Behavior of Prefabricated Composite Arch Coupling Beam for Shear Walls Subjected to Cyclic Loading

Abstract

Under seismic loading, the coupling beam not only connects the wall stems together, but also plays a role in energy dissipation and shock absorption. On the basis of a literature review, this paper presents a new type of prefabricated concrete arch coupling beam with steel connectors. The structure of the prefabricated beam and its connection with the wall stems are introduced. The seismic performance of the beam-wall joint is analyzed through experimental research and finite element analysis. The results show that the arch beam effectively improves the shear bearing capacity of the beam-wall connection, thus improves the overall seismic performance of the coupled wall structures. This research provides a useful reference for developing design codes for prefabricated coupling beams.

1. Introduction

The coupling beam in shear wall structures is a key energy-consuming member that provides earthquake protection for the whole structure. A reasonable structural design scheme is to make the coupling beam yield before the wall stems. Coupling beams should be designed with good ductility to consume seismic energy and reduce damage to the main structure. Traditional reinforced concrete coupling beams have a relatively small span length and poor ductility, which cannot provide effective protection to the shear wall against earthquakes. To solve this problem, scholars all over the world have carried out research on the reinforcement, section design and composite configurations for shear wall coupling beams.

Paulay et al. [1] proposed a reinforcement design with diagonal rebars concealed in beam-column connections. Through experiments, it was found that this reinforcing method can improve the bearing capacity of the beam and avoid the yield of the steel bars and the partial crushing of concrete. Moretti [2] conducted a quasi-static test on this new reinforcing method and the results showed that the ductility and energy dissipation capacity of the reinforced beams supported by the diagonal concealed columns were greatly improved. Tegos et al. [3] proposed a rhomboid reinforcement scheme for the main reinforcement and designed different forms of rhomboid layout of reinforcement to carry out quasi-static tests. The results show that the rhomboid layout of reinforcement can effectively restrain the development of inclined cracks and improve the strength and ductility, with a slight stiffness decrease.

Because steel-concrete composite beams have the advantages of both concrete and steel beams, they have become a new development direction for composite structures. Gong and Shahrooz [4] compared the failure modes of two groups of sectional steel composite beams, and the results showed that the concrete wrapped around the steel section provides constraint to the steel section. Subedi [5] proposed to use steel plate composite beams for the first time. It was verified through tests that steel plate concrete composite beams have better shear performance compared with ordinary beams. Lam et al. [6] studied coupling beams with a vertically embedded steel plate along the whole span, either with or without shear studs. The results showed that embedded steel plates could improve the strength and stiffness of coupling beams. Teng et al. [7] first proposed to use concrete-filled steel tubular beams. Through an experimental study, it was found that the lower tensile zone of the steel pipe cracked and failed.

For prefabricated structures, the vertical joints of the components have an important effect on the seismic performance of the structures. At present, many scholars have carried out a series of studies on the vertical joint connection of prefabricated concrete shear walls.

Sun et al. [8] proposed a vertical joint device for prefabricated shear walls, which is composed of a built-in H-type steel frame and high-strength bolts. Through quasi-static tests, it was found that this connection method could effectively transfer forces. Liu et al. [9] embedded a connector at the joint and connected it by welding. The test results show that the welding joint can fully transfer the shear force at the vertical joint. Twigden [10] studied a precast concrete shear wall system using a slot-bolted connection; the tests evaluated a design procedure with both the global and local force-displacement response parameters.

Jesper et al. [11] improved the connection form of stirrup bolts and added buckle-resisting bars to the post-cast belts. The test found that the steel bars in the specimens could effectively transfer stress. Ramin et al. [12] used a finite element software to compare and analyze two vertical connection forms of stirrup bolts and embedded steel member bolts. The simulation results show that the energy dissipation capacity of the embedded steel bolt connection is 127% higher than that of the stirrup bolt connection. Trevor et al. [13] used glass fiber (GFRP) composite boards and carbon fiber (CFRP) anchors as connection materials to reinforce vertical joints between prefabricated walls. Through static cyclic load tests, it was shown that this composite connection structure had good performance. Li et al. [14] designed connections with long and short section steel embedded in wall stems, respectively. Low cycle repeated loading tests were carried out to study the mechanical properties of the steel beam and concrete shear wall connected by the embedded steel sections.

The results of most experimental studies show that the common failure of reinforced concrete coupling beams starts from the ends of the beams. The existing research on the prefabricated coupling beams is still very limited and has not addressed many important issues, such as the preferable joint design for desired cost-effectiveness, constructability and seismic performance. Based on a broad literature review [15] and the unique characteristics of prefabricated double stem shear walls, a new structure of coupled shear walls is proposed in this paper, in which the two ends of the coupling beam use an arched section to increase the end section area, and the joints between the coupling beam and the wall stems are reinforced with steel connectors and horizontal and inclined steel bars. The new coupled shear wall can improve the shear bearing capacity of the beam-wall connection, optimize the stress distribution, improve the overall ductility, and ensure the convenience for prefabricated construction. The structure of the prefabricated arch beam and its connection with the wall stems are introduced in detail, and the seismic performance of the new arch beam joints are evaluated through experimental research and finite element analyses. As a preliminary but pioneering research, this paper provides reference information for further studies and for practical applications in this field. The paper covers the following sections: first, the structure of the prefabricated arch coupling beam is proposed; then, the test of the coupling beam is introduced and the results are summarized and analyzed; followed by an finite element analysis of the coupling beam; conclusions are made at the end of the paper.

2. Prefabricated Arch Coupling Beam Joint Design Scheme

The joint, as shown in Figure 1, consists of the following parts: prefabricated wall stems, prefabricated arched coupling beam, cast-in-place composite connection layer and shaped steel connectors (steel sections in the wall and steel plate in the beam).

The structure of the prefabricated wall stems is shown in Figure 2. The prefabricated upper and lower wall stems are connected in the field by welding the embedded I-shaped steel sections. A gap remains between the upper and lower wall stems that will be filled with concrete cast in place, together with the top portion of the coupling beam (see the composite layer in Figure 1), to connect the wall and the beam. To further improve the connection, U-shaped rebar hooks are extended out of the precast wall stem and beam, which are connected and cast in concrete (see Figure 2 and Figure 3). The prefabricated arch beam is composed of the following two parts: the bottom part of the beam is precast (Figure 3) and the upper part is cast in place, as shown in Figure 4. An embedded trapezoidal steel plate at the end of the beam is weld-connected to the flange of the I-shaped steel section in the wall. Reinforcements in the upper part of the beam are connected with the reinforcements in the wall. A mechanical threaded connection sleeve is embedded at the end of the coupling beam to connect the horizontal bars between the wall and the beam, as shown in Figure 4. After connecting the reinforcements and the steel members between the upper and lower prefabricated walls and between the wall and the beam, concrete is cast in the gap between the upper and lower walls, as well as the top of the prefabricated beam. The cast-in-place concrete, designated as the composite layer in Figure 1, bonds the wall stems and the beam together. The I-shaped steel sections embedded in the precast wall stems also provide temporary support during the assembling of the walls.

3. Test of the Prefabricated Arch Coupling Beam Joint

3.1. Purpose

This test was performed to evaluate the effect of the composite steel-concrete connectors at the end of the coupling beam on the shear capacity of the prefabricated arch beam. The failure process and seismic performance of the arch coupling beam is evaluated under reciprocating loads.

3.2. Specimen Design

When a coupled shear wall is subjected to horizontal loading, the deformation and force in the coupling beams are second-order symmetric with reversed bending in the beam, as shown in Figure 5. To simplify the test configuration and focus on the beam ends, half of the beam is taken out, as the test specimen and a concentrated force is applied at the mid-span of the beam (the edge of the half-beam) to obtain the equivalent shear and moment force at the end of the beam. The transition process from the real shear wall structure to the test specimen is shown in Figure 5.

The arch beam specimen was designed according to the equivalent deformation model shown in Figure 5c. The beam (half) and wall were rotated 90 degrees to facilitate loading in the horizontal direction. Because the wall is much stiffer than the coupling beam, the wall was fixed on the floor through a footing, as shown in Figure 6. The design of the specimen follows the Code for design of concrete structures (GB 50010-2010). The size of the wall is 1500 mm × 800 mm × 200 mm (length × height × thickness). The half-length of the coupling beam is 600 mm. The cross-section of the half-beam end is 450 mm × 200 mm (depth × thickness). To apply the load, a reinforced concrete loading block is cast together with the beam at the top, with a size of 500 mm × 300 mm × 200 mm (length × height × thickness). The footing was anchored on the ground by two ground anchor bolts. The size of the footing is 2400 mm × 300 mm × 400 mm (length × height × thickness). The coupling beam reinforcements, shaped steel section and steel plate are shown in Figure 6. The coupling beam and the wall were connected by welding the steel plate in the beam and the flange of the I-shaped steel section in the wall before casting concrete.

To improve the connection between the steel plate and concrete, short steel bars are welded on both sides of the steel plate as shear studs (Figure 7). The dimensions and properties of the I-section and steel plate are given in

Table 1. I-shaped steel section size (mm).

Component Name	Depth of Section (h)	Web Thickness (tw)	Flange Width (bf)	Flange Thickness (tf)	Yield Strength (MPa)
I-section steel	80	8	80	8	235

and

Table 2. Trapezoidal steel plate size (mm).

Component Name	Upper Width (mm)	Bottom Width (mm)	Height (mm)	Thickness (mm)	Yield Strength (MPa)
Trapezoidal steel plate	100	200	260	8	235

, respectively. Reinforced bars used are made of HRB335 and HRB400 steel. As shown in Figure 6, HRB335 bars have a diameter of either 8 mm or 10 mm, and HRB400 bars have a diameter of 18 mm; the properties of the steel are listed in Table 1 and Table 2, which are also used for the finite element analysis in Section 5.

3.3. Specimen Preparation

Forming, reinforcing and concrete pouring were carried out in a laboratory. The wall stem and coupling beam were poured separately at two times (Figure 8 and Figure 9) to simulate the actual construction sequence. Three test blocks were cast from the wall concrete and the coupling beam concrete, respectively, for strength tests. The measured compressive strength of the concrete cubes on the test day is summarized in

Table 3. Measured strength of concrete.

Block No.	Concrete Strength (MPa)	AVE. (MPa)
Q1	35.1	36.2
Q2	35.9
Q3	37.6
L1	24.9	24.6
L2	25.4
L3	23.5

, where Q1 to Q3 are from the wall concrete and L1 to L3 are from the beam concrete.

3.4. Loading and Measuring Sensor Layout

3.4.1. Loading Scheme

In this test, an MTS electro-hydraulic servo actuator (1000 kN ± 250 mm stroke) was used to apply a horizontal low cycle reciprocating load on the specimen (Figure 10). The loading process was controlled by displacement. The displacement was increased each cycle with an increment of 1 mm at the initial loading stage, and 2 mm after obvious cracks appeared on the specimen. The loading sequence was designed referring to the typical cyclic reciprocating loading process used in previous research [15].

3.4.2. Measuring Sense Layout

The purpose of the test is to obtain the relationship between force and displacement under horizontal low-cycle reciprocating loading and develop a stress–strain curve. The strain of steel bars, the shaped steel section and steel shear plate were measured at the locations shown in Figure 11. Sensors B1–B4 measure the strain of the shear steel plate at its four conners, H1–H4 measure the strain of the steel section along its flange, and L1–L4 measure the strain of the vertical steel bars of the connecting beam close to its end. The whole process of crack development and specimen failure was recorded. The force and displacement were collected by the sensor in the actuator, which was also checked by displacement meters including D1 and D2 that measure the displacement of the base, and D3 and D4 that measure the displacement of the loading block (as shown in Figure 11).

4. Quasi-Static Tests and Results

4.1. Test Procedure

The load was applied horizontally at the top of the beam, which is defined as positive to the right, and negative to the left, as shown in Figure 12. The loading process was controlled by displacement, while the specimen was alternatively loaded in the positive and negative directions. The loading sequence is shown in Figure 13 and summarized in

Table 4. Critical test stages.

Displacement (mm)	Loading (kN)	Related Beam Changes
−3	−35.0	First horizontal crack appeared at a position about 300 mm away from the end part of the beam. The cracks gradually extended to both sides of the beam;
6	30.7	A new crack appeared on the upper side of the beam, about 200 mm above the end of the beam;
−8	−52.9	Another new crack appeared at the lower side of the coupling beam about 200 mm away from the end of the beam;
10	36.9	A crack appeared on the upper side of the coupling beam about 340 mm above the end of the beam;
−10	−60.78	A crack appeared at a position about 400 mm above the end part of the beam;
14	45.2	A new crack appeared on the upper side of the coupling beam about 80 mm above the end of the beam;
−14	82.54	A new crack appeared at the lower side of the coupling beam about 80 mm above the end of the beam.

As the displacement continued to increase, the existing cracks continued to develop diagonally downward, and the width of the cracks continued to increase, forming cross oblique cracks. When the displacement reached 30 mm, the cracks basically penetrated the whole beam, and the concrete at the beam end showed compression failure (Figure 14).

.

The first loading cycle was as follows: first, pushing the beam 1 mm in the positive direction, holding the load for about 10 min, releasing the load to 0; then switching the load to the negative direction and deforming the beam to −1 mm, holding the load for about 10 min, then releasing the load to 0.

The following loading cycles repeated the same process as the first cycle, with a gradually increased beam deformation: 2 mm, 3 mm, 4 mm and 5 mm, respectively, with 1 mm increment per cycle, then 6 mm to about 30 mm with 2 mm increment per cycle (shown in Figure 13). The beam failed at a deformation of about 30 mm and the test was terminated accordingly.

At the early stage of loading, no obvious cracks appeared in the structure when the horizontal displacement was small. With the increase in load, the condition change was recorded. Some critical stages are summarized in Table 4 (the left side of the beam shown in Figure 12 is referred to as the upper side and the arched right side as the lower side).

4.2. Test Results

Steel Bar and I-Shaped Steel Section Stress Analysis

To compare the stress changes in the I-shaped steel section, steel plate, longitudinal bars and oblique bars of the beam from the beginning of loading to failure, the records of the strain gages on these components are summarized and plotted in Figure 15.

As shown in Figure 15a, the bottom corner of the trapezoidal steel plate (B4) had the largest stress, where it is connected with the I-shaped steel section flange. The large stress in the steel plate indicates that the steel plate is fully stressed and contributes to the shear bearing capacity of the beam-wall connection.

As shown in Figure 15b, the stress in the I-shaped steel section is the largest near the connecting point with the trapezoidal steel plate (H2 and H3), and the stress at its ends is smaller (H1,H2). When the beam failed, the wall stem did not show any obvious damage.

As shown in Figure 15c, the maximum stress was in the oblique bars arranged below the coupling beam, which reached 519 MPa (L4). The oblique bars inhibited the development of oblique cracks.

5. Finite Element Analysis of the Coupling Beam

The applied finite element method simulates the structure by virtually dividing it into small elements that are connected by normal and shear springs positioned at specific contact points around the surface of the elements. This study uses finite element software (ABAQUS) for analysis and typical solid elements for concrete and steel and typical tie constraints are used to build the element models [16,17].

5.1. Establishment of the Finite Element Model

5.1.1. Property of Materials

HRB335 steel was used for the horizontal bars, longitudinal bars and stirrups in the wall, as well as for the oblique bars and stirrups in the coupling beam; HRB400 steel was used for the main longitudinal bars and connecting bars between the beam and the wall stems; Q345 steel was used for the I-shaped steel section and shear steel plate; and C30 concrete was used for the wall and the beam. The properties of concrete and steel are listed in

Table 5. Concrete properties.

Material	Elasticity Modulus (MPa)	Poisson’s Ratio	Compressive Strength (MPa)	Tensile Strength (MPa)
Concrete C30	3.0 × 104	0.2	30	3

and

Table 6. Steel properties.

Material	Elasticity Modulus (MPa)	Poisson’s Ratio	Yield Strength (MPa)	Ultimate Strength (MPa)
HRB335	2.0 × 105	0.3	335	455
HRB400	2.0 × 105	0.3	400	540
Q235	2.0 × 105	0.3	235	490

.

5.1.2. The Constitutive Relationship of Materials

The uniaxial stress–strain relationship of concrete and the double line model for the steel section, steel plate and steel bars are based on Appendix C of the Chinese specification GB 50010-2010 (China Standards Press 2010), as shown in Figure 16.

5.1.3. Establishment of Finite Element Model

The finite element model of the test specimen was established using Abaques 2016 (Figure 17); the dimensions and reinforcement layout of each component are the same as those used in the test specimen. The full Newton –Raphson method was used for the finite element analysis.

C3D8R solid elements are used for the concrete, I-shaped steel section and trapezoidal steel plate in the model. The reinforcement is made of T3D2 truss element.

The interaction between the reinforced steel bar, I-shaped steel section and trapezoidal steel plate and concrete is constrained by the embedded region. The binding constraint (TIE) is used between the coupling beam and the wall stem, and the welding between the shear steel plate and I-steel section is considered reliable, so the binding constraint (TIE) is also used. General contact is used between new and old concrete.

The reference point (RP-1) is set at the center of the end section of the coupling beam. The point is connected with the end section of the connected beam through coupling, and a repeated load is applied to the beam. To show the failure of the connecting beam more accurately, the coupling beam is divided into fine grids.

5.2. Finite Element Analysis Results

5.2.1. Plastic Damage Behavior of Concrete

When the reference point (RP-1) is applied and the cyclic load is low, the upper and lower side of the beam is subjected to tensile stress alternately under the positive and negative load on the beam. Cloud images of each loading step were obtained in the post-processing of the finite element software to explore the failure modes of the fabricated beam (Figure 18).

When the displacement of the top of the coupling beam reaches 30 mm, the concrete at the bottom of the coupling beam damages by compression. The shear steel plate shares part of the load and provides protection to the concrete at the bottom of the beam. The compression damage area of the concrete at the bottom of the beam is around the shear steel plate at both sides of the beam.

The cracks of the arch beam were distributed evenly along the span direction of the beam. No penetrating cracks occurred at the bottom of the coupling beam at the early stage (Figure 18).

5.2.2. Analysis of Concrete Strain

The equivalent plastic compressive strain (PEEQ) represents the accumulation of plastic damage of concrete during the whole loading process. The distribution and failure modes of the main cracks in the model can be determined by the maximum principal plastic strain. At the bottom of the steel plate, the plastic damage of the concrete is large, and at the belly of the steel plate, the plastic damage of the concrete is the largest. Along the direction of the span of the beam, a number of cracks appear, and cracks form about 200 mm above the end of the beam (Figure 19).

5.2.3. Stress Analysis of Steel Bars and Steel Plate

The stress in the oblique bars in the coupling beam is large; the maximum value reaches 522 MPa (Figure 20). After the yield of the oblique bars, the stress increases rapidly and nearly reaches its ultimate strength. The configuration of the oblique bars has good resistance to the development of the oblique cracks in the coupling beam.

The stirrups at the beam bottom yielded (343 MPa) and the constraints on the concrete inside the stirrups reduced. The concrete is gradually damaged in the process of repeated tension and compression. It is concluded that more stirrups at the end of the coupling beam can delay the damage of concrete.

The maximum stress of the trapezoidal steel plate reaches 354 MPa, indicating that the steel plate is fully stressed and yielded. The shear bearing capacity of the wall-coupling beam connection is improved by the contribution of the steel plate.

The stress in the I-shaped steel section near the steel plate is large (210 MPa), and the stress at the locations away from steel plate is smaller (less than 95 MPa), which is in good agreement with the test results.

The results demonstrate that if the length of the I-shaped steel section embedded in the wall increases, a plastic hinge may be formed at the connection joint between the wall stem and the coupling beam when concrete at the joint is damaged; therefore, a longer steel section could play a better role in energy dissipation.

5.2.4. Comparison of Test and Analysis Results

Comparing the finite element analysis with the test results (Figure 21), the change trend of both is similar. Compared with the test results, the initial stiffness and the ultimate load of the finite element analysis is larger, but the ultimate displacement of the finite element analysis is smaller. The possible reason is that despite the restraints used during the test, there is still relative sliding of the foundation pedestal; in addition, the numerical simulation does not consider the slip effect between the steel sections and concrete.

5.3. Discussions

It is verified from the test and simulation results that the stress and strain are intensified at the interface of the wall and the coupling beam. The advantage of this innovative coupling beam design is that the design of the beam end is enhanced by an increased cross-section area and the composite steel-concrete structure that provides more effective and efficient resistance to the unique loading caused by high shear wall structures. The cracks are more evenly distributed over a larger region along the beam, due to the resistance provided by the steel plate and the strategically arranged reinforcement bars. Since more components are used at the connection, the function and the contribution of each component need to be further explored based on this research.

6. Conclusions

In this paper, a new type of prefabricated arch coupling beam is proposed based on the force characteristics of coupling beams and the requirements of prefabricated construction. Pseudo-static tests and finite element analyses were carried out to evaluate the seismic performance of the wall-beam structure. The following conclusions are drawn.

The arch coupling beam provides an effective connection with the wall stem with good shearing resistance and ductile behavior. The cracks of the arch beam are uniformly distributed along the span direction of the beam.

The I-steel section in the wall stem and the steel plate in the beam can effectively connect the wall and the beam; there is no slip at the construction joints. The shear bearing capacity of the connection is improved by the steel connectors between the beam and wall stem.

The oblique bars in the precast beam can effectively prevent the development of oblique cracks. Sufficient stirrups should be provided to restrain the concrete at the wall-beam connection. The amount of stirrups should match other reinforcements, so as to ensure that they work together with the longitudinal bars and the oblique bars before failure.