Quenching vs tempering: two halves of one process
One of the most searched pairs in heat treatment — and the question is framed wrongly. Here is what each step actually does, and which one decides the hardness you ordered.
Quenching vs Tempering
Short answer: quenching creates the hardness, tempering gives back toughness and sets the final HRC
Quenching heats steel into the austenite range — roughly 820–880 °C depending on grade — then cools it fast in oil, water or a polymer so the structure becomes martensite: the hardest, most brittle and most highly stressed state the part will ever be in. Tempering puts that hardened part back in a furnace below the transformation temperature, typically 150–650 °C, so the martensite tempers, carbides precipitate, residual stress relaxes and toughness comes back. They are not alternatives: together they are one route, quench and temper (Q&T). Quenching sets the ceiling — capped by the carbon content of the steel — and the tempering temperature chooses where on the way down you stop. No fastener is delivered in the as-quenched condition.
Quenching versus tempering
| Decision factor | Quenching | Tempering |
|---|---|---|
| What it is | Heat into the austenite range, then cool rapidly | Reheat the hardened part below the transformation temperature, then cool |
| Typical temperature | About 820–880 °C by grade, then oil, water or polymer quench | About 150–650 °C, chosen for the hardness required |
| What changes in the steel | Austenite transforms to martensite: hard, brittle, heavily stressed | Martensite tempers, carbides precipitate, residual stress relaxes |
| Effect on hardness | Takes hardness to its maximum; the ceiling is set by carbon content | Lowers hardness predictably — temperature is the control knob |
| Effect on toughness | The lowest point of the whole route | Restores toughness and impact resistance |
| Residual stress and crack risk | Highest — quench cracks start here, at sharp corners, cross holes and thread roots | Substantially reduces residual stress |
| Can it be used alone? | No — an as-quenched fastener is not a usable part | Only meaningful on a part that was hardened first |
| What controls the result | Grade, section size, quenchant, agitation and part geometry | Tempering temperature above all, then hold time |
| What the standard requires | ISO 898-1 requires quench and temper for class 8.8 and above | ISO 898-1 sets a minimum tempering temperature: 425 °C for 8.8, 9.8 and 10.9; 380 °C for 12.9 |
| Common failures | Quench cracks, soft spots, distortion, retained austenite | Over-tempering (soft parts), tempering below the class minimum to chase a hardness number, holding in an embrittling range |
What each step decides for you
Quenching
Quenching decides the ceiling. A low-carbon steel will never reach the hardness of a high-carbon one whatever you do to it, and quenching also decides whether the core hardens at all — that is hardenability against section size. If the question is "why did the parts not reach the specified hardness", the answer is almost always in this step, not in the temper.
Tempering
Tempering decides what is delivered: the HRC on the drawing, the toughness, and how much residual stress is left. If the question is "why are the parts too hard, too brittle, or cracking on installation", the answer is usually tempering temperature and time. On graded fasteners, never temper below the class minimum to make a hardness number pass — you get the number and lose the toughness the standard exists to guarantee.
Limits before putting it on the drawing
- They are not alternatives, so "quench or temper?" has no answer. If you want a soft, machinable part, you want annealing or normalizing, not tempering.
- Temperatures and hardness ranges here are typical guidance for small fasteners and depend on the grade, the section and the governing specification — verify by hardness test.
- The same tempering temperature gives different hardness on different grades: 400 °C on S45C and on SCM440 are not the same result. Work from the grade’s own tempering curve.
- The delay between quenching and tempering matters — parts left in the as-quenched state carry high stress and can crack in the basket.
- Tempering does nothing for a part that was never hardened, and it cannot repair decarburisation that has already happened.
- Parts that will be zinc plated and sit at or above roughly 320 HV (class 10.9 and up) need de-embrittlement baking per ISO 4042 — a separate operation from tempering.
What to send the heat treater
- 1Steel grade and mill certificate
- 2Required hardness range, test location and method (HRC/HV per ISO 6508 / ISO 6507)
- 3Governing fastener class if one applies, e.g. ISO 898-1 class 8.8 / 10.9 / 12.9
- 4Thickest section and the risk features — sharp corners, cross holes, thread roots
- 5Distortion limits and the downstream step (plating, de-embrittlement baking)
- 6Lot quantity or weight, plus sample pieces for sectioning
Frequently asked questions
Can I choose quenching or tempering?
No — they are consecutive steps of one process. Quenching alone leaves a part that is hard but brittle and highly stressed, so it is not usable. Tempering alone, on a part that was never hardened, changes almost nothing.
Which step sets the HRC I ordered?
The tempering temperature sets the delivered value; quenching sets the maximum the grade can reach. If that ceiling is below your figure, the grade is wrong for the requirement — no tempering adjustment fixes it.
Can we skip tempering to keep the parts hard?
It is not advisable. An as-quenched part carries very high residual stress and can crack in storage or handling, and for graded fasteners ISO 898-1 sets a minimum tempering temperature, so skipping the step means the parts do not meet the class you specified.