Ready for Take-Off

Abstract

Atmospheric free flight scaled flight testing is an affordable way to investigate the dynamic properties of an aircraft, while enabling a wider range of test possibilities than a windtunnel. This research is the first step into the development of a scaled flight testing model and as such will highlighted the difficulties and discrepancies which will be faced when scaling is performed. This study has the objective of investigating the effect of scaling on the aerodynamic properties and of the flow field at the tail location, with the focus on take-off conditions. The full scale case is a regional aircraft, the ATR72, with an unswept and slightly tapered wing. The scaled case is 14.7% geometric scaled The analysis of the aerodynamic properties is split between the clean wing and the wing with an additional surface in the form of a flap. The aerodynamic properties investigated were the lift, moment and drag curve, and the maximum lift coefficient. The clean wing was analysed using a quasi-3D analysis called Q3D, which is a combination of a vortex-lattice method (AVL) and a vortex panel method (XFOIL). A modification of XFOIL, called RFOIL, is used to analyse the difference in maximum lift coefficient between the airfoil of the scaled and full scale case. RFOIL is selected because of the better results near stall. The wing with flap is investigated using a vortex lattice method, in this case again AVL, where the airfoil with flap is investigated using the Euler-solver MSES. Finally the maximum lift coefficient for the wing with flap is analysed with the semi-empirical Pressure Difference Rule. The Pressure Difference Rule states that there is a ratio between the peak pressure and the trailing edge pressure, on a surface, at which maximum lift occurs. The methods are limited in their incorporation of viscous effects, the 2D-analysis tools (XFOIL, RFOIL and MSES) only include viscous effects in the small boundary layer region. AVL does not include viscous effects directly, but only via a correction factor for the lift-curve slope. However, the results found in literature using these methods are satisfactory. An analysis of the wake field is done using methods developed by the ESDU. The wake properties of both the scaled as well as the full scale are calculated using semi-empirical models and use a simplification where to place the vortex sheet. The dimensions and velocity loss of the wake itself can be calculated using either a method by ESDU or a method by Schlichting. Results showed that both these methods provided similar solutions. The scaling of the wing proves to change the aerodynamic properties and the similarity is no longer present for all the investigated aerodynamic properties. For both the wing with and without flap this is the case. These differences are mainly due to Reynolds number effects, the Mach number effects are only minimal on the lift coefficient. Due to the limits on the selected methods the exact magnitude of the difference found can not be guaranteed, but the trends in the found differences are certain. A basis for this is found in the different boundary layer properties, the scaled wing exhibits a thicker boundary layer, leading to a decambering effect, and laminar separation bubbles are found to occur. Both the clean and wing with flap shows more outboard loading for the scaled wing. The scaled wing exhibits a thicker wake, with a larger velocity loss. The difference between the full and scaled clean wing is thus larger than the difference between the full and scaled wing with flap. The reason for difference between the full and scaled wing can be found in the increased drag (coefficient) of the scaled wing. The reason behind the larger increase in wake for the clean wing is due to the fact that the difference in drag coefficient between the full and scaled case is larger for the clean wing than for the wing with flap. An investigation to minimize the differences between the full and scaled wing is done. It is decided to change the airfoil shape. Improvements are visible for the optimized airfoil, on all the aerodynamic properties. The optimized airfoil shapes tends towards a thinner airfoil and thus an investigation into solely optimizing the thickness of the airfoil is also done. This also proves to give better results than the original airfoil, however, not as good as the shape optimized airfoils.. This thinner profile was also investigated in the configuration with a flap, but here it proved to decrease the performance of the airfoil with flap. The difference between the full and scaled case can be reduced by an optimization. If an optimization is to be done, it must be done on the whole of configurations and flight conditions. As the next step into a development of a scaled flight test model, the exact extend of tests must be determined, only then can it be investigated how scaling affects the results of testing. No direct solutions exist to completely overcome the gap between the full and scaled aerodynamic properties of a flight testing model, however airfoil shape optimization does provide a better similarity.