Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Data engineering”

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 613 records · Page 34

Data-Driven Operator Theoretic Methods for Phase Space Learning and Analysis

This paper uses data-driven operator theoretic approaches to explore the global phase space of a dynamical system. In this work, we defined conditions for discovering new invariant subspaces in the state space of a dynamical system starting from an invariant subspace based on the spectral properties of the Koopman operator. When the system evolution is known locally in several invariant subspaces in the state space of a dynamical system, a phase space stitching result is derived that yields the global Koopman operator. Additionally, in the case of equivariant systems, a phase space stitching result is developed to identify the global Koopman operator using the symmetry properties between the invariant subspaces of the dynamical system and time-series data from any one of the invariant subspaces. Finally, these results are extended to topologically conjugate dynamical systems; in particular, the relation between the Koopman tuple of topologically conjugate systems is established. The proposed results are demonstrated on several second-order nonlinear dynamical systems including a bistable toggle switch. Our method elucidates a strategy for designing discovery experiments: experiment execution can be done in many steps, and models from different invariant subspaces can be combined to approximate the global Koopman operator.

42 ENGINEERING↗

Implementation of a realistic artificial data generator for crash data generation

In this paper, a framework is outlined to generate realistic artificial data (RAD) as a tool for comparing different models developed for safety analysis. The primary focus of transportation safety analysis is on identifying and quantifying the influence of factors contributing to traffic crash occurrence and its consequences. The current framework of comparing model structures using only observed data has limitations. With observed data, it is not possible to know how well the models mimic the true relationship between the dependent and independent variables. Further, real datasets do not allow researchers to evaluate the model performance for different levels of complexity of the dataset. RAD offers an innovative framework to address these limitations. Hence, we propose a RAD generation framework embedded with heterogeneous causal structures that generates crash data by considering crash occurrence as a trip level event impacted by trip level factors, demographics, roadway and vehicle attributes. Within our RAD generator we employ three specific modules: (a) disaggregate trip information generation, (b) crash data generation and (c) crash data aggregation. For disaggregate trip information generation, we employ a daily activity-travel realization for an urban region generated from an established activity-based model for the Chicago region. We use this data of more than 2 million daily trips to generate a subset of trips with crash data. For trips with crashes crash location, crash type, driver/vehicle characteristics, and crash severity. The daily RAD generation process is repeated for generating crash records at yearly or multi-year resolution. In conclusion, the crash databases generated can be employed to compare frequency models, severity models, crash type and various other dimensions by facility type - possibly establishing a universal benchmarking system for alternative model frameworks in safety literature.

42 ENGINEERING↗

Physics-informed State-space Neural Networks for transport phenomena

This work introduces Physics -informed State -space neural network Models (PSMs), a novel solution to achieving real-time optimization, flexibility, and fault tolerance in autonomous systems, particularly in transportdominated systems such as chemical, biomedical, and power plants. Traditional data -driven methods fall short due to a lack of physical constraints like mass conservation; PSMs address this issue by training deep neural networks with sensor data and physics -informing using components' Partial Differential Equations (PDEs), resulting in a physics -constrained, end -to -end differentiable forward dynamics model. Further, through two in silico experiments - a heated channel and a cooling system loop - we demonstrate that PSMs offer a more accurate approach than a purely data -driven model. In the former experiment, PSMs demonstrated significantly lower average root -mean -square errors across test datasets compared to a purely data -driven neural network, with reductions of 44 %, 48 %, and 94 % in predicting pressure, velocity, and temperature, respectively. Beyond accuracy, PSMs demonstrate a compelling multitask capability, making them highly versatile. In this work, we showcase two: supervisory control of a nonlinear system through a sequentially updated state -space representation and the proposal of a diagnostic algorithm using residuals from each of the PDEs. The former demonstrates PSMs' ability to handle constant and time -dependent constraints, while the latter illustrates their value in system diagnostics and fault detection.

42 ENGINEERING↗

Operator inference for non-intrusive model reduction with quadratic manifolds

This paper proposes a novel approach for learning a data-driven quadratic manifold from high-dimensional data, then employing this quadratic manifold to derive efficient physics-based reduced-order models. The key ingredient of the approach is a polynomial mapping between high-dimensional states and a low-dimensional embedding. This mapping consists of two parts: a representation in a linear subspace (computed in this work using the proper orthogonal decomposition) and a quadratic component. The approach can be viewed as a form of data-driven closure modeling, since the quadratic component introduces directions into the approximation that lie in the orthogonal complement of the linear subspace, but without introducing any additional degrees of freedom to the low-dimensional representation. Combining the quadratic manifold approximation with the operator inference method for projection-based model reduction leads to a scalable non-intrusive approach for learning reduced-order models of dynamical systems. Further, applying the new approach to transport-dominated systems of partial differential equations illustrates the gains in efficiency that can be achieved over approximation in a linear subspace.

42 ENGINEERING↗

Data-driven learning of nonlocal physics from high-fidelity synthetic data

A key challenge to nonlocal models is the analytical complexity of deriving them from first principles, and frequently their use is justified a posteriori. Here, we extract nonlocal models from data, circumventing these challenges and providing data-driven justification for the resulting model form. Extracting data-driven surrogates is a major challenge for machine learning (ML) approaches, due to nonlinearities and lack of convexity — it is particularly challenging to extract surrogates which are provably well-posed and numerically stable. Our scheme not only yields a convex optimization problem, but also allows extraction of nonlocal models whose kernels may be partially negative while maintaining well-posedness even in small-data regimes. To achieve this, based on established nonlocal theory, we embed in our algorithm sufficient conditions on the non-positive part of the kernel that guarantee well-posedness of the learnt operator. These conditions are imposed as inequality constraints to meet the requisite conditions of the nonlocal theory. We demonstrate this workflow for a range of applications, including reproduction of manufactured nonlocal kernels; numerical homogenization of Darcy flow associated with a heterogeneous periodic microstructure; nonlocal approximation to high-order local transport phenomena; and approximation of globally supported fractional diffusion operators by truncated kernels.

42 ENGINEERING↗

Quantifying the Effects of Idle-Stop Systems on Fuel Economy in Light-Duty Passenger Vehicles

Vehicles equipped with idle-stop (IS) systems are capable of engine shut down when the vehicle is stopped and rapid engine re-start for the vehicle launch. This capability reduces fuel consumption and emissions during periods when the engine is not being utilized to provide propulsion or to power accessories. IS systems are a low-cost and fast-growing technology in the industry-wide pursuit of increased vehicle efficiency, possibly becoming standard features in European vehicles in the near future. In contrast, currently there are only three non-hybrid vehicle models for sale in North America with IS systems and these models are distinctly low-volume models. As part of the United States Department of Energy’s Advanced Vehicle Testing Activity, ECOtality North America has tested the real-world effect of IS systems on fuel consumption in three vehicle models imported from Europe. These vehicles were chosen to represent three types of systems: (1) spark ignition with 12-V belt alternator starter; (2) compression ignition with 12-V belt alternator starter; and (3) direct-injection spark ignition, with 12-V belt alternator starter/combustion restart. The vehicles have undergone both dynamometer and on-road testing; the test results show somewhat conflicting data. The laboratory data and the portion of the on-road data in which driving is conducted on a prescribed route with trained drivers produced significant fuel economy improvement. However, the fleet data do not corroborate improvement, even though the data show significant engine-off time. It is possible that the effects of the varying driving styles and routes in the fleet testing overshadowed the fuel economy improvements. More testing with the same driver over routes that are similar with the IS system-enabled and disabled is recommended. There is anecdotal evidence that current Environmental Protection Agency fuel economy test procedures do not capture the fuel economy gains that IS systems produce in real-world driving. The program test results provide information on the veracity of these claims.

25 ENERGY STORAGE↗

Investigation of the effects of passive pre-chamber nozzle pattern and ignition system on engine performance and emissions

The impact of passive pre-chamber (PC) internal volume, nozzle hole pattern (i.e. with and without a central axial nozzle), and PC igniter plug type on performance and emissions was investigated in an optically accessible, single-cylinder, gasoline research engine. The two PC igniter plugs investigated were a conventional inductive coil spark plug and a nanosecond repetitively pulsed (NRP) plasma discharge system previously demonstrated to accelerate early flame propagation. The baseline PC design featured a funnel shaped internal volume with a PC tip that contained six radial nozzles and a larger central axial nozzle. Two additional PC tip geometries were evaluated where either the baseline internal volume was increased or the axial nozzle was removed and the radial nozzle diameters were increased. A sweep of charge equivalence ratios ( ϕ) from nearly stoichiometric to the lean limit was performed for a fixed engine speed (1300 revolutions per minute), and engine load (3.5 bar gross indicated mean effective pressure). Time-resolved PC and main chamber (MC) pressure data as well as MC emissions data were collected to analyze engine performance and emissions characteristics. Combustion in the MC was further investigated using high-speed excited methylidyne radical (CH*) chemiluminescence imaging. Collected results highlighted that while all PC tips and ignition systems exhibited similar performance and emissions down to ϕ = 0.8, relevant differences in thermal efficiency and emissions for leaner charge mixtures were observed, with the results highly dependent on the nozzle pattern and ignition system. Major deviations were correlated to preferential de-pressurization of the PC through the axial nozzle for lean conditions that was not observed for mixture conditions closer to stoichiometric. Results show that a combination of radial and axial nozzle patterns in the PC extended lean-stability limits at the low-load condition evaluated. Further benefits were observed with the use of NRP ignition systems due to faster combustion within the PC volume provided that the volume was sufficiently large.

Di Sabatino, Francesco↗

Processing ultrasonic inspection data from multiple scan patterns for turbine rotor weld build-up evaluations

The study presents a data processing methodology for weld build-up using multiple scan patterns. To achieve an overall high probability of detection for flaws with different orientations, an inspection procedure with three different scan patterns is proposed. The three scan patterns are radial-tangential longitude wave pattern, axial-radial longitude wave pattern, and tangential shear wave pattern. Scientific fusion of the inspection data is implemented using volume reconstruction techniques. The idea is to perform spatial domain forward data mapping for all sampling points. A conservative scheme is employed to handle the case that multiple sampling points are mapped to one grid location. The scheme assigns the maximum value for the grid location to retain the largest equivalent reflector size for the location. The methodology is demonstrated and validated using a realistic ring of weld build-up. Tungsten balls and bars are embedded to the weld build-up during manufacturing process to represent natural flaws. Flat bottomed holes and side drilled holes are installed as artificial flaws. Automatic flaw identification and extraction are demonstrated. Results indicate the inspection procedure with multiple scan patterns can identify all the artificial and natural flaws.

42 ENGINEERING↗

Tetris-inspired detector with neural network for radiation mapping

Abstract Radiation mapping has attracted widespread research attention and increased public concerns on environmental monitoring. Regarding materials and their configurations, radiation detectors have been developed to identify the position and strength of the radioactive sources. However, due to the complex mechanisms of radiation-matter interaction and data limitation, high-performance and low-cost radiation mapping is still challenging. Here, we present a radiation mapping framework using Tetris-inspired detector pixels. Applying inter-pixel padding for enhancing contrast between pixels and neural networks trained with Monte Carlo (MC) simulation data, a detector with as few as four pixels can achieve high-resolution directional prediction. A moving detector with Maximum a Posteriori (MAP) further achieved radiation position localization. Field testing with a simple detector has verified the capability of the MAP method for source localization. Our framework offers an avenue for high-quality radiation mapping with simple detector configurations and is anticipated to be deployed for real-world radiation detection.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Bridging the length scales on mechanical property evaluation (Final Report)

The development of small-scale mechanical testing in combination with microstructural investigation is of great interest to the nuclear materials community for both materials development and monitoring applications. Dramatically reducing the sample sizes to reduce radioactivity and obtaining mechanical properties of irradiated samples is truly intriguing. Moreover, such studies promise a range of benefits including cost reduction, fundamental insight in structure-property relationships, increased statistics on less sample material, and reinvestigation of prior irradiated and tested reactor samples while simultaneously enabling the generation of mechanical test data on ion beam irradiated materials with limited penetration depths. Small scale materials testing on sub-sized samples has been studied for several decades, though it has only been after the development of micro-testing based on Focused Ion Beam (FIB) sample manufacturing in that orders of magnitude smaller samples could really be investigated in a quantitative manner. In recent years, small scale mechanical testing techniques at a number of length scales has been developed for both unirradiated and irradiated (ion and neutron) materials. Technological advances made in this field have enabled ex-situ and in-situ transmission electron microscopy (TEM) and scanning electron microscopy (SEM) examination, thus leading to more accurate measurements as well as additional mechanistic information. A recent review of the benefits of these techniques show that these techniques are at a stage to tackle multi-scale ranges of materials investigations and can be utilized to obtain fundamental science-based understanding of nuclear materials. Considering the tremendous advances made, one can see how small-scale mechanical testing techniques combined with modeling can enable true small scale to bulk scale mechanical property correlations. However, for the engineering community to adapt this approach fully, one needs to demonstrate that a) that these techniques can produce results with high fidelity and reproducibility, b) generate engineering stress-strain data that one can utilize to understand bulk behavior, and c) generate new insight into relevant phenomena fostering the true understanding of radiation damage and microstructure in materials for nuclear applications. It is the objective of this proposal to bridge the length scale between macro- and micro- scale mechanical testing of unirradiated and irradiated materials. This involves the development and demonstration of procedures for multi-scale mechanical testing that enable high fidelity reproducibility of data and the generation bulk property data from small-scale mechanical tests. Through this, the proposal aims to enhance the confidence in the obtained data at the smaller length scales and enhance the insight provided from these techniques for bulk scale applications on both unirradiated and irradiated materials.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Expanded analysis of machine learning models for nuclear transient identification using TPOT

Industries around the world are becoming more and more data driven. The nuclear field is no exception with several different applications being proposed. One popular area of research is the use of machine learning in transient detection. This paper seeks to build upon a previous study which made use of the AutoML package TPOT to train traditional machine learning models to classify transient events occurring with a reactor. Synthetic data was once again collected using a GPWR reactor simulator. Data on 12 different events was collected using 15 different initial conditions. Here, a dataset consisting of over 100,000 data points was compiled and used to train 7 different machine learning models using a pre-defined TPOT dictionary with 12 different preprocessing techniques. Three of the trained models were able to produce validation results in the 90s with the expanded dataset. Once the models were trained, it was possible to look into where during the simulation, misclassifications occurred. Using these three models, analysis was done to determine if TPOT could be used to train models that were effective if important features were missing. The results from this were positive with the newly trained models scoring close to the original models. Finally, to conclude this study, the three high performing models were retrained using different random states to see if there was any major variation when different states were used.

42 ENGINEERING↗

Hydropower Resilience Database for Assessing Microgrid Formation Capability and Enhancing Power Grid Resilience

In the face of increasing frequency of extreme events, enhancing resilience and reliability of energy infrastructure demands innovative solutions. Hydropower, with its inherent generation flexibility and grid-forming capability, offers significant potential for enhancing power grid resilience through the establishment of microgrids. To enable informed decision-making and strategic planning, we present the Hydropower Resilience Database (HRD) that integrates data from various sources such as Oakridge National Laboratory (ORNL) HydroSource, National Inventory of Dams, and US Western Grid database. By integrating relevant information on hydropower plant characteristics, including dams, reservoirs, and connected electrical grid, this resource enables hydropower plant owners, utilities, and community stakeholders to identify and evaluate the feasibility of using hydropower resources in microgrids to support nearby communities and critical infrastructure in various challenging scenarios. Using HRD, a set of metrics are evaluated to quantify the capability of hydropower plants to support essential microgrid functions. An interactive tool is developed using ArcGIS, enabling the visualization and analysis using HRD. Our research contributes to strategic planning efforts for fortifying energy infrastructure, ensuring reliable power supply during disruptions, and advancing the development of robust and resilient energy systems in regions susceptible to grid vulnerabilities.

13 HYDRO ENERGY↗

Hydropower Resilience Database for Assessing Microgrid Formation Capability and Enhancing Power Grid Resilience

In the face of increasing frequency of extreme events, enhancing resilience and reliability of energy infrastructure demands innovative solutions. Hydropower, with its inherent generation flexibility and grid-forming capability, offers significant potential for enhancing power grid resilience through the establishment of microgrids. To enable informed decision-making and strategic planning, we present the Hydropower Resilience Database (HRD) that integrates data from various sources such as Oakridge National Laboratory (ORNL) HydroSource, National Inventory of Dams, and US Western Grid database. By integrating relevant information on hydropower plant characteristics, including dams, reservoirs, and connected electrical grid, this resource enables hydropower plant owners, utilities, and community stakeholders to identify and evaluate the feasibility of using hydropower resources in microgrids to support nearby communities and critical infrastructure in various challenging scenarios. Using HRD, a set of metrics are evaluated to quantify the capability of hydropower plants to support essential microgrid functions. An interactive tool is developed using ArcGIS, enabling the visualization and analysis using HRD. Our research contributes to strategic planning efforts for fortifying energy infrastructure, ensuring reliable power supply during disruptions, and advancing the development of robust and resilient energy systems in regions susceptible to grid vulnerabilities.

13 HYDRO ENERGY↗

Pavement condition and climatic data in southeast Texas: A dataset for evaluating flood impacts on pavement performance

Effective pavement maintenance is essential for economic stability, optimal network performance, and roadway safety. Achieving this requires thorough evaluation of pavement conditions, including structural integrity, surface roughness, and distress characteristics. Pavement performance indicators play a critical role in influencing vehicle safety and ride quality. Recent advances have emphasized the use of data-driven modeling to anticipate pavement behavior, with the goal of optimizing resource allocation and refining Maintenance and Rehabilitation (M&R) strategies through accurate condition assessment. A foundational requirement for these modeling efforts is the availability of standardized, high-quality datasets that can support robust and reproducible infrastructure analysis. This data article presents a comprehensive dataset assembled to facilitate pavement performance prediction, with a geographic focus on Southeast Texas, particularly the flood-vulnerable area of Beaumont. The dataset encompasses pavement and traffic attributes, meteorological records, flood simulation outputs, ground deformation measurements, and topographic indices, enabling detailed examination of both load-associated and non-load-associated degradation mechanisms. Data preprocessing was performed using ArcGIS Pro, Microsoft Excel, and Python to ensure consistency and usability in data-driven modeling applications, including machine learning workflows. Key contributions of this dataset include its utility in analyzing the climatic and environmental factors affecting pavement conditions, identifying critical predictive features, and enabling in-depth correlation analysis across diverse variables. By filling existing gaps in input variable selection resources, this dataset supports the development of predictive tools for estimating future maintenance demand and enhancing the resilience of pavement networks in flood-impacted areas. The resource highlights the importance of standardized datasets for advancing pavement management practices and provides a robust foundation for ongoing infrastructure performance modeling.

42 ENGINEERING↗

Final Technical Report: Moving ASR Cloud Microphysical Retrievals Beyond the Vertical Column

Clouds of all types play important roles in atmospheric heating, surface heating and emission of radiation, hence energy, to space. Clouds not only affect atmospheric heating via their impacts on solar and terrestrial radiation, they also release kinetic energy to the atmosphere when they form and they extract kinetic energy from the atmosphere when they evaporate. To understand the impacts of clouds on weather and climate, one must be able to make accurate observations of the amount of water associated with them and the impact of this water on atmospheric and surface heating rates. One important cloud type is liquid water clouds. Liquid clouds are often found in the boundary layers of Earth at altitudes from just above the surface to approximately 2000 m above it, at a variety of temperatures, including temperatures below the freezing point of liquid water (0 °C). As a result, they are important constituents over the world’s oceans as well as over cold, ice-covered surfaces such as in the Arctic. Observationally estimating the total amount of liquid contained in a column through a liquid water cloud can now be accomplished to acceptable accuracies (except for the thinnest, lowest liquid containing clouds) using instruments called passive microwave radiometers. However, observationally determining the spatial distribution of liquid within these columns has proven to be difficult. The focus of this project was on the application of two radars operating at two different frequencies to retrieve the distribution of liquid water within columns passing through liquid water clouds. Motivation in this project for using two different radars operating at different frequencies is that as overlapping (in space) beams from the two radars pass through the same column of liquid water cloud absorption of power from one beam (the 94-GHz beam) is greater than absorption of power from the other beam (the 35-GHz beam). Therefore, the difference in returned powers between the two radar beams with distance from the radars is proportional to the amount of liquid water that the two beams have passed through. Therefore, one can use the differences in power between the two beams with distance through a liquid water cloud to determine the distribution of liquid water throughout the cloud. The two radars used in this study are the Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) Program Ka-band Scanning ARM Cloud Radar (Ka-SACR) and the W-band Scanning ARM Cloud Radar (W-SACR). When this projected started in 2013, the Ka- and W-band SACRs were new technology, state-of-the-art sensors. To put together a sufficiently diverse and big set of test cases for the study, data from the SACRs deployed to NSA, SGP, TCAP and MAGIC were used, leading to a test set composed of 65 hours of suitable data spread over approximately 17 case study days. Using the Ka- and W-band SACR data from the case study periods, vertical distributions of liquid water content through liquid water clouds were retrieved. Because there are no other data sets with which to compare the retrievals, distributions of liquid water through a column were summed and compared to the total amounts of liquid water obtained by retrievals applied to passive microwave radiometer observations. The total amounts of liquid water (in units of millimeters, or the thickness in millimeters of the total liquid water in a vertical column if allowed to pool at the bottom of the column) between the two retrievals agreed to within 0.060 mm with a standard deviation of 0.31 mm, or to within 0.016 mm with a standard deviation of 0.12 mm once outliers were removed. The standard deviation of 0.12 mm was fairly independent of the total amount of liquid water in the clouds, indicating that retrieval errors were larger for clouds with smaller amounts of liquid in them. Errors in the retrieved total amounts of liquid were also found to be sensitive to the a priori information that was fed into the retrieval approach. Overall, standard deviations of 0.12 mm are sufficient for some cloud studies but for radiation studies standard deviations on the order of 0.02 mm need to be achieved for clouds with the lowest amounts of liquid water. Many modifications to the retrieval approach were tested subsequently in an attempt to reduce the standard deviations relative to the radiometer retrievals below 0.12 mm but none of them worked reliably. We decided to wait until the DOE ARM Program deployed its second generation SACR2s to its NSA and ENA sites for further retrieval evaluation. The SACR2s were engineered to better synchronize data between the Ka- and W-band radars, reducing potential sources of noise for radar retrieval of liquid water distributions within clouds. The SACR2s have improved in reliability recently and the system deployed to ENA is best positioned to provide the data to advance research on this path. This will be an ongoing research activity beyond the project because of the promise in SACR2 data from the DOE ARM ENA Facility.

58 GEOSCIENCES↗

Frequency Injection Based HVDC Attack-Defense Control Via Squeeze-Excitation Double CNN

Due to the independent controllability and fast power regulation capability, the High Voltage Direct Current (HVDC) system could be a prospective technology to provide multiple ancillary services to the system besides conventional bulk power transmission. However, with the increase of False Data Injection Attacks (FDIAs) on PMU data, the HVDC system could have the wrong response once the collected data that the HVDC system relied on is attacked, thus threatening the system operating security. How to ensure the security of the PMU-based HVDC ancillary service control become an urgent issue. To mitigate the risk, this paper proposed an HVDC attack-defense control based on the FDIAs detection method. Firstly, the Squeeze-Excitation based Double Convolutional Neural Networks (SE-DCNN) is proposed to realize fast identification of the attacking frequency type based on the time and frequency domain signals. Furthermore, the duration time of FDIAs is detected by the local outlier factor. Then, utilizing the results from SE-DCNN, HVDC ancillary service control framework is reorganized and an HVDC attack defense control is proposed for suppressing the potential influence of various types of FDIAs on the HVDC system ancillary service. Different experiments results demonstrate that the proposed method has the ability to significantly mitigate the frequency deviation and oscillation under the FDIA.

42 ENGINEERING↗