Guides / Specs
Hydrogen Embrittlement and the Relief Bake: What to Put on the Drawing
October 8, 2026
Hydrogen embrittlement is the reason a hardened steel part can pass every dimensional and visual check after plating, then crack weeks later under a load it was designed to carry. It is a real and well understood failure mode, not a rare edge case, and it is also fully preventable with a bake step most shops run routinely once it is called out.
What actually happens
Electroplating processes, and the acid cleaning steps that usually precede them, can drive atomic hydrogen into the surface of steel. In lower strength steel this hydrogen mostly diffuses out harmlessly over time. In high strength steel, the hydrogen can collect at grain boundaries and around inclusions, reducing the metal’s ability to deform before fracturing. Under sustained tensile stress, a crack can start and grow slowly with no external sign until the part fails.
The mechanism is specific to hydrogen introduced during processing, not to the coating itself sitting on the part. A hardened bolt with a hydrogen-embrittled substrate can fail even though the zinc or nickel on its surface looks and measures perfectly.
Why hardness is the trigger
Higher strength steel is more sensitive to hydrogen embrittlement because it has less ability to redistribute stress by deforming locally around a hydrogen-affected region. Softer steel tolerates the same hydrogen content without the same crack risk. The threshold commonly used in industry, including in fastener specs such as ASTM F519, is around 39 HRC, though the real variable is a combination of hardness, applied stress and the specific steel’s sensitivity, not hardness in isolation.
If any part in a lot is heat treated above roughly this hardness and will be electroplated, plan for a bake. Treating the threshold as a hard line rather than a rough guide is the safer approach when in doubt.
What the relief bake does
Baking the part in an oven soon after plating, at a moderate temperature for a period of hours, allows the absorbed hydrogen to diffuse back out of the steel before it can concentrate enough to cause cracking. ASTM B850 is a commonly referenced spec for post-plating baking of steel parts, giving time and temperature guidance by hardness range. The bake does not remove hydrogen instantly and does not work if delayed too long after plating, since embrittlement can begin before the part ever reaches the oven.
Timing matters as much as the bake itself. A shop should start the bake within a defined window after plating, commonly within a few hours, not at the end of the day once a rack happens to be free.
What to put on the drawing
State the requirement plainly rather than assuming a shop will infer it from the hardness callout:
- Name the spec, such as ASTM B850, or a customer-specific bake requirement if one applies.
- State the part’s hardness or heat treat condition so the shop selects the correct bake time and temperature.
- Require the bake time, temperature and start time relative to plating to be recorded and included on the certificate of conformance.
A drawing that specifies hardness and a plated finish but says nothing about baking leaves the decision to the shop’s default process, which is a gap worth closing before parts run, not after a failure.
Which parts are most at risk
Hydrogen embrittlement failures concentrate in a predictable set of parts: fasteners above roughly Grade 8 or Class 10.9, springs, landing gear and other high-strength aerospace structural components, and any hardened steel part that carries sustained tensile stress in service. The common factor is not the part’s shape but the combination of high hardness and a load that stays applied, rather than one that cycles or releases.
A part that sees the same high hardness but only intermittent or compressive loading carries lower practical risk, though the safer approach on any hardened, electroplated part is still to bake unless there is a specific reason not to. Fasteners are overrepresented in embrittlement failure reports mainly because they are both hardened and permanently loaded in tension once torqued, which is close to a worst case for the mechanism.
Testing for hydrogen embrittlement
Because embrittlement cracking can take days or weeks to show up, standard inspection at the time of plating cannot catch it directly. ASTM F519 is the specification most commonly used to qualify a plating process for hydrogen embrittlement risk before it runs on production parts: specially prepared notched test specimens are plated alongside the qualification lot, then sustained under load for a defined period, commonly 200 hours, to see whether they crack.
This qualification testing is done periodically to validate a shop’s process and bake parameters, not on every production lot, since the notched specimens are destructive and the test takes days to run. A shop that can show current ASTM F519 qualification data for the plating line running your parts has demonstrated the bake actually works for that process, rather than assuming a generic time and temperature is sufficient.
Alternatives that avoid the mechanism
For hardened fasteners in particular, coatings applied without electroplating current or strong acid activation carry little to no hydrogen embrittlement risk:
- Zinc flake coatings are cured rather than electroplated and are widely used on hardened fasteners for exactly this reason.
- Mechanical zinc plating deposits zinc by tumbling parts with metal media and a chemical promoter rather than electric current.
Neither replaces zinc-nickel or hard chrome for every application, since coating properties differ, but both are worth asking about when a hardened part needs corrosion protection and the bake adds cost or lead time you would rather avoid.
How we help
We are a matching service, not a plating shop. If a part in your RFQ is hardened steel, tell us the hardness and we pass the request to shops that run relief bakes routinely and record them on the certificate, not shops that treat the question as unusual. See the coating RFQ checklist for the rest of what to include, or go straight to the RFQ form.
Frequently asked questions
Does hydrogen embrittlement happen with every plating process?
The risk is highest with electroplating and with acid cleaning or pickling before plating, both of which can drive atomic hydrogen into the steel. Electroless nickel carries lower risk than electrolytic plating but is not zero. Zinc flake and mechanical zinc coatings, applied without electric current or strong acid, avoid the mechanism almost entirely.
What hardness level actually needs a relief bake?
The common threshold cited is about 39 HRC, drawn from ASTM F519 and similar aerospace fastener specs, but the real driver is applied stress and material sensitivity, not hardness alone. Below that threshold many specs waive the bake by default; above it, plan for one unless the coating process specifically avoids hydrogen uptake.
How long does the relief bake actually take?
Commonly a few hours at moderate oven temperature, per specs such as ASTM B850, though exact time and temperature depend on the steel grade, hardness and the spec your drawing calls out. It adds roughly a day to lead time once queue and oven scheduling are included.
Can a part fail from hydrogen embrittlement weeks after it was plated and put into service?
Yes, and that delay is what makes the failure mode dangerous. A part can pass inspection, get assembled and see load for days or weeks before a hydrogen-embrittled crack propagates under sustained stress. This is why the bake window after plating matters, not just whether a bake eventually happens.
Need a quote for this finish?
Send the drawing. We match you with up to 3 finishing shops that run your spec and reply within 24 business hours.
Get matched with a coating shop