Document
Mechanics
Tilt/Tip/Piston Manipulator With Base-Mounted Actuators
The geometry and kinematics of this manipulator would afford advantages for some
applications.
Goddard Space Flight Center, Greenbelt, Maryland A proposed three-degree-of-freedom (tilt/tip/piston) manipulator, suitable P for aligning an optical or mechanical Platform Holding component, would offer several advan- Object To Be q Manipulated tages over prior such manipulators: q • Unlike in some other manipulators, no r actuator would support the weight of P another actuator: All of the actuators P q would be mounted on a base. Hence, there would be less manipulated weight.
• The basic geometry of the manipula- R r r tor would afford mechanical advan- Base-Mounted tage: that is, actuator motions would Prismatic Actuators T be larger than the motions they pro- 2 X duce in the manipulated object. Me- N Y O chanical advantage inherently in- N Base N creases the accuracy and resolution of 1 T Plane R manipulation. T • Unlike in some other manipulators, it R would not be necessary to route power and/or data lines through manipula- Lengths of Links R T , R T , and R T are varied to adjust the piston, tilt, and tip coordinates of the platform.
1 1 2 2 3 3 tor joints.
The proposed manipulator (see fig- would be located at the corners of an equi- ward kinematics of the manipulator (given ure) would include three prismatic actu- lateral triangle of side length q on the plat- the lengths of the variable links, find the ators (T N , T N , and T N ) mounted form holding the object to be manipu- tilt, tip, and piston coordinates) have been 1 1 2 2 3 3 on the base and operating in the same lated. Three inextensible limbs (R P , derived. The equations of inverse kinemat- 1 1 plane. Examples of suitable prismatic ac- R P , and R P ) having length r would con- ics (find the variable link lengths needed 2 2 3 3 tuators include lead-screw mechanisms, nect the spherical joints on the platform to obtain the desired tilt, tip, and piston linear hydraulic motors, piezoelectric to revolute joints (R , R , and R ) at the coordinates) have also been derived.
1 2 3 GSC-14874-1 linear drives, inchworm-movement lin- ends of the actuator-controlled links R T , This work was done by Farhad Tahmasebi of ABPI 1 1 12-14-04 es ear stepping motors, and linear flexure R T , and R T . By varying the lengths of Goddard Space Flight Center . Further in- 2 2 3 3 drives. The actuators would control the these links, one could control the tilt, tip, formation is contained in a TSP (see page 1).
lengths of links R T , R T , and R T . and piston coordinates of the platform. GSC-14874-1 1 1 2 2 3 3 Three spherical joints (P , P , and P ) Closed-form equations for direct or for- 1 2 3
Measurement of Model Noise in a Hard-Wall Wind Tunnel
Spurious noise is suppressed in processing of digitized microphone outputs.
Ames Research Center, Moffett Field, California Identification, analysis, and control of noise-source regions and reject unwanted model in a NASA Ames 7-by-10-ft (about fluid-mechanically-generated sound from reflections or background noise. Although 2-by-3-m) wind tunnel for the purpose of models of aircraft and automobiles in spe- it may be difficult to simulate the total fly- identifying and attenuating airframe cial low-noise, semi-anechoic wind tunnels over or drive-by noise in a closed wind tun- noise sources. Simulated landing, take- are an important research endeavor. Such nel, individual noise sources can be iso- off, and approach configurations were studies can also be done in aerodynamic lated and analyzed. evaluated at Mach 0.26. Using a phased wind tunnels that have hard walls if phased An acoustic and aerodynamic study microphone array mounted in the ceil- microphone arrays are used to focus on the was made of a 7-percent-scale aircraft ing over the inverted model, various NASA Tech Briefs, September 2006 25 phone outputs are computed, then oper- were acquired at 12,321 scan points in a ations are performed on the matrices of plane encompassing the model. From microphone-signal cross-spectra. The en- these data, aerodynamic noise from tire acoustic field at one station in such a sources as small as 6 mm on the model system is acquired quickly and interro- surface could be identified easily.
gated during postprocessing. Beam-form- The microphone signals were digitized ing algorithms are employed to scan a at a rate of 153,600 samples per second on plane near the model surface and locate 104 channels simultaneously by use of noise sources while rejecting most back- analog-to-digital converter circuits and a ground noise and spurious reflections. In computer. The resulting maximum the case of the system used in this study, acoustic frequency was 60 kHz with a previous studies in the wind tunnel have bandwidth of 300 Hz. The data for fre- Figure 1. The Array of Microphones was mounted above the aircraft model in the test identified noise sources up to 19 dB quencies <2 kHz were found to be of mar- section of the wind tunnel. A cloth cover has below the normal background noise of ginal utility because the microphone been removed from under the microphones to ARC-14967 Fig 1 the wind tunnel. Theoretical predictions beam pattern at those frequencies was too make the model visible in this view. ABPI 06-06-06 le of array performance are used to mini- wide. The data for frequencies >32 kHz mize the width and the side lobes of the were found to be of marginal utility be- beam pattern of the microphone array cause at those frequencies, the sources for a given test arrangement. were too weak and the side lobes too To capture flyover noise of the in- strong. The frequency limits of 2 and 32 verted model, a 104-element micro- kHz correspond to limits of 140 and 2,240 phone array in a 622-mm-diameter clus- Hz, respectively, on the full-scale aircraft.
ter was installed in a 19-mm-thick A sound-convection correction was in- poly(methyl methacrylate) plate in the cluded in the processing of the data so ceiling of the test section of the wind that sources appeared to come from the tunnel above the aircraft model (see Fig- model rather than being swept down- ure 1). The microphones were of the stream. The acoustic sources were de- condenser type, and their diaphragms picted, one frequency at a time, as color were mounted flush in the array plate, contours on the scan plane with the which was recessed 12.7 mm into the model outline superimposed, as shown ceiling and covered by a porous aro- in Figure 2. Various integration schemes Figure 2. A Contour Map of Wing-Slat Noise at a matic polyamide cloth (not shown in the have been developed to compute the frequency of 12.6 kHz was computed from meas- figure) to minimize boundary-layer combined effects on a listener and to urements made by the microphone array. The ARC-14967 Fig 2 noise. This design caused the level of generate narrowband and third-octave ABPI color-contour range is 8 dB.
06-06-06 le flow noise to be much less than that of acoustic spectra.
noise sources in the high-lift system, flush-mount designs. The drawback of Ten airframe noise sources that might landing gear, fins, and miscellaneous this design was that the cloth attenuated be important to approach and landing other components were located and sound somewhat and created acoustic noise of the full-scale aircraft were iden- compared for sound level and frequency resonances that could grow to several dB tified in the study. The relative strengths at one flyover location. Numerous noise- at a frequency of 10 kHz. of these sources and their dependences alleviation devices and modifications of A correction methodology has been on the configuration of the aircraft were the model were evaluated. Simultane- developed to account for the signal in- documented. Although the data were ously with acoustic measurements, aero- terference. The first side lobe of the scaled to the frequencies for the full- dynamic forces were recorded to docu- beam pattern was 13.4 dB down from scale aircraft, no extrapolation to full- ment aircraft conditions and any the peak response at 8 kHz and at an scale flyover was performed.
performance changes caused by geomet- angle of 23° from the normal vector: This work was done by Paul T. Soderman ric modifications. these characteristics made it possible to of Ames Research Center . For further in- Most modern microphone-array sys- obtain good acoustic signals from the formation, contact the Ames Technology Part- tems function in the frequency domain model when the model was located at a nerships Division at (650) 604-2954.
in the sense that spectra of the micro- distance of 1.11 m from the array. Data ARC-14967
Loci-STREAM Version 0.9
Marshall Space Flight Center, Alabama Loci-STREAM is an evolving compu- etc. Loci-STREAM implements a pres- robust for flows at all speeds from zero tational fluid dynamics (CFD) software sure-based flow-solving algorithm that to hypersonic. The flexibility of arbi- tool for simulating possibly chemically utilizes unstructured grids. (The bene- trary polyhedral grids enables accurate, reacting, possibly unsteady flows in di- fit of low memory usage by pressure- efficient simulation of flows in complex verse settings, including rocket en- based algorithms is well recognized by geometries, including those of plume- gines, turbomachines, oil refineries, experts in the field.) The algorithm is impingement problems. The present 26 NASA Tech Briefs, September 2006