Search NASA⌕ Search

SEARCH · Search NASA

Results for “Simulation, Modeling, Performance, and Visualization”

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.

At least 397 records · Page 22

Bridging the Gap between Cosmological Simulations with Graph Neural Networks and Domain Adaptation

Deep learning models have been shown to outperform methods that rely on summary statistics, like the power spectrum, in extracting information from complex cosmological data sets. However, due to differences in the subgrid physics implementation and numerical approximations across different simulation suites, models trained on data from one cosmological simulation show a drop in performance when tested on another. Similarly, models trained on any of the simulations would also likely experience a drop in performance when applied to observational data. Training on data from two different suites of the CAMELS hydrodynamic cosmological simulations, we examine the generalization capabilities of Domain Adaptive Graph Neural Networks (DA-GNNs). By utilizing GNNs, we capitalize on their capacity to capture structured scale-free cosmological information from galaxy distributions. Moreover, by including unsupervised domain adaptation via Maximum Mean Discrepancy (MMD), we enable our models to extract domain-invariant features. We demonstrate that DA-GNN achieves higher accuracy and robustness on cross dataset tasks (up to 28% better relative error and up to almost an order of magnitude better χ 2 ). Using data visualizations, we show the effects of domain adaptation on proper latent space data alignment. This shows that DA-GNNs are a promising method for extracting domain-independent cosmological information, a vital step toward robust deep learning for real cosmic survey data.

97 MATHEMATICS AND COMPUTING↗

Estimating the Single-Trial Characteristics of Event-Related Responses: Evaluation of the MCERP Algorithm

Single-trial event-related responses collected during the course of an experiment are typically averaged before analysis resulting in a rather crude picture of event-related brain dynamics. It has been quite clear for some time that these responses exhibit trial-to-trial variability: however, the computational techniques necessary to deal with such responses in noisy conditions have not been available. To this end we have developed the multiple-component, event-related potential model (mcERP), which assumes that the each event-related response consists of a sum of multiple evoked components each described by a stereotypical waveshape. These waveshapes are allowed to vary in amplitude and onset latency from trial to trial, which allows us to capture, to first-order, the trial-dependent variations in event-related brain dynamics. We have constructed many sets of synthetic data designed to simulate intracortical recordings from a 15 channel, linear-array multielectrode implanted acutely in V1 of an awake-behaving macaque undergoing visual stimulation with a red light flash. This synthetic data was used to characterize the performance of the mcERP algorithm. First we quantified the degree to which such trial-to-trial variability aids in the identification of multiple components, and we demonstrate that amplitude variability is a more important factor in component separation than latency variability. Second, we quantified the behavior of the algorithm under two distinct signal-to-noise ratio (SNR) conditions: Gaussian noise independently present in each channel, and highly correlated (1/f distributed), far-field noise presented identically in each channel of the array. The mcERP algorithm was found to be robust to noise accurately identifying all component waveshapes and their associated single-trial characteristics down to SNR levels of -20dB for Gaussian noise and -7dB for 1/f far-field noise. Comparisons of the performance of this algorithm with factor analysis (FA) and independent component analysis (ICA) will be described by Knuth et al. (SFN abstracts, 2002). In addition, the advantages of application of mcERP to real data will be described by Shah et al, (these abstracts, 2002: SFN abstracts, 2002).

Knuth, K. H.↗

Human Stability While Exercising on a VIS Device in Zero Gravity

BACKGROUND: A Vibration Isolation and Stabilization (VIS) system is being designed for use with the European Enhanced Exploration Exercise Device (E4D) [1]. The exercise bar is attached to cables that extend and retract through openings in the E4D platform, which the subject stands on while exercising. Since the E4D is mounted on a moving VIS device, the whole system moves under the subject’s feet. While tension in the cables does help stabilize a crew member in 0g by bracing him or her against the moving platform, flexible cables do not offer full support against falling. This raises the question of whether various exercises that are successfully performed in 1g on a stationary device can be performed without losing balance in 0g on a moving device. Quantifying stability conditions and establishing stability requirements for exercise countermeasure systems is a long-standing challenge and this work helped to inform this specific need for integrated E4D/VIS flight project development. METHODS AND RESULTS: A simplified stability analysis can be attempted based on platform accelerations generated by the VIS device simulation. In this type of analysis, the subject is conceived as standing on the platform in 0g while being held down to it by the tension in the cables. If the sideways acceleration of the platform is such that the cables cannot generate a sufficient moment relative to, e.g., the heels or the toes of the subject to overcome the tipping moment from the inertial forces on the accelerated subject’s body, the subject becomes unstable. This approach, which we refer to as the ‘static’ approximation, indicated failure of all the exercises that were considered in the comprehensive VIS analysis of the E4D. It was realized, however, that platform accelerations are not externally imposed but are themselves induced by the motion of the subject’s body during exercise, and a model calculation confirmed that this drastically altered the tipping moment on the subject, with the potential to even switch the direction in which the body would tip over. This made it imperative to consider in a coupled manner both the motion of the exercising subject’s body and the induced VIS platform motion while considering stability. The coupled dynamics requirement was met by the VIS simulation incorporating time-dependent human mass properties atop the platform and driven by the inertial forces from the prescribed subject motion relative to the platform based on motion-capture recorded human trajectories in 1g on a stationary E4D [2]. To analyze simulation results, a stability criterion was also needed. We do not know how to account for the human ‘control system’ that would take visual and vestibular cues as inputs as the platform moves in 0g. We thus chose to follow an approach we had previously used to analyze the dynamic feasibility of performing a task in lunar gravity along the human body trajectory recorded in 1g [3]. In this approach, one analyzes the center of pressure (COP) between the shoes and the ground (or platform) and checks if the COP remains within the convex hull of the footprints on the ground (or platform), commonly referred to as base of support (BOS). Since pressure is vertical and, in the absence of foot restraints, positive, the COP going outside the BOS indicates that the trajectory recorded in 1g is not dynamically feasible in microgravity and/or on the moving platform of the device. The outcome of the COP-based analysis was that some of the key exercises, especially the deadlift and the back squat exercises, were found to be dynamically feasible for at least some number of the exercise cycles, this number increasing with the cable tension. While this type of stability analysis does not account for the likely alteration of the exercise trajectory from its 1g form in space, it does show that there exist at least some realistic trajectories that pass the dynamic feasibility criterion. This adds confidence that, with further adjustment by the subject of the exercise form in 0g on a moving device, a number of exercises can be performed without losing balance.

D Frenkel↗

Computational Reproducibility of Scientific Workflows at Extreme Scales

We propose an approach for improved reproducibility that includes capturing and relating provenance characteristics and performance metrics, in a hybrid queriable system, the ProvEn server. The system capabilities are illustrated on two use cases: scientific reproducibility of results in the ACME climate simulations and performance reproducibility in molecular dynamics workflows on HPC computing platforms.

provenance, Reproducibility, Climate & Earth Syste↗

Aerodynamic measurements on a finite wing with simulated ice

The effect of a simulated glaze ice accretion on the aerodynamic performance of a three-dimensional straight and swept wing is studied experimentally. A semispan wing of effective aspect ratio five was mounted from the sidewall of the UIUC subsonic wind tunnel. The model uses an NACA 0012 airfoil section on a rectangular planform with interchangeable tip and root sections to allow for 0- and 30-deg sweep. A sidewall suction system is used to minimize the tunnel boundary-layer interaction with the model. A three-component sidewall balance has been designed, built and used to measure lift, drag and pitching moment on the clean and iced model. Fluorescent oil flow visualization has been performed on the iced model and reveals extensive spanwise flow in the separation bubble aft of the upper surface horn. These results are compared to computational results for the surface pressures, span loads and surface oil flow.

Bragg, M. B.↗

The Impact of Behind-the-Meter Heterogeneous Distributed Energy Resources on Distribution Grids: Preprint

The increasing integration of distributed energy resources (DERs) on the electric grid brings new challenges and opportunities for utility grid operations. With the rapid deployment of DERs, there is emerging interest in integrating these controllable devices with utility operations at all levels for monitoring and management. Different types of DERs such as Photovoltaics (PV), energy storage, electric vehicles (EVs), etc. have varying effects on the grid based on how and where they are deployed. Again, the DER deployment and its impact are network-dependent, while the traditional electric grids were not designed to host DERs. To understand the challenges with increasing behind-the-meter (BTM) DERs and to identify the needs in deploying advanced controls, a comprehensive grid impact study is indispensable. This paper investigates the impacts of integrating a mix of DERs in the utility distribution network in Colorado, U.S. Firstly, BTM DERs at the residential scale including distributed PV; battery energy storage systems (BESS); heating, ventilation, and air-conditioning (HVAC) load; electric water heater (EWH) load; and EVs are modeled. The impacts of integrating these resources on the distribution network are evaluated by conducting time-series simulations for different scenarios considering different days to capture the worst-case conditions. Monte Carlo simulations are conducted to generate the realistic EV charging profile. The voltage issues, substation transformer loadings, and critical nodes in the network are identified with the incorporation of uncoordinated EV charging loads and other DER in the network. This analysis helps visualize the DER impact on the grid, identify the challenges, and provides an insight into the new distribution management and control needs to enable reliable and resilient distribution grid operations. Additionally, further analysis is performed to estimate the daily residential electricity cost with the inclusion of DERs under time-of-use tariffs. The result shows the daily residential electricity cost reduced by 20:7% on average with DERs compared to the case without DERs.

27 ARPA - Advanced Research Projects Agency-Energy↗

Domain Adaptive Graph Neural Networks for Constraining Cosmological Parameters Across Multiple Data Sets

Deep learning models have been shown to outperform methods that rely on summary statistics, like the power spectrum, in extracting information from complex cosmological data sets. However, due to differences in the subgrid physics implementation and numerical approximations across different simulation suites, models trained on data from one cosmological simulation show a drop in performance when tested on another. Similarly, models trained on any of the simulations would also likely experience a drop in performance when applied to observational data. Training on data from two different suites of the CAMELS hydrodynamic cosmological simulations, we examine the generalization capabilities of Domain Adaptive Graph Neural Networks (DA-GNNs). By utilizing GNNs, we capitalize on their capacity to capture structured scale-free cosmological information from galaxy distributions. Moreover, by including unsupervised domain adaptation via Maximum Mean Discrepancy (MMD), we enable our models to extract domain-invariant features. We demonstrate that DA-GNN achieves higher accuracy and robustness on cross-dataset tasks (up to $28\%$ better relative error and up to almost an order of magnitude better $\chi^2$). Using data visualizations, we show the effects of domain adaptation on proper latent space data alignment. This shows that DA-GNNs are a promising method for extracting domain-independent cosmological information, a vital step toward robust deep learning for real cosmic survey data.

79 ASTRONOMY AND ASTROPHYSICS↗

Frameworks, Algorithms, and Scalable Technologies for Mathematics (FASTMath) SciDAC Institute

As computational models scale to larger computers, the rate at which they produce data has far outstripped the same computers ability to write that data and further the file systems ability to store that data. Almost all of the SciDAC applications, but especially those related to fusion solve very large scale PDEs whose scientific output his impacted by this problem. To gain access to dynamics in an exascale simulation that are not identifiable a priori and to make that dynamical data available to machine learning requires fundamental research in the area of in situ data data analytics. Here data analytics includes compression, visualization, uncertainty quantification, and machine learning. This in situ data analytics will enable on-the-fly spatial and temporal compression of solution dynamics, expose that space-time compressed field to machine learning algorithms that have been specialized to work with dynamically evolving data (existing machine learning algorithms treat data sets as static), greatly improving the opportunity for machine learning to provide feedback to the compression, all within an ongoing simulation, without the need to write data to files. The same concepts are also being applied to uncertainty quantification and multi-fidelity modeling which have similar needs for spatial and temporal compression of the ongoing exascale simulation to perform either without the typical, unacceptable writing of data to files.

97 MATHEMATICS AND COMPUTING↗

CAPRI (Computational Analysis PRogramming Interface): A Solid Modeling Based Infra-Structure for Engineering Analysis and Design Simulations

CAPRI is a CAD-vendor neutral application programming interface designed for the construction of analysis and design systems. By allowing access to the geometry from within all modules (grid generators, solvers and post-processors) such tasks as meshing on the actual surfaces, node enrichment by solvers and defining which mesh faces are boundaries (for the solver and visualization system) become simpler. The overall reliance on file 'standards' is minimized. This 'Geometry Centric' approach makes multi-physics (multi-disciplinary) analysis codes much easier to build. By using the shared (coupled) surface as the foundation, CAPRI provides a single call to interpolate grid-node based data from the surface discretization in one volume to another. Finally, design systems are possible where the results can be brought back into the CAD system (and therefore manufactured) because all geometry construction and modification are performed using the CAD system's geometry kernel.

Haimes, Robert↗

Ames vision group research overview

A major goal of the reseach group is to develop mathematical and computational models of early human vision. These models are valuable in the prediction of human performance, in the design of visual coding schemes and displays, and in robotic vision. To date researchers have models of retinal sampling, spatial processing in visual cortex, contrast sensitivity, and motion processing. Based on their models of early human vision, researchers developed several schemes for efficient coding and compression of monochrome and color images. These are pyramid schemes that decompose the image into features that vary in location, size, orientation, and phase. To determine the perceptual fidelity of these codes, researchers developed novel human testing methods that have received considerable attention in the research community. Researchers constructed models of human visual motion processing based on physiological and psychophysical data, and have tested these models through simulation and human experiments. They also explored the application of these biological algorithms to applications in automated guidance of rotorcraft and autonomous landing of spacecraft. Researchers developed networks for inhomogeneous image sampling, for pyramid coding of images, for automatic geometrical correction of disordered samples, and for removal of motion artifacts from unstable cameras.

Watson, Andrew B.↗

Present status of the MIT/NASA Langley 6-inch MSBS

The latest improvements are described which were made to the MIT/NASA Langley 6 Inch Magnetic Suspension and Balance System and its operational capabilities. Since its relocation from MIT in 1984, the system has suffered from reliability problems with original equipment. System performance is ultimately limited by noise and antiquated power supplies. Although a limited amount of funding and man hours has been expended, some significant improvements were made to the system. The design and implementation of a new electro-magnetic position sensor has significantly reduced system noise and drift. The system was shown fully operational for the first time since its arrival at NASA Langley. An ogive cylinder test model was suspended while running the wind tunnel at its maximum of approx. Mach 0.5. Other aerodynamic models were suspended with wind on. Also, the simulation of low speed cavity store separation using an ogive cylinder test model is described. Methods used for flow visualization for this test are also discussed. The system is ideally suited for this type of low speed testing.

Schott, Timothy↗

Multiphysics modeling of accelerators through code integration

This work aims to improve the ability of particle accelerator researchers to develop high-performance accelerator cavity designs by creating an overall multiphysics framework that integrates and couples existing application codes. This framework will allow accelerator researchers to build multiphysics models that will optimize cavity design, improve understanding of whole-device performance, and reduce the development and fabrication costs of accelerator research. We utilize the open-source VizSchema data standard as an intermediate data structure interface layer to standardize interfaces between individual application codes. VizScema is extensively documented online, and plugins for VizSchema are available for popular visualization packages, including VisIt and ParaView. Currently, the work focuses on coupling the EM field solver COMSOL and the electron gun code MICHELLE to allow COMSOL field-solve results to be seamlessly used by MICHELLE for particle-solve. Later work will extend this integration to include other fields, particles, and thermodynamics simulation codes.

43 PARTICLE ACCELERATORS↗

Wake Vortex Advisory System (WakeVAS) Evaluation of Impacts on the National Airspace System

This report is one of a series that describes an ongoing effort in high-fidelity modeling/simulation, evaluation and analysis of the benefits and performance metrics of the Wake Vortex Advisory System (WakeVAS) Concept of Operations being developed as part of the Virtual Airspace Modeling and Simulation (VAMS) project. A previous study, determined the overall increases in runway arrival rates that could be achieved at 12 selected airports due to WakeVAS reduced aircraft spacing under Instrument Meteorological Conditions. This study builds on the previous work to evaluate the NAS wide impacts of equipping various numbers of airports with WakeVAS. A queuing network model of the National Airspace System, built by the Logistics Management Institute, Mclean, VA, for NASA (LMINET) was used to estimate the reduction in delay that could be achieved by using WakeVAS under non-visual meteorological conditions for the projected air traffic demand in 2010. The results from LMINET were used to estimate the total annual delay reduction that could be achieved and from this, an estimate of the air carrier variable operating cost saving was made.

Smith, Jeremy C.↗

How close are urban scale building simulations to measured data? Examining bias derived from building metadata in urban building energy modeling

Residential and commercial buildings in the United States accounted for 40% of total energy in 2020. Building energy modeling (BEM) is a useful tool that allows individuals, researchers, companies, or utilities to save energy by optimizing buildings through estimation of building technology savings and performance projection of building energy under various environmental conditions. Urban building energy modeling (UBEM) expands the scope beyond individual buildings to the buildings in a neighborhood, city, utility and more. Yet there is a knowledge gap in the literature as to how these models compare to measured data on an individual and aggregated basis. As UBEM data and methods continue to develop, it is important to consider the accuracy, bias, and limitations of the models. Here, nation-scale data and UBEM software suite named Automatic Building Energy Modeling (AutoBEM) was used to model 50,843 buildings in Chattanooga, Tennessee. The uncalibrated simulation results were compared to aggregated 15-minute electricity data for the year 2019 with visualizations highlighting sources of bias in building data and the AutoBEM framework while considering how they relate to other UBEM methods. Estimation of building type and year of constructions are found to be the major sources of bias. Accounting for the amount of conditioned area per building significantly improves the overall fit of the simulated energy use intensity. it was found that inherent variation in building energy use contributes to R 2 values between 0.008 and 0.095 across building types but slope values near 1 for the total number of buildings. This indicates the need for building aggregation for representative building energy modeling with data sources available at an urban scale while illustrating the need for additional individual building data and model improvement beyond the originally produced UBEM models for individual building analysis.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Analysis of a flare-director concept for an externally blown flap STOL aircraft

A flare-director concept involving a thrust-required flare-guidance equation was developed and tested on a moving-base simulator. The equation gives a signal to command thrust as a linear function of the errors between the variables thrust, altitude, and altitude rate and corresponding values on a desired reference flare trajectory. During the simulator landing tests this signal drove either the horizontal command bar of the aircraft's flight director or a thrust-command dot on a head-up virtual-image display of a flare director. It was also used as the input to a simple autoflare system. An externally blown flap STOL (short take-off and landing) aircraft (with considerable stability and control augmentation) was modeled for the landing tests. The pilots considered the flare director a valuable guide for executing a proper flare-thrust program under instrument-landing conditions, but were reluctant to make any use of the head-up display when they were performing the landings visually.

Middleton, D. B.↗

Deep material network via a quilting strategy: visualization for explainability and recursive training for improved accuracy

Recent developments integrating micromechanics and neural networks offer promising paths for rapid predictions of the response of heterogeneous materials with similar accuracy as direct numerical simulations. The deep material network is one such approaches, featuring a multi-layer network and micromechanics building blocks trained on anisotropic linear elastic properties. Once trained, the network acts as a reduced-order model, which can extrapolate the material’s behavior to more general constitutive laws, including nonlinear behaviors, without the need to be retrained. However, current training methods initialize network parameters randomly, incurring inevitable training and calibration errors. Here, we introduce a way to visualize the network parameters as an analogous unit cell and use this visualization to “quilt” patches of shallower networks to initialize deeper networks for a recursive training strategy. The result is an improvement in the accuracy and calibration performance of the network and an intuitive visual representation of the network for better explainability.

97 MATHEMATICS AND COMPUTING↗

Experimental and Computational Investigation of a Translating-Throat Single-Expansion-Ramp Nozzle

An experimental and computational study was conducted on a high-speed, single-expansion-ramp nozzle (SERN) concept designed for efficient off-design performance. The translating-throat SERN concept adjusts the axial location of the throat to provide a variable expansion ratio and allow a more optimum jet exhaust expansion at various flight conditions in an effort to maximize nozzle performance. Three design points (throat locations) were investigated to simulate the operation of this concept at subsonic-transonic, low supersonic, and high supersonic flight conditions. The experimental study was conducted in the jet exit test facility at the Langley Research Center. Internal nozzle performance was obtained at nozzle pressure ratios (NPR's) up to 13 for six nozzles with design nozzle pressure ratios near 9, 42, and 102. Two expansion-ramp surfaces, one concave and one convex, were tested for each design point. Paint-oil flow and focusing schlieren flow visualization techniques were utilized to acquire additional flow data at selected NPR'S. The Navier-Stokes code, PAB3D, was used with a two-equation k-e turbulence model for the computational study. Nozzle performance characteristics were predicted at nozzle pressure ratios of 5, 9, and 13 for the concave ramp, low Mach number nozzle and at 10, 13, and 102 for the concave ramp, high Mach number nozzle.

Deere, Karen A.↗

Generic Environment for Simulating Launch Operations

GEM-FLO (A Generic Simulation Environment for Modeling Future Launch Operations) is a computer program that facilitates creation of discrete-event simulation models of ground processes in which reusable or expendable launch vehicles (RLVs) are prepared for flight. GEM-FLO includes a component, developed in Visual Basic, that generates a graphical user interface (GUI) and a component, developed in the Arena simulation language, that creates a generic discrete-event simulation model. Through the GUI, GEM-FLO elicits RLV design information from the user. The design information can include information on flight hardware elements, resources, and ground processes. GEM-FLO translates the user s responses into mathematical variables and expressions that populate the generic simulation model. The variables and expressions can represent processing times, resource capacities, status variables, and other process parameters needed to configure a simulation model that reflects the ground processing flow and requirements of a specific RLV. Upon execution of the model, GEMFLO puts out data on many measures of performance, including the flight rate, turnaround time, and utilization of resources. This information can serve as the basis for determining whether design goals can be met, and for comparing characteristics of competing RLV designs

Steele, Martin↗