Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “performance data”

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 1,585 records · Page 88
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Near-Complete Sampling of Forest Structure from High-Density Drone Lidar Demonstrated by Ray Tracing

Drone lidar has the potential to provide detailed measurements of vertical forest structure throughout large areas, but a systematic evaluation of unsampled forest structure in comparison to independent reference data has not been performed. Here, we used ray tracing on a high-resolution voxel grid to quantify sampling variation in a temperate mountain forest in the southwest Czech Republic. We decoupled the impact of pulse density and scan-angle range on the likelihood of generating a return using spatially and temporally coincident TLS data. We show three ways that a return can fail to be generated in the presence of vegetation: first, voxels could be searched without producing a return, even when vegetation is present; second, voxels could be shadowed (occluded) by other material in the beam path, preventing a pulse from searching a given voxel; and third, some voxels were unsearched because no pulse was fired in that direction. We found that all three types existed, and that the proportion of each of them varied with pulse density and scan-angle range throughout the canopy height profile. Across the entire data set, 98.1% of voxels known to contain vegetation from a combination of coincident drone lidar and TLS data were searched by high-density drone lidar, and 81.8% of voxels that were occupied by vegetation generated at least one return. By decoupling the impacts of pulse density and scan angle range, we found that sampling completeness was more sensitive to pulse density than to scan-angle range. There are important differences in the causes of sampling variation that change with pulse density, scan-angle range, and canopy height. Our findings demonstrate the value of ray tracing to quantifying sampling completeness in drone lidar.

47 OTHER INSTRUMENTATION↗
DOE OSTI2021

Performance Evaluation of Comparative Vacuum Monitoring and Piezoelectric Sensors for Structural Health Monitoring of Rotorcraft Components

The costs associated with the increasing maintenance and surveillance needs of aging structures are rising at an unexpected rate. Multi-site fatigue damage, hidden cracks in hard-to-reach locations, disbonded joints, erosion, impact, and corrosion are among the major flaws encountered in today’s extensive fleet of aging aircraft and space vehicles. Aircraft maintenance and repairs represent about a quarter of a commercial fleet’s operating costs. The application of Structural Health Monitoring (SHM) systems using distributed sensor networks can reduce these costs by facilitating rapid and global assessments of structural integrity. The use of in-situ sensors for real-time health monitoring can overcome inspection impediments stemming from accessibility limitations, complex geometries, and the location and depth of hidden damage. Reliable, structural health monitoring systems can automatically process data, assess structural condition, and signal the need for human intervention. The ease of monitoring an entire on-board network of distributed sensors means that structural health assessments can occur more often, allowing operators to be even more vigilant with respect to flaw onset. SHM systems also allow for condition-based maintenance practices to be substituted for the current time-based or cycle-based maintenance approach thus optimizing maintenance labor. The Federal Aviation Administration has conducted a series of SHM validation and certification programs intended to comprehensively support the evolution and adoption of SHM practices into routine aircraft maintenance practices. This report presents one of those programs involving a Sandia Labs-aviation industry effort to move SHM into routine use for aircraft maintenance. The Airworthiness Assurance NDI Validation Center (AANC) at Sandia Labs, in conjunction with Sikorsky, Structural Monitoring Systems Ltd., Anodyne Electronics Manufacturing Corp., Acellent Technologies Inc., and the Federal Aviation Administration (FAA) carried out a trial validation and certification program to evaluate Comparative Vacuum Monitoring (CVM) and Piezoelectric Transducers (PZT) as a structural health monitoring solution to specific rotorcraft applications. Validation tasks were designed to address the SHM equipment, the health monitoring task, the resolution required, the sensor interrogation procedures, the conditions under which the monitoring will occur, the potential inspector population, adoption of CVM and PZT systems into rotorcraft maintenance programs and the document revisions necessary to allow for their routine use as an alternate means of performing periodic structural inspections. This program addressed formal SHM technology validation and certification issues so that the full spectrum of concerns, including design, deployment, performance and certification were appropriately considered. Sandia Labs designed, implemented, and analyzed the results from a focused and statistically relevant experimental effort to quantify the reliability of a CVM system applied to Sikorsky S-92 fuselage frame application and a PZT system applied to an S-92 main gearbox mount beam application. The applications included both local and global damage detection assessments. All factors that affect SHM sensitivity were included in this program: flaw size, shape, orientation and location relative to the sensors, as well as operational and environmental variables. Statistical methods were applied to performance data to derive Probability of Detection (POD) values for SHM sensors in a manner that agrees with current nondestructive inspection (NDI) validation requirements and is acceptable to both the aviation industry and regulatory bodies. The validation work completed in this program demonstrated the ability of both CVM and PZT SHM systems to detect cracks in rotorcraft components. It proved the ability to use final system response parameters to provide a Green Light/Red Light (“GO” – “NO GO”) decision on the presence of damage. In additional to quantifying the performance of each SHM system for the trial applications on the S-92 platform, this study also identified specific methods that can be used to optimize damage detection, guidance on deployment scenarios that can affect performance and considerations that must be made to properly apply CVM and PZT sensors. These results support the main goal of safely integrating SHM sensors into rotorcraft maintenance programs. Additional benefits from deploying rotorcraft Health and Usage Monitoring Systems (HUMS) may be realized when structural assessment data, collected by an SHM system, is also used to detect structural damage to compliment the operational environment monitoring. The use of in-situ sensors for health monitoring of rotorcraft structures can be a viable option for both flaw detection and maintenance planning activities. This formal SHM validation will allow aircraft manufacturers and airlines to confidently make informed decisions about the proper utilization of CVM and PZT technology. It will also streamline future regulatory actions and formal certification measures needed to assure the safe application of SHM solutions.

42 ENGINEERING↗
DOE OSTI2025

Machine learning for improved current-density reconstruction from two-dimensional vector magnetic images

The reconstruction of electrical current densities from magnetic field measurements is an important technique with applications in materials science, circuit design, quality control, plasma physics, and biology. Analytic reconstruction methods exist for planar currents, but break down in the presence of high-spatial-frequency noise or large standoff distance, restricting the types of systems that can be studied. Here, we demonstrate the use of a deep convolutional neural network for current density reconstruction from two-dimensional images of vector magnetic fields acquired by a quantum diamond microscope . Trained network performance significantly exceeds analytic reconstruction for data with high noise or large standoff distances. This machine learning technique can perform quality inversions on lower-signal-to-noise-ratio data, significantly reducing the data collection time and permitting reconstructions of weaker and three-dimensional current sources. Published by the American Physical Society 2025

Reed, Niko R. (ORCID:0009000305222403)↗
DOE OSTI2021

Multiscale Data-Driven Seismic Full-Waveform Inversion With Field Data Study

Seismic full-waveform inversion (FWI), which uses iterative methods to estimate high-resolution subsurface models from seismograms, is a powerful imaging technique in exploration geophysics. In recent years, the computational cost of FWI has grown exponentially due to the increasing size and resolution of seismic data. Moreover, it is a nonconvex problem and can encounter local minima due to the limited accuracy of the initial velocity models or the absence of low frequencies in the measurements. To overcome these computational issues, we develop a multiscale data-driven FWI method based on fully convolutional networks (FCNs). In preparing the training data, we first develop a real-time style transform method to create a large set of synthetic subsurface velocity models from natural images. We then develop two convolutional neural networks with encoder-decoder structures to reconstruct the low- and high-frequency components of the subsurface velocity models, separately. To validate the performance of our data-driven inversion method and the effectiveness of the synthesized training set, we compare it with conventional physics-based waveform inversion approaches using both synthetic and field data. Finally, these numerical results demonstrate that, once our model is fully trained, it can significantly reduce the computation time and yield more accurate subsurface velocity models in comparison with conventional FWI.

58 GEOSCIENCES↗
DOE OSTI2022

Performance trade-offs in reconfigurable networks for HPC

Designing efficient interconnects to support high-bandwidth and low-latency communication is critical toward realizing high performance computing (HPC) and data center (DC) systems in the exascale era. At extreme computing scales, providing the requisite bandwidth through overprovisioning becomes impractical. These challenges have motivated studies exploring reconfigurable network architectures that can adapt to traffic patterns at runtime using optical circuit switching. Despite the plethora of proposed architectures, surprisingly little is known about the relative performances and trade-offs among different reconfigurable network designs. We aim to bridge this gap by tackling two key issues in reconfigurable network design. First, we study how cost, power consumption, network performance, and scalability vary based on optical circuit switch (OCS) placement in the physical topology. Specifically, we consider two classes of reconfigurable architectures: one that places OCSs between top-of-rack (ToR) switches—ToR-reconfigurable networks (TRNs)—and one that places OCSs between pods of racks—pod-reconfigurable networks (PRNs). Second, we tackle the effects of reconfiguration frequency on network performance. Our results, based on network simulations driven by real HPC and DC workloads, show that while TRNs are optimized for low fan-out communication patterns, they are less suited for carrying high fan-out workloads. PRNs exhibit better overall trade-off, capable of performing comparably to a fully non-blocking fat tree for low fan-out workloads, and significantly outperform TRNs for high fan-out communication patterns.

Teh, Min Yee↗
DOE OSTI2022

Determination of the argon spectral function from ( e , e ' p ) data

The E12-14-012 experiment, performed in Jefferson Lab Hall A, has measured the $(e, e'p)$ cross section in parallel kinematics using a natural argon target. Here, we report the full results of the analysis of the data set corresponding to beam energy 2.222 GeV, and spanning the missing momentum and missing energy range $15 \lesssim p_m \lesssim 300$ MeV/c and $12 \lesssim E_m \lesssim 80$ MeV. The reduced cross section, determined as a function of $p_m$ and $E_m$ with $\approx$4\% accuracy, has been fitted using the results of Monte Carlo simulations involving a model spectral function and including the effects of final state interactions. The overall agreement between data and simulations turns out to be quite satisfactory ($\chi^2$/n.d.o.f.=1.9). Furthermore, the resulting spectral function will provide valuable new information, needed for the interpretation of neutrino interactions in liquid argon detectors.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2023

Knowledge-Informed Uncertainty-Aware Machine Learning for Time Series Forecasting of Dynamical Engineered Systems

The high complexity and multiscale nature of many engineered systems—such as those in nuclear power plants—make representing and forecasting their dynamic behavior challenging. Physics-based models can be overly complex and computationally intractable, whereas machine learning (ML) tools are often data-hungry and prone to unphysical solutions. This study proposes a knowledge-informed ML-aided hybrid residual modeling approach that offers accurate and efficient time series forecasting for the operation of dynamical engineered systems. Hybrid residual modeling entails a baseline solution from domain knowledge and known physics expressions about the system dynamics integrated with an ML model to capture undiscovered information from the mismatch (i.e., residuals) between true states from measurements and baseline-predicted outputs. This study further quantifies the ML model uncertainty to provide trustworthy solutions. Real-time operational data from thermal-hydraulic flow loops of the cryogenic moderator system in Oak Ridge National Laboratory’s Spallation Neutron Source facility were used to demonstrate the potential of knowledge-informed uncertainty-aware ML in real-world applications. The state variables of the cryogenic helium loop were modeled with (1) first principles–based system identification (sysID), (2) long short-term memory (LSTM) neural network, and (3) hybrid sysID (baseline) + LSTM (residual). The superior predictive capability of the sysID+LSTM model versus stand-alone sysID and LSTM is confirmed by average performance metrics and individual data points across different prediction horizons. By creating a robust representation of the underlying physical system, the widely applicable hybrid residual modeling approach will enable the future development of digital twins for performance prediction, prognostics, and operation control.

Zhao, Xingang↗
DOE OSTI2021

Rapid estimation of photosynthetic leaf traits of tropical plants in diverse environmental conditions using reflectance spectroscopy

Tropical forests are one of the main carbon sinks on Earth, but the magnitude of CO 2 absorbed by tropical vegetation remains uncertain. Terrestrial biosphere models (TBMs) are commonly used to estimate the CO 2 absorbed by forests, but their performance is highly sensitive to the parameterization of processes that control leaf-level CO 2 exchange. Direct measurements of leaf respiratory and photosynthetic traits that determine vegetation CO 2 fluxes are critical, but traditional approaches are time-consuming. Reflectance spectroscopy can be a viable alternative for the estimation of these traits and, because data collection is markedly quicker than traditional gas exchange, the approach can enable the rapid assembly of large datasets. However, the application of spectroscopy to estimate photosynthetic traits across a wide range of tropical species, leaf ages and light environments has not been extensively studied. Here, we used leaf reflectance spectroscopy together with partial least-squares regression (PLSR) modeling to estimate leaf respiration ( R dark25 ), the maximum rate of carboxylation by the enzyme Rubisco ( V cmax25 ), the maximum rate of electron transport ( J max25 ), and the triose phosphate utilization rate ( T p25 ), all normalized to 25°C. We collected data from three tropical forest sites and included leaves from fifty-three species sampled at different leaf phenological stages and different leaf light environments. Our resulting spectra-trait models validated on randomly sampled data showed good predictive performance for V cmax25 , J max25 , T p25 and R dark25 (RMSE of 13, 20, 1.5 and 0.3 μmol m -2 s -1 , and R 2 of 0.74, 0.73, 0.64 and 0.58, respectively). The models showed similar performance when applied to leaves of species not included in the training dataset, illustrating that the approach is robust for capturing the main axes of trait variation in tropical species. We discuss the utility of the spectra-trait and traditional gas exchange approaches for enhancing tropical plant trait studies and improving the parameterization of TBMs.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Comparative Assessment of U-Net-Based Deep Learning Models for Segmenting Microfractures and Pore Spaces in Digital Rocks

Segmentation of high-resolution X-ray microcomputed tomography (µCT) images is crucial in digital rock physics (DRP), affecting the characterization and analysis of microscale phenomena in the porous media. The complexity of geological structures and nonideal scanning conditions pose significant challenges to conventional image segmentation approaches. Motivated by the recent increasing popularity of deep learning (DL) techniques in image processing, this work undertakes a comparative study of DL models, specifically U-Net and its variants, for segmenting multiple targets with distinguished features in digital rocks, including discrete fracture networks (DFNs), pore spaces, and solid rock. Particularly, DFNs have a smaller volumetric fraction over others, bringing in a substantial challenge of imbalanced segmentation. The primary focus is to evaluate the architecture and feature enhancement strategies of various DL models, including U-Net, attention U-Net, residual U-Net, U-Net++, and residual U-Net++. The models were designed as 2.5D, utilizing a central 2D image and its two adjacent upper and lower 2D images as input to provide a pseudo-3D context. In addition, because the ground truth of segmentation was unknown for real-world digital rocks, we created a benchmark data set following the inverse operations of segmentation. The data synthesis started from the label images (i.e., solid rock, pore spaces, and DFNs), followed by simulating partial volume blurring, adding random background noise, and introducing ring artifacts to mimic real raw X-ray µCT images. The data set, which included various rock types (i.e., sandstone and artificial data), scanning resolution, and magnitudes of noise and artifacts, was divided into training and testing data sets with a 90% and 10% ratio, respectively. Moreover, in addition to the conventional pixel-wise evaluation metrics, the physics-based metric of the lattice-Boltzmann method (LBM) simulated permeability provided more comprehensive assessments. The results demonstrated that the residual connections, nested architectures, and redesigned skip connections contribute to the model performance and give the residual U-Net++ the highest accuracy. The improvements were mainly on the boundaries and small targets, especially the DFNs, which dominate the interconnectivity and therefore affect the permeability greatly. This study also rigorously evaluated the efficiency and generalization of each model, demonstrating that the sophisticated architectures achieved excellent practicability and maintained robust performance on completely unseen data, ensuring their suitability for diverse and challenging DRP applications.

58 GEOSCIENCES↗
DOE OSTI2024

Evaluation of Deep Learning Model Architectures for Point-of-Care Ultrasound Diagnostics

Point-of-care ultrasound imaging is a critical tool for patient triage during trauma for diagnosing injuries and prioritizing limited medical evacuation resources. Specifically, an eFAST exam evaluates if there are free fluids in the chest or abdomen but this is only possible if ultrasound scans can be accurately interpreted, a challenge in the pre-hospital setting. In this effort, we evaluated the use of artificial intelligent eFAST image interpretation models. Widely used deep learning model architectures were evaluated as well as Bayesian models optimized for six different diagnostic models: pneumothorax (i) B- or (ii) M-mode, hemothorax (iii) B- or (iv) M-mode, (v) pelvic or bladder abdominal hemorrhage and (vi) right upper quadrant abdominal hemorrhage. Models were trained using images captured in 27 swine. Using a leave-one-subject-out training approach, the MobileNetV2 and DarkNet53 models surpassed 85% accuracy for each M-mode scan site. The different B-mode models performed worse with accuracies between 68% and 74% except for the pelvic hemorrhage model, which only reached 62% accuracy for all model architectures. These results highlight which eFAST scan sites can be easily automated with image interpretation models, while other scan sites, such as the bladder hemorrhage model, will require more robust model development or data augmentation to improve performance. With these additional improvements, the skill threshold for ultrasound-based triage can be reduced, thus expanding its utility in the pre-hospital setting.

47 OTHER INSTRUMENTATION↗
DOE OSTI2021

Techno-Economic Analysis of a Concentrating Solar Power Plant Using Redox-Active Metal Oxides as Heat Transfer Fluid and Storage Media

We present results for a one-dimensional quasi-steady-state thermodynamic model developed for a 111.7 MW e concentrating solar power (CSP) system using a redox-active metal oxide as the heat storage media and heat transfer agent integrated with a combined cycle air Brayton power block. In the energy charging and discharging processes, the metal oxide CaAl 0.2 Mn 0.8 O 2.9-δ (CAM28) undergoes a reversible, high temperature redox cycle including an endothermic oxygen-releasing reaction and exothermic oxygen-incorporation reaction. Concentrated solar radiation heats the redox-active oxide particles under partial vacuum to drive the reduction extent deeper for increased energy density at a fixed temperature, thereby increasing storage capacity while limiting the required on sun temperature. Direct counter-current contact of the reduced particles with compressed air from the Brayton compressor releases stored chemical and sensible energy, heating the air to 1,200°C at the turbine inlet while cooling and reoxidizing the particles. The cool oxidized particles recirculate through the solar receiver subsystem for another cycle of heating and reduction (oxygen release). We applied the techno-economic model to 1) size components, 2) examine intraday operation with varying solar insolation, 3) estimate annual performance characteristics over a simulated year, 4) estimate the levelized cost of electricity (LCOE), and 5) perform sensitivity analyses to evaluate factors that affect performance and cost. Simulations use hourly solar radiation data from Barstow, California to assess the performance of a 111.7 MW e system with solar multiples (SMs) varying from 1.2 to 2.4 and storage capacities of 6–14 h. The baseline system with 6 h storage and SM of 1.8 has a capacity factor of 54.2%, an increase from 32.3% capacity factor with no storage, and an average annual energy efficiency of 20.6%. Calculations show a system with an output of 710 GWh e net electricity per year, 12 h storage, and SM of 2.4 to have an installed cost of $329 million, and an LCOE of 5.98 ¢/kWh e . This value meets the U.S. Department of Energy’s SunShot 2020 target of 6.0 ¢/kWh e ( U. S Department of Energy, 2012 ), but falls just shy of the 5.0 ¢/kWh e 2030 CSP target for dispatchable electricity ( U. S Department of Energy, 2017 ). The cost and performance results are minimally sensitive to most design parameters. However, a one-point change in the weighted annual cost of capital from 8 to 7% (better understood as a 12.5% change) translates directly to an 11% decrease (0.66 ¢/kWhe) in the LCOE.

Gorman, Brandon T.↗
DOE OSTI2024

Consumer-Oriented Energy Use and Range Metrics for Battery Electric Vehicles

The present study was motivated by a need to expand information for consumers offered through the FuelEconomy.Gov website. To that end, a power-based modeling approach has been used to examine the effect of steady-speed driving on estimated range for model year 2020 – 2023 battery electric vehicles (BEVs). This approach allowed rapid study of a broader range of BEV models than could be accomplished through vehicle tests. Publicly accessible certification test results and other data were used to perform a regression between cycle-average tractive power requirements and the resulting electrical power. Importantly, this regression enabled estimation of electric power and energy use over a range of steady highway speeds. These analyses in turn allowed projection of vehicle range at differing speeds. The projections agree within 6% with available 65 MPH manufacturer test data. Analyses of vehicles from model years 2020 – 2023 show that the 5-cycle range and energy use values from the window stickers of new vehicles are reasonable values for steady-speed driving at 65 MPH. Range decreases by a median value of approximately 15% for each 10 MPH increase in speed. The 5-cycle energy use (in kW-Hr / 100 miles) and energy economy (in miles / kW-Hr) derived from the 5-cycle energy use are reasonable estimates for 65 MPH steady speed driving. Energy economy decreases by about 0.5 miles / kW-Hr for every 10 MPH increase in speed.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2022

Application of Advanced Metering Infrastructure Data to Advanced Utility System Operations: Cooperative Research and Development, CRADA Number CRD-17-00712 (Final Report)

The primary objective of the proposed project is to evaluate the capabilities of Advanced Metering Infrastructure (AMI) based controls for grid operations. This includes monitoring and control at the secondary transformer level using existing AMI infrastructure. In addition, some of the challenges of using a model-less control will also be identified. The lessons learned will be disseminated to other utilities around the country through peer-reviewed publications with the consent of SDG&E. The scope of the work includes using data from SDG&E's existing AMI infrastructure and working with third party vendors selected by SDG&E. The grid edge monitoring systems and controls provided by the selected vendors will perform grid operations using AMI data. The National Renewable Energy Laboratory (NREL) will perform a software and controller-based evaluation of the grid edge controller, intelligent devices and sensors, at the Energy System Integration Facility (ESIF) laboratory.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

A Bayesian model for multivariate discrete data using spatial and expert information with application to inferring building attributes

When modeling sparsely observed multivariate data, strong prior information elicited from experts can be used to bolster predictive accuracy and counteract sampling bias. Similarly, modeling autocorrelation in space can help make use of co-occurrence patterns present in many types of spatial data. To make use of both expert prior information and spatial structure, we propose a novel graphical model for a spatial Bayesian network developed specifically to address challenges in inferring the attributes of buildings from geographically sparse observational data. This model is implemented as the sum of a spatial multivariate Gaussian random field and a tabular conditional probability function in real-valued space prior to projection onto the probability simplex. This modeling form is especially suitable for the usage of prior information in the form of sets of atomic rules obtained from experts. To perform inference with missing data, we implement a Markov chain Monte Carlo scheme composed of alternating steps of Gibbs sampling of missing entries and Hamiltonian Monte Carlo for model parameters. A case study in building attribution is presented to highlight the advantages and limitations of this approach.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Peeking into the next decade in large-scale structure cosmology with its Effective Field Theory

After the successful full-shape analyses of BOSS data using the Effective Field Theory of Large-Scale Structure, we investigate what upcoming galaxy surveys might achieve. We introduce a “perturbativity prior” that ensures that loop terms are as large as theoretically expected, which is effective in the case of a large number of EFT parameters. After validating our technique by comparison with already-performed analyses of BOSS data, we provide Fisher forecasts using the one-loop prediction for power spectrum and bispectrum for two benchmark surveys: DESI and MegaMapper. We find overall great improvements on the cosmological parameters. In particular, we find that MegaMapper (DESI) should obtain at least a 12σ (2σ) evidence for non-vanishing neutrino masses, bound the curvature Ω k to 0.0012 (0.012), and primordial inflationary non-Gaussianities as follows: f NL loc. to ± 0.26 (3.3), f NL eq. to ±16 (92), f NL orth. to ± 4.2 (27). Such measurements would provide much insight on the theory of Inflation. We investigate the limiting factor of shot noise and ignorance of the EFT parameters.

cosmological parameters from LSS↗
DOE OSTI2022

Foundation Models of Scientific Knowledge for Chemistry: Opportunities, Challenges and Lessons Learned

Foundation models pre-trained on large corpora demonstrate significant gains across many natural language processing tasks and domains e.g., law, healthcare, education, etc. However, only limited efforts have investigated the opportunities and limitations of applying these powerful models to science and security applications. In this work we develop foundation models of scientific knowledge for chemistry to augment scientists with the advanced ability to perceive and reason at scale previously unimagined. Specifically, we build large-scale (1.47B parameter) general-purpose models for chemistry that can be effectively used to perform a wide range of in-domain and out-of-domain tasks. Evaluating these models in a zero-shot setting, we analyze the effect of model and data scaling, knowledge depth, and temporality on model performance in context of model training efficiency. Our novel findings demonstrate that (1) model size significantly contributes to the task performance when evaluated in a zero-shot setting; (2) data quality (aka diversity) affects model performance more than data quantity; (3) similarly, unlike previous work (Luu et al., 2021) temporal order of the documents in the corpus boosts model performance only for specific tasks, e.g., SciQ; and (4) models pre-trained from scratch perform better on in-domain tasks than those tuned from general-purpose models like Open AI’s GPT-2.

Foundation Models, Chemistry↗
DOE OSTI2022

Progress on Demonstration of a MOOSE-Based Coupled Capability for Hot Channel Factors in Fast Reactors

Hot channel factors (HCFs) are computed values that account for the impact on predicted peak fuel, cladding, and coolant temperatures due to uncertainties in the as-built reactor’s material properties and geometry as well as uncertainties due to modeling approximations. Reduction in computed HCF values via reduction or elimination of modeling approximations may translate to significant economic savings if the reactor power can be raised due to the extra temperature margin gained. While limited historical datasets exist for sodium-cooled fast reactors (SFRs), there are no available HCF data for lead-cooled fast reactors (LFRs) outside of work generated previously within NEAMS. The computation of HCFs involves insights from reactor physics, thermal fluids and heat conduction calculations to determine how the peak temperatures respond to various uncertainties in the design. Due to the significant advantages for multi-physics coupling offered by the MOOSE framework, Griffin (MOOSE-based reactor physics code), MOOSE Heat Conduction Module, and Cardinal (MOOSE-wrapped multi-physics application which includes the NekRS thermal fluids code) are being coupled together using the MOOSE MultiApp System to develop a highfidelity multi-physics modeling capability for HCF simulations. This high-fidelity coupling workflow may also be beneficial for other fast reactor applications in the future. In previous work, Griffin and NekRS were individually assessed to ensure the necessary capabilities were in place. This work describes initial efforts to couple the codes (including folding in the MOOSE Heat Conduction Module) and determining the workflow for the perturbed calculations which will leverage the Stochastic Tools Module (STM). To our knowledge, this is the first coupling of Griffin and NekRS as well as the first exploratory use of Stochastic Tools Module for Cardinal. In this report, the neutronics code Griffin, the heat conduction solver in MOOSE, and the MOOSE-wrapped application containing NekRS (Cardinal) are linked together to demonstrate the coupled capability. Griffin and Cardinal are linked dynamically by specifying shared libraries. Different coupling hierarchies are tested for selecting the most appropriate coupling strategy. A coupling scheme is selected based on the efficiency of calculation and ease of data communication. Multiple tests are performed to choose suitable mesh structure, model configurations, scheme setup and boundary conditions to avoid loss of energy due to data interpolation between different modules or weak imposition of fluxes in finite element codes. Computational experiments are performed to study the tolerance control of each type of iteration to avoid false convergence. The coupled capability is demonstrated in both single pin and 7-pin models based on LFR materials and geometry. The study finds that the use of too large a time step size in the heat conduction module can lead to temperature oscillation even though the heat conduction equation does not have a time-derivative kernel, but only the time-dependent boundary condition. A 7-pin model without duct region achieved good convergence in the coupled calculation while a 7 pin model with duct region experienced data communication issues which need to be resolved.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2021

Accelerating Binarized Neural Networks via Bit-Tensor-Cores in Turing GPUs

Despite foreseeing tremendous speedups over conventional deep neural networks, the performance advantage of binarized neural networks (BNNs) has merely been showcased on general-purpose processors such as CPUs and GPUs. In fact, due to being unable to leverage bit-level-parallelism with a word-based architecture, GPUs have been criticized for extremely low utilization (1%) when executing BNNs. Consequently, the latest tensorcores in NVIDIA Turing GPUs start to experimentally support bit computation. In this work, we look into this brand new bit computation capability and characterize its unique features. We show that the stride of memory access can significantly affect performance delivery and a data-format co-design is highly desired to support the tensorcores for achieving superior performance than existing software solutions without tensorcores. We realize the tensorcore-accelerated BNN design, particularly the major functions for fully-connect and convolution layers — bit matrix multiplication and bit convolution. Evaluations on two NVIDIA Turing GPUs show that, with ResNet-18, our BTC-BNN design can process ImageNet at a rate of 5.6K images per second, 77% faster than state-of-the-art. Our BNN approach is released on https://github.com/pnnl/TCBNN.

Li, Ang↗