Early attempts in the 1950s to develop rational theories of reinforced concrete in shear neglected the role played by aggregate interlock. Both implicit and explicit in these early theories was the assumption that all the vertical shear force in concrete sections without transverse reinforcement is carried in the uncracked concrete compression zone. This represented a reasonable first approximation of the complex behaviour of these element types. As research progressed, however, a belief gradually emerged that significant shear stress may, in fact, be transferred through the cracked web of a reinforced concrete beam. Fenwick and Pauley (1968) definitively showed this to be the case. Through direct measurement on subassemblies, it was possible to conclude that at least 60% of the vertical shear is carried by aggregate interlock at flexural cracks, with the remaining proportion being carried in the compression zone and through dowel forces.
Despite the successes of early classic studies on aggregate interlock, their results have been forgotten or otherwise neglected by many prominent modern shear researchers. Tureyn and Frosch (2003), for example, have formulated an expression for Vc based on the explicit assumption that all of the vertical shear force is carried in the compression zone, describing the assumption as a “reasonable approximation.” Many others have taken similar approaches, particularly those using fracture mechanics principles (for example, Bazant and Yu, 2005). The fact that the importance of aggregate interlock is not appreciated has slowed the implementation of theoretically-sound design methods for shear. In particular, the size effect in shear can not be adequately accounted for unless aggregate interlock is explicitly considered.