SEARCH · Engineering Papers

Results for “benchmarks”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 433 records · Page 24

All sources NASA NTRS DOE OSTI

NASA NTRS2017

First benchmark of the Unstructured Grid Adaptation Working Group

Unstructured grid adaptation is a technology that holds the potential to improve the automation and accuracy of computational fluid dynamics and other computational disciplines. Difficulty producing the highly anisotropic elements necessary for simulation on complex curved geometries that satisfies a resolution request has limited this technology's widespread adoption. The Unstructured Grid Adaptation Working Group is an open gathering of researchers working on adapting simplicial meshes to conform to a metric field. Current members span a wide range of institutions including academia, industry, and national laboratories. The purpose of this group is to create a common basis for understanding and improving mesh adaptation. We present our first major contribution: a common set of benchmark cases, including input meshes and analytic metric specifications, that are publicly available to be used for evaluating any mesh adaptation code. We also present the results of several existing codes on these benchmark cases, to illustrate their utility in identifying key challenges common to all codes and important differences between available codes. Future directions are defined to expand this benchmark to mature the technology necessary to impact practical simulation workflows.

Ibanez, Daniel↗

NASA NTRS2021

Comparing Commercial and Research Computational Fluid Dynamic Codes Using High-Order Workshop Benchmark Problems

The commercial computational fluid dynamic (CFD) code ANSYS Fluent and multiple research CFD codes (ez4d with uses the conservation element and solution element (CESE) method and codes that use the flux reconstruction (FR) method) were tested using three different benchmark problems from the International Workshop for High-Order CFD Methods. The benchmark problems included the transonic Ringleb flow, vortex transport by uniform flow, and laminar boundary layer on a flat plate. Simulation results from all three benchmark problems showed that the Fluent solutions had less error than the ez4d solutions for a given degree of freedom. As expected, both the Fluent and ez4d solutions had larger errors for a given degree of freedom than the simulations that used the FR method because both Fluent and ez4d utilized a second-order scheme whereas the FR codes utilized a fourth-order scheme.

CFD↗

NASA NTRSDate not supplied

Benchmark Models for Classification of Radiation Type Induced in Immune Cells

NASA Biological and Physical Sciences and the Science Mission Directorate have published a benchmark dataset of mouse immune cells subjected to radiation-induced DNA damage. The dataset comprises ML-ready microscopic imagery of said cells, including labels indicating radiation type and dose. The machine learning team at NASA Interagency Implementation and Advanced Concept Team (IMPACT) created multiple benchmark models. Initially, we conducted a preliminary analysis using thresholding. The algorithm used thresholds on average brightness of the available images to classify them into their respective radiation type. We also tested machine learning approaches. Convolutional Neural Networks (CNN) emerged as the best-performing model. This poster presents the benchmark scores obtained by the models.

Vishal Perekadan↗

NASA NTRS2021

Benchmarking Planning Applications on the Qualcomm Snapdragon

We benchmark several space planning/scheduling applications on the Qualcomm Snapdragon 855 Handheld Development Kit (HDK), a high performance embedded processor used in many mobile phones.We are flying 2 Snapdragon HDKs onboard the International Space Station (ISS) where they are hosted by the the Spaceborne Computer-2 by Hewlett Packard Enterprise linked by USB and 12V power delivery. We run computational benchmarks using three planner/ schedulers that are used for several space missions: Multi- Mission Executive (MEXEC), Compressed Large-scale Activity Scheduling and Planning (CLASP), and M2020 Ground Scheduler (Surrogate). We compare the Snapdragon performance to a performance baseline on Linux workstations. In addition, we are currently working on benchmarking the same applications on other space flight processors, such as the LEON4 Processor on the Sabertooth card, the LEON3 Processor on the Sphinx card, and the RAD750 processor.

Chien, Steve↗

NASA NTRSDate not supplied

Investigations of an Aeroelastic Optimization Benchmark Problem with MPhys and FUN3D

High-fidelity aeroelastic optimization capabilities are becoming more common in recent years. In this field, researchers typically present results on the design problems of greatest interest to them. The lack of consistency with respect to baseline configurations, design variables, objectives and constraints, etc., makes it difficult to compare approaches and creates an obstacle for the community to work together to advance the state of the art. The High-Fidelity Aeroelastic Optimization Benchmark Working Group has been established to create and study a series of benchmark aeroelastic optimization problems that the community can utilize to compare methods on the same optimization problem. This paper presents optimization results from NASA Langley Research Center for the first benchmark case of this series with analysis based on MPhys, a multiphysics library for OpenMDAO, and FUN3D, an unstructured computational fluid dynamics suite of tools. For the optimization case presented, it is demonstrated that a progression of increasing aerodynamic model fidelity reduces the wall time required to complete the optimization.

Aeroelasticity↗

NASA NTRSDate not supplied

Investigations of an Aeroelastic Optimization Benchmark Problem with MPhys and FUN3D

CFD↗

NASA NTRS2026

Benchmarking Bayesian Optimization Frameworks and Acquisition Strategies for Materials Discovery and Autonomous Laboratories

Bayesian optimization (BO) can accelerate materials discovery by guiding expensive experiments toward the most promising processing conditions. We systematically compare five BO surrogate and framework combinations (Gaussian processes in Ax, Gaussian processes and Monte-Carlo neural networks in BayBE, random forests in Lolopy, and tree-structured Parzen (TPE) estimators in Hyperopt) on three benchmarks that mimic common materials design tasks (a discrete solid-electrolyte composition space, a hybrid discrete/continuous laminate-composite design problem solved with micromechanics modeling, and the continuous Ishigami analytic function which is a standard optimization benchmark). Each BO surrogate is paired with posterior mean, probability of improvement, and expected improvement acquisition functions and run for 100 trials from randomized initial samples with uniform random search providing a control. Across five random seeds per setting, BayBE’s Gaussian-process surrogate with expected improvement consistently reached ≥95 % of the known optimum in the fewest evaluations, while Lolopy’s random forest matched or exceeded GP performance on purely categorical or mixed spaces at a higher computational cost. Posterior mean alone often stagnated at local optima, underscoring the need for exploration, whereas probability and expected improvement balanced exploration and exploitation leading to better optimization in fewer trials. Execution times ranged from milliseconds for TPE to minutes for neural-network and random-forest surrogates. These results establish baseline expectations for BO in automated materials laboratories and highlight expected improvement with Gaussian processes as a reliable first choice, with random forests offering a strong alternative when categorical variables dominate. The benchmark suite and code are released to facilitate future surrogate, acquisition, and constraint-handling research in data-driven materials optimization.

Bayesian optimization↗

DOE OSTI2021

MULTIPHYSICS-MODELING OF FIRE-INDUCED URANIUM AEROSOL FORMATION – A-POSTERIORI BENCHMARKING OF EXPERIMENTS

The formulation and solution of a model that properly described the temperature profiles of uranium in fire conditions was prepared, verified and validated, to support a-posteriori benchmarking of historical experiments. Data from the multi-physics model combined with data regression from the experiments provides a useful tool for exploring sources of potential bias in historical experiments. The approaches used are augmented by visual observations and photographic evidence allowing for correlations between various phenomena, independent of whether or not that specific phenomena are part of the validated model. In this manner, the validated model is a tool for a-posteriori benchmarking, and not a model designed to replicate the experimental output exactly for any specific experiment. For experiments like the Clark (2015) experiments, this integrated model (including chemistry) demonstrated that the largest bias was associated in the metallurgy of the coupons. The model developed does serve as a basis for defending which of the historical alloys tested are the most appropriate for any other alloy being considered (e.g., one that has not yet been tested). The model identifies key reactions that impact uranium chemistry and can significantly bias experimental results. Through use of the model and energy balances, data from historical experiments by Elder and Tinkle (1980) were regressed and benchmarked. This study identifies that all uranium alloys are not equal and that there are distinct differences between the dominant alloys representing distinct metallurgical phases.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2021

Evaluation of Oak Ridge National Laboratory Health Physics Research Reactor Operation Data for Criticality Accident Alarm System Benchmark Creation

Over the last few decades, the International Criticality Safety Benchmark Evaluation Project (ICSBEP) Working Group has focused on gathering critical and subcritical benchmark experiment data from different facilities around the world for inclusion in the International Handbook of Evaluated Criticality Safety Benchmark Experiments (ICSBEP Handbook). The compiled data can be used by criticality safety analysts to help validate their calculation tools and cross-section libraries for a variety of applications.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

DOE OSTI2022

Open-source FPGA-ML codesign for the MLPerf Tiny Benchmark

We present our development experience and recent results for the MLPerf Tiny Inference Benchmark on field-programmable gate array (FPGA) platforms. We use the open-source hls4ml and FINN workflows, which aim to democratize AI-hardware codesign of optimized neural networks on FPGAs. We present the design and implementation process for the keyword spotting, anomaly detection, and image classification benchmark tasks. The resulting hardware implementations are quantized, configurable, spatial dataflow architectures tailored for speed and efficiency and introduce new generic optimizations and common workflows developed as a part of this work. The full workflow is presented from quantization-aware training to FPGA implementation. The solutions are deployed on system-on-chip (Pynq-Z2) and pure FPGA (Arty A7-100T) platforms. The resulting submissions achieve latencies as low as 20 $\mu$s and energy consumption as low as 30 $\mu$J per inference. We demonstrate how emerging ML benchmarks on heterogeneous hardware platforms can catalyze collaboration and the development of new techniques and more accessible tools.

Borras, Hendrik↗

DOE OSTI2022

FastML Science Benchmarks: Accelerating Real-Time Scientific Edge Machine Learning

Applications of machine learning (ML) are growing by the day for many unique and challenging scientific applications. However, a crucial challenge facing these applications is their need for ultra low-latency and on-detector ML capabilities. Given the slowdown in Moore's law and Dennard scaling, coupled with the rapid advances in scientific instrumentation that is resulting in growing data rates, there is a need for ultra-fast ML at the extreme edge. Fast ML at the edge is essential for reducing and filtering scientific data in real-time to accelerate science experimentation and enable more profound insights. To accelerate real-time scientific edge ML hardware and software solutions, we need well-constrained benchmark tasks with enough specifications to be generically applicable and accessible. These benchmarks can guide the design of future edge ML hardware for scientific applications capable of meeting the nanosecond and microsecond level latency requirements. To this end, we present an initial set of scientific ML benchmarks, covering a variety of ML and embedded system techniques.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2023

MUSiC: Critical Experiment with Bare Highly Enriched Uranium Shells Benchmark

The measurements of the Uranium Subcritical and Critical (MUSiC) experiment were performed from December 2020 through April 2021. The purpose of the experiment was to test a variety of multiplicity detection systems for a large range of multiplication values and to see how well they perform as the k eff of the MUSiC series of experiments approached the delayed critical window. Out of the ten configurations that were studied, two of them were super critical and were recently approved as benchmarks and they will be included in the International Criticality Safety Benchmark Evaluation Project (ICBESP) handbook. For each of the two experiments, uncertainties were evaluated six categories: (1) critical measurement, (2) mass, (3) composition, (4) positioning, (5) dimensions, and (6) temperature. The largest contribution to the overall uncertainty in each of the cases was due to the uncertainties in the dimensions of the HEU Rocky Flats shells. Simplified detailed models were created with an average bias of -130.5 pcm due to the removal various components surrounding the experiment. For MCNP6.2 models using ENDF-VIII.0 neutron cross section data, the average difference between the calculated and benchmark k effs was -54 pcm for the detailed model and -102 pcm for the simplified model.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2023

DeepSurveySim: Simulation Software and Benchmark Challenges for Astronomical Observation Scheduling

Modern astronomical surveys have multiple competing scientific goals. Optimizing the observation schedule for these goals presents significant computational and theoretical challenges, and state-of-the-art methods rely on expensive human inspection of simulated telescope schedules. Automated methods, such as reinforcement learning, have recently been explored to accelerate scheduling. However, there do not yet exist benchmark data sets or user-friendly software frameworks for testing and comparing these methods. We present DeepSurveySim -- a high-fidelity and flexible simulation tool for use in telescope scheduling. DeepSurveySim provides methods for tracking and approximating sky conditions for a set of observations from a user-supplied telescope configuration. We envision this tool being used to produce benchmark data sets and for evaluating the efficacy of ground-based telescope scheduling algorithms, particularly for machine learning algorithms that would suffer in efficacy if limited to real data for training.We introduce three example survey configurations and related code implementations as benchmark problems that can be simulated with DeepSurveySim.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2016

Benchmarking of the Updated Resolved Resonance Region Evaluations of Copper

A new resolved resonance region (RRR) evaluation for the two naturally occurring isotopes of copper, {sup 63}Cu and {sup 65}Cu was performed at Oak Ridge National Laboratory (ORNL), and the benchmarking results based on the four highly enriched uranium (HEU), intermediate spectrum, copper-reflected Zeus experiments were presented in June 2014. Since then, two significant updates to the evaluations have been performed: first, the data for the experimental measurement of the capture cross section between 220 keV and 300 keV were renormalized and the resonance parameters were refitted. Second, the angular distribution of elastic scattering initially reported in point-wise form in File 4 of the ENDF file format with high fidelity was significantly reduced in size to allow for better computer storage. Twenty-seven criticality safety benchmarks from the International Criticality Safety Benchmark Evaluation Project (ICSBEP) were calculated using these updated resolved resonance region evaluations for the two isotopes of copper. (authors)

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2023

RELAP5-3D Solutions to Exercise 1 of the OECD-NEA HTTF Benchmark

Discussion of RELAP5-3D solutions to Exercise 1 of the OECD-NEA High Temperature Test Facility (HTTF) Benchmark. Results from ongoing benchmark activities supported by the Advanced Reactor Technologies - Gas-Cooled Reactor Program with use of the High Performance Computing Center at Idaho National Laboratory. Also discussion of Ultra Safe Nuclear Corporation and BWXT plans for the decade, and issues of modern thermal hydraulic tools for prismatic HTGR modeling. Open discussion of verification and validation and ART-GCR campaign to spearhead the development of an HTGR thermal hydraulics benchmark based on the HTTF.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2023

RELAP5-3D Modeling in the OECD-NEA HTTF Benchmark

Prismatic gas-cooled reactors are a technologically mature reactor concept of interest for near- to mid-term deployment. To accelerate the deployment of these reactors, the DOE's Advanced Reactor Technologies Gas-Cooled Reactor campaign is spearheading a thermal hydraulics code validation benchmark based on the High Temperature Test Facility at Oregon State University. This presentation provides an introduction to that benchmark and an overview of some of the RELAP5-3D validation activities stemming from that benchmark.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2026

A Benchmarking Framework for Evaluating Large Language Model Capabilities in Nuclear Reactor Safety Applications

Large language models (LLMs) are increasingly capable of answering technical questions, synthesizing domain knowledge, and supporting engineering workflows. For nuclear science and engineering, these capabilities require careful, domain-specific evaluation before they can be credibly incorporated into safety-related activities, regulatory review, or technical decision support. This paper presents preliminary results from benchmarking framework for evaluating LLM capabilities in nuclear contexts. The framework is organized into three evaluation categories: nuclear fundamentals, general dual-use knowledge, and plant specific knowledge. These categories are intended to distinguish general nuclear engineering competence from broader technical reasoning and more context-dependent nuclear knowledge. Initial evaluations focus on nuclear fundamentals using questions representative of the knowledge expected of a nuclear professional engineer. Results indicate that contemporary frontier models perform at a high level and substantially exceed the performance of older model generations, with some models approaching saturation of the current benchmark. These findings suggest both the rapid improvement of LLM capabilities in specialized technical domains and the need for more discriminating evaluation methods. The paper presents the benchmark structure, preliminary model-comparison results, and ongoing work. This work supports development of verifiable, responsible, and safety-conscious methods for assessing AI systems in nuclear engineering applications.

21 - SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLAN↗

DOE OSTI2024

IER 480: TEX-Pu Benchmark (PU-MET-THERM-004) to Test Polyethylene and Lucite Thermal Scattering Laws [Slides]

This presentation finds that PMT-004 has four new benchmark cases highly sensitive to PE TSL (2 cases) and PMMA TSL (2 Cases). PE cases were well predicted using MCNP6.2 and ENDF/B-VIII.0. PMMA cases overpredicted by approximately 0.6-0.7% in k eff at 20°C. Accepted into 2023 version of the ICSBEP Handbook. Temperature had a large impact on reactivity of the critical configuration. Implications for validation work for thermal cases- need to adjust TSL data to correct temperature as it can have hundreds of pcm effects for a few °C. Future thermal experiments should try and measure reactivity at multiple temperatures to aid in data testing and benchmark adjustment.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗