Search NASA⌕ Search

SEARCH · Search NASA

Results for “generalized linear 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 1,711 records · Page 95

A study of solute transport of radiolysis products in crud and its effects on crud growth on PWR fuel pin

Our research examines the concentration of radiolysis species (H 2 , O 2 , and H 2 O 2 ) over the porous crud layer using a three dimensional time dependent solute transport model. A Monte Carlo random walk technique is adopted to simulate the transport behavior of the different species with various parametric studies of source term, crud thickness, and crud porosity. Particularly, this model employs a system of coupled mass transport and chemical interactions as the source term, which makes the problem non-linear. We also demonstrate that a negligible effect on radiolysis species concentrations change due to the consideration of source term. The crud thickness and porosity effect on the concentration distributions are notably observed. In general, higher concentration starts from the intersection of the heating surface with the chimney wall from the beginning and it reaches the equilibrium state within tens of seconds. The concentration profiles of the radiolysis species H 2 , O 2 , and H 2 O 2 can be directly related to corrosion issues. Furthermore, the direct application of this study to nuclear engineering research is to aid in the design of reactors with higher performance without experiencing an Axial Offset Anomaly (AOA), an unexpected measured shift in axial power distribution from predicted values.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Binary spinning black hole Hamiltonian in canonical center-of-mass and rest-frame coordinates through higher post-Newtonian order

The recently constructed Hamiltonians for spinless binary black holes through third post-Newtonian order and for spinning ones through formal second post-Newtonian order, where the spins are counted of zero post-Newtonian order, are transformed into fully canonical center-of-mass and rest-frame variables. The mixture terms in the Hamiltonians between center-of-mass and rest-frame variables are in accordance with the relation between the total linear momentum and the center-of-mass velocity as demanded by global Lorentz invariance. The various generating functions for the center-of-mass and rest-frame canonical variables are explicitly given in terms of the single-particle canonical variables. The no-interaction theorem does not apply because the world-line condition of Lorentz covariant position variables is not imposed.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Viscous instability triggered by layered accretion in protoplanetary disks

Layered accretion is one of the inevitable ingredients in protoplanetary disks when disk turbulence is excited by magnetorotational instabilities (MRIs). In the accretion, disk surfaces where MRIs fully operate have a high value of disk accretion rate ( M-dot ), while the disk midplane where MRIs are generally quenched ends up with a low value of M-dot . Significant progress on understanding MRIs has recently been made by a number of dedicated MHD simulations, which requires improvement of the classical treatment of α in 1D disk models. To this end, we obtain a new expression of α by utilizing an empirical formula that is derived from recent MHD simulations of stratified disks with ohmic diffusion. It is interesting that this new formulation can be regarded as a general extension of the classical α. Armed with the new α, we perform a linear stability analysis of protoplanetary disks that undergo layered accretion, and we find that a viscous instability can occur around the outer edge of dead zones. Disks become stable in using the classical α. We identify that the difference arises from Σ-dependence of M-dot ; whereas Σ is uniquely determined for a given value of M-dot in the classical approach, the new approach leads to M-dot that is a multivalued function of Σ. We confirm our finding both by exploring a parameter space and by performing the 1D, viscous evolution of disks. We finally discuss other nonideal MHD effects that are not included in our analysis but may affect our results.

79 ASTRONOMY AND ASTROPHYSICS↗

High-precision spectra for dynamical Dark Energy cosmologies from constant-w models

Spanning the whole functional space of cosmologies with any admissible DE state equations w(a) seems a need, in view of forthcoming observations, namely those aiming to provide a tomography of cosmic shear. In this paper I show that this duty can be eased and that a suitable use of results for constant-w cosmologies can be sufficient. More in detail, I ''assign'' here six cosmologies, aiming to span the space of state equations w(a) = w{sub o}+w{sub a}(1−a), for w{sub o} and w{sub a} values consistent with WMAP5 and WMAP7 releases and run N-body simulations to work out their non-linear fluctuation spectra at various redshifts z. Such spectra are then compared with those of suitable auxiliary models, characterized by constant w. For each z a different auxiliary model is needed. Spectral discrepancies between the assigned and the auxiliary models, up to k ≅ 2–3 h Mpc{sup −1}, are shown to keep within 1 %. Quite in general, discrepancies are smaller at greater z and exhibit a specific trend across the w{sub o} and w{sub a} plane. Besides of aiming at simplifying the evaluation of spectra for a wide range of models, this paper also outlines a specific danger for future studies of the DE state equation, as models fairly distant on the w{sub 0}-w{sub a} plane can be easily confused.

79 ASTRONOMY AND ASTROPHYSICS↗

Galileons on cosmological backgrounds

We construct four-dimensional effective field theories of a generalized DBI galileon field, the dynamics of which naturally take place on a Friedmann-Robertson-Walker spacetime. The theories are invariant under non-linear symmetry transformations, which can be thought of as being inherited from five-dimensional bulk Killing symmetries via the probe brane technique through which they are constructed. The resulting model provides a framework in which to explore the cosmological role that galileons may play as the universe evolves.

79 ASTRONOMY AND ASTROPHYSICS↗

On the Effect of Small-Angle Scattering by Density Fluctuations on the Efficiency of Linear Transformation of Ordinary and Extraordinary Waves in a Toroidally Inhomogeneous Plasma

We have analyzed the efficiency of tunneling of quasi-optical wave beams through the evanescent region in the vicinity of plasma cutoff in a randomly inhomogeneous magnetoactive plasma. A new theoretical model proposed here makes it possible to study the effect of a random phase modulation induced by density fluctuations in the wave beam path on the efficiency of the linear transformation of waves in the 2D geometry corresponding to the experimental conditions for heating the overdense plasma in toroidal magnetic systems. We have derived a general analytic expression connecting the wave beam coefficient of transformation averaged over the ensemble of random wave field realizations with its phase correlation function. We have analyzed the dependence of the coefficient of transformation on the correlation length for the random phase distribution in the beam and on the traversed path length to the interaction region. The threshold value of the path length above which the fluctuations produce a dominating effect has been determined. The importance of taking into account the 2D effects has been demonstrated.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Performance Comparison of Machine Learning Models for Ultrasonic Nondestructive Evaluation of Alkali-Silica Reaction in Concrete

Alkali-silica reaction (ASR) causes concrete degradation, leading to cracking, rebar corrosion, and reduced structural integrity, which raises safety concerns. Ultrasonic nondestructive evaluation (NDE) effectively assesses concrete properties and monitors ASR progression. However, its deployment and analysis require specialized expertise and subjective interpretation. As computational power increases, artificial intelligence (AI) and machine learning (ML) algorithms are increasingly being used to automate NDE data analysis across various industries for AI-assisted automation. Regulatory agencies are adapting to this technological shift, prompting a need to evaluate current ML technologies’ capabilities and limitations in assessing concrete material properties and damage. This report presents a comparative analysis of four ML regression models for predicting concrete material damage induced by ASR expansion using long-term ultrasonic data monitoring. The models investigated include linear regression (LR), support vector regression (SVR), shallow neural networks (NN), and deep neural networks (DNN). LR, SVR, and shallow NN models use features extracted from ultrasonic signals, whereas the DNN model processes time-domain ultrasonic signals and frequency spectra directly. The study systematically compared the models’ performance from various perspectives, including model input, prediction performance, and generalization ability. The findings indicate significant variability in model performance, with some ML algorithms achieving very high or very low prediction accuracy depending on the preprocessing and feature engineering (extraction and selection) applied. Key insights include the observation that shallow ML models (LR, SVR, and shallow NNs) require meticulous preprocessing and feature extraction to achieve high accuracy. In contrast, the DNN model, although it bypasses the need for feature engineering, necessitates extensive preprocessing to mitigate noise and computational demands. The SVR model emerged as the top performer among the shallow models, and the DNN model exhibited superior performance on specific datasets but struggled with generalization across specimens from different batches. Additionally, the SVR model is sensitive to temperature variations, whereas the DNN model is robust in this regard. Using recurrent neural networks is recommended for future ASR expansion prediction studies. Recurrent neural networks’ inherent ability to capture temporal dependencies and long-term patterns makes them well suited for analyzing sequential ultrasonic monitoring data. Overall, the results and conclusions of this study could provide insights into the capabilities and effectiveness of ML when applied to ultrasonic NDE data and help identify best practices for using ML for ultrasonic NDE of concrete material properties.

36 MATERIALS SCIENCE↗

Linear baroclinic instability in the Martian atmosphere

A spherical quasi-geostrophic model is used in an examination of linear baroclinic instability in such zonal-mean flows as those of the Martian atmosphere in winter, under both relatively nondusty and very dusty conditions. These zonal flows, which possess both vertical and meridional shear, are characterized by baroclinically unstable modes whose growth rates and phase speeds are generally consistent with available observations. The structures of the spherical modes are similar to those obtained for terrestrial zonal flows, if similar zonal wavelengths are compared. Zonally symmetric topography, like that of Mars' northern hemisphere, reduces linear growth rates without changing the most unstable scale. It also increases phase speeds.

Barnes, J. R.↗

A theoretical approach for analyzing the restabilization of wakes

Recently reported experimental results demonstrate that restabilization of the low-Reynolds-number flow past a circular cylinder can be achieved by the placement of a smaller cylinder in the wake of the first at particular locations. Traditional numerical procedures for modeling such phenomena are computationally expensive. An approach is presented here in which the properties of the adjoint solutions to the linearized equations of motion are exploited to map quickly the best positions for the small cylinder's placement. Comparisons with experiment and previous computations are favorable. The approach is shown to be applicable to general flows, illustrating how strongly control mechanisms that involve sources of momentum couple to unstable (or stable) modes of the system.

Hill, D. C.↗

Short-term load forecasting using generalized regression and probabilistic neural networks in the electricity market

For the economic and secure operation of power systems, a precise short-term load forecasting technique is essential. Modern load forecasting techniques - especially artificial neural network methods - are particularly attractive, as they have the ability to handle the non-linear relationships between load, weather temperature, and the factors affecting them directly. A test of two different ANN models on data from Australia's Victoria market is promising. (author)

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Critical look at cosmological perturbation theory techniques

Recently, a number of analytic prescriptions for computing the nonlinear matter power spectrum have appeared in the literature. These typically involve resummation or closure prescriptions which do not have a rigorous error control, thus they must be compared with numerical simulations to assess their range of validity. We present a direct side-by-side comparison of several of these analytic approaches, using a suite of high-resolution N-body simulations as a reference, and discuss some general trends. All of the analytic results correctly predict the behavior of the power spectrum at the onset of nonlinearity, and improve upon a pure linear theory description at very large scales. All of these theories fail at sufficiently small scales. At low redshift the dynamic range in scale where perturbation theory is both relevant and reliable can be quite small. We also compute for the first time the two-loop contribution to standard perturbation theory for cold dark matter models, finding improved agreement with simulations at large redshift. At low redshifts however the two-loop term is larger than the one-loop term on quasilinear scales, indicating a breakdown of the perturbation expansion. Finally, we comment on possible implications of our results for future studies. A software package implementing the methods presented here is available at http://mwhite.berkeley.edu/Copter.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

HYPERELASTIC MODELS FOR GRANULAR MATERIALS

A continuum framework for modeling of dust mobilization and transport, and the behavior of granular systems in general, has been reviewed, developed and evaluated for reactor design applications. The large quantities of micron-sized particles expected in the international fusion reactor design, ITER, will accumulate into piles and layers on surfaces, which are large relative to the individual particle size; thus, particle-particle, rather than particle-surface, interactions will determine the behavior of the material in bulk, and a continuum approach is necessary and justified in treating the phenomena of interest; e.g., particle resuspension and transport. The various constitutive relations that characterize these solid particle interactions in dense granular flows have been discussed previously, but prior to mobilization their behavior is not even fluid. Even in the absence of adhesive forces between particles, dust or sand piles can exist in static equilibrium under gravity and other forces, e.g., fluid shear. Their behavior is understood to be elastic, though not linear. The recent “granular elasticity” theory proposes a non-linear elastic model based on “Hertz contacts” between particles; the theory identifies the Coulomb yield condition as a requirement for thermodynamic stability, and has successfully reproduced experimental results for stress distributions in sand piles. The granular elasticity theory is developed and implemented in a stand- alone model and then implemented as part of a finite element model, ABAQUS, to determine the stress distributions in dust piles subjected to shear by a fluid flow. We identify yield with the onset of mobilization, and establish, for a given dust pile and flow geometry, the threshold pressure (force) conditions on the surface due to flow required to initiate it. While the granular elasticity theory applies strictly to cohesionless granular materials, attractive forces are clearly important in the interaction of micron-sized particles; extension of the theory to account for these effects is also considered. A set of continuum models are proposed for use in the future dust transport modeling.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

The f ( R ) halo mass function in the cosmic web

An important indicator of modified gravity is the effect of the local environment on halo properties. This paper examines the influence of the local tidal structure on the halo mass function, the halo orientation, spin and the concentration-mass relation. We use the excursion set formalism to produce a halo mass function conditional on large-scale structure. Our simple model agrees well with simulations on large scales at which the density field is linear or weakly non-linear. Beyond this, our principal result is that f ( R ) does affect halo abundances, the halo spin parameter and the concentration-mass relationship in an environment-independent way, whereas we find no appreciable deviation from \text(ΛCDM) for the mass function with fixed environment density, nor the alignment of the orientation and spin vectors of the halo to the eigenvectors of the local cosmic web. There is a general trend for greater deviation from \text(ΛCDM) in underdense environments and for high-mass haloes, as expected from chameleon screening.

79 ASTRONOMY AND ASTROPHYSICS↗

Propagation of excitation in long 1D chains: Transition from regular quantum dynamics to stochastic dynamics

The quantum dynamics problem for a 1D chain consisting of 2N + 1 sites (N Much-Greater-Than 1) with the interaction of nearest neighbors and an impurity site at the middle differing in energy and in coupling constant from the sites of the remaining chain is solved analytically. The initial excitation of the impurity is accompanied by the propagation of excitation over the chain sites and with the emergence of Loschmidt echo (partial restoration of the impurity site population) in the recurrence cycles with a period proportional to N. The echo consists of the main (most intense) component modulated by damped oscillations. The intensity of oscillations increases with increasing cycle number and matrix element C of the interaction of the impurity site n = 0 with sites n = {+-}1 (0 < C {<=} 1; for the remaining neighboring sites, the matrix element is equal to unity). Mixing of the components of echo from neighboring cycles induces a transition from the regular to stochastic evolution. In the regular evolution region, the wave packet propagates over the chain at a nearly constant group velocity, embracing a number of sites varying periodically with time. In the stochastic regime, the excitation is distributed over a number of sites close to 2N, with the populations varying irregularly with time. The model explains qualitatively the experimental data on ballistic propagation of the vibrational energy in linear chains of CH{sub 2} fragments and predicts the possibility of a nondissipative energy transfer between reaction centers associated with such chains.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Woven ceramic matrix composite surrogate model based on physics-informed recurrent neural network

A recurrent neural network (RNN) based surrogate model is developed to emulate the nonlinear constitutive behavior of woven ceramic matrix composites (CMCs) driven by matrix damage at multiple length scales. Physics-informed constraints are introduced into the surrogate model through regularization to ground the prediction in physics and improve its predictive capabilities. Training data is generated using the multiscale generalized method of cells (MSGMC) approach coupled with a matrix damage model. This coupling permits simulating the nonlinear behavior of woven CMCs based on constituent response at the micro-, meso-, and macroscales. The multiscale repeating unit cell is loaded under non-monotonic conditions including multiple load / unload cycles and tension / compression. The fiber volume fraction as well as the intra- and intertow void volume fractions are also varied in the generation of training data. Therefore, the RNN-based surrogate model is tasked with predicting, as a function of variable input strain sequence and fiber and void volume fractions, the resulting stress versus strain response while satisfying physical constraints such as positive semi-definiteness of the tangent stiffness matrix and linear elastic unloading. Further, the trained surrogate model effectively matches the stress versus strain response and successfully predicts the tangent modulus throughout the loading regime. Neural network based surrogate models can offer efficient alternatives to running computationally intensive multiscale material models to simulate the nonlinear response of large structural models. Therefore the presented work provides evidence towards the feasibility of developing, training, and running such models for CMCs with complex architectures, nonlinear multiaxial material response, and under non-monotonic loading conditions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Theory of Current Transients in Planar Semiconductor Devices: Insights and Applications to Organic Solar Cells

Time-domain current measurements are widely used to characterize semiconductor material properties, such as carrier mobility, doping concentration, carrier lifetime, and the static dielectric constant. It is therefore critical that these measurements be theoretically understood if they are to be successfully applied to assess the properties of materials and devices. In this paper, we derive generalized relations for describing current-density transients in planar semiconductor devices at uniform temperature. By spatially averaging the charge densities inside the semiconductor, we are able to provide a rigorous, straightforward, and experimentally relevant way to interpret these measurements. The formalism details several subtle aspects of current transients, including how the electrode charge relates to applied bias and internal space charge, how the displacement current can alter the apparent free-carrier current, and how to understand the integral of a charge-extraction transient. We also demonstrate how the formalism can be employed to derive the current transients arising from simple physical models, like those used to describe charge extraction by linearly increasing voltage (CELIV) and time-of-flight experiments. In doing so, we find that there is a nonintuitive factor-of-2 reduction in the apparent free-carrier concentration that can be easily missed, for example, in the application of charge-extraction models. Finally, to validate our theory and better understand the different current contributions, we perform a full time-domain drift-diffusion simulation of a CELIV trace and compare the results to our formalism. As expected, our analytic equations match precisely with the numerical solutions to the drift-diffusion, Poisson, and continuity equations. Furthermore, overall, our formalism provides a straightforward and general way to think about how the internal space-charge distribution, the electrode charge, and the externally applied bias translate into a measured current transient in a planar semiconductor device.

36 MATERIALS SCIENCE↗

U-net architected deep material network training with microstructure local field information

The Deep Material Network (DMN) has recently emerged as a powerful reduced-order modeling framework for simulating the mechanical response of heterogeneous materials such as composites. Unlike most data-driven approaches that directly learn a material’s response under prescribed loading, the DMN acts as a homogenization operator, learning the kinematic constraints and mechanical interactions of the underlying microstructure. However, traditional DMN training relies exclusively on homogenized effective properties derived from Direct Numerical Simulations (DNS), discarding the rich local field data that govern microstructural interactions. In this work, we extend the DMN framework to incorporate such local field information into the offline training process. Utilizing a U-Net architecture, we augment the DMN training objective to include the first and second statistical moments of the local stress fields obtained from linear DNS. This ensures that the learned network topology not only fits the effective stiffness but also accurately reflects the internal local stress and strain partitioning of the microstructure. The results confirm that supervising the localization process during training yields a superior surrogate model, reducing local prediction errors by an order of magnitude and significantly improving generalization to unseen nonlinear constitutive behaviors compared to traditional DMNs.

36 MATERIALS SCIENCE↗

Overall chilled water system energy consumption modeling and optimization

The emergence of increasingly affordable variable-speed drive technology has changed the approach used to control chilled water systems equipped with these drives. The purpose of this research was to develop an integrated chilled water modeling technique that can determine the optimal system setpoints and estimate the energy saving potential of chiller system. The chiller system equipped with Variable Frequency Drives (VFDs) on cooling tower fans and condenser water pumps. To accomplish the objective, physical component models of the centrifugal chiller, cooling tower and condenser water pump were established with the goal of incorporating the system’s condenser water flow rate and cooling tower fan speeds as optimization variables. Furthermore, a cooling load prediction algorithm was developed using a multiple non-linear regression model to approximate the building’s cooling load subject to a range of environmental conditions. The inputs and outputs of the individual component models were linked to estimate how adjusting the cooling tower fan and condenser water pump speed would influence the system’s comprehensive performance. Here, the overall system model was then optimized using a generalized reduced gradient optimization algorithm to determine the potential energy savings through speed control with VFDs and to ascertain a control logic strategy for the building automation system to operate the heating and cooling system. A case-study was performed on a single chiller system at a museum and the model was calibrated according to logged data collected over four months. Results showed that for the system analyzed, the energy saving of optimizing the cooling tower fan system was found to be 12–15%, while the energy saving potential of optimizing the condenser water pump with the cooling tower fan was negligible. Additionally, comparing different cooling tower fan control strategies showed that a wet-bulb approach-based cooling tower control strategy was shown to have the highest correlation to the optimized fan speed with an R 2 of 0.924.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗