Computational Modelling Group

Evaluation of Vortex Shedding effects on Slender Structures using Large-Eddy Simulation

3rd October 2011
5th April 2015
Research Team
Steven Daniels
Zheng-Tong Xie, Ian Castro

Periodic vortex shedding over a rectangular bridge deck

Vortex shedding is a critical design consideration for slender structures such as long-span bridges and high-rise buildings. It can create peak loadings over the surface of the structure, leading to serviceability and fatigue concerns, and is also considered to be the base-mechanism behind some of the familiar structural failures in the last century (e.g. Tacoma Narrows Bridge, Ferrbridge cooling towers). Currently, in industry, analysis of the phenomenon is mainly conducted in wind tunnel laboratories. Numerical approaches are increasingly seen as a viable theoretical alternative, though the long duration of the simulations can make them challenging to deploy effectively within the design cycle.

With the burgeoning growth in computational capability, Computational Fluid Dynamics (CFD) is becoming a powerful tool for the analysis of wind-induced effects of structures. The main issue with this method is the simulation of the turbulence, which is a main characteristic of wind. Wind tunnel tests are somewhat limited in the controlling turbulence in their analysis, provoking an interest in the use of computational methods instead. Currently, the justification of ‘tight-budget’ computing in industry has resulted in the widespread use of statistical-based techniques for modelling turbulence, namely Unsteady Reynolds-Averaged Navier-Stokes (URANS). This approach however is known to be ineffective in simulating the true nature of vortex shedding, and as a result misses key aspects of the flow around the structure. The innovative aspects of this study are based around the use of Large-Eddy Simulation (LES), which from our earlier works has shown to produce a better representation of atmospheric turbulence and the resulting wind loading effects; this is in comparison to URANS. However, LES requires the use of a supercomputer, which is not always at the convenience of industry. With this issue in mind, this research also proposes novel techniques to simulate wind-structure interaction in an economical and efficient manner.


Physical Systems and Engineering simulation: CFD, Structural dynamics, Turbulence

Simulation software: OpenFOAM

Programming languages and libraries: C++

Computational platforms: Linux