Search NASA⌕ Search

SEARCH · Search NASA

Results for “ADAM”

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 181 records · Page 10

Spatial modeling of daily concentrations of ground-level ozone in Montreal, Canada: A comparison of geostatistical approaches

Highlights: • The KED model with road and vegetation density as covariates showed good performance for all six synoptic classes. • The model using KED demonstrated the best results (R{sup 2} = 0.92; RMSE = 3.14 ppb). • The model using OK demonstrated a R{sup 2} = 0.85 and RMSE = 4 ppb. • Our results showed that errors appear to be substantially reduced with the KED model. • When measurements at numerous monitoring stations in a region are available, it may be better to use KED. • The BME may be more useful for prediction when data is not available. Ground-level ozone (O{sub 3}) is a powerful oxidizing agent and a harmful pollutant affecting human health, forests and crops. Estimating O{sub 3} exposure is a challenge because it exhibits complex spatiotemporal patterns. The aim in this study was to provide high-resolution maps (100 × 100 m) of O{sub 3} for the metropolitan area of Montreal, Canada. We assessed the kriging with external drift (KED) model to estimate O{sub 3} concentration by synoptic weather classes for 2010. We compared these results with ordinary kriging (OK), and a simple average of 12 monitoring stations. We also compared the estimates obtained for the 2010 summer with those from a Bayesian maximum entropy (BME) model reported in the literature (Adam-Poupart et al., 2014). The KED model with road and vegetation density as covariates showed good performance for all six synoptic classes (daily R{sup 2} estimates ranging from 0.77 to 0.92 and RMSE from 2.79 to 3.37 ppb). For the summer of 2010, the model using KED demonstrated the best results (R{sup 2} = 0.92; RMSE = 3.14 ppb), followed by the OK model (R{sup 2} = 0.85, RMSE = 4 ppb). Our results showed that errors appear to be substantially reduced with the KED model. This may increase our capacity of linking O{sub 3} levels to health problems by means of improved assessments of ambient exposures. However, future work integrating the temporal dependency in the data is needed to not overstate the performance of the KED model.

54 ENVIRONMENTAL SCIENCES↗

A second order radiative transfer equation and its solution by meshless method with application to strongly inhomogeneous media

A new second order form of radiative transfer equation (named MSORTE) is proposed, which overcomes the singularity problem of a previously proposed second order radiative transfer equation [J.E. Morel, B.T. Adams, T. Noh, J.M. McGhee, T.M. Evans, T.J. Urbatsch, Spatial discretizations for self-adjoint forms of the radiative transfer equations, J. Comput. Phys. 214 (1) (2006) 12-40 (where it was termed SAAI), J.M. Zhao, L.H. Liu, Second order radiative transfer equation and its properties of numerical solution using finite element method, Numer. Heat Transfer B 51 (2007) 391-409] in dealing with inhomogeneous media where some locations have very small/zero extinction coefficient. The MSORTE contains a naturally introduced diffusion (or second order) term which provides better numerical property than the classic first order radiative transfer equation (RTE). The stability and convergence characteristics of the MSORTE discretized by central difference scheme is analyzed theoretically, and the better numerical stability of the second order form radiative transfer equations than the RTE when discretized by the central difference type method is proved. A collocation meshless method is developed based on the MSORTE to solve radiative transfer in inhomogeneous media. Several critical test cases are taken to verify the performance of the presented method. The collocation meshless method based on the MSORTE is demonstrated to be capable of stably and accurately solve radiative transfer in strongly inhomogeneous media, media with void region and even with discontinuous extinction coefficient.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

An algebraic method for constructing stable and consistent autoregressive filters

In this paper, we introduce an algebraic method to construct stable and consistent univariate autoregressive (AR) models of low order for filtering and predicting nonlinear turbulent signals with memory depth. By stable, we refer to the classical stability condition for the AR model. By consistent, we refer to the classical consistency constraints of Adams–Bashforth methods of order-two. One attractive feature of this algebraic method is that the model parameters can be obtained without directly knowing any training data set as opposed to many standard, regression-based parameterization methods. It takes only long-time average statistics as inputs. The proposed method provides a discretization time step interval which guarantees the existence of stable and consistent AR model and simultaneously produces the parameters for the AR models. In our numerical examples with two chaotic time series with different characteristics of decaying time scales, we find that the proposed AR models produce significantly more accurate short-term predictive skill and comparable filtering skill relative to the linear regression-based AR models. These encouraging results are robust across wide ranges of discretization times, observation times, and observation noise variances. Finally, we also find that the proposed model produces an improved short-time prediction relative to the linear regression-based AR-models in forecasting a data set that characterizes the variability of the Madden–Julian Oscillation, a dominant tropical atmospheric wave pattern.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Physics-informed neural networks for solving forward and inverse flow problems via the Boltzmann-BGK formulation

We report the Boltzmann equation with the Bhatnagar-Gross-Krook collision model (Boltzmann-BGK equation) has been employed to describe multiscale flows, i.e., from the hydrodynamic limit to free molecular flow. In this study we employ physics-informed neural networks (PINNs) to solve forward and inverse problems via the Boltzmann-BGK formulation (PINN-BGK), enabling PINNs to model flows in both the continuum and rarefied regimes. In particular, the PINN-BGK is composed of three sub-networks, i.e., the first for approximating the equilibrium distribution function, the second for approximating the non-equilibrium distribution function, and the third one for encoding the Boltzmann-BGK equation as well as the corresponding boundary/initial conditions. By minimizing the residuals of the governing equations and the mismatch between the predicted and provided boundary/initial conditions, we can approximate the Boltzmann-BGK equation for both continuous and rarefied flows. For forward problems, the PINN-BGK is utilized to solve various benchmark flows given boundary/initial conditions, e.g. Kovasznay flow, Taylor-Green flow, cavity flow, and micro Couette flow for Knudsen number up to 5. For inverse problems, we focus on rarefied flows in which accurate boundary conditions are difficult to obtain. We employ the PINN-BGK to infer the flow field in the entire computational domain given a limited number of interior scattered measurements on the velocity without using the (unknown) boundary conditions. Results for the two-dimensional micro Couette and micro cavity flows with Knudsen numbers ranging from 0.1 to 10 indicate that the PINN-BGK can infer the velocity field in the entire domain with good accuracy. Finally, we also present some results on using transfer learning to accelerate the training process. Specifically, we can obtain a three-fold speedup comparted to the standard training process (e.g., Adam plus L-BFG-B) for the two-dimensional flow problems considered in our work.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

B-DeepONet: An enhanced Bayesian DeepONet for solving noisy parametric PDEs using accelerated replica exchange SGLD

Here, the Deep Operator Network (DeepONet) is a neural network architecture used to approximate operators, including the solution operator of parametric PDEs. DeepONets have shown remarkable approximation ability. However, the performance of DeepONets deteriorates when the training data is polluted with noise, a scenario that occurs in practice. To handle noisy data, we propose a Bayesian DeepONet based on replica exchange Langevin diffusion (reLD). Replica exchange uses two particles. The first particle trains a DeepONet to exploit the loss landscape and make predictions. The other particle trains a different DeepONet to explore the loss landscape and escape local minima via swapping. Compared to DeepONets trained with state-of-the-art gradient-based algorithms (e.g., Adam), the proposed Bayesian DeepONet greatly improves the training convergence for noisy scenarios and accurately estimates the uncertainty. To further reduce the high computational cost of the reLD training of DeepONets, we propose (1) an accelerated training framework that exploits the DeepONet's architecture to reduce its computational cost up to 25% without compromising performance and (2) a transfer learning strategy that accelerates training DeepONets for PDEs with different parameter values. Finally, we illustrate the effectiveness of the proposed Bayesian DeepONet using four parametric PDE problems.

97 MATHEMATICS AND COMPUTING↗

Arrhenian to non-Arrhenian crossover in glass melt viscosity

The activation energy for glass melt viscosity is a function of temperature, i.e., non-Arrhenian, within a middle range of viscosity values (typically, between 10 2 and 10 10 Pa s) and constant, i.e., Arrhenian, outside this range. At the low-viscosity end, the Arrhenian to non-Arrhenian transition (termed the crossover) falls within the glass processing range: glass melting and fining occur within the Arrhenian range while glass forming occurs within the non-Arrhenian range. By the Adam-Gibbs equation, the configuration entropy is nearly constant at high temperatures (low viscosities), where the glass melt turns into a simple liquid, and decreases with decreasing temperature (increasing viscosity) as the glass is increasingly polymerized. In the present study, the power-law function and hyperbolic tangent function are used to express the configuration entropy as a function of temperature. The viscosity crossover is defined as the point at which onset of structural grouping is causing the configuration entropy to deviate from its high-temperature value. In conclusion, it appears that the limited range of the configuration entropy of the fully depolymerized melt imposes a restriction on the crossover viscosity; for well-defined glass families, such as float glasses or nuclear waste glasses, the crossover point can be defined using a fixed, i.e., composition-independent, viscosity value.

36 MATERIALS SCIENCE↗

Fast neural network training on a cluster of GPUs for action recognition with high accuracy

In this work, we propose algorithms and techniques to accelerate training of deep neural networks for action recognition on a cluster of GPUs. The convergence analysis of our algorithm shows it is possible to reduce communication cost and at the same time minimize the number of iterations needed for convergence. We customize the Adam optimizer for our distributed algorithm to improve efficiency. In addition, we employ transfer-learning to further reduce training time while improving validation accuracy. For the UCF101 and HMDB51 datasets, the validation accuracies achieved are 93.1% and 67.9% respectively. With an additional end-to-end trained temporal stream, the validation accuracies achieved for UCF101 and HMDB51 are 93.47% and 81.24% respectively. As far as we know, these are the highest accuracies achieved with the two-stream approach using ResNet that does not involve computationally expensive 3D convolutions or pretraining on much larger datasets.

97 MATHEMATICS AND COMPUTING↗

A Comparison of Monte Carlo Particle Transport Algorithms for Binary Stochastic Mixtures

Two Monte Carlo algorithms originally proposed by Zimmerman and Zimmerman and Adams for particle transport through a binary stochastic mixture are numerically compared using a standard set of planar geometry benchmark problems. In addition to previously-published comparisons of the ensemble-averaged probabilities of reflection and transmission, we include comparisons of detailed ensemble-averaged total and material scalar flux distributions. Because not all benchmark scalar flux distribution data used to produce plots in previous publications remains available, we have independently regenerated the benchmark solutions including scalar flux distributions. Both Monte Carlo transport algorithms robustly produce physically-realistic scalar flux distributions for the transport problems examined. The first algorithm reproduces the standard Levermore-Pomraning model results for the probabilities of reflection and transmission. The second algorithm generally produces significantly more accurate probabilities of reflection and transmission and also significantly more accurate total and material scalar flux distributions.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Viscosity of transient glass-forming melt and its relation to foaming during batch-to-glass conversion

Primary foam, which affects the heat transfer into the glass batch and the final glass quality, occurs when a sufficient quantity of transient glass-forming melt evolves with viscosity low enough to close the open porosity of the reacting glass batch. To better understand how the fraction of transient melt and its viscosity affect the primary foam temperature range, we determined, with x-ray diffraction, the fraction and composition of the transient glass-forming melt in a heated waste glass melter feed as a function of temperature. Then we prepared a set of transient melts that occurred within the foaming temperature interval and measured their viscosities with spindle and falling sphere viscometers. Further, the results agree with the Adam-Gibbs and VFT viscosity-composition models, even for transient melt compositions outside of the models compositional validity range. As silica and other refractory particles dissolved in the predominantly borate transient melt while temperature increased, viscosity increased from the initial value of ~500 Pa s at the onset of foaming (~650°C) to a maximum of ~770 Pa s when silica dissolution was almost complete (~700°C). As temperature increased further, transient melt viscosity decreased to ~220 Pa s (~850°C) when the primary foam collapsed.

36 MATERIALS SCIENCE↗

Fast-RF-Shimming: Accelerate RF shimming in 7T MRI using deep learning

Ultrahigh field (UHF) Magnetic Resonance Imaging (MRI) offers an elevated signal-to-noise ratio (SNR), enabling exceptionally high spatial resolution that benefits both clinical diagnostics and advanced research. However, the jump to higher fields introduces complications, particularly transmit radiofrequency (RF) field ($B^{+}_{1}$) inhomogeneities, manifesting as uneven flip angles and image intensity irregularities. These artifacts can degrade image quality and impede broader clinical adoption. Traditional RF shimming methods, such as Magnitude Least Squares (MLS) optimization, effectively mitigate $B^{+}_{1}$ inhomogeneity, but remain time-consuming. Recent machine learning approaches, including RF Shim Prediction by Iteratively Projected Ridge Regression and other deep learning architectures, suggest alternative pathways. Although these approaches show promise, challenges such as extensive training periods, limited network complexity, and practical data requirements persist. In this paper, we introduce a holistic learning-based framework called Fast-RF-Shimming, which achieves a 5000 ​× ​speed-up compared to the traditional MLS method. In the initial phase, we employ random-initialized Adaptive Moment Estimation (Adam) to derive the desired reference shimming weights from multi-channel $B^{+}_{1}$ fields. Next, we train a Residual Network (ResNet) to map $B^{+}_{1}$ fields directly to the ultimate RF shimming outputs, incorporating the confidence parameter into its loss function. Finally, we design Non-uniformity Field Detector (NFD), an optional post-processing step, to ensure the extreme non-uniform outcomes are identified. Comparative evaluations with standard MLS optimization underscore notable gains in both processing speed and predictive accuracy, which indicates that our technique shows a promising solution for addressing persistent inhomogeneity challenges.

Deep learning↗

Calibration of parallel bond parameters in bonded particle models via physics-informed adaptive moment optimisation

This study proposes an automated calibration procedure for bond parameters in bonded discrete element modelling. By exploring the underlying physical correlations between microscopic parameters of bonds and macroscopic strength parameters of the continuum to be modelled, the microscopic shear strength and tensile strength are identified as independent variables for calibration purpose. Then a physics-informed iterative scheme is proposed to automatically approximate the bond parameters by viewing the micro-macro relation as an implicitly defined mathematical mapping function. As a result of highly non-convex features of this implicit mapping, the adaptive moment estimation (Adam), which is especially suitable for problems with noisy gradients, is adopted as the basic iterative scheme, in conjunction with other numerical techniques to approximately evaluate the partial derivatives involved. The whole procedure offers a simple and effective framework for bond parameter calibration. Finally, a numerical example of SiC ceramic is provided for validation. By compared with some existing calibration methods, the proposed method shows significant advantages in terms of calibration efficiency and accuracy.

42 ENGINEERING↗

An error-controlled adaptive time-stepping method for particle advancement in coupled CFD-DEM simulations

Coupled Computational-Fluid-Dynamics (CFD) and Discrete-Element-Method (DEM) models provide an accurate description of multiphase physical systems where a solid granular particle phase exists in an underlying gaseous continuous medium. The time integration of the granular phase in these simulations is typically handled using an explicit scheme with a constant time-step among all particles that is invariant in time to resolve inter-particle collisions. A locally third-order accurate adaptive time integration technique for particles that employs an embedded locally second-order scheme for error determination is presented in this work. The particle time-step size is dynamically adapted based on solution error, thus leading to significant savings in computational time. The efficacy of our scheme is quantified using four test cases of varying complexity (binary collision, homogeneous cooling system, fluidized bed and hopper discharge). The adaptive time-stepping method exhibits improved performance (~ 2–3 times in most of the cases studied) compared to three commonly used non-adaptive time-step methods (first-order Euler-explicit, second-order Adams-Bashforth and third-order Runge-Kutta schemes), while maintaining the same level of accuracy and parallel scalability.

42 ENGINEERING↗

Dose conversion factors for absorbed dose in a mobile phone to absorbed dose in critical organs in an anthropomorphic phantom for emergency dosimetry applications: OSL and TL experimental results, and Monte Carlo simulations

Here, an anthropomorphic phantom was equipped with three smartphones and irradiated in five separate exposure geometries in air using a 137 Cs source. The phantom was fitted internally with thermoluminescence dosimeters (TLD-700 chips) to measure the absorbed dose to the individual organs, while the absorbed doses in the phones were evaluated using optically stimulated luminescence (OSL) from the phones’ surface mount resistors (SMRs), along with reference dosimeters, namely Luxel OSL films and TLD-700 chips. The results indicate a strong dependence on the phone location with respect to the source due to shielding and scattering from the body. Dose conversion factors were calculated for several plausible exposure scenarios in order to correlate the absorbed doses measured in the phones to the average body dose received by the phantom. The measured conversion factors were compared with calculated factors using MCNP6.2 Monte Carlo simulations and an ADAM voxel phantom. Experiments were conducted to assess how the phone orientation and angle impacts the dose received by the phone components. The overall results indicate that the doses measured can vary significantly depending upon exposure geometry (including phone position and angle). However, in rotational irradiation symmetry, the phone doses lead to an over-estimate of the average whole-body dose by an average of ~10%, independent of the phone locations tested in this study. This result, if confirmed in future studies, may be acceptable for triage.

61 RADIATION PROTECTION AND DOSIMETRY↗

Preventing cleavage of Mer promotes efferocytosis and suppresses acute lung injury in bleomycin treated mice

Mer receptor tyrosine kinase (Mer) regulates macrophage activation and promotes apoptotic cell clearance. Mer activation is regulated through proteolytic cleavage of the extracellular domain. To determine if membrane-bound Mer is cleaved during bleomycin-induced lung injury, and, if so, how preventing the cleavage of Mer enhances apoptotic cell uptake and down-regulates pulmonary immune responses. During bleomycin-induced acute lung injury in mice, membrane-bound Mer expression decreased, but production of soluble Mer and activity as well as expression of disintegrin and metalloproteinase 17 (ADAM17) were enhanced . Treatment with the ADAM inhibitor TAPI-0 restored Mer expression and diminished soluble Mer production. Furthermore, TAPI-0 increased Mer activation in alveolar macrophages and lung tissue resulting in enhanced apoptotic cell clearance in vivo and ex vivo by alveolar macrophages. Suppression of bleomycin-induced pro-inflammatory mediators, but enhancement of hepatocyte growth factor induction were seen after TAPI-0 treatment. Additional bleomycin-induced inflammatory responses reduced by TAPI-0 treatment included inflammatory cell recruitment into the lungs, levels of total protein and lactate dehydrogenase activity in bronchoalveolar lavage fluid, as well as caspase-3 and caspase-9 activity and alveolar epithelial cell apoptosis in lung tissue. Importantly, the effects of TAPI-0 on bleomycin-induced inflammation and apoptosis were reversed by coadministration of specific Mer-neutralizing antibodies. These findings suggest that restored membrane-bound Mer expression by TAPI-0 treatment may help resolve lung inflammation and apoptosis after bleomycin treatment. -- Highlights: ►Mer expression is restored by TAPI-0 treatment in bleomycin-stimulated lung. ►Mer signaling is enhanced by TAPI-0 treatment in bleomycin-stimulated lung. ►TAPI-0 enhances efferocytosis and promotes resolution of lung injury.

60 APPLIED LIFE SCIENCES↗

Global stochastic optimization of stellarator coil configurations

In the construction of a stellarator, the manufacturing and assembling of the coil system is a dominant cost. These coils need to satisfy strict engineering tolerances, and if those are not met the project could be cancelled as in the case of the National Compact Stellarator Experiment (NCSX) project. Therefore, our goal is to find coil configurations that increase construction tolerances without compromising the performance of the magnetic field. In this paper, we develop a gradient-based stochastic optimization model which seeks robust stellarator coil configurations in high dimensions. In particular, we design a two-step method: first, we perform an approximate global search by a sample efficient trust-region Bayesian optimization; second, we refine the minima found in step one with a stochastic local optimizer. To this end, we introduce two stochastic local optimizers: BFGS applied to the sample average approximation; and Adam, equipped with a control variate for variance reduction. Numerical simulations performed on a W7-X-like coil configuration demonstrate that our global optimization approach finds a variety of promising local solutions at less than 0.1% of the cost of previous work, which considered solely local stochastic optimization.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Spatially heterogeneous dynamics in a metallic glass forming liquid imaged by electron correlation microscopy

Supercooled liquids exhibit spatial heterogeneity in the dynamics of their fluctuating atomic arrangements. The length and time scales of the heterogeneous dynamics are central to the glass transition and influence nucleation and growth of crystals from the liquid. In this work, we report direct experimental visualization of the spatially heterogeneous dynamics as a function of temperature in the supercooled liquid state of a Pt-based metallic glass, using electron correlation microscopy with sub-nanometer resolution. An experimental four-point space-time correlation function demonstrates a growing dynamic correlation length, ξ, upon cooling of the liquid toward the glass transition temperature. ξ as a function of the relaxation time τ are in good agreement with Adam-Gibbs theory, inhomogeneous mode-coupling theory and random first-order transition theory of the glass transition. The same experiments demonstrate the existence of a nanometer thickness near-surface layer with order of magnitude shorter relaxation time than inside the bulk.

42 ENGINEERING↗

Discontinuous Galerkin Methods and Local Time Stepping for Wave Propagation

Locally refined meshes impose severe stability constraints on explicit time-stepping methods for the numerical simulation of time dependent wave phenomena. To overcome that stability restriction, local time-stepping methods are developed, which allow arbitrarily small time steps precisely where small elements in the mesh are located. When combined with a discontinuous Galerkin finite element discretization in space, which inherently leads to a diagonal mass matrix, the resulting numerical schemes are fully explicit. Starting from the classical Adams-Bashforth multi-step methods, local time stepping schemes of arbitrarily high accuracy are derived. Numerical experiments validate the theory and illustrate the usefulness of the proposed time integration schemes.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

SuperADAM: Upgraded polarized neutron reflectometer at the Institut Laue-Langevin

A new neutron reflectometer SuperADAM has recently been built and commissioned at the Institut Laue-Langevin, Grenoble, France. It replaces the previous neutron reflectometer ADAM. The new instrument uses a solid state polarizer/wavelength filter providing a highly polarized (up to 98.6%) monochromatic neutron flux of 8 Multiplication-Sign 10{sup 4} n cm{sup -2} s{sup -1} with monochromatization {Delta}{lambda}/{lambda}= 0.7% and angular divergence {Delta}{alpha}= 0.2 mrad. The instrument includes both single and position sensitive detectors. The position sensitive detector allows simultaneous measurement of specular reflection and off-specular scattering. Polarization analysis for both specular reflection and off-specular scattering is achieved using either mirror analyzers or a {sup 3}He spin filter cell. High efficiency detectors, low background, and high flux provides a dynamic range of up to seven decades in reflectivity. Detailed specifications and the instrument capabilities are illustrated with examples of recently collected data in the fields of thin film magnetism and thin polymer films.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗