In Situ Calibration of Girder-End Restraint Parameters in Suspension Bridges Jointly Driven by Monitoring Data and Numerical Modeling

In-service girder-end restraint devices in long-span suspension bridges inevitably degrade, which directly alters their mechanical parameters and affects bridge service safety. It is imperative to quantitatively identify girder-end restraint parameters for their performance evaluation. Because current methods typically require laboratory testing after disassembly, this study proposes a novel, hybrid method, driven by monitoring data and numerical modeling, for the in situ online calibration of girder-end restraint parameters in suspension bridges. First, an inversion method for girder-end restraint forces (GERFs) was derived by establishing an explicit mapping relationship between GERFs and vehicle-induced girder-end displacement (VIGED) differences. Second, a physics-informed theoretical reconstruction model of the VIGED under longitudinal unconstrained conditions (LUC-VIGED) was established to estimate VIGED differences. On this basis, a statistical equivalence-based correction method was introduced to refine the reconstructed LUC-VIGED for operational bridges. Third, a decoupling and quantitative identification algorithm for the mechanical parameters of multirestraint devices was proposed. Finally, the proposed method was validated through numerical simulations and a case study on an in-service suspension bridge. It accurately identified girder-end restraint parameters, including damper damping coefficient and expansion joint stiffness, and showed good agreement with short-term field test results. These findings provide valuable insights for the performance evaluation of restraint devices in suspension bridges.