Summary In order to evaluate analytically the ITZ volume fraction (fITZ) in concrete, a three phase model is proposed for the random concrete microstructure using the Voronoï tessellation. Within this model, the ITZ local thickness is a statistical variable depending on the local paste thickness available between each couple of neighbouring aggregates. The fITZ is found to not exceed 7% for typical concretes. Then, the concrete Young's modulus is predicted analytically using a four‐phase generalized self consistent model but in which the proposed fITZ is considered. It is found that the concrete Young's modulus increases when increasing aggregates volume fraction, aggregates maximum size and the proportion of coarse aggregates and when decreasing the ITZ thickness and Young's modulus. Finally, the validity of the proposed model is discussed based on a comparison between its predictions and three sets of experimental results related to normal and high strength concretes taken from literature. Copyright © 2016 John Wiley & Sons, Ltd.

The main purpose of this paper is to predict analytically the concrete basic creep time-dependent deformation according to its composition and microstructure. Thus, concrete was modelled as a three-phase composite formed by aggregates surrounded by interfacial transition zones (ITZs) and embedded randomly in a hardened cement paste. First, the ITZ phase volume fraction within concrete was evaluated using an analytical formula that the authors had recently proposed (Zouaoui et al. 2017). Then, the specific basic creep deformation of concrete was evaluated analytically based on a four-sphere homogenization model where cement paste and ITZ have a linear viscoelastic behaviour whereas aggregates are assumed linear elastic. The predictions of the proposed model show that cement paste basic creep and aggregates volume fraction and maximum packing density are the main parameters governing concrete basic creep. Moreover, these predictions show that aggregates gradation and ITZ thickness have secondary effects on concrete basic creep. Finally, the relevance and the validity of the proposed model have been discussed based on a comparison between its predictions and experimental results taken from literature and related to both normal and high-strength concretes.