Search NASA⌕ Search

SEARCH · Search NASA

Results for “Reference Model 5”

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 127 records · Page 7

Operating Wind Turbine as Synchronous Generator: Modeling and Power-Hardware-in-the-Loop Demonstration

Grid-forming (GFM) control of Type 3 and Type 4 wind turbine generators (WTGs) has attracted substantial attention in power systems research; however, the limited overcurrent capability of power electronics converters continues to deteriorate the grid strength of the evolving power systems. Synchronous wind, also referred to as a Type 5 WTG, offers a unique GFM solution to address grid integration and grid strength issues by keeping the grid largely synchronous at very high integration levels of renewable generation. A Type 5 WTG interfaces with the electric grid via a synchronous generator driven by a variable speed hydraulic torque converter; hence, the wind rotor operates in variable-speed mode for maximum power generation, and the generator shaft remains synchronous to the grid. This paper develops and tests a high-fidelity model of a Type 5 WTG in a power-hardware-in-the-loop testing environment, and it presents its operation characteristics under different grid contingencies. The power-hardware-in-the-loop demonstration shows that a Type 5 WTG inherently behaves as a GFM unit and can obtain similar performance in terms of power response, wind rotor dynamics, and stability enhancement compared to a Type 3 WTG in GFM control mode. Furthermore, the paper provides further insight into how Type 5 WTGs can support the smooth transition to power systems with high integration levels of inverter-based resources.

17 - WIND ENERGY↗

Large-scale seismic waveform quality metric calculation using Hadoop

Here in this work we investigated the suitability of Hadoop MapReduce and Apache Spark for large-scale computation of seismic waveform quality metrics by comparing their performance with that of a traditional distributed implementation. The Incorporated Research Institutions for Seismology (IRIS) Data Management Center (DMC) provided 43 terabytes of broadband waveform data of which 5.1 TB of data were processed with the traditional architecture, and the full 43 TB were processed using MapReduce and Spark. Maximum performance of ~0.56 terabytes per hour was achieved using all 5 nodes of the traditional implementation. We noted that I/O dominated processing, and that I/O performance was deteriorating with the addition of the 5th node. Data collected from this experiment provided the baseline against which the Hadoop results were compared. Next, we processed the full 43 TB dataset using both MapReduce and Apache Spark on our 18-node Hadoop cluster. We conducted these experiments multiple times with various subsets of the data so that we could build models to predict performance as a function of dataset size. We found that both MapReduce and Spark significantly outperformed the traditional reference implementation. At a dataset size of 5.1 terabytes, both Spark and MapReduce were about 15 times faster than the reference implementation. Furthermore, our performance models predict that for a dataset of 350 terabytes, Spark running on a 100-node cluster would be about 265 times faster than the reference implementation. We do not expect that the reference implementation deployed on a 100-node cluster would perform significantly better than on the 5-node cluster because the I/O performance cannot be made to scale. Finally, we note that although Big Data technologies clearly provide a way to process seismic waveform datasets in a high-performance and scalable manner, the technology is still rapidly changing, requires a high degree of investment in personnel, and will likely require significant changes in other parts of our infrastructure. Nevertheless, we anticipate that as the technology matures and third-party tool vendors make it easier to manage and operate clusters, Hadoop (or a successor) will play a large role in our seismic data processing.

58 GEOSCIENCES↗

Preshot Calculations for a Small-Scale HE Experiment. Overview and Results for Symmetric Configurations

Explosively-driven magnetic flux compression generators create substantial currents (10’s of mega-amps) by compressing magnetic fields initially created by injected seed currents. In a Ranchero generator it is the field between two cylinders of aluminum that is compressed when the inner cylinder (armature) is driven across the magnetized gap toward the second cylinder (stator). All Rancheros to date have used the explosive PBXN-110, but future devices are expected to use PBX-9501 because of several advantages of the latter over the former. For Ranchero applications, though, a potentially important disadvantage stems from the requirement that the large PBX-9501 charges (15 to 50 kg) must built up from smaller machined pieces rather than cast into the appropriate shape as with PBXN-110. Calculations and related experiments raise the possibility that jetting may occur at gaps between machined pieces of PBX-9501 and lead to localized failure of the soft aluminum armature causing premature contact of the armature with the stator or, in the most extreme case, a severing of the armature into separate pieces and a subsequent loss of current. A set of small-scale experiments has been designed to provide Ranchero designers and users insight into the effects of gaps and also to provide useful data for the validation of Ranchero calculations. These experiments should be executed in early May 2015. The code Rage was used to model the small-scale experiment and this paper presents the results. The emphasis here is on the calculations and the experimental details are limited, so the interested reader is referred to reference 5 for a fuller description of the experimental configuration and diagnostics. Less-interested readers may be interested in only a summary of results and are directed to the “Summary of key results” section later in this paper.

36 MATERIALS SCIENCE↗

Modeling the potential of screen printed front junction CZ silicon solar cell with tunnel oxide passivated back contact

Carrier selective passivated contacts composed of thin oxide, n + polycrystalline Si and metal on top of a n-Si absorber can significantly lower the recombination current density (J orear ≤8 fA/cm 2 ) under the contact while providing excellent specific contact resistance (5–10 mΩ-cm 2 ); 25.1% efficient small area cells with photolithography front contacts on boron doped selective emitter and Fz wafers have been achieved by Fraunhofer ISE using their tunnel oxide passivated contact (TOPCon) approach. This paper shows a methodology to model such passivated contact cells using Sentaurus device model, which involves replacing the TOPCon region by carrier selective electron and hole recombination velocities to match the measured J orear of the TOPCon region as well as all the light IV values of the cell. We first validated the methodology by modeling a 24.9% reference cell. The model was then extended to assess the efficiency potential of large area TOPCon cells on commercial grade n-type Cz material with screen-printed front contacts. To use realistic input parameters, a 21% n-type PERT cell was fabricated on Cz wafer (5 Ω-cm, 1.5-ms lifetime). Modeling showed that the cell efficiency will improve to only 21.6% if the back of this cell is replaced by the above TOPCon, and the performance is limited by the homogenous emitter. Efficiency was then modeled to improve to 22.6% with the implementation of selective emitter (150/20 Ω/sq). Lastly, it is shown that screen printing of 40-µm-wide lines and improved bulk material (10 Ω-cm, 3-ms lifetime) can raise the single side TOPCon Cz cell efficiency to 23.2%.

14 SOLAR ENERGY↗

Modeling laser wakefield accelerators in a Lorentz boosted frame

Modeling of laser-plasma wakefield accelerators in an optimal frame of reference [1] is shown to produce orders of magnitude speed-up of calculations from first principles. Obtaining these speedups requires mitigation of a high frequency instability that otherwise limits effectiveness in addition to solutions for handling data input and output in a relativistically boosted frame of reference. The observed high-frequency instability is mitigated using methods including an electromagnetic solver with tunable coefficients, its extension to accomodate Perfectly Matched Layers and Friedman's damping algorithms, as well as an efficient large bandwidth digital filter. It is shown that choosing theframe of the wake as the frame of reference allows for higher levels of filtering and damping than is possible in other frames for the same accuracy. Detailed testing also revealed serendipitously the existence of a singular time step at which the instability level is minimized, independently of numerical dispersion, thus indicating that the observed instability may not be due primarily to Numerical Cerenkov as has been conjectured. The techniques developed for Cerenkov mitigation prove nonetheless to be very efficient at controlling the instability. Using these techniques, agreement at the percentage level is demonstrated between simulations using different frames of reference, with speedups reaching two orders of magnitude for a 0.1 GeV class stages. The method then allows direct and efficient full-scale modeling of deeply depleted laser-plasma stages of 10 GeV-1 TeV for the first time, verifying the scaling of plasma accelerators to very high energies. Over 4, 5 and 6 orders of magnitude speedup is achieved for the modeling of 10 GeV, 100 GeV and 1 TeV class stages, respectively.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Improving Hydrological Simulation in the Upper Mississippi River Basin through Enhanced Freeze-Thaw Cycle Representation

Freeze-thaw cycles are important processes relevant to terrestrial hydrological cycling. However, the representation of freeze-thaw cycle has been often simplified in large scale watershed models. The Soil and Water Assessment Tool (SWAT), which has been widely used to understand and assess hydrologic budgets and water resources management, employs a simplified empirical approach to estimate soil temperature and determine the freezing and thawing status of soils. Here, we compared the performance of a physically-based soil temperature module and the built-in empirical approach in SWAT against field measurements at surface and 5, 10, 20, 50, and 100 cm depths at six stations of the U.S. Climate Reference Network (USCRN) within the Upper Mississippi River Basin (UMRB). In general, SWAT consistently underestimated winter soil temperatures and overestimated frozen days at all soil depths, while the modified version of SWAT (equipped with the physically-based soil temperature model; referred to as TSWAT) pronouncedly reduced the bias in estimated winter soil temperatures and frozen days compared with SWAT. Model performance assessment is conducted with three statistical coefficients, i.e., Bias (?), coefficient of determination (R2), and Nash-Sutcliffe coefficient (NS). Statistical analyses show that TSWAT accurately simulated surface and soil temperatures at the five depths with R2 and NS values greater than 0.82 at most sites, and Bias values were generally within the range of -1 to 1? during winter and ranged between -2.09 and 2.58 ? in non-winter seasons. The differences in freeze-thaw cycle representation between SWAT and TSWAT translate into noticeable discrepancies in simulated key hydrologic variables, such as surface runoff, percolation, and baseflow. Compared against long-term observed streamflow (1980-2015), TSWAT outperformed SWAT in capturing variations in monthly streamflow in both winter and non-winter seasons. These results and analyses highlight the value of improving freeze-thaw cycle representation for enhanced hydrologic modeling in large watersheds that are subject to freeze-thaw cycles.

soil temperature, freeze-thaw cycles, streamflow↗

TRISO fuel performance analysis: Uncertainty quantification toward optimization

Tri-structural isotropic (TRISO) fuel particles are a fuel form being considered for potential use in next-generation nuclear reactors (i.e., high-temperature gas-cooled reactors). Though the TRISO fuel manufacturing process has continually advanced in recent years, particle comparisons still reveal statistical variations and uncertainties in terms of geometric configurations and material properties. Given that the physical processes ongoing in TRISO fuel particles during reactor operation are highly correlated with each other, a small degree of uncertainty in one model may lead to significant uncertainty in another. This makes appropriate uncertainty quantification of TRISO fuel particles essential. However, one may wonder about the extent to which the current version of TRISO particles has been optimized, and whether any room remains for further improvements. This paper quantifies TRISO fuel performance model uncertainties that stem from geometric and material data. For this analysis, the BISON code was used, and the Advanced Gas Reactor (AGR)-2 experiment served as a reference case. A total of 10 5 calculations was performed for the uncertainty and optimization analysis, altering the geometric and material data within their uncertainty range. Lastly, the optimization potential of TRISO particles is evaluated from a fuel performance perspective.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

A Landau-Devonshire Approach of Elastic Strain Effects on Ferroelectric Al1-xScxN

Landau-Devonshire thermodynamic modeling enables the phenomenological description of ferroelectrics. 1 The approach has been validated across a wide range of ferroelectric materials from the first ferroelectric discovered, Rochelle Salt, to recent HfO2-based materials. 2-4 Extension to what is now referred to as the Landau-Ginsburg-Devonshire model enables rigorous investigation of polarization boundaries and reorientation, phase transformations, and domain formation/evolution. We present a thermodynamic analysis of the recently discovered nitride ferroelectric materials 5-7 using the classic Landau-Devonshire approach. The electrostrictive and dielectric stiffness coefficients of Al1-xScxN with wurtzite structure (6mm) are determined using a free energy density function assuming a hexagonal parent phase (6/mmm), with the first order phase transition based on the dielectric stiffness relationships (Fig. 1). The results of this analysis show that the strain sensitivity of the energy barrier is one order of magnitude larger than that of the spontaneous polarization in these wurtzite ferroelectrics, yet both are less sensitive to strain compared to classic perovskite ferroelectrics (Fig. 2). These analysis results reported here explain experimentally reported sensitivity of coercive field to elastic strain/stress in Al1-xScxN films, and would enable further thermodynamic analysis via phase field simulation and related methods.

CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSICS↗

TU-EF-204-11: Impact of Using Multi-Slice Training Sets On the Performance of a Channelized Hotelling Observer in a Low-Contrast Detection Task in CT

Purpose: To investigate using multiple CT image slices from a single acquisition as independent training images for a channelized Hotelling observer (CHO) model to reduce the number of repeated scans for CHO-based CT image quality assessment. Methods: We applied a previously validated CHO model to detect low contrast disk objects formed from cross-sectional images of three epoxy-resin-based rods (diameters: 3, 5, and 9 mm; length: ∼5cm). The rods were submerged in a 35x 25 cm2 iodine-doped water filled phantom, yielding-15 HU object contrast. The phantom was scanned 100 times with and without the rods present. Scan and reconstruction parameters include: 5 mm slice thickness at 0.5 mm intervals, 120 kV, 480 Quality Reference mAs, and a 128-slice scanner. The CHO’s detectability index was evaluated as a function of factors related to incorporating multi-slice image data: object misalignment along the z-axis, inter-slice pixel correlation, and number of unique slice locations. In each case, the CHO training set was fixed to 100 images. Results: Artificially shifting the object’s center position by as much as 3 pixels in any direction relative to the Gabor channel filters had insignificant impact on object detectability. An inter-slice pixel correlation of >∼0.2 yielded positive bias in the model’s performance. Incorporating multi-slice image data yielded slight negative bias in detectability with increasing number of slices, likely due to physical variations in the objects. However, inclusion of image data from up to 5 slice locations yielded detectability indices within measurement error of the single slice value. Conclusion: For the investigated model and task, incorporating image data from 5 different slice locations of at least 5 mm intervals into the CHO model yielded detectability indices within measurement error of the single slice value. Consequently, this methodology would Result in a 5-fold reduction in number of image acquisitions. This project was supported by National Institutes of Health grants R01 EB017095 and U01 EB017185 from the National Institute of Biomedical Imaging and Bioengineering.

60 APPLIED LIFE SCIENCES↗

Optimizing Energy Savings from Direct-DC in U.S. Residential Buildings

An increasing number of energy efficient appliances operate on direct current (DC) internally, offering the potential to use DC from renewable energy systems directly and avoiding the losses inherent in converting power to alternating current (AC) and back. This paper investigates that potential for net-metered residences with on-site photovoltaics (PV) by modeling the net power draw of the ‘direct-DC house’ with respect to today’s typical configuration, assuming identical DC-internal loads. Power draws were modeled for houses in 14 U.S. cities, using hourly, simulated PV-system output and residential loads. The latter were adjusted to reflect a 33% load reduction, representative of the most efficient DC-internal technology, based on an analysis of 32 electricity end-uses. The model tested the effect of climate, electric vehicle (EV) loads, electricity storage, and load shifting on electricity savings; a sensitivity analysis was conducted to determine how future changes in the efficiencies of power system components might affect savings potential. Based on this work, we estimate that net-metered PV residences could save 5% of their total electricity load for houses without storage and 14% for houses with storage. Based on residential PV penetration projections for year 2035 obtained from the National Energy Modeling System (2.7% for the reference case and 11.2% for the extended policy case), direct-DC could save the nation 10 trillion Btu (without storage) or 40 trillion Btu (with storage). Shifting the cooling load by two hours earlier in the day (pre-cooling) has negligible benefits for energy savings. Direct-DC provides no energy savings benefits for EV charging, to the extent that charging occurs at night. However, if charging occurred during the day, for example with employees charging while at work, the benefits would be large. Direct-DC energy savings are sensitive to power system and appliance conversion efficiencies but are not significantly influenced by climate. While direct-DC for residential applications will most likely arise as a spin-off of developments in the commercial sector—because of lower barriers to market entry and larger energy benefits resulting from the higher coincidence between load and insolation—this paper demonstrates that there are substantial benefits in the residential sector as well. Among residential applications, space cooling derives the largest energy savings from being delivered by a direct-DC system. It is the largest load for the average residence on a national basis and is particularly so in high-load regions. It is also the load with highest solar coincidence.

14 SOLAR ENERGY↗

Development of a high current density, high temperature superconducting cable for pulsed magnets

Abstract A low AC loss Rare Earth Barium Copper Oxide (REBCO) cable, based on the VIPER cable technology has been developed by Commonwealth Fusion Systems for use in high field, REBCO based tokamaks. The new cable is composed of partitioned and transposed copper ‘petals’ shaped to fit together in a circular pattern with each petal containing a REBCO tape stack and insulated from each other to reduce AC losses. A stainless steel jacket adds mechanical robustness—also serving as a vessel for solder impregnation—while a tube runs through the middle for cooling purposes. Additionally, fiber optic sensors are placed under the tape stacks for quench detection. To qualify this design, a series of experiments were conducted as part of the SPARC tokamak Central Solenoid Model Coil program—to retire the risks associated with full scale, fast ramping, high flux HTS Central Solenoid (CS) and Poloidal Field (PF) coils for tokamak fusion power plants and net energy demonstrators. These risk study and risk reduction experiments include (1) AC loss measurement and model validation in the range of ~5 T/s, (2) an IxB electromagnetic loading of over 850 kN/m at the cable level and up to 300 kN/m at the stack level, (3) a transverse compression resilience of over 350 MPa, (4) manufacturability at tokamak relevant speeds and scales, (5) cable to cable joint performance, (6) fiber optic based quench detection speed, accuracy, and feasibility, and (7) overall winding pack integration and magnet assembly. The result is a cable technology, now referred to as PIT VIPER, with AC losses that measure fifteen times lower (at ~5 T/s) than its predecessor technology; a 2% or lower degradation of critical current (Ic) at high IxB electromagnetic loads; no detectable Ic degradation up to 570 MPa of transverse compression on the cable unit cell; end to end magnet manufacturing, consistently producing Ic values within 7% of the model prediction; cable to cable joint resistances at 20 K on the order of ~15 nΩ; and fast, functional quench detection capabilities that do not involve voltage taps. This cable technology will be tested comprehensively in a Central Solenoid Model Coil to prove its readiness for compact, high field tokamak operation.

Sanabria, Charlie (ORCID:0000000150175309)↗

Dark Energy Survey Year 6 results: Clustering redshifts and importance sampling of self-organized-maps 𝑛⁡(𝑧) realizations for 3 × 2 ⁢pt samples

This work is part of a series establishing the redshift framework for the 3 × 2 ⁢pt analysis of the Dark Energy Survey Year 6 (DES Y6). For DES Y6, photometric redshift distributions are estimated using self-organizing maps (SOMs), calibrated with spectroscopic and many-band photometric data. To overcome limitations from color-redshift degeneracies and incomplete spectroscopic coverage, we enhance this approach by incorporating clustering-based redshift constraints (clustering-z, or WZ) from angular cross-correlations with BOSS and eBOSS galaxies and eBOSS quasar samples. We define a WZ likelihood and apply importance sampling to a large ensemble of SOM-derived 𝑛⁡(𝑧) realizations, selecting those consistent with the clustering measurements to produce a posterior sample for each lens and source bin. The analysis uses angular scales corresponding to 1.5–5 Mpc to optimize signal-to-noise ratio while mitigating modeling uncertainties and marginalizes over redshift-dependent galaxy bias and other systematics informed by the N-body simulation CARDINAL . While a sparser spectroscopic reference sample limits WZ constraining power at 𝑧 >1.1, particularly for source bins, we demonstrate that combining SOM with WZ improves redshift accuracy and enhances the overall cosmological constraining power of DES Y6. As a result, we estimate an improvement in 𝑆 8 of approximately 10% for cosmic shear and 3 ×2⁢pt analysis, primarily due to the WZ calibration of the source samples.

Cosmological parameters↗

Using synthetic infrared spectra derived from n/k optical constants for standoff detection of chemical deposits

We report results from a recent field experiment to test the validity of using physics-based synthetic infrared spectra to serve as endmembers in a spectral database targeted at chemical deposits. Specifically, the optical constants n and k, (the real and imaginary part of the refractive index) were used to first model infrared reflectance spectra for different thicknesses of chemical layers (e.g. acetaminophen, methylphosphonic acid – MPA, etc.) on various conducting and insulating substrates such as aluminum, wood, and glass. In the experimental portion of the research, thin films of the solid and liquid analytes were deposited onto such substrates to form micron-thick layers of the analytes at different thicknesses: Standoff data from an imaging instrument were then recorded and analyzed to not only identify the different analytes, but also quantify the layer/deposit thickness. To gauge success, the detection results using the synthetic data were compared to the results from hemispherical reflectance (HRF) spectra that were collected for the same sample planchets measured in the field via standoff methods. Preliminary results indicate good agreement between the synthetic reference data as compared to the lab-measured HRF data in terms of their ability to quantitatively reduce longwave infrared data. Specifically, modeled IR spectra for acetaminophen on an aluminum planchet at various thicknesses (1, 2, 5, 10, 15, and 20 µm) were synthesized and compared with standoff field reflectance data as well as HRF laboratory reflectance spectra for two samples: a 5.2 µm- and 12.8 µm-thick layer of acetaminophen on aluminum. Using a first-order approximation, analysis of the field data estimates the thicknesses of the samples to be 2 and 10 µm for the two samples, respectively, while the HRF laboratory data yields thickness estimates of between 5-10 µm and 10 µm, respectively. Both yield reasonable estimates, with the uncertainty most likely due to factors yet to be accounted for in the synthetic spectra such as light scattering.

complex index of refraction, optical constants, hy↗

HydraGNN_Predictive_GFM_2026 - Ensemble of predictive graph foundation models for atomistic materials modeling

This release contains data and parameters of HydraGNN-based graph foundation models trained as a result of the work published in the pre-print "Exascale Multi-Task Graph Foundation Models for Imbalanced, Multi-Fidelity Atomistic Data" by M. Lupo Pasini et al. (https://arxiv.org/abs/2604.15380). We jointly train on 16 open first-principles datasets (544+ million structures covering 85+ elements) using a multi-task architecture with per-dataset heads and a scalable ADIOS2/DDStore data pipeline. On Frontier, we execute six large-scale DeepHyper hyperparameter optimization campaigns in FP64 and promote the top-performing message-passing models to sustained 2,048-node training, yielding a PaiNN-based lead model. The version of HydraGNN used to generate the outputs provided in this release is HydraGNN v5.0 (https://github.com/ORNL/HydraGNN/releases/tag/v5.0) The list of datasets used for the training of the graph foundation model is the following: 1) Alexandria [1] 2) ANI1x [2] 3) MPTrj [3] 4) Open Catalyst 2020 (OC20) [4] 5) Open Catalyst 2022 (OC22) [5] 6) Open Catalyst 2025 (OC25) [6] 7) Open Direct ir Capture 2023 (ODAC23) [7] 8) Open Materials 2024 (OMat24) [8] 9) Open Molecules 2025 (OMol25) [9] 10) OMol25-neutral (subset of OMol25 that contains only molecules with zero total charge) 11) OMol25-non-neutral (subset of OMol25 that contains only molecules with non-zero total charge) 12) Open Polymers 2026 (OPoly2026) [10] 13) Nabla2DFT [11] 14) QCML [12] 15) QM7X [reference 13] 16) transition1x [14] Dataset references: [1] J. Schmidt et al., “A dataset of 175k stable and metastable materials calculated with the PBEsol and SCAN functionals,” Scientific Data, vol. 9, p. 64, 2022. [2] J. S. Smith et al., “The ANI-1ccx and ANI-1x data sets, coupled-cluster and density functional theory properties for molecules,” Scientific Data, vol. 7, p. 134, 2020. [Online]. Available: https: //www.nature.com/articles/s41597-020-0473-z [3] A. Jain et al., “Commentary: The Materials Project: A materials genome approach to accelerating materials innovation,” APL Materials, vol. 1, no. 1, p. 011002, 07 2013. [Online]. Available: https://doi.org/10.1063/1.4812323 [4] L. Chanussot et al., “Open catalyst 2020 (oc20) dataset and community challenges,” ACS Catalysis, vol. 11, no. 10, pp. 6059–6072, 2021. [Online]. Available: https://doi.org/10.1021/acscatal.0c04525 [5] K. Tran et al., “Open catalyst 2022 (oc22) dataset and challenges for oxidation electrocatalysts,” ACS Catalysis, vol. 13, no. 5, pp. 3066–3084, 2023. [Online]. Available: https://doi.org/10.1021/acscatal.2c05426 [6] S. J. Sahoo et al., “The open catalyst 2025 (oc25) dataset and models for solid-liquid interfaces,” arXiv preprint arXiv:2509.17862, 2025. [Online]. Available: https://arxiv.org/abs/2509.17862 [7] A. Sriram et al., “The open DAC 2023 dataset and challenges for sorbent discovery in direct air capture,” ACS Central Science, vol. 10, no. 5, pp. 923–941, 2024. [8] L. Barroso-Luque et al., “Open materials 2024 (omat24) inorganic materials dataset and models,” 2024. [Online]. Available: https://arxiv.org/abs/2410.12771 [9] D. S. Levine et al., “The open molecules 2025 (OMol25) dataset, evaluations, and models,” 2025. [Online]. Available: https://arxiv.org/abs/2505.08762 [10] D. S. Levine et al., The open polymers 2026 (OPoly26) dataset and evaluations,” arXiv preprint arXiv:2512.23117, 2025. [Online]. Available: https://arxiv.org/abs/2512.23117 [11] K. Khrabrov et al., “Nabla2dft: A universal quantum chemistry dataset of drug-like molecules and a benchmark for neural network potentials,” in NeurIPS 2024 Datasets and Benchmarks Track, 2024. [Online]. Available: https://openreview.net/forum?id=ElUrNM9U8c [12] S. Ganscha et al., “The QCML dataset, quantum chemistry reference data from 33.5M DFT and 14.7B semi-empirical calculations,” Scientific Data, vol. 12, p. 406, 2025. [13] J. Hoja et al., “QM7-X, a comprehensive dataset of quantum-mechanical properties spanning the chemical space of small organic molecules,” Scientific Data, vol. 8, p. 43, 2021. [Online]. Available: https://www.nature.com/articles/s41597-021-00812-2 [14] M. Schreiner et al., “Transition1x - a dataset for building generalizable reactive machine learning potentials,” Scientific Data, vol. 9, p. 779, 2022. The folder "datasets_ADIOS2_format" contains the set of pre-processed datasets in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used for the development and training of GFMs in this work. The "datasets_ADIOS2_format" directory contains 2 sub-directories, one for the version "v1" of the datasets and one for the version "v2" of the datasets. The version "v1" of the datasets provides values of the total energy as they are extracted from the original data as it was released by the respective institutions. The version "v2" of the datasets provides values of the energy that have been realigned. The realignment was performed by training a linear regression model that predicts the total energy as a function of the chemical composition of the atomistic structure, and then subtract such prediction from the original value of the total energy. Both folders "v1" and "v2" contain 16 sub-directories, each corresponding to an ADIOS2-formatted dataset The folder "DeepHyper-results" contains the configurational files and model's parameters for all the 186 HPO trials that were successfully completed by the scalable hyperparameter optimization (HPO) runs on Frontier. The content of the folder "DeepHyper-results" I structured as follows: 1) task-list.txt: list of mpnn name, jobid, and deephyper task id 2) gfm_${MPNN}_${JOBID}_0.${TASKID}: run directory with checkpoint files 3) gfm_${MPNN}: deephyper summary directory (*.csv) for each specific MPNN type 4) deephyper-experiment-${JOBID}: output and error logs for each job The file "deephyper-sorted.csv" contains the details of each HydraGNN model built and tested by HPO, obtained by merging the (*.csv) filed from each HPO run executed. Out of all the HPO trials, we selected 10 to continue the training of the respective HydraGNN models. Due to limited computational budget available in the LRN070 allocation we could not complete the training till convergence for all these 10 selected models. The folder "models" contains multiple sub-folders, one per each HydraGNN model trained. Each model sub-folder contains the parameters of each HydraGNN model, with multiple checkpoint-restarts. The list of sub-folders are as follows: 1) multidataset_hpo-BEST1-fp64 2) multidataset_hpo-BEST2-fp64 3) multidataset_hpo-BEST3-fp64 4) multidataset_hpo-BEST4-fp64 5) multidataset_hpo-BEST5-fp64 6) multidataset_hpo-BEST6-fp64 7) multidataset_hpo-BEST7-fp64 8) multidataset_hpo-BEST8-fp64 9) multidataset_hpo-BEST9-fp64 10) multidataset_hpo-BEST10-fp64 Within each one of these folders, additional auxiliary log files are provided with descriptions about how the training proceeded. The lead PaiNN-model is contained inside "multidataset_hpo-BEST6-fp64". The file "mlp_branch_weights" contains the parameters of the multi-layer perceptron (MLP) used to reconcile the predictions of the 16 output decoding heads of the HydragNN architectures. The MLP takes in input the chemical composition of the atomistic structure and predicts averaging weights to linearly mix the predictions of each output decoding head toward consolidating them into a single one. The folder "1.1billion-structure-inference" contains 1.1 billion atomistic structures randomly generated. Each structures is associated with energy and forces predicted with the lead-PaiNN model combined with the MLP model for reconciliation of the multi-branch predictions generated by the 16 output decoding heads. The folder "1.1billion-structure-inference" contains 9,300 (*.tar.gz) subdirectories, one per Frontier compute node used to execute the inference at exascale. Once uncompressed, each (*.tar.gz) subdirectory contains an ADIOS2 (*.bp) file container, where each atomistic structure is stored as a PyTorch-Geometric Data object. The file "export_dataset_environment_variables.sh" contains the environment variables that need to be set before running the HydraGNN code to reproduce the results provided in this dataset release. The code that can be used to load the ADIOS2 files, load HydraGNN models, and run inference is available at: https://github.com/ORNL/HydraGNN/releases/tag/v5.0

36 MATERIALS SCIENCE↗

Building America Indoor Temperature and Humidity Measurement Protocol

When modeling homes using simulation tools, the heating and cooling set points can have a significant impact on home energy use. Every 4 years the Energy Information Administration (EIA) Residential Energy Consumption Survey (RECS) asks homeowners about their heating and cooling set points. Unfortunately, no temperature data is measured, and most of the time, the homeowner may be guessing at this number. Even one degree Fahrenheit difference in heating set point can make a 5% difference in heating energy use! So, the survey-based RECS data cannot be used as the definitive reference for the set point for the 'average occupant' in simulations. The purpose of this document is to develop a protocol for collecting consistent data for heating/cooling set points and relative humidity so that an average set point can be determined for asset energy models in residential buildings. This document covers the decision making process for researchers to determine how many sensors should be placed in each home, where to put those sensors, and what kind of asset data should be taken while they are in the home. The authors attempted to design the protocols to maximize the value of this study and minimize the resources required to achieve that value.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

The Building America Indoor Temperature and Humidity Measurement Protocol

When modeling homes using simulation tools, the heating and cooling set points can have a significant impact on home energy use. Every four years, the Energy Information Administration (EIA) Residential Energy Consumption Survey (RECS) asks homeowners about their heating and cooling set points. Unfortunately, no temperature data is measured, and most of the time, the homeowner may be guessing at this number. Even one degree Fahrenheit difference in heating set point can make a 5% difference in heating energy use! So, the survey-based RECS data cannot be used as the definitive reference for the set point for the "average occupant" in simulations. The purpose of this document is to develop a protocol for collecting consistent data for heating/cooling set points and relative humidity so that an average set point can be determined for asset energy models in residential buildings. This document covers the decision making process for researchers to determine how many sensors should be placed in each home, where to put those sensors, and what kind of asset data should be taken while they are in the home. The authors attempted to design the protocols to maximize the value of this study and minimize the resources required to achieve that value.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Model America - data and models of every U.S. building

The 5-year goal of the 'Model America' concept was to generate a model of every building in the United States. This data repository delivers on that goal. Oak Ridge National Laboratory (ORNL) has developed the Automatic Building Energy Modeling (AutoBEM) software suite to process multiple types of data, extract building-specific descriptors, generate building energy models, and simulate them on High Performance Computing (HPC) resources. For more information, see AutoBEM-related publications (bit.ly/AutoBEM). There were 125,714,640 buildings detected in the United States and this dataset contains 122,930,327 (97.8%) buildings which resulted in a successful simulation. Future, annual updates have been proposed that may include additional buildings, data improvements, or other algorithmic enhancements. This dataset of 122.9 million buildings includes: Models (state_county.zip) - OpenStudio (v3.1.0) and EnergyPlus (v9.4) building energy models. Please note that the download requires the free Globus Connect Personal (https://www.globus.org/globus-connect-personal); Each model has approximately 3,000 building input descriptors that can be extracted. Please see the EnergyPlus(v9.4) 2,784-page Input/Output Reference Guide (https://energyplus.net/sites/all/modules/custom/nrel_custom/pdfs/pdfs_v9.4.0/InputOutputReference.pdf) for everything that can be retrieved or simulated from these models. These models were derived from the following metadata, which is not included in this dataset: 1. ID - unique building ID 2. County - county name 3. State - state name 4. CZ - ASHRAE Climate Zone designation 5. Clim_Zone - text label of climate zone 6. est_year - estimated year of construction 7. est_commercial - estimated building type (0=residential, 1=commercial) 8. Centroid - building center location in latitude/longitude (from Footprint2D) 9. Footprint2D - building polygon of 2D footprint (lat1/lon1_lat2/lon2_...) 10. Height - building height (meters) 11. Area2D - footprint area (ft2) 12. BuildingType - DOE prototype building designation (IECC=residential) as implemented by OpenStudio-standards 13. WWR_surfaces - percent of each facade (pair of points from Footprint2D) covered by fenestration/windows (average 14.5% for residential, 40% for commercial buildings) 14. NumFloors - number of floors (above-grade) 15. Area - estimate of total conditioned floor area (ft2) 16. Standard - building vintage. These models are made free and openly available in hopes of stimulating any simulation-informed use case. Data is provided as-is with no warranties, express or implied, regarding fitness for a particular purpose. We wish to thank our sponsors which include Oak Ridge National Laboratory (ORNL) Laboratory Directed Research and Development (LDRD), U.S. Dept. of Energy's (DOE) Building Technologies Office (BTO), Office of Electricity (OE), Biological and Environmental Research (BER), and National Nuclear Security Administration (NNSA). This research used resources of the Argonne Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC02-06CH11357. Please cite as: New, Joshua R., Adams, Mark, Bass, Brett, Berres, Anne, and Clinton, Nicholas (2021). 'Model America - data and models of every U.S. building. [Data set].' Constellation, doi.ccs.ornl.gov/ui/doi/339, April 14, 2021

24 POWER TRANSMISSION AND DISTRIBUTION↗