Nonlinear Analysis of Circular Concrete-Filled Steel Tube Structures Based on the Composite Element Method
Literature Overview and Core Research Contribution
This paper by Liu Wei, Ding Faxing and Xing Feng presents a significant advance in the analytical modelling of concrete-filled steel tube (CFST) structures. The authors propose a "composite element method" that treats the CFST section as a single composite element rather than modelling the steel tube and concrete core separately. This approach is based on practical calculation methods for the axial force-strain relationship and the axial force-moment-curvature relationship of the composite section, which were developed in earlier work by the same research group.
The key innovation is the use of the updated Lagrangian (U.L.) formulation for the incremental equilibrium equations, combined with a nonlinear analysis method for plane beam elements. This allows the authors to perform a double nonlinear analysis — geometric nonlinearity and material nonlinearity — of CFST structures using a composite element approach. The method is implemented in a computer program and validated against experimental results for several types of CFST members, including eccentrically loaded columns, columns with unequal end moments, beam-columns and model arch ribs.
The most important advantage of the composite element method is computational efficiency. The traditional "layered beam element" method requires the section to be divided into many layers, and the stiffness of each layer must be calculated and integrated at each load increment. This is computationally intensive and can be slow for large structures. The composite element method, on the other hand, directly calculates the section stiffness from the end forces, without the need for layer-by-layer integration. This results in a significant reduction in computational time, which is particularly beneficial for large-scale structures or for parametric studies.
Technical Interpretation and Comparison with Layered Beam Element Method
The authors compare the composite element method with the layered beam element method, which is the traditional approach for nonlinear analysis of CFST structures. Both methods are based on the same constitutive relationships for the steel and concrete, and both incorporate the composite effect. The comparison shows that the load-deformation curves and ultimate bearing capacities predicted by the two methods are very close to each other and to experimental results.
| Comparison Aspect | Composite Element Method | Layered Beam Element Method |
|---|---|---|
| Section stiffness calculation | Direct from end forces | Layer-by-layer integration |
| Computational time | Faster | Slower |
| Accuracy | Good agreement with tests | Good agreement with tests |
| Implementation complexity | Moderate | Higher |
| Applicability to large structures | Good | Limited by computational cost |
The layered beam element method is conceptually simpler and more widely understood, but it requires a fine discretisation of the section to achieve acceptable accuracy. This is particularly problematic for sections with complex geometries or for members with varying cross-sections. The composite element method avoids this problem by using a single element for the whole section, with the section properties calculated from the end forces. This is a more elegant and efficient approach, and it is particularly well-suited for the analysis of frames and other structures with many members.
The validation of the composite element method against experimental results for a range of loading cases is a strong point of the paper. The eccentrically loaded columns and the beam-columns are particularly important because they represent the most common loading conditions in real structures. The model arch ribs are also interesting because they demonstrate the applicability of the method to curved members, which are common in bridge structures.
Engineering Practice and Structural Design Implications
From a structural engineering perspective, the composite element method has significant practical implications. The ability to perform nonlinear analysis of CFST structures quickly and accurately is essential for the design of high-rise buildings, bridges and other structures where CFST members are used. In my experience, the design of such structures often involves a large number of load combinations and a detailed assessment of the ultimate limit state, which requires a reliable and efficient nonlinear analysis method.
The composite element method is particularly useful for the analysis of frames, where the interaction between members can have a significant effect on the overall behaviour. In a frame, the redistribution of moments and forces after yielding can be complex, and a nonlinear analysis is required to capture this behaviour accurately. The composite element method allows such an analysis to be performed efficiently, which is a significant advantage over the layered method.
| Loading Case | Composite Element Method | Layered Beam Element Method | Experimental Results |
|---|---|---|---|
| Eccentrically loaded column | Good agreement | Good agreement | Reference |
| Column with unequal end moments | Good agreement | Good agreement | Reference |
| Beam-column | Good agreement | Good agreement | Reference |
| Model arch rib | Good agreement | Good agreement | Reference |
One of the key advantages of the composite element method is that it can be easily extended to three-dimensional analysis. The layered method is primarily suited to two-dimensional analysis, and extending it to three dimensions is complex and computationally expensive. The composite element method, on the other hand, can be extended to three-dimensional beam elements with relative ease, which opens up the possibility of analysing three-dimensional frames and space structures.
Key Questions and Reflections
The paper focuses on the analysis of CFST structures under monotonic loading. However, in real structures, members are often subjected to cyclic loading, such as during earthquakes or wind storms. The behaviour under cyclic loading is more complex because of the Bauschinger effect in the steel and the degradation of the concrete under repeated loading. The composite element method, as presented in this paper, does not address cyclic loading, and I would like to see the method extended to include this important case.
Another question is the treatment of the interface between the steel tube and the concrete core. The composite element method assumes that the steel and concrete act together as a single composite section, which implies a perfect bond between the two materials. In reality, the bond can be lost under high loads, particularly in the presence of cyclic loading or temperature effects. The loss of bond can significantly affect the behaviour of the member, and this is not captured by the composite element method.
I would also note that the composite element method is based on the assumption of plane sections remaining plane, which is a standard assumption in beam theory. This assumption is valid for members with a reasonable slenderness ratio, but it may not be valid for deep beams or for members with significant shear deformation. In such cases, a more sophisticated analysis method, such as a continuum-based approach, would be required.
Study Insights and Practical Recommendations
This paper makes a valuable contribution to the field of nonlinear analysis of CFST structures by proposing a computationally efficient composite element method. The method is validated against experimental results and is shown to be accurate and fast, making it a practical tool for design and analysis.
My recommendation to fellow engineers is to consider using the composite element method for the nonlinear analysis of CFST structures, particularly for large frames and for parametric studies. The method is well-suited for implementation in a general-purpose finite element program, and it can significantly reduce the computational time required for nonlinear analysis.
However, I would also recommend that the method be used with caution for members with significant shear deformation or for structures subjected to cyclic loading. In such cases, the assumptions of the method may not be valid, and a more sophisticated analysis approach may be required.
In conclusion, this paper presents a significant advance in the analytical modelling of CFST structures. The composite element method is efficient, accurate and practical, and it has the potential to become a standard tool for the nonlinear analysis of CFST structures. I would recommend this paper to researchers and engineers involved in the design and analysis of CFST structures, and I would encourage the authors to extend the method to include cyclic loading and three-dimensional analysis in future work.
Zhuojin Pipe Fitting Co., Ltd