Reliability Analysis of Concrete-Filled Steel Tube Arch Bridges Under Construction Deviations

Literature Overview

The paper by Kang Haigui, Zhang Jing, and Yu Dasheng (2011), published in the Journal of Dalian University of Technology, presents a probabilistic reliability analysis of CFST arch bridges accounting for construction-induced geometric deviations. The authors employed the stress superposition method (limit state design method) to simulate the bridge completion process and combined the response surface method with First-Order Reliability Method (FORM) to compute system reliability. This research is highly relevant to engineers involved in the construction monitoring and structural assessment of large-span CFST arch bridges.

Core Technical Methodology

The study adopts a systematic approach to reliability assessment:

Stress Superposition Method: The construction sequence of the arch is simulated by superimposing stress states from each construction stage, accounting for the progressive loading of the CFST arch ribs.
Response Surface Method (RSM): A polynomial approximation of the limit state function is constructed to reduce computational burden while maintaining accuracy.
FORM Analysis: The reliability index is computed by finding the most probable point of failure in the standard normal space.

Key Results

Analysis Condition	Reliability Index (β)	Failure Probability	Relative Risk
Design arch axis height, 1/2 span influence line loading	3.04	Baseline	1.0
Arch axis dropped 0.3 m, 1/2 span influence line loading	2.44	5.2× baseline	5.2
Design arch axis height, other loading conditions	> 3.04	Lower than baseline	< 1.0

The most critical finding is that a mere 0.3 m downward deviation of the arch axis increases the failure probability by 5.2 times, reducing the reliability index from 3.04 to 2.44. This dramatic sensitivity underscores the importance of construction accuracy in CFST arch bridge projects.

Engineering Practice Integration

Construction Control Requirements

For CFST arch bridge construction, the following control measures are essential:

Surveying accuracy: Total station and GPS monitoring should achieve vertical accuracy of ±5 mm or better during arch rib erection.
Temperature compensation: CFST arch ribs are sensitive to temperature variations; construction schedules should account for thermal expansion/contraction effects on arch geometry.
Concrete filling sequence: The concrete filling process must be controlled to prevent asymmetric loading that could induce permanent geometric deviations.
Temporary support design: Temporary towers and stays must be designed to maintain the arch axis within specified tolerances throughout the construction sequence.

Welding and Fabrication Implications

From a steel pipe fabrication and welding perspective:

The CFST arch rib segments must be fabricated to precise geometric tolerances to ensure proper assembly alignment.
Longitudinal seam welds in the arch rib pipes should be inspected via ultrasonic testing (UT) to ensure full penetration and absence of planar defects.
Butt welds between arch rib segments should be performed with controlled angular distortion to maintain the design arch profile.
The welding residual stress distribution should be considered in the reliability model, as it affects the initial stress state of the structure.

Key Questions and Reflections

This study raises important questions for engineering practice. First, the 0.3 m deviation used in the analysis represents a relatively small displacement for a large-span arch bridge, yet it causes a dramatic increase in failure probability. This suggests that current construction tolerances may need to be tightened for CFST arch bridges. Second, the analysis focuses on geometric deviations but does not fully account for material property variability, welding quality variations, and corrosion effects over the service life. A more comprehensive reliability model should incorporate these additional uncertainty sources.

The use of FORM is appropriate for this study, but engineers should be aware that FORM assumes a linear limit state function near the design point. For complex structural systems with multiple failure modes, Second-Order Reliability Method (SORM) or Monte Carlo simulation may provide more accurate results. Nevertheless, the FORM approach offers a practical balance between accuracy and computational efficiency for preliminary design and construction monitoring purposes.