Optimisation of Parameters and Application of Green Recyclable Precast Hollow Steel Pipe Concrete Supports

Abstract

Underground space development is a crucial approach to addressing traffic congestion in China’s cities, especially through underground construction. However, the traditional pit internal support system, particularly the concrete internal support system, has significant drawbacks. These include high energy consumption and carbon emissions, which are increasingly prominent issues. In this paper, a hollow precast concrete-filled steel tube (H-CFST) internal support system is proposed and a node connection scheme is designed. Through numerical simulation, the load-bearing characteristics of H-CFST with different aspect ratios, hollow ratios, hoop thicknesses, and hoop lengths are investigated, and the optimal design parameters are obtained. Finally, following a field application, monitoring data reveal that the precast H-CFST internal support demonstrates superior load-bearing capacity compared to the concrete internal support, successfully meeting the criteria for replacement of the concrete support.

1. Introduction

Urban underground space rail transit stations are usually constructed by means of open excavation [1], as the location of the station is usually characterized by a large number of people, traffic flow, and dense buildings. In order to ensure construction safety, the pit support is usually chosen to be rigid, with a reliable internal support system, and is combined with other support methods [2].

In recent years, the research on the internal support systems of foundation pits has mainly divided into concrete support, steel support, and other new types of internal support in terms of structural form [3], as shown in Figure 1. Concrete internal support has greater stiffness, so it is designed with long spacing and can avoid the need to divide the construction workspace [4]. However, it has significant disadvantages in terms of high consumption of resources, long maintenance time, and inability to be recycled [5]. In contrast to concrete supports, steel supports can be recycled and have significant ecological significance, but their low stiffness and short design spacing can divide the workspace considerably when in use [6]. Concrete-filled steel tube (CFST) bracing has gradually become a research hotspot in recent years due to its good theoretical foundation, high stiffness, and recyclability [7,8,9,10,11]. CFST member is a combined structure formed by filling concrete into thin-walled circular steel tubes, which is a special form of hooped concrete [12,13,14] and has been widely adopted in structural design due to its combination of the advantages of concrete in compression resistance and steel in tensile strength [15,16]. However, the pair of braces has a huge volume and mass, and there is no condition for transport and construction work, so it needs to be installed and transported to the site [10] for splicing [11]. Therefore, the prefabricated CFST internal bracing system is proposed to reduce the mass of a single member and the construction difficulty by splicing multiple members, but this implies more nodes, which adversely affects the reliability of the brace as a whole as well as the ease of construction. Hollow steel tubular concrete members (H-CFST) developed from CFST members further attenuate the disadvantages of CFST members in terms of weight, and they have the advantages of light weight, high stiffness, and reusability [17,18]. Therefore, if H-CFST members are used to fabricate the internal support, the size of the single member can be further increased without changing the quality of the member, effectively avoiding the disadvantages of multiple member splicing. However, most of the existing studies have used CFST members as their research objects, and there is little research on the fabrication of internal supports with H-CFST members. In addition to the members themselves, the strength and reliability of the connection nodes are also very important. At present, casing [19,20,21] and flange [22,23,24] are the main connection forms for solid/hollow steel pipe concrete members. However, due to the huge size and mass of the internally supported monolithic members, the casing node is not able to complete the monolithic member connection with equal strength [25,26].

Due to the presence of an outer steel tube, CFST members enhance the flexural and shear strength of the core concrete [27], and they possess advantages such as high stiffness, corrosion resistance, and reusability [28]. The H-CFST member, an improvement based on this concept, can substantially reduce its individual weight while maintaining the aforementioned benefits [17]. If utilized as a support in foundation pits, it can mitigate many issues inherent in the use of traditional concrete supports, showcasing significant innovation and practical value. Nevertheless, despite substantial research into the load-bearing characteristics and failure modes of H-CFST members, which has yielded notable achievements, the design remains unoptimized for specific parameters, thus making it challenging to straightforwardly apply this structure to subway foundation pit supports.

This study introduces a novel prefabricated H-CFST internal support system featuring H-CFST members and hoop nodes. It concentrates on four parameters that significantly affect the system’s load-bearing performance: the length-to-width ratio, the hollow ratio, the hoop length, and the thickness. Subsequently, a field application study is conducted to evaluate the system’s performance.

2. Design Solution for the H-CFST Support System

This section proposes the design scheme of the prefabricated H-CFST internal support system using “one intermediate node+two end nodes” and combines the current mainstream connection nodes and field experience to summarise the key issues faced by the two nodes in the design process, then puts forward a detailed node connection scheme. The whole prefabricated H-CFST internal support system includes H-CFST internal support members, hoops, and other components. The support after the connection is completed is shown in Figure 2.

2.1. Nodal Connection Scheme for Prefabricated H-CFST Internal Support System

2.1.1. Intermediate Node Connectivity Programme

This study designed a bolted external connector independent of the internal support members of the H-CFST, as shown in Figure 3, and named it a “hoop” based on its connection. The hoop itself consists of a curved plate and a waist plate with bolt holes in the waist plate and a snap ring in the curved plate. Ribs are provided between the waist plate and the curved plate to increase the stability of the structure, to distribute the forces transferred from the supporting members in the H-CFST to the curved plate to the entire waist plate, and to avoid stress concentrations at the connection between the curved plate and the waist plate.

2.1.2. End Node Connectivity Programme

Combining the design scheme of solid/hollow CFST column end nodes and the site construction conditions, a connection node similar to the CFST column end footing node is designed, as shown in Figure 4a. Several equal-strength bars are welded at the end of the support, connected with the reinforcement cage in the crown beam, and tied into a whole. Then, concrete is poured to form the end node of the prefabricated H-CFST internal support system and an end plate is set up at the end of the H-CFST member to avoid the concrete surging into the member during the pouring process, as shown in Figure 4b.

2.2. Prefabricated H-CFST Internal Support Elements

The completed design of the H-CFST internal support element is shown in Figure 5.

2.3. Axial Compressive Capacity of Hollow Steel Pipe Concrete Members

The combined strength design value is calculated by the following formula:

$$f_{\mathit{sc}}=\left(1.212+B\theta +C{\theta }^2\right)f_c$$

(1)

$${\alpha }_{\mathit{sc}}={\displaystyle \frac{A_s}{A_c}}$$

(2)

$$\theta ={\alpha }_{\mathit{sc}}{\displaystyle \frac{f}{f_c}}$$

(3)

where A_S and A_C are the areas of the steel pipe and the concrete in the pipe; ${\alpha }_{sc}$ is the steel content of solid or hollow steel pipe concrete members; θ is the design value of the set factor for solid or hollow steel tubular concrete members; $f_y$ and f are the value of the tensile strength of the steel and the design value; and $f_{ck}$ and $f_c$ are the standard and design values of compressive strength of coagulation. For hollow members, $f_{ck}$ and $f_c$ are both multiplied by 1.1; B and C are the coefficients of the influence of the shape of the cross-section on the effect of the hoop.

According to the technical specification for solid and hollow steel concrete-filled steel tubular structures (CECS254:2012 [29] Appendix B, and Clause 5.1.1), f_sc = 62.8 MPa is obtained.

3. Parameter Optimisation of the Prefabricated H-CFST Internal Support System

In the structural design and bearing capacity calculation of the prefabricated H-CFST internal support system, some parameters will have a significant influence on the bearing capacity. A finite element analysis model of a prefabricated H-CFST internal bracing system was established by using ABAQUS 2022, and the effects of different design parameters on the load-bearing performance of the bracing were investigated.

3.1. Finite Element Modelling

The specific dimensions of the trial model are shown in

Table 1. Model dimensions.

Projects	Define	Sizes
L	Calculated span of the support system	19.0 m
l	Lengths of the two H-CFST members	7.0 m/12.0 m
D	Diameter of the two H-CFST members	700.0 mm
lh	Length of hoop	700.0 mm
T	Thickness of concrete	100.0 mm
t	Wall thickness of steel pipe	6.0 mm
th	Thickness of the hoop	22.5 mm
rh	Radius of hollow part	244.0 mm

. According to the overview of the project on which it is based (a station pit of Qingdao Metro Line 8), the total length of the model, L, is taken to be 19.0 m. The optimal design location of the intermediate node is at 1/3 of the overall span of the support, and the lengths l of the two H-CFST members are taken to be 7.0 m and 12.0 m, respectively, in consideration of the convenience of factory prefabrication. The two sides of the model are fixedly connected by reference points, and both the concrete and steel pipe examples use C3D8R cells with three translational degrees of freedom at each node. The exact meshing and details are shown in Figure 6.

3.2. Different Aspect Ratios

The aspect ratio λ is the ratio of the length of the H-CFST member to the outer diameter of the steel pipe of the H-CFST member. However, in the actual project, the width of the required support (i.e., the span L of the H-CFST inner support) is determined, and there is no possibility to change the length L of the H-CFST member. The change in material strength will affect the member diameter D. Therefore, this section achieves different aspect ratios by selecting three different materials and then varying the member diameters, as shown in the table below (

Table 2. Design scheme of the finite element model for studying the aspect ratio λ.

Projects	Option 1	Option 2	Option 3
	C30; Q235	C80; Q235	C80; Q355
t	6.0 mm	6.0 mm	4.0 mm
T	100.0 mm	37.5 mm	37.5 mm
rh	244.0 mm	110.0 mm	110.0 mm
D	700.0 mm	307.0 mm	305.0 mm
ψ	≈34%	≈34%	≈34%
λ	27.14	61.89	62.30

).

The material parameters of the finite element analysis model of the precast H-CFST internal bracing system with the studied length to aspect ratio λ are detailed in

Table 3. Material parameters of the finite element model for studying the aspect ratio λ.

Materials	Modulus of Elasticity (GPa)	Tensile Strength (MPa)	Compressive Strength (MPa)	Densities (kN/m3)
Q235b steel	200.0	375.0	235.0	78.5
Q355b steel	200.0	470.0	345.0	78.5
C30 concrete	30.0	2.01	20.1	23.9
C80 concrete	38.0	3.11	50.2	25.2

. The H-CFST members were selected from Q235b or Q355b steel, and the filler concrete was selected from C30 or C80 grade concrete. The strength parameters of concrete were measured by field tests.

The results are shown in Figure 7. Under the same load, the bearing with the “medium-strength steel, medium-strength concrete” (option 1) design principle had the best stability, and the maximum vertical displacement was 12.0 mm. The bearing with “medium-strength steel, high-strength concrete” (option 2) design principle had the worst stability, and the maximum vertical displacement was 228.1 mm. Compared with the bearing with the “medium-strength steel, high-strength concrete” (option 3) design principle and the maximum displacement of the bearing with the “high-strength steel, high-strength concrete” design principle, the maximum displacement of the bearing was shifted to the bearing centre as a whole, and it can be inferred that the bearing with the “high strength steel” design principle was more stable than the bearing with the “medium strength steel, high strength concrete” design principle.

3.3. Different Hollow Ratios

For the study of the hollow fraction ψ of H-CFST members, the method of changing only a single parameter of the support hollow fraction ψ without considering the reduction in its load-bearing performance was also avoided. By changing the materials of H-CFST members, three support models with approximately equal compressive load capacities (fsc) and other parameters were designed. They represent the design processes of “lower strength concrete, smaller hollow ratio”, “medium strength concrete, medium hollow ratio”, and “higher strength concrete, medium hollow ratio”. For higher-strength concrete with a larger hollow rate, the specific design scheme shown in

Table 4. Design scheme of the finite element model for studying the hollow ratio ψ.

Projects	Option 1 C30	Option 2 C50	Option 3 C80
t	6.0 mm	6.0 mm	6.0 mm
D	700.0 mm	700.0 mm	700.0 mm
λ	27.14	27.14	27.14
T	100.0 mm	62.0 mm	40.0 mm
rh	244.0 mm	282.0 mm	304.0 mm
ψ	34%	40%	44%

.

The H-CFST members are made of Q235b steel; the filler concrete is made of C30, C50, or C80 grade concrete; and the clamps are all designed according to strengths far beyond those of the H-CFST members.

When the load reaches 15,000.0 kN, the vertical displacement at the support exceeds 100 mm, indicating significant deformation. Consequently, 1500 kN is considered the ultimate load for the purposes of this study. The stress cloud of the precast H-CFST internal support system is subjected to eccentric loading using different hollow ratios ψ when the load is 15,000.0 kN, as shown in Figure 8. The method of increasing the hollow ratio ψ by selecting a higher-strength concrete has a limited effect on the stress distribution during the load-bearing process of the support system.

3.4. Different Hoop Thicknesses

The hoop thickness t_h mainly refers to the thickness of the arc plate and waist plate of the structure, and this parameter mainly affects the eccentric compression and shear load-bearing performance of the intermediate node. For the study of hoop thickness t_h, the model adopts the same size as the field test of the foot-sized finite element model. The hoop thickness has the same stiffness and strength of the H-CFST member as t_h0, and t_h0 = 15 mm in this model. The variable hoop thickness t_h in each test is increased by 1/2 t_h0 = 7.5 mm, respectively. To study of hoop thickness t_h of the prefabricated H-CFST internal support system, the geometric parameters of the FEA model are detailed in

Table 5. Geometric parameters of the finite element model for studying the thickness t_h of the hoop.

Parametric	Sizes	Parametric	Sizes
l	7.0 m/12.0 m	T	100.0 mm
D	700.0 mm	t	6.0 mm
lh	700.0 mm	rh	244.0 mm
th1	15.0 mm	th2	22.5 mm
th3	30.0 mm	th4	37.5 mm

.

The H-CFST members and hoops are selected from Q235b-type steel, and the filler concrete is selected from C30-grade concrete. Under the condition of different hoop thickness t_h, the stress map of the hoop subjected to eccentric loading at a load of 15,000.0 kN is shown in Figure 9. The stresses at each location of the hoop show a general decreasing trend with the increase in hoop thickness t_h, and the maximum stress is always located at the joint of the arc plate and the waist plate at the transverse centre line of the hoop, away from the eccentric loading point.

3.5. Different Hoop Lengths

For the study of hoop length l_h, the model adopts the same size as the field test of the foot-sized finite element model and selects the hoop length of l_h0, which is equal to the diameter of the H-CFST member in the trial model. Then, l_h0 = 700.0 mm in the model, and the variable hoop length l_h in each test increases by 300.0 mm. The geometric parameters of the H-CFST internal support system are shown in

Table 6. Geometric parameters of the finite element model for studying the length l_h of the hoop.

Parametric	Sizes	Parametric	Sizes
l	7.0 m/12.0 m	T	100.0 mm
D	700.0 mm	t	6.0 mm
th	22.5 mm	rh	244.0 mm
lh1	400.0 mm	lh2	700.0 mm
lh3	1000.0 mm	/	/

.

At 12,500.0 kN, the stress cloud of the hoop subjected to eccentric loading under different hoop length l_h conditions is shown in Figure 10. The stress distribution pattern at each location of the hoop is less affected by the increase in hoop length l_h. Under the same load, the maximum stress is always located away from the eccentric loading point and extends from that position to the top of the hoop with a decreasing trend.

3.6. Prefabricated H-CFST Support System Design Parameters

According to the aforementioned optimal design principles of “medium strength steel, medium strength concrete” and “small hollow ratio”, combined with the optimal design value of hoop thickness t_h, i.e., 1.5 t_h0, and the optimal design value of hoop length l_h, i.e., two times the shear ring length, the corresponding precast H-CFST internal support system was designed for a station foundation pit of Qingdao Metro Line 8. The prefabricated H-CFST internal support system was designed for a station pit of Qingdao Metro Line 8.

According to the project overview and site conditions of a station pit of Qingdao Metro Line 8, the geometric parameters of the corresponding prefabricated H-CFST internal support system are designed as shown in

Table 7. Geometric parameters of prefabricated H-CFST strut system for a station on Qingdao Metro Line 8.

H-CFSTmember: C80; Q355		Hoop: Q355
L	19.0 m	l	7.0 m/12.0 m
D	700.0 mm	lh	600.0 mm
T	100.0 mm	th	35.0 mm
t	6.0 mm	Slot Depth	25.0 mm
rh	244.0 mm	Slot width	25.0 mm

, combined with the optimal design principles and optimal design values of important parameters.

4. Field Application Study of Prefabricated H-CFST Internal Support

4.1. Overview of Field Trials

Based on the construction project of an underground station of Qingdao Metro Line 8, the research on the stress response and deformation law of the prefabricated H-CFST internal support system is carried out in combination with the excavation process of the foundation pit.

The main body of the field test monitoring for the Qingdao Metro Line 8, an underground station pit in the test section of the three prefabricated H-CFST internal supports, the test support is set in the test section between the 13~15 axes. The test section and the locations of the three prefabricated H-CFST internal supports in the pit are shown in Figure 11. Based on the site engineering profile combined with the results of the model test, the detailed geometric parameters of the members, hoops, and end nodes of the prefabricated H-CFST internal support system are shown in

Table 8. The geometric parameters of the prefabricated H-CFST strut system during on-site experiments.

Parametric	Sizes	Parametric	Sizes
L	19.0 m	T	100.0 mm
l	7.0 m/12.0 m	t	6.0 mm
D	700.0 mm	th	35.0 mm
lh	700.0 mm	rh	244.0 mm

.

4.2. Monitoring Programme

The monitoring of the field test was divided into two parts: support axial force and support deflection monitoring.

The axial force of the bearing was monitored by a CK-GJJ-type vibrating string bar dynamometer embedded on both sides of the node and a YB-160-type vibrating string surface strain gauge attached to the surface of the steel pipe. The position of the bearing axial force measurement point is shown in Figure 12.

The deflections at the bearing ends, bearing hoops, and bearing centres were monitored using a NovaTS-60 total station and an ADS-10 prism, and the locations of the measurement points are shown in Figure 13.

4.3. Analysis of Monitoring Results

Figure 14 depicts the relationship between support deflection and soil excavation depth. As the excavation depth increases, the vertical displacement of the test support (i.e., deflection) gradually increases as well. The support deflection is positively correlated with the soil excavation depth. At the end of excavation, the axial force of the concrete internal support is even greater, and the maximum value of its axial force measured at the same time after the completion of excavation is 2162.78 kN, which is an increase of about 31.1% compared with the prefabricated support. At the same time, the concrete support, due to its own shrinkage creep, will lead to a larger axial force when the soil body of the pit is not yet excavated, and the corresponding value of the axial force is 672.33 kN.

Figure 15 shows the development of support deflection with the increase in the soil excavation depth. With the increase in the excavation depth, the vertical displacement of the test support, i.e., deflection, gradually increased and the support deflection grew, with a positive correlation with the soil excavation depth. After excavation, the maximum average deflection of the support was measured at the mid-span measurement point of test support #1. This deflection was 23.30 mm. Additionally, the deflection of the support was measured at the hoop measurement point. This deflection was smaller than that at the mid-span measurement point, with a maximum average value of 20.92 mm. Similarly to the depth of excavation-support axial force curves of the supports, the deflections of each support increased rapidly due to the influence of self-weight when the soil body had just been excavated.

5. Conclusions

Based on a station pit project of Qingdao Metro Line 8, this paper proposes a new type of prefabricated H-CFST internal support system, adopts the method of numerical simulation to optimize the design of each parameter of the H-CFST internal support, and verifies the reliability of the H-CFST internal support system through on-site monitoring of the prefabricated H-CFST internal support system during the whole process of foundation pit excavation. The main conclusions drawn are as follows:

(1)

We adopted the H-CFST structure to make the single pressure-bearing member of the pit internal support system and designed an intermediate node in the middle of the support span by adopting the hoop to bite the H-CFST internal support member. We also designed two end nodes by adopting the method of lapping and casting the reinforcement with the crown beam at the end of the support, which is a practical solution to take into account the transport of the product and the support of the pit.

(2)

The increase in the length-to-aspect ratio λ leads to stability of the support system and a rapid decrease in the damage load. The increase in the hollow ratio ψ leads to an increase in the maximum vertical displacement of the support system, and the design should try to consider choosing a plane with a “smaller hollow ratio”. After the hoop thickness t_h exceeds 1.5t_h0, its increase for the prefabricated H-CFST internal support system bearing performance enhancement is no longer obvious, so the optimal design value for the hoop thickness t_h is 1.5t_h0. The optimal design value of the hoop length l_h is two times the shear ring length to avoid causing local damage to the prefabricated H-CFST internal support system.

(3)

After the excavation of the foundation pit, the performance of the precast H-CFST internal support is better than or comparable to that of the concrete internal support in all aspects, which can meet the requirement for replacing the concrete support.

(4)

In this paper, numerical simulation is used to study the optimisation of the parameters of H-CFST in order to arrive at the best parameters suitable for the present study. Currently, in the era of artificial intelligence (AI) and deep learning (DL), artificial intelligence and artificial neural network (ANN) algorithms are widely used in data research [30,31,32]. In the future, relevant databases should be established to carry out efficient optimisation studies of structural design parameters using deep learning (DL) and other methods.