Abstract

Windborne debris can be described as failed building components, loose items and street furniture that are picked up and carried by the wind during severe storms. Studies of building performance during storms have identified windborne debris as the principle cause of building failure and as an essential component of the debris damage chain. The present work focuses on plate type debris – roofing sheets, shingles and tiles – the dominant type in a residential setting.

Current debris risk and damage models often include analytical models of debris flight, based around six degree-of-freedom (6DOF) models. Many of these models assume aerodynamic coefficients which are obtained from time-averaged wind tunnel measurements on static, inclined plates. These coefficients are then modified to take into account unsteady and non-linear, fluid-structure interactions by means of various empirical coefficients. While successful in certain circumstances, these quasi-steady coefficients produce inaccuracies in the analytical models.

The present work uses a 3D Computational Fluid Dynamics (CFD) model, which is sequentially coupled with a 6DOF rigid body dynamics (RBD) model to simulate the flight of plates in high Reynolds number flow. The idea, using this computationally demanding approach, is to better understand the mechanics of the flight and to derive more reliable aerodynamic coefficients for use in simpler models.

The commercial CFD software, ANSYS -Fluent, has been used throughout but the default RDB model, based on the Euler rotational matrix approach, has been replaced by a singularity-free method based on rotational quaternions. The model has been used to investigate the influence of plate initial orientation on the flight trajectories. Different flight modes are identified, depending on the initial conditions, with a clear transition between different modes being demonstrated.