Industry News

Why do bolts break below their rated strength? Fatigue life and stress concentration

2026-11-03

Compiled by JIALI Industrial (164.hk) — a Hong Kong fastener supplier specialising in screws, bolts, nuts and selection know-how.

Eight out of ten bolt failures are not tension failures - they are fatigue

The bolt failures found in the field are rarely pulled apart by a single static overload. Fatigue accounts for roughly 80 percent of mechanical component failures, and fasteners are a textbook case: the working stress sits far below the 800 MPa tensile strength of grade 8.8, yet the bolt snaps at the thread root after a few hundred thousand cycles. There are three recurring crack origins - the root of the first engaged thread (that first thread carries about 30 percent of the load), the head-to-shank fillet, and the shank-to-thread transition. What they share is a geometric discontinuity that concentrates stress: the thread-root stress concentration factor Kt runs about 3-5, and an under-head fillet radius below 0.1d can push it past 5. The notch effect drags the smooth-specimen fatigue limit (roughly 0.3-0.4 times tensile strength, about plus or minus 240-320 MPa for grade 8.8) down to plus or minus 40-60 MPa - just 5 to 8 percent of static strength. This explains a common puzzle: a grade 10.9 bolt often breaks sooner than a grade 4.8 one, because the higher grade raises tensile capacity while barely improving notch fatigue limit, and high-strength steel is more sensitive to defects. Buying on grade alone, without checking notch geometry and surface quality, is how fatigue failures get designed in.

Three routes to longer fatigue life: more preload, less concentration, better threads

Route one is the most effective: apply full preload. The VDI 2230 bolt load factor Phi shows that once a joint is preloaded, only a fraction of the external working load (typically Phi 0.2-0.3) reaches the bolt. An M10 grade 8.8 bolt (stress area As 58 mm2) preloaded to 0.7 of yield carries about 26 kN; an external cyclic load of plus or minus 10 kN therefore puts only about plus or minus 2.5 kN into the bolt, a stress amplitude of roughly 43 MPa. If preload is insufficient or the joint opens, the full fluctuation lands on the bolt, the amplitude multiplies several times, and life drops one to two orders of magnitude along the S-N curve - the quantified version of the saying that bolts break because they came loose first. Route two is reducing concentration: an under-head fillet of at least 0.2d, proper thread run-out to GB/T 2 / DIN 76-1, and for heavily loaded joints DIN 2510 studs with an undercut or MJ threads, which spread load back from the first thread. Route three is thread and surface process: rolled or rolled-formed threads keep the grain flow continuous and leave compressive residual stress at the root, giving 30-50 percent better fatigue strength than cut threads, with thread-root rolling or shot peening for critical parts. Conversely, laps at the thread root, handling dents and post-plating hydrogen embrittlement all become crack initiators.

Assembly and acceptance: turning fatigue life into checkable actions

Three assembly actions decide fatigue life. First, the tightening method: torque control is at the mercy of friction scatter and delivers preload errors of plus or minus 25-30 percent (verified per GB/T 16823.3 / ISO 16047), while critical joints should use torque-to-yield or hydraulic tensioning to hold errors within 10-15 percent. Second, avoid eccentricity and bending: non-parallel bearing faces, thin washers or oversized clearance holes all introduce bending moments that can raise the measured stress amplitude by 20-40 percent and halve fatigue life - do not widen the GB/T 5277 medium-fit hole (10.5 mm at M10) without cause. Third, tighten in a diagonal, multi-pass sequence so one side does not bottom out first and warp the flange into local overload. Put four items into the purchase contract: thread forming method (rolled or rolled-formed, not cut threads for highly loaded parts); surface defects judged to GB/T 5779.1 with no laps or cracks at the root; head soundness by the wedge loading test of GB/T 3098.1; and axial fatigue testing to GB/T 13682 / ISO 3800 for critical joints, supported by EN 10204 3.1 certificates tied to heat numbers. For grade 8.8-12.9 bolts, undercut studs or rolled-thread fasteners selected against a fatigue duty cycle, contact Chiali Industrial (164.hk) with the load spectrum and joint dimensions for a matching proposal and samples.

Related Reading

• Why do fasteners need quench and temper? Heat treatment, grade mapping and acceptance essentials
• How to measure coating thickness? Fastener plating measurement methods and acceptance essentials
• JIALI launches Stainless Steel zone with full 304 / 316 range

← Back to News