Virtual reliability control of precast prestressed concrete elements using parameters identification and statistical simulation

An approach for statistical and reliability analyses of concrete structures is presented. It describes the virtual simulation used on the way from assessment of experimental results to reliability analysis. The whole approach is based on randomization of nonlinear fracture mechanics finite element analysis of concrete structures. Efficient techniques of both nonlinear numerical analysis of concrete structures and stochastic simulation methods of Monte Carlo type have been combined in order to offer an advanced tool for assessment of realistic behaviour of concrete structures. The stochastic response requires repeated analyses of the structure with stochastic input parameters, which reflects randomness and uncertainties in the input values. The procedure utilizes statistical simulation of Monte Carlo type also for preparation of training set of artificial neural network used for fracture mechanical parameters identification. The application of approach to virtual statistical simulation of shear capacity of precast prestressed concrete elements SPIROLL is presented. Design shear capacities calculated according to design codes and alternative procedures are compared with simulation of reality representing virtual reliability control of elements production.