by Damien Burke
The crew were seated in tandem, with large Vitreosil fused-quartz transparencies offering excellent views sideways and downwards. Only the navigator was to be provided with a sprung compartment to improve ride comfort. Behind the crew compartment was a large equipment bay, which also housed the SLR aerials. Aft of this was the ‘bomb cell’, sized to accommodate all required stores including the single Red Beard store or up to four 1,000lb HE bombs (if six were required, the other two would need to be carried on external pylons). Double doors were used; a single set closed the bay off during flight, with an inner set of doors normally retracted against the cell walls holding the stores. These would fold outwards to lower stores into the air stream before release and also close the cell, eliminating the problem of buffeting experienced with so many bomb bay designs. As with so many other submissions the bomb cell, or weapons bay, could accommodate a dedicated reconnaissance pack including cameras and linescan equipment (though the submission was bereft of any details on this), or a buddy refuelling pack (again, no details were provided). The carriage of rockets was complicated by the foreplane position; they would need to be angled down to clear it when fired (one can only imagine the difficulty of sighting the target). Bristol was the only company to deal with the problem of slow burning or late ignition of rockets. If this happened after retraction the resulting mess could be impressive, so the company proposed a metal grid in front of the pack to stop the rocket leaving the launcher, while the exhaust would be vented through blowout panels in the rear part of the bomb-cell doors.
Fuel tanks occupied much of the space above and behind the bomb cell, with the Doppler radar in the fuselage underneath the rear tank. Further fuel was contained within the thick wings, which also housed the main undercarriage units. The undercarriage used a conventional tricycle layout, with the main gear retracting sideways into the wing (rearward retraction having been investigated originally). Ground clearance gave rise to particularly long main legs, a four-wheel bogie enabling the use of smaller, higher-pressure tyres than would otherwise be the case, giving an overall LCN of 28. The legs were shortened by 20in (50cm) during retraction. The nose gear retracted rearwards and used the same wheels and tyres as the main gear.
The wing’s low aspect ratio gave good comfort at low level, inherently high overall stiffness and thus good fatigue life, even with relatively lightweight construction. Conventional skin-stringer construction was to be used, using light aluminium alloy for the most part with limited use of stainless steel and titanium in the hot areas around the engines. Bristol’s preferred engine was the Bristol Olympus 22SR (SR standing for simplified reheat); this was the only submission to choose this engine. Other engines considered and rejected for weight and performance reasons were the Olympus 15R and either two or three RB.142Rs. The intakes were simple wedges, well matched to the Olympus 22SR engine, but auxiliary inlet doors were provided on the upper surfaces to provide extra airflow at slow speeds, and small spill doors for higher-speed flight. Bristol’s simplified reheat system included a simple convergent-divergent nozzle, the divergent portion sliding forward to reduce the area. While this simplified the mechanics considerably, it led to significant reductions in the amount of thrust available when reheat was not being used. This was a serious issue, but Bristol dismissed it as ‘not an embarrassment’ owing to the ‘operational characteristics of the aircraft’. Clamshell thrust reversers mounted above and below the nozzle considerably reduced landing distances.
On the navigation front, Bristol considered the development of an INS unlikely by 1964 and proposed a simpler gyro system combined with Doppler. The FLR was to be used only for target identification and ranging, with SLR (X or J band) for mapping. Unusually, linescan was to be a permanent fit on the aircraft, with live presentation of the results available to the navigator for navigation purposes, as well as being stored and transmitted to ground stations for intelligence purposes. Only the storage and transmitter equipment would be held within the optional reconnaissance pack. Three permanently fitted cameras would also be carried in the nose. Nighttime photography would require a reconnaissance pack to be carried, as it contained the high-intensity flash equipment. Bristol did not consider a reconnaissance radar necessary.
The all-important performance of the aircraft was a rather mixed bag. Bristol cooked the books somewhat in its sortie profiles, in which the supersonic burst (intended for low-level evasion) was carried out at high altitude on the cruise towards the target (not even within the last 200 miles (320km)); an entirely low-level profile was also to be entirely subsonic, initially at Mach 0.85 and only rising to Mach 0.95 closer to the target. As a result, while the combat-radius requirements were met, the sortie profiles were unrealistic and invalidated the figures. Only in the ferry sortie could it be legitimately claimed that the aircraft met the requirements. To include a supersonic burst where it would actually be useful, Bristol predicted a rise in aircraft weight of more than 10,000lb (4,500kg) (for stronger structure and extra fuel), which would have a deleterious effect on take-off and landing distances. The nuclear LABS attack in high ambient temperatures would result in the aircraft slowing up so much that loss of control was a possibility. To combat this it was suggested that water injection be used, to permit higher engine output, or that the aircraft accelerate to Mach 1.2 before the pull-up.
A brochure model of the Bristol Type 204. An attractive but badly flawed design, it displayed some hints of Concorde in its striking lines. BAE Systems via RAF Museum
Unfortunately Bristol had made a serious error in its calculations of supersonic wave drag, and did not realize this until it was told so by the RAE in February, after the submission had gone to the Ministry. This was a disaster for the company’s chances. The resulting performance impact was massive, resulting in an aircraft incapable of attaining supersonic speeds. Rather than accept this and lose out on the chance to get the GOR.339 contract, Bristol quickly redesigned the wing. The leading edge was straightened slightly and its apex moved forward 15in (38cm) (retaining the same overall area). It became thinner in section, and the undercarriage hinge could no longer be accommodated in the original position. The undercarriage was therefore changed to forward retraction, extra room being made by moving the intakes forward 13in (33cm). The changes lost 780gal (3,545L) of fuel tank space; to regain some of this the engine accessories were relocated below the engine, enabling rearward removal of the engines rather than taking them out through the nacelle floor. Thus part of the rear nacelle could now hold fuel, but only 230gal (1,045L). The forward fuselage tanks were extended by 14in (35cm) to give them another 200gal (910L) capacity, but that still left a deficit of 350gal (1,590L), so to fly the 1,000nm sortie the aircraft now needed to carry a pair of 200gal (910L) drop tanks, to be dropped immediately they were empty. Maximum take-off weight rose by 1,000lb (450kg). With this error added to a succession of vague and hopeful statements scattered through the brochure (‘It is believed that…’, ‘…when this [experimental work] becomes available…’, ‘…it should be possible…’, and so on), it was clear that Bristol’s work on the design was at a very early stage. The company had also failed initially to submit any details of its preferred partner company or companies, though it later indicated that Short Brothers & Harland was the likely partner.
de Havilland Aircraft Company GOR.339
De Havilland did not waste any further effort improving its developed Sea Vixen idea, but came up with an entirely new design to try to satisfy GOR.339. As with just about every other submission, the company was keen to keep the aircraft as small as possible, not only to keep down cost, but for a range of reasons including minimizing maintenance, storage and infrastructure requirements; and also keeping the radar signature low and so reducing vulnerability. Because VTOL by jet deflection, direct jet lift or ducted-fan lift, was too risky, de Havilland discarded such ideas very early in the design process. Its chosen layout was an aircraft of fairly conventional appearance with: a thin fuselage; a shoulder-mounted
variable-incidence cranked delta wing, under which were hung the engine pods; and a conventional tail with a mid-mounted tailplane.
The choice of a variable-incidence wing was unusual for a British aircraft company, though it had been briefly considered by some of the other companies submitting to GOR.339. With the wing tilted so that the angle of attack was greater during the takeoff run, not only was the aircraft closer to the take-off attitude to start with (with the benefit that the undercarriage legs could be shorter, as the required pitching moment was reduced), but it also enabled the wing-mounted engines to provide a significant vertical thrust component. In addition, the wing was provided with blown flaps and ailerons and a full-span drooped and blown leading edge. The all-moving tailplane’s position and span was decided by the requirement to keep it clear of the engine exhausts; de Havilland had not carried out any windtunnel tests to ascertain if this position was actually acceptable, however. The tailplane also had trailing-edge flaps that would come into use at low speeds.
Leading particulars: de Havilland GOR.339
Length
67.5ft (20.57m)
Height
17ft (5.18m)
Wing span
34ft (10.36m)
Wing area
440sq ft (40.8sq m)
Wing aspect ratio
2.63
Foreplane area
228sq ft (21.18sq m)
Fin area
183sq ft (17sq m)
Engines
2 × 14,000lb (6,350kg)
RB.142R (22,400lb
(10,170kg) with reheat);
optional 1 × DH Spectre
15,000lb (6,800kg) rocket
Max speed
Mach 1.3 @ sea level,
Mach 2.0 @ 50,000ft
(15,200m)
Empty weight
26,790lb (2,490kg)
Max AUW
60,400lb (5,610kg)
The nose radome was designed to be large enough to accommodate a 30in (76cm) diameter radar dish, but the company said this could be changed easily to cope with alternative equipment. Crew comfort was judged to be good, owing to the high wing loading and hydraulically sprung and damped seats, with a vague prediction that fatigue would be one sixth of that experienced by Gloster Meteor crews. Tandem seating with vertical staggering ensured that both crew members had a good all-round view, though no downward view was provided for the navigator; a periscope was suggested instead. There was also a suggestion that the navigator’s task of ‘groping’ for various switches could be made easier by giving him a master hand-control; an early stab at the hands-on-throttle-and-stick (HOTAS) concept (minus the throttle, naturally!).
The bomb bay would hold the required single Target Marker Bomb, and also accommodate some alternative loads including two 1,000lb HE bombs, a large rocket pack (containing either eighty-five 2in or fourteen 3in rockets), a reconnaissance pack containing radar and photographic equipment or a 200 or 400gal (910 or 1,820L) fuel tank. With a narrow fuselage, the required bomb-bay dimensions resulted in some localized bulging in this area, which also doubled as strong points for the external attachment of bulbous 375gal (1,700L) conformal fuel tanks or small pylons to carry a single 1,000lb HE bomb on each side. Hardpoints under the wings enabled the carriage of 275gal (1,250L) drop tanks (these could be carried while the aircraft was supersonic, whereas the fuselage-mounted drop tanks would be jettisoned while subsonic) or 1,000lb HE bombs. Thus the maximum load of six such bombs would have them evenly distributed between the bomb bay and two external locations, and de Havilland recommended the use of the low-drag Mk 83 weapon being developed for naval use. The proposed large rocket pack was mounted on a trapeze within the bomb bay, and when lowered into the air stream, there was limited clearance to fire the rockets through the gap between the two dive brakes, if these were extended. It was hoped that airflow disturbance from the dive brakes would not affect rocket firing.
Three F.95 cameras were a permanent fit in the nose, with 12in lenses, or 4in if necessary. The reconnaissance pack was to consist of either three 70mm cameras (no such camera existed at the time) with a pair of powerful photo flash units rather than a large array of photo flash cartridges, or two F.96 cameras with 24in or 6in lenses, or three F.117 cameras modified with 18in lenses. Also in the pack would be 12ft (3.65m)-long SLR with moving target indication, recording to photographic film for later processing at home base, and linescan (which the company thought could be used to give the navigator a downward view to verify that his photographic coverage was correct; slightly missing the point that linescan was a useful reconnaissance tool in its own right).
A general-arrangement drawing of the de Havilland GOR.339 of January 1958. Damien Burke
For in-flight refuelling, a retractable probe was to be mounted in the nose. De Havilland gave extremely sketchy and vague details of a buddy refuelling system, mentioning only that the bomb bay could accommodate a 200gal (910L) fuel tank, and the drogue would be located in the tail tunnel (which could otherwise accommodate the rocket used for short take-offs), with the hose reel within the rear fuselage. This would have presumably required the permanent fitting of part of a large portion of the buddy pack equipment. Even assuming this weight penalty was acceptable for the non-buddy-refuelling roles, the company’s claim that it would be possible to change to this role in ‘minutes’ was perhaps overly optimistic.
The ever-popular RB.142R was the engine of choice, mounted in pods under each wing at a distance from the fuselage that would hopefully avoid the need for any strengthening of the fuselage skin and structure to cope with acoustic fatigue. The single-engine-failure case was, remarkably, felt to be no worse than that of a single-engine aircraft, basically involving throttling down the other engine if the aircraft was not at a safe single-engine speed already. De Havilland was not sure of the final form of the intakes, but sketched a simple fixed intake with a conical centre-body and no auxiliary inlets or bypass. A convergent-divergent nozzle was considered unnecessary, it being believed that the aircraft would spend most of its time at subsonic speeds.
The undercarriage comprised a twin-wheel nose unit, a four-wheel main unit in the central fuselage and small twin-wheel outriggers mounted on the undersides of the engine pods. All units retracted rearwards. A large ventral fairing under the tail housed a small tailwheel and a heavy-duty arrester hook. As with the Sea Vixen, the dive brakes were mounted under the fuselage near the wing leading edge to give minimum trim change on deployment, and doubled as access doors to some of the equipment.
Ferry range was 3,390nm (3,900 miles; 6,270km); however, combat radius was a problem. While the 600nm sortie was possible on internal fuel only, the 1,000nm sortie could be carried out only with external drop tanks. In terms of runway performance the design did rather better. With the aircraft loaded for the 600nm mission, the take-off run to 50ft (15m) would be just 660yd (600m); for the 1,000nm mission it would be 1,090yd (1,000m). For shorter strips the best answer would be the use of a portable catapult and arrester gear, and the company had investigated the use of such a system in co-operation with RAE Bedford. Alternatively, provision was made for a jettisonable liquid-fuel rocket of 15,000lb (6,800kg) thrust, mounted in the tail. Climb to height after take-off would be rapid, to say the least. De Havilland predicted that it would take less than 1.4min to ascend from 1,500ft to 50,000ft (460m to 15,000m), having accelerated to Mach 2 in the process. If the climb was continued the aircraft could be at 80,000ft (24,400m) less than a minute after that, though speed would have dropped, and above 60,000ft (18,000m) the aircraft would basically be ballistic. Such impressive performance was the company’s answer to the danger of the nuclear LABS attack. Rather than diving back down to low level, the climb would simply be continued to get away from the nuclear detonation (further suggesting that, coupled with launch delays, the aircraft could out-accelerate surface-to-air missiles!). De Havilland suggested a landing technique that entailed deploying the braking parachute cluster while on the approach
. To overcome this extra drag the engines would need to use higher power settings, which would make available more thrust to be tapped off for flap blowing, and also put the engines into an rpm zone where rapid spooling-up would be easier (when going around after an aborted landing, for instance). The braking parachute’s drag would also be available for the entire ground roll. Landing distance, with braking parachute, would be less than 800yd (75m).
While other companies put some real effort into the question of bombing accuracy, de Havilland’s treatment was cursory, meriting just a few paragraphs, whereas details of land-based catapult gear occupied several pages. The proposed navigation system was dead reckoning using Doppler radar, with radar identification of a prominent feature near the target to correct the dead-reckoning position before attack. Targets of opportunity could only be attacked visually, using the over-the-shoulder loft attack. When it came to terrain clearance the company was of the opinion that no suitable radar equipment was available or likely to be developed in time, and the best it could offer was flight ‘near’ the local safety height (1,000ft (300m) above the highest point in a grid square), based on Doppler radar contour height detection, with the pilot given height steering commands either by a director on a head-up display, or by the navigator talking to him! Radar ranging for rocket attacks was considered unnecessary, as it gave little or no improvement in accuracy compared with a well-practised pilot using a simple fixed sight. Anything other than a basic autopilot and stabilizer was considered unnecessary, and the bombing manoeuvres were to be hand-flown owing to ‘safety problems’ and the fact that pilots were quite proficient at this sort of flying anyway. The autopilot could handle flying to particular altitudes, headings, pitch angles, speeds, etc, and even land the aircraft automatically when linked to the ILS, but this was a world away from the fully automatic flight control systems of other submissions.