The twin-cam M.G. Midget

SIR CLIVE EDWARDS, in his entertaining article in November, referred at length to his R-type Midget with twin o.h.c.

This remarkable cylinder head was the product of Laurence Pomeroy and Michael McEvoy when they were handling Zoller blowers in this country. Pomeroy recently told the story of it, and we thought that it might be interesting to readers.

At the end of 1935 the Q-type M.G. (a Zoller-blown 750 and orthodox chassis) was replaced by the R-type, which had a backbone chassis and four-wheel independent torsion-bar suspension. The engine was the same as before and the whole outfit weighed some 13 cwt.

It was known that Austins were preparing their twin o.h.c. racers for the 1936 season, and that these were likely to weigh only 9 cwt. as well as having a much smaller frontal area than the M.G.

The Austin developed 100 b.h.p. at 8,000 r.p.m., assisted by an 18 lb. Roots-type boost, and while the M.G. gave 120 b.h.p. at 7,000 r.p.m., it cracked its cylinder heads at this output, which anyway wasn’t enough to deal with the Austins. Other troubles of the M.G. were that the forward position of the blower led to freezing of the carburetter, while the long induction pipe produced heavy condensation on the over-run, with consequent plug oiling.

The oil pump, driven from the end of the blower, was subject to failure, because blower pressure forced itself up the pump spindle and blew all the oil out of the pump.

The McEvoy firm therefore approached the Evans brothers in November, 1935, suggesting that they should construct three special cylinder heads, which would furnish their M.G.s with a whole team of extra horses with specially reliable characters.

The fee for this ministration was to be £450 (£150 for each head) and the cars were to be ready to run in the Empire Trophy 1936; this was agreed to, one head being intended for the Evans’s car, one for Briault’s, and one for Connell’s.

Details of the projected design were as follows:—

The forward mounted Zoller was to be replaced by another, mounted on the near side of the engine, and giving a 30 lb. boost, or capable of being increased to 45 lbs. if desired. The old blowers had an unfavourable internal air-flow, the air being forced round one right-angle as it entered, and another as it left the wind-whisk. This was to be avoided by using a different casing, bolted direct on to the inlet pipe at its outlet and inspirating through a large single downdraught S.U. carburetter. The blower was driven through a flexible coupling, and was lubricated by a pump attached to the end of a camshaft.

As time was so short, the design was kept as conservative as possible consistent with the aim in view. The head was, accordingly, made in cast iron, so as to avoid valve seat and other distortion trouble. The camshafts were driven by spur gears and two idler wheels, meshing with a single gear, driven by the same bevels as are ordinarily found at the top of the M.G. shaft. The camshafts were pressure lubricated and chromium plated, having three plain R.R. 53 bearings. At the front of the gear case was a chain case, containing a triple roller chain which drove the blower at 0.7 engine speed.

The valves were at 90° to each other, with a more extreme timing than the standard M.G. to the extent of a 70° overlap, with exhaust valves opening 80° before bottom dead centre. The exhaust valves were made 70 per cent. the area of the inlets, but in order to use the most advantageous size of Brooklands silencer the actual cylinder head ports were much enlarged. All valves were K. 965 and had long stems, to provide good guide cooling, even at the expense of somewhat greater weight than might otherwise have been the case, and very strong triple valve springs were employed. Hardened steel fingers of Bugatti pattern were specified, and the clearance adjustment was also by shims of Bugatti type.

Bottom seating plugs, in the centre of the head, had very masked openings, in order to secure optimum cooling.

The cooling arrangements represented perhaps the most unorthodox part of the design. To begin with, the pump was of specially large pattern, seated on a bracket at the back of, and driven through, the supercharger. From it, water was fed separately into the head and the block. The supply to the head was disgorged straight on to the exhaust valve seats. From there it swirled past the plug bosses, and was prevented from getting on to the inlet side of the engine by an internal baffle, which deflected the water, via four off-take pipes, back to the radiator. This was the sole circuit of this feed, as there was no water passage between the head and block on the exhaust side. A second supply from the pump entered the block and passed through a thermos-syphon to the head, on the inlet side, and thence back to the radiator. The idea was that the restricted flow of water on the inlet side, and the profuse, direct, and turbulent flow on the exhaust side would produce a very even temperature throughout the head, and so avoid distortion and cracked seatings, etc. A provisional patent was taken out for this arrangement. Further to improve the cooling, an air-scoop was fitted to the exhaust cambox and both camboxes had breathers.

Special Aerolite pistons with 8 mm. thick crowns gave an initial compression ratio of 6 to 1.

Working seventy-five hours a week the actual designing took until Christmas, leaving less than three months for casting, machining and tuning.

As McEvoy’s shop equipment was not capable of carrying out all the machining it was partly done in the Singer tool room, which was fairly slack at the time, and actual assembly was started at the end of February. It was then decided that it would be best to push on with one head and tune it, so that any alterations found necessary by experiment could be built direct into the other two before assembly. The Evan’s car was to have the first head.

It was now found that the water spaces had come out wrong owing to a discrepancy in the casting, and these had to be machined back to the intended shape. The inlet pipe and ports, also, did not match up, and the rocker fingers had only been oil-hardened instead of case-hardened—though this was not discovered till later.

On the Monday before the race the first engine was erected on the Evans’s test-bed, and all was set for starting up at 9 p.m. that evening.

But would it start?

Frankly, no!

At the suggestion of Pomeroy’s versatile brain they even wound a rope round the blower-drive shaft to which three men attached themselves and fiercely pulled, as it were a child starting a top; but they didn’t start that engine.

The electric starter was then brought up. Twelve volts—no good. Twenty-four volts—no good. Thirty-six volts excellent fireworks but no effect on the engine, which lay glowering on its testbed, determined to ward off all attacks upon its neutrality.

Men now stood about in corners, muttering darkly and nibbling their fingernails, while Pomeroy all unconsciously rendered several moving passages from Wagner with that degree of fervour to which only a soul in anguish can attain.

Finally the insides of the cylinders were dried out, and with a further blazing of thirty-six volts the engine at last gave way and started somewhat after 2.30 a.m.

The exhausted combatants then went to bed, and on Tuesday the engine was run in, and by the evening it was ready to test. It was then found that chronic pre-ignition set in at 6,500-7,000 r.p.m., though all kinds of plugs were tried without avail. Finally, some very spent Bosch racing plugs, left over from the previous season, were espied on a shelf, and with them the trouble more or less disappeared.

Then, although the works of the blower were the same as had given 30 lbs. with the old head, nothing better than 21 lbs. could now be clocked. It was then discovered that the new porting gave so little back pressure, as compared with the old type head, that this accounted for the drop in blower pressure; but against that the engine already gave 124 b.h.p. on this moderate boost.

It was also found that the optimum ignition advance was 55º. The other engines were now put together, and the cars were shipped to Donington for practice on Friday, which was a desperately cold day, producing very erratic carburation. In the race on Saturday the handicapping was most peculiar. The 750 class were invited to tour at 67.49 m.p.h., and the unlimiteds only had to encompass 69.51 m.p.h. In point of fact, Seaman won at 66.3— something wrong somewhere. Evans’s car ran fairly well for a first time out, and was still motoring at the end of the race, by which time his was the only 750 still operating, having completed 91 laps. Connell did 40 laps before retiring with a broken oil-pipe and Briault’s car, which was running fitfully, retired after 35 laps owing to carburation. At 12 laps “Bira” had led at 67.05 m.p.h., while Evans lay second at 66.95 m.p.h.—only 0.1 m.p.h. slower. It was now seen that even at this low temperature the engines would tick over at 1,200 r.p.m. and pull away cleanly on the hardest plugs without oiling. After the race the cars were returned to their owners and Pomeroy went to prepare Evans’s for the International Trophy at Brooklands.

The valve operating fingers had now worn considerably, but as it was still not known that they were only oil-hardened this wear was attributed to the very strong valve springs. They were accordingly chromium-plated, and this seemed to work quite well for the time being. The pre-ignition was finally suppressed by the use of a very rich mixture, and the engine put out a sustained 130 b.h.p.

The car went well in practice, but Briault’s was still in trouble with its plugs, and persisted in throwing out all its water, despite a spring-loaded overflow pipe, and to overcome this the entry to the pump was restricted, so as to curb its over-exuberance.

At 5 p.m. on Friday Evans missed a gear change coming off the banking, the revs flew up and a valve was bent. Scrutineering was to take place at 11 a.m. next day.

Pomeroy therefore set out to Derby on Doreen Evans’s trials car, and broke into the works store at midnight, found a valve, and left again at 12.30 a.m. 3 a.m. found him back at Wandsworth, having covered the 145 miles in 2½ hours. including an average of 57 m.p.h. from Derby to the end of the Barnet By-Pass.

The valve was at once put in, and the plug holes bored out to revert to normal top-seating plugs, it having been ascertained that the pre-ignition was due to the exposed, unpositioned copper washers at the foot of a bottom-seating plug.

Wilkinson reassembled the engines, and with the help of an hour’s time-extension the car was brought intact to the scrutineer—no mean feat of speedy work and endurance.

Connell’s car turned up very late, having been prepared by his own mechanics. It was wheeled to the starting line, but it refused to start, so they wheeled it away again.

Doreen Evans had a fine scrap with the Austins, frequently leading them, but after re-fuelling, at 52 laps, her car caught fire on the Railway Straight and could not go on racing. Briault retired after 31 laps with chronic misfiring.

Evans’s car was then prepared at greater leisure for the Nuffield Trophy Race at Donington, and now that all the snags had been overcome it had the race all sewn up by a good 10 m.p.h. on its handicap. But alas! the coil ignition burnt out on the first lap and the car again had to retire.

Briault now decided to give up 750 c.c. racing and sent his car back to the Evans’s garage to be put together properly with a view to selling. When his engine was dismantled it was found that the unprecedented turbulence brought about by this type of water circulation had caused pockets of local boiling and distortion and the head had cracked. This was rectified, and a recurrence of the trouble prevented by slightly drilling the baffle in the head, so as to allow some degree of communication between the inlet and exhaust sides. The plug holes were bored out, as on the Evans’s car, and these two engines finally gave 141 b.h.p. at. 7,300 r.p.m. on a 25 lb. boost. Even this output would not have been enough to cope with the Austins, which finally developed 120 b.h.p. but there is little doubt that the 45 lb. boost originally intended would have yielded a reliable 160 b.h.p. or more (213 per litre, and an m.e.p of roughly 390!) and that this would have answered the purpose.

The Evans brothers and Connell now also gave up 750 c.c. racing, and the Evans’s car was sold to Sir Clive Edwards; the Baines brothers bought Briault’s; and Connell’s machine went to some destination in the Midlands.

Pomeroy and McEvoy then went to Germany to pursue their supercharger researches and largely lost touch with the cars, but several interesting reflections on  this may be made.

(1) Even at 141 b.h.p. the output was 10 h.p. more than Kohlrausch’s world record-breaking car, and this was far below the potential maximum which, as stated above, is 160 b.h.p. at least.

(2) Equally, even 141 b.h.p. is believed to represent as high a b.h.p. per litre as has ever been obtained from a petrol motor.

(3) At the price of £450 McEvoy and Pomeroy neither gained nor lost money, so far as the designing and manufacture were concerned.

(4) This engine was used as the basic design for the highly successful Parnell-Cuddon Fletcher M.G., which, however, has an aluminium head and lighter valves.

(5) How thorny is the path of the innovator, be he never so conservative and canny in his design!

(6) Finally, it can be seen that Edwards has a power unit which, despite its age, is still probably the most advanced 750 c.c. engine in existence and which, with a lighter chassis and proper streamlining, is fully capable of taking the world’s class record.