Search NASA⌕ Search

SEARCH · Search NASA

Results for “design space approximation”

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 451 records · Page 25

Noncontact measurement of high temperature using optical fiber sensors

The primary goal of this research program was the investigation and application of noncontact temperature measurement techniques using optical techniques and optical fiber methods. In particular, a pyrometer utilizing an infrared optical light pipe and a multiwavelength filtering approach was designed, revised, and tested. This work was motivated by the need to measure the temperatures of small metallic pellets (approximately 3 mm diameter) in free fall at the Microgravity Materials Processing Drop Tube at NASA Marshall Space Flight Center. In addition, research under this program investigated the adaptation of holography technology to optical fiber sensors, and also examined the use of rare-earth dopants in optical fibers for use in measuring temperature. The pyrometer development effort involved both theoretical analysis and experimental tests. For the analysis, a mathematical model based on radiative transfer principles was derived. Key parameter values representative of the drop tube system, such as particle size, tube diameter and length, and particle temperature, were used to determine an estimate of the radiant flux that will be incident on the face of an optical fiber or light pipe used to collect radiation from the incandescent falling particle. An extension of this work examined the advantage of inclining or tilting the collecting fiber to increase the time that the falling particle remains in the fiber field-of-view. Those results indicate that increases in total power collected of about 15 percent may be realized by tilting the fiber. In order to determine the suitability of alternative light pipes and optical fibers, and experimental set-up for measuring the transmittance and insertion loss of infrared fibers considered for use in the pyrometer was assembled. A zirconium fluoride optical fiber and several bundles of hollow core fiber of varying diameters were tested. A prototype two-color pyrometer was assembled and tested at Virginia Tech, and then tested on the Drop Tube at Marshall Space Flight Center. Radiation from 5 mm diameter niobium drops falling in the Drop Tube was successfully detected, and recorded for later analysis. Subsequent analysis indicated that the imaging of light output from the light pipe onto the detector active areas was not identical for both detectors.

Claus, R. O.↗

Comparative Soot Diagnostics: Preliminary Results

The motivation for the Comparative Soot Diagnostics (CSD) experiment lies in the broad practical importance of understanding combustion generated particulate. Depending upon the circumstances, particulate matter can affect the durability and performance of combustion equipment, can be a pollutant, can be used to detect fires and, in the form of soot, can be the dominant source of radiant energy from flames. The nonbuoyant structure of most flames of practical interest makes understanding of soot processes in low gravity flames important to our ability to predict fire behavior on earth. These studies also have direct applications to fire safety in human-crew spacecraft, since smoke is the indicator used for automated detection in current spacecraft. In the earliest missions (Mercury, Gemini and Apollo), the crew quarters were so cramped that it was considered reasonable that the astronauts would rapidly detect any fire. The Skylab module, however, included approximately 20 UV-sensing fire detectors. The Space Shuttle has 9 particle-ionization smoke detectors in the mid-deck and flight deck and Spacelab has six additional particle-ionization smoke detectors. The designated detectors for the ISS are laser-diode, forward-scattering, smoke or particulate detectors. Current plans for the ISS call for two detectors in the open area of the module, and detectors in racks that have both cooling air flow and electrical power. Due to the complete absence of data concerning the nature of particulate and radiant emission from incipient and fully developed low-g fires, all three of these detector systems were designed based upon l-g test data and experience. As planned mission durations and complexity increase and the volume of spacecraft increases, the need for and importance of effective, crew-independent, fire detection grows significantly. To provide this level of protection, more knowledge is needed concerning low-gravity fire phenomena and, in particular, how they might be detected and suppressed. Prior to CSD, no combustion-generated particulate samples had been collected near the flame zone for well-developed microgravity flames. All of the extant data either came from drop tower tests and therefore only corresponded to the early stages of a fire or were collected far from the flame zone. The fuel sources in the drop tower tests were restricted to laminar gas-jet diffusion flames and very rapidly overheated wire insulation. The gas-jet tests indicated, through thermophoretic sampling, (2) that soot primaries and aggregates (groups of primary particles) in low-gravity may be significantly larger than those in normal gravity (1-g). This raises new scientific questions about soot processes as well as practical issues for particulate size sensitivity and detection alarm threshold levels used in on-orbit smoke detectors. Preliminary tests in the 2.2 second drop tower suggest that particulate generated by overheated wire insulation may be larger in low-g than in 1-g. Transmission Electron Microscope (TEM) grids downstream of the fire region in the Wire Insulation Flammability experiment as well as visual observation of long string-like aggregates, further confirm this suggestion. The combined impact of these limited results and theoretical predictions is that, as opposed to extrapolation from l-g data, direct knowledge of low-g combustion particulate is needed for more confident design of smoke detectors for spacecraft. This paper describes the operation and preliminary results of the CSD, a project conceived and developed at NASA Lewis Research Center. The CSD flight experiment was conducted in the Middeck Glovebox Facility (MGBX) on USMP-3. The project is support by NASA Headquarters Microgravity Science and Applications Division and Code Q. The results presented here are from the microgravity portion of the experiment, including the temporal response of the detectors and average sizes of the primary and aggregate particles captured on the thermophoretic probes.

Urban, David L.↗

Gain-Scheduled Approximations to H-Infinity Controllers for the F/A-18 Active Aeroelastic Wing

The Active Aeroelastic Wing will demonstrate technologies related to aeroservoelastic effects such as wing twist and load minimization. This paper presents a control design based on H-infinity synthesis that simultaneously considers the multiple objectives associated with handling qualities, actuator limitations, and loads. The controller is realized as a filter and gain set approximation to a state-space H-infinity controller. This approximation allows scheduling of the controller over a flight envelope.

Lind, Rick↗

COBRA System Engineering Processes to Achieve SLI Strategic Goals

The COBRA Prototype Main Engine Development Project was an endeavor conducted as a joint venture between Pratt & Whitney and Aerojet to conduct risk reduction in LOX/LH2 main engine technology for the NASA Space Launch Initiative (SLI). During the seventeen months of the project (April 2001 to September 2002), approximately seventy reviews were conducted, beginning with the Engine Systems Requirements Review (SRR) and ending with the Engine Systems Interim Design Review (IDR). This paper discusses some of the system engineering practices used to support the reviews and the overall engine development effort.

Ballard, Richard O.↗

Design of Neural Networks for Fast Convergence and Accuracy: Dynamics and Control

A procedure for the design and training of artificial neural networks, used for rapid and efficient controls and dynamics design and analysis for flexible space systems, has been developed. Artificial neural networks are employed, such that once properly trained, they provide a means of evaluating the impact of design changes rapidly. Specifically, two-layer feedforward neural networks are designed to approximate the functional relationship between the component/spacecraft design changes and measures of its performance or nonlinear dynamics of the system/components. A training algorithm, based on statistical sampling theory, is presented, which guarantees that the trained networks provide a designer-specified degree of accuracy in mapping the functional relationship. Within each iteration of this statistical-based algorithm, a sequential design algorithm is used for the design and training of the feedforward network to provide rapid convergence to the network goals. Here, at each sequence a new network is trained to minimize the error of previous network. The proposed method should work for applications wherein an arbitrary large source of training data can be generated. Two numerical examples are performed on a spacecraft application in order to demonstrate the feasibility of the proposed approach.

Maghami, Peiman G.↗

Approximate Pressure Distribution in an Accelerating Launch-Vehicle Fuel Tank

A detailed derivation of the equations governing the pressure in a generic liquid-fuel launch vehicle tank subjected to uniformly accelerated motion is presented. The equations obtained are then for the Space Shuttle Superlightweight Liquid-Oxygen Tank at approximately 70 seconds into flight. This generic derivation is applicable to any fuel tank in the form of a surface of revolution and should be useful in the design of future launch vehicles

Nemeth, Michael P.↗

Electrical Energy Storage Design Space Exploration for a Hybrid-Electric Six-Passenger Quadrotor

Hybrid-electric architectures are a promising means to achieve clean and efficient aircraft propulsion needed for small, short-range electric vertical takeoff and landing class vehicles. This paper explores the design space for a six-passenger quadrotor hybrid-electric propulsion system and shows that hybrid architectures that are more efficient than engine-only architectures can be built with near-term battery technology (approximately 150 Wh/kg at the pack level). Data obtained shows that hybrid-electric propulsion systems can achieve 37 percent block fuel reduction and 35 percent total energy cost savings over a 120 nautical mile mission compared to conventional turboshaft architectures. This work suggests that hybrids are likely the most efficient architectures that can be produced in the near term for this class of vehicle and motivates further development in this area.

parallel hybrid electric aircraft↗

A manipulator arm for zero-g simulations

A 12-ft counterbalanced Slave Manipulator Arm (SMA) was designed and fabricated to be used for resolving the questions of operational applications, capabilities, and limitations for such remote manned systems as the Payload Deployment and Retrieval Mechanism (PDRM) for the shuttle, the Free-Flying Teleoperator System, the Advanced Space Tug, and Planetary Rovers. As a developmental tool for the shuttle manipulator system (or PDRM), the SMA represents an approximate one-quarter scale working model for simulating and demonstrating payload handling, docking assistance, and satellite servicing. For the Free-Flying Teleoperator System and the Advanced Tug, the SMA provides a near full-scale developmental tool for satellite servicing, docking, and deployment/retrieval procedures, techniques, and support equipment requirements. For the Planetary Rovers, it provides an oversize developmental tool for sample handling and soil mechanics investigations. The design of the SMA was based on concepts developed for a 40-ft NASA technology arm to be used for zero-g shuttle manipulator simulations.

Brodie, S. B.↗

Advanced stationary and nonstationary kernel designs for domain-aware Gaussian processes

Gaussian process regression is a widely-applied method for function approximation and uncertainty quantification. The technique has gained popularity recently in the machine learning community due to its robustness and interpretability. The mathematical methods we discuss in this paper are an extension of the Gaussian-process framework. We are proposing advanced kernel designs that only allow for functions with certain desirable characteristics to be elements of the reproducing kernel Hilbert space (RKHS) that underlies all kernel methods and serves as the sample space for Gaussian process regression. These desirable characteristics reflect the underlying physics; two obvious examples are symmetry and periodicity constraints. In addition, non-stationary kernel designs can be defined in the same framework to yield flexible multi-task Gaussian processes. We will show the impact of advanced kernel designs on Gaussian processes using several synthetic and two scientific data sets. The results of our research show that including domain knowledge, communicated through advanced kernel designs, has a significant impact on the accuracy and relevance of the function approximation.

97 MATHEMATICS AND COMPUTING↗

Multi-element stochastic spectral projection for high quantile estimation

We investigate quantile estimation by multi-element generalized Polynomial Chaos (gPC) metamodel where the exact numerical model is approximated by complementary metamodels in overlapping domains that mimic the model’s exact response. The gPC metamodel is constructed by the non-intrusive stochastic spectral projection approach and function evaluation on the gPC metamodel can be considered as essentially free. Thus, large number of Monte Carlo samples from the metamodel can be used to estimate α-quantile, for moderate values of α. As the gPC metamodel is an expansion about the means of the inputs, its accuracy may worsen away from these mean values where the extreme events may occur. By increasing the approximation accuracy of the metamodel, we may eventually improve accuracy of quantile estimation but it is very expensive. A multi-element approach is therefore proposed by combining a global metamodel in the standard normal space with supplementary local metamodels constructed in bounded domains about the design points corresponding to the extreme events. To improve the accuracy and to minimize the sampling cost, sparse-tensor and anisotropic-tensor quadratures are tested in addition to the full-tensor Gauss quadrature in the construction of local metamodels; different bounds of the gPC expansion are also examined. The global and local metamodels are combined in the multi-element gPC (MEgPC) approach and it is shown that MEgPC can be more accurate than Monte Carlo or importance sampling methods for high quantile estimations for input dimensions roughly below N=8, a limit that is very much case- and α-dependent.

97 MATHEMATICS AND COMPUTING↗

Enabling probabilistic learning on manifolds through double diffusion maps

Here, we present a generative learning framework for probabilistic sampling that extends Probabilistic Learning on Manifolds (PLoM), which is designed to generate statistically consistent realizations of a random vector in a finite-dimensional Euclidean space, informed by a (representative) set of observations. In its original form, PLoM constructs a reduced-order probabilistic model by combining three main components: (a) kernel density estimation to approximate the underlying probability measure, (b) Diffusion Maps to characterize the manifold of the data, and (c) a reduced-order Itô Stochastic Differential Equation (ISDE) to sample from the learned distribution. However, its sampling dynamics are posed in the ambient space and the retained number of reduced coordinates is chosen by projection-reconstruction error. In practice, this often (i) requires more coordinates than the data’s intrinsic dimension to achieve stable sampling and (ii) lacks a smooth, basis-independent lifting back to the data domain; moreover, standard Diffusion Maps emphasize harmonic eigenfunctions and can miss non-harmonic latent structure. We address these limitations by decoupling geometry learning from sampling: a first Diffusion Maps pass identifies non-harmonic coordinates on which we formulate a full-order ISDE directly in the latent space, while Double Diffusion Maps captures multiscale geometric features and Geometric Harmonics (GH) learns a smooth lifting map to the ambient variables that is independent of the particular diffusion basis. This hybrid design preserves the system’s dynamical richness with a compact geometric representation and enables principled out-of-sample inference. The effectiveness and robustness of the proposed method are illustrated through two numerical studies: one based on data generated from two-dimensional Hermite polynomial functions and another based on high-fidelity simulations of a detonation wave in a reactive flow.

Double diffusion maps↗

An Efficient High-Order Solver for Diffusion Equations with Strong Anisotropy on Non-Anisotropy-Aligned Meshes

This paper concerns numerical solution of the diffusion equation with strong anisotropy on meshes not aligned with the anisotropic vector field. In order to resolve the numerical pollution for simulations on a non-anisotropy-aligned mesh and reduce the associated high computational cost we propose an effective preconditioner, extending our previous work. Similar to the anisotropy-aligned mesh case, we apply the auxiliary space preconditioning framework to design a preconditioner where a continuous finite element space is used as the auxiliary space for the discontinuous finite element space. The key component is an effective line smoother that can mitigate the high-frequency errors perpendicular to the magnetic field. We design a graph-based approach to find such a line smoother that is approximately perpendicular to the vector fields when the mesh does not align with the anisotropy. Finally, numerical experiments for several benchmark problems are presented, demonstrating the effectiveness and robustness of the proposed preconditioner when applied to Krylov iterative methods.

97 MATHEMATICS AND COMPUTING↗

Provably conservative approximations to complex reliability models

Complex models can be the bases for derivation of provably conservative and optimistic reliability models that incorporate a reduced state space and fewer transitions; they accordingly possess solutions that are more cost-effective than those of the original complex models. Design space can thereby be extensively explored without incurring the expense of multiple complex model solutions. A conservative-optimistic pair of derived models produces a band that includes the solution to the complex model. Sensitivity analysis can be performed on this pair of models to determine those parameters of the original model that are most sensitive to change and therefore require further expense in obtaining tighter specifications.

Smotherman, M.↗

Data-driven wind turbine wake modeling via probabilistic machine learning

Wind farm design primarily depends on the variability of the wind turbine wake flows to the atmospheric wind conditions and the interaction between wakes. Physics-based models that capture the wake flow field with high-fidelity are computationally very expensive to perform layout optimization of wind farms, and, thus, data-driven reduced-order models can represent an efficient alternative for simulating wind farms. In this work, we use real-world light detection and ranging (LiDAR) measurements of wind-turbine wakes to construct predictive surrogate models using machine learning. Specifically, we first demonstrate the use of deep autoencoders to find a low-dimensional latent space that gives a computationally tractable approximation of the wake LiDAR measurements. Then, we learn the mapping between the parameter space and the (latent space) wake flow fields using a deep neural network. Additionally, we also demonstrate the use of a probabilistic machine learning technique, namely, Gaussian process modeling, to learn the parameter-space-latent-space mapping in addition to the epistemic and aleatoric uncertainty in the data. Finally, to cope with training large datasets, we demonstrate the use of variational Gaussian process models that provide a tractable alternative to the conventional Gaussian process models for large datasets. Furthermore, we introduce the use of active learning to adaptively build and improve a conventional Gaussian process model predictive capability. Overall, we find that our approach provides accurate approximations of the wind-turbine wake flow field that can be queried at an orders-of-magnitude cheaper cost than those generated with high-fidelity physics-based simulations.

Deep neural networks↗

WaFIRS, a Waveguide Far-IR Spectrometer: Enabling Space-Borne Spectroscopy of High-z Galaxies in the Far-IR and Submm

The discovery of galaxies beyond z approximately equal to 1 which emit the bulk of their luminosity at long wavelengths has demonstrated the need for high sensitivity, broadband spectroscopy in the far-IR/submm/mm bands. Because many of these sources are not detectable in the optical, long wavelength spectroscopy is key to measuring their redshifts and ISM conditions. The continuum source list will increase in the next decade with new ground-based instruments (SCUBA2, Bolocam, MAMBO) and the surveys of HSO and SIRTF. Yet the planned spectroscopic capabilities lag behind, primarily due to the difficulty in scaling existing IR spectrograph designs to longer wavelengths. To overcome these limitations, we are developing WaFIRS, a novel concept for long-wavelength spectroscopy which utilizes a parallel-plate waveguide and a curved diffraction grating. WaFIRS provides the large (approximately 60%) instantaneous bandwidth and high throughput of a conventional grating system, but offers a dramatic reduction in volume and mass. WaFIRS requires no space overheads for extra optical elements beyond the diffraction grating itself, and is two-dimensional because the propagation is confined between two parallel plates. Thus several modules could be stacked to multiplex either spatially or in different frequency bands. The size and mass savings provide opportunities for spectroscopy from space-borne observatories which would be impractical with conventional spectrographs. With background-limited detectors and a cooled 3.5 telescope, the line sensitivity would be better than that of ALMA, with instantaneous broad-band coverage. We have built and tested a WaFIRS prototype for 1-1.6 mm, and are currently constructing Z-Spec, a 100 mK model to be used as a ground-based lambda/DELTAlambda approximately equal to 350 submillimeter galaxy redshift machine.

Bradford, C. M.↗

Overview of the relevant CFD work at Thiokol Corporation

An in-house developed proprietary advanced computational fluid dynamics code called SHARP (Trademark) is a primary tool for many flow simulations and design analyses. The SHARP code is a time dependent, two dimensional (2-D) axisymmetric numerical solution technique for the compressible Navier-Stokes equations. The solution technique in SHARP uses a vectorizable implicit, second order accurate in time and space, finite volume scheme based on an upwind flux-difference splitting of a Roe-type approximated Riemann solver, Van Leer's flux vector splitting, and a fourth order artificial dissipation scheme with a preconditioning to accelerate the flow solution. Turbulence is simulated by an algebraic model, and ultimately the kappa-epsilon model. Some other capabilities of the code are 2-D two-phase Lagrangian particle tracking and cell blockages. Extensive development and testing has been conducted on the 3-D version of the code with flow, combustion, and turbulence interactions. The emphasis here is on the specific applications of SHARP in Solid Rocket Motor design. Information is given in viewgraph form.

Chwalowski, Pawel↗

Aerospace applications of SINDA/FLUINT at the Johnson Space Center

SINDA/FLUINT has been found to be a versatile code for modeling aerospace systems involving single or two-phase fluid flow and all modes of heat transfer. Several applications of SINDA/FLUINT are described in this paper. SINDA/FLUINT is being used extensively to model the single phase water loops and the two-phase ammonia loops of the Space Station Freedom active thermal control system (ATCS). These models range from large integrated system models with multiple submodels to very detailed subsystem models. An integrated Space Station ATCS model has been created with ten submodels representing five water loops, three ammonia loops, a Freon loop and a thermal submodel representing the air loop. The model, which has approximately 800 FLUINT lumps and 300 thermal nodes, is used to determine the interaction between the multiple fluid loops which comprise the Space Station ATCS. Several detailed models of the flow-through radiator subsystem of the Space Station ATCS have been developed. One model, which has approximately 70 FLUINT lumps and 340 thermal nodes, provides a representation of the ATCS low temperature radiator array with two fluid loops connected only by conduction through the radiator face sheet. The detailed models are used to determine parameters such as radiator fluid return temperature, fin efficiency, flow distribution and total heat rejection for the baseline design as well as proposed alternate designs. SINDA/FLUINT has also been used as a design tool for several systems using pressurized gasses. One model examined the pressurization and depressurization of the Space Station airlock under a variety of operating conditions including convection with the side walls and internal cooling. Another model predicted the performance of a new generation of manned maneuvering units. This model included high pressure gas depressurization, internal heat transfer and supersonic thruster equations. The results of both models were used to size components, such as the heaters and gas bottles and also to point to areas where hardware testing was needed.

Ewert, Michael K.↗

Learning nonlinear level sets for dimensionality reduction in function approximation

We developed a Nonlinear Level-set Learning (NLL) method for dimensionality reduction in high-dimensional function approximation with small data. This work is motivated by a variety of design tasks in real-world engineering applications, where practitioners would replace their computationally intensive physical models (e.g., high-resolution fluid simulators) with fast-to-evaluate predictive machine learning models, so as to accelerate the engineering design processes. There are two major challenges in constructing such predictive models: (a) high-dimensional inputs (e.g., many independent design parameters) and (b) small training data, generated by running extremely time-consuming simulations. Thus, reducing the input dimension is critical to alleviate the over-fitting issue caused by data insufficiency. Existing methods, including sliced inverse regression and active subspace approaches, reduce the input dimension by learning a linear coordinate transformation; our main contribution is to extend the transformation approach to a nonlinear regime. Specifically, we exploit reversible networks (RevNets) to learn nonlinear level sets of a high-dimensional function and parameterize its level sets in low-dimensional spaces. A new loss function was designed to utilize samples of the target functions' gradient to encourage the transformed function to be sensitive to only a few transformed coordinates. The NLL approach is demonstrated by applying it to three 2D functions and two 20D functions for showing the improved approximation accuracy with the use of nonlinear transformation, as well as to an 8D composite material design problem for optimizing the buckling-resistance performance of composite shells of rocket inter-stages.

Zhang, Guannan↗