Johnnie had been a USAAF bomber pilot in the War, and was proud of his Irish descent. He was a keen, intelligent officer of great driving power, and poured millions of dollars into the Orpheus and the Fiat G91 project at a time when development funds in the UK had virtually dried up. The office of MWDP was an act of great generosity by the American people, and Johnnie and von Kármán were great ambassadors for them. It was a stroke of good fortune for Bristol when the Orpheus was chosen, and I count myself as very fortunate in the friendship that I had with these two outstanding men.
The meetings in Turin at the Fiat works took place only in the morning because after an Italian light lunch nobody was in any condition for further work. On one occasion at the great Fiat works we were summoned to the main boardroom to report progress to the chairman himself, Dr Valletta. The room was enormous, probably 20 ft high, and its giant mahogany double doors were polished like mirrors. We waited with bated breath for the great man to enter. Suddenly, and in total silence, the mighty doors swung open, controlled by a footman in full morning dress, and through them entered the very short and stocky Valletta. But he was clearly a leader, and took charge of proceedings immediately, giving Gabrielli all possible support and encouragement.
Meanwhile Petter, having disqualified himself from the NATO competition, did create a beautiful and exciting little fighter in the Gnat. But it is difficult to sell a fighter that is rejected by its own air force. The Ministry bought six to play with, and this caused great embarrassment when it was found that RAF pilots were queueing up to get their hands on them. Yugoslavia bought two, because it looked like something they might make themselves, and 12 were sold to Finland where they enjoyed long and successful front-line careers. Much later the RAF did buy a two-seater Gnat as a trainer, with a reduced-thrust version of the Orpheus, and millions saw the specially painted group that for many years formed the equipment of the Red Arrows. But it was far-away India that finally decided that Petter, who sadly died quite early in the project, had been right all along. The Gnat appeared to be a tough customer in a dogfight, and also within the technical capacity of Hindustan Aircraft Ltd to build under licence. The same company’s engine division took out a licence for the Orpheus, and the engine has remained in production to this day at Bangalore, for the Gnat, for the HAL-designed HF.24 Marut twin-engined fighter, for the HAL improvement of the Gnat called the Ajeet (unconquerable) and for the Kiran II trainer, as well as for booster pods fitted to heavily laden transports taking-off from high-altitude airstrips. Thus, despite the total disinterest of the British Air Staff, Verdon’s instant agreement to paying 25 per cent of the development bill proved a fine investment.
To conclude this chapter I would like to acknowledge the help given over many years by the various overseas agents of both the Bristol Aeroplane Company and Rolls-Royce. In the case of the Orpheus the chief intermediary role in France was played by the Bristol agents, the Société Franco-Britannique, whose Gerry Morel and Jean Allais were able to open all doors. Gerry and Jean were patriotic and cultured Frenchmen devoted to the cause of Anglo-French co-operation in aviation, and they played a major role not only in the Orpehus but also in the Pegasus and the Bristol/SNECMA Olympus 593 for the Concorde. In Italy we had the Mercantile Italo-Britannica, whose Major Bill Rogerson was on intimate terms with Fiat. He conducted the early Orpheus negotiations with great skill and success, and remained an essential link through the production of the G91. Sadly, Gerry Morel died from injuries received at the hands of the Gestapo before our projects saw the light of day, but Rogerson and Allais remained great friends, ever generous in hospitality and ever watchful of our interests.
1
2
3
The first non-aviation applications of an aircraft gas-turbine (1) fast patrol boats HMS Brave Borderer and Brave Swordsman, (2) Princetown peak-lopping generating station, (3) Bluebird. All were powered by the Proteus.
Chapter 10
The Pegasus
In Paris, in 1957, Johnnie Driscoll, having got the Fiat G91 light fighter firmly launched, turned his attention to vertical take-off. Rolls-Royce had already flown the ‘Flying Bedstead’, which was a metal frame supporting two Nene engines mounted so that their jets acted as lifters, and under the stimulus of A. A. Griffith was busy developing small ultra-lightweight jet engines such as the RB108, so that a useful military aircraft could be lifted off the ground by a number of such engines.
Even in a single-seat fighter, a minimum of eight such engines would be required for safety purposes. The idea of a pilot attempting to monitor all these engines during the dangerous time of take-off beggared the imagination, despite the initial success of the Short SC.1 research aircraft, lifted by four RB108 engines with a fifth for propulsion. Nonetheless, the subject of jet-lift V/STOL (vertical or short take-off and landing) was fast becoming all the rage. Driscoll was one of those who believed in the development of V/STOL close-support aircraft, up with the forward troops where they could replace heavy fuel loads by weapons, and could be dispersed away from main airbases.
Thus the stage was set for Michel Wibault, who between the wars was a leading French designer of fighters and airliners, to ponder a solution to the problem. He was living in some style in Paris as the protégé of Winthrop Rockefeller, who was interested in his project studies. One such was his proposal for a vertical take-off and landing aircraft called Le Gyroptère. It resembled a stumpy jet fighter containing a Bristol Orion engine, our planned successor to the Proteus, with the propeller drive replaced by two cross-shafts driving four large centrifugal blowers arranged like wheels at the sides of the fuselage. The blower casings could be rotated so that at take-off the four jets of compressed air were directed vertically downwards, evenly disposed around the CG (aircraft centre of gravity) to make the Gyroptère rise vertically. The pilot could then slowly rotate the casings directing the jets rearwards to give propulsive thrust, the lift gradually being taken over by the wing as the forward speed built up.
For my part, the Derby solution of the multi-engined aircraft seemed hardly likely to succeed. The engine control systems had to be automatically connected, so that if one lift jet failed to respond, or failed mechanically, its opposite number on the other side of the CG would be immediately shut down to preserve balance. Thus, at least two engines’ worth of spare thrust was needed just for safety reasons. Moreover, after take-off, the lift engines were just dead weight to be carried throughout the flight, before they were all (hopefully) relit for landing. I used to say facetiously “The pilot must press a button for take-off, close his eyes and, if he is alive one minute later, then he will be flying!” Even my friend Frank Robertson, Chief Designer of the SC.1, used to quip (when he was referring to multi engines in order for balance to be preserved) ‘Nothing comes down faster than a vertical take-off aircraft upside down’.
Wibault’s concept seemed more practical and realistic, though the shafts, gearboxes and centrifugal compressors were cumbersome, inelegant and very heavy. This was my verdict to Johnnie Driscoll when he called me to Paris to discuss Wibault’s proposal. But I was strongly in favour of what von Kármán, christened ‘vectored thrust’. A vector can be directed in any direction, enabling one engine to be used for both lift and propulsion.
At Bristol, I discussed the concept with Lewis, Young and Quinn. Lewis proposed one large axial compressor to replace the four centrifugals with two rotating nozzles, one on each side.
It was easy to show that, for a fighter-type aircraft, it would be better to replace the complex and expensive Orion by the cheap, lightweight Orpheus. The obvious place for the extra compressor was on the front. The next big step was to make it an integral part of the engine, large enough for its inner portion to supercharge the air entering the original Orpheus compressor, and thus increasing its efficiency and power. The air from the outer part of the large compressor could be ducted to left and right vectoring nozzles, while the hot jet could also be used for lift.
We
took our proposal back to Driscoll and von Kármán, who were enthusiastic and urged us to proceed to the design stage. At this crucial moment Driscoll was moved back to Norstad’s staff, and Col (later General) Willis Chapman replaced him. Fortunately Chapman could not have been more enthusiastic and positive in his support. We took on Wibault as a consultant, and he and Gordon Lewis were the original vectored-thrust patentees. Sadly, Wibault died soon afterwards and never saw the success that his original idea was to lead to.
At this juncture Sir Sydney Camm took a hand. He had been watching the Derby proposals with growing disbelief, and suddenly sent me a one-line whip: “Dear Hooker, What are you doing about vertical take-off engines? Yours, Sydney”. I sent him our first BE.53 (Pegasus) brochure, and then at once busied myself in Olympus work. It was thus a shock when Camm telephoned me a few weeks later, asking “When the Devil are you coming to see me?” I replied “As soon as you like, of course; but what is the subject?” And he said “It’s vertical take-off; I’ve got an aeroplane for your BE.53”.
We set off hot-foot for Kingston, where Camm showed us a drawing of the P.1127. It looked very like the Harrier of today and we were thrilled when Camm told us that, despite the inability of the Ministry to show much interest (because manned military aircraft were out), he intended to go ahead and fly a prototype. The Hawker Siddeley Board supported him, and put up the money, just as it had done in 1936 in tooling up to make 1,000 Hurricanes. In 1936, however, at least the company knew the Hurricane would be ordered eventually, whereas in 1958 there was absolutely no reason at all to think that the P.1127 would be ordered, even as a prototype.
On this historic visit to Kingston we met Roy Chaplin, whom I knew from Hurricane days pre-war, and Ralph Hooper and John Fozard. The latter two were brilliant young designers who, after the death of Sir Sydney and the retirement of Chaplin, were to bring the revolutionary P.1127 through the intermediate stage called the Kestrel, to service as an operational weapon system for the RAF in early 1969 as the Harrier. To them and to Chief Test Pilot Bill Bedford must go the credit of producing the first practical VTOL jet.
We started on the definitive design of the BE.53 in 1958. The deal with Bill Chapman of MWDP was the same as for the Orpheus; they paid 75 per cent of the initial funding, and Bristol were to pay the rest. As before, Verdon never hesitated, and this proved one of the most important investments ever made by British private industry. At the same time, had it not been for MWDP, and Driscoll and Chapman who administered that fund, it could never have happened. Chapman’s only proviso was that we should build two flight-cleared engines and allocate these to Hawker Aircraft.
Charles Marchant, having handed over his responsibilities for the Proteus and Olympus to Warlow-Davies and Hughie Green, was able to devote his whole effort to the design of the Pegasus. The engine appeared in the flesh in August 1959, and was handed to John Dale to develop to flight standard.
Converting the Orpheus into the Pegasus presented us with new challenges. Basically we took an Orpheus and added an additional LP turbine driving a shaft passing down the centre of the (conveniently large-diameter) HP shaft to an axial fan at the front. At first this had two stages, but we later added a third stage, all these being overhung like a propeller ahead of the front bearing. The air from this fan was split, the inner supercharging the gas-generator, and the outer part discharging air from the vectoring front nozzles.
Camm, remembering his “birfucated” jet pipe in the Sea Hawk, suggested bifurcating the jet pipe on the Pegasus and using a second pair of left/right nozzles rotating in unison with the first pair. A further desirable feature was to make the HP and LP spools rotate in opposite directions, thus almost eliminating the engine’s gyroscopic couple. This is a very important objective for a V/STOL aircraft, enabling it to hover under perfect control.
Thus, the unique Pegasus arrangement evolved and gave its output in the form of four jets, symmetrically arranged in front and rear pairs so that, when the engine was installed in the Harrier, the resultant thrust acted through the aircraft CG. In the first Pegasus the front jets delivered air at only about 2.3 pressure ratio, resulting in a temperature of some 100°C. We called these the ‘cold jets’. The two rear jets delivered gas at about 650°C. They handled a smaller mass flow, but we designed the Pegasus so that the front and rear jets gave approximately the same thrust, thus easing problems in the balance and control of the aircraft.
For minimum weight we made the ‘cold’ ducts and nozzles in glassfibre. Well into the flight programme the test pilot, Bill Bedford, suffered a sudden inflight malfunction on the side unseen by the accompanying chase aircraft. At low level Bedford found he had lost lateral control, and could not stop the aircraft rolling. He managed to eject safely, but the aircraft was wrecked in the crash. It would have taken a long time to pinpoint the cause had not a local farmer near Yeovilton walked in a day or two later saying “Does this belong to you?” He was holding one of the glassfibre nozzles!
We saw at once that in making the nozzle the strong glassfibre had not been taken round into the bolted flange holding it to the mounting ring (which was then not part of the engine but of the aircraft), leading to a fatal weakness. Design in glassfibre demands careful control of accuracy, consistency and the run of the fibres, which alone provide the strength that is absent from the plastic adhesive filler which binds the fibres together.
Subsequent efforts to make reliable cold nozzles verged on the hilarious. Sydney cursed us for being such idiots as to use material he wouldn’t have for a lavatory seat, and said his chaps would make us a set of nozzles in aluminium sheet. When we got them we put them on the Pegasus, to run on the testbed, and within two or three hours they were cracking and breaking up because of air pulsations. So we told Sydney we would make the damned things in titanium, but these also cracked. Sydney then washed his hands of the whole affair, but John Dale pointed out that, during all the argument, the rear nozzles, in heat-resisting steel, were working perfectly, even at 650°C. So we went to heat-resistant steel, and the front nozzles have never given any more trouble. Sydney howled blue murder at the extra weight of 50 lb, but as we started the Pegasus at 11,000 lb and went into production at 21,500 lb he didn’t do too badly in the end.
Many features of the Pegasus were innovations. The original BE. 53 with a two-stage fan was one of the first engines to have no inlet guide vanes, the first part of the engine encountered by the airflow being the rotating fan blades. When we added the third stage, we found that we could overhang this fan ahead of the front bearing like a propeller and this is today the practice on all turbofan engines. This enabled us to do away with the front bearing and all fixed radial struts or vanes upstream of the fan, and thus also to eliminate the hot-air supply previously needed to prevent ice forming on them. Inlet guide vanes were added to all early axial jet engines to swirl the incoming air in the direction of rotation of the rotating blades, in order to keep the relative Mach number between the air and blade below unity. With engines such as the Pegasus we had fans whose blades were supersonic over their outer portions, the tips running at Mach 1.3 to 1.5. We suddenly recognised that inlet guide vanes had become part of gas-turbine folklore and were no longer needed.
Another innovation was contra-rotating spools. On conventional aeroplanes the large gyroscopic forces imparted by the engine — which in effect is a spinning top — are not noticed because they can be continuously counteracted by the aircraft control surfaces. On the Harrier the airflow over these surfaces can be zero while hovering, so RCVs (reaction control valves) fed with HP air bled from the engine must be used instead. Any attempt to yaw the hovering Harrier, swinging the nose left or right, would have resulted in a powerful nose-down or nose-up tilt, because gyroscopic forces act at 90° to the axis of disturbance. Contra-rotating the two spools overcomes this, and the resultant gyroscopic force imparted by the Pegasus is exceedingly small.
When we ran the first Pegasus in August 1959 it failed to give the 13,500
lb for which we had designed it, and we had to advise Sir Sydney that the thrust would have to be limited for the first flight to 11,000 lb. He predictably went off the deep end, but even 11,000 lb was enough to get daylight under the wheels of the prototype P.1127 when stripped of non-essentials and given the minimum of fuel. In any case 11,000 lb was ample for a conventional take-off using the Dunsfold runway. When I suggested this, Sydney replied “Why should we want to do that?” I put my foot right into it. “To prove”, I said, “that the P.1127 has good handling qualities as a conventional aircraft.” “All Hawker aircraft handle perfectly,” he barked back, “there is no need to waste time on that! The first flight will be a VTO.”
Every meeting with Sydney began with a blast at his visitor or the industry in general. He once opened with “Haven’t you got a sufficient sense of grief and shame?” I tried unsuccessfully to think of a recent disaster, but he explained “You come here every Monday hawking your crackpot scheme for VTOL with your damned engine, you’ll cost the country millions!” It was just his sardonic humour and we were soon down to business.
Not Much of an Engineer Page 22
