SEARCH · Engineering Papers

Results for “performance modeling and prediction”

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

SourceResearch connectionFrom yearThrough year

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

At least 487 records · Page 27

All sources NASA NTRS DOE OSTI

DOE OSTI2023

Stream Temperature Prediction in a Shifting Environment: Explaining the Influence of Deep Learning Architecture

Stream temperature is a fundamental control on ecosystem health. Recent efforts incorporating process guidance into deep learning models for predicting stream temperature have been shown to outperform existing statistical and physical models. This performance is in part because deep learning architectures can actively learn spatiotemporal relationships that govern how water and energy propagate through a river network. However, exploration of how spatiotemporal awareness and process guidance influence a model's generalizability under shifting environmental conditions such as climate change is limited. Here, we use Explainable Artificial Intelligence (XAI) to interrogate how differing deep learning architectures affect a model's learned spatial and temporal dependencies, and how those learned dependencies affect a model's ability to maintain high accuracy when applied to unseen environmental conditions. Using the Delaware River Basin in the northeastern United States as a test case, we compare two spatiotemporally aware process–guided deep learning models for predicting stream temperature (a recurrent graph convolution network—RGCN, and a temporal convolution graph model—Graph WaveNet). Both models achieve equally high predictive performance when testing data are well represented in the training data (test root mean squared errors of 1.64°C and 1.65°C); however, Graph WaveNet significantly outperforms RGCN in 4 out of 5 experiments where test partitions represent different types of unseen environmental conditions. XAI results show that the architecture of Graph WaveNet leads to learned spatial relationships with greater fidelity to physical processes, and that this fidelity improves the generalizability of the model when applied to shifting and/or unseen environmental conditions.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2022

Causality guided machine learning model on wetland CH 4 emissions across global wetlands

Wetland CH 4 emissions are among the most uncertain components of the global CH 4 budget. The complex nature of wetland CH 4 processes makes it challenging to identify causal relationships for improving our understanding and predictability of CH 4 emissions. In this study, we used the flux measurements of CH 4 from eddy covariance towers (30 sites from 4 wetlands types: bog, fen, marsh, and wet tundra) to construct a causality-constrained machine learning (ML) framework to explain the regulative factors and to capture CH 4 emissions at sub-seasonal scale. We found that soil temperature is the dominant factor for CH 4 emissions in all studied wetland types. Ecosystem respiration (CO 2 ) and gross primary productivity exert controls at bog, fen, and marsh sites with lagged responses of days to weeks. Integrating these asynchronous environmental and biological causal relationships in predictive models significantly improved model performance. More importantly, modeled CH 4 emissions differed by up to a factor of 4 under a +1°C warming scenario when causality constraints were considered. These results highlight the significant role of causality in modeling wetland CH 4 emissions especially under future warming conditions, while traditional data-driven ML models may reproduce observations for the wrong reasons. Our proposed causality-guided model could benefit predictive modeling, large-scale upscaling, data gap-filling, and surrogate modeling of wetland CH 4 emissions within earth system land models.

54 ENVIRONMENTAL SCIENCES↗

NASA NTRS2017

Model-Based Thermal System Design Optimization for the James Webb Space Telescope

Spacecraft thermal model validation is normally performed by comparing model predictions with thermal test data and reducing their discrepancies to meet the mission requirements. Based on thermal engineering expertise, the model input parameters are adjusted to tune the model output response to the test data. The end result is not guaranteed to be the best solution in terms of reduced discrepancy and the process requires months to complete. A model-based methodology was developed to perform the validation process in a fully automated fashion and provide mathematical bases to the search for the optimal parameter set that minimizes the discrepancies between model and data. The methodology was successfully applied to several thermal subsystems of the James Webb Space Telescope (JWST). Global or quasiglobal optimal solutions were found and the total execution time of the model validation process was reduced to about two weeks. The model sensitivities to the parameters, which are required to solve the optimization problem, can be calculated automatically before the test begins and provide a library for sensitivity studies. This methodology represents a crucial commodity when testing complex, large-scale systems under time and budget constraints. Here, results for the JWST Core thermal system will be presented in detail.

Cataldo, Giuseppe↗

DOE OSTI2020

Understanding Bifacial PV Modeling: Raytracing and View Factor Models

As part of the PV Magazine Webinar 'View-factor vs. ray tracing - which bifacial modelling techniques should you use?,' the second part lead by NREL will explore the different rear-irradiance calculation software, which fall into two categories: view-factor and ray-tracing models. View factor models assume isotropic scattering of reflected rays, allowing for calculation of irradiance by integration. Due-diligence software such as PVSyst or SAM use the view-factor model. Ray-tracing models simulate multipath reflection and absorption of individual rays entering a scene. Raytracing software such as bifacial_radiance, which is the only available open-source toolkit, offers the possibility of reproducing complex scenes, including shading or finite-system edge effects. This can be useful to evaluate specific system geometries, shading and obstructions, areas of different albedo and electrical mismatch, all topics of interest to the prediction of bifacial performance. Model agreement for view factors and bifacial_radiance software is improving and has been shown to be between 2% (absolute) when compared with measured results.

14 SOLAR ENERGY↗

DOE OSTI2020

Mechanistic Models as Framework for Understanding Biomarker Disposition: Prediction of Creatinine‐Drug Interactions

Creatinine is widely used as a biomarker of glomerular filtration, and, hence, renal function. However, transporter‐mediated secretion also contributes to its renal clearance, albeit to a lesser degree. Inhibition of these transporters causes transient serum creatinine elevation, which can be mistaken as impaired renal function. The current study developed mechanistic models of creatinine kinetics within physiologically based framework accounting for multiple transporters involved in creatinine renal elimination, assuming either unidirectional or bidirectional‐OCT2 transport (driven by electrochemical gradient). Robustness of creatinine models was assessed by predicting creatinine‐drug interactions with 10 perpetrators; performance evaluation accounted for 5% intra‐individual variability in serum creatinine. Models showed comparable predictive performances of the maximum steady‐state effect regardless of OCT2 directionality assumptions. However, only the bidirectional‐OCT2 model successfully predicted the minimal effect of ranitidine. The dynamic nature of models provides clear advantage to static approaches and most advanced framework for evaluating interplay between multiple processes in creatinine renal disposition.

Scotcher, Daniel↗

NASA NTRS1980

Impurities in silicon solar cells

The paper investigates the effects of metallic impurities on the performance of silicon solar cells. Czochralski and polycrystalline ingots were employed with boron and phosphorus as primary dopants and with controlled additions of secondary impurities. The data obtained from over 200 crystals indicate that impurity-induced performance loss is primarily due to a reduction of the base diffusion length. Based on this observation, a model is developed which predicts cell performance as a function of secondary impurity concentrations. The model calculations are in good agreement with experimental values except for Cu, Ni, Fe, and to a lesser degree, carbon, which at higher concentrations degrade the cell by junction defect mechanisms.

Davis, J. R., Jr.↗

DOE OSTI2023

Machine learning based simultaneous control of air handling unit discharge air and condenser water temperatures set-point for minimized cooling energy in an office building

In this study, an artificial intelligence based real-time prediction and control model to optimize condenser water temperature and discharge air temperature (DAT) set-points in water-cooled air handling unit (AHU) system has been developed. EnergyPlus-MATLAB co-simulation has been conducted to analyze the developed model's effectiveness. Here, to develop artificial neural networks (ANN) model, embedded neural network objects in MATLAB was utilized. The developed model could decide an optimal temperature set-points based on outdoor air wet-bulb temperature to reflect the Korean climate context. As a result, the developed ANN prediction model showed the predictive performance of Cv(RMSE) of approximately 21%. Compared to the conventional fixed temperature algorithm, which fixes AHU DAT at 14°C and condenser water temperature at 32°C, the ANN based optimized control showed a 22% total cooling energy reduction. These results show that significant energy savings can be achieved by simultaneously controlling condenser water temperature and AHU DAT set-points considering Korean climatic characteristics using AI technologies such as ANN models.

24 POWER TRANSMISSION AND DISTRIBUTION↗

NASA NTRS1992

Engine Performance and Health Monitoring Models Using Steady State and Transient Prediction Methods

This lecture note discusses the role of computer modelling in the design, development, and validation of a performance monitoring system. The basic requirements of an engine health monitoring system are discussed in the context of the user environment. A form of model based on stage characteristics provides the basis for describing engine measurements. Faults are modelled in accordance with empirical data obtained from tests conducted at the stage level. The model is used to investigate various parameters that provide unambiguous identification of the fault in question. Fault libraries have been developed for field use.

B D MacIsaac↗

DOE OSTI2024

Interpretable Machine Learning Models for Practical Antimonate Electrocatalyst Performance

Computationally predicting the performance of catalysts under reaction conditions is a challenging task due to the complexity of catalytic surfaces and their evolution in situ, different reaction paths, and the presence of solid-liquid interfaces in the case of electrochemistry. We demonstrate here how relatively simple machine learning models can be found that enable prediction of experimentally observed onset potentials. Inputs to our model are comprised of data from the oxygen reduction reaction on non-precious transition-metal antimony oxide nanoparticulate catalysts with a combination of experimental conditions and computationally affordable bulk atomic and electronic structural descriptors from density functional theory simulations. From human-interpretable genetic programming models, we identify key experimental descriptors and key supplemental bulk electronic and atomic structural descriptors that govern trends in onset potentials for these oxides and deduce how these descriptors should be tuned to increase onset potentials. Here, we finally validate these machine learning predictions by experimentally confirming that scandium as a dopant in nickel antimony oxide leads to a desired onset potential increase. Macroscopic experimental factors are found to be crucially important descriptors to be considered for models of catalytic performance, highlighting the important role machine learning can play here even in the presence of small datasets.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2020

Phase Equilibria and Thermochemistry of Advanced Fuels: Modeling Burnup Behavior (Final Report)

Achieving the goal of developing advanced fuel concepts that meet the DOE objectives of being robust, demonstrating high performance, and are more tolerant of accident conditions than current fuel systems will require a thorough understanding of the thermophysical and thermochemical properties of the constituent materials. Non-oxide fuel systems are being explored under the Advanced Fuels Program that hold significant promise for improved performance and accident tolerance, including the uranium silicide-based system considered in the current work. Prospective cladding materials currently considered that contribute to improved accident tolerance include silicon carbide composites and ferritic alloys (Fe-Cr-Al base compositions). Thus, the effort developed thermochemical models and values, supported with targeted experiments, to evaluate the ferritic alloy and silicon carbide composite cladding systems in contrast to current zirconium alloy cladding. The developed detailed understanding will serve to aid in relatively early screening of candidate systems to avoid wasted effort, guide development of new fuel forms, and to provide a basis for predicting and modeling fuel performance. Major deliverables for the project included: Thermochemical assessment and models of phases in the U-Si and U-Si-N systems; thermochemical evaluation supported by experimental measurements of fuel-cladding interactions of silicide fuel with baseline zirconium, ceramic composite, and ferritic alloy cladding; thermochemical assessment and models of phases supported by experimental measurements for silicide fuel with key fission products provided in a dataset and reported in refereed publications. Within the project a significantly refined U-Si phase diagram was developed and reported that now includes homogeneity ranges for key phases, such at the U 3 Si 2 proposed fuel phase, and settles issues with regard to uncertainty in the stability of some phases. Computational efforts together with key experiments has determined phase formation in interactions between U 3 Si 2 and Zircaloy-4 cladding material, a ferritic FeCrAlY alloy of interest as an advanced cladding material, and silicon carbide, also of interest as a fiber-reinforced composite cladding. As expected, very significant reactions occur between U3Si2 and Zircaloy-4, with much less interaction at higher temperatures for the ferritic alloy, and finally interactions with SiC only in the region of contact. A potentially major issue is the stability of U 3 Si 2 fuel that has undergone significant burnup. The result is the loss of the uranium metal, liberating silicon, and the formation of concomitant fission product elements that either dissolve in the U 3 Si 2 phase or form independent, and possibly silicide phases. A combination of experimental determinations of phase formation of U 3 Si 2 reacted with representative fission products yttrium, gadolinium, cerium, zirconium, and molybdenum and first principles calculations has helped understand the fuel chemistry. The behavior of the U 3 Si 2 phase and the partitioning of silicon to possible fission product phases with burnup was thus determined, with significant dissolution in U 3 Si 2 of cerium, gadolinium, zirconium, and plutonium predicted along with independent phase formation of a U-Mo-Si ternary phase, yttrium silicide, and elemental selenium. It can be concluded that at significant burnup there will be a very minor amount of the U 3 Si 2 fuel phase that will decompose to a lower silicide or a uranium alloy as silicon preferentially forms a secondary phase.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

NASA NTRS1995

Centrifugal and Axial Pump Design and Off-Design Performance Prediction

A meanline pump-flow modeling method has been developed to provide a fast capability for modeling pumps of cryogenic rocket engines. Based on this method, a meanline pump-flow code PUMPA was written that can predict the performance of pumps at off-design operating conditions, given the loss of the diffusion system at the design point. The design-point rotor efficiency and slip factors are obtained from empirical correlations to rotor-specific speed and geometry. The pump code can model axial, inducer, mixed-flow, and centrifugal pumps and can model multistage pumps in series. The rapid input setup and computer run time for this meanline pump flow code make it an effective analysis and conceptual design tool. The map-generation capabilities of the code provide the information needed for interfacing with a rocket engine system modeling code. The off-design and multistage modeling capabilities of PUMPA permit the user to do parametric design space exploration of candidate pump configurations and to provide head-flow maps for engine system evaluation.

Veres, Joseph P.↗

NASA NTRS1997

Mir Cooperative Solar Array Flight Performance Data and Computational Analysis

The Mir Cooperative Solar Array (MCSA) was developed jointly by the United States (US) and Russia to provide approximately 6 kW of photovoltaic power to the Russian space station Mir. The MCSA was launched to Mir in November 1995 and installed on the Kvant-1 module in May 1996. Since the MCSA photovoltaic panel modules (PPMs) are nearly identical to those of the International Space Station (ISS) photovoltaic arrays, MCSA operation offered an opportunity to gather multi-year performance data on this technology prior to its implementation on ISS. Two specially designed test sequences were executed in June and December 1996 to measure MCSA performance. Each test period encompassed 3 orbital revolutions whereby the current produced by the MCSA channels was measured. The temperature of MCSA PPMs was also measured. To better interpret the MCSA flight data, a dedicated FORTRAN computer code was developed to predict the detailed thermal-electrical performance of the MCSA. Flight data compared very favorably with computational performance predictions. This indicated that the MCSA electrical performance was fully meeting pre-flight expectations. There were no measurable indications of unexpected or precipitous MCSA performance degradation due to contamination or other causes after 7 months of operation on orbit. Power delivered to the Mir bus was lower than desired as a consequence of the retrofitted power distribution cabling. The strong correlation of experimental and computational results further bolsters the confidence level of performance codes used in critical ISS electric power forecasting. In this paper, MCSA flight performance tests are described as well as the computational modeling behind the performance predictions.

Kerslake, Thomas W.↗

DOE OSTI2022

Grain growth kinetics of the gamma phase metallic uranium

We report metallic uranium is a leading fuel form for sodium cooled fast reactors as an enabling technology of future nuclear energy systems. Mechanistic understanding of fuel behaviors and kinetics under thermodynamic equilibrium and highly non-equilibrium conditions are essential for evaluating fuel performance. It is important to understand and predict the grain and pore evolutions of metallic fuels under thermal and irradiation conditions. However, very limited data are available on the grain growth kinetics and mechanisms of pure gamma phase uranium. In this paper, the pure gamma uranium pellets with different grain structures were fabricated by combining high-energy ball milling and spark plasma sintering. Isothermal annealing tests were performed to investigate the grain growth behavior of the pure gamma phase uranium with different initial grain sizes. A parabolic relationship in grain growth with time was identified for the submicron-sized (374 nm) sample. In contrast, for the nano-sized (137 nm) sample, the grain growth shows a linear relationship with time. The activation energies of grain growth were determined as 199.5 KJ/mol and 80.6 KJ/mol for nano-sized and submicron-sized grain structures, respectively. For the nano-sized sample, the rate-control step of grain growth is dominated by the triple-junction migration, in which the grain boundary triple junction drags the grain growth, leading to a higher activation energy than the bulk diffusion. The dominating mechanism for the submicron-sized sample is grain boundary diffusion. The mechanistic understanding and critical data obtained on the kinetics of pure uranium phases will be useful to evaluate fuel behavior under thermodynamic equilibrium conditions and develop a high fidelity model to predict fuel performance.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2021

Information-Theoretic Bounds on Quantum Advantage in Machine Learning

We study the performance of classical and quantum machine learning (ML) models in predicting outcomes of physical experiments. The experiments depend on an input parameter x and involve execution of a (possibly unknown) quantum process E. Our figure of merit is the number of runs of E required to achieve a desired prediction performance. We consider classical ML models that perform a measurement and record the classical outcome after each run of E, and quantum ML models that can access E coherently to acquire quantum data; the classical or quantum data are then used to predict the outcomes of future experiments. We prove that for any input distribution D(x), a classical ML model can provide accurate predictions on average by accessing E a number of times comparable to the optimal quantum ML model. In contrast, for achieving an accurate prediction on all inputs, we prove that the exponential quantum advantage is possible. For example, to predict the expectations of all Pauli observables in an n-qubit system ρ, classical ML models require 2 Ω(n) copies of ρ, but we present a quantum ML model using only O(n) copies. Our results clarify where the quantum advantage is possible and highlight the potential for classical ML models to address challenging quantum problems in physics and chemistry.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

NASA NTRS1998

Performance Modeling and Measurement of Parallelized Code for Distributed Shared Memory Multiprocessors

This paper presents a model to evaluate the performance and overhead of parallelizing sequential code using compiler directives for multiprocessing on distributed shared memory (DSM) systems. With increasing popularity of shared address space architectures, it is essential to understand their performance impact on programs that benefit from shared memory multiprocessing. We present a simple model to characterize the performance of programs that are parallelized using compiler directives for shared memory multiprocessing. We parallelized the sequential implementation of NAS benchmarks using native Fortran77 compiler directives for an Origin2000, which is a DSM system based on a cache-coherent Non Uniform Memory Access (ccNUMA) architecture. We report measurement based performance of these parallelized benchmarks from four perspectives: efficacy of parallelization process; scalability; parallelization overhead; and comparison with hand-parallelized and -optimized version of the same benchmarks. Our results indicate that sequential programs can conveniently be parallelized for DSM systems using compiler directives but realizing performance gains as predicted by the performance model depends primarily on minimizing architecture-specific data locality overhead.

Waheed, Abdul↗

NASA NTRS2009

Application of TRMM PR and TMI Measurements to Assess Cloud Microphysical Schemes in the MM5 Model for a Winter Storm

Seen by the human eye, precipitation particles are commonly drops of rain, flakes of snow, or lumps of hail that reach the ground. Remote sensors and numerical models usually deal with information about large collections of rain, snow, and hail (or graupel --also called soft hail ) in a volume of air. Therefore, the size and number of the precipitation particles and how particles interact, evolve, and fall within the volume of air need to be represented using physical laws and mathematical tools, which are often implemented as cloud and precipitation microphysical parameterizations in numerical models. To account for the complexity of the precipitation physical processes, scientists have developed various types of such schemes in models. The accuracy of numerical weather forecasting may vary dramatically when different types of these schemes are employed. Therefore, systematic evaluations of cloud and precipitation schemes are of great importance for improvement of weather forecasts. This study is one such endeavor; it pursues quantitative assessment of all the available cloud and precipitation microphysical schemes in a weather model (MM5) through comparison with the observations obtained by National Aeronautics and Space Administration (NASA) s and Japan Aerospace Exploration Agency (JAXA) s Tropical Rainfall Measuring Mission (TRMM) precipitation radar (PR) and microwave imager (TMI). When satellite sensors (like PR or TMI) detect information from precipitation particles, they cannot directly observe the microphysical quantities (e.g., water species phase, density, size, and amount etc.). Instead, they tell how much radiation is absorbed by rain, reflected away from the sensor by snow or graupel, or reflected back to the satellite. On the other hand, the microphysical quantities in the model are usually well represented in microphysical schemes and can be converted to radiative properties that can be directly compared to the corresponding PR and TMI observations. This study employs this method to evaluate the accuracy of the simulated radiative properties by the MM5 model with different microphysical schemes. It is found that the representations of particle density, size, and mass in the different schemes in the MM5 model determine the model s performance when predicting a winter storm over the eastern Pacific Ocean. Schemes lacking moderate density particles (i.e. graupel), with snow flakes that are too large, or with excessive mass of snow or graupel lead to degraded prediction of the radiative properties as observed by the TRMM satellite. This study demonstrates the uniqueness of the combination of both an active microwave sensor (PR) and passive microwave sensor (TMI) onboard TRMM on assessing the accuracy of numerical weather forecasting. It improves our understanding of the physical and radiative properties of different types of precipitation particles and provides suggestions for better representation of cloud and precipitation processes in numerical models. It would, ultimately, contribute to answering questions like "Why did it not rain when the forecast says it would?"

Han, Mei↗

DOE OSTI2021

Bayesian learning of orthogonal embeddings for multi-fidelity Gaussian Processes

Uncertainty propagation in complex engineering systems often poses significant computational challenges related to modeling and quantifying probability distributions of model outputs, as those emerge as the result of various sources of uncertainty that are inherent in the system under investigation. Gaussian Processes regression (GPs) is a robust meta-modeling technique that allows for fast model prediction and exploration of response surfaces. Multi-fidelity variations of GPs further leverage information from cheap and low fidelity model simulations in order to improve their predictive performance on the high fidelity model. In order to cope with the high volume of data required to train GPs in high dimensional design spaces, a common practice is to introduce latent design variables that are typically projections of the original input space to a lower dimensional subspace, and therefore substitute the problem of learning the initial high dimensional mapping, with that of training a GP on a low dimensional space. Here in this paper, we present a Bayesian approach to identify optimal transformations that map the input points to low dimensional latent variables. The \projection" mapping consists of an orthonormal matrix that is considered a priori unknown and needs to be inferred jointly with the GP parameters, conditioned on the available training data. The proposed Bayesian inference scheme relies on a two-step iterative algorithm that samples from the marginal posteriors of the GP parameters and the projection matrix respectively, both using Markov Chain Monte Carlo (MCMC) sampling. In order to take into account the orthogonality constraints imposed on the orthonormal projection matrix, a Geodesic Monte Carlo sampling algorithm is employed, that is suitable for exploiting probability measures on manifolds. We extend the proposed framework to multi-fidelity models using GPs including the scenarios of training multiple outputs together. We validate our framework on three synthetic problems with a known lower-dimensional subspace. The benefits of our proposed framework, are illustrated on the computationally challenging aerodynamic optimization of a last-stage blade for an industrial gas turbine, where we study the effect of an 85-dimensional shape parameterization of a three-dimensional airfoil on two output quantities of interest, specifically on the aerodynamic efficiency and the degree of reaction

42 ENGINEERING↗

NASA NTRS1976

A report on the aerodynamic design and wind tunnel test of a Prop-Fan model

Recent transport aircraft system studies have shown that the Prop-Fan propulsion system offers the potential for significant fuel savings over high bypass turbofans employing the same core engine technology for each. This important finding has encouraged more detailed study of the Prop-Fan aerodynamics and has led to NASA sponsorship of a wind tunnel research program to explore the projected high efficiency levels of advanced propeller configurations for 0.80 Mach number, high-altitude cruise operation. The aerodynamic design philosophy for the Prop-Fan model is discussed. The geometric characteristics and predicted performance of the wind tunnel model are presented. Finally, the preliminary test results are reviewed and compared to the performance goals originally projected.

Rohrbach, C.↗