This study aims at improving the analysis of drying tests and taking better advantage of the measurements, by studying the influence of the specimen size. One self-consolidating concrete (SCC) was studied during three years. Drying started 24 hours after casting. The tests were carried on three sizes of cylinders: Φ78mm, Φ113mm and Φ163mm, under the same experimental conditions (20±1°C, 50±5% relative humidity). Total and autogenous shrinkages and mass-loss evolutions were monitored according to RILEM recommendations [1].
Assuming that concrete is a homogeneous and non-aging material, a simple model can be written and used to highlight size effects observed in real concrete. By changing the time variable to the ratio of the square root of the time to the notional size of cross-section, the drying curves of different specimen sizes theoretically follow a master curve [2]. However, experimental curves showed upshifts, thus size effect can be highlighted. Phenomena that occur at early age are actually not taken into account by diffusion equation.
This paper introduces a new concept called drying depth as a layer where water loss occurs during the first days without resulting in shrinkage at the specimen level. This depth did not depend of the specimen size. However the relative volume of this layer increases as the size of the specimen decreases, which explains the size effect experimentally observed. The hydric properties of this zone can be assumed to be different from the inner concrete, with a higher porosity and higher permeability. Several phenomena could explain this effect, like the initial particle size distribution with a higher paste proportion near the surface due to the wall effect during casting [3] [4] [5]. Because of early drying hydration stops and this creates a coarser pores in the external layer. Finally, the micro-cracking at the surface induced by the differential shrinkage can also influence the hydric parameters.