (1993). (2) Kresse, G. and Hafner, J. Ab initio molecular-dynamics simulation of the liquid-metal-amorphous-semiconductor transition in germanium. Phys. Rev. B 49, 14251 (1994) (3) Kresse, G. and Furthmüller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. materials science 6, 15-50 (1996) (4) Kresse, G. and Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. review B 54, 11169 (1996) (5) Kresse, G. and Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. review b 59, 1758 (1999)

36 MATERIALS SCIENCE↗
NASA NTRS2010

Remote Sensing of the Urban Heat Island Effect Across Biomes in the Continental USA

Impervious surface area (ISA) from the Landsat TM-based NLCD 2001 dataset and land surface temperature (LST) from MODIS averaged over three annual cycles (2003-2005) are used in a spatial analysis to assess the urban heat island (UHI) skin temperature amplitude and its relationship to development intensity, size, and ecological setting for 38 of the most populous cities in the continental United States. Development intensity zones based on %ISA are defined for each urban area emanating outward from the urban core to the nonurban rural areas nearby and used to stratify sampling for land surface temperatures and NDVI. Sampling is further constrained by biome and elevation to insure objective intercomparisons between zones and between cities in different biomes permitting the definition of hierarchically ordered zones that are consistent across urban areas in different ecological setting and across scales. We find that ecological context significantly influences the amplitude of summer daytime UHI (urban-rural temperature difference) the largest (8 C average) observed for cities built in biomes dominated by temperate broadleaf and mixed forest. For all cities combined, ISA is the primary driver for increase in temperature explaining 70% of the total variance in LST. On a yearly average, urban areas are substantially warmer than the non-urban fringe by 2.9 C, except for urban areas in biomes with arid and semiarid climates. The average amplitude of the UHI is remarkably asymmetric with a 4.3 C temperature difference in summer and only 1.3 C in winter. In desert environments, the LST's response to ISA presents an uncharacteristic "U-shaped" horizontal gradient decreasing from the urban core to the outskirts of the city and then increasing again in the suburban to the rural zones. UHI's calculated for these cities point to a possible heat sink effect. These observational results show that the urban heat island amplitude both increases with city size and is seasonally asymmetric for a large number of cities across most biomes. The implications are that for urban areas developed within forested ecosystems the summertime UHI can be quite high relative to the wintertime UHI suggesting that the residential energy consumption required for summer cooling is likely to increase with urban growth within those biomes.

Imhoff, Marc L.↗
NASA NTRS1994

Automation of reverse engineering process in aircraft modeling and related optimization problems

During the year of 1994, the engineering problems in aircraft modeling were studied. The initial concern was to obtain a surface model with desirable geometric characteristics. Much of the effort during the first half of the year was to find an efficient way of solving a computationally difficult optimization model. Since the smoothing technique in the proposal 'Surface Modeling and Optimization Studies of Aerodynamic Configurations' requires solutions of a sequence of large-scale quadratic programming problems, it is important to design algorithms that can solve each quadratic program in a few interactions. This research led to three papers by Dr. W. Li, which were submitted to SIAM Journal on Optimization and Mathematical Programming. Two of these papers have been accepted for publication. Even though significant progress has been made during this phase of research and computation times was reduced from 30 min. to 2 min. for a sample problem, it was not good enough for on-line processing of digitized data points. After discussion with Dr. Robert E. Smith Jr., it was decided not to enforce shape constraints in order in order to simplify the model. As a consequence, P. Dierckx's nonparametric spline fitting approach was adopted, where one has only one control parameter for the fitting process - the error tolerance. At the same time the surface modeling software developed by Imageware was tested. Research indicated a substantially improved fitting of digitalized data points can be achieved if a proper parameterization of the spline surface is chosen. A winning strategy is to incorporate Dierckx's surface fitting with a natural parameterization for aircraft parts. The report consists of 4 chapters. Chapter 1 provides an overview of reverse engineering related to aircraft modeling and some preliminary findings of the effort in the second half of the year. Chapters 2-4 are the research results by Dr. W. Li on penalty functions and conjugate gradient methods for quadratic programming problems.

Li, W.↗
DOE OSTI2022

Experimentally validated multiphysics modeling of fracture induced by thermal shocks in sintered UO 2 pellets

Uranium Dioxide (UO 2 ) fuel powers almost all commercial Nuclear Power Plants (NPPs) worldwide, generating carbon-free energy and contributing to the fight against climate change. UO 2 fuel incurs damage and fractures due to large thermal gradients that develop across the fuel pellet during normal and transient operating conditions. A comprehensive understanding of the underlying mechanisms by which these processes take place is still lacking. A combined experimental and computational approach is utilized here to quantify the behavior of UO 2 fuel fracture induced by thermal shock. Here, this work introduces both (1) an experimental study to understand the fuel fracturing behavior of sintered UO 2 pellets when exposed to thermal shock, and (2) a Multiphysics phase-field fracture model capable of simulating this process. Parametric studies were conducted to evaluate the effects of uncertainties in fracture properties on the fracture behavior of UO 2 due to thermal shocking. A set of energy release rate (or equivalently fracture toughness) and contract area (the part of the fuel pellet in direct contact with the cold bath) were able to capture the overall fracture trends of the corresponding experimental data. Our combined approach presents a new method for accounting for the effects of microstructure and sample size on the energy release rate/fracture toughness. The experimental data were collected from multiple experiments that exposed UO 2 pellets to high-temperature conditions (589–676 °C) followed by a quench in sub-zero water. This work demonstrates that joint experimental and computational efforts are able to advance the understanding of thermal fracture in the primary fuel source for existing and future NPPs.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Diffusion-Model-Assisted Supervised Learning of Generative Models for Density Estimation

Here, we present a supervised learning framework of training generative models for density estimation. Generative models, including generative adversarial networks (GANs), normalizing flows, and variational auto-encoders (VAEs), are usually considered as unsupervised learning models, because labeled data are usually unavailable for training. Despite the success of the generative models, there are several issues with the unsupervised training, e.g., requirement of reversible architectures, vanishing gradients, and training instability. To enable supervised learning in generative models, we utilize the score-based diffusion model to generate labeled data. Unlike existing diffusion models that train neural networks to learn the score function, we develop a training-free score estimation method. This approach uses mini-batch-based Monte Carlo estimators to directly approximate the score function at any spatial-temporal location in solving an ordinary differential equation (ODE), corresponding to the reverse-time stochastic differential equation (SDE). This approach can offer both high accuracy and substantial time savings in neural network training. Once the labeled data are generated, we can train a simple, fully connected neural network to learn the generative model in the supervised manner. Compared with existing normalizing flow models, our method does not require the use of reversible neural networks and avoids the computation of the Jacobian matrix. Compared with existing diffusion models, our method does not need to solve the reverse-time SDE to generate new samples. As a result, the sampling efficiency is significantly improved. We demonstrate the performance of our method by applying it to a set of 2D datasets as well as real data from the University of California Irvine (UCI) repository.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Learning curves for drug response prediction in cancer cell lines

Motivated by the size and availability of cell line drug sensitivity data, researchers have been developing machine learning (ML) models for predicting drug response to advance cancer treatment. As drug sensitivity studies continue generating drug response data, a common question is whether the generalization performance of existing prediction models can be further improved with more training data. We utilize empirical learning curves for evaluating and comparing the data scaling properties of two neural networks (NNs) and two gradient boosting decision tree (GBDT) models trained on four cell line drug screening datasets. The learning curves are accurately fitted to a power law model, providing a framework for assessing the data scaling behavior of these models. The curves demonstrate that no single model dominates in terms of prediction performance across all datasets and training sizes, thus suggesting that the actual shape of these curves depends on the unique pair of an ML model and a dataset. The multi-input NN (mNN), in which gene expressions of cancer cells and molecular drug descriptors are input into separate subnetworks, outperforms a single-input NN (sNN), where the cell and drug features are concatenated for the input layer. In contrast, a GBDT with hyperparameter tuning exhibits superior performance as compared with both NNs at the lower range of training set sizes for two of the tested datasets, whereas the mNN consistently performs better at the higher range of training sizes. Moreover, the trajectory of the curves suggests that increasing the sample size is expected to further improve prediction scores of both NNs. These observations demonstrate the benefit of using learning curves to evaluate prediction models, providing a broader perspective on the overall data scaling characteristics. A fitted power law learning curve provides a forward-looking metric for analyzing prediction performance and can serve as a co-design tool to guide experimental biologists and computational scientists in the design of future experiments in prospective research studies.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Accessing the Subsurface Biosphere Within Rocks Undergoing Active Low-Temperature Serpentinization in the Samail Ophiolite (Oman Drilling Project)

The Oman Drilling Project established an “Active Alteration” multi-borehole observatory in peridotites undergoing low-temperature serpentinization in the Samail Ophiolite. The highly serpentinized rocks are in contact with strongly reducing fluids. Distinct hydrological regimes, governed by differences in rock porosity and fracture density, give rise to steep redox (Eh +200 to –750 mV) and pH (pH range 8.5–11.2) gradients within the 300–400 m deep boreholes. The serpentinites and fluids host an active subsurface ecosystem. Microbial cell abundances in serpentinite vary at least six orders of magnitude, from ≤3.5 × 10 1 to 2.9 × 10 7 cells/g. Low levels of biological sulfate reduction (2–1,000 fmol/cm 3 /day) can be detected in rock cores, particularly in rocks in contact with reduced groundwaters with pH < 10.5. Thermodesulfovibrio is the predominant sulfate reducer identified via metagenomic sequencing of adjacent groundwater communities. Here, we infer that transport and reaction of microbially generated sulfide with the serpentine and brucite assemblages gives rise to optical darkening and sulfide overprinting, including the formation of tochilinite-vallerite group minerals, potentially serving as an indicator that this system is inhabited by microbial life. Olivine mesh-cores replaced with ferroan brucite and minor awaruite, abundant veins containing hydroandradite garnet and polyhedral serpentine, and late-stage carbonate veins are suggested as targets for future spatially resolved life-detection investigations. The high-quality whole-round core samples that have been preserved can be further probed to define how life distributes itself and functions within a system where chemical disequilibria are sustained by low-temperature water/rock interaction, and how biosignatures of in situ microbial activity are generated.

58 GEOSCIENCES↗