By Mark D. Hunter, Anca C. Ferche, and Frank J. Vecchio
ACI Structural Journal Vol. 118 Issue 3
Abstract:
Stochastic simulation is used primarily as a basis for the resistance models in a reliability analysis, and it is often used to calibrate structural concrete building codes. This paper outlines the implementation of stochastic simulation techniques into a nonlinear finite element analysis framework. The stochastic modeling capabilities implemented include Monte Carlo (MC) sampling and Latin hypercube sampling for uncorrelated uniform sampling, uniform sampling with correlated random variables,
and spatial variation using random field generation.
Stochastic simulation was conducted for shear-critical beams containing no transverse reinforcement. The simulation results form the basis for a reliability analysis that computes the reliability index for the CSA A23.3-14 code. The calculated reliability index of 2.96 was lower than the target index of 3.5, indicating that the intended performance is not achieved for this type of element. As such, further investigation is required to assess the load factors and safety factors.
Keywords: finite element analysis; reinforced concrete; reliability analysis; stochastic stimulation.