by Calum E Douglas, BEng, MSc FRAeS
The First World War
The manufacture of aero engines in Britain is a surprisingly strange story, in that in many cases it consisted of making engines which were of entirely foreign design origin.
Early engine use was primarily for the Royal Flying Corps (RFC), with aeroplane fuselages manufactured by the Royal Aircraft Factory (‘the Factory’), the engines being almost exclusively of French manufacture. The National Archives: AVIA 46/263, pdf 51 of 71, para 2. This was no accident, since the Wright-Brothers flight of 1903, most developed nations had been very slow to pay attention to powered flight (their flight by unhappy timing coincided with the first massive global boom in motor-car manufacture and so, at the time, surprisingly little interest in flight with engines manifested). France was not tardy in her appreciation for its potentialities however: ‘the importance was fully realised, and the best engineering talent of the country was devoted to the evolution of an engine suitable for flying purposes. Gnome, Le Rhone, Clerget, Canton Unne, Renault and Azano all bent their efforts in this direction … brilliant and original achievements of French designers will be seen…” The National Archives: AVIA 46/160, pdf 5 of 69.
The French influence
The French, stunned by the possibilities, applied themselves fully to developing new engines necessary for the future of flight, the best engineers in France devoted themselves to it, and before long, Gnome, Le Rhone, Clerget, Canton Unne, Renault and Anzani revealed numerous brilliant and original advances in the aero engine. Germany was close behind France, and soon Mercedes, Benz and Austro-Daimler had respectable aero-engines. In Britain virtually nothing of the sort occurred, and most Royal Flying Corps aero engines in the First World War were of French design, as we simply had neither the designs nor the engineers conversant enough with the required disciplines to create a British design in a production-ready form which could be made in sufficient volume.
In 1912 the first contracts from the War Office to private aircraft firms were tendered, a year later the responsibility of airworthiness inspection was moved from the Factory to the Aeronautical Inspection Department (AID), which provided a pivotally crucial role in making the first rigorous standards to establish the safe manufacture of piston aero engines in Britain, at a time where there was not even so much as a set of agreed national standards for heat-treatment of critical engine components such as crankshafts or valves. Early aero engine reliability was utterly appalling and would be regarded today as being so erratic as to be scarcely believable. The AID set about curing this, and two sheds for inspection were erected on Farnborough Common. The National Archives: AVIA 46/263, pdf 51 of 71, para 5. To try to stimulate the British aero-engine industry a military aero engine competition was held at Farnborough in 1914, but despite this, at the outbreak of war Britain was virtually entirely reliant on aero engines either built in France, or having been designed there. The RAF Museum (Hendon): file B951, p.1.
By 1914 the AID had 49 technical staff, by mid-1915 this reached 488. The National Archives: AVIA 46/263, pdf 53 and 54 of 71, para 6 and 1 respectively. From this point on 50 per cent of AID staff were female for the rest of the First World War. The National Archives: AVIA 46/263, pdf 56 of 71, para 2. So awful was the situation that the War Office offered a £5,000 prize to anyone who could design a good aero engine, the ‘winning’ engine, a ‘Green’, was itself a total failure, and the only engine firm in all of Britain in 1914 which even made an aero engine that worked and could be produced was Sunbeam.
In the summer of 1915, the later very famous British aero-engine designer Frank Halford (designer of the Napier Sabre) returned from France and recommended the War Office take out a license to buildmthe new Hispano-Suiza V8 aero engine, his request was ignored. Much later Britain was forced to accept that his original request was correct, and at far higher price, negotiated a license to build it in Britain. The RAF Museum (Hendon): B963-B964, p.46 of 515.
In late 1916, the activities of the AID were both vast and manifold, staff were made resident on site at engine factories, and this was extended to inspectors who dealt purely with controlling quality even at the firms supplying these aeroengines. Factories, foundries and mills were under more or less permanent inspection by the AID. This culminated in ‘Resident Production Officers’ being appointed and stationed on site where the raw materials were being produced, under the AID Engine Branch. The National Archives: AVIA 46/263, pdf 58 of 71. Probably the best three RFC aircraft of the First World War all had French engines. Sopwith Camel with a Clerget 9B engine, the French Sopwith Pup with a Le Rhône 9C and the seminal SE5A with a Hispano-Suiza 8. This engine – which came from the French subsidiary of Hispano – was made under license in England as the ‘Viper’ . The British firm Beardmore had a useful 6-cylinder engine, but this was in fact a copy of the Austro-Daimler engine of the same layout.
A British indigenous industry emerges
As the war progressed, successful procedures were developed which helped transform the aero-engine manufacturing industry in Britain from an essentially jobbing-shop setup to a properly co ordinated mass industrial effort. Typically, about 15 engines from the first run were given a complete inspection of all parts, which included dimensional checks, surface finish, etc, and thereafter selective inspection was used, typical spot checks being the web thickness of a connecting rod. Where any fault was found, it was traced back to the root, and many faults in poor jigging and manufacturing methods were discovered, each firm had a ‘View Room’ where parts were made available for checking – even the revised methods of jigging to solve an error had themselves to be submitted to the AID and through this thorough system, many disastrous errors were solved before any major production or supply to service had commenced. The National Archives: AVIA 46/263, pdf 62 of 71.
The vast organization necessary to carry out these tasks is obvious, and by 1 January 1918, the AID had 5,646 employees, an increase of 100-fold since the outbreak of war just four years prior. By the time the armistice arrived, this number itself almost doubled to 10,657, which included 4,274 female ‘viewers’ who visually checked parts, and almost a thousand inspectors who checked processes and visited suppliers and factories. The National Archives: AVIA 46/263, pdf 67 of 71.
AID staff even visited squadrons in France during hostilities and spent a lot of time with the Expeditionary Force. AID was the most important body in Britain to influence the quality and standards of aero engines (and that of airframes) in the early days of aviation in Britain, and made profoundly critical improvements to aeronautical manufacturing processes and quality control. At the time they were rarely thanked for this, the factories disliked the intrusion, and the pilots often wondered why it took so long for new developments to reach operational service. So important was the AID to the fledgling industry, that AID staff were appointed to the Internal Combustion Engine Sub-Committee of the Advisory Committee for Aeronautics. Such was their knowledge due to having inspected and corrected nearly every facet of aero-engine manufacture that their knowledge had long since far exceeded that of a simple organisation which checked parts matched the drawings. The National Archives: AVIA 46/263, pdf 71 of 71.
By the end of the First World War, the situation in Britain for aero engines had improved significantly, with Rolls-Royce and Bentley producing useful engines in the form of the liquid cooled V12 Eagle and the BR1 and BR2 rotary engines. Although, they had admittedly only come into existence by copying and then further developing the French Clerget 9B, these were also the first and last rotary type engines used in Britain for military purposes. The bulk of the British automotive industry had been turned over to make whole aero engines or parts for aero engines during the First World War, after a very slow start indeed.
The AID, however, did endure and played a similarly significant role during the whole of the Second World War, by which time the problem was even more difficult as the technology had expanded so dramatically: ‘First 200 engines of a type, made at any factory, to be completely striped, thereafter only 1 in 10, 1 in 20, or 1 in 25 to be stripped …dependant on the results obtained on the 200 engines fully stripped.’ As the war progressed, the difficulties became extreme, with even British-designed aero-engine accessories being made abroad under licence then imported for use here, everything from fuel-pumps to magnetos to gauges had to be tested and complete interchangeability with British made products verified. The National Archives: AVIA 46/263, pdf 24 of 71.
Napier showed tremendous talent, and under their chief designer Arthur Rowlegde (who later defected to Rolls-Royce, leaving Napier to meander aimlessly for many years until Halford arrived) designed the Lion. This had the teething troubles ironed out just as the First World War ended, and thereafter became an instant classic, and was made in huge numbers over a period of nearly 25 years. Its success was so great that Napier stopped making motor-car engines entirely to concentrate on the Lion. The National Archives: AVIA 46/160, pdf 11 of 69.
In 1918 a major contender emerged in British aero-engine design, Roy Fedden, with his draughtsman, Leonard Butler, designed and developed the Cosmos Mercury and Cosmos Jupiter, of 450 hp. Confusingly the Cosmos Mercury was also the name given to a later engine released after the Jupiter, called the Bristol Mercury I. The Jupiter was a very significant engine, of which over 7,000 would eventually be made. Extremely unusually for a radial engine it had four valves per cylinder, and was a first-rate aero engine. Fedden however, would always struggle, as despite being an engineer of profound talents – a gifted visionary – he was cantankerous in nature, possessed no patience of any measure and drove his employees as hard as he drove himself. Few survived such an environment unscathed.
The inter-war years
In 1920 the aero engine department of the Bristol firm was created, on the site of the Filton aerodrome. The National Archives: AIR 5/1369, pdf 3 of 88. The firm acquired the Cosmos Engineering Company, which had been started by Roy Fedden. Fedden was retained and, perhaps with the steady influence of a larger firm to shackle his wilder notions, began to unleash a steady stream of truly classic radial aero engines.
Bristol and Napier
In September 1921 the Jupiter II engine became the first air-cooled engine to pass the official British Air Ministry ‘type-test’. Fedden claimed that at the time this was the most severe test of any in the world. With the war over, the civil market became the most logical to impose on and, in de-rated form, the Jupiter gained an outstanding reputation for reliability and economy. The National Archives: AIR 5/1369, pdf 11 of 88. Fedden realised that high altitude flight was likely to become critical and encouraged the use of supercharging and turbocharging. In 1923 a Jupiter fitted with a turbocharger reached 23,000 feet. A string of high-altitude records would subsequently be set by Bristol during the 1920s and 1930s. The Jupiter saw substantial civil use and was used by Imperial Airways De Havilland airliners from about 1926.
Not everything went well after the First World War, however, and a number of serious political and industrial machinations contrived to wrestle the fundamental design and manufacture of aero engines away from the Royal Aircraft Factory, which became the Royal Aircraft Establishment. The National Archives: AVIA 46/160, pdf 8 of 69. This was somewhat overturned by the advent of the gas-turbine as a viable method of propulsion, but in terms of airframe and piston engine design and manufacture the Royal Aircraft Establishment was to a large degree neutered from about 1927 onwards. In 1927 the Napier Lion won the Schneider Trophy race in a Supermarine S.5 by which time the major engine manufacturers in Britain had solidified into four runners: Siddeley, with the Jaguar engine; Bristol, with their range of radial air-cooled poppet valve engines by Fedden; Napier, with the lion engine; and Rolls-Royce, which concentrated on liquid-cooled engines with the Condor and Kestrel V12s. The RAF Museum (Hendon): B963-B964, p.160 of 515.
The next trophy was instead to focus on the Rolls-Royce engine as the main entry, as Rowledge had since departed Napier for Rolls-Royce and it was felt that with him had departed the genius of the Napier firm. In the same year Halford joined Napier but sadly, despite Halford’s unquestioned skill, the three engines which Napier built between 1927 and 1945 did not include a classic of the same impact as the Lion. The first, the Rapier, was too small for military service, the Dagger worked but achieved no great success, and the Sabre, whilst eventually impressive, had absorbed so much resource to overcome its intrinsic difficulties with the sleeve valves, that it did not materially influence the fortunes of the RAF in the Second World War in the way the Lion had dominated the 1920s. The National Archives: AVIA 46/160, p.11 of 69. ‘The story of the Sabre, therefore instead of being like that of the Merlin, one of continuous evolution to ever increasing power outputs, is one of a whole series of setbacks.’ The National Archives: AVIA 46/160, p.40 of 69. Six Napier Sabres were ordered for development purposes in 1937 by the Air Ministry. It was expected to be ready in around 1939 for fitment in the Typhoon fighter, which was at that early stage intended not as a ground attack fighter but as an interceptor to totally replace the Hurricane and Spitfire. The National Archives: AIR 2/2833, p.35 of 58.
Bristol had developed several other engines which were projected to be of use for the mid-sized transport aircraft, the Neptune and Titan, of the 200 to 400 hp class, however a market for these never materialised in Britain, although the Titan was made under license in large numbers in France by Gnome et Rhone and by S.A.B.C.A. in Belgium. The National Archives: AIR 5/1369, p.28. About the same time, in 1927, the jet engine as a method of aircraft propulsion began to take form, when A. A. Griffith submitted his report at the Royal Aircraft Establishment on an axial turbine. This early form was not however a turbojet, more of a turbo-prop, but it predates Whittle’s thesis by around three years. Griffith however, much like Whittle, found scant support for his idea and about a decade passed before any actual prototype apparatus was built. On 8 November 1929, the War Office responded to overtures to develop the jet engine with the following: ‘I am directed to … inform you that Dr A. A. Griffith’s secret invention relating to internal combustion engines, hot air turbines etc, is unlikely to be required for War Department purposes.’
When the Aeronautical Research Council offered support – finally – it was given to both Whittle and Griffith, thus starting serious British work into both axial and centrifugal jet engines. F. W. Armstrong, ‘Farnborough and the Beginnings of Gas Turbine Propulsion’, Journal of Aeronautical History, paper 2020/02, p.33. Eventually Griffith’s work progressed further and was developed more at the Royal Aircraft Establishment by Hayne Constant, culminating in Metropolitan Vickers being brought in, as they had a lot of experience in steam turbines for static ground power generation usage. It is important to note at this stage that there is no evidence at all of any personal animosity between Griffith and Whittle, Griffith simply believed that the axial layout would prove to be the better system and gave his engineering judgment that this was so. In this he was proven entirely correct in most applications. Author’s interview with Griffith’s son, 15 December 2024.
In 1929, production of the ‘new’ Mercury engine, the Bristol Mercury continued, with a view to its application in high-altitude high-performance military fighters. In the form of the Mercury IV, of around 540 hp, it gained maturity as a production-ready engine. In 1932, the Bristol Pegasus air-cooled radial engine was ready in production form, eventually it would reach a rating of 965 hp and an astonishing 32,000 were eventually built. In the same year it set a world altitude record of 43,976 feet, piloted by Capt. C. F. Uwins, the chief test pilot of Bristol Aeroplane Company. Later, in 1937, the Pegasus achieved the first commercial transatlantic flight crossing, in the Caledonia and Cambria flying boats of Imperial Airlines.
Sadly, Fedden, whilst making very few mistakes, saved them for particularly major themes. Ignoring the glorious string of blinding successes of nearly all the major Bristol air-cooled radial engines with standard poppet valves, his forward-looking gaze sometimes reached too greedily. Convinced that the sleeve-valve promised the future, he swung most of the resources at Bristol towards the new Hercules air-cooled radial, the result set the company back half a decade and prevented it from developing any engines to rival those of Rolls-Royce during the late 1930’s for military fighter use. The Hercules III was a twin tow 14-cylinder radial with a two-speed supercharger, it was in production in 1939. A higher performance engine, the Centaurus, was put into the design phase in 1937, essentially as Fedden’s attempt to field an air-superiority military engine of the 2,000 hp class, but it was not ready until almost the end of the Second World War, and only around 2,500 were ever built.
Rolls-Royce
Whilst somewhat of a broken record, the simple fact remains that the single most staggering achievement in British aero-engine manufacturing remains the Rolls-Royce Merlin. However, far from being an immediate success, its early years were plagued with the severest of problems, and the Merlin Mk1 engine, must only be described accurately as dire. Beginning as the PV.12 (private venture No.12) in around 1931; by 1933 it was modified to have a new cylinder head, which was a unmitigated disaster. It ran in modified form, known then as the Merlin, in 1935. Not until just before the outbreak of war was the Merlin a reliable and useful engine, in the form of the Merlin III. Fortunately, the wilderness years had not been in vain, and had ironed out all the really serious flaws, and the Merlin became the engine of choice for almost all the most important front-line RAF aircraft.
The Second World War
Bristol
The British Government had been planning for the Second World War at a date far earlier than generally assumed and planned for a massive expansion of production of military aero engines, known as the ‘shadow scheme’ . Plans to mass manufacture Bristol engines for the war began to be enacted shortly before May 1936. Eventually Austin, Daimler, Rover, Standard and Humber all produced Bristol engines. Initially it was hoped that the automotive industry, being used to producing far more engines than the aircraft industry, would be well set to assist with minimal intervention. This was quickly proven to be wishful thinking, the standards required for aircraft engines were in many cases considerably above those for a car plant, and ultimately the production sites ended up being totally re-tooled for aero engines. ‘At the end of 1938 the Air Ministry had detected an error in their calculations, It had always been assumed that in war the main motor-car firms could be comparatively easily turned over to aero engine production but when an attempt was made for formulate detailed plans it was found that most of the plant used in car manufacture was too specialised to be of the slightest use.’ The National Archives: AVIA 46/159, p.28 of 99 (see para 38). The same applied to the tolerances required. ‘It had at one time been supposed that motor-car manufacturers would be able to apply the lessons of mass production directly to aero-engines. But when the possibilities were closely examined it was found that the consequences of even the smallest error were liable to prove too series and it was necessary to follow exactly the same technique employed by Bristol. The case for uniformity was strengthened by the need for complete interchangeability between engine made by the group.’ The National Archives: AVIA 46/159, p.27 of 99 (see para 35).
Bristol had first agreed to set up the shadow factory scheme, Rolls-Royce had been extremely reticent, citing concerns about the intellectual property of the firm passing into organisations that they distrusted. Neverthless it eventually became obvious that it would be a military necessity to take production outside the core Derby factory. Eventually, production was underway at new purpose-built factories at Crewe and Glasgow and a new plant managed by Ford (UK). It should be noted that Ford did at no stage undertake Merlin production in any of their existing plants, the factory was a new build facility, funded by the Air Ministry. Ford were contracted to act as the plant management for this entirely new factory. The plant at Glasgow (Hillingdon) at its peak employed nearly 25,000 workers, and produced 23,647 Merlin engines. Rolls-Royce Heritage Trust: “Rolls-Royce Hillington, ‘Portrait of a shadow factory’”, Historical Series No.44, p.113.
Packard and Rolls-Royce join forces
Packard were also contracted to build Merlin engines, and at about the same time that Rolls-Royce were setting up the mass-production facilities at factories like Glasgow, which employed unskilled and semi-skilled workers with no history of such work, including large numbers of women, Packard set up new factory in the USA. To assist with this enormous task, Rolls-Royce sent three of their top engineers to live in America for several years to assist, including Colonel Barrington, the Chief Engineer, and J. E. Ellor, the Chief Experimental Engineer. It was an incalculable strain, Barrington failed to stand the pressure and literally died from the stress in America in 1943. Ellor also struggled and became so unwell dealing with the task that he was sent back to England where he was hospitalised for a long period in 1944. He was dead seven years later, aged just 59. John Reid, the Production Engineer, was in some ways the only one of the three Rolls-Royce men sent to America who survived the Packard-Merlin programme. Reid chose to retire the year Ellor died. Despite the tragedy, with stress bordering on horror, it was a success. Packard produced 54,714 Merlin engines, around a third of the global total of 168,494 Merlin engines made from 1939 to 1945, of which 113,780 were made in the UK.
The Packard total was about the same as the total made in the two main shadow plants at Crewe and Glasgow combined (52,296). The production abroad was however fraught with problems, one was that updates to the Merlin were constantly being released, and these were required to be embodied in the engine reaching service squadrons. In the UK this was less troublesome as the modifications were easier to manage and institute, as they were being developed primarily at Derby in England. It proved too difficult to interrupt the flow of production at Packard however, as of course there was a delay in shipping the parts to America and also Packard thought they could not deliver the targets with constant interruptions. Therefore, Rolls-Royce agreed that a substantial part of the operations at their Glasgow plant would be turned over to modification updates to every Packard Merlin engine shipped to Britain. The National Archives: AVIA 10/12, p.41 of 87. Meeting, 26 May 1942, 3pm ‘Packard Merlin Engines’, I.C. House, London, present B.A.C with Hives and Sidgreaves from Rolls-Royce. The engines arrived at Prestwick, and were all transported to the Glasgow plant where the latest modifications from Derby were installed.
Almost total parts interchangeability was, like Bristol, also achieved, the only exceptions to this were in the case of the parts which were actually of entirely different design, for example Britain could not spare the carburettor capacity nor the shipping time to send carburettors for the Merlin to America, only for them to come all the way back again, therefore Packard fitted American Bendix carburettors. There were also differences in the supercharger drive system (primarily for patent reasons from Farman, from whom Rolls-Royce licenced their use of the Farman gearchange mechanism). With a few exceptions like this, all the other engine parts were completely interchangeable, in both directions. A parts interchange book was published so any field personnel operating any Merlin could use a Packard part, and vice versa. Sadly, entirely factually incorrect stories still circulate to this day about the poor tolerances of Rolls-Royce parts made in England compared to those made in the USA, these mostly stem from a misunderstanding of the situation by comparing pre-war British Merlin production technique to that of wartime Packard technique. In fact, nobody had ever mass produced the Merlin before the new factories were set up in Glasgow, Crewe, Ford and in Detroit. Both nations had to transition the pre-war designs and methods to modern practise, and this happened roughly in parallel on both sides of the Atlantic. The real contribution of America, which Britain could not match, was in the supply of modern automatic machine tools. British plants relied on these, just as the Americans did and the Packard-Merlin effort was very much a team collaboration. Five plants in Britain and one in the USA produced Merlins, with just two of those in Britain also producing far smaller numbers of the larger Griffon Engine, Derby and Crewe. An uncertain number, probably around 800 were also apparently made by Continental.
Napier
In October 1939 the Air Ministry approved proposals for a new factory in Liverpool to build 2,000 sabre engines per annum. The type test was eventually passed at 2200 hp in June 1940, an extraordinarily high output. Sadly, the test being completed, and the engine being reliable enough for actual service, proved to be quite discrete and differing achievements. Eventually in early 1941 the first deliveries of engines began but by April 1942 the situation was still dire, the engine was not yet reliable. The matter was discussed by the Aircraft Supply Council: ‘It was apparent that the firm’s development staff was only doubtfully adequate to the task of solving the intricate mechanical end metallurgical problems which Major Halford had set before them before turning most of his attention to the De-Havilland jet projects.’ However, it must be added that Halford had only departed under very high levels of duress expressed to him via the government, as the problems of the new jet engines were seen to require the best minds available.
Rolls-Royce was approached with the idea of taking over the entire Napier firm to get the Sabre back on track, this was decided against; ostensibly because several key Napier staff were not expected to agree to such an arrangement without generating significant and even more damaging animosity. In May 1942 the Sabre was averaging 340 flight hours per engine failure, the Merlin was achieving over 1,000. However, it was expected that eventually the Sabre would inevitably converge to the same level of reliability. Despite trying the gentle approach and merely installing a Deputy Director at Napier to oversee things, the situation became untenable, and English Electric were brought in to take over the entire running of the Napier concern. Eventually, after tremendous pain, by mid1944, the Sabre MkIIA was proving acceptably reliable and very powerful in the Typhoon and Tempest V. The fact that the Sabre had failed to prove a useful high power engine for the first three years of the war could have been calamitous, especially as the other engines in the same power class, the Vulture and Centaurus, proved no less difficult than the Sabre to tame, and the Griffon, whilst not exhibiting calamitous symptoms, was proving very slow to develop as so much of the Rolls-Royce effort was expended on the Merlin.
Thus, British aviation in the first three years of the Second World War was essentially saved from disaster by the wholly unexpected ability of the Merlin to be developed to entirely unexpectedly high-power levels. The Napier story was a tragic one, and the blame cannot be entirely levelled at Napier. The forced removal of Halford, their chief designer, a definite plan to place Napier under Rolls-Royce control and the immense engineering challenge involved, all conspired to doom the project to a shadow of its potential.
The wartime shift to the turbojet
Amstrong-Siddeley, having somewhat fallen adrift in piston-engine design, were instructed by the Air Ministry to refocus entirely on the jet engine in 1941 (as it turned out this was not disadvantageous, and they did well enough to merge with first Bristol, then finally with Rolls-Royce themselves in 1966). That year the design for the axial flow F.2, designed for 2,300 lb thrust, was completed in December 1940 at the Royal Aircraft Establishment, the plans were then passed to Metropolitan Vickers (Metrovik) for detail design and manufacture. This engine ran in November 1941. The National Archives: AIR 20/4013, p.116 of 236. It had 10 compressor stages and two turbine stages, all axial. This is probably the first British axial turbojet which essentially remains in general form unaltered to the present-day axial turbojet engine. Despite the Whittle centrifugal engine being by far the better known, the Metro-Vick axial jets, derived from Griffith’s work in 1927, and further developed by Hayne Constant at the Royal Aircraft Establishment were not far behind, and in terms of the first test flights, the gap between the two teams was about two years, with the Whittle jet flying in mid 1941, and the axial Metro-Vick design flying in the Gloster F.9/40 (the Meteor prototype) in November 1943. The National Archives: AIR 20/4013, p.116 of 236.
These Metro-Vick engines were eventually developed to the F.9 model, named the Sapphire, it became an extremely good turbojet engine. It first ran in 1948 at which time it was, according to former National Gas Turbine Establishment (NGTE) director Frank Armstrong, probably the best axial turbojet in the world. Not only was the Whittle engine design passed on to the USA, in fact essentially everything that existed in Britain on turbojets was sent there in 1941. This included virtually all technical reports, research reports, meeting minutes and test data from every relevant establishment, and 46 sets of parts for the De Havilland Goblin, together with six complete Goblin engines, and every single technical document, engineering manufacturing drawing and manual for the following engines:
- Power Jets W2/500
- Power Jets W2/700
- Rolls-Royce Welland
- Rolls-Royce Derwent
- Armstrong Siddeley ASX
- Metrovik F2 & F3
- De-Havilland Goblin
This provides a reasonable overview of the notable British turbojets underway in late 1941. Remembering that Armstrong Siddeley had been asked to leave piston engine work in 1941, this came good with the Sapphire engine, as Metro-Vick could not mass produce the Sapphire and now such a production was needed they agreed to hand over the project to Armstrong Siddeley. The engine served in RAF service aircraft, and 12,000 were built under licence by the Americans at Wright-Corporation. It was the first ever British jet engine to pass a type-test at over 10,000 lbs thrust.
The cost of the Ministry of Aircraft Production (MAP) taking over Power Jets Ltd in March 1944, and moving production to larger concerns, was however high. The move was not popular with Power Jets staff, and no less than 16 engineers resigned, almost the entire core of Whittle’s design team. The Times, 15 April 1946, see The National Archives, AVIA 46/234 p.234 of 288. So although production had been assured, by transferring the project to Rolls-Royce at Barnoldswick: ‘…some of them will pass out of aviation altogether… Whittle’s successful design team has therefore, been dispersed, and there is little hope of it being reformed.’
However, it cannot be said that the government`s policy in this regard was wholly detrimental or ill considered, for Whittle’s firm so altered, became the National Gas Turbine Establishment (aka ‘Pyestock’), which was for a considerable time one of the premier turbojet scientific centres in the world, eventually able to ground test engines as potent as the Olympus to altitude conditions. Its sad demise was only due to the lack of an indigenous air-superiority turbojet engine programme, which might have been able to use its formidable facilities. Author’s interview with Frank Armstrong CEng, FREng, FRAeS, former NGTE Deputy R&D Director.
The post-war era: the turbojet phase
The post-war period obviously marked the point, at least for mid to large aircraft sizes, of transition from piston to turbo-prop and turbo-jet propulsion. The costs and resources needed to develop these powerplants rapidly spiralled, and this naturally resulted in dramatic consolidation of British aero-engine manufacture into a few key firms and engine types.
Although piston engine manufacture did carry on for a time, with some efforts and minor success from Napier, Rolls-Royce with the civil Merlin service and Bristol with the civil Centaurus, this effectively became the transition point between the two technologies. Rolls-Royce emerged as the dominant force, in no small part through certain somewhat aggressive mergers, while earlier contributions came from de Havilland, Armstrong Siddeley and Bristol. Napier had intended to pivot towards gas turbine development, and had constructed an extremely capable and very large design, build and test facility. This station, the Napier Research Station Liverpool, was built between 1947 and 1953. It was regarded at the time as one of, if not the premier, establishment of its type in all of Europe. It was purchased by Rolls-Royce in 1961 and effectively shut down within about two years. Thus ended the often promising efforts of the firm Napier in the world of the gas turbine powerplant.
Turbojets like the de Havilland Goblin and Rolls-Royce Avon powered early military and civilian jets, enabling Britain's leadership in jet aviation during the 1950s. Turbofans (essentially a high bypass turbojet which was more efficient at civil aviation speeds) starting with the Rolls-Royce Conway, which revolutionised efficiency on long flights, and ongoing with the modern Trent series reflecting ongoing innovation amid global competition. Even Rolls-Royce struggled however, and financial strains, such as the RB211's development, led to Rolls-Royce's 1971 nationalisation.
The primary manufacturers included the de Havilland Engine Company (based in Leavesden, Hertfordshire), which pioneered early turbojets before merging into Bristol Siddeley in 1961. Armstrong Siddeley operated from Coventry and Brockworth, Gloucestershire, focusing on axial-flow designs until its 1959 merger with Bristol. Bristol Aero Engines, later Bristol Siddeley (Filton/Patchway, Bristol), specialised in advanced turbojets and vectored-thrust turbofans. Rolls-Royce, with factories in Derby and later Bristol, absorbed Bristol Siddeley in 1966 and became the sole major player, driving turbofan advancements. Therefore, after this date, the story of all the other players essentially ended.
De Havilland Engine Company
The de Havilland Engine Company, established in 1944 at Stag Lane and relocated to Leavesden in 1946, was an early leader in centrifugal compressor turbojets. It produced the Goblin (originally Halford H-1), a 3,000–3,750 lbf thrust engine that first ran in 1942, but saw major post-war production. Manufactured at Leavesden, approximately 4,400 Goblins were built, including licensed variants, powering the de Havilland Vampire fighter (over 3,000 UK-built units) and experimental aircraft like the DH 108 Swallow. Its significance lay in being the second British jet to fly and the first certified for propulsion, enabling Britain's entry into jet fighter production and exports to Sweden and Italy. The follow-on Ghost (Halford H-2), scaled up from the Goblin, delivered 5,000 lbf thrust and entered production in the late 1940s at Leavesden. Approximately 1,800 units were produced, powering aircraft including the early de Havilland Comet 1 series (the first twelve production aircraft used four Ghost engines each) and 1,311 de Havilland Venoms (single-engined). It powered the world's first jet airliner, the Comet, which despite the well-known fatigue issues was a major national achievement in turbojet application. De Havilland was merged into Bristol Siddeley in 1961, ending its independent operations as an independent aircraft firm.
Armstrong Siddeley
Armstrong Siddeley, based in Coventry, with later production at Brockworth, specialised in axial-flow turbojets from the late 1940s. The Sapphire, acquired from Metropolitan-Vickers in 1948, produced 7,500–11,000 lbf thrust and entered service in the 1950s. Approximately 1,200 British units were built (plus US licensed J65s), powering 436 Gloster Javelins (two each), 140 Hawker Hunter F2/F5s, and 50 Handley Page Victors (four each). It was significant as Britain's first engine over 10,000 lbf, featuring an annular combustor for reliability, though surge issues in clouds required fixes.
The Viper, derived from the Adder turboprop in 1951, offered 1,700–3,500 lbf thrust and was produced for over 50 years. More than 5,500 were manufactured, built initially in Coventry and continuing under Bristol Siddeley and later Rolls-Royce. It powered 741 BAC Jet Provosts, over 800 Aermacchi MB-326s, and HS Dominie trainers. Its expendable design for drones evolved into a durable trainer engine, introducing the 'Power by the Hour' maintenance model that influenced modern leasing. Armstrong Siddeley merged with Bristol in 1959 to form Bristol Siddeley.
Bristol Aero Engines and Bristol Siddeley
Bristol Aero Engines, operating from Filton, developed the Olympus turbojet from 1946, with production starting in the 1950s. Delivering 11,000–20,000 lbf thrust, around 600 units were produced, powering 134 Avro Vulcans (four each) and 20 Concordes (four each, as Olympus 593). Manufactured at Filton (later Patchway), it was the world's second two-spool axial turbojet, eliminating surge via twin-spool design, and enabled supersonic travel on Concorde, whilst also being adapted for marine use.
The Pegasus turbofan, originated by Bristol Siddeley from 1959 and continued by Rolls-Royce after the 1966 acquisition, saw over 1,200 units built at Bristol by 2008. Providing 15,000–23,000 lbf with vectored thrust, it powered all Harrier variants (over 800 aircraft), revolutionising VTOL operations for the RAF and USMC. Rolls-Royce acquired Bristol Siddeley in 1966, integrating its facilities.
Rolls-Royce
Rolls-Royce, with core production in Derby (and early work in Barnoldswick), led from the 1950s onward. The Nene turbojet (5,000 lbf), produced post-1945 in Derby, had limited British output (around 500), powering 500 Hawker Sea Hawks and 180 Supermarine Attackers. It was significant for doubling the power of earlier designs, though it was licensed abroad (notably to the USSR). The Avon, Rolls-Royce's first axial turbojet from 1950, saw over 11,000 units produced until 1974 in Derby, powering over 1,000 English Electric Canberras, 1,970 Hawker Hunters and 337 English Electric Lightnings. It also powered later marks of the de Havilland Comet (from the Comet 2 onwards) and the Sud Aviation Caravelle. It enabled non-stop transatlantic jet flights and remained in RAF service until 2006.
Shifting to turbofans, the Conway (the world's first bypass engine) from 1952 saw around 700 units produced in Derby, powering 50 Victors, 54 Vickers VC10s, 37 Boeing 707-420s, and 32 Douglas DC- 8-40s. With 18,000 lbf thrust, it pioneered air-cooled turbines for efficiency, entering airline service in 1960. The Spey, from 1964, yielded 2,768 units in Derby, powering Hawker Siddeley Tridents, Blackburn Buccaneers, and British Phantoms. Its low-bypass design accumulated 50 million flight hours, and was also adapted for marine use. The RB211, introduced in 1972, after financial turmoil leading to nationalisation, saw over 2,000 units produced in Derby, powering 250 Lockheed TriStars, numerous Boeing 757s (over 400 RB211-equipped) and some 747s. As the first three-spool turbofan, it offered 7% fuel savings and ETOPS certification.
The Trent family, starting in 1995 from Derby (with components in Bristol), has seen over 6,000 units produced across variants including the Trent 700 (Airbus A330), 800 (Boeing 777), 900 (A380), 1000 (787), XWB (A350), and 7000 (A330neo). Delivering 70,000-115,000 lbf, it has established Rolls-Royce as a market leader, powering modern wide-bodies with high reliability (over 200 million flight hours). Ongoing production emphasises sustainability.
The acquisition of Bristol Siddeley in 1966 effectively completed the consolidation of the British aero-engine industry under Rolls-Royce. The nationalisation of Rolls-Royce in 1971 led to the aero division being separated from the car production and diesel divisions of the company. The aero-engine division of Rolls-Royce was later returned to private sector as Rolls-Royce plc in 1987.
The UltraFan demonstrator
The UltraFan demonstrator programme represents Britain's current peak of gas turbine development. Unlike the direct-drive three-spool Trent configuration, the UltraFan employs a geared turbofan architecture with a variable-pitch carbon-titanium fan, representing a significant departure in design philosophy. The first UltraFan demonstrator completed its ground test in May 2023 at Rolls-Royce's Testbed 80 facility in Derby, running on 100% sustainable aviation fuel. By November 2023, it had achieved maximum power of over 85,000 lbf thrust, exceeding its design target, and accumulated around 70 hours of run time. The technology offers approximately 10% fuel efficiency improvement over current engines. Whilst not yet a production engine, the UltraFan programme is maturing technologies that are already being incorporated into the existing Trent fleet, and Rolls-Royce is developing both wide-body and narrow-body variants with flight tests planned before 2030.
Conclusion
This history reflects Britain's shift from wartime urgency to commercial focus. From the pioneering centrifugal turbojets of de Havilland through the consolidation of the industry under Rolls-Royce, British engine manufacturers have maintained a position at the forefront of gas turbine technology. Engines like the Trent sustain global competitiveness today, whilst the UltraFan programme points toward future developments in efficiency and sustainability.
Notable British piston engine powerplants (excludes overseas production)
| Engine | Firm/Factory | Type | Approx. production | Key aircraft | Significance |
| Jupiter | Bristol/Filton | 9-cylinder radial | 7,100+ | Bristol Bulldog, Handley Page H.P.42 | Widely licensed; powered 262 aircraft types by 1929; first with automatic boost control |
| Eagle | Rolls-Royce/Derby | V-12 liquid-cooled | 4,681 | Handley Page O/400, Felixstowe F.2 | First Rolls-Royce aero engine; enabled long-range WWI bombing |
| Kestrel | Rolls-Royce/Derby | V-12 liquid-cooled | ~4,750 | Hawker Hart, Fury | Precursor to Merlin; pioneered cast-block design for lighter weight |
| Merlin | Rolls-Royce/Derby, Crewe, Glasgow; Ford/Manchester | V-12 liquid-cooled | 112,545 | Spitfire, Hurricane, Lancaster | Most produced WWII engine; supercharged variants delivered up to 2,000 hp; key to Battle of Britain |
| Mercury | Bristol/Filton | 9-cylinder radial | 20,700 | Bristol Blenheim, Gloster Gladiator | Reliable interwar radial; licensed in Europe; accumulated high flight hours |
| Pegasus | Bristol/Filton | 9-cylinder radial | ~32,000 | Vickers Wellesley, Short Sunderland | Versatile for flying boats; set height records; long production run |
| Hercules | Bristol/Filton | 14-cylinder radial (sleeve-valve) | 57,400 | Bristol Beaufighter, Short Stirling | High-power WWII radial; overcame early issues for reliability |
| Gipsy Major | de Havilland/Leavesden | 4-cylinder inline air-cooled | 14,615 (majority pre-1945) | de Havilland Tiger Moth, Chipmunk | Standard trainer engine; simple, inverted design for better visibility |
| Cheetah | Armstrong Siddeley/Coventry | 7-cylinder radial | >37,200 | Airspeed Oxford, Avro Anson | Primary RAF trainer engine; durable for high-volume use |
| Lynx | Armstrong Siddeley/Coventry | 7-cylinder radial | ~6,000 | Avro Tutor, Westland Wapiti | Interwar trainer; scaled-down Jaguar design |
| Lion | Napier/Action | W-12 liquid-cooled | ~500 | Vickers Vernon, Gloster III | Powerful WWI/post-war; set multiple speed records; advanced valve gear |
Notable British gas turbine powerplants
| Engine | Firm/Factory | Type | Approx. production | Key aircraft | Significance |
| Goblin | de Havilland/Leavesden | Turbojet | 4400 | Vampire | Early jet certification, fighter exports |
| Ghost | de Havilland/Leavesden | Turbojet | 1,000+ | Comet, Venom | First airliner jet, scaled design |
| Sapphire | Armstrong Siddeley/Coventry | Turbojet | 1200 | Javelin, Hunter | High-thrust axial, surge-resistant |
| Viper | Armstrong Siddeley/Coventry | Turbojet | 5,000+ | Jet Provost, MB-326 | Long-service trainer, maintenance innovation |
| Olympus | Bristol/Filton | Turbojet | 600 | Vulcan, Concorde | Twin-spool pioneer, supersonic enablement |
| Nene | Rolls-Royce/Derby | Turbojet | 500 | Sea Hawk | Power benchmark, international licensing |
| Avon | Rolls-Royce/Derby | Turbojet | 11000 | Canberra, Hunter | Transatlantic flights, long RAF service |
| Conway | Rolls-Royce/Derby | Turbofan | 700 | Victor, VC10 | First bypass, efficiency gains |
| Spey | Rolls-Royce/Derby | Turbofan | 2768 | Trident, Phantom | Versatile military/civil, 50M hours |
| Pegasus | Bristol Siddeley/Bristol | Turbofan | 1200 | Harrier | VTOL revolution, vectored thrust |
| RB211 | Rolls-Royce/Derby | Turbofan | 2,000+ | TriStar, 757 | Three-spool debut, fuel savings |
| Trent | Rolls-Royce/Derby | Turbofan | 6000 | A330, 777, 787 | Market dominance, high-bypass efficiency |
© Calum E. Douglas BEng MSc FRAeS, 2026