Collet nuts, spanners and torque

More tooling problems come from this page than from any dimension chart on this site. An under-torqued nut lets the cutter pull out; an over-torqued one permanently deforms the collet; and assembling the collet into the nut in the wrong order damages the nut's retaining ring before you have cut anything.

ER nut torque

REGO-FIX published values for standard nuts, with the spread from three other published charts shown alongside. ISO 15488 covers collet, holder and nut dimensions — it does *not* specify tightening torque, so these are maker values by definition.
Size	REGO-FIX standard nut	Mini nut	Other published charts
ER8	6 Nm	— (mini only)	5–7 Nm · 8 Nm
ER11	24 Nm	16 Nm	15–18 · 24 · 25 Nm
ER16	56 Nm	24 Nm	30–40 · 50 · 61 Nm
ER20	80 Nm	28 Nm	50–60 · 70 · 91 Nm
ER25	104 Nm	32 Nm	70–100 · 90 · 115 Nm
ER32	136 Nm	—	100–130 · 100 · 152 Nm
ER40	176 Nm	—	140–170 · 130 · 171 Nm
ER50	240 Nm	—	200–250 · 150 · 300 Nm

The charts disagree by up to 2.5×, and we are not going to average them. ER11 at 24 Nm is corroborated across three sources and is safe to use. Above that the spread widens — one well-known chart publishes 300 Nm for ER50 against another's 150 Nm. Use the value published by the maker whose nut you actually own. If you do not know the maker, start at the REGO-FIX column and work up, checking for slip.

One caution worth reading before you trust a chart: one widely-circulated torque table publishes nut threads that conflict with the values we cross-checked elsewhere on this site (M36×2.0 and M44×2.0 for ER32 and ER40, against M40×1.5 and M50×1.5 from four independent sources). If a table gets the thread wrong, treat its torque column with suspicion too. See nut thread chart.

Small bores need less torque

Torque is not constant within a size — the rated value drops sharply for small collet bores. REGO-FIX publishes by bore band, and the numbers are far lower than most people expect:

Rated torque by collet bore, standard nut. This is why tightening a 1/8" collet to the full ER16 value crushes it.
Size & bore	Rated torque	Mini nut
ER11, below 3 mm	8 Nm	—
ER11, 3.0 – 7.0 mm	24 Nm	16 Nm
ER16, 1.5 – 3.5 mm	20 Nm	—
ER16, 5.0 – 10.0 mm	56 Nm	24 Nm
ER20, 1.5 – 6.5 mm	32 Nm	28 Nm
ER20, 7.0 – 13.0 mm	80 Nm	28 Nm
ER32, all bores	136 Nm	—

What “hand tight” actually is

Hand-tightening an ER32 nut delivers roughly 30–60 Nm — well under half the 136 Nm rating, and therefore well under half the rated clamping force. A firm pull on a 250 mm spanner gets you to about 100 Nm, which is adequate. For production work use a calibrated torque wrench; click-type ER spanners exist for exactly this.

Clamping force rises roughly in proportion to torque, so under-torquing costs you grip directly. Over-torquing does something worse — it plastically deforms the collet and pushes runout permanently above the ISO 15488 Class 2 limit. See runout and accuracy classes.

Nut families and the spanner each one needs

The spanner follows the nut family, not the ER size. One published chart also splits by drive type — slotted versus hex — within the same ER number.
Family	Sizes	Spanner	Why you'd choose it
A (standard)	ER8 – ER50	Hook / C spanner · slotted	Default. Widest availability.
M (mini)	ER8 – ER20	Hook spanner, smaller OD	Slimmer nose for clearance. Finer thread, lower torque rating.
UM (slim / ultra-mini)	ER16 – ER50	Hook spanner, UM pattern	Slimmest nose. Check thread — sources disagree on UM in ER16/ER20.
Hex nut	ER11, ER16, ER20, TG75, TG100	Open-end / socket	Easier to torque accurately.
Ball-bearing nut	ER11 – ER50	Same as standard	30–40% less torque for the same clamp; published runout improvement is claimed at 50% and measured closer to 20–30%. 4–8× the cost.

TG nuts are rated separately: one published chart gives 90 ft-lb (122 Nm) for TG75 and TG100 hex nuts, and the same figure for an SYOZ25 / TG100 hook nut. Single source — verify against your supplier.

The fitting order that prevents pull-out

The collet goes into the nut first. Always. The 30° eccentric ring inside the nut must engage the matching groove on the collet before the collet goes anywhere near the holder. Hold the nut, tilt the collet about 15°, push the small end up into the nut until the ring snaps over the groove, then rock it square — you should feel a positive click.

Putting the collet into the holder first and then threading the nut on top damages the eccentric, and that damage shows up later as the tool pulling out under load.

Snap the collet into the nut until it clicks.
Insert the assembly into the holder.
Insert the tool with the right stick-out — cutting flutes clear of the collet face, and roughly 2× shank diameter of stick-out for general work.
Thread the nut on by hand and make sure the threads engage cleanly.
Torque to the value for your bore band, not the value for the nominal size.

Four ways people ruin a collet

Tightening the nut with no tool installed. The slits close uncontrolled and the collet is permanently deformed. If this has happened to a collet, scrap it.
Clamping the wrong size. A 10 mm shank in a 13–12 collet will not grip and will likely shear under load.
Cleaning in solvent and storing wet. Flash rust forms on the precision surfaces. Wipe dry and store with a light oil film.
Mixing brands on precision work. Geometry is standardised but tolerances stack. Match the collet and nut for best results.

Sealed collets and through-coolant

A sealed collet has a rubber or polyurethane sealing ring at the back of the bore. You need one if you run coolant through the tool: standard ER collets leak past the collet at pressures above about 5 bar, and centre-through coolant systems typically run 10–70 bar — at which point the coolant never reaches the cutting edge.

The trade-off is real though: the seal adds a stack-up term, so sealed collets have slightly higher runout than unsealed ones. Use them only where you actually need through-coolant.

Nuts, spanners and accessories

Torque lookup

Pick the series and the collet bore; the lookup returns the standard-nut value in Nm and ft-lb, bore-adjusted where makers publish one.