INTRODUCTION
The mathematical modeling approach to crevice corrosion has been used to develop a test which is capable of predicting the performance of stainless steel components in the marine environment The test requires a single electrochemical data point as input which can be determined rapidly in the laboratory. The test output, via user friendly software distinguishes between different components e.g. flanged joint O-ring, compression fitting, etc. and takes account of temperature and chlorination level in its predictions.
Stainless steels and Ni-base alloys are being “increasingly used in marine environments for a wide range of components and structures because of their intrinsic corrosion resistance. Their successful use however depends on the ability to resist localized attack, in particular crevice corrosion. While there are various tests available for assessing the resistance of materials to crevice corrosion’ which will rank the many different alloys, two major problems still exist. Firstly, different tests produce different rankings, and secondly these rankings do not relate quantitatively to actual service performance.
The objective of the work reported here was to develop a low cost accelerated laboratory test method to assess accurately the resistance of stainless steels and Ni-base alloys to crevice corrosion in marine environments. The new test method would provide both accurate rankings of alloys and also predict the actual service performance of components and structures over a range of temperatures in chlorinated and non-chlorinated conditions. Such test methods did not exist prior to the project. The use of this new test technique will enable industry to select materials for use in marine environments with an improved confidence in their performance both from an economic and a technical point of view.
There are two fundamental drawbacks associated with the majority of current test techniques for crevice corrosion, both concerning crevice geometry. The first is the reproducibility and quantification of artificial crevice gaps, in both accelerated and exposure tests. The other related drawback concerns the geometry of real crevices. Existing test methods produce rankings, and to relate these to service performance other than “by experience” is not possible except by the mathematical modelling technique. However, use of this technique requires accurate specification of in-service geometries in terms of volumes of aggressive solution trapped between the crevice faces. This is a complex matter because it involves the effects of surface finish and roughness, washer and gasket materials, crevice construction and many other factors. An integral part of this programme therefore was to develop the relationship between actual crevices and their significance to crevice corrosion. This is the major step required to develop improved test methods that can relate directly to service performance.
The technique described by Oldfield & Sutton2 has been used for many years as a reliable method of accurately ranking alloys. The method will, for example, respond to small differences in alloy content such as S and Mn3. With the quantification of crevice gaps, the model is also capable of predicting service performance of real components.
