[home] [Personal Program] [Help]

---

Session: Nonlinear optimisation
Session starts: Wednesday 19 June, 16:00
Presentation starts: 16:30
Room: Room 1.4/1.5

---

Francesco M. A. Mitrotta (University of Bristol)
Alberto Pirrera (University of Bristol)
Terence Macquart (University of Bristol)
Jonathan E. Cooper (University of Bristol)
Alex Pereira do Prado (Embraer S. A.)
Pedro Higino Cabral (Embraer S. A.)

Abstract:
The aviation industry's challenge to achieve net-zero carbon emissions by 2050 demands innovative approaches to aircraft design, particularly through the development of lighter, more fuel-efficient aircraft structures. Aeroelastic optimization plays a crucial role in this process by enabling exploration within a design space bounded by numerous constraints, including structural strength and stability. Traditional aeroelastic optimizations rely on linear buckling analyses which, while computationally efficient, tend to over-constrain the design space due to their conservative nature. Such conservatism places a glass ceiling on the sizing loads used within optimization studies, preventing the exploitation of the full load-bearing capacity, especially as aircraft designs trend towards more slender and compliant wing structures where nonlinear effects become significant. This paper builds on previous research by integrating a nonlinear structural stability constraint into the optimization of a wingbox structure under a distributed load. Three optimization scenarios are considered to evaluate the influence of different load introduction methods: distributed non-follower forces, distributed follower forces, and non-follower forces applied to a load reference axis. In addition, the formulation of the nonlinear structural stability constraint is revisited to allow its application in presence of follower forces. Our findings, demonstrated on an idealized version of the Common Research Model wingbox, confirm a substantial mass reduction using the nonlinear constraints compared to the traditional linear ones, ranging between 8 and 9%. Not much difference is found between the structures optimized with distributed non-follower and follower forces, because of their small deflections. Instead, a noticeable difference is found for the structure optimized with non-follower forces applied to the load reference axis, which achieves a smaller mass reduction. If on one hand these results suggest that employing the load reference axis approach might lead to conservative results, on the other hand we show that such approach leads to an inaccurate prediction of the structural deformation.