AIR: Parry Thomas's Aero-Engine

The Welsh engineer/racing motorist will, I hope, always be remembered in our world as the greatest of the Brooklands’ drivers in the vintage years. and creator of the hastily modified Liberty aero-engined Thomas Special “Babs”, that ultimately killed him, after it had gained him the Land Speed Record at 171.02 mph in 1926.

Thomas’s career previous to his devotion to motor racing may be less well-known, apart from the fact that he was the designer of the memorable but short-lived Leyland Eight luxury car, which he saw as challenging the Rolls-Royce as the best car in the world and which earned the sobriquet of “The Lion of Olympia” when it made its debut at the 1920 London Motor Show. It was hardly Parry Thomas’s fault that this very advanced car failed to attain the success which many though it deserved. The Leyland Motor Company abandoned car construction before the design had been properly developed. But the basic concept, with remarkably little alteration, formed the basis of these Leyland Thomas racing cars which scored more Brooklands’ successes for Thomas between 1922 and 1926 than were achieved by any other driver, as well as setting a lap record of 129.36 mph and doing well in other events.

All this has been well recorded but Thomas’s aero-engine less so. In fact, in the four specialist books I have consulted, there is no reference to it; justifiably it does not feature in Airlife’s recent comprehensive book on British aero-engines and the aeroplanes in which they were used, because the Leyland engine never made an airframe. By the time it was completed the war was virtually at an end. But in 1929, in an Air Ministry review, it was listed second of the aero-engines of the previous ten years…

Parry Thomas was born in Wrexham, the son of the Rev J W Thomas, who three years later advanced from Curate to Vicar. John Godfrey Parry Thomas, who had a brother and three sisters, went to London’s City & Guilds Engineering College, where he studied electrical engineering, against his parents’ hope that he would go into the Army or the Church. He then spent some time on research with Prof Ayrton, before apprenticeship to Siemens Bros and later Clayton & Shuttleworth. In 1907 Thomas set up his own business, developing electrical transmissions for road and other vehicle, with offices in Kensington and later at Spring Gardens, London. Needing workshop facilities, he hired a bay at Leyland’s works. The Thomas transmission was successful, but out-dated by 1914. During the war Thomas’s abilities resulted in advisory work for the Government on aeroengines and tank design. In 1917 he was appointed Leyland’s Chief Engineer.

As the war had not ended, Thomas’s first task was to continue development of the Ferguson 18-cylinder aero-engine, but he was already thinking in terms of the Leyland Eight car, assisted by Reid Railton. It is no surprise that, with his individual outlook on engineering matters, Thomas did not proceed with the aero-engine presented to him but instead set to work on a design of his own. It was an X-formation radial with eight cylinders arranged in four paired banks, of 6 x 4½ in bore and stroke, designed to develop 300 hp at 2500 rpm, at 10,000 feet, at a propeller speed of 2500 rpm, on a c r of 5.8 to 1.

The advanced features included four valves per cylinder, operated through rockers by overhead camshaft above each head, driven by vertical shafts and bevels from the crankshaft, and ingenious lubrication and coolant systems.

The aluminium cylinders were cast in one with the crankcase, nickel chrome liners being inserted into them after they had been machined, annealed, and heated to 180deg C. They could be bored in one operation, as they were in blocks of two. The cylinder heads were detachable (unusual if not unique in aero-engine practice? (someone will tell me!), each pair being held down by eight bolts, the head joint a leatheroid gasket. The oh-valve gear was enclosed by oil-tight covers. Thomas had been clever in taking coolant from the centre of the heads to the space around the exhaust valves and guides. Coolant was circulated by a pump driven directly from the crankshaft at some 40 gallons a minute and another very original aspect was that water was drawn through the drilled crankshaft at eight gallons a minute, from a division in the pump, reducing bearing temperature.

From the pump the coolant went to the heads in the lower banks of cylinders and around the exhaust valves, in four parallel flows, feeaing the eight valves, coring within the heads ensuring that the water had to circulate round the vicinity of the valve seats at approximately eight feet per second. It then flowed to the upper cylinder banks via passages cored round the carburettor induction pipes that were common to the reduction-gear casting, to circulate as in the lower banks, but in the reverse direction, leaving from outlets above the exhaust ports, to the radiator. Dry-sump lubrication was used, with an oil tank of about a gallon capacity below the bottom cylinder banks. Oil from the crankcase was returned to this tank by a 4in-dia pipe. The main supply would be from a 6-to-8-gallon tank in the aeroplane. A clever float system ensured that if oil level fell by a small amount in the supply tank the pump would be changed over to feed from the fuselage tank, but if that tank were to be damaged, the engine would run for about an hour on the sump contents.

The oil pump was driven by three spur gears that would also drive a distributor if coil ignition was used. It supplied the engine bearings, gudgeon-pins and valve gear (the latter by 5/32in filtered copper pipes), another outlet feeding the propeller reduction gear. The latter needed enough oil to carry away heat representing 1% to 3% of the total power but the water-cooled crankshaft reduced the quantity of lubricant the main bearings required.

There was control of the amount of oil from the gudgeon-pins being flung onto the undersides of the pistons. The crankshaft had two 360-deg throws and a master con-rod served the paired cylinder banks, and slipper-type pistons were used. It is of interest that Thomas used leaf valve springs, as on all his later Leyland engines. They provided for a valve lift to 16mm as fast as 10,000 feet/sec, for a spring strength of 65Ib compressed, 16lb on the valve seat, because as one valve was lifted tension on the opposite valve was increased.

Two carburettors at the front of the engine fed the cylinders through water-jacketted U-induction pipes formed within the reduction-gear casing. Air was drawn from the crankcase, thus both reducing piston temperature and slightly warming the air to the carburettors. Ignition could be by magneto and coil, twin coils, or twin magnetos, with a trembler coil and handrotated dynamo for starting. Thomas designed his own planetary reduction gearing, with cooling from the crankshaft coolant. There was provision for driving a gun-interrupter from the magneto gears and a “speedometer” at ¼-engine speed. The engine, less battery, weighed 500 to 525Ib and was estimed to give a BMEP of 130lb/sq in at sea-level and a petrol thirst of 0.55 pints/hp-hour and an oil consumption of 1½-gallons/hour.

The specification was not sent to the Air Board in the Strand until early in December 1917, so with the Armistice just over eleven months away, no wonder Leyland’s engine did not see war service. Just before Christmas 1917 Thomas, modestly not adding “Chief Engineer” after his signature, was writing to the Air Board about criticisms they had made. The experts there wanted the inlet and exhaust ports interchanged, presumably because, as the exhaust manifolds were on the insides of the cylinder heads, this might have meant the exhaust pipes having to extend upwards and downwards in an aeroplane, rather like the two exhausts stacks on a BE2c etc. Thomas replied that as the exhaust ports were water-cooled, the inlets not, it would be very inconvenient to have to re-arrange them, especially as the water flow was at the top of the cylinders to avoid air or steam traps in any part of the heads.

Another thing the Board disliked was the proximity of the water inlets to the exhaust ports; Thomas agreed to couple these with external pipes, with a central off-take to the radiator. He pointed out that to core the water exit in the cylinder would almost certainly invite steam traps when an aeroplane was manouvered. The boffins also wanted a drain-cock at the lowest part of the water pump, to which Thomas replied that in draining the lower heads the pump would be drained also, by placing its cock at the lowest point, so that every time the heads were drained the pump would need priming! The boffins also disliked the drawing of air into the crankcase to cool the pistons; Parry Thomas parried this by saying his system could be deleted but that he would carry out a test at the first opportunity. Prototypes usually need development, Lt Evans, an Inspector who had visited the Leyland works, was told of problems with the leaf valve springs, stressed by “the enormous acceleration the cam profile gives.” However, this was cured, as a 20-hour run at 2500 rpm driving a camshaft by electric motor showed, as a 3-million alteration in stress. After which Thomas hoped there would be no more trouble with the single-cylinder test rig.

He hoped that early in 1918 he would be able to report on using aluminium piston rings. The Air Board was still not satisfied. It did not like Thomas’s clever lubrication system (so he offered to use any system they favoured) and thought mixture might condense in the inlet U-pipes. To which they got a terse reminder that the U-pipe gas velocity was about 70 to 80ft/sec, that in the pipes to each cylinder 150 ft/sec, so Thomas could not forsee condensation or bad distribution, especially as the main pipe would always be reasonably warm. By late January Lt Evans had not informed Thomas about the changes he was considering for the induction system but the designer told the Technical Department of the Air Board that although the reduction gear had yet to be run at full load, the tests so far were “very promising”. showing 99 to 99½% efficiency when transmitting 160 hp. Oil had been fed through the outer bearings to the bevel planetary gear bearings but this was insufficient to prevent overheating, so steps had been taken to increase oil flow. At this stage there was a delay, caused by a move to a new experimental shop, perhaps for testing also the Leyland Eight car engine? The test rig engine had been troublesome, five or six inlet valves fracturing, so new, better streamlined ones were made and the Air Board consulted for its opinion. That was on a 6.3:1 c r, when 125Ib/sq in BMEP was recorded, against 126 at 2200 rpm. Interestingly, no advantage was seen with a valve lift of 1½in against 5/8in; different valve openings were being tried.

By this late stage no running X-engine had been completed, but the test rig showed a little carbon on the piston after a 1/2-hour run at 38 to 45 bhp, so Thomas conceded that perhaps induced draught past the pistons of the full-size engine might be abandoned; but he would decide, after the first radial engine had been run.

It was all a long time ago; but I dare say anyone who has built engines to a Government specification will be familiar with the process. Leyland Motors (1914) Ltd, was by then a Controlled Establishment under the Munitions of War Act, 1915. But for Thomas, time was running out. By April 1918 still only the test rig was running and that broke an old-type piston after a five-minute burst at 3500 to 3700 rpm, when 50 to 53 hp was seen; before the bang the little engine did four hours at 2500 to 2600 rpm, producing about 37 hp, equal to the X-engine’s specified output at 10,000ft. Valve trouble had been cured by using 1¾in-dia valves in place of 2¼in; the size intended had been inlet 57mm, exhaust 45mm, with a 3mm wide valve seat. (The smaller valves were as good as the larger ones at 2500 rpm but there was some power loss at 3500 to 4000 rpm). All materials were to AB specification (ARI ), the inlet valves of 3% nickel steel, the exhausts of n c steel. At 80deg C water heat there was no pre-ignition and the exhaust valves were just dull red (c r still 6.3 to 1). After a 10-hour run a tooth broke in the reduction gearing but a new gear was in hand and a 50-hour trouble-free run anticipated.

In April 1918 the full-size engine was “well on its way”, most patterns completed, parts well in hand, a test visualised in ten weeks. But it was too late; Comdr Barrington had failed to call in, perhaps realising this. But Thomas was optimistic, asking if military call-up of his best machinists (Ken Taylor almost certainly one of them) could be cancelled, for those in the “special aero-engine department”.

IGPT’s hastily assembled aero-engine was still-born, the prototype seizing on test in August 1918. But think what it was to lead to!