|
The Comet chapter as related in DH - a history of de Havilland by Martin Sharp published by Airlife (abridged)
When the MacRobertson Trophy race was announced, there was no runner in de Havilland’s stable. High speed combined with long range were the obvious needs and everyone on the technical side was at top pressure to get the DH86 out and certified by January 31, 1934, to fulfil the Qantas contract. The production men were turning out Dragons and Fox Moths and Tiger Moths against urgent orders. And there was no point in considering a fuselage tank arrangement to give the DH86 long range when its top speed could not be stretched above about 170 miles an hour.
A clean monoplane airliner would have presented no difficulty to the designer, but there was no domestic or empire market on which to base an export sales effort - they had been over all that so many times at de Havilland.
But the thought of doing nothing at all became more and more insufferable, and when the two class prizes were announced ideas began to formulate around a small racing monoplane for the speed class only. There was great promise in the new Gipsy Six. Its bench performance and strip had been excellent. More than 200 hp could be extracted from it even on a long flight, perhaps 10 per cent more. Its low frontal area and clean entry represented a positive advantage. Propeller pitch for a fast cruise condition which would get the best range out of the fuel tankage would be critical. This was typical of the need for controllable-pitch propellers which designer Charles Walker had been forcing on every design study, but which had so far shown no commercial justification.
De Havilland himself decided "something" had to be done - the company couldn’t be absent from the race. From that moment the scene changed. In nine months a monoplane could be 'done'. A number could be built straight off the drawing board. An announcement of the intention could be put out in The Aeroplane and Flight magazines, with an undertaking to produce an aircraft that would conform to the technical requirements of the race, with a speed of at least 200 mph, to be ready by the end of September, three weeks before the start.
At least three aircraft should be built, to spread the cost, and it was a pity that there would scarcely be the resources to prepare more. The company expected to spend around £50,000 on the aircraft, but nobody was expected to pay more than £5,000 for a race plane. So the firm might expend £30,000 or more, on what? Prestige? Not much prestige would be theirs if the aircraft did not put up a good show.
But then, not much prestige would accrue from standing out of the race altogether! Besides, why shouldn't they produce in nine months a successful small aeroplane with Gipsy Sixes, to exceed 200 mph and fly 2,600-mile stages? That was not out of the way - except in this matter of propellers.
The Americans had pitch-control mechanisms coming along; one type of airscrew was said to be well ahead in flight development, and there was one French one worth considering. Unfortunately no British design was well enough advanced. The special splining of the Gipsy shaft should be easy enough. To get a certificate of airworthiness for a controllable airscrew in a matter of a few months might be marginal. If it did not materialise would a fixed-pitch formula be capable of winning the race? Could it lift the big fuel load out of the aerodromes that would be used in the tropics?
Not only prestige would come of this, Walker argued; actual experience with the variable-pitch airscrew, designing an airframe to exploit it, would be valuable for the future.
The firm’s entry was announced and work was begun on the aeroplane itself.
As a crew of two was considered the minimum, to take turns in a flight that would last three or four days and nights, it soon crystallised around two Gipsy Sixes. To take advantage of a coarse propeller pitch, cruising as high as possible, a thin wing with a fairly high loading would be necessary, and that would call for devices such as trailing-edge flaps to increase the lift at low speed, for a safe aerodrome behaviour. The crew must sit in tandem to reduce frontal area. Even so the fuselage must house all the fuel because the thin wing would have no capacity; it would need to be located ahead of the crew to get the weight on the centre of gravity. Calculations by Walker's team defined the required design formula.
First layouts drawn were on the lines of four-wire braced fixed-undercarriage monoplane, like a twin-engined version of the D.H.71 Tiger Moth racer, but these were abandoned in favour of a bold principle employing a wooden stressed-skin construction in a very clean form with a cantilever wing and retractable undercarriage.
But it all depended on the availability of a controllable-pitch airscrew, and the inquiries pointed to America and France. A team was sent to America to look for one, and was advised that the Hamilton, with forged duralumin propeller blades, was the only likely bet. Already it had made a reputation for reliability and efficient functioning.
Deciding that the Harvard, Connecticut, made Hamilton was what thye wanted, the team nevertheless carried on with their tour to get as much information out of the trip as possible. The Boeing 247 and Douglas DC2 were then appearing on the coast-to-coast airway, and going from Chicago to the west coast by United Airlines they flew in a new Boeing which had Hamilton variable-pitch propellers. Then, for a part of the journey (there were a fair number of halts in those days) they flew in another Boeing that had fixed-pitch propellers. The difference in take-off and cruising speed was impressive.
Back at Hartford for further talksthey discovered that propellers for the DH88 racer were not an easy problem. The smallest model was too large for the Gipsy - it could be cropped but that would not give a very efficient blade form, and the broomstick shanks suitable for a larger radial engine might well fail to cool the little Gipsy. The pitch-change mechanism was hydraulic, which would entail designing a pump system on the engine, and it was doubtful whether Hamilton could deliver in time for test and approval on the aircraft before October 20.
As for a manufacturing licence, Hamilton had in fact already made certain offers for overseas licences and they could speak only of a second option to de Havilland.
Reporting this situation at Hatfield served only to confirm what had been feared there. At once the order for Hamilton propellers, although not ideally suitable for the racer, was confirmed, but everyone felt that standby action was needed. Contact was made with Ratier in France, who were developing a propeller with a very simple pitch change. It was a small propeller, better suited to the Gipsy Six, having been designed for the Renault 240 hp engine of the interesting little Caudron monoplane which, with this refinement, had made a new speed record - about 220 mph.
Unfortunately the simplicity of the pitch change had to be paid for. Once in coarse pitch it could not be returned to fine pitch before landing the aircraft. Coarse pitch was maintained by a spring on a piston, and the spring was overcome by air pressure in a bag inflated by a bicycle pump, so holding the blades in fine pitch before take-off. After take-off the gathering forward speed of the aircraft caused drag on a disc mounted in front of the hub, pushing it aft and freeing the air pressure so that the spring force would change the blades to coarse pitch. Trial and error were the means for determining the optimum diameter for the disc. Once in coarse pitch the pilot was committed to landing thus, which would not be very comfortable if he decided to go round again.
The Ratiers were ordered, and with them arrived a clever French design man, Monsieur Dreptin, and much effort did he put into detail development on the spot.
The engine boys had come a long way in a few weeks. Removing metal from the cylinder-head flange had been the method of raising the compression ratio, but detonation resulted, making it necessary to change the pistons as well, so obtaining the extra compression partly from the head and partly from the piston. A cast piston already developed for special editions of the Gipsy Major was adopted, giving a compression ratio of 6.5 to 1.
Then the drag of the installed engine had to be reduced-by a modified valve-rocker and push-rod mechanism - quite a task. It made the valve cover shallower and so reduced the overall depth of the engine cowling. The carburettor and air-intake system had to be redesigned. The intake was taken from the nose of the nascelle, so avoiding the usual excrescence.
With variable pitch the increase in power over the standard engine was as follows:
Actually the effective thrust for take-off was increased in even bigger proportion.
To get flight development of the engine without waiting for the DH88 racing aircraft, a Leopard Moth G-ACHD was adapted, and John E. Walker, engine installation designer, got to work on this. By May it was flying with the Gipsy Six R.
Type approval was secured during July and August by tests comprising 60 hours' bench running followed by 20 hours' propeller testing, but test flying in the racer could not begin until September 8, and there was then only time for cooling trials and checks on fuel and oil consumption. Time was really getting short.
Most of the entries for the race were made near to the closing date. As well as the Comets, two Airspeed machines and one unnamed type were announced, followed by two Continental entries published later in May, and a Speed-model Fox Moth.
American airliner and racer entries had been expected from early in the year, and before the end of May KLM filed their entry for a twin-motor Douglas airliner with K. D. Parmentier as captain. Later J. J. Moll was named as co-pilot.
After the closing date, June 1, the full list was published - 64 entries, 21 of them from the United States, 16 from Britain, seven from France, five from Holland, four from Australia, two each from Italy, New Zealand, Sweden, one each from Denmark, Germany, India, Irish Free State, and Portugal.
The KLM. Douglas and the expected Boeing 247D, both with variable-pitch propellers, were looked upon at Hatfield as the most formidable opponents in the speed race as well as probable leaders in the handicap. Both had the background of airline operation in America. The KLM crew were airline pilots who knew much of the route, and would be working with familiar ground services. What a pity, they said at Hatfield, not to be doing - not to have already done - a transport on the Comet principle, with the clean-entry engine.
The Comet airframes were largely built at busy Stag Lane, in what was to become the plant department. The essence of the thin wing was to build up the top and bottom skins by wood planking like a boat, to get a very thin, stiff structure, the skins serving as tension and compression flanges of a spar having three webs. These were not strong, could not be called spars; they were spacer webs. A beautiful thin tapered wing resulted.
The fuselage was wooden, semi-monocoque, with flat sides and curved deck and keel. Longerons carried little stress - skins were stressed highly. Manual retraction of the undercarriage was adopted for quickness, neither hydraulics nor electrics had been sufficiently developed. A tailskid was preferred because a wheel might give shimmy and swing trouble. The skid was lighter, a spare could be carried, it shortened the run and saved the wheel brakes.
This Comet racer was the first British aircraft which combined the three features: variable propeller pitch with a flapped wing and retractable undercarriage.
The first Comet finished was G-ACSR, for Ken Waller and Cathcart Jones. Hubert Broad was ready for it and secrecy was kept about the expected first flight, and only the inner few attended at Hatfield early on the morning of September 8. Broad took off to the south-west. Fuel load was light and take-off run short. A wide circuit took him out of sight for a few minutes and when he approached from the north-east he appeared to be landing-with undercarriage up. Harry Povey whipped the spare wheel off his Morris car and held it aloft. But Broad wasn't landing. He had not been sure whether he had exercised the handwinding gear an odd or an even number of times, having as yet no indicator, and he flew low over the field to try to tell, from the shadow in the low morning sunlight, whether his undercarriage was up or down. He was also glad of the spare-wheel signal, and he then came round again to a perfect landing. The wing-drop on landing was something to beware of; it resulted inevitably from the tapered-wing characteristic. But the aircraft flew well and soon showed that it would attain estimated performance.
Accurate fuel and oil consumption tests were extremely difficult, in fact were not possible in the time available and in the weather that prevailed. They called for either a very long flight or a shorter one at 10,000 ft. between exactly known points.
Competing machines were due at the starting aerodrome on Sunday, October 14, by 4.30 p.m. The racing pilots, with the exception of Amy Mollison, had flown Comets at Hatfleld. Cathcart Jones brought the first Comet G-ACSR into Mildenhall in a strong and gusty wind, landed well and then dropped a wing a little in a gust. Soon afterwards Scott and Black landed in G-ACSS and later on the Mollisons approached, went round again twice (in high pitch inevitably) and landed nicely the third time.
The night before the race there was another last-minute decision - the Comet oil tanks must all be larger. This came when the Mollisons' and Scott and Black's machines were already tucked away in the hangar with cowlings locked and wired. Off came six sets of cowlings, new oil tanks were made and fitted.
On the day the three Comets were out and ready early, polished by a band of DH apprentices, and they were among the first to toe the starter's line.
They were off. Here the industrial pen must retire. Only the pilots, mechanics and organisers could write the story of the next three days and nights, but the flight to Australia was summed up in Charles Scott's immortal phase, at Melbourne: "It was lousy - and that's praising it." In the event 20 aircraft set out on the race, five of which were of de Havilland make, and nine of which had de Havilland Gipsy engines.
The best story of the Comets in the race is that of DH engineer Frank Halford. He was in London’s Regent Street on the Monday morning when Scott and Black were due at Darwin. A bus conductor happened to say to him 'Bad luck about Black and Scott. The papers say they've retired.' Halford dashed into his office in Golden Square and found a Press Association man there waiting to interview him. He said the Press Association had a telephone line actually open to Darwin and promptly put it at the disposal of Halford who was astounded to find himself speaking to Scott, loud and clear, within a few minutes.
Scott explained they had crossed the Timor Sea with one engine throttled right back because the gauge showed no oil pressure. Suspecting the cast pistons, which had been used only because they had made the mistake of believing that they would not be able to get forged pistons in time, Halford asked Scott whether they had a good mechanic there. 'Yes,' said Scott, 'an experienced D.H. man from Sydney.' 'All right,' said Halford, 'what you must do then is to fit standard pistons. Then use both engines for take-off and throttle back the doubtful engine when you are in the air.'
Scott said they would follow this advice. Yet they were on the ground at Darwin only 2 hours 26 minutes, and shortly afterwards Halford saw the London placards 'Scott Resumes Race'. It seemed as if one was holding hands with the other side of the world. Yet he wondered how they could have changed pistons in so little time. But those pistons were not changed. The 'experienced man from Sydney', C. H. Tuckfield, describes the incident as follows:
One of the pilots told me at Darwin that midway over the Timor Sea the oil pressure in the port engine had failed. He switched off and continued the rest of the flight to Darwin on the one engine. He reckoned that the race was lost, as the defective engine had seized. I turned the engine over and found that it was free, and had four fair-to-good compressions and two very weak ones. On inspection I found no oil leaks and the tank half full of oil. I then examined the Autoklean pressure oil filter, which I found completely clogged up and undoubtedly the cause of the oil-pressure failure.
After cleaning the filter and its housing, and changing all the spark plugs I got one of the pilots (who were being given shower baths, beer and chicken) into the cockpit and started the engine and found the oil pressure O.K. After cowling up the engine it was given a run up and ran O.K. except that it was down in revs. as compared to the starboard engine by 150 r.p.m. At the time I put this down to the two very weak compressions. Anyhow, as the engine was running without undue roughness they took off, and I wired Charleville (their next and only landing before Melbourne) to have two cylinder heads ready as I suspected burnt-out valves, etc., in two cylinders. The delay at Darwin was about two and a half hours. About one hour after the Comet had departed it occurred to me that the lack of full rpm on the chocks in the port engine was possibly due to the propeller being in coarse pitch. Just before starting the engines for the take-off, the pilot produced a motorcycle-type pump and applied it to the cone of the propellers, which were both fitted with Schrader-type valves. Whilst pumping the port propeller he had said something about it not responding. I knew nothing about these propellers but concluded that the action of the pumping held the blades in low pitch for the take-off and early part of the climb.'
Later Halford had had a press message 'Scott Returns to Charleville'. The engine trouble must have recurred. Then another message that they had gone back because they had forgotten their maps!
Afterwards Halford learned that after Darwin they had throttled the doubtful engine most of the way, but they let it go on the last leg.
|
