Seismic Performance Simulation of Square Steel Tube Concrete Columns Using OpenSEES

Literature Overview

This paper by Chen Jianwei, Bian Jinliang, Su Youpo, Zheng Lingling, and Gong Wei, published in the World Information on Earthquake Engineering (Vol. 31, No. 3, 2015, pp. 71–77), presents a numerical investigation into the seismic performance of square steel tube concrete (CFST) columns using the OpenSEES finite element software. The research is supported by the National Natural Science Foundation of China (Grant 51278164) and the Hebei Provincial Natural Science Foundation (Grant E2014209221). The study focuses on developing a fiber-based finite element model that accurately captures the interaction between the steel tube and the confined concrete, and uses this model to conduct parametric analyses on key design variables.

Modeling Approach and Methodology

Fiber Section Model

The OpenSEES software employs a fiber-based section formulation for modeling the cross-section of the CFST column. In this approach, the cross-section is discretized into a number of fibers, each assigned a uniaxial material constitutive law. The key modeling decisions include:

1. Steel tube fibers: The steel tube is modeled using a steel material model that captures elastic-plastic behavior, including kinematic hardening.
2. Concrete fibers: The confined concrete is modeled using the Mander concrete constitutive model, which accounts for the enhanced strength and ductility of concrete under confinement.
3. Steel-concrete interaction: The interaction between the steel tube and the concrete is captured through the confinement pressure provided by the steel tube as the concrete expands laterally.

Mander Concrete Constitutive Model

The Mander model is widely used for modeling confined concrete behavior. The model defines the stress-strain relationship as:

• Pre-peak behavior: The concrete exhibits linear elastic behavior up to the confined compressive strength.
• Post-peak behavior: The concrete undergoes strain-hardening or strain-softening depending on the confinement level.
• Confinement pressure: The lateral confinement pressure from the steel tube is calculated based on the hoop tension developed in the steel tube as the concrete expands.

The model parameters include the unconfined concrete compressive strength, the confined compressive strength, and the ultimate strain of the confined concrete.

Parametric Analysis Results

Effect of Steel Yield Strength

The parametric analysis investigated the influence of steel yield strength on the seismic performance of the square CFST column. The results indicate that:

• Increasing the steel yield strength increases the horizontal load-bearing capacity of the column.
• Higher steel yield strength leads to higher initial stiffness and higher yield load.
• The post-yield behavior is also affected, with higher strength steel providing greater residual strength.

Effect of Concrete Strength

The influence of concrete compressive strength on the column performance was also examined. The findings reveal that:

• The effect of concrete strength on the horizontal load-bearing capacity is relatively minor compared to the effect of steel yield strength.
• Higher concrete strength contributes to increased initial stiffness but has limited influence on the overall seismic performance.
• The ductility of the column is primarily governed by the steel tube behavior rather than the concrete strength.

Effect of Model Parameters

The study also examined the sensitivity of the results to various modeling parameters:

Technical Analysis and Engineering Implications

Model Validation

The finite element model was validated by comparing the simulated hysteresis curves with experimental test results. The agreement between the numerical and experimental results was reported to be good, confirming the adequacy of the modeling approach. Key aspects of the validation include:

• Peak load: The simulated peak load matches the experimental peak load within acceptable tolerance.
• Yield displacement: The simulated yield displacement is consistent with experimental observations.
• Hysteresis loop shape: The simulated hysteresis loops capture the shape and energy dissipation characteristics of the experimental loops.
• Stiffness degradation: The simulated stiffness degradation pattern agrees with the experimental degradation pattern.

Design Recommendations

Based on the parametric analysis results, the following design recommendations can be made for square CFST columns in seismic design:

1. Steel grade selection: Higher strength steel grades provide greater load-bearing capacity and should be considered for columns in high seismic zones or for columns requiring high strength.
2. Concrete strength optimization: Since concrete strength has limited influence on seismic performance, moderate concrete strength grades can be used to optimize cost without significantly compromising performance.
3. Steel tube dimensions: The steel tube wall thickness and diameter should be designed to provide adequate confinement and sufficient post-yield deformation capacity.
4. Modeling practice: Fiber-based models with the Mander concrete constitutive law provide a reliable and efficient approach for analyzing the seismic performance of CFST columns.

Comparison with Other Modeling Approaches

The fiber section model offers the best balance between accuracy and computational efficiency for seismic analysis of CFST columns.

Key Questions and Reflections

Several important considerations arise from this numerical study:

1. Steel-concrete bond: The fiber model assumes perfect bond between the steel tube and the concrete. In reality, slip can occur at the interface, particularly under severe seismic loading. The effect of bond degradation on column performance should be investigated.
2. Local buckling: The fiber model may not capture local buckling of the steel tube, which can significantly affect the post-yield behavior of CFST columns. Shell or solid element models may be necessary for capturing local buckling effects.
3. Cyclic loading: The parametric analysis was conducted under monotonic loading conditions. The behavior under cyclic loading, which is more representative of seismic loading, may differ due to Bauschinger effects and accumulated damage.
4. Long-term effects: The model does not account for long-term effects such as creep, shrinkage, and relaxation, which can affect the performance of CFST columns under sustained loads.
5. Code calibration: The numerical model should be calibrated against experimental data from multiple sources to ensure its reliability for code development and practical design applications.

Study Insights and Practical Value

This research provides a valuable contribution to the numerical analysis of square steel tube concrete columns for seismic design. The development and validation of a fiber-based finite element model using the Mander concrete constitutive law offers a practical tool for engineers to evaluate the seismic performance of CFST columns under various design conditions. The parametric analysis results provide clear guidance on the relative importance of steel yield strength versus concrete strength, demonstrating that steel properties dominate the seismic performance of CFST columns. The modeling approach is efficient and can be readily implemented in OpenSEES for structural analysis and design optimization. However, engineers should be aware of the limitations of the fiber model, particularly regarding local buckling and steel-concrete bond degradation, and supplement the fiber model analysis with more detailed modeling when necessary. The study underscores the importance of numerical analysis in complementing experimental testing, enabling engineers to explore a wider range of design parameters and optimize structural performance efficiently.