Mass-Constrained hybrid Gaussian radial basis neural networks: Development, training, and applications to modeling nonlinear dynamic noisy chemical processes
Not Available
Mukherjee, Angan (ORCID:000000025580153X)↗
SEARCH · Engineering Papers
Results for “Application Development”
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.
GLObal Building heights for Urban Studies (UT-GLOBUS) for city- and street- scale urban simulations: Development and first applications
Not Available
54 ENVIRONMENTAL SCIENCES↗
Developing Renewable Energy Applications for Water Treatment Technologies and Transmission
The need for smaller treatment facilities for more isolated or rural municipal entities and industrial water treatment facilities continues to grow. From environmental justice and energy needs in remote communities that have never been serviced by larger scale utilities to smaller onsite treatment requirements to meet discharge regulations, energy efficient treatment systems that can leverage renewable energy sources can be an important asset to owners. Climate change and population dynamics have changed the needs for distributed water treatment systems and the manner in which they function. New treatment technologies that can operate under dynamic conditions and leverage variable energy sources (e.g., renewable energy, RE) can provide substantial efficiencies and cost savings. The National Renewable Energy Laboratory (NREL) is embarking on research projects that will perform multi-scale studies to investigate treatment technologies on varying source water quality that can operate under unsteady state flow conditions (timing) in response to energy availability, supply, and demand. A pilot-scale system will be built to assess modularity and scaling of treatment system operating conditions and performance. Modeling will also be used to determine feasibility and integration into grid emulation platforms to understand the use and value of operation with RE. Optimization will be completed for the treatment and distribution systems to document strengths and weaknesses of using RE power sources. Results of this effort will advance dynamic water treatment understanding under variable conditions and integration of RE sources to power these systems. This will lead to further decarbonization of municipal and private treatment facilities while supporting a paradigm shift in water treatment process operation and system design to achieve net-zero goals and move toward energy-sensitive resilient water treatment systems.
ENERGY PLANNING, POLICY, AND ECONOMY↗
QMCPACK: Advances in the development, efficiency, and application of auxiliary field and real-space variational and diffusion quantum Monte Carlo
We review recent advances in the capabilities of the open source ab initio Quantum Monte Carlo (QMC) package QMCPACK and the workflow tool Nexus used for greater efficiency and reproducibility. The auxiliary field QMC (AFQMC) implementation has been greatly expanded to include k-point symmetries, tensor-hypercontraction, and accelerated graphical processing unit (GPU) support. These scaling and memory reductions greatly increase the number of orbitals that can practically be included in AFQMC calculations, increasing the accuracy. Advances in real space methods include techniques for accurate computation of bandgaps and for systematically improving the nodal surface of ground state wavefunctions. Results of these calculations can be used to validate application of more approximate electronic structure methods, including GW and density functional based techniques. To provide an improved foundation for these calculations, we utilize a new set of correlation-consistent effective core potentials (pseudopotentials) that are more accurate than previous sets; these can also be applied in quantum-chemical and other many-body applications, not only QMC. These advances increase the efficiency, accuracy, and range of properties that can be studied in both molecules and materials with QMC and QMCPACK.
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
Phenomenon Identification and Ranking Table Development for Future Application Figure-of-Merit Studies on Thermal Energy Storage Integrations with Light Water Reactors
There are no standard prioritization criteria for evaluating thermal energy storage (TES) options for use in integrated energy systems. A framework for proposing, analyzing, and presenting energy storage integration with power producers and users is presented along with a specific figure-of-merit (FOM) study based in this framework. This basis for evaluating storage technologies can provide a structure for the energy industry to analyze and prioritize energy storage in different applications and environments. The phenomena identification and ranking table (PIRT) presents a series of design questions specific to energy storage applications. The FOM study, built in this PIRT framework based on a nuclear-renewable hybrid energy system using TES to produce power and provide process energy for a secondary user, successfully identified specific technologies to use based on the project requirements. Expanding the library of projects using this framework will expand the deployable options for energy storage and increase its potential for energy security.
25 ENERGY STORAGE↗
Machine-learned interatomic potentials: Recent developments and prospective applications
Not provided.
Materials Science↗
Conduction-cooled SRF accelerator development for industrial applications at Fermilab [Slides]
Explore the source record for details and available documents.
43 PARTICLE ACCELERATORS↗
Development of Spectroscopic Tools for Nuclear Forensics Signatures Development
Diversion of nuclear materialsfrom legitimate commercial processes for use by nefarious actors is a frightening and surreptitious problem. Smuggled nuclear material can be used to manufacture a simple radiological dispersion device (RDD) and potentially incorporated into the development of a nuclear weapon. With some frequency, smuggled nuclear material is interdicted in foreign countries as actors cross border checkpoints. Nuclear material interdicted in this way can be analyzed using a variety of tools. Plutonium-239 is commonly used as the fissile material in nuclear weapon production as an alternative to uranium-235. Plutonium in a production, refining, or finishing facility will exist in many forms including oxide precursors (PuF4, PuF3, Pu oxalate, etc.), oxide, and metal. The ability to identify plutonium in each of these forms and determine properties, characteristics, or history of these materials is crucial for determining if the processes occurring in a nuclear facility are legitimate or not. This presentation will focus on the development and application ofspectroscopic tools developed at SRNL using doubled-walled cells (DWC) to characterize Pu-bearing compounds. Advances in the use of vibrational spectroscopy techniques (Raman and IR), UV-VIS-SWIR diffuse reflectance spectroscopy, gamma emission spectroscopy, luminescence spectroscopy, laser-induced breakdown spectroscopy in conjunction with XRD and SEM in the characterization of Pu-bearing compounds will be presented with emphasis in material processing history.
Villa-Aleman, Eliel↗
The Effective Field Theories: Recent Developments and their Applications
Explore the source record for details and available documents.
Atomic, Nuclear and Particle Physics↗
Emerging Jets Search, Triton Server Deployment, and Track Quality Development: Machine Learning Applications in High Energy Physics
Machine learning is becoming prevalent in high energy physics, with numerous applications in physics analyses and event reconstruction showing great improvements compared to traditional computing methods. This thesis studies three projects which each propose new avenues for machine learning applications within the high energy physics CMS experiment located at CERN. In the first project, a search for a dark matter signal called “emerging jets” is performed, using graph neural networks to greatly increase sensitivity to the signal’s signature within the data. The result of this dark matter search sets the most stringent exclusion limits to date on theoretical emerging jet models. Motivated by inefficiencies encountered when processing the emerging jet graph neural network at Fermi National Accelerator Laboratory’s computing centers, the second project re-optimizes the computing centers for machine learning inference. This re-optimization uses NVIDIA Triton Inference Servers to process users’ analysis code heterogeneously, therefore achieving high processing throughput and decreasing user time-to-insight. The last project focuses on an upgrade to the CMS experiment’s real-time event selection system which improves physics object reconstruction under harsh processing conditions. A boosted decision tree is used to quickly and efficiently quantify a reconstructed particle’s “track quality” in order to remove particle tracks reconstructed erroneously. In summary, this thesis will not only present examples of how high energy physics can greatly benefit by leveraging machine learning techniques for physics analysis and reconstruction, but will also provide guidance on how the field can prepare for the inevitable increase in machine learning applications.
72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
Neural architecture codesign for fast physics applications
We develop a pipeline to streamline neural architecture codesign for physics applications to reduce the need for ML expertise when designing models for novel tasks. Our method employs neural architecture search and network compression in a two-stage approach to discover hardware efficient models. This approach consists of a global search stage that explores a wide range of architectures while considering hardware constraints, followed by a local search stage that fine-tunes and compresses the most promising candidates. We exceed performance on various tasks and show further speedup through model compression techniques such as quantization-aware-training and neural network pruning. We synthesize the optimal models to high level synthesis code for FPGA deployment with the hls4ml library. Additionally, our hierarchical search space provides greater flexibility in optimization, which can easily extend to other tasks and domains. We demonstrate this with two case studies: Bragg peak finding in materials science and jet classification in high energy physics, achieving models with improved accuracy, smaller latencies, or reduced resource utilization relative to the baseline models.
72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
LDMS-GPU: Lightweight Distributed Metric Service (LDMS) for NVIDIA GPGPUs
GPUs are now a fundamental accelerator for many high-performance computing applications. They are viewed by many as a technology facilitator for the surge in fields like machine learning and Convolutional Neural Networks. To deliver the best performance on a GPU, we need to create monitoring tools to ensure that we optimize the code to get the most performance and efficiency out of a GPU. Since NVIDIA GPUs are currently the most commonly implemented in HPC applications and systems, NVIDIA tools are the solution for performance monitoring. The Light-Weight Distributed Metric System (LDMS) at Sandia is an infrastructure widely adopted for large-scale systems and application monitoring. Sandia has developed CPU application monitoring capability within LDMS. Therefore, we chose to develop a GPU monitoring capability within the same framework. In this report, we discuss the current limitations in the NVIDIA monitoring tools, how we overcame such limitations, and present an overview of the tool we built to monitor GPU performance in LDMS and its capabilities. Also, we discuss our current validation results. Most of the performance counter results are the same in both vendor tools and our tool when using LDMS to collect these results. Furthermore, our tool provides these statistics during the entire runtime of the tool as a time series and not just aggregate statistics at the end of the application run. This allows the user to see the progress of the behavior of the applications during their lifetime.
97 MATHEMATICS AND COMPUTING↗
Facile Synthesis of Pt Carbide Nanomaterials and Their Catalytic Applications
The development of transition metal carbides, for example, Pd carbide, Mo carbide, Co carbide, and Fe carbide for catalytic applications has attracted significant attention. Pt carbide has also been actively pursued, however, no catalytic examinations of Pt carbide have been performed, likely due to its extreme synthesis conditions. In this study, we report a facile synthesis of Pt carbide nanomaterials under mild conditions, for the first time, through ethane treatment of confined Pt nanoclusters under mild temperatures and ambient pressure. Such Pt carbide nanomaterials exhibit remarkable activity in catalytic hydrocarbon conversion, including ethane aromatization and acetylene hydrogenation, and are superior over supported Pt nanoparticles. Our work further reveals that Pt nanoclusters are favored in terms of forming Pt carbide when compared to isolated single Pt atoms and large Pt nanoparticles. This work provides a new approach to prepare metal carbide based catalytic materials through nanoclusters under mild conditions.
36 MATERIALS SCIENCE↗
Adrastea: An Efficient FPGA Design Environment for Heterogeneous Scientific Computing and Machine Learning
We present Adrastea, an efficient FPGA design environment for developing scientific machine learning applications. FPGA development is challenging, from deployment, proper toolchain setup, programming methods, interfacing FPGA kernels, and more importantly, the need to explore design space choices to get the best performance and area usage from the FPGA kernel design. Adrastea provides an automated and scalable design flow to parameterize, implement, and optimize complex FPGA kernels and associated interfaces. We show how virtualization of the development environment via virtual machines is leveraged to simplify the setup of the FPGA toolchain while deploying the FPGA boards and while scaling up the automated design space exploration to leverage multiple machines concurrently. Adrastea provides an automated build and test environment of FPGA kernels. By exposing design space hyper-parameters, Adrastea can automatically search the design space in parallel to optimize the FPGA design for a given metric, usually performance or area. Adrastea simplifies the task of interfacing with the FPGA kernels with a simplified interface API. To demonstrate the capabilities of Adrastea, we implement a complex random forest machine learning kernel with 10,000 input features while achieving extremely low computing latency without loss of prediction accuracy, which is required by a scientific edge application at SNS. We also demonstrate Adrastea using an FFT kernel and show that for both applications Adrastea is able to systematically and efficiently evaluate different design options, which reduced the time and effort required to develop the kernel from months of manual work to days of automatic builds.
Young, Aaron↗
SAM Finite Volume Method Development Status Update: GCR Application, Restart, and MultiApp
The System Analysis Module (SAM) is being developed as a modern system analysis code for advanced non-light-water-reactor safety analysis under the U.S. DOE NEAMS program. Previous feasibility studies have demonstrated that a staggered-grid finite volume method (SG-FVM), implemented under the MOOSE framework, can deliver more than an order of magnitude speedup over the existing continuous Galerkin finite element method (CG-FEM) solver for liquid-cooled, incompressible but thermally expandable flow systems. This work extends the previous effort to compressible, gas-cooled reactor applications, where pressure couples directly into the mass equation adding additional nonlinearity into the equation system. New code capabilities are implemented for pebble bed high-temperature gas-cooled reactor (PB-HTGR) analysis, including a pebble bed CoreChannel component, built-in pebble bed effective thermal conductivity model and channel-to-channel crossflow model. The capabilities are tested, benchmarked, and demonstrated for problems with increased level of model and physical complexities, including the HTTU effective thermal conductivity test, the SANA passive cooling test, and a demonstration case using the GPBR200 reactor design covering steady-state operation, DLOFC and PLOFC transients. Across all cases, the SG-FVM solver demonstrated strong robustness and efficiency, and the solutions agree well with reference results and data. The finding of this work proves that SG-FVM is a viable and efficient solver pathway for compressible, gas-cooled reactor system analysis in SAM. In addition, work has been done to successfully support SAM-FVM recover/restart code feature that is essential to reactor safety analysis applications, and MultiApp code feature that is essential to multi-scale and multi-physics simulations. In summary, this work continued from previous feasibility studies, and further demonstrated that the SG-FVM will serve as a strong foundation for SAM’s advanced solver algorithm for future deployment.
Zou, Ling↗
Laser-assisted embedding of all-glass optical fiber sensors into bulk ceramics for high-temperature applications
We develop a laser-assisted sensor embedding process to embed all-glass optical fiber sensors into bulk ceramics for high-temperature applications. A specially designed two-step microchannel was fabricated on an Al 2 O 3 substrate for sensor embedment using a picosecond (ps) laser. An optical fiber Intrinsic Fabry-Perot Interferometer (IFPI) sensor was embedded at the bottom of the microchannel and covered by Al 2 O 3 slurry which was subsequently sintered by a CO 2 laser. The sensor spectrum was in-situ monitored during the laser sintering process to ensure the survival of the sensor and optimize the laser sintering parameters. By testing in furnace through high temperature, the embedded optical fiber shows improved stability after CO 2 laser sealing, resulting in the linear temperature response of the embedded optical fiber IFPI sensor. To improve the embedded IFPI sensor for thermal strain measurement, a dummy fiber was co-embedded with the sensing fiber to improve the mechanical bonding between the sensing fiber and the ceramic substrate so that the thermal strain of the ceramic substrate can apply on the sensing fiber. The response sensitivity, measurement repeatability and high-temperature long-term stability of the embedded optical fiber IFPI sensor were evaluated in this work.
47 OTHER INSTRUMENTATION↗
Impact of cycling conditions on lithium-ion battery performance for electric vertical takeoff and landing applications
The development of better electrochemical energy storage systems has sparked significant interest in using Li-ion batteries for electric vertical takeoff and landing (eVTOL) applications. To ensure the optimal performance and safety of onboard batteries, their behavior under different charging/discharging protocols and environmental conditions must be understood. Here, this paper presents a comprehensive evaluation of commercial Li-ion batteries for eVTOL applications, focusing on their responses to varying charging/discharging strategies and mechanical vibrations experienced during flight. Through controlled experiments, the effects of rapid cycling on battery performance were investigated, including effects on lifespan, capacity, and internal resistance. Additionally, the impact of mechanical vibrations on battery behavior was assessed to identify potential challenges for onboard batteries. The results of this study revealed intriguing insights into the interplay between temperature, vibration, and battery performance. This work contributes to the broader adoption of electric aerial transportation, promising a greener and safer future for urban mobility.
18650 cells↗