Seismic Vulnerability Analysis of Long-Span Concrete-Filled Steel Tube Stiff Skeleton Arch Bridges
1. Literature Overview and Technical Background
The paper by Wang Zhiyuan, Zhao Renda, Wu Xinrui and Zhao Chenggong from Southwest Jiaotong University, published in Journal of Vibration and Shock in 2023, presents a comprehensive seismic vulnerability analysis of a long-span concrete-filled steel tube (CFST) stiff skeleton arch bridge. This research was supported by the Sichuan Provincial Science and Technology Programme under grant 2020YFH066, reflecting the importance of seismic performance assessment for bridge infrastructure in earthquake-prone regions.
The study focuses on a specific bridge type: the upper-deck high-speed railway stiff skeleton arch bridge. This is a complex structural system where the main arch ribs are constructed as steel trusses that are subsequently filled with concrete, creating a composite CFST structure. The stiff skeleton approach is used for long-span arch bridges where traditional construction methods are impractical, and it is particularly common in China's high-speed railway network.
The paper addresses a critical gap in the seismic assessment of these bridges. Traditional seismic vulnerability analysis has focused on individual components, but the authors recognise that the overall vulnerability of the bridge depends on the interaction between components and the correlation of their damage states. This is where the Copula function approach becomes valuable, as it allows the correlation between different component damage states to be modelled explicitly.
2. Key Technical Points and Analysis
2.1 Numerical Modelling Approach
The paper uses the OpenSEES platform to establish a nonlinear numerical model of the bridge. This is a well-established open-source finite element framework that is particularly suited to nonlinear dynamic analysis of structures. The model includes:
| Component | Modelling Approach | Key Parameters |
|---|---|---|
| Main arch ribs | Fibre-based beam-column elements with distributed plasticity | Material properties of steel and concrete, section geometry |
| Concrete-filled steel tube | Composite section with confined concrete model | Confinement factor, concrete compressive strength |
| Pier columns | Fibre-based beam-column elements | Reinforcement ratio, concrete properties |
| Arch spandrel columns | Fibre-based beam-column elements | Reinforcement ratio, concrete properties |
| Deck system | Elastic beam elements | Stiffness, mass distribution |
| Bearings | Nonlinear spring elements | Stiffness, capacity, gap behaviour |
The use of fibre-based elements is critical for capturing the nonlinear behaviour of the CFST sections, particularly the confinement effect of the steel tube on the concrete core. This is essential for accurate seismic response prediction because the confinement significantly affects the ductility and energy dissipation capacity of the composite section.
2.2 Incremental Dynamic Analysis and Vulnerability Curves
The paper employs incremental dynamic analysis (IDA), which involves subjecting the bridge to progressively scaled ground motion records and recording the structural response at each intensity level. This approach allows the construction of vulnerability curves that relate the probability of exceeding a given damage state to the ground motion intensity, typically measured by peak ground acceleration (PGA).
The damage states are defined based on engineering demand parameters such as:
- Concrete strain (for compression damage)
- Steel strain (for yielding and fracture)
- Curvature ductility (for flexural damage)
- Shear strain (for shear damage)
The vulnerability curves are constructed for individual components (sub-components of the main arch rib, pier columns, and spandrel columns) and for the system as a whole.
2.3 Copula Function for System Vulnerability
The most innovative aspect of this paper is the use of Copula functions to construct the system vulnerability curve for the main arch rib system. The Copula function approach allows the joint distribution of multiple component damage states to be modelled, taking into account the correlation between them.
The paper compares two types of Copula functions:
- Gaussian Copula: Assumes a normal distribution for the underlying variables
- Archimedean Copula (specifically Gumbel and Clayton): Allows for different tail dependence characteristics
The results show that the system vulnerability curve, constructed using Copula functions, differs significantly from both the individual component vulnerability curves and the system vulnerability curve based on the first-order reliability principle. The Copula-based approach provides a more realistic assessment because it captures the correlation between component damage states, which is particularly important for a series system where the failure of any component leads to system failure.
3. Key Findings and Engineering Implications
3.1 Damage Distribution in the Main Arch Rib
The paper reveals that under three-dimensional ground motion, the damage probability and damage degree of the outer concrete (the concrete outside the steel tube) are significantly greater than those of the steel tube and the concrete inside the tube. This is an important finding because it suggests that the outer concrete is the weak link in the composite system.
The vulnerability location of the steel tube changes with increasing PGA: at lower PGA levels, the critical section is at the mid-span, but as PGA increases, the critical section shifts towards the arch springing. This is a complex behaviour that results from the interaction between the arch geometry, the distribution of seismic forces, and the nonlinear response of the composite section.
3.2 System Vulnerability Characteristics
The system vulnerability analysis shows that:
- The vulnerability of the main arch rib system is more strongly influenced by the outer concrete than by the steel tube or the inner concrete
- The L/4 arch rib section has the highest damage probability
- The pier columns and spandrel columns have significantly higher vulnerability in the concrete cover than in the longitudinal reinforcement or the core concrete
- The columns at the L/4 arch rib position have the highest damage probability
These findings have direct implications for seismic design and retrofitting. The L/4 section of the arch rib should be a priority for strengthening, and the concrete cover in the columns should be designed with adequate confinement to prevent spalling.
3.3 Comparison with First-Order Reliability Principle
The paper compares the Copula-based system vulnerability with the system vulnerability based on the first-order reliability principle. The first-order reliability principle assumes that the components fail independently, which leads to a conservative estimate of system vulnerability (i.e., higher failure probability for a given PGA).
The Copula-based approach provides a more accurate estimate because it accounts for the correlation between component damage states. The paper demonstrates that the Copula approach is particularly important when the components have similar damage probabilities, as the correlation effect becomes more significant.
4. Practical Engineering Considerations
4.1 Design Implications
The findings of this paper have several important implications for the seismic design of CFST stiff skeleton arch bridges:
- Strengthening of the L/4 section: The L/4 arch rib section should be designed with additional reinforcement or confinement to resist the high damage probability identified in the analysis.
- Concrete cover protection: The concrete cover in the pier columns and spandrel columns should be designed with adequate confinement to prevent spalling under seismic loading. This may require increased stirrup ratios or the use of fibre-reinforced concrete.
- System-level design approach: The design should consider the system-level vulnerability, not just the individual component vulnerabilities. The Copula-based approach provides a more realistic assessment of system performance.
- Performance-based design: The vulnerability curves can be used directly in performance-based seismic design, where the bridge is designed to achieve a specific performance level (e.g., no collapse, limited damage) under a specified seismic hazard.
4.2 Construction and Inspection Considerations
From a construction and inspection perspective, the findings highlight the importance of:
- Concrete quality: The quality of the concrete in the arch rib and columns is critical for seismic performance. Poor concrete quality can lead to premature spalling and loss of confinement.
- Steel tube fabrication: The steel tube must be fabricated with high precision to ensure proper composite action with the concrete core. Welding quality is particularly important, as weld defects can initiate cracks that propagate under cyclic loading.
- Interface bond: The bond between the steel tube and the concrete core is essential for composite action. Any debonding, as discussed in the related literature on CFST debonding, can significantly reduce the stiffness and strength of the composite section.
4.3 Comparison with Other Bridge Types
| Bridge Type | Seismic Vulnerability Characteristics | Key Design Considerations |
|---|---|---|
| CFST stiff skeleton arch | Outer concrete is most vulnerable, L/4 section critical | Concrete quality, confinement, system-level design |
| Conventional RC arch | Concrete spalling and reinforcement yielding | Ductility, confinement, plastic hinge formation |
| Steel arch | Local buckling and connection failure | Member slenderness, connection design |
| Cable-stayed bridge | Pylon and cable anchorage zones | Cable replacement, pylon ductility |
The CFST stiff skeleton arch bridge has a unique vulnerability profile because of the composite action between the steel tube and the concrete core. The outer concrete is the weak link, and the L/4 section is the critical location.
5. Conclusion and Outlook
This paper by Wang Zhiyuan and colleagues makes a significant contribution to the seismic vulnerability assessment of CFST stiff skeleton arch bridges. The use of OpenSEES for nonlinear modelling, combined with incremental dynamic analysis and Copula functions for system vulnerability, provides a comprehensive and realistic assessment of seismic performance.
The key findings — that the outer concrete is more vulnerable than the steel tube and inner concrete, that the L/4 section is the critical location, and that the Copula-based system vulnerability differs significantly from the first-order reliability approach — have direct implications for seismic design and retrofitting.
For engineers involved in the design, construction or assessment of long-span arch bridges, this paper provides valuable insights into the seismic behaviour of a complex structural system. The methodology developed in this paper can be applied to other bridge types and other structural systems, making it a useful reference for the broader earthquake engineering community.
Future research directions that would build on this work include:
- Consideration of soil-structure interaction effects
- Incorporation of ground motion spatial variability
- Development of fragility curves for different performance levels
- Validation of the numerical model against shake table tests or field observations
The paper is a valuable contribution to the field of seismic vulnerability analysis, and I would recommend it to any engineer working on the seismic design or assessment of long-span arch bridges.
Zhuojin Pipe Fitting Co., Ltd