The Cryogenic Heat Treatment Process: What Happens to Steel, Step by Step
Most people who search for “cryogenic heat treatment” want to know one thing: what actually happens to the part. Strictly, the process is a deep-cold step run inside the heat-treating cycle rather than a form of heat treating — we've laid out the distinction in full. This page is the process itself, in the order it happens, and what changes in the steel at each stage.
The Precondition: The Part Arrives Already Hardened
Cryogenic treatment doesn't create a hardened structure — it refines one. Every part we process has already been through its austenitize-and-quench, which means it arrives with the three things the quench leaves behind: hard martensite, a fraction of unstable retained austenite, and residual stress locked in by rapid cooling. The process below exists to deal with the second and third of those without disturbing the first.
That's also why the heat-treat history matters to us before a part ever goes in. The outcome depends on it.
The Cycle
Five stages. The middle three are the cryogenic process proper; the first and last are the heat-treating steps it sits between.
Austenitize and quench (already done)
The heat treater's work. The steel is heated into the austenite range and quenched fast enough to form martensite. Hardness is set here. So is the retained austenite — the quench never converts all of it — and the residual stress.
Controlled ramp-down
The part is cooled toward cryogenic temperature slowly and under control — over hours, not minutes. Steel contracts as it cools, and if the outside cools far faster than the core, the differential can crack the part or set new stresses in it. The earliest experiments in this field, a century ago, dropped parts straight into liquid nitrogen and got exactly that result.
Skip the control and you're thermally shocking a hardened part. The damage doesn't always show until it's in service.
Deep cryogenic soak
Below −244°F — the threshold the Cryogenic Society of America uses to define cryogenics — and in practice down to between −308°F and −312°F using liquid nitrogen, the part is held. That range is deliberate: it sits close to liquid nitrogen's own boiling point, and we've found the closer the soak gets to it, the stronger the effect on the alloy. Holding there steadily is a function of how well the chamber is insulated, not just how much nitrogen you pour in. This is where the work happens, and it happens slowly: the transformation has to reach through the full cross-section, not just the surface. Two things occur in steel and cast iron as the part sits at depth:
- Retained austenite transforms to martensite. The unstable structure the quench couldn't finish is finished.
- Fine eta carbides begin to precipitate throughout the structure — very small, very hard particles that carry a large share of the wear-resistance gain.
A brief dip to a shallow temperature will convert some austenite. It will not precipitate carbides, and it will not reach the core of anything thicker than a blade.
Controlled return to ambient
Back up to room temperature, as slowly as it went down. The same thermal-differential rule applies in reverse. As the structure warms it settles into a tighter, more uniform crystalline alignment than it had going in — this is the residual-stress relief, and it's what a customer sees later as dimensional stability. Any brittleness the part shows at depth dissipates on the way back; it isn't a property of the finished part.
Temper
The part goes back to a moderate temperature. The martensite created in stage 3 is fresh and untempered, exactly as quench martensite is, and it needs the same tempering to trade a little hardness for the toughness a working part requires. This is the step that makes the sequence — quench, cryogenic, temper — matter. Run the cryogenic stage after the final temper instead and you've left untempered martensite in the part.
Deep and Slow vs. Shallow and Fast
Every step above has a shortcut, and the shortcut is usually what separates a real cryogenic process from something that borrows the name. The difference isn't cosmetic. It's whether the part gets carbide precipitation and stress relief through its full section, or a surface layer of converted austenite and a thermal shock.
Cold treatment or a quick dip
- Around −140°F (dry ice), or a short plunge into liquid nitrogen
- Converts some retained austenite near the surface
- No carbide precipitation
- Uncontrolled cooling risks thermal shock
- Often no temper afterwards
Deep cryogenic treatment
- Below −244°F, held at −308 to −312°F
- Slow, controlled ramp-down and return
- Held long enough to transform the full cross-section
- Retained austenite converted, fine carbides precipitated, residual stress relieved
- Tempered afterwards, as the new martensite requires
Deeper still is possible. Using liquid helium rather than nitrogen we can take parts to around −430°F, which is where certain aerospace stress-relief specifications live. Helium costs orders of magnitude more than nitrogen and the process is reserved for work that genuinely calls for it — but the capability is there when a specification demands it.
Beyond the Cycle: The Surface
The cryogenic cycle changes the steel all the way through. For parts whose failure is about friction and lubrication at the surface — slides, barrels, gears, anything that runs against another part — we pair it with two surface treatments: micro-dimpling, a high-speed particle process that leaves a controlled texture of microscopic peaks and dimples so two surfaces meet at far fewer points, and a dry film lubricant mechanically embedded into that texture rather than coated over it. The cryogenic treatment underneath is what keeps those peaks durable.
Which Steels — and Which Metals
The full mechanism — austenite conversion and carbide precipitation — belongs to steel and cast iron, which makes hardened tool steels the clearest case: industrial tooling, D2 stamping dies, punches, cutting tools, gears, and wear components across the case studies we've published.
Titanium, Inconel, aluminum and carbides don't have retained austenite to convert, but they do carry residual stress and benefit from the same controlled deep-cold cycle through the tighter, more uniform structure it produces — which is why they're on the list of materials we process. What a given alloy gets from the treatment is a materials question, and it's one we'll answer straight for your part rather than promise the steel result to everything.
Frequently Asked Questions
How long does the cryogenic heat treatment process take?
Longer than most people expect, because the ramp-down and return are slow by design and the soak has to reach the full cross-section. It depends on the mass and section thickness of the load. Tell us the part and we'll tell you the schedule.
Can the part be cryogenically treated before it's heat treated?
No. There's no martensite or retained austenite to work on until after the quench. The sequence is quench, cryogenic treatment, temper.
Will it change the part's dimensions?
The process relieves residual stress and converts the unstable austenite that would otherwise transform later in service — which means the part is more dimensionally stable afterwards, not less. Parts prone to drifting or warping over time are among the best candidates.
Does the steel become brittle?
At cryogenic temperature, many metals are temporarily brittle. That dissipates completely as the part returns to ambient, and the temper afterwards addresses the new martensite. It is not a property of the treated part.
Is one cycle enough?
For a properly hardened part run through a full, controlled cycle, yes. Re-treating is usually a sign the first treatment was shallow or rushed.
Send Us the Part and Its Heat-Treat History
Material, hardness, section thickness, and how it fails in service. That's enough for us to tell you what the process will and won't do for it.