SEARCH · Engineering Papers

Results for “gradient methods”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 937 records · Page 52

All sources NASA NTRS DOE OSTI

DOE OSTI2023

Stabilized lithium metal oxide electrode material and method of preparation

A stabilized lithium metal oxide cathode material comprises microparticles of lithium metal oxide in which individual particles thereof a core of lithium metal oxide and a coating of a different lithium metal oxide surrounding the core. There is an interface layer between the cores and the coatings in which there are gradients of metal ions in the direction of coating to core. The materials are made by a three stage process involving coprecipitating precursor metal hydroxide core particles at a controlled pH; coprecipitating a different metal hydroxide coating on the particles without controlling the pH; and then calcining the resulting coated precursor particles with lithium hydroxide to form the stabilized lithium metal oxide material.

Lipson, Albert L.↗

DOE OSTI2024

Stabilized lithium metal oxide electrode material and method of preparation

Lipson, Albert L.↗

DOE OSTI2019

A space-time adaptive finite element method with exponential time integrator for the phase field model of pitting corrosion

In this paper we propose a space-time adaptive finite element method for the phase field model of pitting corrosion, which is a parabolic partial differential equation system consisting of a phase variable and a concentration variable. A major challenge in solving this phase field model is that the problem is very stiff, which makes the time step size extremely small for standard temporal discretizations. Another difficulty is that a high spatial resolution is required to capture the steep gradients within the diffused interface, which results in very large number of degrees of freedom for uniform meshes. To overcome the stiffness of this model, we combine the Rosenbrock–Euler exponential integrator with Crank–Nicolson scheme for the temporal discretization. Moreover, by exploiting the fact that the speed of the corroding interface decreases with time, we derive an adaptive time stepping formula. For the spatial approximation, we propose a simple and efficient strategy to generate adaptive meshes that reduces the computational cost significantly. Thus, the proposed method utilizes local adaptivity and mesh refinement for efficient simulation of the corrosive dissolution over long times in heterogeneous media with complex microstructures. We also present an extensive set of numerical experiments in both two and three dimensional spaces to demonstrate efficiency and robustness of the proposed method.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2022

Nonadiabatic Molecular Dynamics by Multiconfiguration Pair-Density Functional Theory

Herein, we present the first implementation of multiconfiguration pair-density functional theory (MC-PDFT) ab initio molecular dynamics. MC-PDFT is a multireference electronic structure method that in many cases has a similar accuracy (or even better accuracy) the complete active space second-order perturbation theory (CASPT2) at a significantly lower computational cost. In this study, we introduced MC-PDFT analytical gradients into the SHARC molecular dynamics program for ab initio, nonadiabatic molecular dynamics simulations. We verify our implementation by examining the intersystem crossing dynamics of thioformaldehyde, and we observe excellent agreement with recent CASPT2 and experimental findings. Moreover, with MC-PDFT, we could perform dynamics simulations with the 12 electron in 10 orbitals active space that was computationally too expensive for direct dynamics with CASPT2.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2021

Spontaneous and explicit parity-time-symmetry breaking in drift-wave instabilities

A method of parity-time- (PT) symmetry analysis is introduced to study the high-dimensional, complicated parameter space of drift-wave instabilities. Here, we show that spontaneous PT-symmetry breaking leads to the ion temperature gradient instability of drift waves, and the collisional instability is the result of explicit PT-symmetry breaking. A new unstable drift wave induced by finite collisionality is identified. It is also found that gradients of ion temperature and density can destabilize the ion cyclotron waves when PT symmetry is explicitly broken by a finite collisionality.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2023

An adaptive sampling augmented Lagrangian method for stochastic optimization with deterministic constraints

The primary goal of this paper is to provide an efficient solution algorithm based on the augmented Lagrangian framework for optimization problems with a stochastic objective function and deterministic constraints. Our main contribution is combining the augmented Lagrangian framework with adaptive sampling, resulting in an efficient optimization methodology validated with practical examples. To achieve the presented efficiency, here we consider inexact solutions for the augmented Lagrangian subproblems, and through an adaptive sampling mechanism, we control the variance in the gradient estimates. Furthermore, we analyze the theoretical performance of the proposed scheme by showing equivalence to a gradient descent algorithm on a Moreau envelope function, and we prove sublinear convergence for convex objectives and linear convergence for strongly convex objectives with affine equality constraints. The worst-case sample complexity of the resulting algorithm, for an arbitrary choice of penalty parameter in the augmented Lagrangian function, is $\mathscr{O}$(ϵ -3-δ ) , where ϵ > 0 is the expected error of the solution and δ > 0 is a user-defined parameter. If the penalty parameter is chosen to be $\mathscr{O}$(ϵ -1 ), we demonstrate that the result can be improved to $\mathscr{O}$(ϵ -2 ) , which is competitive with the other methods employed in the literature. Moreover, if the objective function is strongly convex with affine equality constraints, we obtain $\mathscr{O}$(ϵ -1 log(1/ϵ)) complexity. Finally, we empirically verify the performance of our adaptive sampling augmented Lagrangian framework in machine learning optimization and engineering design problems, including topology optimization of a heat sink with environmental uncertainty.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Large-eddy simulation of helical- and straight-bladed vertical-axis wind turbines in boundary layer turbulence

Here, turbulent wake flows behind helical- and straight-bladed vertical axis wind turbines (VAWTs) in boundary layer turbulence are numerically studied using the large-eddy simulation (LES) method combined with the actuator line model. Based on the LES data, systematic statistical analyses are performed to explore the effects of blade geometry on the characteristics of the turbine wake. The time-averaged velocity fields show that the helical-bladed VAWT generates a mean vertical velocity along the center of the turbine wake, which causes a vertical inclination of the turbine wake and alters the vertical gradient of the mean streamwise velocity. Consequently, the intensities of the turbulent fluctuations and Reynolds shear stresses are also affected by the helical-shaped blades when compared with those in the straight-bladed VAWT case. The LES results also show that reversing the twist direction of the helical-bladed VAWT causes the spatial patterns of the turbulent wake flow statistics to be reversed in the vertical direction. Moreover, the mass and kinetic energy transports in the turbine wakes are directly visualized using the transport tube method, and the comparison between the helical- and straight-bladed VAWT cases show significant differences in the downstream evolution of the transport tubes.

17 WIND ENERGY↗

DOE OSTI2025

Multiscale investigation of thermomechanical and compositional developments in Ni alloy 718 under laser processing

Laser processing has been widely employed in various applications due to its exceptional spatial resolution. However, the rapid temperature gradients generated in localized areas present significant challenges for experimental characterization using conventional instruments. To characterize Ni alloy 718 during laser processing, we employed in-situ synchrotron X-ray diffraction with a high-speed detector, a method particularly well-suited for probing processes with high temporal and spatial resolution. Through a series of in-situ experiments, we investigated the local variations in the evolution of microstructures and thermomechanical behaviors within a keyhole mode melt pool. The in situ macroscopic thermomechanical behaviors were quantified using an empirical model derived from diffraction patterns, with experimental results showing reasonable agreement with finite element analysis. Various laser parameters were tested to assess their influences on the residual strains in the melt pools. The results revealed that the residual strain in the keyhole mode melt pool is relatively insensitive to variations in the parameters and is smaller than that in the melt pool created under conduction mode laser scanning. Additionally, we analyzed the shapes of individual diffraction spots, providing insights into the plastic behaviors and compositional developments in the resolidified alloy. The analysis confirmed that compositional variations in a dendritic microstructure manifest as asymmetric broadening of the diffraction spots.

Ni alloy 718↗

DOE OSTI2024

D2NO: Efficient handling of heterogeneous input function spaces with distributed deep neural operators

Neural operators have been applied in various scientific fields, such as solving parametric partial differential equations, dynamical systems with control, and inverse problems. However, challenges arise when dealing with input functions that exhibit heterogeneous properties, requiring multiple sensors to handle functions with minimal regularity. To address this issue, discretization-invariant neural operators have been used, allowing the sampling of diverse input functions with different sensor locations. However, existing frameworks still require an equal number of sensors for all functions. We propose a novel distributed approach to further relax the discretization requirements and solve the heterogeneous dataset challenges. Our method involves partitioning the input function space and processing individual input functions using independent and separate neural networks. A centralized neural network is used to handle shared information across all output functions. This distributed methodology reduces the number of gradient descent back-propagation steps, improving efficiency while maintaining accuracy. Here, we demonstrate that the corresponding neural network is a universal approximator of continuous nonlinear operators and present three numerical examples to validate its performance.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2021

Characterize traction–separation relation and interfacial imperfections by data-driven machine learning models

Abstract Interfacial mechanical properties are important in composite materials and their applications, including vehicle structures, soft robotics, and aerospace. Determination of traction–separation (T–S) relations at interfaces in composites can lead to evaluations of structural reliability, mechanical robustness, and failures criteria. Accurate measurements on T–S relations remain challenging, since the interface interaction generally happens at microscale. With the emergence of machine learning (ML), data-driven model becomes an efficient method to predict the interfacial behaviors of composite materials and establish their mechanical models. Here, we combine ML, finite element analysis (FEA), and empirical experiments to develop data-driven models that characterize interfacial mechanical properties precisely. Specifically, eXtreme Gradient Boosting (XGBoost) multi-output regressions and classifier models are harnessed to investigate T–S relations and identify the imperfection locations at interface, respectively. The ML models are trained by macroscale force–displacement curves, which can be obtained from FEA and standard mechanical tests. The results show accurate predictions of T–S relations ( R 2 = 0.988) and identification of imperfection locations with 81% accuracy. Our models are experimentally validated by 3D printed double cantilever beam specimens from different materials. Furthermore, we provide a code package containing trained ML models, allowing other researchers to establish T–S relations for different material interfaces.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Report Summarizing Progress in Digital Image Correlation Analysis of Burst Phenomenon

This report is an update on the progress made towards utilizing in-situ methods to observe and quantify Zry-4 deformation behavior during burst testing. A heavily modified burst test was developed that allows both digital image correlation and infrared thermography to be utilized during testing to quantify strain and axial thermal gradient data during burst. A variety of tests ranging from 6.2-15.9 MPa internal pressure and utilizing both constant and transient pressure conditions were conducted and presently reported on. The test modifications, which are detailed extensively in a previous report, are effectively found to have no discernible impact on burst temperatures when compared to historical Zircaloy burst data. To support future modeling studies, the IR thermography data was used to approximate the thermal gradient down the axial length at burst for all the claddings. Then, DIC is first correlated to thermocouple temperature for a 10.3 MPa Transient pressure test, and then DIC correlated to IR thermal gradient data down the axial length of a 8.2 MPa Transient pressure test at several points.

36 MATERIALS SCIENCE↗

DOE OSTI2023

Investigation of local distortion effects on X-ray absorption of ferroelectric perovskites from first principles simulations

Understanding the role of ferroelectric polarization in modulating the electronic and structural properties of crystals is critical for advancing these materials for overcoming various technological and scientific challenges. However, due to difficulties in performing experimental methods with the required resolution, or in interpreting the results of methods therein, the nanoscale morphology and response of these surfaces to external electric fields has not been properly elaborated. Here, in this work, we investigate the effect of ferroelectric polarization and local distortions in a BaTiO 3 perovskite, using two widely used computational approaches which treat the many-body nature of X-ray excitations using different philosophies, namely the many-body, delta-self-consistent-field determinant (mb-ΔSCF) and the Bethe–Salpeter equation (BSE) approaches. We show that in agreement with our experiments, both approaches consistently predict higher excitations of the main peak in the O–K edge for the surface with upward polarization. However, the mb-ΔSCF approach mostly fails to capture the L 2,3 separations at the Ti–L edge, due to the absence of spin–orbit coupling in Kohn–Sham density functional theory (KS-DFT) at the generalized gradient approximation level. On the other hand, and most promising, we show that application of the GW/BSE approach successfully reproduces the experimental XAS, both the relative peak intensities as well as the L 2,3 separations at the Ti–L edges upon ferroelectric switching. Thus simulated XAS is shown to be a powerful method for capturing the nanoscale structure of complex materials, and we underscore the need for many-body perturbation approaches, with explicit consideration of core-hole and multiplet effects, for capturing the essential physics in these systems.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

DOE OSTI2020

Direct laser writing of volumetric gradient index lenses and waveguides

Abstract Direct laser writing (DLW) has been shown to render 3D polymeric optical components, including lenses, beam expanders, and mirrors, with submicrometer precision. However, these printed structures are limited to the refractive index and dispersive properties of the photopolymer. Here, we present the subsurface controllable refractive index via beam exposure (SCRIBE) method, a lithographic approach that enables the tuning of the refractive index over a range of greater than 0.3 by performing DLW inside photoresist-filled nanoporous silicon and silica scaffolds. Adjusting the laser exposure during printing enables 3D submicron control of the polymer infilling and thus the refractive index and chromatic dispersion. Combining SCRIBE’s unprecedented index range and 3D writing accuracy has realized the world’s smallest (15 µm diameter) spherical Luneburg lens operating at visible wavelengths. SCRIBE’s ability to tune the chromatic dispersion alongside the refractive index was leveraged to render achromatic doublets in a single printing step, eliminating the need for multiple photoresins and writing sequences. SCRIBE also has the potential to form multicomponent optics by cascading optical elements within a scaffold. As a demonstration, stacked focusing structures that generate photonic nanojets were fabricated inside porous silicon. Finally, an all-pass ring resonator was coupled to a subsurface 3D waveguide. The measured quality factor of 4600 at 1550 nm suggests the possibility of compact photonic systems with optical interconnects that traverse multiple planes. SCRIBE is uniquely suited for constructing such photonic integrated circuits due to its ability to integrate multiple optical components, including lenses and waveguides, without additional printed supports.

36 MATERIALS SCIENCE↗

DOE OSTI2023

A data-driven method for modelling dissipation rates in stratified turbulence

We present a deep probabilistic convolutional neural network (PCNN) model for predicting local values of small-scale mixing properties in stratified turbulent flows, namely the dissipation rates of turbulent kinetic energy and density variance, $\varepsilon$ and $\chi$ . Inputs to the PCNN are vertical columns of velocity and density gradients, motivated by data typically available from microstructure profilers in the ocean. The architecture is designed to enable the model to capture several characteristic features of stratified turbulence, in particular the dependence of small-scale isotropy on the buoyancy Reynolds number $Re_b:=\varepsilon /(\nu N^2)$ , where $\nu$ is the kinematic viscosity and $N$ is the background buoyancy frequency, the correlation between suitably locally averaged density gradients and turbulence intensity and the importance of capturing the tails of the probability distribution functions of values of dissipation. Empirically modified versions of commonly used isotropic models for $\varepsilon$ and $\chi$ that depend only on vertical derivatives of density and velocity are proposed based on the asymptotic regimes $Re_b\ll 1$ and $Re_b\gg 1$ , and serve as an instructive benchmark for comparison with the data-driven approach. When trained and tested on a simulation of stratified decaying turbulence which accesses a range of turbulent regimes (associated with differing values of $Re_b$ ), the PCNN outperforms assumptions of isotropy significantly as $Re_b$ decreases, and additionally demonstrates improvements over the fitted empirical models. A differential sensitivity analysis of the PCNN facilitates a comparison with the theoretical models and provides a physical interpretation of the features enabling it to make improved predictions.

42 ENGINEERING↗

DOE OSTI2023

Corrosion Testing Trade Study on Testing Methods in Molten Salts

In assessing molten salt compatibility with structural materials, static testing is often reported because it is easy to conduct. However, in order to assess mass transfer, where material dissolves in the salt in the hot section and may precipitate out of the liquid in colder regions, a flowing salt experiment in a temperature gradient is needed. While a forced convection or pumped loop produces prototypic flowing conditions, it also is a multi-million dollar investment. Among the less expensive dynamic testing options, the thermal convection loop (TCL) was favored during the 1960’s development of molten salt reactor technology. Because the workforce and core facilities are currently available to conduct TCL experiments, it is the least expensive strategy to evaluate candidate structural alloys in flowing molten salt conditions.

36 MATERIALS SCIENCE↗

DOE OSTI2023

On-Demand Aligned DNA Hydrogel Via Light Scanning

DNA is an anisotropic, water-attracting, and biocompatible material, an ideal building block for hydrogel. The alignment of the anisotropic DNA chains is essential to maximize hydrogel properties, which has been little explored. Here, we present a method to fabricate the anisotropic DNA hydrogel that allows precise control for the polymerization process of photoreactive cationic monomers. Scanning ultraviolet light enables the uniaxial alignment of DNA chains through the polymerization-induced diffusive mass flow using a concentration gradient. In conclusion, while studying anisotropic mechanical properties and orientation recovery according to the DNA chain alignment direction, we demonstrate the potential of directionally controlled DNA hydrogels as smart materials.

36 MATERIALS SCIENCE↗

DOE OSTI2021

Correspondence between neuroevolution and gradient descent

Abstract We show analytically that training a neural network by conditioned stochastic mutation or neuroevolution of its weights is equivalent, in the limit of small mutations, to gradient descent on the loss function in the presence of Gaussian white noise. Averaged over independent realizations of the learning process, neuroevolution is equivalent to gradient descent on the loss function. We use numerical simulation to show that this correspondence can be observed for finite mutations, for shallow and deep neural networks. Our results provide a connection between two families of neural-network training methods that are usually considered to be fundamentally different.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Thermal relaxation and the complete set of second-order transport coefficients for the unitary Fermi gas from kinetic theory

We compute the complete set of second-order transport coefficients of the unitary Fermi gas, a dilute gas of spin-1/2 particles interacting via an 𝑠 -wave interaction tuned to infinite scattering length. The calculation is based on kinetic theory and the Chapman-Enskog method at second order in the Knudsen expansion. We take into account the exact two-body collision integral. We extend previous results on second-order coefficients related to shear stress by including terms related to heat flow and gradients of the fugacity. We confirm that the thermal relaxation time is given by the simple estimate 𝜏 𝜅 = 𝜅⁢𝑚/(𝑐 𝑃 ⁢𝑇) even if the full collision kernel is taken into account. Furthermore, 𝜅 is the thermal conductivity, 𝑚 is the mass of the particles, 𝑐𝑃 is the specific heat at constant pressure, and 𝑇 is the temperature.

Kinetic theory↗