Fitting Equations for Spring Stiffness Coefficients in Rectangular Concrete-Filled Steel Tube Columns
1. Literature Overview and Core Problem
This paper by Xu Na and Fu Xueyi tackles a question that has quietly bothered structural engineers for decades: when you put a steel tube around concrete, how much load does each material actually carry? The authors define three spring stiffness coefficients — the vertical spring stiffness of the steel tube, the rotational spring stiffness of the steel tube, and the vertical spring stiffness of the concrete — and then fit regression equations for each one based on finite element parametric studies.
The motivation is straightforward. When a load transfer member is installed inside the tube to force the steel and concrete to work together, the classic elastic foundation beam theory becomes the analytical tool of choice. But to apply that theory, you need the three spring stiffness values, and solving them by finite element analysis for every new design is tedious, non-intuitive, and frankly impractical for routine engineering work. The paper's contribution is to replace that numerical grind with closed-form fitting equations that account for the coupling effects between parameters.
2. Interpretation of Technical Points
2.1 The Three Spring Stiffness Coefficients
The authors define dimensionless stiffness coefficients rather than absolute stiffness values, which is the right move. It makes the equations scale-independent and applicable across a range of section sizes. The three coefficients are:
| Coefficient | Physical Meaning | Key Influencing Factors |
|---|---|---|
| Steel tube vertical spring stiffness coefficient | Axial load sharing capacity of the steel tube | Wall thickness, section width, concrete strength, load transfer member geometry |
| Steel tube rotational spring stiffness coefficient | Moment resistance contribution of the steel tube | Wall thickness, section aspect ratio, concrete confinement |
| Concrete vertical spring stiffness coefficient | Axial load sharing capacity of the concrete core | Concrete strength, section width, load transfer member geometry |
2.2 The Fitting Equation Methodology
The authors first run parametric finite element analyses to identify the individual influence of each factor on each stiffness coefficient. They then fit simple equations for each factor separately, check for coupling effects between factors, and finally regress a total equation that includes interaction terms. The error analysis confirms that the total fitted equations are accurate enough for engineering use.
The practical significance here is that a designer can now compute the three spring stiffnesses from geometric and material parameters alone, without touching a finite element package. That is a real time-saver, but it also raises a caution flag.
3. Engineering Practice Insights and Reflections
3.1 What the Shop Floor Knows That the Paper Does Not Say
The paper treats the steel tube as a perfect elastic member. In reality, the tube you receive from the mill has residual stresses from forming and welding, especially if it is a cold-formed rectangular hollow section. The corner regions of a cold-formed RHS have significantly higher yield strength than the flat faces due to strain hardening, and the residual stress distribution can be highly non-uniform. These effects shift the actual spring stiffness away from the theoretical value.
For a seamless or ERW tube, the wall thickness tolerance alone can be ±10% per API 5L or GB/T 3091. That translates directly into a ±10% variation in the vertical spring stiffness of the steel tube. The fitting equations in the paper are based on nominal dimensions, so the engineer must remember that the real structure will behave somewhere within a band around the predicted value.
3.2 The Concrete Side of the Equation
The concrete vertical spring stiffness coefficient depends on the concrete strength grade, but the paper does not address shrinkage and creep. In a rectangular concrete-filled steel tube, the concrete shrinks away from the steel wall over time, creating a gap that reduces the composite action. This is a well-documented phenomenon in the literature, and it is one of the reasons why load transfer members are needed in the first place. The spring stiffness coefficients fitted in this paper represent the short-term elastic behaviour, not the long-term service condition.
For a transmission tower or a high-rise column, the long-term behaviour matters. The designer should apply a reduction factor to the concrete spring stiffness to account for shrinkage and creep, or at least verify that the load sharing remains acceptable after the concrete has fully dried and shrunk.
3.3 The Weld Quality Question
The paper assumes the load transfer member is welded to the steel tube wall and that the weld is fully effective. On the shop floor, this is where problems start. The load transfer member is typically a stiffener plate or a shear stud welded inside the tube. If the tube is small, the welder has limited access, and the weld quality can be inconsistent. A lack of fusion or an undercut at the weld toe creates a stress concentration that can initiate a fatigue crack under cyclic loading.
I have seen this failure mode in practice. A rectangular concrete-filled steel tube column in a high-rise building developed cracks at the stiffener-to-tube welds after several years of service. The root cause was incomplete fusion at the weld root, which had been accepted at the time of fabrication because the weld was inspected by visual examination only. The lesson is that internal welds in closed sections must be inspected by ultrasonic testing or radiographic testing, and the inspection must be specified in the fabrication procedure.
4. Key Questions and Reflections
4.1 Is the Fitting Equation Approach Too Simplistic?
The paper's total fitting equations are accurate for the parametric range studied, but they are not universally applicable. The authors studied specific ranges of section width, wall thickness, and concrete strength. Outside those ranges, the equations may extrapolate poorly. The engineer should check the validity range before applying the equations to a non-standard section.
4.2 What About the Steel-Concrete Interface?
The paper does not address the interface condition between the steel tube and the concrete core. In practice, the interface can be either bonded or unbonded, depending on the surface preparation of the tube interior and the concrete placement method. A bonded interface provides higher composite action, while an unbonded interface reduces it. The fitting equations implicitly assume a bonded interface. If the tube interior is greasy or has mill scale, the bond will be poor, and the actual spring stiffness will be lower than predicted.
4.3 The Real Value of This Paper
The paper is a useful contribution because it converts a complex numerical problem into a practical design tool. It saves the designer time and reduces the risk of errors in hand calculation. The coupling effect between parameters is properly addressed, which is more than many papers in this field achieve.
5. Study Insights and Implications
The key insight from this paper is that composite action between steel and concrete in a rectangular tube is governed by three spring stiffnesses, and that these stiffnesses can be estimated from simple equations. This is a practical step forward, but it must be used with engineering judgement.
The manufacturing reality is that the steel tube is not a perfect elastic member, the concrete does not stay perfectly bonded to the steel, and the welds are not always perfect. The fitting equations give a good first estimate, but the designer must apply safety factors and verify the behaviour with testing or detailed analysis for critical applications.
For the transmission tower application discussed in Topic 2, the same principles apply. The Q690 steel tube filled with concrete must have reliable composite action, and the spring stiffness coefficients are the basis for that composite action. The full-scale test in Topic 2 confirms that the design approach works, but it also reveals the stress concentration at the stiffener-to-flange junction, which is a manufacturing issue that the design engineer must address.
6. Reference Value and Outlook
This paper is worth reading for any engineer involved in the design of concrete-filled steel tube columns, whether for buildings, bridges, or transmission towers. The fitting equations are a practical tool that can be used in preliminary design and in verification of finite element results.
The future direction should be to extend the fitting equations to cover long-term effects, including shrinkage and creep of the concrete, and to include the effect of weld quality on the composite action. The manufacturing community should also pay attention to the quality of internal welds in closed sections, because that is where the real failures occur.
In summary, this paper provides a useful analytical tool, but the engineer must remember that the tool is only as good as the assumptions behind it. The steel tube is not perfect, the concrete is not perfect, and the welds are not perfect. Use the equations with care, verify with testing where it matters, and never forget that the structure you build is only as good as the workmanship that goes into it.
Zhuojin Pipe Fitting Co., Ltd