Abstract
The results of the present study are summarized as follows : 1) The statistical distributions of the strength of cement-based composite materials can not be necessarily approximated by Weibull distribution, because the relation between ln (-ln(1-P)) and ln (F) can be generally expressed by a convex curve. Pareto distribution can hardly express the lower and the upper portions of statistical distribution curve. 2) The experimental results can be expressed more exactly by the proposed equation for the statistical distribution of strength than by previous equations, and the statistical properties of the strength of cement-based composite materials can be rationally explained by the proposed equations. 3) The value of β is affected by the type of loading, and the sizes of specimen and aggregate in macroscopic heterogeneous material, but is independent of the other factors. The value of β can be expressed by Eq. (11). 4) The value of parameter related to the quantity of defects in the mortar specimen (θ_m) increases with the increase of the sand-cement ratio (S/C), provided that the value of S/C is smaller than about 2, and can be expressed by Eq. (14). 5) The value of parameter related to the quantity of defects in the concrete specimen (θ_c) decreases with the increasing value of size of aggregate, the decreasing value of size of specimen and the increasing value of volume fraction of coarse aggregate, independently of the shape of specimen and the type of aggregate. These phenomena can be expressed by Eq. (16) and Eq. (17).
