Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Comparative Study”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 1,567 records · Page 87

Uncertainty quantification in scientific machine learning: Methods, metrics, and comparisons

Neural networks (NNs) are currently changing the computational paradigm on how to combine data with mathematical laws in physics and engineering in a profound way, tackling challenging inverse and ill-posed problems not solvable with traditional methods. However, quantifying errors and uncertainties in NN-based inference is more complicated than in traditional methods. This is because in addition to aleatoric uncertainty associated with noisy data, there is also uncertainty due to limited data, but also due to NN hyperparameters, overparametrization, optimization and sampling errors as well as model misspecification. Although there are some recent works on uncertainty quantification (UQ) in NNs, there is no systematic investigation of suitable methods towards quantifying the total uncertainty effectively and efficiently even for function approximation, and there is even less work on solving partial differential equations and learning operator mappings between infinite-dimensional function spaces using NNs. In this work, we present a comprehensive framework that includes uncertainty modeling, new and existing solution methods, as well as evaluation metrics and post-hoc improvement approaches. Further, to demonstrate the applicability and reliability of our framework, we present an extensive comparative study in which various methods are tested on prototype problems, including problems with mixed input-output data, and stochastic problems in high dimensions. In the Appendix, we include a comprehensive description of all the UQ methods employed. Further, to help facilitate the deployment of UQ in Scientific Machine Learning research and practice, we present and develop in [1] an open-source Python library (github.com/Crunch-UQ4MI/neuraluq), termed NeuralUQ, that is accompanied by an educational tutorial and additional computational experiments.

11 physics-informed neural networks↗

Comparison of source-location algorithms for atmospheric samplers

Numerous algorithms have been developed to determine the source characteristics for an atmospheric release, e.g., (Bieringer et al., 2017). This study compares three models that have been applied to the data collected by the International Monitoring System operated by the Comprehensive Nuclear-Test-Ban Treaty Preparatory Commission to estimate source event parameters. Each model uses a different approach to estimate the parameters. A deterministic model uses a possible source region (PSR) approach (Ringbom et al., 2014) that is based on the correlation between predicted and measured sample values. An unnamed model developed at Pacific Northwest National Laboratory (herein called BAYEST) uses a Bayesian formulation (Eslinger et al., 2019; Eslinger et al., 2020; Eslinger and Schrom, 2016). The FREAR model uses a different Bayesian formulation (De Meutter and Hoffman, 2020; De Meutter et al., 2021; Pieter De Meutter, 2021). The performance of the three source-location models is evaluated with 100 synthetic release cases for the single xenon isotope, 133 Xe. The release cases resulted in detections in a fictitious network with 120 noble gas samplers. All three source-location models use the same sampling data. The two Bayesian models yield more accurate location estimates than the deterministic PSR model, with FREAR having slightly better location performance than BAYEST. Samplers with collection periods of 3, 6, 8, 12, and 24-h were used. Results from BAYEST show that location accuracy improves with each reduction in sample collection length. The BAYEST model is slightly better for estimating the start time of the release. The PSR model has about the same spread in start times as the FREAR model, but the PSR results have a better average start time. The Bayesian source-location algorithms give more accurate results than the PSR approach, and provide release magnitude estimates, while the base PSR model does not estimate the release magnitude. This investigation demonstrates that a reasonably dense sampling grid will sometimes yield poor location and time estimates regardless of the model. The poor estimates generally coincide with cases where there is a much larger distance between the release point and the first detecting sampler than the average sampler spacing.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

Multiple contributions to electrostrain in high performance PbTiO 3 –Bi(Ni 1/2 Hf 1/2 )O 3 piezoceramics triggered by phase transformation

Piezoelectric materials, which mechanically respond to electrical stimulation and vice versa, typically exhibit high piezoelectricity at a morphotropic phase boundary (MPB). A thorough structural description is essential for understanding the origin of excellent piezoelectric properties. In this work, a comparative study of electric-field-driven structure and phase evolution has been conducted on MPB and non-MPB tetragonal composition in high performance system of PbTiO 3 -Bi(Ni 1/2 Hf 1/2 )O 3 piezoceramics using in situ high-energy synchrotron X-ray diffraction. The results show that the MPB composition exhibits electric-field-driven reversible tetragonal-to-monoclinic transformation, whereas non-MPB composition does not. Direct evidence shows that the extensive phase transformation triggers multiple contributions to enhancing piezoelectric performance. Namely, it not only enables the unit cell parameters of coexisting phases easily change with an external electric field, but also enhances the domain switching and lattice strain. These results offer a unique perspective on MPB associated piezoelectricity enhancement, and will be helpful for developing high performance piezoelectric materials.

36 MATERIALS SCIENCE↗

Effects of surface runoff and infiltration partition methods on hydrological modeling: A comparison of four schemes in two watersheds in the Northeastern US

The ratio according to which the precipitation/snowmelt is divided into surface and subsurface water is of primary importance to watershed scale hydrological modeling. The present study compared four different surface runoff and infiltration partition methods in the Richards-equation-based SWAT model (RSWAT), i.e., daily curve number method (DC-RSWAT), hourly curve number method (HC-RSWAT), Green-Ampt method (GA-RSWAT), and effective infiltration capacity method (EIC-RSWAT). These four versions of RSWAT and a daily-curve-number-based SWAT (DC-SWAT) were used to simulate daily flow rate (from 2001 to 2015) at the outlets of the Tuckahoe Creek Watershed (TCW) and Greensboro Watershed (GW) of the Chesapeake Bay in Maryland. Parameter sensitivity analysis, calibration, and prediction uncertainty analysis were conducted. The results indicate that surface runoff and infiltration partition methods have pronounced impacts on model performance. Comparative analysis showed that DC-RSWAT outperformed other versions of RSWAT for both calibration (from 2001 to2007) and validation (from 2008 to 2015) periods in the two test watersheds. In general, DC-RSWAT achieved more accurate deterministic flow simulation and reliable uncertainty estimation than other versions. As such, daily-curve-number method is recommended for surface runoff and infiltration partition in RSWAT for simulating stream flow.

hydrology, modeling, streamflow↗

CO as a substrate and inhibitor of H + reduction for the Mo-, V-, and Fe-nitrogenase isozymes

Three known nitrogenase isozymes, Mo-, V-, and Fe-, catalyze biological reduction of dinitrogen (N 2 ) to ammonia (NH 3 ). All three utilize the same reductive elimination mechanism: an intermediate with two metal-bound hydrides reductively-eliminates hydrogen gas (H 2 ) in a reaction coupled to binding and activation of N 2 . Nonetheless, the three isozymes show dramatically different relative rates of H 2 formation and N 2 reduction, revealing important differences in reactivity with substrates. Carbon monoxide (CO) has been characterized as both an inhibitor and substrate for Mo- and V-nitrogenases, but not for the Fe-nitrogenase. In this work, we present a comparative study of the reactivity of the three isozymes with CO, examining CO both as a substrate and as an inhibitor of proton (H + ) reduction under steady-state conditions. For Mo-nitrogenase, there is neither detectable reduction of CO nor inhibition of H + reduction. Fe- and V-nitrogenase show CO reduction and inhibition of H + reduction that depends on the CO partial pressure. For V-nitrogenase, ethylene (C 2 H 4 ) is the major reduction product with a maximum specific activity of ~7.5 nmol C 2 H 4 /nmol VFe protein/min at 1 atm CO. The major product of CO reduction for Fe-nitrogenase is methane (CH 4 ) with a maximum specific activity of ~4.8 nmol CH 4 /nmol FeFe protein/min at 0.05 atm CO. The rate of CH 4 production by Fe-nitrogenase progressively increases to a maximum at 0.05 atm CO and then declines to ~10% with increasing CO partial pressure up to 1 atm. CO does not inhibit proton reduction in Mo-nitrogenase but shows 16% inhibition for V-nitrogenase and 35% for Fe-nitrogenase.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Spectroscopy of Mn 4+ in orthorhombic perovskite, LaInO 3

The low temperature spectroscopic properties of Mn 4+ (3d 3 ) in the orthorhombic perovskite, LaInO 3 , is examined in this work. The exchange charge model of crystal-field theory is used to calculate the Mn 4+ energy levels. The spectra can be fitted by the parameters D q = 1776 cm –1 , B = 700 cm –1 and C = 3189 cm –1 , with C/B = 4.5. Discussions pertaining to the low luminescence efficiency of LaInO 3 :Mn 4+ are presented. A comparative study of the variations in the energy of the 2 E g → 4 A 2g emission transition in a series of materials with the perovskite structure is presented. Here, the data suggest a relationship between the energy of the 2 E g → 4 A 2g emission transition and the deviation of the octahedral O-Mn-O bond angle from the ideal 90°. It is concluded that the angular distortion is not large enough to support the development of high luminosity Mn 4+ activated phosphors in the ABO 3 perovskite family of materials.

36 MATERIALS SCIENCE↗

Enhancing workability in sheet production of high silicon content electrical steel through large shear deformation

Enhanced workability, as characterized by the magnitude and heterogeneity of accommodated plastic strains during sheet processing, is demonstrated in high Si content Fe-Si alloys containing 4 and 6.5 wt% Si using two single-step, simple-shear deformation techniques – peeling and large strain extrusion machining (LSEM). The model Fe-Si material system was selected for its intrinsically poor material workability, and well-known applications potential in next-generation electric machines. In a comparative study of the deformation characteristics of the shear processes with conventional rolling, two distinct manifestations of workability are observed. For rolling, the relatively diffuse and unconfined deformation zone geometry leads to cracking at low strains, with sheet structures characterized by extensive deformation twinning and banding. Workpiece pre-heating is required to improve the workability in rolling. In contrast, peeling and LSEM produce continuous sheet at large plastic strains without cracking, the result of more confined deformation geometries that enhances the workability. Peeling, however, results in heterogeneous, shear-banded microstructures, pointing to a second type of workability issue – flow localization – that limits sheet processing. This shear banding is to a large extent facilitated by unrestricted flow at the sheet surface, unavoidable in peeling. With additional confinement of this free surface deformation and appropriately designed deformation zone geometry, LSEM is shown to suppress shear banding, resulting in continuous sheet with homogeneous microstructure. Thus LSEM is shown to produce the greatest enhancement in process workability for producing sheet. In conclusion, these workability findings are explained and discussed based on differences in process mechanics and deformation zone geometry.

36 MATERIALS SCIENCE↗

Importance of doping and frustration in itinerant Fe-doped Cr 2 Al

We performed an experimental and theoretical study comparing the effects of Fe-doping of Cr 2 Al, an antiferromagnet with a N el temperature of 670 K, with known results on Fe-doping of antiferromagnetic bcc Cr. (Cr 1-x Fe x ) 2 Al materials are found to exhibit a rapid suppression of antiferromagnetic order with the presence of Fe, decreasing T N to 170 K for x=0.10. Antiferromagnetic behavior disappears entirely at x≈0.125 after which point increasing paramagnetic behavior is exhibited. Moreover, this is unlike the effects of Fe doping of bcc antiferromagnetic Cr, in which T N gradually decreases followed by the appearance of a ferromagnetic state. Theoretical calculations explain that the Cr 2 Al-Fe suppression of magnetic order originates from two effects: the first is band narrowing caused by doping of additional electrons from Fe substitution that weakens itinerant magnetism; the second is magnetic frustration of the Cr itinerant moments in Fe-substituted Cr 2 Al. In pure-phase Cr 2 Al, the Cr moments have an antiparallel alignment; however, these are destroyed through Fe substitution and the preference of Fe for parallel alignment with Cr. This is unlike bulk Fe-doped Cr alloys in which the Fe anti-aligns with the Cr atoms, and speaks to the importance of the Al atoms in the magnetic structure of Cr 2 Al and Fe-doped Cr 2 Al.

36 MATERIALS SCIENCE↗

Comparison of permeability predictions on cemented sandstones with physics-based and machine learning approaches

Permeability prediction has been an important problem since the time of Darcy. Most approaches to solve this problem have used either idealized physical models or empirical relations. In recent years, machine learning (ML) has led to more accurate and robust, but less interpretable empirical models. Using 211 core samples collected from 12 wells in the Garn Sandstone from the North Sea, this study compared idealized physical models based on the Carman-Kozeny equation to interpretable ML models. We found that ML models trained on estimates of physical properties are more accurate than physical models. The results show evidence of a threshold of about 10% volume fraction, above which pore-filling cement strongly affects permeability.

03 NATURAL GAS↗

Effects of Temperature on the Interactions of Helium-Vacancy Clusters with Gliding Edge Dislocations in α-Fe

The interaction of helium-vacancy (He-V) clusters with a gliding a/2<111>{110} edge dislocation in a-Fe is investigated by molecular dynamics methods under a constant strain rate at temperatures of 100 to 600 K. A number of small HenVm (n/m = 0~4) clusters initially placed at different positions relative to the slip plane are comparatively studied. The results show that the interaction of He-V clusters with gliding edge dislocations depends on the helium-to-vacancy (He/V) ratio, the position of the clusters relative to the slip plane, the cluster size, and also temperature. The obstacle strength of the He-V clusters relevant to the dislocation motion generally increases with increasing He/V ratio at the same temperature, but decreases slightly with increasing temperature for the same He-V cluster. One of the interesting results is that He-V clusters do not move along with the dislocation, even at 600 K.

a-Fe↗

Microstructural evolution in NF616 (P92) and Fe–9Cr–0.1C-model alloy under heavy ion irradiation

In this comparative study, in situ investigations of the microstructure evolution in a Fee9Cr ferritic emartensitic steel, NF616, and a Fee9Cre0.1C-model alloy with a similar ferriticemartensitic microstructure have been performed. NF616 and Fee9Cre0.1C-model alloy were irradiated to high doses (up to ~10 dpa) with 1 MeV Kr ions between 50 and 673 K. Defect cluster density increased with dose and saturated in both alloys. The average size of defect clusters in NF616 was constant between 50 and 573 K, on the other hand average defect size increased with dose in Fee9Cre0.1C-model alloy around ~1 dpa. At low temperatures (50e298 K), alignment of small defect clusters resulted in the formation of extensive defects in Fee9Cre0.1C-model alloy around ~2e3 dpa, while similar large defects in NF616 started to form at a high temperature of 673 K around ~5 dpa. Interaction of defect clusters with the lath boundaries were found to be much more noticeable in Fee9Cre0.1C-model alloy. Differences in the microstructural evolution of NF616 and Fee9Cre0.1C-model alloy are explained by means of the defect cluster trapping by solute atoms which depends on the solute atom concentrations in the alloys.

Ferritic-martensitic alloys↗

An investigation of FeCrAl cladding behavior under normal operating and loss of coolant conditions

Iron-chromium-aluminum (FeCrAl) alloys are candidates to be used as nuclear fuel cladding for increased accident tolerance. An analysis of the response of FeCrAl under normal operating and loss of coolant conditions has been performed using fuel performance modeling. In particular, recent information on FeCrAl material properties and phenomena from separate effects tests has been implemented in the BISON fuel performance code and analyses of integral fuel rod behavior with FeCrAl cladding have been performed. BISON simulations included both light water reactor normal operation and loss-of-coolant accidental transients. In order to model fuel rod behavior during accidents, a cladding failure criterion is desirable. For FeCrAl alloys, a failure criterion is developed using recent burst experiments under loss of coolant like conditions. The added material models are utilized to perform comparative studies with Zircaloy-4 under normal operating conditions and oxidizing and non-oxidizing out-of-pile loss of coolant conditions. The results indicate that for all conditions studied, FeCrAl behaves similarly to Zircaloy-4 with the exception of improved oxidation performance. Here, further experiments are required to confirm these observations.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Direct comparison of nanoindentation and tensile test results on reactor-irradiated materials

Nanoindentation testing has been used for decades to assess materials on a local scale and to obtain fundamental mechanical property parameters. Nuclear materials research often faces the challenge of testing rather small samples due to the hazardous nature, limited space in reactors, and shallow ion-irradiated zones, fostering the need for small-scale mechanical testing (SSMT). As such, correlating the results from SSMT to bulk properties is particularly of interest. Here, this study compares macroscopic tensile test data (yield and flow stresses) to nanoindentation data (hardness) obtained on a number of different neutron-irradiated materials in order to understand the scaling behavior on radiation-damaged samples.

36 MATERIALS SCIENCE↗

Corrosion susceptibility and chromium loss in Austenitic steels and Nickel-based alloys in molten FLiNaK at 700 °C

A comparative study was conducted to evaluate the corrosion susceptibility of 316L, 316H, Ni 200, Inconel 625, and Hastelloy N in molten FLiNaK at 700 °C for 100 h. Top-view and cross-sectional scanning electron microscopy (SEM) imaging, combined with energy-dispersive X-ray spectroscopy (EDS) mapping, was performed to investigate microstructural and compositional changes. SEM images were further processed by introducing a contrast threshold to map cavity distribution. Using EDS mapping, intergranular and intragranular Cr loss were separately characterized. The ranking of mass loss after corrosion, from highest to lowest, is as follows: 316H > 316L > Inconel 625 > Hastelloy N > Ni 200. Cr loss was found to be correlated with Mo concentration, in agreement with many previous studies, with higher Mo content resulting in reduced Cr loss. On the other hand, there is no evidence that the high carbon content in 316H enhances corrosion resistance. The relatively low solubility of carbon in austenite at the testing temperature limits the amount of dissolved carbon. Therefore, the carbon-retarded vacancy diffusion, and consequently the reduced Cr diffusion, were not observed.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Performance evaluation of automated data-driven feature extraction and selection methods for practical and scalable building energy consumption prediction models

Here, this study quantifies the impact of automated feature engineering methods (feature extraction and selection) on the quality and accuracy of machine learning models that predict building energy consumption. The case study compares model performance for three main scenarios: baseline (no feature extraction and selection), feature extraction only, and feature extraction combined with feature selection (filter and/or wrapper methods) for fully trained machine learning models for 200 metered/sub-metered energy measurements across 118 real buildings. For consistency, the same machine learning model architecture (a black box deep learning neural network with probabilistic forecast output) was used for all scenarios. Based on results, all feature engineering methods provided noticeable prediction accuracy improvements (e.g., 29%-68% median prediction improvement) compared to baseline scenarios. However, in this application, feature selection methods provide little practical value due to their limited performance gains and high computational cost. Smarter algorithm development supported by better computational environments will be needed before feature selection methods can reliably and efficiently improve predictive model performance.

97 MATHEMATICS AND COMPUTING↗

Experimental review of the performances of protective coatings for interconnects in solid oxide fuel cells

Ferritic stainless-steel interconnects are used in solid oxide fuel cells; however, coatings are required to improve their performance. Although several types of coatings have been proposed, they have been scarcely investigated under similar conditions. This study compares the characteristics of uncoated Crofer 22 APU and eight different coatings on Crofer 22 APU for up to 3,000 hours at 800°C. The coatings were deposited at various research laboratories around the world, and the experiments were performed at Chalmers University of Technology, Sweden. Cross-sections of the samples were analysed using scanning electron microscopy and energy-dispersive x-ray spectroscopy. The (Co,Mn)-based coated steels showed more than 50-fold lower chromium evaporation and at least 3 times thinner Cr 2 O 3 scale thickness compared to uncoated steel. The coated steel samples showed lower area-specific resistance (ASR) values than the uncoated steel after 3,000 hours of exposure, irrespective of the coating thickness, composition and deposition method.

25 ENERGY STORAGE↗

Superior electrochemical performance and reduced heat generation in 3D printed vs. 2D tape-casted NMC622 electrodes

This study compares the charge storage mechanisms, thermodynamics behavior, ion transport, and heat generation in NMC622 electrodes fabricated using a novel 3D printing process and the conventional 2D tape casting process. First, potentiometric entropy measurements revealed that the charge storage mechanisms for both types of electrodes consisted of lithium deintercalation in a homogeneous solid solution of NMC622 followed by a transition from a hexagonal (H1) phase to another hexagonal (H2) phase through a monoclinic (M) phase. Both types of electrodes had similar thermodynamics behavior with overlapping entropic potential profiles. Furthermore, operando isothermal calorimetry at high C-rates indicated that the 3D printed electrodes featured larger specific capacity and better rate performance than the 2D tape-casted electrodes. The better performance of 3D printed electrodes was attributed to their larger electrode/electrolyte interfacial surface area and electrical conductivity as well as their faster lithium ion transport. As a result, the instantaneous heat generation rates were smaller in 3D printed electrodes than in 2D tape-casted electrodes, thus resulting in lower overall specific electrical energy and thermal energy dissipation per unit charge stored. Overall, additive manufacturing techniques offer great potential in producing electrodes with superior electrochemical performance and reduced heat generation for fast charging batteries.

25 ENERGY STORAGE↗