Search NASA⌕ Search

SEARCH · Search NASA

Results for “Computational Physics”

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 235 records · Page 13

Preliminary Computational Analysis of the (HIRENASD) Configuration in Preparation for the Aeroelastic Prediction Workshop

This paper presents preliminary computational aeroelastic analysis results generated in preparation for the first Aeroelastic Prediction Workshop (AePW). These results were produced using FUN3D software developed at NASA Langley and are compared against the experimental data generated during the HIgh REynolds Number Aero- Structural Dynamics (HIRENASD) Project. The HIRENASD wind-tunnel model was tested in the European Transonic Windtunnel in 2006 by Aachen University0s Department of Mechanics with funding from the German Research Foundation. The computational effort discussed here was performed (1) to obtain a preliminary assessment of the ability of the FUN3D code to accurately compute physical quantities experimentally measured on the HIRENASD model and (2) to translate the lessons learned from the FUN3D analysis of HIRENASD into a set of initial guidelines for the first AePW, which includes test cases for the HIRENASD model and its experimental data set. This paper compares the computational and experimental results obtained at Mach 0.8 for a Reynolds number of 7 million based on chord, corresponding to the HIRENASD test conditions No. 132 and No. 159. Aerodynamic loads and static aeroelastic displacements are compared at two levels of the grid resolution. Harmonic perturbation numerical results are compared with the experimental data using the magnitude and phase relationship between pressure coefficients and displacement. A dynamic aeroelastic numerical calculation is presented at one wind-tunnel condition in the form of the time history of the generalized displacements. Additional FUN3D validation results are also presented for the AGARD 445.6 wing data set. This wing was tested in the Transonic Dynamics Tunnel and is commonly used in the preliminary benchmarking of computational aeroelastic software.

Chwalowski, Pawel↗

A Mesh-Free Method to Predictively Simulate Solid-to-Liquid Phase Transitions in Abnormal Thermal Environments

Particle methods in computational physics are useful for modeling the motion of fluids and solids subject to large deformations. Under these conditions, mesh-based approaches often fail due to decreasing element quality leading to inaccuracy and instability. The developed software package called Moab investigates and prototypes next-generation particle methods, focusing on rigorous error analysis and active error minimization strategies during the computation. The present work discusses examples calculations representative of real engineering problems with quantified and maximized accuracy while demonstrating the potential for meeting engineering performance requirements.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Improved best estimate plus uncertainty methodology including advanced validation concepts to license evolving nuclear reactors

Many evolving nuclear energy programs plan to use advanced predictive multi-scale multi-physics simulation and modeling capabilities to reduce cost and time from design through licensing. Historically, the role of experiments was primary tool for design and understanding of nuclear system behavior while modeling and simulation played the subordinate role of supporting experiments. In the new era of multi-scale multi-physics computational based technology development, the experiments will still be needed but they will be performed at different scales to calibrate and validate models leading predictive simulations. Cost saving goals of programs will require us to minimize the required number of validation experiments. Utilization of more multi-scale multi-physics models introduces complexities in the validation of predictive tools. Traditional methodologies will have to be modified to address these arising issues. This paper lays out the basic aspects of a methodology that can be potentially used to address these new challenges in design and licensing of evolving nuclear technology programs. The main components of the proposed methodology are verification, validation, calibration, and uncertainty quantification. An enhanced calibration concept is introduced and is accomplished through data assimilation. The goal is to enable best-estimate prediction of system behaviors in both normal and safety related environments. To achieve this goal requires the additional steps of estimating the domain of validation and quantification of uncertainties that allow for extension of results to areas of the validation domain that are not directly tested with experiments, which might include extension of the modeling and simulation (M&S) capabilities for application to full-scale systems. The new methodology suggests a formalism to quantify an adequate level of validation (predictive maturity) with respect to required selective data so that required testing can be minimized for cost saving purposes by showing further testing wold not enhance the quality of the validation of predictive tools. The proposed methodology is at a conceptual level. When matured and if considered favorably by the stakeholders, it could serve as a new framework for the next generation of the best estimate plus uncertainty licensing methodology that USNRC developed previously. In order to come to that level of maturity it is necessary to communicate the methodology to scientific, design and regulatory stakeholders for discussion and debates. This paper is the first step to establish this communication.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Computation Directorate 2007 Annual Report

If there is a single word that both characterized 2007 and dominated the thoughts and actions of many Laboratory employees throughout the year, it is transition. Transition refers to the major shift that took place on October 1, when the University of California relinquished management responsibility for Lawrence Livermore National Laboratory (LLNL), and Lawrence Livermore National Security, LLC (LLNS), became the new Laboratory management contractor for the Department of Energy's (DOE's) National Nuclear Security Administration (NNSA). In the 55 years under the University of California, LLNL amassed an extraordinary record of significant accomplishments, clever inventions, and momentous contributions in the service of protecting the nation. This legacy provides the new organization with a built-in history, a tradition of excellence, and a solid set of core competencies from which to build the future. I am proud to note that in the nearly seven years I have had the privilege of leading the Computation Directorate, our talented and dedicated staff has made far-reaching contributions to the legacy and tradition we passed on to LLNS. Our place among the world's leaders in high-performance computing, algorithmic research and development, applications, and information technology (IT) services and support is solid. I am especially gratified to report that through all the transition turmoil, and it has been considerable, the Computation Directorate continues to produce remarkable achievements. Our most important asset--the talented, skilled, and creative people who work in Computation--has continued a long-standing Laboratory tradition of delivering cutting-edge science even in the face of adversity. The scope of those achievements is breathtaking, and in 2007, our accomplishments span an amazing range of topics. From making an important contribution to a Nobel Prize-winning effort to creating tools that can detect malicious codes embedded in commercial software; from expanding BlueGene/L, the world's most powerful computer, by 60% and using it to capture the most prestigious prize in the field of computing, to helping create an automated control system for the National Ignition Facility (NIF) that monitors and adjusts more than 60,000 control and diagnostic points; from creating a microarray probe that rapidly detects virulent high-threat organisms, natural or bioterrorist in origin, to replacing large numbers of physical computer servers with small numbers of virtual servers, reducing operating expense by 60%, the people in Computation have been at the center of weighty projects whose impacts are felt across the Laboratory and the DOE community. The accomplishments I just mentioned, and another two dozen or so, make up the stories contained in this report. While they form an exceptionally diverse set of projects and topics, it is what they have in common that excites me. They share the characteristic of being central, often crucial, to the mission-driven business of the Laboratory. Computational science has become fundamental to nearly every aspect of the Laboratory's approach to science and even to the conduct of administration. It is difficult to consider how we would proceed without computing, which occurs at all scales, from handheld and desktop computing to the systems controlling the instruments and mechanisms in the laboratories to the massively parallel supercomputers. The reasons for the dramatic increase in the importance of computing are manifest. Practical, fiscal, or political realities make the traditional approach to science, the cycle of theoretical analysis leading to experimental testing, leading to adjustment of theory, and so on, impossible, impractical, or forbidden. How, for example, can we understand the intricate relationship between human activity and weather and climate? We cannot test our hypotheses by experiment, which would require controlled use of the entire earth over centuries. It is only through extremely intricate, detailed computational simulation that we can test our theories, and simulating weather and climate over the entire globe requires the most massive high-performance computers that exist. Such extreme problems are found in numerous laboratory missions, including astrophysics, weapons programs, materials science, and earth science.

99 GENERAL AND MISCELLANEOUS↗

Symplectic machine learning model for fast simulation of space-charge effects

Symplectic simulation of space-charge effects is crucial for the design and operation of high-intensity particle accelerators. Traditional methods for simulating these effects are often computationally expensive, resulting in significant overhead. In this work, we introduce a generative model based on a U-Net architecture within a generative adversarial network framework to efficiently simulate space-charge effects. The model is trained to predict the transverse multiparticle space-charge Hamiltonian, which can be physically computed using a gridless spectral method. The one-step symplectic transverse transfer map for the particles is then obtained by differentiating the predicted Hamiltonian. Benchmarking results demonstrate that this generative model achieves an order of magnitude higher computational efficiency compared to the spectral method, providing a highly efficient alternative for simulating space-charge effects with a large number of particles. By maintaining symplecticity, the model effectively preserves the phase-space structure and mitigates nonphysical errors in long-term simulations. This model has been integrated into jutrack, a novel autodifferentiable accelerator modeling code developed in the julia programming language.

Beam code development & simulation techniques↗

Interrelationship between Lab, Space, Astrophysical, Magnetic Fusion, and Inertial Fusion Plasma Experiments (Graduate support for DIII-D Frontier Science Experiments)

The project, to explore the use of the DIII-D facility to investigate questions of value to discovery plasma science and provide a deepened scientific understanding behind basic plasma processes that are outside of the fusion-energy-science research line, began in September 2016 when five discovery-plasma physicists were contacted by General Atomics to form a Frontier Science Experiments (FSE) panel for brainstorming and for implementing the FSE campaign at the DIII-D National Fusion Facility. After advertising the FSE opportunity at the annual plasma physics conference, the panel grew to include six others. By January 2017, the panel for reviewing and ranking proposals was established. Evaluation of FY2017 proposals took place in February 2017. Evaluation of FY2018 proposals took place in January 2018. This proposal began 15 August 2017 and funded two research assistantships for three students who interacted, gained expertise in experimental and computational physics, and shared responsibilities on the following projects within the DIII-D (non-fusion) Frontier Science Experiments (FSE) campaign: (1) "Impact of Magnetic Perturbations on Turbulence - Zonal Flow Interactions and Saturation", (2) "Nonlinear Processes during Toroidal Alfvén Mode-Mode Interaction", (3) "Self-Consistent Chaos in Magnetic Field Dynamics", (4) “Nonmodal Decomposition of Turbulence”, and (5) “Multi-harmonic EMIC Emission -- Species mix and distribution function dependence of instabilities near the ion-cyclotron frequency”. The outcome was a fraction of one doctoral dissertation and the entirety of one master of science thesis. The work took advantage of the unique capabilities of the DIII-D National Fusion Facility to serve as the first phase of the Frontier Science Experiment (FSE) campaign, an innovative concept aiming to advance fundamental understanding in 3 key general plasma science areas - particle acceleration and energetic particles in plasmas, turbulence and transport in plasmas, and nonlinear waves and processes – all at discovery’s cutting edge.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

An Anisotropic Scale-Invariant Unstructured Mesh Generator Auitable for Volumetric Imaging Data

Mesh generation algorithms must consider the computational physics schemes to be adopted insomuch as tessellation should attempt to minimize discretization error metrics a priori, while placing elements judiciously yet economically. Basing local element size and shape on local geometric feature size is a promising approach, as the underlying physics may either be scale-invariant or may vary with scale in a predictable way. We present a boundary-fitted scale-invariant unstructured tetrahedral mesh generation algorithm that enables registration of element size to local geometric scale, given a triangulated mesh surface. The resulting tetrahedra are well-shaped and nearly orthogonal to the boundary. Unlike previous feature-based approaches, our algorithm does not require a background mesh, nor does it rely on the medial-axis. In contrast, as a first step, our algorithm produces a gradientlimited feature-size field over the input surface based on efficient ray casting. We illustrate how this field can be used to produce quality grids for computational fluid dynamics based simulations of challenging, topologically complex surfaces derived from magnetic resonance images. The algorithm is implemented in the Pacific Northwest National Laboratory (PNNL) version of the Los Alamos grid toolbox LaGriT[6].

36 MATERIALS SCIENCE↗

High-performance finite elements with MFEM

The MFEM (Modular Finite Element Methods) library is a high-performance C++ library for finite element discretizations. MFEM supports numerous types of finite element methods and is the discretization engine powering many computational physics and engineering applications across a number of domains. Furthermore, this paper describes some of the recent research and development in MFEM, focusing on performance portability across leadership-class supercomputing facilities, including exascale supercomputers, as well as new capabilities and functionality, enabling a wider range of applications. Much of this work was undertaken as part of the Department of Energy’s Exascale Computing Project (ECP) in collaboration with the Center for Efficient Exascale Discretizations (CEED).

97 MATHEMATICS AND COMPUTING↗

High-efficiency high-energy Ka source for the critically-required maximum illumination of x-ray optics on Z using Z-petawatt-driven laser-breakout-afterburner accelerated ultrarelativistic electrons LDRD .

Under the auspices of the Science of Extreme Environments LDRD program, a <2 year theoretical- and computational-physics study was performed (LDRD Project 130805) by Guy R Bennett (formally in Center-01600) and Adam B. Sefkow (Center-01600): To investigate novel target designs by which a short-pulse, PW-class beam could create a brighter K{alpha} x-ray source than by simple, direct-laser-irradiation of a flat foil; Direct-Foil-Irradiation (DFI). The computational studies - which are still ongoing at this writing - were performed primarily on the RedStorm supercomputer at Sandia National Laboratories Albuquerque site. The motivation for a higher efficiency K{alpha} emitter was very clear: as the backlighter flux for any x-ray imaging technique on the Z accelerator increases, the signal-to-noise and signal-to-background ratios improve. This ultimately allows the imaging system to reach its full quantitative potential as a diagnostic. Depending on the particular application/experiment this would imply, for example, that the system would have reached its full design spatial resolution and thus the capability to see features that might otherwise be indiscernible with a traditional DFI-like x-ray source. This LDRD began FY09 and ended FY10.

43 PARTICLE ACCELERATORS↗

Charliecloud

Charliecloud is a set of scripts to let users run a virtual cluster of virtual machines (VMs) on a desktop or supercomputer. Key functions include: 1. Creating (typically by installing an operating system from vendor media) and updating VM images; 2. Running a single VM; 3. Running multiple VMs in a virtual cluster. The virtual machines can talk to one another over the network and (in some cases) the outside world. This is accomplished by calling external programs such as QEMU and the Virtual Distributed Ethernet (VDE) suite. The goal is to let users have a virtual cluster containing nodes where they have privileged access, while isolating that privilege within the virtual cluster so it cannot affect the physical compute resources. Host configuration enforces security; this is not included in Charliecloud, though security guidelines are included in its documentation and Charliecloud is designed to facilitate such configuration. Charliecloud manages passing information from host computers into and out of the virtual machines, such as parameters of the virtual cluster, input data specified by the user, output data from virtual compute jobs, VM console display, and network connections (e.g., SSH or X11). Parameters for the virtual cluster (number of VMs, RAM and disk per VM, etc.) are specified by the user or gathered from the environment (e.g., SLURM environment variables). Example job scripts are included. These include computation examples (such as a "hello world" MPI job) as well as performance tests. They also include a security test script to verify that the virtual cluster is appropriately sandboxed. Tests include: 1. Pinging hosts inside and outside the virtual cluster to explore connectivity; 2. Port scans (again inside and outside) to see what services are available; 3. Sniffing tests to see what traffic is visible to running VMs; 4. IP address spoofing to test network functionality in this case; 5. File access tests to make sure host access permissions are enforced. This test script is not a comprehensive scanner and does not test for specific vulnerabilities. Importantly, no information about physical hosts or network topology is included in this script (or any of Charliecloud); while part of a sensible test, such information is specified by the user when the test is run. That is, one cannot learn anything about the LANL network or computing infrastructure by examining Charliecloud code.

Priedhorsky, Reid↗

Charliecloud

Charliecloud is a set of scripts to let users run a virtual cluster of virtual machines (VMs) on a desktop or supercomputer. Key functions include: 1. Creating (typically by installing an operating system from vendor media) and updating VM images; 2. Running a single VM; 3. Running multiple VMs in a virtual cluster. The virtual machines can talk to one another over the network and (in some cases) the outside world. This is accomplished by calling external programs such as QEMU and the Virtual Distributed Ethernet (VDE) suite. The goal is to let users have a virtual cluster containing nodes where they have privileged access, while isolating that privilege within the virtual cluster so it cannot affect the physical compute resources. Host configuration enforces security; this is not included in Charliecloud, though security guidelines are included in its documentation and Charliecloud is designed to facilitate such configuration. Charliecloud manages passing information from host computers into and out of the virtual machines, such as parameters of the virtual cluster, input data specified by the user, output data from virtual compute jobs, VM console display, and network connections (e.g., SSH or X11). Parameters for the virtual cluster (number of VMs, RAM and disk per VM, etc.) are specified by the user or gathered from the environment (e.g., SLURM environment variables). Example job scripts are included. These include computation examples (such as a "hello world" MPI job) as well as performance tests. They also include a security test script to verify that the virtual cluster is appropriately sandboxed. Tests include: 1. Pinging hosts inside and outside the virtual cluster to explore connectivity; 2. Port scans (again inside and outside) to see what services are available; 3. Sniffing tests to see what traffic is visible to running VMs; 4. IP address spoofing to test network functionality in this case; 5. File access tests to make sure host access permissions are enforced. This test script is not a comprehensive scanner and does not test for specific vulnerabilities. Importantly, no information about physical hosts or network topology is included in this script (or any of Charliecloud); while part of a sensible test, such information is specified by the user when the test is run. That is, one cannot learn anything about the LANL network or computing infrastructure by examining Charliecloud code.

Priedhorsky, Reid↗

Transfer learning for probabilistic localization of hidden cracks in concrete structures

Abstract The utility of discriminative supervised learning models built using multiple training-data sources is investigated for hidden crack localization in concrete. Feed-forward neural network (FFNN) is chosen as the model architecture, and transfer learning is used to assimilate the information obtained from different sources (computational physics simulations and laboratory experiments). The labeled training data consists of values of a damage index and the known locations of hidden cracks. The classification models need to learn how the presence of damage (hidden cracks) affects the damage index at different sensors for different test conditions. To this end, diagnostic FFNN models are built by sequentially adding and training new hidden layers to assimilate labeled information from computer models (different model geometries, test conditions, crack lengths, crack locations) and laboratory experiments on a plain cement slab. These transfer learning-based models are then used to localize damage in concrete specimens that reflect real-world conditions (i.e., specimens with steel reinforcement and randomly distributed aggregate). The actual damage state in these specimens is determined by extracting cores and performing petrographic studies on the extracted cores. The damage probability estimated by transfer learning-based models is compared with the petrographic damage rating index (DRI) to identify the most suitable approach to train the diagnostic models. The transfer learning-based diagnostic methodology shows promise and could be used in various structural health monitoring applications, where sufficient labeled data are typically not available from a single data source.

Miele, S.↗

Modeling of nonequilibrium space plasma flows

Godunov-type numerical solution of the 20 moment plasma transport equations. One of the centerpieces of our proposal was the development of a higher order Godunov-type numerical scheme to solve the gyration dominated 20 moment transport equations. In the first step we explored some fundamental analytic properties of the 20 moment transport equations for a low b plasma, including the eigenvectors and eigenvalues of propagating disturbances. The eigenvalues correspond to wave speeds, while the eigenvectors characterize the transported physical quantities. In this paper we also explored the physically meaningful parameter range of the normalized heat flow components. In the second step a new Godunov scheme type numerical method was developed to solve the coupled set of 20 moment transport equations for a quasineutral single-ion plasma. The numerical method and the first results were presented at several national and international meetings and a paper describing the method has been published in the Journal of Computational Physics. To our knowledge this is the first numerical method which is capable of producing stable time-dependent solutions to the full 20 (or 16) moment set of transport equations, including the full heat flow equation. Previous attempts resulted in unstable (oscillating) solutions of the heat flow equations. Our group invested over two man-years into the development and implementation of the new method. The present model solves the 20 moment transport equations for an ion species and thermal electrons in 8 domain extending from a collision dominated to a collisionless region (200 km to 12,000 km). This model has been applied to study O+ acceleration due to Joule heating in the lower ionosphere.

Gombosi, Tamas↗

The Applicability of Unit Systems to High-Performance Computing Applications

Dimensional analysis is a key technique used to verify the soundness of scientific models. Most experts agree that engineering and scientific software would be made more reliable by integrating dimensional analysis in their type system. We explored how High Performance Computing (HPC) applications could integrate compile-time dimensional analysis. We started by investigating various implementation of unit systems for C++. Eventually, selecting the latest (and most advanced) one to apply to our test codes. We worked with code of increasing complexity, from a projectile trajectory calculation to the proxy-application Lulesh. This included our code, Springs-3D, which focuses on demonstrating language features while performing simple physic computations. Finally, our main contribution is a source-code analysis which extracts constraints on the dimension of all variables, functions, and constants in an application. This resulting system of equations is solved using the dimensions of a few of these objects. This analysis has the potential to greatly reduce the time spent performing dimensional analysis when refactoring application to use a representation of units.

97 MATHEMATICS AND COMPUTING↗

SERVER DEVELOPMENT FOR NSLS-II PHYSICS APPLICATIONS AND PERFORMANCE ANALYSIS

The beam commissioning software framework of NSLS-II project adopts a client/server based architecture to replace the more traditional monolithic high level application approach. The server software under development is available via an open source sourceforge project named epics-pvdata, which consists of modules pvData, pvAccess, pvIOC, and pvService. Examples of two services that already exist in the pvService module are itemFinder, and gather. Each service uses pvData to store in-memory transient data, pvService to transfer data over the network, and pvIOC as the service engine. The performance benchmarking for pvAccess and both gather service and item finder service are presented in this paper. The performance comparison between pvAccess and Channel Access are presented also. For an ultra low emittance synchrotron radiation light source like NSLS II, the control system requirements, especially for beam control are tight. To control and manipulate the beam effectively, a use case study has been performed to satisfy the requirement and theoretical evaluation has been performed. The analysis shows that model based control is indispensable for beam commissioning and routine operation. However, there are many challenges such as how to re-use a design model for on-line model based control, and how to combine the numerical methods for modeling of a realistic lattice with the analytical techniques for analysis of its properties. To satisfy the requirements and challenges, adequate system architecture for the software framework for beam commissioning and operation is critical. The existing traditional approaches are self-consistent, and monolithic. Some of them have adopted a concept of middle layer to separate low level hardware processing from numerical algorithm computing, physics modelling, data manipulating and plotting, and error handling. However, none of the existing approaches can satisfy the requirement. A new design has been proposed by introducing service oriented architecture technology, and client interface is undergoing. The design and implementation adopted a new EPICS implementation, namely epics-pvdata [9], which is under active development. The implementation of this project under Java is close to stable, and binding to other language such as C++ and/or Python is undergoing. In this paper, we focus on the performance benchmarking and comparison for pvAccess and Channel Access, the performance evaluation for 2 services, gather and item finder respectively.

43 PARTICLE ACCELERATORS↗

Unraveling the surface chemistry processes in lithiated and boronized plasma material interfaces under extreme conditions

The review of recent theoretical and experimental research on the complex surface chemistry processes that evolve from low-Z material conditioning on plasma-facing materials under extreme fusion plasma conditions is presented. A combination of multi-scale computational physics and chemistry modeling with real-time diagnosis of the plasma-material interface in tokamak fusion plasma edge is complemented by ex-vessel in-situ single-effect experimental facilities to unravel the evolving characteristics of low-Z components under irradiation. Effects of the lithium and boron coatings at carbon surfaces to the retention of deuterium and chemical sputtering of the plasma-facing surfaces are discussed in detail. The critical role of oxygen in the surface chemistry during hydrogen-fuel irradiation is found to drive the kinetics and dynamics of these surfaces as they interact with fusion edge plasma that ultimately could have profound effects on fusion plasma confinement behavior. Computational studies also extend in spatio-temporal scales not accessible by empirical means and therefore open the opportunity for a strategic approach at irradiation surface science studies that combined these powerful computational tools with in-vessel and ex-vessel in-situ diagnostics.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Operator learning for predicting multiscale bubble growth dynamics

We report simulating and predicting multiscale problems that couple multiple physics and dynamics across many orders of spatiotemporal scales is a great challenge that has not been investigated systematically by deep neural networks (DNNs). Herein, we develop a framework based on operator regression, the so-called deep operator network (DeepONet), with the long-term objective to simplify multiscale modeling by avoiding the fragile and time-consuming “hand-shaking” interface algorithms for stitching together heterogeneous descriptions of multiscale phenomena. To this end, as a first step, we investigate if a DeepONet can learn the dynamics of different scale regimes, one at the deterministic macroscale and the other at the stochastic microscale regime with inherent thermal fluctuations. Specifically, we test the effectiveness and accuracy of the DeepONet in predicting multirate bubble growth dynamics, which is described by a Rayleigh–Plesset (R–P) equation at the macroscale and modeled as a stochastic nucleation and cavitation process at the microscale by dissipative particle dynamics (DPD). First, we generate data using the R–P equation for multirate bubble growth dynamics caused by randomly time-varying liquid pressures drawn from Gaussian random fields (GRFs). Our results show that properly trained DeepONets can accurately predict the macroscale bubble growth dynamics and can outperform long short-term memory networks. We also demonstrate that the DeepONet can extrapolate accurately outside the input distribution using only very few new measurements. Subsequently, we train the DeepONet with DPD data corresponding to stochastic bubble growth dynamics. Although the DPD data are noisy and we only collect sparse data points on the trajectories, the trained DeepONet model is able to predict accurately the mean bubble dynamics for time-varying GRF pressures. Taken together, our findings demonstrate that DeepONets can be employed to unify the macroscale and microscale models of the multirate bubble growth problem, hence providing new insight into the role of operator regression via DNNs in tackling realistic multiscale problems and in simplifying modeling with heterogeneous descriptions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Preface: Nonclassical Transport

Transport phenomena in highly heterogeneous media can be dramatically different from those in homogeneous media and therefore are of great fundamental and practical interest. Anomalous transport occurs in semiconductor physics, plasma physics, astrophysics, biology, and other areas. It plays an especially important role in hydrogeology because it may govern the rate of migration and degree of dispersion of groundwater contaminants from hazardous waste sites. The series of four articles in this special section of Vadose Zone Journal is devoted to transport phenomena in heterogeneous media in the context of geologic disposal of radioactive waste. It contains the results of joint investigations performed at the Nuclear Safety Institute of the Russian Academy of Sciences and Lawrence Berkeley National Laboratory in California. The work was supported by the U.S. DOE (under Contract No. DEAC02-05CH11231). The problems addressed in this research involve a broad range of space and time scales and were approached using modern methods of theoretical and computational physics, such as scaling analysis and diagrammatic techniques used before in critical phenomena theory. Special attention is paid to the asymptotics of concentration behavior (concentration tails). This issue is exceptionally important for the reliability assessments of radioactive waste disposal because, depending on the structure of the tails, concentrations at large distances from the source can differ by many orders of magnitude. In the first paper of this special section, Bolshov et al. (2008b) present an overview of field and laboratory observations that demonstrate nonclassical flow and transport behavior in geologic media. It is recognized that natural fracture networks as a rule have fractal geometry and can be classified as percolation systems. This is one of the main factors giving rise to anomalous transport in geologic media. Another important factor is the presence of contaminant traps provided by low-permeable rock matrix and dead-ends of fracture percolation clusters. Physical concepts to describe transport phenomena in fractured rocks are discussed. The second paper (Dykhne et al., 2008) is devoted to the analysis of diffusion in heterogeneous media with sharply contrasting properties. The authors show that as time progresses, three different transport regimes can be realized in the model. Here, an intermediate regime corresponds to subdiffusion. The change of regimes results in a complex structure of concentration tails, with the shapes of the more-distant tail segments determined by earlier-time transport behavior. In the third paper (Bolshov et al., 2008a), new elements are developed to generalize the dual-porosity model for moisture infiltration and solute transport in unsaturated rocks, taking into account fractal aspects of the percolation process. It is shown that the solute transport regime is determined by a competition of two mechanisms: random advection through a fracture network and trapping caused by sharply contrasting properties of the medium. As a result, superdiffusive, subdiffusive, or classical diffusive regimes may occur. The complex structure of concentration tails and effects due to medium characteristic fluctuations is also discussed. In the fourth paper, Goloviznin et al. (2008) develop a stochastic random walk numerical model of anomalous diffusion to simulate solute transport in highly heterogeneous media. Solutions of the one- and symmetric two-dimensional stochastic problem are compared with computations performed on the basis of fractional advection--diffusion equation models. The new model is in reasonable agreement with experimental data on solute transport in highly heterogeneous media.

54 ENVIRONMENTAL SCIENCES↗