There is a group of recurring questions asked of The Heat Treat Doctor, all of which are centered around: “How much surface oxidation is allowable on a steel part?” The latest was an innocent enough request for “a ‘decarburization’ chart (or article), which shows the effect on steel, if normalized, in a non-inert-gas environment.” The fulfillment of this request provides some valuable information for the heat treater. Let’s learn more.

While the question appears straightforward, there are a number of important issues that arise from it, and some interpretation is needed to fully understand what is being asked. For example, what type of steel is involved, in what form is it being purchased and what is the manufacturing sequence used to create the component parts? From a metallurgical perspective, are we talking about total or partial decarburization? How will it be measured, and what is the application end use of the product so that an evaluation can be made of the impact of decarburization on the design?

Decarburization varies with material grade, hardenability, furnace atmosphere, carbon potential and the type of heat-treatment process being performed (e.g., temperature, time). Decarburized parts exhibit lower (surface) hardness, reduced wear resistance and lower fatigue life, which affects their service life. One example is an automotive manufacturer whose steering columns were found to be loosening on the assembly line due to the presence of a 0.025- to 0.075-mm (0.001- to 0.003-inch) partial decarburization layer on the shaft retaining ring.

If one simply refers to AMS 2759/1 or /2,[2,3] four part types are defined that, in turn, dictate the class of atmosphere permitted or prohibited when heating parts above 677°C (1250°F). The type of starting surface (e.g., hot-finished, cold-drawn) and the amount of metal to be removed by partial or finished machining, whether greater or less than 0.51 mm (0.020 inch), are the important criteria. These specifications go further by discussing surface contamination and providing limits on such items as partial decarburization, ≤0.13 mm (0.005 inch), and intergranular attack, ≤0.018 mm (0.0007 inch), as well as defining the method of measurement and rejection criteria.

Being told what to do is one thing, but understanding why it must be done is another, which was the real question being asked here. 

A Little Theory[4,5]

Total decarburization (aka type-1 decarburization) is the depth to which the surface microstructure is free-ferrite; that is, the depth to which there has been a 100% loss of carbon (Fig. 1). Partial decarburization (aka type-2 or type-3 decarburization) is that depth from the surface where some loss (greater than or less than 50% respectively) of carbon has occurred, but there is no measurable depth of complete decarburization.

The loss of carbon from the near surface has been found to occur above 700°C (1290°F) when the furnace atmosphere contains carbon dioxide, water vapor, oxygen and hydrogen (Fig. 2).

Carbon present in the steel will interact with the furnace atmosphere, and it will leave the surface in a gaseous phase under the right conditions. This results in a change in concentration causing migration of carbon from the interior to the surface, which continues until a balance has been re-established, thus creating a maximum depth of decarburization. Depending on whether the steel is between the lower (Ac1) and upper critical (Ac3) temperatures or above the upper critical, carbon diffusion rates vary. Temperature and composition are the principal factors involved, and their influence varies depending on the process (e.g., annealing, normalizing) being performed.

A Little Deeper[7]

As stated, decarburization begins to occur as the rate of carbon at the surface decreases due to its reaction with oxygen as this reaction exceeds the growth rate of scale (iron-oxide) formation. The scale contributes to the decarburization depth (e.g., Equation 1 being typical).

FeO + CFe → Fe + CO(g)             (1)

Decarburization is not only dependent on the presence of oxygen but also the interaction with other oxidizing gases in the atmosphere, most notably water vapor and carbon dioxide. In a controlled trial,[7] samples were decarburized and results measured for both low oxygen-level atmospheres and air (Table 1). The loss of hardness was carefully documented (Fig. 3). These tests found the maximum decarburization depth to be around 0.51 mm (0.020 inch), which correlates well with the AMS information.


The choice of the correct furnace atmosphere, temperature and to a lesser extent time at temperature are important variables to offset the effect of decarburization.

As it turned out in this particular instance, the question involved the need for either nitrogen as a blanketing atmosphere during normalizing at 955°C (1750°F) or if the process could be done in a gas-fired box furnace with the work protected by the products of combustion. Since the minimum stock removal was 1 mm (0.039 inch) and the soak times were four hours or less, the deleterious effects of decarburization could be avoided with either atmosphere. Conducting trials using a furnace running the process cycle(s), however, is always strongly recommended to determine the actual decarburization levels.



  1. Herring, Daniel H., Atmosphere Heat Treatment, Volume II, BNP Media, 2015
  2. AMS 2759/1, “Heat Treatment of Carbon and Low-Alloy Steel Parts Minimum Tensile Strength Below 220 ksi (1517 MPa),” SAE International, (Rev. E), 2009
  3. AMS 2759/2, “Heat Treatment of Carbon and Low-Alloy Steel Parts Minimum Tensile Strength 220 ksi (1517 MPa) and Higher,” SAE International, (Rev. F), 2015
  4. VanderVoort, George F., “Understanding and Measuring Decarburization,” Advanced Materials & Processes, February 2015
  5. SGS MSi (
  6. “Furnace Atmospheres No. 8: Sintering of Steels,” Special Edition Booklet, Linde 2011
  7. Mayott, Steven W., “Analysis of the Effects of Reduced Oxygen Atmospheres on the Decarburization Depths of 300M Alloy Steel,” Thesis, Rensselaer Polytechnic Institute, 2010