Numerical modeling of chloride penetration experiments considering concrete heterogeneity

Abstract

The thesis presents the study, enhancement, verification and application of a 2D Finite Element Method (FEM) model for the numerical modeling of chloride penetration experiments. The targeted advanced model is expected to have capable of precise capturing chloride penetration to reinforced concrete structures considering concrete heterogeneity and the ability to reproduce chloride ingress experiments of modern scanning techniques such as Laser-induced Breakdown Spectroscopy (LIBS). An existing deterministic 2D FEM diffusion based model was adopted and extended by the inclusion of the application of random field via the correlation length for the description of the material resistance to chloride penetration including the probability based approach through the histograms of input parameters for other studied parameters. The advanced model was then verified and validated using data from LIBS test and through the comparison with analytical probabilistic model. Parametric study of the model was carried out through numerical experiments to search for a set of appropriate and reliable parameters for the model for its application to the chloride profile sampling. Results showed that the proposed model can address the heterogeneity of concrete in chloride penetration experiments to some extent and it is capable of reproducing such experiments of LIBS at meso scale level better than probabilistic analytical formulation of the model. Parametric study utilizing the advanced model for simulation of NT Build 443 based chloride profile sampling was performed in order to study the effect of laboratory test set ups such as penetration time and profile layers’ thickness. Several important conclusions were drawn for the model usage and application. Recommendation for laboratory experiments were also drawn, e.g. the necessity of significantly longer chloride penetration time for the high performance concrete (HPC) comparing to the ordinary Portland cement (OPC) one. Directions for the continuation of the research topic were also suggested.

Description

Subject(s)

Chloride penetration, correlation length, experiments, FEM, heterogeneity, LIBS, probabilistic approach, random field, reinforced concrete.

Citation