Let us first consider sorption during initial exposure to water. Figure 6 shows the absorbed water vs (time)1/2 for a period of about one hour (for the two concrete mixes). In nearly all cases, we found that the total water absorbed was proportional to t1/2 for a period of a few hours. Note that the higher porosity mix is subject to very strong surface effects as the sorption data is offset from zero when t=0. After a period of about 6 hours, the rate of sorption began to noticeably decrease.
FIG. 6 Early time behavior (up to 1 hour) of sorption for different curing periods and concrete design mixtures. All specimens were bench dried before exposure to water.
Figure 7 compares sorption data for an exposure period of over 200 days for the oven dried HPM and ASTM mortars. Here, the sorption data is plotted versus t1/2 . Note that both the short and long time moisture sorptions for the two mortars are consistent with a t1/2 behavior but that the sorptivity, S, (slope of the curve) is dramatically different in these regimes, differing by factors of about 30 and 80 respectively. Since S ~ r1/2 for a straight tube model, such a difference in sorptivity would correspond to pore sizes differing by a factor of about 900 and 6400 for the HPM and ASTM mortars respectively. Similarly, concrete is noted as having capillary pores of the order of microns and gel pores which may be several orders of magnitude smaller. Simple comparison with sorption theory would then imply that the capillary pores are dominating the sorption process at early times and that at later stages the gel pores limit the rate of flow. However, it is quite possible that, at later stages, the ingress of water may be controlled by moisture diffusion processes  as well.
FIG. 7. Sorption versus square root of days for mortars. The specimens were oven dried until constant weight, just after demolding (1 day curing).