Reliability-based calibration of fatigue load factors for steel bridges under traffic growth uncertainty.

Abstract

A reliability-based calibration of the fatigue load factor was performed as a case study for a Mexican steel bridge (Puente Calapa). Empirical Mexican axle-weight distributions gave stresses, processed by rainflow counting, Palmgren-Miner damage accumulation, and Monte Carlo evaluation of a probabilistic limit state. For 2.5%/year growth over 75 years, the Eurocode bilinear S-N curve yielded factors of 1.34 (β=1.0) and 0.70 (β=3.0). The analysis is conditional on a single bridge and idealized structural model. The study's originality lies in making the annual traffic growth rate an explicit calibration input rather than an implicit assumption. Growth rate emerged as the dominant uncertainty, shifting the reliability index by 6.5 units across the examined range and supporting its explicit formulation in code calibration.