Chapter 18 - Pitting
Robert G. Kelly, Assistant professor, Department of Materials Science and Engineering, University of Virginia, Charlottesville, VA 22903.
Pitting corrosion occurs when discrete areas of a material undergo rapid attack while the vast majority of the surface remains virtually unaffected. This is in sharp contrast to uniform corrosion in which all parts of the exposed surface recede at approximately the same rate. Essentially all metals and alloys undergo pitting corrosion under some set of experimental conditions, though the relative susceptibility varies widely. The basic requirement for pitting is the existence of a passive state for the material in the environment of interest. Pitting of a given material depends strongly upon the presence of an aggressive species in the environment and a sufficiently oxidizing potential (e.g., Cl ion in neutral, aerated aqueous solution for Type 304 stainless steel).
Testing for localized corrosion resistance usually aims at meeting one (or more) of four general goals: (a) alloy ranking for selection or development, (b) failure analysis, (c) determination of the effects of changes in process parameters, and (d) prediction of penetration rates. Of the four, the last is by far the most difficult to accomplish, but also the most important in many cases.
Coupon Testing
The exposure of test coupons to the corrosive solutions to evaluate localized corrosion resistance has a long and successful history. The materials of interest are typically machined into coupon form before being measured, cleaned, weighed, and exposed to the corrosive medium. After a set period of time, the coupons are removed, the surfaces are carefully cleaned and evaluated for attack. ASTM G 1, Practice for Preparing, Cleaning, and Evaluating Corrosion Test Specimens, describes the details of the preparation of test coupons.
Considerations in coupon testing for pitting are surface finish, evaluation methods, and choice of environment. The pitting resistance of a material usually increases with a finer surface finish. Thus, it is especially important in coupon testing for pitting corrosion that the surface preparation method be chosen carefully and applied uniformly to all specimens. If the goal of the testing is material selection for a component, then the surface finish of the actual structure or equipment that is under consideration should be duplicated as closely as possible. In many cases, this is not possible for reasons of time, cost, or variability. Under such circumstances, the use of a 120 grit finish has become standard practice. Such a finish is easily produced using abrasive papers constantly wetted with water to avoid heat-induced metallurgical changes in the coupon.
Whereas mass loss characterizes the rate of uniform corrosion well, such measurements can be extremely misleading for pitting corrosion. Badly pitted specimens can exhibit negligible weight loss if the attack is extremely localized. In fact, the more localized is a given amount of attack, the more severe the pitting problem. ASTM G 46, Practice for the Examination and Evaluation of Pitting Corrosion, describes different methodologies for evaluating pitting corrosion attack. Visual attack of both cross-sections and plan views of pitted specimens can be ranked as shown in the standard. Whereas pit density (number/cm2) is important, pit depth measurement is often the most directly applicable measurement. Values for both average pit size and for the deepest pit observed are useful. To quantitatively evaluate the depth of attack, a variety of measurementmethods can be employed including depth gages, metallographic examination of cross-sections, or change in the focus plane on a metallurgical microscope depending upon the size of the pits. The pitting factor (PF) is the ratio of the average of the ten deepest observed pits to the average metal penetration by uniform dissolution, with higher values indicating a greater susceptibility to pitting. Statistical methods have also been applied to the evaluation of pitting corrosion (see ASTM G 16, Guide for Applying Statistics to Analyses of Corrosion Data).
ASTM G 48, Test Method for Pitting and Crevice Corrosion Resistance of Stainless Steels and Related Alloys by the Use of Ferric Chloride Solution, involves exposure of the material to a highly oxidizing, highly acidic, concentrated metal chloride solution. In essence, it is an attempt to simulate very roughly the composition of the environment within a localized corrosion site in a stainless steel. Briefly, a material is exposed to a 10 wt% FeCl, solution for a relatively short time (24 to 72 h) at either ambient or elevated (usually 500C) temperature. At the end of the test, the sample is examined for weight loss and localized attack (see ASTM G 46 for a recommended practice for evaluating the extent of pitting corrosion). If pitting and crevice corrosion are of concern, artificial crevices can be applied by the use of Teflon® blocks held tightly to the sample surface by rubber bands.
Cyclic Polarization
By far the most common electrochemical test for localized corrosion resistance is cyclic polarization. ASTM G 61, Test Method for Conducting Cyclic Potentiodynamic Polarization Measurements for Localized Corrosion Susceptibility of Iron-, Nickel-, or Cobalt-based Alloys, has been developed to allow experimenters to test their equipment and procedures on systems that are well characterized. The potential is in the anodic direction until localized corrosion initiates as indicated by a large increase in the applied current. At this point, the direction of the scan is reversed, and the current decreases until it changes polarity.
Conventional wisdom states that Ebd (the breakdown potential) is the potential above which pits are initiated, while Erp(the repassivation potential) is the potential below which pits repassivate. The breakdown potential is usually defined as the potential at which there is a large increase in the applied current, while the repassivation potential is the potential on the reverse scan at which the applied anodic current becomes zero (i.e., the current changes polarity). Thus, the higher the value of Ebd, the more resistant is the alloy to the initiation of localized attack. The higher Erp, the more easily the alloy can repassivate. At potentials between Erp and Ebd, sites that have initiated can propagate.
There have also been numerous attempts to use the amount of hysteresis in the cyclic scan as a measure of localized corrosion susceptibility, with varying degrees of success. In this approach, the larger the hysteresis, the more likely a localized corrosion site will propagate once initiated.
While the interpretation of breakdown and repassivation potentials remains controversial, progress towards a consensus is being made. The large scatter in the breakdown potential and its dependence on scan rate are thought to be due to the sensitivity of pit initiation to the initial conditions, and to the time dependence of the localized corrosion site chemistry dependence, respectively. In addition, it is now generally accepted that corrosion-resistant designs should not use E„ as the important parameter, but the appropriate value of Erp since this is the potential below which pits should repassivate.
Predictive Capabilities
The study of metastable pitting can be used to assist in lifetime prediction studies. If pitting is truly a stochastic phenomenon, then one can apply statistics to allow prediction of the likelihood of pit propagation, given sufficient information. The information needed is (a) the probability that a pit will nucleate under a given set of conditions, and (b) the probability that once a pit nucleates, it will survive (i.e., become stable). Such information could be used to estimate component lifetimes, which could then be used to make design decisions based upon the consequences of a failure. For example, a 30% chance of perforation may be acceptable for an easily shut-down and repaired vessel if it allows a cheaper alloy to be used, but such a probability would not be acceptable for a critical component in an inaccessible submersible.