Connection of the prefabricated updeck of road tunnels by a short lap-spliced joint using ultra-high-performance fiber-reinforced concrete

Abstract

Prefabricated internal structures of road tunnels, consisting of precast elements and the connections between them, provide advantages in terms of quality control and manufacturing costs. However, the limited construction space in tunnels creates challenges for on-site assembly. To identify feasible connecting joints, flexural tests of precast straight beams connected by welding-spliced or lap-spliced reinforcements embedded in normal concrete or ultra-high-performance fiber-reinforced concrete (UHPFRC) are first performed and analyzed. With an improvement in the strength grade of the closure concrete for the lap-spliced joint, the failure of the beam transforms from a brittle splitting mode to a ductile flexural mode. The beam connected by UHPFRC100 with short lap-spliced reinforcements can achieve almost equivalent mechanical performance in terms of the bearing capacity, ductility, and stiffness as the beam connected by normal concrete with welding-spliced reinforcements. This favorable solution is then applied to the connection of neighboring updeck slabs resting on columns in a double-deck tunnel. The applicability is validated by flexural tests of T-shaped joints, which, fail in a ductile fashion dominated by the ultimate bearing capacity of the precast elements, similar to the corresponding straight beam. The utilization of UHPFRC significantly reduces the required lap-splice length of reinforcements owing to its strong bonding strength.

Graphical abstract

Keywords

UHPFRC / prefabricated updeck / road tunnel / lap-spliced rebars / flexural tests

1 Introduction

The development of urban underground spaces to solve traffic congestion has increased the demand for double-deck road tunnels, which provide advantages in terms of construction cost and space utilization. These tunnels generally consist of segmental linings and internal structures, such as upper and lower lanes. The segments are typically precast, which is favorable for both quality control and reducing manufacturing costs [1–3]. However, the majority of the internal structures of these tunnels, including the carriageway slabs and columns, are still cast in situ or semi-prefabricated. The limited space of underground structures creates challenges in the organization and implementation of the construction process [4], which requires a transition to a high level of prefabrication for constructing road tunnels.

The mechanical behavior and long-term durability of the entire prefabricated system are significantly influenced by the connections between adjacent precast elements constructed on-site [5–8]. These connections are achieved either by mechanical connectors consisting of steel elements and bolts, defined as dry joints [9,10], or by spliced reinforcements embedded in concrete that is cast in situ, defined as wet joints [11,12]. Splicing of reinforcements can be achieved by mechanical coupling, welding, or lapping of the rebar [8,13]. The latter is characterized by a simplified on-site assembly and a large lapping length [14]. An inadequate lapping length can lead to a reduction in the ultimate bearing capacity and ductility of the prefabricated structure [15]. However, this large lapping length is contradictory to the limited construction space in tunnels.

The development of ultra-high-performance fiber-reinforced concrete (UHPFRC) provides a new solution for spliced connections [15–17], significantly reducing the required lapping length [18,19]. UHPFRC, which contains embedded steel fibers, is characterized by high strength, favorable ductility, and long-term durability [20–24]; moreover, it can reliably bond to both precast concrete elements [25,26] and steel rebar [27,28]. This provides the motivation to use UHPFRC to connect precast elements inside road tunnels to achieve improved constructability and a high level of prefabrication.

In this study, the mechanical behavior of wet joints in the form of lap-spliced reinforcements embedded in UHPFRC is assessed to evaluate their application to the connection of a prefabricated updeck of a road tunnel, as shown in Fig.1(a). The precast tunnel segment has a thickness of 0.60 m and a radius of 14.0 m. The slabs for both the upper and lower decks, as well as the columns between them inside the road tunnel, are precast. The column footing is cast in the field, as shown in Fig.1(a), and is connected to the precast column using grouted splice sleeves. This type of joint is reported to exhibit an almost equivalent performance in bearing capacity and ductility to that of a conventional cast in situ system [29]. However, the connection between the precast column and two neighboring slabs for the updeck, as shown in Fig.1(b), is still designed to utilize welding-spliced reinforcements embedded in concrete cast on-site, which is difficult to construct owing to the large reinforcement ratio and limited space in underground structures. Therefore, the use of lap-spliced reinforcements embedded in UHPFRC to achieve a small lapping length would be favorable for the connection of the prefabricated updeck.

The research program of the present study is structured as follows. Flexural tests of precast straight beam elements connected by four different wet joints are presented in Section 2, followed by computational analyses and discussion of the experimental results in Section 3. A reliable wet joint for the connection of the updeck of the road tunnel is determined. This joint is then applied to the connection of two neighboring slabs resting on columns and validated based on flexural tests of the T-shaped joints in Section 4. Finally, the conclusions are presented in Section 5.

2 Flexural tests of straight beams connected by different wet joints

To identify a feasible solution for a cast on-site joint, four sets of flexural tests are conducted on precast beam elements connected by different wet joints. Three beams are connected by wet joints consisting of lap-spliced reinforcements embedded in either normal-strength concrete or UHPFRC. For comparison, an additional beam is connected using a wet joint consisting of welding-spliced reinforcements embedded in normal-strength concrete.

2.1 Specimens and materials

The tested straight beams, consisting of two precast concrete elements, are connected at the central region using wet joints, as shown in Fig.2. The lengths of the two precast elements are 3500 mm and the length of the connecting region is 500 mm. The dimensions of the beam cross-sections are 1000 and 500 mm.

The longitudinal reinforcement consists of 4Φ20 and 5Φ28 steel bars at the top and bottom layers, respectively, together with 4Φ18 bars at the two lateral layers, mainly for the ease of construction, as shown Fig.2. The transverse reinforcement consists of Φ12 closed stirrups with a spacing of 200 mm. The steel bars are of Grade HRB400 according to the Chinese standard [30]. The yield strength, f_{\mathrm{s},\mathrm{y}}, and ultimate strength, f_{\mathrm{s},\mathrm{u}}, of the steel bars are tested for three samples according to Ref. [31]. The average values are as follows:

(1)f_{\mathrm{s},\mathrm{y}}=456.0\mathrm{M}\mathrm{P}\mathrm{a} ,f_{\mathrm{s},\mathrm{u}}=616.2 \mathrm{M}\mathrm{P}\mathrm{a} .

The concrete used for the precast elements of the beams is normal concrete of grade C50 in accordance with Ref. [30]. Different types of concrete are used for the joints cast on-site, including two types of normal-strength concrete, denoted as C60 and C80, and two types of UHPFRC, denoted as UHPFRC80 and UHPFRC100; the initial weight mixtures are presented in Tab.1. Material tests are conducted on cubic concrete samples according to the requirements of Refs. [32,33] to quantify their average cube compressive strength (see Tab.1).

The construction details for the four wet joints between two precast beam elements are shown in Fig.3. In accordance with the design strategy of the project, the first beam is connected using welding-spliced rebars embedded in normal-strength concrete C60. The splice lengths of the rebar are equal to ten times their diameters: 200 and 280 mm for the rebar at the top and bottom layers, respectively, as shown in Fig.3. This beam is denoted as W1-C60. The other three beams are connected using equivalent connections consisting of lap-spliced rebar embedded in either normal-strength concrete or UHPFRC; these specimens are denoted as L2-C80, L3-UHPFRC80, and L4-UHPFRC100 according to the grade of the cast in situ concrete used in the joint. To facilitate comparison, the splice lengths of the rebar are equal to those of the welding-spliced joint, i.e., ten times the rebar diameters, as shown in Fig.3.

The beam elements are first cast in molds. To ensure the connection between the precast beam elements and the cast in situ joints, a jagged formwork is utilized at the surface close to the joint of the two precast elements. The notch is 90 mm in width and 30 mm in depth, as shown in Fig.4. After hardening and subsequent demolding, the surfaces of the substrate elements are sprinkled with water, followed by the pouring of concrete at the joints. Thereafter, the cast in situ joint is cured with a film coating for the first 24 h, followed by demolding and natural curing until an age of 28 d before testing.

2.2 Testing setup

The prefabricated beams are simply supported over two hinge supports at both extremities. External loading is applied using a hydraulic jack and distributed to two loading points with a steel girder. This creates the configuration of a four-point flexural test, as shown in Fig.5. Therefore, the investigated wet joints are subjected to pure bending. The applied loading is increased monotonically in accordance with the load-control process and is terminated upon failure of the beams.

The deflection at the midspan of the beam is measured using a linear variable displacement transformer (LVDT) throughout the loading process, as shown in Fig.5. The strains of the reinforcement at both the cast in situ joints and precast elements are monitored using strain gauges. The strain gauges are installed on the longitudinal rebar at the bottom layer of the beam, as shown in Fig.6. The applied loading, deflection, and strain are recorded and stored using a data-acquisition system. The initiation and propagation of cracking are verified and recorded after each loading step.

2.3 Experimental results

2.3.1 Strains

The strains of the reinforcements in the precast elements of beams W1-C60, L3-UHPFRC80, and L4-UHPFRC100 increase similarly with increasing bending moment, as shown in Fig.7(a). At the initial stage of the loading process, the strains of the reinforcements increase almost linearly with increasing loading owing to the elasticity of both the concrete and reinforcements. The tangents of these curves decrease when the bending moment reaches approximately 220 kN·m as a result of the initiation of tensile cracking of the concrete. With the propagation of tensile cracking, the reinforcements bear a larger share of the applied loading, exhibiting an almost linear increase in strain until approximately 2300 × 10⁻⁶, i.e., reaching the yield strain. For the strain of the reinforcements in the precast elements of beam L2-C80, a decrease in the tangent is also observed when the bending moment reaches 220 kN·m because of the start of tensile cracking of concrete, as shown in Fig.7(a). Thereafter, the strain increases until a bending moment of approximately 900 kN·m, reaching its ultimate value of 1000 × 10⁻⁶, which is far smaller than the yield strain. By reducing the applied loading, the elastic strain of the reinforcement can be fully recovered.

The strains of the reinforcements within the cast in situ joints are far smaller than the yield strains at the ultimate states for all four beams, as shown in Fig.7(b). This is attributed, on the one hand, to the increased stiffness of the cast in situ concrete compared to that of the precast concrete. On the other hand, the overlapping of the spliced reinforcements in the joints, leading to an increased effective area of the cross-section, also contributes to the smaller magnitudes of the strains of the reinforcements in the cast in situ joints compared to those of the reinforcements in the precast elements.

2.3.2 Deflections

The measured deflections of the four beams are shown in Fig.8(a). First, it can be seen that the beam connected by lap-spliced reinforcements embedded in normal-strength concrete (L2-C80) exhibits a much smaller bearing capacity of 902.3 kN·m. However, deflections of the beams connected by lap-spliced reinforcements embedded in UHPFRC (L3-UHPFRC80 and L4-UHPFRC100) develop similarly to that of the beam connected by welding-spliced reinforcement embedded in normal concrete (W1-C60) with increasing bending moment. The critical feature points of the three aforementioned beams are investigated in detail (see Tab.2). The values of the yielding moments, M_{\mathrm{y}}, characterized by the start of yielding of the tensile reinforcements based on Fig.7, as well as the corresponding deflections, {\delta }_{\mathrm{y}}, are almost the same: M_{\mathrm{y}}\approx 1620\mathrm{k}\mathrm{N}\cdot \mathrm{m} and {\delta }_{\mathrm{y}}\approx 21.0\times {10}^{-3}\mathrm{m}. An almost equivalent ultimate bearing capacity, M_{\mathrm{u}}, characterized by the crushing failure of compressed concrete, is obtained as M_{\mathrm{u}}\approx 1840\mathrm{k}\mathrm{N}\cdot \mathrm{m}. However, the corresponding deflections, {\delta }_{\mathrm{u}}, differ considerably, as summarized in Tab.2. To assess the ductility of the prefabricated beams, the coefficient of ductility, μ, is defined as the ratio between the deflection related to the ultimate bearing capacity, {\delta }_{\mathrm{u}}, and that related to the yielding moment, {\delta }_{\mathrm{y}}, i.e., \mu ={\delta }_{\mathrm{u}}/{\delta }_{\mathrm{y}}. Beam L4-UHPFRC100 exhibits not only an almost equivalent ultimate bending moment as that of beam W1-C60, but also a similar ductility (see the corresponding deflections and coefficients of ductility in Tab.2). However, the ultimate deflection and coefficient of ductility of beam L3-UHPFRC80 are noticeably smaller.

The bending stiffness of the beam, representing its resistance to deformation, can be quantified as the tangent of the experimentally measured load–deflection curve in Fig.8(a), as shown in Fig.8(b). The bending stiffnesses of all four prefabricated beams exhibit a similar slight decrease as the bending moment increases to approximately 220 kN·m, followed by a sudden decrease, which is attributed to the initiation and propagation of the tensile cracking of concrete. Thereafter, the stiffnesses of beams W1-C60, L3-UHPFRC80, and L4-UHPFRC100 slowly decrease to approximately 33 × 10⁻³ kN/m when the yielding bending moment of 1620 kN·m is reached, and then the stiffness decreases rapidly until the ultimate state is reached. Given that the bending stiffnesses of these three beams follow similar trends, their stiffness degradation is dominated by cracking in the precast beam elements. The lap-spliced connections can maintain their integrity prior to failure. For beam L2-C80, a rapid decrease in stiffness is observed when the applied bending moment reaches 750 kN·m.

2.3.3 Cracking and failure modes

The locations of all visible cracks in the beams are identified during the tests, and their widths are measured. The final cracking patterns are shown in Fig.9. Flexural cracks are dominant on the surfaces of beams W1-C60 and L4-UHPFRC100. These are mainly distributed on the precast beam elements, whereas minor cracks are observed within the cast in situ joints. Cracking is initiated at the interfaces between the joint and precast elements. This is followed by cracking of the tension zones within the precast beam elements on both sides, i.e., on the lower part of the elements. The cast in situ joints start to crack at a loading level of approximately 1100 kN·m. The ultimate state is reached at a bending moment of approximately 1840 kN·m. The cracks are concentrated at the interfaces and tension zones of the precast elements in the final configuration, as shown in Fig.9(a) and Fig.9(d).

The final cracking patterns of beams L2-C80 and L3-UHPFRC80 are quite similar, as shown in Fig.9(b) and Fig.9(c), even though their ultimate bending moments differ significantly (see Tab.2). Concentrated cracks are observed at the bottom surface of the cast in situ joints. Apart from the flexural vertical cracks, horizontal and inclined cracks also prevail, accompanied by moderate spalling of the concrete cover.

The differences in the cracking patterns of these four beams are attributed to their different failure modes. Beams W1-C60 and L4-UHPFRC100 exhibit a ductile failure mode characterized by the yielding of tensile rebar (Fig.7), large deflections (Fig.8(a)), and flexural cracks (Fig.9(a) and Fig.9(d)). However, a brittle splitting failure mode is observed for beam L2-C80, characterized by elastic behavior of the tensile rebar (Fig.7), an abrupt drop in the loading–deflection curve (Fig.8(a)), and localized cracks around the splice rebar (Fig.9(b)). This results from bond failure of the splice rebar within the wet joints. Beam L3-UHPFRC80 exhibits a combined failure mode characterized by the yielding of tensile rebar (Fig.7), which is rapidly followed by splitting failure of the concrete (Fig.9(c)).

3 Computational analyses and discussion

To assess the mechanical performance of the wet joints, the initiation of cracking and bearing capacity of the four prefabricated beams are analyzed and discussed in this section. The computational analyses are based on the assumption of the Euler–Bernoulli hypothesis, i.e., the cross-sections of the beam remain virtually planar in the deformed configuration. This leads to a linear distribution of the strains along the height of the cross-section of the beam. The cross-sectional dimensions, including the numerical values of the areas and positions of the steel rebars, are shown in Fig.10.

The material behaviors of concrete and steel are shown in Fig.11 [34]. The constitutive equations for both normal concrete and UHPFRC under compression are as follows:

(2)

where n and f_{\mathrm{c}} represent the exponent and compressive strength of the concrete, respectively; {\epsilon }_{\mathrm{c}0} and {\epsilon }_{\mathrm{c}\mathrm{u}} denote the strains at the elastic limit and ultimate state of the concrete in compression, respectively. The expected values of the strength and Young’s modulus of concrete are estimated based on the experimentally measured cube compressive strength in Tab.1. The characteristics of the strength and deformation of concrete in Eurocode 2 [34] are used for the estimation, and the results are presented in Tab.3. The yield and ultimate strengths of the steel bars are determined experimentally determined according to Eq. (1). The Young’s modulus is assumed to be E_{\mathrm{s}}= 200 GPa [34], resulting in a yield strain of {\epsilon }_{\mathrm{y}}= 2.28\times {10}^{-3}.

3.1 Cracking initiation

The magnitude of the cracking load is important for assessing the durability and serviceability of prefabricated structures. Computational analyses of the cracking load of a beam are related to the tensile strength of the concrete. Because of the higher strength grade of the cast on-site concrete for the joint, tensile cracking initiates either at the precast elements or at the interfaces between the joint and precast elements.

The tensile strength of normal concrete C50 for the precast elements is estimated as f_{\mathrm{c}50,\mathrm{t}}=3.46 MPa, as presented in Tab.3. The tensile strength of the interfacial zones between the substrate concrete and the newly cast concrete is reported to depend on the strength of both the base and overlay concrete, the roughness and moisture condition of the substrate surface, the curing process, the use of bonding agents, etc. [25,35,36]. Empirical values for the cohesion between normal concrete and ultra-high-performance concrete have been reported [37] by categorizing the roughness of the interface as rough, medium, or smooth. Considering the jagged interface between the precast elements and the joint in this study, the tensile strength of the interfacial zone is taken as the reported cohesion of the rough interface for simplification, i.e., f_{\mathrm{i}\mathrm{n}\mathrm{t},\mathrm{t}}=2.66 MPa [37].

In the case of pure bending, cracking of concrete is initiated when the maximum tensile stress, i.e., the stress at the edge of the tension zone, reaches its tensile strength. Therefore, the cracking bending moment, M_{\mathrm{c}\mathrm{r}}, can be calculated as follows:

(3)M_{\mathrm{c}\mathrm{r}}=W_{\mathrm{z}}\cdot f_{\mathrm{t}},

where f_{\mathrm{t}} and W_{\mathrm{z}} represent the tensile strength of concrete and elastic sectional modulus, respectively. The latter is quantified as W_{\mathrm{z}}=\left(I_{\mathrm{c}}+{\displaystyle \frac{E_{\mathrm{s}}}{E_{\mathrm{c}}}}{\sum }_{i=1}^6{I}_{\mathrm{s},i}\right)=9.1243\times {10}^7{\mathrm{m}\mathrm{m}}^3. The cracking bending moments for the precast elements and interface zones are quantified by substituting the corresponding tensile strengths of precast concrete and interfacial zones into Eq. (3); the results are given in Tab.4.

Both the experimental measurements and computational analyses suggest that cracking always starts at the interfaces between the precast elements and cast in situ joints (see Tab.4). The analytical results agree well with the experimental measurements, except for the cracking bending moment at the interfacial zones for beam L4-UHPFRC100, which is much larger than those for the other three beams. This indicates a much stronger bonding of the newly cast UHPFRC100 with the substrate concrete compared to the other cases. This is attributed to the high content of reinforcing fibers in UHPFRC100, thus endowing it with a high strength (see Tab.1 and Tab.3). To improve the computation of the cracking bending moment, the influence of the substrate concrete on the tensile strength of the interfacial zone should be considered. Overall, the wet joint using UHPFRC100 with embedded lap-spliced reinforcements exhibits superior integrity and durability because of its strong bonding with the base concrete.

3.2 Bearing capacity

The experimental measurements indicate that the ultimate states of the four prefabricated beams are related to either splitting failure of the cast in situ joints or flexural failure of the precast beam elements. The corresponding ultimate bending moments are computationally analyzed in this section.

Splitting failure of the joints is characterized by failure of the bonding between the bottom-most layer of rebar and the surrounding concrete. For this scenario, the distribution of the normal stresses is illustrated in Fig.12(a). The axial force of the bottom-most layer of rebar is balanced with the ultimate anchorage stress, {\tau }_{\mathrm{u}}, of the surrounding concrete as follows:

(4){\tau }_{\mathrm{u}}\text{π}dln=A_{\mathrm{s}6} {\sigma }_{\mathrm{s}6} ,

where l=280\mathrm{m} \mathrm{m}, d=28\mathrm{m}\mathrm{m}, and n=5 indicate the splice length, diameter, and number of the rebar, respectively. The ultimate anchorage stress, {\tau }_{\mathrm{u}}, is estimated empirically as follows [38]:

(5){\tau }_{\mathrm{u}}= \left(0.7+2.5\frac{d}{l}\right)\left(0.5+0.6\frac{c}{d}+55{\rho }_{\mathrm{s}\mathrm{v}}\right)f_{\mathrm{t}},

where {\rho }_{\mathrm{s}\mathrm{v}}=0.45\mathrm{\%} and c=77.0\mathrm{m} \mathrm{m} denote the stirrup ratio and thickness of the concrete cover, respectively. The tensile strength of the concrete used for the wet joints, f_{\mathrm{t}}, is listed in Tab.3. By considering the linearity of the strain field and constitutive laws of the materials, the normal force, N, and bending moment, M, can be expressed as follows:

(6)\{\begin{array}{l}N={\int }_A^{}\sigma \left(y\right)\mathrm{d}A={\displaystyle \sum _{i=1}^6{\displaystyle \frac{{\sigma }_{\mathrm{s}6}{A}_{\mathrm{s}i}(h_{\mathrm{s}i}-h_{\mathrm{c}})}{h_{\mathrm{s}6}-h_{\mathrm{c}}}}-{\int }_0^{h_{\mathrm{c}}}{\sigma }_{\mathrm{c}}\left(y\right)b\mathrm{d}y,}\\ M={\int }_A^{}\sigma \left(y\right)y\mathrm{d}A={\displaystyle \sum _{i=1}^6{\displaystyle \frac{{\sigma }_{\mathrm{s}6}{A}_{\mathrm{s}i}{(h_{\mathrm{s}i}-h_{\mathrm{c}})}^2}{h_{\mathrm{s}6}-h_{\mathrm{c}}}}-{\int }_0^{h_{\mathrm{c}}}{\sigma }_{\mathrm{c}}\left(y\right)by\mathrm{d}y,}\end{array}

where the numerical values for the areas and positions of the reinforcement layers, A_{\mathrm{s}i} and h_{\mathrm{s}i}, respectively, are shown in Fig.10. The height of the compressive zone, h_{\mathrm{c}}, is solved by substituting the quantified {\sigma }_{\mathrm{s}6} from Eq. (4) into Eq. (6) and setting N=0. Substituting the result into the second expression provides the bending moment related to the splitting failure, and the results are summarized in Tab.5.

Flexural failure of the precast elements is characterized by crushing of the uppermost concrete, i.e., reaching its ultimate strain. The distribution of normal stresses is shown in Fig.12(b). Trial calculations indicate that when the ultimate strain is reached for concrete at the top surface, the five lower layers of reinforcements yield, i.e., {\sigma }_{\mathrm{s}i}=f_{\mathrm{y}} (i = 2,3,…,6). Therefore, the normal force, N, and bending moment, M, are quantified as follows:

(7)\{\begin{array}{l}\begin{array}{rl}N=& {\int }_A^{}\sigma \left(y\right)\mathrm{d}A=\sum _{i=2}^6{f}_{\mathrm{y}}{A}_{\mathrm{s}i}\\ & +\frac{{\epsilon }_{\mathrm{c}\mathrm{u}}(h_{\mathrm{s}1}-h_{\mathrm{c}})A_{\mathrm{s}1}{E}_{\mathrm{s}}}{h_{\mathrm{c}}}-\alpha \beta f_{\mathrm{c}}b{h}_{\mathrm{c}},\end{array}\\ \begin{array}{rl}M=& {\int }_A^{}\sigma \left(y\right)y\mathrm{d}A=\sum _{i=2}^6{f}_{\mathrm{y}}{A}_{\mathrm{s}i}(h_{\mathrm{s}i}-h_{\mathrm{c}})\\ & +\frac{{\epsilon }_{\mathrm{c}\mathrm{u}}{(h_{\mathrm{s}1}-h_{\mathrm{c}})}^2{A}_{\mathrm{s}1}{E}_{\mathrm{s}}}{h_{\mathrm{c}}}-\alpha \beta \left(1-\frac{\beta }{2}\right)f_{\mathrm{c}}b{h}_{\mathrm{c}}^2,\end{array}\end{array}

where the characteristic parameters for the equivalent rectangular stress diagram are taken as \alpha =1.0 and \beta =0.8 for C50 concrete [39]. Similarly, the bending moment related to flexural failure is computed by solving N=0 for the height of the compressive zone, h_{\mathrm{c}}, and substituting the result into the expression for M; the results are summarized in Tab.5.

Both the experimental measurements and computational analyses suggest that beam L4-UHPFRC100, connected by UHPFRC100 with embedded lap-spliced reinforcements, and W1-C60, connected by normal concrete C60 with embedded welding-spliced reinforcements, exhibit similar ultimate bearing capacities. These two beams fail in a ductile fashion, with the ultimate state controlled by the compressive crushing of concrete at the top of the precast elements. However, the other two beams, which are connected by lap-spliced reinforcements embedded in either normal concrete C80 or UHPFRC80, fail abruptly. Their ultimate states are dominated by the splitting of the lap-spliced reinforcements from the surrounding concrete.

3.3 Lap-splice length

The above analyses indicate that the beam connected by lap-spliced rebar embedded in UHPFRC100 exhibits an almost equivalent mechanical performance in terms of bearing capacity, ductility, and stiffness to the beam connected by welding-spliced rebar embedded in normal concrete C60. The utilization of UHPFRC100 allows for a short splice length of ten times the diameter of the rebar. This length is compared with the length recommended by the design codes GB50010 of China [30] and EN1992-1-1 of Europe [34] under the same conditions, and the results are presented in Tab.6.

The favorable flexural behavior of the connection using UHPFRC100 with short lap-spliced reinforcements is attributed to the high content of reinforcing fibers and high compressive strength. This promotes bonding not only with the steel rebar but also with the base concrete of the precast elements. The failure mode is dominated by the ultimate bearing capacity of the precast elements, which depends on their sectional dimensions and reinforcement layout. Thus, this is the most favorable connection to connect precast elements for the updeck of the road tunnel in the following sections.

4 Validation of the applicability for connecting a carriageway updeck

The above flexural tests of pure bending elements provided a favorable solution consisting of short lap-spliced rebar embedded in UHPFRC100 for the connection of precast elements. This connection is considered as an alternative strategy for the joint between the precast column and two neighboring slabs in the carriageway updeck of the road tunnel in Fig.1, which is presently designed to be achieved by welding-spliced rebar embedded in normal-strength concrete.

Rather than being subjected to pure bending, this joint is loaded by both bending moments and shear forces as a result of the self-weight and traffic loading. Therefore, it is simulated as a T-shaped joint connecting two precast beam elements and a precast column capital. The applicability of this alternative solution is experimentally validated using three sets of flexural tests. For comparison, an additional set of flexural tests is conducted on the T-shaped joint in accordance with the design scheme.

4.1 Specimens and materials

The geometric dimensions of the tested T-shaped joints are shown in Fig.13. The lengths of the two precast beam elements are 1750 mm, and the length of the precast column capital is 700 mm; these elements are connected in the central region with a cast in situ joint. The cross section of the beam elements is 550 mm at the bottom and 500 mm at the top, with a height of 1000 mm. The cross-sectional dimensions of the column are 700 mm in length and 500 mm in width.

The reinforcement layouts of the precast beam elements and precast column are consistent with those of the engineering project, as shown in Fig.13. The steel bars are of Grade HRB400, and their strength is listed as Eq. (1). The concrete cast in situ for the joint is of grade C60 or UHPFRC100, and the corresponding compositions and mechanical properties are listed in Tab.1.

The construction details for the T-shaped joints are illustrated in Fig.14. For comparison, the first T-shaped joint consists of welding-spliced rebar embedded in normal-strength concrete C60 in accordance with the design proposal of the project; this joint is denoted as T-W. The other three T-shaped joints consist of lap-spliced rebar embedded in UHPFRC100, and these joints are denoted as T-L(1), T-L(2), and T-L(3). The splice length of the reinforcements is equal to ten times the diameter of the rebar, i.e., 280 mm.

The conditioning and surface treatment of the precast elements follow the same strategy as that used for the straight beams. Utilization of the corrugated formwork results in a jagged surface of the substrate concrete to ensure the connection between the precast elements and the closing concrete of the joint. The cast in situ connection is cured until the age of 28 d before flexural testing.

4.2 Testing setup

The T-shaped joints are simply supported over two hinge supports at both extremities. Mechanical loading is applied directly to the precast column capital using a hydraulic jack, as shown in Fig.15. The applied loading is increased monotonically in accordance with the load control process and is terminated upon failure of the beams.

The deflection at midspan of the T-shaped joint is measured using an LVDT throughout the loading process, as shown in Fig.15. The applied loading and midspan displacements are recorded and stored using a data-acquisition system. Similarly, the widths and locations of visible cracks in the beams are determined during the test.

4.3 Experimental results and discussion

The measured deflections of the four T-shaped joints are shown in Fig.16. The ultimate bearing capacities of beams T-L(1), T-L(2), and T-L(3) are similar (approximately 3100 kN), which is slightly larger than that of beam T-W (approximately 2840 kN). Before reaching the ultimate state, all of the beams exhibit favorable ductility; see the large deflection and plateau in the graphs in Fig.16. Therefore, the wet joint in the form of lap-spliced reinforcement embedded in UHPFRC100 meets the both the bearing capacity and ductility requirements for the connection of the precast column and neighboring slabs in the updeck of road tunnels.

The final cracking patterns of the four T-shaped joints are shown in Fig.17, following similar patterns. As a result of the combined loading of the bending moment and shear force, flexural cracks dominate on the surfaces of the precast beam elements, whereas minor cracks are observed within the joints. For all four T-shaped joints, cracking always starts from the interfaces between the connections and the precast beam elements when the applied loading reaches approximately 350 kN. The cracking width increases progressively with increasing loading, reading approximately 0.20 mm at a loading level of 700 kN and 0.50 mm at a loading level of 1600 kN. For the precast beam elements, flexural cracking initiates from the tensile zone at a loading level of 400 kN. Because of the additional contribution of the shear force, this cracking develops in an inclined mode from the tensile zone of the precast beam elements toward the precast column capital. Finally, the ultimate state is reached when the primary flexural cracks almost reach the top surface of the beam, as shown in Fig.17. Both the welding-spliced and lap-spliced joints remain relatively integrated during the entire loading process.

The experimental results validate the applicability of the connection consisting of UHPFRC100 with short lap-spliced reinforcements for the assembly of the precast carriageway updeck. This connection can achieve an almost equivalent bearing capacity and ductility as the connection consisting of normal-strength concrete with welding-spliced reinforcements. The use of UHPFRC100, with a compressive strength reaching 95.9 MPa, can avoid bonding failure of the lap-spliced reinforcements with a lapping length of ten times the rebar diameter. The failure mode is dominated by the ultimate bearing capacity of the precast elements. Following Eq. (7), the ultimate bending moment can be analytically quantified as 1715.3 kN·m. Assuming that the cross-section of the precast element at the interface reaches its ultimate state, the corresponding value of the applied force is 2638.9 kN, which is slightly smaller than the experimental measurements (see Tab.7 for the comparison). However, the non-vanishing shear force results in apparent cracks over the interface, as shown in Fig.17. This can threaten the integrity and long-term durability of the prefabricated structure. Apart from roughening and watering the surface of the substrate concrete, which are utilized in this study, the application of bond agents [40] and increasing the strength of the substrate concrete [41], if applicable, can also improve the strength of the interfaces. Overall, this alternative solution simplifies the in situ assembly and saves time in connecting precast carriageway slabs with columns inside road tunnels.

5 Conclusions

This paper presents flexural testing results for precast beam elements connected with lap-spliced reinforcements embedded in different grades of concrete and compares these results with those for an equivalent connection containing welding-spliced reinforcements. This provides access to a simplified and efficient solution for on-site assembly of the prefabricated updeck of road tunnels within a limited working space. The applicability of the alternative solution is experimentally validated by flexural tests of T-shaped joints connecting two neighboring updeck slabs resting on a column. The main conclusions are as follows.

1) The wet joint comprising lap-spliced reinforcements embedded in UHPFRC100 exhibits an almost equivalent mechanical performance in terms of bearing capacity, stiffness, and ductility as the wet joint comprising welding-spliced reinforcements embedded in normal-strength concrete C60. Because of the strong bonding of UHPFRC with substrate concrete, the former can even exhibit an advantage in delaying cracking initiation over the interfaces between the cast in situ joint and precast elements.

2) With the improvement in concrete grade from C80 to UHPFRC80 and UHPFRC100 for the lap-spliced joint, the failure of the prefabricated beam transforms from a brittle splitting failure mode to a combined splitting and flexural failure mode, and finally to a ductile flexural failure mode. The latter is characterized by the yielding of tensile rebar and large deflection of the beam, dominated by the flexural performance of the precast beam elements.

3) The use of UHPFRC100 with a compressive strength of 95.9 MPa significantly reduces the required lap-splice length of the reinforcements to a magnitude of ten times their diameter. UHPFRC is promising for improving the bond between splice reinforcements and the surrounding concrete owing to its advantageous mechanical properties.

4) The joint with lap-spliced reinforcements embedded in UHPFRC100 is applicable for connecting a precast column with two neighboring slabs resting on columns for the updeck of road tunnels, as validated by flexural tests of T-shaped joints. It can achieve a slightly improved ultimate bearing capacity compared with the T-shaped joint connected by welding-spliced reinforcements in normal concrete. This is favorable for simplifying on-site assembly and controlling construction quality.

5) Cracking generally initiates from the interfaces between the joint and precast elements, which can threaten the long-term durability of the prefabricated updeck of road tunnels. Therefore, surface treatment of the precast elements, such as roughening and wetting of the substrate concrete, as well as the application of bond agents, is recommended before pouring the concrete of the joint.

References

[1] GuglielmettiVGrassoPMahtabA XuS. Mechanized Tunnelling in Urban Areas: Design Methodology and Construction Control. London: CRC Press, 2008

[2] WangSFu JZhangCYangJ. Shield Tunnel Engineering: From Theory to Practice. Amsterdam: Elsevier, 2021

[3] Jiang X, Zhang X, Wang S, Bai Y, Huang B. Case study of the largest concrete earth pressure balance pipe-jacking project in the world. Transportation Research Record: Journal of the Transportation Research Board, 2022, 2676(7): 92–105

[4] Jiang X, Zhang Y, Zhang Z, Bai Y. Study on risks and countermeasures of shallow biogas during construction of metro tunnels by shield boring machine. Transportation Research Record: Journal of the Transportation Research Board, 2021, 2675(7): 105–116

[5] Breccolotti M, Gentile S, Tommasini M, Materazzi A L, Bonfigli M F, Pasqualini B, Colone V, Gianesini M. Beam−column joints in continuous RC frames: Comparison between cast-in-situ and precast solutions. Engineering Structures, 2016, 127: 129–144

[6] Choi H K, Choi Y C, Choi C S. Development and testing of precast concrete beam-to-column connections. Engineering Structures, 2013, 56: 1820–1835

[7] Parastesh H, Hajirasouliha I, Ramezani R. A new ductile moment-resisting connection for precast concrete frames in seismic regions: An experimental investigation. Engineering Structures, 2014, 70: 144–157

[8] ElliottK S. Precast Concrete Structures. London: CRC Press, 2019

[9] Liu T, Wang Z, Guo J, Wang J. Shear strength of dry joints in precast UHPC segmental bridges: Experimental and theoretical research. Journal of Bridge Engineering, 2019, 24(1): 04018100

[10] Ahmed G H, Aziz O Q. Shear behavior of dry and epoxied joints in precast concrete segmental box girder bridges under direct shear loading. Engineering Structures, 2019, 182: 89–100

[11] Lu C, Dong B, Pan J, Shan Q, Hanif A, Yin W. An investigation on the behavior of a new connection for precast structures under reverse cyclic loading. Engineering Structures, 2018, 169: 131–140

[12] Eom T S, Park H G, Hwang H J, Kang S M. Plastic hinge relocation methods for emulative PC beam−column connections. Journal of Structural Engineering, 2016, 142(2): 04015111

[13] Ghayeb H H, Razak H A, Sulong N R. Performance of dowel beam-to-column connections for precast concrete systems under seismic loads: A review. Construction & Building Materials, 2020, 237: 117582

[14] RamirezJ ARussell B W. Transfer, Development, and Splice Length for Strand/Reinforcement in High-Strength Concrete. Washington: Transportation Research Board, 2008, 603

[15] Maya L, Zanuy C, Albajar L, Lopez C, Portabella J. Experimental assessment of connections for precast concrete frames using ultra high performance fibre reinforced concrete. Construction & Building Materials, 2013, 48: 173–186

[16] Haber Z B, Graybeal B A. Lap-spliced rebar connections with UHPC closures. Journal of Bridge Engineering, 2018, 23(6): 04018028

[17] Lagier F, Massicotte B, Charron J P. Experimental investigation of bond stress distribution and bond strength in unconfined UHPFRC lap splices under direct tension. Cement and Concrete Composites, 2016, 74: 26–38

[18] Lagier F, Massicotte B, Charron J P. Bond strength of tension lap splice specimens in UHPFRC. Construction & Building Materials, 2015, 93: 84–94

[19] YuanJGraybeal B A. Bond Behavior of Reinforcing Steel in Ultra-High Performance Concrete. FHWA-HRT-14-090. 2014

[20] Yu J, Zhang B, Chen W, He J. Experimental and multi-scale numerical investigation of ultra-high performance fiber reinforced concrete (UHPFRC) with different coarse aggregate content and fiber volume fraction. Construction & Building Materials, 2020, 260: 120444

[21] Li V C. Postcrack scaling relations for fiber reinforced cementitious composites. Journal of Materials in Civil Engineering, 1992, 4(1): 41–57

[22] Zheng X, Wang Y, Zhang S, Xu F, Zhu X, Jiang X, Zhou L, Shen Y, Chen Q, Yan Z, Zhao W, Zhu H, Zhang Y. Research progress of the thermophysical and mechanical properties of concrete subjected to freeze−thaw cycles. Construction & Building Materials, 2022, 330: 127254

[23] Li L, Gong W, Li J. Service life of prestressed high-strength concrete pile in marine environment considering effects of concrete stratification and temperature. Construction & Building Materials, 2020, 253: 119233

[24] Sun R, Han L, Zhang H, Ge Z, Guan Y, Ling Y, Schlangen E, Savija B. Fatigue life and cracking characterization of engineered cementitious composites (ECC) under flexural cyclic load. Construction & Building Materials, 2022, 335: 127465

[25] Farzad M, Shafieifar M, Azizinamini A. Experimental and numerical study on bond strength between conventional concrete and ultra high-performance concrete (UHPC). Engineering Structures, 2019, 186: 297–305

[26] GraybealB ADe la Varga IHaberZ B. Bond of Field-cast Grouts to Precast Concrete Elements. FHWA-HRT-16-081. 2017

[27] Marchand P, Baby F, Khadour A, Battesti T, Rivillon P, Quiertant M, Nguyen H H, Genereux G, Deveaud J P, Simon A, Francois T. Bond behaviour of reinforcing bars in UHPFRC. Materials and Structures, 2016, 49(5): 1979–1995

[28] Saleem M A, Mirmiran A, Xia J, Mackie K. Development length of high-strength steel rebar in ultrahigh performance concrete. Journal of Materials in Civil Engineering, 2013, 25(8): 991–998

[29] Haber Z B, Saiidi M S, Sanders D H. Seismic performance of precast columns with mechanically spliced column-footing connections. ACI Structural Journal, 2014, 111(3): 639–650

[30] GB50010-2010. Code for Design of Concrete Structures. Beijing: Ministry of Housing and Urban-Rural Development of the People’s Republic of China, 2010 (in Chinese)

[31] GB-T228-2002. Metallic Materials—Tensile Testing at Ambient Temperature. Beijing: State General Administration of the People’s Republic of China for Quality Supervision and Inspection and Quarantine, 2002 (in Chinese)

[32] GB-T50081-2002. Standard of Test Method of Mechanical Properties on Ordinary Concrete. Beijing: Ministry of Construction of the People’s Republic of China, 2002 (in Chinese)

[33] CECS13:89. Testing Method for Steel Fiber Reinforced Concrete. Beijing: China Association for Engineering Construction Standardization, 1989

[34] EN1992-1-1. Eurocode 2: Design of Concrete Structures––Part 1–1: General Rules and Rules for Buildings. London: British Standards Institution, 2005

[35] Ueda T. Material mechanical properties necessary for the structural intervention of concrete structures. Engineering (Beijing), 2019, 5(6): 1131–1138

[36] Sadowski L. Multi-scale evaluation of the interphase zone between the overlay and concrete substrate: Methods and descriptors. Applied Sciences (Basel, Switzerland), 2017, 7(9): 893

[37] OuyangNDeng S. Study on the interfacial properties of UHPC-NC composite component. Journal of Chongqing University, 2019, 44(3): 63−74 (in Chinese)

[38] XuYWangH ShenW. Experimental research on force transition through lap-spliced reinforcements. Building Structure, 1993, 4: 20−24 (in Chinese)

[39] GuoZ. Principles of Reinforced Concrete. Oxford: Butterworth-Heinemann, 2014

[40] Diab A, Eldin M. Slant shear bond strength between self compacting concrete and old concrete. Construction & Building Materials, 2017, 130: 73–82

[41] Wang B, Xu S, Liu F. Evaluation of tensile bonding strength between UHTCC repair materials and concrete substrate. Construction & Building Materials, 2016, 112: 595–606

RIGHTS & PERMISSIONS

Higher Education Press