1. Introduction
The construction industry’s industrialisation has fast development [
1], and the advancement of prefabricated buildings is a significant manifestation of this trend. In prefabricated buildings, PC shear wall members are crucial constituents, and numerous scholars have conducted extensive experimental studies on them, providing a solid theoretical foundation, including studies on the mechanical performance of joints and structural measures [
2]. Moreover, with the global trend towards zero-carbon and low-carbon development concepts, the construction industry adopts more eco-friendly materials for PC members, such as shear walls, reflecting the industry’s transition towards sustainable development. Ultimately, achieving industrialisation in the construction industry, energy conservation, emission reduction, and low-carbon environmental protection aligns with the global consensus on sustainable development and represents the development trend of the construction industry [
3]. This study developed a new type of precast concrete sandwich shear wall by replacing ordinary concrete coarse and fine aggregates with high-titanium heavy slag and adding insulation boards. This innovative wall structure possesses advantages such as environmental and energy efficiency. Its widespread adoption in the construction industry is expected to generate significant economic, social, and ecological benefits, thereby promoting the advancement of the construction industry.
In research on precast concrete sandwich shear walls, addressing the traditional problem of external insulation for shear walls has been challenging. Typically, solutions involve retrofitting insulation boards onto the exterior of the shear walls. However, sandwich walls offer a more effective resolution to this issue [
4]. European countries adopted sandwich wall technology early in the construction industry. In response, various countries have developed comprehensive design schemes and issued a series of construction process specifications to regulate its implementation. For instance, detailed provisions exist regarding crack control in sandwich concrete wall panels in the Uniform Building Code and reinforcement placement within the concrete slab panels. The sandwich wall technology in Japan has been in development for a relatively long time and is reasonably well-established. In 1985, the Japanese Architectural Standard Specification JASS 14 Curtain Wall was promulgated and implemented [
5], stipulating the design, fabrication, construction, and acceptance criteria for sandwich walls. The Japanese Architectural Standard Specification JASS 10 [
6] Precast Concrete Work [
6] was also issued specifically for residential construction. Bush and Stain [
7] fabricated precast sandwich insulation panels with truss reinforcement and researched their flexural performance. Their results indicate that when the connectors are parallel to the wall, the wall exhibits good bearing capacity and flexural performance.
Kang [
8] proposed a design model for sandwich wall panels (SWPs) with an insulating core supported by glass-reinforced polymer (GRP) grids. In this model, the ultimate limit state capacity and serviceability limit state capacity are unified into a single capacity factor, meeting the requirements for the ultimate limit state and the serviceability requirements. Choi [
9] studied the composite performance of insulated sandwich concrete wall panels (ISCWPs) under monotonic and wind loads. Their findings suggest that rough surfaces and good adhesive properties of insulation materials enhance resistance to monotonic and cyclic loads. Studying the life cycle energy consumption of sandwich wall panels with different connectors, Dong [
10] concluded that reducing the thermal conductivity of the structural layers can improve the impact of connectors on the lifecycle of the wall. Through a feasibility study on lightweight concrete sandwich wall structures incorporating expanded polystyrene (EPS), Fernando [
11] found that converting non-biodegradable waste materials into high-quality building materials results in foam concrete. Combined with cement fibre boards, it produces lightweight wall panels suitable for load-bearing walls in single-story buildings and non-load-bearing walls in multi-story buildings.
Kumar [
12] investigated the structural performance of polymer concrete sandwich wall panels reinforced with basalt fibre-reinforced polymer (BFRP) grids under concentric axial loads. Their results indicated that the slenderness ratio significantly affects the ultimate axial load of the sandwich wall. At the same time, the longitudinal spacing of connectors does not significantly impact the ultimate load. The theoretical axial load capacity closely matches the experimental axial load capacity. Rosenthal [
13] conducted axial compression tests on composite walls made of lightweight insulation aggregate, finding that the specimens did not fail due to compressive buckling but rather due to vertical cracks in the overlap sections of the supports. Therefore, it is suggested that transverse steel bars be installed at the supports of the panels to mitigate this issue. Mohamad [
14] conducted eccentric compression tests on double-symmetric and single-symmetric truss sandwich walls. Double-symmetric truss sandwich walls exhibit higher bearing capacity and smaller deflection than single-symmetric ones.
Pavese [
15] conducted seismic tests on precast sandwich walls, simulating full-scale shear walls with and without openings. From the performance and failure mode of the wall, it can be inferred that there is a strong coupling effect between bending and shear. During the experiment, both strength and stiffness exhibited slow degradation. Gara [
16] conducted eccentric compression tests, axial compression tests, and diagonal compression tests on precast sandwich walls, showing that the bearing capacity under axial compression is higher, the results for ultimate load buckling and linear buckling are similar, and the ultimate load is lower under eccentric compression. In diagonal compression tests, mesh reinforcement within the concrete leads to a higher cracking load. Amran [
17] conducted axial compression tests on foam concrete sandwich walls and derived a formula for calculating their bearing capacity through finite element analysis. Choi [
18] studied the shear resistance performance of walls with sandwich layers using different insulation materials.
Benayoune [
19] conducted full-scale experiments to study the vertical bearing capacity of sandwich wall panels. Benayoune [
20,
21] studied fabricated sandwich walls with FRP (fibre-reinforced polymer) connectors and their flexural performance. Woltman [
22] researched the design and construction methods of sandwich insulation wall enclosures, which primarily included the design of connectors and the selection and application of reinforcement, concrete, and insulation materials. Li [
23] researched the fire resistance performance of precast concrete connected with FRP connectors and concrete with outer and inner reinforcement. Zhong [
24] conducted experiments on shear walls with different numbers of stirrups in edge members and found that plastic hinges concentrated at the horizontal joints of the specimens. The specimens exhibited sufficient strength and stable hysteretic characteristics. In the later stages of testing, the failure of tensile reinforcement resulted in insufficient utilisation of the concrete’s inherent strength, reducing the ductility of the specimens.
Zhu [
25] conducted out-of-plane static tests on sandwich insulation walls, showing that the walls exhibited high bearing capacity and crack resistance, with significant safety margin and flexural stiffness. Zhong [
26] researched the out-of-plane flexural performance of sandwich insulation walls utilising glass fibre composite connectors. The study revealed that the sandwich insulation walls exhibit a certain degree of composite behaviour, with the degree of composite behaviour decreasing gradually as the insulation board and concrete slab fail. Additionally, Zhong’s [
27] research indicated that using equal-strength lap splices for horizontal connections primarily results in swaying deformation under seismic action. By increasing the area of the lap-spliced steel bars, the connection becomes stronger, thus enhancing the bearing capacity of the wall. He [
28] conducted finite element analysis on sandwich insulation walls using different insulation materials and thicknesses, showing the effectiveness of foam concrete as an insulation layer in reducing wall deformation and stress.
In the research on the material application for precast concrete shear walls, using industrial waste to replace natural aggregates such as crushed stone, sand, and other admixtures in ordinary concrete is a crucial aspect of promoting green and low-carbon development in the construction industry. The ICC regulations enacted in the United States [
29] prohibit the classification of blast furnace slag as waste for disposal. Before 1980, Japan and the United States [
29,
30] primarily used blast furnace slag as a material for subgrade. After reaching saturation, blast furnace slag gradually began to be used as a concrete additive, cement admixture and mix in concrete aggregate. In the Netherlands [
30], concrete structures commonly incorporate slag cement, with a slag content ranging from 65% to 70%. The sales volume of Portland blast furnace slag cement is exceptionally high. It is important to note that slag cement can enhance the quality of structures, with a particularly noticeable improvement in the durability of concrete structures. A research study by Kumar [
31] on various properties of alkali-activated slag concrete and slag cement concrete under high-temperature conditions indicated that under high-temperature conditions, alkali-activated slag concrete exhibits superior properties compared to slag cement concrete in various aspects.
Acıkök’s [
32] study on the impact of fly ash and slag on pavement suggests that pavement concrete incorporating slag performs better than other alternatives. Li [
33] used the phenolphthalein method to study the specific carbonation depth of high-titanium slag concrete. In India, steel production is approximately 50 million tons, and engineers are increasingly focusing their research on blast furnace slag. Replacing fine aggregate with ground granulated blast furnace slag and blast furnace slag powder has shown significant progress in various strength properties [
34,
35,
36], as indicated by experimental studies.
In China, the Pangang Iron and Steel Group has been utilising high-titanium heavy slag during the smelting of vanadium-titanium magnetite as an industrial waste. This is seen as a practical approach to environmental management. High-titanium heavy slag concrete is prepared by transforming it into coarse and fine aggregates and finely ground admixtures. This innovation addresses the waste disposal issue and drives advancements in slag concrete application. Further utilisation of this concrete to create high-titanium heavy slag concrete members marks a new stage in developing slag concrete applications. Through this series of innovative practices, new vitality is injected into environmental protection and the sustainable development of the construction industry. Sun [
37,
38,
39], have extensively researched using high-titanium heavy slag in concrete for construction purposes, resulting in many achievements. Die [
40] studied the durability of high-titanium slag concrete by incorporating basalt fibres and plastic steel fibres. The research showed that fibres enhance high-titanium slag concrete’s durability and mechanical properties. Zhou [
41] studied the influence of fly ash and cement dosage on the heat resistance of high-titanium slag concrete. The results indicate that high-titanium slag concrete can withstand temperatures up to 500 °C by including an appropriate amount of fly ash.
Mu [
42] developed C50 high-titanium slag concrete using 11 bridges, including the Zhonghuashan Bridge and Sanjingtang Bridge, as engineering backgrounds. The main beam of the Jinshajiang Bridge on the Lipan Expressway is constructed using C65 high-titanium slag concrete. Li [
43] used high-titanium slag as a coarse aggregate to produce high-titanium slag columns and studied their seismic performance. The results showed that the seismic performance of high-titanium slag columns is essentially the same as that of ordinary concrete columns. Pang [
44] studied the properties of ultra-high-strength concrete-filled steel tubes prepared using high-titanium slag sand. The research indicates that the workability is enhanced after two hours, and the mechanical properties improve as the water-reducing rate increases. The specimens all exhibited shear failure, and the peak bearing capacity increased.
Wang [
45] conducted uniaxial compression tests and numerical simulations on high-titanium slag concrete. The experimental results showed that its compressive strength is slightly higher than that of ordinary crushed stone concrete. Huang [
46] conducted axial compression tests on high-titanium slag concrete columns. The results indicated that the member has a high ultimate compressive bearing capacity. Gong [
47] studied the influence of varying amounts of high-titanium slag on concrete properties. The results indicated that as the content of high-titanium slag sand and macadam increases, the slump value of the concrete increases. The compressive strength of the concrete initially increases and then decreases. The water-to-binder ratio has a significant impact, while the influence of high-titanium slag sand, high-titanium slag macadam, and fly ash is relatively tiny.
In Liang’s [
48] study, factors such as water-to-binder ratio and concentration of composite salt solution were investigated regarding the durability of high-titanium slag concrete. The results indicate that the concentration of composite salt solution has the most significant impact, and an increase in the number of freeze-thaw cycles leads to a decrease in the dynamic modulus of elasticity of the specimens. Guo [
49] investigated the resistance to chloride ion penetration and high-titanium slag concrete frost resistance. The results revealed that the resistance to chloride ion penetration and frost resistance of high-titanium slag concrete meet the durability specifications for concrete. However, most existing research on high-titanium slag is based on its fundamental properties when used in concrete. Very little research is on using high-titanium slag as concrete aggregate for preparing precast concrete members. Furthermore, no study has been conducted on using high-titanium slag industrial solid waste material to prepare concrete for fabricating sandwich shear walls. This article aims to fill the academic voids.
The research objective of this project is to develop a novel type of precast concrete sandwich shear wall by replacing the coarse and fine aggregates of ordinary concrete with high-titanium heavy slag and incorporating insulation boards. Additionally, it aims to investigate the seismic performance of the shear wall. Through theoretical and experimental research, this study examines the new material’s influence on the seismic performance of sandwich composite shear walls. The goal is to validate that the physical and mechanical properties of the high-titanium heavy slag concrete sandwich composite shear wall are equal to, and may even exceed, those of ordinary concrete composite shear walls. This research will provide theoretical and experimental evidence to advocate for this type of wall’s widespread adoption in construction projects of prefabricated buildings.
The novelty of this research is to study the seismic performance of sandwich shear walls made of high titanium slag concrete. Comparative studies will be conducted through pseudo-static tests and finite element analysis. It contributes to developing a novel composite shear wall that meets specification standards and lays a foundation for its deployment in precast concrete construction projects. Unlike other similar research endeavours, the distinctiveness of the testing protocol in this experimental study lies in the evaluation of various mechanical performance indicators post experiment. These include analysing the wall’s load-bearing process and failure characteristics, studying hysteresis and skeleton curves, assessing bearing capacity and deformation capacity, analysing stiffness degradation curves, and examining the cage of reinforcement curves. Utilising the ABAQUS 2022 finite element software, a finite element analysis was conducted on the precast high-titanium heavy slag concrete composite shear wall. Subsequently, the ABAQUS analysis data were then compared with the results of pseudo-static tests to validate the feasibility and applicability of the finite element model of the precast high-titanium heavy slag concrete sandwich composite shear wall.
2. Seismic Performance Test of High-Titanium Heavy Slag Concrete Sandwich Composite Shear Wall
2.1. Objective
The objective of the test is to examine the seismic resistance of the specimens under both vertical and horizontal loads. A pseudo-static test was conducted by producing one PHCSPW and one CHCW specimen. Measurements and analysis were performed on the bearing capacity, failure mode, deformation characteristics under load, stiffness, and other parameters of the specimens. A comparative study was conducted to identify the differences between CHCW and PHCSPW.
2.2. Test Materials
The test used Pangang industrial solid waste high-titanium heavy slag macadam and ordinary macadam as coarse aggregates and high-titanium heavy slag sand and ordinary river sand as fine aggregates. PC42.5 composite Portland cement from Yunnan of ChinaYimen Dachun tree cement Co., Ltd. (Yuxi, China), was used. In contrast, Class II fly ash produced by Gongyi of China Hengnuo Filter Material Co., Ltd. (Gongyi, China), was used as the fly ash. After studying the effect of high-titanium slag content on concrete properties [
47], the optimal mix ratio was determined to be cement: sand: macadam: water at 1:1.23:3.01:0.41, with high-titanium slag content accounting for 40% of the coarse and fine aggregates. The average compressive strength of the high-titanium slag concrete is 45.872 MPa, indicating good workability and meeting the strength requirements.
2.3. Dimension and Reinforcement of Specimens
According to the specifications [
50,
51,
52], the experiment was designed to test two reduced-scale shear walls: one cast-in-place high-titanium heavy slag concrete wall (CHCW) and one precast high-titanium heavy slag concrete sandwich panel wall (PHCSPW). Both consist of loading beams, walls, and ground beams. The detailed dimensions and specific reinforcement for CHCW are shown in
Figure 1, while the detailed dimensions and specific reinforcement for PHCSPW are shown in
Figure 2.
The test was conducted on two specimens under horizontal and constant vertical loads combined. The main variable parameters of PHCSPW include (1) the arrangement of connectors, as shown in
Figure 3, with a spacing of 300 mm between connectors, using FRP and Z-shaped metal connectors to effectively alleviate the thermal bridge effect at the connection of sandwich walls, and (2) the insulation layer setup is shown in
Figure 4. The insulation layer is installed to achieve energy efficiency while ensuring the wall’s bearing capacity.
The members are designed according to the strong shear and weak bending principle, with differences compared to CHCW. For the edge members, longitudinal reinforcement and vertical distribution reinforcement are used for seismic anchoring. According to the specifications [
52], the length of longitudinal reinforcement seismic anchorage is calculated according to Formula (1):
In the formula, ζaE represents the anchorage length seismic coefficient, which is taken as 1.05 for a seismic grade of three.
la represents the basic anchorage length of longitudinal reinforcement.
The basic anchorage length of longitudinal reinforcement is explicitly calculated using Formula (2):
In the formula, ζa represents the correction coefficient for anchorage length.
a represents the reinforcement shape factor: for ribbed bars, it is typically 0.14.
fy represents the design tensile strength of reinforcement: for HRB400, it is 360 MPa.
ft represents the design compressive strength of concrete: for C40, it is taken as 1.71 Mpa.
d represents the diameter of the anchorage reinforcement.
2.4. Specimen Fabrication
Fabricating the specimens involves the following steps: formwork, application of strain gauges, reinforcement binding, installation of connectors and insulation boards, concrete pouring, and curing. The fabrication process and the final formed wall specimens are depicted in
Figure 5a–j.
CHCW adopts horizontal casting. The fabrication process and the final formed wall specimens are shown in
Figure 6a–f.
2.5. Test Point Arrangement
The strain gauges were attached at the edges of the specimens. The specific arrangement of strain gauges on the PHCSPW is shown in
Figure 7, and the arrangement of strain gauges on the CHCW is shown in
Figure 8.
Figure 9 shows a schematic diagram of the displacement meter arrangement.
Figure 10 depicts the physical layout of displacement meters.
2.6. Data Collection
Additional temperature compensation pieces were set for the reinforcement. Before the test, the wall was painted white and divided into squares with a surface size of 50 mm × 50 mm. The test process evidenced and recorded the distribution of cracks. The static stress-strain meter collected various test values during loading, as shown in
Figure 11. The recording was performed as shown in
Figure 12.
2.7. Loading Device
The girder, loading beam, and a 100-ton jack transmitted the vertical force of the reaction. The horizontal force of the response was composed of a reaction wall, force transmission steel beams, tie rods, and a 50-ton actuator. During the test, to avoid eccentricity caused by vertical load, the jack’s centre should be aligned with the centre of the distributing beam and the centreline of the distributing beam with the centreline of the concrete. The loading device is shown in
Figure 13.
2.8. Loading Scheme
The test adopted a load-displacement mixed control. According to the specifications [
53], preloading should be performed twice reciprocally before formal loading while ensuring the normal functioning of instruments and meters. The load value of the preload should be maintained below 30% of the theoretical cracking load. Since the specimen has yet to yield during load control, each load level only required one cycle. The vertical load N to be applied by the jack was calculated according to Formula (3), and the values of N are shown in
Table 1.
In the formula, nd is the −design axial compression ratio.
A is the −cross-sectional area of the wall.
fc is the −design compressive strength of concrete, calculated based on parameters related to C40.
The axial load is graded, followed by low-cycle repeated loading conducted at the top of the member, with the line of action of the horizontal force coinciding with the centerline of the loading beam. When displacement control is applied, each displacement level necessitates three cycles as the specimen enters yielding. When the horizontal bearing capacity drops below 85% of the peak bearing capacity during the test, it is considered that the member has failed, and the test is stopped [
54]. The schematic diagram of the loading protocol for the test is shown in
Figure 14.
The horizontal displacement Δ is the displacement of the vertex at the height of the loading line and the angle of the vertex displacement. Loading to the north (N) is considered positive while loading to the south (S) is harmful. Loading once to the north and once to the south constitutes one cycle.
Table 1.
Experimental axial load capacities (KN).
Table 1.
Experimental axial load capacities (KN).
Test Item | | | A (Width × Thickness)/mm × mm | |
---|
PHCSPW | 0.15 | 19.1 | 800 × 250 | 478 |
CHCW | 0.15 | 19.1 | 800 × 200 | 382 |
2.9. Loading Process and Failure Characteristics
The test revealed that most cracks in the wall occurred at the base of the wall. Vertical loads of 382 kN and 478 kN were applied on CHCW and PHCSPW, respectively, which remained constant throughout the test. During load control, the applied loads were set to 80 kN, 100 kN, 120 kN, 140 kN, 160 kN, and 180 kN, with each level undergoing one positive-negative cycle. During displacement control, the loading displacement is multiplied by the yield displacement and was set at 16.70 mm, 25.05 mm, 33.40 mm, 41.75 mm, and 50.10 mm, with each level undergoing three cycles and the final level undergoing one cycle, after which the test was stopped. The failure characteristics of the specimens were generally similar, with cracks appearing first at the base of the wall under the repeated action of horizontal forces. As the load increased, the cracks in the wall extended, and the width of the cracks at the base increased.
2.9.1. Failure of CHCW Specimen
When the horizontal load increased to 80 kN, the first crack appeared at the base of the CHCW. When the loading increased and horizontal displacement reached 16.70 mm, multiple cracks gradually appeared below 500 mm of the wall height. At this point, the cracks at the base began to propagate horizontally. As the loading continued, new cracks appeared intermittently along the wall, and the cracks at the base continued to extend and develop into primary cracks. The development trend of the wall cracks was mainly horizontal, and the crack propagation pattern was relatively curved. The development trend of concrete cracks on both sides of the wall was consistent. The length of the main crack on one side of the specimen at the time of failure was 308 mm, while on the other side, it was 302 mm, indicating a fundamental similarity in length. At the time of failure, the distribution of cracks is depicted in
Figure 15. The member’s height is 1.2 m, and the Figure−only illustrates approximately 0.5 m of the wall height. It is observed that the concrete cracks on the left and right sides essentially correspond to each other.
2.9.2. Failure of PHCSPW Specimen
When the horizontal load reached 160 kN, the PHCSPW specimen exhibited the first crack at its base. Upon reaching a horizontal displacement of 16.70 mm, new cracks gradually appeared around the height of 100 mm along the wall. At this point, the cracks at the base started to propagate horizontally. New cracks continued to appear on the wall, while base cracks kept extending and developing into the main crack. Crack development exhibited a curved pattern in the wall, with essentially consistent development trends on both sides. The length of the main crack on one side of the specimen at the time of failure was 347 mm, while on the other side, it was 340 mm, indicating a fundamental similarity in length. At the time of failure, the distribution of cracks in the member is depicted in
Figure 16. The member’s height is 1.2 m, and the Figure−only illustrates approximately 0.5 m of the wall height. It was observed that the concrete cracks on the left and right sides are relatively symmetrical and essentially correspond to each other.
The specimen’s early appearance of base cracks was attributed to the maximum stress concentration at the base. Cracks concentrated in the lower portion, mainly below 500 mm from the base. Following the tensile failure of the base concrete, reinforcement bore the base bending moment. The reinforcement yielded as the load gradually increased, leading to increased deformation and crack widening.
2.10. Analysis of Hysteresis Curve and Skeleton Curve
Reinforced concrete materials exhibit nonlinear behaviour under significant loads, displaying residual strains after unloading. In pseudo-static tests, the load-displacement curve is the hysteresis curve. After removing the influence of ground beam sliding and rotation on the test, the relationship between the horizontal force F of the member and the vertex displacement Δ is obtained. Its hysteresis curve, shown in
Figure 17, commonly exhibits four different shapes. The shuttle shape is fuller and indicates better seismic performance. Due to the influence of slippage, the reverse S-shape is incomplete, indicating poorer seismic performance. The bow shape is also relatively full. However, due to the influence of slippage, the appearance of the pinch effect diminishes its fullness compared to the shuttle shape. Nonetheless, it still exhibits good seismic performance. Due to significant slippage, the Z-shape appears less complete, indicating poor seismic performance of the specimen.
CHCW concrete remained crack-free in the elastic stage, resulting in a minor hysteresis loop area. After unloading, its residual strain was relatively small, showing a shuttle-shaped hysteresis loop. After cracking, vertex displacement increased, enlarging the hysteresis loop area. Despite increased residual strain during uploading, the shuttle shape persisted. After yielding, the hysteresis loop area expanded rapidly as cracks continued to extend, accompanied by a slight bond-slip of the steel bars. At this point, the hysteresis loop exhibited noticeable pinching. The PHCSPW concrete remained crack-free in the elastic stage, resulting in a minor hysteresis loop area. After unloading, its residual strain was relatively minor, and the hysteresis loop exhibited a shuttle shape.
After cracking, as the displacement increased, the hysteresis loop assumed a bow shape, indicating a slight degree of bond-slip in the connectors. With increased displacement and the number of cycles, shear slip increased. At this point, the hysteresis loop exhibited a reverse S shape. Neither specimen exhibited a Z-shaped hysteresis loop, indicating that no significant shear slip occurred in either specimen. During displacement-controlled loading, stiffness and bearing capacity during the third and second cycles of each loading level were lower than during the first cycle, indicating a degradation of stiffness of the shear walls during the test. After reaching the peak bearing capacity, the degradation of stiffness intensified. Comparing the hysteresis loops of CHCW and PHCSPW, PHCSPW exhibited a larger area, indicating superior energy dissipation capacity.
The skeleton curves, shown in
Figure 18, determined the yield point using the equivalent energy method. Point A, obtained by intersecting a horizontal line passing through the peak point and a line passing through the origin, ensured equal areas on both sides. A perpendicular line is drawn through point A, intersecting the curve at point B. This point represents the yield point, where the load equals yield load, and the displacement equals yield displacement.
Figure 18 shows that the trends of the two skeleton curves are similar, with the skeleton curves of the specimens in both directions being roughly symmetrical. They both undergo four stages, from elasticity to failure. Before member cracking, the skeleton curve is linear. After cracking, a clear inflexion point emerges due to decreased member stiffness. As the member gradually yields, its curve slope decreases. In the later stages of the test, the curve becomes smoother, indicating a gradual reduction of bearing capacity, suggesting good ductility. In the initial stage, PHCSPW exhibited a higher initial stiffness, which was attributed to the presence of the outer leaf wall, resulting in a higher initial stiffness for PHCSPW. After cracking, the outermost vertical reinforcement of the specimen began to yield, indicating that the specimen had entered the yield stage with a significant decrease in stiffness. Under low-cycle repeated horizontal loading, the cracks at the base of the shear wall widened progressively. Initial concrete crushing occurred in the outer leaf wall, while in CHCW, concrete crushing was almost simultaneous, resulting in lower bearing capacity for CHCW. Peak bearing capacity under reverse loading was slightly lower than under forward loading, indicating stiffness degradation during reverse loading.
2.11. Analysis of Bearing Capacity and Deformation Capacity
The measured bearing capacities are listed in
Table 2. The characteristic values of the bearing capacity of the specimens under cyclic loading include yield load, cracking load, peak load, and ultimate load, each corresponding to specific displacements. The determination of the cracking load is based on the following conditions: If the first crack appears during the load application, the cracking load is taken as the average of the previous and current load levels. If the first crack appears at the end of the load application, the cracking load is taken as the current load level. The first sudden change in the slope of the skeleton curve indicates the cracking load, at which point the displacement corresponds to the cracking displacement. The apparent inflection point of the skeleton curve determines the yield load. At this point, the load is the yield load, and the displacement corresponds to the yield displacement. If there is no obvious inflection point, approximation methods such as graphical, energy, or double straight-line methods can be used. The load and displacement at yield are determined using the energy method. The peak load is determined by finding the maximum value on the skeleton curve, with the corresponding displacement being the peak displacement. The ultimate load is determined when 85% of the peak load equals the maximum load applied during the test. The corresponding displacement at this load level is considered the ultimate displacement. If the maximum load applied at the end of the test still exceeds 85% of the peak load, end-test displacement is considered.
Table 2 displays each specimen’s peak load Fp, yield load Fy, yield-to-peak load ratio, and ultimate load Fu. Comparative analysis of the bearing capacity of the two specimens provides a clear understanding of the differences between PHCSPW and CHCW in bearing capacity. It can be observed that the yield load of PHCSPW is slightly greater than that of CHCW because the inner and outer leaf walls of PHCSPW are jointly subjected to force through connectors. The yield load ratio to the specimens’ peak load ranges from 0.636 to 0.888, indicating no significant difference in the initial stiffness of the specimens. The peak bearing capacity of CHCW under forward loading is approximately 68.02% of PHCSPW, with an average peak bearing capacity of around 68.31% of PHCSPW. During the loading process, the shear walls did not reach the ultimate load in the negative direction, while the walls had already reached the ultimate load in the positive direction.
A structure must meet bearing capacity and stiffness requirements during its expected service life. In seismic design, structures are typically assessed against frequent and rare earthquakes, accounting for elastic and elastic-plastic deformations. Ductility serves as a safety margin under significant loads. The ductility coefficient is obtained by dividing the ultimate displacement by yield displacement, comprehensively reflecting a member’s performance under various conditions, indicating its deformation capacity during the inelastic phase. The characteristic point displacement, ultimate displacement angle, and ductility coefficient of the specimens are shown in
Table 3. The yield displacement angles of the specimens are close, ranging from 1/120 to 1/50 [
51], indicating that the influence of the sandwich layer and the outer leaf wall on the yield displacement angle is insignificant. The peak displacement angles are close between 1/41 to 1/32. The peak displacement of PHCSPW is 1.125 times that of CHCW, with a slight difference in ultimate displacement, meeting the Chinese specification requirements. A ductility coefficient of the specimen greater than 3 indicates good deformation capacity, and brittle failure will not occur under seismic action. CHCW’s favourable ductility coefficient indicates that its bearing capacity decreases slowly after yielding.
2.12. Stiffness Degradation Curve Analysis
Under cyclic loading with the same applied load, stiffness degradation manifests as increasing displacement with increasing cycles. Stiffness degradation is the cumulative damage experienced by the member under repeated loading, and it is also a concentrated reflection of material plasticity development and member cracking. The secant stiffness of the member is calculated according to Formula (4) [
54].
In the formula, Ki represents the second stiffness of the member for the i-th cycle.
Fi, − Fi represents the positive or negative horizontal load corresponding to the i-th loading cycle.
Δi, − Δi represents the positive or negative horizontal displacement corresponding to the i-th loading cycle.
As displacement increases, the stiffness of the specimens decreases. The faster the rate of stiffness degradation, the poorer the deformation capacity of the member. During the test process, the degradation rate varies at different stages. The stiffness variation of each specimen is shown in
Table 4. The change in the stiffness degradation curve with displacement angle is illustrated in
Figure 19.
Figure 19 shows that PHCSPW exhibits greater stiffness before cracking than CHCW due to outer leaf walls influencing its stiffness. After cracking, the stiffness of both specimens becomes equivalent. The stiffness degradation of the members occurs in three stages: first, before cracking, the stiffness of the specimen decreases rapidly; second, after cracking but before yielding, the decrease in stiffness slows down slightly; and third, after yielding, the stiffness degradation of the member is relatively slow. Upon reaching the ultimate load, the main crack of the specimen is formed, and its stiffness tends to stabilise. Initially, the stiffness degradation curves of both members roughly overlap. However, as the outer leaf wall cracks, PHCSPW experiences accelerated stiffness degradation.
2.13. Analysis of Cage of Reinforcement Curves
To analyse the stress performance of the vertical reinforcement, strain analysis of the reinforcement at the bottom of the specimens is conducted, as shown in
Figure 20.
Figure 20 shows the strain skeleton curve of the reinforcement at a distance of 200 mm from the top surface of the ground beam for both CHCW and PHCSPW specimens.
The variation patterns of the reinforcement strain skeleton curves of PHCSPW and CHCW are similar. The reinforcement remains elastic in the initial loading stage, displaying a linear strain curve. The strain increases rapidly after entering the elastic-plastic stage, indicating significant plastic deformation. When horizontal cracks appear, there is a sudden increase in the reinforcement strain, while strain variation during continuous cyclic loading remains relatively small. The curves show that the vertical reinforcement PHCSPW3 and PHCSPW7 strain values surpass the CHCW specimens’ corresponding values at the corresponding points.
In contrast, the vertical reinforcement PHCSPW1 and PHCSPW5 values are similar to the CHCW values at the corresponding points. Additionally, the strain value of the vertical reinforcement PHCSPW10 exceeds that of PHCSPW12. The connectors’ connection performance meets the requirements, effectively meeting the shared load bearing of the inner and outer walls.
4. Conclusions
The hysteresis and skeleton curves of PHCSPW and CHCW exhibited similar trends in development. After cracking, there was minimal difference between the bearing and deformation capacities. Comparing the hysteresis loops of CHCW and PHCSPW, it was evident that the hysteresis loop area of PHCSPW was more extensive, indicating that PHCSPW had better energy dissipation capacity. The ratio of yield load over peak load for the specimens ranged from 0.636 to 0.888, indicating that the initial stiffness difference of the specimens was relatively tiny. The yield load of PHCSPW was slightly greater than that of CHCW. The maximum bearing capacity of CHCW was approximately 68.31% of PHCSPW’s.
By comparison with the experiment in reference [
18], the failure mode and crack development of sandwich shear walls are similar to those of cast-in-place shear walls, and the cracks are concentrated below 0.5 m and distributed evenly along the wall height. Prefabricated reinforced concrete sandwich panels (RCSPs) under simulated seismic loading were assessed through a large experimental campaign. Tests were carried out on single full-scale panels with or without openings, simulating the behaviour of lateral resisting cantilever and fixed-end walls. Tests were also carried out on a two-story full-scale H-shaped structure constructed by individual panels that were properly joined together. The performance and failure mode of all panels tested revealed strong coupling between flexure and shear due to the squat-type geometry of the panels. However, due to their well-detailed reinforcement, all panels exhibited only a relatively gradual strength and stiffness degradation, and in no case did any panel suffer from sudden shear failure.
The ductility coefficient of CHCW was 3.211, while PHCSPW’s was 3.004, indicating that the ductility of PHCSPW was slightly lower than CHCW’s. The displacement angles corresponding to each load for PHCSPW were somewhat more significant than those of CHCW. The peak displacement of PHCSPW was 1.125 times that of CHCW, with a relatively small difference in ultimate displacement. Both CHCW and PHCSPW exhibited flexural-shear failure, indicating that the connections could effectively transmit forces under cyclic loading, and their connection performance met the requirements.
The ABAQUS simulations of PHCSPW and CHCW, followed by seismic performance testing, resulted in consistent findings with theoretical and experimental analyses regarding the model’s failure mode, stress-strain distribution, and ultimate bearing capacity. Before cracking, PHCSPW exhibited greater stiffness, while after cracking, PHCSPW experienced a faster rate of stiffness degradation. The hysteresis and skeleton curves of the specimens exhibited similar development trends, with the hysteresis loop area of PHCSPW being larger, indicating better energy dissipation capability. Both PHCSPW and CHCW experienced flexural-shear failure, indicating that the connection performance of PHCSPW connectors satisfies the requirements.
Through ABAQUS simulation with different parameters, it was determined that the strength of concrete has little influence on the energy dissipation capacity of the component, while the reinforcement ratio and axial compression ratio of the specimen have greater energy dissipation capacity.