Analysis of Calculation Models for Concrete-Filled Steel Tube Dumbbell Sections: A Practical Review
Overview of the Research
This paper by Zhou Shuixing, Xiong Hongbin, and Zhang Min, published in Civil Engineering and Environmental Engineering in 2009, addresses a specific but important problem: how to analyse a concrete-filled steel tube (CFST) arch rib with a dumbbell-shaped cross-section. This section type is common in long-span arch bridges in China, where two circular CFST chords are connected by steel webs and filled with concrete in the web cavities. The authors propose a spatial composite beam element model that treats the steel tubes, the in-tube concrete, the webs, and the web-cavity concrete as separate beam elements, connected through a transformation matrix based on the plane section assumption. They then develop a modified lattice column method for calculating the axial forces in the upper and lower chords.
Core Technical Points and Interpretation
The Spatial Composite Beam Element
The dumbbell section is geometrically complex, and the paper's contribution is to simplify it without losing the essential mechanics. The model divides the section into four components: the upper steel tube and its concrete core, the lower steel tube and its concrete core, the connecting webs, and the concrete in the web cavity. Each component is modelled as a standard space beam element, and the plane section assumption is used to derive a transformation matrix that links the degrees of freedom of the individual elements to those of the composite section. This is a classic approach in composite structure analysis, and the authors implement it in a Fortran program.
The Lattice Column Method
The key finding from the numerical examples is that the steel tubes and the in-tube concrete carry almost all the axial force, while the bending moment is very small. This is typical for a CFST arch rib, where the axial force dominates. The authors then propose a modified lattice column method, which treats the dumbbell section as a lattice column with two chords and a web system. The modification is to account for the fact that the web cavity concrete contributes to the axial stiffness, but not in the same way as the chords. The modified formula gives results that agree well with the finite element analysis.
| Component | Axial Force Share | Bending Moment Share | Practical Note |
|---|---|---|---|
| Steel tubes (upper + lower) | 55–65% | Small | Dominant load path |
| In-tube concrete | 25–35% | Small | Confined by steel tube |
| Web steel plates | 5–10% | Moderate | Provides shear connection |
| Web cavity concrete | 5–10% | Small | Adds stiffness but not strength |
Integration with Manufacturing and Delivery Practice
Fabrication Challenges of Dumbbell Sections
This is where the paper connects to my world. A dumbbell section is not a standard product. It is fabricated by welding two circular tubes together with web plates, and then filling the cavities with concrete. The fabrication sequence is critical. The webs must be welded to the tubes with full penetration welds, and the welding sequence must be controlled to minimise distortion. I have seen dumbbell sections where the welding distortion was so severe that the two chords were no longer parallel, and the section had to be straightened with heat. This is a quality control issue that the paper does not address, but it is essential for the section to behave as assumed in the analysis.
The Concrete Filling Problem
The web cavity concrete is a particularly difficult element to place. The cavity is narrow, and the concrete must be pumped in from one end. If the web plates are not properly vented, air pockets can form, reducing the effective area and creating a weak point. I have seen this in practice, and it is a common cause of cracking in the web plates during the service life. The paper assumes that the web cavity concrete is fully effective, but in reality, the quality of the concrete in the cavity is highly variable. This is a gap in the model.
The Importance of the Lattice Column Method
The modified lattice column method is a practical tool because it gives the designer a simple way to estimate the axial forces in the chords without running a full finite element analysis. This is useful for preliminary design and for checking the results of more sophisticated analyses. The fact that the bending moments are small means that the chords are primarily in axial compression, which is a favourable condition for the steel tube—it delays local buckling and improves the confinement of the concrete core.
Key Questions and Reflections
- How does the plane section assumption hold up when the section is subjected to significant shear and torsion? The paper does not address this, and it is a real concern for arch ribs under asymmetric loading.
- The model assumes perfect bond between the steel and the concrete. In reality, the bond can degrade due to shrinkage, creep, and cyclic loading. This would reduce the composite action and increase the bending moments in the chords.
- The modified lattice column method is calibrated against the finite element model, but it is not validated against full-scale tests. This is a limitation that should be acknowledged.
Study Insights and Implications
This paper is a good example of how a complex structural problem can be simplified into a practical design tool. The spatial composite beam element model is a rigorous approach, and the modified lattice column method is a useful approximation. For the practising engineer, the key takeaway is that the dumbbell section is primarily an axial load-carrying element, and the design should focus on the axial capacity of the chords. The bending moments are secondary, but they still need to be checked, particularly for the web plates. The fabrication and concrete filling quality are critical to the performance of the section, and these are aspects that the paper does not cover.
Conclusion and Reference Value
This paper provides a solid analytical framework for the analysis of CFST dumbbell sections, and the modified lattice column method is a practical tool that can be used in design offices. The spatial composite beam element model is a good reference for researchers, and the Fortran program is a useful starting point for more advanced analyses. For the practising engineer, the paper is a reminder that the analysis model is only as good as the assumptions behind it. The fabrication quality, the concrete filling, and the bond behaviour are all critical to the actual performance of the section. I would recommend this paper to anyone involved in the design of CFST arch bridges, but I would also advise them to pay close attention to the construction details and the quality control measures.
Zhuojin Pipe Fitting Co., Ltd