PSI - Issue 78

Sabatino Di Benedetto et al. / Procedia Structural Integrity 78 (2026) 1697–1704

1698

Keywords: bond behaviour; foundation; composite hollow sections; numerical modelling; smooth elements

1. Introduction In civil engineering, the adoption of composite steel – concrete sections has been extensively developed over time. Many structural applications have encased steel elements within reinforced concrete columns and walls, achieving higher strength and stiffness compared to standalone steel or reinforced concrete structures, all while maintaining cost efficiency. These advantages rely on effective stress transfer between the steel core and the surrounding concrete, which can be achieved through either bond mechanisms or mechanical shear connectors. A notable example is the use of concrete-encased concrete-filled steel tube (CFST) elements, which demonstrate high load-bearing capacity, ductility, and fire resistance (Li et al. 2018, Lee et al. 2019, Parsa-Sharif et al. 2023). These composite systems are employed in demanding structural applications, such as steel tube concrete-stacked columns in high-rise buildings, exemplified by the 209-meter frame-core tube structure of the Tianjin Modern City Hotel tower (Minqiang, 2020). They are also integral to arch bridge construction, where they serve as robust skeletal frameworks enabling longer spans, as demonstrated by the Beipanjiang Bridge (445 m span). In the investigation of buildings with composite elements, prototype structural configurations compliant with the Uniform Building Code (1997) seismic provisions have been developed for concentrically braced frames featuring composite columns, as well as for mixed steel systems incorporating composite shear walls. Static, response spectrum, and time history analyses were carried out on these prototypes (Emoto, 1996), identifying regions with high bond stress demands and estimating local stress levels. These demands vary based on the structural configuration and location, but they are particularly pronounced in braced frames with composite columns and in systems with composite shear walls. In such systems, axial forces from braces transfer significant vertical loads to the composite columns at the brace-beam-column joints, necessitating substantial bond stress to effectively distribute forces between steel and concrete. Composite technology is also applied in foundation systems, such as micropiles, which consist of concrete-filled steel tubes anchored into concrete plinths. Here too, combined bending and axial forces transmitted by the structure to the foundation induce bond stresses at the steel – concrete interface. When shear studs are not used, the bond behaviour becomes especially significant. The bond strength at the steel – concrete interface is critical for structural performance and arises from three main mechanisms: chemical adhesion, mechanical interlock and friction. Initially, chemical adhesion and mechanical interlock resist slip and initiate diagonal stresses in the concrete. As slip progresses and cracking occurs, these mechanisms weaken and frictional resistance becomes dominant. Over time, continued slip reduces friction until the steel bar is fully pulled out and bond resistance is lost. Research on bond stress capacities of reinforcement embedded in concrete has a long history. Early experimental investigations (Bach, 1911; Abram, 1913) focused on assessing how end anchorage details influence the pull-out resistance of smooth rebars. Around the same time, Saliger’s comprehensive study (Saliger, 1913) offered valuable insights by examining a wide range of test configurations aimed at evaluating anchorage strength. Tests included both straight and hooked bars (with 180° hooks), varying in diameter, bend radius, and transverse reinforcement. By the 1950s, early forms of ribbed rebars were introduced and compared with smooth bars in terms of force – slip behavior (Fishburn, 1947), through over forty tests involving different hook radii, surface types, development lengths, and hook angles. Later studies from the 1960s (Rehm 1969) contributed a substantial dataset on smooth rebars, which became reference material for subsequent research in bond behaviour. Some of these studies have been exploited to define the design rules of anchorage legnths of steel reinforcement bars currently included into the Model Code 1990 (Telford, 1993). Similarly to the case of the bond behaviour of bars, investigations have regarded also the case of steel elements embedded into concrete. Numerous studies (Bryson and Mathey 1962; Hawkins 1973; Roeder 1984; Hamdan and Hunaiti, 1991; Wium and Lebet, 1991) have examined composite columns using push-out tests. These tests apply axial loads to the exposed steel at one end while the concrete at the opposite end resists the force. The bond stress capacity is typically evaluated as the maximum average bond stress, defined as the peak load transferred from steel to concrete divided by the surface area of the steel embedded in the concrete.