Underlying Design Trend Evaluation of Fixed-Wing Remotely Piloted Aircraft (RPA) Through Factor Analysis

Remotely Piloted Aircraft Systems (RPAS) are experiencing a noteworthy development and expansion in the field of aeronautics. The flexibility and range of possibilities that they offer with respect to missions, aerodynamic layouts, and powerplant, among other aspects, result in an expansive and diverse design space. Within the first phases of the aircraft synthesis process, particularly in the conceptual design stage, it is key to explore and manage the space of potential designs effectively, so as to arrive at a suitable initial sizing estimation which can be used as a starting point for further refinement. In order to tackle this task, knowledge of the main design trends for these aircraft, which are notably different from that of manned aircraft, constitutes an invaluable asset for the designer. In this research, the multivariate statistical methodology known as factor analysis is employed, with this approach being based on the assumption that the changes in the observable variables that are tangible to the designer are mainly determined by a set of latent, unobservable factors. These could be interpreted as underlying design trends that describe the changes seen in the observable variables. Factor analysis enables the identification of this set of latent factors, as well the formal quantitative assessment of their statistical significance and relationships with the observable variables. In this work, this technique is applied to the field of RPA design in order to extract the main design trends governing the observed variance and correlation among mass, geometry, and performance variables at a conceptual level. In particular, this methodology is applied to aircraft with conventional tail, T-tail and V-tail layouts, which are widespread among the current landscape of RPA designs. The main challenges resulting from the use of this multivariate technique to RPA design are examined and discussed, with this assessment being key to set up the analysis cases conducted in this research. The use of different tail types to categorize the data and check for differences among the statistical results, and also in the underlying trends, constitutes one of the original approaches of this study. The results from this research allow to derive various conclusions with respect to the ways in which both observable and underlying RPA design trends may change depending on overall layout and powerplant.