Plain oil film bearings have proved their value in motorcycle engines, displacing the rolling-element bearings so widely used as crankshaft main and con-rod bearings before 1970 (with two-strokes continuing with rolling bearings until their mid-’80s emissions-driven fall from grace).

An important fact remains, despite the success of plain bearings. Rolling bearings have much lower starting friction than do plain bearings. That made them essential for the mass market success of the bicycle around 1895. On a ball-bearing-equipped bicycle you can very easily pedal away from rest without having to be overly athletic. Experiments done in the US in 1905 showed that more than half the energy required for the constant stop-and-go operation of streetcars could be saved by use of rolling bearings.

The low starting friction of rolling bearings easily allows the necessary constant slight steer motions.

The low friction of ball or roller bicycle or motorcycle steering heads is essential to sensitive control, as the most frequent objection to certain forkless front ends has been that the friction in their many ball-joints makes steering heavy and insensitive. The low starting friction of rolling bearings easily allows the necessary constant slight steer motions.

Conceptually, rolling bearings date back to people finding it easier to move heavy objects on rollers than to drag them across the ground. Five hundred years ago artist and engineer Leonardo da Vinci drew a rolling bearing with a separator or cage to prevent double-speed rubbing friction between adjacent balls. The industrial revolution (now deplored by environmentalists) eventually provided the material—hard, accurately shaped steel—and the applications that pushed such bearings into existence around 1885.

Friedrich Fischer (1849–1899) of Schweinfurt, Germany, came up with a grinding process for producing accurately round hard steel balls. By 1896 his plant was able to ship 5 million balls a week.

Ball bearings have a much lower starting friction compared to plain oil film bearings.

If you hold an assembled Conrad-type ball bearing in your hands and look at it, it’s far from obvious how it could have been assembled. A number of balls (often seven) are separated by Leonardo’s cage, and are confined by their tracks: grooves in inner and outer races (also made from hardened and ground steel. Because the balls run in grooves, they cannot escape from the bearing. How did they get there? If you remove the ball cage (usually stamped from thin steel ribbon) and crowd all the balls together, it becomes possible to slip the inner race away from the balls, which then drop out. Such a bearing, properly lubricated and protected from abrasive contamination, can carry considerable radial load combined with some axial thrust load.

Rolling bearings for lower-speed applications (wheel bearings, for example) can be grease-lubricated, but for higher speeds oil lubrication has advantages.

Most Conrad ball bearings are made with steel ribbon separators. An experienced bearing engineer once explained to me that at high speed the balls may not just smoothly glide along their inner and outer race grooves but may instead rapidly track from side to side, exerting considerable cyclic loads against their cage. I was able to see the stages of this during my time rebuilding two-stroke crankshafts. The first visible stage was small cracks, often originating in the strained material around the rivet holes, or at the nearly right-angle bends at either end of each press-formed cage pocket. Next, a crack might spread across the width of the separator, but the bearing could continue to function. Riders would sometime complain of “scritching noises” from the engine at low rpm. Finally, a piece of cage would break off that was small enough to get under a ball and bring the ride to a halt.

Other cage designs avoided this kind of failure. One example is the one-piece snap-in injection-molded plastic cages Yamaha adopted for inner main bearings on its TZ250/350 production roadracer twins during 1972–73. More expensive are two-piece riveted phenolic plastic or bronze cages—I’ve heard this type called “grinder bearings” because they are often used at high rpm.

There are also variants of pressed steel “ribbon” separators. Some may be joined over the balls by bending tabs or by resistance-welding.

A pitfall for the home rebuilder was bearing internal clearance. The industry calls its standard internal clearance “C1.” If a C1 bearing is pressed onto a shaft, its inner race is slightly expanded. If the outer race is then either pressed into one half of a vertically split crankcase or clamped between upper and lower parts of a horizontally split case, the outer race is slightly compressed. This can result in loss of all internal clearance or even create a heavy preload that quickly destroys the bearing once in service. The remedy is to specify increased internal clearance for such bearings. Have a close look at the information codes on the faces of the bearing’s outer race and you will see “C3″ after the size and type, indicating this bearing has extra internal clearance.

Eventually,ceramic balls became commercially available…

Bearing precision is graded by ABEC number, so another pitfall is to assume that you’ll get better performance from your vintage engine by specifying ABEC 5 or ABEC 7 when you buy bearings. Alas, higher grades of precision are usually not available with C3 extra internal clearance, so the story can end badly as in the previous paragraph.

I learned a lot about bearings in 1971 when we had a spate of con-rod big-end needle roller failures. I brushed my hair and put on a tweed coat to disguise myself and was able to walk into MIT’s Barker Engineering Library where I spent the day reading rolling bearing literature and taking notes. Among other things I learned that the ball bearings supporting jet engine shafts may receive most of their load from “centrifugal force” caused by high-speed rotation. Remedies were intensively sought because in early days military jet engine life was measured in tens of hours rather than in thousands. One obvious avenue of improvement was to lighten the balls by making them hollow. This was done by friction welding. Because that was expensive and time-consuming, the next step was material lighter than steel, so sapphire balls appeared, with roughly half the density of steel (3.98 versus 7.8). Eventually, ceramic balls became commercially available, allowing a certain kind of motorcyclist to assume his bike’s performance will “take off” if all its 52,100 steel ball bearings are replaced by expensive alternatives with silicon nitride balls. Human self-expression takes many forms.

The service life of rolling bearings was greatly extended in the mid-1960s by use of bearing steel rendered exceptionally free of crack-nucleating inclusions by the practice of remelting it under vacuum (allowing unwanted material such as oxides to evaporate). The letter designations “VIM/VAR” refer to Vacuum Induction Melting and Vacuum Arc Remelting, and such materials are also used in other high-stress applications (such as landing-gear struts for large aircraft) where a long fatigue life is sought.

Although shiny, precise rolling bearings appeal to us, they are actually quite flexible as balls and races deform against each other under load. They also lack shaft damping. In one case I know of, a manufacturer was experiencing excessive and rising engine friction beyond 12,000 rpm. Systematic study traced this to vibrations made possible by flexure of cam support ball bearings, valve operating levers, and camshafts themselves. Stiffening the parts and supporting the cams on self-damping plain bearings corrected the problem.