Pitting Corrosion in Tool and Mold Steel

Why a pit keeps growing once it has started

Every stainless and tool steel that resists corrosion does so through a passive film, a natural oxide layer only nanometres thick. The film is not a dead coating. It carries an extremely high electric field across that thickness and it reacts continuously with the environment, which is why a small local failure can run away. Chloride is usually the species that triggers the failure, and it is a difficult one to keep out of a mold cooling circuit, a washing line or a coastal warehouse. It is small, it diffuses quickly, it interferes with passivation, and it is present nearly everywhere in small amounts.

Once a pit has opened, it maintains itself. The reactions that make up the corrosion separate in space, with the anodic dissolution happening inside the pit and the cathodic oxygen reduction happening on the bright surface outside it, where oxygen is plentiful. That leaves the pit solution short of oxygen, loaded with metal cations, and acidified by the hydrolysis of those cations, and the charge balance pulls more chloride in from the bulk solution. The result is a small pocket of hot, acidic chloride that keeps attacking the metal and prevents the film from repairing itself. The source calls this an autocatalytic process, and it is the reason a pit that starts small does not heal on its own. The same mechanism inside any occluded geometry is called crevice corrosion, which is covered on crevice corrosion in mold steel.

The four stages of pitting as the source sets them out. The cover over a growing pit stays reflective and is the reason a pit is easy to miss until the surface is agitated. Compiled from the pitting stages section of the article.

For a failure investigation the cover is the item that matters. A growing pit is frequently closed over by a remnant of undermined film or by a thick layer of corrosion product, the cover stays reflective, and the pit underneath is far wider than the visible opening suggests. A short exposure to ultrasonic agitation in a cleaning bath removes the cover and reveals the whole diameter, which is why the surface should be agitated before a pit count is recorded, and why a part that shows a scatter of small dark spots is worth examining further rather than brushing off.

The potentials that decide whether a pit starts

Pitting is not a question of whether the steel has a passive film. It is a question of where the potential of the part sits relative to two characteristic values that the environment sets. Stable pits form at potentials above the pitting potential, and they keep growing at potentials above the repassivation potential, which is always the lower of the two. The practical reading of a cyclic polarisation curve is that the difference between the corrosion potential and the pitting potential is the margin of safety, and the difference between the corrosion potential and the repassivation potential is a more conservative version of the same margin, because pits that have already formed can keep going in a range where new ones cannot start.

The characteristic potentials the source works with, including the alternative names used in the same field. Source, the potential section of the article.

Two cautions belong with that table. The characteristic potentials are not material properties. They are empirical values that shift with the scan rate and with the technique, and the potentiodynamically measured pitting potential of many materials scatters over hundreds of millivolts between runs. They are useful for ranking nominally identical tests, not for absolute predictions. The second caution is that the numbers depend on the surface state of the specimen, so data collected on a specially prepared laboratory coupon should not be carried across to a real part without allowing for it.

What makes one steel pit and another not

The composition of the steel is the first order control, and chromium is the dominant element. Chromium confers passivity on iron, and the pitting potential rises sharply once chromium passes the critical level of about 13% that makes the steel stainless in the first place. Nickel helps moderately. Molybdenum is very effective, but only where chromium is also present, which is the reason a molybdenum bearing grade cannot be judged on its molybdenum content alone. Nitrogen and tungsten have a strong influence at small additions, and the industry condenses the whole effect into one number. The pitting resistance equivalent number weights each element by its practical influence, with nitrogen given a multiplier that has been reported as high as 30 in some versions.

Where the pit actually starts is on the second order control, and it is always a heterogeneity. Pits almost always initiate at an inclusion, a second phase particle, a solute segregated grain boundary, a flaw, a mechanical defect or a dislocation, and in stainless steel they are most often associated with manganese sulfide inclusions, which are present in most commercial steels. The mechanism is not that the inclusion itself dissolves. It is that the matrix around it is depleted in the element that carries the passivity, so the depleted zone has a lower pitting potential and pits at that point first. This is why lower sulfur grades are produced specifically to improve pitting resistance, and why a cleaner steel can outperform a higher alloyed one in the same chloride environment. The cleanliness measurements that support that purchasing decision are described on inclusion and grain size measurement and tool steel quality verification, and the remelted routes that reduce the inclusion population are on ESR tool steel.

The factors the source identifies as deciding whether a stainless steel pits, with the direction of each effect. Compiled from the alloy composition, temperature and surface condition sections of the article.

Temperature deserves its own line because of how the transition behaves. Many materials simply do not pit below a certain temperature, and the change at that temperature is sharp and reproducible. Determining the critical pitting temperature is done by holding a fixed potential while the temperature is programmed upward and watching the current needed to hold that potential, or by sweeping the potential at each of a series of fixed temperatures. The value turns out to be largely independent of the applied potential and of the environmental variables over a wide range, which is what makes it a usable ranking number. If the concern is a gasketed joint or a threaded connection rather than an open surface, the same experiment is run with a creviced specimen and reports a critical crevice temperature, which sits below the pitting temperature for the same alloy.

Surface condition is part of the material

The source is blunt about how much the finishing operation matters. A rougher surface pits more readily and pits at a lower potential, and it gives a measured example. Type 302 stainless steel with a 120 grit finish pitted at a potential roughly 150 mV below the same steel with a 1200 grit finish, across a range of chloride concentrations, and the reason given is geometric. A rough surface has more occluded sites, each with a longer diffusion path and a slower rate of diffusion, so those sites can hold the conditions needed for active dissolution at a lower current density and therefore at a lower applied potential.

The same reasoning applies to the heat treatment and to the operations that follow it. Heat treatment in air leaves a chromium oxide scale sitting on top of a chromium depleted layer, and if the scale is taken off mechanically and the depleted layer is not removed chemically, the part is left with exactly the layer that pits first. Grinding and abrasive blasting are both reported as detrimental, while pickling in nitric and hydrofluoric acid removes the depleted layer and passivation in nitric acid increases the chromium content of the surface oxide. Heat tint from welding, embedded iron particles from machining, and exposed manganese sulfide inclusions are the other common defects, and all three are reduced by the same nitric acid passivation step. The pickling and passivation steps are covered from the coating side on PVD and CVD coatings for tool steel and the tool steel nitriding guide, and the general surface requirement for a finished die is on polishing and texturing plastic mold steel.

How pitting resistance is measured

The standard accelerated test uses ferric chloride, and the reason is that a similar electrolyte chemistry develops inside a real pit in a susceptible ferrous alloy, so the test attacks the material with something close to the environment that would form in service. That sameness is also the test’s limit. Ferric chloride is a chloride environment with the ferric and ferrous ion couple supplying an elevated potential, and results on an alloy intended for chloride free service can therefore mislead. The methods differ in whether a crevice former is fitted and in whether the outcome is a pass or fail ranking or a critical temperature.

Accelerated tests for pitting and crevice corrosion as the source describes them in the evaluating article. These tests rank materials and confirm a condition. They are not a prediction of time to failure in service.

Two of those methods are worth separating for a purchasing discussion. A pass or fail ranking in ferric chloride answers whether a heat of steel is at least as good as a reference material, which is the right question for an incoming inspection. A repassivation potential measured by the electrochemical route answers a design question, because the source reports that when the working potential of the part is held below the measured repassivation potential, no initiation occurred out to at least three years in the test it cites. That distinction is the difference between accepting a delivery and setting the operating envelope of a cooling circuit.

What this means on a tool steel order

Four things follow. Specify cleanliness where the part sees chlorides, because the initiation sites are inclusions and not the matrix. Match the grade to the environment rather than to the hardness, and treat molybdenum as useful only in a steel that also has chromium. Write the finishing sequence into the order, since a rough ground surface and a mechanically cleaned heat treat scale both remove resistance that the analysis paid for. And where the geometry includes a gasket, a lap, a thread or a stagnant water passage, expect the crevice version of the same attack at a lower temperature than the pitting data would suggest. The grades Aobo supplies are charted on the stainless steel corrosion resistance chart, the martensitic family is described on the martensitic stainless properties chart and 440C stainless steel, and the mold side is on plastic mold steels and the mold steel property ranking chart.

Source: ASM Handbook, Volume 13A, Corrosion: Fundamentals, Testing, and Protection, ASM International, 2003.