Fracture of concrete is a dominant failure mechanism when steel and concrete interact mechanically. In a wide variety of structures, such as connections involving steel studs embedded in concrete in composite beam structures, or in hybrid steel/concrete structures involving steel beams which penetrate into concrete columns, steel and concrete must interact with each other when the structure is loaded. Due to the high stiffness of steel and brittleness of concrete, failure usually occurs in the concrete in the form of fracture. A large number of RILEM round robin tests of steel anchor bolt pullout from concrete demonstrate experimentally and numerically that concrete fracture is the governing failure mode and that anchor capacity is controlled by the material toughness rather than compressive strength. In the 1995 great Hanshin earthquake, for instance, it was observed that failure of an exposed column base was due to the fracture of the surrounding concrete near the steel bolts. Other examples involving concrete fracture in steel/concrete interaction zones include severe concrete spalling in RC column-to-steel beam (RCS) connections due to the high bearing stress of the steel beam on concrete and concrete cracking in the anchorage zone due to the transfer of prestressing force through steel anchorage device. In the aforementioned scenarios, fracture failure of the brittle concrete at the steel/concrete interaction zones clearly compromises the safety of the structures.
Several approaches have been attempted to address steel/concrete interaction problems, with limited success. These approaches include the use of steel fiber reinforced concrete, steel band plates or fiber reinforced plastic wrapping, enlarged member section and heavy confining reinforcement. While they generally result in improved behavior in steel/concrete interaction zones, the employment of these approaches is often penalized by higher cost, labor intensity, and/or space congestion. A more elegant approach is to directly impart tensile ductility into the concrete material to minimize or suppress the fracture mode of failure all together.
A preliminary investigation was conducted on steel beam/ECC connection under shear loading. A series of pushout specimens (Fig 1) were tested using ECC material along with normal concrete as the reference material. The connection was made with two steel studs welded onto the steel beam flanges and embedded in the concrete or ECC. Overall, the ECC specimens show an average 53% increase in load and 220% increase in slip capacity compared with concrete specimens (Fig 2), even though the two materials have about the same compressive strength. The ductility of this ECC (with tensile strain capacity of 3%) was clearly reflected by distributed microcracking near the head of the shear stud (Fig 3), suppressing the localized fracture mode typically observed in normal concrete (Fig 4). While both ECC and concrete were crushed underneath the stud shank, the failure mode and load capacity were governed by the tensile response of the materials.
In another series of preliminary study, 2D anchor bolt pullout test (Fig 5) with ECC and concrete was conducted to investigate the influence of concrete material ductility on the damage evolution in anchor bolt/ECC connection. As shown in Fig 6, the microcracks in ECC initiated from the head of the anchor bolt and then diffused and grew in both number and length with increasing pullout load, towards the supporting points. Ultimately, as the tensile strain capacity of ECC material was exhausted, one of the microcracks localized and eventually led to the final failure of the ECC pullout specimen. In contrast, the concrete pullout specimen failed in a very brittle fracture manner (one single macrocrack), resulting in a much lower load capacity and structural ductility in comparison to the ECC specimen.