Search NASA⌕ Search

SEARCH · Search NASA

Results for “multiple regression”

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 1,117 records · Page 62

Uncertainty Quantification using Deep Ensembles for Decision Making in Cyber-Physical-Human Systems

In this paper and its companion, Differential Equation Approximation Using Gradient-Boosted Quantile Regression, Robison et al., we examine an approach to quantifying model uncertainty with the aim of increasing the trustworthiness of computational models in human-machine interactions. In Differential Equation Approximation Using Gradient-Boosted Quantile Regression, we focus on gradient-boosted decision trees, while in this one, we give more details about deep ensembles. Uncertainty quantification is crucial for building trustworthy autonomous decision-making agents in human-machine teams. There are two types of uncertainties: aleatoric and epistemic. The former is related to the inherent stochasticity (noise) of the process, whereas the latter is associated with the lack of knowledge or representation capability of models, such as neural networks. By lack of knowledge, we mean the model’s inability to accurately predict outputs for all possible inputs. The aleatory uncertainty can be estimated fairly easily with, for example, filters, whereas epistemic uncertainty is challenging to compute. This paper uses deep ensembles to quantify both aleatory and epistemic uncertainty. It can act as an uncertainty-aware surrogate transition model for decision-making frameworks. "Uncertainty-aware" means that the surrogate transition model should make predictions along with confidence in those predictions. In the context of decision-making, the transition models are ordinary differential equations (ODEs). Since ODEs can be simulated to make one-step or multi-step predictions, a good surrogate model for them should perform reasonably well in both modes. In a multi-step approach, the trajectory sampling method TS∞ was used to propagate uncertainty over multiple steps. The cartpole dynamical system was selected to demonstrate the ability of deep ensembles as good surrogate transition models for decision-making frameworks. The deep ensembles modeled the dynamics of cartpole ODEs and made uncertainty-aware predictions in single-step and multi-step transition modes.

CPH systems↗

Uncertainty quantification for radiation measurements: Bottom-up error variance estimation using calibration information

One example of top-down uncertainty quantification (UQ) involves comparing two or more measurements on each of multiple items. One example of bottom-up UQ expresses a measurement result as a function of one or more input variables that have associated errors, such as a measured count rate, which individually (or collectively) can be evaluated for impact on the uncertainty in the resulting measured value. In practice, it is often found that top-down UQ exhibits larger error variances than bottom-up UQ because some error sources are present in the fielded assay methods used in top-down UQ that are not present (or not recognized) in the assay studies used in bottom-up UQ One would like better consistency between the two approaches in order to claim understanding of the measurement process. The purpose of this paper is to refine bottom-up uncertainty estimation by using calibration information so that if there are no unknown error sources, the refined bottom-up uncertainty estimate will agree with the top-down uncertainty estimate to within a specified tolerance. Then, in practice, if the top-down uncertainty estimate is larger than the refined bottom-up uncertainty estimate by more than the specified tolerance, there must be omitted sources of error beyond those predicted from calibration uncertainty. The paper develops a refined bottom-up uncertainty approach for four cases of simple linear calibration: (1) inverse regression with negligible error in predictors, (2) inverse regression with non negligible error in predictors, (3) classical regression followed by inversion with negligible error in predictors, and (4) classical regression followed by inversion with non-negligible errors in predictors. Our illustrations are of general interest, but are drawn from our experience with nuclear material assay by non-destructive assay. The main example we use is gamma spectroscopy that applies the enrichment meter principle. Previous papers that ignore error in predictors have shown a tendency for inverse regression to have lower error variance than classical regression followed by inversion. Our paper supports that tendency both with and without error in predictors. Furthermore, the paper shows that calibration parameter estimates using error in predictor methods perform worse than without using error in predictor methods in the case of inverse regression, but perform better than without using error in predictor methods in the case of classical regression followed by inversion. Both inverse and classical regression involve the ratio of dependent random variables; therefore, the assumed error distribution(s) will matter in parameter estimation and in uncertainty calculations. Mainly for that reason, calibration using a single predictor is distinct from simple regression, and it has not been thoroughly treated in the literature, nor in the ISO Guide to the Expression of Uncertainty in Measurements (GUM). This refined approach is based on simulation, because we illustrate that analytical approximations are not adequate when there are, for example, 10 or fewer calibration measurements, which is common in calibration applications, each consisting of measured responses from known quantities.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

A Method for Calculating the Probability of Successfully Completing a Rocket Propulsion Ground Test

Propulsion ground test facilities face the daily challenges of scheduling multiple customers into limited facility space and successfully completing their propulsion test projects. Due to budgetary and schedule constraints, NASA and industry customers are pushing to test more components, for less money, in a shorter period of time. As these new rocket engine component test programs are undertaken, the lack of technology maturity in the test articles, combined with pushing the test facilities capabilities to their limits, tends to lead to an increase in facility breakdowns and unsuccessful tests. Over the last five years Stennis Space Center's propulsion test facilities have performed hundreds of tests, collected thousands of seconds of test data, and broken numerous test facility and test article parts. While various initiatives have been implemented to provide better propulsion test techniques and improve the quality, reliability, and maintainability of goods and parts used in the propulsion test facilities, unexpected failures during testing still occur quite regularly due to the harsh environment in which the propulsion test facilities operate. Previous attempts at modeling the lifecycle of a propulsion component test project have met with little success. Each of the attempts suffered form incomplete or inconsistent data on which to base the models. By focusing on the actual test phase of the tests project rather than the formulation, design or construction phases of the test project, the quality and quantity of available data increases dramatically. A logistic regression model has been developed form the data collected over the last five years, allowing the probability of successfully completing a rocket propulsion component test to be calculated. A logistic regression model is a mathematical modeling approach that can be used to describe the relationship of several independent predictor variables X(sub 1), X(sub 2),..,X(sub k) to a binary or dichotomous dependent variable Y, where Y can only be one of two possible outcomes, in this case Success or Failure. Logistic regression has primarily been used in the fields of epidemiology and biomedical research, but lends itself to many other applications. As indicated the use of logistic regression is not new, however, modeling propulsion ground test facilities using logistic regression is both a new and unique application of the statistical technique. Results from the models provide project managers with insight and confidence into the affectivity of rocket engine component ground test projects. The initial success in modeling rocket propulsion ground test projects clears the way for more complex models to be developed in this area.

Messer, Bradley P.↗

SU-E-U-02: The Development of a Practical Ultrasonic System for Cross-Sectional Imaging of Small Animals

Purpose: To test the feasibility of developing a practical medium frequency ultrasound tomography method for small animal imaging. The ability to produce cross-sectional or full body images of a live small animal using a low-cost tabletop ultrasound scanner without any special license would be very beneficial to long term biological studies, where repeated scanning is often required over an extended period of time. Methods: The cross sectional images were produced by compounding multiple B-scans of a laboratory phantom or an animal acquired at different projection angles. Two imaging systems were used to test the concept. The first system included a programmable 64-channel phased array controller driving a 128-channel, 5–10 MHz linear probe to produce 143 B-Mode projections of the spinning object. The second system designed and manufactured in house, produced 64 or 128 B-Mode projections with a single unfocused 8 MHz transducer scanning with a 0.116 mm step size. Results: The phased array system provided good penetration through the phantoms/mice (with the exception of the lungs) and allowed to acquire data in a very short time. The cross-sectional images have enough resolution and dynamic range to detect both high- and low-contrast organs. The single transducer system takes longer to scan, and the data require more sophisticated processing. To date, our images allow seeing details as small as 1–2 mm in the phantoms and in small animals, with the contrast mostly due to highly reflecting bones and air inclusions. Conclusion: The work indicates that very detailed and anatomically correct images can be created by relatively simple and inexpensive means. With more advanced algorithms and improved system design, scan time can be reduced considerably, enabling high-resolution full 3D imaging. This will allow for quick and easy scans that can help monitor tumor growth and/or regression without contributing any dose to the animal. The authors would like to acknowledge the financial and engineering support from Tessonics.

60 APPLIED LIFE SCIENCES↗

Challenges of Using Indistinguishable from Background as a Release Criterion at FUSRAP Projects - 18157

The US Army Corps of Engineers (USACE) utilizes the Multi-Agency Radiation Survey and Site Investigation Manual (MARSSIM) approach to site release on Formerly Utilized Sites Remedial Action Program (FUSRAP) projects. Use of Indistinguishable from Background (IFB) as a release criterion is discussed in MARSSIM as Scenario B; however, the discussion is arguably limited and refers the user to other documents. USACE has utilized the IFB approach at select FUSRAP projects. USACE experience is that use of IFB as a release criterion has been difficult to implement at best. There are many challenges in using the IFB approach. These include: release criterion is required per MARSSIM; multiple reference areas need to be established; and data presentation. Potential methods to overcome these challenges are discussed. USACE has utilized alternative statistical approaches as well as the IFB approach discussed in MARSSIM for specific purposes at FUSRAP sites. Alternatives to the IFB approach discussed in this paper include other statistical tests such as the Tukey Kramer test. These alternatives may be better suited to release an area of a property prior to the development of release criteria. USACE has found using the IFB criterion useful in certain specific circumstances. These include post MARSSIM Scenario A final status survey data evaluations and for review of historical data. The challenges, alternatives, and circumstances to use the IFB approach are discussed. Other environmental remediation projects attempting to utilize IFB as a criterion may benefit from the USACE experiences as discussed herein. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Recommendations for Using Noise Monitors to Estimate Noise Exposure During X-59 Community Tests

A low fidelity simulation approach is used to explore how to place and use noise monitors during X-59 QueSST community tests, where people’s annoyance to the noise produced by the X-59 aircraft will be gathered. Several recommendations are provided including: 1) the desired number of sparsely spaced noise monitor sites within the survey area, 2) whether to group and average measurements across multiple noise monitors located at a site, 3) what spacing should be used if grouped noise monitors are used, 4) an approach to mitigate ambient noise contamination at the measurement sites, 5) a method to combine empirical and predicted dose estimates to provide a single dose estimate for respondents, and 6) assessing how changes in turbulence intensity and array configuration affect dose uncertainty. To make these recommendations, the error that is expected when fitting contrived, smoothly varying sonic boom “reference exposure surfaces” is studied when a spatially sparse and scattered set of samples is used as responses for the fit. The reference exposure surfaces mimic the sonic boom exposure at ground level that might be expected in the X-59 survey area in the absence of atmospheric turbulence, ambient noise, and other localized effects. The spatial extent of these surfaces varies and is representative of the different survey area sizes that might be expected during future X-59 community overflight tests. These contrived reference surfaces are sampled, and those reference samples are then perturbed to mimic atmospheric turbulence, ambient noise and other localized effects that might affect noise monitor measurements within overflown communities. Two different surface fitting methods are investigated when fitting these perturbed samples to approximate the reference surface. The first method uses interpolation between the perturbed data at the scattered sites to compute the fit. The second method fits a polynomial surface model to the perturbed data using ordinary least squares regression analysis. For both fitting methods, the root mean square fit error is computed from the pointwise difference between the fit surface and the reference surface as the count and configuration of the sites is varied while also averaging the error across many different realizations of both the smooth variation of the reference exposure surface and the random, localized perturbations at the sample sites. Different site configurations are compared using this error statistic to make the recommendations noted above. Additionally, the two fitting approaches (interpolation vs linear regression) are compared based on the fit error observed in these simulations. These analyses, comparisons, and recommendations should inform future decisions on the noise monitor placement and the methods used to analyze the noise monitor data that is collected during X-59 community overflights.

sonic boom↗

Biomass Retrieval from L-Band Polarimetric UAVSAR Backscatter and PRISM Stereo Imagery

The forest above-ground biomass (AGB) and spatial distribution of vegetation elements have profound effects on the productivity and biodiversity of terrestrial ecosystems. In this paper, we evaluated biomass estimation from L-band Synthetic Aperture Radar (SAR) data acquired by National Aeronautics and Space Administration (NASA) Uninhabited Aerial Vehicle SAR (UAVSAR) and the improvement of accuracy by adding canopy height information derived from stereo imagery acquired by Japan Aerospace Exploration Agency (JAXA) Panchromatic Remote Sensing Instrument for Stereo Mapping (PRISM) on-board the Advanced Land Observing Satellite (ALOS). Various models for prediction of forest biomass from UAVSAR data were investigated at pixel sizes of 1/4 ha (50 m x 50 m) and 1 ha. The variance inflation factor (VIF) was calculated for each of the explanatory variables in multivariable regression models to assess the multi-collinearity between explanatory variables. In addition, the t-and p-values were used to interpret the significance of the coefficients of each explanatory variables. The R(exp. 2), Root Mean Square Error (RMSE), bias and Akaike information criterion (AIC), and leave-one-out cross-validation (LOOCV) and bootstrapping were used to validate models. At 1/4-ha scale, the R(exp. 2) and RMSE of biomass estimation from a model using a single track of polarimetric UAVSAR data were 0.59 and 52.08 Mg/ha. With canopy height from PRISM as additional independent variable, R(exp. 2) increased to 0.76 and RMSE decreased to 39.74 Mg/ha (28.24%). At 1-ha scale, the RMSE of biomass estimation based on UAVSAR data of a single track was 39.42 Mg/ha with a R(exp. 2) of 0.77. With the canopy height from PRISM, R(exp. 2) increased to 0.86 and RMSE decreased to 29.47 Mg/ha (20.18%). The models using UAVSAR data alone underestimated biomass at levels above approximately 150 Mg/ha showing the saturation phenomenon. Adding canopy height from PRISM stereo imagery significantly improved the biomass estimation and elevated the saturation level in estimating biomass. Combined use of UAVSAR data acquired from opposite directions (odd and even tracks) slightly improved the biomass estimation.Combined use of UAVSAR data acquired from opposite directions (odd and even tracks) slightly improved the biomass estimation at 1/4-ha scale, R(exp. 2) increased from 0.59 to 0.66 and RMSE reduced from 52.08 to 48.57 Mg/ha. Averaging multiple acquisitions of UAVSAR data from the same look azimuth direction did not improve biomass estimation. A biomass map derived from NASA's LVIS (Laser Vegetation Imaging System) wave-form data was used as a reference for evaluation of the biomass maps from these models. The study has also shown that the errors decreased when deciduous, evergreen, and mixed forests were modeled separately but the improvement was not significant

Zhang, Zhiyu↗

Statistical abstraction of high-fidelity CO2 pressure histories in 2-D, uniform, cylindrical domains

Long-term, deep, geologic sequestration of carbon dioxide (CO{sub 2}) is being evaluated as a world-wide strategy for limiting anthropogenic carbon emissions to the atmosphere. A key element of this evaluation is quantification of the ancillary risks associated with this fundamentally new linkage between the global energy economy and the subsurface ecosphere. Quantitative risk assessment methods traditionally enumerate operational scenarios and describe the multiple physical responses that may ensue from each scenario depending on the quality of information that is available to describe identified system dependencies. For example, multiplepoint injection of compressed CO{sub 2} into a geologic reservoir having a nominal stratigraphy will create a pressurized zone of liquid that migrates through the rock. Scenarios that postulate CO{sub 2} encountering previously undetected wells or natural fractures in the caprock that represent leakage paths to the surface must be treated in a probabilistic format that accommodates unknown details in the subsurface geology. Fluid pressure in the reservoir at the location of the potential transport path drives any potential leakage that might occur, so the spatial and temporal distribution of CO{sub 2} overpressure represents an important metric for numeric simulation. State-of-the-art geologic transport models like FEHM, TUFF, and ECLIPSE (Refs. 1, 2, 3) can accurately simulate multi phase gas migration in a fully characterized geologic domain. However, each simulation can require time periods ranging between minutes and hours to achieve acceptable numerical performance, so it is often impractical to link predictive physics models directly in a quantitative risk assessment that will require transport estimates for thousands of scenarios. When direct computation is not possible, a library of high-fidelity calculations can sometimes be distilled to a simplified statistical correlation that spans the variability in all relevant input parameters while retaining acceptable accuracy in the key predicted qu antities. Essentially, numerical calculations can be used as data for an exploratory trend analysis just as one might regress predictive equations against laboratory measurements to determine unknown parameter values. Statistical correlations are derived here to reproduce radial overpressure as a function of time and position for CO{sub 2} injected along the centerline of cylindrical geologic domains. FEHM was used to compute two-phase pressure histories in a suite of simulations that varied (1) initial pressure, (2) vertical reservoir thickness, (3) domain radius, (4) uniform permeability, and (5) mass injection rate. The simulations include both a 50-yr injection phase and a 50-yr relaxation phase. The correlations are based on a two-step fitting paradigm that first captures the shape of an entire radial pressure profile for a given time and input condition, and then correlates the shape parameters as generalized power products of 6 dimensionless parameters. Estimated overpressures are accurate within a factor of 1.25 compared to the suite of simulated values. The correlations can be used to evaluate rapidly local temporal overpressure for any continuous values within the defined space of input parameters. Inversion of the pressure correlation is also demonstrated to determine the maximum injection rate corresponding to an assumed limiting fracture pressure.

54 ENVIRONMENTAL SCIENCES↗

Experience Report for WOPR

One of the purposes of the SQA effort at LLNL is to attempt to determine the 'goodness' of the research codes used for various scientific applications. Typically these are two and three dimensional multi-physics simulation and modeling codes. These legacy research codes are used for applciations such as atmospheric dispersion modeling and analysis and prediction of the performance of engineered systems. These codes are continually subjected to automated regression test suites consisting of verified and validated expected results. Code is managed in repositories. Experience level of developers is high in the knowledge domain, platforms, and languages used. Code size of the multi-physics code used in this study was 578,242 lines excluding comment and blank lines or 5538.7 function points. Languages were 70% C++, 20% C, and 10% Fortran. The code has 130 users and a development team of 14 and an embedded SQE. The code has achieved 100% prime feature test coverage, 73.6% functional test coverage, and 71.5% statement test coverage. The average cyclomatic complexity of the code was 6.25. The codes have evolved over 10 years. Research codes are challenging because there is a desire to balance agility with discipline as well as compliance with DOE standards. Agility is important to allow experimentation with new algorithms and addition of the latest physics features. Discipline is important to increase the quality of the codes. Automation of processes and defect prevention/detection are deployed throughout the software development process. Since resarch codes are a small segment of the software industry, not much information exists in terms of reliability studies on these types of codes. This paper describes attempts to determine the goodness of these research codes. Goodness defined as both correctness of the codes and their fault densities. Correctness is determined by user interviews, peer review; feature based automated testing, and coverage measurement. This paper focuses on the fault density aspect of goodness and reliability of the codes in particular. The approach taken was to use multiple fault density prediction methods and compare results to actual experimentation and other industry studies on fault density. As a result of the predictions and experiments our confidence in the prediction methods was increased and our confidence in the goodness of the code from a fault density perspective was given more context. A large unintended benefit of these experiments was to find defects hidden for years in the codes when using the Monte Carlo reliability testing results to develop heuristic based bug driven tests.

99 GENERAL AND MISCELLANEOUS↗

Cloud-top pressure retrieval with DSCOVR EPIC oxygen A- and B-band observations

An analytic transfer inverse model for Earth Polychromatic Imaging Camera (EPIC) observations is proposed to retrieve the cloud-top pressure (CTP) with the consideration of in-cloud photon penetration. In this model, an analytic equation was developed to represent the reflection at the top of the atmosphere from above cloud, in cloud, and below cloud. The coefficients of this analytic equation can be derived from a series of EPIC simulations under different atmospheric conditions using a nonlinear regression algorithm. With estimated cloud pressure thickness, the CTP can be retrieved from EPIC observation data by solving the analytic equation. To simulate the EPIC measurements, a program package using the double-k approach was developed. Compared to line-by-line calculation, this approach can calculate high-accuracy results with a 100-fold computation time reduction. During the retrieval processes, two kinds of retrieval results, i.e., baseline CTP and retrieved CTP, are provided. The baseline CTP is derived without considering in-cloud photon penetration, and the retrieved CTP is derived by solving the analytic equation, taking into consideration in-cloud and below-cloud interactions. The retrieved CTPs for the oxygen A and B bands are smaller than their related baseline CTP. At the same time, both baseline CTP and retrieved CTP at the oxygen B band are larger than those at the oxygen A band. Compared to the difference in baseline CTP between the B band and A band, the difference in retrieved CTP between these two bands is generally reduced. Out of around 10 000 cases, in retrieved CTP between the A and B bands we found an average bias of 93 mb with a standard deviation of 81 mb. The cloud layer top pressure from Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) measurements is used for validation. Under single-layer cloud situations, the retrieved CTPs for the oxygen A band agree well with the CTPs from CALIPSO, the mean difference of which within 5 mb in the case study. Under multiple-layer cloud situations, the CTPs derived from EPIC measurements may be larger than the CTPs of high-level thin clouds due to the effect of photon penetration.

DSCOVR↗

A data–model approach to interpreting speleothem oxygen isotope records from monsoon regions

Reconstruction of past changes in monsoon climate from speleothem oxygen isotope (δ 18 O) records is complex because δ 18 O signals can be influenced by multiple factors including changes in precipitation, precipitation recycling over land, temperature at the moisture source, and changes in the moisture source region and transport pathway. Here, we analyse >150 speleothem records of the Speleothem Isotopes Synthesis and AnaLysis (SISAL) database to produce composite regional trends in δ 18 O in monsoon regions; compositing minimises the influence of site-specific karst and cave processes that can influence individual site records. We compare speleothem δ 18 O observations with isotope-enabled climate model simulations to investigate the specific climatic factors causing these regional trends. We focus on differences in δ 18 O signals between the mid-Holocene, the peak of the Last Interglacial (Marine Isotope Stage 5e) and the Last Glacial Maximum as well as on δ 18 O evolution through the Holocene. Differences in speleothem δ 18 O between the mid-Holocene and the Last Interglacial in the East Asian and Indian monsoons are small, despite the larger summer insolation values during the Last Interglacial. Last Glacial Maximum δ 18 O values are significantly less negative than interglacial values. Comparison with simulated glacial–interglacial δ 18 O shows that changes are principally driven by global shifts in temperature and regional precipitation. Holocene speleothem δ 18 O records show distinct and coherent regional trends. Trends are similar to summer insolation in India, China and southwestern South America, but they are different in the Indonesian–Australian region. Redundancy analysis shows that 37 % of Holocene variability can be accounted for by latitude and longitude, supporting the differentiation of records into individual monsoon regions. Regression analysis of simulated precipitation δ 18 O and climate variables show significant relationships between global Holocene monsoon δ 18 O trends and changes in precipitation, atmospheric circulation and (to a lesser extent) source area temperature, whereas precipitation recycling is non-significant. However, there are differences in regional-scale mechanisms: there are clear relationships between changes in precipitation and δ 18 O for India, southwestern South America and the Indonesian–Australian regions but not for the East Asian monsoon. Changes in atmospheric circulation contribute to δ 18 O trends in the East Asian, Indian and Indonesian–Australian monsoons, and a weak source area temperature effect is observed over southern and central America and Asia. Precipitation recycling is influential in southwestern South America and southern Africa. Overall, our analyses show that it is possible to differentiate the impacts of specific climatic mechanisms influencing precipitation δ 18 O and use this analysis to interpret changes in speleothem δ 18 O.

speleothem oxygen isotope records↗

The role of spray-enhanced swirl flow for combustion stabilization in a stratified-charge DISI engine

Implementing spray-guided stratified-charge direct-injection spark-ignited (DISI) engines is inhibited by the occurrence of misfire and partial burns. Engine-performance tests demonstrate that increasing engine speed induces combustion instability, but this deterioration can be prevented by generating swirling flow during the intake stroke. In-cylinder pressure-based heat-release analysis reveals that the appearance of poor-burn cycles is not solely dependent on the variability of early flame-kernel growth. Moreover, cycles can experience burning-rate regression during later combustion stages and may or may not recover before the end of the cycle. Thermodynamic analysis and optical diagnostics are used here to clarify why swirl improves the combustion repeatability from cycle to cycle. The fluid dynamics of swirl/spray interaction was previously demonstrated using high-speed PIV measurements of in-cylinder motored flow. It was found that the sprays of the multi-hole injector redistribute the intake-generated swirl flow momentum, thereby creating a better-centered higher angular-momentum vortex with reduced variability. The engine operation with high swirl was found to have significant improvement in cycle-to-cycle variations of both flow pattern and flow momentum. This paper is an extension of the previous work. Here, PIV measurements and flame imaging are applied to fired operation for studying how the swirl flow affects variability of ignition and subsequent combustion phases. PIV results for fired operation are consistent with the measurements made of motored flow. They demonstrate that the spark-plasma motion is highly correlated with the direction of the gas flow in the vicinity of the spark-plug gap. Without swirl, the plasma is randomly stretched towards either side of the spark plug, causing variability in the ignition of the two spray plumes that are straddling the spark plug. Conversely, swirl flow always convects the spark plasma towards one spray plume, causing a more repeatable ignition. The swirl decreases local RMS velocity, consistent with an observed reduction of early-burn variability. Broadband flame imaging demonstrates that with swirl, the flame consistently propagates in multiple directions to consume fuel–air mixtures within the piston bowl. In contrast, operation without swirl displays higher variability of flame-spread patterns, occasionally causing the appearance of partial-burn cycles.

33 ADVANCED PROPULSION SYSTEMS↗

VERIFICATION OF TRISO FUEL BURNUP USING MACHINE LEARNING ALGORITHMS

Pebble Bed Reactors are fueled with fuel pebbles that are circulated multiple times through the reactor vessel before discharge. During the normal operation of a PBR, ejected pebbles are returned to the reactor or discharged depending on the fuel burnup and physical condition of the pebbles. The burnup measurement is usually based on detected radiation signatures of fission products accumulated in the pebble fuel over burnup. Previous research has shown that height of photopeaks of fission products, such as 134Cs, 137Cs, 154Eu, etc., can be used independently or in combination to infer or predict the level of burnup in the fuel. However, it remains challenging to measure such complex sources due to self-shielding effects, strong radiation background and intervening materials. Another operational challenge is the required high throughput of burnup measurement, which necessitates limited measurement time and thus impacts quality of measured gamma-ray spectra. Hence, advanced spectral analysis methods are needed to analyze the noisy gamma spectra and predict the burnup values. We propose to use machine learning (ML) method to interpret gamma-ray spectra and predict the burnup values of the pebbles. ML has achieved widespread success and adoption across a few domains that require pattern recognition and analysis in varied data types. In this work, we apply three proven ML approaches - multilayer perceptrons, convolutional neural networks, and transformers - to the task of predicting fuel burnup from measured gamma spectra, and compile a dataset of simulated spectra for training and validation of the ML models. In this paper, we will discuss the network architecture of these three ML approaches and compare the performance of the simplest of these (MLP) to a standard linear regression.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

CAMERA: A method for cost-aware, adaptive, multifidelity, efficient reliability analysis

Estimating probability of failure in aerospace systems is a critical requirement for flight certification and qualification. Failure probability estimation involves resolving tails of probability distributions, and Monte Carlo sampling methods are intractable when expensive high-fidelity simulations have to be queried. Here, we propose a method to use models of multiple fidelities that trade accuracy for computational efficiency. Specifically, we propose the use of multifidelity Gaussian process models to efficiently fuse models at multiple fidelity, thereby offering a cheap surrogate model that emulates the original model at all fidelities. Furthermore, we propose a novel sequential acquisition function based experiment design framework that can automatically select samples from appropriate fidelity models to make predictions about quantities of interest at the highest fidelity. We use our proposed approach in an importance sampling setting and demonstrate our method on the failure level set and probability estimation on synthetic test functions and two real-world applications, namely, the reliability analysis of a gas turbine engine blade using a finite element method and a transonic aerodynamic wing test case using Reynolds-averaged Navier-Stokes equations. We show that our method predicts the failure boundary and probability more accurately and at a fraction of the computational cost compared with using just a single expensive high-fidelity model. Finally, we show that our sequential approach is guaranteed to asymptotically converge to the true failure boundary with high probability.

97 MATHEMATICS AND COMPUTING↗

Verification of Triso Fuel Burnup Using Machine Learning Algorithms

Pebble Bed Reactors are fueled with fuel pebbles that are circulated multiple times through the reactor vessel before discharge. During the normal operation of a PBR, ejected pebbles are returned to the reactor or discharged depending on the fuel burnup and physical condition of the pebbles. The burnup measurement is usually based on detected radiation signatures of fission products accumulated in the pebble fuel over burnup. Previous research has shown that height of photopeaks of fission products, such as 134 Cs, 137 Cs, 154 Eu, etc., can be used independently or in combination to infer or predict the level of burnup in the fuel. However, it remains challenging to measure such complex sources due to self-shielding effects, strong radiation background and intervening materials. Another operational challenge is the required high throughput of burnup measurement, which necessitates limited measurement time and thus impacts quality of measured gamma-ray spectra. Hence, advanced spectral analysis methods are needed to analyze the noisy gamma spectra and predict the burnup values. We propose to use machine learning (ML) method to interpret gamma-ray spectra and predict the burnup values of the pebbles. ML has achieved widespread success and adoption across a few domains that require pattern recognition and analysis in varied data types. In this work, we apply three proven ML approaches - multilayer perceptrons, convolutional neural networks, and transformers - to the task of predicting fuel burnup from measured gamma spectra, and compile a dataset of simulated spectra for training and validation of the ML models. In this paper, we will discuss the network architecture of these three ML approaches and compare the performance of the simplest of these (MLP) to a standard linear regression.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Defense Waste Processing Facility Nitric-Glycolic Flowsheet Chemical Process Cell Chemistry: Part 2

The conversions of nitrite to nitrate, the destruction of glycolate, and the conversion of glycolate to formate and oxalate were modeled for the Nitric-Glycolic flowsheet using data from Chemical Process Cell (CPC) simulant runs conducted by Savannah River National Laboratory (SRNL) from 2011 to 2016. The goal of this work was to develop empirical correlation models to predict these values from measurable variables from the chemical process so that these quantities could be predicted a-priori from the sludge or simulant composition and measurable processing variables. The need for these predictions arises from the need to predict the REDuction/OXidation (REDOX) state of the glass from the Defense Waste Processing Facility (DWPF) melter. This report summarizes the work on these correlations based on the aforementioned data. Previous work on these correlations was documented in a technical report covering data from 2011-2015. This current report supersedes this previous report. Further refinement of the models as additional data are collected is recommended. The glass REDOX depends on the concentrations of nitrate and manganese (oxidants), and of glycolate, formate, oxalate, carbon, and antifoam (reductants) in the melter feed. The waste sludge contains nitrite, nitrate, manganese (Mn), and oxalate. Virtually all of the nitrite is converted to nitrate or NO+NO 2 +N 2 O gases in the CPC. The portion of the nitrite converted to nitrate increases the amount of nitrate in the sludge. The amount of glycolate in the final melter feed depends on the amount of the glycolic acid feed that is destroyed. Similarly, the amounts of formate and oxalate formed during the decomposition of glycolic acid are required. The material balance on carbon was found to not close in most cases. Generally, there was less carbon at the end of testing compared to the inputs. The most uncertain product variable was glycolate, so material balances were performed where the glycolate concentration was adjusted, usually upward, to close the balance. Correlation versus the original, as-measured, data was generally poor, but correlation against the material balance adjusted values was greatly improved. It was also shown that the correlation of the measured REDOX versus the predicted REDOX was much better when the material balance adjusted glycolate values were used. Three data series were primarily used during the regressions of the data; these series were 1) Sludge Batch 9 NG flowsheet simulant runs NG51-62 (SB9-NG); 2) Scaled Runs + Bounding Hydrogen Runs (SR+BH); and 3) Runs GN43-50 and 57 (43-50,57). The glycolate destruction was found to correlate with acid stoichiometry (AS), percent reducing acid (PRA), and for some data series, headspace to simulant volume ratio (HSV), mercury (Hg), and nitrate. Although glycolate destruction for pairs of data series (e.g., [SB9-NG] and [SR+BH]) were found to depend on HSV, the combination of all three data series was not found to have significant dependence on this variable. The best model for glycolate destruction depended on AS, nitrate, and Hg. This model predicted the product glycolate compositions of the data to within 92-106%. The conversion of glycolate to formate was high when noble metals and Hg were not present, with values up to 100%. When noble metals and Hg were present, this conversion ranged from zero to 7%, and was dependent on AS. Lower AS gave higher conversions to formate. The conversion to oxalate was found to depend on the AS and the initial concentration of nitrite. An alternative fit versus AS and the form of ruthenium (Ru) used is a possible alternative. This fit was somewhat less statistically significant. This second model predicts that more oxalate is formed when Ru-nitrosyl nitrate is used rather than Ru chloride. The conversion of glycolate to oxalate ranged from zero to 6%. The conversion of nitrite to nitrate depended primarily on AS and PRA, with HSV and Hg being significant when these variables were varied. For multiple series of data, nitrite was also needed to SRNL-STI-2017-00172 5HYLVLRQ viL distinguish between data series, and the effect of HSV became insignificant. The best model for nitrite to nitrate conversion depended on AS, PRA, nitrite, and Hg. The 95% confidence intervals on the predicted values of glycolate destruction, glycolate to oxalate conversion, and nitrite to nitrate conversion were used to determine the uncertainty in the predicted REDOX when starting with only the composition of the sludge, AS, and PRA. Using the 95% confidences on an individual value (that is the confidence in getting a particular value for one single test as opposed to what the mean would be for multiple tests), the uncertainty in the predicted REDOX was calculated. The uncertainty in the actual product composition glycolate, oxalate, formate, and nitrate concentrations translated to an uncertainty in the REDOX value of ±0.1,which is approximately the uncertainty claimed in the REDOX model itself.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Physiological Factors Contributing to Postflight Changes in Functional Performance

Astronauts experience alterations in multiple physiological systems due to exposure to the microgravity conditions of space flight. These physiological changes include sensorimotor disturbances, cardiovascular deconditioning and loss of muscle mass and strength. These changes might affect the ability of crewmembers to perform critical mission tasks immediately after landing on lunar and Martian surfaces. To date, changes in functional performance have not been systematically studied or correlated with physiological changes. To understand how changes in physiological function impact functional performance an interdisciplinary pre/postflight testing regimen (Functional Task Test, FTT) has been developed that systematically evaluates both astronaut postflight functional performance and related physiological changes. The overall objectives of the FTT are to: Develop a set of functional tasks that represent critical mission tasks for Constellation. Determine the ability to perform these tasks after flight. Identify the key physiological factors that contribute to functional decrements. Use this information to develop targeted countermeasures. The functional test battery was designed to address high priority tasks identified by the Constellation program as critical for mission success. The set of functional tests making up the FTT include the: 1) Seat Egress and Walk Test, 2) Ladder Climb Test, 3) Recovery from Fall/Stand Test, 4) Rock Translation Test, 5) Jump Down Test, 6) Torque Generation Test, and 7) Construction Activity Board Test. Corresponding physiological measures include assessments of postural and gait control, dynamic visual acuity, fine motor control, plasma volume, orthostatic intolerance, upper and lower body muscle strength, power, fatigue, control and neuromuscular drive. Crewmembers will perform both functional and physiological tests before and after short (Shuttle) and long-duration (ISS) space flight. Data will be collected on R+0 (Shuttle only), R+1, R+6 and R+30. Using a multivariate regression model we will identify which physiological systems contribute the most to impaired performance on each functional test. This will allow us to identify the physiological systems that play the largest role in decrement in functional performance. Using this information we can then design and implement countermeasures that specifically target the physiological systems most responsible for the altered functional performance associated with space flight.

Bloomberg, J. J.↗

Roles of Engineering Correlations in Hypersonic Entry Boundary Layer Transition Prediction

Efforts to design and operate hypersonic entry vehicles are constrained by many considerations that involve all aspects of an entry vehicle system. One of the more significant physical phenomenon that affect entry trajectory and thermal protection system design is the occurrence of boundary layer transition from a laminar to turbulent state. During the Space Shuttle Return To Flight activity following the loss of Columbia and her crew of seven, NASA's entry aerothermodynamics community implemented an engineering correlation based framework for the prediction of boundary layer transition on the Orbiter. The methodology for this implementation relies upon the framework of correlation techniques that have been in use for several decades. What makes the Orbiter boundary layer transition correlation implementation unique is that a statistically significant data set was acquired in multiple ground test facilities, flight data exists to assist in establishing a better correlation and the framework was founded upon state of the art chemical nonequilibrium Navier Stokes flow field simulations. The basic tenets that guided the formulation and implementation of the Orbiter Return To Flight boundary layer transition prediction capability will be reviewed as a recommended format for future empirical correlation efforts. The validity of this approach has since been demonstrated by very favorable comparison of recent entry flight testing performed with the Orbiter Discovery, which will be graphically summarized. These flight data can provide a means to validate discrete protuberance engineering correlation approaches as well as high fidelity prediction methods to higher confidence. The results of these Orbiter engineering and flight test activities only serve to reinforce the essential role that engineering correlations currently exercise in the design and operation of entry vehicles. The framework of information-related to the Orbiter empirical boundary layer transition prediction capability will be utilized to establish a fresh perspective on this role, to illustrate how quantitative statistical evaluations of empirical correlations can and should be used to assess accuracy and to discuss what the authors' perceive as a recent heightened interest in the application of high fidelity numerical modeling of boundary layer transition. Concrete results will also be developed related to empirical boundary layer transition onset correlations. This will include assessment of the discrete protuberance boundary layer transition onset data assembled for the Orbiter configuration during post-Columbia Return To Flight. Assessment of these data will conclude that momentum thickness Reynolds number based correlations have superior coefficients and uncertainty in comparison to roughness height based Reynolds numbers, aka Re(sub k) or Re(sub kk). In addition, linear regression results from roughness height Reynolds number based correlations will be evaluated, leading to a hypothesis that non-continuum effects play a role in the processes associated with incipient boundary layer transition on discrete protuberances.

Campbell, Charles H.↗