Study on quantitative characterization of yield state for locally corroded bridge steel via metal magnetic memory method

Corrosion poses a significant threat to the load-bearing capacity of steel components in bridges. Metal Magnetic Memory (MMM) emerges as a promising technique for detecting corrosion-induced damage. In this paper, the tensile tests for 18 Q345qD specimens with local corrosion were performed and the SMFL signals H_SF(y) on all specimens were measured. The results show that as the tensile load increased, the mutation degree of the H_SF(y) signal in the corroded region increased monotonically before yielding and decreased monotonically after yielding. The difference H_SF(y)_m between the maximum and minimum of H_SF(y) signals, as a magnetic characteristic parameter, showed a maximum at the yield load, which had a potential possibility for the judgment of the yield status for the locally corroded bridge steel. Furthermore, the modified elastoplastic magneto-mechanical coupling model and the magnetization-related 3-D magnetic charge model were proposed to further describe the quantitative relationship between the magnetic characteristic parameter and tensile load for the corroded bridge steel. The theoretical curve agreed well with the experimental data, and the characterization error of yield loads for corroded specimens based on the maximum theoretical magnetic characteristic parameter averaged 2.21%. This demonstrated that the established theoretical calculation models can accurately reflect the yield load of corroded specimens. This research serves as a base for the yield state characterization for corroded bridge members via the MMM.