ZHUOJIN-LOGOZhuojin Pipe Fitting Co., Ltd
Zhuojin Pipe Fitting Co., Ltd
STEEL PIPE · FITTING · WELDING  |  NOTES FROM A 30-YEAR INDUSTRY VETERAN

Regression Analysis of Flexural Stiffness Reduction Coefficient for Concrete-Filled Steel Tube Members

Literature Overview and Why This Paper Matters

The paper by Wu Naisen and colleagues from Northwestern Polytechnical University addresses a fundamental question in the design of concrete-filled steel tube (CFST) members: how much does the concrete contribute to the overall flexural stiffness? The authors conducted bending tests on five rectangular CFST simply supported beams and combined their results with published data to perform a linear regression analysis on the stiffness reduction coefficient used in the superposition theory. The key finding is that the concrete contribution to the overall stiffness is better in rectangular and circular CFST members than in square CFST members. This is a subtle but important distinction that has direct consequences for design. In my experience, CFST columns are used in high-rise buildings, bridge piers, and industrial structures where the stiffness determines the drift and the dynamic response. A 10% error in stiffness can mean a 10% error in predicted lateral deflection, which can be the difference between a code-compliant design and one that needs expensive retrofitting.

Technical Interpretation of the Stiffness Reduction Coefficient

The Superposition Theory and Its Limitations

The standard approach to calculating the flexural stiffness of a CFST member is to add the stiffness of the steel tube and the stiffness of the concrete core, each multiplied by a reduction coefficient. The steel tube is assumed to be fully effective, but the concrete is not, because of cracking, creep, and the imperfect bond between the steel and the concrete. The reduction coefficient accounts for these effects. The paper uses a linear regression approach to determine the value of this coefficient from experimental data. The regression shows that the coefficient is not a constant but depends on the cross-section shape and the loading conditions.

Cross-Section Shape Concrete Contribution to Stiffness Practical Implication
Rectangular Highest Best for beam-columns with significant bending in one direction
Circular High Best for columns with biaxial bending and high axial load
Square Lowest Use with caution; the corner regions are less effective

Why Square Sections Perform Differently

The paper's finding that square sections have a lower concrete contribution is intuitive to anyone who has looked at a square CFST section. In a square section, the concrete at the corners is far from the neutral axis, and the steel tube at the corners is highly stressed. The confinement effect is also less uniform than in a circular section. In a circular section, the steel tube provides uniform confinement to the concrete core, which improves the concrete's compressive strength and ductility. In a square section, the confinement is concentrated at the corners, and the flat sides of the tube are more prone to local buckling. This local buckling reduces the effective stiffness of the steel tube and, consequently, the overall stiffness of the member.

Integration with Engineering Practice and Standards

Comparison of Common Design Codes

The paper discusses the applicability of several common codes, and this is where the practical value lies. The design engineer must know which code to use and what safety margin to apply.

Code / Standard Approach to Stiffness Range of Applicability Practical Note
GB 50936 Uses a unified theory with a combined stiffness General use in China The reduction coefficient is embedded in the formula
AISC 360 Uses the transformed section method with a modular ratio Common in North America The modular ratio accounts for concrete creep and cracking
Eurocode 4 Uses a similar superposition approach with partial factors Common in Europe The partial factors are more conservative for square sections
AIJ (Japan) Uses a simplified stiffness model Used in high-rise construction The model is calibrated for circular sections

The paper's regression results suggest that the current codes may be overly conservative for rectangular sections and insufficiently conservative for square sections. This is a significant finding because it means that a square CFST column designed to a code that assumes a high concrete contribution may be under-stiff, leading to excessive drift under lateral loads.

Real-World Case: A Bridge Pier Retrofit

I recall a project where a bridge pier was designed as a square CFST column, and the stiffness was calculated using a code that assumed a high concrete contribution. During the seismic retrofit, the engineer discovered that the actual stiffness was lower than predicted, and the pier had a higher fundamental period than expected. This meant that the seismic demand was different from what was assumed in the design. The retrofit had to be modified to add additional damping. The lesson is that the stiffness reduction coefficient is not a minor detail; it is a first-order design parameter. The paper's regression analysis provides a more accurate basis for determining this coefficient, and I would recommend that design engineers use it as a check against the code values.

Key Questions and Reflections

The paper raises an important question that it does not fully answer: what is the effect of the steel tube wall thickness on the reduction coefficient? The paper's regression is based on a limited set of test data, and the wall thickness varies across the specimens. In practice, a thin-walled tube is more prone to local buckling, which would reduce the stiffness further. The paper also does not address the effect of long-term loading. Concrete creeps under sustained load, and this creep reduces the effective stiffness over time. A design that is adequate for short-term loading may be inadequate for long-term service. The paper's regression is a step in the right direction, but more research is needed on these parameters.

Study Insights and Implications for the Industry

This paper is a valuable contribution to the design of CFST members. It provides a quantitative basis for the stiffness reduction coefficient, and it highlights the difference between rectangular, circular, and square sections. For the practicing engineer, the message is clear: do not assume that the concrete contribution to stiffness is the same for all cross-section shapes. The shape matters, and the reduction coefficient must be chosen accordingly. I would recommend that design engineers review the test data in this paper and compare it with the code values they are using. If the code is overly conservative for a rectangular section, there may be an opportunity to reduce the section size and save material. If the code is insufficiently conservative for a square section, the design may need to be revised. The paper is a reminder that design codes are not infallible; they are based on a limited set of test data, and the engineer must use judgment. I have seen too many structures where the stiffness was overestimated, leading to excessive vibration and occupant discomfort. The regression analysis in this paper, if used properly, can prevent that problem. I would encourage the authors to extend their work to include more test data and to address the effects of wall thickness and long-term loading, which are the next logical steps.