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.
A Volterra series approach was applied to the identification of nonlinear systems which are described by a neural network model. A procedure is outlined by which a mathematical model can be developed from experimental data obtained from the network structure. Applications of the results to the control of robotic systems are discussed.
Identification, analysis, and control of fluid-mechanically-generated sound from models of aircraft and automobiles in special low-noise, semi-anechoic wind tunnels are an important research endeavor. Such studies can also be done in aerodynamic wind tunnels that have hard walls if phased microphone arrays are used to focus on the noise-source regions and reject unwanted reflections or background noise. Although it may be difficult to simulate the total flyover or drive-by noise in a closed wind tunnel, individual noise sources can be isolated and analyzed. An acoustic and aerodynamic study was made of a 7-percent-scale aircraft model in a NASA Ames 7-by-10-ft (about 2-by-3-m) wind tunnel for the purpose of identifying and attenuating airframe noise sources. Simulated landing, takeoff, and approach configurations were evaluated at Mach 0.26. Using a phased microphone array mounted in the ceiling over the inverted model, various noise sources in the high-lift system, landing gear, fins, and miscellaneous other components were located and compared for sound level and frequency at one flyover location. Numerous noise-alleviation devices and modifications of the model were evaluated. Simultaneously with acoustic measurements, aerodynamic forces were recorded to document aircraft conditions and any performance changes caused by geometric modifications. Most modern microphone-array systems function in the frequency domain in the sense that spectra of the microphone outputs are computed, then operations are performed on the matrices of microphone-signal cross-spectra. The entire acoustic field at one station in such a system is acquired quickly and interrogated during postprocessing. Beam-forming algorithms are employed to scan a plane near the model surface and locate noise sources while rejecting most background noise and spurious reflections. In the case of the system used in this study, previous studies in the wind tunnel have identified noise sources up to 19 dB below the normal background noise of the wind tunnel. Theoretical predictions of array performance are used to minimize the width and the side lobes of the beam pattern of the microphone array for a given test arrangement. To capture flyover noise of the inverted model, a 104-element microphone array in a 622-mm-diameter cluster was installed in a 19-mm-thick poly(methyl methacrylate) plate in the ceiling of the test section of the wind tunnel above the aircraft model (see Figure 1). The microphones were of the condenser type, and their diaphragms were mounted flush in the array plate, which was recessed 12.7 mm into the ceiling and covered by a porous aromatic polyamide cloth (not shown in the figure) to minimize boundary-layer noise. This design caused the level of flow noise to be much less than that of flush-mount designs. The drawback of this design was that the cloth attenuated sound somewhat and created acoustic resonances that could grow to several dB at a frequency of 10 kHz.
A turbofan simulation has been developed for use in aero-propulso-servo-elastic coupling studies, on supersonic vehicles. A one-dimensional lumped volume approach is used whereby each component (fan, high-pressure compressor, combustor, etc.) is represented as a single volume using characteristic performance maps and conservation equations for continuity, momentum and energy. The simulation is developed in the MATLAB/SIMULINK (The MathWorks, Inc.) environment in order to facilitate controls development, and ease of integration with a future aero-servo-elastic vehicle model being developed at NASA Langley. The complete simulation demonstrated steady state results that closely match a proposed engine suitable for a supersonic business jet at the cruise condition. Preliminary investigation of the transient simulation revealed expected trends for fuel flow disturbances as well as upstream pressure disturbances. A framework for system identification enables development of linear models for controller design. Utilizing this framework, a transfer function modeling an upstream pressure disturbance s impacts on the engine speed is developed as an illustrative case of the system identification. This work will eventually enable an overall vehicle aero-propulso-servo-elastic model
The objective of system identification is to determine reliable dynamical models of a structure by systematically using its measured excitation and response. It brings together in an integrated fashion, experimental, analytical, and computational techniques in structural dynamics. Areas of application for system identification include the following: (1) Model Evaluation--assessing assumptions (linearity and equivalent viscous damping) and techniques (finite-element modeling) used to construct theoretical models of a structure; (2) Model Improvement--updating of a theoretical model to enable more accurate response predictions for possible future loads on the structure, or for control of the structure; (3) Empirical Modelling--developing empirical relationships (nonlinear models) or empirical parameter values (modal damping) because the present state of the art does not provide theoretical results; and (4) Damage Detection and Assessment--continual or episodic updating of a structural model through vibration monitoring to detect and locate any structural damage. It can be argued that since the construction or modification of models using test data is subject to inherent uncertainties, the above problems should be properly treated within a Bayesian probabilistic framework. Such a methodology is presented which allows the precision of the estimates of the model parameters to be computed. It also leads to a guiding principle in applications. Namely, when selecting a single model from a given class of models, one should take the most probable model in the class based on the experimental data. Practical applications of this principle are given which are based on the utilization of measured seismic motions in large civil structures. Examples include the application of a computer program MODE-ID to identify modal properties directly from seismic excitation and response time histories from a nine-story steel-frame building at JPL and from a freeway overpass bridge.
The damage location approach employs the control system capabilities for the structure to test the structure and measure the dynamic response. The measurements are then used in a system identification algorithm to produce a model of the damaged structure. The model is compared to one for the undamaged structure to find regions of reduced stiffness which indicate the location of damage. Kabe's 3,4 stiffness matrix adjustment method was the central identification algorithm. The strength of his method is that, with minimal data, it preserves the representation of the physical connectivity of the structure in the resulting model of the damaged truss. However, extensive storage and computational effort were required as a result. Extension of the damage location method to overcome these problems is the first part of the current work. The central system identification algorithm is replaced with the MSMT method of stiffness matrix adjustment which was previously derived by generalizing an optimal-update secant method form quasi-Newton approaches for nonlinear optimization. Validation of the extended damage location method is the second goal.
Lightning occurrence presents safety challenges to people and property. The main challenge with lightning safety is that the majority of guidance is reactive. In other words, lightning has to have already occurred nearby before a person will respond and take shelter. Further, most injuries or fatalities occur as the storm approaches, or as it's moving away, when rainfall may not be present at the time of the flash. Thus, this project develops a physically-based deep learning model to produce lightning probabilities out to 15 minutes. The deep learning model combines a Convolutional Neural Network (CNN) with a Long Short-Term Memory (LSTM) network to capture both the spatial and temporal evolution of storms to predict the probability that lightning initiation will occur in the next 15 minutes. The model combines radar reflectivity, correlation coefficient and differential reflectivity to inferred storm hydrometer type and precipitation phase, which aids in the identification of electrification processes. The model is trained with data from the Geostationary Lightning Mapper (GLM), which is a near infrared sensor onboard the GOES-R series of satellites that measures optical brightness from lightning. This presentation will provide an overview of the project.
This study presents newly developed benchmarks for modeling flow and transport within discrete fracture networks (DFNs) and useful methods for analyzing the results. The new benchmarks are designed to test modeling approaches for use in probabilistic performance assessment models of deep geologic repositories in fractured rock. The benchmarks simulate flow and transport through a 1 km 3 block of fractured rock. The first simulates migration of a short pulse of tracer through a simple network of four intersecting fractures. The second adds 1089 stochastically generated fractures. The third changes the pulse to a continuous point source. Evaluation of model performance relies on moment analysis and comparison of the results of different models. The expected nondimensional first moment of the conservative tracer for each benchmark is 1. The benchmarks were simulated by teams from Canada, Czechia, Germany, Korea, Sweden, Taiwan, and the United States as part of a DECOVALEX-2023 study (decovalex.org). The teams used various approaches, including explicit DFN modeling, DFN upscaling to an equivalent continuous porous medium (ECPM), and a combination of both methods. Transport mechanisms are modeled using either the advection-dispersion equation or particle tracking. Results demonstrate strong agreement among the models in breakthrough behavior up to the 75th percentile. Significant deviations in first moments and well-clustered outputs led to the identification of inaccuracies in several models. Such findings exemplify the benefit of exercising these benchmarks and using the presented methods to test DFN flow and transport models.
Time accurate solutions of the Euler and Navier{Stokes equations are used as an approach to elucidate aerodynamic coefficients that include rigid body motion effects. The Army-Navy Finner geometry is used for work flow development due to its simple shape, inexpensive grid generation, and available literature that include aerodynamic damping derivatives obtained from ight test, wind tunnel tests, and computational fluid dynamics. Supersonic conditions for pitch and roll damping include angles of attack up to 90 deg. Aerodynamic responses due to rigid body maneuvers with prescribed wind incidence angles and body rates are computed using the DoD CREATE Kestrel and NASA FUN3D flow solvers. First, reference numerical and experimental results provide validation of aero- dynamic damping terms computed by traditional periodic motion in roll and pitch. Next, individual, impulse motion inputs provide the canonical responses for general input-output modeling based on classical superposition and convolution concepts. Finally, simultaneous impulse excitation of all inputs provides an efficient system identification training scenario for accurate aerodynamic model construction in state space via the NASA Sys- tem/Observer/Controller Identification Toolbox.
A simplified open-loop dynamic model of the T700 turboshaft engine, valid within the normal operating range of the engine, is developed. This model is obtained by linking linear state space models obtained at different engine operating points. Each linear model is developed from a detailed nonlinear engine simulation using a multivariable system identification and realization method. The simplified model may be used with a model-based real time diagnostic scheme for fault detection and diagnostics, as well as for open loop engine dynamics studies and closed loop control analysis utilizing a user generated control law.
This paper presents recent results studyingthe influence of aero-propulsive interactioneffects on a small quadrotor vehicle using isolated rotor, bareairframe,andpoweredairframe static wind tunnel test results.Aerodynamic interaction model development was explored utilizing system identification tools and incorporatedintoa previously developed NASA quadrotor flight dynamic simulation.The aerodynamic interaction model improvedtheaccuracyof the simulation in comparisonto free-flight and static wind tunnel resultswithheld from model development. Comparisons between models of different fidelity were completed to demonstrate the effects of aerodynamic interactionson model accuracy and provide recommendations for modeling requirements.
Noble gas transport through geologic media has important applications in the characterization of underground nuclear explosions (UNEs). Without accurate transport models, it is nearly impossible to distinguish between xenon signatures originating from civilian nuclear facilities and UNEs. Understanding xenon transport time through the earth is a key parameter for interpreting measured xenon isotopic ratios. One of the most challenging aspects of modeling gas transport time is accounting for the effect of variable water saturation of geological media. In this study, we utilize bench-scale laboratory experiments to characterize the diffusion of krypton, xenon, and sulfur hexafluoride (SF6) through intact zeolitic tuff under different saturations. Here, we demonstrate that the water in rock cores with low partial saturation dramatically affects xenon transport time compared to that of krypton and SF6 by blocking sites in zeolitic tuff that preferentially adsorb xenon. This leads to breakthrough trends that are strongly influenced by the degree of the rock saturation. Xenon is especially susceptible to this phenomenon, a finding that is crucial to incorporate in subsurface gas transport models used for nuclear event identification. We also find that the breakthrough of SF6 diverges significantly from that of noble gases within our system. When developing field scale models, it is important to understand how the behavior of xenon deviates from chemical tracers used in the field, such as SF6 (Carrigan et al., 1996). These new insights demonstrate the critical need to consider the interplay between rock saturation and fission product sorption during transport modeling, and the importance of evaluating specific interactions between geomedia and gases of interest, which may differ from geomedia interactions with chemical tracers.
The production of biofuels from lignocellulose yields a substantial lignin by-product stream that currently has few applications. Biological conversion of lignin-derived compounds into chemicals and fuels has the potential to improve the economics of lignocellulose-derived biofuels, but few microbes are able both to catabolize lignin-derived aromatic compounds and to generate valuable products. While Escherichia coli has been engineered to produce a variety of fuels and chemicals, it is incapable of catabolizing most aromatic compounds. Therefore, we engineered E. coli to catabolize protocatechuate, a common intermediate in lignin degradation, as the sole source of carbon and energy via heterologous expression of a nine-gene pathway from Pseudomonas putida KT2440. Then, we used experimental evolution to select for mutations that increased growth with protocatechuate more than 2-fold. Increasing the strength of a single ribosome binding site in the heterologous pathway was sufficient to recapitulate the increased growth. After optimization of the core pathway, we extended the pathway to enable catabolism of a second model compound, 4-hydroxybenzoate. These engineered strains will be useful platforms to discover, characterize, and optimize pathways for conversions of lignin-derived aromatics. IMPORTANCE Lignin is a challenging substrate for microbial catabolism due to its polymeric and heterogeneous chemical structure. Therefore, engineering microbes for improved catabolism of lignin-derived aromatic compounds will require the assembly of an entire network of catabolic reactions, including pathways from genetically intractable strains. By constructing defined pathways for aromatic compound degradation in a model host would allow rapid identification, characterization, and optimization of novel pathways. Finally, we constructed and optimized one such pathway in E. coli to enable catabolism of a model aromatic compound, protocatechuate, and then extended the pathway to a related compound, 4-hydroxybenzoate. This optimized strain can now be used as the basis for the characterization of novel pathways.
The fundamental challenge in identification of nonlinear dynamic systems is determining the appropriate form of the model. A robust technique is presented which essentially eliminates this problem for many applications. The technique is based on the Minimum Model Error (MME) optimal estimation approach. A detailed literature review is included in which fundamental differences between the current approach and previous work is described. The most significant feature is the ability to identify nonlinear dynamic systems without prior assumption regarding the form of the nonlinearities, in contrast to existing nonlinear identification approaches which usually require detailed assumptions of the nonlinearities. Model form is determined via statistical correlation of the MME optimal state estimates with the MME optimal model error estimates. The example illustrations indicate that the method is robust with respect to prior ignorance of the model, and with respect to measurement noise, measurement frequency, and measurement record length.
This paper presents and compares several system identification techniques proposed for use with higher harmonic control algorithms designed to alleviate helicopter vibration. All method for actively controlling helicopter vibration require the knowledge of how the vibration outputs are related to the control inputs. Off-line or batch identification methods for obtaining this knowledge are presented first. Then the more advanced, adaptive identification techniques proposed to track the helicopter model parameters in flight are discussed. Considerations regarding system identifiability, identification algorithm stability, and computer implementation are also discussed.
The Lightning Mapper Sensor on GOES-R builds on previous measurements of lightning from low earth orbit by the OTD (Optical Transient Detector) and LIS (Lightning Imaging Sensor) sensors. Unlike observations from low earth orbit, the GOES-R platform will allow continuous monitoring of lightning activity over the Continental United States and southern Canada, Central and South America, and portions of the Atlantic and Pacific Oceans. The LMS will detect total (cloud-to-ground and intracloud) lightning at storm scale resolution (approx. 8 km) using a highly sensitive Charge Coupled Device (CCD) detector array. Discrimination between lightning optical transients and a bright sunlit background scene is accomplished by employing spectral, spatial, and temporal filtering along with a background subtraction technique. The result is 24 hour detection capability of total lightning. These total lightning observations can be made available to users within about 20 seconds. Research indicates a number of ways that total lightning observations from LMS could benefit operational activities, including 1) potential increases in lead times and reduced false alarms for severe thunderstorm and tornado Warnings, 2) improved routing of &rail around thunderstorms, 3) support for spacecraft launches and landings, 4) improved ability to monitor tropical cyclone intensity, 5) ability to monitor thunderstorm intensification/weakening during radar outages or where radar coverage is poor, 6) better identification of deep convection for the initialization of numerical prediction models, 7) improved forest fire forecasts, 8) identification of convective initiation, 9) identification of heavy convective snowfall, and 10) enhanced temporal resolution of storm evolution (1 minute) than is available from radar observations. Total lightning data has been used in an operational environment since July 2003 at the Huntsville, Alabama National Weather Service office. Total lightning measurements are obtained by the North Alabama Lightning Mapping Array (LMA) and have successfully been used in warning decisions. Every 2 minutes, total lightning counts in 2 km by 2 km horizontal, 1 km vertical grids are available to forecasters on an AWIPS (Advanced Weather Interactive Processing System) workstation. Experience with the LMA total lightning data is used to illustrate the potential use of LMS data that would be available to forecasters across the US. This abstract is for submission as a presentation to the National Weather Association Annual Meeting to be held 16-21 October 2004 in Portland, OR. This abstract will be published in the conference proceedings.
This paper focuses on parameter identification for time-varying voltage-dependent polynomial load models in distribution systems—namely, the constant impedance (Z), constant current (I), and constant power (P). The paper assumes the presence of measurements at the distribution nodes from various sources, each at varying temporal resolutions. The least- squares estimation method with bounded variables is used to convert the measurements into ZIP parameter estimations for loads at various locations on the feeder. The results and discussions focus on some of the estimation issues faced in a distribution system caused by varying voltage sensitivities of loads along a feeder, measurement resolution, impact of window size and sampling rate of measurement data. Finally, the paper uses dynamic simulations data (synthetic high-resolution measurements) to further improve the estimation results.
The present study is concerned with the mathematical modelling of aircraft dynamics on the basis of an investigation conducted with the aid of the Rotor System Research Aircraft (RSRA). The particular characteristics of RSRA make it possible to investigate aircraft properties which cannot be readily studied elsewhere, for example in the wind tunnel. The considered experiment had mainly the objective to develop an improved understanding of the physics of rotor flapping dynamics and rotor loads in maneuvers. The employed approach is based on a utilization of parameter identification methodology (PID) with application to helicopters. A better understanding of the contribution of the main rotor to the overall aircraft forces and moments is also to be obtained. Attention is given to the mathematical model of a rotorcraft system, an integrated identification method, flight data processing, and the identification of RSRA mathematical models.
Representation and identification of a non-linear aeroelastic pitch-plunge system as a model of the NARMAX class is considered. A non-linear difference equation describing this aircraft model is derived theoretically and shown to be of the NARMAX form. Identification methods for NARMAX models are applied to aeroelastic dynamics and its properties demonstrated via continuous-time simulations of experimental conditions. Simulation results show that (i) the outputs of the NARMAX model match closely those generated using continuous-time methods and (ii) NARMAX identification methods applied to aeroelastic dynamics provide accurate discrete-time parameter estimates. Application of NARMAX identification to experimental pitch-plunge dynamics data gives a high percent fit for cross-validated data.