More About Threaded Fasteners

[Admin]

Rolling and shot peening is the answer to increasing fatigue life of a fastener. (Dietmar Rabich / Wikimedia Commons / “Haltern am See, Sythen, Werkzeughalle der Quarzwerke -- 2015 -- 4985-2” / CC BY-SA 4.0/)

If you ever want to know what motivated the intensive metal-alloys research of the 1920s and ’30s, just have a look at the kind of failures racers of that period suffered through. Major parts broke often: cranks, con-rods, pistons. Spokes snapped, frames cracked, axles twisted off. Ordinary carbon steel was all that was available in early days, though rumor had it there was something special about “Swedish steel.” There was: Swedish iron ore just happened to contain traces of the powerful grain-refining agent vanadium.

One reader commented that by annealing his bolet and studs, he had good luck with stopping breakage. Annealing heats a metal up to its Curie point, destroying most internal structure, and then cools it slowly. The resulting material is soft and malleable. I once made a rear engine bolt from mild steel, but as I tightened it, the bolt just kept getting longer!

Early cylinders were held in place by four studs, then six, then eight, yet despite the increasing count, the studs kept right on breaking.

Carbon steel offers a painful compromise to seekers of strength. Yes, you can make the stuff quite strong, but only at the cost of brittleness. The normal process was to first quench the red-hot part in water or oil, then “draw” part of the resulting very hard temper by reheating the metal to some lower temperature. The result, properly done, would retain fairly high strength without excessive brittleness. I suspect our reader did not fully anneal his bolts and studs, but just drew some of their temper by heating to a lower temperature. Either that, or his bolts are still growing.

As industry researched the many possible alloys and mapped out how best to heat-treat materials to attain desired properties, a major problem began to arise with aircraft engines. How to keep the base-bolted cylinders of radial engines firmly attached? Early cylinders were held in place by four studs, then six, then eight, yet despite the increasing count, the studs kept right on breaking.

RELATED: Threaded Fasteners on Your Motorcycle

Like motorcycle engines, aircraft engines must be built light enough to lift them out of the boat-anchor category. That being so, every time a cylinder fired, the combination of combustion pressure on the piston and head along with the fast-increasing angle of its connecting-rod as the crank rotated produced a powerful combination of direct lift (trying to uproot the whole cylinder) and side force. At each firing cycle the cylinder’s flange was elastically deformed downward and the matching mouth of the crankcase was coned upward. Thus, those base studs or bolts were being stretched, bent, and sheared, all at once. Like any abused worker, they didn’t like it. So they quit, breaking in large numbers.

The sharp-eyed can still see something similar at the vintage races. Look at one of the many lovingly built (overbuilt?) Norton Commando engines after it explodes. The constant flexure of the cylinder base flange starts a crack, and when failure comes, the whole cylinder launches skyward like a space-mission liftoff, leaving some of the base flange still attached to the crankcase and taking the rest along for the ride. But the ride is short, for the gas tank is directly above. Bam—the rocker boxes stamp their initials on the underside of the tank, a perfect replica of their own shape. In the case of the B-29 bomber’s 18-cylinder R-3350 radials, crewmen would occasionally be treated to the sight of a gaping hole in one engine nacelle, through which an entire cylinder had made an unscheduled departure. The cause? Not enemy antiaircraft fire, but a detached valve head, getting pushed against the head by the piston.

Yet it wasn’t the base flanges of those long-ago aircraft cylinders that failed as the cast Norton cylinders did. No, those flanges were upset-forged as integral parts of the cylinder. Instead, it was the base fasteners that failed constantly.

When a problem has been successfully solved, our human minds no longer find it interesting.

Threads cut on a lathe or with a die bear tool marks that act as stress raisers. OK, so we’re willing to use a more expensive threading process if it stops this infernal breakage! Now they tried grinding the threads, a process which can, if pushed, produce really smooth surface finishes. There, that ought to eliminate those tool marks that were acting as crack nucleation sites.

But no; although ground-thread fasteners were some improvement, the failures continued. Finally they decided to try rolling the threads. In one process, the fastener is fed in among three very hard rollers with the thread profile ground into them. In another, the thread form is applied to two plates that compress and roll the blank fastener between each other, much as we might roll a pencil between our palms. The metal of the blank fastener was (usually) cold-squeezed into the thread form. Although originally associated with cheap mass-market fasteners (thread rolling is quick), it had the advantage of putting the deformed metal into compression, especially at the vulnerable root of the thread, where most cracks originated. Because cracking must start from tensile (pulling) stress, the retained compression in rolled threads greatly increased the fatigue life of the resulting fastener; to put the material into tension required first straining it enough to overcome the retained compression. Shot peening does the same with the surfaces of high-duty connecting rods, putting them into compression.

When a problem has been successfully solved, our human minds no longer find it interesting. Yet if all the details of the solution are not carried out correctly every single time, the problem returns as nasty as ever. We may forget, but physics always remembers. And so it is that my TD1-B’s rear engine-mount bolt broke, and H1R and TZ750A cylinder studs broke. All the knowledge of how to make reliable fasteners was easily available when those parts were made (1965, 1970, 1974). We just needed to remember it.

View the full article