Source: American Bomber Aircraft Development in World War 2 by Bill Norton

With the North American B-25 Mitchell prototype (internal company designation NA-40B) already in the hands of the US Army Air Corps for flight testing in 1939, the promise of cabin pressurization offered a leap in bomber performance by being able to fly higher and faster. Accordingly, in August 1939, the Army issued the XC-214 specification which called for a pressurized medium bomber to supplant the medium bomber types that were soon to become operational. Only Martin and North American responded to XC-214. Martin's submission was for the XB-27 but the USAAC felt Martin didn't have a full grasp of the challenges of high altitude pressurization in their design and North American's submission, the XB-28 won the development contract. This took place on 15 November 1939 just three months after the Army issued its specification with North American inking a contract to begin formal design work on the XB-28. To give you an idea of the pace of development and the pressure of the looming clouds of war, the contract for the development of the XB-28 was signed around the same time that the Army ordered the B-25 Mitchell into production! The XB-28 had started out as a pressurized version of the B-25 Mitchell with a circular fuselage and Pratt & Whitney R-2800 Double Wasp engines with GE Type-C turbosuperchargers replacing the Wright R-2600 Twin Cyclone radials used on the B-25. Design work proceeded rapidly since the company was already at work on a pressurized successor to the B-25 at the time of the release of the XC-214 specification in August 1939. As design work progressed, changes were made stating with abandoning the Mitchell's twin fins for a single fin. Eventually the XB-28 as designed bore little resemblance to the Mitchell. The contract for three prototypes was signed on 13 February 1940.

Besides the change to more powerful R-2800 Double Wasp engines with GE turbosuperchargers, a third supercharger was also fitted to provide cabin pressurization. Heaters powered by gasoline warmed the air in the pressurization ducting for the pressure cabin located in the the forward fuselage. The elongated nacelles had an opening in the rear for the turbo-supercharger exhaust which added some forward propulsive power. The four bladed propellers were counter-rotating to cancel each other's torque to ease handling. Integral self-sealing fuel tanks took up most of the wings. Relatively unique for the day nose wheel steering was fitted and controlled by a lever in the cockpit similar to modern nose wheel tillers. Another unique feature was the use of fluorescent paint on the instrument panel and instruments that would glow at night from overhead UV lamps as an aid to night flying.

To simply the structure of a pressurized aircraft, the pressure cabin only occupied the forward fuselage. All the joints were sealed during assembly and the interior sprayed with a plastic sealant before installation of the cabin items. The cabin atmosphere was maintained at the equivalent of 8,000 feet up to an operating altitude of 33,000 feet. The crew of five was crammed into this space- with the pilot and co-pilot sitting side by side, behind them sat the primary gunner and the radio operator/secondary gunner. The bombardier/navigator sat in the nose compartment but could access the cramped flight deck via a floor panel by the co-pilot's feet. The bomb bay could carry up to 4,000 lbs of bombs and the three defensive turrets consisted of twin 50-caliber guns in dorsal, ventral and tail turrets that were operated remotely by the primary gunner and the radio operator/secondary gunner. Each gunner had a hemispheric observation window next to them and sighted the guns via a periscope system that protruded from streamlined twin fairings above and below the fuselage just aft of the flight deck. Initial plans for were for North American-designed turrets tied to a Sperry fire control system, but Sperry's resources were tied up with current production aircraft. It was decided to switch to General Electric for the remote fire control system and to have them be responsible for the turrets as well well. This imposed delays in the development as changes needed to be made to accommodate GE's equipment and systems. There was also a prevailing opinion at North American that Sperry's system was more advanced. A compromise was reached with the XB-28 defensive systems to use GE turrets and the Sperry sighting system. It's worth noting at this point that when the work on the remote turret fire control system on the XB-28 was under development, both Sperry and GE were working on getting the contract for the remote turret fire control system on the Boeing B-29 Superfortress- ironing out the kinks in the XB-28 gave GE valuable experience that helped the B-29 and made its system the production standard on the Superfortress.

The US entry into the Second World War slowed development of the XB-28 as priority shifted to production types and much of North American's resources were devoted to the production of the B-25 Mitchell and the P-51 Mustang. The maiden flight of the first prototype took place on 24 April 1942 at Mines Field (the site of today's Los Angeles International Airport/LAX) and the flight test program showed the XB-28 to be quite fast at high altitude, capable of 372 mph at 25,000 feet. Following the conclusion of North American's flight test program, the USAAF portion of the flight test program took place with the XB-28 operating out of Wright Field outside of Dayton, Ohio, for service trials. It was decided during the service trials that the third XB-28 prototype would be completed as a reconnaissance and photo-mapping aircraft designated XB-28A. North American was instructed to set aside work on the second XB-28 to get the XB-28A variant flying. The speed and altitude performance of the XB-28A was increased by reducing weight as well as installing more powerful versions of the R-2800 engines. The XB-28A made its maiden flight on 24 April 1943 (exactly one year after the first XB-28) but was unfortunately lost in flutter incident during dive testing on 4 August 1943 with the crew able to parachute to safety. At the time of the accident, design work on the production B-28 was underway with the most significant change being some extra scanning windows on the nose compartment for the bombardier/navigator.

By this point, however, it was realized that despite the outstanding performance of the XB-28, the realities of war showed that medium bombers already in service like the B-25 Mitchell and the Martin B-26 Marauder were most effective at low to medium altitudes flying interdiction missions where pressurization wasn't necessary. In the Pacific, B-25s equipped with extra forward firing machine guns were becoming very effective low level anti-shipping weapons while in the European theater, B-25s and B-26s operated most effectively at medium altitude (though some low level anti-shipping missions were flown in the Mediterranean against Axis vessels along the French and Italian coasts). The final nail in the XB-28's coffin was the Douglas product that was also first flown at Mines Field just a few months after the XB-28's maiden flight. The prototype Douglas XA-26 Invader first flew on 10 July 1942. It used the same engines as the XB-28, carried the same bomb load, but lacked pressurization which made it simpler to build and it only had a crew of three versus the crew of five on the XB-28. The sole XB-28 prototype was still at Wright Field at the time of the program's cancellation- it had its outer wings removed and sat out the war as a ground test article for pressurization tests before being scrapped.