Distributed computing for design optimization of the Messina Bridge considering aeroelastic constraints F. Nieto a , S. Hernández a , J.Á. Jurado a a University of A Coruña, School of Civil Engineering, Campus of Elviña, A Coruña, Spain ABSTRACT: Optimization techniques are a very useful tool for engineers when dealing with the design of long span bridges. In this paper the formulation of the optimization problem consider- ing aeroelastic constraints is shown. The optimum design process is a very much computer time demanding task, so distributed computing must be applied. Three examples of optimum design of the future Messina Strait Bridge are presented, where the thicknesses of the cross-sections of each box of the deck are adopted as design variables. It can be seen that very important material savings can be reached while satisfying the considered design constraints. KEYWORDS: Flutter, optimum design, sensitivity analysis, distributed computing, Messina Bridge. 1 INTRODUCTION Mankind is demanding nowadays very long cable supported bridges to link different regions, countries, continents and even cultures. Because of that, engineers must do their best in order to satisfy the technical requirements of such challenging structures. Non-conventional design tech- niques, as optimum design or sensitivity analyses, have proved to be a very useful tool to im- prove designs in aeronautics or car industry. Researchers as Jurado and Hernández [1] or Mosquera et al. [2] have shown the capabilities of the sensitivity analyses of the aeroelastic re- sponse of long span bridges with regard to the mechanical parameters of its deck in order to an- ticipate their aeroelastic performance when modifying any of the design variables. Optimization techniques are an even more powerful tool that allow designers to avoid tedious operations and improve the final design when projecting cable supported bridges. 2 CLASSICAL DESIGN AND OPTIMUM DESIGN Classical design of suspension and cable-stayed bridges has been for many years based upon the experience and expertise of engineers with the aim of obtaining an acceptable design. In that case, designers apply heuristic rules in order to get a modified design that should perform better than the previous one. However, that circumstance is not guaranteed, so the design procedure becomes a trial and error process. There are alternative approaches, such as the design based upon sensitivity analysis or optimum design [3]. Optimization is the process of maximizing or minimizing a desired objective function while satisfying a series of design and behavior con- straints. To obtain that, mathematical algorithms are linked to structural analysis codes to set up an iterative procedure that finally converges to the best solution. In the present work, the weight of the Messina Bridge deck has been minimized while keeping the flutter speed over a required value, simultaneously the vertical movement of the center of the main span has been lower than a stipulated value (see Figure 1). 757