The appearance of the Curtiss D-12, and its importation under license to Britain by Fairey prompted the Air Ministry to ask Rolls-Royce to develop a new aero engine for fighter aircraft, the ministry felt that there were too many engine manufacturers and did not support or encourage the Fairey company to continue. Rolls-Royce developed two distinct types of engine to meet the requirement, the F engine, which became the Kestrel, and the Eagle XVI.

Designed by Henry Royce, the Eagle XVI was a completely new design and unrelated to the earlier Eagle. The engine featured four banks of fore and aft staggered cylinder blocks with the banks arranged at 90 degrees to each other to form a perfect ‘X’. The valve gear was operated by a single overhead camshaft and used four valves per cylinder. The crankcases and cylinder blocks were cast from aluminium alloy. Normal engine bearers could not be used so the unit was held in the test stand by two conical mounts that attached to the crankcase.

Initial test running was carried out in 1925 without a supercharger fitted and an improvised carburettor and induction system. The engine did not run well with this arrangement due to poor fuel distribution, however when a supercharger and matched carburettor were fitted (borrowed from the Kestrel development unit) the engine ran and performed well, producing 500 hp (373 kW) on the dynamometer. Despite this the engine was not received well by the aircraft industry where it was felt that the unusual layout would block the pilot’s forward view in a typical fighter aircraft installation. The project was dropped and development efforts then concentrated on the Kestrel.

Only one was built and the Eagle XVI was one of the few Rolls-Royce projects that did not fly, it is of note that the ‘X’ engine layout was reused in the later Vulture and Exe designs.

Eagle XVI
Type: 16-cylinder liquid-cooled 90° X aircraft piston engine
Bore: 4.5 in (114 mm)
Stroke: 4.75 in (121 mm)
Displacement: 1,208 in³ (19.8 L)
Valvetrain: Overhead camshaft, four valves per cylinder
Supercharger: Single speed, single sided impeller
Fuel system: Single carburettor
Cooling system: Liquid-cooled
Power output: 500 hp (373 kW)
Specific power: 0.41 hp/in³ (18.8 kW/L)

The first Rolls-Royce engine to be named after a river, as are the modern turbojets, was this experimental unit, which was notable in having sleeve valves and pressure-air cooling. Its 24 cylinders were arranged in X form, in four banks of six, and the capacity was 22 litres. Rated output was 1,200 h.p. at 4,000 ft. The Exe was test flown in a Fairey Battle during 1938, but the project was shelved when it became clear that the company’s main productive capacity must be devoted to the Merlin. Another experimental unit related to the Exe was the Pennine, and it may be noted here that in the years preceding the war Rolls-Royce were experimenting with two-stroke as well as four-stroke engines, and continued the compression-ignition experiments started with the Condor. The H layout was also examined.

The Eagle was designed by Henry Royce in collaboration with his chief assistant, A. G. Elliott. It set the pattern for the smaller Falcon and the larger Condor, and in improved forms continued in service from 1917 until the 1930s.

Though designed for only 200 h.p., the first example was run in October 1915 (less than six months after work on the drawings had been started at Derby) at a brake horsepower of 225.

By March 1916 the new engine was delivering 266 b.h.p.; by July 284 h.p.; and by December 322 h.p. By September 1917 the figure had risen to 350 h.p.; in Feburary 1918 it had reached 360 h.p. An official Rolls-Royce instruction book dated December 1917 ascribes the following outputs (at normal r.p.m.) to the various series: Eagle 1, 225 b.h.p.; Eagle If, 266; Eagle Ill, 284; Eagle IV, 284; Eagle V, 322; Eagle VI, 322; Eagle VII, 322; and Eagle VIII, 350. These variants differed in the type and number of carburettors, in jet size, in the pattern of inlet pipes and pistons, and other details. By far the most famous among them was the Eagle VIII, which powered such fine aircraft as the Phomix Cork, Felixstowe F.3 and F.5, the D.H.4, 9A, 9B, 10B and IOC, the Fairey IIIC and IIII), the Handley Page 0/400 and V/1500, the Short Shirl, and the Vickers Vimy and Vernon. Other military installations were made in the Wight seaplane, F.E.2D, REA, R.E.7, Porte Baby, F.2, Felixstowe Fury (F.4), D.H.10, Fairey F.16 and F.17 Campania, Handley Page 0/100 and Vickers Vim. Civil applications included the D.H.16 and Handley Page W.8b.

Irrespective of series number, the Eagle was a twelve-cylinder, water-cooled, 60-deg vee of 20.32 litres capacity. Each cylinder, with its water-jacket, formed a separate unit; the barrel, machined from a steel forging, was retained by a base flange and studs. Inlet and exhaust ports were machined from solid forgings, welded-in at opposite sides of the head, and the water-jacket was fabricated from steel pressings, welded in place.
There was one inlet and one exhaust valve for each cylinder, and a single camshaft, carried in a casing along each bank of cylinders, operated both. There were two sparking plugs per cylinder, served, on the earlier engines, by four six-cylinder magnetos and on the later series by two twelve-cylinder magnetos.

A contemporary Rolls-Royce description continues: “The oil consumption should be taken as 1 gallon per hour. . . . The lubrication system is of the type in which one oil pump supplies pressure oil to the main bearings and other parts, while one scavenger pump evacuates the accumulation of oil in the crankcase to the oil tank. . . . The quantity of water carried in the cylinder water-jackets, water pipes, and pump is 3.1 gallons…. The Rolls-Royce carburettors are fitted with a control so that the delivery of petrol may be adjusted from the pilot’s seat. This not only serves as a means of correcting the effects of decreasing atmospheric pressure with increasing altitude, but can also be used to obtain extremely economical running, and also to obtain a rich mixture for starting…. All controls, e.g., throttle, altitude, and magneto, are brought to one common location on the engine, to facilitate connection to the plane…. The carburettor throttles are fitted with springs, which, in the event of a breakage of the control, are intended to open the throttles to the full extent.”

A special feature of the Eagle was the epicyclic reduction gear, for which the advantage was claimed that it imposed far less strain on the crankcase than would spur gearing. The gear was of fixed sunwheel type, in which an annulus, or internally toothed gearwheel, on the crankshaft meshed with three small spur gearwheels on the extremities of three arms formed integrally with the airscrew shaft. The three small gears were connected to three pinions which meshed with the fixed sunwheel, and the airscrew shaft was bolted to the three-armed spider which carried the planet pinion cage. An Oldharn coupling provided the anchorage of the sunwheel to the gear casing and allowed the gears to be perfectly aligned.

Late in 1922 Flight reported: “A recent model introduced by Rolls-Royce, Ltd., is the Eagle IX, which is generally similar to the famous Eagle VIII but differs from that type in certain details. The new model has been designed to be equally suitable for peace and war purposes, and to that end the following improvements have been effected: Two carburettors are fitted instead of four, so that tuning is considerably facilitated. The placing of the carburettors low down allows of using directgravity feed; the float feeds have been redesigned. The engine will now function satisfactorily with a ‘head’ of only 8in above the centre-line of the crankshaft. In view of the tendency m modern commercial aircraft of employing direct-gravity petrol feed from the main tanks in order to avoid pumps, piping, etc., this should be a great advantage, and should help to ensure for the Eagle IX as great popularity as that enjoyed by the previous model.”

In the Eagle XV the epicyclic gear was further modified by the insertion of an additional friction clutch and gearwheel in the train, making the existing friction anchorage into a manually controllable clutch and converting the gear from a single-speed to a two-speed type. In the absence of a variable-pitch airscrew this arrangement gave a considerable advantage.

The following data for the Eagle VIII are taken from Rolls-Royce Aero Engine Instruction Book, Eagle Series I – VIII and Falcon Series I, II and III. Air Board Technical Department Book No. 244, December 1917: Dry weight, 847 lb; normal r.p.m., 1,800; max. permissible r.p.m. (5 min only), 2,000; gear ratio, 0.60; b.h.p. at normal r.p.m., 350; petrol consumption in gal/hr, at sea level, normal r.p.m. and normal b.h.p., 23; carburettors, four R.R. C.H. (Rolls-Royce, Claudel Hobson), 42 mm; magnetos, four Watford; pistons, high compression, four rings (one at bottom).

In 1925 a little-known Eagle development-the 16-cylinder, 500 h.p. Mk XVI-made its appearance. It had a stroke of 441n instead of 61in. The Mk XX was a projected scaled-up version, but was never built.

Sir Henry Tizard, Chairman of the Aeronautical Research Committee (ARC), was a proponent of a high-powered “sprint” engine for fighter aircraft and had foreseen the need for such a powerplant as early as 1935 with the threat of German air power looming. It has been suggested that Tizard influenced his personal friend Harry Ricardo to develop what eventually became known as the Rolls-Royce Crecy. The idea was officially discussed for the first time at an engine sub-committee meeting in December 1935.

“The Chairman remarked that if it was the desire of the Air Ministry to develop a type of sprint engine for home defence….there was the question as to how far fuel consumption could be disregarded. Mr Ricardo had raised this point in a recent conversation by enquiring whether a high fuel consumption might not be permissible under certain circumstances, for if so, an investigation of the possibilities of the two-stroke petrol engine appeared to be attractive.”
—Henry Tizard, The Rolls-Royce Crecy

Previous experience gained between 1927 and 1930 using two converted Rolls-Royce Kestrel engines through an Air Ministry contract had proven the worth of further research into a two-stroke sleeve-valved design. Both these engines had initially been converted to diesel sleeve-valved operation with a lower power output than the original design being noted along with increased mechanical failures, although one converted Kestrel was subsequently used successfully by Captain George Eyston in a land-speed record car named Speed of the Wind. The second engine was further converted to petrol injection which then gave a marked power increase over the standard Kestrel.

Single-cylinder development began in 1937 under project engineer Harry Wood using a test unit designed by Ricardo. Although originally conceived as a compression ignition engine, by the time Rolls-Royce started serious development, in conjunction with the Ricardo company, the decision had been taken by the Air Ministry to revert to a more conventional spark-ignition layout, although still retaining fuel injection.

The first complete V12 engine was built in 1941, designed by a team led by Harry Wood with Eddie Gass as the Chief Designer. Bore was 5.1 in (129.5 mm), stroke 6.5 in (165.1 mm), compression ratio 7:1 and weight 1,900 lb (862 kg). The firing angle was 30 degrees BTDC, and 15 lbf/sq.in (100 kPa) supercharger boost was typical. First run on 11 April 1941, in bench-testing it produced 1,400 horsepower (1,000 kW); however, there were problems with vibration and the cooling of the pistons and sleeves. The thrust produced by the exceptionally loud two-stroke exhaust was estimated as being equivalent to a 30% increase in power at the propeller on top of the rated output of the engine. The power of the engine was interesting in its own right, but the additional exhaust thrust at high- speed and altitude could have made it a useful stop gap between engines such as the Rolls-Royce Merlin and anticipated jet engines. Serial numbers were even, Rolls-Royce practice being to have even numbers for clockwise rotating engines when viewed from the front.

The reciprocating sleeve valves were open-ended rather than sealing in a junk head – the open end uncovered the exhaust ports high in the cylinder wall at the bottom of the sleeves’ stroke, leaving the ports cut into the sleeve to handle the incoming charge only. They had a stroke of 30% of the piston travel at 1.950 in (49.5 mm) and operated 15 degrees in advance of the crankshaft. The Crecy sleeve valves were of similar construction but differed in their operation compared to the rotary sleeve valve design that was pioneered by Roy Fedden, and used successfully for the first time in an aircraft engine, the Bristol Perseus, in 1932.

Unlike most two-stroke engines, supercharging was used rather than crankcase compression to force the charge into the cylinder – this also allowed for a conventional oil lubrication system instead of the total-loss type found in many two-stroke engines. Stratified charge was used where the fuel was injected into a bulb-like extension of the combustion chamber where the twin spark plugs ignited the rich mixture. Operable air-fuel ratios of from 15 to 23:1 were available to govern the power produced between maximum and 60%. The rich mixture maintained near the spark plugs reduced detonation allowing higher compression ratios or supercharger boost. Supercharger throttling was used as well to achieve idling. The supercharger throttles were novel vortex types, varying the effective angle of attack of the impeller blades from 60 to 30 degrees. This reduced the power required to drive the supercharger when throttled, and hence fuel consumption at cruising power.

Later testing involved the use of an exhaust turbine which was a half-scale version of that used in the Whittle W.1 turbojet, the first British jet engine to fly. Unlike a conventional turbocharger the turbine was coupled to the engine’s accessory driveshaft and acted as a power recovery device. It was thought that using the turbine would lower fuel consumption allowing the engine to be used in larger transport aircraft. This was confirmed during testing however failures due to severe overheating and drive shaft fractures were experienced.

Test summary
The following summarises the test running programme, hours run, and highlights some of the failures experienced.

Crecy 2

11 April 1941
First run. One-piece cylinder block/head. Testing stopped due to piston failure.
Hours run: 69

October 1942 – December 1942
Three rebuilds during this period, testing stopped after 35 hours due to piston seizure.
Hours run: 67

February 1943 – July 1943
Converted to Mk II configuration (separate cylinder heads), three rebuilds during this period. Air Ministry acceptance test passed.
Hours run: 38

March 1944 – July 1944
Five rebuilds during this period. Equal length injector pipes fitted, modified supercharger drive. Two failures, sleeve valve seizure and supercharger drive failure.
Hours run: 82

August 1944 – November 1944
Successful type test passed (112 hours). Post run inspection revealed cracked big-end bearings, pistons, reduction gear housing and sleeve valve eccentric drive bearing.
Hours run: 150

March 1945 – April 1945
Attempted endurance test, piston failure after 27 hours. Two rebuilds during this period.
Hours run: 49

(Total hours: 461)

Crecy 4

November 1941
No report available.
Hours run: 55

July 1942 – August 1942
Three rebuilds, successful 50-hour test, second 50-hour test abandoned after cylinder block failure due to cracking.
Hours run: 80

September 1942 – October 1942
Two rebuilds. Completed 25-hour test successfully, second test halted after four hours running due to sleeve valve failure.
Hours run: 55

(Total hours: 293)

Crecy 6

July 1943 – February 1944
First engine built as Mk II. Eight rebuilds during this period, failures included supercharger drive failure and sleeve valve eccentric drive bolt fracture.
Hours run: 126

May 1944 – September 1944
Four rebuilds. Supercharger flexible drive failure and sleeve valve seizure.
Hours run: 93

November 1944 – February 1945
Three rebuilds, main bearing failure, piston failure.
Hours run: 128

June 1945 – August 1945
One rebuild, endurance test halted after 95 hours due to sleeve valve drive failure, 40 hours run with a propeller fitted.
Hours run: 132

(Total hours: 481)

Crecy 8

September 1943 – March 1944
Eight rebuilds, endurance test successfully completed.
Hours run: 207

April 1944
Supercharger drive failure.
Hours run: 73

June 1944 – September 1944
Five rebuilds, no failures reported.
Hours run: 32

October 1944 – December 1945
Two rebuilds, piston failure, engine fitted with exhaust turbine.
Hours run: 22

(Total hours: 336)

Crecy 10

August 1944 – February 1945
Six rebuilds, melted inlet manifold after seven hours, sleeve valve seizure after a further four hours. Two injector pump failures.
Hours run: 53

March 1945 – June 1945
One rebuild, piston failure.
Hours run: 30

July 1945 – September 1945
Two rebuilds, exhaust turbine fitted, some running without supercharger. Sleeve valve and supercharger drive failure.
Hours run: 82

(Total hours: 166)

Crecy 12

January 1945 – October 1945
Four rebuilds, exhaust turbine fitted. Turbine failure, piston failure and sleeve valve drive failure.
(Total hours: 67)

The progress of jet engine development overtook that of the Crecy and replaced the need for this engine. As a result work on the project ceased in December 1945 at which point only six complete examples had been built, however an additional eight V-twins were built during the project. Crecy s/n 10 achieved 1,798 horsepower (1,341 kW) on 21 December 1944 which after adjustment for the inclusion of an exhaust turbine would have equated to 2,500 horsepower (1,900 kW). Subsequent single-cylinder tests achieved the equivalent of 5,000 brake horsepower (3,700 kW) for the complete engine. By June 1945 a total of 1,060 hours had been run on the V12 engines with a further 8,600 hours of testing on the V-twins. The fate of the six Crecy engines remains unknown.

If the Crecy had flown it would have done so using a Hawker Henley, L3385 which was delivered to Hucknall for conversion on 28 March 1943. This aircraft remained at Hucknall until 11 September 1945 when it was scrapped without ever having the engine fitted.

Two years prior to the Hawker Henley’s arrival (Summer 1941) a Supermarine Spitfire Mk II, P7674 had been delivered to Hucknall and was fitted with a Crecy mock-up to enable cowling drawings and system details to be designed. It had also been agreed that the first production Spitfire Mk III would be delivered to Hucknall in early 1942 minus its Merlin engine for fitment of an airworthy Crecy; this delivery did not occur however. A Royal Aircraft Establishment report (No. E.3932) of March 1942 estimated the performance of the Spitfire fitted with a Crecy engine and also compared this to a Griffon 61-powered variant of the type. The report stated that the Crecy’s maximum power output would be too much for the Spitfire airframe but that a derated version would have considerable performance gains over the Griffon-powered fighter.

Crecy
Type: 12-cylinder supercharged liquid-cooled 2-stroke aircraft piston engine
Bore: 5.1 in (129.5 mm)
Stroke: 6.5 in (165.1 mm)
Displacement: 1,536 cu.in (26 lt)
Dry weight: 1,900 lb (862 kg)
Valvetrain: Crankshaft-driven reciprocating sleeve valves
Supercharger: Gear-driven centrifugal type supercharger with variable angle of attack of the impeller blades providing up to 24 psi (165 kPa) of boost.
Turbocharger: Three engines fitted with exhaust turbine (50% scale version of Whittle W.1 turbine)
Fuel system: Direct fuel injection, 2 x CAV 6-cylinder pumps
Fuel type: 100 Octane petrol
Oil system: Gear pump
Cooling system: Liquid-cooled
Reduction gear: 0.451:1 (Left-hand tractor)
Power output: 2,729 hp (2,035 kW)
Specific power: 1.77 hp/cu.in (78.2 kW/L)
Compression ratio: 7:1
Fuel consumption: 85.4 gal/hr (388.2 L/hr) at 2,500 rpm
Specific fuel consumption: 0.55 pints/hp/hr at 1,800 rpm (with exhaust turbine)
Power-to-weight ratio: 1.43 hp/lb (2.36 kW/kg)