Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “model discovery”

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 307 records · Page 17
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Validation and parameterization of a novel physics-constrained neural dynamics model applied to turbulent fluid flow

We report, in fluid physics, data-driven models to enhance or accelerate time to solution are becoming increasingly popular for many application domains, such as alternatives to turbulence closures, system surrogates, or for new physics discovery. In the context of reduced order models of high-dimensional time-dependent fluid systems, machine learning methods grant the benefit of automated learning from data, but the burden of a model lies on its reduced-order representation of both the fluid state and physical dynamics. In this work, we build a physics-constrained, data-driven reduced order model for Navier–Stokes equations to approximate spatiotemporal fluid dynamics in the canonical case of isotropic turbulence in a triply periodic box. The model design choices mimic numerical and physical constraints by, for example, implicitly enforcing the incompressibility constraint and utilizing continuous neural ordinary differential equations for tracking the evolution of the governing differential equation. We demonstrate this technique on a three-dimensional, moderate Reynolds number turbulent fluid flow. In assessing the statistical quality and characteristics of the machine-learned model through rigorous diagnostic tests, we find that our model is capable of reconstructing the dynamics of the flow over large integral timescales, favoring accuracy at the larger length scales. More significantly, comprehensive diagnostics suggest that physically interpretable model parameters, corresponding to the representations of the fluid state and dynamics, have attributable and quantifiable impact on the quality of the model predictions and computational complexity.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2020

Enlarging the scope of resonant di-Higgs searches: hunting for Higgs-to-Higgs cascades in 4b final states at the LHC and future colliders

We extend the coverage of resonant di-Higgs searches in the $b\bar{b}b\bar{b}$ final state to the process $pp → H_1 → H_2H_2 → b\bar{b}b\bar{b}$, where both $H_{1,2}$ are spin-0 states beyond the Standard Model. Such a process constitutes a joint discovery mode for the new states $H_1$ and $H_2$. We present the first sensitivity study of this channel, using public LHC data to validate our analysis. We also provide a first estimate of the sensitivity of the search for the HL-LHC and future facilities like the HE-LHC and FCC-hh. We analyze the discovery potential of this search for several non-minimal scalar sector scenarios: an extension of the SM with two extra singlet scalar fields, the two-Higgs-doublet model and a two-Higgs doublet model plus a singlet, which captures the scalar potential features of the NMSSM. We find that this channel represents a novel, very powerful probe for extended Higgs sectors, offering complementary sensitivity to existing analyses.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRSDate not supplied

NASA biological and physical sciences databases: who’s the FAIRest of them all?

Conceptual models are a key part of the foundation of scientific study. Scientific data discovery and retrieval are often inaccurate and incomplete because these models are not sufficiently well-incorporated into data retrieval systems. Systems often don’t provide the necessary tools to those producing scientific data to fully and unambiguously annotate them and the result is consumers of the data cannot find them efficiently. The capability of data archives to provide these tools to link data to underlying conceptual models is one of dimensions of the recently developed “FAIR” principles (https://www.go-fair.org/fair-principles/ ), and is key to many automated processes being able to operate on these data, particularly analytics involving artificial intelligence. We used an open-source web service to measure the FAIR compliance of the three data archives operated by NASA for the biological and physical sciences: the Life Sciences Data Archive, the Physical Sciences Informatics database, and GeneLab. The service ingests references to data sets in these archives, and then executes domain-non-specific examinations of these data and metadata that test compliance to the FAIR principles. Of the 22 metrics tested, GeneLab passed 11 (50%), and PSI and LSDA each passed 7 (32%). These data were gathered using only one representative data set from each archive and we anticipate variability in results as we continue to apply these metrics to other data. A preliminary study of the failure traces for each metric suggests there is a wide range of effort and complexity in the enhancements required for each system to elevate FAIR compliance, and this is the subject of continued investigation. This information has been and will likely continue to be important information in planning these enhancements, with the goal of increased readiness of the data for automated processes.

database↗
DOE OSTI2022

Discovering mechanisms for materials microstructure optimization via reinforcement learning of a generative model

Abstract The design of materials structure for optimizing functional properties and potentially, the discovery of novel behaviors is a keystone problem in materials science. In many cases microstructural models underpinning materials functionality are available and well understood. However, optimization of average properties via microstructural engineering often leads to combinatorically intractable problems. Here, we explore the use of the reinforcement learning (RL) for microstructure optimization targeting the discovery of the physical mechanisms behind enhanced functionalities. We illustrate that RL can provide insights into the mechanisms driving properties of interest in a 2D discrete Landau ferroelectrics simulator. Intriguingly, we find that non-trivial phenomena emerge if the rewards are assigned to favor physically impossible tasks, which we illustrate through rewarding RL agents to rotate polarization vectors to energetically unfavorable positions. We further find that strategies to induce polarization curl can be non-intuitive, based on analysis of learned agent policies. This study suggests that RL is a promising machine learning method for material design optimization tasks, and for better understanding the dynamics of microstructural simulations.

36 MATERIALS SCIENCE↗
NASA NTRS1997

Observational Search for Cometary Aging Processes

The scientific objectives of this study were (i) to search for physical differences in the behavior of the dynamically new comets (those which are entering the solar system for the first time from the Oort cloud) and the periodic comets, and (ii) to interpret these differences, if any, in terms of the physical and chemical nature of the comets and the evolutionary histories of the two comet groups. Because outer solar system comets may be direct remnants of the planetary formation processes, it is clear that the understanding of both the physical characteristics of these bodies at the edge of the planet forming zone and of their activity at large heliocentric distances, r, will ultimately provide constraints on the planetary formation process both in our Solar System and in extra-solar planetary systems. A combination of new solar system models which suggest that the protoplanetary disk was relatively massive and as a consequence comets could form at large distances from the sun (e.g. from the Uranus-Neptune region to the vicinity of the Kuiper belt), observations of activity in comets at large r, and laboratory experiments on low temperature volatile condensation, are dramatically changing our understanding of the chemical'and physical conditions in the early solar nebula. In order to understand the physical processes driving the apparent large r activity, and to address the question of possible physical and chemical differences between periodic, non-periodic and Oort comets, the PI has been undertaking a long-term study of the behavior of a significant sample of these comets (approximately 50) over a wide range of r to watch the development, disappearance and changing morphology of the dust coma. The ultimate goal is to search for systematic physical differences between the comet classes by modelling the coma growth in terms of volatile-driven activity. The systematic observations for this have been ongoing since 1986, and have been obtained over the course of approximately 300 nights using the telescopes on Mauna Kea, Kitt Peak, Cerro Tololo, the European Southern Observatory, and several other large aperture facilities. A greater than 2 TB database of broad band comet images has been obtained which follows the systematic development and fading of the cometary coma for the comets in the database. The results to date, indicate that there is a substantial difference in the brightness and the amount of dust as a function of r between the two comet classes. In addition to this major finding, the program has been responsible for several exciting discoveries, including: the P/Halley outburst at r = 14.3 AU, the discovery of Chiron's coma and modelling and observations of the gravitationally bound component, observational evidence that activity continues out beyond r = 17 AU for many dynamically new comets

Meech, Karen J.↗
DOE OSTI2024

Hierarchical Conditioning of Diffusion Models Using Tree-of-Life for Studying Species Evolution

A central problem in biology is to understand how organisms evolve and adapt to their environment by acquiring variations in the observable characteristics or traits of species across the tree of life. With the growing availability of large-scale image repositories in biology and recent advances in generative modeling, there is an opportunity to accelerate the discovery of evolutionary traits automatically from images. Toward this goal, we introduce Phylo-Diffusion, a novel framework for conditioning diffusion models with phylogenetic knowledge represented in the form of HIERarchical Embeddings (HIER-Embeds). We also propose two new experiments for perturbing the embedding space of Phylo-Diffusion: trait masking and trait swapping, inspired by counterpart experiments of gene knockout and gene editing/swapping. Our work represents a novel methodological advance in generative modeling to structure the embedding space of diffusion models using tree-based knowledge. Our work also opens a new chapter of research in evolutionary biology by using generative models to visualize evolutionary changes directly from images. We empirically demonstrate the usefulness of Phylo-Diffusion in capturing meaningful trait variations for fishes and birds, revealing novel insights about the biological mechanisms of their evolution. (Model and code can be found at imageomics.github.io/phylo-diffusion)

Khurana, Mridul↗
DOE OSTI2025

Machine Learning Thermodynamics And Kinetics of Defects For Accelerated Materials Discovery

Atomistic defects play a pivotal role in functional and structural materials’ performance across a myriad of technology applications. Quantitative prediction of the thermodynamics and kinetics of defect formation and migration, respectively, typically requires accurate but expensive first-principles approaches, such as density functional theory (DFT). Their computational expense limits the throughput needed to perform high-throughput materials discovery/screening exercises or to perform materials modeling tasks relying on extensive sampling techniques. Therefore, in this Sandia National Laboratories Laboratory Directed Research and Development (LDRD) project (Project #229366), we developed a variety of machine learning techniques, trained on density functional theory calculations, to accelerate the discovery and modeling of materials in which vacancy and interstitial defects primarily dictate material performance. These include applications such as metal oxides for water-splitting or mixed ionic-electronic conduction, metal hydrides for hydrogen storage, and transition metal dichalcogenides for electronics, and the approaches developed herein can further be applied to many other domains that similarly depend on materials’ thermodynamic and kinetic defect properties for their desired functionality.

36 MATERIALS SCIENCE↗
DOE OSTI2026

SetGo: Metadata Readiness for Scientific AI Datasets

Scientific datasets intended for AI use require both computational readiness for model training and metadata readiness for discovery, sharing, and reuse. The Readiness Engine for Data Integration (REDI) addresses computational readiness, but no corresponding tool evaluates whether a dataset’s metadata are sufficiently complete, governed, and standards-compliant for publication and agent-based consumption. Existing FAIR assessors operate only on published repository records, and no single system covers FAIR compliance, licensing, provenance, governance, reproducibility, and catalog readiness together. We present SetGo, an open-source Python toolkit that assesses and repairs metadata readiness across these six dimensions before a dataset is published or archived. Applied to four scientific corpora, SetGo surfaces deficiencies that general-purpose tools do not detect: ERA5 climate metadata scores 4% on ACDD 1.3 compliance; materials datasets fail OPTIMADE species-definition requirements; and PDB-derived proteomics data carries licensing terms incompatible with standard SPDX identifiers. Guided enrichment raises overall FAIR scores from 52–57% to 81–91%, and a single setgo publish command pushes to Hugging Face Hub, CKAN, or OpenMetadata with ML Commons Croissant 1.0 metadata sidecars. To support interactive and automated workflows, SetGo integrates with coding agents powered by large language models (LLMs) through a /setgo skill that enables natural-language execution of the full assess–enrich–publish loop, with user involvement limited to supplying missing metadata values.

Wilkinson, Sean [ORNL] (ORCID:0000000214437479)↗
DOE OSTI2021

Quasielastic electromagnetic scattering cross sections and world data comparisons in the GENIE Monte Carlo event generator

The usage of Monte Carlo neutrino event generators (MC$\textit{ν}$EGs) is a norm within the high-energy $\textit{ν}$ scattering community. The relevance of quasielastic (QE) energy regimes to $\textit{ν}$ oscillation experiments implies that accurate calculations of $\textit{ν}$A cross sections in this regime will be a key contributor to reducing the systematic uncertainties affecting the extraction of oscillation parameters. In spite of this, many MC$\textit{ν}$EGs utilize highly phenomenological, parametrized models of QE scattering cross sections. Moreover, a culture of validation of MC$\textit{ν}$EGs against prolific electron ($\textit{e}$) scattering data has been historically lacking. In this work, we implement new eA cross sections obtained from nuclear ab initio approaches in GENIE, the primary MC$\textit{ν}$EG utilized by the FNAL community. In particular, we utilize results from quantum MC methods which solve the many-body nuclear problem in the short-time approximation (STA), allowing consistent retention of two-nucleon dynamics which are crucial to explain available nuclear electromagnetic (electroweak) data over a wide range of energy and momentum transfers. This new implementation in GENIE is fully tested against the world QE electromagnetic data, finding agreement with available data below ~2 GeV of beam energy with the aid of a scaling function formalism. The STA is currently limited to study $\textit{A}$ ≤ 12 nuclei, however, its semi-inclusive multibody identity components are exportable to other many-body computational techniques such as auxiliary field diffusion MC which can reach $\textit{A}$ ≤ 40 systems while continuing to realize the factorization contained within the STA’s multinucleon dynamics. Together, these developments promise to make future experiments such as DUNE more accurate in their assessment of MC$\textit{ν}$EG systematics, $\textit{ν}$ properties, and potentially empower the discovery of physics beyond the Standard Model.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Radium Dial Workers: Back to the Future

This report reviews the history of the radium dial workers in the United States, summarizes the scientific progress made since the last evaluation in the early 1990s, and discusses current progress in updating the epidemiologic cohort and applying new dosimetric models for radiation risk assessment. The discoveries of radiation and radioactivity led quickly to medical and commercial applications at the turn of the 20th century, including the development of radioluminescent paint, made by combining radium with phosphorescent material and adhesive. Workers involved with the painting of dials and instruments included painters, handlers, ancillary workers, and chemists who fabricated the paint. Dial painters were primarily women and, prior to the mid to late 1920s, would use their lips to give the brush a fine point, resulting in high intakes of radium. The tragic experience of the dial painters had a significant impact on industrial safety standards, including protection measures taken during the Manhattan Project. The dial workers study has formed the basis for radiation protection standards for intakes of radionuclides by workers and the public. The mortality experience of 3,276 radium dial painters and handlers employed between 1913-1949 is being determined through 2019. The last epidemiologic follow-up was 30 years ago when most of these workers were still alive. Nearly 65% were born before 1920, 37.5% were teenagers when first hired, and nearly 50% were hired before 1930 when the habit of placing brushes in mouths essentially stopped. Comprehensive dose reconstruction techniques are being applied to estimate organ doses for each worker related to the intake of 226Ra, 228Ra, and associated photon exposures. Time dependent dose-response analyses will estimate lifetime risks for specific causes of death. The study of radium dial workers is part of the Million Person Study of low-dose health effects that is designed to evaluate radiation risks among healthy American workers and veterans. Despite being one of the most important and influential radiation effects studies ever conducted, shifting programmatic responsibilities and declining funding led to the termination of the radium program of studies in the early 1990s. Renewed interest and opportunity have arisen. With scientific progress made in dosimetric methodology and models, the ability to perform a study over the entire life span, and the potential applicability to other scenarios such as medicine, environmental contamination and space exploration, the radium dial workers have once again come to the forefront.

61 RADIATION PROTECTION AND DOSIMETRY↗
DOE OSTI2021

Towards testing the theory of gravity with DESI: summary statistics, model predictions and future simulation requirements

Shortly after its discovery, General Relativity (GR) was applied to predict the behavior of our Universe on the largest scales, and later became the foundation of modern cosmology. Its validity has been verified on a range of scales and environments from the Solar system to merging black holes. However, experimental confirmations of GR on cosmological scales have so far lacked the accuracy one would hope for — its applications on those scales being largely based on extrapolation and its validity there sometimes questioned in the shadow of the discovery of the unexpected cosmic acceleration. Future astronomical instruments surveying the distribution and evolution of galaxies over substantial portions of the observable Universe, such as the Dark Energy Spectroscopic Instrument (DESI), will be able to measure the fingerprints of gravity and their statistical power will allow strong constraints on alternatives to GR. In this paper, based on a set of N-body simulations and mock galaxy catalogs, we study the predictions of a number of traditional and novel summary statistics beyond linear redshift distortions in two well-studied modified gravity models — chameleon f(R) gravity and a braneworld model — and the potential of testing these deviations from GR using DESI. These summary statistics employ a wide array of statistical properties of the galaxy and the underlying dark matter field, including two-point and higher-order statistics, environmental dependence, redshift space distortions and weak lensing. We find that they hold promising power for testing GR to unprecedented precision. The major future challenge is to make realistic, simulation-based mock galaxy catalogs for both GR and alternative models to fully exploit the statistic power of the DESI survey (by matching the volumes and galaxy number densities of the mocks to those in the real survey) and to better understand the impact of key systematic effects. Using these, we identify future simulation and analysis needs for gravity tests using DESI.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Intricate short-range order in GeSn alloys revealed by atomistic simulations with highly accurate and efficient machine-learning potentials

GeSn alloys hold promise for silicon-compatible integrated applications in electronics, photonics, and topological quantum devices. However, understanding their intricate structures using density functional theory (DFT) calculations is hindered by spatiotemporal constraints. To overcome this limitation, we develop highly accurate and efficient machine-learning interatomic potentials based on a neuroevolution potential approach with farthest point sampling on a comprehensive DFT data set. The application of the developed machine-learning potential in large-scale atomistic simulations bridges the spatiotemporal gap between modeling and advanced characterizations, and facilitates the discovery of structural intricacies in GeSn alloys. Through extensive statistical sampling, we identify a type of short-range order (SRO) that is distinguished by both its structural signature and electronic band gap from the SRO structure previously predicted. Modeling based on a large simulation cell reveals the coexistence of nano SRO domains with various degrees of ordering, demonstrating a complex spatial heterogeneity of SRO structure. Finally, our study not only reinforces the significance of fine-level structural information in alloys, but it also constitutes an effective framework for exploring SRO in a broad range of complex alloys based on highly accurate and effective machine-learning potentials.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Cryo2StructData: A Large Labeled Cryo-EM Density Map Dataset for AI-based Modeling of Protein Structures

The advent of single-particle cryo-electron microscopy (cryo-EM) has brought forth a new era of structural biology, enabling the routine determination of large biological molecules and their complexes at atomic resolution. The high-resolution structures of biological macromolecules and their complexes significantly expedite biomedical research and drug discovery. However, automatically and accurately building atomic models from high-resolution cryo-EM density maps is still time-consuming and challenging when template-based models are unavailable. Artificial intelligence (AI) methods such as deep learning trained on limited amount of labeled cryo-EM density maps generate inaccurate atomic models. To address this issue, we created a dataset called Cryo2StructData consisting of 7,600 preprocessed cryo-EM density maps whose voxels are labelled according to their corresponding known atomic structures for training and testing AI methods to build atomic models from cryo-EM density maps. Cryo2StructData is larger than existing, publicly available datasets for training AI methods to build atomic protein structures from cryo-EM density maps. We trained and tested deep learning models on Cryo2StructData to validate its quality showing that it is ready for being used to train and test AI methods for building atomic models.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Machine Learning Modeling for Accelerated Battery Materials Design in the Small Data Regime

Abstract Machine learning (ML)‐based approaches to battery design are relatively new but demonstrate significant promise for accelerating the timeline for new materials discovery, process optimization, and cell lifetime prediction. Battery modeling represents an interesting and unconventional application area for ML, as datasets are often small but some degree of physical understanding of the underlying processes may exist. This review article provides discussion and analysis of several important and increasingly common questions: how ML‐based battery modeling works, how much data are required, how to judge model performance, and recommendations for building models in the small data regime. This article begins with an introduction to ML in general, highlighting several important concepts for small data applications. Previous ionic conductivity modeling efforts are discussed in depth as a case study to illustrate these modeling concepts. Finally, an overview of modeling efforts in major areas of battery design is provided and several areas for promising future efforts are identified, within the context of typical small data constraints.

Sendek, Austin D.↗
DOE OSTI2023

Predicting the synthesizability of crystalline inorganic materials from the data of known material compositions

Abstract Reliably identifying synthesizable inorganic crystalline materials is an unsolved challenge required for realizing autonomous materials discovery. In this work, we develop a deep learning synthesizability model ( SynthNN ) that leverages the entire space of synthesized inorganic chemical compositions. By reformulating material discovery as a synthesizability classification task, SynthNN identifies synthesizable materials with 7× higher precision than with DFT-calculated formation energies. In a head-to-head material discovery comparison against 20 expert material scientists, SynthNN outperforms all experts, achieves 1.5× higher precision and completes the task five orders of magnitude faster than the best human expert. Remarkably, without any prior chemical knowledge, our experiments indicate that SynthNN learns the chemical principles of charge-balancing, chemical family relationships and ionicity, and utilizes these principles to generate synthesizability predictions. The development of SynthNN will allow for synthesizability constraints to be seamlessly integrated into computational material screening workflows to increase their reliability for identifying synthetically accessible materials.

36 MATERIALS SCIENCE↗
DOE OSTI2023

MArVD2: a machine learning enhanced tool to discriminate between archaeal and bacterial viruses in viral datasets

Abstract Our knowledge of viral sequence space has exploded with advancing sequencing technologies and large-scale sampling and analytical efforts. Though archaea are important and abundant prokaryotes in many systems, our knowledge of archaeal viruses outside of extreme environments is limited. This largely stems from the lack of a robust, high-throughput, and systematic way to distinguish between bacterial and archaeal viruses in datasets of curated viruses. Here we upgrade our prior text-based tool (MArVD) via training and testing a random forest machine learning algorithm against a newly curated dataset of archaeal viruses. After optimization, MArVD2 presented a significant improvement over its predecessor in terms of scalability, usability, and flexibility, and will allow user-defined custom training datasets as archaeal virus discovery progresses. Benchmarking showed that a model trained with viral sequences from the hypersaline, marine, and hot spring environments correctly classified 85% of the archaeal viruses with a false detection rate below 2% using a random forest prediction threshold of 80% in a separate benchmarking dataset from the same habitats.

Vik, Dean (ORCID:000000027546899X)↗