Search

A paper describes a photogrammetric method for measuring the changing shape of a gossamer (membrane) structure deployed in outer space. Such a structure is typified by a solar sail comprising a transparent polymeric membrane aluminized on its Sun-facing side and coated black on the opposite side. Unlike some prior photogrammetric methods, this method does not require an artificial light source or the attachment of retroreflectors to the gossamer structure. In a basic version of the method, the membrane contains a fluorescent dye, and the front and back coats are removed in matching patterns of dots. The dye in the dots absorbs some sunlight and fluoresces at a longer wavelength in all directions, thereby enabling acquisition of high-contrast images from almost any viewing angle. The fluorescent dots are observed by one or more electronic camera(s) on the Sun side, the shade side, or both sides. Filters that pass the fluorescent light and suppress most of the solar spectrum are placed in front of the camera(s) to increase the contrast of the dots against the background. The dot image(s) in the camera(s) are digitized, then processed by use of commercially available photogrammetric software.

The Mini-Mast testbed was one of the first large scale Controls-Structure-Interaction (CSI) systems used to evaluate state-of-the-art methodology in flexible structure control. Now that all the testing at Langley Research Center has been completed, a look back is warranted to evaluate the program. This paper describes some of the experiences and technology development studies by NASA, university, and industry investigators. Lessons learned are presented from three categories: the testbed development, control methods, and the operation of a guest investigator program. It is shown how structural safety margins provided a realistic environment to simulate on-orbit CSI research, even though they also reduced the research flexibility afforded to investigators. The limited dynamic coupling between the bending and torsion modes of the cantilevered test article resulted in highly successful SISO and MIMO controllers. However, until accurate models were obtained for the torque wheel actuators, sensors, filters, and the structure itself, most controllers were unstable. Controls research from this testbed should be applicable to cantilevered appendages of future large space structures.

A new method is described for assessing the consistency of structural modal parameters identified with the Eigensystem Realization Algorithm. Identification results show varying consistency in practice due to many sources including high modal density, nonlinearity, and inadequate excitation. Consistency is considered to be a reliable indicator of accuracy. The new method is the culmination of many years of experience in developing a practical implementation of the Eigensystem Realization Algorithm. The effectiveness of the method is illustrated using data from NASA Langley's Controls-Structures-Interaction Evolutionary Model.

Identification of structures which exhibit modal interactions is considered, and the difficulties experienced due to these interactions are examined. Free oscillations of quadratically and cubically coupled pairs of oscillators are analytically studied to illustrate nonlinear interactions between structural modes involved in two-to-one and one-to-one frequency relationships. In light of this study, results obtained from application of the eigensystem realization algorithm toward identification of a beam-mass structure with a two-to-one frequency relationship and quadratic coupling are presented and discussed.

A comparative overview is presented of the disciplines of modal testing used in structural engineering and system identification used in control theory. A list of representative references from both areas is given, and the basic methods are described briefly. Recent progress on the interaction of modal testing and control disciplines is discussed. It is concluded that combined efforts of researchers in both disciplines are required for unification of modal testing and system identification methods for control of flexible structures.

The basic concept of the Eigensystem Realization Algorithm for modal parameter identification and model reduction is extended to minimize the distortion of the identified parameters caused by noise. The mathematical foundation for the properties of accuracy indicators, such as the singular values of the data matrix and modal amplitude coherence, is provided, based on knowledge of the noise characteristics. These indicators quantitatively discriminate noise from system information and are used to reduce the realized system model to a better approximation of the true model. Monte Carlo Simulations are included to support the analytical studies.

The basic concept of the Eigensystem Realization Algorithm (ERA) for modal parameter identification and model reduction is extended to minimize the distortion of the identified parameters caused by noise. The mathematical foundation for the properties of accuracy indicators such as the singular values of the data matrix and modal amplitude coherence is provided based on knowledge of the noise characteristics. These indicators quantitatively discriminate noise from system information and are used to reduce the realized system model to a better approximation of the true model. Monte Carlo simulations are included to support the analytical studies.

The accuracy of the Ibrahim time Domain (ITD) identification algorithm in extracting structural model parameters from free response functions was studied using computer simulated data for 65 positions on an isotropic, uniform thickness plate with mode shapes obtained by NASTRAN analysis. Natural frequencies were used to study identification results over ranges of modal parameter values and user selectable algorithm constants. Effects of superimposing various levels of noise onto the functions were investigated. No detrimental effects were observed when the number of computational degrees of freedom allowed in the algorithm was made many times larger than the minimum necessary for adequate identification. The use of a high number of degrees of freedom when analyzing experimental data, for the simultaneous identification of many modes in one computer run are suggested.

The ability of the ITD identification algorithm in identifying a complete set of structural modal parameters using a large number of free-response time histories simultaneously in one analysis, assuming a math model with a high number of degrees-of-freedom, has been studied. Identification results using simulated free responses of a uniform rectangular plate, with 225 measurement stations, and experimental responses from a ground vibration test of the Long Duration Exposure Facility (LDEF) Space Shuttle payload, with 142 measurement stations, are presented. As many as 300 degrees-of-freedom were allowed in analyzing these data. In general, the use of a significantly oversized math model in the identification process was found to maintain or increase identification accuracy and to identify modes of low response level that are not identified with smaller math model sizes. The concept of a Mode Shape Correlation Constant is introduced for use when more than one identification analysis of the same structure are conducted. This constant quantifies the degree of correlation between any two sets of complex mode shapes identified using different excitation conditions, different user-selectable algorithm constants, or overlapping sets of measurements.

A method called the eigensystem realization algorithm is developed for modal parameter identification and model reduction of dynamic systems from test data. A new approach is introduced in conjunction with the singular-value decomposition technique to derive the basic formulation of minimum order realization which is an extended version of the Ho-Kalman algorithm. The basic formulation is then transformed into modal space for modal parameter identification. Two accuracy indicators are developed to quantitatively identify the system and noise modes. For illustration of the algorithm, an example is shown using experimental data from the Galileo spacecraft.

The OAST-1 Solar Array Experiment was flown on Shuttle mission STS-41D on August 31, 1984 for NASA's Office of Aeronautics and Space Technology. It is pointed out that this was the first flight demonstration of the operation of a large solar array which was designed for multiple deployments and retractions. The array represents also a new class of very lightweight, flexible, low-frequency structures. A program was, therefore, implemented with the aim to study the structural dynamics of the array on the basis of a photogrammetric analysis of data obtained with the existing Shuttle closed circuit television (CCTV) system. The present paper has the objective to describe the experiment and its implementation, taking into account the unique process of analyzing the video data.

The NASA Langley Research Center participated in the Solar Array Experiment with two primary objectives: (1) to study the structural and control dynamics of a new class of large, lightweight, low-frequency space structures, and (2) to develop technology for remote video measurement of structural motions. The shuttle orbiter's closed circuit television (CCTV) system was used to provide recorded video images of the solar array from four locations in the payload bay, two on the forward bulkhead and two on the aft bulkhead. White reflective targets were placed on the array to provide discrete points at which to track array motion. A dynamic test consisted of a quiescent period in which orbiter operations were inhibited and crew motion restricted, an excitation period, and a free decay period. The orbiter was placed in free drift while in a gravity gradient orientation and dynamics tests were timed to occur at orbital noon so that the Sun would illuminate one side of the array and Earth albedo the other. The CCTV system was turned on during the quiescent period, approximately 3 minutes prior to excitation. Video from all four CCTV cameras was recorded continuously through the excitation and free-decay periods for a total test record of 8 to 10 minutes.

A method, called the Eigensystem Realization Algorithm (ERA), is developed for modal parameter identification and model reduction of dynamic systems from test data. A new approach is introduced in conjunction with the singular value decomposition technique to derive the basic formulation of minimum order realization which is an extended version of the Ho-Kalman algorithm. The basic formulation is then transformed into modal space for modal parameter identification. Two accuracy indicators are developed to quantitatively identify the system modes and noise modes. For illustration of the algorithm, examples are shown using simulation data and experimental data for a rectangular grid structure.

A modal parameter identification technique referred to as the Eigensystem Realization Algorithm (ERA) was applied to free-response measurements from the Galileo spacecraft modal survey test. The data were recorded following single-point random excitation of the structure. This work is one phase in a research project coordinated by the Jet Propulsion Laboratory to compare the performance of various contemporary identification techniques using Galileo data. Principal emphasis is placed on estimating the accuracy of the ERA-identified modal parameters. Various accuracy indicators, such as Modal Amplitude Coherence and Modal Phase Collinearity, are discussed. More than 20 modes of the spacecraft were identified, demonstrating the ability of the ERA method to determine the dynamics of such complex structures using only a few seconds of test data.

Results of Monte Carlo numerical simulations conducted to study the close-mode performance of the Ibrahim Time-Domain (ITD) modal identification algorithm are presented. The ITD technique is a matrix eigensolution method for obtaining structural modal parameters directly from free-response test data without using the FFT. Thus, the well-known resolution and leakage limitations of the FFT procedure, that are particularly significant with short records, are avoided. As an example, one of several experimental data analyses where close modes have been accurately identified using very short records is shown. Although the identification scatter is found to increase as the square of reductions in frequency separation at small separations, the ability to differentiate modes spaced at fractions of the FFT resolution is substantiated.

Vibration studies carried out in a vacuum chamber are reported for a three-sided membrane with inwardly curved edges. Uniform tension was transmitted by thin steel cables encased in the edges. Variation of ambient air pressure from atmospheric to near vacuum resulted in increased response frequencies and amplitudes. The first few vibration modes measured in a near vacuum are shown to be predictable by a finite element structural analysis over a range of applied tension loads. The complicated vibration mode behavior observed during tests at various air pressures is studied analytically with a nonstructural effective air-mass approximation. The membrane structure is a candidate for reflective surfaces in space antennas.

Performance characteristics of the Ibrahim time-domain algorithms are examined on a statistical basis. The method uses multiple free-response measurements from a test structure directly in the time domain, and displays varying accuracy, with the strongly excited modes identified most accurately and the weakly excited modes least accurately. Sets of computer-generated free-response functions for a simple 1-mode model are used to study the identification performance for several basic changes in the modal characteristics of the data, the user-selectable analysis constants, and the level of superimposed noise. Results indicate the existence of a direct and often linear effect of noise on the scatter of identified parameters, and good identification is maintained at the 1% noise level down to the point in frequency when only 1/4-cycle of data is analyzed.

The ability of the ITD identification algorithm in identifying a complete set of structural modal parameters using a large number of free-response time histories simultaneously in one analysis, assuming a math model with a high number of degrees-of-freedom, has been studied. Identification results using simulated free responses of a uniform rectangular plate, with 225 measurement stations, and experimental responses from a ground vibration test of the Long Duration Exposure Facility (LDEF) Space Shuttle payload, with 142 measurement stations, are presented. As many as 300 degrees-of-freedom were allowed in analyzing these data. In general, the use of a significantly oversized math model in the identification process was found to maintain or increase identification accuracy and to identify modes of low response level that are not identified with smaller math model sizes. The concept of a Mode Shape Correlation Constant is introduced for use when more than one identification analysis of the same structure are conducted. This constant quantifies the degree of correlation between any two sets of complex mode shapes identified using different excitation conditions, different user-selectable algorithm constants, or overlapping sets of measurements.

The results of an exploratory NASA study, leading to a better understanding of the effects of meteorological conditions on the propagation of aircraft noise, are reported. The experimental program utilized a known sound source fixed atop an instrumented meteorological tower. The basic experimental scheme consisted of measuring the amplitude of sound radiated toward the ground along a line of microphones fixed to a tower guy wire. Experimental results show the feasibility of this approach in the acquisition of data indicating the variations encountered in the time-averaged and instantaneous amplitudes of propagated sound. The investigation included a consideration of ground reflections, a comparison of measured attenuations with predicted atmospheric absorption losses, and an evaluation of the amplitude fluctuations of recorded sound pressures.

The full-scale wind tunnel at NASA-Langley Research Center was designed for low-speed aerodynamic testing of aircraft. Sound absorbing treatment has been added to the ceiling and walls of the tunnel test section to create a more anechoic condition for taking acoustical measurements during aerodynamic tests. The results of an experimental investigation of the present acoustical characteristics of the tunnel test section are presented. The experimental program included measurements of ambient nosie levels existing during various tunnel operating conditions, investigation of the sound field produced by an omnidirectional source, and determination of sound field decay rates for impulsive noise excitation. A comparison of the current results with previous measurements shows that the added sound treatment has improved the acoustical condition of the tunnel test section. An analysis of the data indicate that sound reflections from the tunnel ground-board platform could create difficulties in the interpretation of actual test results.

Fluctuating pressures were measured at 30 positions on the surface of a 1/4-scale YC-14 wing and fuselage model during an outdoor static testing program. These data were obtained as part of a NASA program to study the fluctuating loads imposed on STOL aircraft configurations and to further the understanding of the scaling laws of unsteady surface pressure fields. Fluctuating pressure data were recorded at several discrete engine thrust settings for each of 16 configurations of the model. These data were reduced using the technique of random data analysis to obtain auto-and cross-spectral density functions and coherence functions for frequencies from 0 to 10 kHz, and cross-correlation functions for time delays from 0 to 10.24 ms. Results of this program provide the following items of particular interest: (1) Good collapse of normalized PSD functions on the USB flap was found using a technique applied by Lilley and Hodgson to data from a laboratory wall-jet apparatus. (2) Results indicate that the fluctuating pressure loading on surfaces washed by the jet exhaust flow was dominated by hydrodynamic pressure variations, loading on surface well outside the flow region dominated by acoustic pressure variations, and loading near the flow boundaries from a mixture of the two.

A research program is conducted for the study of the fluctuating loads imposed on both upper-surface blown-flap and externally blown-flap powered-lift STOL aircraft configurations by the impingement of the jet engine exhaust flow. Attention is given to the measurement of the unsteady pressures at 30 positions in the vicinity of the jet exhaust on the surface of a NASA 1/4-scale YC-14 boilerplate wing and fuselage section.