The sporting motorist is very conscious of the efficiency of the hemispherical combustion chamber and must derive satisfaction from the thought that this disposition of overhead poppet valves, operated with an equal maxima of mechanical efficiency by twin overhead camshafts, was developed to its optimum pitch by M. Ernest Henri when he was designing racing cars for Peugeot Frères and is found today in such production cars as Jaguar, Alfa Romeo, Aston Martin, Lagonda, Porsche Carrera and M.G.
M. Henri employed four o.h. valves per cylinder, inclined at 45 deg. and operated by two o.h. camshafts, for his successful 1912 7.6-litre G.P. Peugeot, and he brought this valve arrangement to a peak of efficiency in his Coupe de l’Auto 3-litre Peugeot of 1913, the four valves per cylinder of this engine being inclined at 60 deg., and the power output being 90 b.h.p. at 2,900 r.p.m. A larger version of the same engine was used by M. Henri to win for Peugeot the 1913 French Grand Prix.
So outstandindy successful were these 1913 Coupe de l’Auto Peugeots that it is now common knowledge that when Louis Coatalen set his heart on winning the 1914 T.T. race for the Sunbeam Motor Company he bought one of these Peugeots in France and had it sent to England — not much is known of this piece of piracy, “Motoring Entente” (Cassell, 1956) being strangely reticent about it, although tales are told of the car being hoisted through the french windows of Coatalen’s English drawing-room so that his draughtsmen could make an almost literal copy of the car.
The 1914 T.T. Humber was also closely akin to the previous year’s Coupe de l’Auto Peugeot, and it is related (“The History of Brooklands Motor Course”, page 181) how, some fifteen years later, when Wallbank bought one of these Humbers in Folkestone, it was sold to him in good faith as a Peugeot and, brought to Brooklands, continued for some time to deceive both its new owner and most of the racing drivers who inspected it.
This dramatic, if self-confessed, copying of the 1913 Peugeot by Coatalen gave Laurence Pomeroy an opening for a thought-provoking piece of writing in The Motor dated October 1st, 1958.
Under the heading “Technical Topics,” Mr. Pomeroy sought to show that if, for many years, Sunbeam had owed everything to Peugeot for their win in the 1914 Tourist Trophy race, far from history repeating itself when Humber, for the latest Super Snipe engine, adopted a form of valve actuation seemingly pioneered twenty years earlier by Peugeot for the 203 engine, in fact it was now Peugeot who had cribbed a Sunbeam design. Pomeroy put it that “Any debt the Wolverhampton company had to the French concern in respect of World War I design practice has been more than repaid by the debt that Sochaux has incurred in regard to their post-war products.”
The Pomeroy argument is that in 1928 the Wolverhampton Sunbeam Company hit upon the idea of simplifying the actuation of inclined overhead valves by dispensing with the normal one or two overhead camshafts, instead operating the valves from a single camshaft, base-located as in a side-valve engine, through the medium of inclined push-rods passing between the cylinder bores, and unequal-length rockers. This, maintains Mr. Pomeroy, was copied by Tony Lago for the Paris-built Talbot engine when he took over the direction of the Talbot Company at Suresnes before the second World War, M. Lago having been with Sunbeam in 1928. And, to continue this engrossing theme, Pomeroy maintains that Peugeot simply cribbed from Talbot when they wanted the efficiency of inclined o.h. valves without the complication of overhead camshafts in their Peugeot 203. If, as Pomeroy does, we stretch a point and allow that Humber is now in the same group as Sunbeam (Rootes having absorbed both these famous old English companies), the debt Sunbeam owed to Peugeot in 1914 is washed out by the crib Peugeot are said to have made of the 1928 Sunbeam ‘bus engine.
I have taken some steps to trace the 1928 Sunbeam commercial vehicle engine to which Laurence Pomeroy refers. I can find no description of it in the commercial motor journals of the day and Mr. Pomeroy admits he has never been able to unearth a drawing of it. He seems to have based his assumption on a reference, in a contemporary description of the 1928/9 Sunbeam commercial-vehicle engine, to the desirable qualities of hemispherically-shaped combustion chambers, allied to a hazy, and I think false, recollection of someone who once worked at Wolverhampton, that this engine had push-rod actuation, from a single camshaft in the crankcase, of its inclined o.h. valves.
Delving more deeply into the matter, I discovered that there was every reason for contemporary accounts of this 1928/9 Sunbeam engine to refer to the advantages of hemispherical combustion spaces, because Sunbeam was being pretty revolutionary in using heads of this type for the then-new ‘bus engine. But this engine did not have Lago or Peugeot style push-rod valve gear. What it did have were two camshafts carried high up on the sides of the cylinder block, as in the 2-litre Lagonda car engine, short rockers operating the inclined o.h. valves from these camshafts. Moreover, each rocker engaged its valve about mid-way along the stem, the valve spring being located above the rocker, which is a method of isolating the spring from engine heat and facilitating replacement of broken springs. This layout of comparatively long valve stems called for two separate valve covers over inlet and exhaust valves, and a quick glance at the engine, or a drawing of it, could give rise to the erroneous impression that push-rods and rockers were employed to operate inclined o.h. valves.
Incidentally, I have asked M. Lago whether he can recall the sort of Sunbeam engine envisaged by Pomeroy and he says he cannot. He evolved the push-rod arrangement of valve gear for his Lago-Talbot of 1936, he says, because he wanted the advantages of an inclined-valve head without the complexity of “upstairs” camshafts, which in the early days resulted in noise and which at all times increases the cost of manufacturing and servicing the engine.
However, further research reveals that, if we press on two years beyond the date quoted, Mr. Pomeroy is right, and that, indeed, when Peugeot introduced their 203 model in 1948, they could very well have copied or at least have derived inspiration from a Sunbeam ‘bus engine of 1930. This would certainly, as Pomeroy claims, have neutralised the Wolverhampton debt to Sochaux. Our heading picture, by courtesy of Bus and Coach, shows how the valve gear of this Sunbeam engine was arranged.
The whole subject is so complex and obscured in the mists of antiquity that I now propose to take the opportunity of enlarging on it with a view to elucidating the position to present-day students of i.c. engine design.
Early T-head side-valve engines had a camshaft on each side of the crankcase, like a beloved “Alfonso” Hispano-Suiza I once owned. When inclined o.h. valves were introduced the obvious method of operating them was by extending push-rods from these camshafts to rockers above the valves. Pipe did this as early as 1904, followed by Fiat for their Grand Prix engine of 1907, and a production car with this valve gear was the 1908 “Prince Henry” Benz.
The single or double o.h. camshaft was soon recognised as superior, especially for high-speed engines, but those who sought the greater simplicity of push-rod and rocker valve gear found two further solutions. One, much in vogue for 1914-18 aeroplane engines, and used for a time by Salmson for a car engine, was a single “push-pull” rocker actuating both inlet and exhaust valve, a concave depression on the base camshaft enabling a spring to pull the rocker down for the purpose of opening one valve, conventional downward pressure of the extremity of the rocker under the lift of a normal cam being relied on for opening the other valve. This system was practical only for low-speed running and the next step, taken early by Dorman and copied so successfully by Riley in 1926 for the Nine engine, was to locate the two camshafts high up in the cylinder block, so that short, light push-rods could be used.
Although two camshafts were required, the complication of driving and lubricating these was considerably less than with one or more “upstairs” camshafts, while the head could be removed for decarbonising without disturbing the valve timing and without involving lifting a cumbersome o.h.c. cylinder head.
In later years motor-cycle designs used light-alloy push-rods to reduce reciprocating weight but for car engines this was not widely followed, although Georges Roesch saved weight by using steel push-rods of “knitting needle” size, and rockers pivoted on low-friction knife edges, for the Talbot engines of 1930 onwards.
An early attempt to combine these advantages of a base-chamber located camshaft with the better breathing and compact combustion space of the o.h.v. engine was made by Guy Motors in 1919, with their wedge-shaoe inclined side-valve layout, a sectional drawing of which was published in the September 1956 issue of Motor Sport, page 544.
In 1925 Lagonda brought out their 2-litre engine in which two camshafts were located high up, one on each side of the cylinder block, so that bell-crank rockers could actuate the inclined o.h. valves. In this manner, the reciprocating weight of push-rods was eliminated, yet the easy head removal of the high-camshaft-cum-push-rod layout was retained. As I have said, Sunbeam used a similar arrangement for their 1928/9 commercial-vehicle engine, hemispherical heads, incidentally, being revolutionary in this field.
There was yet another method of actuating inclined o.h. valves without recourse to the customary single overhead camshaft and rockers above the valve stems, or a camshaft above each line of valves as in the majority of post-1918 racing engines and such production cars as the 3-litre Sunbeam, 2-litre Maudslay, the six-cylinder versions of Lea-Francis, Vulcan and Calthorpe, the Alta, Alfa Romeo, Type 55 Bugatti, Salmson, Amilcar Six, Beverley, Barnes, Newton, etc. This was evolved for the 1924 Rover 14/45 by P. A. Poppe. Mr. Poppe operated the inlet valves from a camshaft and rockers on the near side of the detachable cylinder head and cross push-rods went through the head to prod the exhaust valves via rockers on the opposite side. By this means Poppe had to provide only one camshaft, but why he mounted this on the head, thus rendering this heavy for removal, is a mystery. Contemporary descriptions of this ambitious car fail to throw any light on the matter, but probably the designer was seeking a compact engine, and he saved, respectively, height and length, by not placing the camshaft above the valves and by driving the camshaft with a vertical shaft and worm gearing set, not ahead or behind the cylinder block, but in the near-side rear corner of it. Poppe may have felt, too, that a camshaft in the head would be easier to lubricate, and quieter than a camshaft(s) enclosed by a metal cover above the block, while he claimed that the different rate of expansion of the metals employed could be balanced out so as to ensure no more than 1.5 to 2 thous. of an inch variation in valve clearance between hot and cold. In other engines this problem of unequal expansion could prove daunting unless provision was made in the drive to the o.h. camshaft, or noisy tappets were endured in push-rod engines. It was also possible to place the sparking plugs in the best position, i.e., in the centre of the combustion spaces, which is not possible with a single o.h. camshaft and rockers above the valves.
In 1930 Sunbeam introduced a far more simple method of actuating inclined o.h. valves without recourse to an o.h. camshaft, as previously described. In 1936 M. Lago evolved a modified version of this system. He used a chain-drive between crankshaft and “basement”‘ camshaft so that he was able to position the camshaft to suit the angle of the push-rods. In conjunction with inclined push-rods he was obliged to use different length rockers but to his delight no difficuties arose, even at 6,500 r.p.m. With central sparking plug power output was equivalent to that of a twin-cam engine and several hundreds of these Lago-Talbot engines were built. As I have said, M. Lago denies emphatically that he copied any other engine in evolving this since widely-copied valve gear.
Also in 1936, B.M.W. introduced their 328 sports model, which also had inclined o.h. valves operated by push-rods. In this case, however, the inlet valves were operated by vertical push-rods and rockers, while the exhaust valves were actuated by both vertical and cross push-rods, the change-of direction being achieved by bell-crank rockers, with additional rockers at the opposite ends of the cross push-rods. After the war this German engine was copied by Bristol and consequently this simple valve gear has become well known in this country, especially as Bristol engines have been used in Frazer Nash, Lister, Tojeiro and other sports cars.
It is interesting to reflect that although these various methods of operating inclined o.h. valves without o.h.c. camshafts were in the minority, each has been used successfully in competition cars. The Riley high-camshafts figured in all cars of this make, from the Nine to the Pathfinder and formed the basis of the pre-war E.R.A. racing engines. Lago-Talbots and 328 B.M.W.s were frequently raced, 2-litre Lagonda cars have been used for competition work, while there was even a Brooklands version of the Poppe Rover.
After the war Lea-Francis, Connaught. and Turner used the Riley high-camshafts system and M. Lago went over to this layout for his “Record” engine, which, in the monoposto Lago-Talbots, had a compression-ratio of 11 to 1 and developed 280 b.h.p. at 6,200 r.p.m. Separate covers were used over the inlet and exhaust valves to provide better cooling for the central plugs than a single valve cover would have provided.
The form of valve gear in conjunction with hemispherical combustion chambers which was pioneered by Sunbeam nearly thirty years ago had several imitators after the last war. Although twin-cam engines were preferable for really high crankshaft speeds and Jaguar has proved that they can be produced relatively inexpensively, the bogey of noise has not been altogether overcome and, indeed, apparently still troubles those B.M.C. engineers who are responsible for developing the new Twin-Cam M.G.
In 1951 the Chrysler Corporation adopted the Talbot-type valve gear for their V8 engines, and Armstrong Siddeley followed suit in 1952 when they introduced the Sapphire, retaining this valve gear for the new Star Sapphire. The Ardun heads fitted to the Ford V8 engines in some Allards also had this form of valve gear (see Motor Sport, May 1950, page 231). Peugeot, as has been explained, had decided on big-output production of this valve layout for the 203 in 1949 and use it also in the alloy-head 403. At the last London Motor Show Humber introduced the new Super Snipe, the notably smooth and willing six-cylinder engine of which is the latest to use the kind of valve actuation under discussion.
This, and the B.M.W./Bristol layout, necessitate twin rocker shafts and, partly to obviate this complication, towards the end of 1954 the Chrysler engineers reverted to conventional push-rod o.h. valve gear, with wedge heads, forfeiting the centrally-located sparking plug.
I must confess that before I had done the necessary research for this article I was under the impression that the pre-war Lago-Talbot engine had the same cross-push-rod valve gear as the B.M.W., and this was endorsed by a not altogether lucid reference in John Stanford’s book “The Sports Car” (B. T. Batsford, 1957). The two systems are, however, quite distinct, as the accompanying diagrams prove.
Here I would also interpose the remark that if I seem to be devoting a good deal of space to the evolution and present state of unusual valve gear, this may be all to the good, judging by the ignorance which surrounds the subject — at the last Earls Court Show the head man on one minicar stand told me emphatically that his cars had normal vertical valves, whereas investigation of a sectioned engine showed them to have the superior inclined o.h.v., while on inquiring of one of the High Priests on the B.M.W. stand whether the V8 B.M.W. uses cross-push-rods it became patently apparent that he just didn’t know!
Incidentally, a third form of inclined push-rod-operated valve gear is found in the Lancia Appia, which has virtually the Riley-type layout but with the two camshafts down in the base-chamber instead of high set to shorten the push-rods, while the positioning of the exhaust-valve rockers differs, no doubt to accommodate the narrow-angle vee-cylinder formation.
In the case of single-cylinder or flat-twin minicar engines a simplification arises because the valves can be placed in line with the crankshaft instead of transversely across the engine, so that inclined o.h. valves can be operated easily from simple push-rod and rocker gear, as in motor-cycle practice; Lancia continues to do this on the vee-six Aurelia engine, as the above illustration shows. Alas, not, every minicar engine takes advantage of this to employ inclined valves, but perhaps any form of poppet valve is better than the two-stroke cycle!
Emerging from the historical to the topical, it is pleasing to find plenty of variety and ingenuity amongst modern valve gears, such as, for example, the use of torsion-bar valve springs in the Dyna Panhard engine and the eccentric drive for the overhead camshaft on the little twin-cylinder N.S.U. Prinz. The latter recalls the triple-eccentric o.h. camshaft drive used by Parry Thomas for the Leyland Eight engine of 1919 — he also used laminated instead of coil valve springs, as a means of obtaining progressive damping and isolation of the springs from engine heat — and the crankshafts-cum-connecting rods which W. O. Bentley instituted in 1925, likewise in search of silent operation, for driving the o.h. camshaft of the Big Six Bentley. The same system of eccentric drive pioneered by Parry Thomas was employed, even more ingeniously, to operate the twin o.h. camshafts of the 2-L-6 16/80 Maudslay in 1923.
In this essay I have sought to study methods whereby inclined o.h. valves can be operated without using o..h. camshafts, and I do not propose to deal with single and twin o.h.c. layouts, i.o.e., e.o.i., rotary valves, sleeve valves, slide valves, desmodromic valve gear or other alternatives. Nor do I propose to embark on such aspects as camshaft drives, valve timing, cam contours, valve size, spring strengths, seat angles or even the many diverse types of valve spring cotter, the design of tappets and cam followers or the pros and cons of salt-filled valve stems, or valve inserts!
This article endorses Mr. Pomeroy in ensuring that credit will now be given to Sunbeam of Wolverhampton for showing such great engineering houses as Talbot, Peugeot Frères, the Chrysler Corporation, Armstrong Siddeley and Humber how to enjoy simplification while retaining efficiency in this department of engine design. I hope I may have opened the eyes of some of my readers to the various methods used to actuate what was once regarded as the most dubious, and therefore likely to be short-lived, item in the i.c. engine — the poppet valve. — W. B.