Shake Table Tests of Partially Filled Steel Tube Concrete Bridge Piers

Literature Overview

This 2019 study by Xu Yan and colleagues, published in the China Journal of Highway and Transport, presents destructive shake table tests on two two-span elevated bridge models (scale ratio 1:6.7) with partially filled circular steel tube concrete piers. The research is supported by the Ministry of Education International Laboratory for Engineering Earthquake Engineering, the National Key R&D Program of China, and the National Natural Science Foundation of China. The study compares the seismic behavior of partially filled piers with no eccentricity against those with eccentricity, providing critical insights for the seismic design of bridge piers using this construction method.

Test Configuration and Parameters

Two bridge models were tested under strong earthquake ground motion inputs. The key distinction between the two models was the presence or absence of dead load eccentricity.

The tests captured the complete seismic response, including pier top displacements, local buckling deformations, crack patterns, energy dissipation, and residual displacements.

Key Findings and Technical Analysis

Failure Mechanism Without Eccentricity

The pier without eccentricity exhibited uniform local buckling deformation near the base and distinct horizontal cracks. The horizontal cracks are indicative of flexural failure in the concrete-filled portion of the pier, where the tensile stresses exceeded the tensile strength of the concrete. The uniform local buckling of the steel tube suggests that the confinement effect was effectively distributed around the circumference. The longitudinal pier top displacement was significantly larger than that of the eccentric pier, which may seem counterintuitive but is explained by the absence of the stiffening effect provided by the eccentric dead load.

Failure Mechanism With Eccentricity

The pier with eccentricity did not develop horizontal cracks, which is a notable difference from the non-eccentric case. The absence of cracking indicates that the eccentric dead load pre-compressed the section in a way that delayed or prevented tensile failure. The transverse pier top displacement and local buckling deformation were concentrated on the side closer to the centerline, where the eccentric moment induced higher stresses. The plastic development region was larger than in the non-eccentric pier, indicating more extensive inelastic deformation.

Displacement Comparison

The longitudinal pier top displacement of the non-eccentric pier was much larger than that of the eccentric pier. However, the transverse pier top maximum displacement of the eccentric pier was greater than that of the non-eccentric pier. The residual displacements in both directions were approximately 3 cm for both piers, confirming that the eccentric pier can meet seismic performance requirements if the cross-section is properly designed.

Energy Dissipation Asymmetry

A critical finding is that the dead load eccentricity caused unequal energy dissipation on the two transverse sides of the eccentric pier. This asymmetry can lead to differential cumulative plastic deformation between the two sides, which is detrimental to the full utilization of the pier's seismic performance and energy dissipation capacity. This finding has important implications for the design of eccentric piers, as it suggests that the conventional symmetric design approach may not be optimal.

Engineering Practice Implications

The study provides several important design guidelines for partially filled steel tube concrete bridge piers:

• The concrete fill height and diameter-to-thickness ratio must be carefully selected to minimize the adverse effects of dead load eccentricity on seismic performance.
• Eccentric piers can achieve acceptable seismic performance with proper cross-sectional design, as confirmed by the comparable residual displacements of approximately 3 cm for both pier types.
• The asymmetric energy dissipation in eccentric piers should be accounted for in design, as it may lead to progressive damage accumulation on one side of the pier.
• Partially filled piers offer a practical construction advantage by reducing concrete volume and self-weight while maintaining adequate seismic performance, making them attractive for bridge applications where construction efficiency and cost are important.

The comparison between eccentric and non-eccentric piers highlights the importance of considering dead load eccentricity in seismic design. In practice, dead load eccentricity is almost always present due to construction tolerances, asymmetrical bridge deck configurations, or intentional design features. The study demonstrates that this eccentricity is not merely a second-order effect but can significantly influence the seismic behavior and failure mechanism of partially filled steel tube concrete piers.

Study Insights and Reflections

The most important insight from this study is the complex interaction between dead load eccentricity and seismic response in partially filled steel tube concrete piers. The absence of horizontal cracks in the eccentric pier, despite larger plastic deformation, suggests that the pre-compression from eccentric dead load provides a beneficial effect by delaying tensile failure. However, this benefit comes at the cost of asymmetric energy dissipation and potentially uneven damage accumulation.

The finding that residual displacements are comparable for both pier types is reassuring from a post-earthquake functionality perspective. It suggests that partially filled steel tube concrete piers with eccentricity can maintain their structural integrity after strong earthquakes, provided that the design parameters are properly optimized. This is particularly important for bridge structures, where post-earthquake functionality is a key performance objective.

From a construction perspective, the partially filled approach offers significant advantages in terms of concrete volume reduction and construction speed. The shake table test results provide the experimental validation needed to support the use of this construction method in seismic design codes and guidelines. Future research could extend to full-scale tests and consider the effects of varying fill heights, different steel tube grades, and multi-story pier configurations.

This study exemplifies the value of shake table testing in revealing complex seismic behaviors that are difficult to predict through numerical analysis alone. The asymmetric energy dissipation and the delayed cracking due to eccentric pre-compression are phenomena that may not be fully captured by conventional numerical models, highlighting the continued importance of physical testing in advancing seismic design knowledge.