---
title: "Steel Decarburization: Causes, Measurement, and Control"
description: "Learn how steel decarburization forms, how to measure it, and how to control it."
canonical: https://steelequivalents.com/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control
language: en
robots: index, follow
og_type: article
type: wiki-article
language_alternates:
  en: https://steelequivalents.com/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control.md
  cs: https://steelequivalents.com/cs/wiki/tepelné-zpracování/dekarbonizace-oceli-příčiny-měření-a-kontrola.md
  de: https://steelequivalents.com/de/wiki/wärmebehandlung/entkohlung-von-stahl-ursachen-messung-und-kontrolle.md
  es: https://steelequivalents.com/es/wiki/tratamiento-térmico/descarburización-del-acero-causas-medición-y-control.md
  fr: https://steelequivalents.com/fr/wiki/traitement-thermique/décarburation-de-l-acier-causes-mesure-et-contrôle.md
  hu: https://steelequivalents.com/hu/wiki/hőkezelés/az-acél-dekarbonizációja-okai-mérése-és-szabályozása.md
  it: https://steelequivalents.com/it/wiki/trattamento-termico/decarburizzazione-dell-acciaio-cause-misurazione-e-controllo.md
  nl: https://steelequivalents.com/nl/wiki/warmtebehandeling/decarburisatie-van-staal-oorzaken-meting-en-beheersing.md
  pl: https://steelequivalents.com/pl/wiki/obróbka-cieplna/dekarburyzacja-stali-przyczyny-pomiar-i-kontrola.md
  pt: https://steelequivalents.com/pt/wiki/tratamento-térmico/descarbonização-do-aço-causas-medição-e-controlo.md
  ru: https://steelequivalents.com/ru/вики/термическая-обработка/обезуглероживание-стали-причины-измерение-и-контроль.md
  tr: https://steelequivalents.com/tr/wiki/ısıl-işlem/çeliğin-dekarbürizasyonu-nedenleri-ölçümü-ve-kontrolü.md
  zh: https://steelequivalents.com/zh/维基/热处理/钢材脱碳-原因-测量与控制.md
  ja: https://steelequivalents.com/ja/ウィキ/熱処理/鋼の脱炭-原因-測定-制御.md
  ar: https://steelequivalents.com/ar/ويكي/المعالجة-الحرارية/نزع-الكربنة-من-الفولاذ-الأسباب-والقياس-والتحكم.md
  ko: https://steelequivalents.com/ko/위키/열처리/강재-탈탄-원인-측정-및-제어.md
  pt-br: https://steelequivalents.com/pt-br/wiki/tratamento-térmico/descarbonetação-do-aço-causas-medição-e-controle.md
  hi: https://steelequivalents.com/hi/विकी/ऊष्मा-उपचार/इस्पात-का-डीकार्बराइजेशन-कारण-मापन-और-नियंत्रण.md
content_hash: sha256:2915b47fcd94f598d312960b6f04273747bd12b76eee175ebeda62e75bcd3367
generated: 2026-09-25
---

> **Machine-readable Markdown rendering.** This is a Markdown representation of a web page,
> served for AI agents and crawlers as a token-efficient alternative to the full HTML.
> Full interactive HTML version: https://steelequivalents.com/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control
> Canonical URL: https://steelequivalents.com/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control
> Available languages: en, cs, de, es, fr, hu, it, nl, pl, pt, ru, tr, zh, ja, ar, ko, pt-br, hi
> Change detection: `content_hash` is the SHA-256 of this document's Markdown body — everything
> below this block, whitespace-trimmed, excluding the front-matter. Compare it to detect a real
> content change; `generated` moves on every render and is not a change signal.

![Steel Decarburization: Causes, Measurement, and Control](/images/uploads/82d4c658-7cb1-4ba5-a309-424d38c59341/wiki-hero-a-polished-steel-cross-section-showing-a-decarburized-ferritic-surface-layer-gra-1920x823.jpg)

Heat Treatment

# Steel Decarburization: Causes, Measurement, and Control

Learn how steel decarburization forms, how to measure it, and how to control it.

![Portrait of Anders Bergström, steel industry reporter](/images/uploads/a7d8e5bc-7cdf-409b-913b-211e63e8c441/anders-bergstr-m-1920x1920.jpg)

 **[Anders Bergström](/news/author/anders-bergstrom "Anders Bergström")** Heat Treatment 60+ min read Updated Aug 14, 2026 Evidence-reviewed

  On this pageOn this page

- [What Steel Decarburization Is—and Is Not](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#what-steel-decarburization-is-and-is-not "What Steel Decarburization Is—and Is Not")
- [The Thermodynamic Driving Force: Carbon Activity at the Surface](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#the-thermodynamic-driving-force-carbon-activity-at-the-surface "The Thermodynamic Driving Force: Carbon Activity at the Surface")
- [Diffusion, Reaction, and the Formation of a Decarburized Layer](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#diffusion-reaction-and-the-formation-of-a-decarburized-layer "Diffusion, Reaction, and the Formation of a Decarburized Layer")
- [Temperature, Time, and the Non-Monotonic Decarburization Response](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#temperature-time-and-the-non-monotonic-decarburization-response "Temperature, Time, and the Non-Monotonic Decarburization Response")
- [Furnace Atmospheres and the Main Process Causes](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#furnace-atmospheres-and-the-main-process-causes "Furnace Atmospheres and the Main Process Causes")
- [How Decarburization Is Measured Under the Standards](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#how-decarburization-is-measured-under-the-standards "How Decarburization Is Measured Under the Standards")
- [Metallographic Measurement: What the Microscope Can and Cannot Show](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#metallographic-measurement-what-the-microscope-can-and-cannot-show "Metallographic Measurement: What the Microscope Can and Cannot Show")
- [Hardness and Chemical-Analysis Methods](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#hardness-and-chemical-analysis-methods "Hardness and Chemical-Analysis Methods")
- [Controlling Decarburization in Heat Treatment and Sintering](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#controlling-decarburization-in-heat-treatment-and-sintering "Controlling Decarburization in Heat Treatment and Sintering")
- [Correction, Removal, Acceptance, and Failure Consequences](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#correction-removal-acceptance-and-failure-consequences "Correction, Removal, Acceptance, and Failure Consequences")
- [A Standards-Based Investigation and Reporting Workflow](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#a-standards-based-investigation-and-reporting-workflow "A Standards-Based Investigation and Reporting Workflow")

## What Steel Decarburization Is—and Is Not

### Definition of decarburization and the decarburized zone

Steel decarburization is the near-surface loss of carbon, or of carbon-rich phases, from a steel product during heating, processing, or service. The affected region is the **decarburized zone**: a layer beneath the original or finished surface in which carbon content, carbon-bearing phase fraction, or both are lower than in the unaffected interior. NIST’s *Damage and Failure Modes of Structural Steel Components* describes this zone as near-surface loss of carbon and carbon-rich phases when surface carbon reacts with oxygen to form gaseous carbon monoxide (CO) or carbon dioxide (CO₂).

Decarburization results from coupled surface reaction and solid-state carbon diffusion rather than oxidation alone. Strong evidence

The usual setting is heat treatment, forging, rolling, sintering, or another high-temperature operation in which the steel surface contacts a gas capable of removing carbon. Oxygen-bearing furnace gases are common causes, but the reaction is controlled by more than the presence of oxygen alone. Carbon must diffuse from the interior toward the surface, surface reactions must proceed at a sufficient rate, and the furnace atmosphere must have a lower effective carbon activity than the steel. If the atmosphere has a higher carbon potential than the steel surface, carbon can enter the steel instead. The surface therefore acts as a reaction boundary coupled to solid-state diffusion.

ASM’s chapter *Steel Decarburization—Mechanisms, Models, Prevention, Correction, and Effects on Component Life* describes carbon diffusion through austenitic iron toward the surface as a central part of the process. At temperature, carbon moves through the austenite lattice while reactions at the surface consume carbon. The resulting concentration gradient can extend farther than the visibly altered surface. A dark scale layer, by itself, does not show how deep carbon loss extends.

Partial and total decarburization describe metallurgical conditions rather than universal depth categories.
| Condition | Meaning | Typical structural indication |
|---|---|---|
| Partial decarburization | Carbon or carbon-rich phase content is reduced but remains present | Reduced pearlite, carbide content, or altered phase fraction |
| Total decarburization | The expected carbon-rich constituent has been removed from the relevant layer | Carbon-poor or fully ferritic surface structure |

Metallurgists commonly distinguish **partial decarburization** from **total decarburization**. Partial decarburization means that some carbon remains in the affected layer, although its concentration or carbon-rich phase content is below that of the unaffected steel. Total decarburization means that the carbon-rich constituent expected under the steel’s composition and heat-treatment condition has been removed from the relevant layer, leaving a carbon-poor structure. These are metallurgical conditions, not universal depth categories. ISO 3887:2023, *Steels — Determination of the depth of decarburization*, specifies three methods for determining decarburization depth in steel products, but the meaning of a reported depth still depends on the specified criterion and method.

The transition from decarburized material to unaffected material may be sharp, gradual, or irregular. A hypoeutectoid steel may show a ferritic surface layer where pearlite has been reduced or eliminated. A higher-carbon steel may show reduced carbide content, altered pearlite, or a fully ferritic layer near the surface. Alloying elements also affect carbon activity, diffusion, phase stability, and the appearance of the transformed zone. For this reason, “decarburization depth” is not always a single obvious boundary visible at low magnification.

### Decarburization versus oxidation, scaling, carburization, and carbon depletion from processing

#### Related surface conditions

Oxidation

Reaction of steel constituents with oxygen, water vapor, or carbon dioxide.

Scaling

Formation and removal of iron-oxide products on the surface.

Carburization

Net entry of carbon into the steel surface.

Decarburization

Near-surface loss of carbon or carbon-rich phases.

Decarburization and oxidation often occur together, but they are not synonyms. **Oxidation** is the reaction of steel constituents with oxidizing species such as oxygen, water vapor, or carbon dioxide. Iron oxidation can produce an oxide film or scale, while carbon oxidation can produce CO or CO₂ and lower the carbon content beneath the surface. A furnace atmosphere can therefore oxidize the surface without producing a practically significant decarburized zone, or it can cause both effects at once.

**Scaling** refers primarily to the formation and removal of adherent or loose oxide products on the steel surface. Scale may contain iron oxides such as wüstite, magnetite, and hematite, depending on temperature and gas conditions. A thick scale layer does not establish the depth of carbon loss. Conversely, decarburization can occur under an atmosphere or coating that produces little visible scale. Cambridge’s *Atmosphere Control for the Protection of Metals During Production Processes* specifically links direct exposure to excess-oxygen combustion atmospheres with oxidation, scaling, and decarburization, while treating the phenomena as related but distinct outcomes.

Carbon activity **Carbon activity** A thermodynamic measure of carbon chemical potential relative to a defined reference state; it depends on concentration, temperature, phase, and alloying elements.

The chemical driving force is better expressed through **carbon activity** or furnace carbon potential than through a simple label such as “oxidizing atmosphere.” The carbon potential during steel heat treatment is governed by the difference in carbon activity between the steel and its surroundings, together with surface-reaction kinetics. ASM’s *Furnace Atmosphere Controls in Heat Treating* identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as furnace variables used to control carbon potential and prevent carbon loss. Eurotherm’s 2020 *2704CP Handbook* supplement describes common estimation of furnace carbon potential from a zirconia-probe oxygen measurement combined with furnace temperature and carbon-monoxide content, or from dew-point measurement.

#### Process limit

The less-than-5% hydrogen guideline applies only to the cited nitrogen/hydrogen sintering conditions. Do not use it as a general furnace limit without checking temperature, dew point, steel composition, geometry, gas flow, and residence time.

**Carburization** is the opposite net carbon-transfer condition: carbon enters the steel surface and raises its carbon activity or carbon concentration. A furnace atmosphere can carburize one steel grade and decarburize another if their starting carbon contents or surface conditions differ. Even in one furnace load, local gas flow, temperature, dew point, soot, leaks, and workpiece spacing can create different results. Linde’s 2023 *Sintering of Steels* guidance states that water vapor in nitrogen/hydrogen atmospheres causes decarburization and reports an experimental guideline that less than 5% hydrogen limited carburizing and decarburizing rates under the cited sintering conditions. That figure is not a universal furnace rule; gas composition, temperature, powder or solid-steel geometry, and residence time remain decisive.

Carbon can also be depleted by **processing without the classic furnace decarburization mechanism**. Mechanical removal, grinding burns, machining, pickling, melting losses, segregation, or dissolution of carbides during an unsuitable thermal cycle may change the carbon-bearing structure or the measured carbon near a surface. Carbon may redistribute during welding or additive processing, and a decarburized layer may be removed by machining after heat treatment. These conditions should not automatically be called decarburization. The term is most defensible when a near-surface carbon deficit is tied to carbon transfer, reaction, or selective loss during the process under examination.

In the cited spring-steel study, decarburized ferrite thickness showed a non-monotonic temperature response between 700 and 1000 °C. Strong evidence

Temperature alone does not predict the final depth. A 2021 *Journal of Materials Research and Technology* study of Fe-0.6C-1.8Si-0.8Mn spring steel tested between 700 and 1000 °C found that decarburized ferrite thickness increased up to approximately 800–825 °C and then decreased. Higher temperature accelerates carbon diffusion, but it can also alter oxide formation, gas reactions, phase constitution, and the duration of the relevant transformation. The observed maximum therefore reflects coupled diffusion and oxidation effects, not a monotonic temperature-only law.

### Why surface carbon loss matters to steel performance

Carbon controls the fraction and morphology of ferrite, pearlite, bainite, martensite, and carbides that develop during cooling. Removing carbon from the surface can produce a softer layer on a steel whose core was designed to harden. In a quenched component, the surface may fail to reach the intended hardness even though the interior meets the specification. In a spring, gear, bearing, shaft, or fastener, that difference changes load transfer and local deformation.

The mechanical penalty is often greatest at the surface because that is where contact stress, bending stress, friction, and crack initiation are concentrated. A ferritic decarburized layer can reduce wear resistance and contact-fatigue life. Under cyclic bending, the lower-strength layer may yield or develop slip earlier than the core. A surface carbon gradient can also change residual-stress development during quenching, while oxides and rough scale provide additional notches or crack-initiation sites. ASM links these effects to component life, prevention, correction, and modeling rather than treating decarburization as a cosmetic defect.

Different methods can report different decarburization depths from the same cross-section.
| Method | Primary observation | Main limitation |
|---|---|---|
| Metallography | Phase fraction, morphology, and visible transition | Boundary may be gradual or etch-dependent |
| Microindentation hardness | Surface-to-core mechanical response | Hardness also depends on phase, grain size, stress, and alloying |
| Chemical analysis | Carbon concentration or gradient | Spatial resolution depends on sampling volume and layer thickness |
| Screening | Rapid indication of a suspect zone | May not support close acceptance decisions |

Measurement must match the claim. ASTM E1077-14(2021), *Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens*, covers screening, microscopic examination, microindentation-hardness, and chemical-analysis procedures. Its referee approach for disputes is rigorous quantitative or lineal analysis. SAE J419\_201801, *Methods of Measuring Decarburization*, likewise gives recommended microscopic, hardness, and chemical-analysis practices for ferrous materials. ISO 3887:2023 specifies three measurement methods, reinforcing the point that a reported depth is method-dependent.

Hardness traverse **Hardness traverse** A sequence of hardness measurements taken at controlled distances from a surface toward the unaffected interior.

A hardness traverse may identify the practical depth at which strength has recovered, while microscopy may identify the disappearance of pearlite or carbide-rich constituents. Chemical analysis may detect a carbon gradient that neither method resolves clearly. Thus two laboratories can report different “decarburization depths” from the same cross-section without either measurement being careless. The result must state the method, criterion, orientation, sampling location, and steel condition. Decarburization is a metallurgical carbon-loss problem first; scale color is only one possible clue.

## The Thermodynamic Driving Force: Carbon Activity at the Surface

![Schematic of carbon diffusing through austenite toward a reacting steel surface](/images/uploads/e1e8e74f-8138-40ba-852f-c01f60539e01/wiki-inline-carbon-exchange-between-austenitic-steel-and-a-furnace-atmosphere-with-carbon-di-1920x1094.jpg)[](/images/uploads/e1e8e74f-8138-40ba-852f-c01f60539e01/wiki-inline-carbon-exchange-between-austenitic-steel-and-a-furnace-atmosphere-with-carbon-di-1920x1094.avif "Enlarge image — Schematic of carbon diffusing through austenite toward a reacting steel surface")Decarburization couples carbon diffusion through the steel with surface reactions controlled by furnace carbon activity.

### Carbon exchange between steel and furnace atmosphere

Decarburization begins when the steel surface is exposed to an atmosphere whose effective carbon activity is lower than the carbon activity of the steel. Carbon then moves from the interior toward the surface by diffusion, while surface reactions convert that carbon into gaseous species that leave the metal. The process is therefore not just “oxidation of the surface.” It is a coupled exchange between carbon in a solid solution or carbide phase and carbon-bearing gases.

The distinction between **carbon activity in steel** and **atmospheric carbon activity** is central. Carbon activity, usually written aC, describes the chemical potential of carbon relative to a defined reference state. In steel, aC depends on carbon concentration, temperature, crystal structure, and alloying elements. Austenite containing 0.60 mass% C does not have the same carbon activity as an unalloyed iron-carbon alloy at the same carbon concentration, because silicon, manganese, chromium, molybdenum, and other elements alter the activity coefficient of carbon. The phase also matters: carbon activity in austenite cannot be transferred directly to ferrite or martensite without considering their different thermodynamic states.

The furnace atmosphere has an effective carbon activity set by equilibria among carbon monoxide, carbon dioxide, hydrogen, water vapor, methane, and other gases. It is not a carbon concentration in the same sense as mass percent carbon in steel. “Atmospheric carbon potential” is a practical expression of the carbon activity of the gas: it is the carbon content that a reference iron or steel would approach if exposed long enough for gas-metal equilibrium at a stated temperature.

When the steel surface and atmosphere are at equilibrium, their carbon chemical potentials are equal. If the atmosphere has the lower carbon potential, carbon leaves the steel. If it has the higher value, carbon enters the steel and carburization occurs. If the two values are close, a small change in gas composition, temperature, or surface condition can reverse the direction of exchange.

Several gas reactions carry carbon away from steel. Oxygen can react with surface carbon through reactions represented approximately by

\\\[ \\mathrm{C + \\tfrac{1}{2}O\_2 \\rightarrow CO} \\\]

and

C+O2→CO2.

Carbon dioxide and water vapor are also important oxidizing components. Their reactions may be represented as

C+CO2→2⁢CO

and

C+H2⁢O→CO+H2.

The products, especially CO and CO₂, can then participate in additional gas-phase reactions. A low concentration of an oxidizing gas can still produce measurable decarburization if the gas continuously removes carbon from the surface and the furnace atmosphere is continuously replenished.

Atmosphere variables must be interpreted together rather than treated as interchangeable indicators.
| Atmosphere variable | Role in carbon-potential control | Measurement or interpretation |
|---|---|---|
| Water-vapor concentration | Can increase the tendency toward carbon loss | Often monitored indirectly through dew point |
| Carbon-dioxide concentration | Can react with surface carbon and alter gas equilibrium | Interpreted with CO and temperature |
| Oxygen partial pressure | Indicates oxidizing tendency and affects oxide formation | Commonly measured with a zirconia probe |
| Carbon-monoxide content | Affects CO/CO₂ equilibrium and carbon potential | Combined with probe oxygen measurement and temperature |

#### Stages of carbon loss

1. **Diffusion** Carbon moves from the interior toward the exposed surface.
2. **Surface reaction** Oxygen-bearing gases consume or transfer carbon at the interface.
3. **Gas removal** CO, CO₂, or other products leave the metal surface.
4. **Profile formation** A concentration and microstructural gradient develops beneath the surface.

The carbon-potential treatment described in *The Carbon Potential During the Heat Treatment of Steel* is useful because it separates the thermodynamic driving force from the rate of the reaction. Carbon exchange is driven by the difference between the carbon activity in the steel and the carbon activity imposed by the atmosphere. Surface-reaction kinetics determine how quickly that difference is expressed. Carbon must first reach the surface, react there, and then enter the gas phase. A slow step at any stage can reduce the observed decarburization rate without eliminating the underlying driving force.

### Carbon activity, oxygen potential, and surface-reaction kinetics

Oxygen partial pressure is a direct measure of the atmosphere’s oxidizing tendency, but it does not independently specify carbon potential. A very low oxygen partial pressure can coexist with a gas mixture that still removes carbon rapidly, particularly when water vapor or carbon dioxide is present and the relevant reaction has favorable kinetics at the treatment temperature.

ASM’s 2013 guidance, *Furnace Atmosphere Controls in Heat Treating*, identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as furnace variables used to control carbon potential and prevent decarburization. These variables are linked, but they are not interchangeable. Dew point is an indirect measure of water vapor. A zirconia probe responds primarily to oxygen potential. Neither measurement, by itself, supplies every quantity needed to calculate the carbon activity imposed at the steel surface.

For example, the equilibrium of

\\\[ \\mathrm{CO\_2 + C \\rightleftharpoons 2CO} \\\]

depends on the ratio of carbon monoxide to carbon dioxide and on temperature. The equilibrium of

\\\[ \\mathrm{H\_2O + C \\rightleftharpoons CO + H\_2} \\\]

depends on water vapor, hydrogen, carbon monoxide, and temperature. Increasing pCO2 or pH2O, while holding other conditions fixed, generally increases the tendency toward decarburization. Increasing the CO-to-CO₂ ratio generally lowers the oxidizing tendency and can raise the effective carbon potential. Hydrogen changes the interpretation of water vapor because the relevant quantity is often the pH2O/pH2 ratio rather than water-vapor concentration alone.

The steel surface may not reach the gas-equilibrium carbon activity immediately. Oxide films, scale, alloying-element oxides, surface roughness, gas velocity, and boundary-layer thickness all affect the transfer of reacting species. A compact oxide can temporarily restrict carbon transfer, while cracked or porous scale can expose fresh metal and create local reaction sites. Carbon diffusion through austenite can then become the controlling resistance, especially as the depleted layer grows. The ASM chapter *Steel Decarburization—Mechanisms, Models, Prevention, Correction, and Effects on Component Life* describes this inward-moving concentration change as carbon diffusion through austenitic iron toward a reacting surface.

This explains why a furnace may show a satisfactory average probe reading while parts still develop local decarburization. Gas composition can vary with furnace position, load geometry, leaks, combustion products, and circulation. A probe measures its own local environment; it does not automatically measure the atmosphere at every steel surface.

A nominal recipe such as nitrogen, hydrogen, and carbon monoxide in stated proportions is consequently not proof of controlled carbon potential. Actual moisture, carbon dioxide, oxygen ingress, gas dissociation, flow rate, furnace pressure, and probe calibration must be known or controlled. Linde’s 2023 sintering guidance reports that water vapor in nitrogen-hydrogen atmospheres causes decarburization and gives an experimental guideline that less than 5% hydrogen limits carburizing and decarburizing rates under the cited sintering conditions. That value is not a universal furnace rule. Its significance depends on the stated gas system, temperature, steel, residence time, and water-vapor level.

Carbon activity also changes with steel composition. Silicon in the Fe-0.6C-1.8Si-0.8Mn spring steel studied between 700 and 1000 °C affects carbon thermodynamics and oxidation behavior, while manganese and the evolving ferrite-austenite balance affect diffusion and phase stability. The surface response cannot be predicted from carbon percentage alone.

### Why atmosphere composition cannot be interpreted independently of temperature

Temperature enters every part of the balance. It changes gas-reaction equilibrium constants, the carbon activity of austenite, the stability of oxides, the rate of surface reactions, and the diffusion coefficient of carbon. The same CO/CO₂ ratio can therefore impose different carbon potentials at different furnace temperatures. A gas mixture that is protective at one temperature may be decarburizing at another.

Temperature also changes which process controls the measured depth. At higher temperature, carbon diffusion through austenite becomes faster, but oxidation reactions and gas transport may change at the same time. The result need not be a monotonic increase in decarburization depth. In a 2021 study of Fe-0.6C-1.8Si-0.8Mn spring steel treated from 700 to 1000 °C, decarburized ferrite thickness increased up to approximately 800–825 °C and then decreased. That maximum is strong evidence against treating temperature as a one-variable rule. Phase transformations, scale formation, oxygen availability, and reaction kinetics were coupled.

Eurotherm’s 2020 *2704CP Handbook supplement* describes common estimation of furnace carbon potential from zirconia-probe oxygen measurements combined with furnace temperature and carbon-monoxide content, or from dew-point measurement. The temperature term is essential in both approaches. A probe signal without the correct temperature, gas analysis, and calibration can produce a plausible but incorrect carbon-potential value.

Cambridge’s *Atmosphere control for the protection of metals during production processes* likewise warns that direct exposure to excess-oxygen combustion atmospheres can cause oxidation, scaling, and decarburization. The practical control target is not a gas label such as “neutral” or “reducing.” It is a measured and temperature-specific surface balance that keeps the steel’s carbon activity from exceeding the atmosphere’s effective carbon activity for long enough to create an unacceptable depleted zone.

## Diffusion, Reaction, and the Formation of a Decarburized Layer

Decarburization is produced by a moving carbon flux, not by oxidation alone. During heating in the austenite range, carbon dissolved in iron migrates toward a surface where the surrounding atmosphere removes it faster than the steel can replace it. The ASM Handbook chapter *Steel Decarburization—Mechanisms, Models, Prevention, Correction, and Effects on Component Life* describes this process as the interaction of carbon diffusion through austenitic iron with surface reactions and furnace-atmosphere chemistry.

The steel surface therefore becomes a sink for carbon. If the atmosphere has a lower carbon activity than the steel, carbon flows outward. The result can include loss of dissolved carbon, dissolution of carbon-rich phases, formation of ferrite at the surface, or a combination of these effects. NIST describes the affected region as a near-surface decarburization zone in which carbon and carbon-rich phases have been removed, commonly after surface carbon reacts with oxygen to form gaseous carbon monoxide (CO) or carbon dioxide (CO₂).

That description also explains why a measured decarburized depth is not a fixed material constant. It depends on temperature, time, section geometry, phase state, alloy chemistry, atmosphere composition, and the method used to identify the boundary.

### Carbon diffusion through austenite

Above the relevant transformation temperature, carbon occupies interstitial sites in austenite, the face-centered-cubic phase of iron. Its concentration is not uniform when the surface is losing carbon. A concentration gradient develops: carbon concentration is lowest near the exposed surface and approaches the bulk composition farther into the section. The gradient supplies the driving force for diffusion.

Fick’s first law **Fick’s first law** A diffusion relation stating that flux is proportional to the negative concentration gradient, with the diffusion coefficient setting the transport scale.

A simplified description uses Fick’s first law:

JC=−DC⁢∂⁢CC∂⁢x

Here, JC is the carbon flux, DC is the carbon diffusion coefficient in austenite, CC is carbon concentration, and x is distance from the surface. The negative sign indicates movement down the concentration gradient. The diffusion coefficient rises strongly with temperature, often represented by an Arrhenius relationship:

DC=D0⁢exp⁢(−QRT)

where D0 is a frequency factor, Q is the activation energy, R is the gas constant, and T is absolute temperature.

A rough diffusion distance scales with DCt, where t is exposure time. This does not predict the measured decarburization depth by itself, because the surface concentration is not automatically zero and the reaction at the surface may be slow. It does show why longer furnace exposure usually allows carbon depletion to extend farther inward, while a higher temperature can increase the affected depth by accelerating transport.

#### Factors that alter carbon transport

Austenite

Carbon diffuses rapidly through interstitial sites in the face-centered-cubic phase.

Ferrite

Different interstitial pathways produce a different diffusion problem.

Section size

Thin products can develop overlapping depletion zones from opposite surfaces.

Edges and corners

Greater exposed surface area and multidirectional diffusion can increase non-uniformity.

The phase state changes the transport problem. Carbon diffuses much more rapidly in austenite than in ferrite because the interstitial sites and diffusion pathways differ. A heat treatment that crosses the austenite-to-ferrite transformation can therefore produce a profile that reflects carbon movement during one phase and transformation during another. Surface ferrite may form because the local carbon concentration falls below that required to stabilize austenite at the treatment temperature. This ferritic layer is not simply an oxide-reaction product; it is also a phase transformation caused by carbon loss.

Section size matters because a thin product can approach a depleted condition from both sides. In a thick bar, the central region may remain near its original carbon concentration while the two surface zones develop independently. In a thin sheet, the diffusion fields can overlap, reducing or eliminating a fully carburized core. Edges and corners add another complication: their greater exposed surface area per unit volume and two-dimensional diffusion paths can produce deeper or less uniform depletion than a broad planar face.

### Surface reactions that remove carbon

Diffusion delivers carbon to the surface, but a surface reaction must consume or transfer it before sustained decarburization can occur. The reaction may involve oxygen, water vapor, carbon dioxide, or other atmosphere constituents. Representative reactions include:

Csteel+12⁢O2→CO

Csteel+O2→CO2

Csteel+H2⁢O→CO+H2

Csteel+CO2→2⁢CO

The actual gas reactions can occur through oxide films and adsorbed surface species, so these equations should be treated as overall reactions rather than a complete sequence of elementary steps. Oxygen may first produce an iron oxide scale, while carbon at or beneath the scale reacts with the oxidizing environment. Cambridge’s *Atmosphere control for the protection of metals during production processes* links excess-oxygen combustion atmospheres with oxidation, scaling, and decarburization.

Carbon activity provides a more useful thermodynamic description than oxygen concentration alone. The carbon potential of the furnace atmosphere is the carbon activity that the atmosphere would impose on steel under specified conditions. When the steel has higher carbon activity than the atmosphere, decarburization is favored. When the atmosphere has higher carbon potential, carbon can enter the steel and carburization may result. The carbon-potential literature stresses that this exchange is controlled by both the activity difference and surface-reaction kinetics.

ASM’s *Furnace Atmosphere Controls in Heat Treating* identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as important variables for controlling furnace carbon potential. Eurotherm’s 2020 *2704CP Handbook* supplement describes common estimation of carbon potential from a zirconia-probe oxygen measurement combined with furnace temperature and carbon-monoxide content, or from dew-point measurement. These measurements describe furnace conditions; they do not directly measure the carbon profile inside every component.

The less-than-5% hydrogen guideline should not be generalized beyond the cited sintering conditions. Limited evidence

Water vapor deserves particular care. Linde’s 2023 *Sintering of steels* guidance states that water vapor in nitrogen/hydrogen atmospheres causes decarburization and cites an experimental guideline that less than 5% hydrogen limited carburizing and decarburizing rates under the stated sintering conditions. That value is not a universal furnace limit. Gas flow, temperature, steel composition, dew point, surface condition, and equipment geometry determine whether a particular atmosphere is oxidizing, neutral, carburizing, or decarburizing.

### Coupling between reaction rate and diffusion rate

The surface and the interior form one coupled system. If the surface reaction is very fast, the near-surface carbon concentration can fall sharply, producing a steep gradient and a diffusion-controlled flux. If carbon transport through austenite is fast but the surface reaction is slow, carbon accumulates near the surface and the overall loss is reaction-controlled. Between those limits, both resistances matter.

A useful conceptual boundary condition is:

−DC∂⁢CC∂⁢x|s=ks⁢(CC,s−CC,eq)

where ks is a surface-reaction coefficient, CC,s is the surface carbon concentration, and CC,eq represents the concentration in equilibrium with the atmosphere. A low ks leaves the surface relatively carbon-rich; a high ks drives the surface concentration toward the atmospheric equilibrium value. Furnace gas velocity, oxide coverage, scale cracking, and gas composition can all alter the effective surface coefficient.

This coupling produces a gradient rather than an abrupt chemical boundary in many specimens. Carbon concentration changes continuously or semi-continuously with depth, and the associated microstructure changes over a zone as ferrite fraction, carbide content, hardness, and etching response vary. A sharp-looking boundary can appear when a test method applies a threshold, but that threshold is a measurement convention, not necessarily a physical discontinuity.

Temperature does not impose a simple monotonic rule either. A 2021 study of Fe-0.6C-1.8Si-0.8Mn spring steel treated from 700 to 1000 °C found that decarburized ferrite thickness increased up to approximately 800–825 °C and then decreased. The result reflects competing effects: diffusion accelerates as temperature rises, while phase stability, oxidation behavior, surface reaction rates, and the time available in a particular phase alter the profile. \[1\] \[1\] [**Steels — Determination of the depth of decarburization**](https://www.iso.org/standard/86689.html). International Organization for Standardization. ISO standard, 2023.

Measurement can change the reported depth. ISO 3887:2023 specifies three methods for determining decarburization depth in steel products. ASTM E1077-14(2021) covers screening, microscopic, microindentation-hardness, and chemical-analysis procedures; where results are disputed, it designates rigorous quantitative or lineal analysis as the referee approach. SAE J419\_201801 likewise gives recommended microscopic, hardness, and chemical-analysis practices. A hardness criterion may identify a broader affected zone than metallographic examination, while chemical analysis can detect carbon loss that produces little visible change. The reported layer is therefore the outcome of furnace exposure and the chosen definition of where decarburization ends.

## Temperature, Time, and the Non-Monotonic Decarburization Response

Temperature affects decarburization through several linked variables, not through diffusion alone. Raising the workpiece temperature generally increases carbon mobility, but it also changes the steel’s phase constitution, the rate of gas–metal reactions, the growth or removal of oxide scale, and the carbon potential at the surface. The measured layer therefore reflects a moving boundary problem: carbon leaves the steel, the atmosphere consumes or supplies carbon-bearing species, and the microstructure used to define the boundary may change during the same thermal cycle.

The distinction matters because “decarburization depth” is a measurement result, not a single physical quantity independent of method. ISO 3887:2023, *Steels — Determination of the depth of decarburization*, specifies three methods for determining depth in steel products. ASTM E1077-14(2021), *Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens*, includes screening, microscopic, microindentation-hardness, and chemical-analysis procedures; it designates rigorous quantitative or lineal analysis as the referee approach when results are disputed. SAE J419\_201801, *Methods of Measuring Decarburization*, likewise provides microscopic, hardness, and chemical-analysis practices. A temperature curve obtained from ferrite thickness is consequently not interchangeable with one obtained from hardness recovery or a carbon concentration profile.

### Heating and soaking effects

During heating, the surface can begin losing carbon before the component reaches its nominal soaking temperature. The relevant driving force is the difference between carbon activity in the steel and the effective carbon potential of the furnace atmosphere. A low-carbon-potential atmosphere extracts carbon through surface reactions involving oxygen-bearing species, while carbon then diffuses from the interior toward the surface. The ASM chapter *Steel Decarburization—Mechanisms, Models, Prevention, Correction, and Effects on Component Life* describes this coupled process as carbon diffusion through austenitic iron toward the surface, followed by reactions that form gaseous carbon monoxide or carbon dioxide.

The simplest diffusion estimate suggests that affected depth scales approximately with the square root of time, x∼Dt, where D is the carbon diffusivity. Since D rises strongly with temperature, a higher temperature and a longer soak can both increase the distance over which carbon concentration changes. That relation is useful, but incomplete. It assumes a stable phase, a fixed surface condition, and a boundary concentration that does not change. Furnace heating violates all three assumptions.

Austenite formation changes carbon solubility and transport. At temperatures below the relevant transformation range, carbon moves through ferrite, whereas above it, carbon is transported through austenite and the surface may develop ferrite, pearlite, or other low-carbon products during cooling. The temperature at which these transformations occur depends on composition and thermal history. Silicon and manganese shift transformation behavior and alter the stability and appearance of the near-surface microstructure. A layer that appears as decarburized ferrite after cooling is therefore the final record of both the high-temperature carbon profile and the transformation path.

Soaking time also changes the atmosphere’s opportunity to react with the surface. A short exposure may produce a carbon-depleted layer without much scale. A longer exposure can deepen the carbon gradient, thicken oxide, crack or shed scale, and expose fresh metal to the gas. Scale may temporarily slow gas access, but it can also create local reaction paths when it becomes porous or detached. The effect of time is not guaranteed to remain proportional to t. \[2\] \[2\] [**Furnace Atmosphere Controls in Heat Treating**](https://dl.asminternational.org/handbooks/edited-volume/8/chapter-abstract/106617/Furnace-Atmosphere-Controls-in-Heat-Treating). ASM International. ASM Handbook, 2013.

Atmosphere control must be included in any temperature comparison. ASM’s 2013 guidance identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as furnace variables used to control carbon potential and prevent decarburization. Eurotherm’s 2020 *2704CP Handbook supplement* describes common inference of carbon potential from zirconia-probe oxygen measurements combined with furnace temperature and carbon-monoxide content, or from dew-point measurement. These readings matter because two tests at the same steel temperature can impose different surface carbon activities. Linde’s 2023 sintering guidance reports that water vapor in nitrogen/hydrogen atmospheres causes decarburization and gives an experimental guideline that less than 5% hydrogen limits carburizing and decarburizing rates under the cited sintering conditions. That guideline belongs to those stated sintering conditions; it is not a general furnace rule.

![Steel specimens showing different decarburized ferrite thicknesses across a temperature range](/images/uploads/9b652d82-661e-4513-be50-52eaa6e4a35a/wiki-inline-a-controlled-laboratory-comparison-of-fe-0-6c-1-8si-0-8mn-spring-steel-specimens-1920x1288.jpg)[](/images/uploads/9b652d82-661e-4513-be50-52eaa6e4a35a/wiki-inline-a-controlled-laboratory-comparison-of-fe-0-6c-1-8si-0-8mn-spring-steel-specimens-1920x1288.avif "Enlarge image — Steel specimens showing different decarburized ferrite thicknesses across a temperature range")In one spring-steel study, measured decarburized ferrite thickness peaked near 800–825 °C rather than rising steadily.

### The 700–1000 °C spring-steel example

A particularly useful counterexample to a temperature-only explanation comes from a 2021 study in the *Journal of Materials Research and Technology* on Fe-0.6C-1.8Si-0.8Mn spring steel. The material was tested between 700 and 1000 °C. Decarburized ferrite thickness increased as temperature rose through the lower part of the tested range, reached a maximum at approximately 800–825 °C, and then declined at higher temperatures within that experiment.

The result is non-monotonic. It does not mean that diffusion became slower above 825 °C. Carbon diffusion generally becomes faster as temperature rises. Rather, several competing effects changed at once. In the lower temperature portion of the test, increasing temperature accelerated carbon transport and surface reactions while providing sufficient exposure for a wider ferritic decarburized region to develop. As the temperature approached and passed approximately 800–825 °C, phase stability and transformation behavior changed. The quantity and morphology of ferrite visible after cooling no longer represented a simple extension of the lower-temperature layer.

Surface oxidation also became more influential. At higher temperature, oxygen-bearing furnace gases can react more rapidly with iron and carbon. Cambridge’s *Atmosphere control for the protection of metals during production processes* notes that direct exposure to excess-oxygen combustion atmospheres can produce oxidation, scaling, and decarburization. A thick or altered scale can change the supply of reactive gas to the metal, while scale formation can consume surface iron and modify the location at which carbon loss is recorded. Faster oxidation can therefore increase the chemical severity of the atmosphere while reducing, redistributing, or obscuring the ferritic zone selected for measurement.

The study’s reported maximum should be read as a consequence of the Fe-0.6C-1.8Si-0.8Mn composition, the 700–1000 °C schedule, the exposure duration, the furnace atmosphere, the specimen geometry, the cooling path, and the chosen definition of decarburized ferrite. It is not a universal law that [spring steels](/categories/spring-steels "spring steels") reach maximum decarburization at 800–825 °C. Another grade, atmosphere, or soak time can shift the maximum or remove it from the tested range entirely.

### Why higher temperature does not always mean a thicker measured layer

Three separate questions are often collapsed into one: how much carbon left the steel, how far the concentration disturbance extended, and how much decarburized material a test method reported. Higher temperature can increase the first two while decreasing the third.

The surface reaction may become so rapid that oxidation, scale growth, or gas-flow limitations control the boundary condition. If the surface is covered by a compact scale, carbon removal may be restricted after an initial period. If the scale becomes porous or flakes away, the reaction can accelerate again. Water vapor, carbon dioxide, oxygen partial pressure, carbon monoxide, and hydrogen each affect the balance between oxidation, carbon removal, and carbon exchange. Carbon potential is therefore a furnace-control variable, not a fixed property of temperature.

Phase change supplies another reason for a smaller measured thickness. A carbon-depleted austenitic region can transform during cooling into ferrite and other products whose appearance depends on cooling rate and local composition. At a higher treatment temperature, the final ferritic morphology may be thinner, discontinuous, or replaced near the surface by oxide and transformed products, even when carbon transport was rapid. A microscopic ferrite criterion can then report a smaller depth than a chemical profile would show.

Measurement method can reverse the apparent ranking of two treatments. Metallography may locate the boundary from ferrite morphology; microindentation hardness may identify a hardness transition at a different position; chemical analysis may show a broader gradient with no sharp interface. ASTM E1077-14(2021) recognizes these distinctions by covering all four procedure types and reserving quantitative or lineal analysis for referee use in disputes. Comparisons must therefore hold specimen orientation, section preparation, etching, magnification, hardness spacing, sampling depth, and reporting criterion constant.

The practical conclusion is direct: temperature should be treated as one coordinate in a coupled thermal, chemical, and microstructural experiment. A valid comparison records heating rate, soak duration, atmosphere composition or carbon potential, scale condition, cooling path, steel designation, and measurement method. The Fe-0.6C-1.8Si-0.8Mn result between 700 and 1000 °C shows why a temperature increase can first thicken and then thin the reported decarburized ferrite layer. It is evidence against a monotonic temperature rule, not evidence that temperature is unimportant.

## Furnace Atmospheres and the Main Process Causes

Decarburization is not caused by “oxidation” in the loose sense of any oxygen entering a furnace. It occurs when the steel surface loses carbon faster than carbon can be supplied from the interior. The controlling balance involves carbon activity in the steel, carbon activity in the gas, surface-reaction kinetics, diffusion through austenite or ferrite, temperature, and the supply of oxidizing species. A furnace can therefore produce scale without producing the same decarburization depth as another furnace, and a small change in gas chemistry can alter the result even when the set temperature and holding time are unchanged.

The ASM Handbook chapter *Steel Decarburization—Mechanisms, Models, Prevention, Correction, and Effects on Component Life* describes carbon diffusion through austenitic iron toward the surface while surface reactions consume carbon. NIST defines a decarburization zone as near-surface loss of carbon and carbon-rich phases when surface carbon reacts with oxygen to form gaseous carbon monoxide or carbon dioxide. The practical question is not simply whether oxygen is present. It is whether the surface reaction lowers the surface carbon activity enough to establish a carbon gradient into the steel.

ASM’s *Furnace Atmosphere Controls in Heat Treating* identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as furnace variables used to control carbon potential and prevent decarburization. Those variables should be treated as operating measurements, not as after-the-fact explanations. Eurotherm’s 2020 guidance describes common carbon-potential inference from a zirconia-probe oxygen measurement combined with furnace temperature and carbon-monoxide content, or from dew point. Each method has limits, so a furnace record should connect gas readings, temperature history, load position, gas flow, and the measured steel condition.

### Water vapor, carbon dioxide, oxygen partial pressure, and combustion products

Water vapor is often the first suspect because it can oxidize carbon at the steel surface. A simplified reaction is:

Csteel+H2⁢O→CO+H2

Carbon dioxide can participate through:

Csteel+CO2→2⁢CO

The products are gaseous, so carbon is removed from the near-surface region rather than remaining as a carbon-rich oxide. Oxygen partial pressure also matters because it governs the tendency to form iron oxides and influences the reactions that establish carbon monoxide and carbon dioxide levels. These reactions are coupled: a gas with low bulk oxygen may still have enough water vapor or carbon dioxide to lower surface carbon activity.

Carbon monoxide is not automatically protective. Its effect depends on the balance between carburizing and decarburizing reactions, gas temperature, steel composition, and the ratio of CO to CO₂. A gas containing CO can add carbon under one carbon-potential condition and remove it under another. Carbon potential is therefore a chemical equilibrium concept, not a label attached to a particular gas. The carbon exchange rate also depends on surface-reaction kinetics; diffusion inside the steel may be rapid while the surface reaction remains controlling, or the reverse may occur.

Combustion products can upset this balance quickly. An uncontrolled burner may produce excess oxygen, high water vapor, or an abnormal CO₂/CO ratio. Fuel-rich combustion can create carbon monoxide and soot-related deposits, while air infiltration can raise oxygen and moisture after the gas has entered the furnace. A wet load, wet refractory, or recently washed fixture can release steam during heating. A leak at a door, seal, burner tile, fan shaft, or cooling section may draw in room air, particularly when furnace pressure falls below ambient.

These observations are process-control hypotheses, not diagnoses. “The furnace leaked” requires confirmation through pressure readings, tracer-gas checks, seal inspection, dew-point trends, or localized gas sampling. “The gas was wet” requires a calibrated dew-point or moisture measurement and a review of dryer performance, piping temperature, and purge practice. A contaminated load may carry oil, cutting fluid, rust, scale, or cleaning residue that decomposes and changes local gas chemistry. Inspecting the load and comparing results by location can distinguish a load-generated problem from a furnace-wide atmosphere failure.

Temperature makes the interpretation less obvious. In a 2021 *Journal of Materials Research and Technology* study of Fe-0.6C-1.8Si-0.8Mn spring steel tested from 700 to 1000 °C, decarburized ferrite thickness increased up to approximately 800–825 °C and then decreased. This non-monotonic result shows why temperature alone is an inadequate explanation. Higher temperature accelerates carbon diffusion and surface reactions, but it can also change oxide formation, gas equilibrium, phase constitution, and the rate at which carbon is replenished from the interior.

### Excess-oxygen atmospheres and scale formation

Cambridge’s *Atmosphere control for the protection of metals during production processes* gives a direct warning: exposure to excess-oxygen combustion atmospheres can produce oxidation, scaling, and decarburization that may require later removal. This is a stronger statement than saying that scale and decarburization are always identical. Scale is an oxide product on or near the surface; decarburization is a change in carbon content and carbon-bearing microstructure. They often occur together because an oxidizing surface consumes carbon while iron reacts with oxygen, but their depths and rates need not match.

An excess-oxygen atmosphere can result from an incorrect air-to-fuel ratio, burner control drift, poor mixing, a failed oxygen-control instrument, or air entering through furnace openings. A furnace may also show acceptable average gas chemistry while a burner jet or load edge experiences a locally oxidizing plume. Load arrangement matters. Tight stacking can block gas circulation and create local pockets; exposed corners and leading edges may react more severely than sheltered surfaces.

Scale can conceal the evidence. Oxide may spall during transfer, shot blasting, straightening, or pickling, leaving a clean-looking surface with a subsurface carbon gradient. Conversely, a dark or thick scale layer does not establish the exact decarburization depth. The furnace investigation should compare scale appearance, surface carbon, hardness profile, metallographic structure, and atmosphere records rather than infer one from another.

Control begins with stable combustion and verified furnace pressure. Burner oxygen trim, fuel pressure, air pressure, damper position, furnace pressure, and exhaust condition should be checked against calibrated instruments. Door opening, charging practice, and quench-transfer intervals can admit air at the most damaging stage. If a result varies across the load, map the affected surfaces and correlate them with burner positions, gas inlets, seals, and flow patterns. If all surfaces show a similar change, review set-point history, gas supply, and carbon-potential control.

### N2/H2 sintering atmospheres and the role of moisture \[3\] \[3\] [**Sintering of steels**](https://static.prd.echannel.linde.com/wcsstore/EE_REN_Industrial_Gas_Store/pdf/Sintering-of-steels-brochure-EN.pdf). Linde. Linde technical guidance, 2023.

Nitrogen-hydrogen atmospheres used in powder-metal sintering require a separate caution. Hydrogen is commonly included to reduce oxides and support a clean sintering environment, but the atmosphere is not automatically neutral toward carbon. Linde’s 2023 *Sintering of steels* guidance states that water vapor in N₂/H₂ atmospheres causes decarburization. Moisture can react with carbon in the compact or steel surface, while hydrogen changes the equilibrium and transport behavior of the gas mixture.

Linde also cites an experimental guideline that less than 5% hydrogen limits carburizing and decarburizing rates under the stated sintering conditions. That figure must remain attached to those conditions. It is not a universal furnace rule for heat treatment, forging, annealing, or every powder-metal composition. Applying it without checking temperature, dew point, carbon content, gas velocity, residence time, and furnace design would turn a context-specific result into a misleading specification.

In sintering, moisture may enter with the supplied nitrogen or hydrogen, form through incomplete drying, condense in cold lines, or be released from refractory and fixtures. A dew-point rise during a production run can indicate a dryer problem, a leaking water-cooled component, inadequate purge, or wet incoming material. The response should verify the source with dew-point measurements at the gas supply and furnace inlet, leak testing, line-temperature checks, and comparison of empty-furnace and loaded-furnace runs. \[4\] \[4\] [**Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens**](https://store.astm.org/e1077-14r21.html). ASTM International. ASTM standard, 2021.

Gas composition alone is insufficient. Carbon-bearing powder, lubricant decomposition, furnace deposits, and compact geometry can alter the local atmosphere around parts. Measure carbon or microstructure at several positions, record dew point through the complete thermal cycle, and compare the result with the furnace’s gas-flow pattern. Where a dispute concerns measured depth, ISO 3887:2023 specifies three methods for determining decarburization depth in steel products. ASTM E1077-14(2021) covers screening, microscopic, microindentation-hardness, and chemical-analysis procedures, with rigorous quantitative or lineal analysis designated as the referee approach in disputes. The atmosphere diagnosis and the measurement method must be reviewed together; changing either can change the reported process outcome.

## How Decarburization Is Measured Under the Standards

A reported decarburization depth is not a single, method-independent property. It is the result of a test system: the standard selected, the section cut from the product, the surface preparation, the resolution of the instrument, and the rule used to decide where decarburization ends. Two laboratories can examine the same bar and obtain different depths without either measurement being careless.

The metallurgical reason is straightforward but important. During heating, carbon moves through austenite toward a surface whose carbon activity is lower than that of the steel. Surface reactions involving oxygen, water vapor, and carbon dioxide can consume carbon by forming carbon monoxide or carbon dioxide. If the furnace atmosphere supplies carbon at a sufficient activity, the reaction may be reduced or reversed; if it does not, diffusion from the interior feeds the surface reaction. ASM identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as furnace variables used to control carbon potential and limit decarburization. Eurotherm describes common carbon-potential estimates based on zirconia-probe oxygen measurements with furnace temperature and carbon-monoxide content, or on dew-point measurement.

The measured zone therefore depends on the thermal and atmospheric history, not merely on the peak temperature. A 2021 study of Fe-0.6C-1.8Si-0.8Mn spring steel tested from 700 to 1000 °C found that decarburized ferrite thickness increased to approximately 800–825 °C and then decreased. That result is a warning against treating temperature as a one-direction control variable: diffusion accelerates with temperature, but oxide formation, phase constitution, surface kinetics, and the available carbon potential also change.

![Steel test specimen beside equipment for microscopy, hardness testing, and chemical analysis](/images/uploads/f28a33c5-9bd7-48ac-97fe-ff5e1d9a7b89/wiki-inline-a-steel-decarburization-test-specimen-prepared-for-metallography-microindentatio-1520x1920.jpg)[](/images/uploads/f28a33c5-9bd7-48ac-97fe-ff5e1d9a7b89/wiki-inline-a-steel-decarburization-test-specimen-prepared-for-metallography-microindentatio-2027x2560.avif "Enlarge image — Steel test specimen beside equipment for microscopy, hardness testing, and chemical analysis")ISO 3887:2023 provides three routes for determining decarburization depth: microscopy, hardness, and chemical analysis.

### ISO 3887:2023 and its three measurement methods

ISO 3887:2023 provides three routes because decarburization can be defined by structure, hardness, or composition.
| ISO method family | What it evaluates | Suitable decision basis |
|---|---|---|
| Metallographic examination | Surface-to-core phase and morphology changes | Complete or partial decarburization |
| Hardness measurement | Mechanical-property gradient with depth | Functional softened-zone criterion |
| Chemical analysis | Carbon concentration profile | Carbon-content acceptance limit |

ISO 3887:2023, *Steels — Determination of the depth of decarburization*, specifies three methods for steel products. They correspond conceptually to metallographic examination, hardness measurement, and chemical analysis. The standard is therefore not a declaration that one universal test defines decarburization. It provides different routes suited to different materials, geometries, acceptance requirements, and laboratory capabilities.

The metallographic route examines a polished cross-section from the original surface inward. Etching reveals changes in microstructure, such as a ferritic layer, reduced pearlite or carbide content, and a transition toward the unaffected core. This method can distinguish complete decarburization, where carbon-rich phases have substantially disappeared, from partial decarburization, where the carbon level or fraction of pearlite and carbide has fallen but the original constituents remain. Its strength is direct visual evidence. Its weakness is that the boundary can be gradual, banded, or affected by etching contrast. A measured line is then an interpretation of a transition, not a naturally sharp interface.

The hardness route uses measurements made at controlled distances from the surface. A surface layer with lower carbon generally has lower hardness than the unaffected material, although alloying elements, grain size, retained phases, tempering response, and residual stress can also alter hardness. Microindentation is especially useful when the affected layer is thin, but indents must be small enough and sufficiently separated to avoid overlapping plastic zones. A spacing that is appropriate for a thick zone can blur a shallow one. Conversely, very small spacing can make individual readings sensitive to local grains and preparation damage.

Chemical analysis measures carbon as a function of distance from the surface, either through individually removed layers or another spatially resolved technique. It is the most direct approach when the acceptance criterion is carbon concentration rather than a visible phase boundary. It also has a serious limitation: chemical resolution. If material is removed in 0.05 mm layers, a reported boundary cannot credibly be assigned to 0.005 mm. Surface contamination, scale, decarburized ferrite mixed with unaffected metal, and the volume sampled by the analytical method can shift the apparent profile.

ISO 3887:2023 should be treated as a reference framework for choosing and reporting one of these methods. A microscopic depth, a hardness-transition depth, and a chemically defined carbon-loss depth may all be valid while describing different boundaries. The report should identify the method and definition used rather than present “depth” as though it were self-explanatory.

### ASTM E1077-14(2021) procedure families

ASTM E1077-14(2021), *Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens*, organizes the problem into screening, microscopic, microindentation-hardness, and chemical-analysis procedures. These are procedure families, not interchangeable names for the same test.

Screening is the quick decision stage. A low-magnification examination or a limited hardness traverse can indicate whether a surface has no obvious decarburization, a substantial zone, or a condition requiring more exact work. Screening is useful for production checks and for selecting specimens, but it is not automatically suitable for a close acceptance limit. A coarse visual examination may miss a thin partial-decarburized layer; a broad hardness interval may conceal a shallow gradient.

Microscopic examination provides a lineal or quantitative estimate when the section is prepared and measured carefully. The specimen should be cut so that the examined face is perpendicular to the original surface. A longitudinal section answers a different question from a transverse section: it can expose variation along the product and may intersect rolling or forging structure differently. The original surface must remain identifiable. Grinding can remove a thin zone, rounding can distort the edge, and over-etching can exaggerate contrast.

Magnification changes the practical result. At low magnification, the operator may report the outermost visibly altered band. At higher magnification, fine carbide depletion or a narrow ferritic region may become measurable. The field of view, illumination, etchant, calibration, and number of measured locations all matter. ASTM E1077-14(2021) identifies rigorous quantitative or lineal analysis as the referee approach in disputes. That designation reflects a sound principle: when a result is contested, the laboratory should replace a visual impression with documented measurements across defined lines or locations.

Microindentation-hardness testing converts the surface-to-core property gradient into a depth estimate. The traverse should begin far enough from the edge to avoid edge effects but close enough to resolve the expected zone. Indent spacing, load, dwell time, surface flatness, and the criterion for returning to core hardness must be reported. If indents are spaced too widely, the stated depth becomes a bracket rather than a precise boundary. If they are too close, deformation from one indent can influence the next.

Chemical analysis is preferable when carbon concentration itself controls the decision, but the sampled volume and detection limit establish the smallest credible feature. A chemically measured carbon profile may show a gradual decline even where microscopy shows a sharp phase change. Neither result should be relabeled as the other.

### SAE J419\_201801 and method selection

SAE J419\_201801, *Methods of Measuring Decarburization*, gives recommended microscopic, hardness, and chemical-analysis practices for ferrous materials. It is best used with the product specification and heat-treatment record, not as a substitute for either. ISO 3887:2023 provides three defined measurement routes for steel products; ASTM E1077-14(2021) sets out estimation and referee concepts; SAE J419\_201801 supplies practical method guidance. Their designations should remain visible in the report because they establish different procedural contexts.

Method selection should follow the failure risk. Choose microscopy when the relevant question concerns complete or partial decarburization and the microstructure can be interpreted reliably. Choose microindentation hardness when a property gradient matters and the layer is thick enough for a defensible traverse. Choose chemical analysis when a carbon-content limit is specified or when alloying and heat treatment make hardness an ambiguous proxy. Use screening only for a preliminary decision unless the governing product requirement explicitly permits it.

#### Minimum measurement-report fields

- **Material identity** Product, grade, heat-treatment condition, and governing specification.
- **Sampling** Section orientation, surface location, and number of examined areas.
- **Preparation** Cutting, mounting, grinding, polishing, and etching details.
- **Measurement** Magnification, hardness load and spacing, or chemical-analysis resolution.
- **Criterion** Definition of complete, partial, or total affected depth.
- **Result** Maximum, average, spread, and limitations where relevant.

Every report should state the product and grade, the applicable specification, section orientation, location and number of examined areas, surface-preparation sequence, microscope magnification or hardness load and spacing, analytical resolution, boundary criterion, and reporting convention. It should also distinguish maximum depth from average depth and state whether the value refers to complete decarburization, partial decarburization, or total affected depth.

That detail is not paperwork around the measurement. It is part of the measurement. A decarburization value without its boundary definition and method can create false disagreement, while a carefully identified method makes results from different laboratories comparable within their actual limits.

## Metallographic Measurement: What the Microscope Can and Cannot Show

### Cross-section preparation and etching

A metallographic examination should begin with a section taken normal, or perpendicular, to the processed surface. For a sheet, bar, wire, or forging, the cut should expose the complete path from the original surface into unaffected material. A section parallel to the surface cannot establish decarburization depth because it shows lateral variation, not the required surface-to-core transition. Sampling must also record the product orientation, heat-treatment condition, surface location, and any machining or scale-removal operation performed before examination.

The cut face should be protected from additional heating. Abrasive cutting with excessive force can temper the surface, smear softened material over the section, or pull out inclusions. These effects can create a false near-surface band. Mounting, when required, should support the edge without rounding it; edge rounding shortens the apparent distance between the surface and the first unaffected structure. Grinding must proceed through successively finer abrasives with enough lubrication and pressure control to remove the previous damage rather than extend it. Final polishing should leave no dragged metal, embedded abrasive, or deep scratches crossing the surface.

The original product surface needs to remain identifiable. If scale is removed before sectioning, the laboratory should record that fact and retain a representative piece when possible. Otherwise, the analyst may mistake the scale–metal interface for the manufactured surface and report an incorrect depth. A visibly dark oxide layer is not itself a measurement of decarburization. Scale records oxidation; decarburization records carbon loss in the steel beneath, and the two may occur together, separately, or at different rates.

Etching reveals phase contrast after polishing. The reagent and etching time must suit the grade and intended examination. Nital is widely used for carbon and low-alloy steels, but concentration, alcohol condition, application time, and rinsing affect the contrast. Over-etching can darken ferrite or make a transition appear broader. Under-etching can hide fine pearlite and produce an apparently carbon-depleted layer. A properly prepared unetched surface remains useful for checking grinding damage, pores, inclusions, and scale penetration; the etched surface is then used to interpret the steel structure.

#### Interpretation note

Metallography shows the structural effects of carbon loss, thermal history, alloying, cooling rate, and surface reactions. Confirm a disputed carbon-loss diagnosis with hardness or chemical evidence where appropriate.

The microscope does not measure carbon directly. It displays how carbon content, thermal history, alloying elements, cooling rate, and surface reactions have affected the structure.

### Ferrite, pearlite, and carbon-gradient interpretation

In a hypoeutectoid carbon steel, carbon loss during austenitic heating commonly produces a near-surface region with more ferrite and less pearlite than the interior. The interior may show the expected ferrite–pearlite structure, while the surface zone contains ferrite grains extending farther inward than they do in the unaffected material. This is the familiar metallographic indication of partial decarburization. If carbon depletion is sufficiently severe, the visible surface region may be almost entirely ferritic, which is often described as complete decarburization under the applicable test criteria.

The transition is rarely a sharp chemical boundary. Carbon diffuses through austenite toward a surface whose carbon activity is lower than that of the steel. At the surface, reactions involving oxygen, water vapor, carbon dioxide, or other furnace gases consume carbon or alter the interfacial carbon balance. The measured structure therefore reflects diffusion, surface-reaction kinetics, temperature, exposure time, and furnace carbon potential together. ASM’s *Steel Decarburization—Mechanisms, Models, Prevention, Correction, and Effects on Component Life* describes this coupled process rather than treating it as oxidation alone.

Ferrite and pearlite are useful indicators, but they are not automatic proof of decarburization. A ferrite-rich band may result from a local change in cooling rate, a transformed heat-affected region, or the original microstructural gradient from forging or rolling. A heat-affected artifact from welding, flame cutting, grinding, or a prior repair can produce altered grain size and phase contrast without a corresponding carbon gradient. Segregation can also change the local pearlite fraction. A manganese or silicon segregation band, for example, may alter transformation behavior and produce a structure that differs from adjacent areas even though the carbon concentration has not fallen at the surface.

The analyst should compare several locations around the perimeter and, where possible, examine an unaffected reference region from the same product and heat. The depth should be checked against a change in phase fraction, grain morphology, or hardness—not inferred from one isolated photograph. A gradual reduction in pearlite toward the surface supports carbon depletion when the surface is correctly identified and the thermal history is known. It does not establish the furnace cause.

Phase transformation can further complicate interpretation. A steel that was austenitized and rapidly cooled may contain martensite or bainite beneath a ferritic surface layer. The apparent boundary could represent a change in carbon content, a change in cooling rate, or both. In medium-carbon grades such as AISI/SAE 1045, a ferritic surface over a harder transformed core may be consistent with decarburization, but the metallographer still needs hardness or chemical evidence if the result is disputed. In alloy grades, chromium, molybdenum, silicon, and manganese influence phase stability and etching response, so a visual comparison with plain-carbon steel is unsafe.

Temperature alone also cannot predict the observed layer. A 2021 study of Fe-0.6C-1.8Si-0.8Mn spring steel tested from 700 to 1000 °C found that decarburized ferrite thickness increased up to approximately 800–825 °C and then decreased. Greater diffusion at higher temperature did not produce a monotonic increase because oxidation and surface-reaction conditions also changed. This is one reason a microscope result must be read alongside furnace atmosphere records. ASM identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as furnace variables relevant to carbon potential. Eurotherm describes carbon-potential inference from zirconia-probe oxygen measurements combined with temperature and carbon-monoxide content, or from dew-point measurement. Those controls explain exposure; they do not replace examination of the steel.

### Lineal or quantitative analysis in referee examinations

A screening examination can identify whether a suspicious surface zone exists, but a reported decarburization depth requires defined criteria. ASTM E1077-14(2021), *Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens*, covers screening, microscopic, microindentation-hardness, and chemical-analysis procedures. It designates rigorous quantitative or lineal analysis as the referee approach when parties disagree. ISO 3887:2023, *Steels — Determination of the depth of decarburization*, specifies three methods for measuring depth in steel products, while SAE J419\_201801, *Methods of Measuring Decarburization*, gives recommended microscopic, hardness, and chemical-analysis practices for ferrous materials.

Lineal analysis converts a visual transition into a recorded measurement. A test line is placed from the identified original surface into the core, and the analyst records the position at which the specified structural criterion is met. Depending on the standard and contract, that criterion may concern the end of complete decarburization, the end of partial decarburization, or the return to the normal core structure. The report must state which boundary was measured. “Decarburization depth” without that definition can describe different distances.

For quantitative microscopy, the laboratory should document magnification, calibrated scale, etchant, section location, number and spacing of measurements, and the rule used to locate each boundary. Several radial or transverse lines are preferable to one convenient field, especially on bars and forgings where exposure may vary around the circumference. Images should include the surface reference and a scale bar. If the transition is irregular, the minimum, maximum, mean, and measurement spread may matter more than a single rounded value.

A referee examination should separate observation from interpretation. It can state that ferrite fraction rises toward the surface, that pearlite disappears at a measured distance, or that hardness changes across the same zone. It should not claim that a furnace had excessive oxygen merely because the micrograph is ferrite-rich. Scale penetration, polishing damage, segregation, and an unrelated transformed layer must be considered and, where relevant, checked by unetched microscopy, hardness traverses, chemical analysis, or a second section.

The microscope is therefore powerful at locating structural evidence, but limited in identifying its cause and exact carbon profile. ASTM’s referee concept matters because it replaces visual confidence with a documented measurement rule. In a dispute, reproducibility is the result: another examiner should know where the section came from, how it was prepared, what boundary was defined, and how the reported depth was calculated.

## Hardness and Chemical-Analysis Methods

Decarburization is often reported as a depth in millimetres, but that number depends on how the depth was measured. A hardness traverse records a change in local mechanical response. A chemical profile records composition within a sampled volume. Metallographic examination records phases and morphology. These are related observations, not interchangeable measurements.

ISO 3887:2023, *Steels — Determination of the depth of decarburization*, specifies three methods for measuring decarburization depth in steel products. ASTM E1077-14(2021), *Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens*, covers screening, microscopic, microindentation-hardness, and chemical-analysis procedures. SAE J419\_201801, *Methods of Measuring Decarburization*, also sets out recommended microscopic, hardness, and chemical-analysis practices for ferrous materials. The appropriate method depends on whether the result is a production screen, a specification measurement, or evidence in a failure or dispute.

### Microindentation-hardness traverses

A microindentation-hardness traverse places a sequence of small indents from the exposed surface toward the unaffected interior, then plots hardness against distance. The test commonly uses Vickers or Knoop indents, with the load and spacing selected to resolve the expected gradient without overlapping plastic zones. The result is a near-surface mechanical-response profile.

A decarburized layer commonly contains less pearlite, less carbide, or a greater fraction of ferrite than the base material. It therefore tends to show lower hardness, particularly in medium- and high-carbon steels whose strength depends strongly on carbon-rich phases. The traverse may show a steep surface-to-core increase, a gradual transition, or no clear change. A specified hardness criterion can then be used to estimate the boundary of the affected zone.

That boundary is not a direct carbon assay. Hardness also responds to phase balance, grain size, martensite or bainite formation, tempering, residual stress, alloying elements, test direction, and local microstructural banding. A soft surface may reflect decarburization, but it may also result from a different thermal history or an improperly prepared surface. Conversely, a carbon-depleted region can retain substantial hardness if its transformation products or alloy content compensate for some loss of carbon.

Surface preparation matters. Scale, grinding burns, rounded edges, and indentation placement too close to the edge can distort the first measurements. The indenter must be small enough to detect the gradient but not so small that individual ferrite grains, carbides, or surface roughness dominate the reading. Excessively close spacing can also make neighbouring indents interact. A traverse should cross the surface normal and continue far enough into material that several readings establish the unaffected reference hardness.

ASTM E1077-14(2021) includes microindentation hardness among its estimation procedures, but it does not turn a hardness profile into a chemical profile. The method is most persuasive when the hardness transition agrees with metallographic evidence and with the expected carbon gradient. It is less persuasive when the steel has undergone complicated quenching, tempering, spheroidizing, or surface treatment.

The furnace history can also produce a non-monotonic result. In a 2021 *Journal of Materials Research and Technology* study of Fe-0.6C-1.8Si-0.8Mn spring steel heated between 700 and 1000 °C, decarburized ferrite thickness increased with temperature to approximately 800–825 °C and then decreased. Temperature alone did not determine the measured layer. Carbon activity at the steel surface, oxidation reactions, diffusion through austenite, and the duration of exposure all changed together. A hardness traverse taken after such a cycle should not be interpreted using a simple rule that higher temperature always means greater decarburization depth.

### Chemical analysis and sampling depth

Chemical analysis addresses a different question: how much carbon is present in the material removed or sampled from a defined location? This makes it the more direct composition-oriented approach, but “direct” does not mean spatially exact. The result represents the volume of material collected, not necessarily the carbon concentration at one geometric plane.

Bulk combustion analysis, for example, can measure carbon accurately in a prepared specimen while averaging carbon through the entire sample. It cannot identify a 20 micrometre surface gradient if the sample is several millimetres thick. To resolve decarburization, the analyst must remove successive layers, take a carefully defined thin slice, or use a microanalytical technique with a known interaction volume. The sampling geometry becomes part of the result.

Surface removal is especially important. If a specimen is analysed with scale, oil, grinding debris, or a layer of material unintentionally removed, the reported carbon value may not correspond to the original steel surface. Chemical stripping, machining, electrolytic removal, or controlled grinding can each alter the depth increment and introduce contamination. The method should document the initial reference surface, the amount removed at each stage, the area or mass collected, and whether the reported value is an average over a layer or a point-like measurement.

A layer-by-layer chemical profile can reveal a carbon concentration that gradually rises toward the core rather than a sharp interface. That profile may not yield the same “depth” as a hardness criterion or a metallographic boundary. One method might define the affected depth where carbon reaches a chosen fraction of the nominal bulk value; another might define it where ferrite or carbide morphology is no longer distinguishable. The values can differ without either test being invalid.

The furnace atmosphere provides useful context for interpreting such profiles. ASM International’s 2013 *Furnace Atmosphere Controls in Heat Treating* identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as variables used to control furnace carbon potential and prevent decarburization. Carbon exchange is governed by the difference in carbon activity between steel and atmosphere, together with the kinetics of the surface reaction. Eurotherm’s 2020 *2704CP Handbook* supplement describes common inference of furnace carbon potential from zirconia-probe oxygen measurements combined with furnace temperature and carbon-monoxide content, or from dew-point measurement.

These readings are process evidence, not proof of the carbon content at a component surface. Linde’s 2023 sintering guidance reports that water vapour in N2/H2 atmospheres causes decarburization and cites an experimental guideline that less than 5% hydrogen limited carburizing and decarburizing rates under the stated sintering conditions. That figure is conditional, not a universal furnace rule. Exposure to excess oxygen can produce oxidation, scaling, and decarburization, as described by Cambridge’s guidance on atmosphere control for metals.

### Correlating hardness, microstructure, and carbon concentration

The strongest interpretation comes from agreement among independent observations. A surface region showing reduced carbon, increased ferrite or reduced carbide fraction, and lower hardness provides a coherent decarburization diagnosis. If only hardness changes, the finding is suggestive. If only a carbon result changes, the analyst must still consider sampling smear, averaging, and surface preparation. If microstructure changes without a carbon gradient, transformation history or alloy segregation may be responsible.

Correlation requires matching locations. The hardness traverse, metallographic section, and chemical samples should be taken from adjacent or serial sections with a common surface reference. A carbon sample that averages the first 0.5 mm cannot be compared directly with a hardness change confined to the first 0.1 mm. Likewise, a micrograph taken from a corner may not represent a broad flat surface because edge geometry changes heat transfer and gas access.

Standards-based interpretation benefits from this method combination when a specification, failure investigation, or dispute demands stronger evidence. ASTM E1077-14(2021) designates rigorous quantitative or lineal analysis as the referee approach in disputes. That provision matters because a quick hardness screen can identify suspicious material, while a more controlled quantitative examination can establish the reported depth. ISO 3887:2023 and SAE J419\_201801 provide the framework for selecting and applying the relevant measurement approaches; the laboratory report should still state the criterion used, sampling location, preparation procedure, traverse spacing, and uncertainty or repeatability where available.

A practical conclusion follows: hardness is often efficient for locating a transition, chemical analysis is valuable for confirming composition, and microscopy explains which phases changed. None should be treated as a universal substitute for the others. The reported decarburization depth is meaningful only when the test method, sampling volume, acceptance criterion, and steel’s thermal history are identified alongside the number.

## Controlling Decarburization in Heat Treatment and Sintering

Decarburization control starts with carbon potential, not with a nominal furnace-gas recipe. A mixture labelled “endothermic,” “nitrogen–methanol,” or “N₂/H₂” can produce different carbon-transfer conditions when temperature, moisture, leakage, gas flow, or combustion balance changes. The steel responds to carbon activity at its surface and to the rates of the surface reactions that consume or supply carbon. Carbon then diffuses through austenite toward the surface. Furnace control must therefore measure the atmosphere variables that govern carbon activity rather than assume that a fixed gas composition will produce a fixed result.

The required carbon potential should be established for the grade, section size, heating and holding cycle, and any subsequent carburizing, neutral-hardening, or bright-annealing operation. A neutral atmosphere for a plain-carbon grade is not automatically neutral for a high-alloy steel, nor is the same set point suitable at every temperature. The aim is to keep the surface carbon activity close enough to the steel’s required equilibrium condition that carbon loss is prevented without unwanted carburization.

### Setting and maintaining furnace carbon potential

Carbon potential is the effective carbon activity that the furnace atmosphere can impose on steel at a specified temperature. If the atmosphere has a lower carbon potential than the steel surface, carbon leaves the steel. If it has a higher potential, carbon enters it. A short excursion can matter when the load is thin, the surface area is large, or the material remains at temperature for a long time.

ASM’s *Furnace Atmosphere Controls in Heat Treating* identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as furnace variables used to control carbon potential and prevent decarburization. These variables are linked through the reactions between carbon monoxide, carbon dioxide, hydrogen, water vapor, and carbon at the steel surface. A small rise in H₂O or CO₂ may therefore change the effective atmosphere even when the displayed flow rates of nitrogen, hydrogen, or enriching gas remain unchanged.

Set points should be established from the steel specification, process qualification, and furnace calibration work. They should not be copied from a gas-flow table without checking temperature and analyzer response. For a hardening cycle, the selected potential may be near neutral or slightly carburizing at the austenitizing temperature; for a sintering cycle, the permissible oxygen and moisture levels may be specified more directly because powder metallurgy parts can have large surface area and interconnected porosity.

The temperature must be controlled as tightly as the atmosphere. Carbon activity, reaction rates, and carbon diffusion all increase or change with temperature, but decarburization depth does not necessarily rise in a simple straight line. A 2021 study of Fe-0.6C-1.8Si-0.8Mn spring steel tested from 700 to 1000 °C found that decarburized ferrite thickness increased up to approximately 800–825 °C and then decreased. This result reflects competing diffusion, oxidation, phase-transformation, and surface-reaction effects. A higher furnace temperature is not, by itself, proof of greater measured decarburization.

Atmosphere control also requires an alarm and response plan. A failed probe, rising dew point, falling carbon potential, loss of enriching gas, or unexplained CO change should trigger a hold, controlled correction, or load quarantine according to the process risk. A controller that merely drives a valve toward its set point while the measuring device is fouled can maintain a displayed number and still produce decarburized steel.

![Heat-treatment furnace controls monitoring oxygen, dew point, temperature, and carbon monoxide](/images/uploads/629badba-7494-4dc9-9fec-f9d297d99159/wiki-inline-a-heat-treatment-furnace-control-station-with-a-zirconia-oxygen-probe-dew-point-1920x1094.jpg)[](/images/uploads/629badba-7494-4dc9-9fec-f9d297d99159/wiki-inline-a-heat-treatment-furnace-control-station-with-a-zirconia-oxygen-probe-dew-point-1920x1094.avif "Enlarge image — Heat-treatment furnace controls monitoring oxygen, dew point, temperature, and carbon monoxide")Carbon-potential control depends on temperature and atmosphere measurements rather than a nominal gas recipe alone.

### Sensors, dew point, oxygen probes, temperature, and carbon monoxide

Eurotherm’s 2704CP Handbook supplement describes two common routes for estimating furnace carbon potential. One combines a zirconia-probe oxygen measurement with furnace temperature and carbon-monoxide content. The other uses dew-point measurement. These routes are related, but they are not interchangeable readings from one physical sensor.

A zirconia oxygen probe measures oxygen potential through the voltage generated across a zirconia electrolyte at furnace temperature. With a known or controlled CO content, the probe signal and temperature can be converted into an estimated carbon potential. Temperature compensation is essential: the same probe voltage does not represent the same oxygen or carbon potential at different temperatures. CO content is also essential in the calculation. If the controller assumes a constant CO value while the actual gas composition changes, the indicated carbon potential can be wrong even though the zirconia cell is functioning.

Probe location matters. The measurement should represent the atmosphere around the load, not a stagnant corner, a cold sampling tube, or a region close to an air leak. Soot, scale, condensate, probe aging, electrical noise, and reference-air problems can distort readings. Operators should verify probe response against a known calibration condition, inspect the probe and protective components, and compare the signal with independent atmosphere data after maintenance or unexplained drift. Calibration cannot correct a probe installed where the furnace atmosphere is not representative.

Dew-point measurement provides a separate monitoring route by measuring the moisture content of the gas, usually after controlled sampling and cooling. Moisture is a direct warning variable because water vapor can oxidize carbon at the steel surface. ASM lists water-vapor concentration among the variables used to control carbon potential, while Linde’s 2023 sintering guidance states that water vapor in N₂/H₂ atmospheres causes decarburization. Linde also gives an experimental guideline that less than 5% hydrogen limits carburizing and decarburizing rates under the cited sintering conditions. That value is process-specific, not a universal limit for every furnace, alloy, temperature, or dew point.

Dew point can reveal water entering through wet refractories, poor gas drying, leaks, furnace cleaning, or incomplete purging. It cannot by itself describe every reaction affecting carbon potential. A low dew point does not prove that the load is safe if air enters through a door seal or if CO₂, CO, temperature, and flow conditions are abnormal. For this reason, a dew-point system should be interpreted with temperature, flow, pressure, and, where applicable, oxygen and CO measurements.

### Load practice, sealing, gas distribution, and process verification

A well-controlled atmosphere cannot compensate for poor load practice. Parts should be arranged so that gas can reach all exposed surfaces and so that tightly packed regions do not form stagnant, moisture-retaining zones. Baskets, fixtures, trays, and protective coatings must not block flow around critical surfaces. Heavy loads can consume enriching gas or delay atmosphere replacement, while cold or wet loads can release moisture during the most sensitive part of the heating cycle.

Furnace seals deserve the same attention as analyzers. Door gaps, muffle cracks, burner controls, fan-shaft seals, sampling lines, and poorly closed transfer openings can draw air into a furnace operating at low pressure. Air ingress supplies oxygen and can create local oxidation and decarburization even when the bulk analyzer reports an acceptable value. Positive pressure, where appropriate for the furnace design, reduces infiltration, but excessive pressure can push process gas through seals and create other hazards. Leak checks should be performed after maintenance and whenever oxygen, dew point, or CO trends change without a credible process explanation.

Gas distribution must be checked rather than inferred from total flow. Verify inlet and exhaust operation, circulation-fan performance, purge time, flowmeter function, and pressure stability. In continuous sintering, inspect zones separately: the atmosphere at preheat, high-temperature, and cooling sections can differ substantially. Water vapor released in one zone may condense or react downstream, and a single analyzer may miss a local condition at the work surface.

Process verification ends with the steel. Representative samples should include the locations most exposed to furnace atmosphere and the areas most likely to be shielded by contact or packing. ISO 3887:2023, *Steels — Determination of the depth of decarburization*, specifies three methods for measuring decarburization depth in steel products. ASTM E1077-14(2021), *Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens*, covers screening, microscopic, microindentation-hardness, and chemical-analysis procedures. In disputes, ASTM designates rigorous quantitative or lineal analysis as the referee approach. SAE J419\_201801, *Methods of Measuring Decarburization*, gives recommended microscopic, hardness, and chemical-analysis practices for ferrous materials.

The test method can change the reported depth. A hardness traverse may identify a functionally softened zone, while metallography may separate ferrite decarburization from partial decarburization or distinguish oxidation damage. Results should therefore be compared using the same standard, section orientation, preparation method, magnification, and acceptance criterion. If the furnace record looks satisfactory but representative steel fails, investigate sensor lag, sample location, load shielding, moisture, air ingress, and the validity of the carbon-potential calculation before simply raising the set point. That response can trade decarburization for carburization without correcting the underlying fault.

## Correction, Removal, Acceptance, and Failure Consequences

Correction and acceptance depend on the governing specification, geometry, service risk, and verified remaining condition.
| Disposition | When it may be appropriate | Required verification |
|---|---|---|
| Accept | Measured condition meets the governing requirement | Document method, criterion, and result |
| Remove by machining or grinding | Affected layer lies within approved stock allowance | Reinspect the corrected surface and remaining geometry |
| Reprocess | A qualified complete thermal cycle can restore required properties | Requalify atmosphere, heating, cooling, and final properties |
| Reject or downgrade | Critical surface or section cannot meet the requirement | Record measured evidence and engineering disposition |

Decarburization is not automatically a rejection condition, but neither is it a defect that machining always fixes. The decision depends on what carbon was lost, how far the affected zone extends, which surfaces carry load, and whether the specified geometry leaves enough material for removal. A thin ferritic layer on a machining allowance may be removed without impairing performance. The same measured depth on a finished spring, gear tooth, bearing race, or fatigue-critical shaft may represent a serious loss of capability. \[5\] \[5\] [**Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens**](https://store.astm.org/e1077-14r21.html). ASTM International. ASTM standard, 2021.

The first decision should be based on a verified measurement, not on scale color or a single hardness impression. ISO 3887:2023, *Steels — Determination of the depth of decarburization*, specifies three methods for determining decarburization depth in steel products. ASTM E1077-14(2021), *Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens*, covers screening, microscopic examination, microindentation hardness, and chemical-analysis procedures. Where results are disputed, ASTM E1077-14(2021) designates rigorous quantitative or lineal analysis as the referee approach. SAE J419\_201801 likewise sets out recommended microscopic, hardness, and chemical-analysis practices for ferrous materials.

These methods do not necessarily report the same boundary. A metallographic method may identify a zone in which pearlite or martensite has changed into ferrite, while a hardness traverse may show a more gradual transition and define the affected depth at a selected hardness criterion. Chemical analysis can measure carbon concentration directly but may average the result over a sampling volume. Consequently, “remove the decarburization” is incomplete unless the required removal depth and the method used to verify it are stated.

### When decarburized material can be removed by machining or grinding

Removal is usually technically feasible when the affected layer lies entirely within an approved machining or grinding allowance. Turning, milling, cylindrical grinding, surface grinding, and controlled stock removal can expose material with the required carbon content and hardness profile. The remaining section must still satisfy the drawing, dimensional tolerance, surface-finish requirement, and any case-depth or heat-treatment requirement. A part cannot be declared corrected merely because a pale or soft surface has disappeared.

The removal calculation must account for the geometry of the affected surface. If a cylindrical shaft loses 0.20 mm of radius during grinding, its diameter decreases by 0.40 mm. On a flat face, removing 0.20 mm changes thickness by 0.20 mm. A decarburized fillet, tooth flank, thread root, or keyway may be difficult to remove uniformly; local grinding can create undercutting, sharp transitions, or residual tensile stress. Those effects may be more damaging than the original carbon loss.

The required stock is also determined by the actual carbon profile, not only by the visible ferrite depth. A partially decarburized zone can contain a carbon gradient extending beneath a fully ferritic layer. If the specified property depends on a minimum hardness or carbon concentration, the removal depth must reach that criterion across the relevant surface. Verification should normally use a representative cross-section from the corrected region, with the method and location agreed before disposition.

Machining is more defensible on a rough-forged or rough-rolled component than on a finished component with little remaining allowance. It can also be appropriate where the affected surface is nonfunctional and the drawing allows stock removal. It is not a general remedy for a component whose critical surface has already reached final size. Grinding cannot restore stock that the design has already consumed.

The cause should be addressed at the same time. ASM’s *Furnace Atmosphere Controls in Heat Treating* identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as furnace variables used to control carbon potential and limit decarburization. Eurotherm’s 2704CP Handbook supplement describes common estimation of carbon potential from zirconia-probe oxygen measurements combined with furnace temperature and carbon-monoxide content, or from dew-point measurement. A corrected part from a furnace that continues to produce low carbon potential is not a process correction; it is a temporary sorting action.

### When correction is not equivalent to restoration

Removing a soft surface does not restore lost material, altered geometry, or the original heat-treatment history. This distinction is central to the ASM chapter *Steel Decarburization—Mechanisms, Models, Prevention, Correction, and Effects on Component Life*. Decarburization results from carbon moving through austenitic iron toward a surface whose carbon activity is lower than that of the steel, while surface reactions with the furnace atmosphere consume carbon. The final zone depends on diffusion, temperature, exposure time, carbon activity, and surface-reaction kinetics.

A reheat may replenish carbon only under controlled carburizing conditions, and it may introduce new problems: grain growth, distortion, oxidation, retained austenite, excessive case carbon, or a hardness profile different from the original specification. Reheating a quenched-and-tempered component is not equivalent to repeating the original treatment unless the complete thermal cycle, atmosphere, cooling rate, and tempering response are qualified. For a precipitation-hardening, tool, spring, or bearing grade, the required correction may involve a full requalification rather than local surface treatment.

Carbon loss also changes microstructure before material is removed. A decarburized region may contain ferrite where the intended treatment would produce pearlite, bainite, or martensite. Grinding away that region does not reverse dimensional changes, decarburization beneath an inaccessible feature, or a heat-treatment condition that was exceeded during processing. Nor does it prove that the remaining steel has the specified core hardness, case depth, toughness, or residual-stress state.

Temperature alone cannot predict the result. A 2021 *Journal of Materials Research and Technology* study on Fe-0.6C-1.8Si-0.8Mn spring steel tested from 700 to 1000 °C found that decarburized ferrite thickness increased up to approximately 800–825 °C and then decreased. That non-monotonic behavior reflects competing diffusion and oxidation effects. Atmosphere records therefore matter. Linde reports that water vapor in N2/H2 atmospheres causes decarburization and cites less than 5% hydrogen as an experimental guideline limiting carburizing and decarburizing rates under specified sintering conditions; that value is not a universal furnace rule.

Acceptance must come from the applicable product specification, drawing, heat-treatment standard, or contractual requirement. No single decarburization depth is valid for every grade or component. The controlling document may specify total, partial, or ferritic decarburization, a hardness profile, a carbon limit, a location-specific requirement, or no limit because subsequent machining removes the zone. If the requirement is silent, engineering disposition should be based on measured evidence and service analysis rather than an assumed industry-wide allowance.

### Effects on fatigue, wear, strength, and component life

Fatigue is often the most sensitive failure mode because cracks initiate at or near surfaces where cyclic tensile stress, roughness, inclusions, notches, and residual stress interact. A decarburized layer can reduce surface hardness and yield strength, increase local plastic strain, and alter the hardness gradient beneath the surface. Under bending or torsion, that layer may lie exactly where the alternating stress is highest. Removing it may restore fatigue performance only if the finished geometry, surface finish, residual stress, and subsurface microstructure remain acceptable.

Wear behavior depends on contact pressure, sliding or rolling conditions, lubrication, and the hardness profile. A soft ferritic layer on a gear flank, cam, rail, shaft journal, or bearing race can deform, smear, gall, or wear rapidly. In rolling contact, the transition between a soft surface zone and harder material can concentrate strain and promote pitting or spalling. A component may pass a tensile test yet fail early in service because its contact surface lacks the specified hardened layer.

Static strength consequences depend on section loss and the location of the defect. For a large, lightly stressed section, a shallow surface zone may have little effect on ultimate tensile strength. In a small spring wire, thin-walled part, sharp fillet, or highly loaded tooth root, the same depth can consume a significant fraction of the load-bearing section. Reduced carbon may lower hardness and strength locally, while grinding can reduce the net section further. A geometry change that appears dimensionally minor can materially increase stress concentration.

Component life therefore cannot be inferred from decarburization depth alone. Grade, carbon content, alloy hardenability, affected depth, hardness gradient, surface stress, loading mode, stress ratio, notch severity, residual stress, environment, and inspection sensitivity all matter. The correct disposition may be accept, remove and re-inspect, reprocess through a qualified heat treatment, downgrade, or reject. The decision should state which condition was measured, which requirement controls, how much material was removed, and what evidence demonstrates that the remaining component—not merely its appearance—can perform for its intended life.

## A Standards-Based Investigation and Reporting Workflow

Decarburization should be investigated as a conformity question, not as a visual impression. A dark scale layer, a ferritic band, a hardness drop, and a measured carbon gradient are related observations, but they are not interchangeable results. The investigation must connect the product requirement, the original surface, the examination method, the thermal exposure, and the furnace atmosphere.

ISO 3887:2023, *Steels — Determination of the depth of decarburization*, specifies three methods for measuring decarburization depth in steel products. ASTM E1077-14(2021), *Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens*, covers screening, microscopic, microindentation-hardness, and chemical-analysis procedures. SAE J419\_201801, *Methods of Measuring Decarburization*, gives recommended microscopic, hardness, and chemical-analysis practices for ferrous materials. These documents provide methods; they do not, by themselves, decide whether a particular heat, bar, forging, sheet, or finished component is acceptable.

### Define the question and applicable acceptance requirement

Begin by stating what must be decided. Possible questions include:

- Is the material fully decarburized, partially decarburized, or free from measurable decarburization?
- What is the maximum allowable depth from the original surface?
- Does the requirement apply to one surface, all surfaces, a specified sampling length, or the entire lot?
- Is the limit expressed as a distance, a carbon concentration, a hardness change, a metallographic boundary, or a permitted machining allowance?
- Does the requirement concern the supplied product or the finished component after grinding, turning, shot blasting, or other surface removal?

The answer must come from the governing purchase specification, product standard, drawing, heat-treatment specification, or customer acceptance document. Record the exact designation and revision. For example, identify a material as **AISI/SAE 1045** only when that designation is supported by the material records, and separately identify the product standard, such as **ASTM A29/A29M**, when that standard governs the bar. Do not replace a specified grade with a similar nominal carbon grade. **SAE J403 Grade 1045**, an ASTM product specification, and an internal grade designation may have related chemistry but different sampling, processing, or acceptance provisions.

The investigation also needs a definition of “depth.” A total-decarburization depth may end where carbon-rich phases have disappeared; a partial-decarburization depth may extend to the point where the carbon level or microstructure returns to the unaffected core. A hardness traverse can show a transition without proving the exact carbon profile. Conversely, a chemical profile can identify carbon loss while a transformed or tempered surface makes the microstructural boundary less obvious.

Write the acceptance rule before examining the specimen. If the requirement says “maximum 0.25 mm partial decarburization,” the report should not substitute a mean value, a visual estimate, or the deepest point found on a differently oriented section. If no acceptance criterion exists, report the measured condition and identify the missing decision rule rather than declaring the material acceptable.

### Select and document the examination method

Preserve the original surface during sampling. Mark the surface before cutting, prevent grinding or abrasive contact that could remove the affected layer, and record whether scale, oxidation, plating, coating, shot residue, or machining damage was present. A specimen taken after surface dressing cannot establish the original decarburization depth unless the removed allowance is known and documented.

Sampling location and orientation matter. For bar, wire, tube, plate, and forgings, identify the product axis, transverse or longitudinal section, outside or inside surface, corner, edge, and distance from the end. Include locations likely to experience different furnace exposure, such as top and bottom faces, radiused corners, ends, or areas shielded by fixtures. When the complaint concerns a localized defect, sample through the reported location and retain a nearby unaffected comparison location if possible. Record the number of specimens and the reason each location was selected.

Choose the method according to the question and the required decision strength. ISO 3887:2023 provides three measurement approaches; the selected approach, edition, and applicable clause should appear in the report. ASTM E1077-14(2021) permits screening, microscopic examination, microindentation-hardness examination, and chemical analysis. Screening is useful for locating suspect regions, but it should not be presented as a referee result when the acceptance limit is close to the measured value. ASTM E1077-14(2021) designates rigorous quantitative or lineal analysis as the referee approach in disputes.

Microscopy is often the most practical first quantitative method. Prepare a section normal to the surface, preserve the edge, and state the cutting, mounting, grinding, polishing, and etching sequence. Identify the etchant and examination magnification. The operator should define the boundary criterion before measuring: disappearance of pearlite or other carbon-rich constituents, change in ferrite morphology, a specified image threshold, or another criterion required by the selected standard. Report whether the result represents total decarburization, partial decarburization, or both.

Microindentation hardness provides a profile rather than a direct carbon analysis. State indenter type, test force, spacing, distance from the surface, orientation of the traverse, calibration status, and the rule used to identify the transition. Indents placed too close to the edge, too near one another, or across a steep microstructural boundary can distort the result. Hardness is also affected by grain size, phase transformation, tempering, residual stress, and retained austenite, so a hardness decrease should be correlated with microscopy.

Chemical analysis may be required when the dispute concerns carbon concentration rather than a visible boundary. State the technique, sampling increment, analyzed area or mass, depth resolution, calibration material, detection limit, and whether the result is a point, layer-average, or reconstructed profile. A carbon gradient measured with coarse layers cannot support a fine depth claim.

Calibration and preparation records are part of the result. Record instrument identification, calibration date, reference standards, magnification verification, hardness verification, analyst, and measurement uncertainty or repeatability. Measure multiple fields or traverses where the standard and acceptance plan permit it, and report individual values, mean, range, and maximum when each has a different decision significance.

### Link furnace records to metallurgical evidence

A measured surface zone is evidence of the final material condition, not a complete explanation of its origin. Review the full thermal history: furnace identification, charge number, loading pattern, heating rate, austenitizing or sintering temperature, soak time, transfer time, cooling conditions, and any reheating or normalizing operation. Match specimen identity to the furnace chart and product traceability record. A furnace record from the correct date but the wrong charge is not supporting evidence.

Atmosphere data must be read with temperature and time. ASM’s *Furnace Atmosphere Controls in Heat Treating* identifies water-vapor concentration, carbon-dioxide concentration, and oxygen partial pressure as variables used to control furnace carbon potential and prevent decarburization. Eurotherm’s *2704CP Handbook supplement* describes common inference of carbon potential from zirconia-probe oxygen measurements combined with furnace temperature and carbon-monoxide content, or from dew-point measurement. Record the actual sensor values, alarm events, calibration checks, probe recovery, gas flow, and known sensor outages; do not treat a controller set point as proof of the atmosphere at the load.

The mechanism is coupled. Carbon diffuses through austenitic iron toward a surface where carbon activity is reduced by reactions involving oxygen-bearing furnace gases. Surface-reaction kinetics, diffusion distance, temperature, exposure time, and carbon potential all contribute. NIST describes the decarburization zone as near-surface loss of carbon and carbon-rich phases when surface carbon reacts with oxygen to form gaseous carbon monoxide or carbon dioxide. Water vapor in N2/H2 atmospheres can also cause decarburization; Linde’s 2023 sintering guidance reports an experimental guideline that less than 5% hydrogen limited carburizing and decarburizing rates under the stated conditions, not a universal furnace rule.

Temperature alone cannot explain every result. A 2021 *Journal of Materials Research and Technology* study on Fe-0.6C-1.8Si-0.8Mn spring steel tested from 700 to 1000 °C found that decarburized ferrite thickness increased to approximately 800–825 °C and then decreased. Diffusion and oxidation therefore must be considered together.

A report should contain the specimen identity and heat or lot; grade and governing specification; original surface condition; sampling location and orientation; furnace and thermal history; atmosphere records; examination standard and clause; preparation details; instrument calibration; measurement traverse or chemical profile; total and partial depths; uncertainty, repeatability, and limitations; photographs or maps; comparison with the acceptance requirement; and the disposition rationale.

The reproducible workflow is direct: identify the grade and thermal history, preserve the original surface, select and document locations and orientation, choose an ISO, ASTM, or SAE method, define preparation and measurement criteria, compare the measured profile with the governing requirement, and correlate the finding with atmosphere, temperature, time, and sensor records. Only then should the material be accepted, reworked, subjected to further examination, or rejected.

## References

1. \[1\] International Organization for Standardization. [Steels — Determination of the depth of decarburization](https://www.iso.org/standard/86689.html). ISO standard, 2023. [](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#wiki-cite-ref-1) https://www.iso.org/standard/86689.html
2. \[2\] ASM International. [Furnace Atmosphere Controls in Heat Treating](https://dl.asminternational.org/handbooks/edited-volume/8/chapter-abstract/106617/Furnace-Atmosphere-Controls-in-Heat-Treating). ASM Handbook, 2013. [](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#wiki-cite-ref-2) https://dl.asminternational.org/handbooks/edited-volume/8/chapter-abstract/106617/Furnace-Atmosphere-Controls-in-Heat-Treating
3. \[3\] Linde. [Sintering of steels](https://static.prd.echannel.linde.com/wcsstore/EE_REN_Industrial_Gas_Store/pdf/Sintering-of-steels-brochure-EN.pdf). Linde technical guidance, 2023. [](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#wiki-cite-ref-3) https://static.prd.echannel.linde.com/wcsstore/EE\_REN\_Industrial\_Gas\_Store/pdf/Sintering-of-steels-brochure-EN.pdf
4. \[4\] ASTM International. [Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens](https://store.astm.org/e1077-14r21.html). ASTM standard, 2021. [](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#wiki-cite-ref-4) https://store.astm.org/e1077-14r21.html
5. \[5\] ASTM International. [Standard Test Methods for Estimating the Depth of Decarburization of Steel Specimens](https://store.astm.org/e1077-14r21.html). ASTM standard, 2021. [](/wiki/heat-treatment/steel-decarburization-causes-measurement-and-control#wiki-cite-ref-5) https://store.astm.org/e1077-14r21.html

 **At a glance**

Primary phenomenon

Near-surface loss of carbon or carbon-rich phases

Affected region

Decarburized zone

Common gaseous products

CO and CO₂

Main controlling factors

Carbon activity, atmosphere chemistry, diffusion, temperature, time, and test method

 [Back to Heat Treatment](/wiki/heat-treatment "Heat Treatment — Steel Wiki")