Creep Analysis of Steel Tube Concrete Arch Bridges Considering Geometric Nonlinearity and Construction

Literature Overview

Wu Wenjie, Wang Yuanfeng, and Ma Yishuo (2013, Journal of Southwest Jiaotong University, Vol. 48, No. 4, pp. 645-650) investigated the long-term creep behavior of large-span steel tube concrete (STC) arch bridges, with particular attention to the coupling effects of geometric nonlinearity and staged construction. Supported by the National Natural Science Foundation of China (50778020), this study employed the B3 creep model for concrete and the age-adjusted effective modulus method for structural creep analysis. The work was conducted at Beijing Jiaotong University's School of Civil Engineering.

Technical Framework

The analytical approach combined three key elements:

B3 creep model: A sophisticated constitutive model for concrete creep that accounts for the time-dependent nonlinear behavior of concrete under sustained loading. The B3 model incorporates basic creep, drying creep, and the transition from linear to nonlinear creep behavior.
Age-adjusted effective modulus method: A well-established technique for analyzing time-dependent effects in concrete structures, which adjusts the elastic modulus to account for the age-dependent creep and shrinkage of concrete.
Co-rotational method for geometric nonlinearity: A large-displacement formulation that captures the P-Delta effects and second-order behavior inherent in large-span arch structures.
Birth-and-death element technique: A finite element method for simulating the staged construction sequence, where elements are activated (born) at the time of their actual construction and can be deactivated (die) if needed.

Key Quantitative Results

Analysis Condition	Change in Arch Rib Deflection	Change in Steel Tube Stress	Change in Arch Rib Concrete Stress
Elastic analysis (baseline)	Reference	Reference	Reference
With geometric nonlinearity	< 10%	< 10%	Up to 50%
With staged construction	< 10%	< 10%	Up to 50%
With both effects	< 10%	< 10%	Up to 50%

The most striking finding is that while the effects on steel tube stresses and overall deflections are modest (within 10%), the effects on concrete stresses in the arch rib can reach 50%. This disparity arises because the concrete, being the primary creep-prone material, undergoes significant stress redistribution over time that is highly sensitive to the initial stress state established during construction.

Engineering Practice Implications

For steel pipe fabrication and welding engineers involved in STC arch bridge projects, several practical considerations emerge:

Welded joint integrity under creep conditions: The concrete stress redistribution of up to 50% means that the interface between the steel tube and concrete undergoes significant shear stress changes over the service life. Welded joints in the steel tube, particularly butt welds and circumferential welds, must be designed to accommodate these changing stress states. The welding residual stress superimposed on the creep-induced stress redistribution can lead to unexpected stress concentrations.

Fabrication sequence effects: The birth-and-death element technique captures the reality that different segments of the arch are constructed at different times. This means that concrete placed early in the construction sequence experiences longer creep periods and undergoes greater stress redistribution. From a fabrication standpoint, this implies that steel tubes for early-constructed segments may need to be designed for higher long-term stress levels than those for later-constructed segments.

Geometric nonlinearity and weld distortion: The co-rotational method captures large-displacement effects, which in practice manifest as significant geometric changes in the arch under load. These geometric changes can induce additional stresses in welded joints, particularly at segment connections. The welding residual stresses and HAZ properties become more critical when the structure undergoes large displacements.

Study Insights

The finding that concrete stress changes of up to 50% can occur while steel tube stresses change by less than 10% is physically intuitive but often underestimated in practice. The concrete, being the creep-prone material, transfers stress to the steel tube over time, but the steel tube's high stiffness means it absorbs these stress changes with minimal strain increase. The result is a significant stress redistribution within the composite section that can affect both the short-term and long-term structural behavior.

From a quality control perspective, this study highlights the importance of accurately modeling the construction sequence in the design phase. If the construction sequence is not properly captured in the analysis, the predicted long-term stresses may be significantly different from the actual stresses, potentially leading to under-designed welded connections or over-conservative designs.

The coupling of geometric nonlinearity and construction effects is particularly relevant for large-span STC arch bridges where the initial geometry significantly influences the load path. The 10% change in deflection and steel tube stress may seem modest, but for a large-span arch with a total deflection of several meters, a 10% change represents a significant displacement that can affect the structural behavior and the demands on welded connections.

Summary

This study demonstrates that the creep analysis of large-span STC arch bridges must account for both geometric nonlinearity and the staged construction sequence to achieve accurate predictions of long-term behavior. The most significant effect is on concrete stresses, which can change by up to 50% compared to elastic analysis, while steel tube stresses and overall deflections are affected by less than 10%. For fabrication and welding engineers, these findings underscore the importance of incorporating realistic construction sequences and geometric nonlinearities in the design of welded connections, as the long-term stress redistribution can significantly alter the demands on steel tube welds.