The bracket fell off. Not the whole bracket — the bolts held. The threads stripped. M10 bolts into a 20 mm thick aluminum plate, holding a 300 kg load with a 150 mm lever arm. The load was 450 N at the thread. The threads pulled out of the aluminum after eight months of vibration. The calculation on the drawing said “M10 x 25 deep.” It was wrong for the material.
Thread shear is the failure mode nobody checks until it happens. A bolt in tension rarely fails at the shank — it fails where the threads meet the tapped material. The nut or the tapped hole gives way first. The reason is simple: the thread engagement length determines how much shear area carries the load. Too short, and the threads strip. The math is straightforward, and skipping it is how brackets end up on the floor.
The shear area calculation
The thread shear area depends on the minor diameter of the bolt and the engagement length. For a steel bolt into a softer material (aluminum, cast iron, plastics), the tapped material shears first. The formula used for ISO metric threads:
A_s = π × d × L × 0.8
Where d is the nominal bolt diameter (10 mm for M10), L is the engaged thread length, and 0.8 is the empirical factor accounting for the thread profile (the actual shear plane sits at the pitch line, not the full diameter). The 0.8 factor is standard in VDI 2230 and most machine design texts.
For M10 into aluminum with L = 20 mm: A_s = π × 10 × 20 × 0.8 = 502 mm².
The shear strength of the tapped material matters more than the bolt grade. Aluminum 6061-T6 has a shear strength around 190 MPa. The load capacity: 502 mm² × 190 MPa = 95 kN. That’s plenty for 450 N. So why did it strip?
Where the calculation went wrong
Three things. First, the 450 N was the static load. The bracket was on a vibrating pump. The dynamic load was 5x the static. Now the thread sees 2.25 kN. Still far below 95 kN. Second — and this is the real issue — the aluminum was not 6061-T6. It was a cast aluminum alloy, probably ADC12 or similar. Cast aluminum has a shear strength around 90-110 MPa, roughly half of 6061-T6. That drops capacity to ~50 kN. Still enough. Third, the tap drill hole. The installer tapped the hole by hand with a standard tap, which produced a loose fit. With a worn tap, the threads are shallow and weak. A poorly tapped hole can cut the effective shear area by 40%. Now the capacity is 30 kN. Still enough for 2.25 kN dynamic.
So the numbers say it should hold. It didn’t. The remaining factor: fatigue. Vibration at 50 Hz for 8 months — roughly 10 million cycles. Aluminum threads under repeated load don’t fail by shear. They fail by fretting and fatigue at the root of the first engaged thread. The first thread carries about 40% of the load. It fatigues, cracks, and the crack propagates thread by thread. The threads “strip” over time, but it’s fatigue, not overload. This is why the static calculation never predicts it.
What actually fixed it
Three options, we used two. First, the aluminum plate was replaced with a steel one. Steel tapped holes have 3-4x the fatigue strength of aluminum for the same thread. The bracket went into steel. Second, a threaded insert (Heli-coil or similar) was installed in the aluminum where it couldn’t be replaced. The insert provides a steel thread surface inside the aluminum. The fatigue resistance improves dramatically. The insert costs about 30 cents and takes 5 minutes to install. The third option — a through-bolt with a nut on the back — wasn’t possible (blind hole), but it’s the best solution when geometry allows.
The rule we now apply: for aluminum tapped holes under dynamic load, use inserts. For static loads, the VDI formula with 0.5x safety on cast aluminum. For steel, thread engagement of 1x the bolt diameter is enough for full strength (the bolt shank will break before the threads strip).
Minimum engagement by material
| Tapped material | Engagement (multiples of bolt diameter) | Notes |
|---|---|---|
| Steel (Grade 8.8 bolt) | 1.0x | Bolt shank fails first |
| Stainless steel (annealed) | 1.5x | Lower strength, galling risk |
| Cast iron | 1.5x | Brittle, watch the first thread |
| Aluminum 6061-T6 | 2.0x | Use inserts for dynamic loads |
| Cast aluminum | 2.5x | Fails by fatigue first |
| Plastics (nylon) | 3.0x+ | Creep dominates, use inserts |
The practical checklist
When a threaded joint is spec’d on a drawing, check these before it goes out: (1) what material is being tapped — if it’s aluminum or plastic, the engagement length on the drawing is probably too short; (2) is the load dynamic — vibrating equipment needs inserts or steel; (3) is the tapped hole drilled to the correct tap drill size — a worn tap or oversize hole halves the strength; (4) is there a lock feature — thread-locking compound or a serrated flange nut for vibration.
The stripped bracket wasn’t a bolt failure. It was a design assumption — “M10 into aluminum, 25 deep” — made without checking the material’s shear strength and the load’s dynamic nature. The formula is on page one of any machine design textbook. The fatigue behavior is what the textbooks gloss over.
Thread shear is a calculation, not a guess. 1x diameter engagement for steel, 2x for aluminum, inserts for anything dynamic. The bracket fell because a cast-aluminum thread was fatigued to death, not because the bolt was undersized. Check the tapped material before you trust the drawing.
