Multi-Scale Microstructural Modelling of Concrete Diffusivity: Identification of Significant Variable, Cement, Concrete and Aggregates, 1998 Reference: D.P. Bentz, E.J. Garboczi, and E.S. Lagergren, Cement, Concrete, and Aggregates, 20 (1), 129-139 (1998).
PDF Version of Original Paper

Go back to Part I Chapter 7 Sec. (3a) Go back to Table of Contents


Multi-Scale Microstructural Modelling of Concrete Diffusivity: Identification of Significant Variables

Dale P. Bentz and Edward J. Garboczi
Building and Fire Research Laboratory
National Institute of Standards and Technology
Gaithersburg, MD 20899 USA

Eric S. Lagergren
Information Technology Laboratory
National Institute of Standards and Technology
Gaithersburg, MD 20899 USA

Abstract

The ability to predict the expected chloride diffusivity of a concrete based on its mixture proportions and field-curing conditions would be of great benefit in both predicting service life of the concrete and in developing durability-based design codes. Here, a multi-scale microstructural computer model is applied to computing the chloride diffusivities of concretes with various mixture proportions and projected degrees of hydration. A fractional factorial experimental design has been implemented to study the effects in the model of seven major variables: water-to-cement (w/c) ratio, degree of hydration, aggregate volume fraction, coarse aggregate particle size distribution, fine aggregate particle size distribution, interfacial zone thickness, and air content. Based on this experimental design, w/c ratio, degree of hydration, and aggregate volume fraction have been identified as the three major variables influencing concrete diffusivity in the model. Following identification of the significant variables, a response surface design has been executed and least squares regression utilized to develop an equation for predicting chloride ion diffusivity in concrete based on these three parameters. This simple equation essentially summarizes the complicated simulations involved in computing the model response. Since degree of hydration is not typically measured in experimental diffusivity studies, further simulations have been performed to examine the efficiency of different spatial sampling techniques in providing a representative elementary volume of concrete for analysis. Finally, simulations have been conducted to examine the extent of the surface layer in cast-in-place concrete, where the local aggregate volume fraction near the surface is less than that to be found in the bulk of the concrete.

Keywords: concrete diffusivity, durability, experimental design, interfacial transition zone, microstructure, modelling, performance prediction, spatial sampling.




Go back to Part I Chapter 7 Sec. (3a) Go back to Table of Contents