boltwright

Fastener engineering reference — calculators, geometry, and standards.

Preload and tightening torque

Rev C, 2026-09-26

T Tightening torque
62.6N·m
Fi Preload
31.3kN
Fp Proof load
48.1kN

Within proof load

As and Sp, ISO 898-1:2013 · K = 0.20, assumed · 65 % of proof load · Not hot-dip galvanized or structural · T = K·Fi·d

Show the formula

Fp = As·Sp; Fi = (target / 100)·Fp; T = K·Fi·d

As
Tensile stress area, as ISO 898-1:2013 defines it, computed from the ISO 68-1:2023 basic profile, mm²
Sp
Nominal stress under proof load for the property class, ISO 898-1:2013, at the thread's diameter, MPa
Fp
Proof load, N
target
Target preload as a percentage of proof load, %
Fi
Preload, N
K
Nut factor. Assumed, not derived; common shop practice, not standardized
d
Nominal diameter, mm
T
Tightening torque. With d in mm, K·Fi·d is in N·mm and is divided by 1000, N·m

How it computes

The proof load comes first: the tensile stress area of the thread times the nominal stress under proof load for the class, both from ISO 898-1:2013.

Fp = As · Sp

The target preload is a fraction of it, and the torque follows from the short-form torque equation:

Fi = (target / 100) · Fp
T = K · Fi · d

The verdict compares Fi with Fp. Proof load is the load a fastener must withstand without permanent set, so a preload equal to it is within, and only a preload above it exceeds.

Below the result, the page also shows the proof load ISO 898-1:2013 prints for the same thread and class, in Table 5 (coarse) or Table 7 (fine), with the difference. The two are not the same number: the calculator uses the exact As, while the table rounds As to three significant figures and rounds the load again. The difference is small but not zero, and the verdict uses the calculated value.

What it assumes

  • K is assumed. The nut factor lumps thread friction, underhead friction and geometry into one number. It is not a property of the bolt, and it is the largest uncertainty in the result. The presets beside the field are common shop practice, not standardized: 0.20 for dry, unplated steel (the default), 0.15 for oiled, unplated steel, and 0.11 for steel lubricated with molybdenum disulphide paste. Where the torque matters, measure K on the actual parts, finish and lubricant.
  • T = K·F·d is common shop practice, not standardized. It does not separate thread and underhead friction, and it takes no account of the joint.
  • Nominal values. As is the defined tensile stress area, and Sp is the nominal proof stress. Neither is a measured property of a particular bolt.
  • Not hot-dip galvanized, not structural bolting. ISO 898-1:2013 Table 5 qualifies two groups of its proof loads, and the calculator uses the general value for both. Hot-dip galvanized M8 and M10 bolts are cut to thread tolerance 6az to make room for the zinc, and reduced values from ISO 10684:2004, Annex A apply instead (footnote c): here the calculator overstates the proof load. Structural class 8.8 bolts from M12 take the higher-strength band, so their proof loads are higher, 50 700 N for M12 rather than 48 900 N (footnote d): here it understates it.
  • Tension only. The verdict compares the preload that remains after tightening with proof load. It does not include the torsional stress the bolt carries while it is being tightened, which adds to the tensile stress at that moment.
  • ISO metric threads, steel property classes. The thread and class tables cover M3 to M24, and the classes ISO 898-1:2013 specifies for carbon and alloy steel. Stainless steel (ISO 3506) is out of scope.

Worked example

M10 × 1.5, class 10.9, K = 0.20, target 65 % of proof load:

  • As = 57.99 mm², from the ISO 68-1:2023 profile
  • Sp = 830 MPa, class 10.9, ISO 898-1:2013
  • Fp = 57.99 × 830 = 48 131 N
  • Fi = 0.65 × 48 131 = 31 285 N, shown as 31.3 kN
  • T = 0.20 × 31 285 × 10 = 62 571 N·mm, shown as 62.6 N·m

The same case is pinned as a test, with each step worked by hand, in src/lib/calc/preload.test.ts.

Sources

RevDateByDescription
C2026-09-26SDAdded K presets for dry, oiled and MoS₂-lubricated steel, common shop practice, not standardized
2 earlier revisions
B2026-09-26SDAdded M7, the one ISO 262:2023 size from M3 to M24 that was missing
A2026-09-25SDInitial issue