Study on Mechanical Properties of Corroded Reinforced Concrete Using Support Vector Machines

The mechanical properties of corroded reinforced concrete under repeated load are investigated. The maximum crack width, mid-span deflection and reduction factor are predicted by using support vector machines. The maximum crack width and deflection are predicted by the black-box modeling based on support vector machines with the radial basis function kernel function. The reduction factor is predicted by using piecewise regression formula, whole regression formula and black-box modeling, respectively. The proposed prediction method is verified by comparing all prediction results with the experimental values. It is shown that the proposed method has high prediction accuracy, extensive applicable range and many predictive strategies.