Control systems for active flutter suppression (AFS) of a two-dimensional airfoil are designed by means of two methods in which the frequency shaping is taken into consideration. The one is the frequency-shaped weighting LQG (FWLQG). In the method, the performance variables are considered in the frequency domain in order that the controlled system may be robust. The another is the mixed sensitivity reduction problem in H∞ control theory (H∞-mix). Aeroelastic systems, in general, change their internal parameters with free stream velocity. When the system becomes unstable, flutter occurs. To be robust for the parameter perturbation, AFS control systems are designed so that the sensitivity may be low in the low frequency region. As a result, FWLQG and H∞-mix compensators can increase the flutter velocity more than LQG ones. In particular, a H∞-mix compensator of two-output case in which the vertical displacement of the airfoil and its acceleration are considered can increase the flutter velocity by 43.2 percent. Comparing both methods, H∞-mix is superior to FWLQG because the former can shape the frequency response more closely.
In order to investigate the atomic/molecular radiation from hypersonic flowfields, a small plastic projectile weighing 0.4g is accelerated up to higher than 5km/s using a ballistic range. A spectral measurement is done when a projectile passes through the test section, emitting radiation from its strong shock layer: The spectrographic data are obtained in the wavelength range 330 to 650nm. To study the radiation emission with high time-resolution, a high-speed shutter system using an image intensifier and a pulse generator is set up, giving a spatial distribution of radiation intensity in the shock layer. Another measurement is also done, taking framing pictures of the projectile during its flight. For the purpose of convenience, a transformation from the photographic to numerical data is performed, using a digital data-processing system composed of a CCD camera and an image processor. The measured results indicate that such measurement systems can be used in the study of reentry problems in a laboratory environment, after improving of the ballistic range performance.
Coprime factorization controller reduction with frequency weighting (CFW) is extended to the case where controllers are designed by using other than the LQG control theory. In particular, this paper shows that the internal stability of a re-formulated system is equivalent to the one of the original system. This fact indicates that the reduced-order controllers which stabilize the original system can be synthesized and order reduction in the re-formulated system is reasonable. Then, CFW is formulated with using the equivalence of the stability between original and re-formulated systems and with frequency weighting. In numerical simulation, CFW is applied to the controller reduction of active flutter suppression (AFS) system for two-dimensional airfoil. The result shows that CFW is superior to the method in which truncated balanced reduction is directly applied to the controller reduction. Use of frequency weighting is effective for the controller reduction because the stability robustness for the perturbation associated with the controller reduction is taken into consideration.
A numerical method to treat the flow around an arbitrarily moving body is developed in such a way that the incompressible Navier-Stokes equations are solved in the unsteady generalized coordinates. As the numerical scheme, the space is discretized by the finite volume method with a regular mesh, where the convective terms are calculated by the Generalized QUICK method and the others are approximated by the second order symmetric discretization. The time integration is performed by the semi-implicit two step method of second order, where the pressure Poisson equation is solved between the two sub-steps. The boundary condition for pressure is of Neumann type, which represents the balance between the pressure gradient and the unsteady term of motion at boundary. Three test calculations are performed: the steady flow between two coaxial rotating cylinders, the unsteady flow around an advancing cylinder with rotation, and the unsteady flow around an oscillating cylinder. These results are compared with the exact solution of the viscous flow, the experiment, and the potential flow solution, in which the new numerical method proposed here has been validated to have good accuracy.
The “explanatory variables error method” (EVE method) proposed by this report is one of the three main analysis methods under the concept of the “in-flight wind-tunnel test”. The EVE method estimates the values of aerodynamic derivatives, using a multi-regression analysis under the assumption that the noises exist only in the explanatory variables data. This way of thinking about noises is same as the way in the output error method. This suggests that the estimation characteristics obtained from the EVE method can also interpret the contents of the results from the output error method. One of the main characteristics in the EVE method is that the multicollinearity caused by the inherent aerodynamic characteristics of aircrafts may occur in addition to the multicollinearities by the maneuvers and by the constraints due to the equation of motion.