MULTI-FIDELITY WING MASS ESTIMATIONS BASED ON A CENTRAL MODEL APPROACH

Abstract

Although computational power is constantly increasing and Moore’s Law is still not falsified, computational cost remains an essential barrier in aircraft design especially when a high number of design evaluations is necessary. This is especially true at the conceptual design stage of aircraft. While determining the characteristics of a new configuration the number of iterations and the low level of detail in the available data limit the analyses to simple empiric methods. Nevertheless, at a later point in the design it is necessary to determine parameters like the wing mass with higher-fidelity analysis modules. Especially when assessing configurations that lie outside of the well-known design space of conceptual design, empiric methods become unreliable. Examples to name include high aspect ratio and forward-swept wings. In this study a combination of an empiric method, a beam model and vortex lattice model for aerodynamic loads is introduced. While multi-fidelity approaches are already well known, this study focuses on the fact that all analysis modules derive their data from the same data model. Working on a central data model decreases the number of required interfaces and guarantees that all models relate to the same input data, i.e. a compliant geometry definition. This paper includes a design chain starting from the conceptual design tool VAMPzero as initiator for the more advanced models PESTwing and TRIMvl. Using the PESTwing tool a large design space will be explored. An equation for determination of the wing mass based on a physical model is then derived and compared to existing methods in conceptual design.