Sizeeffectinsteel-concretebond:testresultsandmodellingforsmoothbars

Size effects are studied here with reference to the bonding of smooth bars to both ordinary and high-performance concrete (NSC and HPC). To this purpose, 24 moderately-long anchorages (Udb = 10), consisting of an artificially-roughened bar embedded in a concrete cylinder, were cast and tested up to the pull-out of the bars, which had 4 different diameters (db = 5, 12, 18 and 26 mm). For each of the 8 cases examined here (4 diameters x 2 mixes), 3 nominally-identical specimens were tested. Furthermore, to have some information on the roles of specimen supports and bar roughness, 3 more specimens were cast and ging the specimen number to 27. Though the primary objective of this study is to investigate tested, brin whether a general-type size-effect law applies to bond in high-performance silica-fume concrete, the of an anchorage by means of a local elastic-fracturing-frictional bond-slip law is also carried out. modelling l INTRODUCTION is the most important aspect of every physical theory". Consequently, scaling concerns also bar- Size effects in the structures made of quasi-brittle concrete bond, even more since the bond stress-slip materials have been one of the most cherished topics law of a smooth bar exhibits a strong softening. In in the last twenty years of the second millennium, such a case however, the bar-concrete interface and many experimental and theoretical studies have should be considered as a sort of preoriented crack been carried out on a variety of plain-concrete and ("interfacial crack", length "a", Fig.2a, Stang et al., RIC structures, such as shear- and torsion-critical 1990), provided that "cracking" and "fracture members, deep beams, slabs, fastenings and splices process zone" are replaced with "debonding" and (Bazant and Chen, 1997; Ozbolt and Eligehausen, "u·ansitional zone", the latter meaning that '"hazy" 1996).