Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Monte Carlo Method”

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 775 records · Page 43

Stan : A Probabilistic Programming Language

Stan is a probabilistic programming language for specifying statistical models. A Stan program imperatively defines a log probability function over parameters conditioned on specified data and constants. As of version 2.14.0, Stan provides full Bayesian inference for continuous-variable models through Markov chain Monte Carlo methods such as the No-U-Turn sampler, an adaptive form of Hamiltonian Monte Carlo sampling. Penalized maximum likelihood estimates are calculated using optimization methods such as the limited memory Broyden-Fletcher-Goldfarb-Shanno algorithm. Stan is also a platform for computing log densities and their gradients and Hessians, which can be used in alternative algorithms such as variational Bayes, expectation propagation, and marginal inference using approximate integration. To this end, Stan is set up so that the densities, gradients, and Hessians, along with intermediate quantities of the algorithm such as acceptance probabilities, are easily accessible. Stan can also be called from the command line using the cmdstan package, through R using the rstan package, and through Python using the pystan package. All three interfaces support sampling and optimization-based inference with diagnostics and posterior analysis. rstan and pystan also provide access to log probabilities, gradients, Hessians, parameter transforms, and specialized plotting.

97 MATHEMATICS AND COMPUTING↗

A Brownian dynamics study on ferrofluid colloidal dispersions using an iterative constraint method to satisfy Maxwell’s equations

Ferrofluids are often favored for their ability to be remotely positioned via external magnetic fields. The behavior of particles in ferromagnetic clusters under uniformly applied magnetic fields has been computationally simulated using the Brownian dynamics, Stokesian dynamics, and Monte Carlo methods. However, few methods have been established that effectively handle the basic principles of magnetic materials, namely, Maxwell’s equations. An iterative constraint method was developed to satisfy Maxwell’s equations when a uniform magnetic field is imposed on ferrofluids in a heterogeneous Brownian dynamics simulation that examines the impact of ferromagnetic clusters in a mesoscale particle collection. This was accomplished by allowing a particulate system in a simple shear flow to advance by a time step under a uniformly applied magnetic field, then adjusting the ferroparticles via an iterative constraint method applied over sub-volume length scales until Maxwell’s equations were satisfied. The resultant ferrofluid model with constraints demonstrates that the magnetoviscosity contribution is not as substantial when compared to homogeneous simulations that assume the material’s magnetism is a direct response to the external magnetic field. This was detected across varying intensities of particle-particle interaction, Brownian motion, and shear flow. Ferroparticle aggregation was still extensively present but less so than typically observed.

42 ENGINEERING↗

Improving the sampling efficiency of the Grand Canonical Simulated Quenching approach

Most common atomistic simulation techniques, like molecular dynamics or Metropolis Monte Carlo, operate under a constant interatomic Hamiltonian with a fixed number of atoms. Internal (atom positions or velocities) or external (simulation cell size or geometry) variables are then evolved dynamically or stochastically to yield sampling in different ensembles, such as microcanonical (NVE), canonical (NVT), isothermal-isobaric (NPT), etc. Averages are then taken to compute relevant physical properties. At least two limitations of these standard approaches can seriously hamper their application to many important systems: (1) they do not allow for the exchange of particles with a reservoir, and (2) the sampling efficiency is insufficient to allow the obtention of converged results because of the very long intrinsic timescales associated with these quantities. To fix ideas, one might want to identify low (free) energy configurations of grain boundaries (GB). In reality, grain boundaries are in contact the grains which act as reservoirs of defects (e.g., vacancies and interstitials). Since the GB can exchange particles with its environment, the most stable configuration cannot provably be found by sampling from NVE or NVT ensembles alone: one needs to allow the number of atoms in the sample to fluctuate. The first limitation can be circumvented by working in the grand canonical ensemble (TV ) or its derivatives (such as the semi-grand-canonical ensemble useful for the study of substitutional alloys). Monte Carlo methods have been the first to adapt to this kind of system where the number of atoms is allowed to fluctuate. Many of these methods are based on the Widom insertion method [Widom63] where the chemical potential of a given chemical species can be inferred from the potential energy changes upon random insertion of a new particle within the simulation cell. Other techniques, such as the Gibbs ensemble Monte Carlo [Panagiotopoulos87] where exchanges of particles are attempted to equilibrate the chemical potential between two cells and hence allow for the calculation of coexistence curves, exploit the same idea: particle insertion (or exchange) is attempted and accepted with a Metropolis-like rule that depends exponentially on the energy change upon insertion. A well known limitation of this kind of approach is that the probability of accepting such a move decreases extremely rapidly with increasing density, due to the extremely large short-range repulsion between atoms. In response to these difficulties it became apparent that a solution to the problem might be to avoid abrupt insertions but instead to proceed gradually, so as to allow the system to react and make way for the incoming particle. In this view of things, the 'occupation' of a certain atomic site can be viewed as a continuous variable, ranging between 0 and 1, representing 'how much' of the particle is present at any given time. These ideas proved ideal in Molecular Dynamics (MD) settings because equations of motions for these occupation variables can sometimes be obtained. For example, in the case of Grand Canonical Molecular Dynamics [Cagin91], one special particle is allowed to have a fractional occupation. This can lead to either its destruction (occupation = 0) or its complex creation (occupation = 1) so as to enforce a given chemical potential. These approaches proved useful, but mostly in the liquid state where the probability of successfully inserting a new particle is sufficiently high. At higher densities, convergence proved to be hampered by very inefficient sampling. In this work, we explore the use of a related MD-based grand canonical technique, the Grand Canonical Simulated Quenching (GCSQ) of Phillpot and Rickman [Phillpot92,Phillpot94], and explore its application to the grand canonical sampling of solid state systems. We show that, in conjunction with advanced sampling techniques, GCSQ can be a useful tool to sample conformations of complex systems, such as GBs, and assist in the identification of their most stable states and/or most likely defective states.

36 MATERIALS SCIENCE↗

Benchmark analysis of high temperature engineering test reactor core using McCARD code

A benchmark calculation has been performed for a startup core physics test of Japan's High Temperature Engineering Test Reactor (HTTR). The calculation is carried out by the McCARD code, which adopts the Monte Carlo method. The cross section library is ENDF-B/VII.0. The fuel cell is modeled by the reactivity-equivalent physical transform (RPT) method. Effective multiplication factors with different numbers of fuel columns have been analyzed. The calculation shows that the HTTR becomes critical with 19 fuel columns with an excess reactivity of 0.84% Δk/k. The discrepancies between the measurements and Monte Carlo calculations are 2.2 and 1.4 % Δk/k for 24 and 30 columns, respectively. The reasons for the discrepancy are thought to be the current version of cross section library and the impurity in the graphite which is represented by the boron concentration. In the future, the depletion results will be proposed for further benchmark calculations. (authors)

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Secondary electron emission of Y{sub 2}O{sub 3}-Mo cermet cathode

Y{sub 2}O{sub 3}-Mo cermet cathodes were prepared by mechanical mixing of Y{sub 2}O{sub 3} and Mo powder. The prepared cermet cathode provides a certain secondary emission yield (3.09), about 1.8 times lower than that for the cathode prepared by sol-gel method. It was found that the cathode prepared by sol-gel method had a smaller grain size, which could be attributed to its higher emission property. The energy distributions of primary electrons in the cathodes with different grain sizes have been simulated by Monte Carlo method. Based on the calculation results, the emission models of the cathodes are established and presented. The cathodes with different grain sizes exhibited different ways of emission. Two types of emission, of reflection emission and transmission emission, existed in the cathode with small grain sizes whereas only transmission emission existed in the cathode with large grain sizes.

36 MATERIALS SCIENCE↗

BCS-BEC Crossover in a Two-Dimensional Fermi Gas

We investigate the crossover from Bardeen-Cooper-Schrieffer (BCS) superfluidity to Bose-Einstein condensation (BEC) in a two-dimensional Fermi gas at T=0 using the fixed-node diffusion Monte Carlo method. We calculate the equation of state and the gap parameter as a function of the interaction strength, observing large deviations compared to mean-field predictions. In the BEC regime our results show the important role of dimer-dimer and atom-dimer interaction effects that are completely neglected in the mean-field picture. Results on Tan's contact parameter associated with short-range physics are also reported along the BCS-BEC crossover.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Benchmark results for few-body hypernuclei

Here, the Non-Symmetrized Hyperspherical Harmonics method (NSHH) is introduced in the hypernuclear sector and benchmarked with three different ab-initio methods, namely the Auxiliary Field Diffusion Monte Carlo method, the Faddeev–Yakubovsky approach and the Gaussian Expansion Method. Binding energies and hyperon separation energies of three- to five-body hypernuclei are calculated by employing the two-body ΛN component of the phenomenological Bodmer–Usmani potential, and a hyperon-nucleon interaction simulating the scattering phase shifts given by NSC97f. The range of applicability of the NSHH method is briefly discussed.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Characterization and Monte Carlo Simulations of a 4-Liter NaI Detection System for Use During Nuclear and Radiological Emergencies and for the Detection of Nuclear Materials

SCKCEN is engaged in research related to the use of a large volume NaI detection system for use during nuclear and radiological emergencies and for the detection of nuclear materials. As part of this effort a research activity is being carried out to determine the characteristics and build a Monte Carlo model of the available 4-liter NaI detector. These activities consist of (i) determination of energy resolution efficiency lower threshold background in laboratory conditions using well-characterized point sources; (ii) modeling of the detector by Monte Carlo methods and validation; (iii) determination of the response function by combining measurement and simulation results; (iv) characterization of the detector for contaminated surfaces by measurements and modeling; and (v) calibration of detector for dose rate measurements. This paper will focus on the results obtained so far concerning the Monte Carlo model and its validation with experimental data.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Modelling of SAR polarisation phase difference from trees

The data for polarization phase difference Delta Phi between the HH- and VV-polarized backscattered waves from tree-covered fields were obtained with an airborne synthetic aperture radar at 1.225 GHz. The mean values over tree-covered fields were derived from the images of the phase difference and were examined as a function of incident beam angle from 15 to 55 deg. A theoretical model for simulating these data, based on the electromagnetic wave scatterings from the tree trunk and its branches, both of which are assumed as very long dielectric cylinders was developed. The radius and direction of a tree branch are taken as random variables and are chosen by a Monte Carlo method to encounter the incident waves in producing the scattering events. The Monte Carlo simulated results are in good agreement with the observations within experimental uncertainty.

Mo, Tsan↗

Microwave radiative transfer through horizontally inhomogeneous precipitating clouds

Recent advances in cloud microphysical models have led to realistic three-dimensional distributions of cloud constituents. Radiative transfer schemes can make use of this detailed knowledge in order to study the effects of horizontal as well as vertical inhomogeneities within clouds. This study looks specifically at the differences between three-dimensional radiative transfer results and those obtained by plane parallel, independent pixel approximations in the microwave spectrum. A three-dimensional discrete ordinates method as well as a backward Monte Carlo method are used to calculate realistic radiances emerging from the cloud. Analyses between these models and independent pixel approximations reveal that plane parallel approximations introduce two distinct types of errors. The first error is physical in nature and is related to the fact that plane parallel approximations do not allow energy to leak out of dense areas into surrouding areas. In general, it was found that these errors are quite small for emission-dominated frequencies (37 GHz and lower) and that physical errors are highly pronounced only at scattering frequencies (85 GHz) where large deviations and biases up to 8 K averaged over the entire cloud were found. The second error is more geometric in nature and is related to the fact that plane parallel approximations cannot accommodate physical boundaries in the horizontal dimension for off-nadir viewing angles. The geometric errors were comparable in magnitude for all frequencies. Their magnitude, however, depends on a number of factors including the scheme used to deal with the edge, the nature of the surface, and the viewing angle.

Roberti, Laura↗

A deterministic particle method for one-dimensional reaction-diffusion equations

We derive a deterministic particle method for the solution of nonlinear reaction-diffusion equations in one spatial dimension. This deterministic method is an analog of a Monte Carlo method for the solution of these problems that has been previously investigated by the author. The deterministic method leads to the consideration of a system of ordinary differential equations for the positions of suitably defined particles. We then consider the time explicit and implicit methods for this system of ordinary differential equations and we study a Picard and Newton iteration for the solution of the implicit system. Next we solve numerically this system and study the discretization error both analytically and numerically. Numerical computation shows that this deterministic method is automatically adaptive to large gradients in the solution.

Mascagni, Michael↗

Accelerate Nuclear Research and Development by Reducing Time and Cost Spend in the Pre-conceptual Design Phase of Advanced Reactor Experiments

The design process of every new concept, such as advanced nuclear reactors or associated experiments, starts with the pre-conceptual design phase. In this phase, the viability of a wide range of design options needs to be assessed quickly, to understand the operating envelope and its feasibility. A variety of physics models (thermal-hydraulics, neutronics, mechanical design, etc.) has to be considered at this very first design stage and optimum component sizes and materials (e.g. heat exchangers, piping, turbomachinery, coolant type, etc.) have to be chosen for a given set of boundary conditions (e.g. heat source, heat sink, flow rate, etc.). Detailed solutions such as provided by high fidelity methods like computational fluid dynamics (CFD), Monte Carlo methods, etc. and even lower fidelity tools such as system or subchannel codes, etc. are usually not used during the pre-conceptual design due to the relatively long time needed to create input models, the computational time to obtain a solution and the lack of flexibility to quickly investigate different combinations of components, individual component sizes and material properties. High fidelity tools are usually only employed in the conceptual design and later phases once a base concept has been identified during the pre-conceptual design stage. The current practice during the pre-conceptual design stage is that analysts collect the needed equations, material properties, closure laws, etc. and create ad-hoc solutions form scratch for every new problem. There clearly is a lack of a flexible scoping tool that can be used during pre-conceptional design before higher fidelity tools (as described above) come into play. To reduce user errors in ad-hoc solutions and increase fidelity and efficiency, this project aims to investigate and develop a user-friendly scoping tool to address the thermal-hydraulic designing needs during preconceptual experiment design, i.e. Thermal-hydraulic Research Universal Scoping Tool (TRUST). The success of TRUST will provide the nuclear engineers with an easy-to-use and affordable calculator for early reactor system design and optimization.

42 ENGINEERING↗

Final Technical Report for DOE Grant, number DE-FG02-05ER15701; Probing Surface Chemistry Under Catalytic Conditions: Olefin Hydrogenation,Cyclization and Functionalization.

The specific goal of this work was to understanding the catalytic reactions pathways for the synthesis of vinyl acetate over Pd, Au and PdAu alloys. A combination of both experimental methods (X-ray and Auger spectroscopies, low-energy ion scattering (LEIS), low-energy electron diffraction (LEED) and theory (Density Functional Theory (DFT) calculations and Monte Carlo methods under various different reactions) were used to track the surface chemistry and the influence of alloying. The surface intermediates involved in the various reactions were characterized using reflection-absorption infrared spectroscopy and LEED to identify the nature of the surface species and temperature-programmed desorption (TPD) to follow the decomposition pathways and measure heats of adsorption. These results along with those from density functional theoretical calculations were used determine the kinetics for elementary steps. The results from this work showed that the reaction proceeds via the Samanos mechanism over Pd surfaces whereby the ethylene directly couples with acetate to form an acetoxyethyl intermediate that subsequently undergoes a beta-hydride elimination to form the vinyl acetate monomer. The presence of Au was found to modify the adsorption energies and surface coverages of important surface intermediates including acetate, ethylidyne and ethylene which ultimately influences the critical C-H activation and coupling steps. By controlling the surface alloy composition or structure one can begin to control the steps that control the rate and even the mechanism.

36 MATERIALS SCIENCE↗

IMPROVING SENSITIVITY TO WEAK PULSATIONS WITH PHOTON PROBABILITY WEIGHTING

All {gamma}-ray telescopes suffer from source confusion due to their inability to focus incident high-energy radiation, and the resulting background contamination can obscure the periodic emission from faint pulsars. In the context of the Fermi Large Area Telescope, we outline enhanced statistical tests for pulsation in which each photon is weighted by its probability to have originated from the candidate pulsar. The probabilities are calculated using the instrument response function and a full spectral model, enabling powerful background rejection. With Monte Carlo methods, we demonstrate that the new tests increase the sensitivity to pulsars by more than 50% under a wide range of conditions. This improvement may appreciably increase the completeness of the sample of radio-loud {gamma}-ray pulsars. Finally, we derive the asymptotic null distribution for the H-test, expanding its domain of validity to arbitrarily complex light curves.

79 ASTRONOMY AND ASTROPHYSICS↗

Boyd Report1

Andrew Crow arrived at Lawrence Livermore National Laboratory with the intention of continuing work on the Complex Particle Kinetic (CPK) method developed D. Larson and D. Hewett. Andrew Crow had previously worked on duplicating the results of D. Hewett in his previous work. Since arrival, A. Crow has been working with D. Larson on a slightly different project. The current method, still under development, is a Particle in Cell (PIC) code with the following features: (1) all particles begin each timestep at a gridpoint; (2) particles are then advanced in time using a standard special advancement method. The exact method has not been decided upon, but there are many reliable methods from which to choose. (3) All particles within each cell undergo a simultaneous implicit collision step. This is the current area of focus. Currently, A. Crow is not aware of any method of performing implicit collisions over a large number of charged particles. Implicit methods for charged particle movement and electron-electron collisions, have been developed. The work of L. pareschi and G. Russo on the Time Relaxed Direct Simulation Monte Carlo method, also appears to be a good basis for implicit particle collisions. (4) Each individual particle will be divided into a set of particles with a Gaussian velocity distribution. This will collect some of the thermal effects created by the collisions. This algorithm has not been created. (5) Particles will be projected on to the grid points. Currently, a linear weighting technique is intended to be used, but has not settled upon. (6) Once on the gridpoints the particle number will be reduced using a set of quadrature points based on the third order velocity moments of the particles. The method proposed by R. Fox has been programmed and shown to conserve energy, momentum and mass to machine precision. In addition to reducing the number of particles this method will work to quiet the simulation it will behave as a higher order version of the Quiet DSMC method proposed by B. Albright et al. (7) These quadrature points then become the new particles for the next timestep. the advantage of this method can be many: The self force on ions can be easily removed since all particles begin on grid points. The size of the timesteps should not be limited by collision rate, and should only be impacted by particle travel time through the cell. The particle reduction technique should keep many of the higher order features of the particle distribution while reducing the number of particles in the system. It should also quite the variance in the system. The two largest unknowns, at this time are, how large a part numerical diffusion will play in the scheme and how computationally expensive each timestep will be.

42 ENGINEERING↗

MODELING THE MULTI-WAVELENGTH EMISSION OF THE SHELL-TYPE SUPERNOVA REMNANT RX J1713.7-3946

Emission mechanisms of the shell-type supernova remnant (SNR) RX J1713.7-3946 are studied with multi-wavelength observational data from the radio, X-ray, GeV {gamma}-ray, and TeV {gamma}-ray bands. A Markov Chain Monte Carlo method is employed to explore the high-dimensional model parameter space systematically. Three scenarios for the {gamma}-ray emission are investigated: the leptonic, the hadronic, and a hybrid. Thermal emission from the background plasma is also included to constrain the gas density, assuming ionization equilibrium, and a 2{sigma} upper limit of about 0.03 cm{sup -3} is obtained as far as thermal energies account for a significant fraction of the dissipated kinetic energy of the SNR shock. Although systematic errors dominate the {chi}{sup 2} of the spectral fit of all models, we find that (1) the leptonic model has the best constrained model parameters, whose values can be easily accommodated with a typical supernova, but gives a relatively poor fit to the TeV {gamma}-ray data; (2) the hybrid scenario has one more parameter than the leptonic one and improves the overall spectral fit significantly; and (3) the hadronic one, which has three more parameters than the leptonic model, gives the best fit to the overall spectrum with relatively poorly constrained model parameters and very hard spectra of accelerated particles. The uncertainties of the model parameters decrease significantly if the spectral indices of accelerated electrons and protons are the same. The hybrid and hadronic models also require an energy input into high-energy protons, which seems to be too high compared with typical values for a supernova explosion. Further investigations are required to reconcile these observations with SNR theories.

79 ASTRONOMY AND ASTROPHYSICS↗