Performance-based design of FRP retrofitting of existing RC frames by means of multi-objective optimisation

Safeguard of built heritage often involves seismic retrofitting of reinforced concrete (RC) frame structures. When strength and ductility of columns must be enhanced, an effective technique is given by the use of Fibre-Reinforced Polymer (FRP) in forms of wraps glued to the columns, which are able to improve performance thanks to confinement action. In this paper, an optimisation-based procedure for the design of FRP retrofitting of existing RC frames is described. The design aims at finding the most competitive solution in terms of cost and performance, while satisfying the damage levels imposed by Performance-Based Design for serviceability and ultimate seismic hazard levels. The resulting multi-objective optimisation problem, in which the design variables are represented by the thickness of FRP wraps, is then solved by means of Genetic Algorithms. The application to a realistic case study shows how the analysis of the resulting Pareto Front, i.e., the set of non-dominated solutions, clearly describes the threshold between cost and performance. Furthermore, interesting considerations about the sensitivity of the best solutions to the design variables can be made, improving the assessment of the optimisation results.