Long-term state-driven atmospheric corrosion prediction of carbon steel in different corrosivity categories considering environmental effects

This study investigates an environment-centered, state-driven corrosion prognosis framework to predict the long-term atmospheric corrosion loss of metal materials, and this paper takes carbon steel as an example to show the establishment process of the framework. Unlike traditional power-linear prediction models that seldomly consider environmental impacts, the proposed model quantitatively establishes the correlations between corrosion loss and dynamic atmospheric environmental factors. A comprehensive power-linear function model integrating multiple atmospheric environmental factors is constructed, following the corrosion kinetics robustness. Under the proposed framework, the steady-state start time is evaluated, followed by the long-term corrosion loss prediction under different corrosivity categories and test sites. The applicability is justified via a case study of long-term field exposure tests of metal materials in China, as well as the experimental results of the ISO CORRAG program. By comparing with the traditional power model and ISO model, the experimental results demonstrate the capability and effectiveness of the proposed prognosis methodology in acquiring accurate corrosion state information and corrosion loss prediction results with less input corrosion information.