Zhuojin Pipe Fitting Co., LtdFinite Element Analysis of Initial Stress Effects on Axial Compression Performance of Concrete-Filled Steel Tube Members
Literature Overview
This foundational 1997 paper published in the Journal of Harbin University of Architecture by Zha Xiaoxiong and Zhong Shantong presents the first theoretical finite element analysis of the influence of steel tube initial stress on the compressive mechanical performance and axial load-bearing capacity of concrete-filled steel tube (CFST) members. The research treats CFST as a composite material and employs a finite element approach to analyze the interaction between the steel tube and the infilled concrete under axial compression. To validate the theoretical results, the authors conducted a series of comparative tests on CFST members with and without initial stress in the steel tube. This pioneering work laid important groundwork for understanding the complex interaction mechanics in CFST composite members.
Core Technical Framework
The concept of initial stress in the steel tube is central to this study. Initial stress arises from various fabrication and construction processes, including welding residual stress, cold-forming residual stress, thermal residual stress from concrete curing, and prestress applied during construction. Understanding how these initial stresses affect the structural performance is critical for accurate design and safe construction.
Initial Stress Sources and Magnitudes
| Stress Source | Typical Magnitude | Distribution Pattern | Effect on Performance |
|---|---|---|---|
| Welding residual stress | 0.3-0.8 fy | Tensile near weld, compressive away | Reduces buckling resistance, may initiate cracks |
| Cold-forming residual stress | 0.2-0.5 fy | Tensile on outer surface, compressive on inner | Alters yield behavior, affects concrete confinement |
| Thermal curing stress | 0.1-0.3 fy | Compressive in steel tube | Enhances concrete confinement, beneficial |
| Construction prestress | Variable | Depends on application method | Can be beneficial or detrimental |
The finite element model developed in this study treats the CFST member as a composite of two materials: the steel tube and the concrete core. The interaction between these two materials is modeled through appropriate contact algorithms that allow for slip, separation, and friction. The steel material model includes elastic-plastic behavior with von Mises yield criterion and isotropic hardening, while the concrete model incorporates the confinement effect from the steel tube through a modified stress-strain relationship.
Influence of Initial Stress on Load-Bearing Capacity
The study demonstrates that the initial stress state in the steel tube significantly affects the load-bearing capacity of the CFST member. Compressive initial stress in the steel tube enhances the confinement effect on the concrete, leading to increased concrete strength and improved overall member capacity. Tensile initial stress, conversely, reduces the effective confinement and can lead to premature buckling of the steel tube.
The quantitative relationship between initial stress and load-bearing capacity can be expressed as follows: for compressive initial stress levels up to 0.3 times the yield strength, the load-bearing capacity increases by approximately 5 to 15 percent. For tensile initial stress levels up to 0.3 times the yield strength, the load-bearing capacity decreases by approximately 3 to 10 percent. These findings have direct implications for the design and construction of CFST members, as they highlight the importance of controlling residual stresses during fabrication.
Finite Element Methodology
Model Configuration and Validation
The finite element model employs three-dimensional solid elements for both the steel tube and the concrete core. The mesh density is refined in the regions where stress gradients are expected to be highest, particularly near the loading ends and in the mid-span region where maximum deflection occurs. The boundary conditions simulate the actual test setup, with pinned-pinned or fixed-fixed end conditions depending on the test configuration.
The material properties used in the finite element model are based on coupon test results from the same batch of steel and concrete used in the experimental tests. The steel properties include Young's modulus, Poisson's ratio, yield strength, ultimate strength, and the stress-strain curve parameters. The concrete properties include compressive strength, tensile strength, elastic modulus, and the confinement enhancement factor.
Key Results from Finite Element Analysis
| Parameter | No Initial Stress | Compressive Initial Stress (0.3 fy) | Tensile Initial Stress (0.3 fy) |
|---|---|---|---|
| Ultimate load capacity | Baseline | +5% to +15% | -3% to -10% |
| Peak strain | Baseline | Reduced by 2-5% | Increased by 3-8% |
| Concrete confinement pressure | Baseline | Enhanced | Reduced |
| Steel tube buckling load | Baseline | Increased | Decreased |
| Post-peak ductility | Baseline | Slightly improved | Slightly reduced |
The experimental validation confirms the finite element predictions with good agreement, typically within 10 percent of the theoretical values. The slight discrepancies are attributed to imperfections in the experimental specimens, variations in material properties, and the inherent simplifications in the finite element model.
Welding and Fabrication Implications
The findings of this study have significant implications for the welding and fabrication of CFST members. Welding residual stresses are one of the most common sources of initial stress in steel tube members, and their control is essential for achieving the predicted structural performance.
Welding Residual Stress Control
| Control Method | Effectiveness | Practical Feasibility | Cost Impact |
|---|---|---|---|
| Preheat and interpass temperature control | Moderate | High | Low |
| Post-weld heat treatment (PWHT) | High | Moderate | Moderate |
| Mechanical stress relief (hammering, rolling) | Moderate | Moderate | Low |
| Shot peening | Low for bulk stress | Low | Moderate |
| Symmetric welding sequence | Moderate | High | Low |
The recommended approach for controlling welding residual stresses in CFST members includes: using a symmetric welding sequence to minimize angular distortion, maintaining interpass temperatures below 250 degrees Celsius to limit hydrogen diffusion, and applying post-weld heat treatment at 550 to 620 degrees Celsius for steels with carbon equivalent values above 0.40%. The PWHT should be performed before the concrete is placed in the steel tube to avoid thermal cracking of the concrete.
Construction Sequence Considerations
The construction sequence of CFST members also influences the initial stress state. The recommended sequence includes:
- Fabricate the steel tube with minimum welding distortion through symmetric welding and proper fixture design.
- Perform post-weld heat treatment if required by the steel grade and thickness.
- Install the steel tube in position with proper alignment and support.
- Place the concrete in stages to minimize thermal gradients and segregation.
- Allow adequate curing time before applying structural loads.
Engineering Practice Integration
The findings of this study are directly applicable to the design and construction of CFST columns in building structures, bridge piers, and industrial equipment supports. The key practical recommendations include:
- Design calculations should account for the expected initial stress state in the steel tube, particularly for members subjected to significant welding residual stresses.
- Welding procedures should be qualified to minimize residual stresses, with particular attention to the longitudinal weld seam in the steel tube.
- For members where compressive initial stress is beneficial, consider applying controlled prestress to the steel tube before concrete placement.
- Quality control should include residual stress measurement using the hole-drilling method or X-ray diffraction method for critical members.
Standards Compliance
The design of CFST members should comply with relevant standards including GB 50935 (Technical Code for Concrete-Filled Steel Tubular Structures), EN 1993-1-1 (Eurocode 3), and AISC 360. These standards provide design equations that implicitly account for the interaction between steel and concrete, but they do not explicitly address the influence of initial stress. Engineers should apply appropriate safety factors or modification factors when the initial stress state is known to be unfavorable.
Key Questions and Reflections
This pioneering study raises several important questions that remain relevant to current practice. First, the long-term effect of sustained initial stress on the creep and shrinkage behavior of the concrete core was not addressed. Long-term creep can lead to stress redistribution between the steel and concrete, potentially altering the initial stress state over time. Second, the effect of initial stress on the seismic performance of CFST members under cyclic loading was not investigated. Cyclic loading can lead to fatigue cracking in the HAZ, and the initial stress state may influence crack initiation and propagation. Third, the applicability of the findings to different steel grades and concrete strengths was not systematically explored.
The finite element model developed in this study, while groundbreaking for its time, has been superseded by more sophisticated models that include advanced material models, contact algorithms, and damage mechanics. However, the fundamental insights regarding the influence of initial stress remain valid and continue to inform modern design practice.
Study Insights and Implications
This study represents an important milestone in the understanding of CFST composite member behavior. By being the first to systematically analyze the influence of initial stress on CFST performance, the authors provided valuable insights that have been incorporated into subsequent research and design codes. The finding that compressive initial stress is beneficial while tensile initial stress is detrimental has practical implications for fabrication and construction quality control.
From a metallurgical perspective, the control of welding residual stresses is essential for achieving the predicted structural performance. The use of low-hydrogen welding consumables, proper preheat and interpass temperature control, and post-weld heat treatment are critical quality control measures. The hydrogen embrittlement susceptibility of the HAZ should be assessed through hydrogen diffuse tests per ISO 3676, and the PWHT parameters should be optimized to achieve full stress relief without over-tempering the weld metal.
In conclusion, this study provides a foundational understanding of how initial stress affects the performance of CFST members under axial compression. The practical recommendations for controlling initial stress through fabrication and construction practices are directly applicable to current engineering projects. Future research should extend these findings to cyclic loading, long-term creep, and advanced steel grades to create a comprehensive design framework for CFST structures.