WHY SHORT FASTENERS COME LOOSE MORE EASILY
When Harley-davidson brought out the new design known as the “Evolution” engine, a major goal was to eliminate oil leaks and gasket failures that had made the previous model the butt of light remarks. Earlier engines employed cylinders bolted to the crankcase by short bolts through a base flange. Heads were then bolted to the tops of the cylinders.
A major part of the “fix” was to replace the short base and head bolts with long through-studs that held down both cylinders and heads.
Why would this change keep engines leak-free? Think of bolts and studs as springs. When you tighten a bolt or the nut on a stud, you are stretching that bolt or stud, and the resulting tension keeps parts in place and gaskets tight. Because there are limits to how much you can stretch metal before it yields (the usual mistake here is “I just tightened it until it started to get loose…”), you can put more stretch into a long fastener than you can into a short one. A long fastener can better tolerate the heat expansion of the parts they are clamping, whereas a short fastener can more easily be permanently stretched so that it loses part of its clamp load.
Other effects contributing to this difference are settling of gaskets (which is why re-torqueing of head or cylinder fasteners after initial engine operation is sometimes called for) and loss of metal from the slight relative movements between fastener and part caused by vibration. A good example of this is the loosening of the large, thin nuts often used to retain engine output sprockets on their shaft splines or of the several bolts used to retain rear wheel sprockets.
Despite the usual presence of a torsional shock absorber, built into the clutch basket, the drive from engine to rear wheel is not smooth. All the tiny motions that result cause surface scrubbing, each such movement making a zillion micro-welds that are broken by the next movement. In steel-to-steel-contacts of this kind you may find a reddish discoloration or even red powder-the iron oxide that results from “frettage,” the slight vibratory weld-and-break action between surfaces.
Thin sprocket nuts are so notorious for coming loose that many manufacturers provide tab washers to prevent a gradually loosening nut from coming off entirely. Others avoid this by sliding the sprocket onto its splined shaft and retaining it with a circlip— no nut to loosen.
I learned long ago that when the rear sprocket on the bike after last week’s event is the same one needed today, I must still go through the process of unbending the tab washers or safety wire retaining the sprocket bolts then re-torqueing and securing with tab washers or wire. Why? Because I’d found that a weekend of racing mileage had loosened all the bolts. A look at where the sprocket pressed against the wheel showed evidence of back-and-forth motion that had gradually resulted in loss of bolt tension.
At the Canadian GP in 1967 I saw the late Arturo Magni make a kind of “rite” out of changing the rear sprocket on Agostini’s MV 500 Triple. There were no bolts! Instead, the sprocket fitted onto several short drive pins in the wheel hub and was held in place by a single large circlip (which was subsequently wired in place).
Then there was the case of Kawasaki’s F5 off-road 350 single, which was for a time legal in AMA’S 250 GP class. Driving itself forward as it did by a series of large combustion thumps, there was trouble keeping the primary pinion on the crankshaft. This was the result of a couple of factors. Primary was the thumping, which inevitably produced relative motion between pinion and shaft, and secondary was the shortness of the
fastener—resulting in too little “stretch” to maintain tightness. A first-try tab washer just sheared off; the vibratory “unscrewing torque” was too much for the thin metal. The next try: a thicker washer.
The response of veteran racer/ builder (and 1969 250 GP World Champion) Kel Carruthers was, “It could unscrew once for any of us, and we might not get the fix right the first time, but by God I’d weld the thing to the shaft before I’d let it come loose again!”
Perhaps that’s why Nagato Sato, who designed Kawasaki’s next 250 class entry—the tandem-cylinder Kr250—gave it pressed-on crank phasing gears. No vulnerable keys, splines, or tab washers giving up at 12,000 rpm for him.
Polaris/indian may have had to deal with something similar in securing the external flywheel (made in three weights to tailor rear-tire hookup) of its FTR750 race-only dirt-track engine. The slightest backlash in a splined connection encourages frettage and loss of fastener torque.
Then why did the last generation of two-stroke motocross engines attach their cylinders and heads not with advisable long elastic through-studs (as found on all 1970s TZ roadrace engines) but with short base bolts, with the head then bolted to the cylinder? In this case, designers had previously found those long stud tunnels right in the way of where they needed to route next year’s transfer ports or sub-exhausts. In some cases tuners had resorted to first modifying the ports then pressing thin steel sleeves into the stud tunnels to seal the places where they’d cut into them. Short fasteners—despite their propensity for coming loose—could more easily avoid interference with ports.
In many other cases when short fasteners positively must not come
AT THE CANADIAN GP IN 1967 I SAW THE LATE ARTURO MAGNI MAKE A KIND OF “RITE” OUT OF CHANGING THE REAR SPROCKET ON AGOSTINI’S MV 500 TRIPLE.
loose, locking devices such as Palnuts, safety wire (where there’s room to use it!), or red Loctite must be used. Each of the 18 46-pound aircooled cylinders on a WWII P-47’s R-2800 radial piston engine was held to the forged aluminum crankcase by 20 short studs and nuts, each secured by a Palnut (a type of locknut formed from spring steel). No wonder aerospace has pretty much abandoned safety wire in favor of self-locking fasteners; imagine reaching 20 inches into that forest of cylinders to wire all 360 nuts.
The first time I tried to secure a slotted cam sprocket to its cam I foolishly applied the otherwise usual drop of red Loctite. It came loose because 1) the sprocket bolts are very short and 2) cam drive is a very “lumpy load.” Point taken: I should have drowned those bolts in red. No shortcuts! Build to be sure.
Learn and use best practice, and update it with your own hard-won experience.