ABSTRACT

This paper presents a study on the surface morphology of corroded steel in the artificial marine atmosphere. Three groups of specimens in three different supported angles were obtained in the environmental cabinet. The 3D scanner was used to digitalize the corrosion surface. The geometry characteristics contributing to the random field model of the corrosion surface were studied statistically. Results show that the corrosion depth obeys the normal distribution as corrosion proceeds and the corrosion surface random field can be regarded as a weak stationary random field. The better fitting model for autocorrelation function is cosine exponential model and the distribution of parameter correlation length lc is related to the supported angles.