My first impulse in writing this piece was to say “new ideas” in piston ring design, but the ideas here are not new. They date back to Honda’s third era in F1, which ended in 2008, which is now 13 years ago. Honda engineers faced the difficulty of sealing pistons at engine speeds near 20,000 rpm, and they found existing piston ring concepts inadequate.
Related: Hot Air, Cold Air, Delta-T
Since English inventor John Ramsbottom first conceived of them 169 years ago for use on railroad steam engines, conventional piston rings have evolved significantly. Because it is not feasible to fit pistons into cylinders so tightly that they achieve a reliable seal (except for certain tiny model engines), designers fit a springy circular metal split ring of roughly rectangular section into a groove cut near the piston’s head. Since they are normally larger than the cylinder’s bore, the split ring is compressed to fit into it and then expands to seal against it; the ring’s flexibility is able to accommodate the cylinder’s expansion as it heats to operating temperature.
Mind the Gap
It seems intuitively correct to fret over possible leakage through the ring’s gap at the split. Many and various are the inventions proposed to eliminate this gap, such as stacking two vertically thin split rings in a single groove, each with its end gap pegged to keep the two at 180 degrees to each other.
It turns out that ring-end-gap leakage is quite small, provided that the gap itself measures close to the usual rule of thumb: about 0.004 inch for every inch of piston diameter. As a result, stacked rings and other methods of stopping ring-gap leakage have passed out of use.
One early application in which the Ramsbottom ring emphatically did not work was the rotary-radial piston engines used in large numbers of World War I aircraft. Because early aircraft flew too slowly to push adequate cooling air through their engines’ fins, cooling was improved by bolting the crankshaft to the aircraft’s firewall and having the crankcase and cylinders rotate around it (hence “rotary”). A typical rotary engine was about 37 inches in diameter, so when it operated at 1,200 rpm, the heads whirled through the air at 165 feet per second (112 mph)—even with the airplane sitting still.
The big problems were 1) the cylinders’ leading faces were well cooled, but their trailing sides remained very hot, and 2) the cylinders were made from steel, which conducts heat poorly. As a result of this nonuniform cooling, the trailing side of each cylinder bulged from heat expansion while the cooler leading side remained smaller. No simple piston ring could conform to such a heat-distorted cylinder. The best solution at that time was the “obturator,” a brass split ring with an L-shaped cross-section. The base of the L was quite narrow while its vertical part was relatively tall. A groove machined into the piston retained the ring’s base in the traditional fashion, but the resemblance ended there. It was the bending flexibility of the L’s vertical portion which allowed combustion pressure to easily push it into good contact with the cylinder wall—rather like the leather washer in an old bicycle pump. Mainly because of short obturator life, the time-before-overhaul (TBO) of such engines was between 10–25 hours.
We have all read about how hot-rodders achieve truly round cylinders by honing them with thick plates bolted to the cylinder deck under standard head-bolt tension. In some cases, they even circulate hot water through the coolant passages to simulate in-service thermal distortion.
In WWII-era air-cooled radial aircraft engines (in which it was the crankshaft that rotated, not the crankcase), the cylinders were “choked,” or tapered to be smaller toward the top, in the expectation that once warmed up the hotter upper regions would expand enough to eliminate the taper.
Testing Ring Seal
What if all that isn’t enough? In 1981 rider Rich Schlachter and I somehow came into a hoard of piston rings for the Yamaha TZ250H we were running. After deburring the ends, I placed them into a new cylinder, one ring at a time, and squared them up with a piston. Then I held the result up to the sun. Some of the rings had no visible gap to the cylinder, but past others I could see tiny slivers of light.
Related: Cooler Heads Prevail
Now I carefully examined the rings that showed light, imagining they might have little manufacturing burrs that were holding the rings away from the cylinder wall. I fussed and fiddled, but finally had to accept that some of our rings were not going to seal, at least not in that cylinder.
The pressure of combustion gas, flowing down the top land and entering the clearance above the ring (ring axial clearance), can get behind the ring and push it out against the cylinder wall. That might have made my “bad” rings seal up tight, but I had no way of testing the possibility. So we ran the “good” rings in our engine and used the others in the traditional way: as “paddock money,” the same way the incarcerated have always used cigarettes.
Unlike Schlachter and myself, engine development departments do in fact have means of testing ring seal in running engines. They put a flowmeter on the crankcase breather and directly measure blowby volume. After that, it’s just arithmetic to work out how much horsepower is being lost to leakage. When the leakage is serious, you have to do something about it. It’s not only a matter of losing power; hot combustion gas leaking past the ring and flowing between piston and cylinder is going to have unpleasant consequences.
Honda’s Novel Ring-Seal Solution
In developing their V-10 F1 engines to peak at close to 20,000 rpm, Honda engineers decided that conventional rings might be too stiff to conform well to cylinder-wall profiles, so they reduced ring stiffness by cutting radial width. To further increase ring local flexibility, they milled a number of shallow round-bottomed vertical grooves in the inner vertical face of each ring. These changes reduced the ring’s ability to press itself against the cylinder wall, so a compact wave-spring expander, fit behind the ring, supplemented the ring’s preload. Blowby volume was reduced.
While there can be value in continuing to refine existing and well-understood concepts, at some point new thinking is required.