SEARCH · Search NASA

Results for “Knowledge transfer”

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 145 records · Page 8

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Baseflow Identification via Explainable AI With Kolmogorov‐Arnold Networks

Abstract Hydrological models often involve constitutive laws that may not be optimal in every application. We propose to replace such laws with the Kolmogorov‐Arnold networks (KANs), a class of neural networks designed to identify symbolic expressions. We demonstrate KAN's potential on the problem of baseflow identification, a notoriously challenging task plagued by significant uncertainty. KAN‐derived functional dependencies of the baseflow components on the aridity index outperform their original counterparts; they demonstrate that water availability, rather than potential evapotranspiration, drives baseflow by constraining actual evapotranspiration under arid conditions. On a test set, they increase the Nash‐Sutcliffe efficiency (NSE) by 65%, decrease the root mean squared error by 29%, and increase the Kling‐Gupta efficiency by 34%. This superior performance is achieved while reducing the number of fitting parameters from three to two. Next, we use data from 378 catchments across the continental United States to refine the water‐balance equation at the mean‐annual scale. The KAN‐derived equations based on the refined water balance outperform both the current aridity index model, with up to a 105% increase in NSE, and the KAN‐derived equations based on the original water balance. While the performance of our model and tree‐based machine learning methods is similar, KANs offer the advantage of simplicity and transparency and require no specific software or computational tools. This case study focuses on the aridity index formulation, but the approach is flexible and transferable to other hydrological processes. Plain Language Summary Equations used in hydrologic model are often suboptimal, resulting in reduced prediction accuracy and efficiency. We implemented Kolmogorov‐Arnold networks (KAN), a machine learning algorithm for deriving symbolic formulations, to estimate groundwater recharge and showed that it outperforms an existing state‐of‐the‐art semi‐empirical formulation. In hydrology, Nash‐Sutcliffe efficiency (NSE), root mean squared error (RMSE), and Kling‐Gupta efficiency (KGE) are commonly used to evaluate model performance. Higher NSE and KGE values indicate better performance, while lower RMSE values are preferable. Our results show that NSE increased by 71%, RMSE decreased by 32%, and KGE improved by 25%. In addition, KAN identifies an optimal functional form and can be used to derive new analytical formulas using the prior knowledge. The KAN‐inspired equation outperformed the original formulation and reduced the fitting parameters. Furthermore, we refined the water‐balance equation at the mean‐annual scale and showed that, based on the new water‐balance equation, KAN can derive new formulations that are superior to the original aridity index formulations (up to 105% increase in NSE) and KAN‐derived equations based on the original water balance. These findings highlight the significant potential of KAN to advance the scientific understanding of a wide range of hydrologic processes. Key Points Kolmogorov‐Arnold networks (KANs) enhance interpretability of machine‐learned hydrological models KAN‐derived symbolic formulations outperform state‐of‐the‐art semi‐empirical aridity indices KAN‐identified functional form yields an analytical index with fewer fitting parameters and improved performance

baseflow↗

DOE OSTI2025

Field validation of dynamic mechanical torque measurements using fiber-optic strain sensors for geared wind turbines

Abstract Accurate knowledge of the mechanical loads of wind turbine gearboxes has become essential in modern, highly loaded gearbox designs, as maintaining or even improving gearbox reliability with increasing torque density demands is proving to be challenging. Unfortunately, the traditional method of measuring dynamic mechanical torque using strain gauges placed on the outer surface of a rotating shaft and transmitting the resulting signal is unsuitable for serial deployment due to technical and economic constraints. An alternative method based on fiber-optic strain sensors placed on the stationary outer surface of the gearbox ring gear has been proposed. Like shaft torsion, the radial deformation of the ring gear is proportionate to the rotor torque. Placing the sensors on a stationary component is a cost-effective alternative for serial implementation because the need for complex and expensive data transfer via wireless transmission or a slip ring is eliminated. In this paper, we present the results of an extensive field experiment conducted to evaluate the torque measurement accuracy of this novel sensing solution installed on the gearbox of a Gamesa G97 2-MW wind turbine at the National Renewable Energy Laboratory’s Flatirons Campus. Torque measurements derived from fiber-optic strain sensors placed on the ring gear of the planetary stage are compared to conventional torque measurements from strain gauges placed on the main shaft. Two different torque estimation data processing methods were evaluated, with the method based on operational deflection shapes providing the most accurate results with an average normalized root mean square error below 0.7% for a load revolution distribution analysis. The effect of operating conditions on the torque estimate was also investigated, and the third planet-passing operational deflection shape was found to be the least sensitive to nontorque load-related effects. The fiber-optic strain sensors’ successful operation during the complete test campaign has demonstrated a robust and accurate solution for fleet-wide enhanced gearbox remaining useful life estimation.

17 WIND ENERGY↗

DOE OSTI2025

Field Validation of Dynamic Mechanical Torque Measurements for Geared Wind Turbines

Accurate knowledge of the mechanical loads of wind turbine gearboxes has become essential in modern, highly loaded gearbox designs, as maintaining or even improving gearbox reliability with increasing torque density demands is proving to be challenging. Unfortunately, the traditional method of measuring dynamic mechanical torque using strain gauges placed on the outer surface of a rotating shaft and transmitting the resulting signal is unsuitable for serial deployment due to technical and economic constraints. An alternative method based on fiber-optic strain sensors placed on the stationary outer surface of the gearbox ring gear has been proposed. Like shaft torsion, the radial deformation of the ring gear is proportionate to the rotor torque. Placing the sensors on a stationary component is a cost-effective alternative for serial implementation because the need for complex and expensive data transfer via wireless transmission or a slip ring is eliminated. In this paper, we present the results of an extensive field experiment conducted to evaluate the torque measurement accuracy of this novel sensing solution installed on the gearbox of a Gamesa G97 2-MW wind turbine at the National Renewable Energy Laboratory's Flatirons Campus. Torque measurements derived from fiber-optic strain sensors placed on the ring gear of the planetary stage are compared to conventional torque measurements from strain gauges placed on the main shaft. Two different torque estimation data processing methods were evaluated, with the method based on operational deflection shapes providing the most accurate results with an average normalized root mean square error below 0.7% for a load revolution distribution analysis. The effect of operating conditions on the torque estimate was also investigated, and the third planet-passing operational deflection shape was found to be the least sensitive to nontorque load-related effects. The fiber-optic strain sensors' successful operation during the complete test campaign has demonstrated a robust and accurate solution for fleet-wide enhanced gearbox remaining useful life estimation.

17 WIND ENERGY↗

DOE OSTI2025

Phylogenomic discovery and engineering of nitrogen fixation into the bioenergy woody crop poplar

Biological nitrogen fixation (BNF) is a key process enabling plants in specific lineages to convert atmospheric dinitrogen (N₂) into bioavailable ammonia through symbioses with diazotrophic microbes. Expanding this capability beyond native nitrogen-fixing clades into non-nodulating crops would reduce synthetic fertilizer use, lowering energy inputs and environmental impacts in agriculture. Supported by DOE Funding Award DE-SC0018247, the NitFix project advanced foundational knowledge required to engineer root-nodule symbioses in new host species. The team generated the most comprehensive phylogenomic analysis to date of all known nodulating lineages, resolving the evolutionary history of nitrogen-fixing symbiosis and identifying core gene suites retained across nodulating taxa. Through multimodal genomics, transcriptomics, and functional analyses in Medicago truncatula and related species, the project mapped regulatory networks underlying nodule organogenesis, bacterial infection, and nitrogen-fixation efficiency. Key discoveries include the identification of conserved signaling modules for rhizobial recognition, transcription factors controlling nodule differentiation, and metabolic pathways integrating fixed nitrogen into plant growth. The project also developed enabling tools—including optimized transformation pipelines, gene-editing workflows, and imaging-based phenotyping—to accelerate engineering efforts in emerging models. Together, these results refine the mechanistic framework of symbiotic nitrogen fixation and highlight transferable components essential for rewiring these traits into non-nodulating crops.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2025

TRUST Sensors in Environments: Thermocouples (SE-TC), Release FY25

The Delivery Environments Testbeds to Reduce Uncertainty in Simulations and Tests (TRUST) project is a broad project intended to analyze simplified problems experimentally and with modeling and simulation. The purpose of analyzing these simplified problems is to extend solution methods to more complex problems, as well as understand deficiencies and gaps in knowledge of methods currently used in more complex analyses. The TRUST project encompasses several smaller testbeds intended to isolate individual phenomena. The testbed under consideration in this report is the Sensors in Environments: Thermocouples testbed. In previous years, the purpose of this testbed was to quantify uncertainty of thermocouple sensors. To accomplish this, an aluminum plate was placed in a thermal chamber and subject to various types of thermal loading. Thermocouples were placed in various locations on the aluminum plate in various configurations (e.g., embedded in the plate, placed under Kapton tape), and an effort was made to quantify uncertainty in these measurements. Finite element simulations were performed to investigate how sensitive these measurements were to parameters such as the boundary conditions on the plate and material properties. However, a fundamental source of uncertainty in this analysis was the convective heat transfer from the plate. Convective heat transfer is a complex physical phenomenon comprised of a number of interacting sub-processes and is difficult to predict accurately a priori. As such, the main purpose of this testbed in FY25 was to better understand, both experimentally and numerically, the convective heat transfer from the plate. This is a highly applicable problem to several more complex problems, as convective heat transfer occurs in nearly all problems where a body is moving through air. Numerically, this required a two-step approach. First, the air flow in the thermal chamber was in vestigated using computational fluid dynamics. The commercial solver Fluent was used to perform these simulations. From these simulations, a heat transfer coefficient over the surface of the plate was calculated. This heat transfer was then used as boundary conditions for finite element heat transfer simulations within the plate, which were performed using Abaqus. Significant effort was devoted to automating the handoff between these two solvers. Experimentally, previous thermocouple results in the plate were used to validate the time-dependent thermal profiles produced from Abaqus. Further experimental efforts were performed both to help validate the Fluent simulations and to inform its boundary conditions. For example, hot-wire anemometers were used to measure the velocity in the chamber, which would be particularly useful in understanding the chamber inlet velocity. Thermocouple measurements were also taken in the chamber, instead of only on the plate, to serve as validation evidence for the Fluent simulations. Numerical results showed that the Fluent to Abaqus workflow matched previous plate thermocouple measurements well. This type of handoff is useful for more complex experiments, or those that are not able to be examined in as great of detail as this testbed, as it was performed without any experimental input. Experimental results, however, were more mixed. The anemometers proved unreliable, with inconsistent measurements across all anemometers, even at locations that were nearly identical. On the other hand, the thermocouples provided a relatively rich view of the temperature field in the chamber.

42 ENGINEERING↗

DOE OSTI2025

Support for the 32 nd Annual Meeting of the Inter-American Photochemical Society

This award offset the registration and travel costs for researchers from U.S. institutions to attend the 32 nd Inter-American Photochemical Society meeting held on Jan 2-5, 2025, in Miramar Beach, Florida. The Inter-American Photochemical Society (I-APS) is a non-profit organization dedicated to promoting and disseminating knowledge and encouraging the development of fundamental and allied photochemistry throughout the Americas. The I-APS annual meeting is a particularly important tool for fulfilling this goal as it provides a welcoming venue to highlight research progress in theoretical, experimental, and applied aspects of photochemistry and molecular photophysics by researchers in all stages of their careers. The participation of graduate students and postdocs at this conference is particularly important as it aids in training the next generation of scientists in photochemistry. This award provided travel awards to three Assistant Professors and fourteen graduate and two undergraduate students to attend the I-APS Meeting in 2025. A majority of the science presented/discussed at the I-APS meeting is in direct alignment with the mission of the DOE and the Solar Photochemistry program, including studying fundamental phenomena, encompassing light absorption, photoinduced energy/electron transfer, charge separation, and charge and energy transport across mesoscale dimensions, as well as their application in generating electricity and chemical bonds using light.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2025

Bayesian prior construction for uncertainty quantification in first-principles statistical mechanics

First-principles statistical mechanics enables the prediction of thermodynamic and kinetic properties of materials, but is computationally expensive. Many approaches require surrogate models to calculate energies within Monte Carlo or molecular dynamics simulations. Inexpensive surrogates such as cluster expansions enable otherwise intractable calculations by interpolating data from higher accuracy methods, such as Density Functional Theory (DFT). Surrogate models introduce uncertainty into downstream calculations, in addition to any uncertainty inherent to DFT calculations. Bayesian frameworks address this by quantifying uncertainty and incorporating expert knowledge through priors. However, constructing effective priors remains challenging. This work introduces and describes practical strategies for building Bayesian cluster expansions, focusing on basis truncation, hyperparameter selection, and ground state replication. We analyze multiple basis truncation schemes, compare cross-validation to the evidence-approximation for hyperparameter optimization, and provide methods to find and enforce ground-state-preserving models through priors. Additionally, we compare the uncertainties between different approximations to DFT (LDA, PBE, SCAN) against the uncertainty introduced with the use of cluster expansion surrogate models. These approaches are demonstrated on the BCC Li x Mg 1-x and Li x Al 1-x alloys, which are both of interest for solid-state Li batteries. Our results provide guidelines for constructing and utilizing Bayesian cluster expansions, thereby improving the transparency of materials modeling. Furthermore, the approaches and insights developed in this work can be transferred to a wide range of cluster expansion surrogate models, including the atomic cluster expansion and related machine-learned interatomic potential architectures.

Alloy theory↗

DOE OSTI2025

CFD modeling of turbulent air flow in self-heated gyroid TPMS structures: Thermal-hydraulic performance and validation

The application of mathematically derived geometries, such as triply periodic minimal surface (TPMS) lattices, has garnered significant interest across various fields, including the nuclear sector, due to their superior thermal-hydraulic characteristics for heat transfer compared to traditional plain or finned tubes. Here, this study validates a computational fluid dynamics (CFD) model, evaluates different turbulence models and CFD model settings, and performs uncertainty quantification to provide a comprehensive analysis. Despite extensive research on CFD modeling of TPMS lattices, such as gyroid and diamond geometries, there is a notable lack of publicly available literature providing comprehensive details on numerical analysis aspects, including convergence and methodological best practices. This study embarks on a benchmark analysis of a gyroid geometry to evaluate its thermal-hydraulic performance under turbulent flow conditions and scrutinize various CFD model configurations. The main contributions of this work include validating the CFD model, assessing and comparing different turbulence models, and enhancing pressure drop and temperature prediction capabilities. The results aim to support the development of methodologies needed to benchmark and enhance numerical analysis techniques for TPMS lattices. This work seeks to complement the existing body of knowledge, support the development of TPMS reactor concepts, and improve best practices for CFD modeling of TPMS lattices, ultimately advancing methodologies to support future applications in this domain.

22 - GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2024

Coastal bacteria and protists assimilate viral carbon and nitrogen

Abstract Free viruses are the most abundant type of biological particles in the biosphere, but the lack of quantitative knowledge about their consumption by heterotrophic protists and bacterial degradation has hindered the inclusion of virovory in biogeochemical models. Using isotope-labeled viruses added to three independent microcosm experiments with natural microbial communities followed by isotope measurements with single-cell resolution and flow cytometry, we quantified the flux of viral C and N into virovorous protists and bacteria and compared the loss of viruses due to abiotic vs biotic factors. We found that some protists can obtain most of their C and N requirements from viral particles and that viral C and N get incorporated into bacterial biomass. We found that bacteria and protists were responsible for increasing the daily removal rate of viruses by 33% to 85%, respectively, compared to abiotic processes alone. Our laboratory incubation experiments showed that abiotic processes removed roughly 50% of the viruses within a week, and adding biotic processes led to a removal of 83% to 91%. Our data provide direct evidence for the transfer of viral C and N back into the microbial loop through protist grazing and bacterial breakdown, representing a globally significant flux that needs to be investigated further to better understand and predictably model the C and N cycles of the hydrosphere.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2025

TPCpp-10M: Simulated proton-proton collisions in a time projection chamber for AI foundation models

Scientific foundation models hold great promise for advancing nuclear and particle physics by improving analysis precision and accelerating discovery. Yet, progress in this field is often limited by the lack of openly available large scale datasets, as well as standardized evaluation tasks and metrics. Furthermore, the specialized knowledge and software typically required to process particle physics data pose significant barriers to interdisciplinary collaboration with the broader machine learning community. This work introduces a large, openly accessible dataset of 10 million simulated proton-proton collisions, designed to support self-supervised training of foundation models. To facilitate ease of use, the dataset is provided in a common NumPy format. In addition, it includes 70,000 labeled examples spanning three well defined downstream tasks: track finding, particle identification, and noise tagging, to enable systematic evaluation of the foundation model's adaptability. The simulated data are generated using the Pythia Monte Carlo event generator at a center of mass energy of $\sqrt{s}$ = 200 GeV and processed with Geant4 to include realistic detector conditions and signal emulation in the sPHENIX Time Projection Chamber at the Relativistic Heavy Ion Collider, located at Brookhaven National Laboratory. This dataset resource establishes a common ground for interdisciplinary research, enabling machine learning scientists and physicists alike to explore scaling behaviors, assess transferability, and accelerate progress toward foundation models in nuclear and high energy physics. The complete simulation and reconstruction chain is reproducible with the sPHENIX software stack. All data and code locations are provided under Data Accessibility.

Data Analysis, Statistics and Probability (physics↗

DOE OSTI2026

Mark-18A Target Material Recovery Program: Initial Hot Startup

The Savannah River National Laboratory (SRNL) has been tasked by the National Nuclear Security Administration to recover highly valued isotopes from irradiated Mark-18A (Mk-18A) targets. The Savannah River Site (SRS) has sixty-five Mk-18A targets available for the recovery of the high valued materials. The Mk-18A targets were manufactured with Pu-242 then irradiated under high neutron flux in K-Reactor at the SRS from 1968 to 1978. The sixty-five Mk-18A targets are currently stored in the SRS L-Area Basin and will be removed one at a time and individually transported to SRNL. Upon receipt at SRNL, the Mk-18A target material will be removed from the confinement, dissolved, chemically separated, and calcined to a stable oxide. The flowsheet is designed to recover the plutonium as well as the trivalent actinides. The original targets were manufactured with varying quantities of Pu-242, ranging from 5.2 grams to 121.5 grams. Taking a graded approach to process start up, the lowest loaded target (FT-80-03) was selected as the first target to be received and processed at SRNL. As operational experience and knowledge is gained from processing targets, higher loaded targets will be selected for processing to increase the quantities of valuable isotopes recovered. Due to dose concerns, the receipt and processing of the Mk-18A targets is performed in the SRNL Shielded Cells Facility. The targets are stored in a double J-can confinement in the L-Area Basin. A specially designed cask was procured for transport of the targets from L-Area to SRNL. Once received at SRNL, the targets are loaded into the back of Cell 7 and resized as they enter the cell. The resized targets (1/4 length) are then processed one at a time through the following processes: caustic dissolution and filtration, acidic dissolution and filtration, elutable Reillex anion exchange, diglycolamide (DGA) resin extraction, and DGA calcination. This processing results in two product streams. The first is an aqueous plutonium solution which is removed from the shielded cells and taken to a glovebox for further purification and conversion to an oxide. The second is a calcined oxide product containing the Am and Cm as well as other lanthanide fission products which is removed from the shielded cells using a bagless transfer system. Both materials are packaged for shipment to Oak Ridge National Laboratory (ORNL). This paper will discuss the operating experience, lessons learned, and results from initial process hot startup.

Armstrong, Christopher [Savannah River National La↗

DOE OSTI2025

Anomaly Detection in Seismic Data with Deep Learning: Application for Instrument Failure Detection and Forecasting

Seismic data quality assessment (QA) is the first and one of the most important steps before conducting any further data analysis. Traditional methods involve checking various metrics, such as spike detection and power spectral density, by setting strict thresholds or comparing data against synthetic benchmarks. However, these approaches often rely on pre-existing knowledge and assumptions about data anomalies, leading to potential misclassification of unusual cases. Here, in this study, we propose a deep autoencoder model, an unsupervised learning approach that evaluates data quality without making assumptions about normal and anomalous data, which can be used to identify deviations in recorded data that may indicate nascent instrument failure. We test the model with the U.S. International Monitoring System (IMS) seismic stations and demonstrate the capability of detecting anomalies on a monthly scale. This could prompt station operators to examine potential problems early, allowing sufficient time for instrument maintenance to prevent data outages. In addition, we use a new manually selected testing dataset to compare our model performance against two supervised machine learning (ML) approaches and a standard QA package, as baseline models. When applied to the dataset containing known data anomalies, performance of the supervised and unsupervised ML approaches is similar, with an accuracy of 88.1% for our model compared to ∼90% for the supervised ML approach and 78.2% for the standard QA package. Our model outperforms the baseline models when applied to new stations, where new types of data anomalies can be station-specific and not included in the training dataset. Finally, we show model transferability by training the model with data from the Global Seismograph Network only and applying it to the IMS network data. The results suggest that our model is generalizable and can be applied to new stations with good accuracy.

Lin, Jiun-Ting [Lawrence Livermore National Labora↗

DOE OSTI2024

Durable and High-Performance SOECs Based on Proton Conductors for Hydrogen Production

Proton-conducting solid oxide electrolysis cells (P-SOECs) are a promising technology for cost-effective and efficient production of green hydrogen. Breakthroughs in materials development, optimization of cell structure, and achievement of high performance and durability are essential to significantly increase the commercial competitiveness of these technologies. The main objective of this project is to gain scientific knowledge for the rational design, fabrication, and demonstration of a robust, highly efficient, and low-cost SOEC technology based on a proton-conducting electrolyte membrane for hydrogen production. We focused on better understanding the degradation mechanisms of proton-conducting electrolytes, air electrodes, and catalyst materials under electrolysis mode to develop an effective strategy for rationalizing new materials that are vital for enhancing cell performance and durability. The scope includes enhancing the performance and durability of the electrolyte and electrode materials under realistic operating conditions, developing highly active and robust catalysts to minimize electrode losses while improving tolerance to contaminant poisoning, revealing the mechanism of enhanced activity and stability of the catalyst, and understanding the underlying degradation mechanisms. In addition, various characterization techniques were employed to gain a fundamental understanding of the materials’ behavior and their impact on cell performance, providing vital information to guide materials discovery and cell design. After defect chemistry engineering, the optimized donor and acceptor co-doped electrolytes BaMo/W 0.03 Ce 0.71 Yb 0.26 O 3-δ (BM/W03) showed substantially improved chemical stability against high concentrations of CO 2 and H 2 O compared to the state-of-the-art electrolyte (BaZr 0.1 Ce 0.7 Y 0.1 Yb 0.1 O 3-δ , BZCYYb1711) while maintaining comparable ionic conductivity and ionic transference number. To bypass the inherent trade-off between conductivity and chemical stability, we fabricated a bi-layer electrolyte composed of BZCYYb1711 coated with a highly-stable thin layer of BaHf 0.83 Yb 0.17 O 3-δ (BHYb). This bi-layer electrolyte displayed excellent chemical stability against high concentration CO 2 ; there was no detectable formation of BaCO 3 after exposure to 97% CO 2 (with 3% H 2 O) at 500 °C for 1000 hours and the rate of degradation in resistance was about 0.4% per 1,000 hours (kh). In contrast, the same BZCYYb1711 electrolyte without a BHYb coating degraded significantly under the same testing conditions; the degradation rate was increased to 5.1%/kh. In addition, a triple conducting air electrode Ba 0.9 Pr 0.1 Hf 0.1 Y0.1Co 0.8 O 3-δ (BPHYC) was developed by heavily doping transition metal ions into a proton-conducting material. This air electrode material, composed of 3 distinct phases, exhibits superior electrocatalytic activity due to the synergistic effect from the three component phases. Moreover, an active and durable catalyst, La 2 Ni 0.5 Fe 0.5 O 4+δ (LNF), was developed, showing excellent catalytic activity and contaminant tolerance, with a degradation rate of only 0.49%/kh when exposed to high concentrations of steam and Cr. Finally, single cells were constructed from the best electrolytes, electrodes, and catalyst coatings developed in this project. These cells demonstrated superior high current density at a given cell voltage, high roundtrip efficiency, and remarkable durability (up to 1000 hours of operation).

08 HYDROGEN↗

DOE OSTI2025

Physics-based hybrid machine learning for critical heat flux prediction with uncertainty quantification

Critical heat flux (CHF) is a key quantity in nuclear system modeling due to its impact on heat transfer, safety margins, and reactor performance. This study develops and validates an uncertainty-aware hybrid modeling approach that combines machine learning with physics-based models to predict CHF in cases of dryout. The Biasi and Bowring empirical correlations were paired with three ML uncertainty quantification (UQ) techniques: deep neural network (DNN) ensembles, Bayesian neural networks (BNNs), and deep Gaussian processes (DGPs). A pure ML model without a base model was evaluated for comparison. Model performance was assessed under plentiful (7,350 points) and limited (9 points) training data scenarios using parity, uncertainty distributions, and calibration curves. Results show that the Biasi hybrid DNN ensemble achieved the best overall performance, with a mean absolute relative error of 1.846%, and well-calibrated uncertainty estimates. The BNN-based hybrids showed slightly higher error (2.14%) but superior uncertainty calibration. DGP models underperformed, with over 6% error and poor uncertainty calibration. All hybrid models outperformed pure machine learning configurations, demonstrating resistance against data scarcity. These findings indicate that hybrid modeling significantly improves predictive accuracy, interpretability, and resilience to data scarcity. The integration of uncertainty awareness provides actionable confidence in CHF predictions, which is vital for safety-critical decisions in nuclear applications. This hybrid approach offers a viable pathway for deploying ML models in reactor analysis tools while preserving domain knowledge and physical consistency.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2025

CAFE AU LAIT: Compute-Aware Federated Augmented Low-Rank AI Training

Federated finetuning is crucial for unlocking the knowledge embedded in pretrained Large Language Models (LLMs) when data are geographically distributed across clients. Unlike finetuning with data from a single institution, federated finetuning allows collaboration across multiple institutions, enabling the utilization of diverse and decentralized datasets while preserving data privacy. Given the high computing costs of LLM training and the emphasis on energy efficiency in Federated Learning (FL), Low-Rank Adaptation (LoRA) has emerged as a widely adopted algorithm due to its significantly reduced number of trainable parameters. However, this assumes that all data silos have the necessary computing resources to compute local updates of LLMs. Nevertheless, in practice, the computing resources across clients are highly heterogeneous: while some may have access to hundreds of GPUs, others might have limited or no GPU access. Recently, federated finetuning using synthetic data has been proposed, allowing clients to participate in a collaborative training run without training LLMs locally. However, our experimental results reveal a performance gap between models trained using synthetic data and those trained using local updates. Motivated by the observed heterogeneity in computing resources and the performance gap, we propose a novel two-stage algorithm that leverages the storage and computing capabilities of a strong server. In the first stage, under the coordination of the strong server, clients with limited computing resources collaborate to generate synthetic data, which is transferred to and stored on the strong server. In the second stage, the strong server uses this synthetic data on behalf of the resource-constrained clients to perform federated LoRA finetuning alongside clients with sufficient computing resources. This approach ensures that all clients can participate in the finetuning process. Experimental results demonstrate that incorporating local updates from even a small fraction of clients improves performance compared to using synthetic data for all clients. Furthermore, we incorporate the Gaussian mechanism in both stages to guarantee client-level differential privacy.

Wang, Jiayi [ORNL]↗

DOE OSTI2024

Unsupervised learning-enabled pulsed infrared thermographic microscopy of subsurface defects in stainless steel

Metallic structures produced with laser powder bed fusion (LPBF) additive manufacturing method (AM) frequently contain microscopic porosity defects, with typical approximate size distribution from one to 100 microns. Presence of such defects could lead to premature failure of the structure. In principle, structural integrity assessment of LPBF metals can be accomplished with nondestructive evaluation (NDE). Pulsed infrared thermography (PIT) is a non-contact, one-sided NDE method that allows for imaging of internal defects in arbitrary size and shape metallic structures using heat transfer. PIT imaging is performed using compact instrumentation consisting of a flash lamp for deposition of a heat pulse, and a fast frame infrared (IR) camera for measuring surface temperature transients. However, limitations of imaging resolution with PIT include blurring due to heat diffusion, sensitivity limit of the IR camera. We demonstrate enhancement of PIT imaging capability with unsupervised learning (UL), which enables PIT microscopy of subsurface defects in high strength corrosion resistant stainless steel 316 alloy. PIT images were processed with UL spatial–temporal separation-based clustering segmentation (STSCS) algorithm, refined by morphology image processing methods to enhance visibility of defects. The STSCS algorithm starts with wavelet decomposition to spatially de-noise thermograms, followed by UL principal component analysis (PCA), fine-tuning optimization, and neural learning-based independent component analysis (ICA) algorithms to temporally compress de-noised thermograms. The compressed thermograms were further processed with UL-based graph thresholding K-means clustering algorithm for defects segmentation. The STSCS algorithm also includes online learning feature for efficient re-training of the model with new data. For this study, metallic specimens with calibrated microscopic flat bottom hole defects, with diameters in the range from 203 to 76 µm, were produced using electro discharge machining (EDM) drilling. While the raw thermograms do not show any material defects, using STSCS algorithm to process PIT images reveals defects as small as 101 µm in diameter. To the best of our knowledge, this is the smallest reported size of a sub-surface defect in a metal imaged with PIT, which demonstrates the PIT capability of detecting defects in the size range relevant to quality control requirements of LPBF-printed high-strength metals.

36 MATERIALS SCIENCE↗

DOE OSTI2026

Scaling dynamics in low-salt-rejection reverse osmosis for high-salinity produced water desalination: Mechanistic modeling and membrane autopsy

Membrane scaling remains a critical barrier to the reliable operation of desalination systems, particularly for hypersaline produced water (PW) treatment. This study fills the knowledge gap of autopsy-based model validation for PW desalination by elucidating scaling mechanisms in a Low-Salt-Rejection Reverse Osmosis (LSRRO) system through the integration of pilot-scale experimentation and complementary modeling approaches. A semi-empirical modeling framework was developed and applied to a multistage pilot LSRRO system equipped with nanofiltration and RO membranes treating high-salinity PW from the Permian Basin. Water quality analysis showed that total dissolved solids decreased from ~130,000 mg/L to ~1900 mg/L in the permeate, then further reduced to ~300 mg/L by a second-pass RO. Two different thermodynamic modeling approaches were evaluated: the first extends the LSRRO framework by incorporating system complexity and scaling phenomena, whereas the second method explicitly captures concentration polarization in localized supersaturation. Both methods illustrate the tendency for carbonate and sulfate scaling throughout the stages. Membrane autopsies revealed a silica-dominated deposit matrix, localized CaSO 4 at Stage 2, and minor barite/celestite despite their prominence in model predictions. Quantum-chemical calculations indicated silica scaling can be rationalized by favorable adsorption of H 4 SiO 4 on Fe-oxide surfaces (ΔG ≈ −44 kJ/mol), providing a kinetic pathway for interfacial inorganic polymerization even when bulk equilibrium predictions are conservative. Overall, the thermodynamic scaling modeling and membrane autopsy revealed heterogeneous, localized deposits with limited impact on LSRRO performance, while quantum analysis rationalized the thermodynamically unfavorable precipitation formation under bulk equilibrium, reconciling model–autopsy discrepancies. These insights support targeted pretreatment and silica-specific antiscalants to extend membrane lifetime and increase recovery, providing a transferable framework for hypersaline water desalination systems. The combined experimental–computational approach provides new mechanistic insight into scaling in hypersaline membrane systems and establishes a transferable framework for predicting and mitigating scaling in next-generation desalination technologies.

Low-salt-rejection reverse osmosis↗

DOE OSTI2025

University Data Management Pilot Utilizing the Nuclear Research Data System

Background In 2022, the Office of Science and Technology Policy (OSTP) issued a memo that significantly reshaped the landscape of access to federally funded research. The memo mandated that all taxpayer-funded research be made available to the public without delay upon publication, without an embargo period, superseding the 2013 OSTP public access policy. This public access policy promotes transparency and the democratization of knowledge, ensuring that the fruits of scientific endeavors funded by federal agencies could be immediately accessed and built upon by scientists, educators, students, and the public at large. To implement the requirements of the OSTP guidance and DOE Public Access Plan, the Office of Nuclear Energy (NE) has implemented public access plan guidance and has identified several areas where better data management practices would further expand public access to important nuclear energy related scientific data, reports, and other technical products. Significant NE supported efforts are already underway for data management and public access to important nuclear energy related data.1 2 To address gaps in data management practices, and improve retention and accessibility of data, NE is actively exploring enhanced data management options utilizing its high-performance computing resources administered by its Nuclear Scientific User Facility Program. A newly piloted system, the Nuclear Research Data System (NRDS) acts as a portal for data collection and dissemination. Nuclear Energy University Program Research and Development Portfolio According to Web of Science, NEUP has produced 2,345 journal publication that have been cited more than 61,000 times3 and countless conference proceedings. These publications are publicly available through OSTI.gov and in the open literature. Additional scientific and technical products including project milestones that are not publications and NEUP project final reports are vetted through OSTI.gov and released once reviewed and approved by DOE. Since 2009, NEUP has awarded close to 1,000 different R&D projects in technical areas across the NE research programs. As of June 2023, 512 NEUP reports are publicly available on OSTI. The underlying data for projects is still held at universities, and data transfer, co-location, and dissemination has not occurred in a systematic way. NEUP data is currently accessible through myriad university-based data repositories, or through direct requests to PIs. The program identified this patchwork of repositories, or often lack of publicly available data, as a significant barrier to an organized, accessible, and comprehensive solution to sharing data with the larger nuclear energy community. Approach The goal of this pilot project is to establish a pathway to a consolidated long-term repository for NEUP project data. To accomplish this goal, the pilot strives to accomplish the following objectives: Establish data collection standards, including a standard set of required supplementary information to contextualize and support raw data files. Work with the HPC group collect and upload information and to modify the NRDS system, as needed, to support a standardized approach. Resolve potential barriers to successful roll out of an expanded data collection strategy, including modifying data management plan guidelines and establishing a document and data release process that accounts for potential intellectual property and/or export control concerns. Results Overall, the pilot was successful in collecting 8,982 raw and processes data files, 220 reports, 56 calibration files, and 5,931 other supplementary documents. Supplementary documents included experimental plans, methods, journal publications and conference proceedings, milestone reports, and final reports. Figure 2 shows the number of data sets and supplementary project information provided by each project. Projects has significantly different input, depending on experimental data produced and completeness of the datasets provided.

Data collection↗