Experimental analysis, numerical and analytical modeling of shear strength mechanisms in Hybrid Steel Trussed Concrete Beams

The purpose of the present dissertation is the study of the shear behavior of the so-called, in the recent scientific literature, Hybrid Steel Trussed Concrete Beams (HSTCBs). Such beams represent a structural typology which usually consists of a steel truss embedded into a concrete core so that, after curing and maturation, the two materials behave as a unique structural system, the steel members working as the reinforcement of the beam itself. Since the Seventies, the HSTCBs are widely employed in civil constructions because they allow to industrialize the building process, avoiding substantial alterations in the construction processes and organizational protocols of the industries. With regard to the introduction of this beam typology within seismic framed structures, it is necessary to develop specific design criteria based on the capacity design approach, ensuring both an adequate shear resistance in order to prevent brittle failure modes and a cyclic dissipative behavior of beam-to-column joints. While the issues concerning the flexural behavior has been widely investigated in the literature, particularly focusing on the connection deformability and strength, the problems related to the shear behavior still represent an open issue. Within this framework, the present thesis aims at investigating the shear response of HSTCBs and the stress transfer mechanism between the steel members and the surrounding concrete. With reference to the stress transfer mechanisms, theoretical analyses with the aim of interpreting the experimental results of push-out tests on pieces of HSTCBs are developed. Such models for the prediction of the maximum slip force refer both to the classical truss models with variable inclination of the compressed concrete strut, typically adopted for the classical Reinforced Concrete (R.C.) structures, and models with a failure mechanism governed by the dowel effect, generally used for composite structures. The application of these models, according to their original formulation, generally leads to an underestimation of the maximum load experimentally obtained by various authors from the push-out tests. For this reason, the formulation of further strut and tie models, on one hand, and dowel-mechanism models, on the other hand, has been developed. In those models proper changes have been introduced in order to take into account the geometrical and mechanical characteristics of the beam object of study. Besides the analytical interpretation of the maximum load obtained in the push-out tests, a two-dimensional (2D) non-linear finite element (FE) model is also developed with the aim to simulate, under few simplified hypotheses, the mechanical response of the beam identifying the stresses transfer mechanisms. The result of the 2D modeling highlights the difficulty of grasping, with an extremely simplified model, the large variety of parameters on which the transferring of the stresses and the failure modes depend. Among these parameters, the ones playing a preeminent role are the three-dimensional (3D) geometry and the actual bond between the surfaces of the steel bars (smooth or ribbed) and the concrete in which they are embedded. Therefore, after the developing of the simplified modeling, a detailed 3D FE model containing solid elements is realized by means of the software Abaqus 6.10. The model was developed in collaboration with the research group of Prof. Gianvittorio Rizzano of the Department of Civil Engineering, University of Salerno. Particularly, the developed models are representative, on one hand, of cases in which the diagonals of the steel truss are ribbed and, on the other hand, cases in which they are made up of smooth steel. The simulation concern cases in which the hypothesis of perfect bond between the surfaces is assumed or, similarly, cases in which there is no bond between the steel and the concrete as well as the more realistic case in which a specific bond stress-slip relationship at the interface is introduced. Besides the modeling of the experimental tests, also a parametric numerical analysis is provided with the aim of evaluating the influence of the geometrical and mechanical features of the various components of the HSTCB, such as the deformability of the bottom steel plate, the type of steel constituting the diagonal web bars (smooth or ribbed) as well as the mechanical characteristics of the materials. In addition to the study of the local problems of stresses transfer, some theoretical and experimental studies are carried out in order to investigate the global behavior of the structural elements. In particular, an experimental campaign is performed on simply supported HSTCB specimens loaded with a concentrated force in the midspan and designed to exhibit a shear failure. For the execution of the tests, a particular type of steel truss produced by the industry Sicilferro Torrenovese Torrenova (ME) is employed. Six specimens have been manufactured and classified into two series, "A" and "B". Particularly, the specimens of series "A" have been tested inducing a positive bending moment; on the contrary, the specimens of series "B" have been tested so that a negative bending moment arises. Before the concrete casting, electric strain gauges have been placed on the specimens in correspondence of the tensile and compressed diagonal bars (in the section near the welding to the inferior plate) and in the bar of the upper chord in correspondence with the central mesh of the truss in the shear span. After casting and curing of the concrete, strain gauges were placed even on the bottom steel plate. The obtained experimental results are compared with the detailed numerical FE model representative of the abovementioned tests, showing a good agreement in terms of load-displacement curve as well as crack pattern evolution. The numerical analysis is followed by the analytical interpretation for the assessment of the shear strength of the beams. In the first instance, the prediction models existing in the literature and typically employed for the classic R.C. structures have been applied. They can be mainly classified into "additive models" and "strut and tie models". In the additive models the value of shear strength is calculated as the sum of the contribution due to the concrete and the additional contribution provided by the shear reinforcement. The strut and tie models, instead, are primarily truss models in which the hypothesis of the variable inclination or 45° inclination of the compressed concrete strut is assumed. In addition to these classical formulations, also other computational models recently developed by some authors for the HSTCBs are taken into account. Successively, also a specific model able to interpret the shear strength mechanism in the tested beam typology is proposed. Considering the three-point bending tests performed on the HSTCBs, a further 3D model, realized with the software Abaqus 6.11, is developed in a simplified way, with the aim of managing a model sufficiently accurate in the estimation of the maximum load that, in the same time, would allow computational efforts appropriate for the generation of a certain number of different cases for the study of the size effect on beams with similar geometry. The model has been developed under the guide of Professors Roberto Ballarini and Jialiang Le of the Department of Civil Engineering, University of Minnesota. Starting from specific scaling criteria, three different sizes of beams are considered and the numerical load-displacement curve is obtained also interpreting the failure mechanisms and the evolution of the cracks. The numerical analyses have been developed with the aid of computers and software provided by the Minnesota Supercomputing Institute.