As a relief from old-car topics we have pleasure in presenting a review of a paper which M. Lory delivered before the war, translated and commented upon by J. Lowrey, B.Sc. It contains much of intense value and interest, but has never previously been given prominence in this country. – Ed.
The basis of this article, translation, review, or what will you, is a paper which M. Lory read before the Société des Ingenieurs de l’Automobile before the war. It bore the title “Le moteur de la voiture de course,” and dealt with engines for cars complying with the last G.P. formula of the A.I.A.C.R. M. Lory is, of course, well known in this country for his work with Delage.
From consideration of the weight restrictions imposed on cars of various sizes by the formula, M. Lory came to the obvious conclusion that the only types of car having any prospects of success would be those of 3 litres supercharged and of 4 1/2 litres unsupercharged. Whichever of these two types was adopted, he considered the best form of engine to be a 45° V-16, having the induction system on the inside of the V and the exhaust systems on the outside. Such an engine is reasonably light and compact, and it is scarcely practicable to use any greater number of small cylinders in the search for maximum power per litre.
Considering first the unsupercharged 4 1/2-litre engine, 16 cylinders would require to be of 281 c.c. each; 66.8 and 80 mm. are quoted as the best bore and stroke to obtain the desired swept volume, though no reason is given for choosing a stroke/bore ratio of 1.2 for both this engine and the 3-litre considered later.
With a cylinder of this size, and a hemispherical combustion chamber, two valves of 40 mm. head diameter, 37 mm. throat diameter and 9.5 mm. lift can be accommodated. Peak power is anticipated at 6,200 r.p.m., with a piston speed of 3,250 feet per minute and a gas speed through the inlet valve of 203 feet per second. In case any reader should attempt to check these figures, I will say that piston speed is O.K., but gas speed comes to about 176 feet per second, unless you allow for an unduly large diameter valve stem, so that I should expect slightly higher peak engine r.p.m.
The estimated b.m.e.p. is 156 lb. per sq. in. at peak torque, falling to 142 lb. per sq. in. at peak power, giving 310 chevaux-vapeur or a matter of 305 b.h.p. at 0,200 r.p.m. Interpolating a purely personal comment, M. Lory would certainly be doing well to get this b.m.e.p. unblown; with a flat top piston under his hemispherical head, the compression ratio would be under 5 to 1, and with a fancy piston shape to get the necessary compression ratio the combustion chamber shape would be more than slightly imperfect.
Passing on to consider the supercharged 3-litre engine, still of V-16 form, the size of each individual cylinder is 185 c.c., and M. Lory suggests bore and stroke of 58 mm. and 70 mm. respectively. A hemispherical head is again assumed, with valves of 35 mm. head diameter, 32 mm. throat diameter and 7.9 mm. lift. Allowing again, apparently, for obstruction of the port by the valve stern, a mean gas speed of 203 feet per second is found to correspond to 7,500 r.p.m.; this gas speed is assumed to give peak power output, the corresponding piston speed being 3,450 feet per minute.
With an unspecified boost pressure, the b.m.e.p. is estimated as 270 lb. per sq. in. at peak torque, falling to 240 lb. per sq. in. at peak power, the resulting b.h.p. being 415 at 7,500 r.p.m. This is about 140 b.h.p. per litre, not unreasonable when it is considered that the cylinder size is almost the same as on such cars as the twin o.h.c. Austin or the 200 odd m.p.h. M.G. Magnette of Major Gardner.
M. Lory estimates that although the 3-litre engine develops 35 per cent. more power than the 4 1/2-litre, at a speed about 20 per cent. greater, the smaller valves and reduced valve lift result in the inertia load on the valve-operating mechanism being 35 per cent. less on the supercharged engine.
After thus estimating the relative power outputs for the two types of engine, an attempt is made to assess the comparative performances of G.P. cars fitted with the different types of engine. An all-up weight of 2,450 lb. on the starting line, a frontal area of 13 sq. ft. and a drag coefficient appropriate to the rather imperfect streamlining of a road racing machine, are assumed. Balancing available power against rolling and air resistances, the maximum speed of the unsupercharged 4 1/2-litre car is estimated as 183 m.p.h. and that of the supercharged 3-litre car as 205 m.p.h.
Not content with this maximum speed comparison, acceleration curves are plotted for the two cars and interpreted in terms of the Rheims circuit used for the French G.P.
Assuming both cars left the Virage de Thillois together at 25 m.p.h., and accelerated away towards the timing box 1,850 yards distant, the 3-litre should pass the timing box at 186 m.p.h.; the 4 1/2-litre car would follow it past the box at 177 m.p.h., having lost 2 1/2 secs. In a complete lap of the 4.85 mile course, the unsupercharged 4 1/2-litre car will lose 11 secs. on its rival. Unfortunately, I have not at present got access to the necessary records, but perhaps the Editor can give the actual differences in lap times between V-12 Delahayes and Mercédès on this course? [The best Delahaye lap in practice was 2 min. 57.2 secs., and of the Darracq, 2 min. 46.3 secs. These times are some 18.2 secs. and 17.3 secs. slower, respectively, than an average of three best practice laps by Mercédès and Auto-Union. – Ed.]
Having established to his own, and probably everyone else’s satisfaction, the superiority of a 3-litre blown engine over a 4 1/2-litre unblown, M. Lory passes on to some more detailed matters of design.
In regard to general layout, he confirms his partiality to the 45° V-16 type of unit. Owing to the small angle of the V, chosen to obtain even firing intervals, there is not room for two overhead camshafts per block of cylinders; instead, one camshaft must operate all 16 inlet valves, there being a separate exhaust camshaft for each block. To relieve the valves of side thrust, without introducing excessive reciprocating weight, fingers between cam and valve (as on Alta and other engines) are suggested.
To obtain the desired peak speed of 7,500 r.p.m. the inlet valve would have to open 15°–20° early and close 55°-60° late. Probably multiple valve springs would be used, though a form of torsion spring is suggested, and the author makes passing reference to his interest in some form of valveless engine. The suggestion of a torsion bar valve spring is an interesting development from the hairpin valve spring used on certain motor-cycle engines.
Turning to consideration of the bottom half of the engine, it is thought highly desirable to make the engine as short as the bore of the cylinders permits. This restricts crankshaft design considerably, leaving very restricted space for eight crankpins, nine main bearings and sixteen webs. The only thing to do is to stagger the two banks of cylinders and use side-by-side roller bearing big-ends.
In order to maintain reasonable crankshaft rigidity this component must inevitably be made in one piece, and this means splitting the roller bearing big-ends. A nickel-chrome steel connecting rod is suggested, with a length between centres of twice the stroke and with the roller races case-hardened to a Brinell number of about 700. The big-end rollers are separated by cages of forged R.R. alloy.
The use of rollers of 1/4″ diameter and 1/2″ length is suggested, and it is stated that each active roller should withstand a load of about 1,800 lb. per inch length. The suggested crankpin diameter is 18 mm. (0.71″), in conjunction with the most generous possible crank webs and main bearing journals. On appearances this 18 mm. diameter crankpin seems positively indecent to me; maybe M. Lory stressed everything carefully before quoting that figure, but I should have anticipated an appalling lack of torsional rigidity.
To conclude his paper, M. Lory quoted the probable weight of his proposed supercharged 3-litre engine as about 650 lb., representing little more than 1 1/2 lb. per b.h.p.
Likewise, concluding my very free report of it, I feel obliged to apologise for certain rather obvious omissions and inconsistencies. They mostly originate in the original paper, which gave the impression of having been cut down to a definite length. Interest remains in a project by a leading French designer, which was definitely far more advanced than anything which that nation ever actually produced in the immediately pre-war period.