Ultimate Bearing Capacity of Concrete-Filled Steel Tube Dumbbell-Shaped Eccentrically Loaded Long Columns

Literature Overview

The research by Chen Baochun and Sheng Ye, published in Engineering Mechanics (2008, Vol. 25, No. 12, pp. 98-105), presents an experimental and numerical study on the ultimate bearing capacity of concrete-filled steel tube (CFST) dumbbell-shaped eccentrically loaded long columns. The study involves 16 test specimens with varying slenderness ratios and eccentricity ratios, and the authors analyze the failure modes, load-deformation curves, and ultimate bearing capacity. Finite element models are also developed to simulate the full loading process and ultimate bearing capacity. The study concludes that the slenderness ratio and eccentricity ratio have essentially independent effects on the ultimate bearing capacity, and a stability coefficient formula and eccentricity reduction factor formula are proposed.

Core Technical Points

The dumbbell-shaped cross-section is a distinctive geometry for CFST columns, characterized by two enlarged end sections connected by a narrower middle section. This geometry is often used in bridge pier design to optimize material usage while maintaining adequate stiffness at the critical end regions. The eccentrically loaded condition introduces bending in addition to axial compression, creating a complex stress state that combines flexural and compressive demands. The slenderness ratio (lambda) and eccentricity ratio (eta) are the two primary parameters governing the structural behavior, and the study demonstrates that their effects on ultimate bearing capacity are essentially independent, which allows for the use of multiplicative partial factors to account for their combined influence.

Test Results and Analytical Findings

The experimental results indicate that the CFST dumbbell-shaped eccentrically loaded long columns exhibit overall failure, meaning that the failure initiates and propagates through the entire cross-section rather than at a localized weak point. The ultimate bearing capacity and the tangent stiffness in the elastoplastic stage both decrease with increasing slenderness ratio and eccentricity ratio. The finite element analysis confirms these experimental observations and provides detailed insight into the stress distribution and deformation mechanisms throughout the loading process.

The stability coefficient formula for the eccentrically loaded case can adopt the same formula as for the axially loaded dumbbell-shaped long column, which simplifies the design process. The eccentricity reduction factor is derived through regression fitting of both experimental and finite element data, providing a practical formula for design calculations.

Standards and Design Implications

From a steel pipe and welding perspective, the dumbbell-shaped CFST column requires careful fabrication of the transition sections between the enlarged ends and the narrow middle. These transition regions are critical because they experience high stress concentrations due to the geometric discontinuity. The welding joints at these transitions must be of the highest quality, with full-penetration butt welds inspected using phased array ultrasonic testing (PAUT) or time-of-flight diffraction (TOFD) techniques to detect any planar defects. The steel tubes used for the enlarged sections and the narrow middle section may have different diameters and wall thicknesses, requiring careful selection of welding procedures and consumables to ensure metallurgical compatibility.

The design of dumbbell-shaped CFST columns should comply with relevant standards such as GB 50017 for steel structures, JGJ/T 7 for CFST structures, or Eurocode 4 for composite structures. The proposed stability coefficient and eccentricity reduction factor formulas should be validated against these standards before being adopted in design practice.

Integration with Engineering Practice

In engineering practice, the fabrication of dumbbell-shaped CFST columns involves several critical steps that require careful quality control. The steel tubes for the enlarged sections are typically fabricated from large-diameter seamless pipes or LSAW pipes, while the narrow middle section may use smaller-diameter tubes. The transition sections are fabricated through welding, and the weld quality at these locations is paramount. The concrete infill must be poured in a controlled manner to ensure complete filling of the entire cross-section, including the transition regions, without creating voids or honeycombing. The post-pouring inspection should include ultrasonic testing to verify the integrity of the concrete infill.

For welding engineers, the key challenge is ensuring that the weld joints at the transition sections have adequate strength and toughness to withstand the combined axial and bending stresses. The heat-affected zone (HAZ) of these welds should be examined metallographically to ensure that no brittle microstructures such as martensite or tempered martensite have formed, which could reduce the ductility and fracture resistance of the weld. Post-weld heat treatment (PWHT) may be necessary for thick-walled sections to relieve residual stresses and improve the HAZ microstructure.

Key Questions and Reflections

An important consideration is the interaction between the dumbbell geometry and the eccentric loading condition in terms of local buckling. The narrow middle section, being the weakest region in terms of local stiffness, may be prone to local buckling under high compressive stresses, particularly when the eccentricity ratio is large. The study focuses on overall failure, but local buckling could govern the design in some cases, especially for columns with high slenderness ratios and large wall thickness. Engineers should perform local buckling checks in accordance with relevant codes, and the concrete infill should be considered as providing lateral support to the steel tube against local buckling.

Another question is the effect of construction tolerances on the eccentricity ratio. In practice, the actual eccentricity of the applied load may differ from the design value due to construction tolerances, misalignment of column bases, or imperfect load application. The study assumes precise eccentricity control, but real-world structures must accommodate these uncertainties. Engineers should incorporate appropriate safety factors or sensitivity analyses to account for eccentricity variations.

Study Insights and Implications

This study provides a comprehensive understanding of the ultimate bearing capacity of CFST dumbbell-shaped eccentrically loaded long columns, with both experimental and numerical validation. The finding that the slenderness ratio and eccentricity ratio have independent effects on ultimate bearing capacity is a significant simplification for design purposes, as it allows the use of multiplicative partial factors. The proposed stability coefficient and eccentricity reduction factor formulas offer practical tools for engineers. For steel pipe and welding engineers, the key implication is that the fabrication quality of the transition sections and the integrity of the concrete infill are critical to achieving the predicted structural performance. Strict welding quality control, comprehensive non-destructive testing, and careful concrete pouring procedures are essential to ensure that the as-built column meets the design requirements.