SEARCH · Search NASA

Results for “Best”

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.

SourceResearch connectionFrom yearThrough year

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

At least 343 records · Page 19

All sources NASA NTRS DOE OSTI

DOE OSTI2020

Predicting the proximity to macroscopic failure using local strain populations from dynamic in situ X-ray tomography triaxial compression experiments on rocks

Predicting the proximity of large-scale dynamic failure is a critical concern in the engineering and geophysical sciences. Here we use evolving contractive, dilatational, and shear strain deformation preceding failure in dynamic X-ray tomography experiments to examine which strain components best predict the proximity to failure. We develop machine learning models to predict the proximity to failure using time series of three-dimensional local incremental strain tensor fields acquired in rock deformation experiments under stress conditions of the upper crust. Three-dimensional scans acquired in situ throughout triaxial compression experiments provide a distribution of density contrasts from which we estimate the three-dimensional incremental strain that accumulates between each scan acquisition. Training machine learning models on multiple experiments of six rock types provides suites of feature importance that indicate the predictive power of each feature. Comparing the average importance of groups of features that include information about each strain component quantifies the ability of the contractive, dilatational and shear strain to predict the proximity of macroscopic failure. A total of 24 models of four machine learning algorithms with six rock types indicate that 1) the dilatational strain provides the best predictive power of the strain components, and 2) the intermediate values (25th-75th percentile) of the strain population provide the best predictive power of the statistics of the strain populations. In addition, the success of the predictions of models trained on one rock type and tested on other rock types quantifies the similarities and differences of the precursory strain accumulation process in the six rock types. These similarities suggest the potential existence of a unified theory of brittle rock deformation for a range of rock types.

58 GEOSCIENCES↗

DOE OSTI2023

Selecting representative geological realizations to model subsurface CO 2 storage under uncertainty

Carbon capture and storage (CCS) is one of the quickest and most effective solutions for reducing carbon emissions. The majority of subsurface storage occurs in saline aquifers, for which geological information is lacking which in turn results in geological uncertainty. To evaluate uncertainty in CO 2 injection projections, the use of multiple geological realizations (GRs) has been practiced very commonly. In this approach, hundreds or thousands of high-resolution GRs is used that quickly becomes computationally expensive. This issue can be addressed with representative geological realizations (RGRs) that preserve the uncertainty domain of the ensemble GRs. Here, in this study, we propose the use of unsupervised machine learning (UML) frameworks, including dissimilarity measurement, dimensionality reduction, clustering and sampling algorithms ta select a predetermined number of RGRs. We compare the simulation outputs of the RGR sets and the ensemble using the Kolmogorov–Smirnov (KS) test to select the best UML. The UML frameworks and their associated selection processes are evaluated using a saline aquifer with a single CO 2 injection well and 200 GRs with varying uncertain petrophysical characteristics. The best UML framework is selected to use only 5% of the GRs while maintaining the uncertainty domain of the ensemble GRs. In addition, the best UML framework is tested using a saline aquifer with three CO 2 injection wells and varied GRs. The results show that our proposed UML framework can be used to choose RGRs, capturing the whole uncertainty domain. Our approach leads to a significant reduction in the computational cost associated with scenario testing, decision-making, and development planning for CO 2 storage sites under geological uncertainty.

58 GEOSCIENCES↗

DOE OSTI2020

Evaluating three evapotranspiration estimates from model of different complexity over China using the ILAMB benchmarking system

The land surface models range in complexity of terrestrial evapotranspiration, yet it is unknown how model complexity translates to accuracy of modeled evapotranspiration estimates. Here, we use the International Land Model Benchmarking system to assess ET estimates from three models of varying complexity driven by the same forcing datasets: an earth system model, a terrestrial biosphere model, and a stand-alone ET model. The performance assessment includes both temporal and spatial evaluation, and different plant functional types across China. Our results indicate that the most complex model, an earth system model, performed best against the benchmarking datasets and metrics. Terrestrial biosphere model performed best in simulating inter-annual variability of ET, while earth system model performed best in simulating the seasonal cycle. The more complex models (earth system model and terrestrial biosphere model) perform better in forest, shrub and crop ecosystems, while the simpler model (stand-alone ET model) perform better in grass ecosystems. Our study demonstrates the impact of model complexity on ET estimates and highlights directions for future ET model improvements.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2021

Propagating neutronic uncertainties for FFTF LOFWOS Test #13

The safety evaluation conducted for licensing a Sodium-cooled Fast Reactor (SFR) may require modeling transients with best-estimate calculation tools that must first be validated against real-world measurements. To provide the community with a valuable benchmarking opportunity for validating SFR analysis tools and methods, the International Atomic Energy Agency (IAEA) initiated a coordinated research project (CRP) in 2018 for the analysis of the Fast Flux Test Facility (FFTF) Loss of Flow Without Scram (LOFWOS) Test #13. The impact of nuclear data uncertainties on neutronics parameters was previously investigated based on the COMMARA-2.0 covariance matrix. Since the transient simulation results are very sensitive to certain reactivity coefficients, it was decided to employ rigorous uncertainty propagation methods to quantify the impact of nuclear data uncertainties on the best-estimate predication of FFTF LOFWOS Test #13. The DAKOTA code is used to propagate neutronic uncertainties through SAS4A/SASSYS-1 transient simulations, while taking into account spatial and reaction-wise correlations within these uncertainties. This study shows that the remaining discrepancies observed between the Argonne National Laboratory (ANL) best-estimate results and the experimental measurements can be partly explained by the uncertainty associated with Gas Expansion Module (GEM) worth, which contributes the majority of the overall nuclear data uncertainty on the output from the FFTF LOFWOS Test #13 transient simulation. This study also confirmed the importance of including spatial and reaction-wise correlations of nuclear data uncertainties on feedback coefficients in the uncertainty propagation to avoid under-estimating their impact during the transient simulations.

42 ENGINEERING↗

DOE OSTI2021

Material Interactions in Severe Accidents – Benchmarking the MELCOR V2.2 Eutectics Model for a BWR-3 Mark-I Station Blackout: Part II – Uncertainty Analysis

Single case comparisons between severe accident simulations can provide detailed insights into severe accident model behavior, however, they cannot offer insights into model uncertainty, sensitivity to uncertain parameters, or underlying model biases.Here in this analysis, the single case benchmark comparison of the MELCOR material interaction models for a station blackout (SBO) scenario of a boiling water reactor (BWR) using representative Fukushima Daiichi Unit 1 boundary conditions is expanded to include an uncertainty analysis. As part of this uncertainty analysis, 1200 simulations are performed for each material interaction model (2400 total), with random sampling of 14 uncertain MELCOR input parameters. Input parameters are selected for their impact on models representing core degradation processes. These include candling, fuel rod failure, debris quenching and dryout. The analysis performed here is not a traditional “best-estimate” uncertainty analysis that uses best-estimate parameters or identifies best-estimate figure of merit distributions. Instead, it is an exploratory uncertainty analysis that identifies and interrogates underlying model form biases of the two material interaction models (eutectics and interactive materials models). Uniform distributions are applied to all uncertain parameters to ensure coverage of the model parameter uncertainty space. Key findings from this study include underlying model form biases exhibited by material interaction models, and notable differences in accident progression outcomes between the material interaction models. This uncertainty study extends and confirms the conclusions from the first part of this study, which compared the impact of material interaction modeling on simulation of a short-term station blackout scenario with representative Fukushima Daiichi Unit I boundary conditions. In particular, this study confirms that the eutectics model generally exhibits accelerated degradation and failure of fuel components, the core plate, and the lower head. The eutectics model also has a tendency to exhibit a greater degree of core degradation, greater debris mass formation, and larger debris mass ejection. Finally, the eutectics model exhibits higher maximum temperatures for fuel, cladding, particulate debris, oxidic molten pool, and metallic molten pool components than the interactive materials model; interactive materials model simulations exhibit a soft “limitation” on maximum temperatures that is related to the temperature at which material relocation occurs.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2022

Aboveground biomass density models for NASA’s Global Ecosystem Dynamics Investigation (GEDI) lidar mission

NASA's Global Ecosystem Dynamics Investigation (GEDI) is collecting spaceborne full waveform lidar data with a primary science goal of producing accurate estimates of forest aboveground biomass density (AGBD). This paper presents the development of the models used to create GEDI's footprint-level (~25 m) AGBD (GEDI04_A) product, including a description of the datasets used and the procedure for final model selection. The data used to fit our models are from a compilation of globally distributed spatially and temporally coincident field and airborne lidar datasets, whereby we simulated GEDI-like waveforms from airborne lidar to build a calibration database. We used this database to expand the geographic extent of past waveform lidar studies, and divided the globe into four broad strata by Plant Functional Type (PFT) and six geographic regions. GEDI's waveform-to-biomass models take the form of parametric Ordinary Least Squares (OLS) models with simulated Relative Height (RH) metrics as predictor variables. From an exhaustive set of candidate models, we selected the best input predictor variables, and data transformations for each geographic stratum in the GEDI domain to produce a set of comprehensive predictive footprint-level models. We found that model selection frequently favored combinations of RH metrics at the 98th, 90th, 50th, and 10th height above ground-level percentiles (RH98, RH90, RH50, and RH10, respectively), but that inclusion of lower RH metrics (e.g. RH10) did not markedly improve model performance. Second, forced inclusion of RH98 in all models was important and did not degrade model performance, and the best performing models were parsimonious, typically having only 1-3 predictors. Third, stratification by geographic domain (PFT, geographic region) improved model performance in comparison to global models without stratification. Fourth, for the vast majority of strata, the best performing models were fit using square root transformation of field AGBD and/or height metrics. There was considerable variability in model performance across geographic strata, and areas with sparse training data and/or high AGBD values had the poorest performance. These models are used to produce global predictions of AGBD, but will be improved in the future as more and better training data become available.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2021

Predicting measures of soil health using the microbiome and supervised machine learning

Soil health encompasses a range of biological, chemical, and physical soil properties that sustain the commercial and ecological value of agroecosystems. Monitoring soil health requires a comprehensive set of diagnostics that can be cost-prohibitive for routine analyses. The soil microbiome provides a rich source of information about soil properties, which can be assayed in a high-throughput, cost-effective way. We evaluated the accuracy of random forest (RF) and support vector machine (SVM) regression and classification models in predicting 12 measures of soil health, tillage status, and soil texture from 16S rRNA gene amplicon data with an operationally relevant sample set. We validated the efficacy of the best performing models against independent datasets and also tested best practices for processing microbiome data for use in machine learning. Soil health metrics could be predicted from microbiome data with the best models achieving a Kappa value of ~0.65, for categorical assessments, and a R2 value of ~0.8, for numerical scores. Biological health ratings were better predicted than chemical or physical ratings. Validation with independent datasets revealed that models had general predictive value for soil properties, including yield. The ecological profiles of several taxa important for model accuracy matched the observed relationships with soil health, including Pyrinomonadaceae, Nitrososphaeraceae, and Candidatus Udeaobacter. Models trained at the highest taxonomic resolution proved most accurate, with losses in accuracy resulting from rarefying, sparsity filtering, and aggregating at higher taxonomic ranks. Furthermore, our study provides the groundwork for developing scalable technology to use microbiome-based diagnostics for the assessment of soil health.

16S rRNA gene↗

DOE OSTI2025

Boosting Noise2Inverse via enhanced model selection for denoising computed tomography data

Synchrotron-based x-ray tomographic imaging enables the examination of the internal structure of materials at high spatial and temporal resolution. Experimental constraints can impose dose and time limits on the measurements, introducing a higher level of noise and artifacts in the reconstructed images. Deep learning has emerged as a powerful tool to remove noise from reconstructed images. Recently, the Noise2Inverse method was designed specifically for denoising reconstructed images without requiring paired noisy and clean images. This method creates multiple statistically independent reconstructions used to pair the data in which training involves transforming one reconstruction into the other, and vice versa. Originally designed to be used after a fixed number of epochs, we see in practice that this approach may not produce the optimal model and may unnecessarily waste computational resources. Therefore, we propose an alternative method of identifying the best model during training that aligns with the Noise2Inverse method. During validation, we compare the model output of the multiple reconstructions among each other. We hypothesize that the best model is the one that produces images with the highest similarity, implying a convergence in the predicted material properties and absorption values. To compare model outputs, we consider the absolute error, square error, structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), and cosine similarity. We evaluate our method on two simulated tomography datasets and two, real-world, low-contrast, high-energy x-ray tomography datasets. We show our approach is more effective at determining the best model, up to an increase of 12.50% and 12.53% in SSIM and PSNR, respectively, while only requiring a fifth of the training time compared to the original approach.

CT↗

DOE OSTI2023

Iodine Removal from Carbonate-Containing Alkaline Liquids Using Strong Base Resins, Hybrid Resins, and Silver Precipitation

The ability of several material types to remove aqueous iodine from a mildly alkaline, carbonate-rich nuclear waste stream was evaluated: strong base anion exchange resins (SBAR), hybrid resins, Ag-containing materials and Bi-containing hybrid resins. A combination of batch testing and flow through column testing were used in the evaluation. In batch testing, hybrid resins CHM-20, SIR-110-CE and RTBI were shown to have high efficiency for the removal of both iodide and iodate simultaneously, while Ag-containing materials and SBAR demonstrated high capacity for iodide removal. One example of each material type (CHM-20, A532E and Ionex 400) were further evaluated for their sorption isotherms and column performance. The Langmuir isotherm, or a Langmuir-Freundlich hybrid isotherm, best described the sorption of iodide to the CHM-20 hybrid resin and Purolite A532E. The Freundlich isotherm best described the uptake of iodate to CHM-20 and A532E and for both iodide and iodate to the Ag-containing Ionex-400. In column testing the Purolite A532E had exceptional performance for overall iodide removal. With the capacity demonstrated the A532E resin would exceed Class C waste classification before breakthrough initiated, and column change outs in processing would be dictated by eventual waste classification, not breakthrough. The Ionex 400, a Ag-zeolite, was observed to degrade over time in the column in the mild alkaline conditions whereas the hybrid CHM-20 was limited in the single pass through design and would be best suited for applications where iodide and iodate are present and recirculation of the column effluent is feasible. Here this work highlights the feasibility of commercially available materials to separate radioiodine from liquid environments.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2022

Investigating the Influence of Hexanuclear Clusters in Isostructural Metal–Organic Frameworks on Toxic Gas Adsorption

The efficient capture of toxic gases, such as ammonia (NH 3 ) and sulfur dioxide (SO 2 ), can protect the general population and mitigate widespread air pollution. Metal–organic frameworks (MOFs) comprise a tunable class of adsorbents with high surface areas that can meet this challenge by selectively capturing these gases at low concentrations. In this work, we explored how modifying the metal ions in the node of an isostructural MOF series from a transition metal to a lanthanide or actinide influences the electronic environment of the node-based active site. Next, we investigated the adsorption properties of each MOF toward the relatively basic NH 3 and relatively acidic SO 2 gases. Within the NU-907 family of MOFs, we found that Zr 6 -NU-907 exhibits the best uptake toward NH 3 at low pressures, while Th 6 -NU-907 demonstrates the best low-pressure performance for SO 2 adsorption. Here, tracking the infrared (IR) stretching frequency of the node-based μ 3 -OH groups provides insights into the electronegativity of the metal ion and suggests that the most electronegative metal ion (Zr) affords the node with the best NH 3 uptake at low pressures. In contrast, the Th 6 node contains additional coordinated water groups relative to the other M 6 nodes, which appears to yield the MOF with the greatest affinity for SO 2 uptake that occurs predominately through reversible physisorption interactions. Finally, in situ NH3 IR spectroscopic studies indicate that both NH 4 + and Lewis-bound NH 3 species form during adsorption. Combined, these results suggest that tuning the electronic properties and structure of the node-based active site in an MOF presents a viable strategy to change the affinity of an MOF toward toxic gases.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2024

Exceptional Electrical Detection of Trace NO 2 via Mixed Metal MOF-on-MOF Film-Based Sensors

The tunability of metal–organic frameworks (MOFs) makes them exceptional materials for the development of highly selective, low-power sensors for toxic gas detection. Herein, we demonstrate enhanced detection of NO 2 gas by a MOF-based electrical impedance sensor made using a unique mixed metal MOF-on-MOF synthesis. For this work, a combined experimental and computational study was performed using the exemplar Ni x Mg 1–x -MOF-74 to understand the fundamental structure–property relationships behind metal mixing and MOF film synthesis methods on sensor performance. Density functional theory results indicated that the presence of Ni in Mg-MOF-74 increased framework stability and increased the electron density of states at lower energies near the HOMO, as well as enhanced the NO 2 –Mg adsorption interaction. Impedance data of the Ni x Mg 1–x -MOF-74 films with larger Ni contents showed greater impedance change after exposure to 1 ppm of NO 2 gas. Furthermore, when synthesized through either a drop-cast or direct solvothermal film growth approach, the monometallic Ni-based sensors had the best performance. However, the mixed metal Ni x Mg 1–x -MOF-74 sensors synthesized through a MOF-on-MOF approach resulted in the highest impedance change, outperforming all monometallic Ni-based sensors. In particular, the mixed metal Ni-on-Mg-MOF-74 film was the best-performing sensor with an impedance change of 309 upon trace NO 2 exposure. Change in impedance response after NO 2 exposure was improved by 52% compared to the best monometallic Ni-on-Ni-MOF-74 sensor. Structural analysis of the Ni-on-Mg film showed that the first Mg-MOF-74 layer acts as a structural template controlling the structural features of the final film after metal exchange with Ni. This led to improved film quality, evidenced by the greater crystallinity and larger MOF grain sizes, and resulted in enhanced sensor performance which was not achievable through other metal mixing methods. Altogether, this study identifies structure–property relationships and synthetic templating methods that inform MOF-based sensor design, allowing for improved detection of toxic compounds.

36 MATERIALS SCIENCE↗

DOE OSTI2024

Phenology of Photosynthesis in Winter‐Dormant Temperate and Boreal Forests: Long‐Term Observations From Flux Towers and Quantitative Evaluation of Phenology Models

Abstract We examined the seasonality of photosynthesis in 46 evergreen needleleaf (evergreen needleleaf forests (ENF)) and deciduous broadleaf (deciduous broadleaf forests (DBF)) forests across North America and Eurasia. We quantified the onset and end (Start GPP and End GPP ) of photosynthesis in spring and autumn based on the response of net ecosystem exchange of CO 2 to sunlight. To test the hypothesis that snowmelt is required for photosynthesis to begin, these were compared with end of snowmelt derived from soil temperature. ENF forests achieved 10% of summer photosynthetic capacity ∼3 weeks before end of snowmelt, while DBF forests achieved that capacity ∼4 weeks afterward. DBF forests increased photosynthetic capacity in spring faster (1.95% d −1 ) than ENF (1.10% d −1 ), and their active season length (End GPP –Start GPP ) was ∼50 days shorter. We hypothesized that warming has influenced timing of the photosynthesis season. We found minimal evidence for long‐term change in Start GPP , End GPP , or air temperature, but their interannual anomalies were significantly correlated. Warmer weather was associated with earlier Start GPP (1.3–2.5 days °C −1 ) or later End GPP (1.5–1.8 days °C −1 , depending on forest type and month). Finally, we tested whether existing phenological models could predict Start GPP and End GPP . For ENF forests, air temperature‐ and daylength‐based models provided best predictions for Start GPP , while a chilling‐degree‐day model was best for End GPP . The root mean square errors (RMSE) between predicted and observed Start GPP and End GPP were 11.7 and 11.3 days, respectively. For DBF forests, temperature‐ and daylength‐based models yielded the best results (RMSE 6.3 and 10.5 days).

Environmental Sciences & Ecology↗

DOE OSTI2021

The LSBmax algorithm for boosting resilience of electric grids post (N‐2) contingencies

Abstract A computationally improved algorithm is presented to find the best transmission switching (TS) candidate for boosting resilience of electricity grids subject to ( N ‐2) contingencies. Here, resilience is computed as the reduction in load shed after the above‐mentioned ( N‐ ) contingencies. TS is a planned line outage, and past research shows that changing the transmission system's topology changes the power flow and removes post contingency violations. Finding the best TS candidate in a computationally suitable time for effectively boosting resilience is a challenge. The best TS candidate is found using a novel heuristic method by decreasing the search space based on proximity to the bus with the maximum load shedding (LSB). The LSB algorithm is faster than existing algorithms in the literature; and, it is compatible with both the AC and DC optimal power flow formulations. To validate the authors' claims of speedup and accuracy, two metrics are used to analyze the results from the IEEE 39‐bus and 118‐bus systems. Finally, the inherent parallelism of the LSB algorithm is leveraged on a high‐performance computing platform and applied to the large‐scale Polish 2383‐bus test system to validate scalability in both size and speedup in computation time.

24 POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2022

Search of strong lens systems in the Dark Energy Survey using convolutional neural networks

We present our search for strong lens, galaxy-scale systems in the first data release of the Dark Energy Survey (DES), based on a color-selected parent sample of 18 745 029 luminous red galaxies (LRGs). We used a convolutional neural network (CNN) to grade this LRG sample with values between 0 (non-lens) and 1 (lens). Our training set of mock lenses is data-driven, that is, it uses lensed sources taken from HST-COSMOS images and lensing galaxies from DES images of our LRG sample. A total of 76 582 cutouts were obtained with a score above 0.9, which were then visually inspected and classified into two catalogs. The first one contains 405 lens candidates, of which 90 present clear lensing features and counterparts, while the other 315 require more evidence, such as higher resolution imaging or spectra, to be conclusive. A total of 186 candidates are newly identified by our search, of which 16 are among the 90 most promising (best) candidates. The second catalog includes 539 ring galaxy candidates. This catalog will be a useful false positive sample for training future CNNs. For the 90 best lens candidates we carry out color-based deblending of the lens and source light without fitting any analytical profile to the data. This method is shown to be very efficient in the deblending, even for very compact objects and for objects with a complex morphology. Finally, from the 90 best lens candidates, we selected 52 systems with one single deflector to test an automated modeling pipeline that has the capacity to successfully model 79% of the sample within an acceptable computing runtime.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2025

River Dissolved Oxygen Prediction Using Machine Learning Models and Wireless Sensor Measurements

Simultaneous flooding&heat and droughts&heat events can potentially destabilize hydro-meteorological conditions to deteriorate the water quality of Neches River. Machine learning (ML) models utilizing wireless sensor measurements have been applied to predict water quality and optimize various water management strategies. This study aims to develop ML models to predict dissolved oxygen (DO) prediction under various hydro-meteorological conditions and enhance water management decision-making. Wireless sensor measurements of DO, water temperature, sample depth, conductivity, turbidity, and pH, along with discharge from the United States Geological Survey stations, are collected for model inputs at the Pine Island Bayou C749 station (PIB-C749) and Neches River Saltwater Barrier (SWB). Multilayer perceptron neural networks, recurrent neural networks, long short-term memory (LSTM), and bidirectional LSTM (BiLSTM) with and without attention mechanism (AT) are tested to determine the best model, which is applied the rolling forecast method to predict 14-day DO. Traditional and recurrent transfer learning (TL and RTL) methods are adopted to overcome insufficient data at the SWB. The input feature importance analysis using the integrated gradients (IG) algorithm is applied to determine dominant inputs. The results show LSTM-based models are capable handling long sequential data. AT-BiLSTM and RTL-LSTM demonstrate the best performance at the PIB-C749 (RMSE=0.054) and the SWB (RMSE=0.028), respectively. TL and RTL methods significantly improve model performance at the SWB. DO, temperature, and pH show higher importance, consistent with hydrodynamics and water chemistry. Both best models are applied to predict 14-day DO and demonstrate reasonable performance for decision-making. Hydro-meteorological conditions of 2017 flood and 2012 drought events are simulated and reveal that possible hypoxia occurs after flooding due to increasing temperature and turbidity, and DO concentration decreases significantly under heat and drought conditions. In conclusion, LSTM-based models utilizing wireless sensor data can be a timely and effective approach to make appropriate decisions on water resource management.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2025

A comparison of probabilistic generative frameworks for molecular simulations

Generative artificial intelligence is now a widely used tool in molecular science. Despite the popularity of probabilistic generative models, numerical experiments benchmarking their performance on molecular data are lacking. Here, in this work, we introduce and explain several classes of generative models, broadly sorted into two categories: flow-based models and diffusion models. We select three representative models: neural spline flows, conditional flow matching, and denoising diffusion probabilistic models, and examine their accuracy, computational cost, and generation speed across datasets with tunable dimensionality, complexity, and modal asymmetry. Our findings are varied, with no one framework being the best for all purposes. In a nutshell, (i) neural spline flows do best at capturing mode asymmetry present in low-dimensional data, (ii) conditional flow matching outperforms other models for high-dimensional data with low complexity, and (iii) denoising diffusion probabilistic models appear the best for low-dimensional data with high complexity. Our datasets include a Gaussian mixture model and the dihedral torsion angle distribution of the Aib9 peptide, generated via a molecular dynamics simulation. We hope our taxonomy of probabilistic generative frameworks and numerical results may guide model selection for a wide range of molecular tasks.

Artificial intelligence↗

DOE OSTI2020

Comparison of Shadowgraphy and X-Ray Phase Contrast Methods for Characterizing a DT Ice Layer in an Inertial Confinement Fusion Target

Shadowgraphy and x-ray phase contrast (XPC) imaging are two techniques that are used for characterizing the deuterium-tritium ice layer in Inertial Confinement Fusion targets. Each technique has limitations that affect how accurately they can characterize small crystalline defects and measure the ice-thickness non-uniformities that may be only a few micrometers in height. The concern is that shadowgraphy may be overly sensitive to the shape and depth of defects in the surface and insufficiently sensitive to the shape of longer-wavelength roughness, while x-ray phase contrast may be too insensitive to defects in the ice surface. Multiple ice layers with different thicknesses (40 to 63 μm), thickness uniformities (peak-valley variations that range from <2- to 12 μm), and crystal defects were analyzed using shadowgraphy and x-ray phase contrast techniques. The results from each method agree when the ice layer is uniformly thick and the crystal lacks defects. That agreement worsens as the number of defects in the surface of the ice layer increases and the roughness that is determined from a shadowgram image of the target’s limb is greater than can be justified by the number of defects that are seen in the target’s front and rear surfaces. The XPC technique is considerably less sensitive to surface defects, in part due to the poorer dynamic range and image resolution compared to shadowgraphy. Localized regions of the ice layer that are thicker or thinner than the average thickness of the layer are reported by shadowgraphy to be smaller in height and footprint (by up to 30%) than by XPC. As a result, the two techniques report different ice layer thicknesses that can vary by up to 10%. Shadowgraphy results from two caustics that trace different paths through the target, and, in theory, image the same ice-vapor surface (but reflect from either the vapor or ice side of the interface) did not consistently characterize the size or shape of ice features to be the same magnitude. Here, the XPC technique provides the best assessment of low-mode (l < 7) roughness in the ice layer. Shadowgraphy results using the strongest caustic is best for detecting the presence of grooves in the ice, though not for quantifying the size of them. If multiple grooves are present it is best to discard and reform the ice layer.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2024

Did we hear the sound of the Universe boiling? Analysis using the full fluid velocity profiles and NANOGrav 15-year data

Abstract In this paper, we analyse sound waves arising from a cosmic phase transition where the full velocity profile is taken into account as an explanation for the gravitational wave spectrum observed by multiple pulsar timing array groups. Unlike the broken power law used in the literature, in this scenario the power law after the peak depends on the macroscopic properties of the phase transition, allowing for a better fit with pulsar timing array (PTA) data. We compare the best fit with that obtained using the usual broken power law and, unsurprisingly, find a better fit with the gravitational wave (GW) spectrum that utilizes the full velocity profile. Even more importantly, the thermal parameters that produce the best fit are quite different. We then discuss models that can produce the best-fit point and complementary probes using CMB experiments and searches for light particles in DUNE, IceCUBE-Gen2, neutrinoless doubleβ-decay, and forward physics facilities (FPF) at the LHC like FASERν, etc.

Astronomy & Astrophysics↗