Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “burgers”

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 145 records · Page 8
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Non-intrusive reduced-order modeling for dynamical systems with spatially localized features

This work presents a non-intrusive reduced-order modeling framework for dynamical systems with spatially localized features characterized by slow singular value decay. The proposed approach builds upon two existing methodologies for reduced and full-order non-intrusive modeling, namely Operator Inference (OpInf) and sparse Full-Order Model (sFOM) inference. We decompose the domain into two complementary subdomains that exhibit fast and slow singular value decay. The dynamics of the subdomain exhibiting slow singular value decay are learned with sFOM while the dynamics with intrinsically low dimensionality on the complementary subdomain are learned with OpInf. The resulting, coupled OpInf-sFOM formulation leverages the computational efficiency of OpInf and the high resolution of sFOM, and thus enables fast non-intrusive predictions for conditions beyond those sampled in the training data set. A novel regularization technique with a closed-form solution based on the Gershgorin disk theorem is introduced to promote stable sFOM and OpInf models. We also provide a data-driven indicator for subdomain selection and ensure solution smoothness over the interface via a post-processing interpolation step. We evaluate the efficiency of the approach in terms of offline and online speedup through a quantitative, parametric computational cost analysis. We demonstrate the coupled OpInf-sFOM formulation for two test cases: a one-dimensional Burgers’ model for which accurate predictions beyond the span of the training snapshots are presented, and a two-dimensional parametric model for the Pine Island Glacier ice thickness dynamics, for which the OpInf-sFOM model achieves an average prediction error on the order of 1% with an online speedup factor of approximately 8$\times$ compared to the numerical simulation.

42 ENGINEERING↗
DOE OSTI2025

Optimizing the optimizer for physics-informed neural networks and Kolmogorov-Arnold networks

Physics-Informed Neural Networks (PINNs) have revolutionized the computation of PDE solutions by integrating partial differential equations (PDEs) into the neural network’s training process as soft constraints, becoming an important component of the scientific machine learning (SciML) ecosystem. More recently, physics-informed Kolmogorv-Arnold networks (PIKANs) have also shown to be effective and comparable in accuracy with PINNs. In their current implementation, both PINNs and PIKANs are mainly optimized using first-order methods like Adam, as well as quasi-Newton methods such as BFGS and its low-memory variant, L-BFGS. However, these optimizers often struggle with highly nonlinear and non-convex loss landscapes, leading to challenges such as slow convergence, local minima entrapment, and (non)degenerate saddle points. In this study, we investigate the performance of Self- Scaled BFGS (SSBFGS), Self-Scaled Broyden (SSBroyden) methods and other advanced quasi-Newton schemes, including BFGS and L-BFGS with different line search strategies. These methods dynamically rescale updates based on historical gradient information, thus enhancing training efficiency and accuracy. We systematically compare these optimizers – using both PINNs and PIKANs – on key challenging PDEs, including the Burgers, Allen-Cahn, Kuramoto-Sivashinsky, Ginzburg-Landau, and Stokes equations. Additionally, we evaluate the performance of SSBFGS and SSBroyden for Deep Operator Network (DeepONet) architectures, demonstrating their effectiveness for data-driven operator learning. Our findings provide state-of-the-art results with orders-of-magnitude accuracy improvements without the use of adaptive weights or any other enhancements typically employed in PINNs. More broadly, our work reveal insights into the effectiveness of quasi-Newton optimization strategies in significantly improving the convergence and accurate generalization of PINNs and PIKANs.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Learning subgrid-scale models with neural ordinary differential equations

We propose a new approach to learning the subgrid-scale model when simulating partial differential equations (PDEs) solved by the method of lines and their representation in chaotic ordinary differential equations, based on neural ordinary differential equations (NODEs). Solving systems with fine temporal and spatial grid scales is an ongoing computational challenge, and closure models are generally difficult to tune. Machine learning approaches have increased the accuracy and efficiency of computational fluid dynamics solvers. In this approach neural networks are used to learn the coarse- to fine-grid map, which can be viewed as subgrid-scale parameterization. We propose a strategy that uses the NODE and partial knowledge to learn the source dynamics at a continuous level. Our method inherits the advantages of NODEs and can be used to parameterize subgrid scales, approximate coupling operators, and improve the efficiency of low-order solvers. Numerical results with the two-scale Lorenz 96 ODE, the convection-diffusion PDE, and the viscous Burgers' PDE are used to illustrate this approach.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

ECCI analysis of shear accommodations at grain boundaries in commercially pure alpha titanium

This study uses a novel approach that couples electron channeling contrast imaging (ECCI) with controlled removal of surface material to investigate the slip system/grain boundary interactions by assessing the geometry of slip planes and grain boundaries as a function of depth. The comparison of spacing, relative distributions, and correlations of slip bands across grain boundaries reveals surface biasing on the slip band behavior in the grain boundary region. Furthermore, using analysis of the compatibility factor m’, the Schmid factor of each slip system, and the angles θ between slip planes, a more comprehensive understanding of the details of the mechanisms of strain transfer and heterogeneous deformation is developed. In general, the misalignment of primary slip systems on the grain boundary planes necessitates the activation of secondary accommodating slip systems close to the grain boundary area that are better aligned (higher m’ and lower θ) with respect to the incoming slip system across the grain boundary. Such systems have the same Burgers vector as the primary slip systems within the respective grains, but are typically on low mobility slip planes.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

A data driven approach for cross-slip modelling in continuum dislocation dynamics

Cross-slip is a thermally activated process by which screw dislocation changes its glide plane to another slip plane sharing the same Burgers vector. The rate at which this process happens is determined by a Boltzmann type expression that is a function of the screw segment length and the stress acting on the dislocation. In continuum dislocation dynamics (CDD), the information regarding the length of the screw dislocation segment and local stress state on dislocations are lost due to the coarse-grained representation of the density. Here, in this work, a data driven approach to characterize the lost information by analyzing the discrete dislocation configurations is proposed to enable cross-slip modeling in the CDD framework in terms of the coarse-grained dislocation density and stress fields. The analysis showed that the screw segment length follows an exponential distribution, and the stress fluctuations, defined as the difference between the stress on the dislocations and the mean field stress in CDD, follows a Lorentzian distribution. A novel approach for cross slip implementation in CDD employing the screw segment length and stress fluctuation statistics was proposed and rigorously tested by comparing the CDD cross-slip rates with discrete dislocation dynamics (DDD) rates. This approach has been applied in conjunction with three cross-slip models used in DDD simulations differing mainly in the functional form of cross slip activation energy. It was found that different cross-slip activation energy formulations yielded different cross-slip rates, yet the effect on mechanical stress-strain response and dislocation density evolution was minimal for the [001] type loading.

42 ENGINEERING↗
DOE OSTI2020

Adaptive Activation Functions Accelerate Convergence in Deep and Physics-informed Neural Networks

We employ adaptive activation functions for regression in deep and physics-informed neural networks (PINNs) to approximate smooth and discontinuous functions as well as solutions of linear and nonlinear partial differential equations. In particular, we solve the nonlinear Klein-Gordon equation, which has smooth solutions, the nonlinear Burgers equation, which can admit high gradient solutions, and the Helmholtz equation. We introduce a scalable hyper-parameter in the activation function, which can be optimized to achieve best performance of the network as it changes dynamically the topology of the loss function involved in the optimization process. The adaptive activation function has better learning capabilities than the traditional one (fixed activation) as it improves greatly the convergence rate, especially at early training, as well as the solution accuracy. To better understand the learning process, we plot the neural network solution in the frequency domain to examine how the network captures successively different frequency bands present in the solution. We consider both forward problems, where the approximate solutions are obtained, as well as inverse problems, where parameters involved in the governing equation are identified. Our simulation results show that the proposed method is a very simple and effective approach to increase the efficiency, robustness and accuracy of the neural network approximation of nonlinear functions as well as solutions of partial differential equations, especially for forward problems. We theoretically prove that in the proposed method, gradient descent algorithms are not attracted to suboptimal critical points or local minima.

machine leaning, Bad minima, Inverse problems, Phy↗
DOE OSTI2020

Implicit fast sweeping method for hyperbolic systems of conservation laws

Implicit time-accurate methods are often used to integrate stiff problems where explicit schemes impose severe time step restrictions. This paper presents an efficient numerical framework based on the Fast Sweeping Method (FSM) for solving linear and nonlinear hyperbolic systems of conservation laws. The solution at each discrete location is computed by sweeping the numerical domain in several predetermined directions that follow the causality of the characteristic families. The use of a fractional step strategy eliminates the need for a solution selection criterion while one-sided stencils limit the number of sweeps to at most 2 d for d space dimensions. This work focuses on the first-order implicit upwind method since it constitutes the building block for high-order conservative schemes. For problems where the degree of stiffness evolves over time, implicit-explicit hybridization can be accomplished with the same algorithm by simply switching the stencil at each time level. As opposed to traditional implicit solvers, the sweeping method does not require a local time linearization of the fluxes thereby preserving the nonlinear stability properties of the original implicit scheme. It also avoids the large computational and memory requirements associated with solving large block-diagonal systems of equations. Here, a series of one- and two-dimensional test cases are presented for the inviscid Burgers' equation and the reactive Euler equations. The results indicate that the implicit FSM can allow a major reduction in the number of time steps even in the presence of discontinuous solution profiles.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2023

gLaSDI: Parametric physics-informed greedy latent space dynamics identification

A parametric adaptive physics-informed greedy Latent Space Dynamics Identification (gLaSDI) method is proposed for accurate, efficient, and robust data-driven reduced-order modeling of high-dimensional nonlinear dynamical systems. In the proposed gLaSDI framework, an autoencoder discovers intrinsic nonlinear latent representations of high-dimensional data, while dynamics identification (DI) models capture local latent-space dynamics. Here, an interactive training algorithm is adopted for the autoencoder and local DI models, which enables identification of simple latent-space dynamics and enhances accuracy and efficiency of data-driven reduced-order modeling. To maximize and accelerate the exploration of the parameter space for the optimal model performance, an adaptive greedy sampling algorithm integrated with a physics-informed residual-based error indicator and random-subset evaluation is introduced to search for the optimal training samples on the fly. Further, to exploit local latent-space dynamics captured by the local DI models for an improved modeling accuracy with a minimum number of local DI models in the parameter space, a -nearest neighbor convex interpolation scheme is employed. The effectiveness of the proposed framework is demonstrated by modeling various nonlinear dynamical problems, including Burgers equations, nonlinear heat conduction, and radial advection. The proposed adaptive greedy sampling outperforms the conventional predefined uniform sampling in terms of accuracy. Compared with the high-fidelity models, gLaSDI achieves 17 to 2,658× speed-up with 1 to 5% relative errors.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

ECLEIRS: Exact conservation law embedded identification of reduced states for parameterized nonlinear conservation laws from sparse and noisy data

Multi-query applications such as parameter estimation, uncertainty quantification and design optimization for parameterized partial differential equation (PDE) systems are expensive. While reduced/latent state dynamics approaches for parameterized PDEs offer a viable alternative, these approaches rely on high-quality data and struggle with highly sparse spatiotemporal noisy measurements typically obtained from experiments. Furthermore, there is no guarantee that these models satisfy governing physical conservation laws. In this article, we propose a reduced state dynamics approach, referred to as ECLEIRS, that embeds exact conservation in the solution and flux representation by utilizing a space-time divergence-free neural network formulation. We compare ECLEIRS with other reduced state dynamics approaches, those that do not enforce any physical constraints and those with physics-informed loss functions, for three shock-propagation problems: 1-D advection, 1-D Burgers and 2-D Euler equations. In conclusion, the numerical experiments conducted in this study demonstrate that ECLEIRS provides the most accurate prediction of dynamics for unseen parameters even in the presence of highly sparse and noisy data.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

X-ray topography characterization of gallium nitride substrates for power device development

Gallium nitride substrates grown by the hydride vapor phase epitaxy (HVPE) method using a patterned growth process have been characterized by synchrotron monochromatic beam X-ray topography in the grazing incidence geometry. Images reveal a starkly heterogeneous distribution of dislocations with areas as large as 0 3 mm 2 containing threading dislocation densities below 10 3 cm -2 in between a grid of strain centers with higher threading dislocation densities (> 10 4 cm -2 ). Basal plane dislocation densities in these areas are as low as 104 cm -2 . By comparing the recorded images of dislocations with ray tracing simulations of expected dislocations in GaN, the Burgers vectors of the dislocations have been determined. The distribution of threading screw/mixed dislocations (TSDs/TMDs), threading edge dislocations (TEDs) and basal plane dislocations (BPDs) is discussed with implications for fabrication of power devices.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Disclination-dislocation based model for grain boundary stress field evolution due to slip transmission history and influence on subsequent dislocation transmission

This work demonstrates how the structure of a grain boundary (GB) and its evolution due to slip transmission history influences subsequent dislocation transmission. First, a model for the evolution of stress fields within grain boundaries that accounts for the effects of coherent dislocation transmission is introduced. Starting with a disclination-based construct of GBs at minimum energy (equilibrium), the model describes the evolution of the GB stress field to a state characteristic of excess energy (non-equilibrium) due to the incorporation of residual Burgers vector content following sequential slip transmission events. Several essential features of this model are verified via molecular dynamics simulations of lattice dislocation absorption. Second, this model is implemented into a discrete dislocation dynamics (DDD) code and simulations are performed to understand the influence of Non-equilibrium GB stress fields, conditioned by the slip transmission history, on subsequent dislocation transmission. DDD simulations reveal that the critical resolved stress necessary for slip propagation can be reduced by with continued absorption of residual dislocation content. Moreover, DDD simulations prove that a comprehensive consideration of both the binding and driving stresses, and the evolution of the transmission configuration is necessary to quantify the influence of the mechanical state of the GB on slip propagation. In general, this work provides important insights into the role of GB structure evolution, conditioned by prior deformation history, on intergranular plasticity.

42 ENGINEERING↗
DOE OSTI2024

Generalized grain boundary constitutive description implemented in a strain-gradient large-strain FFT-based formulation: Application to nano-metallic laminates

Herein this paper presents a general treatment of grain boundary constitutive behavior in the context of strain-gradient (SG) plasticity, and its numerical implementation in a large-strain (LS) elasto-viscoplastic (EVP) fast Fourier transform (FFT)-based micromechanical model. Two novel grain boundary constitutive equations are proposed, allowing for more accurate description of the Burgers vector flow at the grain boundary. The capabilities of the generalized SG-LS-EVPFFT formulation are illustrated for the case of kink-band formation during layer-parallel compression of nano-metallic laminates (NMLs), requiring consideration of the interaction between dislocations and interfaces.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Revealing the synergistic effects of sequential and simultaneous dual beam irradiations in tungsten via in-situ TEM

Elucidating the synergistic effects of different energetic beams on the radiation response of nuclear materials is critical for developing an improved methodology for their evaluation when exposed to extreme environments. This article describes in-situ sequential (He implantation followed by Kr irradiation and vice-versa) and simultaneous (heavy ion Kr irradiation and He implantation) dual beam irradiations performed on tungsten at 1223 K. Dislocation loop density, average area, and total loop damage as a function of irradiation history and dose/fluence are quantified. The loop Burgers vectors and cavity damage (cavity density, size and total change in volume) are also determined at the final maximum dose for each condition. The loop damage evolution was different in all cases, with the smallest loop damage observed in the simultaneous experiment. Annihilation of Kr generated dislocation loops during He implantation in the Kr + He experiment was an unanticipated observation that may be explained by the dynamic evolution of dislocation loop sink strengths and time-dependent defect fluxes. Dislocation loop raft formation, denuded zones near extended defects, and cavity damage are compared across the different conditions. The phenomena observed and discussed in this work will stimulate further experimental and computational modeling activities leading to improved fundamental understanding of the irradiation response of nuclear materials under reactor-similar environments.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Dislocation loop evolution in F/M steel T91 under in-situ ion irradiation: Influence of the presence of initial dislocations

Ferritic/Martensitic (F/M) steel T91 was irradiated in-situ to 4 and 10 dpa at 470 °C using 1 MeV Kr 2+ . Additionally, the microstructure evolution under irradiation was followed and characterized by in-situ Transmission Electron Microscopy (TEM), in terms of dislocation loop Burgers vector, average size and density as a function of dose. Furthermore, the influence of the presence of dislocations in the pre-irradiated matrix was investigated by comparing the radiation induced defect evolution in grains with and without initial dislocations.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Effect of Confinement of SIA Cluster Diffusion by Impurities on Radiation Defect Accumulation due to 14MeV neutrons in Tungsten

Impurities acting as traps, reduce the self-interstitial atom (SIA) cluster mobility, thereby increasing recombination and in turn enhancing radiation-resistance of polycrystalline materials. However, at elevated temperatures, depending on SIA-trap (impurity) binding energy, SIA clusters may trap and detrap (detach) numerous times. In tungsten, SIA clusters of all sizes glide one-dimensionally (1D) along their Burgers vector direction. Hence, detrapped SIAs would retrace their path, confining their 1D-glide between traps. However, small SIA clusters can change the direction of their 1D-glide by overcoming a rotation barrier. In addition, impurities or solutes are known to reduce this barrier. A lower rotation barrier can result in detrapped SIA clusters that diffuse in a random 1D direction, effectively leading to 3D (net-3D) diffusion. Here we investigate the effect of two cases of SIA diffusion, namely confined-1D diffusion and net-3D diffusion, on the damage accumulation in polycrystalline tungsten under 14 MeV neutron irradiation using the object kinetic Monte Carlo method. Simulations were performed using cascade debris obtained from molecular dynamics simulations with primary knock-on atom (PKA) energies corresponding to 14 MeV neutrons. In the simulations with net-3D diffusion, SIA clusters of all sizes diffuse in a random 1D direction after the clusters detrap. While for the confined 1D diffusion case, SIA clusters larger than size 5 retain their original direction of 1D-glide. In both types of simulations, trapped vacancies are assumed to be permanently immobilized. Unexpectedly, the damage accumulation was lower with confined 1D diffusion than that with net-3D. We present a systematic comparison of the influence of the two cases of SIA diffusion on the radiation damage accumulation as a function of dose rate, detrapping barrier, and impurity concentration at 1025 K.

Accumulation and recovery of lattice damage, tungs↗
DOE OSTI2026

In situ study on radiation response of tungsten manufactured by laser powder bed fusion

Tungsten (W) produced by laser powder bed fusion (LPBF) was examined by in situ Krypton (Kr) ion irradiation at 400 °C up to 2.52 displacements per atom (dpa) to investigate its radiation response. Dislocation loops with identical Burgers vectors form aligned raft structures, inducing significant grain misorientation accumulation. Defect saturation was observed beyond 0.36 dpa, marked by constant loop density and raft spacing. WO 3 nanoparticles are found in the as-printed matrix and served as efficient defect sinks. Dislocation loops were absorbed at the W/WO 3 interface, facilitating defect annihilation and suppressing defect accumulation. In conclusion, these findings highlight the role of LPBF microstructure and oxide interfaces in mediating radiation-induced defect evolution, offering insights for designing radiation tolerant W-based materials.

Defect sink↗
DOE OSTI2023

Bicrystallography-informed Frenkel–Kontorova model for interlayer dislocations in strained 2D heterostructures

In recent years, van der Waals (vdW) heterostructures and homostructures, which consist of stacks of two-dimensional (2D) materials, have risen to prominence due to their association with exotic quantum phenomena originating from correlated electronic states harbored by them. Atomistic scale relaxation effects play an extremely important role in the electronic scale quantum physics of these systems, providing means of manipulation of these materials and allowing them to be tailored for emergent technologies. We investigate such structural relaxation effects in this work using atomistic and mesoscale models, within the context of twisted bilayer graphene — a well-known heterostructure system that features moiré patterns arising from the lattices of the two graphene layers. For small twist angles, atomic relaxation effects in this system are associated with the natural emergence of interface dislocations or strain solitons, which result from the cyclic nature of the generalized stacking fault energy (GSFE), that measures the interface energy based on the relative movement of the two layers. Here, in this work, we first demonstrate using atomistic simulations that atomic reconstruction in bilayer graphene under a large twist also results from interface dislocations, although the Burgers vectors of such dislocations are considerably smaller than those observed in small-twist systems. To reveal the translational invariance of the heterointerface responsible for the formation of such dislocations, we derive the translational symmetry of the GSFE of a 2D heterostructure using the notions of coincident site lattices (CSLs) and displacement shift complete lattices (DSCLs). The workhorse for this exercise is a recently developed Smith normal form bicrystallography framework. Next, we construct a bicrystallography-informed and frame-invariant Frenkel–Kontorova model, which can predict the formation of strain solitons in arbitrary 2D heterostructures, and apply it to study a heterostrained, large-twist bilayer graphene system. Our mesoscale model is found to produce results consistent with atomistic simulations. We anticipate that the model will be invaluable in predicting structural relaxation and for providing insights into various heterostructure systems, especially in cases where the fundamental unit cell is large and therefore, atomistic simulations are computationally expensive.

2D heterostructures↗