Multilevel Computational Model for Cost and Carbon Optimisation of Reinforced Concrete Floor Systems

The cost and carbon efficiency of building structures could be enhanced by the current developments in design automation and optimisation techniques. The study focuses on a common structural system seen in the majority of mid- and high-rise buildings in the UK: flat slabs with reinforced concrete columns. A multilevel optimisation approach is established combining BIM data and FEM with a genetic algorithm offering engineers new ways to systematically assess structural design alternatives based on cost and carbon parameters. The optimisation is performed in three main levels: 1) Structural Grids, 2) Design Elements, 3) Detailing. A constrained genetic algorithm is applied to find structural design alternatives that minimise the cost and carbon functions in the structure. A prototypical building floor is analysed. Structural grid configurations, floor thicknesses and columns sizes and reinforcement details are specified. The cost and carbon performance for the obtained combinations are assessed using the genetic algorithm component. For the tested floor system the results showed that the cost optimum design is 3% cheaper than the carbon optimum design but it has 7% more carbon. In addition, the concrete in the floor is the biggest contributor in both total cost and carbon. Relationships between cost- and carbon- optimum designs for the tested structural configuration are also discussed.