SIDESLIPS ” BALADEUR “
AS-seen as I saw that a correapondent, Mr. G. Arnott„ and the Editor were discussing desmodromie valve mechanism in the: October number of MOTOR SPORT, it occurred to me t hat. to intervene in this dismission would provide t he oppartunity to get a couple or words in ” sideslips ” printed in Greek, which would lift-this series or articles into a very highbrow category indeed and continund those crities who are inelitied to (sew plain about the occasinmal Ilippaiwies in them. Mr. Arnett characterises ” desmodromique “-which is ;teat:illy I i kink the French spelling of desmodromie—-aa a ” weird anima,” and in order to dispel the supernatural atmosphere I would explain that this apparently mysterious term is derived from two quite simple Greek words; 8aettesa a ” halter,” and 5petata., a ” running,” so that the whole thing means “running on the end of a halter or tether.” What this has-got to do with valve mechanism will, I hope, become clearer in the course if this article.
In 1914 the French Grand Prat WitS for the first time run wider a capacity limit formula. the maximum engine size permitted being 4i litres ; am tilt .hint ary of that year the French journal Omnia set itself the problem of decidilig wlett. sort of cars this formula was likely to produee. The engine, it assumed, would be a four-cylinder, and the permissible (limera siona, it calculated, would lie between 80 by 224 mm. and 120 by 100 mm. In order to get rapid combination of the Mixture, it argued—and this was the authentic voice of the long-stroke era speaking—the stroke should be as long as possible in relation to the bore, in order to restrict the area of the combustion chamber ; but in practice engine dimensions had to be determined byreference to two limits, a ma:Sin-aim speed for the piston of 15 metres per second (Say 3,000 feet, per minute), and a maxilla-um speed for the crankshaft of .3,000 r.p.m. Accordingly, Onmia calculated, the dimensions were likely to be 95 by 147 ram., and if but.e.p. were six kilograms (taw 90 lb.), which it thought would be very good, the output would be 85 lap. at 3,000 ap.m. In actual fact the winning Mercedes had dimensions of 03 by 105 mm.., and maximum power was developed at 2,800 r.p.m., but the lam.e.p. at 120 lb. was so much higher than Omnia expected that the output was 115 lap. Whatever may have led Omnia into error over the h.m.e.p. to be expected, it Was entirely justified over piston speed Which was actually 3,050 feet per minute, and Over crankshaft speed ; and it is illuminating, for our present purpose, to gee what, considerations in 1914 placed the limit for the latter at 3,000 ap.in. 10 the first place, there were doubts about the ability of a magneto to give reliable service at Itigater speeds. doubts Which after the war were for a short lime to induce builders of small Itiglaspeed racing engines, such as the 1,50
and of closing, including starting, full speed and stopping, does not exceed one two-hundredth of a second. To achieve this result with a lift of 12 mm., the maximum speed of the valve Intuit not be less than four metres 800 a second, or 17 kilometres an hour [say 10 m.p.h.], in one four-hundredth of a second therefore, the valve must pass from a speed of o to a speed of 17 k.p.h., and in the next four-hundredth of a second from a speed of 17 k.p.h. to a speed of 0. The momentary power necessary to produce this change is 0.4 h.p. per kilogram of the valve weight. These figures show what .difficulties One comes up against when one approaches such crankshaft speeds.”
Whether in feet these figures did show many of Omnia7s readers what sort of difficulties they were likely to dome up against if they insisted On driving their valves at more than 10 miles an hoar I de not know, but in 1920 La Vie Automobile returned to the subject in rather more practical vein. In any ordinary valve system, it explained, the cam only attends to the opening of the valve, and leaves the business of closing it to the spring. The latter can only communicate to it, an acceleration which is constant, depending solely on the strength of the spring and One Weight of the valve, phis the weight of any intermediaries there may be between it and the cans, such as tappets, path-rods, rockers, etc. The speed or Closing, therefore, is constant, and does not vary with the speed of the engine, so that ” it can happen, when engine speed is very high, that the time taken to close the valve is greater than the time the cams takes to traverse the segment of the circle. corresponding to the closed position of the valve: The latter no longer follows the profile of the earn, it lags behind and the timing is upset.” The situation is aggravated in this respect because, in order to avoid the stratification of the gases, the valves Must be made to open and close as sharply aS possible, which involves the use of steep earn faces. As a result, the Valve reaches the full lift position with eonaiderable RIOTRentura, which, if the valve is large and therefore heavy, and the engine speed is high, may overcome the resistance of the spring, with the result that the valve leaves the cam lace and opens too far. It then has further to ravel when shutting, which again causes il to lag and tip aas the timing. All of which, cola:lin led La Vie Automobile, ” explains why an engine must have stronger valve springs the faster it turns, Even so, when you get to very high speeds, it is impossible to avoid some derangement of the tinning.” The difficulty is, as at .least everyone who has had experience of the older ears is aware, that however strong the valve springs are, they cannot prevent the valves from bouncing, and it is this factor which in so many Edwardian and vintage cars sets the limit to engine speed. Indeed, the stronger the spring, and the greater the force with which it bangs the valve down on its seat, the greater the reaction waieli seeks to bounce it off again. The problem. I take it, is almost exactly the same as in the case of road springs. When a wheel is thrown oft
the road by a }k Stiff nano put it back again very sharply ; but, the more sharply it does so, the greater the tendency of the wheel to rebound. The solution, as was realised as loag as fifty years ago, is to supplement the main swing with another having a different rate “—in this ease a shock-absorber–whiela while incapable of dealing with the primary business of returning the wheel . to the road, can cope with the after-effects of the main suspension unit’s unduly brusque efforts to do so. Similarly, by fitting valves with double springs, one for the prim nary purpose or dosing the valve, the ether for the secondary parpose of preventing it boancieg off its seat again, designers were to find that improving technique was to permit of engines With ordinary valve mechanism being run at. Speeds which made 3,000 ratan. look like little more th.an a brisk tick-over.
For the raement, however, this appeared to be merely in palliative, while a much more radical solution of the problem presented itself. The trouble caused by the valve ceasiug to follow the cam face, continued La Vie Automobile in its 1920 article, does not arise if the valve is linked continuously to its operating meelatnism„ so that neither can move without drawing the other in its train. “This is called a desinortromie link, and is similar to that, for exanaple, between a piston and its crankshaft. A Knight engine has a desmodromic drive for its sleeves ; that is one of the reasons for the advantages of BUS type of engine ; whatever its speed may be, the tinning rent tins correct.” Not altogether unnaturally, therefore, designers had been inspired with the desire to de the atone sort of thing with poppet valves ; seeing these latter bucking and bounding during that part of each revolation when they were not under the command of the earn face, they desired to control them all the time,. ” desmodromically “—to “run them on the end of-a tether.” It is usually suppased that the first efforts. to give practical effect to this desire were ma ule in 1014, on the occasion of that Grand Prix the ‘technical prospects for which were so ably diaaissed Iry Omaha with on however, tiep of this development being in outer!. As usual, however, the app trent pioneers were afterwards found o hi i.e been anticipated by several years. In in letter to La Vie Automobile published in 1991, lintron PcI iet claimed that the singlecylinder Aries racer built for the Grand Prix des I/ultimates of I Mai not only had ” iaduction by mixture previously compressed to a pressure higher than that of the atmosphere “—in (a her winds some kind of supercharger–but also ” four valves per cylinder desmodromieally operated by an overhead camshaft.” This engine used the maximum permitted bore of too nun. combined with a stroke Of 180 mm., Which Scents to have been longer than that used by any other competitor in the race ; and as, in addition to the features already mentioned, it had four sparking plugs, and exhaust valves at the bottom of the stroke, great things were doubtless expected of it,. Unfortunately, however, the design was perhaps a little too far in advance of its time, with the result that, of the three Aries which started in the race, one went out on the first lap and the other two on the second, without any of them figuring very prominently in the competition while they were t here. As a result, 1st contemporary, is far as I CAM find, seems to have bothered to describe eit her I his exceedingly interesting engine in general or it s desmodromic valve gear in particular. NN hen will technical conunentators conic to realise that the oars which fail are oft en more worthy of their attention than those Nvitich succeed Y
The single-cylinder Deluge which won this 1908 race was said to develop its maximum power at 2,800 r.p.m., and during the next few years there was extraordinarily little advance in crankshaft. speeds. Instead, bore limitation led designers to increasing strokes and therefore piston speeds, and even when this formula gave way to a capacity limit they still found themselves unable to seek increased power by making their engines turn materially faster. Even the 3-litre Peugeot which won the Coupe de CAW() ill 1913 still developed its ntaximum power at 2,1100 r.p.m., a speed doubtless attained by some of the fourcylinder engines in the 1908 race, and at the beginning of 1014 ()nada, as we have seen, still placed t I a !Mattieuil IMM Of crankshaft. speed at 3,000 r.p.m. At last, however, there were some designers who were resolved to break front this thraldom, and the weapon which they proposed to use for the purpose was desmodromic valve operation. I have no doubt that in some minds this development promised to opendoors as wide as those which had been passed through when automatic inlet valves gave place to inlet valves mechanically operated. The two makers who in 1914 decided to embrace the desmodroinie faith. were
Deluge and Schneider, and, Curiously enough, each of them resolved to climb onto the oppiisite horn of the dilemma with which tItis ft Lit U confronts its adherents. ‘1’lle d hal-MIA in qtiestiOn arises front the fact that it’ a valve is to be positively placed on its seating, it is almost impossible to find an adjustment for the meclutnisin which in all (imamstances will do so with the necessary precision. If the valve is exactly drawn onto its seat and held tliere when the engine is cold, it will not be exactly placed there when the engine is hot, and vice versa. Either the mechanism. Innat be sO adjusted that in certain circumstances the valve is not completely closed, in which case there will lie loss of compression, Or else the mechanism will seek to continue drawing the valve onto its seat when it is already on it., in which ease something will break. Of these two alternatives, Deluge chose the second and Th. Sehneider the first, both of them, however, at the same time making arrangements designed to avert the ill effects of their choice. In the 1911 Coupe de PAW°, and again
in the 1913 Grand Prix, Delage had used horizontal valves, operated by a necessarily complicated—and therefore heavy— system of push-rods and rockers. Now, however, he decided on two overhead camshafts, with four valves per cylinder, mounted at-45 deg. to the vertical. Each pair of valves was operated by the same tappet, having a stirrup-shaped extension, inside which the camshaft revolved. The latter carried two cants inside each stirrup, one of which, with a fairly norinal contour, bore against the flat bottom of the stirrup and opened the pair of valves t While the other bore against the arched section of the stirrup during the greater Part of each revolution and while doing so held the valves onto their seatings. In order to do so it actually pulled the valves too far towards their scats, but the pull was exerted on a cross-piece incorporated in the tappet, which in turn bore against two small springs. in order to provide the requisite play.
The Th. Schneider system was entirely different. There were only two valves per cylinder, they were mounted at 10 deg. to the vertical, and were operated by it single overhead camslutft. This, however, carried two earns for each valve, One of which opened the valve by means of a rocker in a quite conventional manner. The other cam, however, which Was mounted just behind the first one, bore during the greater part of each revolution on a third -arm of the roeker, which exerted. its leverage on the other side of the fulcrum and thus, since the attacking end of the rocker was situated between two bosses on an extension of the valve stem, drew the valve towards its seating. Just before it got there, however, the rocker left the eam face, and the valve was given ” the last few tenths of a millimetre ” of its closure by an ordinary valve spring. In this respect the system was thus the antithesis of that adopted by Delage, and was not, perhaps, in the strictest sense, completely desmodromic. Nevertheless, said La tie Automobile. ” this is certainly a new and original solution which is wortity of’ all our attention.” One would have expected of a novelty of this sort either Butt it would have entirely disappointed its sponsors, as, presuthably, did the Aries engine of 1908: or that it would have been a howling success front the first, as was the Pettgat engine with two overhead camshafts in 1912 ; or that it would have been a brilliant failure, as were the supercharged FIATS in the 1923 Grand Prix, which dominated the opening stages of the race and all of which went out before the end of it. But in actual fact it proved to be none of these things. The Deluges, it arwears, were brilliant in practice, and there was some despondency in the Sunbeam (‘amp, beeanse the ears were to be started in pairs, and each Sunbeam was paired wit Ii ii Delage. On the eve of the race. It OS ever, I Iu)se responsible for t hese promising lavintrites apparently got cold feet, and made some last-minute adjustment to the valve gear, presumably to the tension of the cushioning spring, which so far robbed the Dclages of their brilliance that their acceleration proved inferior to that of the Sunbeams. Even so, they did not become dismal failures, although the performance of the three
cars of the team Seems to have been markedly unequal, the car driven by Minty being decidedly faster than those handled by the two outstanding Deluge drivers, Guyot and I3ablot. For the first six of the twenty laps, Duray ran third to a Mercedes and •Boillot’s Peugeot, for the next three laps he was fourth behind a Mercedes and two Peugeots, and although he fell back just before half distance, it was his sixteenth lap that was his fastest. He covered this lap at 0.85 m.p.h., it speed which was only improved on by three of the Mercedes, Boillot’s Peugeot and Resta’a Sunbeam. Duray’s Delage was thus certainly one of the fastest cars on the course, and also was the only one of the team to finish. Since, however. only It finished out of 37 starters, and of these eleven -three were Mercedes, the pelage record was it least better than the average.
Curiously enough, Champoiseates Schneider, which was also the only one of its team to finish, was likewise the fastest. Champoiseatt’s best lap was covered at 64.92 m.p.h.., a speed which was exceeded by all live Mercedes, two of the Peugeots, all three Sunbeams, Duray’s Deluge and Fagnano’s F.I.A.T.; but which exceeded that of the remaining Peugeot, the other two Deluges, the other two F.I.A.T.s and the best representatives of Nagant, Opel, Piceard-Pictet, Alda, Nazzaro, Vauxhall and Aquila-ltaliana. In fact the performance of the ears in the 1914 Grand Prix with desmodromie valve control was just about what one would have expected Of them if they had not leld desmodromic valve control.
The intervention of the war at this juncture renders obscure to posterity, with so much else, what, as a result of their experiences in 1914, those who had used it thought of desmodromy (if there is such a word). If Delage and Th. Schneider had built racing cars in 1915, would they have used positive valve closing? This must remain an historical mystery, because when Grand Prix racing was resumed in 1921 neither of these makers it first took part. in it, and, as far as their potential disciples were concernetl, two inoiortant iteW factors hail entere,i into their calculations : the capacity limit, had been reduced to three litres and the straight-eight engine had beciinte the vogue. The resultant decrease in the ske of each indiviilual valve had greatly rediteed their weight, and, partly as a result, no doubt, it was now found possible to run engines with ordinary valve gear at over 4,000 r.p.m. histead of 3,000. Positive valve operation, however, had not been forgotten, and with the further reduction of the capacity limit in 1922 to Iwo litres, designers were faced with the iirospe.ct of engines running not at 4,000 r.p.m. but at 5,000. ” There will be difficulty over timing meelianisth,” prophesied NI. Charles Faroux, writing in La Automobile in February, ” and valve control will be -assured in both directions ; the term applied to this is destnodrott,ic ‘.” Delage designed a car for this race which was, I believe, a four-cylinder, but as it was not ready to appear at Strasbourg I do not. know whether its valves were controlled ” in both directions “; and the Deluge which Continned on page 6:11 SIDESLIPS—continued from page 624
did appear at Tours in 1923 had twelve cylinders and ordinary valve mechanism. In the meantime, however, there had been more titan one convert to the desmOdeoznie principle. In May, 1922, it was announced that Rolland-Pilain Were building cars for the Grand Prix Which, in spite of having straight-eight erigines, were to have ” positive closing Of the valves by means of a tam “; but almost itruntsliately afterwards it was announced that the designer of this engine had provided it with two alternative heads, one desmodromic and one not, and when the cars actually appeared at Strasbourg they were found to have ordinary spring-closed valves.. They retained this conventional arrangement when they appeared again the next year at Tours, and as a result I have never,
I think, seen a description of the desmodromie system used in the discarded design. Mr. Arnott says that the mottonism used involute cams ; but, apart from this information, I remain, alas as ignorant of it as I am Of the 1908 Aries engine. Curiously etalagh, however, the devotees of positive valve closing who journeyed to Strasbourg in 1922 for the Grand Prix were not disappointed in their hope of seeing a desmedrontie system in action, in spite of the recantation of Rollaral-Pilain. In conjunction with the big event for 2-litre ears, the A.C.F. that year ran a Grand Prix de Tourisme on a limited fuel consumption basis, the ears having to cover 444 miles at the equivalent of 106 miles to the gallon, and among the starters were three 2-litre Bighans With positive valve control. The system used in this engine is exceedingly hard to describe, especially without a diagram, but I suppose that, having embarked on this theme, I intua make an attempt to de so. The eight vertical overhead valves of the fearcylinder engine, then, were arranged in two squares, each square comprising the inlet and exhaust valves of two cylinders. In the centre of each square, and driven through bevel pinions by a horizontal overhead shaft, were two plates, circular In plan and revolving in a horizontal plane in opposite directions. On an extension of each valve stem were two conical rollers, which fitted one above and the other below one of the plates already mentioned, which were in fact curiously contoured swash-plates. As they revolved, therefore, they imptrted an up-and-down motion to the valves, which were thus positively controlled throughout their cycle of operations. No springs whatever Were used in this system, which was, indeed, more like having a bull on a pole
than a halter, and as far as I can see no provision was made for expansion and eontraction due to changes in temperature. Nevertheless this engine undoubtedly worked, and could be regularly run at 4,000 apan„ with a maximtun crankshaft speed of 5,000 Nam. Admittedly none of the ears which started in the Grand Prix de Tourisme in July survived the full 444 Miles, but it was claimed that their failure was not due to their desmodromic valve gear ; and in August one of them, driven by Gros, -ran third in the Belgian Greed Prix at Spa to a 3-litre Imperia-Abadal and a 8-litre Chenard et Walcker. Jaciptes Bignan and his colla borators in any ease were so much impressed Wi tit I he StICCt•titi of the design that the desmodamtie 2-litre was shown as a. standard model at the Paris Salon of 1922 and at the %%late City, where it attracted a -great deal of attention. As far as I know this is the only .engine with pusitively-closed poppet valves that has ever been offered to the public, but I have no idea how many cars fitted with it were Actually sold, or how their owners got on with them. I have ray suspicions in the matter, because the model seems to have been dropped in 1923 ; and since
then the word desmodrornique,” which in those days was on practically every enthusiast’s lips, has evidently subsided into the limbo of forgotten things.