Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “SCALING”

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 973 records · Page 54
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

Managing Risks via Small/Pilot Scale Testing for INL's Integrated Waste Treatment Unit (IWTU) - 19268

In 2005, the THOR steam reforming fluidized-bed reactor (FBR) technology was selected by the U.S. Department of Energy (DOE) as the technology of choice for treatment of sodium- bearing waste (SBW) at the INL. The basis of the process was to use fluidized bed technology to convert an acidic radioactive nitrate-rich liquid waste into a solid carbonate product suitable for long-term disposal and storage. During initial commissioning, as well as subsequent start-ups, the plant faced many operational challenges, such as stagnant/de-fluidized bed, and wall-scale formation. Such operational issues prompted the original contractor to seek expertise in pilot-scale testing, thereby reaching out to Hazen Research Inc. (HRI, Golden, Co). Several tests were conducted, but the results did not seem to address the underlying issues observed in the full-scale plant, as evidenced by continued operational issues. A systematic approach was then taken by the Fluor-Idaho team (new contractor as of June 2016) with respect to solving these technical issues. First, two different bench-scale, 2-inch fluidized bed reactors (stainless-steal and quartz), were utilized to test several theories to better understand the underlying chemical reactions that may be taking place under expected normal operating conditions. Then, the team designed and built an 18-inch FBR unit, mimicking the full-scale main vessel at IWTU - Denitration Mineralization Reformer (DMR). This was used to test various aspects of operational challenges observed at IWTU, namely particle size control, wall-scale growth-rate control, long-term operational stability, process optimization, and process off- gas filtration challenges. An example of information obtained from bench-scale testing, which was then tested during pilot-scale testing, was the phenomena that an addition of 5 vol % of CO{sub 2} gas into the fluidizing gas system nearly eliminated the formation of agglomerates that had typically led to bed de-fluidization. This change proved to be correct in both the pilot and full-scale testing and will be implemented as part of a long-term operational strategy of IWTU. Several other findings that have provided answers with respect to IWTU operational issues will be discussed in this report. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2016

Analysis on the Heat Transfer Phenomena Inside the Scaled-Down VHTR RCCS Riser Experiment

Very High Temperature gas-cooled Reactor (VHTR) that is one of GEN-IV reactors has been developed for the purpose of hydrogen production. VHTR is characterized by high operating temperature and stepped-up inherent safety given by passive safety systems such as Reactor Cavity Cooling System (RCCS). In case the active heat removal system losses its capability, RCCS significantly contributes to the inherent safety of VHTR by removing the decay heat transferred from the reactor core to the reactor vessel. Therefore, the performance of RCCS is the key parameter to ensure the safety of the reactor. Several computation and experiment researches on the performance of RCCS have been conducted. There are two types of RCCS depending on its coolant; one is the air-cooled type and the other is water-cooled type. This study focused on the former type. Due to the excessive height of the prototype RCCS, experiment facilities were constructed in reduced scale. Korea Atomic Energy Research Institute (KAERI) constructed the quarter-scale RCCS experiment facility, NACEF and Argonne National Laboratory (ANL) has constructed half-scale experiment facility, NSTF. University of Wisconsin (UW) also has the quarter-scale experiment facility and Texas A and M University has conducted separate effect test on the jet effect at the upper plenum of RCCS. From the reduced scale experiment result, the performance of RCCS such as a heat removal rate of RCCS and a peak temperature of reactor vessel were evaluated. When the reduced-scale result is applied to the prototype, however, a distortion could be brought about due to the reduced height of RCCS in the experiments. Especially, if the heat transfer regime inside the riser which roles as a heat exchanger to discharge the decay heat into the atmosphere is not identical, the riser wall temperature and the consequent reactor vessel temperature of the prototype could be incorrectly predicted. Thus, the convective heat transfer phenomena inside the riser duct both in the reduced scale and in the prototype is necessary to be investigated. In this study, the preliminary analysis on the heat transfer phenomena inside the RCCS riser was conducted. The subjects of the analysis were the prototype RCCS riser of PMR 200, and its scaled-down experiment facilities, NACEF. First, the geometry, flow and thermal conditions of the riser in PMR200 RCCS and NACEF were introduced with a scaling method. And then, the heat transfer regimes and the heat transfer coefficient corresponding to each condition were predicted using correlation and CFD calculation. Finally, planned activities including the experiment was introduced. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2015

Validation of Scale 6.2 criticality calculations using KENO V.A and KENO-VI - 14603

SCALE 6.2 provides enhancements for criticality calculations relative to SCALE 6.1, especially with updated nuclear data libraries, resonance self-shielding options for multigroup (MG) calculations, and problem-dependent temperature treatments for continuous-energy (CE) calculations. Benchmark experiments from the International Criticality Safety Benchmark Evaluation Project that are available in the Oak Ridge National Laboratory Verified, Archived, Library of Inputs and Data (VALID) were applied in this validation effort to examine the impact of the SCALE 6.2 enhancements. Over 300 cases are drawn from the VALID library for testing KENO V.a, and a more limited set of approximately 50 configurations is used for validation of KENO-VI. A total of 401 unique critical experiments are used in this validation effort. In some cases, the KENO V.a models were converted to KENO-VI models so that the performance of both codes could be assessed for the same systems. Validation results for the SCALE 6.2 ENDF/B-VII.1 nuclear data libraries are presented, including MG neutron data in 56- and 252-group structures as well as CE data. The performance of all three multipurpose neutron libraries in calculating k{sub eff} is examined with critical experiment models covering a wide range of fuels, moderators, reflectors, absorbers, and neutron spectra. The SCALE 6.2 ENDF/B-VII.1 coupled neutron-photon library, primarily developed for shielding applications with 200 neutron groups and 47 gamma groups, provides additional energy resolution for fast systems relative to the MG neutron libraries which were developed to produce low biases for thermal neutron systems. The applicability of this library to criticality calculations for fast neutron spectrum systems is not investigated here but will be addressed in the final SCALE 6.2 validation report. The results of these calculations demonstrate that SCALE 6.2 generally provides lower biases than SCALE 6.1 and that the two versions of KENO provide equivalent results on the same suite of problems. SCALE 6.2 also includes an updated neutron cross-section covariance library based on the ENDF/B-VII.1 covariance data. The effect of these updated covariances on the uncertainty in k{sub eff} due to cross section uncertainties is also presented for some cases. A more thorough investigation of the updated covariance data is presented in a companion paper. (authors)

42 ENGINEERING↗
DOE OSTI2015

Pore-Scale Process Coupling and Effective Surface Reaction Rates in Heterogeneous Subsurface Materials

This manuscript provides a review of pore-scale researches in literature including experimental and numerical approaches, and scale-dependent behavior of geochemical and biogeochemical reaction rates in heterogeneous porous media. A mathematical equation that can be used to predict the scale-dependent behavior of geochemical reaction rates in heterogeneous porous media has been derived. The derived effective rate expression explicitly links the effective reaction rate constant to the intrinsic rate constant, and to the pore-scale variations in reactant concentrations in porous media. Molecular simulations to calculate the intrinsic rate constants were provided. A few examples of pore-scale simulations were used to demonstrate the application of the equation to calculate effective rate constants in heterogeneous materials. The results indicate that the deviation of effective rate constant from the intrinsic rate in heterogeneous porous media is caused by the pore-scale distributions of reactants and their correlation, which are affected by the pore-scale coupling of reactions and transport.

Environmental Molecular Sciences Laboratory↗
DOE OSTI2019

Multi-scale topology optimization of multi-material structures with controllable geometric complexity — Applications to heat transfer problems

This study presents a topology optimization method to design assemblies of periodic cellular materials with controllable geometric complexity. The framework is based on a novel multi-scale and multi-material design model in which the structure, the layout of the material subdomains, and their micro-structures are optimized concurrently. To allow a tight control over its geometrical complexity, the layout at the macro-scale is described by level-set fields, parameterized by geometric primitives. The micro-scale geometry is represented through a density approach. A nonlinear programming algorithm drives the optimization process using design sensitivities computed by the discrete adjoint method. The proposed design framework is studied with heat transfer problems. Practical design problems, such as a heat sink and a thermal storage unit with phase change are discussed. The macro-scale analysis model relies on a generalized transient diffusion equation. At the micro-scale, homogenization is used to compute equivalent material properties. The numerical examples in this study show that the optimized multi-scale multi-material layouts outperform the corresponding mono-scale structures while maintaining geometric simplicity at the macro-scale level.

42 ENGINEERING↗
DOE OSTI2018

Prevention and management of silica scaling in membrane distillation using pH adjustment

Membrane scaling by silica is a major challenge in desalination, particularly for inland desalination of brackish groundwater or geothermal resources, which often contain high concentrations of silica and dissolved solids. Adjustment of feed pH may reduce silica scaling risk, which is important for inland facilities that operate at high water recoveries to reduce brine disposal costs. However, water recovery of reverse osmosis is also limited due to increased osmotic pressure with feed water concentration. Membrane distillation (MD) is a thermally driven membrane desalination technique that is not limited by increased osmotic pressure of the feed. In this investigation, pH adjustment was tested as a strategy to reduce silica scaling risk in the MD process. With feed water pH less than 5 or higher than 10, scaling impacts were negligible at silica concentrations up to 600 mg/L. Scaling rates were highest at neutral pH between 6 and 8. Cleaning strategies were also explored to remove silica scale from membranes. Cleaning using NaOH solutions at pH higher than 11 to induce dissolution of silica scale was effective at temporarily restoring performance; however, some silica remained on membrane surfaces and scaling upon re-exposure to supersaturated silica concentrations occurred faster than with new membranes.

15 GEOTHERMAL ENERGY↗
DOE OSTI2017

Using high spatial resolution satellite imagery to map forest burn severity across spatial scales in a Pine Barrens ecosystem

As a primary disturbance agent, fire significantly influences local processes and services of forest ecosystems. Although a variety of remote sensing based approaches have been developed and applied to Landsat mission imagery to infer burn severity at 30 m spatial resolution, forest burn severity have still been seldom assessed at fine spatial scales (≤ 5 m) from very-high-resolution (VHR) data. Here we assessed a 432 ha forest fire that occurred in April 2012 on Long Island, New York, within the Pine Barrens region, a unique but imperiled fire-dependent ecosystem in the northeastern United States. The mapping of forest burn severity was explored here at fine spatial scales, for the first time using remotely sensed spectral indices and a set of Multiple Endmember Spectral Mixture Analysis (MESMA) fraction images from bi-temporal — pre- and post-fire event — WorldView-2 (WV-2) imagery at 2 m spatial resolution. We first evaluated our approach using 1 m by 1 m validation points at the sub-crown scale per severity class (i.e. unburned, low, moderate, and high severity) from the post-fire 0.10 m color aerial ortho-photos; then, we validated the burn severity mapping of geo-referenced dominant tree crowns (crown scale) and 15 m by 15 m fixed-area plots (inter-crown scale) with the post-fire 0.10 m aerial ortho-photos and measured crown information of twenty forest inventory plots. Our approach can accurately assess forest burn severity at the sub-crown (overall accuracy is 84% with a Kappa value of 0.77), crown (overall accuracy is 82% with a Kappa value of 0.76), and inter-crown scales (89% of the variation in estimated burn severity ratings (i.e. Geo-Composite Burn Index (CBI)). Lastly, this work highlights that forest burn severity mapping from VHR data can capture heterogeneous fire patterns at fine spatial scales over the large spatial extents. This is important since most ecological processes associated with fire effects vary at the < 30 m scale and VHR approaches could significantly advance our ability to characterize fire effects on forest ecosystems.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Multi-scale integration of satellite remote sensing improves characterization of dry-season green-up in an Amazon tropical evergreen forest

In tropical forests, leaf phenology-particularly the pronounced dry-season green-up-strongly regulates biogeochemical cycles of carbon and water fluxes. However, uncertainties remain in the understanding of tropical forest leaf phenology at different spatial scales. Phenocams accurately characterize leaf phenology at the crown and ecosystem scales but are limited to a few sites and time spans of a few years. Time-series satellite observations might fill this gap, but the commonly used satellites (e.g. MODIS, Landsat and Sentinel-2) have resolutions too coarse to characterize single crowns. To resolve this observational challenge, we used the PlanetScope constellation with a 3m resolution and near daily nadir-view coverage. We first developed a rigorous method to cross-calibrate PlanetScope surface reflectance using daily BRDF-adjusted MODIS as the reference. We then used linear spectral unmixing of calibrated PlanetScope to obtain dry-season change in the fractional cover of green vegetation (GV) and non-photosynthetic vegetation (NPV) at the PlanetScope pixel level. We used the Central Amazon Tapajos National Forest k67 site, as all necessary data (from field to phenocam and satellite observations) was available. For this proof of concept, we chose a set of 22 dates of PlanetScope measurements in 2018 and 16 in 2019, all from the six drier months of the year to provide the highest possible cloud-free temporal resolution. Our results show that MODIS-calibrated dry-season PlanetScope data (1) accurately assessed seasonal changes in ecosystem-scale and crown-scale spectral reflectance; (2) detected an increase in ecosystem-scale GV fraction (and a decrease in NPV fraction) from June to November of both years, consistent with local phenocam observations with R 2 around 0.8; and (3) monitored large seasonal trend variability in crown-scale NPV fraction. Finally, our results highlight the potential of integrating multi-scale satellite observations to extend fine-scale leaf phenology monitoring beyond the spatial limits of phenocams.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Wafer-Scale Synthesis of 2D Materials by an Amorphous Phase-Mediated Crystallization Approach

The interest in the wafer-scale growth of two-dimensional (2D) materials, including transition metal dichalcogenides (TMDCs), has been rising for transitioning from lab-scale devices to commercial-scale systems. Among various synthesis techniques, physical vapor deposition, such as pulsed laser deposition (PLD), has shown promise for the wafer-scale growth of 2D materials. However, due to the high volatility of chalcogen atoms (e.g., S and Se), films deposited by PLD usually suffer from a lack of stoichiometry and chalcogen deficiency. To mitigate this issue, excess chalcogen is necessary during the deposition, which results in problems like uniformity or not being repeatable. This study demonstrates a condensed-phase or amorphous phase-mediated crystallization (APMC) approach for the wafer-scale synthesis of 2D materials. This method uses a room-temperature PLD process for the deposition and formation of amorphous precursors with controlled thicknesses, followed by a post-deposition crystallization process to convert the amorphous materials to crystalline structures. This approach maintains the stoichiometry of the deposited materials throughout the deposition and crystallization process and enables the large-scale synthesis of crystalline 2D materials (e.g., MoS 2 and WSe 2 ) on Si/SiO 2 substrates, which is critical for future wafer-scale electronics. We show that the thickness of the layers can be digitally controlled by the number of laser pulses during the PLD phase. Optical spectroscopy is used to monitor the crystallization dynamics of amorphous layers as a function of annealing temperature. The crystalline quality, domain sizes, and the number of layers were explored using nanoscale and atomistic characterization (e.g., AFM, STEM, and EDS) along with electrical characterization to explore process–structure–performance relationships. Finally, this growth technique is a promising method that could potentially be adopted in conventional semiconductor industries for wafer-scale manufacturing of next-generation electronic and optoelectronic devices.

2D electronics↗
DOE OSTI2018

Reducing Coercive-Field Scaling in Ferroelectric Thin Films via Orientation Control

The desire for low-power/voltage operation of devices is driving renewed interest in understanding scaling effects in ferroelectric thin films. As the dimensions of ferroelectrics are reduced, the properties can vary dramatically, including the robust scaling relationship between coercive field ( E c ) and thickness ( d ), also referred to as the Janovec-Kay-Dunn (JKD) law, wherein E c ∝ d -2/3 . Here, we report that whereas (001)-oriented heterostructures follow JKD scaling across the thicknesses range of 20-330 nm, (111)-oriented heterostructures of the canonical tetragonal ferroelectric PbZr 0.2 Ti 0.8 O 3 exhibit a deviation from JKD scaling wherein a smaller scaling exponent for the evolution of E c is observed in films of thickness ≤ 165 nm. X-ray diffraction reveals that whereas (001)-oriented heterostructures remain tetragonal for all thicknesses, (111)-oriented heterostructures exhibit a transition from tetragonal-to-monoclinic symmetry in films of thickness ≤ 165 nm as a result of the compressive strain. First-principles calculations suggest that this symmetry change contributes to the deviation from the expected scaling, as the monoclinic phase has a lower energy barrier for switching. This structural evolution also gives rise to changes in the c/a lattice parameter ratio, wherein this ratio increases and decreases in (001)- and (111)-oriented heterostructures, respectively, as the films are made thinner. In (111)-oriented heterostructures, this reduced tetragonality drives a reduction of the remanent polarization and, therefore, a reduction of the domain-wall energy and overall energy barrier to switching, which further exacerbates the deviation from the expected scaling. Altogether, this work demonstrates a route toward reducing coercive fields in ferroelectric thin films and provides a possible mechanism to understand the deviation from JKD scaling.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Unexpected Changes of Aerosol Burdens With Decreased Convection in the Context of Scale-Aware Convection Schemes

Atmospheric aerosols in global climate models (GCMs) can be removed by both large-scale and convective precipitation. As the horizontal resolution of GCMs increases, it is expected that large-scale precipitation increases and convective precipitation decreases if scale-aware convection schemes are well developed. Here, to explore its impact on aerosol burdens, this study develops a novel method and applies it to the National Center for Atmospheric Research Community Atmosphere Model version 5.3 to mimic the behavior of scale-aware convective schemes. Surprisingly, it is found that aerosol burden increases globally as convective precipitation decreases and large-scale precipitation increases, which can be attributed to the reduced light rain. In the light-rain intensity category, the decreased frequency of convective precipitation is not offset by the increased frequency of large-scale precipitation. Therefore, for aerosol simulations in high-resolution GCMs using scale-aware convection schemes, subgrid convection and large-scale condensation should be coordinated to assure light rain stays the same.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Allometric Scaling of Hyporheic Respiration Across Basins in the Pacific Northwest United States

Abstract Hyporheic zones regulate biogeochemical processes in streams and rivers, but high spatiotemporal heterogeneity makes it difficult to predict how these processes scale from individual reaches to river basins. Recent work applying allometric scaling (i.e., power‐law relationships between size and function) to river networks provides a new paradigm for understanding cumulative hyporheic biogeochemical processes. We used previously published model predictions of reach‐scale hyporheic aerobic respiration to explore patterns in allometric scaling across two climatically divergent basins with differing characteristics in the Pacific Northwest, United States. In the model, hydrologic exchange fluxes (HEFs) regulate hyporheic respiration, so we examined how HEFs might influence allometric scaling of respiration. We found consistent scaling behaviors where HEFs were either very low or very high, but differences between basins when HEFs were moderate. Our findings provide initial model‐generated hypotheses for factors influencing allometric scaling of hyporheic respiration. These hypotheses can be used to optimize new data generation efforts aimed at developing predictive understanding of allometries that can, in turn, be used to scale biogeochemical dynamics across watersheds.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2026

Empirical scaling of the L–H threshold power for metal wall tokamaks using a multi-device database

The empirical scaling for the H-mode power threshold in tokamaks has been revisited using a database with threshold data from machines with a metallic first wall as part of International Tokamak Physics Activity (ITPA) task TC-26. The database contains discharges from ASDEX Upgrade (AUG) (W), JET (Be/W) and Alcator C-Mod (Mo). This was motivated by reports that in like-for-like discharges the power threshold was reduced by approximately 30% after the change from carbon based to metallic first wall materials on AUG (Ryter et al 2013 Nucl. Fusion 53 113003) and JET (Maggi et al 2014 Nucl. Fusion 54 023007). The database contains L–H transition data for all hydrogen isotopes and mixtures, including T and DT from the recent JET campaigns. Compared to the ITPA 2008 scaling (Martin et al 2008 J. Phys.: Conf. Ser. 123 012033), the metal wall scaling has a smaller magnetic field exponent but a larger density exponent. We present an additional parameter to capture the strong dependence of the L–H power threshold (approx. factor 2) on the magnetic configuration in the divertor on JET. The scaling recovers the approximate inverse isotope mass scaling of the threshold power. Alternative scalings involving the plasma current and poloidal magnetic field are explored. Despite the reduction in threshold observed earlier, the scalings based on the metal wall database do not necessarily extrapolate to a lower threshold for ITER compared to the ITPA 2008 scaling, especially at high density. The divertor configuration effect induces the largest uncertainty in the extrapolation.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2023

Non-local contribution from small scales in galaxy–galaxy lensing: comparison of mitigation schemes

ABSTRACT Recent cosmological analyses with large-scale structure and weak lensing measurements, usually referred to as 3 × 2pt, had to discard a lot of signal to noise from small scales due to our inability to accurately model non-linearities and baryonic effects. Galaxy–galaxy lensing, or the position–shear correlation between lens and source galaxies, is one of the three two-point correlation functions that are included in such analyses, usually estimated with the mean tangential shear. However, tangential shear measurements at a given angular scale θ or physical scale R carry information from all scales below that, forcing the scale cuts applied in real data to be significantly larger than the scale at which theoretical uncertainties become problematic. Recently, there have been a few independent efforts that aim to mitigate the non-locality of the galaxy–galaxy lensing signal. Here, we perform a comparison of the different methods, including the Y-transformation, the point-mass marginalization methodology, and the annular differential surface density statistic. We do the comparison at the cosmological constraints level in a combined galaxy clustering and galaxy–galaxy lensing analysis. We find that all the estimators yield equivalent cosmological results assuming a simulated Rubin Observatory Legacy Survey of Space and Time (LSST) Year 1 like set-up and also when applied to DES Y3 data. With the LSST Y1 set-up, we find that the mitigation schemes yield ∼1.3 times more constraining S8 results than applying larger scale cuts without using any mitigation scheme.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Multi-Time Scale Coordinated Control and Scheduling of Inverter-Based TCLs With Variable Wind Generation

To address microgrid tie flow errors caused by wind generation variability, here we propose and develop a multi-time scale coordinated control and scheduling strategy for inverter-based thermostatically controlled loads (TCLs). First, in hour-time scale, inverter-based TCLs with adjusting temperature set-point are modeled as virtual generators to compensate tie flow deviations in the day-ahead plan. Next, in minute-time scale, virtual batteries representing operating behaviors of inverter-based TCLs with frequency control are scheduled determined by the control of virtual generators in hour-time scale. The virtual batteries are scheduled to smooth out tie flow errors corresponding to day-ahead plan and hour-time scale schedules. The multi-time scale control methods are coordinated to employ the response potential of inverter-based TCLs and response curve-based methods are proposed to control inverter-based TCLs considering the customer privacy. The multi-time scale stochastic schedules which are based on response curves of inverter-based TCLs are coordinated to accommodate wind generation variability. Simulation results demonstrate that the microgrid tie flow errors are effectively mitigated by the proposed multi-time scale coordinated control and scheduling of inverter-based TCLs.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2022

SMALE: Enhancing Scalability of Machine Learning Algorithms on Extreme-Scale Computing Platforms

Deployment and execution of machine learning tasks on extreme-scale computing platforms face several significant technical challenges: 1) High computing cost incurred by dense networks – The computing workload of deep networks with densely-connected topology increases rapidly with the network size, imposing a non-scalable computing model of extreme-scale computing platforms; 2) Non-optimized workload distribution – Many advanced deep learning algorithms, e.g., sparsification and irregular net-work topology, produce very unbalanced workload distribution on extreme-scale computing platforms. The computation efficiency is greatly hindered by the incurred data and computation redundancies as well as long tails of the node with extensive workload; 3) Constraints in data movement and I/O bottle-neck – Inter-node data movement in extreme-scale computing platforms are associated with high energy and latency costs, and subject to the constraints of I/O bandwidth; and 4) Generalization of algorithm realization and acceleration on computing platforms – The large varieties of machine learning algorithms and structures of extreme-scale computing platforms make the derivation of a generalized algorithm realization and acceleration method very challenging, which, however, is the requirement by domain scientists and interested users. We call the above challenges Smale’s Problems in Machine Learning and Understanding for High-Performance Computing Scientific Discovery. The objective of our three-year research project is to develop a holistic innovation set at structure, assembly, and acceleration layers of machine learning algorithms to address the above challenges in algorithm deployment and execution. Three tasks are particularly performed, including: At the algorithm structure level, we investigate the techniques that can structurally sparsify on the topology of deep networks for computing workload reduction. We also study clustering and pruning techniques that can optimize the workload distributions over the extreme-scale computing platforms; At the algorithm assembly level, we derive a unified learning framework for unsupervised transfer learning and dynamic growing capabilities. Novel training methods are also exploited to enhance the training efficiency of the proposed framework; At the algorithm acceleration level, we will develop a series of techniques that can accelerate the computation of sparse matrix operations, which are one of the core executions in deep learning and optimize memory access of the concerned platforms. Our proposed techniques attack the fundamental problems in machine learning algorithms running on extreme-scale computing platforms by vertically integrating the solutions at three closely entangled layers, paving the long-term scaling path of machine learning applications under DOE context. Three tasks corresponding to the above respective research orientations are performed during the three-year project period with our collaborators at ORNL. The outcome of the proposed project is anticipated to form a holistic solution set of novel algorithms and network topologies, efficient training techniques, and fast acceleration methods to promote the computing scalability of the machine learning applications of particular interest to DOE.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Calcium Sulfate and Calcium Carbonate Scaling of Thin-Film Composite Polyamide Reverse Osmosis Membranes with Surface-Tethered Polyacrylic Acid Chains

The gypsum and calcite scaling propensities of the thin-film composite polyamide (PA-TFC) reverse osmosis (RO) membrane, modified with a tethered surface layer of polyacrylic acid (PAA) chains, was evaluated and compared to the scaling of selected commercial RO membranes. The tethered PAA layer was synthesized onto a commercial polyamide membrane (i.e., base-PA) via atmospheric pressure plasma-induced graft polymerization (APPIGP). The PAA nano-structured (SNS) base-PA membrane (SNS-PAA-PA) was scaled to a lesser degree, as quantified by a lower permeate flux decline and surface imaging, relative to the tested commercial membranes (Dow SW30, Toray SWRO, and BWRO). The cleaning of gypsum-scaled membranes with D.I. water flushing achieved 100% water permeability recovery for both the SNS-PAA-PA and Dow SW30 membranes, relative to 92–98% permeability restoration for the Toray membranes. The calcium carbonate scaling of SNS-PAA-PA membranes was also lower relative to the commercial membranes, but permeability recovery after D.I. water cleaning was somewhat lower (94%) but consistent with the level of surface scale coverage. In contrast, the calcite and gypsum-scaled membrane areas of the commercial membranes post-cleaning were significantly higher than for the SNS-PAA-PA membrane but with 100% permeability recovery, suggesting the potential for membrane damage when mineral scaling is severe.

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

Error Characteristics and Scale Dependence of Current Satellite Precipitation Estimates Products in Hydrological Modeling

Satellite precipitation estimates (SPEs) are promising alternatives to gauge observations for hydrological applications (e.g., streamflow simulation), especially in remote areas with sparse observation networks. However, the existing SPEs products are still biased due to imperfections in retrieval algorithms, data sources and post-processing, which makes the effective use of SPEs a challenge, especially at different spatial and temporal scales. In this study, we used a distributed hydrological model to evaluate the simulated discharge from eight quasi-global SPEs at different spatial scales and explored their potential scale effects of SPEs on a cascade of basins ranging from approximately 100 to 130,000 km 2 . The results indicate that, regardless of the difference in the accuracy of various SPEs, there is indeed a scale effect in their application in discharge simulation. Specifically, when the catchment area is larger than 20,000 km2, the overall performance of discharge simulation emerges an ascending trend with the increase of catchment area due to the river routing and spatial averaging. Whereas below 20,000 km 2 , the discharge simulation capability of the SPEs is more randomized and relies heavily on local precipitation accuracy. Our study also highlights the need to evaluate SPEs or other precipitation products (e.g., merge product or reanalysis data) not only at the limited station scale, but also at a finer scale depending on the practical application requirements. Here we have verified that the existing SPEs are scale-dependent in hydrological simulation, and they are not enough to be directly used in very fine scale distributed hydrological simulations (e.g., flash flood). More advanced retrieval algorithms, data sources and bias correction methods are needed to further improve the overall quality of SPEs.

DTVGM↗