Extruded aluminum decks for pedestrian bridges Design optimization using genetic algorithm

Abstract

Material efficiency in bridge structures is an important research topic to reduce the climate footprint and initial costs. One way of increasing material efficiency is to use appropriate materials where their properties benefit the structure. Aluminium offers a high stiffness-to-weight ratio, high durability, and is to a large degree re cyclable. These properties make aluminium an interesting material for bridge deck applications. The initial cost of the bridge deck has high priority amongst bridge authorities in Europe. To minimize the initial cost, one target is to minimize material con sumption. This thesis aims to develop an optimization procedure with the objective to minimize the material consumption of extruded aluminium profiles for pedestrian bridge deck applications. In the design of the deck, requirements stated in Eurocode are followed. The optimization is made using a genetic algorithm function from the global optimization toolbox in the software MATLAB. Cross-sectional geometries generated by the optimization procedure was evaluated separately by the imple mentation of a FE-module. The FE-module is controlled by parameterized Python scripting to create a FE-model and execute an analysis in the software ABAQUS CAE for each iteration. The optimization provided a cross-sectional geometry for the bridge deck. These results were used for a cost comparison between the optimized cross-section and more conventional alternatives in C-Mn and duplex steel. The comparison showed that the weight per square meter of the resulting optimized profile was significantly lower compared to the deck alternatives in steel. It also showed that saving from 8% up to 27% can be made on the initial investment cost if choosing aluminium instead of a conventional deck alternative in stainless steel. Thus, aluminium is shown to be a potential alternative to steel.