SEARCH · Search NASA

Results for “grey-box predictive energy model”

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.

1,692 records · Page 5

All sources NASA NTRS DOE OSTI

DOE OSTI2026

High-resolution modeling of indoor radon exposure with uncertainty quantification in Utah

Indoor radon accounts for 37% of population-level exposure to ionizing radiation in the United States. However, radon metrics are typically reported at coarse spatial scales, potentially obscuring meaningful local variation. We developed a high-resolution modeling framework to estimate indoor radon concentrations across Utah while explicitly quantifying predictive uncertainty. A total of 19,497 residential radon measurements collected between 2006 and 2017 were combined with environmental and housing characteristics and analyzed using a geospatial neural network that accommodates spatial dependence and nonlinear associations. Predictions were generated on a uniform hexagonal grid at 0.73 km2 resolution (H3 level 8). Out-of-sample predictions aggregated to the H3 level 8 grid showed good agreement with observed concentrations (Pearson r=0.64), while household-level predictions exhibited more moderate agreement (r=0.45). The model produced well-calibrated uncertainty estimates, with 24.1% of held-out observations exceeding the predicted 75th-percentile threshold. Maps of predicted radon concentrations and the probability of exceeding the U.S. EPA action level of 148 Bq/m3 (4 pCi/L) revealed substantial fine-scale spatial heterogeneity that was not apparent in conventional coarse-resolution summaries, with greater local variability observed in densely monitored urban counties than in sparsely sampled regions. High-resolution radon models that explicitly quantify uncertainty provide a useful framework for characterizing the spatial distribution of indoor radon and identifying areas of elevated exceedance risk. These findings highlight the value of fine-scale monitoring data and uncertainty-aware modeling approaches for radon exposure assessment, environmental risk characterization, and radon-related health research.

Wu, Yunhan [ORNL] (ORCID:0000000178842994)↗

DOE OSTI2025

Machine Learning for Predicting Multipactor Susceptibility in Planar RF Structures

Multipactor discharge is a persistent challenge in high-power microwave (HPM) and accelerator systems, where secondary electron avalanches can cause heating, vacuum degradation, and failure. This work presents the first supervised machine learning (ML) framework for multipactor prediction, trained on high-fidelity 3D Particle-in-Cell (PIC) simulation data in planar geometries. The model maps operational, geometric, and material-dependent secondary electron yield (SEY) parameters to the time-averaged electron growth rate, enabling rapid reconstruction of susceptibility charts. Among the models evaluated, tree-based ensemble methods such as Random Forest and Extra Trees demonstrate superior generalization to unseen materials compared to neural networks such as multilayer perceptron (MLP). Performance metrics, including Intersection over Union (IoU), Structural Similarity Index Measure (SSIM), and Pearson correlation, show close agreement with simulation benchmarks. Principal Component Analysis attributes generalization limits to material feature-space disjointedness.

43 PARTICLE ACCELERATORS↗

DOE OSTI2025

Micrometer: Micromechanics transformer for predicting full field mechanical responses of heterogeneous materials

Predicting mechanical responses of heterogeneous materials across scales remains a significant challenge. Traditional computational methods often struggle with complex and multiscale nature of these materials, limiting their effectiveness in real-world applications. Here, in this paper, we introduce Micrometer, a vision transformer based deep learning model designed to predict full field mechanical responses of heterogeneous materials, bridging the gap between computer vision and solid mechanics problems. We show that Micrometer, trained on a large-scale high-resolution dataset of 2D fiber-reinforced composites, can achieve state-of-the-art performance in predicting microscale strain fields across a wide range of material properties and loading conditions. Our model demonstrates accuracy and computational efficiency in applications such as computational homogenization and multiscale modeling, reducing computational time by up to two orders of magnitude compared to conventional numerical solvers while maintaining less than 1 % errors in predicting macroscale stress fields. Furthermore, we showcase Micrometer’s adaptability through transfer learning experiments on new materials with limited data, highlighting its potential to tackle diverse scenarios in computational solid mechanics. These results represent a significant step towards AI-driven innovation in materials science, addressing the limitations of traditional numerical methods and paving the way for more efficient simulations of heterogeneous materials across various industrial applications.

Composite materials↗

DOE OSTI2026

Sparse non-Markovian Noise Modeling of Transmon-Based Multi-Qubit Operations

The influence of noise on quantum dynamics is one of the main factors preventing current quantum processors from performing accurate quantum computations. Sufficient noise characterization and modeling can provide key insights into the effect of noise on quantum algorithms and inform the design of targeted error protection protocols. However, constructing effective noise models that are sparse in model parameters, yet predictive can be challenging. In this work, we present an approach for effective noise modeling of multi-qubit operations on transmon-based devices. Through a comprehensive characterization of seven devices offered by the IBM Quantum Platform, we show that the model can capture and predict a wide range of single- and two-qubit behaviors, including non-Markovian effects resulting from spatiotemporally correlated noise sources. The model’s predictive power is further highlighted through multi-qubit dynamical decoupling demonstrations and an implementation of the variational quantum eigensolver. As a training proxy for the hardware, we show that the model can predict expectation values within a relative error of 0.5%; this is a sevenfold improvement over default hardware noise models. Through these demonstrations, we highlight key error sources in superconducting qubits and illustrate the utility of reduced noise models for predicting hardware dynamics.

open quantum systems & decoherence↗

DOE OSTI2025

High-Fidelity Building Emulator for Integrated Comfort and Energy Analysis using EnergyPlus and Radiance

The growing need for smart, energy-efficient, and occupant-centric buildings has created a demand for advanced control systems that can optimize building operations to balance energy savings, demand flexibility, and comfort. However, current building energy simulation tools, such as EnergyPlus, have limitations that hinder the development and evaluation of these complex control systems. To address this challenge, we introduce a high-fidelity building emulator that dynamically couples EnergyPlus with Radiance for enhanced daylight modeling. The introduced workflow allows researchers and practitioners to rapidly develop and evaluate innovative control solutions. An example study looking at a south-facing office zone revealed up to 67% deviation in predicted light levels, which can significantly impact building assessment.

Yu, Tammie↗

NASA NTRS2013

Electrochemistry-based Battery Modeling for Prognostics

Batteries are used in a wide variety of applications. In recent years, they have become popular as a source of power for electric vehicles such as cars, unmanned aerial vehicles, and commericial passenger aircraft. In such application domains, it becomes crucial to both monitor battery health and performance and to predict end of discharge (EOD) and end of useful life (EOL) events. To implement such technologies, it is crucial to understand how batteries work and to capture that knowledge in the form of models that can be used by monitoring, diagnosis, and prognosis algorithms. In this work, we develop electrochemistry-based models of lithium-ion batteries that capture the significant electrochemical processes, are computationally efficient, capture the effects of aging, and are of suitable accuracy for reliable EOD prediction in a variety of usage profiles. This paper reports on the progress of such a model, with results demonstrating the model validity and accurate EOD predictions.

battery↗

NASA NTRS1992

Monte Carlo modeling of atomic oxygen attack of polymers with protective coatings on LDEF

Characterization of the behavior of atomic oxygen interaction with materials on the Long Duration Exposure Facility (LDEF) will assist in understanding the mechanisms involved, and will lead to improved reliability in predicting in-space durability of materials based on ground laboratory testing. A computational simulation of atomic oxygen interaction with protected polymers was developed using Monte Carlo techniques. Through the use of assumed mechanistic behavior of atomic oxygen and results of both ground laboratory and LDEF data, a predictive Monte Carlo model was developed which simulates the oxidation processes that occur on polymers with applied protective coatings that have defects. The use of high atomic oxygen fluence-directed ram LDEF results has enabled mechanistic implications to be made by adjusting Monte Carlo modeling assumptions to match observed results based on scanning electron microscopy. Modeling assumptions, implications, and predictions are presented, along with comparison of observed ground laboratory and LDEF results.

Bruce A. Banks↗

NASA NTRS1989

Nickel cadmium battery performance modelling

The development of a model to predict cell/battery behavior given databases of temperature is described. The model accommodates batteries of various structural as well as thermal designs. Cell internal design modifications can be accommodated as long as the databases reflect the cell's performance characteristics. Operational parameters can be varied to simulate any number of charge or discharge methods under any orbital regime. The flexibility of the model stems from the broad scope of input variables and allows the prediction of battery performance under simulated mission or test conditions.

Clark, K.↗

DOE OSTI2026

Solar Heat for Industrial Processes: Integration with Chemical Reactors

The integration of solar thermal systems with chemical reactors has been proposed as part of a larger effort to develop and deploy solar heat for industrial processes (SHIP) technologies. A strong motivation for SHIP processes and technologies is the potential for high thermal efficiency coupled with low-cost thermal energy storage (TES) which can enable commercial deployment of such systems. While there are different ways to categorize SHIP technologies, one important such distinction is between directly irradiated systems and indirect off-sun process driven by a SHIP system. While directly irradiated systems can provide high thermal efficiencies and high fluxes, they usually require complex engineering solutions due to the need for redesigning the established processes and unit operations. In most cases, it is also more challenging to couple such a process to a TES system, losing some of the benefits of SHIP. On the other hand, using a SHIP system to drive an industrial process off-sun can allow better integration with existing process chains, easier TES capabilities, and potential for more applications fitting a specific SHIP technology. However, the integration of SHIP systems with the industrial processes is not fully explored in detail, especially in the case of high-temperature processes such as reforming, cracking, cement manufacturing, and iron/steelmaking. Many of these systems require heating fluxes of >50 kW/m^2, supplied via combustion of hydrocarbons in a fire box and benefitting from radiative heat transfer between the flue gases and the reaction zones. As such, using SHIP systems for such processes is more complex than providing the same thermal input in the form of a heat transfer medium (HTM) entering the reactor, kiln, or furnace. Moreover, in case convective heat transfer using SHIP is envisioned, for example using supercritical CO2 as the HTM from a particle receiver, the thermal integration might be more challenging than initially envisioned: lower heat transfer coefficients and limited approach temperature might require large flow rates, causing a mismatch between the process thermal requirements and the thermal capacity of the SHIP system. In addition, even if the heat exchange between SHIP and reactor is effective, there is still a cold leg HTM at the reaction temperature or slightly below it. Chemical plants usually include a set of heat exchangers, heat recovery steam generators, and even power generation units - in a tightly integrated design - to recover the flue gases which are eventually vented. With SHIP systems mostly operating on a closed HTM loop, bottoming the cold leg is crucial. In this talk, we will present different modeling results for a variety of syngas production reactions, using catalytic and chemical looping processes, and discuss some of the challenges and design considerations for off-sun chemical reactors using SHIP systems.

14 SOLAR ENERGY↗

NASA NTRSDate not supplied

Li-ion Battery Aging with Hybrid Physics-Informed Neural Networks and Fleet-wide Data

In this work, we propose a hybrid model for Li-ion battery discharge and aging prediction that leverages fleet-wide data to predict future capacity drops.The model is built upon an hybrid approach merging physics-based and empirical equations, as well as neural network models in a recurrent neural network cell. The hybrid physics-informed neural network can predict voltage discharge cycles given the loading profile, and estimate the used capacity of the battery under random-loading conditions by tracking aging parameters connected to the residual capacity of the battery. By merging information on the battery aging parameters with existing fleet-wide aging data, the model can predict the future residual capacity of the battery that is being monitored, and therefore enable predictions of voltage discharge curves far ahead in the battery life cycle. We validated the approach using the NASA Prognostics Data Repository Battery data-set, which contains experimental data on Li-ion batteries discharged at random loading conditions in a controlled environment. The approach also allows the identification of discrepancies between the battery aging trend and the trend observed at the fleet level, so that batteries behaving differently from the rest of the fleet can be subject to closer monitoring and further testing to refine predictions.

PINN↗

NASA NTRS1995

Effect of Turbulence Models on Criticality Conditions in Swirling Flows

The critical state of vortex cores downstream of vortex breakdown has been studied. Base vortical flows were computed using the Reynolds-averaged, axisymmetric Navier-Stokes equations. Standard K - epsilon , RNG and second-order Reynolds stress models were employed. Results indicate that the return to supercriticality is highly dependent on the turbulence model. The K - epsilon model predicted a rapid return of the vortex to supercritical conditions, the location of which showed little sensitivity to changes in the swirl ratio. The Reynolds stress model predicted that the vortex remains subcritical to the end of the domain for each of the swirl ratios employed, and provided results in qualitative agreement with experimental work. The RNG model produced intermediate results, with a downstream movement in the critical location with increasing swirl. Calculations for which area reductions were introduced at the exit in a subcritical flow were also performed using the Reynolds stress model. The structure of the resulting recirculation zone was altered significantly. However, when area reductions were employed within supercritical flows as predicted using the two-equation models, no significant influence on the recirculation zone was noted.

Robert E Spall↗

NASA NTRS2026

Model Correlation and Thermal Analysis of xEMU Boot at Lunar South Pole Temperatures

The National Aeronautics and Space Administration (NASA) Artemis program plans to send astronauts to the lunar south pole, a region of the moon that is colder than previous lunar missions and low earth orbit operations. The spacesuit boots that will be used on these missions will be directly impacted by these extremely cold temperatures (down to ~50 K). To assess the performance of the Exploration Extravehicular Mobility Unit (xEMU) lunar boot in these extreme temperatures, the boot was tested at the Jet Propulsion Lab (JPL) in the Cryogenic Ice Transfer, Acquisition Development, and Excavation Laboratory (CITADEL) thermal vacuum (TVAC) chamber. This data was used to correlate thermal models to predict operational performance of the boots on the lunar south pole. This paper documents the data analysis, thermal boot model correlation, and operational predictions conducted using data from the xEMU TVAC test. Data from the test series was used to determine expected thermal conductances within the boot and between the boot and environment. These conductances were used to correlate a Thermal Desktop model of the xEMU boot across 10 different test points which varied external temperature, internal boot ventilation flowrate, and contact pressure. The correlated model was then used to predict operational performance in the lunar south pole. While the predictions indicate promising evidence for performance of the boots at the 100K test point, there is still substantial uncertainty in performance, particularly at the 48K test point. The results of this test series and model correlation stress the importance of improved testing for characterizing the expected thermal resistance between the outside of the boot and the lunar surface.

xEMU↗

DOE OSTI2026

Sensitivity Analysis of Numerical Modeling Input Parameters on Wind Turbine Loads in Deterministic Transient Load Cases

Aero-hydro-elastic-servo numerical models used to design and analyze wind turbines are based on thousands of variable input parameters that dictate the inflow, aerodynamic, structural, and control characteristics of the system as well as sea state, hydrodynamic, and mooring characteristics for fixed-bottom and floating offshore wind turbines. Each of these parameters has some level of uncertainty, which can significantly impact the predicted loads. Understanding the uncertainty in the inputs is critical to understanding the uncertainty in the outputs. This work demonstrates a screening technique to identify which parameters ultimate loads are most sensitive to so that more focus can be given to quantifying the possible range of those parameters. This technique has been demonstrated previously for different turbine and load case types and is extended here for a floating offshore wind turbine in design load cases with transient events both in the inflow and operations. Each load case features a deterministic gust, including variations in wind speed, direction, and shear. Load cases are considered with an operating turbine as well as with prescribed fault, startup, and shutdown procedures. The study found that key input parameters with a large impact on loads include the length of the gust, the magnitude of direction change and speed in the gust, the initial wind speed, and the shape of the gust profile.

17 WIND ENERGY↗

NASA NTRSDate not supplied

Model Correlation and Thermal Analysis of xEMU Boot at Lunar South Pole Temperatures

thermal analysis↗

NASA NTRS1993

Prediction of prompt NO(x) in hydrocarbon air flames

The gas turbine industry is directing particular attention to very low NOx combustors, whether for aircraft or land based CCGT systems. These low NOx combustors frequently use liquid fuels or natural gas burning under very lean premixed conditions with air or under rich-lean conditions, although only the first case is studied here. In land based systems, diluted steam or nitrogen are sometimes injected into the combustion chambers to reduce flame temperature. The NOx emissions from such systems are the product of three chemical mechanisms which are interrelated: the hydrocarbon prompt NO, the thermal NO (extended Zeldovich mechanism), and the nitrous oxide route to NO. Formation of NO2 from NO also occurs, as well as emission of carbon monoxide and unburnt hydrocarbons. When the fuel-oxidant proportion decreases towards leaner conditions, flame temperatures are lowered, resulting in the total NO being reduced and the thermal-NO contribution greatly diminished to the benefit of the remaining two mechanisms of NO formation. While knowledge of the elementary reactions and their chemical kinetics concerning methane and simple hydrocarbons combustion has existed for a number of years, its use for computer modeling is limited to simple flow dynamics configurations. Nevertheless, understanding of such combusting flows under a wide range of experimental conditions allows for analogies or speculations with more complex actual systems. Such understanding can be achieved by means of one dimensional laminar premixed flame modeling, with a full chemical mechanism which incorporates the three routes of NO formation. Complementary to this understanding is the modeling of the actual combustion system using a full description of the fluid dynamics coupled with a reduced chemical scheme, which is then compared against the first model. The objective of this investigation is to evaluate the relative importance of the three mechanisms of NO formation in lean premixed methane-air combustion with increasing pressure using the one dimensional plug flow package PREMIX, and to test the validity of a three dimensional model with a global chemical mechanism against the one dimensional model in the atmospheric pressure case. Methane is chosen because it is the only mechanism which is reasonably well known and is a good guide to the behavior of other hydrocarbons. The mixture ratio chosen is richer than that in lean gas turbines, but the combustion of this mixture with the low preheat gives realistic gas turbine final flame temperatures. Conditions of NO2 formation are also analyzed in the one dimensional model and results extrapolated to the case of gas turbines.

Valerie Dupont↗

DOE OSTI2026

Machine Learning for Multipactor Susceptibility Prediction in Planar RF Gaps

Multipactor discharge is a nonlinear electron avalanche that limits the performance of high-power radio-frequency (RF) and vacuum electronic devices. Predicting multipactor susceptibility traditionally relies on Monte Carlo or particle-in-cell (PIC) simulations, which become computationally expensive for large parametric studies. In this work, we present a supervised machine-learning (ML) framework for prediction of multipactor susceptibility in a two-surface planar geometry. The models are trained using high-fidelity PIC simulation generated susceptibility data and learn the relationship between operational parameters, geometry, and material-dependent secondary electron emission properties. The proposed approach enables rapid reconstruction of susceptibility charts while preserving the physical structure of multipactor growth regions.

43 PARTICLE ACCELERATORS↗

DOE OSTI2026

Multitask graph neural networks for elastoplastic response prediction in dual-phase polycrystals

Microstructure-sensitive prediction of elastoplastic response remains a recurring bottleneck in multiscale damage and fatigue modeling, where large ensembles of statistically distinct polycrystals are required to quantify variability and extreme-value behavior. In this work, we develop a multitask graph neural network (GNN) surrogate that maps dual-phase ferrite–martensite polycrystal microstructures to Statistical Volume Element (SVE)-level elastoplastic Quantities of Interest (QoIs). Each SVE is represented as a grain-adjacency graph, with node features encoding phase, geometry, and crystallographic orientation, and edge features encoding relative misorientation. A message-passing graph convolution generates node embeddings, which are pooled into a graph representation and passed to a multitask regression head that jointly predicts 10 scalar QoIs and vector-valued stress–strain responses in orthogonal loading directions across multiple martensite volume fractions and SVE sizes. Results show high accuracy for scalar QoIs and strong agreement for full stress–strain trajectories, with population envelopes reproducing both median behavior and finite-SVE variability across compositions and partition scales. A unified model trained on pooled volume-fraction data preserves most within-regime accuracy relative to regime-specific models while also capturing the broader cross-regime variation reflected in the pooled test set. Distributional comparisons further demonstrate that the surrogate preserves heterogeneity under SVE partitioning, enabling statistically consistent block-wise random-field construction for mesoscale analyses. Overall, the proposed grain-graph surrogate provides a practical pathway to accelerate ensemble-based studies of SVE-level constitutive variability in dual-phase polycrystals.

Crystal plasticity↗

NASA NTRS2026

Development and Experimental Validation of a Path-Dependent Spin Forming Finite Element Model

Spin forming is an advanced manufacturing process widely used in the aerospace and defense sectors to produce lightweight, high-strength cylindrical components with tight dimensional tolerances. This study explores the applicability of the path-dependent Mechanical Threshold Stress (MTS) constitutive model by simulating the evolution of geometry, machining forces, and plastic deformation during the spin forming of a 10-mm thick 6061-O aluminum cylinder. While numerical modeling of spin forming has advanced substantially over the past decade, systematic verification and experimental validation of material models remain limited, particularly in predicting through-thickness process evolution. The MTS model, incorporating a Voce hardening rule, is employed for its ability to represent cyclic loading, rapidly varying temperature fields, and strain rates characteristic of spin forming. Numerical convergence analysis indicates discretization uncertainties between 0.3% and 9.2% for key quantities of interest. Experimental validation demonstrates that the MTS model, when implemented with a verified mesh, accurately reproduces both elastic and plastic behavior of 6061-O aluminum, predicting peak roller loads within 11–18% of measurements, geometric tolerances within 3%, and plastic strain distributions within 10% of experimental values. Collectively, these results establish a validated computational framework for predictive spin-forming simulations with quantified confidence, providing a foundation for extension to other alloys, geometries, and forming conditions.

Spin forming↗