Distortion after heat treatment — bend, ovality, pitch drift
Bent screws, out-of-round nuts and threads that stop gauging after hardening: released residual stress, uneven quenching and loading faults, with the before-and-after measurements that prove it.
| Defect family | Dimensional |
|---|---|
| Related service | Heat Treatment |
| Reviewed |
Grades this affects
What you see
Parts that were straight and in tolerance before the furnace no longer are. Long screws bow, thin shanks bend, nuts and washers go out-of-round or dish, flatness and perpendicularity fail, and threads that gauged correctly now fail the GO gauge because pitch diameter and helix have moved. Distortion is systematic — the whole lot leans the same way — where handling damage is random.
How to confirm it
Measure the same pieces before and after, not against a memory of the drawing: run-out on centres or a V-block for straightness, geometric tolerances to ISO 4759-1, and GO/NO GO thread gauges to ISO 1502 with pitch diameter to ISO 965-1. Sort the failures by position and orientation in the basket — if the deviation always points the same way, the cause is support, loading or quench flow rather than material. A section is rarely needed, since distortion is a dimensional problem, but a hardness check confirms the transformation itself was normal.
Related calculator: Thread fit after plating — clearance checkRoot causes — most common first
- 1Residual stress released by heating: cold heading, thread rolling and machining lock stress into the part, and the first austenitising soak lets it out.
- 2Non-uniform quenching — one side cools first, so transformation happens at different times across the section and pulls the part out of shape.
- 3Unsupported geometry: long slender screws, thin washers and spring parts laid flat in a deep basket sag under their own weight at temperature.
- 4Volume change on transformation — martensite occupies more space than the structure it replaces, so sections grow, and grow unevenly wherever hardening depth varies.
- 5Heating or quenching faster than the section change tolerates, especially where a thin thread meets a thick head.
- 6Loading and stacking weight, which imprints its own pattern on the load and changes with every operator unless it is specified.
Process controls that prevent it
- Stress-relieve or normalise before hardening when parts arrive heavily worked, so the movement happens before the finished dimensions exist.
- Fix the loading: defined layer depth, no stacking on slender parts, trays or fixtures for long screws and flat parts.
- Preheat and heat more gradually where section changes are severe, and keep the load uniform so the whole basket sees the same cycle.
- Control quench entry and agitation so the part cools evenly — a symmetric part cooled asymmetrically is the classic cause.
- Agree the allowance on the drawing: thread and dimensional tolerances checked after heat treatment and after coating, not before.
- Measure a first-off after treatment and again after plating, so movement is attributed to the process that actually caused it.
What to send us for diagnosis
- Matched samples — the same parts measured before and after treatment, marked so the two states can be compared.
- The drawing with the geometric tolerances that failed, and the gauge or instrument used.
- Loading information: how the parts sit in the basket, layer depth, and whether they were stacked.
We reply to quote requests within 24 hours.
Part types this shows up on
Questions engineers ask
Can distortion be straightened afterwards?+
Sometimes, on ductile parts and within limits agreed on the drawing, but straightening a hardened fastener introduces its own stress and can crack it. On threaded parts straightening rarely restores the thread. Preventing the movement is cheaper than correcting it.
Is distortion the same as growth?+
No. Growth is the uniform volume change when austenite transforms to martensite, and it is predictable enough to allow for in the pre-treatment dimensions. Distortion is the non-uniform part — bend, taper, ovality — from uneven heating, uneven cooling and released residual stress.
Why do threads fail gauging only after hardening?+
Pitch diameter grows with the volume change and the helix moves with any bending. A thread rolled to the top of its tolerance band before treatment can be pushed out by growth alone. Rolling to the middle of the band — or after treatment where the design allows — solves most of these cases.
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.