penetration method (Mindess et al.). The crack profile of the air cured specimens was

similar to the crack profile of Landis and Shah (1993).

 Based on these observations it was reasonable to assume that the critical length of

the micro columns was not constant but varied through the cylinder becoming smaller

towards the center of the cylinder. The length change was more severe in the outer layers

and close to the exterior surface. The length change was significantly smaller for the

interior concrete layers (see Fig. 8-4).

 Estimated crack profile

 S\/ I A A
 50\ i / 640


70 -
011o



 4 i


 *


 S Region 2
 Region 1
 Cross-section A-A Load test set-up

Figure 8-3. Distribution of acoustic emissions for mortar beams

 Therefore, it was assumed that the combination of radial and longitudinal cracks

generated a series of micro columns with shorter lengths and smaller cross sections as

they approached the center of the cylinder (see Fig. 8-5). For this study it was assumed

that the micro column length variation was bilinear from the outside face of the cylinder

to its center. The lengths of the micro columns varied by an angle 01 from the exterior

concrete layers and by an angle 02 thereafter to the center of the cylinder (see Fig. 8-4).