Probabilistic analysis of reinforced concrete columns

Abstract

The subject of this work is the probabilistic finite element analysis of reinforced concrete columns. Concrete properties are represented as homogeneous Gaussian random fields. The yield stress and position of steel reinforcement, dimensions of the column cross-section and axial load are considered as random variables. The Monte Carlo method is employed to obtain expected values and standard deviations of the rupture load. The partial safety factors method is used for columns design and structural safety is evaluated by means of the reliability index, which is obtained through simulations. The effects of main parameters on the reliability index are investigated. It is shown that the correlation length of random fields for concrete properties may have a significant effect on reliability. Therefore, simplified procedures, which do not consider spatial variations of concrete properties are inappropriate for safety analysis.