The Avro CF-105 Arrow was a delta-wing interceptor aircraft, designed and built by Avro Canada during a short period of time in the 1950s. Likely the most advanced interceptor of its day, the design was entering the middle stages of testing when it was cancelled in 1959, after a long and bitter political debate. The prototypes were then destroyed, creating an enduring piece of Canadian mythology.

Avro Arrow rollout

Table of contents
1 The Plane
2 The Politics
3 The Controversy
4 Further Reading
5 External links

The Plane

Raison d'etre

Avro's first project, the Avro CF-100 Canuck, had not even entered RCAF service in 1951 when Air Force planners started looking for its replacement. Lead times in getting new designs into service appeared to be growing at a rapid clip; designs like the P-51 Mustang were in service only three years after introduction, but the CF-100 was at six years and counting. Similar problems were occurring with designs around the world, it was not simply a problem with the CF-100. Unless a new design effort started immediately, the CF-100 would have no replacement by 1960 when it would be outdated. In March 1952 the RCAF's Final Report of the All-Weather Interceptor Requirements Team was submitted to Avro.

Information about World War II era research on swept-wing designs in Germany started reaching design teams around the world in the late 1940s. The simple solution of sweeping the wings to the rear dramatically reduced the drag of a wing as it approached the speed of sound (the so-called wave drag), making trans-sonic and supersonic aircraft powered by existing jet engines a real possibility. Avro engineers had already explored a number of paper projects on modifications to the CF-100 using swept wings (and tail) as the C-103. Although it beat the CF-100 in terms of performance, it appeared that the small performance gain was not worth the effort.

For the new project the engineers instead turned to another piece of German wartime research, Alexander Lippisch's thin delta-wing designs. This planform had a number of advantages in high speed flight, notably at high altitudes, because the leading edge of the wing stayed clear of the shock wave off of the nose of the aircraft (which adds drag). In addition it had a number of practical advantages too, including large chord at the root (making it strong), excellent internal space for any given drag (great for storing fuel), and a good performance at high angles of attack (great for landing). However it also had a number of disadvantages, primarily high drag at lower speeds, and a bad behaviour known as "pitch-up" due to much of the wing being behind the center of gravity, if the wing stalls the lift suddenly moves forward, forcing the nose up.

Both the US and British were already experimenting with the delta platform. Lippich had moved to the US after the war, and started working at Convair, where the local engineers became very excited in his World War II point-interceptor design. They proposed to build a modified version as the Convair F-92, and in order to test the flight dynamics of the delta-wing configuration they built the 7003 in 1948 as a test-bed. In England the RAE contracted for a series of delta-wing aircraft, including designs both with and without tails, the Gloster Javelin and Fairey FD.2 for example, and flew them throughout the 1950s. In France Marcel Bloch studied the delta and used it to develop the famous Mirage series of fighters in the mid-1950s.

All of this test data started flowing out in the early 1950s, and along with it a huge debate on the merits of the delta design. The delta proved to have all of the promised qualities at high speeds and high altitudes, but at lower speeds and altitudes the performance was considerably worse than conventional planforms. Avro's designers took advantage of all of this research and conducted a great deal of their own, in American wind tunnels. They selected the tail-less delta based primarily on its excellent high-speed, high-altitude performance, exactly where an interceptor spends most of its time.

They created two versions of a design known as the C-104, the C-104/1 with a single engine, and the C-104/2 with twin engines. The planes were otherwise similar, using a low-mounted delta-wing, powered by the new Orenda TR.9 engines, armed only with Velvet Glove missiles (an RCAF design) stored in an internal bay, crewed by one, and guided with a completely automatic interception system that would track down and attack the target after it was selected by the pilot (similar to the F-86D). The primary advantage of the twin engine /2 version was that it was larger overall, including a much larger weapons bay which could have been used for camera packs, long range stand off missiles, side-looking radar and other such uses. The results were submitted to the RCAF in June 1952.

Design is finalized

Intensive discussions between Avro and the RCAF examined a wide range of possible sizes and configurations, culminating in RCAF Specification AIR 7-3 in April 1953.

AIR 7-3 now called specifically for a twin-engined aircraft, since no engine then built would be able to lift the fuel load needed for the long-range missions the RCAF demanded. This was to be 300 nautical miles for a normal low-speed mission, and 200nm for a high-speed interception mission. It was to fight at Mach 1.5 at an altitude of 50,000 feet, and be able to pull 2g in maneuvers with no loss of speed or altitude under those conditions. The time from a signal to start the engines to the aircraft's reaching an altitude of 50,000 feet and a speed of Mach 1.5 was to be less than five minutes. The turn around time on the ground was to be less than ten minutes. The new specification also called for a crew of two, as it was considered unlikely that even a fully automated system would be easy enough to operate by the pilot while covering as much ground as the new plane could in full flight. An RCAF team led by Ray Footit visited US and European aircraft producers and declared that no existing, or planned, aircraft could fulfill these requirements.

In response to the updated requirements, Avro returned their modified C-105 design in May 1953. It remained similar to the C-104/2, but was even larger since the RCAF had concluded that a two man crew would be needed. It was also decided to move the wing to the upper part of the fuselage from its former low-mounted point, in order to improve access to the internals of the plane, weapons bay, and engines. The high-wing on this design also allowed the wing to be a single structure across the plane, which simplified construction and added strength. However this also required long landing gear that still needed to fit within the thin delta-wing, an engineering challenge. Five different wing sizes were outlined in the report, from 1000 to 1400 square feet. The 1200 sq.ft. version was eventually selected. Three engines were considered as well; the Rolls-Royce RB-106, the Bristol B.0L.4 Olympus, and the Curtiss-Wright J67 (a license-built version of the Olympus). The RB-106 was selected with the J67 as a backup.

The weapons bay was even larger than the 104/2, situated in a large thin box running from the front to the middle of the wing. The weapon system originally selected was the Hughes MX-1179, which was the pairing of the existing MA-1 fire-control system, firing Falcon missiles of radar and heat seeking variants. This system was already under development for proposed use in the US's WS-201 1954 Interceptor (dating from 1949, which would lead to the Convair F-102). The Velvet Glove radar-guided missile was considered unsuitable for supersonic launch, and further work on that project was cancelled in 1956.

In July 1953 the work was accepted and Avro was given the go-ahead to start a full design study. In December $27 million was provided to start flight modelling. At first the project was limited in scope, but the introduction of the Soviet Myasishchev M-4 Bison jet bomber and their testing of a hydrogen bomb dramatically changed priorities. In March 1955, the contract was upgraded to a $260 million contract for five Arrow Mark 1 flight-test aircraft, to be followed by 35 Arrow Mark 2s with production engines and fire-control systems.

Production starts

Most aircraft designs start with the construction of a small number of hand-built prototypes. These are test-flown, and the inevitable problems are discovered and fixed. Once everyone is happy with the results, a set of jigs is constructed which are laid out in the assembly hall. Parts coming off the assembly line are clamped to the jigs while they are fastened to each other. This is a slow and expensive process, but a safe one.

For the Arrow project it was decided to adopt the Cook-Cragie system. Developed in the 1940s, Cook-Cragie skipped the prototype phase and built the first test-airframes on the production jigs. Any changes could be incorporated into the jigs while testing continued, so production started as soon as the test program was complete. The downside of Cook-Cragie is that changing jigs is expensive, so if the number of changes needed is large, it's more expensive than the hand-built prototypes. If you have high confidence that the plane will enter production largely "as designed" it can save considerable time and money, but if you're wrong it can cost you a lot.

Given this it's somewhat surprising that the Cook-Cragie system was selected for the Arrow program. The plane was Avro's first delta, first supersonic, and practically no parts of the aircraft design (weapons, fire-control or engines) existed when work started. The chance that something would go wrong and the Cook-Cragie system would backfire was tremendous.

In order to have any confidence in an advanced design like the Arrow, a massive testing program was started. By mid-1954 the first production drawings were issued and wind tunnel work began. In another program, large, instrumented free-flight models were mounted on Nike solid rockets and launched over Lake Ontario for aerodynamic drag and stability tests.

Experiments showed the need for only a small number of changes to the design, mostly involving changes to the wing profile and positioning. In order to improve high-alpha performance the front of the wing was drooped, especially on the outer sections, a dog-tooth was introduced to control spanwise flow, and the whole wing was given a slight negative camber to help control trim drag and pitch-up.

Further data on the area rule became available during the design stage, and several changes were made to the layout of the plane to incorporate this. These are largely hidden from casual observation on the Arrow, but you can see the design in the rapid narrowing of the cockpit spine (which originally ran the length of the plane) and the addition of a tailcone in order to make it "pointy" at both ends.

Several portions of the evolving Arrow design were unique at the time. The plane used a large measure of magnesium and titanium in the fuselage, the latter limited largely to the area around the engines and for fasteners. At the time titanium was an expensive material and not widely used due to it being difficult to machine. The construction of the airframe itself was fairly conventional however, with a semi-monocoque frame and two-spar wing.

The Arrow's thin wing demanded aviation's first 4000 psi hydraulic system that could supply enough power while using small actuators. This resulted in the problem of there being no control "feel" for the pilot, and to solve this the control stick input was "disconnected" from the hydraulic system. The pilot's input was sensed by a series of force transducers in the stick, and their signal was sent to an electronic control servo that operated the valves on the hydraulic system to move the various flight controls. In addition the same box fed pressure back into actuators in the stick itself, making it move. This happened quickly enough that it appeared as if the pilot was moving the stick directly. A particularly advanced stability augmentation system was added as well, as long, thin aircraft have a number of coupling modes that can lead to departure if not damped out quickly. Since the centre of lift moved with speed, this also assisted stability and manoeuvre.

In 1954 the RB.106 program was cancelled, so plans were made to use the backup J67 instead. In 1955 this engine was also cancelled, leaving the plane with no engine. At this point the new Pratt & Whitney J-75 was selected for the initial test-flight models (a move mirroring what many US companies were forced into when the J-67 died), while the new TR.13 (soon PS-13 Iroquois) engine was developed at Orenda for the production Mk.2's. Ironically it was the rejected Bristol Olympus design that would actually go into production, and continues in use today on the Concorde and other designs. The Iroquois was a fairly standard high-speed engine design, but incorporated a number of titantium parts for high-heat areas and had a very good power-to-weight ratio for the era.

In 1956 the RCAF demanded an additional change, the use of the very-advanced RCA-Victor Astra fire control system in place of the MX-1179, firing the equally advanced US Navy Sparrow II in place of the Falcon. Avro objected to this choice on the grounds that neither of these pieces were even in testing at that point, whereas both the MX-1179 and Falcon were almost ready to go. The RCAF planners felt that the greatly improved performance of the Sparrow was worth the gamble.

The Astra proved to be one of the few serious problems in the Arrow design. The system ran into a lengthy period of delays, and the US Navy eventually cancelled all work on the Sparrow II in 1956. This left the Arrow weaponless, although Canadair was quickly brought in to continue the Sparrow program in Canada.

A rush study looked at alternatives, including resurrecting the Velvet Glove for use with the Astra, or the use of the original MX-1179 system with its Falcons. Even the MX-1179 had run into difficulties, and the F-102 eventually settled on the older MG-1 system originally used in the F-86D. Work was continuing on the MX however, as it was planned to be used in the upgraded F-102B (later renamed as the Convair F-106 Delta Dart) so this was selected for the Arrow as well. Luckily the huge weapons bay of the Arrow could hold several of any of these missiles, so it was really a matter of selecting the first system that actually started working.

Mark 1

Given the number of technical advancements in the Arrow design, as well as the continuing problems with a suitable powerplant, the Arrow was completed in a surprisingly short period of time. Go-ahead on the production was given only in 1955, and the rollout of the first prototype, RL-101, took place in late 1957.

The J-75 was slightly heavier than the PS-13, which required ballast to be placed in the nose to move the center of gravity back to the correct position. In addition the Astra fire-control system was not ready, and it too was replaced by ballast. The otherwise-unused weapons bay was loaded with test equipment. This means that the Arrow was one of a very small number of planes that would actually get lighter when entering production.

RL-101 first flew on March 25, 1958. Four more J-75 powered Mk.1's were delivered in the next two years. The test flights went surprisingly well; the plane demonstrated excellent handling at all extremes of the flight envelope – quite in contrast to the majority of fighter designs of the era, which proved to have wicked handling. Much of this is due to the natural qualities of the delta-wing, but an equal amount is due to the stability augmentation system.

The only major problems encountered during the testing phase were that the landing gear tended to "skid", and the stability augmentation system needed considerable tuning.

The former problem was due to the gear being very thin in order to fit into the wings. To do this they consisted of two tires in front and back of the gear leg, and the leg retracted in length and twisted as it was stowed. Under some circumstances the tires could hit the ground slightly twisted, and would start to skid. Two landings failed in this manner, but it appeared easy enough to correct.

The stability augmentation system was a matter of tuning, tuning, tuning. Although the Arrow was not the first plane to use such a system, it was one of the first, and the concept had not yet developed into the science it is today. There was little worry that this would not be corrected in the future.

Mk.2

The primary differences between the Mk.1 and Mk.2 were the fitting of the Iroquois engine and the planned fitting of the Astra fire control system. Nevertheless the Astra was still not ready when the Iroquois was reaching the stage of flight-testing, so the decision was made to go ahead with flight tests with the new engine anyway. The first plane of this Mk.2 run, RL-206, was almost complete in 1959 when the project was cancelled.

The Politics

Until 1955, the Arrow project had been rather cost effective. Only $27 million had been earmarked for the studies, and $260 million for the initial production line. However in September 1955, Avro told Cabinet that it needed an additional $59 million to keep the program on schedule. In December 1955, Cabinet limited Avro to eleven prototypes and put a spending cap on the overall program of $170 million over three years. By this time, Avro had become the third largest corporation in Canada, and was employing 14,000 people.

It was around 1955 that the notion began to surface, not only in Canada but most of the western world, that the era of the manned interceptor was over, and that the age of guided missiles had arrived. Britain and the US both scaled back most of their interceptor development programs, leaving only one in Britain, the English Electric Lightning, and only two in the US, the Convair F-102 (and F-106) and the recently started North American F-108 Rapier. Yet by 1956 the British were very interested in the Arrow and tried to buy 200 of them, while cancelling the Thin Wing Javelin since the Arrow was promising to be such a superior performer. This purchase was cancelled, along with most of the British aviation industry, in Duncan Sandys infamous White Paper of 1957.

In February 1957, Cabinet ordered the spending cap increased to $216 million. There is some evidence that the Liberals were losing faith in the project, but it would be impossible to cancel it in an election year. In June the Liberals lost the election, and a Conservative government under John Diefenbaker took power.

In August Diefenbaker signed the NORAD (North American Air Defence) agreement with the United States, which required the subordination of the RCAF Air Defence Command to American command and control. The USAF was in the process of completely automating their air defense system with the SAGE project, and insisted that the RCAF had to use it as well. One aspect of the SAGE system was the BOMARC nuclear-tipped anti-aircraft missile, which when intercepting bombers over Ontario and Quebec would be exploding right over major Canadian cities. This led to studies on basing BOMARCs in Canada in order to push the line futher north, away from the cities. The Canadian politicans seemed to come to the conclusion that they could not afford SAGE and Bomarc AND Arrows.

But perhaps the most fateful event in the Arrow project was on its day of triumph on October 4, 1957, when it was first rolled out to a crowd of 12,000 in front of the Avro plant. That same day Sputnik 1 was launched. Now the age of the missile was clear even to the public, and soon the outcry over the cost of the project was spilling into the press. The Arrow program was already the most expensive in Canadian history, representing a considerable fraction of all government spending, and was seen by many outside Ontario as an industrial welfare program.

In 1958, the Department of Defence Production estimated that $300 million had been spent on the Arrow, and that a further $871 million would need to be spent to have it enter service in 1962. The number of planes to be produced was dropped to 169 from 500, (and later, under US advice, was dropped to 60) at a price of $12 million per unit. This last, grossly inflated, price included lifetime spare engines, runway construction and improvement, simulators, ground support equipment, weapons and other items.

In-fighting soon reached the top of the military, as both the army and navy required new equipment of their own. Wven groups within the air force were worried that the introduction of the Arrow would leave no money for a new tactical aircraft, desperately needed for use in Europe. In August 1958 the CSC advised the government to cancel the Arrow, and buy two Bomarc and 100 interceptors from the US, as well as constructing two SAGE control installations in Canada.

The decision to cancel was a difficult one for the Diefenbaker cabinet, and there was much discussion over the impact it would have on Canadian morale. Nevertheless, on February 20, 1959 Diefenbaker announced to the Canadian House of Commons that the Arrow and Iroquois programs were immediately cancelled. Avro received the word later. This date became known as Black Friday.

The boss of Avro, Crawford Gordon, had long argued with Diefenbaker, and was known to make snap decisions out of spite. That day he did both, and laid off the entire staff of the main plant, over 14,000 workers. After the announcement, Avro employees wandered around the factory in a state of shock. They had expected slowdowns, not cancellation. Defence Minister George Pearkes and Minister of Defence Production Raymond O'Hurley had repeatedly given his assurance to management that the program would not be cancelled. The timing seemed particularly odd considering that a reassesment of the project was scheduled for March, and had been since the previous year. Meanwhile the first Mk.2 was getting ready to fly, and it was expected to take the world speed records at over Mach 2, surpassing the world speed record of 1,404.19 mph set on May 18, 1958 by a United States Lockheed F-104 Starfighter.

Two months later a wrecking crew moved into the largely abandoned Avro plants and starting sawing the six existing airframes apart. They were trucked away and melted down. All plans, engines, parts and even notebooks were also confiscated and destroyed. The blame for this bizarre action has long been placed on Diefenbaker and his cabinet, but this appears to be untrue. Several attempts were made to sell (or even give) the five completed planes to various parties, but in the end, none accepted. With no-one willing to accept security for the planes, and the government clearly unwilling to do so politically, the planes were destroyed by the defense minister. Diefenbaker denied any involvement or knowledge to his death.

Although almost everything connected to the program was destroyed, the forward fuselage and some sections of the wings and control surfaces of the first Mark 2 Arrow were saved and are on display at the National Aviation Museum in Ottawa.

The Controversy

Considered to be one of the finest achievements in Canadian aviation history, the Avro Arrow aircraft program, many say, was cancelled by short-sighted political leaders. Many Canadians consider these leaders to have had little vision or understanding of the technological world unfolding at the time, and the Arrow remains a lasting symbol of a Canada that could have been, but would never be.

Reality is less clear on the fate of the Arrow. As the defense world had expected, the strategic bombing role was passed onto the missile, although perhaps not at the expected rate. The interceptors designed to fight a potential Soviet bomber attack all retired in the 1980s. Today the only remaining interceptor is, ironically, the Soviet MiG-31, built to counter a USAF bomber attack. Nevertheless the Arrow would have filled an important role in the 1960s before the bomber finally passed away.

Many have also suggested that the aviation industry in Canada was destroyed with the cancellation of the Arrow. This claim is rather suspect, considering that Canada is the 3rd largest aircraft producer in the world (behind the US and France). It is true that design of fighter aircraft in Canada ended with the Arrow, but the same is true for most countries of similar means. The rapidly rising costs of fighter aircraft have driven almost everyone out of the business, there are only three companies in the Western world designing them today, when at the time there were dozens.

The US is also often blamed for the demise, often with claims that the US aerospace industry was upset about the 'upstarts' in Canada that were making them look foolish, or alternatively that they were hunting for Avro employees. Quite to the contrary, the US military was distressed at the prospect of losing a first-rate staff in their own North American ally, and even considered buying 50 Arrows to give back to the RCAF in order to ensure production.

Other theories grow even more outlandish. In one instance it had been suggested that it was a good thing the Arrow was cancelled, otherwise Canada would have ended up bombing Viet Nam with them.

To this day, some Canadians see the Arrow as an example of Canadian industrial ingenuity that might have resulted in a world-class product. Instead it became one of their country's worst strategic blunders of the 20th century.

The company's talented design and production teams dispersed, and their talents were used by other countries in the aerospace field, mostly in the United States and Britain. Some of the principal members of the Arrow design and engineering teams headed programs in the Mercury, Apollo and Space Shuttle programs with NASA, others worked for the Anglo-French Concorde project and some of the large private American aircraft companies.

See also: Fighter aircraft

Further Reading

External links