Engineering Papers⌕ Search

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.

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

At least 1,045 records · Page 58
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Comparing GGA, GGA+ U , and meta-GGA functionals for redox-dependent binding at open metal sites in metal–organic frameworks

Metal–organic frameworks (MOFs) with open metal sites have been widely investigated for the selective adsorption of small molecules via redox mechanisms where charge transfer can take place between the binding site and the adsorbate of interest. Quantum-chemical screening methods based on density functional theory have emerged as a promising route to accelerate the discovery of MOFs with enhanced binding affinities toward various adsorbates. However, the success of this approach is linked to the accuracy of the underlying density functional approximations (DFAs). In this work, we compare commonly used generalized gradient approximation (GGA), GGA+U, and meta-GGA exchange-correlation functionals in modeling redox-dependent binding at open metal sites in MOFs using O 2 and N 2 as representative small molecules. We find that the self-interaction error inherent to the widely used Perdew, Burke, and Ernzerhof (PBE) GGA predicts metal sites that are artificially redox-active, as evidenced by their strong binding affinities, short metal–adsorbate bond distances, and large degree of charge transfer. The incorporation of metal-specific, empirical Hubbard U corrections based on the transition metal oxide literature systematically reduces the redox activity of the open metal sites, often improving agreement with experiment. Additionally, the binding behavior shifts from strong chemisorption to weaker physisorption as a function of U. The M06-L meta-GGA typically predicts binding energies between those of PBE-D3(BJ) and PBE-D3(BJ)+U when using empirically derived U values from the transition metal oxide literature. Despite the strong sensitivity of the binding affinities toward a given DFA, the GGA, GGA+U, and meta-GGA approaches often yield the same qualitative trends and structure–property relationships.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Accelerating defect predictions in semiconductors using graph neural networks

First-principles computations reliably predict the energetics of point defects in semiconductors but are constrained by the expense of using large supercells and advanced levels of theory. Machine learning models trained on computational data, especially ones that sufficiently encode defect coordination environments, can be used to accelerate defect predictions. Here, we develop a framework for the prediction and screening of native defects and functional impurities in a chemical space of group IV, III–V, and II–VI zinc blende semiconductors, powered by crystal Graph-based Neural Networks (GNNs) trained on high-throughput density functional theory (DFT) data. Using an innovative approach of sampling partially optimized defect configurations from DFT calculations, we generate one of the largest computational defect datasets to date, containing many types of vacancies, self-interstitials, anti-site substitutions, impurity interstitials and substitutions, as well as some defect complexes. We applied three types of established GNN techniques, namely crystal graph convolutional neural network, materials graph network, and Atomistic Line Graph Neural Network (ALIGNN), to rigorously train models for predicting defect formation energy (DFE) in multiple charge states and chemical potential conditions. We find that ALIGNN yields the best DFE predictions with root mean square errors around 0.3 eV, which represents a prediction accuracy of 98% given the range of values within the dataset, improving significantly on the state-of-the-art. We further show that GNN-based defective structure optimization can take us close to DFT-optimized geometries at a fraction of the cost of full DFT. The current models are based on the semi-local generalized gradient approximation-Perdew–Burke–Ernzerhof (PBE) functional but are highly promising because of the correlation of computed energetics and defect levels with higher levels of theory and experimental data, the accuracy and necessity of discovering novel metastable and low energy defect structures at the PBE level of theory before advanced methods could be applied, and the ability to train multi-fidelity models in the future with new data from non-local functionals. The DFT-GNN models enable prediction and screening across thousands of hypothetical defects based on both unoptimized and partially optimized defective structures, helping identify electronically active defects in technologically important semiconductors.

Rahman, Md Habibur (ORCID:000000027705984X)↗
DOE OSTI2024

How well do one-electron self-interaction-correction methods perform for systems with fractional electrons?

Recently developed locally scaled self-interaction correction (LSIC) is a one-electron SIC method that, when used with a ratio of kinetic energy densities (z σ ) as iso-orbital indicator, performs remarkably well for both thermochemical properties as well as for barrier heights overcoming the paradoxical behavior of the well-known Perdew–Zunger self-interaction correction (PZSIC) method. In this work, we examine how well the LSIC method performs for the delocalization error. Our results show that both LSIC and PZSIC methods correctly describe the dissociation of $H$$^{+}_{2}$ and $H$$^{+}_{2}$ but LSIC is overall more accurate than the PZSIC method. Likewise, in the case of the vertical ionization energy of an ensemble of isolated He atoms, the LSIC and PZSIC methods do not exhibit delocalization errors. For the fractional charges, both LSIC and PZSIC significantly reduce the deviation from linearity in the energy vs number of electrons curve, with PZSIC performing superior for C, Ne, and Ar atoms while for Kr they perform similarly. The LSIC performs well at the endpoints (integer occupations) while substantially reducing the deviation. The dissociation of LiF shows both LSIC and PZSIC dissociate into neutral Li and F but only LSIC exhibits charge transfer from Li + to F – at the expected distance from the experimental data and accurate ab initio data. Overall, both the PZSIC and LSIC methods reduce the delocalization errors substantially.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Optical diagnostics of laser-produced plasmas

Laser-produced plasmas (LPPs) engulf exotic and complex conditions ranging in temperature, density, pressure, magnetic and electric fields, charge states, charged particle kinetics, and gas-phase reactions, based on the irradiation conditions, target geometries, and the background cover gas. The application potential of the LPP is so diverse that it generates considerable interest for both basic and applied research areas. The fundamental research on LPPs can be traced back to the early 1960s, immediately after the invention of the laser. In the 1970s, the laser was identified as a tool to pursue inertial confinement fusion, and since then, several other technologies emerged out of LPPs. These applications prompted the development and adaptation of innovative diagnostic tools for understanding the fundamental nature and spatiotemporal properties of these complex systems. Although most of the traditional characterization techniques developed for other plasma sources can be used to characterize the LPPs, care must be taken to interpret the results because of their small size, transient nature, and inhomogeneities. The existence of the large spatiotemporal density and temperature gradients often necessitates non-uniform weighted averaging over distance and time. Among the various plasma characterization tools, optical-based diagnostic tools play a key role in the accurate measurements of LPP parameters. The optical toolbox contains optical spectroscopy (emission, absorption, and fluorescence), passive and active imaging and optical probing methods (shadowgraphy, Schlieren, interferometry, Thomson scattering, deflectometry, and velocimetry). Each technique is useful for measuring a specific property, and its use is limited to a certain time span during the LPP evolution because of the sensitivity issues related to the selected measuring tool. Therefore, multiple diagnostic tools are essential for a comprehensive insight into the entire plasma behavior. In recent times, the improvements in performance in the lasers and detector systems expanded the capability of the aforementioned passive and active diagnostics tools. Furthermore, this review provides an overview of optical diagnostic tools frequently employed for the characterization of the LPPs and emphasizes techniques, associated assumptions, and challenges. Considering most of the industrial and other applications of the LPP belong to low to moderate laser intensities (10 8 – 10 15 Wcm –2 ), this review focuses on diagnostic tools pertaining to this regime.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2026

Surface Water Quality Data from Beaver-Impacted Streams; Trail Creek and East River, Colorado 2025

This data package contains surface water chemistry measurements collected in 2025 to evaluate how beaver damming and low-tech process-based stream restoration influence water quality and metal mobility in mountainous headwater systems of the Upper Colorado River Basin. Sampling was conducted at Trail Creek (Taylor Park watershed, Colorado), a tributary undergoing restoration through installation of low-tech process-based structures (i.e., beaver dam analogs), and at off-channel beaver ponds within the East River floodplain (East River watershed, Colorado). Samples were collected along longitudinal transects spanning upstream control reaches, beaver-influenced ponded reaches, and downstream segments. Additional samples were collected from near-surface pore waters within a beaver dam seepage face. The dataset includes concentrations of major and trace elements measured by inductively coupled plasma–mass spectrometry (ICP-MS) and inductively coupled plasma–optical emission spectrometry (ICP-OES), major anions measured by ion chromatography (IC), and dissolved organic carbon (DOC; reported as non-purgeable organic carbon, NPOC). Samples were size-fractionated at 0.45 micrometers (µm), 0.22 µm, and 0.02 µm to distinguish particulate (>0.45 µm), colloidal (0.22–0.02 µm), and dissolved (<0.02 µm) fractions. The data package consists of comma-separated value (.csv) files containing tabulated chemical concentration data, sample metadata (site identifiers, geographic coordinates, sampling dates, fraction type), and quality control flags. All files are provided in open, non-proprietary formats that can be accessed using standard data analysis software such as Microsoft Excel, R, Python, MATLAB, or other programs capable of reading .csv files. Units, detection limits, and analytical methods are documented in accompanying metadata files. The dataset is designed to support analyses of (1) how beaver impoundment and restoration structures alter elemental partitioning and transport, (2) the role of iron and organic carbon in mediating trace metal mobility, and (3) reach-scale changes in water quality across restoration gradients. This work was supported by the Watershed Function Science Focus Area at Lawrence Berkeley National Laboratory funded by the US Department of Energy, Office of Science, Biological and Environmental Research under Contract No. DE-AC02-05CH11231. Part of this work was performed at SLAC Accelerator Laboratory funded by the US Department of Energy, Office of Science, Biological and Environmental Research under Contract No. DE-AC02-76SF00515.

Anions↗
DOE OSTI2024

The Information Length Concept Applied to Plasma Turbulence

A methodology to study statistical properties of anomalous transport in fusion plasma is investigated. Three time traces generated by the full-f gyrokinetic code GKNET are analyzed for this purpose. The time traces consist of heat flux as a function of the radial position, which is studied in a novel manner using statistical methods. The simulation data exhibit transport processes with both medium and long correlation length along the radius. A typical example of a phenomenon with long correlation length is avalanches. In order to investigate the evolution of the turbulent state, two basic configurations are studied, one flux-driven and one gradient-driven with decaying turbulence. The information length concept in tandem with Boltzmann–Gibbs and Tsallis entropy is used in the investigation. It is found that the dynamical states in both flux-driven and gradient-driven cases are surprisingly similar, but the Tsallis entropy reveals differences between them. This indicates that the types of probability distribution function are nevertheless quite different since the higher moments are significantly different.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2020

Scalable Deep-Learning-Accelerated Topology Optimization for Additively Manufactured Materials

Topology optimization (TO) is a popular and powerful computational approach for designing novel structures, materials, and devices. Two computational challenges have limited the applicability of TO to a variety of industrial applications. First, a TO problem often involves a large number of design variables to guarantee sufficient expressive power. Second, many TO problems require a large number of expensive physical model simulations, and those simulations cannot be parallelized. To address these issues, we propose a general scalable deep-learning (DL) based TO framework, referred to as SDL-TO, which utilizes parallel schemes in high performance computing (HPC) to accelerate the TO process for designing additively manufactured (AM) materials. Unlike the existing studies of DL for TO, our framework accelerates TO by learning the iterative history data and simultaneously training on the mapping between the given design and its gradient. The surrogate gradient is learned by utilizing parallel computing on multiple CPUs incorporated with a distributed DL training on multiple GPUs. The learned TO gradient enables a fast online update scheme instead of an expensive update based on the physical simulator or solver. Using a local sampling strategy, we achieve to reduce the intrinsic high dimensionality of the design space and improve the training accuracy and the scalability of the SDL-TO framework. The method is demonstrated by benchmark examples and AM materials design for heat conduction. The proposed SDL-TO framework shows competitive performance compared to the baseline methods but significantly reduces the computational cost by a speed up of around 8.6x over the standard TO implementation.

Bi, Sirui↗
DOE OSTI2023

Deep reinforcement learning assisted co-optimization of Volt-VAR grid service in distribution networks

With the increasing penetration of distributed energy resources in distribution networks, Volt-VAR control and optimization (VVC/VVO) have become very important to ensure an acceptable quality of service to all customers. System operators can rely on slow-responding utility devices, including capacitor banks and on-load tap changing transformers, along with fast-responding battery and photovoltaic (PV) inverters for the VVC/VVO implementation. Because of variations in response time of these two classes of devices, and different control actions (discrete versus continuous), coordinated and optimal scheduling and operation have become of utmost importance. Here, this paper develops a look-ahead deep reinforcement learning (DRL)-based multi-objective VVO technique to improve the voltage profile of active distribution networks, decrease network and inverter power loss, and save the operational cost of the grid. It proposes a deep deterministic policy gradient (DDPG)-based approach to schedule the optimal reactive and/or active power set-points of fast-responding inverters, and a deep Q-network (DQN)-based DRL agent to schedule the discrete decisions variables of slow-responding assets. The reactive power output of PV and battery smart inverters are scheduled at 30-minute intervals and the capacitors’ commitment status is scheduled with several hour intervals. The proposed framework is validated on the modified IEEE 34-bus and 123-bus test cases with embedded PV and PV-plus-storage. To validate the efficacy of the proposed VVO, it is compared with several scenarios, including the base case without VVO, localized droop control of DERs, DDPG-only, and twin delayed DDPG (TD3) agent-based DRL techniques. The results justify the superior performance of the proposed method to improve the voltage profile, reduce network power loss, and minimize the look-ahead grid operational cost while minimizing the undesirable power losses in inverters as a result of power factor adjustments.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2022

Understanding the role of segmentation on process-structure–property predictions made via machine learning

Here, the present study investigated the effect of porosity surface determination methods on performance of machine learning models used to predict the tensile properties of AlSi10Mg processed by laser powder bed fusion from micro-computed tomography data. Machine learning models applied in this work include support vector machines, neural networks, decision trees, and Bayesian classifiers. The effects of isosurface thresholding and local gradient approaches for porosity segmentation, as well as image filtering schemes, on model precision were evaluated for samples produced under differing levels of global energy density.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Adaptive, problem-tailored variational quantum eigensolver mitigates rough parameter landscapes and barren plateaus

Abstract Variational quantum eigensolvers (VQEs) represent a powerful class of hybrid quantum-classical algorithms for computing molecular energies. Various numerical issues exist for these methods, however, including barren plateaus and large numbers of local minima. In this work, we consider the Adaptive, Problem-Tailored Variational Quantum Eiegensolver (ADAPT-VQE) ansätze, and examine how they are impacted by these local minima. We find that while ADAPT-VQE does not remove local minima, the gradient-informed, one-operator-at-a-time circuit construction accomplishes two things: First, it provides an initialization strategy that can yield solutions with over an order of magnitude smaller error compared to random initialization, and which is applicable in situations where chemical intuition cannot help with initialization, i.e., when Hartree-Fock is a poor approximation to the ground state. Second, even if an ADAPT-VQE iteration converges to a local trap at one step, it can still “burrow” toward the exact solution by adding more operators, which preferentially deepens the occupied trap. This same mechanism helps highlight a surprising feature of ADAPT-VQE: It should not suffer optimization problems due to barren plateaus and random initialization. Even if such barren plateaus appear in the parameter landscape, our analysis suggests that ADAPT-VQE avoids such regions by design.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2022

Machine learning for impurity charge-state transition levels in semiconductors from elemental properties using multi-fidelity datasets

Quantifying charge-state transition energy levels of impurities in semiconductors is critical to understanding and engineering their optoelectronic properties for applications ranging from solar photovoltaics to infrared lasers. While these transition levels can be measured and calculated accurately, such efforts are time-consuming and more rapid prediction methods would be beneficial. Here, we significantly reduce the time typically required to predict impurity transition levels using multi-fidelity datasets and a machine learning approach employing features based on elemental properties and impurity positions. We use transition levels obtained from low-fidelity (i.e., local-density approximation or generalized gradient approximation) density functional theory (DFT) calculations, corrected using a recently proposed modified band alignment scheme, which well-approximates transition levels from high-fidelity DFT (i.e., hybrid HSE06). Further, the model fit to the large multi-fidelity database shows improved accuracy compared to the models trained on the more limited high-fidelity values. Crucially, in our approach, when using the multi-fidelity data, high-fidelity values are not required for model training, significantly reducing the computational cost required for training the model. Our machine learning model of transition levels has a root mean squared (mean absolute) error of 0.36 (0.27) eV vs high-fidelity hybrid functional values when averaged over 14 semiconductor systems from the II–VI and III–V families. As a guide for use on other systems, we assessed the model on simulated data to show the expected accuracy level as a function of bandgap for new materials of interest. Finally, we use the model to predict a complete space of impurity charge-state transition levels in all zinc blende III–V and II–VI systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Western Sector In-Situ Chemical Oxidation Project: Additional Injection Pilot Testing Supplemental Results (U)

Trichloroethylene (TCE) and tetrachloroethylene (PCE) are persistent at elevated concentrations in the Western Sector of the M-Area Hazardous Waste Management Facility (HWMF) groundwater plume. Migration of these chlorinated ethenes as Dense Non-Aqueous Phase Liquids (DNAPLs) occurred from the M-Area Settling Basin along geologic strata resulting in a high concentration groundwater plume in Western Sector. Pump-and-treat has been active for over 30 years at the M-Area HWMF to hydraulically contain the primary sources of the plume in addition to other remediation strategies throughout the contaminated zone, like soil vapor extraction, dynamic underground stripping, groundwater recirculation wells, and bioremediation. In-situ chemical oxidation (ISCO) has been identified as a viable treatment method for removal of these volatile organic compounds (VOCs) due to the complex lithology and high concentrations in this part of the aquifer. In addition to ISCO, a new recovery well (i.e., RWM018) was also installed in Western Sector to contain the high concentration VOC plume. The ISCO projects have utilized the gradient induced by RWM018 to help plan the injection and monitoring locations. The combination of ISCO and the additional pump and treat in Western Sector with the installation of RWM018 will help to control the plume and maximize remediation efforts.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2025

Product ion distributions using H 3 O + proton-transfer-reaction time-of-flight mass spectrometry (PTR-ToF-MS): mechanisms, transmission effects, and instrument-to-instrument variability

Abstract. Proton-transfer-reaction mass spectrometry (PTR-MS) using hydronium ion (H3O+) ionization is widely used for the measurement of volatile organic compounds (VOCs) both indoors and outdoors. H3O+ ionization, as well as the associated chemistry in an ion–molecule reactor, is known to generate product ion distributions (PIDs) that include other product ions besides the proton-transfer product. We present a method, using gas-chromatography pre-separation, for quantifying PIDs from PTR-MS measurements of nearly 100 VOCs of different functional types including alcohols, ketones, aldehydes, acids, aromatics, organohalides, and alkenes. We characterize instrument configuration effects on PIDs and find that reactor reduced electric field strength (E/N), ion optic voltage gradients, and quadrupole settings have the strongest impact on measured PIDs. Through an interlaboratory comparison of PIDs measured from calibration cylinders, we characterized the variability of PID production from the same model of PTR-MS across seven participating laboratories. Product ion variability was generally smaller (e.g., < 20 %) for ions with larger contributions to the PIDs (e.g., > 0.30) but less predictable for product ions formed through O2+ and NO+ reactions. We present a publicly available library of H3O+ PTR-MS PIDs that will be updated periodically with user-provided data for the continued investigation into instrument-to-instrument variability of PIDs.

Link, Michael F. (ORCID:0000000218412455)↗
DOE OSTI2022

A finite micro-rotation material point method for micropolar solid and fluid dynamics with three-dimensional evolving contacts and free surfaces

This paper introduces an explicit material point method designed specifically for simulating the micropolar continuum dynamics in the finite deformation and finite microrotation regime. The material point method enables us to simulate large deformation problems while circumventing the potential mesh distortion without remeshing. To eliminate rotational motion damping and loss of angular momentum during the projection, we introduce the mapping for microinertia and angular momentum between particles and grids through the affine particle-in-cell approach. The microrotation and the curvature at each particle are updated through zero-order forward integration of the microgyration and its spatial gradient. We show that the microinertia and the angular momentum are conserved during the projections between particles and grids in our formulation. We verify the formulation and implementation by comparing with the analytical dispersion relation of micropolar waves under the small strain and small microrotation, as well as the analytical soliton solution for solids undergoing large deformation and large microrotation. Additionally we also demonstrate the capacity of the proposed computational framework to handle a wide spectrum of simulations that exhibit size effects in the geometrical nonlinear regime through three representative numerical examples, i.e., a cantilever beam torsion problem, a fragment-impact penetration problem, and a micropolar fluid discharging problem.

42 ENGINEERING↗
DOE OSTI2021

Multiscale design of nonlinear materials using a Eulerian shape optimization scheme

Motivated by recent advances in manufacturing, the design of materials is the focal point of interest in the material research community. One of the critical challenges in this field is finding optimal material microstructure for a desired macroscopic response. This work presents a computational method for the mesoscale-level design of particulate composites for an optimal macroscale-level response. The method relies on a custom shape optimization scheme to find the extrema of a nonlinear cost function subject to a set of constraints. Three key “modules” constitute the method: multiscale modeling, sensitivity analysis, and optimization. Multiscale modeling relies on a classical homogenization method and a nonlinear NURBS-based generalized finite element scheme to efficiently and accurately compute the structural response of particulate composites using a nonconformal discretization. A three-parameter isotropic damage law is used to model microstructure-level failure. An analytical sensitivity method is developed to compute the derivatives of the cost/constraint functions with respect to the design variables that control the microstructure's geometry. The derivation uncovers subtle but essential new terms contributing to the sensitivity of finite element shape functions and their spatial derivatives. Several structural problems are solved to demonstrate the applicability, performance, and accuracy of the method for the design of particulate composites with a desired macroscopic nonlinear stress-strain response.

42 ENGINEERING↗
DOE OSTI2025

Effect of artificial viscosity on shocked particle-laden flows for staggered grid Lagrangian methods

Abstract Shocked particle-laden flows are important to many natural and industrial processes. When simulating these systems, artificial viscosity is often required to prevent numerical artifacts, such as ringing, from arising in the pressure and density fields. The linear and quadratic coefficients of the artificial viscosity determine the amount of smoothing that occurs in these fields. For particle-laden flows, however, many of the fluid–particle interaction forces, for example, the pressure gradient force and unsteady forces, depend on gradients in the fluid fields. Furthermore, while the shock passes over a particle, these forces can be more dominant than drag. This means that the artificial viscosity coefficients affect how a particle and fluid interact when simulating shocked particle systems. Here this effect is investigated for isolated particles and for a particle curtain using a staggered grid Lagrangian approach. The artificial viscosity coefficients have a significant impact on the maximum force that a fluid imparts to a particle, which is important for determining whether a particle will break up in response to the shock. Furthermore, it is found that the density ratio between the particle and the fluid is important in determining whether the artificial viscosity coefficients have a significant impact on the particle’s motion.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2022

Shear strain gradient in Cu/Nb nanolaminates: Strain accommodation and chemical mixing

Disentangling the intragranular and interfacial plasticity contribution to the overall strain accommodation is crucial to understanding the microstructural evolution and mass transport upon deformation in materials with the nanoscale feature size. In this study, we devise a new approach to tackle the issue by introducing shear strain gradients into Cu/Nb nanolaminates of different layer thicknesses with the shear perpendicular to the laminate interfaces. The measurement of the strain gradient and the resultant lattice disorientation enables a quantitative understanding of the intragranular and interfacial plasticity contribution. We found that intragranular slip entirely governs the deformation in the 300 nm-layer laminate and, unexpectedly, contributes ~80% of the total plasticity in the 30 nm-layer laminate. The high intragranular plasticity in the thin laminate is attributed to the large width of confined slip planes and their remnant potential for storing dislocations. In addition, substantial forced chemical mixing is observed in the top region of the 30 nm-layer laminate where the effective layer thickness is reduced below 8 nm. The transition of deformation mechanism from confined layer slip to dislocation transmission is largely responsible for the initiation of substantial mixing. Our method and findings shed light on the deformation mechanism and deformation-induced mass transport behavior in nanostructured materials.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Differentiable Multiphysics Codes: A Breakthrough Technology for Simulation and Computing

This document summarizes the findings of a strategic planning exercise commissioned by the Weapons Simulation and Computing, Computational Physics (WSC/CP) program at the Lawrence Livermore National Laboratory (LLNL) in FY24. During the year, the committee met with multiple stakeholder communities to gather input, opinions, suggestions and concerns which have been incorporated throughout this document. The key findings from this exercise are summarized: • The development of multiphysics modelling and simulation (mod/sim) codes and software technologies, their deployment on exascale compute platforms, and their broad adoption across the NNSA is a major success of the Advanced Simulation and Computing (ASC) program and the Exascale Computing Project (ECP). Sustained investment in these core technologies is essential. • Today’s state of the art involves running ensembles of O(100K) simulations to perform uncertainty quantification (UQ) and design studies using multiple statistical methods such as Bayesian optimization to understand sensitivities of our models and explore parameterized design spaces. Even with exascale computing, we are practically limited to O(10) parameters in these studies since the number of simulations required to sample the space scales exponentially with the number of design parameters. • The data from these simulation ensembles is increasingly being used to train machine learned (ML) surrogates (or reduced order models, ROMs) which can then be used for optimization or real time design exploration. However, the trained surrogates are still limited in the number of parameters they can represent due to the sampling limitations previously noted. • Augmenting our suite of integrated multiphysics simulation codes, both current and emerging, with the ability to compute gradients (solution derivatives) of arbitrary simulation outputs with respect to (some or all) simulation inputs would be a breakthrough technology, opening the door to a new era of efficient and automated inverse design based on verified and validated mod/sim capabilities. • This capability, which we refer to as differentiable multiphysics codes (DMCs), would revolutionize both UQ and optimization studies by breaking the curse of dimensionality that presently limits our “gradient-free” ensemble based computing approach. A similar breakthrough occurred in the AI/ML community once the ability to compute gradients of arbitrary loss functions using back-propagation became commonplace. Gradient information from the multiphysics codes can also be used to dramatically improve the efficiency and scale of training of ML/ROM surrogates for rapid assessments. • Achieving this in our suite of codes will be a grand challenge, similar to the amount of effort that was required to transition from CPU to GPU computing. It will require buy-in from the entire WSC/CP program and beyond, including all integrated codes, physics and engineering models, third-party library dependencies and performance portability abstractions. It will also require investment in research and development of numerical methods for computing adjoints of coupled physics across multiple adaptively refined moving meshes and of stochastic (Monte Carlo) and mesh free (SPH) methods. • New software and numerical techniques, largely pioneered by the AI/ML community, make this feasible. Chief among these is automatic differentiation (AD), the ability to employ AD at point-wise locations in a physics calculation (instead of traditional black-box approaches) and the ability to perform “back-propagation in time” (or reverse mode AD) for non-linear partial differential equations (PDEs). Fundamentally, the conclusion of this strategic planning exercise is that the time is right to undertake a large scale effort in WSC, centered on the existing integrated codes, to continue the natural evolution of mod/sim in the age of AI/ML. Instead of attempting to replace mod/sim with purely data driven AI/ML models, we believe the key to success is to integrate AI/ML by building on top of the decades of hard-won knowledge and the verified/validated multiphysics modelling capability that is the hallmark of the ASC program.

97 MATHEMATICS AND COMPUTING↗