Reliability Analysis of Circular Hollow Concrete-Filled Steel Tube Short Columns Under Axial Compression

Literature Overview

This paper by Wang Hongwei, Chen Shaojie, and Zhou Yun from Guangzhou University, published in Journal of Civil, Architectural and Environmental Engineering (2018, Vol. 40, No. 6, pp. 98-107), conducts a comprehensive reliability analysis of circular hollow CFST short columns under axial compression. The study combines experimental data statistics with Monte Carlo simulation to evaluate the reliability indices of both the standard formula (GB 50936-2014) and a modified formula proposed by the authors.

The research addresses a fundamental question in structural engineering: how reliable are the design formulas used for CFST columns, and how should resistance partial factors be determined considering various sources of uncertainty?

Core Technical Methodology

Statistical Analysis Framework

The use of four probability distribution models (Normal, Lognormal, Weibull, Gamma) represents a rigorous approach to characterizing the resistance variability, acknowledging that different distribution assumptions may lead to different reliability estimates.

Key Parameters Analyzed

Key Findings and Technical Interpretation

Reliability Index Results

The study reveals several important findings:

1. Both formulas meet target reliability: The reliability indices calculated using both the GB 50936-2014 formula and the modified formula satisfy the target reliability index of 3.7.
2. Standard formula is more conservative: The reliability index from the standard formula is approximately 1.05 times that of the modified formula, indicating the standard provides a larger safety margin.
3. Parametric sensitivity:

• Reliability index increases with concrete strength and load effect ratio
• Reliability index decreases with steel ratio
• Steel strength and hollow ratio have negligible effects on reliability index

Resistance Partial Factor Recommendations

Based on the reliability analysis, the authors propose resistance partial factors for different load effect ratios:

These values account for the actual load combinations commonly encountered in engineering practice and provide more accurate safety assessments than the fixed partial factors currently used in codes.

Connection to Steel Pipe Engineering Practice

Quality Assurance Implications

The reliability analysis directly informs steel pipe quality control requirements:

Standards Compliance

The study validates the GB 50936-2014 design formula from a reliability perspective, which is important for:

• Justifying the current code provisions to regulatory authorities
• Identifying areas where the code may be overly conservative
• Providing a basis for potential code revisions

Material Specification Requirements

For CFST columns designed according to this reliability framework, the following material specifications are recommended:

• Steel tubes: Minimum yield strength of Q345 for general applications, Q390 for high-strength applications
• Wall thickness tolerance: ±10% of nominal thickness, with minimum thickness verified by UT
• Concrete: Minimum grade C30 for structural applications, with compressive strength verified by cube tests
• Fill ratio: Minimum 95% for hollow CFST columns to ensure adequate confinement

Welding Quality Requirements

The welding of steel tubes in CFST columns must meet the following requirements:

1. Weld procedure qualification per AWS D1.1 or ISO 15614
2. 100% UT inspection of circumferential welds per ASME B31.3
3. MT or PT inspection of longitudinal welds
4. Hydrostatic testing at 1.5 times design pressure per API 5L
5. Visual inspection of all welds per AWS D1.1 Table 6.1

Critical Reflection and Study Insights

The reliability analysis methodology employed in this study is rigorous and follows established probabilistic assessment practices. However, several aspects merit further consideration:

• Model error characterization: The study considers model error as a random variable but does not deeply investigate its physical sources. From a materials science perspective, model error may arise from:
• Simplified constitutive models that do not capture full material behavior
• Assumptions about steel-concrete interface behavior
• Neglect of secondary effects such as creep and shrinkage
• Spatial variability: The study treats material properties as random variables but does not account for spatial correlation of material properties along the column length.
• Long-term behavior: The reliability analysis focuses on short-term load effects and does not consider time-dependent degradation mechanisms such as corrosion, which is particularly relevant for steel pipe structures.

Practical Recommendations

For engineers implementing CFST column designs based on this reliability framework:

1. Document all material test results and traceability information
2. Implement statistical process control (SPC) for tube manufacturing
3. Establish calibration procedures for measurement equipment
4. Maintain construction quality records for future reliability assessment
5. Consider condition monitoring for critical structures

Reference Value and Outlook

This research provides a solid probabilistic foundation for the design of CFST columns and validates the current code provisions. The proposed resistance partial factors offer a more refined approach to partial factor calibration that could be incorporated into future code revisions.

For steel pipe manufacturers and construction contractors, the reliability framework emphasizes the importance of consistent material quality and construction workmanship. Investment in quality control systems that reduce material property variability directly translates to improved structural reliability and potentially lower lifecycle costs.

The methodology presented here can be extended to other CFST structural elements and loading conditions, providing a systematic approach to reliability-based design that bridges the gap between experimental research and practical engineering codes.