A contributor to one of our contemporaries has been deploring, in a recent article, the trend, or rather the stagnation, in the design of racing engines. For twenty years, he says in effect, designers have been concentrating on the chassis to the exclusion of the power-plant. I agree; and, while agreeing, I have been glowing with a sense of quite unjustifiable pride.
Rather over twenty years ago, in April, 1927, to be exact, there appeared in Motor Sport an article signed by “E. K.”, who was in fact none other than “Baladeur” under a different nom de plume, in the course of which the author complained that “the tendency at present . . . is to develop engines at the expense of the rest of the chassis,” and entered the plea, “let us by all means avoid emulating the three-speed American track racer, in which nothing but the engine is studied.” After this interval of time, the complaint, it seems to me, was fully justified; and the plea, to judge by the reactions of our contemporary’s contributor, was answered to an almost embarrassing degree.
Not, of course, that anyone in the racing world took the slightest notice of “E. K.” His proposed solution, in any case, was to remove the French and Italian Grands Prix from Montlhèry and Monza, to concentrate on the Targa Florio and to find an equally difficult course over which to run the big race of the year in France. Such a course, incidentally, although the fact was not mentioned, had already been mapped out, in the shape of the Puy de Dome circuit near Clermont-Ferrand, a really good mountain course in the massif central which was to have been used for the 1914 Coupe de l’Auto, if Europe had not begun committing suicide instead. As the circuit is over 17 miles in length, it would, I suppose, be considered much too long in these latter days, even though the Italians seem to prefer the much longer tour of Sicily to the old Madonie Circuit. In the meantime, racing round the houses, particularly in Monte Carlo, has, I suppose, something of the same effect, which may perhaps be taken as an example of extremes meeting.
However, the reason why designers took “K K.’s” advice and switched their attention from engines to chassis had very little to do with the courses over which races were run, and was really the result of a fundamental change in racing rules. But before going into this question, it seems advisable to go back a bit from 1927 and examine the grounds of the then complaint that it was engines rather than chassis that were monopolising attention. I am, I know, frequently accused of taking any opportunity of going back to Methuselah, but on the present occasion I propose to limit my incursion into the past to the year 1910.
In that year Paul Zucarelli won the Coupe de l’Auto on an Hispano-Suiza with a side-valve, T-head four-cylinder engine, of 65 by 200 mm. bore and stroke, giving a capacity of 2,653 c.c. This engine is said to have developed over 60 h.p., but by 1914, Peugeot, using a slightly smaller engine of 75 by 140 mm. bore and stroke (2,472 c.c.) to fit the new 2 1/2-litre capacity limit, had pushed the output up to 80 h.p. In the meantime, Delage had proved successful in 1911 with a horizontal-valve engine, Sunbeam in 1912 with a side-valve L-head engine; but by 1914 both had come round to the overhead valve design with two overhead camshafts as introduced by M. Henry on the Peugeot of 1912. It is a design which, for racing engines, has remained unchallenged to this day.
By the time the French Grand Prix was revived in 1921, Henry had moved from Peugeot to Ballot, and in the race for 3-litre cars at Le Mans, a 2-litre Ballot actually finished third. The engine of this car closely followed the lines of the 1914 Peugeot, although with a bore and stroke of 69.9 by 130 mm., the capacity was some 20 per cent. smaller. On the other hand, its speed had been advanced from 3,000 r.p.m. to 3,400 r.p.m., and the output in consequence maintained at 80 h.p.
It was, moreover, rapidly coming to be realised that if engine speed was to be progressively increased, the weight of the reciprocating parts must be reduced, and the easiest means of doing so seemed to be by the use of a multiplicity of cylinders. The 1921 race was won by a straight-eight 3-litre Duesenberg, and the Italian Grand Prix that same autumn by a straight-eight 8-litre Ballot. Then, in 1922, came the reduction in the capacity limit to two litres, and at Strasbourg, F.I.A.T., chiefly, it would seem, for reasons of manufacturing convenience, substituted a six-cylinder for the previous year’s eight-cylinder engine. The speed, nevertheless, was pushed up to between 4,000 and 5,000 r.p.m., the output to 92 h.p. In 1923 the successful, and very similar, Sunbeam engine was getting up to nearly 6,000 r.p.m., and developing 108 h.p., or more than the 105 h.p. of the 3-litre Duesenberg of only two years earlier.
But at these extravagant engine speeds it was found that, even with best design of inlet pipe, valves and cylinderhead, it was impossible for the pistons to suck in a full charge of gas. The only answer to that was to blow it in, and in 1923 at Tours, F.I.A.T. appeared with a straight-eight 2-litre, fitted with a supercharger. The new device proved the Italians’ undoing, but while it lasted, the new engine developed 118 h.p., 10 h.p. more than the successful Sunbeam and 16 h.p. more than the F.I.A.T. engine of the previous year. And in 1924 the rather similar Alfa-Romeo engine not only developed 134 h.p., but got away with it; by 1927, the year in which “E. K.’s” article was written, the successful straight-eight Delage was attaining an engine speed of 7,500 r.p.m., and, although its capacity was only 1,500 c.c., was developing 150 h.p. Thus in the twelve active seasons, 1910-14 and 1921-27, the successful racing engine had developed from a side-valve four-cylinder 2 1/2-litre turning at 2,500 r.p.m., or a little more, and giving 60 h.p., or 24 h.p. per litre, into a double-overhead camshaft supercharged eight-cylinder 1 1/2-litre turning three times as fast and giving four times the output per litre. It was enough to take your breath away; no wonder that designers and commentators alike felt that they, too, were in need of a supercharger.
In the meantime, what had happened to the chassis? Well, really only one thing of note. The choice of the hilly Lyons course for the 1914 Grand Prix and of the mountainous Auvergne course for the 1914 Coupe de l’Auto had convinced the more perspicacious designers of the necessity of providing their racing cars with front wheel brakes; and in 1927 there was hardly a touring car left without them. But apart from this, the chassis of the 1927 Delage hardly differed in any essentials from that of the 1910 Hispano-Suiza; and as a result of the enormous increase in the power available, racing cars were becoming exceedingly difficult to control. “I have always insisted,” declared Louis Delage, in an interview with The Autocar at the end of 1925, “that the present type of racing car is too fast for safety; it has become so fast, indeed, that I do not consider there are more than half a dozen men at the present time capable of driving a modern racing car at its maximum speed on the open road. With cars equipped with a 2-litre engine we now obtain speeds of 130 miles an hour or more . . . It does not appear to be sufficiently realised that we have now developed a type of racing car which is not tested to the utmost because the driver is humanly incapable of handling it, on the road at any rate, at its maximum speed.” And he advocated some change in racing rules which would limit the speed of racing cars to, say, 110 miles an hour.
But the authorities paid no more attention to Louis Delage than they did to “E. K.”; in fact, rather less, for while they did return races from the track to rather awkward road circuits, they so arranged the rules that designers were able to build cars that would do a great deal more than 130 m.p.h.; and yet they found a good many more than half a dozen drivers who could handle them. In the process they had to give the most intensive study to chassis design, while, from the point of view of technical progress, the engine department could be left to take care of itself. Of this, two facts stand out in striking illustration. The first is that while the 1927 Delage is still able, after more than 20 years, to give an exceedingly good account of itself in the 1,500-c.c. class, it has been found desirable, at any rate in some quarters, to modify its suspension. The second is that M. Lory, who designed it, has made very little change in the basic design of his latest product, the C.T.A.-Arsenal engine, except that the latter is a V-eight instead of a straight-eight; and I suspect that this change has been made more for the sake of the chassis than the engine.
Having had my way over chassis, therefore, it is time, it seems to me, that we went back and had a look at engines. Admittedly, I should have expected in 1927 that we should have got four-wheel drive by now, but that, perhaps, will have to wait until we switch back from engines to chassis again. As far as the former are concerned, it is illuminating to turn back and see what lines of development seemed to be opening out when interest in engines virtually ceased.
In 1911, the formula for the Coupe de l’ Auto was changed from the limited bore rule to the limited capacity rule. The former had left designers the choice between increasing the capacity per revolution (or rather per two revolutions) by increasing the stroke, or increasing the number of revolutions per minute. The latter restricted their aims to increasing the number of revolutions per minute, or rather the number of power strokes per minute. This immediately led to two schools of thought, of which one plumped for higher engine speeds, among its disciples being Marc Birkigt, the Hispano-Suiza engineer, and Georges and Maurice Sizaire of Sizaire-Naudin. Both of them pretty quickly came up against the difficulty of inadequate cylinder filling at high engine speeds, and both turned to the supercharger as the solution, a solution, incidentally, which had been suggested by Louis Renault as long before as 1902. In the case of the Hispano-Suiza engine designed for the 1912 race, the supercharger consisted of two pumping cylinders, with pistons operated by the crankshaft, and placed in front of the four working cylinders. According to M. Charles Faroux, this 3-litre engine developed 100, h.p., which, if accurate, means that its output was almost equal to the 105 h.p. developed by the successful unsupercharged 3-litre Duesenberg of 1921. There must have been snags, however, for the Hispano-Suizas did not appear in the race, and by 1923, when F.I.A.T. caused a sensation by reintroducing the supercharger in French racing, M. Birkigt expressed himself as most hostile to its use.
The Sizaires, eschewing pumping cylinders, chose a rotary blower, but although the Sizaire-Naudins did start in the 1912 race, it was with ordinary atmospheric induction, the blowers, apparently, having been abandoned because no plugs could be found to stand up to them.
The second school of thought, represented by MM. Côte and Koechlin, set out not just to increase piston speed, but to reduce the unproductive proportion of piston travel by employing a two-stroke cycle. There was, of course, nothing very novel about this — the very first Benz petrol engine was, I believe, a two-stroke, but it was considered too complicated for cars, and in practice it had (and has) been virtually left for use in motor-cycles. The Koechlin, like the Hispano-Suiza, was a non-starter both in 1911 and in 1912; the Côte appeared on both occasions — and proved itself even less successful than the Sizaire-Naudins, which without their blowers were most disappointing.
The curious thing is that these two progressive schools of thought did not combine. The organisers of the Coupe de l’Auto were from the first rather doubtful as to whether the use of superchargers infringed at least the spirit of a limited capacity rule, and in 1913 they made the position quite clear by banning the device — except as an adjunct to two-stroke engines. This apparently so greatly discouraged Hispano-Suiza and Sizaire-Naudin that both of them gave up building racing cars altogether. Koechlin made a rather half-hearted appearance in 1913 with a two-stroke but without a supercharger; and very soon both seemed to have been forgotten.
And yet as far as the two-stroke was concerned, the supercharger, one would have thought, was just what the doctor ordered. The development of the former had by this date already progressed far beyond what the French call the”rustic” stage of the ordinary single-cylinder design, in which a larger proportion of the charge only fails to escape through the exhaust port because so large a proportion of the exhaust gases also fails to escape through it. The engine of the Valveless car of about this period, for example, had two cylinders with a common combustion chamber, in which respect it resembled the Violet two-stroke engine of the 1920’s, and two pistons with their crankshafts geared together. The exhaust port was in one cylinder, the inlet in the other, and the cylinder wall between the two pistons acted as a considerably more effective baffle than the spur on the ordinary two-stroke piston.
But the Valveless still employed crankcase compression, although it compressed only air, to which petrol vapour was subsequently added, and its scavenging was admittedly poor. The supercharger, by providing an external compressor, solved one of the most obstinate problems of two-stroke design, and in the Silent Snow, designed by Major Snow for the first 200 Mile Race at Brooklands, a pumping cylinder, concentric with the working cylinder, and containing an annular piston oscillated by eccentrics, forced the charge into the working cylinder through ring valves. Of even greater significance, and of considerably earlier date, was the N.E.C. two-stroke aero engine of 1910, which employed what was contemporaneously described as a “Roots blower,” and rotary valves. The Roots blower, though not necessarily the rotary valves, presumably reappeared in the stillborn F.I.A.T. Grand Prix engine of 1924, which was reported to be a 2-litre six-cylinder two-stroke, developing 150 h.p. Actually when the racing cars appeared at Lyons, they were found to have modified versions of the 1923 straight-eight four-stroke supercharged engines, developing about 130 h.p., and the two-stroke, as far as I know, never saw the light of day.
It may be conjectured, however, that if the cylinders were arranged in a V, they may have resembled the Valveless design in being in pairs with a common combustion chamber, the V formation obviating the necessity for gearing the crankshafts together. By arranging the pistons in an engine of this type so that one has a lead on the other it is possible to combine some of the niceties of valve timing without the use of valves, and the idea has, if I remember rightly, been put into practice both by Ricardo and Arnold Zoller.
Probably, however, the two-stroke would be a very much better engine if its designers did not yield to the temptation to dispense with valves, just because they are not absolutely necessary. Sleeve valves, or even better, rotary valves, would seem likely to suit it admirably.
The objections to their use in four-stroke engines centre round the difficulties of lubricating them at high engine speeds; but obviously one of the main objects of a two-stroke is to reduce crankshaft speed without reducing power.
At any rate, by 1927, the date when “E. K.” was pleading for less attention to be paid to engines, A. Caputo was writing in Omnia, “the progress of tomorrow will be with the two-stroke engine. Forced induction by a light rotary blower has added to it one of the elements it lacked . . . the two-stroke is racing’s most immediate promise . . ” Unfortunately, perhaps, the world preferred, it would seem, to listen to “E. K.” instead of to the erudite Monsieur Caputo; and to-day, after more than twenty years, mighty little more seems to have been heard of the two-stroke racing car. It is high time, it seems to me, that it was.