IT is curious that just when we have devoted some space to modern racing fuels (see “Rumblings”) a fantastic claim in respect of high power output on ordinary petrol should come in from America. The most advanced American racing car is the four-wheel-drive 3-litre Miller, one of which crashed in practice for Indianapolis, the other being withdrawn from the race. In a post-race article in “Automotive Industries” some details of the Indianapolis cars are given by R. T. Jackson. It is claimed that the six-cylinder 2,956 c.c. engine of the Miller runs on commercial “No-Nox” ethyl petrol and that, boosted at about 20 lbs. per square inch by a double-entry centrifugal blower running at 32,000 r.p.m., the output is nearly 300 b.h.p. at 7,000 r.p.in. This is less than the output of the 3-litre German Grand Prix cars and represents a b.h.p. per litre about equal to that of the Mercedes-Benz and Auto-Union cats built to the previous formula. We have no quarrel with the power claim, but we cannot believe that this was attained on a commercially obtainable leaded fuel having an octane number of 80.
The last time ordinary petrol was used in a first line racing car was in 1923, when the unsupercharged, 2-litre six-cylinder Sunbeam with which Segrave won the French G.P., ran on No. 1 petrol. This Sunbeam had a compression ratio of probably 6 to 1, certainly not more than 7 to 1, and gave about 54 b.h.p. per litre. Even so, it finished the race at Tours with its valves so badly burned that it is doubtful if it could have run much further. When Sunbeam built the 2-litre supercharged cars for 1924, which developed about 67 b.h.p. per litre, they used a 50 per cent. ethanol, 50 per cent. petrol-benzole fuel, and they used the same fuel for the 4-litre V12 Sunbeam with which Segrave held the Land Speed Record for a time, and which gave around 75 b.h.p. per litre. The increasing output from small engines has led to an ever increasing alcohol content in racing fuels. Certainly the Germans were using specialised fuel in 1938, when their Grand Prix engines were developing about 90 b.h.p. per litre in 5½-6 litre form. At Indianapolis, the winning Maserati, blown at not more than 15 lbs. per square inch, ran on an alcohol base fuel, as did the Sampson, with the Frank Lockhart supercharged V16 3-litre engine, and the 4½-litre unblown Thorn. Certainly other cars at Indianapolis used leaded petrol fuels, notably the 3-litre eight-cylinder centrifugally-blown Bowes, which finished second, and the Elgin, which finished third. The Bowes developed about 116 b.h.p. per litre, and it ran on Shell 100 octane aviation petrol— which is not commercially obtainable fuel. The Elgin used similar fuel for its 4½-litre unblown Offenhauser engine, which gave about 68 b.h.p. per litre. Our Royal Air Force uses petrol as a fuel, but of an octane value of 100 with a b.h.p. per litre from its big engines of only 35. The old 1¾-litre racing Alfa-Romeo, with a compression-ratio of about 6 to 1, blown at about 3-5 lbs. per square inch, used to overheat seriously unless 30 per cent. benzole was mixed with No. 1 petrol to give an octane number of approximately 72. Compare such an engine with the 90 b.h.p. per litre Miller, boosted at 20 lbs., and ask yourself whether 80 octane fuel seems adequate. Indianapolis is a track race wherein engine speeds only vary by 1,000-1,500 r.p.m. per lap, and under such conditions the cooling effect of alcohol is extremely desirable. Sunbeam found the limit of petrol as a fuel seventeen years ago, and even granting Harry Miller modern head design, with bronze valve seats in an aluminium alloy head, extreme finning of sump and crankcase walls, and oil cooling, we just cannot believe that he is able to run satisfactorily on petrol at an output per litre nearly twice as great, even granted an octane improvement of about 20 by reason of ethyl lead. If he does do so, either he knows more about cylinder head design than anyone in Britain, France, Italy or Germany, or the possibilities of leaded fuel have never been properly appreciated in those countries. The correct answer seems important.