ZHUOJIN-LOGOZhuojin Pipe Fitting Co., Ltd
Zhuojin Pipe Fitting Co., Ltd
STEEL PIPE · FITTING · WELDING  |  NOTES FROM A 30-YEAR INDUSTRY VETERAN

Creep Effect Analysis of CFST Arch Bridges

Introduction and Context

Concrete-filled steel tube (CFST) arch bridges have become increasingly popular for long-span applications due to their high load-carrying capacity and excellent seismic performance. However, one of the most challenging aspects of designing these bridges is the time-dependent behaviour of the concrete core. Concrete creeps under sustained load, and this creep causes a redistribution of stress between the steel tube and the concrete core. Over time, the creep can lead to significant changes in the bridge geometry, including increased deflections and changes in the internal force distribution. This paper by Li Shengyong and colleagues addresses the creep problem using the age-adjusted effective modulus method and compares three commonly used creep coefficient models: CEB-FIP 78, CEB-FIP 90 and ACI 209R. The authors then apply the most suitable model to a 380 m span railway arch bridge, providing valuable insights for the design of long-span CFST arch bridges.

Age-Adjusted Effective Modulus Method

Theoretical Background

The age-adjusted effective modulus method (AEMM) is a well-established approach for analysing the time-dependent behaviour of concrete structures. The method accounts for the fact that the concrete is loaded at different ages and that the creep deformation is not linear with respect to the stress history. The key parameter in the method is the age-adjusted effective modulus, which is calculated using the creep coefficient and the aging coefficient.

Derivation of the Creep Coefficient Formula

The authors derive a formula for the creep coefficient of a CFST axially loaded member using the AEMM. The derivation considers the interaction between the steel tube and the concrete core, including the effects of confinement and the different elastic moduli of the two materials. The resulting formula is expressed in terms of the creep coefficient of the concrete, the modular ratio between steel and concrete, and the confinement factor.

Validation Against Test Data

The authors validate the derived formula by comparing the predicted creep strains with experimental data from published tests. The comparison shows good agreement, confirming the validity of the formula for axially loaded CFST members. This is an important result because it provides a simple and reliable method for predicting the creep behaviour of CFST members without the need for complex finite element analysis.

Comparison of Creep Coefficient Models

Model Description

Three creep coefficient models are compared in the study:

Model Origin Key Features
CEB-FIP 78 Comité Euro-International du Béton Based on a hyperbolic creep function; widely used in European practice
CEB-FIP 90 Comité Euro-International du Béton Improved version with better consideration of relative humidity and member size
ACI 209R American Concrete Institute Based on a power function; widely used in North American practice

Comparison Results

The comparison shows that the ACI 209R model provides the most accurate prediction of the creep behaviour of CFST members when compared with the experimental data. The CEB-FIP 78 model tends to overestimate the creep, while the CEB-FIP 90 model underestimates it. The ACI 209R model captures the long-term creep behaviour more accurately, particularly for the first few years after loading.

Engineering Significance

The choice of creep model has a significant impact on the design of long-span CFST arch bridges. If the creep is overestimated, the designer will specify too much prestressing or too high a camber, leading to unnecessary costs. If the creep is underestimated, the bridge may experience excessive deflections and cracking. The finding that ACI 209R is the most suitable model for CFST members is valuable because it provides a clear recommendation for designers working on such bridges.

Case Study: 380 m Span Railway Arch Bridge

Bridge Description

The authors apply the ACI 209R model to a 380 m span railway arch bridge, which is one of the longest CFST arch bridges in the world. The bridge has a CFST arch with a rise-to-span ratio of approximately 1/5, and the arch is filled with high-strength concrete. The bridge is designed for double-track railway loading.

Creep Analysis Results

The creep analysis shows that the creep deformation of the arch is significant, with a maximum creep deflection of approximately 150 mm at the crown after 10 years. The creep also causes a redistribution of stress between the steel tube and the concrete core, with the steel tube carrying a larger share of the load over time. This is because the concrete creeps and sheds load to the steel tube, which has a higher elastic modulus and does not creep.

Comparison with CEB-FIP 78 Model

The authors compare the results obtained with the ACI 209R model with those obtained using the CEB-FIP 78 model. The comparison shows that the CEB-FIP 78 model predicts a creep deflection that is approximately 20% higher than the ACI 209R model. This difference is significant for a bridge of this span, as it affects the camber and the prestressing force required to compensate for the creep.

Practical Implications for Design and Construction

Camber and Prestressing

The creep analysis results have a direct impact on the design of the bridge camber and the prestressing force. The designer must account for the creep deflection when setting the initial camber of the arch. If the creep is underestimated, the bridge will sag below the design profile after a few years, which can affect the track alignment and the ride quality of the railway. The ACI 209R model provides a more realistic estimate of the creep, which allows the designer to set the camber more accurately.

Construction Sequence

The construction sequence of a CFST arch bridge also affects the creep behaviour. If the concrete is filled in stages, the age of the concrete at the time of loading will vary along the arch, which affects the creep deformation. The AEMM can be used to analyse the creep during the construction sequence, allowing the contractor to optimise the filling sequence to minimise the creep effects.

Quality Control of Concrete

The creep behaviour of the concrete depends on its mix design, particularly the water-cement ratio and the aggregate type. For a long-span CFST arch bridge, the concrete must have a low creep coefficient to minimise the long-term deformation. This requires careful mix design and quality control during production. In my experience, the use of high-performance concrete with a water-cement ratio of 0.35 or lower and the addition of silica fume can significantly reduce the creep coefficient.

Lessons Learned and Recommendations

What the Industry Should Take Away

The paper provides a clear recommendation for the creep model to be used in the design of CFST arch bridges. The ACI 209R model is more accurate than the CEB-FIP models for this type of structure, and its use should be encouraged in design practice. The paper also demonstrates the importance of considering the interaction between the steel tube and the concrete core in the creep analysis, which is often overlooked in simplified design methods.

Practical Recommendations

  1. Use the ACI 209R creep coefficient model for the design of CFST arch bridges.
  2. Account for the stress redistribution between the steel tube and the concrete core in the creep analysis.
  3. Set the initial camber of the arch based on the predicted creep deflection, using the ACI 209R model.
  4. Optimise the concrete filling sequence during construction to minimise the creep effects.
  5. Specify low-creep concrete for the core, with a water-cement ratio of 0.35 or lower and the use of silica fume.

Conclusion

This paper makes a valuable contribution to the design of long-span CFST arch bridges by providing a reliable method for predicting the creep behaviour. The age-adjusted effective modulus method, combined with the ACI 209R creep coefficient model, provides accurate predictions of the creep deformation and the stress redistribution between the steel tube and the concrete core. The case study of the 380 m span railway arch bridge demonstrates the practical significance of the creep analysis, showing that the creep deflection can be as large as 150 mm and that the choice of creep model can affect the results by 20%. For the practising engineer, the key takeaway is that creep must be carefully considered in the design of CFST arch bridges, and the ACI 209R model should be used as the default choice. I would also like to see further research on the effect of temperature and humidity on the creep behaviour of CFST members, as these environmental factors can significantly affect the long-term deformation.