Search NASA⌕ Search

SEARCH · Search NASA

Results for “data reduction techniques”

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 793 records · Page 44

Scalable training of graph convolutional neural networks for fast and accurate predictions of HOMO-LUMO gap in molecules

Abstract Graph Convolutional Neural Network (GCNN) is a popular class of deep learning (DL) models in material science to predict material properties from the graph representation of molecular structures. Training an accurate and comprehensive GCNN surrogate for molecular design requires large-scale graph datasets and is usually a time-consuming process. Recent advances in GPUs and distributed computing open a path to reduce the computational cost for GCNN training effectively. However, efficient utilization of high performance computing (HPC) resources for training requires simultaneously optimizing large-scale data management and scalable stochastic batched optimization techniques. In this work, we focus on building GCNN models on HPC systems to predict material properties of millions of molecules. We use HydraGNN, our in-house library for large-scale GCNN training, leveraging distributed data parallelism in PyTorch. We use ADIOS, a high-performance data management framework for efficient storage and reading of large molecular graph data. We perform parallel training on two open-source large-scale graph datasets to build a GCNN predictor for an important quantum property known as the HOMO-LUMO gap. We measure the scalability, accuracy, and convergence of our approach on two DOE supercomputers: the Summit supercomputer at the Oak Ridge Leadership Computing Facility (OLCF) and the Perlmutter system at the National Energy Research Scientific Computing Center (NERSC). We present our experimental results with HydraGNN showing (i) reduction of data loading time up to 4.2 times compared with a conventional method and (ii) linear scaling performance for training up to 1024 GPUs on both Summit and Perlmutter.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Lensing reconstruction in post-Born cosmic microwave background weak lensing

The study of the cosmic microwave background (CMB) lensing potential has established itself by now as a robust way of probing the physics of large-scale structure growth. The most common estimators of the lensing potential are derived under the assumption of Gaussianity of the matter distribution and in the Born approximation of the photon diffusion. In this paper we study the performance of quadratic estimators when applied to realistic sky maps extracted from multiple-lens ray tracing techniques in cosmological $N$-body simulations. These are expected to model accurately the effects due to the non-Gaussianity of the matter distribution induced by its nonlinearity and the deviation from the Born approximation. We show that both these effects on their own lead to reconstruction biases, but these tend to partially cancel each other when both these effects are considered together. We forecast the impact of these biases on the estimation of cosmological parameters for future high-sensitivity CMB experiments like CMB-S4. We find that the cold dark matter density, $Ω_{cdm}$, the optical depth to reionization $τ$, the amplitude of primordial inflationary perturbations, $A_s$, and the sum of neutrino masses, $M_ν$, could be biased at the $1–2σ$ level, if no external data set is used. We also observe a reduction of the bias if external data like baryon acoustic oscillations is included.

79 ASTRONOMY AND ASTROPHYSICS↗

Enhancing I/O Throughput via Efficient Routing and Placement for Large-scale Parallel File Systems

As storage systems get larger to meet the demands of petascale systems, careful planning must be applied to avoid congestion points and extract the maximum performance. In addition, the large data sets generated by such systems makes it desirable for all compute resources to have common access to this data without needing to copy it to each machine. This paper describes a method of placing I/O close to the storage nodes to minimize contention on Cray s SeaStar2+ network, and extends it to a routed Lustre configuration to gain the same benefits when running against a center-wide file system. Our experiments using half of the resources of Spider the center-wide file system at the Oak Ridge Leadership Computing Facility show that I/O write bandwidth can be improved by up to 45% (from 71.9 to 104 GB/s) for a direct-attached configuration and by 137% (47.6 GB/s to 115 GB/s) for a routed configuration. We demonstrated up to 20.7% reduction in run-time for production scientific applications. With the full Spider system, we demonstrated over 240 GB/s of aggregate bandwidth using our techniques.

99 GENERAL AND MISCELLANEOUS↗

A comparison of neural network architectures for data-driven reduced-order modeling

The popularity of deep convolutional autoencoders (CAEs) has engendered new and effective reduced-order models (ROMs) for the simulation of large-scale dynamical systems. Despite this, it is still unknown whether deep CAEs provide superior performance over established linear techniques or other network-based methods in all modeling scenarios. To elucidate this, the effect of autoencoder architecture on its associated ROM is studied through the comparison of deep CAEs against two alternatives: a simple fully connected autoencoder, and a novel graph convolutional autoencoder. Through benchmark experiments, it is shown that the superior autoencoder architecture for a given ROM application is highly dependent on the size of the latent space and the structure of the snapshot data, with the proposed architecture demonstrating benefits on data with irregular connectivity when the latent space is sufficiently large.

42 ENGINEERING↗

Hybrid data-driven cement-stabilized soil design: An integration of machine learning, multi-objective optimization, and life cycle assessment

Soil stabilization is crucial in geotechnical engineering, yet conventional methods are often time-consuming, resource-intensive, and environmentally unsustainable. Despite growing interest in Machine Learning (ML) and optimization tools for mix design, few studies integrate these methods with decision-making techniques and environmental assessment to support practical implementation. This study proposes a hybrid data-driven framework for predicting strength, optimizing mix compositions, and evaluating environmental impacts via life cycle assessment of cement-stabilized soft soils. Six ML models were evaluated, and the top-performing eXtreme Gradient Boosting (XGB) model was further improved using the Grey Wolf Optimizer (GWO). The optimized XGB-GWO model, integrated with a polynomial cost function, served as the objective function in a multi-objective optimization problem solved via the Non-Dominated Sorting Genetic Algorithm II (NSGA-II), with final mix selection guided by the entropy-weighted TOPSIS method. Validation through a case study produced mix designs offering superior strength-cost trade-offs, with the optimal mix achieving 2243.2 kPa unconfined compressive strength and a 16.07 % reduction in carbon emissions compared to the highest-cost design. In conclusion, this study offers a sustainable, scalable approach to soil stabilization and supports informed decision-making in construction.

Life cycle assessment↗

Observation of transient iron(II) formation in dye-sensitized iron oxide nanoparticles by time-resolved x-ray spectroscopy.

The reduction of ferric iron in solid phase minerals leads to the mobilization of ferrous iron in the environment and is thus a crucial component of the global iron cycle. Despite the importance of this process, a mechanistic understanding of the structural and chemical changes that are caused by this electron transfer reaction is not established because the speed of the fundamental chemical steps renders them inaccessible to conventional study. Ultrafast time-resolved X-ray spectroscopy is a technique that can overcome this limitation and measure changes in oxidation state and structure occurring during chemical reactions that can be initiated by a fast laser pulse. We use this approach with {approx}100 ps resolution to monitor the speciation of Fe atoms in iron oxide nanoparticles following photoinduced electron transfer from a surface-bound photoactive dye molecule. These data represent the first direct real-time observation of the dynamics of ferrous ion formation and subsequent reoxidation in iron oxide.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

APPLICATION OF A DAMPED LOCALLY OPTIMIZED COMBINATION OF IMAGES METHOD TO THE SPECTRAL CHARACTERIZATION OF FAINT COMPANIONS USING AN INTEGRAL FIELD SPECTROGRAPH

High-contrast imaging instruments are now being equipped with integral field spectrographs (IFSs) to facilitate the detection and characterization of faint substellar companions. Algorithms currently envisioned to handle IFS data, such as the Locally Optimized Combination of Images (LOCI) algorithm, rely on aggressive point-spread function (PSF) subtraction, which is ideal for initially identifying companions but results in significantly biased photometry and spectroscopy owing to unwanted mixing with residual starlight. This spectrophotometric issue is further complicated by the fact that algorithmic color response is a function of the companion's spectrum, making it difficult to calibrate the effects of the reduction without using iterations involving a series of injected synthetic companions. In this paper, we introduce a new PSF calibration method, which we call 'damped LOCI', that seeks to alleviate these concerns. By modifying the cost function that determines the weighting coefficients used to construct PSF reference images, and also forcing those coefficients to be positive, it is possible to extract companion spectra with a precision that is set by calibration of the instrument response and transmission of the atmosphere, and not by post-processing. We demonstrate the utility of this approach using on-sky data obtained with the Project 1640 IFS at Palomar. Damped LOCI does not require any iterations on the underlying spectral type of the companion, nor does it rely on priors involving the chromatic and statistical properties of speckles. It is a general technique that can readily be applied to other current and planned instruments that employ IFSs.

79 ASTRONOMY AND ASTROPHYSICS↗

Spectral and Spatial Classification of Hyperspectral Images Based on ICA and Reduced Morphological Attribute Profiles

The availability of hyperspectral images with improved spectral and spatial resolutions provides the opportunity to obtain accurate land-cover classification. In this paper, a novel methodology that combines spectral and spatial information for supervised hyperspectral image classification is proposed. A feature reduction strategy based on independent component analysis is the main core of the spectral analysis, where the exploitation of prior information coupled to the evaluation of the reconstruction error assures the identification of the best class-informative subset of independent components. Reduced attribute profiles (APs), which are designed to address well-known issues related to information redundancy that affect the common morphological APs, are then employed for the modeling and fusion of the contextual information. Four real hyperspectral data sets, which are characterized by different spectral and spatial resolutions with a variety of scene typologies (urban, agriculture areas), have been used for assessing the accuracy and generalization capabilities of the proposed methodology. The obtained results demonstrate the classification effectiveness of the proposed approach in all different scene typologies, with respect to other state-of-the-art techniques.

42 ENGINEERING↗

Signatures of enhanced superconducting properties in niobium cavities

Superconducting radio-frequency (SRF) niobium cavities are critical for modern particle accelerators, as well as for advancing superconducting quantum systems and enabling ultrasensitive searches for new physics. In this work, we report a systematic observation of an anomalous frequency dip in Nb cavities, which occurs at temperatures just below the critical temperature (𝑇𝑐 ), indicative of enhanced superconducting properties at 𝑇 ≪ 𝑇 𝑐 . The magnitude of this dip is strongly correlated with the rf surface resistance, impurity distribution near the surface, and 𝑇𝑐 . Additionally, we report measurements of the coherence peak in the ac conductivity of two Nb SRF cavities processed using distinct methods. By comparing recent theories developed to model this experimental data, we show that the frequency-dip feature, larger coherence peak height, and reduction in the temperature-dependent surface resistance with rf current occur at minimal but finite levels of disorder.

43 PARTICLE ACCELERATORS↗

Reduced Order Modeling conditioned on monitored features for response and error bounds estimation in engineered systems

Reduced Order Models (ROMs) form essential tools across engineering domains by virtue of their function as surrogates for computationally intensive digital twinning simulators. Although purely data-driven methods are available for ROM construction, schemes that allow to retain a portion of the physics tend to enhance the interpretability and generalization of ROMs. However, physics-based techniques can adversely scale when dealing with nonlinear systems that feature parametric dependencies. This study introduces a generative physics-based ROM that is suited for nonlinear systems with parametric dependencies and is additionally able to provide numerical error bounds associated with the respective estimates. A main contribution of this work is the conditioning of these parametric ROMs to features that can be derived from monitoring measurements, feasibly in an online fashion. This is contrary to most existing ROM schemes, which remain restricted to the prescription of the physics-based, and usually a priori unknown, system parameters. Our work utilizes conditional Variational Autoencoders to continuously map the required reduction bases to a feature vector extracted from limited output measurements, while additionally allowing for a probabilistic assessment of the ROM-estimated Quantities of Interest. An auxiliary task using a neural network-based parametrization of suitable probability distributions is introduced to re-establish the link with physical model parameters. We verify the proposed scheme on a series of simulated case studies incorporating effects of geometric and material nonlinearity under parametric dependencies related to system properties and input load characteristics.

Conditional VAEs↗

Discovery of Energy Storage Molecular Materials Using Quantum Chemistry-Guided Multiobjective Bayesian Optimization

Redox flow batteries (RFBs) are a promising technology for stationary energy storage applications due to their flexible design, scalability, and low cost. In RFBs, energy is carried in flowable redoxactive materials (redoxmers) which are stored externally and pumped to the cell during operation. Further improvements in the energy density of RFBs necessitates redoxmer designs with wider redox potential windows and higher solubility. Additionally, designing redoxmers with a fluorescence-enabled self-reporting functionality allows monitoring of the state of health of RFBs. To accelerate the discovery of redoxmers with desired properties, state-of-the- art machine learning ( ML) methods, such as multiobjective Bayesian optimization (MBO), are useful. Here, we first employed density functional theory calculations to generate a database of reduction potentials, solvation free energies, and absorption wavelengths for 1400 redoxmer molecules based on a 2,1,3-benzothiadiazole (BzNSN) core structure. From the computed properties, we identified 22 Pareto-optimal molecules that represent best trade-off among all of the desired properties. We further utilized these data to develop and benchmark an MBO approach to identify candidates quickly and efficiently with multiple targeted properties. With MBO, optimal candidates from the 1400-molecule data set can be identified at least 15 times more efficiently compared to the brute force or random selection approach. Importantly, we utilized this approach for discovering promising redoxmers from an unseen database of 1 million BzNSN-based molecules, where we discovered 16 new Pareto-optimal molecules with significant improvements in properties over the initial 1400 molecules. Furthermore, we anticipate that this active learning technique is general and can be utilized for the discovery of any class of functional materials that satisfies multiple desired property criteria.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Technical Note: Phantom study to evaluate the dose and image quality effects of a computed tomography organ-based tube current modulation technique

Purpose: This technical note quantifies the dose and image quality performance of a clinically available organ-dose-based tube current modulation (ODM) technique, using experimental and simulation phantom studies. The investigated ODM implementation reduces the tube current for the anterior source positions, without increasing current for posterior positions, although such an approach was also evaluated for comparison. Methods: Axial CT scans at 120 kV were performed on head and chest phantoms on an ODM-equipped scanner (Optima CT660, GE Healthcare, Chalfont St. Giles, England). Dosimeters quantified dose to breast, lung, heart, spine, eye lens, and brain regions for ODM and 3D-modulation (SmartmA) settings. Monte Carlo simulations, validated with experimental data, were performed on 28 voxelized head phantoms and 10 chest phantoms to quantify organ dose and noise standard deviation. The dose and noise effects of increasing the posterior tube current were also investigated. Results: ODM reduced the dose for all experimental dosimeters with respect to SmartmA, with average dose reductions across dosimeters of 31% (breast), 21% (lung), 24% (heart), 6% (spine), 19% (eye lens), and 11% (brain), with similar results for the simulation validation study. In the phantom library study, the average dose reduction across all phantoms was 34% (breast), 20% (lung), 8% (spine), 20% (eye lens), and 8% (brain). ODM increased the noise standard deviation in reconstructed images by 6%–20%, with generally greater noise increases in anterior regions. Increasing the posterior tube current provided similar dose reduction as ODM for breast and eye lens, increased dose to the spine, with noise effects ranging from 2% noise reduction to 16% noise increase. At noise equal to SmartmA, ODM increased the estimated effective dose by 4% and 8% for chest and head scans, respectively. Increasing the posterior tube current further increased the effective dose by 15% (chest) and 18% (head) relative to SmartmA. Conclusions: ODM reduced dose in all experimental and simulation studies over a range of phantoms, while increasing noise. The results suggest a net dose/noise benefit for breast and eye lens for all studied phantoms, negligible lung dose effects for two phantoms, increased lung dose and/or noise for eight phantoms, and increased dose and/or noise for brain and spine for all studied phantoms compared to the reference protocol.

60 APPLIED LIFE SCIENCES↗

Determining Reduced Order Models for Optimal Stochastic Reduced Order Models

The use of parameterized reduced order models(PROMs) within the stochastic reduced order model (SROM) framework is a logical progression for both methods. In this report, five different parameterized reduced order models are selected and critiqued against the other models along with truth model for the example of the Brake-Reuss beam. The models are: a Taylor series using finite difference, a proper orthogonal decomposition of the the output, a Craig-Bampton representation of the model, a method that uses Hyper-Dual numbers to determine the sensitivities, and a Meta-Model method that uses the Hyper-Dual results and constructs a polynomial curve to better represent the output data. The methods are compared against a parameter sweep and a distribution propagation where the first four statistical moments are used as a comparison. Each method produces very accurate results with the Craig-Bampton reduction having the least accurate results. The models are also compared based on time requirements for the evaluation of each model where the Meta- Model requires the least amount of time for computation by a significant amount. Each of the five models provided accurate results in a reasonable time frame. The determination of which model to use is dependent on the availability of the high-fidelity model and how many evaluations can be performed. Analysis of the output distribution is examined by using a large Monte-Carlo simulation along with a reduced simulation using Latin Hypercube and the stochastic reduced order model sampling technique. Both techniques produced accurate results. The stochastic reduced order modeling technique produced less error when compared to an exhaustive sampling for the majority of methods.

97 MATHEMATICS AND COMPUTING↗

Solvent-Induced Swelling Behaviors of Microphase-Separated Polystyrene- block -Poly(ethylene oxide) Thin Films Investigated Using In Situ Spectroscopic Ellipsometry and Single-Molecule Fluorescence Microscopy

Block copolymers have attracted considerable interest in the fields of nanoscience and nanotechnology, because these polymers afford well-defined nanostructures via self-assembly. An in-depth understanding of solvent effects on the physicochemical properties of these microdomains is crucial for their preparation and utilization. Herein, we employed in situ spectroscopic ellipsometry and single-molecule fluorescence techniques to gain detailed insights into microdomain properties in polystyrene-block-poly(ethylene oxide) (PS-b-PEO) films exposed to ethanol- and water-saturated N 2 . We observed a quick increase and a subsequent gradual decrease in the ellipsometric thickness of PS-b-PEO films upon exposure to ethanol-saturated N 2 . This observation was unexpected because ethanol-saturated N 2 induced negligible thickness change for PS and PEO homopolymer films. The similarity in maximum thickness gain observed under ethanol- and water-saturated N 2 implied the swelling of PEO microdomains. Ethanol vapor permeation through the PEO microdomains was supported by the red-shift of the ensemble and single-molecule fluorescence emission of Nile red in PS-b-PEO films. Single-molecule tracking data showed the initial enhancement and subsequent reduction of the diffusion of hydrophilic sulforhodamine B molecules in PS-b-PEO films upon exposure to ethanol-saturated N 2 , consistent with the spectroscopic ellipsometry results. Furthermore, the higher ethanol susceptibility of the PEO microdomains was attributable to their amorphous nature, as shown by FTIR data.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Strain controlled systematic variation of metal-insulator transition in epitaxial NdNiO{sub 3} thin films

We report here the strain dependent structural and electrical transport properties of epitaxial NdNiO{sub 3} thin films. Pulsed laser deposition technique was used to grow the NdNiO{sub 3} thin films on c-axis oriented SrTiO{sub 3} single crystals. Deposited films were irradiated using 200 MeV Ag{sup 15+} ion beam at the varying fluence (1 Multiplication-Sign 10{sup 11}, 5 Multiplication-Sign 10{sup 11}, and 1 Multiplication-Sign 10{sup 12} ions/cm{sup 2}). X-ray diffraction studies confirm the epitaxial growth of the deposited films, which is maintained even up to the highest fluence. Rise in the in-plane compressive strain has been observed after the irradiation. All the films exhibit metal-insulator transition, however, a systematic decrease in the transition temperature (T{sub MI}) has been observed after irradiation, which may be attributed to the increase in the in-plane compression. Raman spectroscopy data reveal that this reduction in T{sub MI}, with the irradiation, is related to the decrease in band gap due to the stress generated by the in-plane compressive strain.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Development of a Two-fluid Drag Law for Clustered Particles using Direct Numerical Simulation and Validation through Experiments

This study presents a new drag model, based on the cohesive inter-particle forces, implemented in the MFIX code. This new drag model combines an existing standard model in MFIX with a particle-based drag model based on a switching principle. Switches between the models in the computational domain occur where strong particle-to-particle cohesion potential is detected. Three versions of the new model were obtained by using one standard drag model in each version. Later, performance of each version was compared against available experimental data for a fluidized bed, published in the literature and used extensively by other researchers for validation purposes. In our analysis of the results, we first observed that standard models used in this research were incapable of producing closely matching results. Then, we showed for a simple case that a threshold is needed to be set on the solid volume fraction. This modification was applied to avoid non-physical results for the clustering predictions, when governing equation of the solid granular temperate was solved. Later, we used our hybrid technique and observed the capability of our approach in improving the numerical results significantly; however, improvement of the results depended on the threshold of the cohesive index, which was used in the switching procedure. Our results showed that small values of the threshold for the cohesive index could result in significant reduction of the computational error for all the versions of the proposed drag model. In addition, we redesigned an existing circulating fluidized bed (CFB) test facility in order to create validation cases for clustering regime of Geldart A type particles.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Ni5Ga3 Catalysts for CO2 Reduction to Methanol: Exploring the Role of Ga Surface Oxidation/reductions on Catalytic Activity

Ad-Ni5Ga3/SiO2 catalyst, which is highly active and stable for thermal CO2 hydrogenation to methanol, was investigated to understand its surface dynamics during reaction conditions. The catalyst was prepared, tested and characterized using a multitude of techniques, including ex-situ XRD, TEM, H2-TPR, CO chemisorption and in-situ ETEM and APXPS. The formation of Ga2O3 upon air exposure was observed and the consequences on catalytic activity were investigated. Results revealed that upon air exposure Ga migrates from the subsurface region to the surface of the nanoparticles forming a Ga-oxide shell surrounding a metallic core. The oxide shell can be reduced completely at high temperatures and the temperature of the reduction activation treatment plays a crucial role on the catalytic activity. This air exposure-reduction procedure results in a redistribution of the Ga species in the surface/subsurface region, promoted by a higher Ga affinity for oxygen. XRD analysis of the post-reaction samples provide evidence for the partial transformation of a Ni5Ga3 intermetallic phase into a Ni-rich Ni3Ga phase. APXPS data indicate that the Ni-rich phase forms in the subsurface region. A Ga-rich phase exists on the surface, where the Ni/Ga ratio reaches 2/3 upon activation in hydrogen. At low reduction temperatures an amorphous Ga2O3 shell remains, indicating a promotion effect on methanol synthesis activity.

Ni-Ga, intermetallic, methanol synthesis, in-situ ↗

Review of heat transfer enhancement techniques for single phase flows

The thermal energy exchange between a flowing fluid and its confining channel is a ubiquitous process in modern society. To enhance the fluid-to-wall or wall-to-fluid heat transfer, several techniques have been developed to maximize the contact area between the fluid and the inner wall and/or disrupt the flow to enhance circulation or induce turbulence. Deployment of channels having features capable of enhancing heat transfer enables the reduction of heat exchanger size while maintaining performance. Reduction in equipment size is critical due to the ability to minimize the required volume of costly working fluids and to mitigate potential safety concerns associated with total system fluid volume. Here, a comprehensive review of single-phase heat transfer enhancement techniques is presented. The article provides a thorough comparison by analyzing the heat transfer rate, pressure drop, and other operational aspects. Single-phase heat transfer enhancement methods are divided into active and passive techniques. Active methods such as electrohydrodynamic (EHD), magnetohydrodynamics (MHD), or mechanical motion require external power to create enhancement. Passive methods such as dimples, fins, or tape inserts do not require external input and rely only on surface modification. Although active methods are more expensive and difficult to implement compared to passive techniques, it enables active control of heat transfer augmentation. Finally, this review develops and summarizes key learning data for design optimization enabled by additive manufacturing and machine learning algorithms, helping to inform these next-generation heat exchanger design methodologies for a plethora of modern applications such as electrification of vehicles, computing, and classical industries.

42 ENGINEERING↗