Preliminary investigation on the bending capacity of a novel connection between the concrete slab and steel beam

Abstract

Using precast concrete floors can greatly improve the construction efficiency of steel structures, and the development of the connection between precast slabs affects the application of precast concrete floors in steel structures. In this paper, a new connection between the precast concrete floor and steel beam is proposed. The bending capacity of the connection was investigated through experiments, and compared with that of the cast-in-place floor. The test results showed that the reliability of the new connection was excellent, and the bending capacity of the floor with this connection is basically the same as that of the cast-in-place floor. The finite element software ABAQUS was employed to simulate the bending behavior of the new connection floor and cast-in-place floor, and the feasibility of the model was verified by comparing it with the experimental results. The model was used to analyze the structural response of the two types of composite beams composed of new connections and precast floor, cast-in-place floor, and steel beam, respectively. The load-bearing capacity and failure modes of the two composite beams were compared. The results showed that the bearing capacity of the composite beam fabricated with the new connection can reach that of the cast-in-place floor composite beam. This is only a preliminary investigation about the new connection and there are some limitations as well.

Keywords

concrete slab steel beam composite beam connection load-bearing capacity

Introduction

Precast concrete structures present good structural, economic, environmental, and social benefits, and are currently widely used in economically developed areas, and the proportion of prefabricated concrete structures is increasing year by year (Yee, 2001). Cast-in-place concrete slabs are mostly used in traditional buildings, and the construction period is longer. Prefabricated floor slabs could greatly reduce the workload of wet work on-site compared with cast-in-place concrete slabs, and can realize the standardization of production and the assembly of construction (Liu, 2010; Wu et al., 2013). However, the integrity of the prefabricated structure is not as good as that of the cast-in-place structure, and the joint is an important part to ensure the integrity of the prefabricated structure and the continuity of the stiffness in the floor plane.

The connection of assembled concrete composite floor slabs mainly improves the force transfer performance of slab joints by changing the overlapping and bending forms of the steel bar at the connection. Composite floor slab has the advantages of industrial production, fast construction speed, and good quality of the fully prefabricated floor, so it has been widely studied by scholars (Liu et al., 2014; Nie et al., 2003; Yan et al., 2016). But there is still a lot of on-site work for concrete pouring and maintenance. Generally, the extended steel bars are reserved in the assembled integral floor slab, and the slabs are connected by post pouring concrete at the joint. There are many common forms of the prefabricated integral floor slab, including precast hollow slab and prestressed hollow slab. Small weight and low cost are the advantages of this kind of floor, and this kind of floor can ensure that the structure has enough stiffness and strength. Precast solid slabs are usually used in small span structures, and can be made into prestressed members, and the structural strength and stiffness can also be guaranteed, so in recent years, they have been more and more studied and applied (Han et al., 2016; Liu et al., 2018; Zheng et al., 2015). Prefabricated steel-concrete composite beam with multi-bolt shear connectors have been investigated by Yang et al. (2020) and Gao et al. (2021).

In this paper, a new type of connection between the assembled floor slab and steel beam is presented. The connection is realized by the hexagonal double-end screw, sleeve, and perforated steel plate. The perforated steel plate is welded on the steel beam, and the hexagonal double-end screw is fixed together with the steel plate through the hexagonal hole of the steel plate, and then the hexagonal double-end screw is connected with the reinforcement in the precast floor slab through the sleeve, to realize the connection between the precast floor slab and the steel beam. The connection guarantees the transmission of the reinforcement between the precast slabs by the double sleeve, which improve the poor mechanical behavior at the connection of assembled concrete slab. Precast slabs can be produced in batches in the workshop to ensure the quality of concrete maintenance, and then lifted and mechanically connected on site, which can simplify and speed up the construction, improve the production efficiency. It can be considered that this new type of connection, while ensuring load-bearing capacity, greatly facilitates construction and is in accordance with the concept of assembled construction. The limitation of this novel connection is using the special tool to install the sleeve automatically.

The diagram of the connection is shown in Figure 1. The installation process of the connection is as follows: (1) Perforated steel plates are welded to steel beams, and hexagonal double-end screw in place; (2) Placing precast concrete slabs in position; (3) Rotate the sleeve into position, connecting the reinforcement and the hexagonal double-end screw, and then pour concrete in the reserve section.

Figure 1.

Diagram of the connection: (a) Diagram of hexagonal double-end screw; (b) Diagram of the sleeve; (c) The installation process of the connection.

To study the mechanical properties of the new type of connection, the flexural capacity tests of two assembled floor specimens using the new type of connection and two cast-in-place floor specimens were carried out. The nonlinear three-dimensional finite element (FE) models of assembled floor specimen and cast-in-place floor specimen were developed using ABAQUS software, and the comparison with the test results verifies the feasibility of the models. The models are used to analyze the structural responses of composite beams with new connections and cast-in-place floor composite beams, and the bearing capacity and failure modes of the two kinds of composite beams are compared.

Test program on connection

Specimen details

A vertical perforated steel plate is welded along the top surface of the steel beam and the fillet weld size is 8 mm. The height of the perforated steel plate is 100 mm and the thickness is 10 mm. The perforated steel plate is provided with a hexagonal hole at the position of the longitudinal reinforcement of the floor slab. The center of the hole is in a straight line with the longitudinal reinforcement of the floor plate. The dimension and detail of the perforated steel plate are shown in Figures 2(a, b).

Figure 2.

Dimension of perforated steel plate, hexagon double-end screw, sleeve, single precast concrete slab and assembled concrete floor: (a) Dimension of perforated steel plate; (b) Detail of perforated steel plate; (c) Dimension of hexagonal double-end screw; (d) Dimension of the sleeve; (e) Sleeve connection of assembled concrete floor; (f) Plan view of single precast concrete slab; (g) Side view of single precast concrete slab; (h) Plan view of assembled concrete floor; (i) Side view of assembled concrete floor.

The hexagonal double-end screw is fixed with the steel plate through the middle hexagonal section. The side length of the hexagon is 6 mm and the thickness of the middle of the hexagonal double-end screw is 9 mm. The length of the thread at both ends of the hexagonal double-head screw is 20 mm. The sleeve has an inner diameter of 10 mm, an outer diameter of 20 mm, and a length of 40 mm. The outer surface of the sleeve is grooved to facilitate mechanical connection. The dimension of the hexagonal double-end screw and sleeve are shown in Figures 2(c, d). The sleeve connection of assembled floor slab is shown in Figure 2(e).

The assembled floor adopts the new connection style to connect two precast slabs and a steel beam. The thickness of a single precast slab is 100 mm, the thickness of the concrete protective layer is 15 mm, the longitudinal reinforcement extends 40 mm from the precast slab, and screw threads are processed. The dimension of a single precast concrete slab is shown in Figures 2(f, g). After the sleeve connects the hexagonal double-end screw and the longitudinal reinforcement, concrete is poured on the connecting joint. The height of the joint concrete part is the same as the thickness of the floor, which is 100 mm, and the width is130 mm. The dimension of the assembled concrete floor is shown in Figures 2(h, i).

The cast-in-place floor adopts the common form of stud connectors. Grade 4.6 studs are set on the steel beam along the central axis every 200 mm. The reinforcement at the joint is not cut off. The detailed connection of the cast-in-place floor and steel beam is shown in Figure 3(a). The dimension of the cast-in-place concrete floor, which is the same as the assembled concrete floor, is shown in Figures 3(b, c). The thickness of the cast-in-place floor is 100 mm, and the thickness of the concrete protective layer is 15 mm.

Figure 3.

Connection and dimension of cast-in-place concrete floor: (a) Connection of cast-in-place concrete slab and steel beam; (b) Plan view; (c) Side view.

For the experimental investigation, as introduced above, two assembled concrete floors (YZ-1 and YZ-2) and two cast-in-place concrete floors (XJ-1 and XJ-2) were designed following the GB 50010-2010 (2011). HRB400 reinforcement and C30 concrete were used for all floors. HN300 × 150 × 8 × 10 narrow flange H steel beam (depth 300mm, top flange width = bottom flange width 150mm, flange thickness 8mm, and web thickness 10mm) with Q345 grade steel was adopted. Grouting material was used for the post-pouring section. The arrangement of longitudinal reinforcement of the floor is as 10@200, and the transverse reinforcement is arranged as 8@200.

When a concrete slab rests on the flange of a steel beam, the lap length can be calculated from the self-weight of the slab, the construction load and the strength of the concrete to ensure that the concrete does not fail. The dimensions of the reinforcement, screws and sleeves in the floor slab are designed to suit the forces in the floor slab and can be changed to meet the strength requirements of slabs with different depths and higher load capacities.

Material properties

The steel specimens for reinforcement and steel beam are reserved and the material properties are tested. The material properties of reinforcement and steel beam are obtained by averaging the values. For reinforcement, the elastic modulus E_s = 222 × 10³ MPa, the yield strength f_y = 462 MPa, the tensile strength f_u = 643 MPa, and yield strain ε_y = 2.1 × 10⁻³. For steel beam, the elastic modulus E_s = 217 × 10³ MPa, the yield strength f_y = 353 MPa, the tensile strength f_u = 492 MPa, and yield strain ε_y = 2.2 × 10⁻³.

As the small space reserved for the post-cast section did not facilitate the vibrating of the concrete when it was placed, a better flowing and stronger grout material was used to meet its requirements for adequately filling the reserved section and wrapping the sleeve and reinforcement. The cast-in-place concrete floor and precast slabs of assembled concrete floor are concretely cast, and reserve 6 standard cubic concrete specimens (150 mm × 150 mm × 150 mm). Three standard cubic concrete specimens are reserved for post-cast sections of the assembled concrete floor. Concrete cubic specimens are cured under the same conditions as floors, and material properties tests are carried out one day before the experiment following the GB/T 50081-2002 (2002) to determine the compressive strength. The average compressive strength is 40 MPa and 71 MPa, respectively.

Test setup

All floor specimens were simply supported with a roller on the left and a pin on the right over a span of 2200 mm and subjected to a single point load. Figure 4(a) shows an overall view of the test setup for the floor, and the schematic diagram is shown in Figure 4(b). Simple supports were provided at both ends of the floor through reaction frames. Steel beams were placed under the floor and vertical loads were applied upwards from the mid-span position by jacks under the steel beams. When the floor was stressed, the upper reinforcement and concrete were pulled and the lower reinforcement and concrete were compressed, which is consistent with the stress state of the floor in actual engineering.

Figure 4.

Test setup: (a) View of the floor test setup; (b) Schematic diagram of loading and displacement gauging; (c) Distribution of loading point and displacement gauging points.

Test devices and arrangement

A force sensor was used to measure the applied vertical loads. The force sensor was placed between the jack and the specimen at the mid-span of the specimen and at the centerline of the steel beam as shown in Figure 4(c). The vertical displacement of the specimen is measured by six linear voltage displacement transducers (LVDTs), which were arranged at mid-span, 1/4 span, and support of the specimen. The specific positions of LVDTs are shown in Figures 4(b, c). D1 and D2 located at both ends of the steel beam measured the vertical deflection at the middle of the floor span. D3 and D4 measured the vertical deflection at 1/4 span of the specimen, and D5 and D6 were located at the specimen support.

Test procedure

At the beginning of loading, the force control method was adopted, and it was changed to the displacement control method after the specimen yields. Before performing the test, three preloads were performed to check that the instrument and device were working properly. In terms of the force control method, the specimens were loaded step by step monotonously, and 1 kN and 3 kN were adopted before and after the cracking, respectively. In the displacement control method, the displacement of each stage is increased by 2–3 mm. The duration of each load or displacement increment was about two minutes.

Before the specimens were installed in place, the floor surface was painted white, and grids (100 mm × 100 mm) were drawn on the floor surface. During the test, the position, development, and distribution of the cracks were observed and recorded after the cracks occurred, and the dominant crack width and load value of the cracks were recorded during the duration of each increment.

Test results and discussion on the connection

Failure phenomenon and pattern

Figure 5 illustrates the failure phenomenon and pattern of cast-in-place floor specimens and assembled floor specimens. During the test, the following characteristics of crack development on the surface of floor specimens were captured: (1) The failure modes of the specimens were all bending failures. The major cracks of the cast-in-place floor were located near the flange edges of the steel beam, and the major cracks of the assembled floor were located at the interface of the post-cast section and precast slabs; (2) Cracks occurred earlier than the cast-in-place floor due to poor integrity of assembled floors; (3) Brittle cracking sound accompanied the early cracking of cast-in-place floors and the vertical load drops abruptly; (4) The concrete near the flange of steel beam was crushed and peeled off under compression, as shown in Figure 6(a). It was considered that the concrete in the compression zone had reached its compressive bearing capacity.

Figure 5.

Failure phenomenon of specimens: (a) XJ-1 plan view; (b) XJ-2 Plan view; (c) XJ-1 Side view; (d) XJ-2 Side view; (e) YZ-1 plan view; (f) YZ-2 plan view; (g) YZ-1 side view; (h) YZ-2 side view.

Figure 6.

Failure phenomenon: (a) Concrete crushing and peeling off; (b) Sleeve and reinforcement after the test; (c) Broken reinforcement.

After the test, observe the sleeve as shown in Figure 6(b). The sleeve was well connected and no cracks are observed on the surface. In the YZ-2 specimen, there was reinforcement that was broken at the threads near the sleeve as shown in Figure 6(c). Other reinforcements were also yielded and necking phenomenon had occurred at the same position.

Load-deflection curves

The load-deflection (P-δ) curves of four specimens are shown in Figure 7(a). The vertical load P is obtained from the data collected by the force sensor, and the mid-span deflection is obtained by averaging the vertical displacement measured by the displacement meters D1 and D2 at both ends of the steel beam.

Figure 7.

Load-deflection curves and yield point determination method: (a) Load-deflection curves of four specimens; (b) Yield point determination method.

It can be seen from Figure 7(a) that at the initial loading stage, the P-δ curves of the cast-in-place floors XJ-1 and XJ-2 develop linearly until the mid-span deflection reaches about 1.7 mm. The concrete in the tension zone exited work due to sudden cracking, the vertical load dropped suddenly, and the slope of the continuous loading curve also decreased. The P-δ curves of the assembled floors YZ-1 and YZ-2 have a shorter linear development stage than the cast-in-place floors, and the slope of the curves decreases with the increase of the load. This is due to the effect of adhesion between the post-cast section and the precast slab is very poor. Cracks continue to occur at the interface between the post-cast section and the precast slab when the assembled floors were loaded, and the concrete in the tension zone was out of work.

After two cracks in XJ-1 and XJ-2 floors, the development trend of the P-δ curve is basically the same as that of YZ-1 and YZ-2 until the specimen yields. After the specimen yields, as the mid-span deflection increases, the vertical load presents a different growth trend. According to GB/50152-2012 (2012), the ultimate bearing capacity of the specimens is judged when the mid-span deflection reaches 1/50 of the span of the specimens, that is, 44 mm. At this stage, the ultimate bearing capacity of the assembled floor specimens is 7.6% higher than that of the cast-in-place floor specimens, which may be due to the high compressive strength of the grouting material used in the post-pouring section of the assembled floor specimens. The mid-span deflection continues to increase thereafter, while the vertical loads of the four specimens show a downward trend after a small and stable increase, and the specimens show good ductility.

Table 1 summarizes the vertical load, mid-span deflection, rotation angle, and ductility ratio of four specimens in different states. The rotation angle θ is the ratio of the mid-span deflection to half the span of the specimen, and the ductility ratio $β_{Δ}$ is defined according to $β_{Δ} = \frac{Δ_{u}}{Δ_{y}}$ (Guo and Shi, 2003), where $Δ_{u}$ and $Δ_{y}$ are the ultimate and yield displacements of the member, respectively, as shown in Figure 7(b). The cracking state is when the first crack is observed. The nominal yield state is determined according to the graphical method (Feng et al., 2017; Guo and Shi, 2003). As shown in Figure 7(b), a straight line OA tangent to the initial section is plotted and a horizontal line past the peak point at point A is intersected. A vertical line AB is plotted and a curve at point B is intersected. O and B is connected and the line OB is extended to intersect the horizontal line at point C. The displacement at point C is taken as the yield displacement and the corresponding load on the curve is taken as the yield load. The ultimate state of bearing capacity is when the mid-span deflection reaches 1/50 of the specimen span, that is, 44 mm.

Table 1.

Vertical load, mid-span deflection, and rotation angle in different states.

Specimen	Cracking			Yield			Ultimate			β _Δ
Specimen	P_cr/kN	δ_cr/mm	θ _cr	P_y/kN	δ_y/mm	θ _y	P_u/kN	δ_u/mm	θ _u	β _Δ
XJ-1	15.3	1.85	1/595	26.6	10.66	1/103	36.8	44.0	1/25	5.0
XJ-2	17.2	1.58	1/696	25.6	8.55	1/128	36.1	44.0	1/25	7.7
YZ-1	8.6	1.77	1/621	28.6	12.22	1/90	39.6	44.0	1/25	4.7
YZ-2	7.6	1.69	1/651	29.7	14.78	1/74	38.8	44.0	1/25	4.3

Table 1 indicates that: (1) The cracking load (P_cr) of the assembled floor (average 8.1 kN) is much lower than that of the cast-in-place floor (average 16.25 kN), and the difference is 49.8%. (2) Nominal yielding load (P_y) and ultimate load (P_u) are relatively close, and with a difference of 10.5% and 7.0%, respectively. (3) Ductility ratio $β_{Δ}$ of the assembled floors is lower than that of the cast-in-place floors, due to the greater initial stiffness of the cast-in-place floor slab and the lower yield displacement, but it can meet the seismic requirements of reinforced concrete structures with a ductility ratio β = 3–4 (Guo and Shi, 2003). This suggests that the new connection method guarantees good ductility of the specimens without significantly affecting the bearing capacity of the floor.

Cracking problems are common for concrete bending members. The proposed connection mainly considers the reliable force transmission between the reinforcement to ensure the mechanical behavior of the floor slabs. Moreover, the concrete bending members are allowed to work with cracks. When the crack width of the assembled slab specimens reach the maximum crack width limit of 0.3 mm according to GB50010-2010 (2011) and 0.4 mm according to EN1992-1-1 (2004), the cast-in-place slab specimens have also cracked. At this time, the load-deflection curves of the floor specimens are generally consistent. The maximum allowable deflection for serviceability is also given in GB50010-2010 and EN1992-1-1 for floor slabs, which are 1/200 and 1/250 of the span, that is, 11 mm and 8.8 mm, respectively. At this time, the floor specimens are close to yielding.

Stiffness deterioration

The secant stiffness-deflection (K-δ) curves of the floors are shown in Figure 8(a). The secant stiffness K is determined by the ratio of the vertical load during to the corresponding mid-span deflection.

Figure 8.

Secant stiffness and deflection curves: (a) Secant stiffness-mid span deflection curves; (b) Deflection curves.

At the beginning of loading, the stiffness of cast-in-place floors XJ-1 and XJ-2 are significantly higher than that of assembled floors YZ-1 and YZ-2. This is due to the poor adhesion effect between the post-cast section and the interface of the precast slabs, which leads to the poor integrity of the assembled floor specimens. Cracking occurs at the interface at very low loads, resulting in a low initial stiffness of the assembled floor. After cracking of cast-in-place floors, the stiffness of the specimens decreases rapidly due to the withdrawal of tensioned concrete. After the mid-span deflection reaches about 3.3 mm, the stiffness-mid-span deflection curves of four floor specimens tend to be consistent with the increase of deflection. Therefore, the difference between the stiffness of the assembled floors and that of the cast-in-place floors is mainly reflected in the stage before the cracking of the cast-in-place floors.

Deflection curves

Taking the XJ-2 and YZ-2 specimens as examples, the deflections of the specimens under different load levels are compared as shown in Figure 8(b). The deflection of the assembled floor is significantly higher than that of the cast-in-place floor before the vertical load reaches 0.8P_u, especially at the mid-span position, due to the poor integrity of the post-cast section of the assembled floor. As the load approaches the ultimate load, the deflection curves of the specimens do not differ significantly, which indicates that there is no obvious difference between the deformations of the assembled and cast-in-place floors when the ultimate bearing capacity is reached.

Finite element modeling on connection

The extended numerical study was developed by nonlinear three-dimensional finite element (FE) models using commercial software ABAQUS. Based on the experimental phenomenon, the sleeve is effectively connected with the hexagon double-end screw and reinforcement. Therefore, the hexagon double-end screws were omitted in the assembled floor model and simplified as the direct connection between the reinforcement and the sleeve. The post-cast section was not considered, and the remaining parts were modeled according to the actual dimensions of the specimens. Only half of the floor was modeled because of the symmetry of the specimen and loading condition. Detailed descriptions of material properties, element type and mesh, and boundary conditions and constraints are outlined in the following sections.

Material models

Concrete

The nonlinear behavior of concrete material is defined by the uniaxial stress-strain curve, and the stress in the unidirectional bending test is dominated by uniaxial stress. The concrete damaged plasticity (CDP) model with uniaxial tensile and compressive strength is available in the ABAQUS material library, and it was proved by Nie and Wang (2013) that the CDP model can accurately simulate the macroscopic response of concrete members or composite members under monotonous loads. Therefore, a CDP model was used in the FE model to represent the concrete behavior. The uniaxial tensile and compressive stress-strain relationships can be obtained according to GB50010-2010 (2011), as shown in Figure 9.

Figure 9.

Stress-strain relationship for concrete: (a) Concrete in compression; (b) Concrete in tension.

The parameters of concrete include the density of concrete ρ_c = 2400 Kg/m³; the elastic modulus of concrete E_c =3.25 × 10⁴ MPa; the Poisson’s ratio ν = 0.2; the dilation angle ψ = 30°; the flow potential eccentricity ε = 0.1; the ratio of initial equi-biaxial compressive yield stress to initial uniaxial compressive yield stress σ_b0/σ_c0 = 1.16; the ratio of the second stress invariant on the tensile meridian K_c = 0.6667; the viscosity parameter μ =0.0005.

Table 2 lists the damage parameters input in the ABAQUS. The parameters were obtained according to GB50010-2010 (2011) and the energy equivalence principle proposed by Sidoroff (Liu and Hao, 2011), and it is related to the inelastic strain.

Table 2.

Damage parameters.

Compressive strength/MPa	Inelastic strain/(10⁻³)	Plastic damage factor d_c	Tensile strength/MPa	Inelastic strain/(10⁻³)	Plastic damage factor d_t
23.43	0.00	0.000	1.95	0.000	0.000
25.04	0.34	0.168	2.17	0.017	0.105
26.06	0.47	0.206	2.39	0.031	0.159
26.75	0.77	0.280	2.18	0.058	0.267
25.75	1.12	0.356	1.88	0.087	0.369
23.60	1.50	0.429	1.26	0.169	0.568
21.19	1.89	0.494	0.88	0.264	0.696
16.89	2.66	0.596	0.62	0.397	0.787
12.45	3.75	0.696	0.50	0.505	0.828
7.38	6.13	0.811	0.38	0.717	0.874
3.82	11.01	0.897	0.21	1.555	0.936
0.96	39.71	0.973	0.10	4.156	0.973

Steel beam, perforated steel plate, sleeves, steel reinforcements, and studs

In the model established in this paper, steel mainly includes steel beam, perforated steel plate, sleeve, reinforcement, and studs. Compared with concrete, steel is a homogeneous and isotropic material, so the stress-strain relationship of steel is usually simplified in numerical simulation. The von Mises yield criterion, which is suitable for metallic materials, was adopted for all the steel parts. The material properties of steel beam, perforated steel plate, sleeve, and studs were simulated using the ideal elastic-perfectly plastic stress-strain curves. The stress-strain curves of steel reinforcement were measured by material properties tests. For simplicity, the bilinear stress-strain relationship was used in the FE model, with consideration of the strain hardening. Based on the material properties test results, the modulus of elasticity E_p = 220 GPa, yield stress f_y = 460 MPa, ultimate strength f_su = 640 MPa, and ultimate strain ε_pu = 84,000 × 10⁻⁶ were used for the steel reinforcement in the FE model. The density of all steel components was taken as 7800 kg/m³, and the Poisson’s ratio for steel in all of the components was taken as 0.3, as accepted generally.

Element type and mesh

All the components were modeled by three-dimensional eight-node elements (C3D8R) with three translational degrees of freedom, reduced integration, and hourglass control. The overview of the FE model mesh for steel beams, concrete slab, sleeve, and studs is shown in Figure 10(a–c). A good compromise between accuracy and computational efficiency was provided through sensitivity analysis, which determines the size of the mesh. To obtain accurate results, a fine mesh was adopted for the sleeves, studs, and the regions around them. The overall mesh size of elements was 50 mm and the minimum mesh size was 7.5 mm.

Figure 10.

FE model: (a) Mesh of steel beam and studs; (b) Mesh of steel beam, perforated steel plate, and sleeves; (c) Mesh of concrete slab; (d) Boundary conditions.

Constrains and contact interactions

Appropriate interactions and constraints were defined between components. The nodes at the root of the studs were tied to the nodes of the steel beam and the studs were embedded into the concrete. The surfaces of reinforcement in contact with the sleeve were tied to the inner surface of the sleeve and the rest of the reinforcement was embedded inside the concrete slab by embedded constraint. The effect of relative slip between reinforcement and concrete slab was neglected. The sleeve had defined the constraint that was tied to the concrete according to the experimental results. The surface-to-surface contact interaction available in ABAQUS software was applied at the interfaces between the components, including the steel beam-concrete slab, perforated steel plate-concrete slab, perforated steel plate-sleeve were considered in the FE model. The HARD contact property was employed in the direction normal to the interface planes for depicting the contact behavior, while the friction was not been considered for the tangential response.

Boundary condition

The Cartesian coordinate system used to create the FE model is shown in Figure 10(d). Considering the symmetry, a fixed boundary condition was applied to the YZ symmetrical plane to restrain the degree of freedom U1 = U2 = U3 = 0. For the assembled floor model, the longitudinal reinforcement was connected by sleeves at the symmetrical plane, so no boundary conditions were applied. Displacement load in the Y-direction was applied through the reference point, which was defined as coupling to the surface of the support steel plate and through which the reaction force can be obtained.

Model validation

The load-deflection curves predicted by the FE models are compared with the experimental results in Figure 11. For all specimens, the initial stiffness of the floors is consistently overestimated by the FE models. However, the ultimate load-carrying capacity of the floors and the trend of the load-mid-span deflection predicted by the FE models show a very good agreement with the experimental results, and the development of cracks has been adequately captured by the FE models (manifested by the changes in the curve slope). For cast-in-place and assembled floor specimens, the ultimate bearing capacity between the experimental (36.2 kN and 39.2 kN) and FE models (36.9 kN and 39.7 kN) was within a 5% difference is observable.

Figure 11.

Comparison of experimental and analytical results: (a) Cast-in-place concrete floor; (b) Assembled concrete floor.

Figure 12 shows the failure mode of the assembled floor estimated by tensile damage in ABAQUS, which agrees well with the experimental observation. The assembled floor YZ-2 in the experiment was selected as a control sample. It can be concluded that the proposed FE models are reliable and can adequately capture the fundamental structural behavior of the floor specimens, and can be used to analyze the structural response of the cast-in-place floor composite beams and composite beams composed of new connections.

Figure 12.

Comparison of failure mode in experiment and FE analysis.

Finite element modeling on composite beam

The FE models were extended to analyze the structural response of the cast-in-place floor composite beam (CB-XJ) and the composite beam composed of new connections (CB-YZ). The composite beam FE models were developed as simply supported over a span of 3600 mm and a single point load was applied at the position of mid-span. The number of stud shear connectors of the cast-in-place floor composite beam was determined by the sheer force and was designed to be totally shear connections following the relevant provisions of GB50017-2017 (2017). And according to the shear strength of steel material, the perforated steel plate thickness of the composite beam composed of new connections, which is 3 mm, was designed to resist the shear force at the interface between the concrete slab and steel beam. Based on the symmetry of the model and the loading conditions, one-quarter of the composite beam was modeled.

Load capacity comparison of composite beam

The load-bearing capacity of the two composite beam types is compared as shown in Figure 13. In the initial stage of loading, the stiffness of the two composite beams is basically equal. When the composite beams yield, the vertical force applied to the composite beam fabricated with the new connection is slightly higher than that of the cast-in-place floor composite beam. And the ultimate bearing capacity (when the deflection reaches 1/100 of the span, 419.5 kN and 414.3 kN, respectively) of the two composite beam types is almost the same (the difference is 1.3%). The results showed that the bearing capacity of the composite beam fabricated with the new connection can reach the bearing capacity of the cast-in-place floor composite beam. This is because while ensuring the effective connection between the concrete slab and the steel beam, the presence of the perforated steel plate provides an enhancement of the bearing capacity.

Figure 13.

Comparison of analytical results of composite beams.

Failure pattern of composite beam

Figures 14(a–d) present the failure pattern of composite beams predicted by FE analysis, which is the typical flexural failure. It is seen that the developed FE models can predict the local response and the global response of the composite beams. The comparison shows that when the two kinds of composite beams reach the ultimate bearing capacity, the concrete slab and steel beam exhibited similar structural responses. Steel beams are tensioned at mid-span, resulting in large deflection and deformation, and the perforated steel plate is subjected to compressive stress. The concrete floor is mainly subjected to compressive stress, and the damaged concrete is located near the mid-span.

Figure 14.

Failure pattern of composite beams: (a) Von Mises stresses of CB-XJ; (b) Von Mises stresses of CB-YZ; (c) Compression damage in the concrete floor of CB-XJ; (d) Compression damage in the concrete floor of CB-YZ; (e) Shear stress distribution in stud shear connectors of CB-XJ; (f) Shear stress distribution in perforated steel plate of CB-YZ.

The shear stress distribution in stud shear connectors and perforated steel plate is presented in Figures 14(e, f), where one stud shear connector is selected per 400 mm and five stud shear connectors are shown. It is noted that the shear stress is small at mid-span and distributed along the entire shear span. The shear stress of the perforated steel plate concentrates near its connection to the steel beam, so the perforated steel plate as a shear connector can effectively resist the shear stress between the concrete slab and the steel beam.

Parametric studies

Parametric studies have been performed using the developed FE model of the composite beam composted of the new connections to evaluate the effects of some pertinent variables on the behavior of the composite beam proposed in this paper. This paper focuses on the performance of this new connection, considering several parameters that affect the mechanical performance of the connection.

The effects of three parameters related to the new connection configuration, that is, the thickness of the perforated steel plate (t = 2, 3, 4, 6 mm), hole spacing (d = 100, 200, 300 mm), and sleeve diameter (D = 16, 20, 24 mm), were investigated. A total of 10 FE analyses were performed with one parameter varied each time. Except for the parameters that are being studied, all the other characteristics in the FE models are identical.

The relationship between load and mid-span deflection response is appraised, and the key performance properties of the composite beam composed of new connections are evaluated based on the results of parametric studies. The main results of the FE model parametric studies are discussed as below:

Figure 15(a) shows the comparison of load-deflection curves at mid-span with varying thickness of the perforated steel plate. It is interesting to note the difference in static behavior with an increasing thickness of perforated steel plate. When t is increased from 2 mm to 3 mm, the yield load and ultimate load increase significantly, but continue to increase t to 4 mm and 6 mm, the curve changes slightly, and a downward trend of load capacity is observed. Therefore, after the thickness of the perforated steel plate meets the shear requirements, the increase of the bearing capacity cannot be proportional to the increase of the thickness.

It is clear that increasing hole spacing will decrease the stiffness and yield load. Figure 15(b) shows the load-deflection curves at mid-span for different hole spacings. It is noted that the stiffness and yield load decrease significantly with the increasing hole spacing; however, the ultimate bearing capacity of the composite beam is only slightly reduced.

The minor influence of the sleeve diameter on the load-deflection response and peak load-carrying capacity of the new connection composite beam is evident from Figure 15(c). Hence, the sleeve diameter only needs to meet the requirements for connecting reinforcements.

Figure 15.

Parametric analysis results of the composite beam: (a) Effects of perforated steel plate thickness; (b) Effects of hole spacing; (c) Effects of sleeve diameter.

Conclusions

Based on the preliminary investigation about the new connection, the following conclusion can be drawn:

The major cracks of the cast-in-place floor are located near the flange edges of the steel beam, and the major cracks of the assembled floor are located at the interface of the post-cast section and precast slabs.

Cracks in the assembled floor with new connections occur earlier than that of the cast-in-place floor due to the poor integrity; however, the ultimate bearing capacity of the assembled floor is slightly higher. It is shown that the reliability of the new connection is excellent.

The bearing capacity of the composite beam fabricated with the new connection can reach that of the cast-in-place floor composite beam.

The increase of the bearing capacity cannot be proportional to the increase of the thickness after the thickness of the perforated steel plate meets the shear requirements.

The stiffness and yield load decrease significantly with the increasing hole spacing; however, the ultimate bearing capacity of the composite beam is only slightly reduced.

Sleeve diameters have little effect on load-deflection response and load-bearing capacity of new connection composite beams.

Footnotes

Acknowledgements

The authors wish to acknowledge the support of the Chongqing Talents Plan for Young Talents (Grant No.: cstc2021ycjh-bgzxm0009). Any opinions, findings, and conclusions or recommendations expressed in this paper are those of the authors and do not necessarily reflect the views of the sponsors.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Chongqing Talents Plan for Young Talents (Grant No.: cstc2021ycjh-bgzxm0009).

ORCID iD

Weiyong Wang

References

BS EN1992-1-1 (2004) Eurocode2: Design of Concrete Structures-Part 1-1: Gerneral Rules and Rules for Buildings. Brussels: European Committee for Standardization.

Feng

Qiang

(2017) Discussion and definition on yield points of materials, members and structures. Engineering Mechanics 34(3): 36–46 (in Chinese).

Gao

Wang

, et al. (2021) Shear-slip behaviour of prefabricated composite shear stud connectors in composite bridges. Engineering Structures 240: 112148.

GB 50010-2010 (2011) Code for design of concrete structure. Beijing: China Architecture & Building Press.

GB/T 50081-2002 (2002) Standard for test method of mechanical properties on ordinary concrete. Beijing: China Architecture & Building Press.

GB/T 50152-2012 (2012) Standard for test methods for concrete structures. Beijing: China Architecture & Building Press.

GB50017-2017 (2018) Code for design of steel structures. Beijing: China Architecture & Building Press.

Guo

Shi

(2003) Theory and Analysis of Reinforced Concrete. Beijing: Tsinghua University Press (in Chinese).

Han

Zhao

Qian

, et al. (2016) Experimental study on composite RC beam-slab joints of precast concrete hollow floor with rebars lapping in pressed sleeve. Building Structure 46(10): 68–73 (in Chinese).

10.

Liu

Zhou

Xiong

, et al. (2018) Mechanical properties of prefabricated assembly slab with anchorage of pier head. Journal of Tianjin University (Science and Technology) 51(S1): 103–109 (in Chinese).

11.

Liu

(2010) Development overview and trend of steel residential floors for multi-story and high-rise buildings. China Hi-Tech Enterprises 10: 190–192 (in Chinese).

12.

Liu

Hao

(2011) Continuous Media Damage Mechanics. Beijing: National Defense Industry Press (in Chinese).

13.

Liu

Ding

, et al. (2014) Experimental study of load transfer behavior of reinforced joints between superposed slabs. Industrial Construction 44(05): 43–46 (in Chinese).

14.

Nie

Wang

(2013) Comparison study of constitutive model of concrete in Abaqus for static analysis of structures. Engineering Mechanics 30(4): 59–67 (in Chinese).

15.

Nie

Chen

, et al. (2003) Experimental study on shear behavior of R.C. aminated slabs. Industrial Construction(12): 43–46 (in Chinese).

16.

Zhu

Liang

, et al. (2013) Nonlinear FE analysis on in-plane load-bearing behavior of the new-type precast fabricated concrete diaphragms. Building Structure 43(S1): 677–680 (in Chinese).

17.

Yan

Gao

Tian

, et al. (2016) Experimental study on full-scale superimposed concrete slab with joints. Building Structure 46(10): 56–60 (in Chinese).

18.

Yang

Liu

Qin

, et al. (2020) Experimental study on multi-bolt shear connectors of prefabricated steel-concrete composite beams. Journal of Constructional Steel Research 173: 106260.

19.

Yee

(2001) Social and environmental benefits of precast concrete technology. PCI Journal 46(3): 14–19.

20.

Zheng

Zhao

(2015) Study on the connection shear behavior of the new prefabricated floor. Science Technology and Engineering 15(27): 180–183 (in Chinese).