Effect of environmental conditions on corrosion rates

Existing methods of estimating long-term corrosion loss are almost always based on time, with separate models for different atmospheric conditions (i.e., industrial, urban, rural, marine). Such models do not account for variation in environmental conditions, even though it is known that corrosion loss varies significantly with different environmental factors. This research involved the formulation, calibration, and validation of models that show the effects of the environment on corrosion for four materials (carbon steel, zinc, copper, and aluminum) and two types of specimens (flat and helix). The models were formulated with a structure that yields rational effects for four environmental variables: time-of-wetness, sulfur dioxide, salinity, and temperature. ISO CORRAG data obtained from 51 test sites in 13 countries was used to calibrate the coefficients of the eight models. The prediction accuracy of the models was assessed using goodness-of-fit statistics. The correlation coefficients ranged from 0.532 to 0.834 for the models. Using data independent of the calibration data, the models were validated. The results of validation indicate that the models are reliable for use in a broad range of conditions or locations. The temporal variation in corrosion loss for each of the four environmental factors is also discussed. Both calibration and validation suggest that the environmental components of the models can be used with other time-dependent corrosion models to assess the effects of variation in environmental conditions.