The Artificial Divisions of Knowledge
When Chemistry Met Engineering: The Worm-Holes
by Kevin Cameron
Tuesday, August 25, 2026
When in the mid-1960s I began trying to help my racer friends I had never seen detonation damage to pistons and cylinder heads. That deficiency was soon corrected. What could be eating tiny worm-holes into the solid metal of heads and piston crowns, out near the cylinder wall?
This was also a chronic problem much earlier, as the already hot-running air-cooled radial aircraft engines of the late 1920s were being boosted even further by use of supercharging.
Let’s not overlook the articles, written during WW II by Norton’s race engineer Joe Craig. He described, step-by-step, the death of an engine suffering chronic detonation (that is the engineers’ term for engine knock or ping). Earlier, an Aircraft engine designer used nearly the same words to describe piston failure from detonation in big round engines.
Here are the steps:
1. Outer edges of the piston crown begin to look as if they have been lightly sand-blasted.
2. In various locations this damage goes deeper into the metal, especially just above the top piston ring.
3. Eventually the top ring is exposed by this metal erosion, and its ability to seal is destroyed.
4. Now the erosion can reach the second ring, and when that ring, too, loses its seal, hot combustion gas begins to flow down the face of the piston, then inward through the oil-scraper ring’s drain-back holes. This rapidly heats the metal exposed to it, causing it to soften and fail.
5. The erosion becomes general, with short arcs of piston ring now no longer supported by anything.
The behavior of fuels as they are heated before ignition, then are ignited and burned, possibly with late-in-the-cycle auto-ignition of bits of still-unburned charge, which having been chemically altered by their history of heat exposure, behave like a sensitive explosive, burning at supersonic speed to create the shock waves we hear as knock, and the mystifying piston and head damage described above.
This is chemistry.
But the temperatures of engine parts, which are responsible for steadily heating the unburned parts of the charge and altering them chemically such that they may detonate, are mechanical engineering.
The result, in the 1930s, was a pointless blame-fest in which the fuel chemists blamed the engine designs for the hostile environments that so heated the mixture. And vice-versa.
In effect, this was the classic unwinnable argument of two groups, each shouting at the other in different mutually incomprehensible languages.
Artificial intelligence (AI) theorists are right now imagining how the new form of analysis will have equal access to all the forms of information to which it is connected—potentially eliminating such organizational barriers to understanding.
At this point, S.D. “Sam” Heron, an Englishman who specialized in the problems of air-cooled engines, decided he must learn some chemistry so that he could talk to both the chemists and the mechanical engineers.
Meanwhile, a mass of basic research was under way. Dr. Graham Edgar was compiling a list of petroleum hydrocarbons and their knock behavior, using the newly-available variable compression CFR knock rating test engine (CFR = Cooperative Fuel Research). In the early 1920s Charles Kettering of Delco had set Thomas Midgley the task of searching for possible fuel additives with strong anti-knock action.
There are people who cast Midgley as a demon for discovering the highly poisonous tetraethyl lead (TEL) anti-knock compound and the ozone-layer-destroying Freon series of refrigerants. If there is blame, it must be directed at the needs these compounds answered – the need to operate aircraft engines at higher power but without detonation, and the desirability of finding refrigerants which were neither toxic (the widely-used ammonia) nor inflammable.
Dr. Edgar devised a useful scale for rating the knock resistance of fuels. It remains in use to this day.
As the “zero” on his scale he chose the normal heptane, a straight-chain hydrocarbon whose fragility made it very knock-prone. As the “100” on his scale, he chose 2, 2, 4 trimethyl pentane, a naturally durable, rolled-up-in-a ball branched hydrocarbon whose familiar name is ‘octane’ (“octo” being Latin for the 8 carbons in its structure). Thinking about octane, my mind blinked to Gary Fisher, who upon coming off his bike in a pack of other machines, rolled up into a ball so he would be less likely to be hit by the speeding bikes.
They run the CFR engine on the fuel to be knock-rated, varying the compression ratio as it runs to find the highest usable compression ratio. They then try various mixtures of the two reference fuels, octane and n-heptane, until they find the one that has the same knock threshold as the fuel under test. Its rating is then given as the percentage of octane in that mixture.
Meanwhile, S.D. Heron had learned a useful amount of chemistry and found he could talk to both the chemists and the engine people.
Using this kind of combined understanding, fuels and engines were developed together to result in extraordinary wartime aircraft engine performance. After the war, the vast alkylation plants that had so boosted av-gas performance numbers were kept at work making higher-octane commercial gasolines for high-compression postwar automobile engines.
When I was puzzled by worm-holes in piston and combustion chamber edges, all the above wealth of information was in place. I had a grand time learning all this stuff that was new to me. Lots more understanding had been added by engine designers who learned to accelerate combustion by creating turbulence in the mixture about to be ignited. The faster the charge could be burned, the greater the chance that combustion would become complete before the heat-driven pre-flame chemical reactions in it could lead to knock. Honda’s testing before 1965 showed that above 12,000 rpm, an engine’s octane requirement began to fall. At something like 27,000 rpm, 37 octane fuel would burn knock-free in Honda’s test engine. This is why F1 teams are not bothered by the low octane number limits specified in their tech regulations.
Where can I order a Dr. Graham Edgar T-shirt?
— ends —
