by Damien Burke
The layout of the TSR2 navigator’s cockpit as of February 1965. The hooded FLR display sits above the moving-map display, with the SLR display to the left, with magnifying lens overlaid. The binocular-shaped shroud to the left of the SLR display is for the downward sight. BAE Systems via Warton Heritage Group
The navigator’s cockpit under night lighting conditions. BAE Systems via Warton Heritage Group
Cockpit access was via dedicated ladders; the specification called for integral ladders for use at dispersed sites. BAE Systems via Brooklands Museum
The front cockpit of XR220 at Cosford in the mid-1990s, showing gaps where various items are missing. BAE Systems
The risk of bird strike at low level and high speed, allied with the heating expected at high speed, meant that special formulations of glass were examined for the windscreen, including formulations that did not expand much with temperature increases, such as alumina-silicate glass. Fused-silica glass was better at dealing with heat but not as strong, so a sandwich of the two was proposed. Given the amount of time the aircraft was to spend at low level and high speed, erosion of the windscreen by impact with rain, sand or insects was going to be a problem. A Hawker Hunter windscreen, for instance, would survive only 6hr of use at low level in sandy environments. The Lightning had an efficient rain-dispersal device at the front edge of the windscreen, which blasted air upwards to shatter or deflect droplets, but this was only thought to be effective at up to 350kt (400mph; 650km/h), not the 750kt (860mph; 1,387km/h) being aimed at with the TSR2.
Experience with types such as the Canberra and de Havilland Sea Vixen had also shown that low-level flight at high speeds through rain or salt spray could result in serious impairment of the view forward in flight, to the detriment of both the mission and flight safety, and salt accumulation could be experienced up to 50 miles (80km) from the coast. Tests were carried out at the RAE with Hunter and Javelin aircraft, both on the ground and in the air, to ascertain the best method of keeping the windscreen clear. These tests simulated not only low-level flight over the sea using salt solutions sprayed on the windscreen, but more exotic contaminants such as dead insects (fruit flies, in particular; smeared on, crushed on, baked on). The results showed that a weak detergent solution flooded over the screen in combination with a hot-air-blast system would keep the windscreen clear of salt and other contaminants, and also provided adequate rain clearance. Consequently such a system was destined to be installed in production TSR2s.
The windscreen, showing the twin pipes and slots for the windscreen clearance and rain-dispersal system. Damien Burke
The canopy design came in for some stiff criticism, as American aircraft of similar performance were using clear canopies that were much less restrictive of both view and access. Vickers deflected these objections, arguing that the currently available transparent materials were unable to deal with the heat of Mach 2 flight, and suggested that later in the aircraft’s life it might be possible to retrofit clear canopies when appropriate materials became available. Flash screens to block the flash of a nuclear weapon were also proposed; manually erected curtains for the canopy sides, and an automatic screen that would cover the windscreen area, linked to the weapons-release circuit so that it would be erected in time for the detonation. In the end the much simpler (and cheaper) option of smoked helmet visors was chosen.
The pilot’s canopy. Transparency quality was a recurring problem with the first aircraft, and a redesign was under way at the time the project was cancelled. BAE Systems via Brooklands Museum
The navigator’s canopy. BAE Systems via Brooklands Museum
Various types of transparencies were experimented with for the TSR2’s windscreen and canopies, and a Triplex multi-layer transparency with integral gold film deicing and demisting (with the bonus that it provided some radar signature reduction, though this was not the primary aim) was finally selected. Aircrew remarked that this gold film had an unexpected bonus in that it turned the most dismal of grey days into a cheery golden-hued occasion, until the canopies were raised! Of less use as an anecdote was the fact that the overall experience with both the windscreen and canopy transparencies was by any standards pretty poor. The first windscreen panel on XR219 delaminated and was subject to severe mottling that made the forward view into glare almost totally impossible, and had to be replaced before flight testing could continue. The replacement was little better. Ammonia-based cleaning products were suspected to be the cause, and instructions were issued to use a Lux soap/alcohol/ chalk mixture instead, to try to avoid future delamination. The laminated Perspex in the canopies also came in for sustained complaint owing to optical distortions and delamination, but Napier, which was manufacturing the canopies, had already produced a new canopy design using monolithic Perspex by August 1964, and had begun tests on it. These dragged on somewhat, and by the time the project was cancelled the canopy had yet to be fitted to any airframe. The development effort put into the TSR2’s transparencies was not wasted, as Concorde ended up using similar material.
Air conditioning
Heat was a major matter of concern from the outset. At high Mach numbers airframe heating would quickly lead to intolerable temperatures within the cockpit, degrading aircrew performance and soon causing unconsciousness and death. Therefore a highly efficient air-conditioning system was going to be a must, along with the provision of air-ventilated suits. The canopies were to have limited glazing to reduce the ‘greenhouse effect’ from sunshine into the cockpit, and also provide some protection against nuclear flash.
The air-conditioning system proved to be a serious challenge. Research on the provision of air conditioning in aircraft of similar performance included an examination of the system used in the Vigilante, and of anecdotal evidence of the system in the cancelled Avro Canada CF-105 Arrow interceptor. The latter had similar cockpit cooling needs in a slightly larger cockpit area, and ventilation at a rate of 27.5lb (12.5kg) of air per minute had initially produced ‘gales’ in the cockpit, only alleviated by modified louvres which then gave rise to a highly unpleasant whistling noise. The TSR2 was going to need 30lb (13.5kg) per minute, so this was an immediate source of concern.
The RAE at Farnborough had an existing air-conditioning test rig that had been put together for the Bristol Type 188 project, using a fuselage section from the aircraft, and this was modified to simulate the TSR2 pilot’s compartment. Electric heaters within the rig would bring the temperature up to the expected limits, while an airflow system pumped cooling air in various configurations to try to bring the temperature down to the levels acceptable to the Institute of Aviation Medicine (33°F on the aircrews’ skin). It was found that the optimum performance with an aircraft skin temperature of 125° was attained with air to 0° pumped at 16lb (7.25kg) per minute. As this level of cooling was impractical, air at 6° was tried, and found to provide acceptable cooling, though with occasional complaints of hot feet! The rig was only of limited use for dealing with heating. Temperatures as low as –20° might be encountered in slow or cruising-speed flight at high altitudes, so tests under these conditions were to be made in a real aircraft.
The air-conditioning and cooling system layout. BAE Systems via Brooklands Museum
Once TSR2 construction was under way, a complete TSR2 forebody (nose section) was earmarked for thermal environment tests (along with ejection seat testing and the aforementioned windscreen clearance tests) and delivered to the RAE at Farnborough. There, various configurations of air-conditioning piping based on previous research was tried, in order to come up with a final arrangement that could be built into the airframes then on the production line. In the end the system’s performance in the first aircraft, XR219, was only a qualified success, with air distribution considered excellent but the temperature often uncomfortably cold. This was one system that would definitely have benefited from some additional work.
Seating and escape system
Despite efforts in the early stages of the design to allevia
te the rough ride at low level, the crew were expected to be subjected to fierce vibration, and a sprung seat was tentatively mentioned as one possible way of dealing with this, along with minimum weight and bulk of personal clothing and, possibly, extra trunk restraints on top of existing harnessing. In September 1960 RAE Bedford performed some experiments with a sprung seat inside Canberra WH975. It was found that that crew comfort could be significantly improved, but no attempt was made to measure crew performance, such as their efficiency in performing tasks. Vickers Research Ltd designed a vibrating seat to try and replicate the conditions crews would experience, and this was programmed to simulate the sort of turbulence experienced at low level in various aircraft types, including the Scimitar, Canberra, P.1B and Hunter, along with predicted values for the TSR2. However, while the sprung seat helped crew comfort it was found that the crew’s ability to perform tasks was not significantly improved, so the idea of the sprung seat was eventually dropped, despite opposition from the Institute of Aviation Medicine.
It was recognized early on that existing ejection-seat technology was woefully insufficient to deal with the problems of ejecting at high speed or low altitude. An escape capsule was an attractive prospect, as it eliminated many of the highly complex timing and protection issues involved in ejection seats, but lack of full-scale test facilities in the UK contributed to this possibility being deemed beyond the timescales required.
Ejection seats, therefore, were the only real choice. At the time no data existed on the forces to which the human body could be subjected and survive, but, based on American tests using chimpanzees, these were assumed to be in the order of 30g deceleration and 1,000lb/ft2 (4882kg/m2) loading from wind blast, with tumbling of no more than one revolution per second to start with. No existing British ejection seat could satisfy these conditions beyond speeds of 540kt (620mph; 1,000km/h), and the Americans had already suffered fatalities in unsuccessful high-speed ejections. Added to that was the need to eject at ground level and boost the seat to a height sufficient for safe parachute deployment. Existing ‘shotgun’-style seats that used an explosive to fire the seat out of the aircraft had just about reached the limit of their capability, so rocket propulsion would be needed to enable the seat to clear the aircraft’s tail at high speeds. This also gave a smoother acceleration out of the aircraft and a smoother deceleration relative to the outside air.
Only English Electric considered the problems of escape at high speeds in detail in its submission, the other firms leaving it up to an independent manufacturer, such as Martin-Baker. English Electric sketched out a design for a seat using the existing gun-style mechanism but with a rocket sustainer, pop-out stabilizing fins, arm and leg restraints, an integrated lox system for high-altitude escapes, wind-blast shields and a head restraint. The parachute was to be fitted into the seat rather than worn by the pilot. The RAE agreed with all of these measures, and recommended development of such a seat. Deadlines for development were to be March 1962 for a basic seat of limited capability, 1963 for more capable seats and 1965 for the completely developed and capable items. It was accepted at an early stage that a fully capable escape system would not be provided in the first aircraft.
At the time Martin-Baker was using a Hunter for high-speed ejection-seat trials, and there was a rocket track at Pendine, but neither were capable of the 750kt (860mph; 1,390km/h) speed being aimed at for the trickiest part of the ejection envelope. (This was soon reduced to 650kt (750mph; 1,200km/h) when it became clear just how challenging the 750kt requirement was.) The Americans had been carrying out some research on high-speed ejection for their ‘Century series’ fighters, and efforts were made to obtain the resulting data to see if it could help.
The vibrating seat rig at the Vickers Research Laboratories. The lucky subject is performing basic navigation-related tasks while various performance and biological functions are monitored and recorded. BAE Systems via Brooklands Museum
The proposed Convair ‘B’ seat installation used an impractical pivoting sequence in an attempt to provide high-speed blast protection for the ejecting crew members. Damien Burke
It soon became obvious that the seat currently being considered for the American F-106, the Convair ‘B’, could be a possibility for use in the TSR2, and considerable effort went into investigating this option fully. The Convair ‘B’ used an unusual method of dealing with the stresses of supersonic ejection. It would only partly leave the aircraft in the initial stage of ejection, before rotating to lie flat along the upper surface. Thus foot pans at the base of the seat were in a position to protect the occupant from the worst of the wind blast. At this point stabilizing booms would pop out, explosive bolts would separate the seat from the aircraft, and rocket motors would fire to push the seat on a trajectory upwards and forwards to reduce the deceleration to acceptable levels. Simultaneous ejection of pilot and navigator would not be possible, as the navigator’s seat would be struck by the pilot’s seat on rotation, so a significant delay between ejection of the navigator and pilot was necessary. Thankfully the seat was not chosen, a decision vindicated by several fatalities resulting from unsuccessful F-106 ejections. The F-106 was later modified with a replacement seat that could successfully eject the pilot at low altitudes and speeds as well as high ones. It did not use the rotational method employed by the earlier seat.
The next seat to come up for consideration was the North American HS-1, in use on the new YA3J-1 Vigilante prototype. ‘HS’ stood for high speed, and the similarity of the Vigilante’s crew escape problems to those of the TSR2 were not lost on those tasked with solving them. As with the Convair ‘B’, new features such as automatic arm and leg restraints were present on the HS-1. A plate underneath the seat added lift to help push it clear of the aircraft’s tail, but protection from wind-blast at the high-speed end of the envelope relied upon the aircrew wearing full pressure suits.
Folland, which had successfully designed its own ejection seat for the Gnat lightweight fighter, was also interested in providing a seat for the TSR2. It proposed a rocket sustainer to give adequate fin clearance, angled to give some forward as well as upward thrust to reduce deceleration forces during ejection. Even so, English Electric felt that the decelerations likely to be experienced on exit and when the rocket burned out were dangerous. Folland also wanted to make use of English Electric’s windtunnels during development, which could be tricky with the company’s own projects needing them so often.
Martin-Baker, after some initial dragging of feet, put forward its Mk 5 seat, then in development for the F-4 Phantom II, as a possibility for the TSR2, suggesting some modifications to make it better able to deal with the more extreme environment. These basically boiled down to the use of rocket propulsion and a fully automatic body and leg restraint system. Compared with the HS-1 it had poorer stability, comfort and restraints, and its acceleration profile was likely to be more damaging to the occupant. However, it would be lighter, a quarter of the price, less likely to fall prey to supply problems and also, importantly, more politically acceptable.
Martin-Baker had also supplied details of its patented ejection method using a canopy split along a fore-and-aft line, which would open sideways (as used on the Avro Canada Arrow). Before the advent of miniature detonating cord that shattered Perspex canopies before ejection, this was one way of quickly getting the canopy clear without the problems of jettisoning it (and possibly interfering with the ejection of the crew member behind). This idea was of some interest initially, but the top-opening design and a variation on it, in which the two opening sections were of unequal size, were both tested on a Hunter and the pilots were unimpressed, considering there was a serious risk of mid-air collision because of the poor standard of view. The open-top design was soon ditched, and despite a small penalty in escape time owing to the requirement to jettison the entire assembly, and continued criticism of the amount of view available, more conventional clamshell solid canopies were decided on in April 1960.
The f
inal decision was due to be made in November 1960, but the Martin-Baker submission was still lacking in detail at the time (in fact it was basically a verbal submission from James Martin), so the decision was postponed. James Martin met with the Air Staff in December and pushed his case once more. He was obviously persuasive, because in January 1961 Martin-Baker was given the nod to develop an improved Mk 5 for the TSR2. The HS-1’s considerable weight penalty (a total of 390lb (177kg) more compared with two Martin-Baker Mk 5s), discomfort from the parachute (which would be worn on the occupant’s back) and larger size were factors that meant the HS-1 fell by the wayside. Most important, though, was the fact that co-operation between Vickers and Martin-Baker in matters of canopy separation and ejection sequencing would be considerably more efficient than any dealings with a US-based manufacturer; as would UK manufacture of the seat, instead of the possible difficulties of licensed production of an American seat. On the Vigilante the HS-1 proved to be an excellent seat and gained a reputation for reliability, saving many lives.
Martin-Baker Mk 8VA ejection seat
Martin-Baker took the opportunity to develop its Mk 5 into an entirely new seat, unrestrained by the need to fit it into a number of existing airframes. By August 1961 the seat for the TSR2 was no longer known as the developed Mk 5, but had been given a new designation: Mk 8VA (for Vickers-Armstrongs). As well as the problems of ejecting at high speeds at both high and low altitudes, Martin-Baker had to deal with the difficulty of ejecting an unprepared occupant, as the pilot was intended to be able to use command ejection, meaning that when he ejected, the navigator would be ejected too. Correct posture is an important part of surviving an ejection without serious injury, so the seat needed to incorporate some means of automatically restraining its occupant very quickly before it was fired. The resulting seat was a revolutionary design embodying several advances that would become part of many future ejection seats.
