Search NASA⌕ Search

SEARCH · Search NASA

Results for “Models, Statistical”

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 469 records · Page 26

Validation of MCNP Critical Benchmark Models of PU-MET-FAST-016 [Slides]

The validation of PU-MET-FAST-016 contributed to the centralized LANL benchmark repository currently under development. The revisions made the MCNP models statistically, significantly more similar to the benchmark models. The overall impact on the USL is negligible. The revisions to the PU-MET-FAST-016 models provide value to the Los Alamos Benchmark Suite without invalidating past Whisper results.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Predictive Indicators of the Performance of Large Language Models

In several mission contexts, it is desirable to estimate the performance of large language models (LLMs) on tasks that we cannot run directly. In light of published “scaling laws” our hypothesis is that some tasks should be consistently more challenging than others based on characteristics of the task. The goal of this project was to begin quantifying how much information about LLM performance can be gained from the features of a model and a task. Two of our statistical models struggled to converge. Pass/fail test results may provide limited information for inference beyond model quality and task difficulty, but we see no evidence at this time for significant feature interaction effect sizes, arguing for simple models. Future work extending the models to capitalize on perplexity of ground truth answers is suggested. This project also introduces “Depth of Knowledge Variant Testing” as a strategy for more finely assessing language models on open domain question and answer tasks. We developed sets of questions that ask a language model to produce similar information while demonstrating increasing depth of knowledge, and also relabeled existing Q&A test questions with their depth of knowledge. Our results suggest further consideration of Bloom’s taxonomy and further refinement of prompts to properly elicit information at varying depths. In the course of this work, we set up a basic infrastructure for standardizing tasks and testing many language models on these tasks. In addition to testing the predictive quality of model features and performance across test suites, with this project we have introduced two new task features to contextualize each test question: the Dewey Classification main category of information covered, and the Bloom’s taxonomy level that corresponds to the depth of knowledge probed by the question. Splits across these and other features produced over five hundred task subtypes with distinct feature vectors, which we tested on half a dozen models.

97 MATHEMATICS AND COMPUTING↗

Bayesian chain graph models to characterize microbe-environment dynamics

Microbiome data require statistical models that can simultaneously decode microbes' reaction to the environment and interactions among microbes. While a multiresponse linear regression model seems like a straight-forward solution, we argue that treating it as a graphical model is problematic given that the regression coefficient matrix does not encode the conditional dependence structure between response and predictor nodes. This observation is especially important in biological settings when we have prior knowledge on the edges from specific experimental interventions that can only be properly encoded under a conditional dependence model. Here, we propose a chain graph model with two sets of nodes (predictors and responses) whose solution yields a graph with edges that indeed represent conditional dependence, thus agreeing with the experimenter's intuition on the average behavior of nodes under treatment. The solution to our model is sparse via the Bayesian linear regression (LASSO). In addition, we propose an adaptive extension so that different shrinkages can be applied to different edges to incorporate edge-specific prior knowledge. Our model is computationally inexpensive through an efficient Gibbs sampling algorithm and can account for binary, counting, and compositional responses via an appropriate hierarchical structure. We test the performance of our model in a variety of simulated datasets, thereby showing superior performance to state-of-the-art approaches. We further apply our model to human gut and soil microbial compositional datasets, and we highlight that CG-LASSO can estimate biologically meaningful network structures in the data.

compositional data↗

Quantifying Uncertainty in Epidemiological Models

Modern epidemiology has made use of a number of mathematical models, including ordinary differential equation (ODE) based models and agent based models (ABMs) to describe the dynamics of how a disease may spread within a population and enable the rational design of strategies for intervention that effectively contain the spread of the disease. Although such predictions are of fundamental importance in preventing the next global pandemic, there is a significant gap in trusting the outcomes/predictions solely based on such models. Hence, there is a need to develop approaches such that mathematical models can be calibrated against historical data. In addition, there is a need to develop rigorous uncertainty quantification approaches that can provide insights into when a model will fail and characterize the confidence in the (possibly multiple) model outcomes/predictions, when such retrospective analysis cannot be performed. In this paper, we outline an approach to develop uncertainty quantification approaches for epidemiological models using formal methods and model checking. By specifying the outcomes expected from a model in a suitable spatio-temporal logic, we use probabilistic model checking methods to quantify the probability with which the epidemiological model satisfies the specification. We argue that statistical model checking methods can solve the uncertainty quantification problem for complex epidemiological models.

model checking↗

Effects of nuclear orientation on fusion and fission process for reactions using actinide target nuclei

Fission fragment mass distributions in the reaction of {sup 30}Si+{sup 238}U were measured at the energies around the Coulomb barrier. At the above-barrier energies, the mass distribution showed Gaussian shape. At the sub-barrier energies, triple-humped distribution was observed, which consists of symmetric fission and asymmetric fission peaked at A{sub L}/A{sub H}approx =90/178. The asymmetric fission should be attributed to quasifission from the results of the measured evaporation residue (ER) cross-sections produced by {sup 30}Si+{sup 238}U. The cross-section for {sup 263}Sg at the above-barrier energy agree with the statistical model calculation which assumes that the measured fission cross-sections are equal to the fusion cross-sections, whereas the one for {sup 264}Sg measured at the sub-barrier energy is smaller than the calculation, indicating the presence for quasifission. We also report the results on the fragment mass distributions for {sup 36,34}S+{sup 238}U and {sup 40}Ar+{sup 238}U.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Measurement of the (p,{gamma}) Cross Sections of {sup 46}Ti, {sup 64}Zn, {sup 114}Sn and {sup 116}Sn at Astrophysically Relevant Energies

The total cross sections for the {sup 116}Sn(p,{gamma}){sup 117}Sb, {sup 114}Sn(p,{gamma}){sup 115}Sb, {sup 64}Zn(p,{gamma}){sup 65}Ga, and {sup 46}Ti(p,{gamma}){sup 47}V reactions have been measured using an activation technique for protons with center-of-mass energies between about 1 and 4 MeV. Radiative proton captures on {sup 116}Sn and {sup 64}Zn are compared to previous measurements while extending the measured energy range. From these measurements, astrophysical S-factors have been deduced by integrating NON-SMOKER results which have been scaled to the deduced S-factors in the measured energy range. Comparisons are made to thermonuclear reaction rates from the MOST and NON-SMOKER statistical model calculations.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Lithium-ion battery physics and statistics-based state of health model

A pseudo-2d model using COMSOL Multiphysics® software is developed to simulate performance and performance degradation of Li-ion batteries consisting of layered and olivine cathodes with graphite anode when subjected to peak shaving grid service. Multiple degradation pathways are considered, including solid electrolyte interphase (SEI) formation and breakdown at the anode, cathode dissolution and its synergistic effect on SEI formation at the anode. The model is validated by simulating commercial cylindrical cell performance. A global model is developed to simulate performance across all chemistries, along with individual chemistry models using global model parameters as initial values. There is good agreement between these models for various optimization parameters such as SEI equilibrium potential, cathode dissolution exchange current density, solvent diffusivity in the SEI and SEI ionic conductivity. To circumvent time constraints related to the COMSOL model, a 0d global model is developed which fits data well and provides more clarity on differences in cathode dissolution exchange current density. Again, good agreement for various optimization parameters is obtained among the COMSOL global & individual chemistry models and the 0-d model. The lessons learned from the physics-based model is used to develop a top down statistics-based model using current, voltage and anode volumetric change per mole lithium intercalated, along with their interactions as degradation predictors. This model predicts out of sample degradation for multiple grid services and electric vehicle drive cycle with high accuracy and provides the pathway to develop an efficient battery management system combining machine learning and findings from physics-based computationally intensive algorithms.

Crawford, Aladsair J.↗

Evaporation residue excitation function from complete fusion of {sup 19}F with {sup 184}W

Evaporation residue cross sections for {sup 19}F+{sup 184}W have been measured at beam energies in the range of 90-130 MeV using the Heavy Ion Reaction Analyzer at the Inter University Accelerator Centre. A new approach, based on Monte Carlo calculations, has been adopted for estimating evaporation residue (ER) transmission efficiency. The challenges in the detection of ERs caused by the presence of isomeric states with longer half-lives have been thoroughly discussed. Measured cross sections have been fitted with statistical model calculations treating the fission barrier as an adjustable parameter. Comparison with other neighboring systems with nearly similar entrance-channel mass asymmetry indicates the onset of instability in the compound nucleus {sub 83}{sup 203}Bi{sub 120} after crossing of the Z=82 shell.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Isotopic dependence of the cross section for the induced fission of heavy nuclei

The cross sections for the induced fission of {sup 211-223}Ra, {sup 203-211}Rn, and {sup 221-231}Th nuclei undergoing peripheral collisions with {sup 208}Pb nuclei are calculated on the basis of the statistical model. The role of the N = 126 neutron shell is studied. The level density in excited nuclei is determined within the Fermi gas model and a model that takes into account the collective enhancement of the level density. The inclusion of a particle-hole excitation in addition to a collective Coulomb excitation makes it possible to obtain a satisfactory description of experimental cross sections for the fission of radium isotopes. The calculated ratios of the cross sections for the induced fission of {sup 236}U ({sup 237}U) and {sup 238}U ({sup 239}U) nuclei agree with experimental data.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Studies of nuclei under the extreme conditions of density, temperature, isospin asymmetry and the phase diagram of hadronic matter

The main emphasis of the entire project is on issues having to do with medium energy and ultra-relativistic energy and heavy ion collisions. A major goal of both theory and experiment is to study properties of hot dense nuclear matter under various extreme conditions and to map out the phase diagram in density or chemical potential and temperature. My studies in medium energy nuclear collisions focused on the liquid-gas phase transition and cluster yields from such transitions. Here I developed both the statistical model of nuclear multi-fragmentation and also a mean field theory.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Progress in Developing Finite Element Models Replicating Flexural Graphite Testing

This report documents the status of flexural strength evaluations from current ASTM procedures and of developing finite element models predicting the probability of failure. This work is covered under QLD REC-00030. Flexural testing procedures of the American Society for Testing and Materials (ASTM) assume a linear elastic material that has the same moduli for tension and compression. Contrary to this assumption, graphite is known to have different moduli for tension and compression. A finite element model was developed and demonstrated that accounts for the difference in moduli tension and compression. Brittle materials such as graphite exhibit significant scatter in tensile strength, so probabilistic design approaches must be used when designing components fabricated from brittle materials. ASTM procedures predicting probability of failure in ceramics were compared to methods from the current version of the ASME graphite core components rules predicting probability of failure. Using the ASTM procedures yields failure curves at lower applied forces than the ASME rules. A journal paper was published in the Journal of Nuclear Engineering and Design exploring the statistical models of fracture in graphite.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

TH-CD-207A-05: Lung Surface Deformation Vector Fields Prediction by Monitoring Respiratory Surrogate Signals

Purpose: In this study, we developed and evaluated a method for predicting lung surface deformation vector fields (SDVFs) based on surrogate signals such as chest and abdomen motion at selected locations and spirometry measurements. Methods: A Patient-specific 3D triangular surface mesh of the lung region at end-expiration (EE) phase was obtained by threshold-based segmentation method. For each patient, a spirometer recorded the flow volume changes of the lungs; and 192 selected points at a regular spacing of 2cm X 2cm matrix points over a total area of 34cm X 24cm on the surface of chest and abdomen was used to detect chest wall motions. Preprocessing techniques such as QR factorization with column pivoting (QRCP) were employed to remove redundant observations of the chest and abdominal area. To create a statistical model between the lung surface and the corresponding surrogate signals, we developed a predictive model based on canonical ridge regression (CRR). Two unique weighting vectors were selected for each vertex on the surface of the lung, and they were optimized during the training process using the all other phases of 4D-CT except the end-inspiration (EI) phase. These parameters were employed to predict the vertices locations of a testing data set, which was the EI phase of 4D-CT. Results: For ten lung cancer patients, the deformation vector field of each vertex of lung surface mesh was estimated from the external motion at selected positions on the chest wall surface plus spirometry measurements. The average estimation of 98th percentile of error was less than 1 mm (AP= 0.85, RL= 0.61, and SI= 0.82). Conclusion: The developed predictive model provides a non-invasive approach to derive lung boundary condition. Together with personalized biomechanical respiration modelling, the proposed model can be used to derive the lung tumor motion during radiation therapy accurately from non-invasive measurements.

60 APPLIED LIFE SCIENCES↗

Deformed Materials: Towards a Theory of Materials Morphology Dynamics

This grant supported work on the response of crystals to external stress. Our primary work described how disordered structural materials break in two (statistical models of fracture in disordered materials), studied models of deformation bursts (avalanches) that mediate deformation on the microscale, and developed continuum dislocation dynamics models for plastic deformation (as when scooping ice cream bends a spoon, Fig. 9). Glass is brittle -- it breaks with almost atomically smooth fracture surfaces. Many metals are ductile -- when they break, the fracture surface is locally sheared and stretched, and it is this damage that makes them hard to break. Bone and seashells are made of brittle material, but they are strong because they are disordered -- lots of little cracks form as they are sheared and near the fracture surface, diluting the external force. We have studied materials like bone and seashells using simulations, mathematical tools, and statistical mechanics models from physics. In particular, we studied the extreme values of fracture strengths (how likely will a beam in a bridge break far below its design strength), and found that the traditional engineering tools could be improved greatly. We also studied fascinating crackling-noise precursors -- systems which formed microcracks of a broad range of sizes before they broke. Ductile metals under stress undergo irreversible plastic deformation -- the planes of atoms must slide across one another (through the motion of dislocations) to change the overall shape in response to the external force. Microscopically, the dislocations in crystals move in bursts of a broad range of sizes (termed 'avalanches' in the statistical mechanics community, whose motion is deemed 'crackling noise'). In this grant period, we resolved a longstanding mystery about the average shape of avalanches of fixed duration (using tools related to an emergent scale invariance), we developed the fundamental theory describing the shapes of avalanches and how they are affected by the edges of the microscope viewing window, we found that slow creep of dislocations can trigger an oscillating response explaining recent experiments, we explained avalanches under external voltage, and we have studied how avalanches in experiments on the microscale relate to deformation of large samples. Inside the crystals forming the metal, the dislocations arrange into mysterious cellular structures, usually ignored in theories of plasticity. Writing a natural continuum theory for dislocation dynamics, we found that it spontaneously formed walls -- much like models of traffic jams and sonic booms. These walls formed rather realistic cellular structures, which we examined in great detail -- our walls formed fractal structures with fascinating scaling properties, related to those found in turbulent fluids. We found, however, that the numerical and mathematical tools available to solve our equations were not flexible enough to incorporate materials-specific information, and our models did not show the dislocation avalanches seen experimentally. In the last year of this grant, we wrote an invited review article, explaining how plastic flow in metals shares features with other stressed materials, and how tools of statistical physics used in these other systems might be crucial for understanding plasticity.

36 MATERIALS SCIENCE↗

Extra gamma-ray strength for {sup 116,117}Sn arising from pygmy dipole resonance

Photoneutron cross sections were measured for {sup 117}Sn and {sup 116}Sn near neutron thresholds with quasi-monochromatic laser Compton scattering gamma-rays. The measured cross sections for {sup 117}Sn and {sup 116}Sn are strongly enhanced from the threshold behavior expected for L = 1 neutron emissions after E1 photoexcitation. This suggests the presence of extra gamma-ray strength in the low-energy tail of the giant dipole resonance. The present cross sections were analyzed together with radiative neutron capture cross sections for {sup 116}Sn within the framework of the statistical model calculation. It is shown that the extra gamma-ray strength can be interpreted as pygmy E1 resonance which was previously reported in the nuclear resonance fluorescence experiment for {sup 116}Sn and {sup 124}Sn.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Revisiting the photodissociation dynamics of the phenyl radical

We have reinvestigated the photodissociation dynamics of the phenyl radical at 248 nm and 193 nm via photofragment translational spectroscopy under a variety of experimental conditions aimed at reducing the nascent internal energy of the phenyl radical and eliminating signal from contaminants. Under these optimized conditions, slower translational energy (P(E{sub T})) distributions for H-atom loss were seen at both wavelengths than in previously reported work. At 193 nm, the branching ratio for C{sub 2}H{sub 2} loss vs. H-atom loss was found to be 0.2 ± 0.1, a significantly lower value than was obtained previously in our laboratory. The new branching ratio agrees with calculated Rice-Ramsperger-Kassel-Marcus rate constants, suggesting that the photodissociation of the phenyl radical at 193 nm can be treated using statistical models. The effects of experimental conditions on the P(E{sub T}) distributions and product branching ratios are discussed.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Using data-science approaches to unravel insights for enhanced transport of lithium ions in single-ion conducting polymer electrolyte

Solid polymer electrolytes have yet to achieve the an ionic conductivity > 1 mS/cm at room temperature for realistic applications. This target implies the need to reduce the effective energy barriers of ion transport in polymer electrolytes to around 20 kJ/mol. In this work, we combine information extracted from existing experimental results with theoretical calculations to provide insights into ion transport in single-ion conductors (SICs) with a focus on lithium ion SICs. Through the analysis of temperature-dependent ionic conductivity data obtained from the literature, we evaluate different methods of extracting energy barriers for lithium transport. The traditional Arrhenius fit to the temperature-dependent ionic conductivity data indicates that the Meyer-Neldel rule holds for SICs. However, the values of the fitting parameters remain unphysical. Our modified approach based on recent work (Macromolecules, 56, 15, 6051(2023)), which incorporates a fixed pre-exponential factor, reveals that the energy barriers exhibit temperature dependence over a wide range of temperatures. Using this approach, we identify a series of anions leading to the energy barriers less than 30 kJ/mol, which include trifluoromethane sulfonimide (TFSI), fluoromethane sulfonimide (FSI), and boron-based organic anions. In our efforts to design the next generation of anions, which can exhibit the energy barriers less than 20 kJ/mol, we focused on boron-containing SICs, and performed density functional theory (DFT) based calculations to connect the chemical structures via the binding energy of cation (lithium)-anion pairs with the experimentally derived effective energy barriers for ion transport. Not only have we identified a correlation between the binding energy and the energy barriers, but we also propose a strategy to design new boron-based anions by using the correlation. This combined approach involving experiments and theoretical calculations is capable of facilitating the identification of promising new anions, which can exhibit ionic conductivity $> 1$ mS/cm near room temperature, thereby expediting the development of novel superionic single-ion conducting polymer electrolytes. The published datasets include all the temperature-dependent ionic conductivity collected from the literature with literature DOIs, DFT calculated binding energies, and python scripts to analyze data, construct statistical models, and generate plots.

36 MATERIALS SCIENCE↗

Uncertainty Quantification and Management for Multi-scale Nuclear Materials Modeling

Understanding and improving microstructural mechanical stability in metals and alloys is central to the development of high strength and high ductility materials for cladding and cores structures in advanced fast reactors. Design and enhancement of radiation-induced damage tolerant alloys are facilitated by better understanding the connection of various unit processes to collective responses in a multiscale model chain, including: dislocation nucleation, absorption and desorption at interfaces; vacancy production, radiation-induced segregation of Cr and Ni at defect clusters (point defect sinks) in BCC Fe-Cr ferritic/martensitic steels; investigation of interaction of interstitials and vacancies with impurities (V, Nb, Ta, Mo, W, Al, Si, P, S); time evolution of swelling (cluster growth) phenomena of irradiated materials; and energetics and kinetics of dislocation bypass of defects formed by interstitial clustering and formation of prismatic loops, informing statistical models of continuum character with regard to processes of dislocation glide, vacancy agglomeration and swelling, climb and cross slip.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Effect of Holding Pressure on Microstructure and Mechanical Properties of A356 Aluminum Alloy

In this study, the effect of holding pressure on microstructure and mechanical properties of low-pressure die cast A356 aluminum alloy was investigated. The results showed that the application of high holding pressure (300 kPa) generated castings with denser structure and superior mechanical properties. By increasing the holding pressure up to 300 kPa, the size of secondary dendrite arm spacing greatly reduced by 22.7% at the cooling rate of 1°C/s and decreased by 12.8% at 10°C/s. The Feret’s diameter and aspect ratio of eutectic silicon particles decreased by 8.4 and 5.1% at the cooling rate of 1°C/s and decreased by 9.3 and 6.4% at 10°C/s, respectively. Meanwhile, the density of A356 aluminum alloy increased to 2.678 g/cm{sup 3} and the area fraction of porosity decreased to 0.035%. Thus, tensile properties of A356 aluminum alloy obtained at high holding pressure were enhanced, especially the ductility. All these could be associated with the better filling capability and faster cooling rate caused by high holding pressure. In the analytical range of experimental conditions, the correlation of mechanical properties with process parameters was established by statistical models to predict the ultimate tensile strength and elongation of low-pressure die cast A356 aluminum alloy.

36 MATERIALS SCIENCE↗