Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “model inference”

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 991 records · Page 55
All sourcesNASA NTRSDOE OSTI
NASA NTRS2019

A NICER view of PSR J0030+0451: Implications for the Dense Matter Equation of State

Both the mass and radius of the millisecond pulsar PSRJ0030+0451 have been inferred via pulse-profile modeling of X-ray data obtained by NASA’s Neutron Star Interior Composition Explorer (NICER) mission. In this Letter we study the implications of the mass–radius inference reported for this source by Riley et al. for the dense matter equation of state (EoS), in the context of prior information from nuclear physics at low densities. Using a Bayesian framework we infer central densities and EoS properties for two choices of high-density extensions: a piecewise-polytropic model and a model based on assumptions of the speed of sound in dense matter. Around nuclear saturation density these extensions are matched to an EoS uncertainty band obtained from calculations based on chiral effective field theory interactions, which provide a realistic description of atomic nuclei as well as empirical nuclear matter properties within uncertainties. We further constrain EoS expectations with input from the current highest measured pulsar mass; together, these constraints offer a narrow Bayesian prior informed by theory as well as laboratory and astrophysical measurements. The NICER mass–radius likelihood function derived by Riley et al. using pulse-profile modeling is consistent with the highest-density region of this prior. The present relatively large uncertainties on mass and radius for PSR J0030+0451 offer, however, only a weak posterior information gain over the prior. We explore the sensitivity to the inferred geometry of the heated regions that give rise to the pulsed emission, and find a small increase in posterior gain for an alternative (but less preferred) model. Lastly, we investigate the hypothetical scenario of increasing the NICER exposure time for PSRJ0030+0451.

G Raaijmakers↗
DOE OSTI2016

Exploiting jet binning to identify the initial state of high-mass resonances

If a new high-mass resonance is discovered at the Large Hadron Collider, model-independent techniques to identify the production mechanism will be crucial to understand its nature and effective couplings to Standard Model particles. In this paper, we present a powerful and model-independent method to infer the initial state in the production of any high-mass color-singlet system by using a tight veto on accompanying hadronic jets to divide the data into two mutually exclusive event samples (jet bins). For a resonance of several hundred GeV, the jet binning cut needed to discriminate quark and gluon initial states is in the experimentally accessible range of several tens of GeV. It also yields comparable cross sections for both bins, making this method viable already with the small event samples available shortly after a discovery. Theoretically, the method is made feasible by utilizing an effective field theory setup to compute the jet cut dependence precisely and model independently and to systematically control all sources of theoretical uncertainties in the jet binning, as well as their correlations. Finally, we use a 750 GeV scalar resonance as an example to demonstrate the viability of our method.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS2012

The Structure and Emission Model of the Relativistic Jet in the Quasar 3C 279 Inferred From Radio To High-Energy Gamma-Ray Observations in 2008-2010

We present time-resolved broad-band observations of the quasar 3C 279 obtained from multiwavelength campaigns conducted during the first two years of the Fermi Gamma-ray Space Telescope mission. While investigating the previously reported gamma-ray/optical flare accompanied by a change in optical polarization, we found that the optical emission appears delayed with respect to the gamma-ray emission by about 10 days. X-ray observations reveal a pair of 'isolated' flares separated. by approx. 90 days, with only weak gamma-ray/optical counterparts. The spectral structure measured by Spitzer reveals a synchrotron component peaking in the mid-infrared band with a sharp break at the far-infrared band during the gamma-ray flare, while the peak appears in the mm/sub-mm band in the low state. Selected spectral energy distributions are fitted with leptonic models including Comptonization of external radiation produced in a dusty torus or the broad-line region. Adopting the interpretation of the polarization swing involving propagation of the emitting region along a curved trajectory, we can explain the evolution of the broad-band spectra during the gamma-ray flaring event by a shift of its location from approx. 1 pc to approx. 4 pc from the central black hole. On the other hand, if the gamma-ray flare is generated instead at sub-pc distance from the central black hole, the far-infrared break can be explained by synchrotron self-absorption. We also model the low spectral state, dominated by the mm/sub-mm peaking synchrotron component, and suggest that the corresponding inverse-Compton component explains the steady X-ray emission.

Source record↗
DOE OSTI2024

Enhancing the Quality and Reliability of Machine Learning Interatomic Potentials through Better Reporting Practices

Recent developments in machine learning interatomic potentials (MLIPs) have empowered even nonexperts in machine learning to train MLIPs for accelerating materials simulations. However, reproducibility and independent evaluation of presented MLIP results is hindered by a lack of clear standards in current literature. In this Perspective, we aim to provide guidance on best practices for documenting MLIP use while walking the reader through the development and deployment of MLIPs including hardware and software requirements, generating training data, training models, validating predictions, and MLIP inference. We also suggest useful plotting practices and analyses to validate and boost confidence in the deployed models. Finally, we provide a step-by-step checklist for practitioners to use directly before publication to standardize the information to be reported. Altogether, we hope that our work will encourage the reliable and reproducible use of these MLIPs, which will accelerate their ability to make a positive impact in various disciplines including materials science, chemistry, and biology, among others.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Unifying Advective and Diffusive Descriptions of Bedform Pumping in the Benthic Biolayer of Streams

Abstract Many water quality and ecosystem functions performed by streams occur in the benthic biolayer, the biologically active upper (~5 cm) layer of the streambed. Solute transport through the benthic biolayer is facilitated by bedform pumping, a physical process in which dynamic and static pressure variations over the surface of stationary bedforms (e.g., ripples and dunes) drive flow across the sediment‐water interface. In this paper we derive two predictive modeling frameworks, one advective and the other diffusive, for solute transport through the benthic biolayer by bedform pumping. Both frameworks closely reproduce patterns and rates of bedform pumping previously measured in the laboratory, provided that the diffusion model's dispersion coefficient declines exponentially with depth. They are also functionally equivalent, such that parameter sets inferred from the 2D advective model can be applied to the 1D diffusive model, and vice versa. The functional equivalence and complementary strengths of these two models expand the range of questions that can be answered, for example, by adopting the 2D advective model to study the effects of geomorphic processes (such as bedform adjustments to land use change) on flow‐dependent processes and the 1D diffusive model to study problems where multiple transport mechanisms combine (such as bedform pumping and turbulent diffusion). By unifying 2D advective and 1D diffusive descriptions of bedform pumping, our analytical results provide a straightforward and computationally efficient approach for predicting, and better understanding, solute transport in the benthic biolayer of streams and coastal sediments.

Grant, Stanley B.↗
DOE OSTI2026

DELVE Milky Way Satellite Galaxy Census. I. Satellite Population and Survey Selection Function in DES, DELVE, and Pan-STARRS

The properties of Milky Way satellite galaxies have important implications for galaxy formation, reionization, and the fundamental physics of dark matter. However, the population of Milky Way satellites includes the faintest known galaxies, and current observations are incomplete. To understand the impact of observational selection effects on the known satellite population, we perform rigorous, quantitative estimates of the Milky Way satellite galaxy detection efficiency in three wide-field survey datasets: the Dark Energy Survey Year 6, the DECam Local Volume Exploration Data Release 3, and the Pan-STARRS1 Data Release 1. Together, these surveys cover ∼13,600 deg 2 to g ∼ 24.0 and ∼27,700 deg 2 to g ∼ 22.5, spanning ∼91% of the high-Galactic-latitude sky (∣b∣ ≥ 15°). We apply multiple detection algorithms over the combined footprint and recover 49 known satellites above a strict census detection threshold. To characterize the sensitivity of our census, we run our detection algorithms on a large set of simulated galaxies injected into the survey data, which allows us to develop models that predict the detectability of satellites as a function of their properties. We then fit an empirical model to our data and infer the luminosity function, radial distribution, and size–luminosity relation of Milky Way satellite galaxies. Our empirical model predicts a total of $265^{+79}_{-47}$ satellite galaxies with −20 ≤ M V ≤ 0, half-light radii of 15 ≤ r 1/2 , (pc) ≤ 3000, and galactocentric distances of 10 ≤ D GC (kpc) ≤ 300. We also identify a mild anisotropy in the angular distribution of the observed galaxies, at a significance of ∼2σ, which can be attributed to the clustering of satellites associated with the LMC.

Tan, Chin Yi [Univ. of Chicago, IL (United States)↗
DOE OSTI2022

A Search of the Full Six Years of the Dark Energy Survey for Outer Solar System Objects

We present a search for outer solar system objects in the 6 yr of data from the Dark Energy Survey (DES). The DES covered a contiguous 5000 deg 2 of the southern sky with ≈80,000 3 deg 2 exposures in the grizY filters between 2013 and 2019. This search yielded 812 trans-Neptunian objects (TNOs), one Centaur and one Oort cloud comet, 458 reported here for the first time. We present methodology that builds upon our previous search on the first 4 yr of data. All images were reprocessed with an optimized detection pipeline that leads to an average completeness gain of 0.47 mag per exposure, as well as improved transient catalog production and algorithms for linkage of detections into orbits. All objects were verified by visual inspection and by the “sub-threshold significance,” the signal-to-noise ratio in the stack of images in which its presence is indicated by the orbit, but no detection was reported. This yields a pure catalog complete to r ≈ 23.8 mag and distances 29 < d < 2500 au. The TNOs have minimum (median) of 7 (12) nights’ detections and arcs of 1.1 (4.2) yr, and will have grizY magnitudes available in a further publication. We present software for simulating our observational biases for comparisons of models to our detections. Initial inferences demonstrating the catalog’s statistical power are: the data are inconsistent with the CFEPS-L7 model for the classical Kuiper Belt; the 16 “extreme” TNOs (a > 150 au, q > 30 au) are consistent with the null hypothesis of azimuthal isotropy; and nonresonant TNOs with q > 38 au, a > 50 au show a significant tendency to be sunward of major mean-motion resonances.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
NASA NTRS2003

Mechanosensitive Ion Channels in Bacteria: Functional Domains and Mechanisms of Gating

The past funding period was productive for the group. The progress in the mechanosensitive channel field was critically affected in the end of 1998 by the solution of the crystal structure of the mycobacterial homolog of MscL by our colleagues from Caltech. Having the structure of TbMscL in the closed state, we developed a detailed homology model of EcoMscL, and related the structural model with the wealth of functional phenomenology available for the E. coli version of the channel (EcoMscL). The biophysical properties of the open MscL helped to model the open conformation and infer the pathway for the entire gating transition. The following experiments provided strong support to the atomic model of the gating process, and allowed to make further predictions. The work has advanced our understanding of tension-driven conformational transitions in membrane-embedded mechanosensory proteins, determine major energetic contributions and set the stage for further exploration of the whole family of mechanosensitive channels. The results have been published in seven experimental and theoretical papers, with three other papers currently in press or in preparation.

Sukharev, Sergei↗
DOE OSTI2020

On the quantification and efficient propagation of imprecise probabilities with copula dependence

This paper addresses the problem of quantification and propagation of uncertainties associated with dependence modeling when data for characterizing probability models are limited. Practically, the system inputs are often assumed to be mutually independent or correlated by a multivariate Gaussian distribution. However, this subjective assumption may introduce bias in the response estimate if the real dependence structure deviates from this assumption. In this work, we overcome this limitation by introducing a flexible copula dependence model to capture complex dependencies. Here, a hierarchical Bayesian multimodel approach is proposed to quantify uncertainty in dependence model-form and model parameters that result from small data sets. This approach begins by identifying, through Bayesian multimodel inference, a set of candidate marginal models and their corresponding model probabilities, and then estimating the uncertainty in the copula-based dependence structure, which is conditional on the marginals and their parameters. The overall uncertainties integrating marginals and copulas are probabilistically represented by an ensemble of multivariate candidate densities. A novel importance sampling reweighting approach is proposed to efficiently propagate the overall uncertainties through a computational model. Through an example studying the influence of constituent properties on the out-of-plane properties of transversely isotropic E-glass fiber composites, we show that the composite property with copula-based dependence model converges to the true estimate as data set size increases, while an independence or arbitrary Gaussian correlation assumption leads to a biased estimate.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2017

Turbulent mixing and removal of ozone within an Amazon rainforest canopy

Simultaneous profiles of turbulence statistics and mean ozone mixing ratio are used to establish a relation between eddy diffusivity and ozone mixing within the Amazon forest. A one–dimensional diffusion model is proposed and used to infer mixing time scales from the eddy diffusivity profiles. Data and model results indicate that during daytime conditions, the upper (lower) half of the canopy is well (partially) mixed most of the time and that most of the vertical extent of the forest can be mixed in less than an hour. During nighttime, most of the canopy is predominantly poorly mixed, except for periods with bursts of intermittent turbulence. Even though turbulence is faster than chemistry during daytime, both processes have comparable time scales in the lower canopy layers during nighttime conditions. Nonchemical loss time scales (associated with stomatal uptake and dry deposition) for the entire forest are comparable to turbulent mixing time scale in the lower canopy during the day and in the entire canopy during the night, indicating a tight coupling between turbulent transport and dry deposition and stomatal uptake processes. Because of the significant time of day and height variability of the turbulent mixing time scale inside the canopy, it is important to take it into account when studying chemical and biophysical processes happening in the forest environment. In conclusion, the method proposed here to estimate turbulent mixing time scales is a reliable alternative to currently used models, especially for situations in which the vertical distribution of the time scale is relevant.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2013

A strategy for improved computational efficiency of the method of anchored distributions

This paper proposes a strategy for improving the computational efficiency of model inversion using the method of anchored distributions by “bundling” similar model parametrizations in the likelihood function. Inferring the likelihood function typically requires a large number of forward model simulations for each possible model parametrization; as a result, the process is quite expensive. To ease this prohibitive cost, we present an approximation for the likelihood function – called “bundling” – that relaxes the requirement for high quantities of forward model simulations. This approximation redefines the conditional statement of the likelihood function as the probability of sets of similar model parametrizations – “bundles” – replicating field measurements. To evaluate the effectiveness of these modifications, we compare the quality of predictions and computational cost of bundling relative to a baseline method of anchored distributions inversion of three-dimensional flow and transport model parameters. For our synthetic experiment, bundling achieved a 35% reduction in overall computational cost and had a limited negative on predicted probability distributions of the model parameters. Strategies for minimizing error in the bundling approximation, for enforcing similarity amongst the sets of model parametrizations, and for identifying convergence of the likelihood function are also presented.

Bayesian Inversion, SVD, computational efficiency↗
DOE OSTI2017

Investigation of Homogeneous Relaxation Model Parameters and their Implications for Gasoline Injectors

Flash boiling is known to be a common phenomenon for gasoline direct injection (GDI) engine sprays. The Homogeneous Relaxation Model has been adopted in many recent numerical studies for predicting cavitation and flash boiling. The Homogeneous Relaxation Model is assessed in this study. Sensitivity analysis of the model parameters has been documented to infer the driving factors for the flash-boiling predictions. The model parameters have been varied over a range and the differences in predictions of the extent of flashing have been studied. Apart from flashing in the near nozzle regions, mild cavitation is also predicted inside the gasoline injectors. The variation in the predicted time scales through the model parameters for predicting these two different thermodynamic phenomena (cavitation, flash) have been elaborated in this study. Turbulence model effects have also been investigated by comparing predictions from the standard and Re-Normalization Group (RNG) k-ε turbulence models.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2021

Spectroscopy outperforms leaf trait relationships for predicting photosynthetic capacity across different forest types

Leaf trait relationships are widely used to predict ecosystem function in Terrestrial Biosphere Models (TBMs), in which leaf maximum carboxylation capacity (V c,max ), an important trait for modelling photosynthesis, can be inferred from other easier- to-measure traits. However, whether trait-V c,max relationships are robust across different forest types remains unclear. Here we used measurements of leaf traits, including one morphological trait (leaf mass per area), three biochemical traits (leaf water content, area-based leaf nitrogen content, and leaf chlorophyll content), one physiological trait (V c,max ), as well as leaf reflectance spectra, and explored their relationships within and across three contrasting forest types in China. We found weak and forest type-specific relationships between V c,max and the four morphological and biochemical traits (R 2 ≤ 0.15), indicated by significantly changing slopes and intercepts across forest types. In contrast, reflectance spectroscopy effectively collapsed the differences in the trait-V c,max relationships across three forest biomes into a single robust model for V c,max (R 2 = 0.77), and also accurately estimated the four traits (R 2 = 0.75-0.94). Furthermore, these findings challenge the traditional use of empirical trait-V c,max relationships in TBMs for estimating terrestrial plant photosynthesis, but also highlight spectroscopy as an efficient alternative for characterizing V c,max and multi-trait variability, with critical insights into ecosystem modeling and functional trait ecology.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

A comparison of histopathology imaging comprehension algorithms based on multiple instance learning

Whole slide imaging (WSI), also called digital virtual microscopy, is a new imaging modality. It allows for the application of AI and machine learning methods to cancer pathology to help establish a means for the automatic diagnosis of cancer cases. However, designing machine-learning models for WSI is computationally challenging due to its required ultra-high resolution. The current state-of-the-art models use multiple instance learning (MIL). MIL is a weakly-supervised learning method in which the model uses an array of inferences from many smaller instances to make a final classification about the entire set. In the context of WSI, researchers divide the ultra-high-resolution image into many patches. The model then classifies the slide based on an array of inferences from the patches. Among several ways of making the final classification, attention-based mechanisms have resulted in superb accuracy scores. The Transformer, one attention-based algorithm, has reported substantial improvements for WSI comprehension tasks. In this project, we studied and compared several WSI comprehension algorithms. We used the following three datasets: CAMELYON16+17, TCGALung, and TCGA-Kidney. We found that attention-based MIL algorithms performed better than standard MIL algorithms for classifying WSI images, achieving a higher mean accuracy and AUC. However, none of the attention-based algorithms performed significantly better than the others, reporting accuracy scores that varied widely. Presumably, it is due to the limited availability of training samples in the data corpus. Since it is not easy to increase the samples from human subjects, some machine learning techniques like transfer learning could help mitigate this issue.

Saunders, Adam↗
DOE OSTI2024

Automated Bayesian high-throughput estimation of plasma temperature and density from emission spectroscopy

Here, this paper introduces a novel approach for automated high-throughput estimation of plasma temperature and density using atomic emission spectroscopy, integrating Bayesian inference with sophisticated physical models. We provide an in-depth examination of Bayesian methods applied to the complexities of plasma diagnostics, supported by a robust framework of physical and measurement models. Our methodology is demonstrated using experimental observations in the field of magneto-inertial fusion, focusing on individual and sequential shot analyses of the Plasma Liner Experiment at LANL. The results demonstrate the effectiveness of our approach in enhancing the accuracy and reliability of plasma parameter estimation and in using the analysis to reveal the deep hidden structure in the data. This study not only offers a new perspective of plasma analysis but also paves the way for further research and applications in nuclear instrumentation and related domains.

Bayesian inference↗
DOE OSTI2013

A BAKE-OFF BETWEEN CMFGEN AND FASTWIND: MODELING THE PHYSICAL PROPERTIES OF SMC AND LMC O-TYPE STARS

The model atmosphere programs FASTWIND and CMFGEN are both elegantly designed to perform non-LTE analyses of the spectra of hot massive stars, and include sphericity and mass-loss. The two codes differ primarily in their approach toward line blanketing, with CMFGEN treating all of the lines in the co-moving frame and FASTWIND taking an approximate approach which speeds up execution times considerably. Although both have been extensively used to model the spectra of O-type stars, no studies have used the codes to independently model the same spectra of the same stars and compare the derived physical properties. We perform this task on 10 O-type stars in the Magellanic Clouds. For the late-type O supergiants, both CMFGEN and FASTWIND have trouble fitting some of the He I lines, and we discuss causes and cures. We find that there is no difference in the average effective temperatures found by the two codes for the stars in our sample, although the dispersion is large, due primarily to the various difficulties each code has with He I. The surface gravities determined using FASTWIND are systematically lower by 0.12 dex compared to CMFGEN, a result we attribute to the better treatment of electron scattering by CMFGEN. This has implications for the interpretation of the origin of the so-called mass discrepancy, as the masses derived by FASTWIND are on average lower than inferred from stellar evolutionary models, while those found by CMFGEN are in better agreement.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Understanding the fusion yield dependencies in OMEGA DT-layered implosion experiments using a physics-based statistical mapping model

Improving the performance of inertial confinement fusion implosions requires physics models that can accurately predict the response to changes in the experimental inputs. Good predictive capability has been demonstrated for the fusion yield using a statistical mapping of simulated outcomes to experimental data [Gopalaswamy et al., Nature 565(771), 581–586 (2019)]. In this paper, a physics-based statistical mapping approach is used to extract and quantify all the major sources of degradation of fusion yield for direct-drive implosions on the OMEGA laser. Here, the yield is found to be dependent on the age of the deuterium tritium fill, the ℓ = 1 asymmetry in the implosion core, the laser beam-to-target size ratio, and parameters related to the hydrodynamic stability. A controlled set of experiments were carried out where only the target fill age was varied while keeping all other parameters constant. The measurements were found to be in excellent agreement with the fill age dependency inferred using the mapping model. In addition, a new implosion design was created, guided by the statistical mapping model by optimizing the trade-offs between increased laser energy coupling at larger target size and the degradations caused by the laser beam-to-target size ratio and hydrodynamic instabilities. When experimentally performed, an increased fusion yield was demonstrated in targets with larger diameters.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

Landmark-embedded Gaussian process with applications for functional data modeling

In practice, we often need to infer the value of a target variable from functional observation data. A challenge in this task is that the relationship between the functional data and the target variable is very complex: the target variable not only influences the shape but also the location of the functional data. In addition, due to the uncertainties in the environment, the relationship is probabilistic, that is, for a given fixed target variable value, we still see variations in the shape and location of the functional data. To address this challenge, we present a landmark-embedded Gaussian process model that describes the relationship between the functional data and the target variable. A unique feature of the model is that landmark information is embedded in the Gaussian process model so that both the shape and location information of the functional data are considered simultaneously in a unified manner. Gibbs-Metropolis-Hasting algorithm is used for model parameters estimation and target variable inference. The performance of the proposed framework is evaluated by extensive numerical studies and a case study of nano-sensor calibration.

42 ENGINEERING↗