Since 2015, China has witnessed explosive growth in the mileage of constructed utility tunnels. Tunnels are the lifeline of cities (Valdenebro and Gimena, 2018; CANTO-PERELLO and CURIEL-ESPARZA, 2001; O’ROURKE et al., 2001; DARLI et al., 2021), and the stress on an underground structure is more complex than that on an above-ground structure (Bai et al., 2019; Luo et al., 2020; Bai et al., 2021a), so its safety is essential. From the perspective of the construction mode, the integral cast-in-place mode is applied to most of the tunnels. The integral cast-in-place mode provides good waterproof performance but has a longer construction duration, a greater environmental impact, and requires more field workers. Thus, it is an inevitable trend to adopt the fabricated construction method.

The composite slab-type concrete structure has been extensively studied and applied in constructing above-ground buildings (Gu et al., 2020; Bai et al., 2021b; Jiang et al., 2021; Zhou et al., 2021; Bai et al., 2023). As a structural member, it consists of the precast concrete arranged on both sides as the formwork connected through truss bars. In the construction process, concrete is poured into the cavity to form the integral structure.

Some scholars have studied composite slab-type underground utility tunnels. Yanmin Yang (YANG et al., 2020) conducted static loading tests on a fabricated composite tunnel and analyzed the phenomenon of its failure, bearing capacity and load-strain curve, and load-displacement curve of each measuring point on the axillary angle under the influence of simulated overburden load. They indicated that the axillary angle exhibits no obvious cracks under the normal service condition, which suffers significant deformation and presents good mechanical performance only under ultimate load. Weichen Xue (XUE et al., 2018) et al. Investigated the seismic performance outside the composite shear wall plane, and their results showed that the ultimate bearing capacity outside the composite shear wall plane is 10.7% lower than that of the cast-in-place shear wall. However, there are still some limitations in applying this technology to underground engineering, which are primarily illustrated by the following points:

(1) It is not easy to detect the molding quality of post-cast concrete. The prefabricated slab for existing underground fabricated composite concrete utility tunnels is generally made of ordinary concrete with a minimum thickness of 50 mm. In addition, there are steel bars within the post-cast concrete, making it difficult to detect the molding quality directly.

The application of high-performance cement-based materials in construction engineering has become more mature (Roy et al., 1972; Wang et al., 2012; Zheng et al., 2015; Chong et al., 2016; De et al., 2016; Zheng et al., 2016). The prefabricated slab made of ordinary concrete is replaced with formwork made of high-performance material to reduce the thickness of the prefabricated component, thereby assisting in settling the problems mentioned above. Based on this concept, a new technical system of underground fabricated concrete utility tunnels is proposed—underground utility tunnels of prefabricated composite shell system (after this referred to as composite shell utility tunnels). High-strength materials can provide the following advantages:

1) The out-of-plane mechanical performance is better. The loaded and connecting steel bars are all located in the cavity between the formwork shells, and the steel bars at the joints are connected by a lapping connection. The connection method is consistent with the cast-in-place structure, and the distance between the two rows of connecting steel bars is significant, resulting in improved out-of-plane mechanical performance.

This study addresses the design structure of the system first, determines the component structure, joint construction, the ratio of formwork shell materials, and others of the composite shell utility tunnels, and tests the fundamental mechanical properties of formwork shell materials. Then, an experimental investigation is conducted on the composite shell roof specimens to analyze the primary performance indices and failure forms of cast-in-place and composite shell specimens. On the other hand, the test phenomena and results are utilized to derive the calculation formula for the bearing capacity of the composite shell roof specimens, providing a reference for the system’s application.

The test phenomena and results showed that the first crack on the specimen occurred below the loading point, and the formwork shell and concrete cracked simultaneously during the normal section cracking mode, while the joint surface of the formwork shell-concrete did not crack. Based on this phenomenon, the formwork shell and concrete can be investigated as a unit when the force analysis is performed on the specimen’s cracking point. Since the concrete stress was small in the compression zone, the compressive stress on the concrete was regarded as the linear distribution mode.

For ordinary concrete, it is believed that cracking occurred when the tensile strain on concrete reached two times the peak axial tensile strain ε t 0 ^[21]. Based on the literature ^[22], the concrete used for specimen fabrication in this study has an elastic modulus E ₁ of 31,500 MPa, the tensile strength f t of 2.2 MPa, and the peak tensile strain ε t 0 of 100 × 10⁻⁶. According to the fundemental properties of the formwork shell material measured by tests, the elastic modulus E ₂ was 34,500 MPa, the tensile strength f t ′ was 6.7 MPa, and the peak tensile strain ε t 0 ′ was 200 × 10⁻⁶. Based on force equilibrium and bending moment equilibrium on the section, the formula for calculating the bearing capacity of the cracking point is as follows: 0.5 σ c 0 b x c r + A s ′ σ c s ′ = N + 0.75 b f t h ′ − x c r + A s σ c s + f t ′ h s b (1) V c r L 4 + N h 2 − h s + A s σ c s a s + 7 b f t h ′ − x c r 2 24 = A s ′ σ c s ′ h ′ − a s ′ + b σ c 0 x c r h ′ − x c r 3 2 (2)

In calculating the yield point’s bearing capacity, the tensile strength of the concrete and formwork shell was disregarded, and the compressive stress of concrete in the compression zone was used as the mode of linear distribution. The state point of the yield of lower steel bars was determined to be the yield point of specimens, and the yield strain ε y for steel bars was taken as 2000 × 10⁻⁶. The following formula can calculate the bearing capacity of the yield point: 0.5 σ y 0 b x y + A s ′ σ y s ′ = N + A s f y (3) V y L 4 + N h 2 − h s − a s = A s ′ σ y s ′ h ′ − a s − a s ′ + 0.5 b σ y 0 x y h ′ − a s − x y 3 (4)

In calculating the peak point’s bearing capacity, the tensile strength of the concrete and formwork shell was ignored, and the compressive stress on concrete in the compression zone was considered the uniformly distributed load mode. Based on the previous analysis, at the peak point, the upper steel bars changed from the compressive state to the tensile state, and the stress value was lower; therefore, the acting force on the upper steel bars was not accounted for in the equilibrium formula. The lower steel bars experienced greater deformation and entered the reinforcing stage, so the stress at the peak point was assumed to be 1.1 times the yield stress f y . The formula for calculating the bearing capacity of the peak point is as follows: f c 1 b x = 1.1 A s f y + N (5) V L 4 + N h 2 − h s − a s = f c 1 b x h ′ − x 2 − a s (6) where f t is tensile strength of concrete; f t ′ is the tensile strength of the formwork shell; f c 1 is the compressive strength of concrete; h ′ is the result of subtracting the thickness h s of formwork shell on the tensile side from the section thickness h; b is section width; σ c 0 and σ y are maximum compressive stress of the concrete in the compression zone of the cracking point and yield point; x c r , x y and x are the height of the compression zone of the cracking point, yield point, and peak point; A s and A s ′ are sectional areas of longitudinal steel bars in tension and compression zones; σ c s and σ c s ′ are stress of longitudinal steel bars in the tension and compression zones of the cracking point; σ y s ′ is stress of longitudinal steel bars in the compression zone of the yield point; V c r , V y , and V are concentrated force at the midspan of the cracking point, yield point, and peak point; L is span; N is axial compression; a s   a n d   a s ′ are the distance from the acting point of the resultant force on longitudinal steel bars in the tension and compression zones to the edge of the tension and compression zones of concrete; ε t 0 is the peak axial tensile strain of concrete.

Figure 17 depicts the detailed calculation of various load points, with the calculation results illustrated in Table 7. The calculated bearing capacity of the cracking, yield, and peak points was 134, 430, and 490 kN, respectively, with errors of 6.5%, 11.9%, and 12.5%, respectively, compared with the test values. The calculated values were all lower than the test values. The proposed formula can be employed to accurately determine the bearing capacity of composite shell roof specimens in different states. It has certain safety redundancy, providing a reference for performance design in different operating conditions.

Due to the use of high-performance cement-based materials in the prefabricated parts, a new fabricated underground concrete utility tunnel technology system is introduced—underground utility tunnels in a fabricated composite shell system. In addition, the system’s component structure and connection structure are proposed, and the properties of high-performance cement-based materials are tested. Based on the proposed structure, two roof specimens for utility tunnels are designed and fabricated, and an experimental study is conducted. The key findings are as follows:

(1) The failure processes and modes of composite shell specimen CS-TF and cast-in-place specimen CIS-TF were identical: lower steel bars suffered tensile yield, upper steel bars gradually transformed from compression to tension, and concrete on the compression side was crushed.