A pioneer in the scientific testing of automobiles
Cecil Clutton writes-up a remarkable book, which is of great interest to students of Edwardian performance, and which, even to-day, shows car-testing in a new light.
IT is maddeningly difficult to obtain any definite figures about the performance of early automobiles. Of course, if you don't happen to feel the need of this sort of information the difficulty ceases to be quite so maddening, and you can really do without this article as well.
Such information as I possess (and it is now fairly considerable) was entirely the result of tabulating times and speeds recorded in Edwardian classes at speed trials and races, until I came across a remarkable work entitled "The Scientific Determination of the Merits of Automobiles," by Dr. A. Riedler, of the Royal Technical University at Charlottenburg. The English translation seems to have been published in 1912 and the German original only a trifle earlier.
The doctor starts off by complaining that previous standards and comparisons of performance were entirely subjective; and as his tidy Teutonic brain found this very annoying and unsatisfactory, he set about rectifying it with characteristically Teutonic thoroughness. Apart from sundry excursions into the realms of army lorries and electric broughams the important part of the book records the testing of a 1905 Renault, Prince Henry competition models of Benz and Adler, and a chapter devoted to sleeve-valve motors. He seems to have made about 200 different tests on each machine, and the book is illustrated by 227 figures of which a few are photographs of the cars tested and over 200 are graphs of their performance. There is every sort of graph—graphs to show the motor power at given revs. and speed; losses from different causes, excess power available for gradients and acceleration, and a host of other things beside. His apparatus was a series of revolving drums, brakes, dynamos and clutches. In fact, Dr. Riedler laid the foundations of scientific automobile testing and the work was evidently of a high order, as his calculations were invariably borne out within a minute proportion on actual test.
For the doctor was sufficient of a motorist to realise that bench tests are not everything, and that the intangible "feel'' of a car counts for much as well. He accordingly took each machine tested for a comprehensive road trial as well.
However, the present-day interest of the book lies not so much in the methods of testing as in the tabulated results obtained, and the intention of this article is to record some of the more important examples.
The 1905 Renault was taken as typical of the conservative touring car of the day, and the doctor remarks that it had altered little by 1910, although these are now regarded as amongst the most important years in the evolution of automobile design. It is rather interesting that the engine dimensions of this car are identical with the 4½-litre Bentley, namely 100 x 140. Running up to a maximum of 1,450 r.p.m. the engine developed 33 b.h.p. with an unobstructed exhaust, which is equivalent to 6.1 b.h.p. per litre, or all m.e.p. of about 50. The three forward gears had ratios of 10.8, 5.83 and 3.14 to I. and the top speed of the car was 43.75 m.p.h. Top gear would tackle a gradient of 1 in 18, and it may therefore be concluded that the acceleration from 10 to 30 in top would take 18 secs. The best gradient which bottom would surmount was 1 in 5.
In the case of each car there is given what is described as a "Characteristic Speed" which is "to be regarded as that speed at which the greatest car efficiency is obtained, fuel consumption and running times being taken into account." This certainly sounds rather subjective in itself, but I suppose it really amounts to what we call the cruising speed of a car, and in the case of the Renault it is put down at 37½ m.p.h.
The car weighed 31 cwt., of which 15 cwt. rested on the back wheels and 10 on the front.
It is characteristic of the doctor's thoroughness that he calculated the ignition advance and compression ratio separately for each cylinder—and they were all different ! The average ignition advance was 10.5° and the mean compression ratio was 4 to 1, although the four compression spaces differed so widely as to contain 360, 380, 346 and 37 cubic cms. respectively! It is even recorded that the back wheels were not truly round.
Next comes a most exciting car in the shape of a "100 h.p." 1910 Prince Henry Competition Benz car. This fine machine had a four-cylinder 7,272 c.c. engine (115 x 175), weighed 26½ cwt (12½ front, 14 back), and attained a speed of 83.75 m.p.h. The body was a four-seater torpedo type with fairly ample wings and no windscreen. The back axle ratio was the rather low one of 2.48 to 1 and the compression ratio the surprisingly high one of 4.7 to 1. The engine is described as "high speed type" and ranged from 500 to 2,400 r.p.m., peaking at 2,050 r.p.m., when it produced 118 h.p. unsilenced or 104 with the silencer attached. This is equivalent to the commendable figure of 14.2 per litre, or 7 per 1,000 r.p.m. per litre, which represents an m.e.p. of 90, and is still the staple figure for modern production cars. Incidentally, it is upon this basis that the R.A.C. formula works, and if engine exceeded 1,000 r.p.m. (as few did when the formula was devised), it would be remarkably accurate.
The car cannot have been very tractable to drive as the stalling speed on top was 31.3 m.p.h. (800 r.p.m.) It is, however, characteristic of engines of this type that the top gear acceleration curve reached its optimum at 37½ m.p.h.— only 6 m.p.h. above the stalling speed—when it attained the creditable advance of 3 feet per second per second. The maximum gradient to be tackled in top was also the quite steep hill of 1 in 10½.
The engine had the cylinders cast in pairs, with four valves per cylinder (described as "two admission and two exhaust "), but the doctor seemed to regard this as a somewhat unnecessary complication. They are slightly inclined in the head and the inlets are larger than the exhaust and have the characteristic Benz triple seating. The lift on the inlet valves was 8 mm. and on the exhaust 9 mm., the operation being by push-rods with separate camshafts on each side of the engine. The rockers gave a fantastic step-up, as was another habit of Benz, and the lift of the cams themselves cannot have been more than 4 mm.! Why this terrific strain was so unnecessarily imposed on the long pushrods is a mystery.
The dual ignition permitted a maximum advance of 23.5º. The copious drawings this particular item give the impression of being slightly bogus, as they do not always agree, two entirely different back axles (live, by the by, and not chain, as would be anticipated with so large a car at this date) being simultaneously attributed to it.
The Benz may have been magnificent, but the doctor's fancy was clearly taken by his next subject, the 1910 Prince Henry Adler. This he found was much less of a real racer than the Benz, and practically indestructible. Furthermore, despite the rather unfavourable arrangement of inlet over exhaust the Adler attained the same m.e.p. as the Benz. In this connection, however, it does not seem to have struck the doctor that it is not so easy to obtain good m.e.p. in large cylinders as in small, the Adler having the substantially smaller capacity of 5,192 c.c. and cylinders measuring 101 x 150.
The weight was 21 cwt., almost equally divided between front and rear; the compression ratio was 4.76 to 1 and top gear 2.895 to 1. It is surprising that there were only three forward ratios. The engine peaked at 2,100 r.p.m. when it gave off 76 b.h.p., equal to 14.7 per litre. The most rapid acceleration was 3.2 ft.-sec.² at 25 m.p.h. and the maximum speed was 71 m.p.h. with quite a touring sort of body, but no windscreen.
We now pass to what is grandiloquently described as the "Metabolism and Pathology of Sleeve Valve Motors."
One of the troubles seems to have been that no one would lend a car to test, but eventually Riedler wheedled an engine out of Mercedes, and a complete 1911 40 h.p. car out of the English Daimler Company who, as he points out, originally pioneered and popularised the Knight engine—not so much on account of any inherent merit, as by very clever and comprehensive publicity. Other manufacturers were even forced into the sleeve-valve market against their better judgment, to meet the popular demand.
The English car lent had a four-cylinder 101.1 x126 engine and the whole outfit weighed 28 cwt. (13 front, 15 rear). The frame was of very light construction, which undoubtedly assisted performance, and the top speed was 53 m.p.h. Top gear would surmount a gradient of 1 in 12½ and bottom 1 in 4. The engine had the creditable m.e.p. of 90 at low and medium revs., although it is observed that the earlier 1909 engines (96 x130) had attained 7.6 b.h.p. per 1,000 r.p.m. per litre, or 99 m.e.p. The reduction in efficiency was therefore presumably due to cooling problems in the earlier engines, which were certainly not lacking in smoothness or flexibility. The Mercedes engine, however, was only fractionally less efficient than the 1909 Daimler, and better than either of the English engines at higher revs., the b.h.p. per litre of the 1909 and 1911 Daimler engines and the 1911 Mercedes at 1,900 r.p.m. being 11.2, 9.85 and 11.3 respectively.
Compression ratios were surprisingly high on early sleeve-valve engines—5 to 1 in the case of the 1909 Daimler— and it was doubtless this that gave them their supremacy over contemporary poppet-valve engines. But there was little scope for development in the Knight engine, so that by 1911 the poppet valve had again attained superior efficiency.
Dr. Riedler goes on to remark that he could not make any of the sleeve-valve engines perform with reliability at prolonged high revs., and adds that their limitation is due to cooling difficulties. Thus, the sleeve-valve engine could only be made reliable in one of three ways:
(1) by cutting down the choke tube ;
(2) by deteriorating the mixture to encourage slow turning, or
(3) by deliberate over-lubrication.
He drily remarks that the first two are the worst—and those adopted by the English Daimler Company. On the other hand, he admits that the third system, which was favoured by Mercedes, leads to oiled plugs, even worse starting and rapid carbonization, although it favours maximum efficiency. In any case, however, it seems doubtful if he deals fairly with the Daimler, which was primarily a luxury carriage, concerned with smoothness and flexibility, as opposed to high speeds; and a small choke is, of course, the easiest way of attaining this end. He does, however, conclude by saying that "The English Daimler car owes its reputation mainly to its good running under light and medium conditions."
Generally speaking, the doctor concluded that upon no single point is the sleeve-valve superior to the poppet-valve engine.
Dr. Riedler then goes on to recapitulate his findings, and here he certainly lays his finger on all essentials of modern design, as the two following passages show: —
"The widespread view that high-speed motors afford no important specific improvement in power, because at a higher speed of revolution a higher actual loss is unavoidable, is already refuted by these results. . . . In reality, high speed motors represent the most important progress made in the building of motors and motor vehicles."
And again: "If special cooling for the piston be not provided for, increase of motor speed requires a diminution of the piston diameters."
Curiously enough, in no part of the book is there any mention of the limiting factor of piston speed.
He then goes on to trace how, in the early '80s, the normal steam engine revolutions of 150-200 per minute were increased to 400 or 500, and before 1890, 800-1,000 had been reached (news to me, I confess, and I wonder who did it). Further advance was stopped, mainly owing to trouble arising from improper valve design, and in adhering for many years to maximum r.p.m. of 1,000-1,500, designers were aiming principally at reliability. By 1912, however, 1,600-1,800 r.p.m. had become normal, with permitted increases to 2,100 r.p.m. for short bursts. He names German Daimler, Benz and Adler as pioneers in the development of the high-speed motor, but ignores the late 19th century advances of De Dion, who was already attaining 3,500 r.p.m., and the successful Bugatti light cars (Type 13) of 1910 onwards. He also adds that the high-speed motor had brought about the use of the gears to aid performance, and it is my suggestion that this first became current practice on the advanced Grand Prix cars of 1908.
The book is of the utmost technical and historical interest throughout—I wish I could find a copy to buy for myself.