The story of the TSR.2, and specifically the cancellation of the TSR.2, is perhaps the one most likely to bring heat to the heart of the British aviation enthusiast. It is still regretted as a dark moment in the history of the UK, a magnificent lost opportunity.
The story of TSR.2 fits with a number of important British narratives. For a nation that never hesitated to remember or even celebrate a failure or a defeat, TSR.2 provides a significant failure in aeronautical space. For a nation in which many people that can still feel some nostalgia for the days when its empire controlled much of the world, the cancellation of TSR.2 coincides with the abandonment of the last global imperial ambitions. The projection of power “East of Suez” would no longer be on the agenda.
In addition, the prevailing narrative of British aeronautical engineering since WWII has been one in which brilliant engineers and private enterprise were stymied by foolish officialdom and politicians. That narrative is historically incorrect and even harmful (as collaboration between the state and private enterprise was and is important in aircraft development) but it serves to conveniently and neatly assign blame for the failure of TSR.2. The real history was considerably more complex, with plenty of blame to go around.
The tale of the TSR.2 began in March 1957, when the Air Staff announced its operational requirement GOR.339. The GOR.339 document (which later became OR.339, and after review OR.343) called for a tactical strike and reconaissance aircraft, to enter service in 1964. The GOR.339 aircraft was to be able to operate in all weather conditions, without relying on external systems (such as navigation beacons) to complete its mision. The mission pattern assumed was an attack at very low level, at high subsonic speed on most of the way to the target, with a supersonic dash over the target. At high altitude, the aircraft should be able to fly at Mach 2 and to “supercruise”, although this term had not been invented yet. Despite being a tactical strike aircraft, it had to be able to reach targets up to 1000 nautical miles (1850km) away. The load would be a tactical nuclear free-fall bomb, or conventional bombs. A crew of two, pilot and navigator, was to fly the aircraft. A good take-off and landing performance was required.
From the start, some fog surrounded GOR.339. The introduction to this document stated that the requirement contained the “broad outlines” of the project. It also stated that the aircraft would be able to fly missions to the USSR, with in-flight refuelling, and thus added an unofficial strategic dimension to the GOR.339 requirement. This suggests strongly that, although the GOR.339 was officially intended to replace only the Canberra, it was also being regarded by the Air Staff as a possible replacement and enhancement for the V-bomber force. After the cancellation of the Skybolt air-launched ballistic missile, it had been decided to transfer the nuclear deterrent to the Royal Navy and its Polaris submarines. But the RAF wanted to retain a nuclear capability. The GOR.339 was considered to be suitable for a condition of “limited” nuclear war.
Proposals were submitted by Avro, Blackburn, Bristol, Fairey, English Electric, Handley Page, de Havilland, Short Brothers, and Vickers. The proposal from Blackburn was a modified version of the NA.39 Buccaneer naval attack aircraft, which it was going to build for the Royal Navy. Although the NA.39 did not meet the requirements of GOR.339, the Secretary of State for Air suggested that the Buccaneer could be required as an interim replacement for the Canberra. The GOR.339, he suggested, would not enter service before 1968. But this idea was dismissed by the RAF: The Blackburn NA.39 was unsuitable because it was subsonic, had a too short range, required a too long runway, and had an insufficiently advanced bombing system. The RAF would rather prefer the keep the Canberra (subsonic, and with a primitive bombing system) in service a few years longer! Obviously, there was a fear that adoption of the NA.39 as a palliative would lead to abandonment of the GOR.339. Nobody can have expected that in the end, it would be the other way around. The RN of course liked the idea of the RAF accepting a derivative of its NA.39, because this would reduce the costs. The First Sea Lord advocated the NA.39 as much superior to the Canberra, and available in 1961. He even offered to share the ordered pre-production aircraft with the RAF. His suggestion was again rejected by the Air Ministry. As the Air Ministry would state later, for the purposes of the RAF the NA.39 would be obsolescent by the time it could enter service!
On 17 June 1958, the Deputy Chief of Air Staff defended the OR.339 in the Defence Research Policy Committee (DRPC). AM G. Tuddle argued that the OR.339 was also vital to the aircraft industry, the position of the UK in the NATO, the Army, and (surprise!) the Navy. He added that it would probably be “the last military fighting aircraft developed in the UK.” This statement was probably based on the 1957 “Defence White Paper”, that rather boldly had declared all manned aircraft obsolescent in favour of missiles. GOR.339 was making an exception on this rule.
According to Tuddle, the development of OR.339 would cost 35 million, and it would be ready in 1965. The representative of the Navy correctly dismissed these predictions as overly optimistic, and advocated NA.39 again to fill the gap between the Canberra and OR.339. It was even possible, he suggested, to delay OR.339 by two years. However, when it was later suggested that the RN could adopt the aircraft developed for the RAF, the Admiralty was quick to answer that NA.39 and OR.339 were not in competition and designed for different roles.
On 10 September, the Minister of Defence authorized a production order for the NA.39, and indicated that he would not yet set a limit on production. The suggestion that in the future some surplus NA.39 might be available upset the RAF, and it began putting pressure on the Minister, H. Watkinson, to decide on OR.339. In November the Ministry of Defence gave authorizaton to ask the Treasury for the money for the development. But then the estimated development costs were already up to 70 million, twice the amount mentioned in June. When the Air Ministry asked the Treasury to allocate 62 million to them, with an option on an additional 15 to 25 million, this provoked a sharp reaction. The response from the Treasury amounted to a thinly veiled accusation that the RAF had deliberately underestimated the costs to safeguard the programme, and suggested to look at the NA.39 option again. The Treasury also complained that the MoD asked for large amounts of money, while it was not yet itself convinced of the need for the OR.339!
On 1 January 1959, a press statement was issued, using for the first time the name TSR.2. The acronym TSR indicated the aircraft’s function of Tactical Strike / Reconaissance, although the first documents described it as Tactical Support / Reconaissance. TSR.1 was supposed to be the Canberra. It is also possible that someone had remembered the original TSR.II: The prototype of the Fairey Swordfish! The air staff requirement was revised and now became OR.343. It was required that one squadron would be fully equipped at the end of 1965. Thus began the long fight of the RAF against delays of the TSR.2 program. The 1965 goal could only be met, it was agreed, if no changes were made to the specifications during development and by a very concentrated effort.
There was still anxiety about the very high costs, and on 7 March the Chiefs of Staffs were asked to review the need for the TSR.2 again. They submitted a paper in May. In a later meeting, Minister of Defence mentioned that there were two cures to the cost problem. The first was selling the TSR.2 to the USAF, which he intended to try during his next visit to the USA. This was a reasonable hope, because the USAF did already have a licence-built version of the Canberra in service, the Martin B-57. The other option was giving the TSR.2 an extended strategic role by fitting it with a missile. The latter option would of course not reduce the cost, but it would make the TSR.2 more useful, and thus help justifying the expenditure. But the Minister’s suggestion that this could be a weapon “like Bulpup” revealed his lack of techical knowledge: The US Bullpup had a range of about 10km. The development of a new cruise missile for the TSR.2 was under consideration. Blue Steel, the missile carried by the V-bombers, was much too large.
Contracts, Management, and Cancellation
The first TSR.2 contract was placed on 3 June 1959, but this was only a development contract, running until 30 July 1960. It was not until the autumn of 1960 that all branches of the government finally agreed with the development of the TSR.2, and the contract was signed on 6 October. The Treasury agreed with an expenditure of 61.7 million pounds. A Progress Review Committee, a Development Progress Committee, a Management Board and a Steering Committee would supervise the TSR.2 development. In the long run, these committees would show themselves complete incapable of controlling the TSR.2 project. (To those of us who have professional experience with projects controlled by a hierarchical structure of committees, this is entirely unsurprising. If you are ever assigned to a project with a Steering Committee, run!)
After the 1959 elections, Duncan Sandys had begun to reorganize what was left of the British aviation industry. There was a feeling that there were too many and too small companies. (The financial weakness of some of them was due to Sandys’ own 1957 Defence White Paper.) Not only were these thought to be uncompetitive, it was also an expensive situation for the taxpayer. The creation of the V-bomber force had resulted in contract for two interim types (the Sperrin and the Valiant) and two definitive types (the Victor and the Vulcan) at great expense. In the future there would certainly be fewer, but more expensive development contracts, and the creation of larger conglomerates was logical. To achieve this reorganisation, the government made it a condition in the contracts for GOR.339 and other projects, and even cancelled orders for a batch of Victor bombers because Handley Page refused to cooperate.
Hence the contract for GOR.339 was placed with a conglomerate of Vickers-Armstrong, English Electric, and Bristol Aeroplane. BAC was created speedily, and a final agreement was signed in in June 1960. The three merging companies had 40:40:20 shares. At the engine side a similar merger occured: The Olympus 22R Mk.320 engines would be developed by Bristol-Siddeley Engines, also recently created by a merger of Bristol Engines and Armstrong Siddeley. This choice of engine had not been the one of BAC, which had preferred Rolls-Royce engines. It was easier to sign a formal merger than to achieve integration between design team and work forces that were used to compete sharply with each other. Not everyone was enthusiast about the enforced merger, and the cooperation within BAC often lacked effectiveness.
The allocation of the contract was by no means the end of the political dicussions. On 27 March 1962 the MoD called for a new review of the TSR.2 project, because estimates of the development cost had risen to 137 million. The DRPC decided that the TSR.2 development could continue, but only on condition that the project would be reviewed again when new cost estimates became available. In July 1962 the Managament Board for the TSR.2 project heard more bad news: The first flight date was being shifted back to August 1963. The reason was highly embarrassing: BAC had failed to produce a number of drawings, and some parts of the prototypes had therefore not been manufactured!
In December 1962 the cost estimates were up to 175-200 million, with a rather large margin of error. The estimated unit cost had risen to 2.1 million, and the predicted service entry of the TSR.2 had slipped back to late 1967. The Management Board was highly dissatisfied, and suggested that BAC could improve its own management of the programme. The director of BAC, George Edwards, agreed and made some suggestions. BAC was also willing to accept “incentive clauses” in their contract. Officials began to take a cynical view of the cost of the TSR.2 program, but in February 1963 the DRPC agreed again with the continuation of it. Suggestions were made by the MoD and the Admiralty that the number of TSR.2s could be reduced to 50 or 60 instead of 138, but the RAF and MoA rejected these. In June the estimates were up again, to 197-222 million.
In June 1963 an order was placed for eleven pre-production TSR.2s, to reinforce the development batch of 9 aircraft. Important was also a visit by an Australian delegation, because the RAAF wanted 24 strike aircraft. (The RAAF would later select and received the F-111.) Confidence was now growing, and in October the MoA gave a press release on the TSR.2. The reaction from the press was considered favorable by the MoA, but “inevitably there was a tendency to dwell upon the cost.” The RAF began making plans for the training of TSR.2 crews.
In December 1963 the TSR.2 program was debated in the House of Commons. The Times commented sarcastically that there were “rumours of cancellation”, and official denials of this “strenuous enough to spread panic through the arms industry”. There were also questions about the strategic role of the TSR.2, because of an earlier decision that the RAF would abandon this role to the RN and its Polaris missiles. In reply the MoD stressed that the strategic role of the TSR.2 was only a bonus, not the reason for its existence.
As far as the TSR.2 was concerned, 1964 was just more of the same: Rising costs, slipping time schedules, and disagreements between government officials and BAC. It was now obvious that the RAF, MoA and MoD had lost their confidence in the management of BAC. The TSR.2 had still not flown in July 1964, but cost estimates were now at 240 million, with an unit price of 2.3 to 2.8 million. It made its first light on 27 September.
In October 1964, a Labour government came in power. It wanted to reduce the defence budget to 2000 million a year by 1969/1970. And because TSR.2 was the most expensive development contract running, it was a logical target. The R&D costs were now estimated at 272 million, and the production costs for 158 aircraft at 469 million. This prompted yet another review of the TSR.2 program, and a number of foreign types were considered to replace it, including the TFX, later known as the F-111. This was, sadly, also a mismanaged project, and an Air Staff team sent to Washington reported that “There is some reason to believe that there are difficulties about aerodynamics, weight growth and rising cost.” Nevertheless, it was estimated that 158 TFX aircraft could be acquired for 332 million.
Several means of cutting the costs were also considered, and a proposal to negotiate an agreement on a fixed-price contract was approved in March 1965. Meanwhile, it was decided to update the specifications of the TSR.2 by taking into account the first flight test results.
On 15 January 1965, the Secretary of State for Defence expressed for te first time his view that the TSR.2 should be cancelled, along with the P.1154 and the HS681. The TFX would be cheaper, although it was admitted that costs might rise, as this aircraft was still in an early stage of development. At the end of January, it was decided to defer a decision until a closer comparison of the TFX and the TSR.2 had been made. The Prime Minister announced this decision on the House of Commons in early February. He estimated that this would cost the taxpayer 4 million per month.
At the end of March, the Defence and Overseas Policy Committee reviewed its position. BAC had refused to accept a fixed price contract, although it had been willing to accept a limited financial responsability. The final decision was made in two Cabinet meetings on 1 April. An option for the F-111A was secured before announcing the cancellation, because the government did not want to be “in the hands of the Americans”. The USA was willing to sell the F-111 at an unit price of 2.125 million pound for the first ten and 2.32 million for later aircraft. The UK did not yet have to commit itself to actually buying the F-111.
On 6 April 1965, the Prime Minister, Harold Wilson, announced the cancellation of the TSR.2 in the House of Commons. The Secretary of State for Defense, Denis Healey, explained that the government had been very reluctant to do so, but that the cost of the TSR.2 program was becoming an “intolerable burden”, because the total cost of 750 millions pounds amounted to 5 million per aircraft, assuming a production of 150 aircraft. Such an expenditure, he declared, could not be cost-effective. He added that even the best efforts of BAC and the government could not provide any “assurance that the Government’s ultimate financial responsibility would be limited”.
Healey continued to say that Britain could no longer afford to produce combat aircraft for its own armed forces only. He went one important step further when he announced the governments intention to reduce the number of strike aircraft that would be bought for the RAF. “It might even be possible to reshape our defences in such a way as to dispense with this type of aircraft altogether.” When questioned about this remarkable statement, Healey admitted that this was very unlikely. It would only occur if Britain would decide to retire its forces from the Middle East and Asia. He mentioned that the government did indeed intend to buy a new strike aircraft, the General Dynamics F-111A, and that this could be done for less than half the price of the TSR.2 program.
To that date, BAC had received 125 million for the development of the TSR.2, and about 70 million had still to be paid.
The TSR.2 was intended to cruise at Mach 2.05 at an altitude of 36,000ft (10,100m), and to achieve supersonic speeds at treetop height. This had to be combined with a long range and a good take-off and landing performance. These requirements called for powerful engines, a large fuel load, and a relatively small wing.
The most remarkable characteristic of the TSR.2 was its enormous length. From front to tail the fuselage contained a small radar, the cockpit, an avionics bay, a large fuel tank, the jet intakes and another fuel tank, the bomb bay, and the large engines with two more fuel tanks. The fuselage was crammed with fuel, and even the engines were almost completely surrounded by fuselage fuel tanks 3 and 4. About 80% of the fuel was in the fuselage, the rest was in integral wing tanks, which filled almost the entire wing. The internal fuel capacity was 5,588 imperial gallons (25,425 liter). The wing was of delta planform with cropped wingtips. Instead of giving the entire wing anhedral, the wingtips were turned sharply downwards. Two pylons could be fitted under each wing, to carry missiles, bombs, or even more fuel. The external ferry tanks would contain another 2,400 imp gallons (10,920 liter).
A small, highly loaded wing gives the smoothest ride at very low altitude, but to achieve the requested good take-off and landing performance, the entire wing trailing surface had to be fitted with blown flaps. They could be turned down to 50 degrees for landing, the take-off setting was 35 degrees. There were no ailerons, and the TSR.2 relied on differential operational of the tailplanes for roll control.
The tailfin and tailplanes had no fixed parts, but the elevators were split in two sections. Four airbrakes were fitted on the aft fuselage, in the narrow gap between the wing trailing edge and the tailfins.
The engine intakes were of half-circular type, with movable shock cones. There were auxiliary intake doors behind the lips of the main intakes.
The Olympus 22R Mk.320 engines were twin-spool axial-flow engines, with variable afterburners and water injection. They were designed for sustained cruise at Mach 2+, a feature which would be used later in the Concorde, which was also powered by the Olympus. At least the money for the engine development was not wasted.
The development of the Olympus 22R engine was not without problems. That there was a serious problem was demonstrated in December 1962, when a Vulcan testbed exploded during ground running. The event was traced back to a resonant vibration of the turbine, excited by the coolant air flow for the turbine blades. In July 1964 the shaft of an Olympus engine failed during ground testing, requiring more modifications. And after the engines had been installed in the first TSR.2, the vibration problems returned.
The two crew members sat far forward in the fuselage, on Martin Baker Mk.8A ejection seats. The location of the cockpit was chosen to reduce the vibrations during low-level flights as much as possible.
The view from the front cockpit was very good, without any glare even in bright sunlight conditions. Only during the initial climb had the seat to be raised. The navigator only had a view on the outside world by two small windows in the side of the hood. Directly in front of him was nothing but a large electronics panel. The most serious complaint about the cockpit design was related to the very unreliable and temperamental air conditioning.
The Autonetics Verdan computer system was an American product, developed for the A-5 Vigilante carrier-based jet bomber, an aircraft which had many similarities with the TSR.2. It combined data from the nose radar, the Doppler navigation radar, and an intertial navigation system. The TSR.2 was also fitted with a radar altitude meter and systems to keep the TSR.2 automatically at an altitude of 90m. The system was designed for blind attacks with an error of less than 100ft (30m).
The nose radar had limited search capability, only over sea or flat terrain, and an air-to-air capacity suitable to assist in the connection with tanker aircraft. It was intended primarily for navigation and bomb aiming.
For the reconaissance role, a pallet could be installed in the bomb bay. This contained a side-looking radar, cameras, and a line scanner.
The TSR.2 stood high above the ground, and a not too tall person could walk under it. It had a backward retracting nose leg with twin wheels. The main legs had two wheel in tandem, and retracted into the fuselage. During tests, there were several occurences of heavy vibrations upon touchdown, and modifications to the landing gear were planned.
A tail parachute was fitted and routinely used during the test flights.
The original contract called for 9 prototypes. This was later extended by the order of 11 pre-production aircraft.
The first TSR.2 was rolled out at Weybridge (Vickers) on 4 March 1964, but the first flight was repeatedly delayed. It was found that the engines required modification because they did not fit into the TSR.2, a sign of the poor communications between BAC and Bristol Sidddeley. After relocation of the engine accesories they were installed, but during ground tests there were serious vibrations. A palliative was implemented, and the TSR.2 was cleared for its first flight, but with an engine power limitation. Afterwards, test pilot Roland Beamont commented: “It is clear that the current engine ratings leave the aircraft critically short of thrust, and this situation is likely to dictate the rate of flight development.” New engines would have to be fitted, and this caused again a delay of months.
There was pressure to have the TSR.2 ready to fly at Farnborough’s SBAC show in September. But the first flight of XR219 was finally made on 27 September, with Roland P. Beamont at the controls and D. J. Bowen in the navigator’s seat. For this flight, the undercarriage was left extended and the engine intake configuration fixed. Beamont found that the TSR.2 was pleasant to fly, without any major deficiencies. The behavior was very close to that predicted by design studies and a simulator.
The second flight was not made until the end of December, because the engines had to be replaced. The undercarriage was also troublesome, and not until the tenth flight was a successful undercarriage retraction achieved. This was on 6 February 1965. Supersonic speed was achieved on the 14th flight, on 22 February, when XR219 achieved supersonic speed at “Max. Dry Intermediate” power. The full performance envelope could not be investigated during this flight, because of an error in the No.2 engine (port) which prevented the selection of afterburning. Nevertheless Beamont flew the aircraft to Mach 1.12, with only the No.1 engine at 1/3 afterburner. The XR219 remained completely controllable, with only a small trim change required.
The TSR.2 was directionally very stable, stable in pitch, and somewhat unstable laterally at subsonic speeds. Supersonic flight was reached with only mild buffet between Mach 0.93 and Mach 1, and no trim change was needed. At supersonics speeds it became laterally stable, and behaved flawlessly.
While everything was fine on the aerodynamics side, there were numerous theething troubles with other elements of the TSR.2. The undercarriage continued to cause problems, with serious (1.5g) oscillations on touch-down. The cabin air conditioning failed to operate in any reasonable way. The engines were a source of constant problems.
After the cancellation on 6 April 1965, the three TSR.2s built were immediatly grounded. They were never to fly again.
The TSR.2 was a magnificent technical achievement, but the management of the project turned sour. It would certainly not be the last aircraft project burdened by overly complicated, poorly performing bureaucratic structures. (More recently the way the US government managed the development of the F-35 has come in for severe criticism, but the aircraft was delivered nevertheless.) And spending way beyond the initial budget was almost the norm in the 1950s and 1960s, when increasingly complex technology was incorporated. The TSR.2 suffered from starting with a cost estimate of 35 million pounds at the beginning that climbed to 750 million at the end. Probably these were wilful under- and over-estimates, respectively, but it didn’t help that every time someone prepared a new cost estimate, it appeared to have doubled.
The big failures of the project were not governmental. The forced marriage that was BAC did not work well, as was evidenced by problems with engines that did not fit and aircraft parts that were not manufactured. The customer lost confidence in the ability of BAC to deliver, and by refusing to accept financial responsibility, the management of BAC signalled that it did not have confidence in itself either. While the TSR.2 prototype as flown impressed its test pilots favourably, it was also clear that there were significant problems with engines, landing gear and equipment that remained to be resolved. In effect the government was asked to write a blank cheque to finance the TSR.2. Realistically, it is difficult to blame it for refusing to do so.
In January 1968 the UK cancelled its option on the F-111. At that time the favorite aircraft of McNamara was in serious trouble. It fell below the specifications, was seriously delayed and had become even more expensive than the TSR.2. Although the USN abandoned its F-111B carrier-borne fighter version as totally unsuitable, the USAF version of the F-111 was not cancelled. Later the F-111 would prove that it was an effective strike aircraft, but it took a long time to mature. Its complex electronic suite would only become fully reliable late in its career.
For the RAF was decided to buy the F-4 Phantom for the tactical attack role, and the Buccaneer as long-range strike aircraft. This was a remarkable reversal of fortunes. A single Buccaneer, XK487, had been used as a testbed for the TSR.2’s radar in 1963, and had attracted new interest from the RAF because of its excellent performance at low altitudes. In 1969 the RAF received the first example of the aircraft it had rejected twelve years earlier. In July 1970 No.12 squadron became the first operational Buccanneer unit of the RAF. The land-based version of the Buccaneer was the S.2A. The S.2B was a modified version, which had a bulging bomb bay door with an additional fuel tank, and provision for the Martel missile.
The Buccaneer would continue in service until 1994 and still serve effectively in the 1991 Gulf War. Its longevity and success points to the major flaw in the reasoning behind the TSR.2 project: The advances in raw aircraft performance that the TSR.2 offered over the Buccaneer were probably not worth the extra cost of the bigger, heavier and more expensive platform. The money would have been better spent on improving the equipment and armament fit of the Buccaneer.
Here four aircraft are compared: The TSR.2, the Buccaneer S.2 that effectively replaced it, the General Dynamics F-111A that was once selected to replace it, and the North American A-5 Vigilante. The mention of the latter may be a surprise in this context. But this carrier-borne attack aircraft resembled the TSR.2 in many ways, and actually preceded it by seven years. It had the same all-moving tail surfaces, the long fuselage, the blown flaps, the inertial navigation system, the twin engines and the internal bomb bay. It failed as an attack aircraft because of a complicated and ineffective bomb bay design. (The data listed here are for the reconnaissance version, which carried more fuel.)
|Type||BAC TSR.2||Blackburn Buccaneer S.2||General Dynamics |
|North American RA-5C Vigilante|
|Engines||Bristol Siddeley Olympus Mk.320||Rolls-Royce RB168 Spey Mk.101||P&W TF30-P-3||General Electric J79-GE-10|
|Wing Span||11.28m||13.41m||19.20m / 9.74m||16.17m|
|Fuel||25404 l||19090 l||13633 l|
|Load internal||6 * 454kg||4 * 454kg||2 * 340kg|
|Max. Speed||2185km/h||2337km/h at 16290m||2229km/h at 12190m|
|Low Level Speed||1352km/h||1040km/h||1470km/h|
- RAF Nuclear Deterrent Forces.
Humphrey Wynn, Ministry of Defence Air Historical Branch.
- Testing Early Jets.
- Airplane Nr.16.
Published by De Agostini (Netherlands).
- Handley Page Victor.
Ian Allan Ltd., 1988.
- RAF Buccaneer
Peter R. Foster
Ian Allan Ltd., 1987.