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 summary report provides an overview of all LDRD projects at Argonne National Laboratory that concluded in Fiscal Year 2025. Many projects are funded for multiple years, the fiscal year for each project is indicated by the first four digits of the LDRD project number.
We present the results from the first axionlike particle search conducted using a dish antenna. The experiment was conducted at room temperature and sensitive to axionlike particles in the 44 – 52 μ eV range (10.7–12.5 GHz). The novel dish antenna geometry was proposed by the BREAD Collaboration and previously used to conduct a dark photon search in the same mass range. To allow for axionlike particle sensitivity, the BREAD dish antenna was placed in a 3.9 T solenoid magnet at Argonne National Laboratory. In the presence of a magnetic field, axionlike dark matter converts to photons at the conductive surface of the reflector. The signal is focused onto a custom coaxial horn antenna and read out with a low-noise radio-frequency receiver. No evidence of axionlike dark matter was observed in this mass range and we place the most stringent laboratory constraints on the axion-photon coupling strength, g a γ γ , in this mass range at 90% confidence.
The CHICOX detection system is an array of position sensitive parallel plate avalanche counters used for heavy ion detection in in-beam particle-gamma coincidence experiments. CHICOX, an upgrade of the CHICO2 array, is geometrically compatible with the GRETINA and GRETA gamma-ray arrays and provides an angular resolution sufficient to utilize the excellent position resolution of GRETINA/GRETA. CHICOX was successfully commissioned and fielded as an auxiliary detector to GRETINA in a campaign of stable-beam Coulomb excitation experiments at the ATLAS facility of Argonne National Laboratory. We report here on the design and in-beam performance of the CHICOX array.
As part of the sensor characterization task of the SMART 2.0 project, this dataset includes raw data from three spinning lidars ([Ouster OS2-128](https://ouster.com/products/scanning-lidar/os2-sensor/), [Velodyne Puck (VLP-16)](https://velodynelidar.com/products/puck/), and [Velodyne Ultra Puck (VLP-32)](https://velodynelidar.com/products/ultra-puck/)), one camera ([Mako G-319](https://www.alliedvision.com/en/camera-selector/detail/mako/g-319/)), and one present weather sensor ([Vaisala FD-70](https://www.vaisala.com/en/products/weather-environmental-sensors/forward-scatter-fd70)). All data were synchronized, with the log start time indicated in the file name (HHMMSS). The data can be filtered by date, log time (HHMMSS), sensor, frame ID, and weather classification. These data were gathered statically at the Argonne Testbed for Multiscale Observational Science (ATMOS). Two target stop signs were placed in view of the sensors to contribute a target for comparing sensor data under different conditions. The weather data for each day are stored in netCDF “.nc” files. The lidar data contain the X, Y, Z, intensity, reflectivity, and ring from Ouster OS2-128 rev6, Velodyne VLP-16, and Velodyne VLP-32 lidars. 
Understanding how new vehicle technologies affect fuel economy and energy use is critical to the regulatory work performed by the U.S. Department of Transportation’s National Highway Traffic Safety Administration (NHTSA), which sets Corporate Average Fuel Economy (CAFE) standards under the Energy Policy and Conservation Act of 1975. In order to support this work, Argonne National Laboratory uses Autonomie, a full-vehicle simulation tool, to evaluate advanced powertrain architectures and their effects on vehicle energy consumption and performance. A wide range of vehicle classes has been assessed (i.e., internal combustion engine vehicles, hybrid electric vehicles, plug-in hybrid electric vehicles, battery-electric vehicles, and fuel cell electric vehicles), as well as the effects of various technology improvements such as lightweighting, aerodynamic refinements, and low-rolling-resistance tires. Simulations have been run across multiple drive cycles to capture fuel and electricity use under realistic operating conditions. The resulting datasets include detailed vehicle-level results, model assumptions, and validation reports, all of which have been made publicly available through NHTSA in support of the 2023 notice of proposed rulemaking covering light-duty vehicles for model years 2027 to 2035. These data are critical to stakeholders working in fuel economy regulation, vehicle technology assessment, and energy policy analysis.
The patent-pending ARG-US system uses customizable high-tech sensors and monitoring systems in conjunction with a remote network connection to provide easy and quick access to the status of nuclear and radiological facilities and storage containers. In a partnership with the Argonne National Laboratory and the Idaho National Laboratory, the research team are finding new ways to customize and test the limitations the ARG-US system. This presentation is for the student poster session.
This review provides a personal account of career-long efforts, interwoven with the contributions of others, to exploit the [B 2 ]O 4 framework of an A[B 2 ]O 4 spinel structure for lithium-ion battery applications, where A = Li, and B is one or more metal cations. The narrative begins at the Council for Scientific and Industrial Research (CSIR) in South Africa in the mid-1970s when, in response to the Middle East oil crisis, worldwide efforts were turned to develop high-temperature sodium and lithium batteries to power electric vehicles. In 1981, results in CSIR's program on a high-temperature lithium battery led to a collaborative project on room-temperature lithium cells with Prof. John B. Goodenough at Oxford University, where studies of Fe 3 O 4 , Mn 3 O 4 , and LiMn 2 O 4 cathodes led to the recognition of the broad utility of the [B2]O4 spinel framework as an anode, cathode, or solid electrolyte for rechargeable lithium cells. This finding has had a profound and lasting impact, influencing further ideas and discoveries at the CSIR, Argonne National Laboratory, and elsewhere. The narrative emphasizes the compositional versatility of the spinel structure and the opportunities to tailor the electrochemical potential and stability of a Li-ion cell.
Abstract Machine learning (ML) continues to grow in importance across nearly all domains in modeling to learn from data. Often a tradeoff exists between a model's ability to minimize bias and variance. In this article, we utilize ensemble learning to combine linear, nonlinear, and tree‐/rule‐based ML methods to cope with the bias‐variance tradeoff and result in more accurate models. We use the datasets collected for two parallel cancer deep learning CANDLE benchmarks, NT3 and P1B2, to build performance and power models based on hardware performance counters using single‐object and multiple‐objects ensemble learning to identify the most important counters for improvement on the Cray XC40 Theta at Argonne National Laboratory. Based on the insights from these models, we improve the performance and energy of P1B2 and NT3 by optimizing the deep learning environments TensorFlow, Keras, Horovod, and Python under the huge page size of 8 MB. Experimental results show that ensemble learning not only produces more accurate models but also provides more robust performance counter ranking. We achieve up to 61.15% performance improvement and up to 62.58% energy saving for P1B2 and up to 55.81% performance improvement and up to 52.60% energy saving for NT3 on up to 24,576 cores.
Energy-efficient scientific applications require insight into how high performance computing system features impact the applications' power and performance. This insight can result from the development of performance and power models. Here, in this article, we use the modeling and prediction tool MuMMI (Multiple Metrics Modeling Infrastructure) and 10 machine learning methods to model and predict performance and power consumption and compare their prediction error rates. We use an algorithm-based fault-tolerant linear algebra code and a multilevel checkpointing fault-tolerant heat distribution code to conduct our modeling and prediction study on the Cray XC40 Theta and IBM BG/Q Mira at Argonne National Laboratory and the Intel Haswell cluster Shepard at Sandia National Laboratories. Our experimental results show that the prediction error rates in performance and power using MuMMI are less than 10% for most cases. By utilizing the models for runtime, node power, CPU power, and memory power, we identify the most significant performance counters for potential application optimizations, and we predict theoretical outcomes of the optimizations. Based on two collected datasets, we analyze and compare the prediction accuracy in performance and power consumption using MuMMI and 10 machine learning methods.
As we enter the exascale computing era, efficiently utilizing power and optimizing the performance of scientific applications under power and energy constraints has become critical and challenging. We propose a low-overhead autotuning framework to autotune performance and energy for various hybrid MPI/OpenMP scientific applications at large scales and to explore the tradeoffs between application runtime and power/energy for energy efficient application execution, then use this framework to autotune four ECP proxy applications—XSBench, AMG, SWFFT, and SW4lite. Our approach uses Bayesian optimization with a Random Forest surrogate model to effectively search parameter spaces with up to 6 million different configurations on two large-scale HPC production systems, Theta at Argonne National Laboratory and Summit at Oak Ridge National Laboratory. The experimental results show that our autotuning framework at large scales has low overhead and achieves good scalability. Using the proposed autotuning framework to identify the best configurations, we achieve up to 91.59% performance improvement, up to 21.2% energy savings, and up to 37.84% EDP (energy delay product) improvement on up to 4096 nodes.
Abstract Imaging the collapse of a single void that creates a hot spot initiation site in an otherwise defect‐free explosive is challenging given the spatial and temporal scales involved in explosive systems. This work presents our attempt to examine a single hot spot mode (void collapse) in single‐crystal octahydro‐l,3,5,7‐tetranitro‐l,3,5,7‐tetrazocine (HMX) embedded in Sylgard. The hot spot heating mechanisms involved with pore collapse include adiabatic heating, jetting, and viscoplastic dissipation. Quantifying the dynamics of a pore collapse is a crucial step to understanding which mechanisms dominate during ignition events. Our experiments were conducted with a single‐stage, light‐gas gun at Argonne National Laboratory's Advanced Photon Source, applying the phase contrast imaging technique while collecting high‐speed video. The details of HMX single crystal production, defect (pore) engineering, and sample construction, along with experimental results are presented here. These results demonstrate that detailed collapse dynamics can be obtained from homogeneous, single‐crystal explosives with this approach. Qualitative comparisons are made with simulation data which show good agreement in the transition between a quasi‐symmetric pore collapse and an asymmetric collapse with jetting across the pore as measured with normalized pore area and pore circularity.
This article describes the history of the Beer Game. By triangulating information from literature, archival analysis, and interviews with experts in the field, the main changes in the game over its almost 70-year history are identified. The article discusses three aspects of the game: 1) its structure (phases of its history, stocks and flows, parameters, etc.); 2) the process for playing the game; and 3) the game debrief. The structure of the Beer Game, and the process for running it, have stabilized over the years into what is now a de facto standard approach. Additional work is needed in the game debrief, specifically in the clarification of key insights and messages (depending on the context of the use of the game), in how to communicate such messages to different audiences, and in the development of support materials for its delivery. © 2024 UChicago Argonne, LLC. System Dynamics Review published by John Wiley & Sons Ltd on behalf of System Dynamics Society.
GeoSoilEnviroCARS (GSECARS) is a comprehensive analytical laboratory for Earth and environmental science research using X-ray beams from the Advanced Photon Source, Argonne National Laboratory. State-of-the-art instruments are available for (1) high-pressure/high- or low-temperature diffraction, total scattering, and spectroscopy (Brillouin, Raman, and VIS-IR) using the laser heated diamond anvil cell (DAC); (2) high-pressure/high-temperature diffraction, scattering, and imaging as well as acoustic emission (AE) and ultrasonics using the large-volume press (LVP); (3) powder, single crystal, and surface/interface diffraction; (4) X-ray absorption fine structure spectroscopy; (5) X-ray fluorescence microprobe analysis; and (6) microtomography. As shown here, experiments are facilitated by senior level staff who collaborate on all aspects of the analytical work including experiment design, sample preparation, data collection, data interpretation, and publication preparation. Both technical and scientific synergies occur as a result of the intimate association of the various techniques and scientists experienced in the applications of synchrotron radiation to Earth, environmental, and planetary science problems. The facility includes state-of-the-art instrumentation designed and built in-house, including custom X-ray optics, online and offline laser-based systems, specialized sample environments and positioning systems, as well as pixel-array and multi-crystal energy dispersive X-ray detectors, which are available to be shared among the experimental stations.
We investigate the crosstalk between Transition-Edge Sensor (TES) pixels in a 24-pixel hard X-ray spectrometer array fabricated at the Advanced Photon Source, Argonne National Laboratory. Analysis shows thermal cross talk, possibly associated with insufficient thermalization, and rare but larger in magnitude electrical crosstalk between specific perpetrator-victim pixel combinations, potentially due to defects in the bias wiring or microwave multiplexing circuit. We use a method based on group-triggering and averaging to isolate the crosstalk response despite only having access to X-ray photon illumination uniform across the entire array. This allows us to identify thermal and electrical crosstalk between pixel pairs in repeated measurements to the level of 1 part in 1000 or better. In the array under study, the magnitude of observed crosstalk is small but comparable to the resolving power of this pixel design (E/Delta E similar to\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$E/\Delta {}E \sim$$\end{document} 1000 at 20 keV) and so potentially responsible for a degradation in energy resolution of the array at high incident photon rates. Having proven the methods to identify and quantify crosstalk in our setup, we can now consider mitigations.
High Performance Computing (HPC) systems are critical enablers for groundbreaking scientific research across various domains. Efficient resource allocation, facilitated by job scheduling, is paramount for maximizing the utilization of HPC systems. However, the variability in wait times for queued jobs poses challenges for users, necessitating accurate job wait time estimation. This paper explores the influence of job characteristics, including job size (the number of nodes requested and walltime), the queue to which the job is submitted and other resource requirements, on job wait times in leadership-class HPC systems. Focusing on the Theta Cray XC40 and Polaris machines at Argonne National Laboratory, the study evaluates the performance of different supervised learning algorithms in predicting job wait times. It also evaluates the impact of data preprocessing, including outlier detection, Principal Component Analysis (PCA), and feature selection, on the performance of wait time prediction models. The findings reveal insights into the relationship between job characteristics and wait times, offering a foundation for optimizing resource allocation and enhancing user experience. The methodologies and tools developed in this study are adaptable to other leadership-class HPC systems, providing a valuable contribution to the broader HPC community aiming to improve job scheduling efficiency and user satisfaction.
Here, the ability to non-destructively map the residual strain field inside an engineering component is important for predicting its fatigue life or developing processing methods to prevent failure or enhance performance. In this paper, we describe a new residual strain mapping program at the Advanced Photon Source, Argonne National Laboratory. The new program is based on energy dispersive x-ray diffraction (EDXRD). It is capable of non-destructively penetrating a several- cm thick polycrystalline sample fabricated from engineering alloys using high-energy x-rays and measuring the residual strain field with mm or better spatial resolution and approximately ±1 x 10 -4 strain resolution. A multi-element detector array is employed to measure multiple strain components simultaneously. The residual strain mapping setup is augmented with a high-energy tomography capability, allowing precise alignment of the material volume of interest for residual strain mapping and providing a complementary view of the structure to understand the measured strain field. These measurement capabilities are demonstrated using several strain mapping examples ranging from polycrystalline structural alloys to biological materials. We also provide some guidance for the future users of the program for a successful residual strain mapping experiment. We are expanding the capabilities of the new setup with various in situ capabilities including thermo-mechanical loading.
In laser-based directed energy deposition (DED) additive manufacturing, interactions among the laser beam, particle flow, and melt pool influence the properties of the solidified final part. Two separate DED systems, one with high powder flow rates to represent industrial-scale DED processing and the other with low powder flow rates for individual particle tracking, were synchronized with the high-speed imaging setup at the Advanced Photon Source in Argonne National Laboratory. In-situ x-ray imaging of the DED process using both systems highlighted the influence of powder flow rates. Increased powder flow rates resulted in less laser absorption into the melt pool, leading to a transition from a keyhole mode to a melt pool without a keyhole but with surface fluctuations due to powder flow. Increased velocities of particles during powder flow resulted in a decrease in particle melting times and a greater propensity for porosity formation. Altogether, better understanding of the interactions that occur during various scales of the DED process will enable flexibility, control, and new materials development in DED-based additive manufacturing.
Sulfur accumulation during biofuel production is pollutive and toxic to conversion catalysts and causes the premature breakdown of processing equipment. Air classification is an effective preprocessing technology for ash and sulfur reduction from biomass feedstocks. Here, a life cycle assessment (LCA) sought to understand the environmental impact of implementing air classification as a sulfur-mitigation technique to improve feedstock quality for pine residues using a grid electricity scenario (GES) versus a bioelectricity scenario (BES). Global warming potential (GWP) for preprocessing was simulated using inventory databases embedded in SimaPro and the Argonne National Laboratory’s GREET model, specifically focusing on comparing the GWP of a GES versus a BES. Overall, the GES had a GWP impact over seven times that of the BES (136 versus 18 kg CO 2 equivalent per tonne of usable feedstock), with steam generation during rotary drying accounting for 57% of the GES’s GWP. Air classification represents 0.4% and 1.6% of the total GWP impact for the GES and BES, respectively. Therefore, air classification can facilitate a 30% reduction in feedstock sulfur content to improve feedstock quality for biofuel conversion and lessen corrosion of equipment while contributing minimal GWP impact during preprocessing.