Decarburization and scale in knife forging
Mechanism, depth, and how carbon loss ruins edge stability
By Yashar Mousavand
What this page does
This is a technical reference on two surface damage mechanisms that quietly destroy blade performance: decarburization (carbon loss) and oxide scale (iron oxidation). Both are driven by the hot atmosphere around the steel. Both can happen during forging, normalizing, and heat treatment. And both target the exact material you later sharpen into. The goal here is to explain what is happening at the surface, how deep it goes, how to detect it, and which fixes are real.
Key takeaways
- Scale and decarburization are different. Scale is iron oxide that consumes metal. Decarburization is carbon leaving the steel as CO or CO2, producing a carbon gradient under the surface.
- Edge stability is controlled by the first fraction of a millimeter. A thin decarburized skin can dominate edge behavior because the edge is thin and sharpening exposes that skin again and again.
- Decarburization depth is diffusion controlled. It increases roughly with the square root of time, but the diffusion coefficient increases exponentially with temperature, so a small temperature increase can multiply the damage.
- Oxide scale growth is often diffusion controlled and follows parabolic kinetics (thickness proportional to square root of time). In forging temperature ranges, scale commonly forms as layered hematite, magnetite, and wustite, with wustite dominating at high temperature.
- Decarburization is not ‘fixed’ by later heat treating. Once carbon is gone at the surface, the only recovery is to remove the damaged layer or prevent it from forming.
1. Why decarburization ruins edge stability
A bushcraft knife fails at the edge, not in the middle of the blade. The edge is a high-stress contact zone where the steel must resist three things at once: plastic deformation (rolling), fracture (microchipping), and abrasive wear (loss of bite). Decarburization attacks all three by changing the near-surface microstructure and lowering the maximum hardness that can be achieved after quenching.
If carbon is depleted near the surface, the steel in that region transforms differently during heat treatment. Instead of a fully hardened structure, you can end up with a ferrite-rich layer (complete decarburization) or a lower carbon microstructure (partial decarburization). In medium and high carbon steels, complete decarburization is often called ‘free ferrite’ because the surface becomes ferrite in the absence of carbon. This layer is readily identifiable in metallography and corresponds to a large drop in near-surface mechanical properties.
Why this matters in the field
A decarburized edge often feels sharp at first but quickly loses bite. It may roll during carving, mushroom during batoning, or develop a ragged apex after contact with hard knots. Sharpening may not solve it because you can stay inside the decarburized zone for many sharpenings if the damage is deep.
2. Scale versus decarburization: do not confuse them
Scale is oxidation of iron. It consumes steel and forms oxides that can flake off or get pressed into the surface. Decarburization is carbon loss. You can have heavy scale with modest decarb, or little visible scale with significant decarb, depending on the atmosphere chemistry. In real forging shops both often occur together because the same oxidizing gases drive both reactions.
3. The chemistry that pulls carbon out of steel
Decarburization happens when the atmosphere at the steel surface has a carbon potential lower than the steel. Carbon at the surface reacts into gas and the remaining steel becomes carbon depleted. Thermal Processing describes decarburization as carbon leaving steel in a gaseous state when atmosphere equilibrium is not maintained. In controlled atmosphere heat treating, key equilibria include carbon in steel reacting with CO2 and H2O to produce CO and H2.
Two commonly referenced decarburizing reactions are:
C(Fe) + CO2(g) <-> 2CO(g)
C(Fe) + H2O(g) <-> CO(g) + H2(g)
These reactions are reversible and shift with gas composition and temperature.
3.1 Why water vapor is especially aggressive
Water vapor is strongly decarburizing. In industrial furnace control, it is tracked through dew point because the H2O/H2 ratio is coupled to the CO2/CO ratio through the water-gas reaction:
CO2 + H2 <-> CO + H2O
Herring explains that water vapor and CO2 both oxidize and decarburize steel and that preventing decarburization requires control of these constituents.
Why this matters for gas forges
A forge can look ‘hot’ and still be chemically decarburizing. If the flame is oxygen rich or water vapor rich, carbon potential can be low even while the work is at forging temperature. This is why judging only by steel color or scale amount can mislead you.
4. How deep does decarburization go
Once the surface carbon potential is low, the surface carbon content drops. Then carbon from the interior diffuses toward the surface down the concentration gradient. That transport is diffusion controlled. Mackenzie (Thermal Processing) models decarburization using Fick’s Second Law and provides the classic error-function solution used for carburizing and decarburizing cases.
For a semi-infinite solid with constant surface carbon concentration Cs and uniform initial carbon C0, one form of the solution is:
(C(x,t) – C0) / (Cs – C0) = 1 – erf( x / (2 * sqrt(D * t)) )
For a strongly decarburizing surface where Cs approaches 0, this reduces to C(x,t) / C0 = erf( x / (2 * sqrt(D * t)) ).
4.1 Diffusion coefficient is the real amplifier
The key danger is not that depth grows with sqrt(time). The key danger is that D increases rapidly with temperature. In the same Thermal Processing column, a temperature dependent diffusion coefficient for carbon in austenite is given in Arrhenius form:
D = 0.162 * exp( -137800 / (R*T) ) (D in cm^2/s)
This means the same extra minutes at higher temperature can cause much deeper carbon loss than you expect.
4.2 A worked example using published values
Mackenzie includes an example at 1600°F (871°C) with a diffusion coefficient D = 8.222 x 10^-8 cm^2/s (8.222 x 10^-12 m^2/s). If you hold steel at that temperature for 1 hour under a fully decarburizing surface condition (Cs = 0), then the model predicts:
Depth to 50% of base carbon (C/C0 = 0.50): about 0.16 mm
Depth to 90% of base carbon (C/C0 = 0.90): about 0.40 mm
These are model depths defined by a chosen carbon threshold, not a single sharp boundary. Real blades can show deeper effective damage due to alloying, grain boundaries, and atmosphere changes.
Figure 1 shows how the carbon gradient moves inward with time at constant temperature in this example.
Why this matters for knife edges
A tenth of a millimeter matters at the apex. If you have a decarburized zone hundreds of microns deep, then even careful sharpening can keep you inside it. That is why a knife can refuse to hold an edge even when the core steel is excellent.
5. What decarburization looks like in microstructure
Industrial practice separates decarburization into complete, partial, and effective depths. Rothleutner (Thermal Processing) explains that complete decarburization is also called free ferrite because a ferrite layer forms at the surface in the absence of carbon. Partial decarburization is the region with a measurable carbon gradient without fully carbon-free ferrite. Effective decarburization is the depth at which the carbon loss reduces properties below what the component requires, and it can be defined by hardness traverses if hardness correlates with the required properties.
For blades, ‘effective’ is the only definition that matters. If the near-edge layer cannot reach the hardness and microstructure you designed for, the knife behaves like a softer steel no matter what the spine hardness reads.
Figure 2 illustrates a typical hardness drop near the surface when a decarburized layer is present.
6. The mechanics: why a carbon gradient destabilizes the apex
6.1 Lower carbon means lower achievable hardness and different transformation products
Hardness after quench is strongly influenced by carbon content because carbon controls martensite strength. If the surface carbon is reduced enough, the surface may not form the same hardened structure as the interior. In extreme cases the surface becomes ferrite-rich (free ferrite), which is dramatically softer than tempered martensite. Even partial decarburization shifts the edge toward lower hardness and lower abrasion resistance.
6.2 A gradient creates a weak layer that concentrates strain
A decarburized skin is a low-strength layer bonded to a higher-strength core. Under cutting, the surface sees high shear and compressive stresses. The softer layer yields first, which rounds the apex and promotes rolling. If the edge is very thin, the strain mismatch between surface and core can also promote microcrack initiation at inclusions or pits, especially if scale damage is present.
6.3 Decarb hides inside sharpening and ‘mystery soft edges’
A common failure pattern is this: the knife sharpens easily, feels sharp, then loses bite quickly and develops a rolled wire-like apex. That behavior matches a low hardness surface layer. The core can still measure high hardness if tested away from the edge. Without a cross-section check, makers and users may blame the steel grade or the quench even when the real cause is surface carbon loss.
7. Scale: the oxide crust that consumes steel and seeds defects
At forging temperatures, iron oxidation is fast. The oxide crust is not a single compound; it is typically layered. Shimadzu describes the classic three-layer structure in hot-rolled steel: outer hematite (Fe2O3), middle magnetite (Fe3O4), and inner wustite (FeO). They note that around 1000°C the FeO layer is the dominant portion of the scale.
ISIJ International reports that in early stages of oxidation in air, wustite was the predominant phase on the surface of steel over the temperature range 800 to 1200°C, and that magnetite and hematite fractions increased as oxidation proceeded.
7.1 Scale growth kinetics: why it speeds up with heat and time
When oxide growth is controlled by diffusion through the scale, thickening often follows a parabolic rate law: x^2 is proportional to time. A short derivation from Washington University notes that parabolic kinetics indicates diffusion of reactants through a growing oxide scale is rate determining and leads to x^2 = 2*k’*t. University lecture notes on the Wagner model explain that oxidation involves ionic and electronic diffusion through the oxide film, producing parabolic growth under diffusion control.
Figure 3 shows the parabolic trend in a schematic way.
7.2 How scale interacts with decarburization
Scale and decarburization are coupled because both are driven by oxidizing species at the surface. A thick scale layer is not a protective coating in a forge. It cracks, becomes porous, and can spall. Those defects provide pathways for oxygen, CO2, and water vapor to reach the steel and continue pulling carbon out. Scale also increases surface roughness, which makes later grinding remove more metal and can expose deeper decarburized layers.
7.3 Scale damage becomes edge defects
If scale is hammered into the surface during forging, it can become a non-metallic inclusion. Later grinding can open that inclusion as a pit. At an edge, pits behave as stress concentrators. Under impact or hard knot contact, they raise local stress and promote microchipping. This is one reason that a blade can show ‘random’ microchips even when hardness is reasonable.
8. Detecting and measuring decarburization in a blade shop
8.1 Visual cues are weak
Heavy scale tells you the atmosphere is oxidizing, but it does not tell you the carbon gradient depth. A clean surface does not prove the absence of decarburization either, because carbon loss can occur with modest visible oxide depending on gas chemistry.
8.2 Metallography and hardness traverses are the technical answer
ASTM E1077 covers procedures for estimating the depth of decarburization, ranging from screening methods to microscopical methods, microindentation hardness methods, and chemical analysis methods. For blades, the most useful approach is often a polished cross-section plus a microhardness traverse to define an ‘effective’ depth where hardness meets your requirement.
8.3 Practical shop workflow: a sacrificial coupon beats guessing
If you need certainty, forge and heat treat a small coupon from the same bar and run it through the same thermal history as the blade. Then polish and etch a cross-section. If you have access to microhardness, run a traverse. This gives you real depth and lets you set grinding allowance based on evidence rather than habit.
9. Reducing decarburization and scale in real forging workflows
This section is not about secret tricks. The physics is simple: decarburization needs time, temperature, and an atmosphere that is decarburizing. Control means removing at least one of those inputs.
- Minimize time above critical temperatures. The diffusion coefficient rises fast with temperature, so long soaks at high heat are expensive in carbon loss.
- Avoid oxygen rich and wet atmospheres. CO2 and especially H2O drive decarburization through reversible surface reactions. Dew point and gas ratios are why some forges scale aggressively.
- Use barriers when practical. In controlled furnaces, foil wrap or protective atmospheres reduce oxidation and decarb. In open forging this is harder, but muffles and controlled atmosphere boxes can help when used correctly.
- Plan grinding allowance deliberately. If you forge near final thickness and leave no material to remove, you force the edge to live inside whatever decarb layer forms.
- Separate operations: do the most oxidizing work early and remove scale and decarb before final austenitize when possible. Final heat treat should start with a clean, low-damage surface.
Why this matters for bushcraft knives
Bushcraft knives see repeated field sharpening and lateral loading. A knife that loses bite quickly in wood is often suffering from a near-edge property problem, not a core steel problem. Reducing surface damage is one of the simplest ways to make performance more repeatable across knives of the same steel.
10. What to do when decarb already happened
There are only two real options once carbon is depleted at the surface: remove the damaged layer or accept the performance loss. Thermal cycling does not add carbon back. A new quench does not add carbon back. The carbon must come from the steel itself, and diffusion during decarb moved it out.
Practically, removal means grinding past the effective decarburization depth. Because depth varies with time, temperature, and local geometry, a coupon and a cross-section check is the fastest way to stop guessing.
FAQ
- Does heavy forge scale prove I have deep decarburization?
No. Scale indicates oxidation of iron. Decarburization is carbon loss. They often occur together, but depth depends on atmosphere chemistry and time at temperature. Use a cross-section to know.
- Can heat treating fix a decarburized edge without grinding?
No. Once carbon has left the surface, later heat treatment can only transform what carbon remains. Effective recovery requires removing the depleted layer or preventing it from forming.
- How much material should I grind off to remove decarb?
There is no universal number because diffusion depth depends strongly on temperature and time. The technical approach is to run a coupon through the same cycle and measure effective depth by etch plus hardness traverse (ASTM E1077 style methods).
About the author
Yashar Mousavand is a bushcraft and survival instructor and the founder of Yashar Survival Academy. He teaches practical field skills and writes evidence based guides on knives, gear, and outdoor technique, focused on what reliably works in real conditions.
References
Primary sources used for equations, definitions, and oxide scale structure:
R1: D. Scott Mackenzie, Ph.D., FASM. ‘Calculating decarburization’ (Thermal Processing, July 2023). Includes Fick’s law solution, example diffusion coefficient and Arrhenius expression for carbon diffusion in austenite, and notes FeO predominance at typical heat treating temperatures.
https://thermalprocessing.com/wp-content/uploads/2023/07/0723-HS.pdf
R2: Lee Rothleutner. ‘Understanding decarburization’s fundamentals is vital to product performance’ (Thermal Processing, May 2019). Defines decarburization in heat treating, lists key reversible reactions with CO2 and H2O, and explains complete, partial, and effective decarburization and their identification.
https://thermalprocessing.com/wp-content/uploads/2019/05/0519-MU.pdf
R3: Daniel H. Herring. ‘Considerations in Heat Treatment’ (Industrial Heating, Oct 2009, Heat Treat Doctor PDF). Discusses oxidation and decarburization by oxygen, water vapor and CO2, and the water-gas reaction used for atmosphere control.
https://heat-treat-doctor.com/documents/considerations%20in%20ht1.pdf
R4: ASTM E1077-14 (Reapproved 2021). Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens. Covers screening, microscopical, microindentation hardness, and chemical analysis methods.
https://www.astm.org/e1077-14r21.html
R5: ASTM E1077-14 (2021) PDF copy. Useful for understanding the scope and method categories when access to ASTM pages is restricted.
https://img.antpedia.com/standard/files/pdfs_ora/20230612/astm/E/E%201077%20-%2014%20%282021%29.pdf
R6: Shimadzu. ‘State Analysis of Iron Oxide Scale’ application note. Describes three-layer scale structure: hematite, magnetite, wustite, and notes FeO dominance around 1000°C.
https://www.shimadzu.com/an/sites/shimadzu.com.an/files/pim/pim_document_file/applications/application_note/13976/an_p116-en.pdf
R7: ISIJ International (2004) ‘Growth Rate and Phase Composition of Oxide Scales during Hot Rolling of Low Carbon Steel’ (J-STAGE PDF). Reports wustite predominance in early oxidation over 800 to 1200°C and describes how oxide composition evolves with time and temperature.
https://www.jstage.jst.go.jp/article/isijinternational1989/44/9/44_9_1554/_pdf
R8: G. Visscher (Washington University). ‘Derivation of the Parabolic Rate Law’ PDF. Explains why diffusion through oxide scale produces parabolic kinetics and derives x^2 proportional to time.
https://epsc.wustl.edu/~visscher/research/paraboliclaw.pdf
R9: University of Utah. ‘Lecture 29: Kinetics of Oxidation of Metals: Part 2: Wagner Parabolic Model’ PDF. Explains coupled ionic and electronic diffusion in oxide films and parabolic growth behavior.
https://my.eng.utah.edu/~lzang/images/lecture-29.pdf
R10: Oxide-Scale Structures Formed on Commercial Hot-Rolled Steel Strip and Their Formation Mechanisms (Springer). Notes initial oxide scale commonly shows three-layer hematite, magnetite, wustite structure.
https://link.springer.com/article/10.1023/A:1010395419981
