Characterisation of the splitting behaviour of concrete slabs in steel concrete composite beams with shear stud connection

Abstract

Steel-concrete composite members combine the high tensile strength of steel with the high compressive strength of concrete, to create stronger and stiffer members in bending. For a composite member to act effectively, internal loads must be transferred from the steel to the concrete via the interface. When the steel-concrete interface starts to fail, the composite load carrying capacity of the structural member can be significantly compromised. This type of failure must be prevented for satisfactory performance of complex structural assemblies, such as composite beams in buildings and bridges, with respect to both the strength and serviceability limit states. To address this, the primary aim of this research has been to characterize the behaviour of the composite beam when it fails by splitting of the concrete along the line of the studs, known as longitudinal splitting. This was conducted through careful experimental testing. This testing allowed the exploration of several currently unknown effects of various loads on components of steel-concrete composites, such as transverse loading and low-level cyclic loading on shear stud capacity. Additionally, the differences and limitations of EuroCode 4 and NZS standards, and their effects on composite beam failure characteristics were also studied. The results of these tests provide data to validate numerical models, to further explore the behaviour of composite beams. Geometries of numerical models have been created, but the implementation of them and exploration of composite beam behaviour using these models falls outside the scope of this work. The results and analysis of 15 push-off tests to determine the longitudinal splitting characteristics of composite beams with headed shear stud connectors were carried out. A modified test rig was used, instead of the standard push-off test. The advantages of the modified push-off tests were; change in rib width in a primary T-beam and the effect of transverse compression across the base of the stud could be investigated visibly and experimentally to a high degree of resolution, thus tracking the change in width throughout the whole loading range. Also, there is only one interface where slip will occur, compared to the two interfaces in a typical push-off test. In a typical push-off test, when the behaviour transitions from linear to nonlinear, the force distribution in each interface is unknown and suitable data cannot be drawn once nonlinear behaviour is observed. Both monotonic and cyclic loadings were considered. The results allowed the splitting failure mode to be characterised in unprecedented detail. It was shown that a large number of cyclic loads will not lead to premature failure of the composite beam, but instead increase the capacity of the beam against longitudinal splitting. This is significant for composite members under variable loading such as floor supporting beams in car parks. It was also shown that the transverse compression forces across the base of the studs would increase the capacity of the beam against longitudinal splitting. When looking at differences in design codes, it was shown that the presence of anti-splitting reinforcement would result in a gradual loss of load carrying capacity, compared to a sudden loss when it is not present. It was also shown that even with sufficient side cover from the stud to the edge of the concrete, according to NZS standards, longitudinal splitting will still be observed. Finite element models have been developed in order to explore the behaviour of these steel-concrete composites further. Two different finite numerical models have been created. Firstly, a macro model comprising a steel beam, shear studs and a concrete deck to simulate the adopted push-off test. Secondly, a mesoscopic model of concrete to simulate standard compression and tension tests of a concrete cylinder. Whilst validating, utilising and using these models as an exploratory tool falls outside the scope of this research, details of the modelling to date has been provided with the idea that future research will be conducted to continue the study of understanding the behaviour of steel-concrete composite behaviour.