Internally Constrained Square Steel Tube Concrete Columns: Seismic Performance Study Note
Literature Overview and Core Findings
Zhang Tao and colleagues from Zhengzhou University of Aeronautics, Tianjin University, and Central South University address a problem I have seen on many seismic retrofit projects: square steel tube concrete columns have weak confinement of the concrete core, and under seismic loading the steel tube wall at the column base tends to buckle outward. Their solution is a new type of internally constrained column, where welded tie bars are placed inside the tube to restrain the concrete and, indirectly, the steel tube wall. Using ABAQUS finite element analysis with a concrete damaged plasticity model and a cyclic hardening steel model, they built a three-dimensional solid model that captures the interaction between the steel tube, the concrete, and the tie bars. The model was validated against test results and then used to optimise the internal constraint configuration.
The key findings are threefold. First, the tie bar configuration with diagonal ties (拉箍筋, which I interpret as diagonal or cross-tie bars) performs significantly better than circular ring ties in terms of seismic performance. Second, the authors proposed optimal tie bar lengths and volumetric tie ratios for different axial load ratios. Third, increasing the steel ratio and slenderness ratio improves the ultimate bearing capacity, while the axial load ratio has a beneficial effect up to a certain range.
Technical Points and Interpretation
The fundamental issue is the confinement mechanism. In a circular steel tube, the hoop stress in the tube provides uniform confinement to the concrete core. In a square tube, the flat walls are less effective because they tend to bend outward rather than carry hoop stress. This is a well-known problem. The internal tie bars act as a mechanical interlock: they tie the opposite walls together, preventing the wall from buckling outward and simultaneously confining the concrete. The authors' finding that diagonal ties outperform circular ring ties makes sense from a mechanics standpoint: a diagonal tie creates a triangulated load path that resists both wall bending and wall separation, whereas a circular ring tie only resists wall separation.
The parametric study on axial load ratio is particularly interesting. The authors found that increasing the axial load ratio up to a certain range improves the bearing capacity. This is counterintuitive to many engineers who assume that higher axial load always reduces ductility. In composite columns, a moderate axial load pre-compresses the concrete, which improves its confinement and delays cracking. Beyond a certain ratio, however, the concrete enters the softening regime and the column loses ductility. This is a classic trade-off that I have seen in practice: a column designed for a high axial load ratio may have adequate strength but poor ductility, and under a strong earthquake it could fail in a brittle manner.
Comparison of Internal Constraint Configurations
| Constraint Type | Seismic Performance | Ultimate Capacity | Ductility | Practicality |
|---|---|---|---|---|
| No internal constraint | Poor (wall buckling at base) | Baseline | Low | Standard construction |
| Circular ring ties | Moderate | Slightly improved | Moderate | Requires welding rings inside tube |
| Diagonal / cross ties | Best (significantly better) | Improved | High | Requires welding diagonal bars; more labour |
| Longitudinal ties only | Moderate | Improved | Moderate | Easier to install but less effective |
The authors' conclusion that diagonal ties are the best configuration is supported by their finite element results. However, from a fabrication standpoint, I have a practical concern. Welding diagonal ties inside a square steel tube is difficult because the welder has limited access. In a 400 mm x 400 mm tube with 10 mm wall thickness, you can reach inside with a welding torch, but it is not comfortable. The tie bars need to be welded to the tube wall, and the weld quality is critical because the tie bars are under high stress during an earthquake. I would recommend using full penetration welds or at least fillet welds with a minimum leg size of 0.7 times the tie bar diameter. In my experience, the most common defect in this type of fabrication is lack of fusion at the tie bar to wall connection, which can lead to premature failure.
Engineering Practice and Lessons Learned
The proposed internally constrained column is a promising solution for seismic regions. In my experience, the standard solution for square steel tube concrete columns in high seismic zones is to either increase the wall thickness or add external stiffeners. Both increase the steel tonnage and the fabrication cost. The internal tie bar approach is more efficient because it uses the concrete core as part of the load path. However, the fabrication cost is higher due to the additional welding. I have seen projects where the cost of internal ties added 15-20% to the column fabrication cost. Whether this is worth it depends on the seismic demand and the importance of the structure.
The authors' recommendation on the optimal tie bar length and volumetric tie ratio is valuable. They found that the optimal volumetric tie ratio is around 1.5-2.0%, and the optimal tie bar length is such that the tie bars extend from one wall to the opposite wall, with a slight overhang to ensure proper anchorage. This is a practical guideline that I would pass on to any fabricator. I would also recommend that the tie bars be placed at a spacing of no more than 200 mm along the column height, to ensure uniform confinement.
One issue I have with the paper is that it relies solely on finite element analysis. The authors state that their model was validated against test results, but they do not present the test data in detail. For a practicing engineer, it would be more convincing to see actual test results of the proposed column under cyclic loading. I have seen many papers where the finite element model looks good but the actual test fails to confirm the predicted behaviour. The reason is usually that the finite element model does not capture the weld defects, residual stresses, or the actual boundary conditions. I would recommend that the authors conduct a full-scale test of the proposed column before recommending it for widespread use.
Study Insights and Implications
The main insight from this paper is that the confinement of square steel tube concrete columns can be significantly improved by internal ties, and that the tie configuration matters. This is a practical solution that can be implemented with existing welding technology. The cost increase is moderate, but the seismic performance improvement is substantial. For a hospital or a school in a high seismic zone, the additional cost is justified. For a warehouse, it may not be.
The paper also highlights the importance of the axial load ratio. In my experience, many designers specify a high axial load ratio for composite columns to reduce the column size, but they do not check the ductility. This paper provides a quantitative basis for that check. The authors found that the optimal axial load ratio is around 0.5-0.6 for the proposed column. Beyond 0.7, the ductility drops sharply. This is a useful guideline for design.
Reference Value and Outlook
For the industry, this paper provides a new concept for seismic-resistant square steel tube concrete columns. The internal tie bar approach is innovative and practical. The main challenge is fabrication, but this can be overcome with proper welding procedures and quality control. I would recommend that the design community adopt this concept in pilot projects, with full-scale testing to validate the finite element results. The potential benefit is significant: a 20-30% improvement in ductility and energy dissipation capacity, with only a 15-20% increase in fabrication cost.
In the meantime, I would advise engineers to use the existing code provisions for square steel tube concrete columns, but to add a ductility check based on the axial load ratio. If the axial load ratio exceeds 0.7, either increase the wall thickness or add internal ties. This is a conservative but safe approach. The paper gives you the tools to do better, but the tools need validation. I have been in this industry for three decades, and I have learned that a new idea is only as good as the test that proves it. This paper is a step in the right direction, but it is not the last word.
Zhuojin Pipe Fitting Co., Ltd