SEARCH · Engineering Papers
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.
This presentation is a deep dive into cyber-physical challenges and opportunities. How NREL's cybersecurity tools enable better quantification, management and governance for securing inverter-based resources.
This report presents a series of implicit–explicit (IMEX) variable stepsize algorithms for the incompressible Navier–Stokes equations (NSE). Here, IMEX means the nonlinear term is treated fully explicitly while the remaining terms are treated implicitly. With the advent of new computer architectures there has been growing demand for low memory solvers of this type. The addition of time adaptivity improves the accuracy and greatly enhances the efficiency of the algorithm. We prove energy stability of an embedded first–second order IMEX pair. For the first order member of the pair, we prove stability for variable stepsizes, and analyze convergence. We believe this to be the first proof of this type for a variable stepsize IMEX scheme for the incompressible NSE. We then define and test a variable stepsize, variable order IMEX scheme using these methods. Our work contributes several firsts for IMEX NSE schemes, including an energy argument and error analysis of a two-step, variable stepsize method, and embedded error estimation for an IMEX multi-step method.
This paper resolves a perennial point of confusion regarding the source-weighting normalization factor recommended in the MCNP manual (𝜋𝑟 2 ) to stochastically estimate the volume of a region within an enclosing inward-directed spherical surface source with radius 𝑟. The normalization factor arises from the relationship between a sphere’ s mean chord length, its volume, and the values estimated by MCNP track-length tallies. A brief derivation is given that relates these quantities and results in the stated normalization. The correctness of this factor is demonstrated by estimating the volume of a variety of convex and nonconvex volumes. A heuristic demonstration of how biasing the inward-directed source reduces the statistical uncertainty of the stochastic volume estimate is also given, but a rigorous analysis of this improvement is left as future work.
This study shows that in recent years, active subspace methods (ASMs) have become a popular means of performing subspace sensitivity analysis on black-box functions. Naively applied, however, ASMs require gradient evaluations of the target function. In the event of noisy, expensive, or stochastic simulators, evaluating gradients via finite differencing may be infeasible. In such cases, often a surrogate model is employed, on which finite differencing is performed. When the surrogate model is a Gaussian process (GP), we show that the ASM estimator is available in closed form, rendering the finite-difference approximation unnecessary. We use our closed-form solution to develop acquisition functions focused on sequential learning tailored to sensitivity analysis on top of ASMs. We also show that the traditional ASM estimator may be viewed as a method of moments estimator for a certain class of GPs. We demonstrate how uncertainty on GP hyperparameters may be propagated to uncertainty on the sensitivity analysis, allowing model-based confidence intervals on the active subspace. Our methodological developments are illustrated on several examples.
As we move into the exascale era, the power and energy footprints of high-performance computing (HPC) systems have grown significantly larger. Due to the harsh power and thermal conditions the system, components are exposed to extreme operating conditions. Operation of such modern HPC systems requires deep insights into long term system behavior to maintain its efficiency as well as its longevity. To help the HPC community to gain such insights, we provide a dataset that records the long-term power and thermal behavior of the 200PF pre-exascale supercomputer at the Oak Ridge Leadership Computing Facility (OLCF), Summit. This system is an IBM AC922 based system that has 9,252 IBM Power9 CPUs and 27,756 Nvidia V100 GPUs and can consume up to 13MW power at peak. Heat removal is performed using medium temperature direct liquid cooling and rear-door heat exchanger based secondary cooling loop. Originally extracted from a high-resolution (1Hz) per-component (GPUs, CPUs) measurements from the system, we primarily provide a dataset that has 10-second and 1-minute mean power and thermal measurements selected from five month-long segments over the course of 2020 (January and August), 2021 (February and August), and 2022 (January). For convenience, we also provide various sub datasets randomly sampled from the time and space (hosts) of the cluster. Further details and example code for analysis can be found in the following GitHub repository: https://github.com/at-aaims/summit_power_and_thermal_data
This project leverages advances in machine learning based data analysis techniques and untargeted omic analytical methods to progress nuclear nonproliferation technologies beyond current capabilities. The developed approaches can be used to identify and detect complex chemical fingerprints of facilities of interest. These techniques have been developed for fields such as metabolomics and genomics but have not been applied to nuclear nonproliferation applications. Adaptation of these techniques for volatile organic compound analysis has far reaching application within the scientific community including environmental chemistry, atmospheric physics, and climate sciences.
Focal Area(s): We will use unsupervised machine learning methods to identify and quantify the influence of large scale natural modes of climate variability to gain insight into the observed and simulated seasonal average and extreme precipitation changes. Science Challenge: A recent paper, led by co-PI Mark Risser, finds that although much of the variability in seasonal average and extreme precipitation over CONUS is unforced, the effect of large-scale modes of circulation variability (such as ENSO, AMO, PNA, etc.) can be detected and attributed. However, it is unclear whether or not unsupervised learning methods can (a) replicate this finding or (b) yield insight into possible nonlinear behavior that was not captured in the initial statistical analysis. Further work would entail extending this framework to other global land areas.
This report summarizes results from a 2 year effort to improve the current automated seismic event processing system by leveraging machine learning models that can operated over the inherent graph data structure of a seismic sensor network. Specifically, the GraphAlign project seeks to utilize prior information on which stations are more likely to detect signals originating from particular geographic regions to inform event filtering. To date, the GraphAlign team has developed a Graphical Neural Network (GNN) model to filter out false events generated by the Global Associator (GA) algorithm. The algorithm operates directly on waveform data that has been associated to an event by building a variable sized graph of station waveforms nodes with edge relations to an event location node. This builds off of previous work where random forest models were used to do the same task using hand crafted features. The GNN model performance was analyzed using an 8 week IMS/IDC dataset, and it was demonstrated that the GNN outperforms the random forest baseline. We provide additional error analysis of which events the GNN model performs well and poorly against concluded by future directions for improvements.
This project provides a methodology for retrieving data from both existing public data and municipal open data sets to characterize changes in the building stock, socioeconomic, demographic, and climate indicators at a neighborhood scale. This methodology forms the basis for a geospatial dashboard to analyze and share the data. The developed methodology may be used to collect data for further energy equity analyses. Additional data sets, such as from other municipalities, may be incorporated in the future to expand the extents of the possible analysis.
This presentation covers a few key points. First and foremost, that ORIGEN is a consumer of FPY data. Secondly, TRITON and POLARIS are assets which can be used to validate new FPY data, e.g., from existing radiochemical benchmarks and new irradiation experiments. Lastly, this presentation finds that new capabilities with covariance data should be developed, and V and V methods and analysis are also needed for the covariance data.
Support customizations of algorithms and raw data file structures to enhance software workflows for liquid chromatography (LC), mass spectrometry (MS) and ion mobility mass spectrometry (IM-MS)-based metabolite characterization. Evaluate and improve the integration of ion mobility to existing MS analysis methods of the Mass Profiler Professional workflow (Mass Profiler, ID Browser and Mass Profiler Professional).
Infrasound, characterized by low-frequency sound inaudible to humans (<20 Hz), emanates from natural and anthropogenic sources. Its efficacy for monitoring phenomena necessitates robust sensing networks. Traditional ground-based infrasound sensors have limitations due to atmospheric dynamics and noise interference. Balloon-bore sensors have emerged as an alternative, offering reduced noise and improved capabilities. This study bridges clustering algorithms with balloon borne infrasound data, a domain yet to be explored. Employing K-Means, DBSCAN, and GMM algorithms on normalized and reshaped data and only normalized data from a New Zealand-based NASA balloon flight, insights into background noise at stratospheric altitudes were revealed. Despite challenges arising from distinguishing signals amid unique background noise, this research provides vital reference material for noise analysis and calibration. Beyond infrasound event capture, the dataset enriches comprehension of background noise characteristics in the southern hemisphere.
This final report briefly reviews the main technical accomplishments of the INVERT award, summarizes existing or planned publications or transitions from the effort, and lastly reviews T2M strategies resulting from the program. Specifically, under the award, our team studied three main technical areas and performed one preliminary T2M study on the cost-benefit analysis of Inverse Design Methods and one major software release (the Maryland Inverse Design Benchmark Suite).
This tutorial will be a deep dive into considerations for battery modeling and demonstrating how to model them in SAM, including battery chemistry, thermal modeling, degradation/lifetime, dispatch, interconnection limits and curtailment, and their associated impacts on project profits and battery lifetime. By the end of the tutorial attendees will know how to size and model both behind-the-meter and front-of-meter battery systems, including financial analysis and pairing with other PV models (including pvlib) via PySAM.
With the increasing penetration of intermittent renewable energy sources into the grid, there is a growing need for process systems-based strategies that integrate dispatchable and variable energy systems for supplying the demand while maintaining grid reliability. The proposed framework corresponds to a dynamic mixed-integer linear programming optimization approach that integrates coal-fired and natural gas-fired power plants, NaS batteries for energy storage, and solar/wind energy to supply the demand. This optimization approach considers an economic goal and constraints to provide power balance while maintaining the overall damage of the natural gas combined cycle (NGCC) power plant drum under a maximum stress as well as avoiding the overheating of the NGCC superheater and reheater. Renewable curtailment levels are also retained at minimum levels. Case studies are analyzed considering different loads and renewable penetration levels. The results show that the demand was met for all cases. Grid flexibility was mostly provided by the NGCC, while the batteries were used sparingly. In addition, considering a CO 2 equivalent analysis, the environmental performance was intrinsically connected to grid flexibility and the level of renewable penetration. Stress analysis results reinforced the necessity for an equipment health-related constraint.
The present invention relates to methods for analyzing a chemical sample. For instance, the methods herein allow for global analysis of spectroscopy data in order to extract useful chemical properties from complicated multidimensional data. Such analysis can optionally employ data compression to further expedite computer-implemented computation. In particular, the methods herein provide global analysis of data matrices explained by both linear and non-linear terms.
Powerpoint presentation for INSTAR annual meeting. Data and results are hypothetical and do not represent any actual nuclear plant.
Arizona State University, in collaboration with University of South Carolina, worked on a project aimed at development of a new high temperature, high pressure CO 2 perm-selective membrane reactor for water-gas-shift reaction (WGS) with simulated gasifier syngas to produce a high concentration H 2 stream with CO 2 capture. The membrane reactor is made of a CO 2 semi-permeable ceramic-carbonate dual-phase (CCDP) membrane with high CO 2 perm-selectivity/permeance and thermal/mechanical stability for application in WGS reaction. The objectives of this project were to (1) synthesize the chemically/thermally stable tubular CCDP membranes with CO 2 permeance and selectivity (with respect to H 2 , CO or H 2 O) larger than 6.5×10-7 mol/m2·s·Pa and 500, respectively; (2) establish CCDP membrane reactor setup and study high pressure CO 2 permeation and WGS reaction with CO 2 capture using the setup; and (3) identify conditions for WGS in the CCDP membrane reactor that produce CO 2 and H 2 streams with purity of >99% and >90% respectively at CO conversion >95% and overall carbon capture >90%. The work in this project included both membrane development and membrane reactor process study. The membrane development efforts were focused on investigating a H 2 S resistant and highly oxygen-ionic conducting metal oxide material and membrane for CO 2 separation, fabrication of tubular samaria-doped-ceria/molten-carbonate CCDP membrane with high mechanical strength, and experimental and modeling study of high-pressure CO 2 permeation of the CCDP membranes. Mathematical models were developed to describe WGS in the CCDP membrane reactor without a catalyst or packed with a commercial high temperature WGS catalyst. Experiments on WGS in the CCDP membrane reactor with the commercial WGS catalyst, guided by the model analysis, were performed to identify optimum conditions for achieving the CO conversion, carbon capture, and the purity of the H 2 and CO 2 streams mentioned above. At 30 atm feed pressure, 750°C operation temperature, space velocity of 250 h-1, and with steam sweep, a single-stage CCDP membrane reactor with average CO 2 permeation flux of 0.5 cm3(STP)/min.cm2 can achieve CO 2 conversion of 95% and overall carbon capture of 94%, and produce CO 2 and H 2 streams with dry-based purity of >99% and 92% respectively. The project also included process design and techno-economic analysis (TEA) for a CCDP membrane reactor process for WGS reaction with CO 2 capture for a 550 MW coal-fired IGCC power plant, and its comparison with the conventional fixed-bed reactor system for WGS with follow-up CO 2 capture by an amine absorption process. The target performance for the reactor for WGS with CO 2 capture includes CO conversion >95%, hydrogen stream purity >90%, CO 2 stream purity >95%, and total carbon capture >90%. The CCDP membrane developed in this project can achieve the performance target, without subsequent CO 2 capture process at the optimum conditions identified in this project. The outcome of the process design and TEA analysis shows that the membrane reactor for WGS with in-situ CO 2 capture has an operating cost about 40% lower than that for the conventional fixed-bed reactor with a separate amine absorption process for CO 2 capture. However, the capital cost of the membrane reactor process is about twice that of the conventional process because of the higher cost of the CCDP membrane. Modeling analysis shows that a membrane reactor using a CCDP membrane with higher CO 2 permeance (about three times the current value) can deliver the targeted performance for WGS reaction with CO 2 capture at a much higher space velocity and lower membrane surface area to catalyst volume ratio, leading to a smaller catalyst amount and/or membrane area and hence significantly reduced membrane reactor capital costs.