Performance of a transfer beam with hybrid reinforcement of CFRP bars and steel bars under reversed cyclic loading

2.1 Specimen design and fabrication

The model of specimen was based on the W10-1 model in previous studies [29] and partly modified according to the requirements of the test equipment. The reduced scale ratio was 1:3. For simulating the actual stress situation, the transfer beam was supported by the frame supporting column. One floor of the short-leg shear wall was set on the transfer beam. Three specimens of the transfer beam model were identical in dimension: the height was 3050 mm, the width was 3100 mm, the wall height was 800 mm with the width of 100 mm, the net height of frame supporting column was 800 mm, the cross-section was 300 mm×300 mm, the section height of the transfer beam was 300 mm with the width of 300 mm, and the section height of the ground beam was 500 mm with the width of 500 mm. The specimens were fabricated by a horizontal framework. The design strength of the concrete was C20. For the three specimens, the reinforcements of the frame supporting column and short-leg shear wall were identical. The reinforcements of the transfer beam were hybrid CFRP bars and steel bars, while all the stirrups were still steel bars. The longitudinal reinforcement with hybrid bars differed in the form of reinforcement, reinforcement ratio, and reinforcement root number. In the transfer beam of ZHL-2, CFRP bars were involved in the bottom longitudinal reinforcement. In ZHL-3, they were applied in upper longitudinal bars. In ZHL-4, the CFRP bars were utilized in both bottom and upper longitudinal bars. The material properties of model components are listed in Tables 1–3. The scheme and reinforcement diagram of model components are shown in Figure 1.

Figure 1:

Specimen details (components scale and reinforcement detailing). F was designated for CFRP bars.

The design strength grade of the concrete for specimens was C20, while the actual concrete strength of different components in the experiments was as shown in Table 1, and the material properties of involved steel bars are shown in Table 2.

Based on the test results of longitudinal reinforcement material properties, the material property indicators of CFRP longitudinal reinforcement have also been given, including tensile strength, tensile modulus, and so on (Table 3).

2.2 Loading equipment and loading system

Both the scheme (Figure 2) and actual picture (Figure 3) of experimental equipment are shown. Two 1000 kN hydraulic jacks were utilized in applying a vertical load. A 500 kN hydraulic servo actuator was utilized as the loading device for horizontal reversed cyclic loading.

Figure 2:

Loading equipment.

Figure 3:

The actual loading equipment.

2.2.2 The horizontal load

The total horizontal load was provided by a 500 kN hydraulic servo actuator. According to [30], the loading method in this experiment was a two-point loading program. With the distribution beam, the total horizontal loads were allocated to the stress transmission beam and transfer beam in a ratio of 1.625:1. The specific horizontal loading was redistributed on the distribution beam by a hydraulic servo actuator. Four anchor rods around the stress transmission beam and transfer beam were connected with the distribution beam, and the horizontal force could be applied on the components. The steel plates were embedded in the ends of the stress transmission beam and transfer beam to prevent local crush.

The loading in the experiment could be divided into two loading steps. In the first loading step, a vertical load of 490 kN was exerted, so as to reach a required axial compression ratio of 0.3. The performance of the transfer beam under a vertical load was mainly explored.

In the second loading step, the vertical load was kept constant, and horizontal loads were applied with different grades. In the horizontal loading system, the control method combined both loading and displacement [30]: the loading control was applied before the specimen yielding and the displacement control was applied after the specimen yielding. The control value of yield displacement Δy was selected to be the maximum displacement when longitudinal reinforcement or CFRP bar in transfer beam was yielding. The loading of multiple displacements was controlled to apply [30] until the destruction of the specimen. In the experiments of this study, before yielding, loading of per level reciprocated once, while it reciprocated twice after yielding. The scheme of the loading system for the pseudo static test is shown in Figure 4.

Figure 4:

Loading steps.

3.1 The rule of crack propagation

The crack appeared in the loading process. The location, order, and loading size of cracks were observed during loading at each level. The whole process of the experiment was recorded in detail. In the whole process of crack propagation, similarities were found in all three specimens of ZHL-2, ZHL-3, and ZHL-4. The similar items included the crack of the transfer beam, crack of the frame supporting column, the first appeared crack on the short-leg shear wall, and the propagation and cut-through of the cracks. In the vertical loading process, no crack was found within the upper short pier shear wall and pillar box regardless of the transfer beam.

When a horizontal 80–90 kN load was applied, the transfer beam cracked. Increasing the horizontal load, the crack was propagated and gradually extended at both ends of the transfer beam. With continually increased horizontal load, the left and right capitals cracked, followed by the left and right column foot, and the upper short pier shear. Further increasing the horizontal load, the cracks on the column continued to expand at both ends. The concrete at both ends of the column was crushed. Finally, the beam-column joints plastic hinge failed, leading to the eventual damage of the specimen.

The final failure modes of ZHL-2, ZHL-3, and ZHL-4 were as shown in Figure 5. The positions of the cracks were also similar. On the transfer beam, most of the cracks appeared on the bearings that were close to the frame supporting column and at the bottom of the short-leg shear wall. Few cracks appeared in the middle of the transfer beam. The cracks of the frame supporting column were mainly concentrated in the upper and lower ends of the column, while few cracks were in the middle of the column. Most cracks on short-leg shear wall were on the section where the wall crossed with the transfer beam. There were hardly any cracks appearing in the upper part of the wall. Moreover, whether it appeared in transfer beam, frame supporting column or short-leg shear wall, the cracks all showed strong anti-symmetry characteristics.

Figure 5:

Specimens after failure: (A) the specimen of ZHL-2 after destruction; (B) the specimen of ZHL-3 after destruction; (C) the specimen of ZHL-4 after destruction.

The cracks were diagonal cracks with the joined interaction of bending moment and shear force. The final destructed specimens are shown in Figure 5. From the phenomena and data in this experiment, the specimens of ZHL-2, ZHL-3, and ZHL-4 showed similarities in the crack position, crack propagation process, and failure mechanism. In this related failure mechanism, the beam hinge first appeared, and then the column hinge was generated. The differences were as follows: more cracks of ZHL-2 were concentrated on the lower part of the transfer beam; most cracks of ZHL-3 were mainly concentrated on the upper part of the transfer beam; the cracks of ZHL-4 were relatively few, with a uniform distribution. This suggested that under the horizontal reversed cyclic loading, the asymmetric reinforcement of CFRP bars and steel bars in the transfer beam would bring non-uniform stress to the upper and lower part of the transfer beam. On the other hand, symmetrically allocating CFRP bars would help in increasing horizontal loading and deformation capacity of the components.

3.2 The rule of reinforcement strain development

The longitudinal reinforcement strain distribution of three transfer beam specimens was described (Figure 6). The specimens were reinforced with hybrid CFRP bars and steel bars (ZHL-2, ZHL-3, and ZHL-4). In the displacement control, one time of yield displacement cycle to six times of yield displacement cycle were carried out on these specimens.

Figure 6:

Strain curves for bars of a test piece. (A) strain curves for bars of ZHL-2; (a) strain curves for lower bars pushed; (b) strain curves for lower bars pulled; (c) strain curves for middle bars pushed; (d) strain curves for middle bars pulled; (e) strain curves for upper bars pushed; (f) strain curves for upper bars pulled; (B) strain curves for bars of ZHL-3; (a) strain curves for lower bars pushed; (b) strain curves for lower bars pulled; (c) strain curves for middle bars pushed; (d) strain curves for middle bars pulled; (e) strain curves for upper bars pushed; (f) strain curves for upper bars pulled; (C) strain curves for bars of ZHL-4; (a) strain curves for lower bars pushed; (b) strain curves for lower bars pulled; (c) strain curves for middle bars pushed; (d) strain curves for middle bars pulled; (e) strain curves for upper bars pushed; (f) strain curves for upper bars pulled.

Due to the failure of some strain gauges, the longitudinal reinforcement strain distribution was not continuous. However, from the longitudinal reinforcement strain curves of the three specimens (ZHL-2, ZHL-3, and ZHL-4), a rule could be concluded that under the positive and negative loading, there was an anti-symmetric distribution in the upper and lower longitudinal reinforcement strain. The strain distribution of the middle longitudinal reinforcement was identical. The stain distribution rule has been consistent with the crack propagation rule.

3.3 The comparison and analysis of loading capacity

The data of measured crack loading, yield loading, and ultimate loading are listed in Table 4. According to the previous study [30], the values of crack loading and displacement were the corresponding loading and displacement when the first crack appeared. The values of yield loading and displacement were the corresponding loading and displacement when the longitudinal reinforcement of the transfer beam reached a yield strain. The values of ultimate loading and displacement were the loading and displacement of the structure that withstands a maximum loading.

Because the equipment was out of order, cracks of ZHL-2 were not observed. The influence of the concrete strength was also taken into account. The strength of the concrete for ZHL-2 was 1.34 times that of ZHL-4. In the positive loading, the yield loading and ultimate loading of ZHL-2 was just 1.09 and 1.18 times that of ZHL-4, respectively. In the negative loading, the two data were almost the same. The concrete strength of ZHL-3 was 1.16 times that of ZHL-4. In the positive loading, the yield loading and ultimate loading of ZHL-3 were 1.09 and 1.06 times that of ZHL-4, respectively. However, in the negative loading, the yield loading and ultimate loading of ZHL-4 were higher than that of ZHL-3. For ZHL-3 and ZHL-4, the data of crack loading for positive and negative loadings were completely equal. From the comparison and analysis, it was suggested that under a vertical load, the transfer beam was an eccentrical tension member, while under horizontal reversed cyclic loadings, symmetrically allocated CFRP bars in the upper and lower part of the transfer beam would help in improving the loading capacity of the structure.

3.4 The comparison of ductility and elastic-plastic deformation

The positive and negative elastic-plastic deformation displacement and ductility factor of the samples were measured (Table 5). u_y , Δu_y , θ_y were the centerline displacement of the stress transmission beam/transfer beam, storey displacement, and inter-storey displacement angle during the obvious yield, respectively. u_p , Δu_p , θ_p were the centerline displacement of the stress transmission beam/transfer beam, storey displacement, and inter-storey displacement angle during a maximum elastic-plastic deformation, respectively. μ was the ductility factor [14, 15, 30].

The ductility factors of ZHL-2 under positive and negative loadings were 5 and 5, respectively. The average ductility factors of ZHL-3 were 7 and 6, respectively. The average ductility factors of ZHL-4 were 6 and 7, respectively. Comparing the ductility factors of the three specimens, it could be seen that the structural ductility factors of all the specimens were >4, which could meet the requirement from seismic regulations [31, 32]. The ductility of the structure was good.

The ductility factor at the position of the stress transmission beam was generally lower. All the data were smaller than 4 except ZHL-2. It was mainly obtained from the “column hinge destruction” after 4Δ_y, and the storey displacement of the second floor was smaller than that of the first floor.

3.5 Hysteresis curve analysis

The hysteresis curve of P (horizontal force)-Δ (horizontal displacement) for the transfer beam is shown in Figure 7. P was the total loading applied by tension and compression jack, and Δ was the horizontal displacement of the center in the left end of the transfer beam.

Figure 7:

Hysteresis curves of three specimens (ZHL-2, ZHL-3, and ZHL-4): (A) ZHL-2 hysteresis curve; (B) ZHL-3 hysteresis curve; (C) ZHL-4 hysteresis curve.

It could be observed that in the initial loading process, the P-Δ curve of the transfer beam was generally a straight line with small residual deformation, indicating an elastic state of the specimen. Until cracks appeared, the curve became bent. The plastic performance of the concrete began to emerge. Under repeated cyclic loading, the stiffness of components would gradually decrease after the yield of longitudinal reinforcement for the transfer beam. After this point, a slight increased loading would bring relatively significant displacement and the cycle began to appear in the hysteretic curve. Therefore, the yield loading and yield displacement of transfer beam could be determined based on whether the longitudinal reinforcement reached yield strain and whether there was a significant bend in the hysteresis curve. The combined methods would be more accurate.

In the hysteretic curve of displacement control, under cyclic loading, a pinching phenomenon was found in the hysteretic curve during both positive and negative loading process. However, the two ends of the hysteresis curve were full, especially in the final reverse loading process. The loading capacity decline for specimens was not significant with increased displacement. When the loading reached 5Δ_y, the loading capacity of the specimen began to decline by 15%–20% and the component failed. From the comparison of the three hysteretic curves, a certain seismic performance could be observed in the transfer beam with hybrid reinforcements of CFRP bars and steel bars (ZHL-2, ZHL-3, and ZHL-4).