Failure Mode Analysis of Square Steel Tube Concrete Column-Steel Truss Structure

Literature Overview

This paper by Li Zhiqiang, Wang Wei, and Chen Yiyi, published in the Journal of Tongji University (Natural Science) (2015, Vol. 43, No. 4, pp. 513-520), presents a simplified analytical model for predicting the failure modes and load-displacement behavior of composite structures consisting of square steel tube concrete (STC) columns connected to steel trusses. The research, supported by the National Natural Science Foundation of China (Grant 51038008), is conducted at the State Key Laboratory of Disaster Prevention in Civil Engineering at Tongji University. The work addresses an important practical problem: how to predict and control the failure sequence in composite structures where steel and concrete elements are combined.

Core Analytical Methodology

The authors propose a simplified analysis model based on the superposition principle, assuming that the deformation of the steel truss and the STC column are uncoupled. The model consists of two components:

Steel Truss Component: The load-displacement curve is derived by superimposing two mechanisms:

Hinged mechanism (pin-jointed behavior)
Rigid-jointed mechanism (frame action)

STC Column Component: The load-displacement curve is calculated based on:

Moment-curvature relationship
Equivalent shear stiffness

The failure mode of the overall structure is determined by comparing the load-displacement curves of the two components. The component that reaches its limit state first governs the structural failure mode.

Failure Mode Classification and Comparison

The study identifies several possible failure modes for the composite structure:

Failure Mode	Governing Component	Characteristic Behavior	Seismic Implication
Column-dominated failure	STC column	Concrete crushing or steel tube local buckling	Poor ductility; potential sudden collapse
Truss-dominated failure	Steel truss	Member yielding or joint fracture	Better ductility; progressive collapse possible
Balanced failure	Both components	Simultaneous yielding of column and truss	Optimal energy dissipation; ductile behavior
Joint-dominated failure	Connection zone	Weld fracture or bolted joint failure	Brittle; requires careful design

The simplified model predictions showed good agreement with experimental results in terms of failure mode, load-displacement curve shape, and ultimate capacity.

Welding and Connection Design Considerations

The failure mode analysis has direct implications for the welding design of the connections between the steel truss and the STC column:

Connection type selection: The choice between welded, bolted, and hybrid connections affects the failure mode. Fully welded connections tend to create rigid-jointed behavior, while bolted connections with slotted holes allow some degree of hinged behavior.
Weld design for ductility: If the design objective is to achieve truss-dominated failure (which provides better seismic ductility), the weld connections must be designed to be stronger than the truss members themselves. This requires:

Full-penetration groove welds with complete fusion
Post-weld heat treatment to maintain HAZ toughness
Adequate weld toe geometry to minimize stress concentration

Residual stress management: The welding of the truss to the STC column creates a complex residual stress field that can affect the failure mode. If residual tensile stresses are present at the weld toe, they can promote early crack initiation under cyclic loading, potentially changing the failure mode from ductile yielding to brittle fracture.

Weld Parameter	Target Value	Verification Method
HAZ hardness	≤ 350 HV (for Q345 base metal)	Hardness survey per ASTM E18
Residual stress	< 0.3 σ_y in critical zones	X-ray diffraction or hole drilling
Dilution ratio	20-30% for optimal toughness	Chemical analysis of weld metal
Microstructure	Fine-grained acicular ferrite preferred	Metallographic examination

Seismic Performance Implications

The paper discusses the relationship between failure modes and seismic performance, which is of paramount importance for structures in earthquake-prone regions:

Truss-dominated failure is generally preferred for seismic design because steel trusses exhibit better ductility and energy dissipation capacity than concrete elements. The steel members can undergo large inelastic deformations without sudden strength loss.
Column-dominated failure is undesirable because concrete crushing is typically brittle and provides limited energy dissipation. If this failure mode must be accepted, the STC column must be designed with adequate confinement to maintain ductility.
Joint-dominated failure is the most critical scenario because it represents a sudden loss of structural integrity. The weld connections must be designed to be the last element to fail, ensuring that all other ductile mechanisms are fully mobilized before the joints reach their limit state.

Engineering Practice Integration

The simplified analysis model presented in this paper provides a practical tool for engineers to predict failure modes during the design phase. However, its application requires careful attention to several assumptions:

Uncoupled deformation assumption: The assumption that the truss and column deform independently is valid only when the connection flexibility is properly characterized. In practice, the connection stiffness significantly affects the load distribution between the two components.
Material property accuracy: The model relies on accurate material property inputs, including the concrete confinement model and the steel stress-strain curve. For welded structures, the weld metal properties and HAZ properties must also be considered.
Geometric nonlinearity: The simplified model may not capture geometric nonlinear effects such as P-Δ effects, which become significant under large deformations during seismic events.

Study Insights and Concluding Remarks

This paper makes a valuable contribution to the understanding of failure modes in composite steel-concrete structures, and the simplified analysis model provides a practical design tool. However, from a welding and fabrication perspective, I believe the study could benefit from more detailed consideration of weld joint behavior. The failure mode predictions assume idealized connections, whereas real welded joints may exhibit reduced stiffness, strength degradation under cyclic loading, and fatigue crack initiation at stress concentration points. Future research should integrate detailed weld joint models into the failure mode analysis framework, incorporating the effects of welding residual stresses, microstructural variations in the HAZ, and potential weld defects on the predicted failure sequence. The practical implication is clear: the failure mode of a composite structure is not determined solely by the structural configuration but is also influenced by the quality and design of the weld connections that tie the system together. Achieving the desired failure mode requires careful coordination between structural engineering design and welding engineering practice.