A spatially explicit evolutionary algorithm for the spatial partitioning problem
Not Available
Liu, Yan Y. (ORCID:0000000322984728)↗
SEARCH · Engineering Papers
Results for “Inspires”
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.
Ten questions concerning agent-based modeling of occupant behavior for energy and environmental performance of buildings
We report the complexity of occupant behavior is one of the major contributors to uncertainty in building performance simulation. Agent-based modeling (ABM), a computational simulation technique, has gained attention in the occupant modeling field due to its capability and flexibility to capture the heterogeneity and dynamics of human behavior and the emergent effects. While multiple efforts in the past decade have demonstrated the usefulness of the ABM approach for simulating occupants and their impacts on building performance, several crucial matters in the ABM research still remain unexplored. This paper presents ten questions that highlight the most important issues regarding ABM research and applications for occupant behavior in the context of building performance simulation. The questions and answers aim to provide insights into current and future ABM research, and more importantly to inspire new significant questions from young researchers in the field. This research is part of the IEA EBC Annex 79 project, occupant-centric building design and operation.
32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
A scalable domain decomposition method for FEM discretizations of nonlocal equations of integrable and fractional type
Nonlocal models allow for the description of phenomena which cannot be captured by classical partial differential equations. The availability of efficient solvers is one of the main concerns for the use of nonlocal models in real world engineering applications. Here, we present a domain decomposition solver that is inspired by substructuring methods for classical local equations. In numerical experiments involving finite element discretizations of scalar and vectorial nonlocal equations of integrable and fractional type, we observe improvements in solution time of up to 14.6x compared to commonly used solver strategies.
97 MATHEMATICS AND COMPUTING↗
Inverse design of two-dimensional graphene/h-BN hybrids by a regressional and conditional GAN
Design of materials with desired properties is currently laborious and heavily relies on intuition of researchers through a trial-and-error process. To tackle this challenge, in this work we propose a novel regressional and conditional generative adversarial network (RCGAN) for inverse design of representative two-dimensional materials, the graphene and boron-nitride (BN) hybrids. RCGAN incorporates a supervised regressor network, thus overcoming the common technical barrier in the traditional unsupervised GANs, which cannot generate data when fed with continuous and quantitative labels. RCGAN can autonomously generate graphene/BN hybrids given any target bandgap values. These structures are distinguished from the ones used for training and exhibit high diversity for a given bandgap. Moreover, they exhibit high fidelity, yielding bandgaps within ~10% MAE F of the desired bandgaps as validated by density functional theory (DFT) calculations. Analysis by the principle component analysis (PCA) and modified locally linear embedding (MLLE) reveals that the generator has successfully generated structures following the statistical distribution of the real structures. It implies the possibility of the RCGAN in recognizing physical rules hidden in the high-dimensional data. The novel strategy for designing regressional GAN architecture together with the successful application to inverse design of materials would inspire further exploration in research fields beyond materials.
36 MATERIALS SCIENCE↗
Influence of graphite geography on the yield of mechanically exfoliated few-layer graphene
Some of graphene's exciting properties that inspired countless studies over the last two decades are exclusive to defect-free, few-layer graphene (FLG, ≤4 layers) produced by mechanical exfoliation. Despite graphite exfoliation having been thoroughly studied, the influence of the graphite source on the successful exfoliation of FLG remains unexplored. Here, in this work, we examine the physicochemical properties and exfoliation performance of various graphite types (e.g., natural crystalline flake, natural vein, and synthetic) from around the globe (e.g., Alaska, Tanzania, China, etc.). We first established a normalization method for the facile comparison of FLG (Φ) yields between separately exfoliated graphites. Using this approach, we find that the relative yield of FLG ranges from 0% to 22% across graphite types, with regular natural crystalline flakes (NCF-R) performing the best overall (Φ NCF-R = 16% ± 5%). FLG yield was also found to depend on source geography, evidenced by Φ NCF-R ranging from 8% to 22% for NCF-R sourced from Alabama and Canada, respectively. We employed machine learning and Pearson correlation analysis to determine the graphite characteristics that govern FLG yield. Graphite source-dependent properties such as graphite surface area and mineral impurities, including aluminum and calcium, were found to be critically important. Our findings highlight the importance of the graphite source when trying to maximize the yield of mechanically exfoliated FLG and represent an important step toward producing higher quantities of defect-free FLG for graphene-based research.
42 ENGINEERING↗
Protein-based models offer mechanistic insight into complex nickel metalloenzymes
There are ten nickel enzymes found across biological systems, each with a distinct active site and reactivity that spans reductive, oxidative, and redox–neutral processes. We focus on the reductive enzymes, which catalyze reactions that are highly germane to the modern-day climate crisis: [NiFe] hydrogenase, carbon monoxide dehydrogenase, acetyl coenzyme A synthase, and methyl coenzyme M reductase. The current mechanistic understanding of each enzyme system is reviewed along with existing knowledge gaps, which are addressed through the development of protein-derived models, as described here. Finally, this opinion is intended to highlight the advantages of using robust protein scaffolds for modeling multiscale contributions to reactivity and inspire the development of novel artificial metalloenzymes for other small molecule transformations.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Tackling issues of lithium metal anodes with a novel polymeric lithicone coating
Lithium metal (Li) has been hindered from as anodes in commercial batteries for over 50 years, due to two main issues: continuous formation of solid electrolyte interphase (SEI) and lithium dendritic growth. In this work, we report a new strategy to tackle these issues, i.e., using molecular layer deposition (MLD) to grow an ionically conducting but electrically insulating polymeric lithicone coating, an Li-containing triethanolamine (LiTEA). Our electrochemical tests revealed that this LiTEA coating could serve as an exceptional protection layer over Li anodes. Consequently, the LiTEA-coated Li electrodes could achieve a superior cyclability of > 10000 Li stripping/plating cycles at a current density of 5 mA cm -2 and a long cyclability of > 5500 cycles at 2 mA cm -2 in Li||Li symmetric cells without failures, under a fixed areal capacity of 1 mAh cm -2 . Characterizations using scanning electron microscopy and X-ray photoelectron microscopy verified that Li + ions could be easily extracted through and deposited under the LiTEA coating during the stripping/plating processes. Consequently, this LiTEA coating significantly inhibited the formation of SEI and Li dendrites. This underlies the long cycling lifetime of the LiTEA-coated Li||Li cells. Coupling with the nickel-rich LiNi 0.8 Mn 0.1 Co 0.1 O 2 (NMC811) cathodes, more encouragingly, the LiTEA-coated Li anodes could remarkably extend the cyclability and sustainable capacity of the resultant Li||NMC full cells. Finally, we also demonstrated that the performance of Li||NMC cells could be further improved through combining an LiTEA-coated Li anode with an Li 2 S-modified NMC811 via atomic layer deposition (ALD). Thus, this study is inspiring for developing high-energy Li||NMC lithium metal batteries.
25 ENERGY STORAGE↗
Advances in atomistic modeling and understanding of drying shrinkage in cementitious materials
Despite more than a century of research, drying shrinkage still causes unwanted stresses and cracks in our concrete infrastructures. This is partly due to the lack of understanding of the underlying mechanisms governing the drying shrinkage in cementitious materials. Inspired by the tremendous impact of atomistic simulations on materials science, we expect such innovative simulation techniques to hold a key to unlocking the enigmatic nature of drying shrinkage at the nanoscale. In this treatise, we first introduce basic physics concepts and atomistic simulation methods in detail. Afterward, we discuss the progress brought by atomistic simulations in understanding the drying shrinkage in cementitious materials and other nanoporous and nanolayered systems. This review also highlights important remaining fundamental questions and practical issues regarding drying shrinkage and delineates how atomistic simulations can help resolve them.
36 MATERIALS SCIENCE↗
Trends in electrocatalytic activity and stability of transition-metal nitrides
Transition metal nitrides (TMNs) are a class of electrocatalyst support materials similar to transition metal carbides (TMCs) with the advantage of avoiding the issues arising from graphitic carbon surface deposits during synthesis. Inspired by previous studies suggesting that TMCs could be used to reduce Pt loading for the hydrogen evolution reaction (HER), this work explored the feasibility of TMN-supported Pt and Au as HER electrocatalysts. Here this study established a volcano-like trend between electrochemical HER activity and hydrogen-binding energy (HBE) calculated from density functional theory. The Pt/TiN and Au/TiN materials were used to extend knowledge from well-characterized thin films to powder catalysts. In situ X-ray absorption spectroscopy (XAS) measurements provided additional characterization of the Pt/TiN and Au/TiN catalysts under HER conditions. Trends in the electrochemical stability of TMNs were also investigated over a wide range of potentials and pH values, which can be used to guide future studies for TMN-supported electrocatalysts.
58 GEOSCIENCES↗
Sealing fractures to increase underground storage security: Lessons learned from a multiscale multim
Explore the source record for details and available documents.
Hajirezaie, Sassan↗
A biohybrid strategy for enabling photoredox catalysis with low-energy light
Natural systems drive the high-energy reactions of photosynthesis with efficient and broadband energy capture. Transition-metal photocatalysts similarly convert light into chemical reactivity, and yet suffer from light-limited operation and require blue-to-UV excitation. In photosynthesis, both light capture and reactivity have been optimized by separation into distinct sites. In this work, inspired by this modular architecture, we synthesized a biohybrid photocatalyst by covalent attachment of the photosynthetic light-harvesting protein R-phycoerythrin (RPE) to the transition-metal photocatalyst tris(2,2'-bipyridine)ruthenium(II) ([Ru(bpy) 3 ] 2+ ). Spectroscopic investigation found that absorbed photoenergy was efficiently funneled from RPE to [Ru(bpy) 3 ] 2+ . The utility of the biohybrid photocatalyst was demonstrated via an increase in yields for a thiol-ene coupling reaction and a cysteinyl-desulfurization reaction, including recovered reactivity at red wavelengths where [Ru(bpy) 3 ] 2+ alone does not absorb.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Porous materials for hydrogen storage
Hydrogen is a clean energy carrier and a promising candidate to realize carbon-neutral systems for a sustainable future because the combustion of hydrogen produces only water. The large-scale implementation of hydrogen, particularly in zero-emission hydrogen-powered fuel-cell automobiles, requires efficient hydrogen storage systems; however, the current on-board hydrogen storage technologies mainly rely on costly and potentially unsafe high-pressure compression strategies that reach pressures of up to 700 bar. Recently, solid-state porous materials, such as metal-organic frameworks, have emerged as sorbents that can conceivably store comparable amounts of hydrogen in a safer and more efficient manner relative to the current high-pressure storage technologies. Here this review summarizes the current progress on the development of state-of-the-art porous materials and their composites for hydrogen storage. We hope this work will inspire efforts toward creating next-generation hydrogen storage materials.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Learning high-dimensional parametric maps via reduced basis adaptive residual networks
We propose a scalable framework for the learning of high-dimensional parametric maps via adaptively constructed residual network (ResNet) maps between reduced bases of the inputs and outputs. When just few training data are available, it is beneficial to have a compact parametrization in order to ameliorate the ill-posedness of the neural network training problem. By linearly restricting high-dimensional maps to informed reduced bases of the inputs, one can compress high-dimensional maps in a constructive way that can be used to detect appropriate basis ranks, equipped with rigorous error estimates. A scalable neural network learning framework is thus to learn the nonlinear compressed reduced basis mapping. Unlike the reduced basis construction, however, neural network constructions are not guaranteed to reduce errors by adding representation power, making it difficult to achieve good practical performance. Inspired by recent approximation theory that connects ResNets to sequential minimizing flows, we present an adaptive ResNet construction algorithm. This algorithm allows for depth-wise enrichment of the neural network approximation, in a manner that can achieve good practical performance by first training a shallow network and then adapting. We prove universal approximation of the associated neural network class for $L^2_v$ functions on compact sets. Our overall framework allows for constructive means to detect appropriate breadth and depth, and related compact parametrizations of neural networks, significantly reducing the need for architectural hyperparameter tuning. Numerical experiments for parametric PDE problems and a 3D CFD wing design optimization parametric map demonstrate that the proposed methodology can achieve remarkably high accuracy for limited training data, and outperformed other neural network strategies we compared against.
42 ENGINEERING↗
Surrogate construction via weight parameterization of residual neural networks
Surrogate model development is a critical step for uncertainty quantification or other sample-intensive tasks for complex computational models. Here, in this work, we develop a multi-output surrogate form using a class of neural networks (NNs) that employ shortcut connections, namely Residual NNs (ResNets). ResNets are known to regularize the surrogate learning problem and improve the efficiency and accuracy of the resulting surrogate. Inspired by the continuous, Neural ODE analogy, we augment ResNets with weight parameterization strategy with respect to ResNet depth. Weight-parameterized ResNets regularize the NN surrogate learning problem and allow better generalization with a drastically reduced number of learnable parameters. We demonstrate that weight-parameterized ResNets are more accurate and efficient than conventional feed-forward multi-layer perceptron networks. We also compare various options for parameterization of the weights as functions of ResNet depth. We demonstrate the results on both synthetic examples and a large scale earth system model of interest.
97 MATHEMATICS AND COMPUTING↗
Dedication to James A. Miller
This special memorial issue pays tribute to James (Jim) A. Miller, a giant of combustion science who died in 2021, with a celebration of his enormous influence on the field. We were touched by the responses we received after we sent out the invitations for it. Jim inspired several generations of scientists, who viewed him as a mentor, a father figure, and a friend. Together with Nils Hansen and Peter Glarborg, we have written a detailed account on his life and work. Furthermore, it appeared in this journal shortly after his death; and so here we focus on the scientific areas he had interest in and influence on, and how they relate to the 34 papers in this issue. The topics of these papers span a variety of Jim's interests including nitrogen chemistry, polycyclic aromatic hydrocarbon (PAH) chemistry, oxidation chemistry, energy transfer, prompt dissociations, and codes to facilitate combustion chemistry simulations.
96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗
Development of an interatomic potential for the W–Ta system
A physics-inspired, data-driven interatomic potential framework for multi-element system is presented. This potential is based on the generalization of the embedded atom method potential and systematically incorporates two-, three- and many-body effects. Different atomic environments are described by atom-centered Gaussian basis sets which provides sufficient flexibility to capture diverse atomic environments and allows for easy optimization of the free parameters. An interatomic potential model for the tungsten–tantalum (W–Ta) system is developed using this framework by training on data from ab initio density functional theory (DFT) calculations. The potential is thoroughly tested at various compositions by comparing bulk and defect properties from experimental and DFT data. It is shown that the potential predicts the elastic constants, defect properties, such as vacancy and interstitial formation energies, core structures of dislocations, and melting points for both pure and tantalum and tungsten with an accuracy comparable to other single element machine learning based potentials. In conclusion, the potential model is used to investigate the formation energies of ordered alloys and vacancy formation energies for chemically random W–Ta alloys.
75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
Toucan: A performance portable, scalable implementation of the DECA algorithm
In the field of additive manufacturing (AM), cellular automata (CA) is extensively used to simulate microstructural evolution during solidification. However, while traditional CA approaches are relatively fast, they still require a substantial number of time steps, are limited to moderate volumes, and are relatively difficult to improve through parallelism due to the highly localized nature of the solidification front. Here, to address these issues of time to solution and load balancing, we introduce Toucan, a parallel, performance-portable, and scalable code written in C++ with the Kokkos library that leverages the discrete event inspired cellular automata (DECA) algorithm to perform parallel-in-time (PinT) grain growth simulations. Toucan effectively mitigates load balancing issues by distributing the computational workload more evenly across processors, enhancing scalability and efficiency. We conduct both strong and weak scaling studies on up to 64 GPUs on the Frontier supercomputer, demonstrating that Toucan significantly outperforms the current state-of-the-art, time-stepped CA code, ExaCA, on both single and multi-GPU simulations. Even in AM-specific weak scaling scenarios, Toucan maintains near-ideal scaling, in contrast to the linear increase observed with ExaCA due to the moving laser raster pattern. This study highlights Toucan’s potential to transform microstructural simulations in AM by radically improving both efficiency and scalability over existing methods.
36 MATERIALS SCIENCE↗
Qualitative trend analysis based on a mixed-integer representation
Shape constrained spline fitting is a useful method to impose prior knowledge onto flexible semi-parametric models during parameter estimation. Most typically, the function shape is imposed through order restrictions on the regression coefficients. The intended shape is considered known or selected based on heuristic rules. In this study, we present a method to estimate the optimal set of order restrictions to segment a univariate data series into episodes with distinct shapes. This is also known as the qualitative trend analysis (QTA) problem. The obtained solution uses a trade-off between lack-of-fit and model complexity. Further, our practical implementation takes inspiration from the generalized order restricted information criterion (GORIC) for inequality-constrained model selection. From this, one learns (a) that QTA can be formulated as a mixed-integer quadratic program (MIQP) and (b) that the newly proposed mixed order restricted information criterion (MORIC) enables optimal segmentation. This is illustrated through didactic case studies.
42 ENGINEERING↗