T-Shaped Square Steel Tube Concrete Composite Special-Shaped Columns: Eccentric Compression Performance and Practical Implications
Overview and Significance of the Study
This study by Li Quan, Zhou Xuejun, Li Guoqiang and their colleagues at Shandong Jianzhu University and Tongji University addresses a structural form that is increasingly common in high-rise residential and office buildings in China: the T-shaped concrete-filled square steel tube (CFSST) composite special-shaped column. The work is supported by the National Key R&D Program and the Shandong Provincial Natural Science Foundation, which signals its relevance to current engineering practice. The paper reports an experimental program of nine specimens with varying length, eccentricity and eccentric direction, tested under eccentric compression to failure.
From my years in the steel pipe business, this topic matters because the square hollow section (SHS) is a workhorse product in structural steel. We ship thousands of tonnes of SHS to fabricators who weld them into composite columns. The question of how these columns behave when the load is not perfectly concentric is not academic — it is the difference between a column that performs as designed and one that fails prematurely on site. The authors have done the industry a service by quantifying this behaviour and comparing it against three design codes.
Test Program and Key Parameters
The test matrix is well conceived. Nine specimens were designed with three lengths (600, 1500 and 1800 mm) and three eccentricities. The key parameters are summarised below.
| Parameter | Values | Remarks |
|---|---|---|
| Specimen length (mm) | 600, 1500, 1800 | Covers short, medium and slender columns |
| Eccentricity (mm) | Varied per specimen | Represents different load paths |
| Eccentric direction | Two orthogonal directions | Tests section asymmetry effect |
| Steel tube section | Square hollow section | T-shaped arrangement of three tubes |
| Concrete grade | Not specified in abstract | Likely C30–C50 for composite columns |
| Loading type | Monotonic eccentric compression | Quasi-static, displacement controlled |
The choice of 600 mm as the shortest length is deliberate — it produces a stocky column that fails by material crushing rather than instability. The 1500 and 1800 mm lengths push into the slender range where second-order effects dominate. This range is exactly what a designer needs to calibrate a slenderness-dependent design formula.
Failure Modes and Load-Deformation Behaviour
The experimental results are clear and consistent with composite column mechanics. The 600 mm specimens failed by strength failure — concrete crushing and local steel tube yielding at the compression face. The 1500 and 1800 mm specimens failed by bending instability, with visible lateral deflection and eventual outward buckling of the steel tube on the tension side.
The load-strain and load-deflection curves show the expected three-stage behaviour: elastic, elastic-plastic transition, and post-peak softening. The authors note that longer specimens exhibit more pronounced bending failure characteristics, which is intuitive but worth stating quantitatively. The eccentricity effect is monotonic — larger eccentricity reduces the axial load capacity, as one would expect from the P-M interaction diagram. The eccentric direction has a relatively small effect, which is a useful finding because it suggests the T-shaped section is not overly sensitive to loading direction, simplifying design.
Comparison with Design Codes
The most practically significant contribution is the comparison of experimental capacities with predictions from three codes: DBJ/T 13-51-2010 (Fujian Provincial Standard), GB 50936-2014 (National Standard for Concrete-Filled Steel Tube Structures), and CECS 254-2012 (Technical Specification for Concrete-Filled Steel Tube Structures). The results are summarised below.
| Code | Prediction Accuracy | Remarks |
|---|---|---|
| DBJ/T 13-51-2010 | Best agreement | Closest to experimental values |
| GB 50936-2014 | Conservative | Under-predicts capacity in some cases |
| CECS 254-2012 | Reasonable | Slightly conservative for slender columns |
The fact that the Fujian provincial standard performs best is not surprising — it was developed with a large database of composite column tests and incorporates a more refined stability factor. The national standard is more conservative, which is safe but may lead to over-design in some cases.
Engineering Practice Implications
From a fabricator's and contractor's perspective, this study reinforces several practical points.
- Welding quality of the T-shaped section is critical. The three square tubes are welded together along their longitudinal seams. If these welds are not full penetration and sound, the composite action between the tubes is compromised, and the column will not behave as a single unit. This is a shop floor issue — the welder must have access to the joint from both sides, and the weld must be inspected by UT or RT.
- Concrete filling quality matters more than the steel tube alone. The composite action depends on intimate contact between the steel and concrete. If the concrete is not properly compacted, or if there is shrinkage-induced voiding at the steel-concrete interface, the load transfer mechanism changes. This is a site issue — the concrete must be placed in lifts, vibrated properly, and the filling height monitored.
- The eccentric direction effect being small is good news for erection. It means the column does not need to be oriented with extreme precision on site. But it does not eliminate the need for proper alignment — the column must still be plumb and the load path must be as designed.
- The design code selection matters. If a designer uses GB 50936-2014, they will get a conservative result. If they use DBJ/T 13-51-2010, they may achieve a more economical section. This is a commercial decision that the fabricator should be aware of when quoting.
Critical Questions and Reflections
The study has some limitations that the reader should note. The number of specimens is small — nine is enough for a preliminary study but not for a full parametric calibration. The concrete grade is not specified in the abstract, which makes it difficult to compare with other studies. The eccentric direction was tested in only two directions, and the interaction with slenderness was not fully explored.
From a practical standpoint, I would ask: what happens when the column is subjected to cyclic loading, as in an earthquake? The eccentric compression test is monotonic, and the seismic behaviour of T-shaped CFSST columns is not addressed here. This is a gap that the industry needs to fill, especially for high-rise buildings in seismic zones.
Another question: what is the effect of the weld between the tubes on the local buckling behaviour? The authors note that the tubes work together, but the weld itself is a stress concentration. If the weld is under-matched or has defects, the local buckling capacity could be reduced. This is a quality control issue that the fabricator must address.
Study Insights and Implications
This study is a solid contribution to the understanding of T-shaped CFSST composite columns under eccentric compression. It confirms that the section has good ductility, that the tubes work together, and that the eccentric compression performance is satisfactory. The comparison with design codes provides useful guidance for practising engineers.
For the steel pipe industry, the key takeaway is that the square hollow section is a versatile and reliable product for composite column applications. The fabrication requirements are well understood, and the quality control measures are standard. The main risk is not in the steel tube itself but in the concrete filling and the welding of the tubes together. As long as these are done properly, the column will perform as designed.
The practical value of this study lies in its confirmation that the design approach in DBJ/T 13-51-2010 is appropriate for T-shaped CFSST columns. This gives fabricators and designers confidence to use this structural form in real projects, knowing that the capacity predictions are reliable and the behaviour is ductile and predictable.
Zhuojin Pipe Fitting Co., Ltd