3.3.1. Cast-in-Place RACSWs

The transfer mechanism of the lateral load of steel frames with infilled cast-in-place RACSWs is illustrated in Figure 10a. In the initial loading stage, the horizontal load was mainly resisted by the compressive strips of the wall along the diagonal direction under the effect of the extrusion pressure of the steel frame and the horizontal shear force transferred by the shear studs. With the increase in horizontal load, the wall was divided into multiple diagonal compressive strips. Then, the concrete at the compressive strips was gradually crushed. The wall gradually failed. The horizontal load was then mostly borne by the steel frame, and the bearing capacity and lateral stiffness of the specimens decreased sharply. The structure of the infilled cast-in-place RACSWs satisfied the requirements of double seismic fortification.

**Figure 10.** Transfer mechanism of the lateral load.

#### 3.3.2. Prefabricated RACSWs

The transfer mechanism of the lateral load of steel frames with infilled prefabricated RACSWs is illustrated in Figure 10b. In the early loading stage, the horizontal load was transferred to the wall by the ear plates, T-shape connectors, and bolts, and the wall mainly bore the horizontal shear force. With the increase in horizontal load, the wall also bore oblique compression and tension along the diagonal direction, in addition to the horizontal shear force. The wall began to crack when the stress reached the tensile strength of concrete. In the later loading stage, the wall was divided into multiple diagonal compressive strips. After the cracks in the horizontal direction formed at the bottom of the embedded T-shape connector, a large area of concrete fell <sup>o</sup>ff, and the wall gradually failed. The horizontal load was then mainly borne by the steel frame, and the bearing capacity and lateral stiffness of the specimens degraded rapidly.

#### **4. Results and Discussions**

#### *4.1. Hysteretic Curves*

The load-displacement hysteretic curves of the specimens are presented in Figure 11.

#### 4.1.1. Cast-in-Place RACSWs

Figure 11a–c demonstrate the following:

(1) The specimens (Figure 11a–c) are in the elastic stage at the initial loading stage, the hysteretic curves are linear, and the loops are narrow. The hysteresis loops become spindle-shaped, and the loops open gradually with the expansion and connection of cracks in the cast-in-place RACSWs. A significant "pinch effect" occurs at the zero point. The hysteretic curves become fully arched and have a reverse S shape after the peak load because of the large area of concrete falling off in diagonal compressive strips and the plastic deformations of the steel beam and columns. The areas enclosed by the loops increase. The bearing capacity of the specimens decreases noticeably under the same load.

(2) Comparison of the hysteretic curves of the SPE1 infilled ordinary concrete wall (Figure 11a) and SPE2 infilled RACSWs with a 100% replacement rate of recycled coarse aggregate (Figure 11b) show that the area and shape surrounded by hysteretic curves and peak loads of the structure are close, thereby indicating that the performance of RACSWs is close to that of the ordinary concrete wall.

(3) The hysteretic curves of specimens SPE2 (Figure 11b) and SPE3 (Figure 11c) are relatively close, thereby showing that the connecting stiffness of BCJs slightly influences the hysteretic behavior of the specimens of infilled cast-in-place RACSWs.

**Figure 11.** Hysteretic curves.

#### 4.1.2. Prefabricated RACSWs

Figure 11d,e depict the following:

(1) The stiffness of the specimens (Figure 11d,e) is high at the initial loading stage, and the hysteretic curves are linear. Moreover, no residual deformation occurs after unloading. With the expansion and connection of cracks in the prefabricated RACSWs, the stiffness of the specimens starts to decline and the loops open gradually. Next, a significant "pinch effect" occurs at the zero point. The energy dissipation capacity of the structure is increased, the enclosed areas of the loops increase, and the hysteresis loops are spindle-shaped because of the crushing and collapse of the corner concrete on the wall, the local buckling of the steel frame, and the relative slip between the connectors. Residual deformation occurs after unloading. The hysteretic curves become fully arched after reaching the peak load. The bearing capacity of the specimens decreases under the same load due to the large relative slip among the connectors.

(2) The hysteresis curves of SPE4 (Figure 11d) and SPE5 (Figure 11e) are nearly coincidental, indicating that the connecting sti ffness of BCJs has an insignificant e ffect on the hysteretic behavior of the specimens of infilled prefabricated RACSWs.

#### *4.2. Skeleton Curves*

The load-displacement skeleton curves of the specimens are illustrated in Figure 12. The loads of the main characteristic points are summarized in Table 4, where *P*cr, *<sup>P</sup>*y, *P*max, and *P*u are the cracking, yield, peak, and damage loads of the specimen, respectively, and *P*u = 0.85 *P*max. Figure 12 and Table 4 demonstrate the following:

(1) The skeleton curve of the pure steel frame is relatively smooth. The skeleton curves of the specimens are S-shaped when the cast-in-place RACSWs are infilled. Compared with SPE2 infilled RACSWs with a 100% replacement rate of recycled coarse aggregate, the cracking load of SPE1 infilled ordinary concrete wall increases by 37%, the average yield load decreases by 22%, and the bearing capacity is nearly the same.

(2) The bearing capacity of SPE2 is 2.4 times higher than that of the pure steel frame. The load decreases faster in SPE2 and SPE3 than in the pure steel frame after the peak load, demonstrating that the ductility of the specimens of infilled cast-in-place RACSWs decreases slightly.

(3) The comparison of SPE2 and SPE3 demonstrates that the concrete cracking load of the specimen is approximately 1.25 times higher in end-plate joints than in welded–bolted joints. The yield and peak loads decrease by 13% and 8%, respectively, showing that the connecting sti ffness of BCJs slightly influences the bearing capacity of the specimens of infilled cast-in-place RACSWs.

(4) The bearing capacity is 1.44 times higher in SPE4 than in the pure steel frame when the prefabricated RACSWs are infilled, thereby indicating that the prefabricated RACSWs can e ffectively improve the bearing capacity of the structure.

(5) The skeleton curves of SPE4 and SPE5 are coincidental at the initial loading stage, and the peak load is only 4% lower in SPE5 than in SPE4, emphasizing that the connecting sti ffness of BCJs slightly influences the bearing capacity of the specimens of infilled prefabricated RACSWs. The load of SPE4 and SPE5 decreases smoothly after the peak load, thereby indicating that the structure of the infilled prefabricated RACSWs has a high safety reserve.


**Table 4.** Loads of main characteristic points on skeleton curves.

**Figure 12.** Skeleton curves.

### *4.3. Sti*ff*ness Degradation*

The secant stiffness of the first cycle under the same load is calculated to reflect the degradation law of the stiffness of the specimen under cyclic loading. The formula is

$$K = \frac{|P+| + |P-|}{|\Delta+| + |\Delta-|} \tag{1}$$

where *P*+ and *P*− are the positive and negative horizontal loads at the vertex under the same load, respectively; and Δ+ and Δ− are the corresponding positive and negative horizontal displacements at the vertex under the same load.

The stiffness degradation curves of the specimens are presented in Figure 13. The values of stiffness on the main stages are provided in Table 5, where θ is the horizontal drift angle of the specimen, and *K*0 is the initial stiffness of the specimen. Figure 13 and Table 5 present the following:

(1) The stiffness degradation curve of the pure steel frame is relatively smooth. The initial stiffness of SPE2 is 4.3 times higher than that of the pure steel frame when the cast-in-place RACSWs are infilled and approximately 7% lower than that of SPE1 infilled ordinary concrete wall. The comparison of SPE2 and SPE3 implies that the initial stiffness of the specimen is approximately 13% lower in end-plate joints than in welded–bolted joints, and the degradation trend of the stiffness of two specimens is basically the same.

(2) The stiffness of the specimens of infilled cast-in-place RACSWs degrades rapidly at the initial loading stage. With the increase in horizontal load, wall cracks occur and continue to expand. BCJs exhibit a slight rotation, and the stiffness degradation rate of the specimens decreases. The walls are severely damaged and gradually fail after the peak load. The drift angle of the BCJs increases, and the steel frames are used as the second seismic fortification lines to dissipate the seismic energy. The stiffness degradation of the specimens of infilled cast-in-place RACSWs stabilizes.

(3) The initial stiffness is 2.8 times higher in SPE4 than in the pure steel frame when the prefabricated RACSWs were infilled. The stiffness degradation curves of SPE4 and SPE5 are coincidental, indicating that the connecting stiffness of BCJs has an insignificant influence on the stiffness of the specimens of infilled prefabricated RACSWs. The stiffness of the specimens is degraded rapidly at the initial loading stage. The stiffness degradation rate of the specimens decreases, and the stiffness of the specimens of infilled prefabricated RACSWs is steadily reduced by expanding and connecting the cracks in the walls.

**Figure 13.** Stiffness degradation curves.


**Table 5.** Values of stiffness on main stages.
