Representativeness of small coupons for corrosion of large plates

Earlier work (Melchers and Jeffrey 2002) showed that coupon size and shape was not a major contributor to the variability in corrosion loss estimates. However these conclusions were based on the relatively small coupons used typically in corrosion field studies. In order to estimate whether such small coupons produced realistic results for much larger plates, mild steel plates 1.2 m x 0.8 m x 3mm thick were exposed for up to 2.5 years in immersion, tidal and splash zone conditions at Taylors Beach, NSW. The plates were recovered at 6, 12, 18, 24 and 30 months. At each stage the plates recovered were cleaned, dried and guillotined into 12 equal rectangles suitable for fitting into a mechanical 3D scanner. The digitised scan results show a high degree of similarity of corrosion across a surface, and a similar pattern of corrosion loss with time of exposure. Estimates for variability of corrosion loss were made from the digitized data. Differences between immersion, tidal and splash zone corrosion were observed. The results validate the use of smaller coupons for field studies and loss and variability estimates.