Decarburization — surface carbon loss after heat treatment
Soft skin on hardened screws and bolts: how partial and complete decarburization fail ISO 898-1, how to measure it by metallographic section or microhardness traverse, and the furnace controls that prevent it.
| Defect family | Heat treatment |
|---|---|
| Related service | Heat Treatment |
| Reviewed |
Standards commonly cited
Grades this affects
What you see
Hardness is correct in the core but low at the surface — a file bites the thread crest of a bolt that tests in range on the head. Threads gall, strip or round off at nominal torque, self-tapping screws stop cutting, and after plating the finish can look dull or patchy. On a polished and etched section, the thread flank carries a light, ferrite-rich rim that does not darken with the martensite behind it.
How to confirm it
Section a thread axially, mount, polish and etch, then judge it against ISO 898-1: a fully ferritic surface layer (complete decarburization, G) is not permitted, and the height of the unaffected zone E must stay above the minimum for the property class. Two methods are recognised — metallographic examination at the specified position on the thread flank, and a microhardness traverse to ISO 6507-1 / ASTM E384 from the surface inwards, comparing the near-surface reading with the reading at the reference depth. ISO 3887 and ASTM E1077 describe the general measurement. Always measure at the thread, not on a convenient flat: the thread is where the carbon leaves and where the load is carried.
Related calculator: Hardness by steel grade — reference tableRoot causes — most common first
- 1Furnace atmosphere carbon potential below the carbon level of the steel, so an air or weakly protective atmosphere strips carbon at austenitising temperature.
- 2Too long at temperature — affected depth grows roughly with the square root of time, so an overloaded belt run slow to “make sure it hardens” also makes sure it decarburizes.
- 3Leaks and dilution: a worn door curtain, low gas flow, an unstable dew point or air ingress at the charge end.
- 4Wire rod that arrived decarburized — the skin was lost in hot rolling, patenting or spheroidize annealing, and cold heading does not remove it.
- 5No stock removed from the loaded surface: threads rolled before heat treatment keep the affected rim on the flank that carries the load.
- 6Scale and rust burned in during preheat, which consumes carbon and hides the loss until the part is sectioned.
Process controls that prevent it
- Run a controlled protective atmosphere with carbon potential matched to the grade, and verify it by dew point, oxygen probe or shim stock rather than trusting a controller set-point.
- Keep time at temperature to what the section actually needs — small fasteners through-harden quickly, and long soaks buy nothing but scale and carbon loss.
- Inspect incoming wire rod for decarburization before heading: a mill certificate stating the limit plus one sectioned sample stops the problem entering the plant.
- Where the design allows, roll threads after heat treatment, or leave grinding stock so the affected rim is removed from the loaded surface.
- Put a sectioned thread into lot release for hardened class 10.9 and 12.9 parts instead of relying on one surface hardness reading.
- Maintain seals, curtains and gas flows and log the work — decarburization failures cluster around maintenance events.
What to send us for diagnosis
- Six to ten parts from the suspect lot, untumbled and unplated where possible, plus parts from a lot that passed.
- The drawing and property class, the material grade, and the wire-rod mill certificate if you have it.
- Where the low hardness was measured — head, shank, thread crest or flank — with the method, scale and load used.
We reply to quote requests within 24 hours.
Part types this shows up on
Questions engineers ask
Does decarburization always mean the heat treatment was wrong?+
No. A large share of cases arrive with the wire rod. Carbon lost in rolling or annealing is still missing after heading, and no furnace puts it back during a short hardening cycle. Sectioning an as-received sample separates the two causes and decides who owns the problem.
Can decarburization be reversed?+
Not practically on a finished fastener. Restoring carbon means a carburizing cycle that changes dimensions and core properties, and the value of the material rarely justifies it. Incoming inspection and atmosphere control are the economical answers.
How does it show up in service?+
As stripped or rounded threads, torque-tension results that miss target, and early fatigue cracks starting at the thread root — the soft rim yields first and transfers load onto fewer engaged threads.
Guidance based on published standards and general fastener metallurgy. Values are typical ranges that depend on the part, the material and the governing specification — the customer drawing and the applicable standard always prevail. This page does not replace a qualified failure analysis of your own parts.