Grey relational analysis of static tensile properties of structural steel subjected to urban industrial atmospheric corrosion and accelerated corrosion

The correlation between the accelerated corrosion and atmospheric corrosion is needed in order to study the degradation of mechanical properties, based on the former. To achieve that end, static tensile tests were performed on 78 specimens under three different environments: urban industrial atmospheric corrosion, dry/wet cyclic accelerated corrosion in a laboratory as simulated from the urban industrial atmosphere, and machine-simulated corrosion; and the effects of corrosion method and corrosion degree on the constitutive law of corroded steel plates were investigated based on the corrosion morphology. Moreover, the degree of correlation for the corrosion level assessment and mechanical properties indices were analyzed between the atmospheric corrosion and accelerated corrosion, based on the grey relational analysis. As a result, the mechanical properties decrease and the dispersion increases with the increasing corrosion degree. When the mass loss is ≥16%, the yield plateau of the stress-strain curve of corroded steel plate is significantly shortened. The mass loss of 30% can be defined as the critical point at which the yielding plateau disappears under the dry/wet cyclic accelerated corrosion test. The critical corrosion ratios for the main mechanical indices not meeting the specification requirements are proposed; and the relationship between the etch pit depth and mechanical property degradation is established along with a deterministic constitutive model related to the corrosion ratio. The degree of correlation for either the corrosion level assessment index or the mechanical property index is found to be good. The mechanical properties of corroded steel under the acid rain corrosion environment may be better simulated by the dry/wet cyclic accelerated corrosion or the machine-simulated corrosion when both pitting corrosion and uniform corrosion are considered.