Eccentric Compression Failure Modes of Hollow CFST Short Columns: Experimental Investigation

Literature Overview

This experimental study by Liu Xiaping, Tang Shu, Yang Zuoyong, Cai Kahong, and Liu Airong from Guangzhou University, published in the Journal of Guangzhou University (Natural Science Edition) in 2010 (Vol. 9, No. 3, pp. 40–44), investigates the failure behavior of short concrete-filled steel tube (CFST) columns with internal voids (hollow spaces between the steel tube and concrete core) under eccentric compression. The research was supported by multiple funding sources including the National Natural Science Foundation (5097806), Guangdong Provincial Natural Science Foundation (9151065004000002), and Guangzhou Construction Science and Technology Program (200617).

Experimental Configuration and Parameters

The test specimens are short columns with specific geometric and material parameters designed to simulate the condition of incomplete concrete filling in CFST members—a common construction defect in practice.

The void ratio (脱空率) of 0–1.24% represents the cross-sectional area of the hollow space relative to the total cross-sectional area. While this may appear small, even minor voids can significantly affect structural performance, particularly under eccentric loading where bending stresses amplify local effects.

Failure Mode Analysis

The study identifies two distinct failure mechanisms depending on the eccentricity direction relative to the void location:

Case 1: Eccentric loading on the void side

When the eccentric load is applied toward the void, the void-side steel tube wall experiences local inward deformation (denting). Two sub-cases emerge:

• Small void ratio: The void-side steel tube wall dents inward until it contacts the concrete core. This local denting creates a geometric defect in the column cross-section, reducing the effective confining pressure and ultimately decreasing bearing capacity.
• Large void ratio: The void-side steel tube wall does not dent through to the concrete core. Instead, the concrete in the void region fails due to loss of lateral confinement, causing overall column capacity reduction.

Case 2: Eccentric loading on the non-void side

When the eccentric load is applied away from the void, no inward denting of the void-side steel tube wall is observed. The failure mechanism differs fundamentally from Case 1, with the column behaving more like a conventional CFST member with a minor cross-sectional defect.

Engineering Significance for Pipe Manufacturing and Quality Control

This research has profound implications for steel pipe quality control and CFST construction practices:

1. Concrete filling quality: The study demonstrates that even small voids (as low as 0.5% of cross-sectional area) can significantly reduce bearing capacity under eccentric compression. This reinforces the critical importance of complete concrete filling during CFST construction.
2. Steel tube straightness and dimensional accuracy: Voids often develop when the steel tube has local ovality or dimensional deviations that prevent complete concrete contact. For ERW or HFW welded pipes, maintaining strict dimensional tolerances (particularly for ovality) is essential to prevent void formation.
3. Welding quality impact: Longitudinal weld seams in steel tubes can create local stiffness variations that influence void formation. A slightly convex weld bead may create a local gap between the tube wall and concrete core, initiating a void that grows during concrete placement.
4. NDT methodology: The study suggests that void detection is critical for CFST quality assurance. While conventional NDT methods (UT, RT) may not directly detect internal voids in CFST members, indirect methods such as impact echo testing or ultrasonic velocity measurements can identify regions of poor concrete-to-steel bond.

Defect Classification and Countermeasures

Practical Recommendations for Construction

• Implement strict quality control during concrete pouring to ensure complete filling of the steel tube interior, using internal vibrators or post-pouring inspection methods.
• For steel pipes supplied for CFST applications, specify tight dimensional tolerances for ovality and straightness to minimize void formation risk.
• When welding CFST columns in the field, ensure that the longitudinal seam is flush or slightly recessed to avoid creating local gaps that promote void formation.
• For existing structures, consider non-destructive evaluation methods to assess the condition of CFST members suspected of having internal voids.

Study Insights and Reflections

The research by Liu and colleagues provides compelling evidence that voids in CFST members, even at seemingly negligible volumes, can have disproportionate effects on structural performance under eccentric loading. The asymmetric response—where void-side eccentric loading causes significantly more damage than non-void-side eccentric loading—reveals the importance of load path continuity in composite members.

From a materials science perspective, the void creates a discontinuity in the composite action between steel and concrete. The confining pressure that normally enhances concrete strength (the well-known confinement effect in CFST members) is locally eliminated at the void, creating a weak zone that governs failure. This finding underscores the fundamental principle that CFST performance depends on continuous composite action, and any disruption to this interaction compromises the structural system.

For engineers involved in steel pipe supply chains, this research reinforces the need to specify quality requirements that address not only the mechanical properties of the steel pipe but also its geometric accuracy and surface condition, as these factors directly influence the quality of the final CFST composite member. The study serves as a valuable reminder that in composite structures, the performance of the final assembly is only as good as the weakest interface, and quality control must extend beyond individual component specifications to encompass the complete assembly process.