Physical Splines for Aerodynamic Modelling of Innovative Control Effectors

Physical Splines for Aerodynamic Modelling of Innovative Control Effectors

van der Peijl, I.V.

de Visser, C.C. (mentor)

Aerospace Engineering

Control & Operations

Control & Simulation

2017-04-07

Multivariate simplex B-splines as a modelling technique provide more accurate models than standard polynomial models, which makes them an excellent modelling tool. Their coefficients, which are expressed in a barycentric coordinate system, are abstract. In this paper a physical transformation matrix is derived which can be applied to transform B-coefficients into physical coefficients. The result is a set of per-simplex polynomials in Cartesian space. Direct computation of physical coefficients is also possible, but undesired because the physical transformation matrix is ill-conditioned and causes inaccurate regression. A case is made for using Kuhn’s triangulation algorithm which is preferred for perfectly gridded data structures, instead of the more conventional Delaunay method for triangulation of the model domain. The physical spline modelling technique is applied on Lockheed Martin’s Innovative Control Effector concept aircraft. Two models are created: a large set of simplices hosting polynomials of degree one, and a smaller set of simplices with higher order polynomials. These both are analytical formulations which can be used for control allocation. The resulting models are accurate and C0 continuous, but have local areas of error which could have an undesirable effect on control allocation. The high degree model more is affected by this more than the degree one model.

Multivariate Simplex B-Splines
Physical Splines
Innovative Control Effectors
ICE
Kuhn
Triangulation
Modelling

http://resolver.tudelft.nl/uuid:7730dd69-2bb8-4047-8f6d-5681cfd04cee

2020-04-07

Student theses

master thesis

(c) 2017 van der Peijl, I.V.