Damage detection in steel bridges using modal strain energy and improved β-index method

Bridges are critical elements of transportation networks, and their long-term safety is essential for public welfare and economic growth. However, many steel railway bridges in service are ageing and operate under increasing loads, yet they lack systematic health assessments. Conventional vibration-based damage detection techniques often require large datasets and extensive computational resources, making them impractical for routine use on single-span steel bridges. This study introduces a streamlined damage detection framework built on the Modal Strain Energy (MSE) concept, designed to be both efficient and reliable. A validated finite element model of a representative steel bridge was developed to test the approach under various damage conditions. To enhance its reliability, a modified damage index (β′) was developed that increases sensitivity to real damage while reducing the influence of low-level noise. This adjustment minimizes false alarms, which are common in traditional β-based methods. Results from multiple damage scenarios and bridge spans demonstrate that the proposed approach can accurately pinpoint damage locations and substantially lower the rate of false detections. The findings provide a practical and resource-efficient solution for structural health monitoring of steel railway bridges, supporting safer and more sustainable infrastructure management.