Scheme: Industry Fellowship
Organisation: Mott MacDonald
Dates: Nov 2006-Sep 2010
Summary: The design of long-span bridges becomes ever more ambitious, as demonstrated by the recently opened Viaduc de Millau in France and the proposed crossing of the Messina Straits in Italy, and each new design stretches the available technology further. A key limiting factor in the performance of such slender structures is their tendency to move under the influence of aerodynamic forces in a manner governed by a complex interplay of the wind and structure. This interaction was famously demonstrated in the wind-induced destruction of the Tacoma Narrows bridge in the USA in 1940 which acted as a wake up call to the industry and underlined the importance of wind engineering as a discipline. Such wind effects are not limited to bridges – they can also be significant for other landmark structures such as the arch of the new Wembley stadium and the Spinnaker Tower in Portsmouth.
There is strong interest in the use of computer simulation methods to assess the aerodynamic performance of bridges and other slender structures. Though computational tools for simulating the dynamics of structures themselves are now quite advanced and are regularly applied on real design projects, tools for simulating the air flow and the crucial fluid/structure interaction have yet to enter routine industrial practice. For this reason, information concerning the aerodynamic performance of the structure is at present derived almost exclusively from various forms of wind-tunnel simulations which are expensive and carried out at much less than full scale. This significantly limits the possibilities for parametric optimisation during the early design stages.
Though the simulation of bridge aerodynamics has been the focus of considerable academic research over the last ten to fifteen years, with the publication of quite promising results, it is notable that very little of this research has so far reached industry or significantly influenced the design process for new bridges or other structures. Some use has been made of the discrete-vortex method though this has significant limitations. The central goal of the present project is to close this gap and take the simulation of fluid-structure interaction out of the realm of university research and into the design office as a practical tool.