In-Situ Experimental Study of Closed-Diaphragm Wall Foundations for Cross-Sea Suspension Bridges

Abstract

This study examines the in-situ lateral static load behavior of a closed-diaphragm wall foundation, aiming to better understand its load–displacement response, structural behavior, and soil interaction under horizontal loading. An in-situ static load test was conducted with a maximum applied load of 70 MN, revealing that the diaphragm wall initially exhibits a linear load–displacement response, which becomes increasingly nonlinear as the load increases. The horizontal displacement of the lateral walls is nearly identical to the overall displacement of the diaphragm wall, making it a reliable indicator of the wall’s load state, particularly when it is challenging to measure total displacement. The wall behaves as a rigid body with minimal relative displacement between sections, and overturning failure is identified as the primary failure mode. Earth pressure distribution varies around the wall: passive earth pressure is observed at the front edge, while active and passive pressures alternate at the rear edge. These findings provide valuable insights into the design of diaphragm wall foundations, emphasizing the importance of lateral displacements.

1. Introduction

With the increasing span length of suspension bridges, the load carried by the main cables also increases. Although direct foundations or large-diameter bored pile foundations can satisfy the load-bearing requirements, they often result in a significant increase in the foundation’s plan dimensions. In contrast, closed-wall foundations, as a three-dimensional foundation type, offer superior horizontal load-bearing performance and, combined with the distinct advantages of diaphragm wall foundations such as high construction efficiency, strength, stiffness, and resistance to water and soil pressure, are increasingly being adopted as a key foundation type in modern bridge engineering [1,2]. It is widely used in bridge projects and plays a critical role in the design and construction of cross-sea bridges. The closed-diaphragm wall foundation, formed by diaphragm wall units, improves both the stability and bearing capacity of the foundation while effectively handling complex seabed terrain and enormous soil and water pressures, becoming one of the critical technologies for overcoming various technical challenges in marine engineering. In this context, the closed-diaphragm wall is hereafter referred to as the “closed wall”.

The Hakuchō Bridge [3,4], a cross-sea bridge project in Japan, is the first one to incorporate an ultra-deep closed-diaphragm wall foundation into bridge engineering. The main tower foundation uses a cylindrical diaphragm wall with an outer diameter of 37 m, a wall thickness of 1.5 m, and bottom elevations at TP73 m and TP57 m. The foundation was successfully built under seabed conditions with an inverted bell-shaped terrain. Three-dimensional finite element analysis [5] shows that the polygonal section of the closed-diaphragm wall foundation results in about a 25% increase in stress compared to the circular section, emphasizing the crucial role of a cross-sectional shape in stress distribution. This move offers new insights and approaches for the design of marine bridge foundations.

The Akashi Kaikyō Bridge [6,7,8,9], crossing the Akashi Strait, utilizes a closed-diaphragm wall foundation for the Kobe-side anchorage, with a diameter of 85 m, a wall thickness of 2.2 m, and a depth of 64.5 m, effectively withstanding the soil and water pressures experienced during excavation. The construction of this foundation was divided into multiple phases, with strict control of verticality and horizontality to ensure the stability of the anchorage. No deviations occurred during the construction process, showcasing the application potential of diaphragm wall technology in super-large bridge projects.

The Yi Sun-Sin Bridge [10,11], linking Maodao-dong in Yeosu, South Jeolla Province, with Geumho-dong in Gwangyang, also employs circular diaphragm walls as the anchorage foundation on the Gwangyang side, with a construction depth of 33 m. The diaphragm wall was initially built as a retaining structure, followed by excavation and concrete filling, ensuring the stability and safety of the foundation throughout the construction process. This design effectively copes with the complex geological conditions, offering useful references for similar projects.

The design of the southern anchorage foundation for the Izmit Bridge [12,13,14,15] in Turkey, spanning the Bay of Izmit, employs an innovative combination of diaphragm walls and a three-cable saddle pier, successfully integrates both the anchorage and pier into one cohesive foundation system. The project utilized an island-building method to fill soil and create a dry environment initially and then proceeded with the construction of diaphragm walls and excavation of the foundation pit, ensuring both the stability of the anchorage and smooth construction progress. The anchorage features a banjo-shaped layout, made up of two circular diaphragm walls with a diameter of 58 m and a rectangular section, effectively handling the complex geological and hydrological conditions of the marine environment.

The anchorage foundation on the Asian side of the 1915 Çanakkale Bridge [16,17], spanning the Bay of Izmit in Turkey, uses an innovative barrette foundation system consisting of diaphragm wall units. The design addresses large horizontal tensile forces and bending moments by employing a foundation system composed of seven rows of barrettes, integrating impermeable cohesive soil layers to effectively provide water tightness and seepage prevention. This innovative design not only solves geological challenges but also significantly reduces the use of concrete and steel reinforcement, showcasing the economic benefits under resource limitations and environmental constraints.

Engineering practices from existing cross-sea bridge projects demonstrate that diaphragm wall foundations improve both the stability and load-bearing capacity of the foundation while reducing construction complexity through optimized design, advancing the development of marine bridge foundation technology. This study is based on the Kahaluo Bridge project in China, where an in-situ lateral static load test was carried out on a closed-diaphragm wall foundation consisting of diaphragm wall units, and aims to investigate the stress behavior of the closed-diaphragm wall foundation and its interaction mechanism with the soil.

2. Geotechnical Profile and Method of Construction

2.1. Geotechnical Profile

The experimental site is located on a slope along the Jinsha River in the Kahaluo Township, Sichuan Province, China, characterized by a thick overburden layer. Geological surveys reveal that the experimental closed wall is situated in Upper Pleistocene collapse slope alluvium, consisting primarily of block stones, gravel, and breccia. The stone is mainly composed of strongly to moderately weathered mudstone. The deep collapse body experiences minimal weathering and geological movement, resulting in dense boulders, while the upper to middle portions are more weathered, degrading into gravel. The breccia is mainly derived from interlayer fractures within the collapsed strata.

The geological profile consists of Quaternary Upper Pleistocene colluvial deposits, with deep collapse bodies composed of dense and poorly permeable boulders made of silty mudstone and some argillaceous limestone, with particle sizes greater than 200 mm. The shallower collapse bodies are moderately dense gravel layers, with better permeability, composed of similar materials and particle sizes ranging from 20 mm to 200 mm. Breccia layers, commonly found in fracture zones, can be soft and easily crushed, forming granular fragments due to bedrock compression.

The overburden layer is water-deficient and impermeable, with scarce groundwater resources, and the foundation-bearing layer does not contain groundwater. Geological profiles along the bridge direction are shown in Figure 1.

2.2. Method of Construction

A closed-diaphragm wall is constructed using the diaphragm wall construction technique. Upon completion of construction, the internal soil core is preserved, and a rigid top slab is placed on top to form an integrated structure. As illustrated in Figure 2, the closed-diaphragm wall foundation has the following dimensions: Bx is 11.8 m, By is 18 m, and Bz is 26 m, with a wall thickness of 1.2 m. The closed wall is formed by L-shaped diaphragm walls at each of the four corners, a rectangular cross-section diaphragm wall along the Bx side, and two rectangular cross-section diaphragm walls along the By side. Adjacent diaphragm walls are connected using rigid joints to form a closed-diaphragm wall foundation.

Guide walls were constructed before trench excavation to ensure precise positioning and vertical alignment of the diaphragm wall. Bentonite slurry wall stabilization was employed during excavation. Each diaphragm wall section took approximately 36 h to excavate, followed by immediate concrete pouring. The diaphragm wall reinforcement cage consisted of hot-rolled ribbed steel bars as the primary reinforcement, with both the vertical and transverse bars having a diameter of 32 mm and a yield strength of 400 MPa. The concrete used for the diaphragm wall was underwater concrete with a compressive strength grade of 40 MPa. According to Chinese standards JGJ 106-2014 [18], it meets the Class I pile standards, indicating the diaphragm wall is complete and of good quality, without significant defects.

3. Lateral Load Test

The specific arrangement of the test is shown in Figure 3. Horizontal loads are applied using seven hydraulic jacks in parallel, each capable of applying a maximum load of 11 MN. Due to the large cross-sectional dimensions of the foundation, a rigid distribution platform is utilized to evenly distribute the load on the closed wall, with circular steel bars placed under the distribution platform to reduce friction. The test’s planar arrangement is shown in Figure 3a. Under horizontal loading, the two closed walls of the split box-type diaphragm wall foundations interact to provide mutual reaction forces, with one closed wall serving as the test object.

Two inclinometer casings were installed within the front wall and the rear wall. At the midpoints of lateral walls, one inclinometer casing was installed at each location where the rigid joints of the single diaphragm wall connect. Each inclinometer casing is 26 m long, and the inclinometers record horizontal displacement changes at various points of the foundation, as illustrated in Figure 3a.

A row of earth pressure cells was installed at the midpoints along the depth direction on both the inner and outer sides of the walls at points A, B, C, and D of the closed wall to monitor the changes in earth pressure, as illustrated in Figure 3c.

As shown in Figure 3a,c, Linear Variable Differential Transformers (LVDTs) were installed on the top of the diaphragm wall to monitor the displacement variations during the test. The LVDTs for monitoring horizontal displacements were spaced 1 m apart.

This experiment followed the Chinese standard JTG/T 3512-2020 [19], using a slow-loading and maintained-load method, with the final load reaching 70 MN. The initial load was 4 MN, increasing in increments of 2 MN over 34 stages. This method is similar to ASTM D1143-81, both specifying the use of maintained constant load, implemented through a series of equal load increments, and ensuring that each load level is maintained for a period of time.

4. Test Result and Analysis

4.1. Load–Displacement Response

During the horizontal load application process, the load–displacement response at the top was measured by the upper LVDTs at point A of the closed wall and is shown in Figure 4a. The load–displacement response is nearly linear, gradually exhibiting nonlinear characteristics as the load increases. Figure 4b illustrates the relationship between the ratio of displacement increment to load increment at 0 m and 1 m from the top of the closed wall as the load increases. As the load increases, the ratio of displacement increment to load increment gradually decreases, indicating that the displacement produced by a unit load gradually decreases. The nonlinear interaction between the closed wall and the soil becomes more pronounced.

4.2. Lateral Displacement

Figure 5 illustrates the displacement variation of the closed wall along its depth under a horizontal load. The top of the closed-diaphragm wall experiences the greatest displacement, which gradually decreases with depth and shows a slight overall bend. Specifically, at the front wall and rear wall, the displacement of the diaphragm wall does not decrease to zero near the foundation bottom, whereas at the lateral walls, the displacement approaches zero near the foundation bottom. The difference in displacement variation with depth at various positions is that the nonlinearity of displacement variation with depth at the front wall and rear wall is more pronounced than at the lateral walls. This is because the local cross-sectional bending stiffness of the diaphragm wall at the front wall and rear wall is smaller than at the lateral walls, resulting in a higher susceptibility to deformation under a horizontal load.

Figure 6 illustrates the pattern of relative deformation with depth at various positions of the closed-diaphragm wall. Specifically, the maximum relative deformation of the lateral and front walls is approximately 0.4 mm, and it is approximately 1.2 mm between the rear wall and lateral walls and approximately 1.6 mm between the rear wall and the front wall. The maximum relative deformation of different sections of the closed wall is only 0.13% of the wall thickness, and the relative deformation is much smaller than the smallest dimension of the closed-diaphragm wall. Therefore, it can be concluded that the deformations of all parts of the closed-diaphragm wall are synchronous, allowing it to be treated as a rigid body model for force analysis. This conclusion is consistent with the assumption of the closed-diaphragm wall foundation as a rigid body, as proposed by Kaino [1].

4.3. Vertical Displacement

Figure 7a shows the vertical displacement at the top slab of the closed wall, where positive values represent upward vertical displacement and negative values represent downward vertical displacement. Under a horizontal load, the vertical displacement at the top slab of the closed wall is upward at the location’s rear wall and downward at the location’s front wall. This behavior is consistent with the vertical displacement variation under a horizontal load, as described by Cheng [20].

Based on the displacement variation analysis of the closed-diaphragm wall under a horizontal load, its motion characteristics combine both translation and rotation, indicating a composite motion state. Under a horizontal load, the closed-diaphragm wall will rotate around a certain point (i.e., the rotation center). Given the coordinates of the two LVDTs at the top of the closed-diaphragm wall as a (x_a, y_a) and b (x_b, y_b), and assuming the coordinates of the rotation point are (x₀, y₀), the horizontal rotation angle and the vertical coordinate of the rotation point can be calculated using Equation (1) based on geometric relationships. Similarly, the vertical rotation angle can be calculated as the horizontal rotation angle using this equation.

$$\begin{array}{l}{\theta }_1={\tan }^{-1}\left(\frac{\Delta x_a}{y_a-y_0}\right)\\ {\theta }_2={\tan }^{-1}\left(\frac{\Delta x_b}{y_b-y_0}\right)\\ {\theta }_1={\theta }_2\end{array}$$

(1)

θ₁ and θ₂ represents the horizontal rotation angle, Δx_a is the horizontal displacement at the top of the closed wall at sensor a, and Δx_b is the horizontal displacement at the top of the closed wall at sensor b.

The horizontal and vertical rotation angles at the top slab of the closed wall are shown in Figure 7b. Under the same load, the horizontal rotation angle is greater than the vertical rotation angle. As the load increases, the difference between them gradually becomes more pronounced, indicating that under a horizontal load, the closed wall primarily exhibits horizontal displacement.

4.4. Earth Pressure

The distribution of the horizontal earth pressure at different positions of the closed-diaphragm wall under a horizontal load is depicted in Figure 8. The variations in earth pressure shown in the figure are based on the increments and decrements of the horizontal earth pressure of the closed-diaphragm wall in a static state.

Figure 8a shows that the earth pressure increment at point A, located outside the front edge of the closed-diaphragm wall, is positive, indicating that earth pressure increases with the load. In addition, earth pressure gradually decreases with depth.

As depicted in Figure 8b, the horizontal earth pressure distribution at point B, located inside the front edge of the closed-diaphragm wall, exhibits the following trend with depth: earth pressure initially increases and then decreases, reaching its peak at a depth of 11 m, followed by a subsequent decrease. At a depth of 3 m, the earth pressure increment becomes negative, whereas below 5 m, the earth pressure increment is positive. This indicates that under a horizontal load, the closed-diaphragm wall and surrounding soil may tend to separate within the 3 m depth range, while below 5 m, the closed-diaphragm wall and internal soil core remain in contact.

As depicted in Figure 8c, the variation in the horizontal earth pressure of the soil core inside the closed-diaphragm wall at point C shows the following characteristics: under a horizontal load, within a depth range of 5 m below the top, the increment of earth pressure is positive, indicating that earth pressure increases with the load. However, when the depth exceeds 5 m, the increment of earth pressure in the internal soil core becomes negative, suggesting that earth pressure gradually decreases with the increasing load.

As illustrated in Figure 8d, the earth pressure at point D, located on the external side of the closed-diaphragm wall, shows a negative increment, indicating that earth pressure decreases as the load increases. The trend of changes in earth pressure is divided into two sections: in the upper section, as the depth increases, the decrease in earth pressure becomes more pronounced; in the lower section, as the depth increases, the decrease in earth pressure becomes less pronounced. The turning point under the maximum load occurs at a depth of about 7 m.

5. Discussion

5.1. Motion State

The vertical and horizontal displacements at the top of the closed wall are used to determine the motion trajectory of its rotation point through geometric relationships, as illustrated in Figure 9a. The trajectory of the rotation point shows minor variation along the vertical axis (Bz: depth range), changing from 16.8 m to 17.5 m, with the vertical variation range being 0.67 Bz to 0.7 Bz. Along the horizontal axis (By: length range), it varies significantly from 4.4 m to 10.5 m, with the horizontal variation ranging from 0.24 By to 0.58 By. This indicates that the rotation point’s position is greatly influenced by horizontal displacement, and the displacement of the closed wall is primarily horizontal. This observation is consistent with the analysis of the top angle of the foundation (Figure 6b). As the load increases, the horizontal movement of the rotation point gradually decreases, suggesting that it may stabilize at a certain position in the ultimate state. The trajectory of the rotation point is shown in Figure 9b. Combined with the displacement analysis of the closed wall in Figure 4, it is clear that the horizontal load-bearing failure mode of the interaction between the closed wall and the soil is an overall overturning failure.

5.2. Soil–Foundation Interaction

As shown in Figure 5 and Figure 6, the analysis of displacements and relative deformations at different positions of the closed wall indicates that under the overall force state, the closed wall can be regarded as a rigid body model. In practical engineering, in addition to the overall movement of the closed wall, the reinforced concrete material will deform under load, and this deformation will further influence the interaction mechanism between the foundation and the soil. Therefore, when analyzing the interaction between the interior of the closed wall and the soil, the relative deformation of the closed wall under load needs to be considered.

By combining the relative displacements of AB and CD of the closed wall in Figure 6c and the earth pressure distribution with depth, shown in Figure 8a,d, it is evident that under a horizontal load, the closed wall undergoes horizontal displacement, compressing the external soil at the front edge and forming a passive pressure zone. Conversely, the back edge of the closed wall exhibits a trend of separation from the surrounding soil, creating an active earth pressure zone.

Figure 6a illustrates the relative deformation characteristics of the closed wall at positions EF and AB, while Figure 8b depicts the distribution pattern of earth pressure along the depth inside the front edge of the closed wall. The distribution pattern of earth pressure with depth at point B inside the closed wall is consistent with the trend of its relative deformation with depth. Specifically, under a horizontal load, the diaphragm wall at the front edge of the closed wall produces relative inward displacement, resulting in the internal soil being compressed and forming a passive pressure zone.

Comparative analysis of Figure 6b and Figure 8c indicates that the variation trend of earth pressure with depth at point C inside the closed wall highly matches the relative deformation trend at that point. Therefore, a stratified pressure characteristic is formed at point C inside the closed wall, with significant passive soil pressure in the upper part and reduced active soil pressure in the lower part.

Based on the soil pressure distribution and the relative deformation of the closed wall, the interaction model between the closed wall and the soil is determined, as illustrated in Figure 10. At point A on the outer side of the front edge of the closed wall, a passive pressure zone is formed. Similarly, a passive pressure zone is formed inside point B on the front edge of the closed wall due to the deformation of the diaphragm wall. At point C on the inner side of the back edge of the closed wall, within a depth of 5 m, a passive pressure zone is formed, while an active soil pressure zone is formed at the lower part of point C’. At point D on the outer side of the back edge of the closed wall, the diaphragm wall separates from the soil, creating an active soil pressure zone.

6. Conclusions

In this study, we conducted an in-situ lateral static load test on the closed-diaphragm wall foundation, and the following conclusions are drawn from the analysis of the test:

• The closed-diaphragm wall exhibits an initially linear, but increasingly nonlinear load–displacement response as the horizontal load increases. The horizontal displacement of the lateral walls closely mirrors the overall displacement of the diaphragm wall, making it a reliable indicator of the closed wall’s load state, especially in cases where it is difficult to measure total displacement.
• The closed-diaphragm wall exhibits rigid-body deformation characteristics under horizontal loading, with minimal relative displacement between sections or components. Its primary failure mode is tilting, emphasizing the importance of considering rotational effects in its design. The rotation point varies, with horizontal variation ranging from 0.24 By to 0.58 By and vertical variation ranging from 0.67 Bz to 0.7 Bz.
• Earth pressure around the closed-diaphragm wall varies, as follows: at the front edge, both inner and outer soils experience passive earth pressure. At the rear edge, the upper soil is under active pressure, while the lower soil is subjected to passive pressure. This differential soil behavior is critical for ensuring the stability and resistance of the foundation.

Field tests were employed in this study to investigate the closed-wall foundation suitable for long-span suspension bridges. The motion characteristics of the closed-wall foundation and its interaction with the surrounding soil were clarified in this study. Nevertheless, an in-depth investigation into the calculation method of the bearing capacity of the closed-wall foundation remains insufficient. Future investigations may combine field tests under horizontal loads with established theoretical methods to refine the bearing capacity model, offering a scientific basis for the practical application of closed-wall foundations.