Kent Karslake discusses the piston speeds and bore dimensions as judging engine efficiency
To the readers of MOTOR SPORT at least, all things it seems, come to him who waits. After spending the 1983 season reading accounts in other periodicals of Grand Prix racing which paid far more attention to squabbles between drivers than to trials of technical strength between makers, I have been favoured in the January issue with D.S.J.’s table showing particulars of Formula One engines.
I had been eagerly awaiting this information, first, because I was anxious to catch up with an old hobby-horse known as piston speeds. In the early years of the century it was widely held by the experts that this was the factor which ultimately limited the output of any engine of a given static capacity. Instead of limiting such capacity, therefore, it seemed good to them to limit the bore only of cylinders. Designers seeking more power could then choose between increasing the stroke or increasing crankshaft speeds. In either case they would, sooner or later, be checked by the problem of intolerable piston speed. In the meantime, they would have demonstrated which was the more profitable road to follow. The 3 1/2 hp 2-cylinder Daimler engine fitted to the Panhard et Levassor cars in 1894 had dimensions of 75 x 140 mm. I have never seen any suggestion as to why Daimler plumped for what would subsequently look like a decidedly long-stroke design. He can’t, I suppose, have had much more than guesswork to go on. In any case, Levassor evidently decided that he had gone too far, and in his Phenix engine of 1895 he both increased the bore and reduced the stroke, so that the dimensions were 80 x 120 mm. Moreover, having taken this first step, he evidently decided that he was on the right road, so that the 6 hp 2-cylinder engine of 1897 had dimensions of 90 x 130 mm, and thereafter the stroke-bore ratio of Panhard et Levassor engines was progressively reduced, until the “square” engine, 170 x 170 mm, was reached in 1904 and the “over-square” engine, 185 x 170 mm, in 1906. I doubt whether, in this process, the desire to limit Piston speeds was the dominant factor. As racing cars became more and more powerful, the overriding problem, particularly after the imposition of a weight limit of 1,000 kilos in 1902, was how to mount larger and larger engines without incurring an undue weight penalty. As the the capacity of a cylinder increases directly with stroke, but as the square of the bore, the latter dimensions must have seemed to be the one to enlarge if you wanted more engine for a given weight.
The people who were interested in piston speeds were the organisers of the Coupe de L’Auto races for voiturettes; and in 1906 they decreed that the bore of single-cylinder engines eligible for the race should be limited to 120 mm, with no limit set to the stroke used. It was an indication, perhaps, of the vogue for “square” engines at the time that the Sizaire-Naudin, which proved the most successful voiturette in the competition, used a stroke which, at 120 mm, was only equal to the bore. The designers, in fact, seem to have concentrated on increasing crankshaft speed, which attained 2,000 rpm, which was more than twice as great as that used at least in most larger engines and which at the time was regarded as dizzy. As a result, the piston reached a velocity of 480 metres per minute. At this point a word of apology for eccentricity would not perhaps be out of place. It has been customary in this country, and maybe still is, to measure piston speed in terms of feet per minute. This was fine in the days of steam engines, when strokes were also measured in feet. But, in spite of a gallant attempt by some British manufacturers in the early days of motor cars to measure engine dimensions in inches, a cause still upheld by Americans long after it was lost, the French habit of using millimetres for the purpose soon commanded international acceptance. Measuring piston speed in feet per minute when the stroke was measured in millimetres thus involved a tiresome and unneccesary conversion. How much happier, one might have thought, were Continental commentators who need have no truck with these unseemly feet. But almost infinite, it seems, is the capacity of humanity for making rods for its own back. Although crankshaft speeds have always, and universally as far as I know, been measured in revolutions per minute, these Continental commentators insisted on measuring piston speed in metres per second thus saddling themselves with a quite unnecessary division by 60. If by measuring it in metres per minute I am setting a new fashion, I am not in the least bit ashamed of it. From now on all those who adopt this method have to do is to double the stroke, multiply by the revs, and hope he has got the damned dot in the right place. For 1907 the bore limit for single-cylinder engines in the Coupe de L’Auto race was reduced to 100 mm. Sizaire et Naudin duly complied, but, by way of double compensation, they not only put the stroke up to 150 mm but also advanced the crankshaft speed to 2,400 rpm, so that piston speed was now 720 metres per minute. When it came to 1908, they seemed to think that they had reached the limit as far as the second variable was concerned and left crankshaft speed unchanged; but they had the audacity to put the stroke up to 250 mm, producing a piston speed of 1,200 metres per minute. Nor was this quite the end of the story, for in 1910 the 2-cylinder Lion-Peugeot, while only using 2,200 rpm, had dimensions of 80 x 280 mm, so that piston speed, at 1,232 metres per minute, was marginally higher than in the 1908 Sizaire-Naudin. It had set a record which was destined to last for a very long time. One reason for this was that by now the organisers of the Coupe de L’Auto had come to the conclusion that their theory about piston speed being the limiting factor governing speed was mistaken. Bearing loadings seemed to be so much more critical that they abandoned the bore limit for 1911 in favour of a capacity limit of 3 litres, and for some inscrutable reason, a maximum stroke/bore ratio of 2 to 1. I have always been rather sorry about this, as it would have been interesting to see just how much further piston speeds could have been advanced if designers had continued with the incentive to do so. Perhaps in practice not very much, not because they were up against any physical limit in the matter, but because these long-stroke engines were becoming so intolerably top-heavy that it was difficult to know how to mount them satisfactorily in a chassis.
As it was, Peugeot, as if to show that piston speeds were no problem to them, plumped for as long a stroke as possible, and the dimensions of their 3-litre 4-cylinder engine were 78 x 156 mm. Most of their competitors fell only just short of this extreme and favoured dimensions of 80 x 149 mm. Why this was so is not entirely clear except that perhaps, while crankshaft speeds were still so strictly limited, the wide power band produced by a long stroke engine was found to favour the best overall performance, particularly on a varied road circuit. A small bore, it is true, inevitably added to aspiration problems, but in 1912 Peugeot took a mammoth step towards their solution when it adopted the Henry design of 4 overhead valves per cylinder operated by 2 overhead camshafts. More than 70 years later this design, modified only in detail, was still used by the Cosworth engineers when providing a special high-performance head for Mercedes-Benz. From 1911 onwards various designers had claimed crankshaft speeds of up to 3,000 rpm and even if some of these claims must be discounted, that in favour of the 1913 3-litre Peugeot, which was perhaps the apotheosis of the Henry design, may probably be admitted. Even so, with a stroke of only 156 mm, piston speed at 936 metres per minute was well down on the record set by the 2-cylinder engine of 1910.
After the First World War, two important factors militated against any early attack on it: the first was a progressive reduction in the capacity limit for Grand Prix races during the 1920s from 3 litres to 2 litres and then to 1500 cc, and the second was the vogue for multi-cylinder engines, particularly the straight-eight. By 1921 the crankshaft speed of the 3-litre straight-eight Fiat engine was up to 4,400 rpm, twice that of the 1910 Lion-Peugeot, but with dimensions of 65 x 112 mm, piston speed at 985 metres per minute was still substantially lower. By 1924, with the aid of supercharging, the crankshaft speed of the 2-litre straight-eight Alfa-Romeo was up to 5,500 rpm, but with dimensions of 61 x 85 mm, piston speed at 935 metres per minute was almost exactly the same as in the 1913 Peugeot. According to Pomeroy, the 1 ½-litre straight-eight Delage engine of 1927 could attain 8,000 rpm, but he admits that it would not do it for very long and suggests that 6,500 rpm was more like its useable maximum. At peak revs, this small 8-cylinder engine with dimensions of 55.8 x 76 mm could boast a piston speed of 1,216 metres per minute, which nudged that of the 1910 Lion-Peugeot without actually exceeding it.
During the 1930s, the use of larger engines was usually combined with lower crankshaft speeds, at least until the end of the period when the 1939 3-litre Mercedes-Benz engine attained 7,800 rpm. By now, however, it had 12 cylinders and the days of the long-stroke engine had at last given way to almost “square” dimensions, of 67 x 70 mm, so that piston speed at 1,092 metres per minute, was still well short of the 1910 record which had, I think to wait until quite recent years to be broken. In 1983, however, the “long-stroke” DFV dimensions of 85.6 x 64.8 mm was doing 11,000 rpm, and its pistons speed was 1,425.6 metres per minute. At the same time the 4-cylinder 1500 cc turbocharged BMW engine, with dimensions of 89.2 x 60 mm, using the same crankshaft speed, was not far behind it with a piston speed of 1,320 metres per minute. At last, then, we have broken new ground as far as this factor is concerned; but few commentators seems to be as keenly interested in the matter as were Charles Faroux and his colleagues when they drew up the rules for the Coupe de L’Auto race in 1906.
Perhaps some recollection of their preoccupations was lodged in the minds of the designers of the Renault turbocharged Formula One engine. The Lion-Peugeot of 1910 used a short bore ratio of 3.5 to 1, which may not have constituted a second but which was at least sufficiently remarkable after only a few years since moderately “over-square” engines had been in vogue for Grand Prix racing. By contrast, the Renault engineers chose dimensions of 86 x 42.8 mm, giving a stroke/bore ratio of just under 0.5 to 1. I wonder whether it represents a record in the other direction. As a result, the piston speed of the Renault engine, even at 12,000 rpm, is a very moderate 1,027.2 metre per minute well within the territory explored in the past. If this consideration was what inspired the Renault design, their caution in the matter seems to have been misplaced, as several of their competitors use substantially high piston speeds with immunity.
In my ignorance I had for long assumed that the object of turbocharging was to attain higher crankshaft speeds and to make up for the smaller static capacity of a 1500 cc engine compared with a 3-litre. The fact that this assumption is completely illusory stares at one from D.S.J.s table when it is seen that the 12-cylinder 3-litre Alpha Romeo engine does 12,300 rpm and the same maker’s 8-cylinder 1,500 cc engine (with, incidentally, a shorter stroke than that of the 12-cylinder) only 11,500 rpm. Nevertheless the 1500 cc engine develops 640 bhp and the 3-litre only 540 bhp. At the same crankshaft speed, the 3-litre disposes of twice the capacity of the 1,500 cc, and the lower output can only be due to the fact that if the smaller engine’s cylinders are completely filled, then those of the larger one must be worse than half empty. In the absence of a second Ernest Henry to revolutionise engine breathing, it really does look as though the development of the atmospheric pressure induction engine is at the end of its tether.
It has been fascinating in recent years to watch the conflict between it and the turbocharged engine of half the size. Now, unfortunately, the battle seems to be over. The only Formula One contestants with 3-litre engines in future will be those without access to turbocharged power. This suggests, I am afraid, that the present Formula One has become meaningless. The authorities, I believe, are stuck with it for some time to come, but when they can change it they might consider pitting the 1500 cc turbocharged engine against engines of unlimited capacity but without forced induction. After all, if you can’t get near filling a “I-Atm induction” 3-litre engine you would have even less luck, I suppose, with one twice the size. If in practice nobody would use the bigger engines, however big they were permitted to be, my suggestion I am afraid is not a very good one. All I can do, therefore, is to make another, even more eccentric one. If we are not committed to turbocharged engines only, why not scrap the 1,500 cc capacity limit in favour of a bore limit, say 90 mm for 4-cylinder (compared with the present BMW engine of 89.2 mm.) 80 mm for 6 cylinders (as in the Honda), and 74 mm for 8-cylinders (as the Alfa-Romeo)? With no limit on the stroke, we might then see where piston speeds to get to! – Kent Karslake.