| Objective To propose a new method based on data analysis to refine the atmospheric corrosion categories. Methods The method of ordered sample clustering was used in this paper to analyze the data of carbon steel corrosion rate accumulated in atmospheric corrosion test stations in China during the past years. As a result, the atmospheric corrosion degree of carbon steel was divided into 10 categories. On this basis, the inverse distance weighted interpolation method combined with ISO 9223—2012 standard was used to obtain the estimation table of atmospheric corrosivity categories according to the concentration of sulfur dioxide, concentration of chlorine ion, time of wetness data and the corrosion rate data of carbon steel in the corresponding atmospheric environment. Results The result of refined 10 categoried was 0~8 μm/a(C1), 8~25 μm/a(C2), 25~40 μm/a(C3), 40~50 μm/a(C4), 50~60 μm/a(C5), 60~ 75 μm/a(C6), 75~90 μm/a(C7), 90~100 μm/a(C8), 100~200 μm/a(C9), and 200~700 μm/a(C10). Conclusion This 10-category method is more refined in describing atmospheric corrosion. |