Multi-Dimensional Stationary Random Seismic Response of Concrete-Filled Steel Tube Arch Bridges

Literature Overview

This paper by Zhao Canhui and Zhou Zhixiang, published in the Journal of Southwest Jiaotong University (2005, Vol. 40, No. 2), investigates the seismic response characteristics of concrete-filled steel tube (CFST) arch bridges under multi-dimensional stationary random earthquake excitation. The research was funded by the National Basic Research Program of China (2003CCA01500) and represents an important contribution to the seismic design methodology for CFST arch bridge structures. The study combines random vibration theory with a novel virtual excitation method to efficiently compute the seismic response under complex loading conditions.

Core Methodology

Random Vibration Theory Framework

The paper applies stationary random vibration theory to analyze the seismic response of CFST arch bridges. The governing equation of motion for a multi-degree-of-freedom (MDOF) system under random excitation is expressed as:

[M]{ü}(t) + [C]{u̇}(t) + [K]{u}(t) = {F}(t)

where [M], [C], and [K] are the mass, damping, and stiffness matrices respectively, {u}(t) is the displacement vector, and {F}(t) is the random excitation force vector.

For stationary random excitation, the response statistics are characterized by the power spectral density function (PSDF) and the auto-correlation function. The mean square response (root mean square, RMS) is obtained by integrating the response PSDF over all frequencies.

Virtual Excitation Method

The key innovation of this paper is the application of the virtual excitation method to multi-dimensional seismic excitation. This method transforms the random vibration problem into a deterministic time-domain problem by representing the stationary random process as a sum of sinusoidal components with random phases. The virtual excitation method offers several advantages over traditional frequency-domain methods:

The virtual excitation method was shown to have the same computational accuracy as the traditional random vibration theory but with significantly improved computational efficiency. This makes it particularly suitable for parametric studies involving multiple design variables.

Key Technical Findings

Effect of Multi-Dimensional Excitation

The study examined the effects of horizontal (two orthogonal directions) and vertical seismic excitation on the internal forces in the arch ribs. The key findings are:

1. Lateral excitation effect: The lateral (horizontal) excitation perpendicular to the arch plane has a significant influence on the internal forces in the arch ribs. This is because the arch ribs are generally less stiff in the lateral direction compared to the in-plane direction.
2. Vertical excitation effect: The vertical excitation also has a notable influence on the arch rib internal forces. This is somewhat counterintuitive, as the arch is primarily designed to carry vertical loads through compression. However, the dynamic interaction between the vertical excitation and the arch's vibration modes can produce significant additional forces.
3. Combined effect: The combined multi-dimensional excitation produces response characteristics that cannot be adequately captured by considering each excitation direction independently. The correlation between different excitation directions is critical for accurate response prediction.

Effect of Transverse Stiffener (Cross-Tie) Stiffness

The paper investigated the influence of the transverse stiffener (cross-tie) stiffness on the structural response. The cross-ties connect the two arch ribs and provide lateral stability and load distribution between the ribs.

The key finding is that the internal force response in the arch ribs tends to increase with increasing cross-tie stiffness. This is an important and somewhat counterintuitive result that has significant implications for the design of CFST arch bridges:

The explanation for this phenomenon lies in the dynamic interaction between the arch ribs and the cross-ties. When the cross-tie stiffness is low, the two arch ribs can move independently, and the system behaves more like two separate arches. As the cross-tie stiffness increases, the ribs are forced to move together, which changes the vibration modes and can lead to higher dynamic amplification factors for certain vibration modes.

Seismic Response Characteristics of CFST Arch Bridges

The CFST arch bridge exhibits several distinctive seismic response characteristics:

• High stiffness and strength: The CFST arch ribs provide high axial and flexural stiffness, resulting in relatively short natural periods. This means the structure is primarily excited by high-frequency components of the earthquake motion.
• Non-linear behavior: Under strong seismic excitation, the CFST arch ribs may exhibit non-linear behavior due to steel yielding, concrete crushing, or steel-concrete interface slip. The virtual excitation method can be extended to handle non-linear systems, making it suitable for this analysis.
• Energy dissipation: The concrete infill provides additional energy dissipation through micro-cracking, which can reduce the overall seismic response. However, this also introduces non-linearity and potential degradation of the structural system.

Standards and Design Implications

Seismic Design Code Compliance

The findings of this paper have direct implications for the seismic design of CFST arch bridges according to relevant standards:

Most current design codes use the response spectrum method as the primary seismic design tool, supplemented by time history analysis for important structures. The random vibration approach presented in this paper provides an alternative methodology that can be used for preliminary design, parametric studies, and verification of the response spectrum method results.

Design Recommendations

Based on the findings of this paper, the following design recommendations are proposed for CFST arch bridges:

1. Multi-dimensional excitation consideration: For important CFST arch bridges, the seismic design should consider the combined effects of horizontal and vertical excitations, as well as the correlation between different excitation directions.
2. Cross-tie design: The cross-tie stiffness should be optimized rather than simply maximized. Excessive cross-tie stiffness can lead to higher dynamic internal forces in the arch ribs. A balanced design that provides adequate lateral stability while minimizing dynamic amplification is recommended.
3. Vibration isolation: For large-span CFST arch bridges, the use of seismic isolation bearings at the arch feet can significantly reduce the seismic response. The isolation system should be designed to accommodate the large displacements expected under strong earthquakes.
4. Damping enhancement: Additional damping devices (e.g., viscous dampers, friction dampers) can be installed to reduce the seismic response of CFST arch bridges. The effectiveness of these devices should be evaluated using the random vibration methodology.

Integration with Engineering Practice

Case Study Application

Consider a typical CFST arch bridge with the following characteristics:

Using the virtual excitation method, the seismic response of this bridge can be analyzed under multi-dimensional stationary random excitation. The analysis would reveal:

• The fundamental vibration modes and their contributions to the overall response.
• The distribution of internal forces (axial force, shear force, bending moment) in the arch ribs under seismic excitation.
• The effect of cross-tie stiffness on the response, allowing optimization of the cross-tie design.
• The influence of vertical excitation on the arch rib forces, which may be significant for this arch rise ratio.

Quality Control Considerations

From a fabrication and construction quality control perspective, the seismic design of CFST arch bridges requires particular attention to:

• Concrete infill quality: The concrete infill must be uniformly compacted to ensure full composite action between the steel tube and concrete. Voids or honeycombing in the concrete can significantly reduce the seismic capacity of the arch rib.
• Steel tube integrity: The steel tube must be free from manufacturing defects (weld seams, cracks, surface imperfections) that could initiate failure under cyclic seismic loading.
• Steel-concrete bond: The bond between the steel tube and concrete is critical for the composite action. Any debonding or slip at the interface can lead to premature failure under seismic loading.
• Weld quality: All welds in the arch ribs and cross-ties must be inspected using appropriate non-destructive testing (NDT) methods (UT, MT, PT) to ensure they can withstand the cyclic loading expected during earthquakes.

Key Questions and Reflections

Several important questions arise from this study:

1. Non-stationary random excitation: The paper focuses on stationary random excitation, which is a simplification of real earthquake motion. Real earthquakes are non-stationary processes with evolving frequency content and amplitude. How does the non-stationary nature of earthquake excitation affect the seismic response of CFST arch bridges?
2. Non-linear response: The linear elastic analysis presented in the paper may not capture the full seismic response of CFST arch bridges, especially under strong earthquakes. The non-linear behavior of the steel-concrete composite material, including concrete crushing, steel yielding, and interface slip, can significantly affect the response.
3. Fatigue and cumulative damage: CFST arch bridges are subjected to repeated seismic loading during their service life. The cumulative damage from multiple earthquakes, as well as from traffic and wind loading, can affect the long-term seismic capacity of the structure.
4. Post-earthquake assessment: After a major earthquake, the seismic damage assessment of CFST arch bridges is challenging because the internal forces in the arch ribs are difficult to measure directly. The development of non-destructive testing methods for assessing the post-earthquake condition of CFST members is an important research need.

Study Insights and Implications

This paper makes a valuable contribution to the seismic analysis methodology for CFST arch bridges. The virtual excitation method provides an efficient and accurate tool for analyzing the multi-dimensional seismic response, and its application to CFST arch bridges demonstrates the importance of considering all excitation directions in the seismic design of these structures.

The finding that cross-tie stiffness affects the arch rib internal forces in a non-trivial manner is particularly important for practical design. It suggests that the cross-tie design should be optimized through dynamic analysis rather than simply following empirical rules. This optimization can lead to more efficient and economical designs while maintaining adequate seismic performance.

From a materials engineering perspective, the seismic performance of CFST arch bridges depends critically on the quality of the steel tube and concrete infill. The steel tube must have adequate ductility to accommodate the inelastic deformations expected during strong earthquakes, while the concrete must have sufficient strength and toughness to provide confinement and energy dissipation. The selection of appropriate steel grades (e.g., Q345 with adequate elongation) and concrete grades (e.g., C50 with supplementary cementitious materials for improved toughness) is essential for achieving the desired seismic performance.

The research also highlights the importance of interdisciplinary collaboration in structural engineering. The effective analysis and design of CFST arch bridges requires expertise in materials science (steel and concrete properties), structural dynamics (vibration analysis), earthquake engineering (seismic hazard assessment and design), and fabrication engineering (welding, concrete placement). Only through such collaboration can the full potential of CFST technology be realized in bridge engineering.