SCREWKING
Where the fastener standards stop
The short version: the international fastener standards were written for ordinary sizes. Work downwards and you meet, in turn, a test that cannot be run, a method that does not apply, and a tolerance with nothing set aside in it. None of this is an oversight — each limit is stated in the scope. But if your plan is “we follow ISO”, below a certain diameter there is less to follow than you think, and that usually surfaces in production rather than at sample stage.
Four unrelated standards, one pattern
| What you want to do | Standard | Lower limit | In coarse thread |
|---|---|---|---|
| Torque / clamp force test | ISO 16047 | M3 | No reference conditions below M3 |
| Low-temperature impact | ISO 898-1 (Charpy) | Nominal dia. 16 mm | Impossible at any size we make |
| Decarburisation, hardness method | ISO 898-1 Table 18 | Pitch P ≥ 1.25 mm | Not reached until M8 |
| Tolerance allowance for plating | ISO 965-2 | 6g runs M1.6–M100 | M1–M1.4 use 6h — zero upper deviation |
Add the scope itself: ISO 898-1 covers coarse thread from M1.6 to M39, so M1, M1.2 and M1.4 fall outside it entirely.
The decarburisation limits and the microscopy rule belong to the decarburisation and carburisation regime for property classes 8.8–12.9. They are not blanket requirements for every fastener in ISO 898-1.
A stated limit usually comes with a route around it
This matters more than the limits themselves, because it determines what you should be asking. In most cases the standard names the alternative:
- Decarburisation. Below P = 1.25 mm the hardness method is not permitted, and ISO 898-1 requires the microscopic method instead — longitudinal section, mounted, ground, polished, etched. The route is open; it is simply a different one.
- Tensile. Short parts cannot be tensile tested, so ISO 898-1 substitutes minimum hardness for products under 2.5×d in nominal length.
- Torsional strength. ISO 898-7 covers 1–10 mm. But it answers a different question. ISO 898-7 measures the torque that breaks a screw. ISO 16047 measures the clamp force a given torque produces. The first is strength; the second is what you actually control.
Note what “no reference conditions” does not mean. Torque/clamp testing below M3 is not impossible or forbidden — there is simply no standard fixture, sensor specification or acceptance criterion, so those have to be agreed between you and the supplier rather than cited.
Fixed limits bite harder as parts shrink
Standards carry two kinds of limit: those expressed as a proportion, which scale, and those expressed as an absolute figure, which do not.
| Item | Limit | Type |
|---|---|---|
| Complete decarburisation depth G | max 0.015 mm | Absolute — does not scale |
| Non-decarburised height E | ½, ⅔ or ¾ × H1 | Proportional — scales with thread height |
Plating behaves the same way. A single-side coating of t on a 60° thread raises pitch diameter by 4t, and the coating thickness does not shrink with the screw. The Chinese article covers this in detail →
What to ask
“Do you comply with ISO?” is a weak question in this size range. Ask instead what happens where the standard stops:
- How is decarburisation checked — microscopy or hardness? If hardness, at what pitch? (Below 1.25 mm it should not be.)
- How is it controlled below M1.6, which ISO 898-1 does not cover — by contractually specifying another metallographic standard, or through the process: incoming wire, furnace atmosphere, batch discipline?
- Where is the plating allowance taken from? Is the drawing tolerance before or after coating, and is acceptance a thickness report or a GO gauge after plating?
- Where do the torque figures come from — measured or calculated? And how is scatter in the nut factor managed? At this size scatter matters more than the mean.
- What performs 100% inspection below M2? Commercial optical sorters are typically specified M2–M16.
One quick test of a supplier: ask which standard tests cannot be run at your size. Anyone who has actually produced it can answer. “We comply with all ISO standards” usually means the scope clauses have not been read.
References
- ISO 898-1 — Mechanical properties of fasteners made of carbon steel and alloy steel — Part 1 (scope; Table 3 items 14–15; Table 18 note a; clause 9.10)
- ISO 898-7 — Torsional test and minimum torques for bolts and screws with nominal diameters 1 mm to 10 mm
- ISO 16047 — Fasteners — Torque/clamp force testing (incl. Amd 1:2012)
- ISO 965-2 — ISO general purpose metric screw threads — Tolerances — Part 2
- ISO 4042 — Fasteners — Electroplated coating systems (an Amendment 1:2026 exists; its technical content is not public and is not reflected here)
Standards are paid documents. Figures quoted here follow the clauses cited; the current edition of each standard governs. Purchase and acceptance should rest on the drawing agreement and the standard itself, not on this page.
Enquiries
We produce M1.0–M5.0 in volume — the range where these standards start going quiet. Send the drawing and we will tell you which of these cannot be done by the book at your size, and how it is handled in practice, before quoting.