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 white paper briefly summarized key conclusions of the recent US Community Study on the Future of Particle Physics (Snowmass 2021) workshop on Software and Computing for Small High Energy Physics Experiments.
There is an increased interest in instrument-computing ecosystems (ICEs) that support science workflows empowered by AI-automated experiments and computations in diverse areas. In particular, electrochemistry ICEs are promising for accelerating the design and discovery of electrochemical systems for energy storage and conversion, by automating significant parts of workflows that combine synthesis and characterization experiments with computations. They require the integration of flow controllers, solvent containers, pumps, fraction collectors, and potentiostats, all connected to an electrochemical cell, as illustrated in Fig. 1. These are specialized instruments with custom software that is not originally designed for network integration. We developed network and software solutions for electrochemical workflows that adapt system and instrument settings in real-time for multiple rounds of experiments. In particular, we developed Python wrappers for Application Programming Interfaces (APIs) of instrument commands and Pyro client-server modules that enable them to be executed from remote computers. The entire workflow is orchestrated by a Jupyter notebook running on a remote computer.
Turbulent separated flow over a two-dimensional hump is computed by solving the RANS equations with k - omega (SST) turbulence model for the baseline, steady suction and oscillatory blowing/suction flow control cases. The flow equations and the turbulent model equations are solved using a fifth-order accurate weighted essentially. nonoscillatory (WENO) scheme for space discretization and a third order, total variation diminishing (TVD) Runge-Kutta scheme for time integration. Qualitatively the computed pressure distributions exhibit the same behavior as those observed in the experiments. The computed separation regions are much longer than those observed experimentally. However, the percentage reduction in the separation region in the steady suction case is closer to what was measured in the experiment. The computations did not predict the expected reduction in the separation length in the oscillatory case. The predicted turbulent quantities are two to three times smaller than the measured values pointing towards the deficiencies in the existing turbulent models when they are applied to strong steady/unsteady separated flows.
A computer model is developed using infinitely long concentric cylindrical shells to represent the neutral atoms, ions and electrons in the barium cloud. The neutral shells are given a distribution of positions and velocities whose parameters are chosen to be consistent with the dynamics of the release. From this distribution, the ion and electron shells are generated at random using the observed time constant for photoionization. The ion and electron shells thus formed are followed using self-consistent equations of motion. Various averages which could be compared with observation of the actual cloud are calculated at regular time intervals. An unexpected result is the predicted very early return of the magnetic field within the cloud to its ambient value.
The Department of Meteorology at the University of Maryland is developing one of the first computer systems in meteorology to take advantage of the new networked computer architecture that has been made possible by recent advances in computer and communication technology. Elements of the department's system include scientific workstations, local mainframe computers, remote mainframe computers, local-area networks,'long-haul' computer-to-computer communications, and 'receive-only' communications. Some background is provided, together with highlights of some lessons that were learned in carrying out the design. In agreement with work in the Unidata Project, this work shows that the networked computer architecture discussed here presents a new style of resources for solving problems that arise in meteorological research and education.
Computer vision models have great potential as tools for international nuclear safeguards verification activities, but off-the-shelf models require fine-tuning through transfer learning to detect relevant objects. Because open-source examples of safeguards-relevant objects are rare, and to evaluate the potential of synthetic training data for computer vision, we present the Limbo dataset. Limbo includes both real and computer-generated images of uranium hexafluoride containers for training computer vision models. Here, we generated these images iteratively based on results from data validation experiments that are detailed here. The findings from these experiments are applicable both for the safeguards community and the broader community of computer vision research using synthetic data.
Examination of the literature shows that the comparison between experiment and computation for highly separated unswept compression ramp flows is generally poor, irrespective of the turbulence model used. In general, the upstream influence is not correct, the wall pressure rise through separation is too steep, and the pressures under the separated shear layer are too high. In the current study, the objective is to determine if these discrepancies might be attributed more to other factors such as flowfield unsteadiness or three-dimensionality, rather than to inadequate turbulence modeling. To examine this possibility, multichannel wall pressure fluctuations were measured under the unsteady separation shock wave in a 28-deg unswept compression ramp flow at Mach 5. The results show that the large scale, low frequency separation shock unsteadiness controls the distribution of time-averaged surface properties and that neglect of the unsteadiness is probably the primary cause of the discrepancy between experiment and computation.
Turbulent separated flow over a two-dimensional hump is computed by solving the RANS equations with k-omega (SST) turbulence model for the baseline, steady suction and oscillatory blowing/suction flow control cases. The flow equations and the turbulent model equations are solved using a fifth-order accurate weighted essentially nonoscillatory (WENO) scheme for space discretization and a third order, total variation diminishing (TVD) Runge-Kutta scheme for time integration. Qualitatively the computed pressure distributions exhibit the same behavior as they are observed in the experiments. The computed separation regions are much longer than that are observed. However, the percentage reduction in the separation region in the steady suction case is closer to that was measured in the experiment. The computations did not predict the expected reduction in the separation length in the oscillatory case. The predicted turbulent quantities are two to three times smaller than that are measured and it points towards the deficiencies in the existing turbulent models when they are applied to strong steady/unsteady separated flows.
An aerodynamic computer code, capable of predicting unsteady and C sub m values for an airfoil undergoing dynamic stall, is used to predict the amplitudes and frequencies of a wing undergoing torsional stall flutter. The code, developed at United Technologies Research Corporation (UTRC), is an empirical prediction method designed to yield unsteady values of normal force and moment, given the airfoil's static coefficient characteristics and the unsteady aerodynamic values, alpha, A and B. In this experiment, conducted in the PSU 4' x 5' subsonic wind tunnel, the wing's elastic axis, torsional spring constant and initial angle of attack are varied, and the oscillation amplitudes and frequencies of the wing, while undergoing torsional stall flutter, are recorded. These experimental values show only fair comparisons with the predicted responses. Predictions tend to be good at low velocities and rather poor at higher velocities.
Future scientific discoveries will rely on flexible ecosystems that incorporate modern scientific instruments, high performance computing resources, parallel distributed data storage, and performant networks across multiple, independent facilities. In addition to connecting physical resources, such an ecosystem presents many challenges in logistics and accessibility, especially in orchestrating computations and experiments that span across leadership computing systems and experimental instruments. Past efforts have typically been application-specific or limited to interfaces for computing resources. This paper proposes a general framework for integrating computation resources and instrument operations, addressing challenges in code development/execution, data staging and collection, software stack, control mechanisms, resource authorization and governance, and hardware integration. We also describe a demonstration use case wherein a Bayesian optimization algorithm running on an edge computing resource guides a scanning probe microscope to autonomously and intelligently characterize a material sample. This science edge ecosystem framework will provide a blueprint for federating multi-institutional, disparate resources and orchestrating scientific workflows across them to enable next-generation discoveries.
X-ray photon correlation spectroscopy (XPCS) allows for the resolution of dynamic processes within a material across a wide range of length and time scales. X-ray speckle visibility spectroscopy (XSVS) is a related method that uses a single diffraction pattern to probe ultrafast dynamics. Interpretation of the XPCS and XSVS data in terms of underlying physical processes is necessary to establish the connection between the macroscopic responses and the microstructural dynamics. To aid the interpretation of the XPCS and XSVS data, we present a computational framework to model these experiments by computing the X-ray scattering intensity directly from the atomic positions obtained from molecular dynamics simulations. We compare the efficiency and accuracy of two alternative computational methods: the direct method computing the intensity at each diffraction vector separately, and a method based on fast Fourier transform that computes the intensities at all diffraction vectors at once. Furthermore, the computed X-ray speckle patterns capture the density fluctuations over a range of length and time scales and are shown to reproduce the known properties and relations of experimental XPCS and XSVS for liquids.
Abstract The field of low-temperature plasmas (LTPs) excels by virtue of its broad intellectual diversity, interdisciplinarity and range of applications. This great diversity also challenges researchers in communicating the outcomes of their investigations, as common practices and expectations for reporting vary widely in the many disciplines that either fall under the LTP umbrella or interact closely with LTP topics. These challenges encompass comparing measurements made in different laboratories, exchanging and sharing computer models, enabling reproducibility in experiments and computations using traceable and transparent methods and data, establishing metrics for reliability, and in translating fundamental findings to practice. In this paper, we address these challenges from the perspective of LTP standards for measurements, diagnostics, computations, reporting and plasma sources. This discussion on standards, or recommended best practices, and in some cases suggestions for standards or best practices, has the goal of improving communication, reproducibility and transparency within the LTP field and fields allied with LTPs. This discussion also acknowledges that standards and best practices, either recommended or at some point enforced, are ultimately a matter of judgment. These standards and recommended practices should not limit innovation nor prevent research breakthroughs from having real-time impact. Ultimately, the goal of our research community is to advance the entire LTP field and the many applications it touches through a shared set of expectations.
Computer models play a key role in many scientific and engineering problems. One major source of uncertainty in computer model experiments is input parameter uncertainty. Computer model calibration is a formal statistical procedure to infer input parameters by combining information from model runs and observational data. The existing standard calibration framework suffers from inferential issues when the model output and observational data are high-dimensional dependent data, such as large time series, due to the difficulty in building an emulator and the nonidentifiability between effects from input parameters and data-model discrepancy. To overcome these challenges, we, in this work, propose a new calibration framework based on a deep neural network (DNN) with long short-term memory layers that directly emulates the inverse relationship between the model output and input parameters. Adopting the “learning with noise” idea, we train our DNN model to filter out the effects from data-model discrepancy on input parameter inference. We also formulate a new way to construct interval predictions for DNN using quantile regression to quantify the uncertainty in input parameter estimates. Through a simulation study and real data application with the Weather Research and Forecasting Model Hydrological modeling system (WRF-Hydro), we show our approach can yield accurate point estimates and well-calibrated interval estimates for input parameters.
Remote sensing relative to applications involving vegetation cover and land use is reviewed to consider the potential benefits to the Earth Observing System (Eos) of a proposed Information Sciences Experiment System (ISES). The ISES concept has been proposed as an onboard experiment and computational resource to support advanced experiments and demonstrations in the information and earth sciences. Embedded in the concept is potential for relieving the data glut problem, enhancing capabilities to meet real-time needs of data users and in-situ researchers, and introducing emerging technology to Eos as the technology matures. These potential benefits are examined in the context of state-of-the-art research activities in image/data processing and management.
A preconditioning scheme has been implemented into a three-dimensional viscous computational fluid dynamics code for turbomachine blade rows. The preconditioning allows the code, originally developed for simulating compressible flow fields, to be applied to nearly-incompressible, low Mach number flows. A brief description is given of the compressible Navier-Stokes equations for a rotating coordinate system, along with the preconditioning method employed. Details about the conservative formulation of artificial dissipation are provided, and different artificial dissipation schemes are discussed and compared. The preconditioned code was applied to a well-documented case involving the NASA large low-speed centrifugal compressor for which detailed experimental data are available for comparison. Performance and flow field data are compared for the near-design operating point of the compressor, with generally good agreement between computation and experiment. Further, significant differences between computational results for the different numerical implementations, revealing different levels of solution accuracy, are discussed.
A direct algorithm is presented for computing the Euler load of a column from experimental data. The method is based on exact inextensional theory for imperfect columns, which predicts two distinct deflected shapes at loads near the Euler load. The bending stiffness of the column appears in the expression for the Euler load along with the column length, therefore the experimental data allows a direct computation of bending stiffness. Experiments on graphite-epoxy columns of rectangular cross-section are reported in the paper. The bending stiffness of each composite column computed from experiment is compared with predictions from laminated plate theory.
Limitations concerning the possibility to simulate all the significant flow and thermal phenomena occurring during the entry of a space vehicle into a planetary atmosphere make it necessary to rely on computational analyses to obtain the required data for the design of the spacecraft needed for the NASA missions planned for the next two decades. 'Benchmark' computer programs concerned with complete, detailed, and accurate computational solutions of entry problems are considered along with programs representing engineering approximations for cases in which the accuracy provided by the benchmark programs is not needed. The information obtainable by computational analysis has to be supplemented by actual flight experience in order to meet the goals of the NASA entry-technology program. The individual space missions planned for the coming years are examined together with the possibilities for obtaining the data needed to satisfy the entry requirements in each case.
We report that linear systems are occurring in many applications, and solving them can take a large amount of the total simulation time. The high performance library hypre provides a variety of interfaces and linear solvers, including various multigrid methods, that have achieved good scalability on a variety of homogeneous parallel computer architectures. Heterogeneous architectures with nodes that have both CPUs and accelerators provide new challenges, since they require more fine-grained parallelism and reduced data movement between different memories on a single node as well as across nodes. We will discuss our experiences and strategies to port hypre to heterogeneous computers with accelerators, including the design of a new memory model, the use of abstractions, the BoxLoop macros in the structured and semi-structured interfaces, and the restructuring of algebraic multigrid (AMG) into modular components. We present numerical experiments comparing CPU and GPU performance for several test problems.