This paper studies three-axis attitude control problems of a class of flexible spacecraft having elastic appendages. A Linear Quadratic Regulator (LQR) associated with a state estimator is designed to control its attitude and the appendage vibration, simultaneously. In order to evaluate and demonstrate the controller performance, ground-based control experiments are conducted, using a ETS (Engineering Test Satellite)-VI laboratory model supported on a single-axis air-table. The evaluation is made concerning mainly with (a) LQR realization by using a thruster and a reaction wheel, and (b) stability robustness against the truncated vibration modes of flexible appendages.
The density-weighted Navier-Stokes equations with k-ε turbulence closure were applied in confined swirling turbulent reacting flows. Extensive comparisons were made with the benchmark test data. The cold boundary difficulty was tentatively suppressed by a temporary lowering of activation energy constant. The two-step reaction model of propane-air diffusion flame could approximate the flame zone. The chemical reaction time based on turbulence scale gave better agreements with the measured radical distribution than that on the mean flow did. The prediction capability of temperature fields was improved by the inclusion of radiative heat losses.
This paper concerns with the problem on aerodynamic loadings over kinked planform wings, when lifting surface theories are used. In using “conventional” mode methods, kinked edges of the planform should be rounded in some way. The effects of roundings on loadings are investigated through BIS-QS, which is a discrete method developed in our laboratory. Another discrete method, DLM, is also used supplementarily, the results of which are shown in the appendix. Two kinds of roundings are used. It is found that the influence of roundings remains spanwise considerably far from the rounded region. Therefore it may be said that “conventional” mode methods are less efficient than discrete methods, for investigating kinked planform wings. “Modified” aerodynamic loading ΔCp mod is introduced, to be kept free from the conventional square-root edge-singularities. It is noteworthy that quasi-conical distributions of ΔCp mod appear near kinks. Thus the well-known peculiar loadings of swept-back, or -forward wings can be explained essentially and naturally. This quasi-conical ΔCpmod would serve for rapid convergence of mode methods in lifting surface theory.
Aerodynamic characteristics of the Weis-Fogh mechanism which is a lift generating mechanism of hovering flight of small insects were studied by the discrete vortex method. Two flat wings were approximated by a finite number of bound vortices, and the unsteady force due to the change of the strengths of these vortices was estimated by a contour of integration including newly introduced nascent vortices. A sufficient lift is generated in all the stages of the wing motion. Especially, at the final stage of the closing motion, the moment for closing the wing becomes very large while the lift remains moderate. At the same stage, a jet is produced between the two wings and a region of high speed flow also appears in the back side of each wing.
A duct flow experiment was conducted to define the behavior of engine inlet distortion and to find the parameter which could be used for one of engine control inputs. In a rectangular duct with cross section of 110×140mm and length of 2m, six levels of distortion were generated by different distortion screens which was placed in the upper half portion of duct inlet. Pressure and temperature distribution caused by the inlet distortion were measured along the duct, and data was evaluated as follows. (1) The distortion levels at each station along the duct do not necessary respond to the levels of inlet distortion. (2) The higher level distortion at the inlet causes the lower average total pressure at each station and the thicker boundary layer on the bottom surface which is washed by the free flow of lower half portion of the duct. (3) In case of high level distortion, it takes long way along the duct to recover the pressure drop on the wall at immediately after the duct inlet. (4) The best parameter to sense the inlet distortion levels is the wall pressure gradient in a specific region.