This presentation was held at the 2020 NAFEMS UK Conference "Inspiring Innovation through Engineering Simulation". The conference covered topics ranging from traditional FEA and CFD, to new and emerging areas including artificial intelligence, machine learning and EDA.

Resource Abstract

The advent of distributed propulsion system for a greener aviation has led to novel aircraft concepts with an integrated airframe / propulsion system, such as the tilt wing Electric Vertical Take-Off/Landing (EVTOL) concept, in which a number propulsors are attached to the wing. While the cruise configuration is effectively identical to conventional aircraft, the vertical configuration with the wing tilted upward requires a different system.



In hover and vertical flight, aircraft control can be achieved by varying the propulsor thrusts that produce most of the lift to augment the conventional control surfaces. A relatively simple blown wing model that rely on classical linear aerodynamic could be used for modelling the flight mechanics. The flow over the wing can remain attached, even if the wing is set at perpendicular angle relative to the fuselage axis, when the propulsor slipstream, which is aligned with the wing, is the most dominant flow, such as the case of hovering in a calm atmosphere. However, the accuracy of such model degrades significantly when the EVTOL is hovering in a strong cross flow condition since flow separation occurs over a portion of the wing due to incoming wind that arrives at high incidence relative to the wing exerting comparable influence to the propulsor slipstream.



An improved flight mechanics model of hovering EVTOL in cross flow can be developed using RANS (Reynolds Averaged Navier-Stokes) CFD. However, CFD computation over the whole 3D aircraft model, even with simplified tilt wing geometry, is too expensive for a practical flight mechanics model. Hence a surrogate CFD model with significantly lower computational cost is desirable. In this work, a Reduced Order Model (ROM) that is based upon Singular Value Decomposition of a 3D RANS CFD field is being investigated with encouraging initial results.