Non-Uniform Wall Thickness Rectangular Concrete-Filled Steel Tube: Flexural Behaviour Under Pure Bending
Introduction to the Problem
This paper by Lu Fangwei, Chen Jianbin, Li Siping and Li Dawang addresses a question that I have often had to answer in the design office: is it worth using non-uniform wall thickness in rectangular concrete-filled steel tube (CFST) members? The authors tested three specimens of non-uniform wall thickness rectangular CFST under pure bending and compared their performance with that of uniform wall thickness specimens with the same steel content. Their conclusion is that the non-uniform wall thickness members have good ductility and later-stage bearing capacity, and their flexural performance is better than that of uniform wall thickness members with the same steel weight.
Now, I have been manufacturing steel tubes for decades, and I can tell you that non-uniform wall thickness is not a free lunch. It costs money to produce, it complicates the welding, and it makes the quality control more difficult. But the paper makes a strong case that the performance benefits may justify the cost in certain applications. Let me break down what the paper actually found and what it means for the shop floor.
Test Setup and Key Findings
The authors tested three non-uniform wall thickness rectangular CFST specimens under pure bending. The non-uniformity was achieved by varying the wall thickness along the cross-section, presumably with thicker walls in the regions of higher stress and thinner walls in the regions of lower stress. This is a classic optimisation approach: put the material where the stress is. The test results showed that these members exhibited good ductility and significant later-stage bearing capacity, meaning that they did not fail suddenly but continued to carry load after yielding. The flexural performance was better than that of uniform wall thickness members with the same steel content.
Let me put this in a table to make the comparison clear.
| Parameter | Non-Uniform Wall Thickness | Uniform Wall Thickness (Same Steel Weight) |
|---|---|---|
| Wall thickness distribution | Varies along cross-section | Constant |
| Steel content | Same | Same |
| Flexural capacity | Higher | Lower |
| Ductility | Good | Good but lower |
| Later-stage bearing capacity | Significant | Less significant |
| Failure mode | Gradual, ductile | More sudden |
| Material utilisation | Better | Worse |
The key insight is that by placing the steel where the bending stress is highest, you get more capacity for the same amount of steel. In a rectangular section under pure bending, the maximum stress is at the top and bottom fibres, and the stress is zero at the neutral axis. So a non-uniform wall thickness with thicker walls at the top and bottom and thinner walls at the sides makes physical sense.
Manufacturing Considerations
Now, from my perspective, the question is not whether the concept works but whether it can be manufactured economically and reliably. Let me walk you through the options.
For a rectangular CFST member, the typical manufacturing route is to form a rectangular hollow section (RHS) from a steel plate or coil, weld the longitudinal seam, and then fill it with concrete. If you want non-uniform wall thickness, you have several options. The first is to use a rolled or formed section with varying thickness, which is difficult to achieve with standard rolling processes. The second is to use two different plate thicknesses and weld them together, which introduces a longitudinal weld at the junction of the two thicknesses. The third is to use a cold-formed section and then reinforce the high-stress regions with additional plates or stiffeners.
Each of these options has its own set of problems. The welded joint between different thicknesses is a classic source of stress concentration and potential fatigue or fracture issues. The weld must be properly transitioned, and the quality control must be rigorous. In my experience, the most practical approach is to use a uniform wall thickness section and then add reinforcement in the high-stress regions, but this increases the fabrication cost and complicates the concrete filling process.
Performance Under Fire and Other Extreme Loads
One of the things that this paper does not address, but which I would want to know, is how these non-uniform wall thickness members perform under fire and other extreme loads. In a fire, the steel tube heats up and the concrete core provides thermal mass and insulation. If the wall thickness is non-uniform, the heat transfer will also be non-uniform, which could create differential thermal expansion and additional stresses. In my experience, this is a potential problem that needs to be investigated before the concept is adopted in fire-resistant structures.
Similarly, under seismic loading, the ductility of the member is critical. The paper shows good ductility, but the cyclic behaviour under reversed loading may be different. I would want to see tests under cyclic loading before I would recommend this concept for seismic applications.
Practical Applications and Design Implications
Despite the manufacturing challenges, there are practical applications where non-uniform wall thickness rectangular CFST members make sense. The most obvious is in long-span beams where the bending moment varies along the span. In such cases, the wall thickness could be varied along the length of the member to match the moment diagram, saving steel where the moment is low and adding steel where the moment is high. This is analogous to the way a plate girder is designed, but with the added benefit of concrete filling.
Another application is in columns with high eccentricity, where the bending moment is significant. By making the wall thicker on the tension side and thinner on the compression side, you can improve the flexural capacity without increasing the steel weight. This is a clever design approach, but it requires careful analysis and a thorough understanding of the composite action between the steel and the concrete.
Reflections and Lessons Learned
Let me share some of the lessons I have learned from my years in the industry that are relevant to this paper. The first lesson is that material efficiency is not always the same as cost efficiency. A non-uniform wall thickness member may use less steel, but if it costs more to fabricate, the overall cost may be higher. You have to look at the whole life cycle cost, not just the material cost.
The second lesson is that composite action is only as good as the bond between the steel and the concrete. If the bond is poor, the member will not behave as a composite, and the benefits of non-uniform wall thickness will be lost. The paper does not discuss the bond behaviour, but in my experience, the bond in a rectangular section is generally good because the flat surfaces provide good contact, but the corners can be problematic.
The third lesson is that ductility is not just about the material; it is about the whole system. A non-uniform wall thickness member may have good ductility in a static test, but under dynamic loading, the behaviour can be different. I would want to see more tests under dynamic and cyclic loading before I would be confident in this concept.
Concluding Thoughts
This paper makes a solid case for the use of non-uniform wall thickness rectangular CFST members in flexural applications. The test results show that the concept works, and the performance benefits are real. However, the manufacturing challenges are significant, and the long-term behaviour under fire, seismic and fatigue loading needs further investigation. From my perspective, the most promising application is in long-span beams where the moment diagram is non-uniform, and the wall thickness can be tailored to match the loading. This is a concept that deserves further research and development, but it is not yet ready for widespread adoption. The paper is a valuable contribution to the field, and I would recommend it to any engineer who is designing CFST structures and looking for ways to improve material efficiency.
Zhuojin Pipe Fitting Co., Ltd