Veteran Types-XXXIII
A 1922 T.T. Sunbeam by Cecil Chilton
No single person has so completely dominated racing design in his time as did Monsieur Henri during the decade from 1912 to 1922. His admirers and imitators were many and the 1914 T.T. Sunbeam was an almost nut-by-bolt copy of the 1913 Peugeot. Having set the pace for high-speed, 2-o.h.c. engines before the Kaiser war, when racing started again in 1919, Henri pioneered the straight eight engine which has remained the most consistently successful layout until, at the present time, it seems likely at last to give way to the V12 and V16.
Despite much head-shaking by the pundits, the majority of cars entered for the 1921 Grand Prix had eight-cylinder in-line engines and the marked external similarity between the engines of the Henri-designed Ballot, and the Coatalen designed S-T-D cars showed that Coatalen had not forsaken his hero. Nevertheless, the 1921 cars were no mere slavish copies as their predecessors of 1914 had been, and while the S-T-D team were not, perhaps, as fast as the Ballots, they were in many ways more modern in outlook, and if one contemplates the similarity of layout and design between them and a Type 57 Bugatti, it is seen that here was a model for sports car designers for the next 15 years.
It appears that seven of these cars were prepared for the 1921 Grand Prix, it being intended that two of them should run as Sunbeams, two as Talbots, and three as Talbot-Darracqs. The Sunbeams did not start in the race, but the Talbots all finished and Boillot’s car gained fifth place. In the same year the team competed at Indianapolis, and one of the cars finished fifth at an average of 83 m.p.h. At Indianapolis the front brakes were omitted and it was said that the chassis was offset on the axles—a trick practised by Peugeot for the 1911 race.
In 1921-22 the Sunbeams were considerably rebuilt for entry in the 1922 T.T. race over the mountain circuit in the Isle of Man, and various modifications were carried out. In the G.P. the foot brake was reported to operate on all four wheels and the handle on the rear drums only. For the T.T., the pedal operated the rear brakes only and the front were controlled by the hand lever.
Four carburetters were fitted in 1921, and different sources of evidence give them as being variously Claudel Hobson, and horizontal Zenith instruments, fitted with forward-facing air scoops, which were supposed to provide a mild supercharging effect. Whether the French cars actually used four carburetters in the Grand Prix seems open to doubt. It was said at the time that two produced greater power at either end of the power curve, while four were more effective in the middle range. As this is the opposite to what one would expect it seems not unlikely that the carburation in both cases may have been rather hazy and, indeed, most carburetters during the Henri decade seem to have had unnecessarily large chokes.
Delco coil and distributors were used for the G.P., but for 1922 these were supplanted by two B.T.H., four cylinder magnetos. With the very poor spark at high speeds offered by coils in the early twenties, it is difficult to see what practical advantages can have been attributed to their use on a racing car at that time.
In the T.T., of the three Sunbeam entries, Kenelm Lee Guinness was a non-starter owing to clutch slip, Segrave retired on the 5th lap with magneto trouble, and the race was won by Chassagne at 55.75 m.p.h. The race was noteworthy for the first appearance in racing of a Bentley team, being practically standard production models, all of which finished, the best securing second place. The 1 1/2-litre race was won by Algernon Lee Guinness on a Talbot-Darracq.
The multi-plate Hele-Shaw clutches were a rational departure by Coatalen from Henri practice, for although multiplate clutches had been popular in racing for the previous 15 years, Henri retained, to the last, an almost pathetic faith in cones. It did, however, happen that the Sunbeam clutches gave a lot of trouble, being unequal to transmitting the power supplied.
After the T.T., the cars had quite a long racing career and, even previously, on March 28th, 1921, Segrave drove one with an Indianapolis chassis to victory in a 3-litre scratch race at Brooklands, averaging 94,64 m.p.h., with a fastest lap of 101.64 m.p.h. This car had four Zenith carburetters. The best Brooklands lap recorded by any of the team was 102.9 m.p.h., by Dario Resta in the 38th, 100 m.p.h. short handicap, Easter, 1924. One handled by C.J. Jackson also performed successfully at Southport for many years.
The specification of the Sunbeams in T.T. form was as follows : Engine.-8-cylinders in line, 65 by 112 equals 2,973 c.c. Two blocks of four cylinders, non-detachable heads, with aluminium blocks and steel liners. Four valves per cylinder operated by twin overhead camshafts driven by train of straight-cut gears at front of engine. Crankshaft is a nickel-chrome forging, with counterbalance weights, carried in five white metal main bearings. The front four webs are parallel to each other, being at right angles to the rear four, giving a firing order of 1, 8, 3, 6. 4, 5, 2, 7. Connecting rods H-Section, white metal big-end bearings. Aluminium pistons. dry-sump lubrication ( separate tank under mechanics seat) effected by gear-type pump. Minimum safe oil pressure 85 lb. per square inch, two vertical Claudel-Hobson carburetters with water-jacketed manifolds. Fuel feed by pressure from a bolster type tank at rear. Two B.T.H., 4-cylinder magnetos on off side of engine. Weight of engine dry 520 lb., 108 b.h.p. at .1,000 r.p.m. equals 4.8 lb. per b.h.p. Fuel consumption .7 pint per b.h.p./hour at 3.000 r.p.m. ; .65 pint per b.h.p./hour at 3.500 r.p.m.
Transmission.—Hele-Shaw clutch with 6 plates, alternate phosphor-bronze and steel, driving to seperate four-speed gearbox having forward ratios 1, 1.165, 1.725, 2.6 to 1. Open propellor shaft. Bevel driven rear axle with alternative ratios of 3.69 or 4 to 1 ; 4 to 1 was used for T.T. and a higher ratio still was probably used at Brooklands. Wheels, Budge-Whitworth wire detachable with straight-sided tyres, rear 32 ins. by 4 1/2 ins. : front 32 ins. by 4 ins.
Chassis.—Wheelbase : 8 ft. 9 ins. ; track : 4 ft. 7 ins. Upswept over front and rear axles. Engine and gearbox carried in sub-frame. Front end of crankshaft is below centre-line of rear axle to give lower centre of gravity. H-section front axle. Semi-elliptic springs front and rear.
Brakes.—Ribbed steel drums on four wheels. Original linings were cast iron. Foot brake to rear drums ; hand lever to front drums. Ratchet adjustment for both front and rear operated from mechanic’s seat. Operation of front brakes is on the Isotta Fraschini principle.
Body.—Narrow two-seater. Staggered seats ; 39-gallon, bolster-type copper petrol tank. Flared front wings used in T.T. Approximate frontal area without wings, 111 sq. ft.
Weight as for T.T. race 2,595 lb. (weight in G.P. form was 2,184 lb.).
Of the T.T. team, all three chassis still exist and most or all of the remaining engine parts are owned by Anthony Heal, who also has Chassis No. 3. This car probably ran in the 1921 Grand Prix as a Talbot. (with, of course, a Talbot radiator) and finished 10th, at 62.6 m.p.h., driven by Segrave. Segrave again drove it (after rebuilding as a Sunbeam) in the 1922 T.T., retiring with magneto trouble on the 5th of the 8 laps. Segrave, however, led for the first three laps and put up the fastest lap of the race, in 39 min. 15 sec., at 57.7 m.p.h. This was the first lap and was, therefore, from a standing start. After passing through various private ownerships the car finally came into Hears hands in 1942.
During the past eighteen months it has been pulled to pieces, cleaned up, reassembled in T.T. form and painted (British racing-green) by Len Gibbs, of Slade’s Garage, Penn, who is also responsible for Heal’s 1910 F.I.A.T., which he drove so effectively at Prescott and Brighton during 1947. The Sunbeam is now thoroughly reliable and serviceable, but no attempt has so far been made to regain its original performance. The axle ratio now fitted is 3.69 (the low ratio of 4 to 1 was used for the T.T.) giving overall ratios of 3.69, 4.325, 6.4 and 9.6 to 1. While, at the permitted maximum of 4,000 r.p.m. the road speeds are 104, 89, 70 and 42 m.p.h. This indicates that at 104 m.p.h. there would still be power in hand. Although, with T.T. wings and a mechanic, the frontal area might well be as much as 131 sq. feet, which would almost exactly absorb 108 b.h.p. at 104 m.p.h.
In its present tune, and on pool petrol, the maximum on either 3rd or top seems to be about 80 m.p.h., which suggests that the power output may be down to as little as 30 b.b.p. Indeed, probably due to rather tired sparking plugs and valve springs, it is difficult to push the engine above 3,700 r.p.m., even in bottom. The power-curve, as will be seen from the accompanying graph, is remarkably straight-line for the period and as a natural consequence, the b.m.e.p. is almost constant. from 2,000 r.p.m. upwards. In this respect, the Sunbeam differs from the 1921 G.P. Ballot which had a more flat-topped power curve, peaking at 3,800 r.p.m., as against the 3,890 r.p.m. of the Sunbeam. As a result, the Ballot, which was also lighter than the Sunbeam, was able to pull both higher and more widely spaced ratios, namely, 3, 4.2, 5.4 and 7 to 1. Since 4,000 r.p.m. on the Sunbeam represents a piston speed of only 3,000 feet per minute, the engine could be driven practically without reserve, and even the classical touring speed of 2,500 f.p.m. is as much as 3,350 r.p.m., being equal to 88 m.p.h. on the present gearing.
The contemporary T.T. Vauxhall, designed by Ricardo, was far more adventurous in the matter of piston speed and power output, developing 129 b.h.p. at nearly 4,000 feet per minute, but, in usual Vauxhall fashion, it had not been finished in time to show its potential form in the T.T. race.
The firing order adopted by Henri and Coatalen amounts to two four-cylinder engines joined end to end, each having a firing order of 1, 3, 4, 2. The back set of four has, as its “No. 1” what is, in fact, No. 8 of the engine, so that the two firing orders start at opposite ends of the engine and work in towards the middle. The arrangement is particularly favourable to gas-distribution from two carburetters, but is subject to an unbalanced couple, which is said to produce excessive crankshaft vibration. Tests conducted by Sunbeams showed that maximum vibration occurred at 1,750 r.p.m., which is, in any ease, outside the speed range usually employed. Moreover, it is quite indiscernible when driving the car. It may, however, be that the present nickel-chrome crankshaft is of later manufacture than 1922. The good mixture distribution is clearly evidenced by the car’s ability to run smoothly at 800 r.p.m. in top gear.
This car, like nearly all racing cars of the early twenties, has always possessed a tremendous fascination for me, so that I was immensely excited when Heal invited me to quite an extensive run on the last day of basic motoring. The party was completed by C.R. Abbott of 1904 Mercedes fame, and we took it in turns to drive his Lancia Augusta which came as tender.
On this occasion, the de-tuned engine may have been something of a safeguard, since we soon found ourselves in quite a dense fog, and the temptation to discover terminal velocities, had they been available, might have been fraught with some peril. As it was, 75-80 m.p.h. seemed quite fast enough for safety.
The driver is protected by a metal cowl, which is so effective that it is hardly necessary even to wear goggles. The passenger, however, enjoys no such protection, and on a cold or wet day his lot is not a happy one.
From either seat the long narrow bonnet stretching out ahead, and the pattering wheels and axles afford a kind of thrill peculiar to this type and era of car. while the driver gains a feeling of immediate confidence from the light, sensitive, high-geared and extremely accurate steering (incidentally, the steering wheel is now of rigid pattern, but in G.P. form they were made with laminated, spring-steel spokes). The pedals are conveniently placed (accelerator in the middle) and the brake and gear levers are outside. The brake comes comfortably to hand, but the gear lever seems unnecessarily far forward (uncomfortably placed gear levers seem a speciality of Sunbeam racing cars).
Getting away from rest requires careful use of the clutch, and until the gearbox oil is warmed up, gear changing is excessively tricky. Once warm, a slight pause is required between bottom and second, and subsequent upward changes are made as quickly as the lever can be moved. Indeed, the movement across the gate being considerable, it is quite an effort to get from 2nd to 3rd before the revs, have died. Double-declutching assists downward changes, though, the ratios being so close, and the engine being innocent of a flywheel, it is difficult not to over-rev. With a little practice a single, semi-declutch would probably produce better results. With the engine so much below par, acceleration times become of little interest. while the usual tests from 0-60 m.p.h. are hardly representative, since it takes so long before the engine bites in bottom. Even so, 0-60 was accomplished in 19 sec., which is good enough for most small modern sports cars 45-60 m.p.h. in 2nd, covering the most effective part of the range, from 2.500-3.500 r.p.m., occupied 8 sec. Poor as these figures are, they do represent what an extremely useful performance would be available with a full, properly fed stable.
The general engine characteristics are quite extraordinarily Henri-like. This is the fourth Henri or Henri-inspired car I have had the privilege of driving (1914 T.T. Humber, and 1919 and 1921 Ballots) and all of them hunt about and rattle prodigiously below 2,500 r.p.m., at which speed the whole engine seems to gather itself together ; the engine note changes ; the threshing of the timing gears merges into a single note, and the car starts to accelerate in an entirely different manner. As also seems true of most of this school of design, the engine appears to be remarkably insensitive to the ignition control. The unenthusiastic performance below 2,500 r.p.m. is, I feel reasonably sure, largely attributable to the immense chokes and barrel-type throttles. I should much like to see how one of these engines would perform on two or four S.U.s. If the engine was as unresponsive at low speeds in the T.T. as it is today, it is difficult to imagine how the cars dragged themselves up the first part of the mountain, coming out of Ramsay.
The engine runs extremely cool, and on the occasion of our run more than half the radiator was blanked off. An amusing fitting is an overflow pipe from the radiator, emerging through the cockpit just by the driver’s elbow. An emission of steam thus enabled him to determine when the kettle was boiling. Despite the very homely set of sparking plugs installed, there was no tendency at any time to oil up—a very different state of affairs to the Ballots where, unless M. Henri’s precarious system of lubrication was adjusted to a nicety, plugs oiled with distressing frequency. On the Sunbeam it is, however, possible momentarily to rich up a plug at low engine speeds. The car provides a very comfortable ride and the springs, Hartford-damped, are long for a racing car, especially those at the front. To lock the front brakes, rock the car to and fro, and watch what happens to the springs and axle, gives no confidence at all in the brakes, and on the 1921 Ballot they were, in fact, notoriously treacherous. But, in practice, the Sunbeam brakes are wonderfully effective, while the ability to balance the pressure on front and rear shoes enables the driver to obtain the best braking for any given set of conditions. Despite the powerful retardation and the fact that (apart from the plainly visible effect) the driver can feel the axle turning somersaults through the reaction of the hand lever, there is no evil effect apparent on the steering of the car. The H-section axle, incidentally, was described, in its day, as being exceptionally strong in order to withstand braking torque, but to us, today, it looks painfully flimsy and it does in fact, twist to a noticeable extent.
The cornering characteristics are extremely modern in style and have none of the pronounced oversteer so fashionable in cars of the vintage era. This is no doubt due to the fact that there is so little disparity in spring-rate between front and rear and that the spring-centres are so nearly similar at each end. This gives an almost equal roll-resistance so that the car can be taken round corners in a steady four-wheel slide, with no tendency at either end to a premature breakaway. The limiting factor is provided by the chassis which, being almost entirely unbraced from end to end (the engine and gearbox are mounted on a subframe), and narrow as well, is excessively flexible. So that, if a bump of any severity were to be encountered when the front wheels were sliding, it would be all too easy to lose them completely. But on a smooth road, it would be hard to match the Sunbeam’s cornering powers, either with independent suspension or without.
Taken as a whole, it is as pleasurable and satisfying a car to drive as I know, and with full power available it would be as fast from place to place as most.
Anthony recently took the car to call on Sir Algernon Guinness who was extremely pleased to see it after so many years, in such handsome condition. The accompanying photograph shows Sir Algernon sitting in the car and also gives a very good notion of what a tremendously handsome machine it is. Perhaps the most inspiring view of all is from the back, where the narrow body and spring-centres, and the huge polished bolster tank recall to the full a very grand age of motor-racing.
I am tremendously indebted to Anthony Heal for providing much of the historical data contained in this article, but far more for allowing me to drive his very beautiful motor car.