Speaker
Description
Accurate fatigue life prediction for steel connections is vital for structural safety under cyclic loading. While engineers increasingly rely on commercial finite element (FE) software, variations in modeling assumptions and stress definitions can lead to inconsistent results.
This study evaluates the reliability of FE-based fatigue assessments by comparing software predictions against experimental data and established design standards. A comprehensive database of 505 test specimens including base materials, various welded joints (butt, fillet-welded cruciform, and T-shaped), and riveted connections, was compiled from literature. Fatigue lives predicted by Idea StatiCa were compared with experimental outcomes and Eurocode 3 standards. Additionally, representative specimens were modeled in ANSYS Workbench for cross-verification.
The findings reveal that discrepancies primarily stem from differences in stress evaluation methods, weld classification, and damage modeling assumptions. While both software packages generally captured fatigue trends, significant variations in stress ranges and life expectancy occurred across different connection types. This research identifies the strengths and limitations of current FE methodologies, offering practical recommendations for their application in structural design and assessment.
| Topic Area | Fatigue, Corrosion and Structural Integrity of Engineering Structures and Mechanical Systems |
|---|