Dynamic Characteristics Test and Finite Element Analysis of Steel Tube Concrete Column Frame Structures

Literature Overview

This study by Du Guofeng, Xu Lihua, Xu Chengxiang, and Fan Hong, published in the Journal of Wuhan University (Engineering Sciences) (2007, Vol. 40, No. 5, pp. 98-102), investigates the dynamic characteristics of steel tube concrete (CFST) column-steel beam frame structures through both experimental modal testing and finite element analysis. Funded by the Hubei Provincial Natural Science Foundation (2003ABA059), the research involves the design and fabrication of a single-bay, two-span, eight-story CFST column-steel beam frame model, followed by dynamic testing using the Single Input Single Output (SISO) method to obtain natural frequencies, mode shapes, and damping ratios. The experimental results are compared with numerical predictions from ANSYS 8.1 finite element analysis.

Experimental Methodology

Model Design and Fabrication

The physical model represents a typical multi-story CFST frame structure with the following configuration:

• Structural type: Single-bay, two-span, eight-story frame.
• Columns: Steel tube concrete (CFST) members providing high axial and bending capacity.
• Beams: Steel members connected to CFST columns.
• Scale: The model is designed at a reduced scale to fit within testing facilities while maintaining geometric and dynamic similarity.

The fabrication of CFST columns in the model involves:

• Steel tube manufacturing with precise dimensional control.
• Concrete placement with vibration to ensure full compaction within the steel tube.
• Proper curing to achieve design strength before dynamic testing.

Dynamic Testing Procedure

The SISO (Single Input Single Output) method involves:

1. Excitation: Applying a known impulsive or harmonic force at a single point on the structure.
2. Response measurement: Recording the displacement or acceleration response at multiple points.
3. Frequency domain analysis: Converting time-domain signals to frequency domain using Fast Fourier Transform (FFT).
4. Parameter extraction: Identifying natural frequencies, mode shapes, and damping ratios from the frequency response functions.

Finite Element Analysis

The ANSYS 8.1 finite element model employs:

• Beam elements for steel beams and steel tubes.
• Solid or shell elements for concrete cores.
• Interface elements to model the steel-concrete bond.
• Appropriate boundary conditions representing the base support.

The numerical model must accurately capture:

• The composite action between steel tube and concrete core.
• The stiffness contribution of both materials to the overall structural response.
• The connection behavior between beams and CFST columns.

Comparison of Test and Analysis Results

The study reports good agreement between experimental and numerical results for:

• Natural frequencies: The FE model predicts frequencies within acceptable deviation from test values.
• Mode shapes: The qualitative patterns of vibration modes match between test and analysis.
• Damping ratios: The analytical damping assumption provides reasonable estimates.

The agreement validates the finite element model for use in seismic design and performance prediction of CFST frame structures.

Engineering Practice Implications

Structural Design Considerations

The dynamic characteristics obtained from this study have direct implications for seismic design:

• Fundamental period: Determines the seismic force demand through the response spectrum. CFST frames typically exhibit shorter fundamental periods than all-steel frames due to the increased stiffness and mass from the concrete core.
• Mode participation: The first few modes typically account for the majority of seismic energy. Engineers should verify that the FE model captures sufficient modes (typically 90% cumulative mass participation).
• Damping ratio: The composite nature of CFST frames may provide slightly higher damping than pure steel frames due to interfacial friction between steel and concrete. However, the study's damping values should be used cautiously as they depend on test conditions and excitation amplitude.

Fabrication Quality and Dynamic Performance

The accuracy of dynamic predictions depends on the actual structural properties, which are influenced by fabrication quality:

• Steel tube dimensions: Variations in wall thickness and diameter affect the flexural rigidity of CFST columns. A 5% reduction in wall thickness can reduce the column's bending stiffness by approximately 19% (since I ∝ t³ for thin-walled tubes).
• Concrete fill quality: Incomplete concrete fill or voids within the steel tube reduce the effective composite stiffness. Post-filling ultrasonic testing can verify concrete fill integrity.
• Connection details: The beam-column connections in the model represent simplified idealizations. In practice, bolted or welded connections introduce local flexibility that affects dynamic response.

Welding and Material Quality

For CFST frame structures, the following welding and material quality aspects are critical:

• Steel tube longitudinal welds: ERW or HFW welds in the steel tubes must be free of defects, as these represent potential weak planes under cyclic seismic loading.
• Beam-column connection welds: The welds connecting steel beams to CFST columns are subject to complex stress states and should be designed with appropriate weld details (e.g., extended end plates, reinforced flange connections).
• Material certification: Both the steel tubes and the structural steel beams should have mill test certificates verifying yield strength, tensile strength, elongation, and impact energy.

Key Questions and Reflections

The study provides valuable validation of FE modeling approaches for CFST frames, but several limitations warrant consideration:

• Scale effects: The reduced-scale model may not fully capture the behavior of full-scale CFST columns, particularly regarding the steel-concrete interface behavior and the concrete confinement effect, which are size-dependent.
• Nonlinear behavior: The dynamic testing and FE analysis appear to focus on linear elastic response. Seismic design requires understanding of inelastic behavior, including steel tube yielding, concrete crushing, and potential debonding.
• Connection behavior: The model's connections are likely simplified. Real connections with bolts, welds, and end plates introduce local nonlinearities that affect overall dynamic response.

The good agreement between test and analysis results provides confidence in the FE modeling approach, but engineers should recognize that such validation is specific to the tested configuration. Extrapolation to different structural configurations, material grades, or connection details requires additional verification.

Summary

This study successfully demonstrates the feasibility of characterizing the dynamic behavior of CFST column-steel beam frame structures through combined experimental testing and finite element analysis. The SISO modal testing methodology provides reliable dynamic parameters, while the ANSYS 8.1 FE model offers a validated tool for predicting structural response. The good agreement between test and analysis results supports the use of FE modeling in the seismic design of CFST frames. For practitioners in steel pipe fabrication and structural engineering, the key insights are that fabrication quality directly influences dynamic performance, that the composite action between steel tubes and concrete cores significantly affects structural stiffness and period, and that validated FE models can serve as essential design tools for optimizing CFST frame structures under seismic loading.