Computational modelling of the aerodyanmics and flight mechanics of a boomerang

Abstract

The returning boomerang is a singularly interesting device: when thrown spinning in a near-vertical plane, rather than travelling along a straight line, the boomerang turns continuously to the left and ultimately returns to its thrower. The return of the boomerang is caused by two types of gyroscopic precession, the result of an interaction between the lift generated by the boomerang’s arms, and its initial angular velocity.

This project aims to apply an understanding of the physics of boomerang flight, and the principles of Computational Fluid Dynamics in order to model the boomerang as a rigid body with six-degrees of freedom, within the framework of the STAR-CCM+ software.

In order to initialise the six degree-of-freedom simulation, a steady-state simulation of the boomerang, at the instant just before it leaves the thrower’s hand, is carried out. The steady-state pressure distribution reveals an imbalance in the lift produced by the two arms of the boomerang, which reinforces the gyroscopic precession of its angular velocity vector.

Using the results of the steady-state simulation, a six degree-of-freedom simulation is initialised. This simulation solves the governing equations to simulate the motion of the boomerang in response to pressure and shear forces exerted by the air surrounding it. From this, the trajectory of the boomerang is obtained and is found to be typically curved and returning. The nature of the trajectory is analysed in greater depth.

The changes in the boomerang’s orientation are also studied, and its aerodynamic loading reveals how the rapid translation and rotation of its rotor blade-like arms results in the generation of lift. Lastly, the development of vortical structures in the wake of the boomerang is studied. It is observed that the motion of air around the boomerang is influenced governed largely by the interaction of tip vortices generated and shed by its arms, with flows induced by pressure differences on and across its surfaces.

In additional to the results above, this report discusses the constraints encountered in setting up a computational model, and offers recommendations for further work.