Mon–Sat 08:30–18:00 / +971 6 767 2469 / sales@mfindllc.com
MF FASTENER INDUSTRIES LLC
Emirates Modern Industrial Area, Umm Al Quwain, U.A.E. · P.O. Box 7071 · ISO 9001:2015 · Technical Library
TECHNICAL LIBRARY

Why bolts fail

Six mechanisms account for nearly every fastener failure investigated. Each one looks different, has a different cause, and is prevented at a different stage — specification, purchase, installation or service.

Hydrogen embrittlement — the delayed snap

Looks like: a high-strength bolt breaks days or weeks after tightening, with no overload, often under the head or at the first thread. Brittle, flat fracture with little deformation.

Cause: hydrogen absorbed during acid pickling or electroplating, trapped in a hard microstructure (above about 320 HV, so 10.9, 12.9, A490M), driven to crack tips by the tensile stress of preload.

Prevention: bake plated high-strength bolts within hours of plating (ISO 4042 Annex); do not hot-dip galvanize A490M or 12.9; use zinc-flake or mechanical galvanizing where a coated high-strength bolt is unavoidable; ask the supplier for the baking record.

Fatigue — the crack that grew

Looks like: a smooth, curved "beach-marked" fracture face with a small final rough zone, at the first engaged thread or the head radius. The bolt was cycling for a long time before it let go.

Cause: insufficient preload. A properly preloaded bolt sees only a small fraction of the external load fluctuation; a loose one sees all of it. Also sharp thread roots, poor head radius, bending from unsquare seating.

Prevention: achieve and verify preload (turn-of-nut, DTI, tensioning); rolled threads rather than cut on fatigue-loaded bolts; hardened washers and machined seats so the bolt is not bent.

Thread stripping — the silent one

Looks like: the nut turns and turns; the bolt is intact; the threads on one or both parts are sheared. Often found only when the joint is dismantled or when it slips.

Cause: nut class below bolt class; insufficient thread engagement (nut too thin, or a bolt in a tapped hole engaging less than about 1 × d in steel, 1.5–2 × d in aluminium); oversize-tapped galvanized nuts with a lower class than required.

Prevention: match nut class to bolt class per ISO 898-2; full nut height; check engagement length in tapped holes; specify class 10 nuts on galvanized 8.8 assemblies where the standard calls for it.

Overload and over-torque — the yielded bolt

Looks like: necked, stretched threads; the nut will not turn back; or a cup-and-cone tensile fracture with visible deformation.

Cause: a torque figure applied to the wrong friction condition (oiled bolt, dry table); impact wrenches; a lower grade than the torque table assumed.

Prevention: torque tables tied to the actual finish and lubrication; calibrated tools; never torque to more than 85–90% of proof; if a bolt has been taken past yield, replace it — it does not recover.

Galvanic corrosion — the wrong neighbour

Looks like: the fastener corrodes rapidly while the surrounding plate is fine, or a stainless bolt stains and pits the aluminium around it.

Cause: two metals far apart in the galvanic series in contact with an electrolyte (salt spray, condensation). The anodic metal corrodes; if it is the small part — the bolt — it goes fast.

Prevention: keep the fastener at least as noble as the plate (stainless bolts in carbon steel are fine; carbon steel bolts in stainless are not); isolate aluminium from stainless with washers and sleeves; in C5 coastal exposure match coatings across the joint.

Stress corrosion cracking and chloride attack — the coastal one

Looks like: branching cracks in a stainless bolt under tension in a warm, chloride environment; rust-coloured pitting on "stainless" that was 304 in a marine location.

Cause: austenitic stainless under sustained tensile stress above about 50–60 °C in chlorides; 304/A2 has little resistance to it; 316/A4 better; duplex better again.

Prevention: A4 (316) minimum on the coast; duplex or super-duplex for hot chloride service and splash zones; keep stainless clean of carbon-steel contamination; do not assume "stainless" means "marine".