Nonlinear Finite Element Analysis of Static Performance of Rectangular Steel Tube Concrete Trusses

Literature Overview

This 2011 paper by Lin Hongyu, Liu Yongjian, and Ren Xiang, published in the Journal of Zhengzhou University (Engineering Science), presents a nonlinear finite element analysis of rectangular steel tube concrete (SC) trusses commonly used in bridge superstructures. The authors developed two different bar-system models—a single-material model and a dual-material model—and compared their analytical results. The research was supported by the National Western Transportation Construction Science and Technology Project and the Shaanxi Provincial Department of Education Research Plan Project. For steel pipe and welding professionals, this study is highly relevant because it directly addresses the structural integrity of rectangular steel tube members and, critically, identifies joint failure as the primary failure mode rather than member failure.

Core Technical Findings

The study establishes two fundamental conclusions that have direct implications for pipe fabrication and welding quality control. First, the single-material model provides more reliable analysis results for rectangular SC trusses, but the material constitutive relationship becomes the critical factor determining the accuracy of the finite element analysis. Second, and more importantly from a fabrication standpoint, the failure of rectangular SC trusses is not caused by member yielding or buckling but by joint failure, specifically occurring at joint locations with relatively large deformation demands.

The limit bearing capacity obtained from the analysis corresponds to the load level at which truss joints undergo plastic deformation, which is slightly lower than the overall truss ultimate bearing capacity. The authors conclude that the analysis model is sufficient to meet the design requirements for trusses in bridge structures.

Welding and Joint Integrity Implications

The identification of joint failure as the governing failure mode is of paramount importance to welding engineers. Rectangular steel tube trusses require extensive welding at nodes where multiple tubes intersect. These joints are typically fabricated using one of the following methods:

1. Tubular joints with welded plates: Gusset plates are welded to the tube surfaces using full-penetration fillet welds or groove welds. The weld quality at these multi-directional joints is extremely sensitive to heat input, weld sequence, and fit-up tolerances.
2. Direct tube-to-tube connections: Where tubes are directly welded to each other at nodes, the geometry creates complex stress states that are difficult to control through welding alone.
3. Hybrid connections: A combination of bolted and welded connections, where welding is used for the primary load path and bolting provides redundancy.

The study's finding that joint failure occurs at locations with large deformation demands directly correlates with areas of high welding residual stress. In rectangular steel tubes, the corner regions experience significantly higher residual stresses than flat surfaces due to the geometric discontinuity at the corner welds. When these corner welds are located at truss joints, the superposition of welding residual stress and service stress can lead to premature failure.

Material Constitutive Relationships and Welding Metallurgy

The paper emphasizes that the material constitutive relationship is the key to accurate finite element analysis. From a welding metallurgy perspective, this translates to the need for precise characterization of the welded joint properties. The stress-strain behavior of the heat-affected zone (HAZ) differs significantly from both the base metal and the weld metal. For Q345 or Q390 grade rectangular tubes commonly used in bridge trusses:

• The base metal typically exhibits a yield strength of 345-390 MPa with an elongation of 20-25%.
• The weld metal (using E7018 or equivalent electrodes) may have a yield strength of 480-520 MPa but with reduced ductility.
• The HAZ, particularly the intercritical zone, can exhibit a yield strength variation of ±30% from the base metal, depending on the cooling rate achieved during welding.

This variability in material properties means that any finite element model using a uniform constitutive law will underestimate the complexity of the actual structural behavior at welded joints. Engineers should consider incorporating weld-specific material models in detailed analyses of critical truss joints.

Engineering Practice and Quality Control Recommendations

Based on the findings of this study, I recommend the following quality control measures for rectangular SC truss fabrication:

1. Weld sequence optimization: Develop a welding sequence that minimizes angular distortion at joint locations, using back-step welding or symmetric welding patterns for corner joints.
2. Pre-weld stress relief: Apply intermediate stress relief heat treatment at 550-620°C for tubes thicker than 25 mm, particularly at multi-tube intersection joints.
3. Non-destructive testing (NDT) protocol: Implement 100% UT (phased array ultrasonic testing, PAUT) at all truss joint welds, supplemented by MT for surface crack detection.
4. Dimensional tolerance control: Maintain tube squareness within ±0.5% of nominal dimension and joint fit-up gap within 2 mm to reduce stress concentration at welded corners.

Study Insights and Reflections

This paper provides valuable insights into the structural behavior of rectangular SC trusses, but I believe it could be strengthened by incorporating experimental data on welded joint behavior. The finite element analysis, while rigorous, does not account for the welding-induced imperfections that are inevitable in practice—such as weld undercut, incomplete fusion, and micro-cracking in the HAZ. These imperfections can significantly reduce the fatigue life of truss joints, even if the static bearing capacity is adequate.

In my experience with large bridge truss fabrication projects, the transition from member design to joint design is where most engineering challenges arise. The assumption of perfectly welded, stress-free joints in analytical models is optimistic at best. I would encourage future research to develop hybrid analytical-experimental approaches that combine nonlinear finite element analysis with physical testing of representative welded joints, particularly focusing on the interaction between welding residual stress and the cyclic loading experienced in bridge service.

The practical implication for welding engineers is clear: joint design and welding quality must be treated as primary design variables, not secondary fabrication details. The failure mode identified in this study—joint plastic deformation—can be directly influenced by welding procedures, and therefore, welding engineers have a direct role in ensuring the structural integrity of rectangular SC truss systems.