Search NASA⌕ Search

SEARCH · Search NASA

Results for “Performance Model”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 811 records · Page 45

Xyce Case Study

With the elimination of underground nuclear testing and declining defense budgets, science-based stockpile stewardship requires increased reliance on high performance modeling and simulation of weapon systems. Today's weapon systems are comprised of various electrical components and systems. As a result, there is a need for tools that will allow the use of massively parallel modeling and simulation techniques on high performance computers in existing and future weapons' electrical systems models. The Xyce Parallel Electronic Simulator is a SPICE (Simulation Program with Integrated Circuit Emphasis)- compatible circuit simulator designed to run on large-scale parallel computing platforms, though it can also execute efficiently on a variety of architectures including single processor workstations. As a mature platform for large-scale parallel circuit simulation, Xyce supports standard capabilities available in commercial simulators, in addition to various devices and models specific to Sandia's needs. Specifically, Xyce aids in the design and verification of electrical and electronic circuits and systems prior to weapons' manufacturing and deployment.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

Teleseismic Network Association with GENIE

In this report we investigate adapting the Graph Neural Interpretation Engine (GENIE), an associator developed for three-component dense monitoring networks, to regional to teleseismic association using a sparse network of array stations. We expand GENIE’s input features to include first-P detection time, azimuth, and slowness estimates. Additionally, we include a probability of detection (PDET) term which measures a station’s likelihood of detecting an event. To assess each feature’s relative importance, we train four models, each using an increasing set of node features and find that the PDET models perform the best. We define two measures of event complexity which demonstrate that all GENIE model versions perform better than the standard backprojection stack.

47 OTHER INSTRUMENTATION↗

Subhourly Clipping Correction Model Comparison

This work will compare the Allen method and Walker method of accounting for subhourly inverter clipping power losses in hourly PV performance models. The Allen method uses a matrix lookup based on DNI clearness and clipping potential to assign a clipping correction loss at each simulation timestep. The Walker method models the PV DC power input to the inverter as a distribution over the hourly timestep and uses integration over the timestep to determine the amount of clipping that occurs within the timestep. Both these models have been recently implemented in the System Advisor Model's (SAM) open-source code, and will be applied to hourly SURFRAD datasets to analyze the subhourly clipping loss predicted by each model for different system designs and inverter loading conditions. Both models will be compared to "true" 1-minute SURFRAD data simulations to see their accuracy against more accurate 1-minute clipping correction loss predictions. This model comparisons will be investigated in more detail at the PVSC conference in Seattle, Washington June 2024.

ENERGY PLANNING, POLICY, AND ECONOMY,MATHEMATICS A↗

Modeling needs for very large systems.

Most system performance models assume a point measurement for irradiance and that, except for the impact of shading from nearby obstacles, incident irradiance is uniform across the array. Module temperature is also assumed to be uniform across the array. For small arrays and hourly-averaged simulations, this may be a reasonable assumption. Stein is conducting research to characterize variability in large systems and to develop models that can better accommodate large system factors. In large, multi-MW arrays, passing clouds may block sunlight from a portion of the array but never affect another portion. Figure 22 shows that two irradiance measurements at opposite ends of a multi-MW PV plant appear to have similar irradiance (left), but in fact the irradiance is not always the same (right). Module temperature may also vary across the array, with modules on the edges being cooler because they have greater wind exposure. Large arrays will also have long wire runs and will be subject to associated losses. Soiling patterns may also vary, with modules closer to the source of soiling, such as an agricultural field, receiving more dust load. One of the primary concerns associated with this effort is how to work with integrators to gain access to better and more comprehensive data for model development and validation.

14 SOLAR ENERGY↗

Physics-informed neural network with transfer learning (TL-PINN) based on domain similarity measure for prediction of nuclear reactor transients

Nuclear reactor safety and efficiency can be enhanced through the development of accurate and fast methods for prediction of reactor transient (RT) states. Physics informed neural networks (PINNs) leverage deep learning methods to provide an alternative approach to RT modeling. Applications of PINNs in monitoring of RTs for operator support requires near real-time model performance. However, as with all machine learning models, development of a PINN involves time-consuming model training. Here, we show that a transfer learning (TL-PINN) approach achieves significant performance gain, as measured by reduction of the number of iterations for model training. Using point kinetic equations (PKEs) model with six neutron precursor groups, constructed with experimental parameters of the Purdue University Reactor One (PUR-1) research reactor, we generated different RTs with experimentally relevant range of variables. The RTs were characterized using Hausdorff and Fréchet distance. We have demonstrated that pre-training TL-PINN on one RT results in up to two orders of magnitude acceleration in prediction of a different RT. The mean error for conventional PINN and TL-PINN models prediction of neutron densities is smaller than 1%. We have developed a correlation between TL-PINN performance acceleration and similarity measure of RTs, which can be used as a guide for application of TL-PINNs.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

RDMA vs. RPC for Implementing Distributed Data Structures

Distributed data structures are key to implementing scalable applications for scientific simulations and data analysis. In this paper we look at two implementation styles for distributed data structures: remote direct memory access (RDMA) and remote procedure call (RPC). We focus on operations that require individual accesses to remote portions of a distributed data structure, e.g., accessing a hash table bucket or distributed queue, rather than global operations in which all processors collectively exchange information. We look at the trade-offs between the two styles through microbenchmarks and a performance model that approximates the cost of each. The RDMA operations have direct hardware support in the network and therefore lower latency and overhead, while the RPC operations are more expressive but higher cost and can suffer from lack of attentiveness from the remote side. We also run experiments to compare the real-world performance of RDMA- and RPC-based data structure operations with the predicted performance to evaluate the accuracy of our model, and show that while the model does not always precisely predict running time, it allows us to choose the best implementation in the examples shown. We believe this analysis will assist developers in designing data structures that will perform well on current network architectures, as well as network architects in providing better support for this class of distributed data structures.

distributed data structures↗

Use of multiscale zirconium alloy deformation models in nuclear fuel behavior analysis

Accurate prediction of cladding mechanical behavior is a key aspect of modeling nuclear fuel behavior, especially for conditions of pellet-cladding interaction (PCI), reactivity-initiated accidents (RIA), and loss of coolant accidents (LOCA). Current approaches to fuel performance modeling rely on empirical constitutive models for cladding creep, growth and plastic deformation, which are limited to the materials and conditions for which the models were developed. To improve upon this approach, a microstructurally-based zirconium alloy mechanical deformation analysis capability is being developed within the United States Department of Energy Consortium for Advanced Simulation of Light Water Reactors (CASL). Specifically, the viscoplastic self-consistent (VPSC) polycrystal plasticity modeling approach, developed by Lebensohn and Tomé [1], has been coupled with the BISON engineering scale fuel performance code to represent the mechanistic material processes controlling the deformation behavior of light water reactor (LWR) cladding. A critical component of VPSC is the representation of the crystallographic nature (defect and dislocation movement) and orientation of the grains within the matrix material and the ability to account for the role of texture on deformation. A future goal is for VPSC to obtain information on reaction rate kinetics from atomistic calculations to inform the defect and dislocation behavior models described in VPSC. The multiscale modeling of cladding deformation mechanisms allowed by VPSC far exceed the functionality of typical semi-empirical constitutive models employed in nuclear fuel behavior codes to model irradiation growth and creep, thermal creep, or plasticity. This paper describes the implementation of an interface between VPSC and BISON and provides initial results utilizing the coupled functionality.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Forecasting influenza activity using machine-learned mobility map

Human mobility is a primary driver of infectious disease spread. However, existing data is limited in availability, coverage, granularity, and timeliness. Data-driven forecasts of disease dynamics are crucial for decision-making by health officials and private citizens alike. In this work, we focus on a machine-learned anonymized mobility map (hereon referred to as AMM) aggregated over hundreds of millions of smartphones and evaluate its utility in forecasting epidemics. We factor AMM into a metapopulation model to retrospectively forecast influenza in the USA and Australia. We show that the AMM model performs on-par with those based on commuter surveys, which are sparsely available and expensive. We also compare it with gravity and radiation based models of mobility, and find that the radiation model’s performance is quite similar to AMM and commuter flows. Additionally, we demonstrate our model’s ability to predict disease spread even across state boundaries. Our work contributes towards developing timely infectious disease forecasting at a global scale using human mobility datasets expanding their applications in the area of infectious disease epidemiology.

60 APPLIED LIFE SCIENCES↗

Duplications and losses of genes encoding known elements of the stress defence system of the Aspergilli contribute to the evolution of these filamentous fungi but do not directly influence their environmental stress tolerance

The contribution of stress protein duplication and deletion events to the evolution of the Aspergilli was studied. We performed a large-scale homology analysis of stress proteins and generated and analysed three stress defence system models based on Saccharomyces cerevisiae , Schizosaccharomyces pombe and Aspergillus nidulans . Although both yeast-based and A. nidulans -based models were suitable to trace evolutionary changes, the A. nidulans -based model performed better in mapping stress protein radiations. The strong Mantel correlation found between the positions of species in the phylogenetic tree on the one hand and either in the A. nidulans -based or S. cerevisiae -based models on the other hand demonstrated that stress protein expansions and reductions contributed significantly to the evolution of the Aspergilli. Interestingly, stress tolerance attributes correlated well with the number of orthologs only for a few stress proteins. Notable examples are Ftr1 iron permease and Fet3 ferro-O 2 -oxidoreductase, elements of the reductive iron assimilation pathway, in the S. cerevisiae -based model, as well as MpkC, a HogA-like mitogen activated protein kinase in the A. nidulans-based model. In the case of the iron assimilation proteins, the number of orthologs showed a positive correlation with H 2 O 2 -induced stress tolerance while the number of MpkC orthologs correlated positively with Congo Red induced cell wall stress, sorbitol induced osmotic stress and H 2 O 2 induced oxidative stress tolerances. For most stress proteins, changes in the number of orthologs did not correlate well with any stress tolerance attributes. As a consequence, stress tolerance patterns of the studied Aspergilli did not correlate with either the sets of stress response proteins in general or with the phylogeny of the species studied. These observations suggest that stress protein duplication and deletion events significantly contributed to the evolution of stress tolerance attributes of Aspergilli. In contrast, there are other processes, which may counterbalance the effects of stress gene duplications or deletions including (i) alterations in the structures of stress proteins leading to changes in their biological activities, (ii) varying biosynthesis of stress proteins, (iii) rewiring stress response regulatory networks or even (iv) acquiring new stress response genes by horizontal gene transfer. As a result, all these multilevel changes are indispensable for the successful adaptation of filamentous fungi to altering environmental conditions, especially when these organisms are entering new ecological niches.

59 BASIC BIOLOGICAL SCIENCES↗

Thrust and efficiency model for electron-driven magnetic nozzles

A performance model is presented for magnetic nozzle plasmas driven by electron thermal expansion to investigate how the thrust coefficient and beam divergence efficiency scale with the incoming plasma flow and magnetic field geometry. Using a transformation from cylindrical to magnetic coordinates, an approximate analytical solution is derived to the axisymmetric two-fluid equations for a collisionless plasma flow along an applied magnetic field. This solution yields an expression for the half-width at half-maximum of the plasma density profile in the far-downstream region, from which simple scaling relations for the thrust coefficient and beam divergence efficiency are derived. It is found that the beam divergence efficiency is most sensitive to the density profile of the flow into the nozzle throat, with the highest efficiencies occurring for plasmas concentrated along the nozzle axis. Increasing the expansion ratio of the magnetic field leads to efficiency improvements that are more pronounced for incoming plasmas that are not concentrated along the axis. This implies that the additional magnet required to increase the expansion ratio may be worth the added complexity for plasma sources that exhibit poor confinement.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

PUMA:POWDER UTILIZATION MODELING APPLICATION

SF-25-084 PUMA a high performance modeling framework to simulate powder processing. It provides a scalable tool for manufacturers to simulate powder pre- and post-processing. The tool can predict the distortion, residual stress, and (for reactive processes) reaction completion fraction of complex parts after curing/debinding, sintering, and infiltration processes. These predictions are key metrics industry uses to optimize these processes to produce dense, defect-free, stable components.

HU, TIANCHEN (GARY)↗

SAM Photovoltaic Model Technical Reference

This manual describes the photovoltaic performance model in the System Advisor Model (SAM). The U.S. Department of Energy’s National Renewable Energy Laboratory maintains and distributes SAM, which is available as a free download from https://sam.nrel.gov. These descriptions are based on SAM 2015.1.30 (SSC 41).

14 SOLAR ENERGY↗

Future Projections of the Large-Scale Meteorology Associated with California Heat Waves in CMIP5 Models

Previous work showed that climate models capture historical large-scale meteorological patterns (LSMPs) associated with California Central Valley heat waves including both ways these heat waves form. This work examines what models predict under the Representative Concentration Pathway (RCP) 4.5 and RCP8.5 scenarios. Model performance varies, so a multimodel average weights each model based on its historical performance in four parameters. An LSMP index (LSMPi) defined using upper atmosphere variables captures dates of past extreme surface temperature maxima. LSMPi correlates well with all values of California Central Valley-average surface maximum temperature. LSMPi distributions in future simulations shift ~0.6 standard deviations higher between 1961–2000 and 2061–2100 for RCP8.5 data. Based on the historical climatology, future scenarios show a large increase in the frequency and duration of heat waves in every model. Four times as many heat waves occur and their median duration doubles, using historical thresholds. Of the two ways heat waves form, Type 1 has similar frequency in the future. But, Type 2 becomes much more common because Type 2 has a preexisting hot anomaly in Southwestern Canada, much like the historical to future climatological change in that region (a “global warming” signal). The 20-year return value anomaly increases by 30–40%. The average of the 50 hottest temperatures increases 3.5–6 K depending on the scenario. When extreme values are defined using the future climatology, the models and their average have no consistent increase or decrease of distribution properties such as shape, scale, and return values of the extremes compared to historical values.

54 ENVIRONMENTAL SCIENCES↗

Machine learning framework for predicting uranium enrichments from M400 CZT gamma spectra

A machine learning framework was developed for predicting uranium enrichments from M400 CZT gamma spectra. This framework leverages the availability of a large amount of measured M400 gamma spectra and uses a recently updated version of Gamma Detector Response and Analysis Software (GADRAS) for gamma spectrum analysis and generation. It also leverages the existing machine learning modules in Python for gamma spectrum data processing, curation, model training, benchmarking, and optimization of the deep machine learning models. The framework is used to develop a deep learning model to analyze gamma spectra from a set of U 3 O 8 samples with enrichments ranging from 0.31 to 93.17% and UF 6 cylinders with enrichments ranging from 0.2 to 4.95%, and the model performance is tested using a set of measured spectra and the respective declared enrichment values. Results show that the model can correctly classify 99.35% of the U 3 O 8 sample enrichments, and can predict the samples’ enrichments within an average absolute error of 0.099% (in percentage points of enrichment). For the UF 6 cylinders, the average absolute error was approximately 0.03%, with an accuracy of 98% in classifying discrete enrichment values of UF 6 samples. Finally, the results also show that the model has performed significantly better in terms of predicting enrichments in UF 6 cylinders based on measured gamma spectra than the GEM code, with a standard deviation (of the relative errors) of 2.23% (compared with the 11.51% value for the GEM code) based on results from a set of test data.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Comparison of Machine Learning-Based Predictive Models of the Nutrient Loads Delivered from the Mississippi/Atchafalaya River Basin to the Gulf of Mexico

Predicting nutrient loads is essential to understanding and managing one of the environmental issues faced by the northern Gulf of Mexico hypoxic zone, which poses a severe threat to the Gulf’s healthy ecosystem and economy. The development of hypoxia in the Gulf of Mexico is strongly associated with the eutrophication process initiated by excessive nutrient loads. Due to the complexities in the excessive nutrient loads to the Gulf of Mexico, it is challenging to understand and predict the underlying temporal variation of nutrient loads. The study was aimed at identifying an optimal predictive machine learning model to capture and predict nonlinear behavior of the nutrient loads delivered from the Mississippi/Atchafalaya River Basin (MARB) to the Gulf of Mexico. For this purpose, monthly nutrient loads (N and P) in tons were collected from US Geological Survey (USGS) monitoring station 07373420 from 1980 to 2020. Machine learning models—including autoregressive integrated moving average (ARIMA), gaussian process regression (GPR), single-layer multilayer perceptron (MLP), and a long short-term memory (LSTM) with the single hidden layer—were developed to predict the monthly nutrient loads, and model performances were evaluated by standard assessment metrics—Root Mean Square Error (RMSE) and Correlation Coefficient (R). The residuals of predictive models were examined by the Durbin–Watson statistic. The results showed that MLP and LSTM persistently achieved better accuracy in predicting monthly TN and TP loads compared to GPR and ARIMA. In addition, GPR models achieved slightly better test RMSE score than ARIMA models while their correlation coefficients are much lower than ARIMA models. Moreover, MLP performed slightly better than LSTM in predicting monthly TP loads while LSTM slightly outperformed for TN loads. Furthermore, it was found that the optimizer and number of inputs didn’t show effects on the LSTM performance while they exhibited impacts on MLP outcomes. This study explores the capability of machine learning models to accurately predict nonlinearly fluctuating nutrient loads delivered to the Gulf of Mexico. Further efforts focus on improving the accuracy of forecasting using hybrid models which combine several machine learning models with superior predictive performance for nutrient fluxes throughout the MARB.

54 ENVIRONMENTAL SCIENCES↗

Study of Gamma-ray Production from Neutron-induced Activity on Spacecraft for DIORAMA

The Distributed Infrastructure Offering Real{Time Access to Modeling and Analysis (DIORAMA) software provides performance modeling capabilities of the United States Nuclear Detonation Detection System (USNDS) with a focus on the characterization of Space{Based Nuclear Detonation Detection (SNDD) instrument performance [1]. A study was conducted investigating the neutron activation on spacecraft caused by cosmic{ray and trapped proton{induced spallation, and by direct irradiation from hypothetical nuclear detonations as a background component to be incorporated into DIORAMA. The neutron activation background study amasses a database of pre-calculated Monte Carlo data that reflects the delayed gamma-ray production from activated nuclei as a function of incident cosmic or neutron flux, incident energy, material volume, etc. CINDER2008 coupled with MCNPX and the Activation Analysis Tools in the Activation in Accelerator Radiation Environments (AARE) distribution along with a suite of Python scripts are used to calculate and assemble the Monte Carlo database [2{6]. The neutron activation background study improves the fidelity of DIORAMA by providing the neutron-induced activation component to the background of the instruments on-board the spacecraft, and has implications for future instrument design studies.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Detonation performance experiments and modeling for the DAAF-based high explosive PBX 9701

Detonation performance experiments and modeling are reported for the explosive PBX 9701, which is composed of 97% 3,3$\acute{-}$diamino-4,4$\acute{-}$azoxyfurazan (DAAF) and 3% vinylidene uoride-chlorotriuoroethylene copolymer (Kel-F 800) binder by weight. PBX 9701 is a newly-developed reducedsensitivity explosive with increased performance relative to the triaminotrinitrobenzene or TATB-based PBX 9502 while still retaining low sensitivity to mechanical insult. The first detonation performance measurements for this formulation are presented, including front-curvature ratesticks and cylinder expansion test data. Prior shock initiation data is also reviewed. These data are used to develop programmed burn and reactive flow calibrations for existing commonly-used performance models which allow engineering calculations with PBX 9701. The calibration process involves several enhancements relative to conventional approaches including the use of an analytical scaling correlation to speed the calibration process and development of a new methodology to improve the consistency between the programmed burn and reactive model calibration process and the resulting calculations. Overall, PBX 9701 is seen to exhibit improved performance relative to insensitive explosives, with a trinitrotoluene (TNT) equivalence of 1.24. The detonation performance properties are found to be well captured by existing models.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

Trade Study for Neutron Transport at Low Earth Orbit: Adding Fidelity to DIORAMA

The Distributed Infrastructure Offering Real-Time Access to Modeling and Analysis (DIORAMA) software provides performance modeling capabilities of the United States Nuclear Detonation Detection System (USNDS) with a focus on the characterization of Space-Based Nuclear Detonation Detection (SNDD) instrument performance [1]. A case study was done to add the neutron propagation capabilities of DIORAMA to low earth orbit (LEO), and compare the back-calculated incident energy from the time-of- ight (TOF) spectrum with the scored incident energy spectrum. As the scoring altitude lowers, the time increase due to scattering takes up much more of the fraction of total TOF; whereas at geosynchronous earth orbit (GEO), the time increase due to scattering is a negligible fraction of the total TOF [2]. The scattering smears out the TOF enough to make the back-calculation of the initial energy spectrum from the TOF spectrum very convoluted.

79 ASTRONOMY AND ASTROPHYSICS↗