Zhuojin Pipe Fitting Co., LtdNumerical Analysis of Eccentric Compression Capacity for Concrete-Filled Steel Tube Lattice Columns
Literature Overview
This paper by Jiang Lizhong, Zhou Wangbao, and Tang Bin, published in the Journal of Computational Mechanics in 2010 (Volume 27, Issue 1, pages 127–131), addresses a critical yet often overlooked aspect of concrete-filled steel tube (CFST) lattice column design under eccentric compression. The authors develop a novel numerical method based on Taylor series interpolation, considering equilibrium conditions at multiple cross-sections and the confining effect on the stress-strain behavior of the concrete core. The work is funded by the National Natural Science Foundation of China (Grants 50438020 and 50778177) and the Hunan Provincial Outstanding Youth Fund (Grant 07JJ1009).
Core Technical Approach
The fundamental innovation lies in replacing the conventional piecewise linear interpolation with Taylor series functions as the interpolation scheme for segment-wise synthesis of the column's deformation and equilibrium. This choice is not merely mathematical elegance but reflects a physically meaningful improvement: Taylor series naturally capture the curvature of the elastic-plastic moment-rotation relationship at each cross-section, whereas linear interpolation introduces artificial kinks that distort the true load-displacement response.
The method accounts for three critical physical phenomena that traditional approaches often neglect:
- Shear deformation contribution — The classical Euler-Bernoulli beam assumption is relaxed to incorporate shear flexibility, which becomes significant in short-to-medium-span lattice columns where the shear-to-bending ratio is non-negligible.
- Non-uniform eccentricity — The formulation is extended to handle unequal end eccentricities, a condition frequently encountered in asymmetric braced frames and lattice towers.
- Confinement-enhanced concrete behavior — The stress-strain model for the concrete core explicitly incorporates the lateral confinement pressure exerted by the steel tube, following the well-established Mander-type model but adapted for the specific geometry of lattice column limbs.
Comparison with Existing Methods
| Method | Shear Deformation | Unequal Eccentricity | Confinement Effect | Accuracy |
|---|---|---|---|---|
| Existing segmented synthesis | Not considered | Not applicable | Simplified or absent | Conservative, large error |
| Proposed Taylor-series method | Fully included | Supported | Explicitly modeled | Good agreement with tests |
| Code-based design (GB 50936) | Implicit via empirical formulas | Limited | Empirical | Overly conservative |
The authors validated their program against experimental data for four-limb CFST lattice columns with equal end eccentricities. The results demonstrate that existing design codes yield conservative estimates with errors exceeding 15–20% in certain parameter ranges, while the proposed method achieves deviations within 5–8% of measured ultimate loads.
Engineering Practice Implications
From a fabrication and quality control perspective, this research has several practical implications for pipe and fitting manufacturers:
- Welding sequence planning — Lattice columns require precise geometric control at each diagonal brace connection. The numerical method's accuracy in predicting plastic hinge formation locations can guide the pre-deformation compensation strategy during welding, reducing residual distortion.
- Material selection — The confinement effect analysis implies that higher-grade steel tubes (e.g., Q345 vs. Q235) provide disproportionately greater confinement benefits, justifying the cost premium for critical structural applications.
- Inspection priorities — Since the ultimate capacity is governed by the interaction between steel tube yielding and concrete crushing, ultrasonic testing (UT) of the tube wall thickness uniformity and hydrostatic pressure testing of the concrete fill become the most critical quality control steps.
Key Reflections
The adoption of Taylor series interpolation is a subtle but powerful methodological choice. In my decades of experience with numerical modeling of welded structures, I have observed that the accuracy of any segmented analysis is fundamentally limited by the interpolation order. For members exhibiting pronounced plastic curvature — such as CFST columns approaching their ultimate state — higher-order interpolation is not optional but essential. This work provides a rigorous justification for that principle within the specific context of lattice columns.