Title: Probabilistic analysis of a pull-out test

Author(s): J. Humbert, J. Baroth, L. Daudeville

Publication: RILEM

Volume: 43

Issue: 3

Appears on pages(s): 345-355

Keywords: Pull-out test, Failure modes, Stochastic finite element method, Monte Carlo simulation, Probabilistic sensitivity analysis, Nonlinear damage mechanics

DOI:

Date: 3/2/2010

Abstract:
This paper presents a sensitivity analysis of the pull-out strength of reinforcement embedded in concrete. Considering both European and French design codes, this failure strength depends on the variability of uncertain parameters such as Young’s modulus of concrete and yield stresses of materials (concrete and steel); moreover, two failure modes can be observed in the studied experimental test. A methodology allowing the characterization of the sensitivity of the pull-out strength to these uncertain parameters is derived. These parameters are modeled by Lognormal random variables. Results show the evolution of the pull-out strength for different anchorage lengths. Probability density functions of the random variable modeling the failure strength are computed using probabilistic methods. A finite element model is also built to quantify uncertainties concerning failure modes, computing 95% confidence intervals.