Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Stochastic modelling”

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 1,657 records · Page 92
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

MLMOD: Machine Learning Methods for Data-Driven Modeling in LAMMPS

MLMOD is a software package for incorporating machine learning approaches and models into simulations of microscale mechanics and molecular dynamics in LAMMPS. Recent machine learning approaches provide promising data-driven approaches for learning representations for system behaviors from experimental data and high fidelity simulations. The package facilitates learning and using data-driven models for (i) dynamics of the system at larger spatial-temporal scales (ii) interactions between system components, (iii) features yielding coarser degrees of freedom, and (iv) features for new quantities of interest characterizing system behaviors. MLMOD provides hooks in LAMMPS for (i) modeling dynamics and time-step integration, (ii) modeling interactions, and (iii) computing quantities of interest characterizing system states. The package allows for use of machine learning methods with general model classes including Neural Networks, Gaussian Process Regression, Kernel Models, and other approaches. Here we discuss our prototype C++/Python package, aims, and example usage. For related papers, examples, updates, and additional information see https://github.com/atzberg/mlmod and http://atzberger.org/.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

A Vibrational Energy Harvesting Sensor Based on Linear and Rotational Electromechanical Effects

In this investigation, a magnetically coupled double-spring design is presented for harvesting low-level non-stationary random vibrational energy. The sensor relies on multimodal coupling between the translation and rotation of a two-spring magnet and coil system to widen the harvesting bandwidth. Energy methods are used to develop a model to characterize the electromechanical response of the system, the solution of which is obtained using stochastic techniques based on a particle swarm algorithm. This approach provides an efficient method to estimate system parameters that otherwise are difficult or impossible to determine with independent measurements. The experimental results demonstrate agreement with the theoretical predictions over a limited bandwidth. The sensor can effectively harvest non-stationary vibration energy down to 10 -4 g within a limited bandwidth of 130–150 Hz. The sensor prototype has an operational volume of 2.6 cm 3 with a calculated power density of 0.2 W/cm 3 . The sensor’s small size results in a coupling efficiency of approximately 6% across the tested bandwidth.

42 ENGINEERING↗
NASA NTRS1984

Solar-flare neutrons and gamma-rays

The numbers and spectra of the accelerated protons and nuclei that produce the neutrons and gamma-rays observed in solar flares are derived, and the results are compared with interplanetary observations of flare protons. The two most widely studied flare acceleration mechanisms, stochastic and diffusive shock acceleration, are discussed, and the arguments favoring the thick-target interaction model for neutron and gamma-ray production at the sun are briefly reviewed. The pertinent results of the theory of neutron and gamma-ray production are presented. The number and spectrum of the accelerated particles are derived from observations of nuclear deexcitation lines and the 2.223 MeV line from several flares. The June 21, 1980 and June 3, 1982 flares, from which a wealth of neutron, gamma-ray and energetic-particle data has recently become available, are discussed.

Murphy, R. J.↗
NASA NTRS1987

Implementation of variable time step stochastic dynamics for electronically inelastic gas-surface collisions

A variable time step algorithm has been implemented for solving the stochastic equations of motion for gas-surface collisions. It has been tested for a simple model of electronically inelastic collisions with an insulator surface in which the phonon manifold acts as a heat bath and electronic states are localized. In addition to reproducing the accurate nuclear dynamics of the surface atoms, numerical calculations have shown the algorithm to yield accurate ensemble averages of physical observables such as electronic transition probabilities and total energy loss of the gas atom to the surface. This new algorithm offers a gain in efficieny of up to an order of magnitude compared to fixed time step integration.

Garrett, Bruce C.↗
NASA NTRS1988

Particle acceleration in impulsive solar flares

A model for second-step electron acceleration in impulsive solar flares is presented. The theory of stochastic particle acceleration is extended to include Coulomb energy losses which become important at low coronal heights. This inclusion successfully explains the observed steepening of interplanetary electron spectra below about 3 MeV following impulsive solar flares taking place at low coronal heights. It also explains the observed spectral differences of relativistic electrons in long-duration and impulsive flares.

Steinacker, J.↗
NASA NTRS1997

Particle Acceleration in Active Galactic Nuclei

The high efficiency of energy generation inferred from radio observations of quasars and X-ray observations of Seyfert active galactic nuclei (AGNs) is apparently achieved only by the gravitational conversion of the rest mass energy of accreting matter onto supermassive black holes. Evidence for the acceleration of particles to high energies by a central engine is also inferred from observations of apparent superluminal motion in flat spectrum, core-dominated radio sources. This phenomenon is widely attributed to the ejection of relativistic bulk plasma from the nuclei of active galaxies, and accounts for the existence of large scale radio jets and lobes at large distances from the central regions of radio galaxies. Reports of radio jets and superluminal motion from galactic black hole candidate X-ray sources indicate that similar processes are operating in these sources. Observations of luminous, rapidly variable high-energy radiation from active galactic nuclei (AGNs) with the Compton Gamma Ray Observatory show directly that particles are accelerated to high energies in a compact environment. The mechanisms which transform the gravitational potential energy of the infalling matter into nonthermal particle energy in galactic black hole candidates and AGNs are not conclusively identified, although several have been proposed. These include direct acceleration by static electric fields (resulting from, for example, magnetic reconnection), shock acceleration, and energy extraction from the rotational energy of Kerr black holes. The dominant acceleration mechanism(s) operating in the black hole environment can only be determined, of course, by a comparison of model predictions with observations. The purpose of the work proposed for this grant was to investigate stochastic particle acceleration through resonant interactions with plasma waves that populate the magnetosphere surrounding an accreting black hole. Stochastic acceleration has been successfully applied to the problem of ion and electron energization in solar flares, and is capable of accounting for a wide range of both neutral and charged particle emissions. It is also a component in diffusive shock acceleration, since pitch-angle scattering (which is necessary for multiple shock crossings) is accompanied by diffusion in momentum space, which in turn yields a net systematic energy gain; however, stochastic energization will dominate the first-order shock process only in certain parameter regimes. Although stochastic acceleration has been applied to particle energization in the lobes of radio galaxies, its application to the central regions of AGNs has only recently been considered, but not in detail. We proposed to systematically investigate the plasma processes responsible for stochastic particle acceleration in black hole magnetospheres along with the energy-loss processes which impede particle energization. To this end we calculated acceleration rates and escape time scales for protons and electrons resonating with Alfven waves, and for electrons resonating with whistlers. Assuming either a Kolmogorov or Kraichnan wave spectrum, accretion at the Eddington limit, magnetic field strengths near equipartition, and turbulence energy densities approx. 10% of the total magnetic field energy density, we find that Alfven waves accelerate protons to Lorentz factors approx, equals 10(exp 4) - 10(exp 6) before they escape from the system. Acceleration of electrons by fast mode and whistler waves can produce a nonthermal population of relativistic electrons whose maximum energy is determined by a competition with radiation losses.

Miller, James A.↗
NASA NTRS2004

Sensitivity of Simulated Warm Rain Formation to Collision and Coalescence Efficiencies, Breakup, and Turbulence: Comparison of Two Bin-Resolved Numerical Models

Numerical models that resolve cloud particles into discrete mass size distributions on an Eulerian grid provide a uniquely powerful means of studying the closely coupled interaction of aerosols, cloud microphysics, and transport that determine cloud properties and evolution. However, such models require many experimentally derived paramaterizations in order to properly represent the complex interactions of droplets within turbulent flow. Many of these parameterizations remain poorly quantified, and the numerical methods of solving the equations for temporal evolution of the mass size distribution can also vary considerably in terms of efficiency and accuracy. In this work, we compare results from two size-resolved microphysics models that employ various widely-used parameterizations and numerical solution methods for several aspects of stochastic collection.

Fridlind, Ann↗
DOE OSTI2025

Low‐dimensional manifold learning for uncertainty quantification in complex multi‐scale stochastic systems

Broadly speaking, the goals of the project are to develop techniques to use manifold learning to develop reduced‐order and surrogate models for "hyper‐reduction" of very high‐dimensional complex multi‐scale systems. This is being achieved by employing a newly proposed form of manifold projection and learning that leverages recent advancements in computational geometry and data‐driven modeling. In particular, we are applying a manifold projection technique to project the solutions of very high‐dimensional systems onto the so‐called Grassmannmanifold, a Reimannian manifold comprised of orthonormal matrices. We then apply data‐driven machine learning techniques to classify the solutions on the manifold (e.g. clustering techniques) according to their proximity on the manifold and leverage a further nonlinear dimension reduction to organize the structured data on the manifold. Finally, we are developing novel techniques that enable us to directly interpolate the hyper‐reduced data such that we can predict the solution of the complex, high‐ dimensional system without need to call the full expensive computational model. Given their adherence to the underlying structure of the solution of the physical system, it is expected that these approximate solutions will be sufficiently constrained so as to (approximately) adhere to physical principles.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1982

Some critical questions about deterministic and stochastic adaptive control algorithms

The purpose of this informal paper is to discuss certain robustness issues associated with existing adaptive control algorithms. A modeling framework for incorporating high-frequency unknown dynamics in the adaptive control framework is suggested. Possible fundamental limitations of existing adaptive control framework is suggested. Possible fundamental limitations of existing adaptive control algorithms are also discussed, with emphasis upon their closed-loop stability properties in the presence of unmodeled high-frequency dynamics.

Athans, M.↗
DOE OSTI2020

Linking classical PRA models to a dynamic PRA

Here, this paper presents a series of methods designed to incorporate classical Probabilistic Risk Assessment (PRA) models such as Event Trees (ETs) and Fault Trees (FTs) into dynamic PRA. In contrast to classical PRA, dynamic PRA couples stochastic methods with system simulators to determine the risks associated with complex systems such as nuclear power plants. Compared with classical PRA methods, they can evaluate with higher resolution the safety impact of timing and sequencing of events on the progression of the accident. As part of a dynamic PRA analysis, it is not uncommon that parts of the system to be analyzed might not require a computationally expensive simulation model. These parts could be in fact modeled by employing classical PRA models (e.g., a FT). Here, we present a set of methods and tools that can be used to link the most common classical PRA models (ETs, FTs, reliability block diagrams and Markov models) to simulation codes such as RELAP5-3D: creating a “hybrid PRA.” In order to show the potential of such an hybrid PRA we employ this method to verify ET modeling assumptions (e.g., success criteria) using a large break loss of coolant accident initiating event as a test case. In this respect, we link a set of FTs from the original PRA to the RELAP5-3D code and perform a hybrid PRA. The FTs are employed to model the control logic of several safety systems and to propagate component failures throughout the system. Provided the generated dynamic PRA data, we show how conservative assumptions in the original PRA can be identified and how such original PRA can be modified by updating success criteria captured by the set of RELAP5-3D simulation runs.

97 - MATHEMATICS AND COMPUTING↗
DOE OSTI2021

End-to-End Differentiable Modeling and Management of the Environment

Focal Area: (2) Data acquisition and assimilation enabled by machine learning, AI, and advanced methods including experimental/network design/optimization. We emphasize the importance of leveraging optimization techniques from AI/machine learning (ML) to solve challenging problems in Earth system modeling. Science Challenge: Automatic differentiation has had a transformative effect on ML by allowing the calculation of gradients of arbitrary functions in an incredibly large class of models. We can potentially realize similar improvements in parameter estimation and control for Earth system models (ESMs) by reimplementing them in computational frameworks from ML. Practitioners working with large (>10 7 parameters) models in ML and AI can obtain good predictive performance in a range of spatiotemporal tasks by making use of optimization via stochastic gradient descent and incorporating prior knowledge at multiple levels. We propose writing ESMs in open-source computational frameworks such as Torch, Tensorflow, and JAX to greatly expand the scope of environmental forecasting and management challenges, which can be addressed by leveraging automatic differentiation and gradient descent-like algorithms. We do not call for a wholesale replacement of physical models with data-driven surrogates, but rather advocate for interleaving physical and empirical equations in a manner that is most faithful to the extent of our scientific knowledge and observational data. Central to this topic is the merging of differentiable physical simulations with differentiable optimization layers, which are now both beginning to come to the forefront.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Multiscale-Informed Modeling of High Temperature Component Response with Uncertainty Quantification

This report summarizes a joint effort between Argonne National Laboratory, Idaho National Laboratory, and Los Alamos National Laboratory to develop and deploy constitutive models targeted at predicting the life of Grade 91 alloy components subjected to high temperature environments typical of those that structural components in advanced nuclear reactors would experience. Two distinct, but complementary constitutive modeling approaches have been taken here. The first employs a phenomenological viscoplastic model for which parameters have been calibrated based on experimental data for a wide range of Grade 91 alloy that has undergone a variety of processing. A Bayesian approach was used to derive distributions of uncertain parameters for this model based on this data set. The second approach is a reduced order model suitable for engineering-scale analysis that is based on the results of a large set of mesoscale simulations. Mesoscale models allow for the microstructure and composition of a particular alloy to be directly taken into account in the computation of the viscoplastic response, but are computationally expensive, which makes it impractical to directly call those models for the material constitutive response in an engineering-scale simulation. The reduced-order representation of the response of the underlying model used here allows for an engineering-scale model to take into account the characteristics of the underlying microstructure, while only incurring a reasonable computational expense. Both of these approaches have different strengths, and are applicable for different parts of the design/analysis process. The phenomenological models can be readily parameterized based on a set of experimental data for a given class of materials and used for scoping calculations. Once a specific material is chosen and adequately characterized, the reduced order models can accurately predict the response of that specific alloy, and because the models are based on predictive models of the underlying microstructure, they can be used to more confidently predict the response under conditions in regions where there is limited experimental data. Both of these models have been integrated in the Grizzly code, which is used here to perform proof-of-concept uncertainty quantification analyses of a simple component under prototypical conditions. The built- in stochastic analysis capabilities in the MOOSE framework that Grizzly is built on are used here to run large sets of simulations for this uncertainty quantification analysis. As would be expected, because the reduced order models are developed for a much more tightly defined alloy, they predict tighter distributions of the time to failure than the phenomenological models, which are calibrated to a broader set of data. Also important is that these simulations demonstrate that a reduced order modeling approach can be successfully deployed to propagate uncertainties from the material scale to practical engineering-scale component simulations.

42 ENGINEERING↗
DOE OSTI2023

Solving the nuclear pairing model with neural network quantum states

In this work, we present a variational Monte Carlo method that solves the nuclear many-body problem in the occupation number formalism exploiting an artificial neural network representation of the groundstate wave function. A memory-efficient version of the stochastic reconfiguration algorithm is developed to train the network by minimizing the expectation value of the Hamiltonian. We benchmark this approach against widely used nuclear many-body methods by solving a model used to describe pairing in nuclei for different types of interaction and different values of the interaction strength. Despite its polynomial computational cost, our method outperforms coupled-cluster and provides energies that are in excellent agreement with the numerically-exact full configuration interaction values.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
NASA NTRS1990

Stochastic estimation of tropospheric path delays in global positioning system geodetic measurements

Water vapor radiometric and surface meteorological measurements taken during three GPS geodetic experiments are used to calculate process noise levels for random walk and first-order Gauss-Markov temporal models of tropospheric path delays. Entire wet and combined wet and dry zenith delays at each network site are estimated simultaneously with the geodetic parameters without prior calibration. The path delays and corresponding baseline estimates are compared to those obtained with calibrated data and stochastic residual delays. In this manner, the marginal utility of a priori tropospheric calibration is assessed given the ability to estimate the path delays directly using only the GPS data. Estimation of total zenith path delays with appropriate random walk or Gauss-Markov models yields baseline repeatabilities of a few parts in 100 million. Results suggest that GPS data alone have sufficient strength to resolve centimeter-level zenith path delay fluctuations over periods of a few minutes.

Tralli, David M.↗
DOE OSTI2021

Gravitational waves from first-order phase transitions in Majoron models of neutrino mass

We show how the generation of right-handed neutrino masses in Majoron models may be associated with a first-order phase transition and accompanied by the production of a stochastic background of gravitational waves (GWs). We explore different energy scales with only renormalizable operators in the effective potential. If the phase transition occurs above the electroweak scale, the signal can be tested by future interferometers. We consider two possible energy scales for phase transitions below the electroweak scale. If the phase transition occurs at a GeV, the signal can be tested at LISA and provide a complementary cosmological probe to right-handed neutrino searches at the FASER detector. If the phase transition occurs below 100 keV, we find that the peak of the GW spectrum is two or more orders of magnitude below the putative NANOGrav GW signal at low frequencies, but well within reach of the SKA and THEIA experiments. We show how searches of very low frequency GWs are motivated by solutions to the Hubble tension in which ordinary neutrinos interact with the dark sector. We also present general calculations of the phase transition temperature and Euclidean action that apply beyond Majoron models.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Z 2 topological order and first-order quantum phase transitions in systems with combinatorial gauge symmetry

We study a generalization of the two-dimensional transverse-field Ising model, combining both ferromagnetic and antiferromagnetic two-body interactions, that hosts exact global and local Z 2 gauge symmetries. Using exact diagonalization and stochastic series expansion quantum Monte Carlo methods, we confirm the existence of the topological phase in line with previous theoretical predictions. Our simulation results show that the transition between the confined topological phase and the deconfined paramagnetic phase is of first order, in contrast to the conventional Z 2 lattice gauge model in which the transition maps onto that of the standard Ising model and is continuous. We further generalize the model by replacing the transverse field on the gauge spins with a ferromagnetic X X interaction while keeping the local gauge symmetry intact. We find that the Z 2 topological phase remains stable, while the paramagnetic phase is replaced by a ferromagnetic phase. The topological-ferromagnetic quantum phase transition is also of first order. For both models, we discuss the low-energy spinon and vison excitations of the topological phase and their avoided level crossings associated with the first-order quantum phase transitions.

2-dimensional systems↗
NASA NTRSDate not supplied

Recommended DSMC Collision Model Parameters for Planetary Entry

Hybrid simulation frameworks combining Computational Fluid Dynamics (CFD) and Direct Simulation Monte Carlo (DSMC) are frequently employed to efficiently perform high-fidelity simulations of environments consisting of both continuum and rarified flow. DSMC is a stochastic, particle-based method which solves the fundamental Boltzmann equation and is therefore necessary for high-Knudsen flow where continuum-based assumptions governing CFD break down. However, the DSMC methodology is generally computationally inefficient to model the continuum regime. In a CFD/DSMC hybrid approach, obtaining an accurate, high-fidelity solution hinges on the consistent treatment of transport properties and the thermo-chemical models employed within the two solvers. In principle, in regions where CFD and DSMC are both employed, the same gas mixture under the same conditions should have the same properties, regardless of simulation type. Observed differences should be due to non-equilibrium processes, rather than differences in physical models. The goal of this work is to provide a comprehensive DSMC transport database encompassing the vast majority of known gas species encountered during Earth or other planetary atmospheric entry. This goal is largely possible due to recently performed ab initio quantum chemistry calculations. Combined with other high-fidelity data, the planned database will consist of collision integral data for over 200 neutral and ionized species and over 20000 binary collisions. From these collision integrals, Nelder-Mead optimization is used to compute collision-specific Variable Soft Sphere (VSS) collision model parameters, fitted from 300 K to 20000 K. Initial comparisons of transport properties of relevant equilibrium gas mixtures show great agreement between CFD and DSMC-derived results. The completed database can be readily applied to model binary collisions of any gas mixture containing the included species over the specified temperature range, making it a valuable tool for future planetary probe modeling efforts. An example is shown below in Fig. 1. Equilibrium mixture transport properties for a 35-species mixture composed originally of 10% air and 90% pyrolysis species of a carbon-phenolic ablator material [4] are computed using both fitted VSS Parameters and the original CFD collision integral values. Deviations in computed properties between the two approaches are less than 5% for the entire temperature range.

M. R. Gosma↗
DOE OSTI2022

Decoding Golden Eagle Movement Behavior from High-Resolution, Variable-Rate Telemetry Data Through Bayesian Filtering

The recent advances in animal tracking technology have enabled the collection of a vast amount of in situ data regarding the movement of wildlife at high spatiotemporal resolution. These data are usually available at variable time resolutions and contains noise (error) originating from GPS fixes. Decoding movement characteristics, particularly of flying animals, from telemetry data while handling these factors is a challenging yet important task for conservation purposes. Typically, this task is broken into two subtasks: resampling, and model calibration. The resampling subtask converts the variable rate positional data into a constant time interval data, while the model calibration subtask uses the resampled data to tune time-invariant parameters of the proposed models. For telemetry data at high temporal resolutions (order of 1 second), it is very challenging to decouple noise from actual movements using interpolation-based resampling techniques. Any errors introduced during resampling can significantly alter the the calibration and prediction attributes of the movement model. We address this problem through a unified Bayesian state-space framework that can handle both the resampling and calibration tasks in a single step. In addition, we use the speed and heading of the bird from telemetry data to regularize the position information of the bird. We use a Kalman filtering approach to include these nonlinearly related motion parameters within the state space framework. We cross-validated to quantify how this inclusion affects the model performance in estimating true bird movements. The relationship between the true state of the bird and environmental and topographical covariates is then represented parametrically. These parameters are then tuned using stochastic sampling strategies like Markov Chain Monte Carlo (MCMC). We use the telemetry data collected from golden eagles in the western USA to demonstrate the applicability of this approach to build a predictive, probabilistic movement model. Our preliminary results show that this approach provides improved predictive performance in terms of capturing higher-order motion parameters such as angular and horizontal accelerations, which may have simpler and more direct relationships with environmental covariates than corresponding speeds. In this talk, we will demonstrate how this state-space approach benefits the prediction capabilities of a movement model in simulating golden eagle paths through a wind power plant in Wyoming given certain atmospheric conditions. The model outcomes are aimed at informing mitigation strategies that can minimize the potential for collisions of golden eagles with wind turbines.

Bayesian methods↗