Creep Calculation Methods for Steel Tube Concrete Axially Compressed Members

Literature Overview

This paper by Zhao Jintang, Zhao Renda, and Zhan Yulin from the School of Civil Engineering at Southwest Jiaotong University, published in Highway and Transportation Research in 2013, Volume 30, Issue 4, presents a systematic study on creep calculation methods for steel tube concrete (SRC) axially compressed members. The research is funded by the National Natural Science Foundation of China (51208431), the Ministry of Education Doctoral Discipline Special Research Fund (20090184120033), and the Central University Basic Research Business Fee Special Fund (SWJTU12CX064). The paper is classified under TU398 and addresses the theoretical derivation of creep stress-strain relationships, the proposal of an equivalent temperature load method, and the comparative evaluation of three typical concrete creep coefficient models.

Theoretical Framework and Derivation

The authors begin by deriving the creep stress-strain calculation formula for ordinary concrete under long-term loading based on the generalized creep coefficient. A key theoretical result is that strain can be used as the fundamental variable for concrete creep calculations, and under reasonable assumptions, the free creep strain of the core concrete can serve as the fundamental variable for SRC creep calculations. This theoretical foundation enables the development of practical calculation methods for SRC members.

The equivalent temperature load method is then proposed as a practical approach to calculate SRC creep. This method converts the creep strain into an equivalent thermal strain, which can be readily incorporated into finite element analysis software. The approach leverages the well-established capabilities of finite element software to handle thermal effects, thereby avoiding the need for specialized creep analysis modules.

The following table summarizes the theoretical framework and calculation approach:

Theoretical Component	Description	Application
Generalized creep coefficient	Basis for creep stress-strain relationship	Ordinary concrete analysis
Strain as fundamental variable	Enables simplified calculation framework	SRC member analysis
Free creep strain of core concrete	Fundamental variable for SRC creep	Practical calculation method
Equivalent temperature load	Converts creep to thermal effect	Finite element implementation

Comparative Analysis of Creep Models

The study evaluates three typical concrete creep coefficient models: MC90, EC2, and GL2000. These models are implemented using the equivalent temperature load method in the ANSYS finite element software to analyze the creep performance of SRC members under long-term axial compressive loading. The calculated results are compared with experimental test data to assess the accuracy of each model.

The following table presents the comparative evaluation of the three creep models:

Creep Model	Accuracy of Prediction	Over/Under Estimation	Applicability
MC90	Good agreement with test data	Accurate	Recommended for SRC analysis
EC2	Good agreement with test data	Accurate	Recommended for SRC analysis
GL2000	Poor agreement with test data	Overestimates creep strain	Not recommended for SRC analysis

The results clearly demonstrate that the MC90 and EC2 models provide predictions that are in good agreement with the experimental test data for SRC members under long-term axial compressive loading. In contrast, the GL2000 model significantly overestimates the creep strain of SRC members, making it unsuitable for practical engineering applications involving SRC structures.

Engineering Practice Applications

The equivalent temperature load method offers a practical and efficient approach for analyzing the long-term behavior of SRC members. For engineers involved in the design of SRC bridges, towers, and other long-term loaded structures, the following practical guidance emerges from this research:

The MC90 or EC2 creep coefficient models should be adopted for SRC member analysis, as they provide accurate predictions of long-term deformation
The equivalent temperature load method can be implemented in standard finite element software such as ANSYS without requiring specialized creep analysis capabilities
The free creep strain of the core concrete should be used as the fundamental variable for SRC creep calculations
The GL2000 model should be avoided for SRC applications due to its significant overestimation of creep strain
Long-term serviceability assessments of SRC structures should account for the interaction between steel tube and concrete core creep behavior

The study also demonstrates that the equivalent temperature load method is applicable to both ordinary SRC and steel tube expansive concrete members, broadening its utility for different types of SRC structures. Expansive concrete, which is used to improve the bond between the steel tube and concrete core, exhibits different creep characteristics that are also accurately captured by this method.

Study Insights and Reflections

This paper makes a significant contribution to the understanding of long-term behavior of SRC members by providing both theoretical foundations and practical calculation methods. The development of the equivalent temperature load method is particularly valuable for practicing engineers, as it enables the use of widely available finite element software without requiring specialized knowledge of creep analysis theory. The comparative evaluation of creep models provides clear guidance on model selection, with the MC90 and EC2 models emerging as reliable choices for SRC applications. For engineers designing SRC bridges and other long-term loaded structures, this research provides the tools and confidence to accurately predict long-term deformation and ensure serviceability performance throughout the design life of the structure. The findings also highlight the importance of model validation against experimental data, as the significant overestimation by the GL2000 model demonstrates the risks of using inappropriate creep models in engineering practice.