Pseudo-Dynamic Testing and Finite Element Analysis of Steel Tube Concrete Substructures Under Variable Axial Force

Literature Overview

This paper by Zhang Guowei, Chen Boshan, Zhan Yuan, Guo Yurong, and Gao Yuchun, published in Steel Construction (Vol. 32, No. 6, 2017), presents a comprehensive investigation into the seismic performance of circular steel tube concrete (CFT) frame columns subjected to variable axial force conditions. The study employs substructure pseudo-dynamic testing methodology on six CFT column specimens under four levels of seismic excitation, followed by finite element analysis using OpenSees software. The research specifically addresses the effect of variable axial force on the hysteresis characteristics, stiffness degradation, and energy dissipation capacity of CFT columns, which is a critical consideration for realistic seismic performance evaluation of steel tube concrete structures.

Core Technical Points

Substructure Pseudo-Dynamic Testing Methodology

Substructure pseudo-dynamic testing is an advanced experimental technique that combines physical testing of critical structural components with numerical simulation of the remaining structure. In this approach, the physical substructure (the CFT column specimens) is tested in the laboratory while the global structural response is computed numerically in real-time. This methodology allows for realistic simulation of the interaction between the column and the rest of the structure, including the effects of P-Δ moments, axial force variations due to gravity load redistribution, and the influence of structural period on seismic demand.

The six CFT column specimens were designed with varying parameters including axial load levels, diameter-to-thickness ratios (D/t), and column types. The four seismic excitation levels represent progressively more severe ground motions, corresponding to frequent, fortification, and rare earthquake intensities as defined in Chinese seismic design codes (GB 50011). The axial force variation during testing simulates the realistic condition where gravity loads are redistributed as the structure deforms during seismic response, leading to cyclic changes in the axial compression force on individual columns.

Effect of Variable Axial Force on Seismic Performance

Performance Parameter	Fixed Axial Force Column	Variable Axial Force Column	Key Observation
Hysteresis loop shape	Full, stable loops	Pinched loops, especially under rare earthquake	Pinching effect more pronounced with variable axial force
Maximum lateral displacement	Larger	Smaller	Variable axial force reduces drift capacity
Ultimate load	Higher	Lower	Axial force variation reduces load capacity
Stiffness degradation	Gradual	Accelerated	Variable axial force accelerates stiffness loss
Energy dissipation	Higher cumulative	Lower cumulative	Reduced hysteretic energy absorption

The most significant finding is the pronounced "pinching effect" observed in the hysteresis curves of variable axial force columns under rare earthquake conditions. This pinching effect arises because the reduction in axial force during certain phases of the seismic cycle reduces the confining pressure on the core concrete, allowing the steel tube to buckle more readily and the concrete to crush more easily. When the axial force increases again, the column must re-establish the confining mechanism, leading to a loss of stiffness and strength during each cycle.

Failure Mode and Ductility

The primary failure mode observed in the CFT columns was local buckling of the steel tube in the plastic hinge region, which is consistent with the expected behavior of CFT columns designed for seismic applications. The steel tube, acting as external confinement, experiences outward radial pressure from the confined concrete as it crushes, leading to local buckling when the hoop stress exceeds the buckling capacity of the tube wall. The diameter-to-thickness ratio is the primary geometric parameter controlling this buckling behavior, with higher D/t ratios leading to earlier buckling and reduced ductility.

The OpenSees finite element analysis was calibrated against the experimental results, with good agreement confirmed between numerical predictions and test observations. The numerical model incorporated concrete confined behavior using the Mander model, steel tube plasticity with kinematic hardening, and contact interaction between the steel tube and concrete core. This validation confirms the reliability of OpenSees for seismic performance analysis of CFT structures.

Standards and Design Code Considerations

The findings of this paper have direct implications for the seismic design of steel tube concrete structures governed by GB 50011 (Code for Seismic Design of Buildings), GB 51249 (Technical Code for Concrete-Filled Steel Tubular Structures), and JGJ/T 198 (Technical Specification for Steel-Concrete Composite Structures). Current design codes typically assume constant axial force during seismic analysis, which simplifies the analysis but may underestimate the adverse effects of axial force variation on column performance.

The paper's findings suggest that design provisions should account for the reduction in ductility and energy dissipation capacity caused by variable axial force. This may require:

More conservative ductility demands for columns subjected to significant axial force variation.
Lower allowable D/t ratios for columns in high-seismicity zones where axial force variation is expected.
Enhanced confinement requirements (thicker steel tubes or additional internal confinement) for columns expected to experience large axial force fluctuations.
Performance-based design approaches that explicitly account for the variable axial force effect on column behavior.

Engineering Practice Integration

From a practical engineering standpoint, the variable axial force effect is most pronounced in structures where:

The gravity load distribution is asymmetric or has significant eccentricity.
The structure has a large number of stories, leading to significant P-Δ effects.
The foundation has differential settlement or rotational stiffness variations.
The structure is designed with significant overstrength in some columns relative to others.

For these cases, engineers should consider performing nonlinear time-history analysis with variable axial force rather than relying on constant axial force assumptions. The substructure pseudo-dynamic testing methodology demonstrated in this paper provides a rigorous experimental validation approach that can be applied to critical structures where the seismic performance of CFT columns is a primary design concern.

Study Insights and Implications

This paper makes a significant contribution to the understanding of CFT column seismic behavior under realistic loading conditions. The demonstration that variable axial force substantially degrades the hysteretic performance of CFT columns, particularly under severe seismic excitation, challenges the conventional design assumption of constant axial force. The good agreement between OpenSees numerical analysis and experimental results provides confidence in using this software for performance-based seismic design of CFT structures.

The practical implications are clear: engineers designing CFT structures for seismic zones should not rely solely on constant axial force assumptions, particularly for critical columns that may experience significant axial force variation during earthquake response. The pinching effect and reduced ductility observed under variable axial force conditions may lead to premature failure of columns that appear adequately designed under constant axial force analysis. Future research should focus on developing simplified design methods that account for the variable axial force effect, and on exploring reinforcement strategies that can mitigate this adverse effect without excessive cost increase.