1. Introduction
Compared to cast-in-situ concrete structures, precast concrete structures have the advantages of design-build efficiency, aesthetical versatility, reduced costs, low maintenance, and superior resistance to disasters [
1,
2,
3,
4]. However, many long-standing issues of precast structures, such as the defects of the connectors between precast components, always cause difficulties in the construction procedure and safety hazard under service as well as disasters. Although these issues need comprehensive research, most of the existing investigations are related to cast-in-situ structures and the studies on precast structures, especially those that focus on the connectors of precast structures, which have yet to be sufficient.
To guarantee the performance of precast concrete structures, the reinforcement bars in components of the structures need to be well spliced. Thus, the grouted splice connector has been widely studied by researchers from all over the world [
5,
6,
7,
8]. For instance, Parks et al. [
9] investigated the grouted splice connector, which connected a reinforced precast concrete bridge pier cap and a precast column in America. Li et al. [
10] studied the dynamic response of precast concrete beams connected by the grouted splice connector in China. Tullini and Minghini [
11] conducted an experimental research on the behavior of precast reinforced concrete column-to-column connections made with grouted sleeve splices in Italy. Ling et al. [
12] studied the behavior of the precast concrete wall panels, which was connected by a grouted splice connector in Malaysia.
Figure 1 is a sketch of a vertically-connected precast concrete shear wall system with a grouted splice connector. A sleeve is embedded at the bottom of the upper precast shear wall in advance during the fabrication process. In the field, the sleeve is placed onto the bar protruding from the top of the lower precast shear wall. Then, grout is poured into the sleeve through the grouting hole until the sleeve is full [
13]. The spliced bars must be aligned and positioned at the center of the sleeve for the tensile capacity of the connector to develop fully [
14]. The force in one bar is transferred from the grout to the sleeve, and, by the same way, to the other bar [
15]. Thus, the grout in the sleeve plays a critical role in force transfer. Therefore, it is essential to completely fill the sleeve inner cavity with grout.
According to the specification
JGJ 355-2015 [
16], for bars with diameters of 12–25 mm and 28–40 mm, the minimum differences between the sleeve inner diameter and the bar diameter are 10 mm and 15 mm, respectively. However, most sleeve inner diameters designed by practitioners are less than 50 mm [
17,
18,
19,
20], since a smaller sleeve diameter will provide sufficient confinement to achieve a higher bond strength between the bar and the grout, which increases the tensile capacity [
21,
22]. However, a small sleeve inner diameter always creates difficulties in on-site construction. One problem is that insufficient space between the bar and the sleeve cavity will cause an incomplete fill of grout (
Figure 2), which weakens the ability of the connector to transfer force. There is a more serious secondary problem. The protruded bar inevitably leans during transportation and construction. Usually, construction workers need to exert force to correct the inclination of the bar so that the bar can be inserted into the sleeve to align with the other bar. However, when the angle of the incline of the bar is relatively large, and the sleeve inner diameter is relatively small, some workers choose to cut off the inclined bar for ease of construction, which severely weakens the connection performance of the splice.
Grout-sleeve bond failure is a typical but undesirable failure mode in grout splice connector [
23,
24]. Sufficient bond strength is required between the grout and the sleeve to avoid the slippage of grout from the sleeve. However, a strong enough bond cannot form for a smooth sleeve inner surface because of the relatively weak chemical bond between the sleeve and the grout [
15]. Hence, numerous fabrication methods for sleeves have been developed to improve the bond strength between the grout and sleeve in grouted splice connectors, such as welding bars to the sleeve inner surface and tapering the sleeve head [
14], threading the inner surface of the sleeve [
25], and producing ribs on the interior surface of the sleeve [
18,
19,
26]. The complexity of these sleeve configurations requires advanced casting technology to fabricate the sleeve. In addition, the manufacturing cost of these sleeves is relatively high.
To avoid the difficulties caused by the small inner diameter of the sleeve and lower the production cost of the sleeves illustrated above, Yu and Xu [
27] developed a grouted sleeve lapping connector. The sleeve of this connector is cut from a standard pipe section and does not require additional fabrication, which is both economical and simple to produce. In addition, its inner diameter is up to 70 cm. In the grouted sleeve lapping connector, the two main bars are lapped, and the pipe is set along with the grout around the lapping splice to provide transverse confinement. A schematic of the grouted sleeve lapping connector is shown in
Figure 3. The construction process of the grouted sleeve lapping connector is similar to that of the grouted splice connector. The sleeve in the upper precast shear wall is placed onto the bar protruding from the lower precast shear wall. Thus, the two main bars are lapped. Then, grout is poured into the sleeve through the grouting hole until the pipe cavity is full.
Although the grouted sleeve lapping connector offers the advantages of a large sleeve inner diameter and low sleeve manufacturing cost, it is unknown whether the tensile capacity of the grouted sleeve lapping connector is larger or smaller than that of the grouted splice connector and if there is a difference between the failure modes of the two connectors. Thus, it is meaningful and essential to conduct experimental research to study the differences in tensile capacity and failure modes between the two connectors.
There are many significant typical mechanical properties of grouted splice connectors, such as the mechanism of force transfer [
28], the bond stress distribution along the embedded length [
13], and the confining mechanism [
18]. In-depth studies of these mechanical properties have been conducted for the grouted splice connector [
29,
30,
31,
32,
33,
34], which significantly promotes the application of this connector. Similarly, research on the mechanical properties of the grouted sleeve lapping connector must be conducted to provide a theoretical basis for its popularization. A previous study on the grouted sleeve lapping connector focused on its feasibility and working mechanism, and two equations were developed to calculate the average lapping bond stress and the critical lap length using linear regression [
27]. Many principal mechanical properties—the load transfer mechanism, the force state on the cross section, and the bond stress distribution along the lapped bars—were not investigated.
To fill in the gaps that were left by the previous study, 16 grouted sleeve lapping connectors and three grouted splice connectors were tested under monotonic conditions. Since the subject of this research is to study the mechanical behavior of the grouted sleeve lapping connector, the test on three grouted splice connectors can be considered as a reference, in which the purpose is to conduct a comprehensive analysis of the grouted sleeve lapping connector. The differences in the tensile capacity and the failure modes between the two connectors were studied, and the origin of these differences was explored. The mechanical properties of the grouted sleeve lapping connector were analyzed in terms of the bond stress distribution between the bar and the grout, the force state in the middle section, and the strain of the reinforcement bars and the sleeve. In addition, an approximate mechanical model was put forward to describe the mechanical properties of the grouted sleeve lapping connector.
3. Test Results
For the grouted sleeve lapping connector, the misalignment of the main bars generated secondary moments that resulted in the rotation of the sleeve and a slight bending of the bars outside the sleeve, as shown in
Figure 8. It is well-known that bar kinking produces local and early failure of the bond between the bar and grout [
19] and the results are conservative. If the specimen passes a test in which there is no restraint on rotation, the specimen will perform better under restrained rotation [
35]. In addition, in the field, the connector is surrounded and constrained by the concrete and stirrups in the precast shear wall or column such that the deflection of the connector is impeded. However, further studies in which the connector is constrained to avoid deflection need to be conducted.
The ultimate tensile capacity (
), the ultimate tensile strength (
), and the failure mode of all the specimens are listed in
Table 1.
3.1. Strength Evaluation
It is evident from the test results in
Table 1 that the tensile capacity of the specimens that failed by bar tensile fracture was close to that of the bare bar in tension.
According to
ACI-318 [
36] specifications, the tensile strength of a splice should be at least 125% of the nominal yield strength of the spliced bar. Therefore, the ratio of the tensile strength of the splice to the yield strength of the spliced bar,
Rs, should satisfy:
Rs ≥ 1.25. The strength ratings following this criterion are shown in
Table 1. It is shown that the specimens of the grouted sleeve lapping connector with lap lengths greater than or equal to 150 mm satisfied the previously mentioned criterion.
3.2. Failure Mode
Figure 9 illustrates the typical failure modes of the specimens. The typical failure modes for the grouted sleeve lapping connectors were bar tensile fracture and bar-grout slip. All of the grouted splice connectors failed by the mode of grout-sleeve slip.
The tensile capacity of the specimens was governed by the tensile capacity of the main bars, the bond capacity between the bar and the grout, and the bond capacity between the grout and the sleeve.
For the grouted sleeve lapping connectors of 150-4, 200-2, 200-3, the 250 series, and the 300 series, the lap length was long enough to provide a sufficient bond stress between the bar and the grout. Thus, the bond capacity between the bar and the grout for these specimens was larger than the tensile capacity of the main bars, which results in the failure mode of the bar fracture. The grouted sleeve lapping connectors of the 100 series, 150-1, 150-2, 150-3, and 200-1 had a short lap length, and the ultimate bond stress between the bar and the grout was not sufficient to prevent the bar from slipping out of the grout. Thus, these specimens failed the bar-grout slip. The grout-sleeve slip did not occur for the grouted sleeve lapping connectors.
For all of the grouted splice connectors, the ultimate bearing capacity depended on the bond strength between the grout and the sleeve. All of the specimens failed by the grout-sleeve slip as long as the load exceeded the ultimate bond between the grout and the sleeve.
3.3. Load-Displacement Curve
Figure 10a,b demonstrated the load-displacement responses of the grouted sleeve lapping connectors. For the specimens of the 100 series that failed by the mode of bar-grout slip, the displacement varied about linearly with the load increment initially and then failed suddenly when the load exceeded the tensile capacity of the specimen. Since the tensile strength of the specimens was smaller than the yield strength of the overlapped bars, the curve exhibited no plastic regime. Therefore, the specimens failed in a brittle manner.
Although the specimens 150-1, 150-2, 150-3, and 200-1 also failed by the mode of the bar-grout slip, the tensile strength of these specimens was larger than the yield strength of the bar. Thus, the bars were in the post-yielding state when the specimens failed, and their load-displacement curves exhibited yielding and hardening stages, which was different from the behavior of the specimens of the 100 series. Therefore, these specimens failed in a ductile manner.
Specimens of the 150-4, 200-2, 200-3, 250 series, and 300 series failed by bar tensile fracture and exhibited a ductile response. The load–displacement curves of these specimens that failed by the mode of bar fracture was basically the same as that of the bare bar. The curves exhibited an approximately elastic response at the initial stage. The stiffness gradually degraded as the rotation of the sleeve and the development of internal micro-cracks [
14]. However, the magnitude of this change was insignificant. As the bar yielded, a large displacement was produced for a small load increment. The curve began to descend after the load exceeded the tensile capacity of the specimens.
The load-displacement curves of the grouted splice connectors are shown in
Figure 10c. Since the ultimate tensile strength was smaller than the bar yield strength, all of the specimens failed in a brittle manner. For the specimens S-200-1, S-200-2, and S-200-3, the tensile load first increased to an initial peak and then decreased to a certain value. Subsequently, the tensile load increased to a second peak as the displacement increased. Lastly, the tensile load decreased suddenly, and the loading process ended. The curve can be explained by the following destruction mechanism: immediately after the load reached the first peak, the grout was completely pulled off at the middle section and split into two pieces. Subsequently, the load was carried by the bond between the separated grout and the sleeve until the grout was pulled out of the sleeve. The second peak corresponded to the maximum bond strength between the separated grout and the sleeve.
6. Conclusions
Experimental studies were performed on 16 grouted sleeve lapping connectors and three grouted splice connectors under tensile loading. The differences in the tensile capacity and failure modes between the grouted splice connector and the grouted sleeve lapping connector were studied. Analysis of the mechanical properties of the grouted sleeve lapping connector was conducted. The following conclusions can be drawn based on this research.
(1) When the inner surface of the sleeve was smooth and all of the parameters other than the bar structural form were the same, the tensile capacity of the grouted sleeve lapping connector was 2.45 times that of the grouted splice connector, which was explained in terms of the different load transfer mechanisms of the two connectors.
(2) Different bar configurations resulted in different failure modes for the connectors. All of the grouted sleeve lapping connectors with lapped bars failed by a bar tensile fracture or bar-grout slip, whereas the only failure mode of the grouted splice connectors with aligned bars was a grout-sleeve failure. Specific construction measures must be taken to generate a mechanical interlock force to improve the bond force between the grout and the sleeve in the grouted splice connector.
(3) The bond stress distribution around the inserted bar in the grouted sleeve lapping connector was similar to that around a single bar anchored in concrete.
(4) For the grouted sleeve lapping connectors, the slopes of the load-displacement curves of strain gauges installed at the middle bar inside the sleeve decreased as the load increased. As the lap length increased, the ultimate hoop compressive strain of the sleeve and the corresponding load increased.
(5) An approximate mechanical model was put forward to describe the mechanical properties of the grouted sleeve lapping connector and was proved to be highly reliable.
Studying the feasibility of a grouted sleeve lapping connector under a cycling load is a future research direction. It is also necessary to test the seismic behavior of precast shear walls with vertical reinforcements that are spliced by grouted sleeve lapping connectors to evaluate the operating performance of grouted sleeve lapping connectors when used in structures.