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

Equivalent Constitutive Relations for Concrete-Filled Steel Tubes: A Practical Review

Literature Overview and Industrial Significance

The paper by Yuan Weibin and Jin Weiliang, published in the Journal of Zhejiang University (Engineering Science) in 2004, addresses a fundamental problem that has occupied structural engineers for decades: how do we accurately model the composite behavior of concrete-filled steel tubes (CFST)? The authors propose an equivalent constitutive relationship by simplifying the CFST into a transversely isotropic body of uniform scale, using the principle of strain energy equivalence to derive macroscopic elastic coefficients, and assuming quadratic parabolic variation of moduli after entering the elastoplastic stage. This is not merely an academic exercise. In my thirty-plus years in the steel pipe industry, I have seen countless CFST columns, arches, and bridge piers designed and fabricated. The single most common question from design engineers is: what properties do I use for the composite section? This paper provides a rigorous yet practical answer.

The significance of this work lies in its ability to bridge the gap between detailed micro-mechanical modeling and simplified design formulas. Traditional approaches either treat the steel tube and concrete as separate components with interaction forces, or use empirical formulas that lack theoretical foundation. The equivalent constitutive model offers a middle path: it captures the essential composite action while remaining computationally efficient for large-scale finite element analysis. From my perspective as someone who has supplied thousands of tonnes of steel pipe for CFST applications, I can attest that this type of model is exactly what the industry needs to move from conservative empirical design to more optimized, performance-based design.

Theoretical Approach and Key Assumptions

The authors begin by summarizing existing constitutive models for CFST, noting the limitations of each. They then propose a novel approach: treat the CFST as a single homogenized material with transversely isotropic properties. The key assumptions are:

Assumption Description Practical Implication
Perfect bond between steel and concrete No slip at the interface Requires adequate surface preparation and concrete compaction
Uniform cross-section properties Composite treated as homogeneous Valid for circular and square sections with regular concrete fill
Strain energy equivalence Macroscopic coefficients derived from energy balance Ensures global behavior is captured correctly
Quadratic parabolic moduli in elastoplastic range Simplified nonlinearity representation Adequate for monotonic loading; needs verification for cyclic
Small deformation theory Linear kinematics assumed Valid for serviceability and moderate overload conditions

The strain energy equivalence principle is particularly elegant. By equating the strain energy of the composite section to that of an equivalent homogeneous material, the authors derive macroscopic elastic coefficients that automatically account for the steel-to-concrete area ratio, the elastic moduli of both materials, and the geometric configuration. This approach is fundamentally sound and aligns with the energy-based methods I have seen successfully applied in pipeline design and composite member analysis.

One critical assumption that deserves scrutiny is the perfect bond between steel and concrete. In practice, achieving perfect bond is challenging. The concrete shrinks as it cures, and the steel tube may expand or contract due to temperature variations. In my experience, the bond quality depends heavily on the internal surface condition of the pipe, the concrete mix design, and the curing regime. For pipes supplied with internal ribs or shear connectors, the bond is more reliable. For smooth pipes, the bond may degrade over time, especially under cyclic loading. The authors acknowledge this limitation but do not fully address it. This is a point I will return to in the critical reflections section.

Interpretation of the Equivalent Model

The derived macroscopic elastic coefficients include the equivalent elastic modulus in the longitudinal direction (E1), the equivalent elastic modulus in the transverse direction (E2), and the equivalent shear modulus (G12). For a circular CFST, the longitudinal modulus is essentially a weighted average of the steel and concrete moduli based on area fractions. The transverse modulus, however, is more complex because it involves the interaction between the steel tube and the concrete core under lateral loading.

The authors then extend the model to the elastoplastic range by assuming that the moduli vary as quadratic parabolas of the strain. This is a reasonable approximation for monotonic loading, but I have concerns about its applicability under cyclic or seismic loading. In seismic design, the unloading and reloading behavior is critical, and a quadratic parabola model may not capture the Bauschinger effect or the stiffness degradation that occurs with increasing plastic strain. From my experience with CFST bridge piers in seismic zones, the hysteretic behavior is complex and requires more sophisticated models.

The parametric study in the paper examines the effects of material strength variation and steel ratio on the equivalent moduli. The results show that increasing the steel ratio significantly increases the longitudinal modulus, as expected. However, the effect on the transverse modulus is less pronounced. This is consistent with physical intuition: the steel tube provides strong confinement in the transverse direction, but the concrete core dominates the transverse stiffness. The authors also compare their results with previous research and find good agreement, validating the model's theoretical foundation.

Practical Implications for Manufacturing and Design

From a manufacturing perspective, this paper has direct implications for how we specify steel pipe for CFST applications. The steel ratio, which is a key parameter in the equivalent model, is determined by the pipe wall thickness and diameter. For a given diameter, increasing the wall thickness increases the steel ratio, which enhances the composite stiffness and strength. However, there is a practical limit: excessively thick walls are difficult to form, especially for seamless pipe, and may not be economical. In my experience, the optimal steel ratio for CFST columns is typically in the range of 4% to 10%, depending on the application and loading conditions.

The paper also highlights the importance of concrete strength. For high-strength concrete (C60 and above), the concrete modulus is higher, which increases the composite stiffness. However, high-strength concrete is more brittle and may exhibit less ductility in the composite section. This is a critical consideration for seismic applications. In my experience with CFST columns in high-rise buildings, we often use C60 or higher concrete, but we must carefully design the steel tube to provide adequate confinement and ductility.

From a delivery perspective, the equivalent model allows designers to specify a single set of material properties for the composite section, simplifying the design process. This is particularly valuable for international projects where the design team may be in one country and the fabrication in another. Clear specification of the equivalent properties reduces the risk of miscommunication and ensures that the fabricated section meets the design intent.

Critical Reflections and Lessons from the Field

While the paper provides a solid theoretical foundation, I have several concerns based on my field experience:

  1. Bond degradation over time: The perfect bond assumption may not hold in practice, especially for pipes with smooth internal surfaces. I have seen cases where CFST columns exhibited reduced stiffness after several years of service due to bond degradation caused by concrete shrinkage and creep. This is particularly problematic in dry, low-humidity environments where concrete shrinkage is more pronounced.
  2. Temperature effects: The model does not account for temperature gradients across the section. In fire scenarios, the steel tube and concrete core expand at different rates, creating thermal stresses that can cause bond failure and spalling. I have witnessed the aftermath of fires in CFST structures, and the failure modes are often more severe than predicted by models that assume perfect bond.
  3. Cyclic loading behavior: The quadratic parabola model for the elastoplastic range is not suitable for seismic analysis. In my experience with CFST bridge piers subjected to earthquake loading, the hysteretic behavior exhibits significant pinching and stiffness degradation that is not captured by this model. More sophisticated models, such as those based on fiber elements or plasticity theory, are needed for accurate seismic assessment.
  4. Square and rectangular sections: The paper focuses on circular CFST, but many applications use square or rectangular sections. The equivalent model may need modification for these geometries, as the confinement effect is less uniform in square sections. In my experience, square CFST columns are more susceptible to local buckling of the steel tube at the corners, which can lead to premature failure.

Summary and Recommendations

This paper makes a valuable contribution by providing a theoretically sound and practically usable equivalent constitutive model for CFST. The strain energy equivalence approach is elegant and captures the essential composite behavior. The parametric study provides useful insights into the effects of steel ratio and material strength on the composite properties. However, the model has limitations that must be acknowledged in practical applications. For monotonic loading and serviceability design, the model is adequate. For seismic design and long-term performance assessment, more sophisticated models are needed.

Based on my experience, I recommend the following for practicing engineers:

  1. Use the equivalent model for preliminary design and serviceability checks, but verify with more detailed analysis for critical members.
  2. Ensure that the steel pipe is supplied with adequate surface preparation to promote bond with the concrete core.
  3. Consider temperature effects and fire scenarios in the design, as the equivalent model does not account for these.
  4. For seismic applications, use fiber element models or other advanced methods that capture the nonlinear cyclic behavior.
  5. Validate the model against experimental data for the specific steel and concrete grades being used.

The paper lays a solid foundation for further research and practical application. I would encourage the authors to extend the model to account for bond degradation, temperature effects, and cyclic loading, as these are the areas where the current model falls short of real-world requirements.