The Martin B-57 Canberra was a rare example of a foreign-designed military aircraft being built under license by an American manufacturing company for use by the US armed forces. The last previous such example was the de Havilland DH-4 of World War I.
The first B-57 was manufactured in 1953, and the Air Force had accepted a total of 403 B-57s before production ended in early 1957. Overall, the B-57 was not easy to fly. Prior to modification of its longitudinal control and stabilizer systems, the B-57 was actually uncontrollable if one of its two engines were to fail during takeoff or landing. The number of B-57 accidents was rather high, but the accident rate actually compared favorably with that of the B-47 and some other aircraft of that era.
The B-57 had its origin in the Korean War, which broke out on June 25, 1950. At that time, the only bombers that the USAF could commit to battle were types that were left over from World War 2. In particular, the only light bomber available was the Douglas B-26 Invader. The Invader had proven to be the only bomber suited to the night interdiction role in Korea. Unfortunately, the Invader was capable of visual-only operations and was available only in dwindling numbers. At the current attrition rate, it was projected that the B-26 fleet would be entirely gone by 1954. A modern replacement was urgently needed.
In pursuit of a B-26 replacement, on September 16, 1950 the USAF Board of Senior Officers issued a preliminary requirement for a light jet bomber with a service ceiling of 40,000 feet, a cruising speed of about 400 knots, a maximum speed of 550 knots and a range of 1000 nautical miles. The aircraft had to be capable of operating from unimproved airfields, had to be able to search for targets at low altitude and low speed, and had to be capable of destroying mobile or stationary targets at night or in bad weather with either conventional or nuclear weapons. The aircraft was also to be capable of performing the high-altitude reconnaissance role. The Air Force wanted the aircraft as soon as possible.
To assure that the new aircraft would be available to the Air Force in a timely fashion, it was decided that the winning candidate would have to be selected exclusively from a list of existing designs, since the creation of an entirely new design would add years to the development schedule. Designs that would be considered were the Martin XB-51, the North American B-45 Tornado, and the North American AJ-1 Savage. In a rare move, foreign designs were also to be considered, namely the Avro-Canada CF-100 and the English Electric Canberra.
The North American AJ-1 Savage carrier-based attack bomber was powered by a pair of R-2800 radial engines, augmented by a J33 turbojet in the tail. It had the advantage of already being in fleet service with the US Navy, but it was too slow and showed relatively little growth potential. In addition, it was felt that the complex hydraulic system of the AJ-1 would make the aircraft too vulnerable to antiaircraft fire. Another negative was the fact that the Savage did not carry any defensive armament nor any type of forward-firing armament at all. An extensive redesign would be needed to adapt it to the night interdiction role.
The North American B-45 Tornado (which was already in service with the USAF's 47th Bombardment Wing) showed some promise in being able to fulfill the night intruder role, but its aerodynamic structure was based on World War II technology and the aircraft was considered as being too large and heavy to make it sufficiently maneuverable in the night low-level role.
The Avro Canada CF-100 was a heavy, long-range all-weather interceptor that was nearing production status in Canada, but it was felt that the CF-100 was too small to be able to carry the required bomb load without a major redesign of its structure. In addition, the maximum range that could be designed into the aircraft was far short of that which would be required for a bomber.
The Martin XB-51 was a large, three-jet bomber, with two J47 turbojets mounted on the lower forward side of the fuselage and a third J47 in the tail. A variable-incidence swept wing was fitted. The XB-51 had the advantage of being highly maneuverable for its size, and it was fast. Nevertheless, its low limit load factor of 3.67 Gs severely limited its maneuverability during tactical operations and the range and endurance were considered to be too small to meet the night intruder loiter requirements. The aircraft had a novel rotatable bomb bay door, but the weapons bay was too small to permit the carriage of a wide variety of ordnance types.
For a while, the XB-51 appeared to be the front runner in the contest, but the English Electric Canberra, a twin-jet three-seat bomber built in England soon emerged as an important contender.
At first sight, the English Electric company would appear to be an unlikely candidate to have built a successful bomber. The Aircraft Division of the English Electric Company Ltd was located at Preston in Lancashire. The company had been formed in 1918 as an amalgamation of five Midlands electrical engineering firms. English Electric had been briefly and relatively unsuccessfully involved in aviation during the 1920s, but the company had gotten out of the aircraft business entirely in 1926. However, it got back into the business again in a big way as a result of the British arms buildup that led up to World War 2. In 1938, English Electric got a contract from the British governmemt to build Handley Page Hampden bombers for the RAF. During the war, English Electric built no less than 2145 Halifax heavy bombers under license, and continued on after the war to build de Havilland Vampire jet fighters under license. In support of these endeavors, English Electric had built three large assembly plants at Samlesbury, located between Preston and Blackburn.
In the last months of World War 2, jet-propelled aircraft had just entered operational service, and British manufacturers of bombers (Avro, Handley Page, Vickers) were instructed to begin work on jet-propelled bombers. Although the company had never produced a successful combat aircraft design of its own, English Electric was also invited to participate, mainly because it had built up a good reputation for itself through the quality of the aircraft that they had built for the RAF.
At that time, the Chief Engineer at the English Electric Company was William Edward Willoughby Petter, who had come over to English Electric from Westland Aircraft in 1944. While at Westland, he had worked on the Lysander army cooperation aircraft, the Whirlwind long-range escort fighter, and the Welkin high-altitude interceptor. English Electric's managing director, George Nelson, chose Petter to lead the company's bomber design effort. Petter immediately recruited F. W. (Freddy) Page, who had worked on the Tempest fighter while at Hawker.
English Electric's preliminary design effort began in mid-1945. Petter's first concept for the new jet bomber was a mid-wing monoplane powered by a single very-large turbojet, an idea which he had been exploring back when he was at Westland. The entire design was to be built around a proposed enormous (5 feet 6 inches in diameter) two-stage centrifugal turbojet engine to be built by Rolls Royce. This engine would, it was hoped, be able to deliver 12,000 pounds of static thrust. At that time, to have used two or even three of any existing type of engine would have been prohibitive in terms of fuselage size and weight. The large engine would be mounted inside the fuselage, and would be fed by a pair of intakes in the wing roots, and would exhaust via a long tailpipe extending to the rear of the aircraft. The weapons load and fuel would be carried entirely in the wings. The aircraft would weigh about 40,000 pounds and would cruise at 500 mph at 35-40 thousand feet.
The Royal Air Force pictured the new jet bomber as a Mosquito replacement, to operate primarily in the radar bombing role at high altitude. It would carry no defensive armament, relying on its high performance to evade interception. In May of 1945, the Ministry of Aircraft Production formalized the concept in Specification E.3/45 (changed soon afterwards to B.3/45).
Shortly thereafter, advances in turbojet design caused Petter to abandon his original single-engined design and to switch to a twin-engined configuration. This made it possible for the fuel load to be accommodated inside the fuselage, with extra fuel to be carried in the wing leading edges if needed. Swept-back wings were briefly considered but rejected since their attendant aerodynamic problems were at that time still largely unsolved.
The engines were to be a pair of Rolls-Royce AJ.65 axial-flow turbojets of 6500 lb.s.t. each. This engine, later to be known as the Avon, had a protracted development phase and it was not until seven years later that it was operationally acceptable to the RAF. Initially, the two engines were to be buried in the wing roots, but at an early stage they were moved outward to about quarter wing span and were semi-buried inside the wing's thickness.
In the autumn of 1945, English Electric submitted their design. On January 7, 1946, a contract was issued which called for four prototypes to be available by 1949. As insurance in case the AJ.65 engine encountered delays, the second prototype was to be equipped with two 5000 lb.s.t. Rolls Royce Nene turbojets.
English Electric got to work on the detailed design of the aircraft in early 1946. The team set up shop in a wartime-acquired garage in Preston, and it was there that the first design took shape. The project was known as the A1 by the English Electric factory.
The aircraft that took shape had two jet engines semi-buried in the wing, which was of symmetrical section with the center of pressure well aft for a high usable Mach number. The controls were manually operated with an electrically-operated trimmer and an electrically-operated variable incidence tailplane to compensate for strong Mach trim changes. The aileron and rudder were mass balanced and aerodynamically horn balanced. The fuselage was of stressed skin semi-monocoque construction, with a pressurized cockpit accommodating a crew of two sitting on individual ejector seats. Three fuel tanks (with a total of 1375 gallon capacity) were in the upper center section of the fuselage, with the lower part of the fuselage comprising the bomb bay. Twin nosewheels retracted rearwards into the forward fuselage, and the single mainwheels retracted inwards into the wings. The aircraft was fitted with hydraulically-operated dive brakes of the finger spoiler, drag channel section types, and split type flaps were provided. The maximum bombload was envisaged at 10,000 pounds, but the normal bomb load would be one 5000-lb, two 4000 lb, or six 1000 lb bombs. The bomb bay could also be fitted with a 300-gallon auxiliary tank for long-range operations.
It was envisaged that the aircraft would be bombing by radar rather than visually, so the aircraft originally carried a crew of only two, with the pilot sitting to the left underneath a large and wide transparent one-piece canopy, and the navigator/radar operator sitting below and to the rear of the pilot. The bombing radar was to be housed in the forward part of the nose. The two crewmen entered the aircraft via a large door cut into the right side of the fuselage.
In 1947, Specification B.3/45 was revised somewhat, forming the basis of a new Specification, known as B.5/47. B.5/47 added a requirement for a visual bombing system and added a third crew member as a bomb aimer. The pilot still sat towards the left underneath a large single-piece bubble canopy, but there ware now two crewmen sitting side-by-side below and to the rear. During approach to the target, the bomb aimer would move forward into the nose where the bomb aiming equipment was housed inside a small transparent nose cap. This change was made necessary in part because the ARI 5829 G-H radar bombing system originally proposed for the Canberra was seriously behind schedule. Also added was a provision for two jettisonable wingtip fuel tanks of 250 gallons each. Like the bomber version of the De Havilland Mosquito, the aircraft was to be devoid of defensive armament, relying on its speed and maneuverability to evade interception.
In 1948, Petter moved his design facility to Warton, a former USAAF maintenance base located five miles from Preston, and final assembly of the A1 first prototype took place there.
The A1 first prototype (VN799) was rolled out of English Electric's Warton facility on May 2, 1949. It was powered by a pair of 6000 lb.s.t. Rolls Royce Avon RA.1 engines. The first flight took place on May 13, with English Electric chief test pilot Roland Beamont at the controls. Some flutter and directional snaking were encountered, which resulted in a change in the shape of the rudder, with the horn balance area being reduced in size and the rounded rudder giving way to a more squared-off appearance. This rudder shape remained with the production models.
VN799 was demonstrated at Farnborough in September of 1949, where it was the star attraction. Three more prototypes flew before the end of 1949. The second (VN813) was powered by Rolls Royce Nene turbojets, which was the backup engine in case the Avon failed. VN813 was later used by Rolls-Royce at Hucknall in 1950 for the Nene engine development program. The third prototype (VN828) was retained by English Electric for development flying and underwent radio and radar trials. The last prototype (VN850) was the first to have jettisonable wingtip tanks.
British bombers were traditionally named for cities. The name "Canberra" was chosen for the new bomber, in honor of the capital of Australia. This became official in January of 1950.
Canberra VN850 was displayed at the Belgian International Air Show in June of 1950, and appeared twice at Farnborough during that year, first at an RAF display in July and then at the SBAC display in August.
The first four prototype Canberras were known as B.1. They were two-seaters, being based on the provisions of the original specification, B.3/45. The Canberra B.2 was the definitive production version for the RAF Bomber Command, being based on the provisions of Specification B.5/47. The first B.2 (VX165) flew for the first time on April 21, 1950, followed by the second aircraft (VX169) on August 2. These aircraft were both powered by 6500 lb.s.t. Rolls-Royce Avon RA.3 (Mk 101) engines. The first production B.2, WD929, flew on October 8, 1950. All three aircraft were used for trials which were aimed at clearing the aircraft for RAF service by the spring of 1951.
The next six production B.2s (WD930/WD935) were all used for various trials. The first delivery of a Canberra B.2 to the RAF was made in May of 1951, when it began to replace the Lincolns of No. 101 Squadron at Binbrook in Lincolnshire
The Canberra had attracted the attention of the USAF almost from the date of its very first flight. A US mission had come to Boscombe Down in August of 1950 to take a look at the VN850 Canberra prototype. The USAF was sufficiently interested that they requested that a demonstration of the Canberra be carried out. Roland Beamont, English Electric's chief test pilot, paid a visit to Burtonwood in Cheshire (the USAF facility responsible for the maintenance of USAF aircraft in Europe and the supply of USAF bases in the United Kingdom) in VX169, the second prototype B.2, and carried out a flying demonstration on August 17, 1950. The American officers had an opportunity to inspect the aircraft at close quarters before Beamont returned the aircraft to Warton. In September, a group of USAF test pilots visited Warton, and the pilots were given the opportunity to fly the Canberra. They were impressed with its performance. As a result of this flight, the USAF expressed a desire to evaluate the Canberra as a possible candidate for the replacement of the B-26 Invader.
On December 15, 1950, the board of Senior Officers organized a committee chaired by Brig. Gen. S. P. Wright to evaluate the performance of the various candidates. Very early on, the choice narrowed down to either the Martin XB-51 or the English Electric Canberra. While neither design was deemed to fully meet the night intruder requirements, the Wright Committee suggested that both types should be acquired, with two groups being equipped with each of the two competitors.
This recommendation was short-lived. The Senior Board was not all that enthusiastic about the Wright Committee's suggestion, and found itself favoring the Canberra. The Board conceded that the B-51 was 100-knots faster than the Canberra, but they concluded that the Canberra was much better in flight endurance. The Canberra could loiter for 2 1/2 hours over a target 780 nautical miles from its base, but the B-51 could loiter only one hour over a target only 350 nautical miles from its base. The Board was also concerned that the wingtip wheels of the B-51 would be a problem at hastily-prepared forward bases, and that the Canberra would be much easier to maintain.
Although the Board of Senior officers generally favored the Canberra, the evaluation committee wanted to be certain that the Canberra would be made available to the USAF in sufficient numbers. They sent Lt. Gen. Kenneth B. Wolfe to the United Kingdom to obtain more information from the British about their ability to meet their own needs and at the same time to supply over 300 additional aircraft to the USAF to equip four bomb groups.
Although the British government was willing to supply the Canberra to the USAF, they candidly admitted that they could not make deliveries at a fast enough rate to satisfy both USAF and RAF needs. The only option would be for an American aircraft company to manufacture the Canberra under license in the USA. The Glenn L. Martin Company was approached with the proposal that it build the Canberra under license should the XB-51 lose out in the tactical bomber contest.
As suggested by General Boyd, the Board felt that no final determination could be made until they obtained a Canberra for trials, so that competitive trials could be carried out. At USAF request, the RAF agreed to loan WD932, a B.2 trials aircraft, to the USAF for six months for evaluation. On February 21, WD932 left Aldergrove, Northern Ireland and landed at Gander, Newfoundland, becoming the first jet aircraft to make a direct, unrefuelled crossing of the Atlantic. It arrived at Andrews AFB on February 24, 1951.
The flyoff was held at Andrews AFB on February 26, and after a ten-minute display with Roland Beamont at the controls, the Canberra was the clear favorite. On March 2, 1951, the Air Staff directed that the Air Materiel Command arrange for the domestic production of the Canberra. Since Martin had a considerable amount of experience with the XB-51, the Baltimore-based company was the contractor chosen to carry out the license production of the Canberra.
The Air Force decreed that the B-57 was to go directly into production, with relatively few changes from the Canberra B.2 RAF version. Letter Contract AF 33(038)022617 was issued on March 24, 1951 to the Martin Aircraft Company, which authorized the manufacture of 250 Canberras under license. The USAF designation would be B-57, with the company designation being Martin Model 272. It was anticipated that 50 planes a month would be delivered by Martin between November 1952 and October 1953.
The license agreement between Martin and English Electric was concluded on April 3. This was the first foreign-designed military aircraft to be bought by the USA in 35 years, dating back to the S.E.5 and DH-4 of World War I vintage.
On July 1, Martin agreed to have 60 percent of the work on the B-57 performed by subcontractors. The principal subcontractors were Kaiser Products of Bristol, Pennsylvania which was to build the wings and special weapons bomb-bay doors, and the Hudson Motors Corporation of Detroit, Michigan which was responsible for the aft portions of the plane.
The B-57A was to be essentially identical to the Canberra B.2. However, the Canberra B.2 had been powered by a pair of Rolls Royce Avon 101 turbojets, rated at 6500 lb.s.t. each, and it was agreed that the US version of the Canberra would be powered by two 7200 lb.s.t. Armstrong Siddeley Sapphire turbojets, which would be built under license by the Wright Aeronautical Corporation under the designation YJ65-W-1. Like the B.2, the B-57A would lack any defensive armament, relying on its speed and maneuverability to evade interception.
Canberra B.2 WD932 was allocated to Martin as its first pattern aircraft, and it was delivered to Baltimore in March of 1951. It was used for a series of extensive flying trials. The set of Canberra drawings arrived by air at the beginning of June 1951. In June, Martin requested a second Canberra and was allocated WD940. WD940 set an official trans-Atlantic speed record during its delivery flight on August 31, 1951. The two planes WD940 and WD932 acquired USAF serial numbers (51-17352 and 51-17387 respectively), but they were carried in the USAF inventory as Canberras, not as B-57s, and they were never painted in USAF colors either.
Martin tested the first British Canberra from April to October 1951. The second British Canberra was flown only four hours, and then disassembled and shipped in pieces to Martin's main subcontractors.
On December 21, 1951, WD932 (51-17387)lost a wing during a 4.8g maneuver at 10,000 feet over Centerville, Maryland, on the Delmarva Peninsula. Both crewmembers ejected, but one of them was killed when his parachute failed to open. An investigation later showed that the cause of the accident was an incorrect use of the fuel load, with the forward fuselage tanks having been used first, making the aircraft tail heavy and causing it to be longitudinally unstable and pitch up in a tight turn, overstressing the wings.
The first production B-57A (52-1418) was rolled out of Martin's Middle River plant on July 20, 1953, and took of on its first flight that same day. The first flight was an uneventful 46-minute hop. The B-57A was externally almost identical to the Canberra B.2 except that the navigator's window was repositioned aft of the cockpit on the starboard side, the windows on the port side of the fuselage being deleted. In addition, the pitot tube, which on the B.2 was at the center of the glazed nose cone, was relocated beneath and aft of the glazed nose. The bomb bay was shortened and a cooling system air intake was added underneath each engine nacelle. The cockpit canopy was slightly modified to afford better visibility and to provide more room for the crew, which was reduced from three to only two. A pair of wing tip fuel tanks were provided to increase the loiter time. The British clamshell-type bomb-bay doors were replaced by a Martin-developed rotating bomb bay door. Originally developed for the XB-51, this new rotary bomb bay door rotated 180 degrees and eliminated the drag caused by an opened bomb bay compartment during the bomb run.
The first B-57A underwent its official Air Force flight acceptance at the Middle River field on August 20. The first B-57A was accepted by the Air Force that same day, but the plane was immediately bailed back to Martin. USAF testing did not begin until December of 1953, when all the other B-57As were delivered.
The B-57As were used exclusively as test airframes for evaluation stability and control, airframe structures, and for tests of systems such as Shoran bombing, navigation, and radio systems. They were not considered as being combat-ready.
The second of the eight B-57As (52-1419) was operated by the National Oceanic and Atmospheric Administration (NOAA) of the US Department of Commerce out of Miami International Airport. Operating under the civilian registration of N1005 and fitted with a large nose radome replacing the transparent nosecone, it was used in the tracking of hurricanes in . When it was retired by the NOAA, N1005 was acquired by the Gorge T. Baker Aviation School at Miami International Airport for flight training.
52-1418/1425 Martin B-57A-MA 1418 transferred to NASA as 218 6/19/57. 1419 transferred to Edwards AFB for stability tests. Later operated by NOAA as N10005 for weather research.