Experimental Investigation of Flexural Mechanical Properties of Rectangular Concrete-Filled Steel Tube Members

Literature Overview

This paper by Yang Youfu and Han Linhai, published in Journal of Earthquake Engineering and Engineering Vibration (2001, Vol. 21, No. 3), presents a systematic experimental study on the flexural behavior of rectangular concrete-filled steel tube (CFST) members. The research was funded by the Fok Ying Tung Education Foundation (0501064). Han Linhai is a renowned researcher in the field of CFST structures, and this paper represents a foundational contribution to the understanding of rectangular CFST bending behavior. The study tested 8 rectangular CFST pure bending specimens and compared the results with design provisions from four major design codes.

Technical Significance of Rectangular CFST Members

Rectangular CFST members are widely used in building structures, bridges, and industrial facilities because of their architectural flexibility and efficient use of space. Unlike circular CFST members, rectangular sections can be easily connected to beams and columns at right angles, making them suitable for multi-story building frames. However, the flexural behavior of rectangular CFST members is more complex than that of circular sections due to:

• Corner effects: The corners of rectangular sections experience triaxial stress states that enhance concrete confinement but also create stress concentrations in the steel tube.
• Plate buckling: The flat plates between corners are susceptible to local buckling under flexural compression, which can reduce the effective section modulus.
• Anisotropic behavior: The flexural stiffness and strength differ between the strong axis and weak axis, requiring careful consideration in design.

Experimental Program

Eight rectangular CFST pure bending specimens were tested under four-point bending. The specimens varied in key geometric and material parameters:

The pure bending test configuration ensures that the specimens are loaded in a region of constant bending moment, allowing clear observation of the flexural failure mechanism without interference from shear effects.

Key Experimental Findings

High flexural capacity: The rectangular CFST members exhibited significantly higher flexural capacity compared to empty steel tubes of the same section. The enhancement is attributed to the composite action between the steel tube and core concrete, where the concrete contributes directly to flexural resistance and the steel tube confines the concrete, enabling it to sustain higher compressive stresses.

High flexural stiffness: The flexural stiffness (EI) of the CFST members was substantially higher than that of equivalent empty steel tubes. This is critical for serviceability design, as it means that CFST members can achieve smaller deflections under the same loading, potentially allowing longer spans or lighter sections.

Good curvature ductility: The specimens demonstrated good curvature ductility, meaning they could sustain large rotations before failure. This is essential for seismic design, as ductile members can dissipate energy through inelastic deformation. The ductility was measured as the ratio of curvature at maximum load to curvature at yield, and values of 3–5 were typical for the tested specimens.

Failure modes: The primary failure modes observed were:

• Local buckling of the compressed flange plate, initiating at the mid-span region of constant moment.
• Concrete crushing at the compressed edge of the section, following steel tube buckling.
• In some specimens, corner cracking of the concrete, indicating the triaxial stress state at the corners.

Design Code Comparison

One of the most valuable aspects of this paper is the systematic comparison of experimental results with design provisions from four major codes:

The comparison revealed several important observations:

1. EC4 tends to be conservative: The European code's approach to flexural design of composite members is based on partial factor methods that result in conservative predictions, particularly for rectangular CFST sections with high steel ratios.
2. LRFD shows scatter: The American code's predictions exhibited more scatter, with some sections showing unconservative predictions. This may be related to the code's simplified treatment of local buckling effects in composite sections.
3. AIJ provides the best overall agreement: The Japanese code's predictions were generally closest to the experimental results, suggesting that its approach to composite flexural design is well-calibrated for rectangular CFST sections.
4. Chinese code needs refinement: The Chinese code (GJB) showed conservative predictions for low steel ratio sections but was less accurate for thin-walled sections, where local buckling effects are more pronounced.

Engineering Practice Implications

1. Code selection matters: For international projects involving rectangular CFST members, the choice of design code can significantly affect the predicted capacity and stiffness. Engineers should be aware of the differences between codes and select the most appropriate one based on project location and regulatory requirements.
2. Local buckling is the governing failure mode: The experimental results confirm that local buckling of the compressed flange plate is the primary failure mechanism in rectangular CFST flexural members. Design codes should adequately account for this failure mode, particularly for thin-walled sections where the buckling stress may be lower than the material yield strength.
3. Stiffness prediction is critical for serviceability: The underestimation of flexural stiffness by some codes can lead to overly conservative serviceability design, resulting in unnecessarily heavy sections. Accurate stiffness prediction is essential for optimizing the structural design of CFST members.
4. Steel ratio is a key parameter: The steel ratio (ratio of steel area to total cross-sectional area) significantly influences both capacity and stiffness predictions. Codes should provide clear guidance on the applicability range of their provisions with respect to steel ratio.

Key Questions and Reflections

The study focuses on pure bending, which is an idealized loading condition. In practice, CFST members are subjected to combined bending and shear, and sometimes combined bending and axial force. The interaction between these load components may alter the failure mode and reduce the effective flexural capacity. Future research should investigate the flexural behavior of rectangular CFST members under combined loading conditions, particularly bending-shear interaction and bending-axial force interaction.

The study also does not address the effect of residual stresses from the steel tube manufacturing process on the flexural behavior. Residual stresses can reduce the effective section modulus and alter the buckling behavior of the compressed flange plate. For hot-rolled rectangular tubes, residual stresses are relatively well-characterized, but for cold-formed rectangular tubes, the residual stress distribution may be more complex and require dedicated investigation.

The comparison with design codes is limited to elastic and ultimate capacity predictions. The codes' provisions for ductility, which is critical for seismic design, are not explicitly evaluated. The experimental ductility data should be compared with the ductility requirements of seismic design codes to assess the adequacy of code provisions for seismic applications.

Study Insights and Reference Value

This paper provides a comprehensive experimental database for the flexural behavior of rectangular CFST members, which is essential for the calibration and validation of design codes. The systematic comparison with four major codes is particularly valuable for international engineers who need to understand the differences between code approaches. The key takeaway is that rectangular CFST members offer excellent flexural performance—high capacity, high stiffness, and good ductility—but the design code selection significantly affects the predicted performance. Engineers should carefully evaluate the code provisions against experimental data for the specific section geometry and material properties of their project. The paper also highlights the need for continued research on combined loading behavior and the effect of manufacturing-induced residual stresses on flexural performance.