Development
The B-47 arose from a 1943 U.S. Army Air Forces requirement for a jet bomber and reconnaissance aircraft that could reach Nazi Germany in the event that Great Britain fell. The next year, the requirement evolved into a formal request for a bomber with a specified speed of 500 mph (800 km/h) or more, a range of 3,500 mi (5,600 km), and a service ceiling of 40,000 ft (12,200 m). It envisioned using the General Electric TG-180 turbojet engine, then in development. The subsequent von Kármán mission inspected Nazi Germany laboratories from May 1 - July 1945. One of the members was Boeing's chief aerodynamicist, George Schairer, who became so convinced of the merits of such a design that in May, 1945[clarification needed] he wrote a letter to Boeing management suggesting the matter be investigated.[citation needed]


[edit] Preliminary designs
North American, Convair, and Boeing submitted proposals. The first Boeing proposal, the Model 424, was a modification of a conventional propeller-driven bomber design, basically a scaled-down version of the Boeing B-29 fitted with four jet engines.

Meanwhile, the USAAF had awarded study contracts to all three aircraft manufacturers working on the jet bomber project, as well as to Martin, which had also decided to join the fray. All of the competing bombers, including the North American B-45, Convair XB-46 and Martin XB-48 would have conventional straight wings with four to six engines, and would lack the performance of the swept wing B-47.


[edit] NACA design tests
The National Advisory Committee for Aeronautics (NACA, the ancestor of NASA) performed wind tunnel tests on a composite model of the designs submitted by the manufacturers. (The three submissions were generally similar.)

The NACA wind tunnel tests showed that the Boeing model suffered from excessive drag. Boeing engineers then tried a revised design, the Model 432, with the four engines buried in the forward fuselage, but although it had some structural advantages there was little effect on drag. At this point Boeing engineers turned to the German swept-wing data. A little design work by Boeing aerodynamicist Vic Ganzer led to an optimum sweepback of 35 degrees.

Boeing modified the Model 432 design with a swept wings and tail, resulting in the Model 448, which was presented to the USAAF in September 1946. The Model 448 retained the four TG-180 engines in the forward fuselage and, at the instigation of project manager George Martin, added two more TG-180s buried in the rear fuselage to provide greater range and performance.

Boeing submitted the Model 448 to the USAAF, only to have it rejected immediately. The Air Force strongly disliked fitting the engines in the fuselage, since that made engine fire or disintegration catastrophic. The engines would have to be moved back out on the wings.

That led straight back to the drag problem, but the engineering team came up with a clean, elegant solution, with the engines in streamlined pods attached to the wings. This innovation led to the next iteration, the Model 450, which featured two TG-180s in a single pod mounted on a pylon about a third of the way outboard on each wing, plus another engine slung from the wingtip.

The Air Force liked the new configuration, and so the Boeing team continued to refine it. One problem was landing gear. There was no space for landing gear in the thin wings, and trying to put conventional tricycle landing gear in the fuselage would have ruined the aircraft's streamlining and degraded its performance. Furthermore, the USAAF was now also insisting that the bomber be able to carry an atomic bomb. As nuclear weapons were very large at the time, that meant a long bomb bay, further limiting space for landing gear.

The solution was a "bicycle" landing gear configuration, with the two main gear assemblies arranged in a tandem, not a side by side, configuration. Outrigger landing gear was to be fitted to the inboard engine pods. The concept had already been tested on a modified Martin B-26 Marauder aircraft.

However, bicycle landing gear made it difficult for a pilot to "rotate" an aircraft into a nose-up position for takeoff. Again, the solution was simple: the landing gear was designed so that the nose-up position was the default. This little change would have a very pleasing effect on an aircraft that was already shaping up to be very elegant, giving the machine the appearance of being ready to leap into the air even when it was sitting still.

There were some other tweaks to the design, such as a wingtip extension to improve range. This had the effect of moving the outboard engines from a wingtip position to an underwing position towards the end of the wings.


[edit] USAAF selects Boeing
The USAAF was very pleased with the refined Model 450 design, and in April 1947, the service ordered two prototypes, to be designated "XB-47". Assembly began in June 1947. People involved with the project were very excited, since they believed (correctly as it turned out) they were working on a breakthrough in aircraft design.

On 18 September 1947, the USAAF became a separate service as the newly-established U.S. Air Force. A few months later, the XB-47 prototype flew its first flight on 17 December 1947, with test pilots Robert Robbins and Scott Osler at the controls. The aircraft flew from Boeing Field in Seattle to the Moses Lake Airfield in central Washington state, in a flight that lasted 52 minutes. There were no major problems, except that Robbins had to pull up the flaps with the emergency hydraulic system and the engine fire warning lights kept popping on, the sensor technology being very unreliable at the time. Robbins reported that the flight characteristics of the aircraft were good.