Evaluation of the V-tail configuration on a machine learning and CFD based design framework

Recent works show that unconventional tail configurations like Vee-Tails might be good alternatives to conventional tails in terms of providing less weight and reduced drag in commercial airplanes. In this study, a conventional tail configuration consisted of an horizontal and a vertical stabilizer has been replaced by a Vee-Tail. The baseline airplane is the CSR-01, which is an hypothetical aircraft from the CeRAS repository. Various sizing cases for the Vee-Tail configuration have been considered. The sizing variables have been selected as the following 4 geometric parameters that define the Vee-Tail: The span length, the root chord, the taper ratio and the dihedral angle. The flows over the whole airplane have been computed using a Reynolds-Averaged Navier-Stokes solver. The Spalart-Allmaras turbulence model has been employed to compute the eddy viscosity. All the flow computations have been done in a parallel computation environment. Once the flows have been computed, the aerodynamic performance of the Vee-Tail has been established using the Support Vector Regression, which is a machine learning algorithm. A Vee-Tail geometry has been determined matching the lift, drag and pitching characteristics of the original configuration as close as possible.