Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Scaling Studies”

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 235 records · Page 13
All sourcesNASA NTRSDOE OSTI
NASA NTRS2016

Conceptual Design and Structural Optimization of NASA Environmentally Responsible Aviation (ERA) Hybrid Wing Body Aircraft

Simultaneously achieving the fuel consumption and noise reduction goals set forth by NASA's Environmentally Responsible Aviation (ERA) project requires innovative and unconventional aircraft concepts. In response, advanced hybrid wing body (HWB) aircraft concepts have been proposed and analyzed as a means of meeting these objectives. For the current study, several HWB concepts were analyzed using the Hybrid wing body Conceptual Design and structural optimization (HCDstruct) analysis code. HCDstruct is a medium-fidelity finite element based conceptual design and structural optimization tool developed to fill the critical analysis gap existing between lower order structural sizing approaches and detailed, often finite element based sizing methods for HWB aircraft concepts. Whereas prior versions of the tool used a half-model approach in building the representative finite element model, a full wing-tip-to-wing-tip modeling capability was recently added to HCDstruct, which alleviated the symmetry constraints at the model centerline in place of a free-flying model and allowed for more realistic center body, aft body, and wing loading and trim response. The latest version of HCDstruct was applied to two ERA reference cases, including the Boeing Open Rotor Engine Integration On an HWB (OREIO) concept and the Boeing ERA-0009H1 concept, and results agreed favorably with detailed Boeing design data and related Flight Optimization System (FLOPS) analyses. Following these benchmark cases, HCDstruct was used to size NASA's ERA HWB concepts and to perform a related scaling study.

Quinlan, Jesse R.↗
DOE OSTI2026

Unravelling fast-charging degradation in NMC/Gr pouch cells: Lithium plating and SEI properties

As fast-charging technology expands across the electric vehicle and emerging energy-storage applications, understanding its impact on battery performance and longevity is critical. In this study, 1.8 Ah LiNi 0.6 Mn 0.2 Co 0.2 O 2 /graphite pouch cells were charged at various charging rates (0.5C, 2C, 4C, and 6C) to investigate the degradation mechanisms. Our results showed that well-designed NMC/Gr pouch cells could reach over 1000 cycles with a 2C charging rate, while only reaching around 500 cycles with 4C and 6C charging rates. Fast-charging effects on NMC and graphite electrodes were obtained through a series of post-mortem characterizations, including electrochemical impedance spectroscopy (EIS), Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), and scanning electron microscopy (SEM) with energy-dispersive X-ray spectroscopy (EDS). Although higher charging rates cause pulverization of NMC secondary particles, the dominant degradation mechanism driving the fading of fast-charging-related performance lies in the graphite anode, where lithium plating and LiF-rich solid electrolyte interphase (SEI) formation result in Li inventory loss and impedance growth. The postmortem results suggest that the formation of a LiF-rich SEI, which exacerbates anode impedance and some irreversible Li + ion loss, is likely driven by the substantial decomposition of PF 6 − during fast charging, an effect often overlooked in smaller laboratory-scale studies.

Luo, Mei [Argonne National Laboratory (ANL), Argon↗
DOE OSTI2024

Optimizing Distributed Training on Frontier for Large Language Models

Large language models (LLMs) have demonstrated remarkable success as foundational models, benefiting various downstream applications through fine-tuning. Loss scaling studies have demonstrated the superior performance of larger LLMs compared to their smaller counterparts. Nevertheless, training LLMs with billions of parameters poses significant challenges and requires considerable computational resources. For example, training a one trillion parameter GPT-style model on 20 trillion tokens requires a staggering 120 million exaflops. This research explores efficient distributed training strategies to extract this computation from Frontier, the world's first exascale supercomputer. We enable and investigate various model and data parallel training techniques, such as tensor parallelism, pipeline parallelism, and sharded data parallelism, to facilitate training a trillion-parameter model on Frontier. We empirically assess these techniques and their associated parameters to determine their impact on memory footprint, communication latency, and GPU's computational efficiency. We analyze the complex interplay among these techniques and find a strategy to combine them to achieve high throughput through hyperparameter tuning. We have identified efficient strategies for training large LLMs of varying sizes through empirical analysis and hyperparameter tuning. For 22 Billion, 175 Billion, and 1 Trillion parameters, we achieved GPU throughputs of 38.38%, 36.14%, and 31.96%, respectively. For the training of the 175 Billion parameter model and the 1 Trillion parameter model, we achieved 100% weak scaling efficiency on 1024 and 3072 Mi250X GPUs, respectively. We also achieved strong scaling efficiencies of 89% and 87% for these two models. We trained these models only tens of iterations instead of training till completion.

Yin, Junqi↗
DOE OSTI2020

Ecohydrology of wetland plant communities along an estuarine to tidal river gradient

Abiotic filters that interact with wetland plant communities along tidal–fluvial gradients are highly dynamic, and understanding their quantitative thresholds and relationships to interspecific competition is important during an era of sea-level rise and watershed hydrologic change. Yet, landscape-scale studies of major coastal rivers from the river mouth to the head of tide, such as this study, remain rare. Here, we develop a new predictive framework for estuarine–tidal river research and management using a river-specific low-water datum and the wetland inundation indicator SEVg, the growing-season sum exceedance value of hourly surface-water depth. The distribution and variability of the wetland species pool (n = 203) on the 234 river kilometer (rkm) lower Columbia River and estuary floodplain are described for the first time. 4,940 quadrats at 50 marshes were surveyed (2005–2016). Throughout the estuarine–tidal river system, SEVg was well suited to describe the wetland inundation regime and its variability based on the combination of longitudinal river position and elevation. SEVg increased significantly landward. Two primary wetland inundation regimes were identified: the seaward-tidal, usually greatest during the winter months, and landward-fluvial, greatest during the growing season. Nearest the ocean, salinity is the abiotic factor limiting species richness and non-native species. Farther upriver, the daily, seasonal, and interannual variability of the wetting and drying cycle encourage disturbance-tolerant species and non-natives and limit the number of hydrophytes and total vegetative cover. Hence, the average between-year similarity of site-scale areal cover significantly decreased landward. Hierarchical cluster analysis indicated five vegetative groups and five ecohydrologic zones between rkm 0 and 234 were discriminated with 76 significant species–zone associations. All zones had unique indicator species. Species with high indicator values were Carex lyngbyei throughout the estuarine zones, and Eleocharis palustris, Sagittaria latifolia, and the invasive non-native Phalaris arundinacea in the upper estuarine and lower, middle, and upper tidal river zones (IV > 0.90). Competition from C. lyngbyei nearest the ocean and P. arundinacea in the tidal river was associated with reduced species richness when total cover was >65%. This framework of filters informs the design and prediction of future wetland plant communities on coastal river floodplains.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Regional-scale soil carbon predictions can be enhanced by transferring global-scale soil–environment relationships

Accurate modelling and mapping soil organic carbon are crucial for supporting soil health restoration and climate change mitigation at both regional and global scales. However, regional soil predictions often suffer from data scarcity and high prediction uncertainty. Utilizing a pre-trained global-to-regional soil carbon predictive model can be a potential solution to address this challenge. Despite its promise, how to construct and apply the global-scale model to enhance regional-scale soil carbon mapping remains largely unexplored. Here, we propose the Global Soil Carbon Pre-trained Model (GSoilCPM), a deep-learning-based domain adaptative model, to enhance regional-scale soil carbon predictions. Based on large amount of environmental covariate data and 106,167 soil samples across the globe, we verify our hypothesis of the effectiveness of this 'global-to-regional' modelling strategy. The pre-trained model can be then transferred and fine-tuned to bridge the regional- and global-scale soil–environment relationships. We applied and validated this modelling strategy in four regional-scale study areas, three in the Northern Hemisphere and one in the Southern Hemisphere, each with distinct environmental background. Compared to traditional modelling approaches as a baseline, four case studies all demonstrated significant improvement in prediction accuracy across diverse environments and varying data availabilities. The average percentage improvement across all regions is 10.93% (absolute values decreased by 1.20 g kg−1 averagely) in MAE and 29.04% (absolute values increased by 0.10 averagely) in CCC. The applicability and future horizons of using GSoilCPM were further discussed. We further reveal that regions with fewer soil samples or lower baseline accuracy benefit more from the pre-trained global model. Our findings highlight the advantages of leveraging the generalized knowledge from global models to enhance specifically localized soil modelling, positioning a potential paradigm shift in digital soil mapping, and far-reaching implications for soil monitoring and land management.

Deep learning↗
DOE OSTI2023

Snow-corrected vegetation indices for improved gross primary productivity assessment in North American evergreen forests

North American evergreen forests cover large areas and influence the global carbon cycle. Satellite remote sensing has been used to track the phenology of ecosystem photosynthesis of these forests by detecting variation in vegetation optical properties associated with physiological and structural features, and most of these methods have been closely tied to vegetation greenness. However, in evergreens, the application of satellite data to monitor photosynthetic phenology is often limited by the lack of sensitivity of greenness-based indices. In this study, we identified 47 evergreen forest flux sites in North America that had MODIS observation overlapping with the flux tower records. We then calculated four vegetation indices using MODIS MAIAC data (MCD19A1), including NDVI, CCI, NIRv, and kNDVI, for the 47 flux sites and evaluated relationships between gross primary productivity (GPP) and vegetation indices across the North American evergreen forests. Our results showed that snow had substantial effects on the performance of all vegetation indices in tracking GPP phenology, particularly in the early spring when rapid changes occurred to both GPP and snow cover. Furthermore, different vegetation indices were affected differently, indicating contradictory and confounding effects of snow on these indices. After correcting for the snow effects, both CCI and NIRv performed well in tracking GPP phenology, albeit for different reasons. CCI is sensitive to seasonal changes in the relative levels of chlorophyll and carotenoid pigments, which are closely tied to GPP phenology in evergreens. NIRv is sensitive to the absorbed photosynthetically active radiation and to the contribution of deciduous components to the overall optical properties. We also found that correlations between GPP and vegetation indices varied among ecoregions and climate classes. In general, regions with pronounced seasonal GPP patterns had stronger correlations between GPP and greenness-based indices than regions with weaker seasonal GPP patterns. These biome differences were less pronounced for CCI. The snow artifacts and complementary vegetation index effects reported here should be considered in any large-scale studies of GPP using reflectance-based indices from optical satellites.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Subcomponent Validation of Composite Joints for the Marine Energy Advanced Materials Project

The Marine Energy Advanced Materials project is an ongoing multi-year, multi-lab project with the main goals of addressing barriers and uncertainties facing marine energy developers in adopting advanced materials for structural applications. NREL's goals of the project were to address subcomponents testing needs for marine energy materials, to improve understanding of design allowables at the full-scale and provide near net-scale static and fatigue data of composite subcomponents using materials applicable to the marine energy industry. In the long term, the test method development and data generated would be used to inform standards development. This report outlines perhaps one of the largest-scale studies conducted with regards to saltwater conditioning of various composite material subcomponents and their subsequent structural validation, specifically directed at the marine renewable energy industry. A variety of fiberglass composite panels with epoxy and vinyl ester epoxy resin systems were manufactured at Montana State University, which were then used to manufacture an array of different types of subcomponent test specimens at the National Renewable Energy Laboratory's Flatirons Campus. These subcomponents were in the form of T-bolt and double-ended-insert specimens, which were intended to represent bonded and mechanical bolted connections for thick composite laminates, metal-metal and composite lap shear specimens to evaluate adhesion of constituent materials, and adhesive beam-shear specimens as part of an effort to better evaluate the characteristics of thick adhesive bondlines. Overall, the materials used were fiberglass reinforced epoxy and vinyl ester matrix composites, epoxy and methacrylate adhesives, and 316 and 2507 stainless steels. Specimens were then conditioned in salt water at various temperatures and for various periods of time at Florida Atlantic University and Pacific Northwest National Laboratory. All specimens were then mechanically characterized and validated using various test methods under static and fatigue loading conditions at NREL's Structural Technology Laboratory. Throughout the conditioning and mechanical validation process, valuable experience was gained, which will help guide future test method development for marine energy materials. In many instances, the results indicated similar observations as to what had been observed during previous coupon scale characterization efforts that provided a vital understanding of the scale up process. However in some instances, unexpected phenomena were observed, such as interactions between the adhesives and 316 steel. Furthermore, some materials exhibited significant degradation due to the saltwater conditioning. Ultimately, this report provides a detailed summary of the specimens that were designed, the subcomponent test methods that were developed, and the results that were generated, which will serve as important guidance for marine renewable energy developers and researchers for future structural designs and validation.

16 TIDAL AND WAVE POWER↗
DOE OSTI2024

Agent-based modeling for multimodal transportation of CO 2 for carbon capture, utilization, and storage: CCUS-agent

Here, to understand the system-level interactions between the entities in Carbon Capture, Utilization, and Storage (CCUS), an agent-based foundational modeling tool, CCUS-Agent, is developed for a large-scale study of transportation flows and infrastructure in the United States. Key features of the tool include (i) modular design, (ii) multiple transportation modes, (iii) capabilities for extension, and (iv) testing against various system components and networks of small and large sizes. Five matching algorithms for CO 2 supply agents (e.g., powerplants and industrial facilities) and demand agents (e.g., storage and utilization sites) are explored: Most Profitable First Year (MPFY), Most Profitable All Years (MPAY), Shortest Total Distance First Year (SDFY), Shortest Total Distance All Years (SDAY), and Shortest distance to long-haul transport All Years (ACAY). Before matching, the supply agent, demand agent, and route must be available, and the connection must be profitable. A profitable connection means the supply agent portion of revenue from the 45Q tax credit must cover the supply agent costs and all transportation costs, while the demand agent revenue portion must cover all demand agent costs. A case study employing over 5500 supply and demand agents and multimodal CCUS transportation infrastructure in the contiguous United States is conducted. The results suggest that it is possible to capture over 9 billion tonnes (GT) of CO 2 from 2025 to 2043, which will increase significantly to 22 GT if the capture costs are reduced by 40 %. The MPFY and SDFY algorithms capture more CO 2 earlier in the time horizon, while the MPAY and SDAY algorithms capture more later in the time horizon.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2022

Luminescence dosimetry for evaluation of the external exposure in Metlino, upper Techa River valley, Southern Urals, Russia: Analysis of new results

Luminescence dosimetry was applied in the former settlement of Metlino, Southern Urals, Russia as part of a full-scale study to validate the Techa River Dosimetry System (TRDS) 2016 for the upper Techa River region. The village, which was evacuated in 1956, was located 7 km downstream of the release point of liquid radioactive waste by the Mayak plutonium facility. Several brick samples were taken from north-eastern and south-eastern walls of the granary, facing the former Techa river shoreline and floodplain. Samples were all taken at the same height and measured at different depths into the brick. For the majority of brick samples, good Optically Stimulated Luminescence properties of the quartz grains were observed. In some cases, however, strong levels of sensitization and/or signal recuperation were encountered which necessitated adjustment in the measurement protocols. Anthropogenic doses in bricks varied from 1.5 to 6.6 Gy and the horizontal profiles along both walls showed significant variation, which is explained on a qualitative basis. Further, a dose depth profile is observed for selected samples, which is different from the dose depth profile measured and simulated for samples from the north-western wall of the granary in previous studies. This is qualitatively explained by the differences in source configuration.

61 RADIATION PROTECTION AND DOSIMETRY↗
DOE OSTI2026

Insights into mixing of non-isothermal multi-polymer melts for complex plastics recycling

Catalytic recycling or upcycling of plastics is often limited not by catalyst performance, but by transport, arising from highly viscous, non-Newtonian polymer melts. In this work, we develop a reactor-scale framework that integrates rheological measurements, constitutive modeling, computational fluid dynamics (CFD), and experiments to quantify mixing, heat transfer, and dispersion in surrogate hydrocarbon melts representing mixed plastics systems. Temperature- and shear rate-dependent viscosity of low-density polyethylene (LDPE) and high-density polyethylene (HDPE) is measured to create two surrogate polymers (PLD and PHD) that capture the dominant shear-thinning flow behavior while neglecting strong elastic effects, enabling tractable simulation of non-isothermal, polymer-melt mixing using a Carreau-Arrhenius generalized Newtonian framework. Three-dimensional CFD simulations are employed to evaluate impeller performance in PLD using mixing time, cavern volume, thermal uniformity, and interfacial area for regimes in which viscoelastic effects are not dominant. We show that magnetic stir bars commonly used in lab-scale studies produce large thermal gradients (~60 °C) and poor mixing, even under idealized power delivery and polymer flow conditions. In contrast, close-clearance anchor impellers achieve near-isothermal operation, reduce mixing times by up to 5×, and provide >90% active circulation volume. We further demonstrate that, at low pseudo-Deborah number (De*), motor power requirements can be predicted directly from shear rate-dependent rheology using the Carreau-Arrhenius framework, enabling rational selection of operating conditions. Extension to surrogate immiscible multi-polymer systems based on PLD and PHD shows that interfacial area is highly sensitive to operating conditions and impeller design, with coaxial anchor-turbine configurations enhancing dispersion by up to 4 × .

Close-clearance impellers↗
DOE OSTI2020

Eulerian thermo-mechanical simulations of heterogeneous solid propellants using an approximate projection method

We present a mathematical model that describes thermo-mechanical deformations and thermal gradients on the unsteady burning of a heterogeneous solid propellant. A scaling study shows that the deformations in solid at combustion timescales can be treated as quasi-static. The resulting thermo-mechanical formulation is formulated on a Cartesian grid and makes use of a weak form of Chorin-type projection method to deal with large difference in shear modulus of constituent materials. Further, a one-dimensional verification study is carried out by comparing numerical simulations with those of an analytical model. In addition, convergence studies for a two-dimensional propellant sandwich configuration are presented for the stress, velocity, and reference map components. Finally, simulations are carried out for a two-dimensional random propellant pack and the time-averaged burn rate and strains are reported.

42 ENGINEERING↗
DOE OSTI2020

Towards the scale-up production of cellulose nanocomposites using melt processing: A critical review on structure-processing-property relationships

Cellulose nanomaterials (CNMs) naturally exist in plant biomass. The success of extraction of CNMs opened up a new era of using plant biomass for innovative industrial applications. Because CNMs are abundant, renewable, biodegradable, transparent, light weight and low in cost, they are ideal materials for large volume applications such as packaging, automotive, building and infrastructure. In many potential application areas, CNM-enable products appear in a composite form, mostly polymer composites. Additionally, the industrial-scale manufacturing of CNM/thermoplastic composites remains as a set of unsolved problems for academia and industry. A prime challenge in applications is the nanoscale dispersion of CNMs in thermoplastic matrices during melt processing. Both bench-scale and pilot-scale studies have been conducted to solve the dispersion issue of CNMs. In this article, research related to the dispersion of CNMs in thermoplastic matrices during melt processing were critically reviewed. All research papers were classified into three groups: chemically-aided dispersion, physically-aided dispersion and mechanically-aided dispersion. Numerous factors affect the CNM dispersion and the mechanical performance of its nanocomposites. There are material-related factors, including CNM types and forms, polymer matrices, surface modification, coupling agents, etc. Furthermore, extrusion processing parameters also play a significant role, covering screw rotation speed, extrusion barrel temperature settings and screw design. In addition, the material-related factors interact with the processing-related factors. Understanding all factors and their interactions are important for moving CNM nanocomposites research a step further towards industrial-relevant production, which is the final ambitious goal of this manuscript.

42 ENGINEERING↗
DOE OSTI2025

Computing material volume fractions on a superimposed mesh as applied to Monte Carlo particle transport simulations

Here, we present a newly implemented ray tracing algorithm in OpenMC for efficiently computing material volume fractions on superimposed meshes in complex geometries. By firing rays along each coordinate direction through the geometry, the approach accumulates track-length data in each mesh element, thereby determining the fractional composition of each material. Scaling studies on three different models—a random tetrahedra configuration, the Frascati Neutron Generator ITER dose rate benchmark, and a stellarator design—show excellent parallel performance, with nearly linear speedup on modern multi-threaded and distributed-memory systems. An analysis of the residual error relative to high-resolution reference solutions demonstrated that under optimal conditions it decreases as 1/R, where R is the number of rays fired, making it straightforward to achieve user-prescribed accuracy. This new functionality enables practical, mesh-based approaches for detailed nuclear analyses in production Monte Carlo workflows without resorting to expensive, fully conformal or unstructured meshing.

Monte Carlo↗
DOE OSTI2020

Massively parallel transport sweeps on meshes with cyclic dependencies

When solving the first-order form of the linear Boltzmann equation, a common misconception is that the matrix-free computational method of “sweeping the mesh”, used in conjunction with the Discrete Ordinates method, is too complex or does not scale well enough to be implemented in modern high performance computing codes. This has led to considerable efforts in the development of matrix-based methods that are computationally expensive and is partly driven by the requirements placed on modern spatial discretizations. In particular, modern transport codes are required to support higher order elements, a concept that invariably adds a lot of complexity to sweeps because of the introduction of cyclic dependencies with curved mesh cells. In this article we will present a comprehensive implementation of sweeping, to a piecewise-linear DFEM spatial discretization with particular focus on handling cyclic dependencies and possible extensions to higher order spatial discretizations. We find that these methods are implemented in a new C++ simulation framework called Chi-Tech (). We present some typical simulation results with some performance aspects that one can expect during real world simulations, we also present a scaling study to >100k processes where Chi-Tech maintains greater than 80% efficiency solving a total of 87.7 trillion angular flux unknowns for a 116 group simulation.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2022

Atomistic and cluster dynamics modeling of fission gas (Xe) diffusivity in TRISO fuel kernels

TRISO fuel particles are candidates for use in next generation reactors including gas reactors, fluoride salt-cooled high temperature reactors, and micro-reactors. The UCO fuel kernel consists of a uranium dioxide (UO) and uranium carbide mixture. The addition of UC helps suppress the formation of carbon monoxide gas, which led to failures during initial TRISO development. The addition of uranium carbide alters the chemistry of the UO kernel, which is known to influence performance parameters such as fission gas diffusivity, although the impact has not been quantified and no models exist that take the change in chemistry into account. Therefore, better understanding and more accurate models of the impact of chemistry on fuel performance are of high priority. In this paper, a first-principles density functional theory (DFT) and empirical potential based multi-scale study has been carried out to model the diffusivity of fission gas xenon (Xe) in UCO TRISO fuel kernels. The focus is on the UO component in the UCO fuel kernels, as that represents the largest volume fraction of the fuel kernels. The study relies on DFT and empirical potential calculations to determine Xe and point defect properties, which are then used in thermodynamic and kinetic models to predict diffusion for intrinsic conditions. In addition, the information is utilized in cluster dynamics simulations using the Centipede code to estimate the impact of irradiation on defect transport. Additionally, the presence of UC or UC in the UCO fuel kernels is shown to have a substantial impact on the UO non-stoichiometry by inducing oxygen vacancies and driving UO sub-stoichiometric, which causes much slower Xe diffusion in UCO compared to light water reactor UO fuel. The application of this model in fuel performance simulations using the Bison code is also demonstrated.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2024

Accelerating Multivariate Functional Approximation Computation with Domain Decomposition Techniques⋆

Modeling large datasets through Multivariate Functional Approximations (MFA) provide an elegant way to handle many visualization and scientific analysis workflows. The process necessitates scalable data partitioning methods to compute MFA representations efficiently without compromising the accuracy or continuity of the reconstructed solution. We propose a domain -decomposed method for computing the MFA with B -spline bases, which reduces the total work per task and uses a restricted Additive Schwarz (RAS) method to converge the control point data degrees -of -freedom along subdomain boundaries. We provide an in-depth analysis of the parallel approach with domain decomposition solvers, aiming to minimize local subdomain error residuals and recover high -order continuity at subdomain interfaces with appropriate choices of knot overlaps. The communication cost, determined by the overlap regions in the RAS implementation, is optimized to recover the numerical error profile of the single subdomain case. Our proposed method stands in contrast to previous methods, which typically only recover either C 0 or at best C 1 continuity for arbitrary B -spline degree expansions, or those that require post -processing to blend discontinuities in the reconstructed data. We demonstrate the effectiveness of our approach using analytical and real -world datasets in 1D, 2D, and 3D through both strong and weak scaling studies. The performance results indicate that the overall cost of computing the approximation is directly proportional to the underlying nearest -neighbor communication implementation, and is only weakly dependent on the overlap region size that determines the size of the messages. This finding underscores the efficiency and scalability of our proposed method, making it a promising solution for handling large datasets in scientific workflows.

additive Schwarz solvers↗
DOE OSTI2020

Revealing Molecular Mechanisms in Hierarchical Nanoporous Carbon via Nuclear Magnetic Resonance

Hierarchical nanoporous carbons (HNC) have been proven to be an effective adsorbent for the adsorption of volatile organic compounds (VOCs) and CO 2 . However, questions remain regarding the hierarchical structure regulation, the adsorption mechanisms of adsorbate uptake, and interactions within the HNC. Here, we synthesize HNC from wood, using a microwave-induced heating method incorporating K 2 CO 3 activation. Our HNC exhibit Murray's law multiscale structures, prompting a molecular-scale study of adsorbate adsorption via nuclear magnetic resonance (NMR). NMR chemical shifts are consistent with ring-current effects from the adsorbent. Our NMR technique provides a convenient way to quantitate adsorption of adsorbate in HNC. VOC vapor adsorption results show NMR chemical-shift changes with time, suggesting initial adsorption into mesopores, followed by diffusion into micropores. Schroeder's paradox is demonstrated by differences in observed shifts for adsorbed liquid vis-à-vis vapor phase in these HNC. These HNC show high CO 2 adsorption capacity, portending applications to carbon capture.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Scalable line and plane relaxation in a parallel structured multigrid solver

The efficient solution of sparse, linear systems that arise through the discretization of partial differential equations remains a key challenge for a range of high performance scientific simulations. One approach for reducing data movement and improving performance is by exposing and exploiting structure in a problem through the use of robust structured multilevel solvers. By choosing coarsening that preserves the structure of the problem, these methods maintain efficient structured computation and communication throughout the multigrid hierarchy. However, when coarsening is not permitted to be dependent on the operator, anisotropy must be addressed by the smoother — producing error compatible for coarse-grid correction with structured coarsening. Here, the components required in a scalable parallel structured solver are described with a focus on memory and communication efficiency of robust smoothers. While the implementation of communication and memory reduction techniques in smoothers integrated in a complete 3D solver present a significant engineering challenge, a novel approach is proposed that addresses these challenges systematically through a change to the solver’s execution model. Enabled by user-level threading paired with a set of data and communication abstractions, this approach permits seamless aggregation of communication in plane smoothers — directly reusing code for a 2D distributed multilevel cycle. Results show an effective reduction in communication costs for coarse-grid problems, and result in a speedup of 8.7x in smoothing routines shown in Fig. 12 using this approach. This produces a significant improvement to strong scalability while maintaining favorable weak scaling behavior. Finally, a parallel scaling study using a series of refined meshes is included that demonstrates the effectiveness of this approach in an application of interest.

97 MATHEMATICS AND COMPUTING↗