A Hybrid Approach for Damage Identification of Steel Bridges with Discrete Cracks

Many steel bridges are reaching the end of their service life due to increased traffic loads and outdated design practices. To predict their future condition, a hybrid approach for identifying fatigue-related damage, particularly discrete cracks, is applied to steel bridges with orthotropic decks. This method combines a data-driven selection of potential damage locations with highly detailed, physics-based simulation models using the Static Condensation Reduced Basis Element method (SCRBE). Synthetic data from static load tests, based on real-world loading scenarios, demonstrates the approach’s effectiveness on a steel arch bridge with various damage states. The results show that the method can identify damage using global displacement measurements. However, higher loads are required for practical damage detection due to the small effect of discrete damage on the global load-bearing behaviour. The study highlights the importance of focusing on realistic, localised damage, rather than extensive, fictitious scenarios in all damage identification methods. Incorporating innovative simulation techniques and Structural Health Monitoring can improve predictive maintenance for steel bridges, especially given the challenges of complex damage in large structures.