Abstract
Corrosion, particularly in marine and offshore environments, often leads to material loss and surface irregularities that compromise structural integrity. Traditional non-destructive testing methods, such as ultrasonic thickness measurements, are limited in regard to detecting widespread or irregular corrosion damage. This study explores the use of vibration-based analysis to assess the effects of corrosion by examining changes in the dynamic behaviour— specifically, the natural frequencies—of steel plates. Numerical simulations are conducted using Abaqus that include random surface irregularities, modelled with Gaussian random fields, to represent generalised corrosion. Experimental validation involves steel plates subjected to accelerated electrochemical corrosion, with degradation assessed based on mass loss and ultrasonic thickness measurements. Changes in modal parameters due to progressive corrosion are recorded using impact hammer excitation and accelerometers. The results show a clear relationship between corrosion-induced thickness reduction and shifts in modal characteristics. The findings demonstrate that modal analysis offers a viable, non-invasive method for detecting and evaluating global corrosion damage in large-scale steel structures.