Optimum design of reinforced concrete columns employing teaching-learning based optimization

In structural engineering, the design of reinforced concrete (RC) structures needs an initial de-sign for cross sectional dimensions. After these dimensions are defined, the design constraints and the required reinforcement bars are calculated. But the required reinforcement area is not exactly provided since the size of rebars are fixed. At the end of the design, the security measures are provided, but the designer has no idea for the optimization of the design in mean of economy. For that reason, a powerful search methodology can be programed by using metaheuristic algorithms. In this study, optimum design of reinforced concrete columns was investigated by using an education based metaheuristic algorithm called teaching-learning based optimization (TLBO). In the methodology, the slenderness of the columns is also taken into consideration by using a simple approach given in the ACI 318 design code. In this approach, the factored design flexural moments are defined according to the buckling load and axial load of columns. The design variables of the problem include cross section dimension of the column and the detailed reinforcement design and the optimization objective is the minimization the maximum material cost of the column. Differently from the other metaheuristic algorithms, the decision of the optimization type (global or local search) is not defined by using a probability parameter in TLBO. In optimization, two phases of TLBO; teacher (global search) and learner (local search) phases are consequently applied in search of best design variables. The proposed approach is effective for the structural optimization problem