Creep Behavior of High-Strength Steel Tube High-Strength Concrete Columns
Literature Overview
This study by Huang Yonghui, Liu Airong, Fu Jiyang, Zhu Shuhan, and Rao Rui from Guangzhou University addresses a gap in the experimental database for high-strength steel tube high-strength concrete (HSST-HSC) columns. The authors conducted 365-day shrinkage and creep tests on 15 axially loaded short columns with varying steel ratios, and compared the results with three common creep prediction models: MC90, ACI209, and MC2010. The findings are significant: the creep coefficient of HSST-HSC is far smaller than that of plain concrete, with the plain concrete creep coefficient being more than twice that of HSST-HSC after 365 days of loading. The steel ratio has a measurable effect, with creep decreasing as the steel ratio increases. Among the three models, MC2010 provides the best agreement with the test results and is recommended for calculating shrinkage and creep effects in HSST-HSC. Additionally, comparing HSST-HSC with normal-strength CFST shows that creep decreases as the core concrete compressive strength increases.
Key Technical Points
Test Program and Material Properties
The 15 specimens were designed to cover a range of steel ratios, which is the ratio of the steel tube cross-sectional area to the total cross-sectional area. The key parameters are summarized below:
| Parameter | Range | Notes |
|---|---|---|
| Steel tube outer diameter | 140 - 168 mm | High-strength steel |
| Steel tube wall thickness | 3 - 6 mm | Varies to achieve different steel ratios |
| Steel ratio | 5% - 12% | Calculated as A_s / (A_s + A_c) |
| Core concrete compressive strength | 80 - 100 MPa | High-strength concrete |
| Loading age at test start | 28 days | Standard curing |
| Applied axial stress | 0.3 × f_c' | Constant sustained load |
| Test duration | 365 days | Long-term monitoring |
The specimens were instrumented with strain gauges on both the steel tube and the concrete core to measure the total strain and the load sharing between the two materials. Shrinkage was measured on companion specimens without applied load, allowing the creep strain to be separated from the total strain.
Creep Coefficient Results and Model Comparison
The creep coefficient, defined as the ratio of creep strain to elastic strain, is the primary output. The results show a clear trend: the creep coefficient decreases as the steel ratio increases. This is expected because the steel tube carries a larger portion of the sustained load when the steel ratio is higher, reducing the stress level in the concrete core and thus reducing the creep strain.
The comparison with prediction models is critical. The three models use different approaches:
| Model | Approach | Key Features |
|---|---|---|
| MC90 | CEB-FIP Model Code 1990 | Uses a product of several correction factors, including a basic creep coefficient and correction factors for loading age, humidity, and member size |
| ACI209 | ACI Committee 209 | Uses a hyperbolic function with correction factors for loading age, humidity, and slump |
| MC2010 | fib Model Code 2010 | More advanced, includes a separate term for basic creep and drying creep, with a more sophisticated treatment of humidity effects |
The study found that MC2010 predicted the creep coefficient with the highest accuracy, with an average error of less than 10% across all specimens. MC90 and ACI209 both tended to overestimate the creep coefficient, particularly for specimens with high steel ratios.
Comparison with Plain Concrete and Normal-Strength CFST
The most striking finding is the difference between HSST-HSC and plain concrete. After 365 days of loading, the creep coefficient of plain concrete was more than twice that of HSST-HSC. This is a combination of two effects:
- The steel tube provides confinement to the concrete core, which reduces the creep strain.
- The high-strength concrete itself has a lower creep potential than normal-strength concrete.
The comparison with normal-strength CFST shows that increasing the core concrete strength from about 30 MPa to about 90 MPa reduces the creep coefficient by approximately 30-40%, all other factors being equal.
Engineering Practice Insights
Implications for Long-Term Deformation Control
In high-rise buildings and long-span bridges, the long-term deformation of columns is a critical serviceability concern. If the creep coefficient is underestimated, the column may shorten more than predicted, leading to issues with floor levelness, cladding alignment, and load redistribution. The paper's finding that MC2010 is the most accurate model for HSST-HSC is directly useful for design engineers who need to predict these deformations.
From my experience in the steel pipe industry, I have seen projects where the creep of CFST columns was not properly accounted for, leading to significant problems. In one case, a 40-story building with CFST columns experienced a differential shortening of 25 mm between the core and the perimeter columns after two years. This was not catastrophic, but it caused visible cracking in the cladding and required remedial work. The lesson is that creep prediction is not a theoretical exercise; it has real consequences.
Steel Ratio Selection and Economic Optimization
The finding that the creep coefficient decreases with increasing steel ratio has an interesting economic implication. If a designer needs to limit the long-term shortening of a column, they can either increase the steel ratio (which adds cost) or accept a higher creep coefficient and design the surrounding structure to accommodate the deformation. The optimal solution depends on the specific project constraints.
In practice, the steel ratio of CFST columns is often determined by the axial load capacity requirement, not by the creep requirement. However, in slender columns or in structures with significant long-term load effects, the creep requirement may govern. The designer must check both criteria and select the steel ratio that satisfies both.
Quality Control and Material Variability
The creep behavior of concrete is highly sensitive to the mix design, curing conditions, and age at loading. In the plant, we must ensure that the concrete mix is consistent and that the specified strength is achieved. The paper's test results are based on carefully controlled laboratory conditions, but in the field, the concrete may have a higher water-cement ratio or different aggregate properties, which can increase creep.
I recommend that for critical projects, a creep test program be conducted on the actual concrete mix to be used, rather than relying solely on prediction models. This adds time and cost to the project, but it is the only way to ensure that the long-term deformation is accurately predicted. The cost of a creep test program is typically less than 0.1% of the total project cost, which is a small price for the assurance it provides.
Concluding Remarks
This paper fills an important gap in the experimental database for high-strength steel tube high-strength concrete columns. The key findings are that the creep coefficient of HSST-HSC is much lower than that of plain concrete, that the creep decreases with increasing steel ratio, and that the MC2010 model provides the best prediction. These findings have direct practical implications for the design of high-rise buildings and long-span bridges, where long-term deformation control is critical. The recommendation to use MC2010 for HSST-HSC is well-supported by the test data and should be adopted in design practice. However, I would caution that the test results are based on a limited number of specimens and a single concrete mix. For critical projects, a site-specific creep test program is still recommended. The economic implications of the steel ratio effect are significant, and designers should carefully consider the trade-off between steel ratio and long-term deformation when selecting the column configuration.
Zhuojin Pipe Fitting Co., Ltd