Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “reasonable accuracy”

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 721 records · Page 40
All sourcesNASA NTRSDOE OSTI
NASA NTRS2002

Higher Order Time Integration Schemes for the Unsteady Navier-Stokes Equations on Unstructured Meshes

The rapid increase in available computational power over the last decade has enabled higher resolution flow simulations and more widespread use of unstructured grid methods for complex geometries. While much of this effort has been focused on steady-state calculations in the aerodynamics community, the need to accurately predict off-design conditions, which may involve substantial amounts of flow separation, points to the need to efficiently simulate unsteady flow fields. Accurate unsteady flow simulations can easily require several orders of magnitude more computational effort than a corresponding steady-state simulation. For this reason, techniques for improving the efficiency of unsteady flow simulations are required in order to make such calculations feasible in the foreseeable future. The purpose of this work is to investigate possible reductions in computer time due to the choice of an efficient time-integration scheme from a series of schemes differing in the order of time-accuracy, and by the use of more efficient techniques to solve the nonlinear equations which arise while using implicit time-integration schemes. This investigation is carried out in the context of a two-dimensional unstructured mesh laminar Navier-Stokes solver.

Jothiprasad, Giridhar↗
NASA NTRS2009

Measurement of Satellite Impact Test Fragments for Modeling Orbital Debris

There are over 13,000 pieces of catalogued objects 10cm and larger in orbit around Earth [ODQN, January 2009, p12]. More than 6000 of these objects are fragments from explosions and collisions. As the earth-orbiting object count increases, debris-generating collisions in the future become a statistical inevitability. To aid in understanding this collision risk, the NASA Orbital Debris Program Office has developed computer models that calculate quantity and orbits of debris both currently in orbit and in future epochs. In order to create a reasonable computer model of the orbital debris environment, it is important to understand the mechanics of creation of debris as a result of a collision. The measurement of the physical characteristics of debris resulting from ground-based, hypervelocity impact testing aids in understanding the sizes and shapes of debris produced from potential impacts in orbit. To advance the accuracy of fragment shape/size determination, the NASA Orbital Debris Program Office recently implemented a computerized measurement system. The goal of this system is to improve knowledge and understanding of the relation between commonly used dimensions and overall shape. The technique developed involves scanning a single fragment with a hand-held laser device, measuring its size properties using a sophisticated software tool, and creating a three-dimensional computer model to demonstrate how the object might appear in orbit. This information is used to aid optical techniques in shape determination. This more automated and repeatable method provides higher accuracy in the size and shape determination of debris.

Hill, Nicole M.↗
NASA NTRS2017

Tidal Prediction

Tides are the most predictable of oceanographic phenomena, due both to the simplicity and predictability of the astronomical forcing and to the near linearity of the ocean's dynamical response. In the classical and simplest scenario, tidal prediction is based on harmonic analysis of past measurements at a fixed location. Limits to predictability arise because isolated astronomical spectral lines are broadened into "cusps" of incoherent energy, for example through interactions with non-tidal flows. Tidal prediction at locations without past measurements has historically been a major challenge, but, owing to near-global observations of modern satellite altimeters, the empirical harmonic approach now yields reasonably accurate predictions throughout most of the open ocean. Advances in numerical modeling and data assimilation allow these predictions to be refined (especially in shallow seas where observations remain insufficient to directly constrain tidal wave structure) and extended to include tidal currents. We review recent progress in the development of global and regional-scale tidal prediction capabilities, summarize accuracy of available charts, and briefly consider outstanding issues. Satellite altimetry has also helped unravel the global tidal energy budget and has clarified the role of internal tides as a sink of tidal energy. We summarize these results, and then turn to the challenging problem of predicting internal tides. To the extent that low modes remain coherent with the surface tide, elevations can be directly mapped. We review current efforts in this direction, which are already producing charts with predictive capability. However, internal tides (especially higher modes) interact much more strongly with lower-frequency ocean flows, so a significant fraction of this tidal signal is intermittent and incoherent. We close with a brief review of ongoing efforts to model global tides in combination with wind-forced ocean motions.

Harmonic Constants↗
NASA NTRS1973

Large amplitude flutter of a low aspect ratio panel at low supersonic speeds comparison of theory and experiment

Flutter boundaries, as well as flutter limit cycle amplitudes, frequencies and stresses were computed for a panel of length-width ratio 4.48 exposed to applied in-plane and transverse loads. The Mach number range was 1.1 to 1.4. The method used involved direct numerical integration of modal equations of motion derived from the nonlinear plate equations of von Karman, coupled with linearized potential flow aerodynamic theory. The flutter boundaries agreed reasonably well with experiment, except when the in-plane loading approached the buckling load. Structural damping had to be introduced, to produce frequencies comparable to the experimental values. Attempts to compute panel deflections or stress at a given point met with limited success. There is some evidence, however, that deflection and stress maxima can be estimated with somewhat greater accuracy.

Ventres, C. S.↗
NASA NTRS1989

Floating point arithmetic in future supercomputers

Considerations in the floating-point design of a supercomputer are discussed. Particular attention is given to word size, hardware support for extended precision, format, and accuracy characteristics. These issues are discussed from the perspective of the Numerical Aerodynamic Simulation Systems Division at NASA Ames. The features believed to be most important for a future supercomputer floating-point design include: (1) a 64-bit IEEE floating-point format with 11 exponent bits, 52 mantissa bits, and one sign bit and (2) hardware support for reasonably fast double-precision arithmetic.

Bailey, David H.↗
NASA NTRS1993

Testing the frozen flow approximation

We investigate the accuracy of the frozen-flow approximation (FFA), recently proposed by Matarrese, et al. (1992), for following the nonlinear evolution of cosmological density fluctuations under gravitational instability. We compare a number of statistics between results of the FFA and n-body simulations, including those used by Melott, Pellman & Shandarin (1993) to test the Zel'dovich approximation. The FFA performs reasonably well in a statistical sense, e.g. in reproducing the counts-in-cell distribution, at small scales, but it does poorly in the crosscorrelation with n-body which means it is generally not moving mass to the right place, especially in models with high small-scale power.

Lucchin, Francesco↗
NASA NTRS2009

Simulation of an Isolated Tiltrotor in Hover with an Unstructured Overset-Grid RANS Solver

An unstructured overset-grid Reynolds Averaged Navier-Stokes (RANS) solver, FUN3D, is used to simulate an isolated tiltrotor in hover. An overview of the computational method is presented as well as the details of the overset-grid systems. Steady-state computations within a noninertial reference frame define the performance trends of the rotor across a range of the experimental collective settings. Results are presented to show the effects of off-body grid refinement and blade grid refinement. The computed performance and blade loading trends show good agreement with experimental results and previously published structured overset-grid computations. Off-body flow features indicate a significant improvement in the resolution of the first perpendicular blade vortex interaction with background grid refinement across the collective range. Considering experimental data uncertainty and effects of transition, the prediction of figure of merit on the baseline and refined grid is reasonable at the higher collective range- within 3 percent of the measured values. At the lower collective settings, the computed figure of merit is approximately 6 percent lower than the experimental data. A comparison of steady and unsteady results show that with temporal refinement, the dynamic results closely match the steady-state noninertial results which gives confidence in the accuracy of the dynamic overset-grid approach.

Lee-Rausch, Elizabeth M.↗
NASA NTRSDate not supplied

Runway Configuration Management with Offline Reinforcement Learning

Runway configuration management (RCM) is a challenging task, and it affects the efficiency of the National Airspace System (NAS) and airport surface operations significantly. Each airport, depending on the geometry, capacity, local climate patterns, etc. has multiple configurations for the runway usage for arriving and departing flights. Many factors such as the incoming/outgoing traffic load, wind direction and speed, convective weather, cloud ceiling and other environmental factors might affect the choice of a runway configuration at any point in time. However, other factors such as safety measures and regulations, noise abatement, capacity of each configuration, and preference of the air traffic controllers (ATCs) can also play a significant role in selecting the configuration. A sub-optimal selection of the runway configuration, or delay in making configuration changes might result in significant increase in taxi times for aircraft on the surface of the airport, fuel and energy use of the aircraft, and maintenance costs. It can also lead to safety concerns, such as an aircraft performing one or more go-arounds before being able to land. All these factors make RCM an extremely important and challenging decision-making process for the ATCs. The current state of practice sets the runway configuration by the ATCs based on relevant information available at the time including weather, traffic, noise abatement, safety bounds, etc. This makes the decision-making process subjective based on the accuracy of the available information and the bias in human decision making. Unfortunately, this approach yields poor results (e.g., significant delays) if the predicted outcomes are uncertain and their relative impact is not well understood. This is especially evident when the uncertainty increases the size of possible predicted outcomes (combinatorial explosion in possible scenarios) that cannot be handled by human reasoning. On the other hand, an automated approach based on machine intelligence can make use of historical data and search through all (or significant amount of) possible scenarios under uncertainty and make well-informed decisions.

Milad Memarzadeh↗
DOE OSTI2020

FPDetect: Efficient Reasoning About Stencil Programs Using Selective Direct Evaluation

We present FPDetect, a low-overhead approach for detecting logical errors and soft errors affecting stencil computations without generating false positives. We develop an offline analysis that tightly estimates the number of floating-point bits preserved across stencil applications. This estimate rigorously bounds the values expected in the data space of the computation. Violations of this bound can be attributed with certainty to errors. FPDetect helps synthesize error detectors customized for user-specified levels of accuracy and coverage. FPDetect also enables overhead reduction techniques based on deploying these detectors coarsely in space and time. Experimental evaluations demonstrate the practicality of our approach.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2006

Estimating Temperature Retrieval Accuracy Associated With Thermal Band Spatial Resolution Requirements for Center Pivot Irrigation Monitoring and Management

This study explores the use of synthetic thermal center pivot irrigation scenes to estimate temperature retrieval accuracy for thermal remote sensed data, such as data acquired from current and proposed Landsat-like thermal systems. Center pivot irrigation is a common practice in the western United States and in other parts of the world where water resources are scarce. Wide-area ET (evapotranspiration) estimates and reliable water management decisions depend on accurate temperature information retrieval from remotely sensed data. Spatial resolution, sensor noise, and the temperature step between a field and its surrounding area impose limits on the ability to retrieve temperature information. Spatial resolution is an interrelationship between GSD (ground sample distance) and a measure of image sharpness, such as edge response or edge slope. Edge response and edge slope are intuitive, and direct measures of spatial resolution are easier to visualize and estimate than the more common Modulation Transfer Function or Point Spread Function. For these reasons, recent data specifications, such as those for the LDCM (Landsat Data Continuity Mission), have used GSD and edge response to specify spatial resolution. For this study, we have defined a 400-800 m diameter center pivot irrigation area with a large 25 K temperature step associated with a 300 K well-watered field surrounded by an infinite 325 K dry area. In this context, we defined the benchmark problem as an easily modeled, highly common stressing case. By parametrically varying GSD (30-240 m) and edge slope, we determined the number of pixels and field area fraction that meet a given temperature accuracy estimate for 400-m, 600-m, and 800-m diameter field sizes. Results of this project will help assess the utility of proposed specifications for the LDCM and other future thermal remote sensing missions and for water resource management.

Ryan, Robert E.↗
DOE OSTI2021

Particle image velocimetry analysis with simultaneous uncertainty quantification using Bayesian neural networks

Particle image velocimetry (PIV) is an effective tool in experimental fluid mechanics for extracting flow fields from images. Recently, convolutional neural networks (CNNs) have been used to perform PIV analysis with accuracy on par with classical methods. Here we extend the use of CNNs to analyze PIV data while providing simultaneous uncertainty quantification on the inferred flow field. The method we apply in this paper is a Bayesian convolutional neural network (BCNN) which learns distributions of the CNN weights through variational Bayes. In order to demonstrate the utility of BCNNs for the PIV task, we compare the performance of three distinct BCNN models with simple architectures. The first network estimates flow velocity from image interrogation regions only. Our second model learns to infer velocity from both the image interrogation regions and interrogation region cross-correlation maps. Finally, our best performing network infers velocities from interrogation region cross-correlation maps only. We find that BCNNs using interrogation region cross-correlation maps as inputs perform better than those using interrogation windows only as inputs and discuss reasons why this may be the case. Additionally, we test the best performing BCNN on a full synthetic test image pair and a real image pair from the 1st International PIV Challenge. We show that ~98% of true particle displacements from the full synthetic image pair can be captured within the BCNN's 95% confidence intervals, and that the BCNN's performance on the real image pair is quantitatively similar to that of algorithms tested in the 1st International PIV Challenge. Finally, we show that BCNNs can be generalized to be used with multi-pass PIV algorithms with a moderate loss in accuracy, which may be overcome by future work on finetuning and training schemes. So to our knowledge, this is the first use of Bayesian neural networks to perform PIV.

47 OTHER INSTRUMENTATION↗
NASA NTRS1985

Blended SST fields combined SMMR/ship data (summary of technique)

All remote sensors have bias problems which vary in space and time. The infrared measurements are very sparse in cloudy areas: the SMMR has coverage limitations due to the interference of land areas and only half time operation. If one simply averaged the data from these various sources, the combination of the biases and varying sampling would introduce artifacts into the analyzed product which would reflect the sampling and which could be quite misleading. The remote sensors provide reasonable estimates of gradients of SST over length scales greater than the sensor resolution but smaller than hemispheric. Ships can be considered unbiased but are poorly distributed on small scales except in the densest shipping lanes. An analysis scheme based on a suggestion by Holl was developed which exploits just this approximately complementarity to provide an improved SST analysis. The resulting accuracy appears to be in the required approx. 0.5 C range for the particular case of using SMMR and ship data to produce two degree latitude by two degrees longitude monthly analyses.

Wilheit, T. T.↗
NASA NTRS1996

Verification of Orthogrid Finite Element Modeling Techniques

The stress analysis of orthogrid structures, specifically with I-beam sections, is regularly performed using finite elements. Various modeling techniques are often used to simplify the modeling process but still adequately capture the actual hardware behavior. The accuracy of such 'Oshort cutso' is sometimes in question. This report compares three modeling techniques to actual test results from a loaded orthogrid panel. The finite element models include a beam, shell, and mixed beam and shell element model. Results show that the shell element model performs the best, but that the simpler beam and beam and shell element models provide reasonable to conservative results for a stress analysis. When deflection and stiffness is critical, it is important to capture the effect of the orthogrid nodes in the model.

Steeve, B. E.↗
NASA NTRS1997

Improving Discrete-Sensitivity-Based Approach for Practical Design Optimization

In developing the automated methodologies for simulation-based optimal shape designs, their accuracy, efficiency and practicality are the defining factors to their success. To that end, four recent improvements to the building blocks of such a methodology, intended for more practical design optimization, have been reported. First, in addition to a polynomial-based parameterization, a partial differential equation (PDE) based parameterization was shown to be a practical tool for a number of reasons. Second, an alternative has been incorporated to one of the tedious phases of developing such a methodology, namely, the automatic differentiation of the computer code for the flow analysis in order to generate the sensitivities. Third, by extending the methodology for the thin-layer Navier-Stokes (TLNS) based flow simulations, the more accurate flow physics was made available. However, the computer storage requirement for a shape optimization of a practical configuration with the -fidelity simulations (TLNS and dense-grid based simulations), required substantial computational resources. Therefore, the final improvement reported herein responded to this point by including the alternating-direct-implicit (ADI) based system solver as an alternative to the preconditioned biconjugate (PbCG) and other direct solvers.

Baysal, Oktay↗
NASA NTRS2000

Desert Dust Layers Over Polluted Marine Boundary Layers: ACE-2 Measurements and ACE-Asia Plans

Aerosols in ACE-Asia are expected to have some commonalties with those in ACE-2, along with important differences. Among the commonalities are occurrences of desert dust layers over polluted marine boundary layers. Differences include the nature of the dust (yellowish in the East Asia desert outflow, vs. reddish-brown in the Sahara Outflow measured in ACE-2) and the composition of boundary-layer aerosols (e.g., more absorbing, soot and organic aerosol in-the Asian plume, caused by coal and biomass burning, with limited controls). In this paper we present ACE-2 measurements and analyses as a guide to our plans for ACE-2 Asia. The measurements include: (1) Vertical profiles of aerosol optical depth and extinction (380-1558 nm), and of water vapor column and concentration, from the surface through the elevated desert dust, measured by the 14-channel Ames Airborne Tracking Sunphotometer (AATS-14); (2) Comparisons of airborne and shipborne sunphotometer optical depths to satellite-retrieved values, with and without desert dust; (3) Comparisons between airborne Sunphotometer optical depth and extinction spectra and those derived from coincident airborne in situ measurements of aerosol size distribution, scattering and absorption; (4) Comparisons between size distributions measured in situ and retrieved from sunphotometer optical depth spectra; (5) Comparisons between aerosol single scattering albedo values obtained by several techniques, using various combinations of measurements of backscatter, extinction, size distribution, scattering, absorption, and radiative flux. We show how analyses of these data can be used to address questions important to ACE-Asia, such as: (1) How do dust and other absorbing aerosols affect the accuracy of satellite optical depth retrievals? How important are asphericity effects? (2) How important are supermicron dust and seasalt aerosols to overall aerosol optical depth and radiative forcing? How well are these aerosols sampled by aircraft inlets and instruments? (3) How consistent are suborbital in situ and remote measurements of aerosols, among themselves and with satellite retrievals? What are the main reasons for observed inconsistencies?

Russell, Philip B.↗
DOE OSTI2022

Neuromorphic Information Processing by Optical Media

Classification of features in a scene typically requires conversion of the incoming photonic field int the electronic domain. Recently, an alternative approach has emerged whereby passive structured materials can perform classification tasks by directly using free-space propagation and diffraction of light. In this manuscript, we present a theoretical and computational study of such systems and establish the basic features that govern their performance. We show that system architecture, material structure, and input light field are intertwined and need to be co-designed to maximize classification accuracy. Our simulations show that a single layer metasurface can achieve classification accuracy better than conventional linear classifiers, with an order of magnitude fewer diffractive features than previously reported. For a wavelength λ, single layer metasurfaces of size 100λ x 100λ with aperture density λ -2 achieve ~96% testing accuracy on the MNIST dataset, for an optimized distance ~100λ to the output plane. This is enabled by an intrinsic nonlinearity in photodetection, despite the use of linear optical metamaterials. Furthermore, we find that once the system is optimized, the number of diffractive features is the main determinant of classification performance. The slow asymptotic scaling with the number of apertures suggests a reason why such systems may benefit from multiple layer designs. Finally, we show a trade-off between the number of apertures and fabrication noise.

47 OTHER INSTRUMENTATION↗
DOE OSTI2021

Retrospective Analysis of Backwater Habitat Availability Using Remote Sensing

Low-velocity channel-margin habitats serve as important nursery habitats for the endangered Colorado pikeminnow (Ptychocheilus lucius) in the middle Green River between Jensen and Ouray, Utah. These habitats, known as backwaters, are associated with emergent sandbars, and are shaped and reformed annually by peak flows. Our recent knowledge about backwater characteristics and dynamics that is summarized in the synthesis report (Grippo et al. 2015) was based on detailed annual survey data from a relatively small sample of backwaters that were collected from 2003 to 2014 and reach-wide evaluations of backwater surface area based on manual interpretation of aerial and satellite imagery. Methods that bridge the gap between the detailed surveys from a small number of backwaters and the reach-wide assessment of their surface area would enable an assessment of the availability of backwater habitats that meet the minimum depth requirements for suitable habitat for Colorado pikeminnow. In 2015 Argonne National Laboratory (Argonne) tested three regression models—linear, multiple, and partial least square (PLS) regression models—for estimating backwater depth using National Agriculture Imagery Program (NAIP) imagery collected in July 2006 that covered the Jensen-Ouray reach of the Green River (Hamada and LaGory 2016). The results suggested that a PLS regression model showed high correlation with the reference depth (R 2 = 0.69) and had the most unbiased and consistent estimate of backwater depth. The results also indicated that the PLS model would provide reasonable estimates of the amount of habitat providing a minimum suitable depth of 30 cm for young-of-the-year Colorado pikeminnow, even though absolute depth estimates may be uncertain for backwater areas deeper than approximately 40 cm. The study also provided insights regarding the amount and the selection of calibration and validation (cal-val) data needed for improving the accuracy of depth prediction.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1992

Application of artificial neural networks to thermal detection of disbonds

A novel technique for processing thermal data is presented and applied to simulation as well as experimental data. Using a neural network of thermal data classification, good classification accuracies are obtained, and the resulting images exhibit very good contrast between bonded and disbonded locations. In order to minimize the preprocessing required before using the network of classification, the temperature values were directly employed to train a network using data from an on-site testing run of a commercial aircraft. Training was extremely fast, and the resulting classification also agreed reasonably well with an ultrasonic characterization of the panel. The results obtained using one sample show the partially disbonded vertical doubler. The vertical lines along the doubler correspond to the original extent of the doubler obtained using blueprints of the aircraft.

Prabhu, D. R.↗