Thread Engagement and Why Two Threads Are Not Enough

Why this matters

A fastener with too little thread in the hole hits its torque value, feels right, gets written up as done, and strips out later under a load the bolt itself would have shrugged off. The wrench cannot see engagement. Neither can the ticket. And the number people carry in their heads, some version of "a couple of threads is fine," comes from a nut on a through-bolt, where it is roughly true, applied to a tapped hole in soft metal, where it is badly wrong. This is a worksheet for getting the number instead of guessing it.

Before drilling or tapping deeper into any casting, housing or vessel wall: know what is on the other side. A housing can contain oil, refrigerant, water or a live winding, and breaking through is not recoverable in the field. Isolate the equipment and relieve pressure to zero at the component under 29 CFR 1910.147 in general industry for mechanical and stored energy, or 29 CFR 1910.333(b)(2) where a conductor or winding is involved, since 1910.147 excludes exposure to electrical hazards from work on conductors and equipment in electric utilization installations at (a)(1)(ii)(C). Support any part the fastener is holding on an independent stand before it comes out, and stand clear of it rather than beside it. Drilling and tapping throw chips: eye and face protection under 29 CFR 1910.133, and where cutting fluid is used as a mist, ventilate rather than working in the plume.

What engagement is actually balancing

Two failure paths compete in every threaded connection, and the design intent is to make one of them win on purpose.

• The bolt breaks in tension. Its capacity is the tensile stress area times its material strength.
• The threads strip. Either the female thread shears out of the hole or the male thread shears off the bolt. Capacity here is the cylindrical shear area of the engaged threads times the shear strength of whichever material is weaker.

You want the bolt to win the race to failure, meaning the threads should be stronger than the bolt. A broken bolt is a bad afternoon. A stripped tapped hole in a housing is a component replacement, and if the housing is a casting it may be the whole assembly.

Thread shear area grows with engagement length. Bolt tensile area does not. So the required engagement is set by how much weaker the female material is than the bolt, and it is a ratio, not a fixed number of threads.

Why the published multipliers vary the way they do

The field shortcuts everyone has met come straight out of that ratio:

Those multipliers assume a standard-series thread of matched pitch, full-form engaged threads, and a bolt at ordinary structural grades. Raise the bolt grade and the requirement goes up, because you have made the male side stronger without touching the female side, so a Grade 8 or class 10.9 screw into aluminum wants more engagement than a Grade 5 or class 8.8 screw into the same hole. Where the joint is engineered, the drawing or the equipment manufacturer owns the actual figure and these multipliers are a sanity check on it, not a substitute.

Where the "two threads" belief comes from, and where it breaks

It is not pure invention. On a standard nut of matched strength, the load along the engagement is very uneven: the first engaged thread carries the largest share and each successive thread carries less, which is why a standard nut height of roughly 0.8 times the diameter is enough and making the nut taller adds little. Somebody watched that and concluded that only the first few threads do any work.

Here is the part that inverts the conclusion. That steep load taper exists because the threads are stiff and strong enough not to yield. Put the same bolt into a much softer female material and the first threads yield locally, and yielding sheds load down the stack to the threads behind them. The load distributes further along the engagement, which is precisely why the soft material needs the extra length. The mechanism that makes extra threads pointless in a matched steel nut is the same mechanism that makes them load-bearing in aluminum.

So run it at both ends: in a matched-strength pair, adding engagement past about one diameter buys very little. In a soft female material, engagement is doing real work all the way down, and cutting it short removes capacity roughly in proportion.

Three more things eat engagement without anyone noticing:

• The chamfer at the hole entrance costs about one pitch of full-form thread.
• The incomplete threads at the bottom of a blind tapped hole cost roughly two pitches with a common taper or plug tap, more with a taper tap.
• The runout at the base of the bolt's own thread is not engagement either, and neither is a shank that is larger than the thread.

The worksheet, filled in

A fan assembly mounts into a tapped aluminum housing. Screws are M8 x 1.25, class 8.8. The tapped hole in the casting is 16 mm deep. Everything below is in millimetres.

Line 1, target engagement. Aluminum female, ordinary-grade steel screw: 2 x nominal diameter = 2 x 8 = 16 mm.

Line 2, hole depth available. 16 mm as drilled and tapped.

Line 3, deductions.

• Entry chamfer, about one pitch: 1.25 mm
• Incomplete threads at the bottom of the blind hole, about two pitches: 2.50 mm

Line 4, full-form engagement actually available. 16 - 1.25 - 2.50 = 12.25 mm.

Line 5, compare. 12.25 against a 16 mm target is 12.25 divided by 16, about 77 percent of what the material wants. In diameters, 12.25 divided by 8 is 1.53 x D, which is the cast-iron figure being applied to aluminum.

That is the whole finding, and it was invisible before the deductions were written down. A tech measuring the hole would have called it 16 mm deep, which is exactly 2 x D, and signed it off.

Line 6, put the myth in the same units. Two threads at a 1.25 mm pitch is 2.50 mm of engagement. Against the 16 mm target that is about 16 percent. Not marginal. Not most of the way there. One sixth.

Line 7, the options, and which term each one moves.

• Longer screw and a deeper tap. Moves line 2. It also requires knowing what is behind the casting wall, and on a housing containing oil or a winding that is often a hard no. This is where the instruction itself creates the hazard, so it is not a default.
• Through-bolt with a nut and washer on the far side. Removes the female aluminum thread from the problem entirely, since the nut is steel. Requires access to the back, which a housing usually does not give.
• A threaded insert. Converts the female thread to steel, which drops the target on line 1 from 2 x D to about 1 x D, or 8 mm. Against the 12.25 mm of available depth, that leaves 4.25 mm of margin instead of a 3.75 mm shortfall.

The choice, and the reason. The insert wins because it moves the term that actually set the requirement. Longer and deeper fights the symptom by adding length; the insert changes the strength ratio the length was compensating for, and it does it inside the depth that already exists. Note that installing an insert enlarges the hole, so the casting has to have the wall thickness for it, and that is the check to make before committing.

The failure mode if nobody does the arithmetic. The screws are torqued to the value for M8 class 8.8, which is a value derived for a matched-strength female thread. The aluminum threads at 77 percent of required engagement start yielding during tightening. The wrench still clicks, because a yielding thread still resists rotation. Nothing looks wrong. Then the first vibration cycle or the first thermal cycle finds the already-damaged threads, one screw pulls out, the remaining screws pick up its share and go the same way, and the housing needs the insert anyway, now with a damaged hole and a fan assembly that has moved.

Reading a connection that has already failed

• Aluminum swarf or thread-shaped shavings around the hole means the female thread stripped. Engagement, material ratio, or torque value applied to the wrong material.
• A bolt with clean threads that pulled out with no damage to itself is the same finding from the other side.
• A bolt whose own threads are stripped while the hole is intact means the female side was the stronger one, which usually means the bolt was the wrong grade or was already damaged.
• A bolt that snapped with the threads intact on both sides is the design intent working. The joint was overloaded or the bolt was fatigued, and the sibling article on how threads fail covers how to read that surface.

Checking you got this right

• Measure the hole depth and then subtract the chamfer and the incomplete threads before comparing anything. Depth is not engagement.
• Name the female material out loud before choosing a multiplier. A tapped hole in a steel bracket and a tapped hole in an aluminum housing take the same screw and need different depths.
• Check the bolt grade against the multiplier. Raising the grade in an existing tapped hole raises the engagement requirement, which is a common way a well-intentioned upgrade strips a housing.
• Confirm the screw is not bottoming. A screw that reaches the bottom of a blind hole stops clamping and starts jacking, and it will hit a torque value while the joint is loose.
• On any repair that installs an insert, record it on the equipment history. The next tech who meets a steel thread in an aluminum casting needs to know it was engineered and not original.

References

• 29 CFR 1910.147, control of hazardous energy, and 29 CFR 1910.333(b)(2) for the electrical case that 1910.147 excludes at (a)(1)(ii)(C)
• 29 CFR 1910.133, eye and face protection, for drilling, tapping and thread repair
• Equipment and fastener manufacturer documentation for specified engagement depth, screw length, and any qualified insert
• Published fastener engineering references for thread shear area and the relationship between engagement length and female material strength
• See related: How a Thread Fails and What the Failure Tells You; What Grade and Class Markings on a Fastener Mean; Cross-Threading Prevention