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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 397 records · Page 22

All sources NASA NTRS DOE OSTI

DOE OSTI2021

Differentiable programming for online training of a neural artificial viscosity function within a staggered grid Lagrangian hydrodynamics scheme

Lagrangian methods to solve the inviscid Euler equations produce numerical oscillations near shock waves. A common approach to reducing these oscillations is to add artificial viscosity (AV) to the discrete equations. The AV term acts as a dissipative mechanism that attenuates oscillations by smearing the shock across a finite number of computational cells. However, AV introduces several control parameters that are not determined by the underlying physical model, and hence, in practice are tuned to the characteristics of a given problem. We seek to improve the standard quadratic-linear AV form by replacing it with a learned neural function that reduces oscillations relative to exact solutions of the Euler equations, resulting in a hybrid numerical-neural hydrodynamic solver. Because AV is an artificial construct that exists solely to improve the numerical properties of a hydrodynamic code, there is no offline ‘viscosity data’ against which a neural network can be trained before inserting into a numerical simulation, thus requiring online training. We achieve this via differentiable programming, i.e. end-to-end backpropagation or adjoint solution through both the neural and differential equation code, using automatic differentiation of the hybrid code in the Julia programming language to calculate the necessary loss function gradients. A novel offline pre-training step accelerates training by initializing the neural network to the default numerical AV scheme, which can be learned rapidly by space-filling sampling over the AV input space. We find that online training over early time steps of simulation is sufficient to learn a neural AV function that reduces numerical oscillations in long-term hydrodynamic shock simulations. These results offer an early proof-of-principle that online differentiable training of hybrid numerical schemes with novel neural network components can improve certain performance aspects existing in purely numerical schemes.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Blanco DECam Bulge Survey (BDBS) Ⅳ: Metallicity distributions and bulge structure from 2.6 million red clump stars

We present photometric metallicity measurements for a sample of 2.6 million bulge red clump stars extracted from the Blanco DECam Bulge Survey (BDBS). Similar to previous studies, we find that the bulge exhibits a strong vertical metallicity gradient, and that at least two peaks in the metallicity distribution functions appear at b < -5°. We can discern a metal-poor ([Fe/H] ~ -0.3) and metal-rich ([Fe/H] ~+0.2) abundance distribution that each show clear systematic trends with latitude, and may be best understood by changes in the bulge’s star formation/enrichment processes. Both groups exhibit asymmetric tails, and as a result we argue that the proximity of a star to either peak in [Fe/H] space is not necessarily an affirmation of group membership. The metal-poor peak shifts to lower [Fe/H] values at larger distances from the plane while the metal-rich tail truncates. Close to the plane, the metal-rich tail appears broader along the minor axis than in off-axis fields. We also posit that the bulge has two metal-poor populations – one that belongs to the metal-poor tail of the low latitude and predominantly metal-rich group, and another belonging to the metal-poor group that dominates in the outer bulge. We detect the X-shape structure in fields with |Z| >0.7 kpc and for stars with [Fe/H] < -0.5. Stars with [Fe/H] > -0.5 may form a spheroidal or ‘thick bar’ distribution while those with [Fe/H] ≳ -0.1 are strongly concentrated near the plane.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2025

Microstructural evaluation of thin-wall sections of 316L stainless steel produced by laser powder-bed fusion processing

Laser powder bed fusion (LBPF) is a method of additive manufacturing (AM) that offers a means to fabricate complex components with features which rely upon the mechanical integrity of thin-wall (less than 0.4 mm thick) structures. The microstructure of a 316L AM thin wall is investigated as metallographically prepared from a cylindrical shell. Optimization is pursued through mechanical polishing and chemical etching to minimize the effects of surface roughness. Microanalysis measurements of the sectioned component reveal the sample has high microstructural fidelity, uniform composition and phase content across grain as well as powder layer boundaries within the thin wall. The results are attributed to a uniform thermal gradient at the part scale during rapid solidification.

36 MATERIALS SCIENCE↗

DOE OSTI2020

Experimental system for studying temperature gradient-driven fracture of oxide nuclear fuel out of reactor

Temperature gradients in ceramic light water reactor (LWR) uranium dioxide (UO 2 ) nuclear fuel pellets generate thermal stresses that cause fractures in the fuel beginning early in the life of fresh fuel. The combination of heating due to fission and forced convective cooling on the exterior of LWR fuel rods generates a temperature profile that is difficult to replicate outside the reactor environment. In the present study, a state-of-the-art experimental set-up using electrical heating to replicate fission heating was built and surrogate fuel materials such as ceria (CeO 2 ) were used to validate the system. Cracking experiments were conducted on these surrogates by inducing reactivity-initiated-accident (RIA) like temperature gradients in the pellets via induction and direct resistance heating. Induction heating was done using copper coils and molybdenum susceptors which heated the surrogates to a threshold temperature that is sufficiently high for the fuel material to conduct current. Thereafter, direct resistance heating was used by a D.C. power supply to introduce volumetric heating to replicate LWR operating conditions analogous to fission heating. The pellets were held against nickel electrodes and mounted on a boron nitride test-stand. All the tests were carried out in a stainless-steel vacuum chamber. Simultaneous real-time dual imaging of the surrogate pellet surface has been implemented using an optical and infrared camera system which will be mounted along axial and perpendicular directions to the pellet surface respectively. A beam-splitter was used to split the incoming radiation from the sample into two halves. While one of the beams is transmitted from the splitter through a bandpass filter to obtain optical images, the other beam is reflected from the splitter to the thermal camera to capture full field temperature gradients of the as fabricated pellet surface during crack initiation and propagation. In the current series of tests, a 2-color pyrometer was used for recording and comparing the surface and centerline temperatures of the surrogate pellets in lieu of the thermal camera. A LabVIEW data acquisition system has been set up for collecting useful data during experiments.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2024

Automated path planning for functionally graded materials considering phase stability and solidification behavior: Application to the Mo-Nb-Ta-Ti system

Functionally graded materials have the potential to improve upon monolithic parts by locally tailoring compositions to surrounding environmental conditions. Difficulties arise when designing composition gradients as incompatible materials can result in detrimental phase formation and failure of the gradient joint. As many alloys are multi-component, designing a composition gradient free of detrimental phases is difficult due to the large composition space available to explore. A framework was developed that improves the path planning algorithm and surrogate models with adaptive sampling schemes specific to their problem definition. A cost function was created to minimize a property (such as cracking susceptibility) along a path. This framework was applied to the Mo-Nb-Ta-Ti system as a case study to showcase the efficiency in building the surrogate models and in iterating different optimal compositionally graded paths.

36 MATERIALS SCIENCE↗

DOE OSTI2024

Investigation of W-SiC compositionally graded films as a divertor material

W-SiC composite material is a promising plasma-facing material candidate alternative to pure W due to the low neutron activation, low impurity radiation, and low tritium diffusivity of SiC while leveraging the high erosion resistance of the W armor. Additionally, W and SiC have high thermomechanical compatibility given their similar thermal expansion rates. The present study addresses the synthesis and performance of compositionally graded W-SiC films fabricated by pulsed-DC magnetron sputtering. Compositional gradients were characterized using transmission electron microscopy (TEM) and energy-dispersive X-ray spectroscopy (EDS), and crystallographic information was obtained using electron diffraction and X-ray diffraction (XRD). Samples were exposed to L-mode deuterium plasma discharges in the DIII-D tokamak using the Divertor Material Evaluation System (DiMES). Post-mortem characterizations were performed using scanning electron microscopy (SEM) and XRD. Electron diffraction and XRD showed that the compositionally graded W-SiC films were composed of polycrystalline W and amorphous SiC with amorphous W+SiC interlayers. No macroscopic delamination or microstructural changes were observed under mild exposure conditions. Here, this study serves as a preliminary examination of W-SiC compositionally graded composites as a potential candidate divertor material in future tokamak devices.

36 MATERIALS SCIENCE↗

DOE OSTI2024

Perspective on Integrative Simulations of Bioenergetic Domains

Bioenergetic processes in cells, such as photosynthesis or respiration, integrate many time and length scales, which makes the simulation of energy conversion with a mere single level of theory impossible. Just like the myriad of experimental techniques required to examine each level of organization, an array of overlapping computational techniques is necessary to model energy conversion. Here, in this work, a perspective is presented on recent efforts for modeling bioenergetic phenomena with a focus on molecular dynamics simulations and its variants as a primary method. An overview of the various classical, quantum mechanical, enhanced sampling, coarse-grained, Brownian dynamics, and Monte Carlo methods is presented. Example applications discussed include multiscale simulations of membrane-wide electron transport, rate kinetics of ATP turnover from electrochemical gradients, and finally, integrative modeling of the chromatophore, a photosynthetic pseudo-organelle.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2023

Magnetic phase diagram mapping in Fe 1- x Rh x composition-spread thin films

We have fabricated high-quality polycrystalline Fe 1- x Rh x composition-spread thin films by cosputtering Fe and Rh, and investigated their structural and magnetic transformations as a systematic function of composition. With increasing Rh concentration, Fe 1-x Rh x thin film undergoes from an α' phase to a disordered γ phase and also shows a magnetic transition from a ferromagnetic phase to a paramagnetic phase. Vibrating-sample magnetometry and x-ray magnetic circular dichroism measurements show an antiferromagnetic-ferromagnetic transition in the range of 0.52 < x < 0.58 in the Fe 1- x Rh x composition gradient at room temperature. Based on our structural and magnetic property mapping, we construct a thin-film phase diagram of Fe 1- x Rh x . Compared to reported results in bulk alloys, the antiferromagnetic-ferromagnetic transition in the Fe 1- x Rh x thin films was found to occur at slightly higher Rh concentrations, while the boundary between the pure γ phase and the α'/ γ mixed phase region is shifted to the lower concentration Rh.

36 MATERIALS SCIENCE↗

DOE OSTI2026

Mapping the soil microbiome functions shaping wetland methane emissions

Accounting for only 8% of Earth’s land cover, freshwater wetlands remain the foremost contributors to global methane emissions. Yet the microorganisms and processes underlying methane emissions from wetland soils remain poorly understood. Over a five-year period, we surveyed the microbial membership and in situ methane measurements from over 700 samples in one of the most prolific methane-emitting wetlands in the United States. We constructed a catalog of 2,502 metagenome-assembled genomes (MAGs), with more than half of the 70 bacterial and archaeal phyla sampled containing novel lineages. Integration of these data with 133 soil metatranscriptomes provided a genome-resolved view of the biogeochemical specialization and versatility expressed over wetland soil spatial and temporal gradients. Centimeter-scale depth differences best explained patterns of microbial community structure and transcribed functionalities, even more than land cover or temporal information. Moreover, while extended flooding restructured soil redox, this perturbation failed to reconfigure the transcriptional profiles of methane-cycling microorganisms, contrasting with theoretically expected responses to hydrological perturbations. Co-expression analyses, coupled with depth-resolved methane measurements, revealed the metabolisms and trophic structures most predictive of methane hotspots. Mapping the spatiotemporal transcriptional patterns on this compendium of biogeochemically classified soil-derived genomes begins to untangle the microbial carbon, energy, and nutrient processing contributing to wetland methane production.

MAG↗

DOE OSTI2022

Median bed-material sediment particle size across rivers in the contiguous US

Abstract. Bed-material sediment particle size data, particularly the median sediment particle size (D50), are critical for understanding and modeling riverine sediment transport. However, sediment particle size observations are primarily available at individual sites. Large-scale modeling and assessment of riverine sediment transport are limited by the lack of continuous regional maps of bed-material sediment particle size. We hence present a map of D50 over the contiguous US in a vector format that corresponds to approximately 2.7 million river segments (i.e., flowlines) in the National Hydrography Dataset Plus (NHDPlus) dataset. We develop the map in four steps: (1) collect and process the observed D50 data from 2577 U.S. Geological Survey stations or U.S. Army Corps of Engineers sampling locations; (2) collocate these data with the NHDPlus flowlines based on their geographic locations, resulting in 1691 flowlines with collocated D50 values; (3) develop a predictive model using the eXtreme Gradient Boosting (XGBoost) machine learning method based on the observed D50 data and the corresponding climate, hydrology, geology, and other attributes retrieved from the NHDPlus dataset; and (4) estimate the D50 values for flowlines without observations using the XGBoost predictive model. We expect this map to be useful for various purposes, such as research in large-scale river sediment transport using model- and data-driven approaches, teaching environmental and earth system sciences, planning and managing floodplain zones, etc. The map is available at https://doi.org/10.5281/zenodo.4921987 (Li et al., 2021a).

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2026

Classical-Quantum Algorithm for Solving Stochastic Programs

Stochastic programming provides a rigorous mathematical framework for making decisions under uncertainty in a risk-aware manner. Two-stage stochastic programming is, perhaps, the simplest form of this framework. Here the first-stage variables represent decisions that must be made "here and now" in the face of uncertainty, while the second-stage variables are decisions made after uncertain events. However, the broad adoption of stochastic programming has been hindered by computational challenges caused by the two-stage stochastic programming formulation which requires solving an ensemble of optimization problems. Using quantum amplitude estimation (QAE), quantum computers have shown the theoretic ability to compute expectations with Monte-Carlo methods with quadratically fewer samples than classical methods. In this work, we present a quantum algorithm for computing the expectation term using QAE for given first-stage decisions. Further, we detail methods of computing gradient information from the quantum calculation enabling the application of classical gradient-based optimization techniques. The result is a classical-quantum hybrid method of solving two-stage stochastic programs. These techniques are demonstrated with computational experiments based an engineering optimization problem.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Mapping Rare Earths and Toxics in E-Waste via Hyperspectral Imaging and Machine Learning

Electronic waste (e-waste) presents a mounting challenge to environmental sustainability due to its complex composition, which includes high-value rare earth elements, hazardous organic compounds, and non-recyclable plastics. Accurate and scalable material classification is essential for enabling efficient resource recovery and safe recycling practices. This study introduces a confidence-aware classification pipeline that combines mid-infrared hyperspectral imaging (HSI), spectral angle mapping (SAM), and iterative machine learning to perform pixel-level material identification across e-waste devices. A curated spectral library encompassing artificial materials (e.g., plastic iron oxide, galvanized metals), minerals (e.g., allanite, hematite), and organic compounds (e.g., benzanthracene, toluene) was used to generate pseudo-labels, each assigned a confidence score based on SAM-derived spectral similarity. High-confidence samples from seven consumer electronics—digital cameras, keyboards, laptop fans, modems, motherboards, TV remotes, and speakers—were iteratively expanded and classified using models such as Support Vector Machine (SVM), Random Forest, Gradient Boosting Classifier, Partial Least Squares Discriminant Analysis (PLSDA) and Logistic Regression. The best-performing classifiers achieved macro F1 scores approaching 1.0. Results revealed widespread plastic content (dominated by plastic iron oxide), the presence of rare earth-bearing minerals like cerium-containing allanite, and pervasive detection of hazardous organics such as benzanthracene. Principal Component Analysis (PCA) visualizations and confusion matrices confirmed high separability and robust classification performance. This methodology enables precise, non-destructive, and scalable classification of heterogeneous e-waste streams. It supports automated, hazard-aware sorting in recycling workflows, facilitating selective recovery of critical materials and compliance with circular economy goals. The confidence-aware framework provides a foundation for real-time deployment in industrial settings, offering significant implications for smart e-recycling infrastructure and policy-driven material stewardship.

Circular economy↗

DOE OSTI2022

Global stochastic optimization of stellarator coil configurations

In the construction of a stellarator, the manufacturing and assembling of the coil system is a dominant cost. These coils need to satisfy strict engineering tolerances, and if those are not met the project could be cancelled as in the case of the National Compact Stellarator Experiment (NCSX) project. Therefore, our goal is to find coil configurations that increase construction tolerances without compromising the performance of the magnetic field. In this paper, we develop a gradient-based stochastic optimization model which seeks robust stellarator coil configurations in high dimensions. In particular, we design a two-step method: first, we perform an approximate global search by a sample efficient trust-region Bayesian optimization; second, we refine the minima found in step one with a stochastic local optimizer. To this end, we introduce two stochastic local optimizers: BFGS applied to the sample average approximation; and Adam, equipped with a control variate for variance reduction. Numerical simulations performed on a W7-X-like coil configuration demonstrate that our global optimization approach finds a variety of promising local solutions at less than 0.1% of the cost of previous work, which considered solely local stochastic optimization.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2025

Molecular Modeling and Molecular Dynamics Simulation of a Packed and Intact Bacterial Microcompartment

Bacterial microcompartments (BMCs) are protein-bound organelles found in some bacteria which encapsulate enzymes for enhanced catalytic activity. These compartments spatially sequester enzymes within semipermeable shell proteins and are packed full of enzyme cargoes and metabolites as they fulfill their function. Coupling together recent SAXS and proteomics work, it is possible to develop molecular models for these microcompartments and interrogate enzyme and metabolite dynamics within. Our primary goal of this study is to quantify the permeability of metabolite glyceraldehyde-3-phosphate (G3P) and dihydroxyacetone phosphate (DHAP) across the BMC shell through classical molecular dynamics simulation. The Haliangium ochraceum model of BMC shell (PDB: 6MZX) was used to model an intact BMC of approximately 10 million atoms. Working at this scale presented its own challenges in managing large data sets, with multiple challenges and hardware advances discussed that facilitated this work. Over approximately 750 ns of aggregate simulation, we see multiple permeation events for these metabolites that were added at high concentration through the pores present within BMC shell tiles. When compared to independent permeability estimates for the same metabolites determined through replica exchange umbrella sampling simulations, the permeabilities varied by approximately 3 orders of magnitude. Regardless, the permeability coefficients for both G3P and DHAP are highly similar and very high, such that only very small concentration gradients can be maintained across the BMC shell between the cytosol and BMC interior. The large simulation systems also facilitated comparisons for molecular diffusivity in the crowded environment within the BMC shell. By our estimates, the viscosity within a packed BMC shell is at least 10-fold higher than it would be in neat solution and is the real driver for varying permeability estimates we obtained through simulation. These findings will be used as design inputs for future bioengineering efforts to make products from BMCs, highlighting how permeable BMC shells can be.

Diffusion↗

DOE OSTI2025

Decayheatml

This code is designed to predict and analyze the decay heat generated in molten salt reactors (MSRs) using a hybrid approach that combines machine learning and segmented polynomial fitting. The accurate prediction of decay heat is essential for reactor safety and the optimization of spent fuel storage. The code operates through several key components: 1) Data Architecture: It incorporates a modular data architecture that handles various MSR-specific operational parameters such as power density, humidity content, and air ingress. These parameters are sampled using Sobol sequences to ensure comprehensive coverage of operational uncertainties. 2) Machine Learning Framework: The code employs a diverse set of machine learning models, including polynomial regression, decision trees, random forests, gradient boosting, support vector regression, k-nearest neighbors, multi-layer perceptrons, and symbolic regression. These models are trained to predict decay heat over a wide temporal range, from immediate shutdown up to 10,000 years. 3) Region-Optimized Training: The temporal domain is divided into multiple regions, each modeled separately to capture distinct decay heat characteristics across different time scales. This approach significantly improves the accuracy and interpretability of predictions. 4) Segmented Polynomial Interpretation (SPI): The SPI method translates machine learning predictions into piecewise polynomial equations. These equations are physically interpretable and can be directly integrated into existing engineering workflows and safety analyses. 5) Front-End Interfaces: The code includes both a Jupyter notebook interface for research development and a Streamlit web application for operational deployment. These interfaces allow users to interactively explore decay heat predictions, adjust operational parameters, and visualize results in real-time. 6) Applications: The framework supports various applications, including safety system validation and spent fuel container optimization. It enables real-time evaluation of worst-case decay heat scenarios, informing the design of passive safety systems and optimizing container designs for long-term storage. Overall, this code provides a robust, accurate, and user-friendly tool for predicting decay heat in MSRs, enhancing reactor safety, and optimizing spent fuel management.

Retamales, Mauricio Eduardo Tano [Idaho National L↗

DOE OSTI2021

Revisiting Dropout: Escaping Pressure for Training Neural Networks with Multiple Costs

A common approach to jointly learn multiple tasks with a shared structure is to optimize the model with a combined landscape of multiple sub-costs. However, gradients derived from each sub-cost often conflicts in cost plateaus, resulting in a subpar optimum. In this work, we shed light on such gradient conflict challenges and suggest a solution named Cost-Out, which randomly drops the sub-costs for each iteration. We provide the theoretical and empirical evidence of the existence of escaping pressure induced by the Cost-Out mechanism. While simple, the empirical results indicate that the proposed method can enhance the performance of multi-task learning problems, including two-digit image classification sampled from MNIST dataset and machine translation tasks for English from and to French, Spanish, and German WMT14 datasets.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2023

The effect of secondary phases on microstructure and irradiation damage in an as-built additively manufactured 316 L stainless steel with a hafnium compositional gradient

Additive manufacturing (AM) or rapid prototyping has become a crucial tool for reducing both cost and time while increasing efficiency in qualifying structural materials for reactor use. In this study, directed energy deposition (DED) was used to develop an as-built 316 L stainless steel sample with three regions of increasing Hf-dopant to study the effects of Hf on the irradiation response of the material. Morphological and microstructural changes were analyzed before and after 2 MeV proton irradiation at 360 °C to a damage of 2.5 dpa at ~ 10 µm below the surface. The addition of Hf effectively suppressed radiation-induced damage (dislocation loops, radiation-induced segregation) due to enhanced point defect recombination. The radiation damage seen in the as-built sample was further compared to a thermo-mechanically treated counterpart of the same fabrication and irradiation parameters which was found to behave superiorly. In conclusion, the increased radiation resistance of this material may be attributed to the as-built microstructure, which includes undissolved Hf particles, delta ferrite grains and cellular sub-grain boundaries that can hinder defect motion.

316L Stainless Steel↗

DOE OSTI2022

Heterogeneous phase transformation pathways in additively manufactured Al-Ce-Mn alloys

Heat treatment of additively manufactured Al-Ce based multicomponent alloys leads to complex microstructure evolution. In this research, the ability to extend the phase transformation theories involving nucleation of a product phase from a heterogeneous multi-phase microstructure typical to that of additively manufactured samples is explored. The Al-10Ce-8Mn (wt%) was used as a model alloy system. Under additive manufacturing conditions different solidification microstructures were obtained due to spatial and temporal variations of thermal gradients (G) and liquid-solid interface velocities (R) within a given melt pool. Near the melt pool boundary (high G and low R, referred as MPB region), initially, Al 20 Mn 2 Ce forms from the liquid followed by a eutectic of FCC Al and Al 11 Ce 3 . In the melt pool interiors (low G and high R referred as ES region) a eutectic structure between FCC Al and Al 20 Mn 2 Ce is observed. During subsequent heat treatments, the MPB and ES regions transform into different sets of microstructures. In the MPB region, a fine globular microstructure containing FCC Al, Al 11 Ce 3 , Al 6 Mn, and Al 12 Mn results from the decomposition of Al 20 Mn 2 Ce. In the ES region a faceted Al 51 Mn 7 Ce 4 plate phase results from the decomposition of Al 20 Mn 2 Ce. The formation of the Al 51 Mn 7 Ce 4 phase within the eutectic microstructure at the boundaries of FCC Al and Al 20 Mn 2 Ce has not been reported in the literature. Further, these two distinct phase transformation pathways are rationalized based on the role of driving force on the nucleation of (Al 6 Mn) and/or metastable intermetallic (Al 51 Mn 7 Ce 4 ) phases at the interface of aluminum (FCC) and the non-equilibrium intermetallic (Al 20 Mn 2 Ce) phases.

36 MATERIALS SCIENCE↗