Search NASA⌕ Search

Engineering topics

Germann, Timothy Clark

Publications and source records attributed to Germann, Timothy Clark.

At least 19 records

Hydrogen Diffusion in Slit Pores: Role of Temperature, Pressure, Confinement, and Roughness

Diffusion of hydrogen (H 2 ) is important to understand the leakage risk and transport behavior for H 2 geologic storage. We applied molecular dynamics simulations to investigate the influencing factors of H2 diffusion in the slit pores of calcite, hematite, and quartz, owing to their abundance. It is revealed that the H2 self-diffusion coefficient increases with the temperature, regardless of the type of pore minerals. The diffusion of H 2 in the 20 nm slit pores falls into the bulk diffusion regime when the pressure is 10 MPa. The self-diffusion of H 2 decreases with pressure in all three types of slit pores, following a power law model with the exponents ranging from -0.825 to -0.964. Furthermore, the impact of confinement on H 2 diffusion is more pronounced for the slit pores with stronger interactions with H 2 -like calcite. The role of surface roughness in H 2 diffusion depends on the slit aperture. The rough surface enhances H 2 diffusion in the larger slit pores due to the enlarged effective pore space, whereas it weakens H 2 diffusion in the small slit pores due to stronger adsorption. These findings will fill the knowledge gap on the coupling effect of different factors influencing H 2 diffusion.

08 HYDROGEN↗

High Pressure Suppression of Plasticity in RDX [Slides]

RDX shear banding nucleates without other defects present and is important for both mechanical properties and for mechano-chemical effects. Above ~1.0 km/s particle velocity, the shear band mechanism ceases to occur. It is not replaced by a phase transformation or other plastic event. Shear band initial sites over-nucleate and lower the VM tress drastically. This results in a fully reversible, high pressure mechanism that suppresses the formation of shear bands.

36 MATERIALS SCIENCE↗

Interplay of Mechanochemistry and Material Processes in the Graphite to Diamond Phase Transformation

The manifestation of intramolecular strains in covalent systems is widely known to accelerate chemical reactions and open alternative reaction paths. This process is moderately well understood for isolated molecules and unimolecular processes. However, in condensed matter processes such as phase transformations, material properties and structure may influence typical mechanochemical effects. Therefore, we utilize steered molecular dynamics to induce out of plane strains in graphite and compress the system under a constant strain rate to induce phase transformation. We show that the out of plane strain allows phase transformations to initiate at small amounts of compressive strain. Yet, in contrast to typical mechanochemical results, the sum of compressive and out of plane work needed to form a diamond has a local minimum due to altered defect formation processes during phase transformation. Additionally, these altered processes slow the kinetics of the phase transformation, taking longer from initiation to total material transformation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Understanding racial and ethnic disparities in COVID-19 outcomes using an agent-based simulation [Poster]

Infectious disease outbreaks pose a major threat to public health, economic stability, and human security. In order to mitigate disease impacts, we need to understand drivers that contribute to its spread. Latinx, Black, and American Indian racial and ethnic groups experienced disproportionate health outcomes during the COVID-19 pandemic. Current modeling approaches often fail to capture the high level of heterogeneity present in the population needed to measure the drivers of these disparities.

97 MATHEMATICS AND COMPUTING↗

Using limited neural networks to assess relative mechanistic influence on shock heating in granular solids

The rapid compaction of granular media results in localized heating that can induce chemical reactions, phase transformations, and melting. However, there are numerous mechanisms in play that can be dependent on a variety of microstructural features. Machine learning techniques such as neural networks offer a ubiquitous method to develop models for physical processes. Limiting what kind of microstructural information is used as an input and assessing normalized changes in network error, the relative importance of different mechanisms can be inferred. Here we utilize binned, initial density information as network inputs to predict local shock heating in a granular high explosive trained from large-scale molecular dynamics simulations. In this study, the spatial extent of the density field used in the network is altered to assess the importance and relevant length scales of the physical mechanisms in play, where different microstructural features result in different predictive capabilities.

36 MATERIALS SCIENCE↗

Influence of Pore Surface Structure and Contents on Shock-Induced Collapse and Energy Localization

The majority of computational efforts in the shockwave driven collapse of porosity focus on idealized pore geometries and structures. These give significant insight into the physical mechanisms in play, but omit some potentially important effects of real energetic materials. Here, to address this, we simulate pore collapse in 1,3,5-triamino-2,4,6-trinitrobenzene (TATB) for four types of pore surfaces and fillings: an idealized cylindrical pore, a pore surface coated with polymer, a melted/amorphous pore surface, and a pore filled with gas molecules. We find that these perturbations to the pore structure result in only minor changes to the highly complex collapse mechanisms. However, significant differences in hotspot temperature are observed, with peak temperatures ranging over several hundred Kelvin. The polymer-coated system has the most effect, lowering the hotspot temperature by 300 K, and also limiting the manifestation of intramolecular strain energy that is known to induce mechanochemistry.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

t22_person_interact viewgraph

This document contains a viewgraph from data generated for the t22_person_interact project. Calculations were performed using Intuitional Computing resources.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Molecular Dynamics 1

Explore the source record for details and available documents.

Germann, Timothy Clark↗

Energy localization efficiency in 1,3,5-trinitro-2,4,6-triaminobenzene pore collapse mechanisms

Atomistic and continuum scale modeling efforts have shown that the shock-induced collapse of porosity can occur via a wide range of mechanisms dependent on pore morphology, the shockwave pressure, and material properties. The mechanisms that occur under weaker shocks tend to be more efficient at localizing thermal energy but do not result in high, absolute temperatures or spatially large localizations compared to mechanisms found under strong shock conditions. However, the energetic material 1,3,5-trinitro-2,4,6-triaminobenzene (TATB) undergoes a wide range of collapse mechanisms that are not typical of similar materials, leaving the collapse mechanisms and the resultant energy localization from the collapse, i.e., hotspots, relatively uncharacterized. Therefore, we present the pore collapse simulations of cylindrical pores in TATB for a wide range of pore sizes and shock strengths that trigger viscoplastic collapses that occur almost entirely perpendicular to the shock direction for weak shocks and hydrodynamic-like collapses for strong shocks that do not break the strong hydrogen bonds of the TATB basal planes. The resulting hotspot temperature fields from these mechanisms follow trends that differ considerably from other energetic materials; hence, we compare them under normalized temperature values to assess the relative efficiency of each mechanism to localize energy. The local intra-molecular strain energy of the hotspots is also assessed to better understand the physical mechanisms behind the phenomena that lead to a latent potential energy.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

GLUE Code: A framework handling communication and interfaces between scales

Many scientific applications are inherently multiscale in nature. Such complex physical phenomena often require simultaneous execution and coordination of simulations spanning multiple time and length scales. This is possible by combining expensive small-scale simulations (such as molecular dynamics simulations) with larger scale simulations (such continuum limit/hydro solvers) to allow for considerably larger systems using task and data parallelism. However, the granularity of the tasks can be very large and often leads to load imbalance. Traditionally, we use approximations to streamline the computation of the more costly interactions and this introduces trade-offs between simulation cost and accuracy. In recent years, the available computational power and the advances in machine learning have made computing these scale-bridging interactions and multiscale simulations more feasible. One driving application has been plasma modeling in inertial confinement fusion (ICF), which is fundamentally multiscale in nature. This requires deep understanding of how to extrapolate microscopic information into macroscopically relevant scales. For example, in ICF one needs an accurate understanding of the connection between experimental observables and the underlying microphysics. The properties of the larger scales are often affected by the microscale behavior incorporated usually into the equations of state and ionic and electronic transport coefficients (Liboff, 1959; Rinderknecht et al., 2014; Rosenberg et al., 2015; Ross et al., 2017). Instead of incorporating this information using reliable molecular dynamics (MD) simulations, one often needs to use theoretical models, due to the inability of MD to reach engineering scales (Glosli et al., 2007; Marinak et al., 1998). One approach to resolve this issue is by coupling two MD simulations of different scales via force interpolation, e.g., the AdResS method (Krekeler et al., 2018; Nagarajan et al., 2013). Another approach, which we will pursue in the scope of this work, is by enabling scale bridging between MD simulations and meso/macro-scale models through the development and support of application programming interfaces that these different applications can interact through.

54 ENVIRONMENTAL SCIENCES↗

Enabling Predictive Scale-Bridging Simulations through Active Learning (Institutional Computing Annual Report (Project w21_alscalebridging)) [Slides]

The goal of this project was to develop, demonstrate, and provide a new capability to achieve greater physical fidelity in large-scale simulations, rather than the usual brute-force increases in the number of mesh elements or particles. Transport in nanoporous media, critical to hydrocarbon extraction from tight shale formations, is affected by molecular-level interactions. Several coarse-scale Lattice Boltzmann model (LBM) parameters cannot be directly computed, so instead we calibrate them to molecular dynamics (MD) simulations, by building emulators that mimic MD and LBM behavior and then training an upscaler. The resulting machine learning (ML)-based scale-bridging framework is up to 7 orders of magnitude faster than direct MD.

74 ATOMIC AND MOLECULAR PHYSICS↗

Enabling Predictive Scale-Bridging Simulations through Active Learning (Institutional Computing Annual Report (Project w20_alscalebridging)) [Slides]

The goal of this project was to develop, demonstrate, and provide a new capability to achieve greater physical fidelity in large-scale simulations, rather than the usual brute-force increases in the number of mesh elements or particles. This was done by using machine learning (ML) techniques to develop emulators for subscale physics that can be used in coarse-scale continuum simulations, trained on fine-scale molecular dynamics (MD) simulations that are launched on-the-fly via active learning. raditional inertial confinement fusion (ICF) simulations rely on numerical diffusion to simulate molecular effects such as non-local transport and mixing without truly accounting for molecular interactions; our approach directly accounts for this physics.

74 ATOMIC AND MOLECULAR PHYSICS↗

Machine learning of consistent thermodynamic models using automatic differentiation

In this study, we propose a data-driven method to describe consistent equations of state (EOS) for arbitrary systems. Complex EOS are traditionally obtained by fitting suitable analytical expressions to thermophysical data. A key aspect of EOS is that the relationships between state variables are given by derivatives of the system free energy. In this work, we model the free energy with an artificial neural network and utilize automatic differentiation to directly learn the derivatives of the free energy. We demonstrate this approach on two different systems, the analytic van der Waals EOS and published data for the Lennard-Jones fluid, and we show that it is advantageous over direct learning of thermodynamic properties (i.e., not as derivatives of the free energy but as independent properties), in terms of both accuracy and the exact preservation of the Maxwell relations. Furthermore, the method implicitly provides the free energy of a system without explicit integration.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

A physics-informed and hierarchically regularized data-driven model for predicting fluid flow through porous media

This paper presents a new deep learning data-driven model for predicting structure dependent pore-fluid velocity fields in rock. The model is based on a Convolutional Auto-Encoder (CAE) artificial neural network capable of learning from image data generated by direct numerical simulations of fluid flow through pore-structures, such as by Lattice Boltzmann or molecular dynamics methods. The main novelty of the model in comparison to previous CAE-based data-driven approaches consists of three parts. The first is a methodology for decomposing the full-domain of the porous media into sub-regions, or “sub-domains”, in order to reduce the overall size of the CAE, batch process the sub-domains in parallel, and enable the CAE to learn local and generalizable nonlinear mappings of pore-fluid velocities. The second consists of embedding the finite difference solutions of the incompressible Navier-Stokes and continuity equations into convolutional layers prior to the CAE in order to provide the CAE with knowledge of fluid dynamics physics (PhyFlow). The third main novelty is that the training of the CAE is regularized with a hierarchical loss function that encourages the learning of fluid flow patterns (in a way similar to ranked modes in principal component analysis), ranking from most to least important. This is shown to increase the stability in learning, reduce over-fitting, and promote interpretability of the CAE neural network layers (HierCAE). The comprehensive new data-driven model, which we call the PhyFlow-HierCAE model, is shown to exhibit improved accuracy and generalizability of flow field predictions over conventional CAE models, attributable to the embedded physical knowledge and the hierarchical regularization, as well as realize orders of magnitude speed-ups in computation times as a surrogate for the direct numerical simulations. Examples of training and forward predictions on unseen pore-structures are provided and evaluated for data from Lattice Boltzmann and molecular dynamics simulations of pore-fluid flow. The model is shown to be a fast and accurate emulator (or “surrogate”) for predicting effective permeability of unseen pore-structures based on learning from relatively small direct numerical simulation datasets.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

EpiCast: Simulating Epidemics with Extreme Detail

In early 2020, COVID-19 swept the globe. Governments attempted to “flatten the curve” through business shutdowns and stay-at-home orders, but the United States was hit hard. By the end of March, mere months after the virus first emerged in humans 7,000 miles away, the U.S. had recorded 192,300 cases and 5,300 deaths. While this unprecedented disaster sent shockwaves through every level of society and clouded an uncertain future, state and local governments turned to computational and mathematical epidemiology researchers to help formulate intervention strategies to limit the spread of the disease. Traditional forecasting models provided a reasonable understanding of how the near future was likely to look, but local policy makers and public health communities still struggled to understand how potential mitigations ought to be implemented. Decision makers needed a way to measure the impact of their policy choices—they needed better technology. EpiCast answered the call, bringing urgently needed answers to policymakers grappling with how to adjust school and business schedules. EpiCast is modeling software that generates a synthetic, representative population to simulate infectious disease transmission in the United States with extreme detail and granularity. The software models human behavior combined with community-specific information to provide a fine-grained preview of the effect of potential mitigation strategies for decision makers.

60 APPLIED LIFE SCIENCES↗

Automated discovery of a robust interatomic potential for aluminum

Machine learning, trained on quantum mechanics (QM) calculations, is a powerful tool for modeling potential energy surfaces. A critical factor is the quality and diversity of the training dataset. Here we present a highly automated approach to dataset construction and demonstrate the method by building a potential for elemental aluminum (ANI-Al). In our active learning scheme, the ML potential under development is used to drive non-equilibrium molecular dynamics simulations with time-varying applied temperatures. Whenever a configuration is reached for which the ML uncertainty is large, new QM data is collected. The ML model is periodically retrained on all available QM data. The final ANI-Al potential makes very accurate predictions of radial distribution function in melt, liquid-solid coexistence curve, and crystal properties such as defect energies and barriers. We perform a 1.3M atom shock simulation and show that ANI-Al force predictions shine in their agreement with new reference DFT calculations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗