A series of regression analyses were made on the one-year corrosion measurements of flat panel and wire helix specimens. There was a very strong relationship between the wire and panel results with the wire specimens yielding higher rates than the panels, but approaching the panel rates at the highest rates for all four metals. A second series of regression analyses between the residuals from the panel-helix regression was run against environmental variables. The residuals showed a significant positive relationship with sulfur dioxide for zinc, chloride for copper, and both chloride and sulfur dioxide for aluminum. This indicates that increases in the environmental variable increased the corrosion rate of the wires more than flat panels. An estimate of the gas phase convective mass transfer rate of sulfur dioxide to wires as compared to flat surfaces showed that sulfur dioxide would diffuse to wire surfaces about three times the rate to a flat panel. The collection efficiency of chlorides to wire surfaces would likewise be greater than fiat surfaces, but the actual rates would depend on the particle size and wind velocity. These results show that wire helix specimens are not equivalent to flat panels in estimating atmospheric corrosivity.
Key words: atmospheric corrosion, time of wetness, chloride deposition, regression analyses, ISO CORRAG program, wire corrosion rate.