by Damien Burke
An artist’s impression of the de Havilland GOR.339. In this view the aircraft is not burdened by the bulbous conformal fuel tanks it would have needed to meet the combat radius requirements in GOR.339; nor, for that matter, by the Red Beard store. BAE Systems via RAF Museum
De Havilland suggested a couple of additional alternative versions, a naval version and a fighter. The aircraft’s wing span was small enough to need no wing folding to fit on the deck lifts of current British carriers, though the nose and a substantial portion of the tail would need to fold. The normal AUW, even with folding mechanisms taken into account, was also within the deck limits. For catapulting at this weight, wind speed over the deck would need to be 24kt (28mph; 45km/h); combat radius would be 645nm (740 miles; 1,190km) with a five-minute combat allowance, and 920nm (1,060 miles; 1,700km) without. This could be increased to 1,205/1,425nm (1,385/ 1,640 miles; 2,230km/2,640km) if the aircraft was refuelled immediately after takeoff. For the fighter role the aircraft’s great thrust was its primary selling point. With a reduced fuel load it could go from the end of the runway to Mach 2 at 60,000ft (18,000m) in four minutes; excellent interception performance. A more economical climb and cruising speed could allow a long-range patrol of up to 20min to be undertaken 1,100 miles (1,770km) away from base (including combat time of 7.5min), or a four-hour patrol carried out much closer to home.
In structural terms, de Havilland made a bold claim that at the AUW the aircraft’s structure would form just 24 per cent of the total weight. This was said to be due to the extremely simple and efficient structure, with none of the complications introduced by a fuselage engine installation. The wing’s large central box would allow bending loads to be distributed evenly, with stiffness benefits from the engine pod arrangement. Aluminium-copper alloys would be the primary material used, with sandwich construction.
Referring to the MoS’s requirement for company mergers, de Havilland suggested building the GOR.339 under the banner of the Airco consortium being set up to deal with ‘the new BEA aeroplane’ (the DH121, or Trident, as it became). This consortium included Hunting and Fairey, and de Havilland believed that the combination of its own Sea Vixen experience and Fairey’s supersonic experience from the Fairey Delta project would be perfect for dealing with the RAF’s tactical strike bomber. (In the end Airco never got off the ground, and de Havilland became part of Hawker Siddeley in 1960).
English Electric with Short Brothers & Harland, Project 17
The early discussions on a Canberra replacement enabled English Electric to get a serious head start on meeting the requirement, and its advanced projects office began work on a design labelled Project 17 (or P.17 for short). The P.17 design effort was substantial, and English Electric looked at a large variety of configurations. Several layouts with podded engines had looked promising, giving the ability to readily change the chosen powerplant without having to redesign the fuselage, but there were difficulties in designing a suitably strong pylon/wing joint, and the effects of exhaust on the tailplane and noise fatigue on the rear fuselage were going to be serious. Finally, the control problems of losing an engine killed off these configurations. With a calculated minimum control speed of 200kt (230mph; 370km/h) on a single engine, loss of an engine clearly meant the loss of the aircraft if it happened during landing or take-off.
Several canard layouts were also drawn up. With a rear-fuselage engine installation a very short intake could be used, the straightening effect of the airflow under the wing being used to advantage, along with the wing’s bow wave. The design was easy to balance, but a big problem was introduced by the pitching moment from the large wing flaps; the longitudinal trim problem proved insurmountable, so the canard idea was abandoned. A high T-tail, as used on the new Lockheed F-104 Starfighter was tried, even though English Electric believed the F-104’s layout ‘looked all wrong’. In windtunnel tests it matched English Electric’s predictions, performing poorly, even when the tailplane was mounted on a ridiculously high and impractical fin. A low-set tailplane, as on the P.1, seemed the only possible way to proceed.
The eventual P.17 design chosen bore some passing similarities to the final TSR2, having shoulder-mounted delta wings, a long fuselage with the crew seated in tandem and engines buried side-by-side, a large vertical tail and a low-set tailplane. As previously mentioned, English Electric had issued report P/103, entitled Possibilities of a multi-purpose Canberra replacement – P.17, in early 1957. Outlining the company’s ideas, it contained an earlier version of this P.17 design, differing mainly in the intake configuration (at that time a quarter circle set well back and underneath the wing). As time had passed and discussions with the Air Ministry and MoS firmed up the RAF’s requirements, so English Electric had modified its work to match and hopefully exceed the requirements in some areas.
This early P.17 windtunnel model with ‘butterfly’ twin tail surfaces looks more like an ancestor of the YF-23 than of TSR2! BAE Systems via Warton Heritage Group
An early P.17 design with straight wings, and engines in underwing pods. The high tailplane could not be made to work, even when positioned higher than this. Damien Burke
In December 1957 English Electric and Shorts invited Handel Davies of the MoS to Warton to have a final look at their work on Project 17 before official submission, with the aim of gaining some further insight into what the Ministry really wanted from GOR.339. It was an illuminating meeting, and some unexpected new information was exchanged, affecting English Electric’s submission in several areas. Handel Davies had let English Electric know that the requirement to use existing engines was due to Treasury pressure, and that the Rolls-Royce RB.141 now looked very likely to be used for the new BEA airliner, so a military version, the RB.142, could now be considered. (An earlier meeting at the Ministry had also revealed that the Ministry’s definition of ‘existing’ was considerably more vague than English Electric’s, but still excluded the RB.133 development that the company had wanted to use.) It was thought unlikely that the high-resolution SLR would be developed in time, so an existing less-capable unit, Blue Shadow, could be used as a stopgap. Davies felt that the P.17 was too heavy at 70,000lb (32,000kg), and asked if English Electric had considered a smaller aircraft. There was still some thought within the Ministry that GOR.339 could fulfil a naval requirement, ironically as a successor to the NA.39 when it became obsolete, and English Electric’s suggestion that P.17 could at least use American Forrestal-class carriers as staging posts was welcomed. Other subjects included winged bombs, the STOL requirement and the Shorts ‘Flying Base’ lifting platform (described below). Finally the meeting moved on to the expected timescale for examination and selection of a winning design.
Davies was clearly a keen supporter of English Electric’s work, and led the company to believe that if it won the GOR.339 competition (which Davies and English Electric both saw as pretty much a sure thing, given the amount of work it had put in), and subject to the three Ministries (MoD, MoS and Air) agreeing, an order would be placed for the aircraft within a month or two, though there would be intense opposition from the Civil Service Administration to this short-cut in procedure. (One wonders why Davies ever suggested such a possible sequence of events in the face of the ‘amalgamate or die’ requirement introduced by his boss.) Should an OR have to be drawn up first(!), up to three firms would then be approached to build the resulting aircraft, but Davies wondered what would be the meaning of this in the face of the amount of work already done by English Electric. The meaning was, of course, that such co-operation was the whole point of GOR.339 from the MoS’s viewpoint.
Having been thoroughly misled as to the purpose of the whole GOR.339 process, English Electric had put a great deal more work into the project than any other firm. In the expectation of the P.17 being ordered in due course, it had advanced well beyond theoretical work into detailed windtunnel testing and assembly of sample structures as part of the preliminary design stage, something a company would not normally do until in receipt
of a design contract. The resulting GOR.339 submission was a massive three-volume brochure, explaining in great detail not only what English Electric was offering, but how it had got to that point, along with the various alternative solutions looked at, and mostly discarded, along the way. English Electric’s aviation arm was already part of a larger business with varied interests, but, recognizing that this alone would not exempt it from the edict that the contract to build GOR.339 would only be awarded to a cooperative, they had already turned to Short Brothers & Harland, with whom they had already worked on Canberra production. Shorts was working on VTOL, and its SC.1 research aircraft was shortly to fly. It had also come up with an unusual proposal to add a VTOL capability to the P.17 by means of a lifting platform, and was responsible for the P.17’s tailplane design, basing it on that of the P.1.
Another podded early P.17 design, this time with a delta wing and lower tailplane. This layout’s tailplane lacked effectiveness, as it could not be lowered into the wake of the engine exhausts. Damien Burke
One of several canard P.17 designs that were tried out. Damien Burke
P.17A
This was really what the brochure was all about: English Electric’s Canberra replacement, unadulterated by the more exotic wishes of the Air Staff, such as VTOL. The company’s succession of possible designs had gradually evolved into a layout with a 60-degree swept delta wing and a low-set P.1B-style tailplane and conventional fin. It was in its fuselage that the design really parted company with the P.1B, this being much longer and having side intakes under the wing, leading to a side-by-side engine installation in the centre fuselage with a reheat installation in the rear fuselage. The long fuselage gave an increased tail-arm compared with the P.1B, which allowed the use of a full delta wing rather than the cutout version employed on the P.1B. Full-span blown flaps assisted in providing acceptable take-off and landing performance, and roll control was achieved by differential movement of the tailplane.
A general-arrangement drawing of the English Electric P.17A of January 1958. Damien Burke
An artist’s impression of the English Electric P.17A. BAE Systems via Warton Heritage Group
A breakdown of the P.17A’s nose. Note the large equipment bay with access from below and room enough for groundcrew to stand up inside it, sheltered from the weather, while working. BAE Systems via Warton Heritage Group
Leading particulars: English Electric P.17A
Length
84.5ft (25.75m)
Height
22ft (6.7m)
Wing span
35ft (10.67m)
Wing area
610sq ft (56.67sq m)
Wing aspect ratio
2
Wing anhedral
10°
Tailplane span
22.7ft (6.91m)
Tailplane area
185.8sq ft (17.26sq m)
Tailplane aspect ratio
2.77
Fin area
187sq ft (17.37sqm)
Fin aspect ratio
0.97
Engines
2 × 14,000lb (6,350kg)
RB.142R or 15,460lb
(7,020kg) Olympus 15R
Max speed
750kt (860mph; 1,380km/h)/Mach 2.0
at altitude
Empty weight
38,250lb (17,360kg)
AUW
73,400lb (33,300kg)
Crew comfort at low level was primarily provided by the relatively small, highly-loaded wing of low aspect ratio, attention being paid to structural design so that resonant frequencies did not coincide with human body resonance. The crew sat in tandem, giving a particularly good view for the pilot and a slightly more restricted one for the navigator, who was also supplied with a magnifying periscope that could be tilted and rotated for forward, downward and rearward viewing. Unlike just about every other firm, English Electric had looked in detail at the requirements of safe escape at high speeds and altitudes. It outlined an ejection seat with automatic leg and arm restraints, and the possibility of using the instrument panel as a deflector plate to protect the occupant from wind blast in the early stages of ejection. Both seats would be mounted on ‘vibration insulators’ to improve comfort further.
The aircraft would be built of aluminium alloy (L73), which would give satisfactory fatigue characteristics though it would impose limitations on maximum speed (a steel tailplane had been investigated as part of an investigation of the requirements of flight beyond Mach 2). Stiffness was important to overcome aero-elastic problems, and the general concept was that the flying surfaces would have large skin panels unbroken by anything but the smallest of access panels. The fuselage would be a different matter, with so much equipment concentrated within it that access panels needed to be more numerous and larger. The engines, either Rolls-Royce RB.142Rs or Bristol Olympus 15Rs, would be accessed and installed through the roof of the main undercarriage bays. Shear panels would be installed after the engines were in place, completing the bottom skin of a rigid torsion box running throughout the fuselage length above the various cut-outs in the lower fuselage for undercarriage, equipment and weapon bays. The quarter-conical intakes originally envisaged gave rise to complicated boundary layer bleeds and were difficult to fair into the fuselage lines, so a simpler fixed vertical ramp intake was used instead, still positioned behind the wing leading edge.
Production components for the P.17A. The breakdown was designed for subcontracting, with major items to be manufactured at English Electric’s own factories and smaller subassemblies able to be easily transported from elsewhere. BAE Systems via Warton Heritage Group
Most of the fuel was contained in the fuselage, in six of eight tanks, with flow proportioning to ensure that each tank was emptied in proportion to its size to keep the c.g. stable. A retractable flight-refuelling probe was mounted in the port side of the nose below the cockpit floor. Unlike the P.1B, where cable ducts had needed to be tacked on to the fuselage exterior, provision was made for cable ducts within the fuselage layout, thus concentrating access to the cables and permitting easier electrical maintenance. A conventional tricycle undercarriage was mounted on the fuselage, the twin-wheel nose gear retracting forwards and the tandem twin-wheel main units retracting to the rear. A braking parachute was to be fitted, though thrust reversers had also been investigated. The main gear bays doubled as engine access areas, and the nose gear was similarly used for equipment access, along with a large underside door aft of the nose gear bay that allowed ground crew ‘stand-up’ access to the main equipment bay while sheltering them (and the equipment) from the elements.
The bomb bay was one of the larger ones among the various GOR.339 submissions, able to accommodate the full load of six 1,000lb HE bombs, the Target Marker Bomb, twenty-four 3in Mk 5 rockets in canisters or 370 2in rockets in canisters. The bomb door was of the rotating type, activated by a jack and lever at each end. Retractable locking pins along the door’s length would keep the entire assembly stiff in the open or closed positions. The entire door assembly, to which the weapons would be attached, was to be removable so that arming could be carried out on the door, away from the aircraft. More traditional loading techniques could also be used. English Electric had investigated a more conventional bomb bay with hingeing doors each side, or doors that retracted into the bay, but these would make rearming difficult or make poor use of bomb bay space.
The P.17A could successfully meet the 1,000nm sortie requirement, including a high supersonic dash and the final 200nm to and from the target being flown at sea level, and would weigh in at 73,400lb (33,320kg) for this sortie. Take-off distance would be 1,500yd (1,370m) (communications after the issue of the GOR had made it clear that the longest sortie did not have to be paired with the shortest take-off roll). To stay within the suggested 1,000yd (915m) take-off roll the take-off weight would need to be reduced to 66,000lb (29,930kg), in which case the aircraft still retained enough range to handle the 600nm sortie. Ferry range would be 3,000nm
(3,450 miles; 5,550m) clean, or up to twice that with bomb bay and underwing tanks plus buddy flight refuelling en route.
A windtunnel model showing the final P.17A configuration. Much of the P.17 windtunnel research was applied to the eventual TSR2 design. BAE Systems
Uniquely, English Electric’s brochure actually outlined a complete sortie from start to finish, describing the crew’s actions throughout. A typical attack sortie was to be flown almost entirely by the autopilot, only take-off and landing being down to the pilot. In effect, the navigator would be doing more ‘flying’ than the pilot, by entering fix points on the navigation system to ensure the aircraft was on track and allowing the autopilot to correct its course. Only if the pilot saw something of concern, such as unexpected defences or obvious errors in line-up on the final run-in to the target, would he take control and fly the aircraft manually. Pull-up into the LABS manoeuvre, bomb release and recovery to low level would all be a job for the autopilot. Non-nuclear attacks would need more crew input, with rocket and bomb attacks all requiring visual sighting. Bombing could be accomplished either by blind bombing from medium altitude (very Second World War) or via an over-the-shoulder loft attack (overflying the target, seeing it through the periscope and beginning a programmed climb into a loop to throw the bombs back towards the target).
