206 M. T. Hughes et al. with degrees of freedom in pitch and heave. Further studies by Bryant and Garcia [13] built on the previous system and experimentally characterized the power output and flutter response over a large range of freestream conditions and with both a flat plate and an airfoil. Zhao and Yang [14] showed that the addition of a beam stiffening device resulted in effective power generator for aeroelastic systems experiencing galloping instabilities, vortex-induced vibrations, and flutter. Other methods of power enhancement of such aeroelastic systems have focused on wing-wake interactions resulting from upstream bodies. Kirschmeier and Bryant [15] showed that tandem oscillating wings can be tuned to enhance the energy transfer from the wake into the downstream wing. This chapter will discuss the application of a state-machine control scheme to automatically control aeroelastic behavior in a two degree-of-freedom (2DOF) aeroelastic wing using vortices produced by an upstream variable frequency disturbance generator. The remainder of this chapter will walk through the development of the aeroelastic wing and the disturbance generator experimental hardware, discuss the state-machine control scheme, present preliminary results, and discuss progress toward incorporating a recurrent neural network (RNN) machine learning model in the aeroelastic control approach. 27.2 Methodology Experimental testing was performed in the North Carolina State University (NCSU) Subsonic Wind Tunnel, which is a closed-return tunnel with a test section measuring 0.81 m by 1.14 m by 1.17 m. A variable pitch fan driven by a three-speed electric motor provides flow in the tunnel, measured by dynamic pressure, which is controlled by changing the blade pitch at each of the three motor speeds. The maximum dynamic pressure in the tunnel test section is 720 Pa, corresponding to a flow velocity of approximately 40 m/s at nominal temperature and atmospheric pressure. The aeroelastic wing apparatus presented in this work was initially developed by Gianikos et al. [16] and described in more detail there and in Kirschmeier et al. [17]. The apparatus (please refer Fig. 27.1) includes the main wing section, two mounting carriages, and a support structure. The main wing section consists of a decambered SD7003 airfoil with a chord length of 15 cm and a span of 60 cm. The wing was constructed from 3D printed ABS plastic with two internal support rods to improve the strength of the wing and reduce deflection along the span. The ends of the wing were capped with flat, elliptical plates of length 45 cm to reduce tip effects and allow the wing section to emulate 2D flow based on the work of Visbal and Garman [18]. The mounting carriages were placed above and below the wind tunnel test section along a sliding rail mounted to the support structure. The wing section was mounted vertically in the test section and attached to the carriages Fig. 27.1 CAD rendering of the aeroelastic wing apparatus showing the wing section and mounting carriages with key components labeled. The external support structure and linear rails are not shown
RkJQdWJsaXNoZXIy MTMzNzEzMQ==