Bolt Tightening Torque and Preload Reference — Metric 8.8, 10.9, 12.9 and Stainless A2/A4 (VDI 2230 basis, µ = 0.12 and 0.14)

Assembly preload and tightening torque for metric coarse-thread bolts M4 to M36 in property classes 8.8, 10.9 and 12.9 at 90 % utilisation of the minimum yield stress, for total friction coefficients µ = 0.12 and µ = 0.14 (VDI 2230 basis), a stainless A2-70 / A4-70 table at reduced utilisation, and the rules that make the numbers valid: friction dominates, lubrication and finish change torque by 30–50 %, hot-dip galvanised joints tighten by turn-of-nut, calibrated tools (ISO 6789), re-use limits for 10.9 and 12.9, and how prevailing-torque nuts add to the reading.

Bolt Tightening Torque and Preload Reference — Metric 8.8, 10.9, 12.9 and Stainless A2/A4 (VDI 2230 basis, µ = 0.12 and 0.14)

Catalogue data and standard-part designations in this guide follow the cBallast C-Parts range, supplied through Bufab (Värnamo, Sweden). Values are nominal reference data; confirm against the controlling standard revision before procurement.

Torque is only a proxy for what the joint actually needs: clamping force (preload). Of the torque applied to a bolt, roughly 50 % is spent against friction under the head or nut face, 40 % against friction in the thread, and only about 10 % produces preload. That is why the same M12 × 8.8 bolt can require 60 Nm with a lubricated zinc-flake finish and 100 Nm dry and plain to reach the same clamp force, and why the friction coefficient must be stated with every torque value. The tables below follow the VDI 2230 method: preload F_M at 90 % of the minimum yield stress (Rp0.2 min from ISO 898-1) and the torque M_A that produces it for a total friction coefficient µ_G = µ_K = 0.12 and 0.14.

Stainless A2-70 / A4-70 (≤ M20) — typical values

Austenitic stainless galls; it is tightened to a lower utilisation (about 70 % of Rp0.2 min = 450 MPa) and always with an anti-seize or the supplier's lubricant. Using a K-factor of 0.20 (µ ≈ 0.14–0.16 lubricated): M5 ≈ 4.5 kN / 4.5 Nm; M6 ≈ 6.3 kN / 7.6 Nm; M8 ≈ 11.5 kN / 18 Nm; M10 ≈ 18.3 kN / 37 Nm; M12 ≈ 26.6 kN / 64 Nm; M14 ≈ 36 kN / 100 Nm; M16 ≈ 49.5 kN / 158 Nm; M20 ≈ 77 kN / 310 Nm. For A2-80 / A4-80 (Rp0.2 ≥ 600 MPa) multiply by about 1.3. Above M20 the material is normally class 50 (Rp0.2 ≥ 210 MPa) unless class 70 is confirmed — never take a carbon-steel 8.8 table for stainless.

What the numbers assume

• Friction — µ 0.12 corresponds to lightly oiled zinc-plated or zinc-flake-with-lubricant bolts on machined steel; µ 0.14 to plain or black-oxide bolts with light oil, or zinc plated dry. Dry unlubricated plain steel can reach 0.20 (torque +40 %); MoS₂ paste can drop to 0.08 (torque −30 %, risk of over-stretching if a 0.14 table is used).
• Utilisation 90 % — the standard VDI 2230 assembly target, leaving 10 % of yield for the working load and scatter. Some industries (pressure equipment, cranes) specify 70–75 %.
• Coarse thread, ISO 4014/4017 head, ISO 4032 nut, plain washer — flange bolts, larger head bearing diameters and fine pitch shift the torque by 5–15 %.
• Torque wrench accuracy — ISO 6789 class ±4 % (type II, class A) or better; calibrate at least annually or every 5 000 cycles. Impact wrenches are not torque tools.

Tightening methods compared

• Torque control — preload scatter ±25–30 % because friction scatter passes straight through. Adequate for most general joints.
• Torque + angle (turn-of-nut) — snug torque then a defined rotation (e.g. 90° or 120° per structural bolting rules EN 1090-2 / EN 14399); scatter ±10–15 %; used for hot-dip galvanised and HV structural bolts where friction is unpredictable.
• Yield-controlled — electronic tool detects the yield point; scatter ±5–10 %; engine and gearbox lines.
• Direct tension — hydraulic tensioners or load-indicating washers; scatter ±5 %; flange studs, large-diameter and critical joints.

Special cases

• Prevailing-torque nuts (nylon insert, all-metal) — the locking torque (typically 1–10 % of the table value, larger for small sizes) is spent before any preload is generated; add the measured prevailing torque to the table value.
• Hot-dip galvanised — friction 0.16–0.25 and scattered; use turn-of-nut, or a lubricated (waxed) oversize nut with a verified µ.
• Re-use — 8.8 bolts tightened to 90 % may be re-used a few times if undamaged; 10.9 and 12.9 bolts tightened by angle or to yield are single-use; stainless should be re-lubricated every time.
• Soft or gasketed joints — retighten after the first thermal cycle or after 24 h; embedding relaxes 5–15 % of the preload.
• Threadlockers — anaerobic adhesives (medium strength) change friction little when wet but add breakaway torque; do not compensate with a lower tightening torque unless the adhesive manufacturer specifies it.

Quick check of any torque value

The K-factor method gives a sanity check: M_A ≈ K × F_M × d, with K ≈ 0.16–0.20 for lubricated to dry steel. For an M16 × 8.8 bolt at 84.7 kN: 0.16 × 84.7 × 16 ≈ 217 Nm, which brackets the table value of 195–215 Nm. If a supplied torque value is far outside this bracket, the friction assumption or the property class behind it is different — find out which before tightening.

Browse the C-Parts range · Bufab · Related: fastener standards cross-reference · torque wrench calibration (ISO 6789) · ASTM A193 B7 stud bolts.

Compiled by cBallast for identification and cross-reference. Part numbers, designations and trademarks belong to their respective manufacturers. Confirm ratings and dimensions against the manufacturer's current documentation and the nameplate of the installed equipment before ordering or installation.