Localized corrosion damage prediction of steel plates in marine applications using quadrilateral inverse-shell elements based on iFEM

Abstract

This research effort focuses on a damage detection strategy for identifying pitting corrosion locations in corroded marine structures using numerical approaches. To this end, a four-node quadrilateral inverse shell element (iQS4) is implemented to apply the inverse finite element method (iFEM). At first, a high-fidelity finite element model of a corroded specimen with a semi-elliptical pit and a curved plate with a rectangular pit in the center are developed using the commercial software Abaqus to produce the strain-sensor data. These strain data are used as an input of the iFEM formulation to reconstruct the displacement and continuous strains of the corroded samples. Then, the corrosion damage index is defined based on the difference between the equivalent strain of the corroded and intact model. A single pit in a critical point can cause a great deal of damage; therefore, higher values of the damage index reveal the location of the pits. Moreover, to increase the practicality of the proposed method, sensors are removed from the corroded zone to estimate the damage distribution with sensors mounted in the intact parts of the specimen. Finally, in both full and reduced sensor cases, the predicted pit locations show high consistency with the reference numerical solutions.