Investigation of a modified low-drag body for an alternative wing-body-tail configuration

Abstract

A wing-body-tail (WBT) configuration has been proposed without the conventional tailplane, allowing for new design objectives for the aerodynamic shape of the fuselage. A shorter, low-drag body (LDB) can be employed with a lower structural mass and lower drag per unit volume. There is an additional possibility that a carefully-designed deflector flap (Kutta edge or KE) can modify the flow around the body to allow for a more uniform circulation between the body and the wing. This basic idea has been tested and supported in wind tunnel tests but there have been no systematic attempts to quantify or explore the design space for these WBT geometries. All prior investigations were experimental on the full configuration at a low Reynolds number (Re) range where the aerodynamics of simple, rigid, fixed wings becomes extraordinarily sensitive to small changes in geometry and the environment. Therefore, the purpose of this study was to first investigate a simple case of the NACA0012 airfoil and wing, which becomes non-trivial when making baseline comparisons for experimentation and computation purposes. A diagnostic procedure then guides comparisons and predictions in subsequent more complex cases. In a preliminary attempt to bring specifically-designed low-drag bodies (LDB) into the design space, a numerical study was conducted to investigate the effect of adding a KE to two LDBs whose properties have been well established in the technical literature. Initial experimental tests on KE deflection were always accompanied by aftbody deflection, and the same procedure was adopted in the numerical studies, so tail deflection was achieved through deflection of the entire aftbody. It was further noted that such measures were quite sensitive to details of separation over the body and tail and that, paradoxically, a preferred arrangement would be to locate the KE entirely within the bounds of the viscous wake. Finally, the notion that the body-tail combination can be used in conjunction with the wing to improve the net downwash of the configuration was considered. The two bodies used previously with the KE were fitted with an NACA0012 wing at a fixed angle of attack of 6°. The overall purpose of this study was to provide some insight into the differences between the previous experimental and numerical studies and to investigate a potential design space of the initially tested WBT configurations. Then further exploration of the design space was conducted by using 10 discrete WBT configurations (five deflection angles with and without KE) for the original experimental WBT at chord-based, Rec = 105 and two LDBs at their specific design length-based, Rel (1.2 x 106 for F-57 LDB and 107 for Myring LDB). This work confirms that the KE can influence the WBT wake structure as initially estimated in experiment. The results here suggest that the original KE concept requires careful matching in the reality of viscous flows over bodies and wings at finite Re. In particular, if the KE is wholly or partially immersed in a wake that derives from earlier upstream separation or wing-induced wake effects, then the KE cannot operate effectively and the body termination conditions must already be judged to be sub-optimal. Adding the trailing-edge increased the total drag coefficient and the expected improvement of induced drag did not lead to a net benefit. If there is an optimal WBT configuration that leads to significant benefits in lift-to-drag, L/D, then it presumably would have to live in a domain where separation is almost completely avoided. The second modifying consideration is that if an entire system is designed for a certain lifting objective, then the option of providing that weight support through a modified geometry that includes a KE might not be well described by a single number such as L/D. When system benefits of reduced wing length, area and weight are included (and subsequently fed back into the new design set-point for the lift coefficient), the interlocked design benefits of each component might be difficult to isolate. The properties of the idealized WBT project with real bodies at finite Re are not easy to predict, this work offers useful guidelines and a helpful start, with recommendations for further numerical and experimental investigations.