The XB-48, like the more fortunate XB-45, originated in 1944, when the War Department concluded that jet propulsion was promising enough to warrant extension of the program, thus far centered on fighters and light bombers, to heavier aircraft with gross weights ranging from 80,000 to more than 200,000 pounds. Realizing that such an ambitious project could be fraught with difficulties, Army Air Forces (AAF) headquarters informed the Materiel Command and Air Services Command on 10 Augusts that in the beginning contracts for jet bombers of the medium and heavy categories would have to be let on a phased basis so that they could be readily terminated upon completion of any one stage of development. This cautious procedure was formalized on 15 August. About 2 weeks later the 2 commands merged to form the AAF Technical Service Command, which was redesignated Air Technical Service Command on 1 July 1945. This organization became the Air Materiel Command on 9 March 1946.
On 17 November 1944, the AAF issued military characteristics calling for a bomber with a range of 3,000 miles (minimum acceptable, 2,500); a service ceiling of 45,000 feet and a tactical operating altitude of 40,000 feet (minimums acceptable, 40,000 and 35,000 feet, respectively); and an average speed of 450 miles per hour with a high speed of 550. These characteristics were amended on 29 January 1945 to reemphasize that such aircraft needed to carry specific types of bombs, including the conventional M-121, a 10,000-pound "dam-buster" developed during World War II. The M-121, sometimes called the "Earthquake" bomb, was more often referred to as the "Grand Slam" bomb, a totally misleading nickname. Actually "Grand Slam" was the code name of a highly classified modification project strictly concerned with atomic matters. The "Grand Slam" modifications would allow the Convair B-36 to carry atomic bombs, which the Air Force believed might weigh more than 40,000 pounds. Since the 10,000-pound M-121, when properly dropped, could inflict the damage of a 40,000-pound bomb, curiosity and rumors most likely explained the ensuing confusion. As a matter of fact, the "Grand Slam" designation was also loosely applied to other conventional bombs of the M-121 category.
In accordance with the AAF's endorsement of "phase" contracts and based on the military characteristics of November 1944, a Martin proposal, submitted to the Air Technical Service Command on 9 December 1944, led to Letter Contract W33-038 ac-7675. Approved on 29 December, this initial document covered certain engineering services and completion by 1 May 1945 of 1 mockup of Martin's Model 223, designated XB-48 by the Air Technical Service Command. Tentative costs were set at $574,826. The letter contract of December 1944 was replaced on 27 March 1945 by a definitive contract, which reduced estimated costs to $569,252, including Martin's fixed-fee of $16,500.
Procurement of the XB-48 overcame many vicissitudes. In June 1945, 2 months after inspection of the XB-48 mockup, Martin submitted a proposal for 1 stripped and 1, 2, or 3 complete XB-48s. Accompanying cost figures, however, were immediately questioned. The Air Technical Service Command's surprise, it was soon ascertained that the estimated cost of $80.09 per pound for the XB-48 compared favorably to the $105.68 for the XB-45, but the AAF remained dissatisfied because the XB-48's engineering lagged behind the XB-45 and XB-46. Despite these concerns, the XB-48 project survived, and the initial contract was supplemented many times while negotiations went on. In March 1946, the contractor introduced a new proposal and offered to furnish 1 stripped and 1 complete XB-48 for about $10 million. This proposal was made on a fixed-price rather than a cost-plus-fixed-fee basis in order to conform to the policy set forth by the Air Technical Service Command in December 1945 on the procurement of experimental airplanes. Just the same, the Martin proposal of March 1946 had to be revised, and negotiations were not consummated until the end of the year. The final contract (W33-038 ac-13492), approved on 13 December 1946, superseded Contract W33-038 ac-7675 which, as amended, had reached an estimated future cost of $10.9 million. Only some $500,000, covered by the initial letter contract, were unaffected. For the same amount, the new contract promised 2 XB-48s, spare parts, and a bomb-bay mockup. Also, the first XB-48 was to be flight tested and delivered by 30 September 1947; the second one, by 30 June 1948. Finally, all wind tunnel tests were to be completed by 1 January 1947.
Development and testing of the 2 XB-48s were delayed by engine difficulties. General Electric turbojet engines were installed, the first XB48 being powered by 6 J35-GE-7 (TG-180-B1) engines; the second, by 6 J35-GE-9s (TG-180-Cls). Since the engines were in an even more experimental stage than the airplanes, it took time to get them to operate properly. Also, like every new engine, the J35s were in short supply. Still, the first XB-48 would go through 14 engines during its first 44 flights.
The sleek, all-metal, high-wing XB-48 presented many special features, the most outstanding one being the tandem bicycle landing gear necessitated by the airplane's wings, too thin to house conventional landing gear with bulky retracting mechanisms. Martin had experimented with a 4-wheel bicycle landing gear on an XB-26H and concluded that such an arrrangement was feasible. Bicycle-type landing gears were later used by other jet bombers, including the B-47. Other novel features were the number of engines, 6 as compared to 4 on the other proposed medium bombers; the turbojet engine's installation, encased in pods (3 under each wing) in a lift section with air ducts between the pods; and also adjustable tail pipes on the engines. The 3-crew arrangement was also unusual. The pilot and co-pilot were seated in tandem under a canopy-type inclosure, similar to that found in high-speed fighter planes, while the bombardier-navigator was seated in the aircraft's nose. The XB-48 had retractable bomb-bay doors, a feature that sprang from the fact that all new medium and heavy bombers had to be capable of carrying the so-called "Grand Slam" bombs, as well as the cumbersome atomic bombs of the period.
The XB-48, the first US. 6 jet bomber to fly, made its initial flight on 22 June 1947. The experimental plane took off from Martin's airfield at Baltimore and landed some 80 miles away at the Patuxent Naval Air Station, also in Maryland. The 38-minute flight was not a great success. At 10,000 feet, the Martin pilot discovered that the right spoiler aileron snapped up too rapidly. On landing, the XB-48 drifted across the runway. Rudder steering was attempted, but the rudder was ineffective with the full use of brakes. In addition, the brakes overheated and stopped working. The aircraft finally came to a halt off the runway with no damage, even though both tires were worn through.
The second XB-48 did not fly until 16 October 1948, some 3 months behind schedule. The 30-minute flight was satisfactory, but of relative unimportance since the future of the experimental program had already been decided. The Air Force gave the contractor the option to eliminate all flaws or to pay a lump-sum penalty of $25,000. In January 1950, Martin agreed to pay the penalty.
Martin pilots tested the first XB48 52 times, for a total of 41 hours; the Air Force, 50 times for a total of 64 hours. The second XB48 was also thoroughly tested. The contractor put in 14 hours, accumulated in 15 flights; the Air Force, 49 hours, reached in 25 flights. Results of the first XB-48's flight test program revealed that the aircraft did not meet the Martin guarantees. The XB48 was 14,000 pounds overweight; the nose wheel was too sensitive; turbulence occurred in the bomb bay when the doors were open; and metal chips, deposited by disintegrated test stand hydraulic pumps, shattered the hydraulic system.29
The experimental B-48 program agreed upon in December 1946 was not curtailed. Yet, in spite of the contractor's efforts, no production program followed. Although no firm commitment would be made before many months, planning for the procurement of B-47 production models began in December 1947, right after the XB-47's first flight-a poor omen for the B-48, initially flown in June of the same year.
In the spring of 1948, after early experimental flight information had been obtained for both the XB-47 and the XB-48, the Air Force conducted an evaluation to determine which of the 2 planes could best satisfy the urgent need for a high-speed, high-altitude medium bomber. The evaluation confirmed that the performance of the XB-47 was appreciably better than that of the XB-48. It was also apparent that the XB-47 design provided possibilities for growth which surpassed those of the XB-48. The XB-47's swept-back wing would enable it to attain higher speeds, and its simpler pod-nacelle arrangement minimized the problem of incorporating newer and more efficient jet engines as they became available. The end of the B-48 production program became official in September 1948, when the Air Force ordered the first lot of 10 B-47s.
Early in 1949 Martin attempted to rescue the B-48 production program and proposed to modify the second XB-48 by removing the J35 engines and nacelles and installing 4 XT-40A propeller turbines in new and repositioned nacelles, at an estimated cost of $1.5 million. Actually, the reconfigured XB48 would become a prototype of the Martin Model 247-1, an airplane, the contractor insisted, capable of competing with the B-47, B-50, and B-54. On paper, Model 247-1's performance looked good, but the Air Force did not believe the proposed reconfiguration could be accomplished for the amount of money estimated by the contractor. In addition, since the XT-40A turboprop was a Navy-developed engine, it was doubtful that Martin could obtain enough engines to complete the reconfiguration on schedule. Finally, and of overriding importance, senior Air Force officials believed that turbojet aircraft "currently offered greater promise than turboprop installations." Thus, on 31 March 1949, Martin was formally told that the Model 247-1, like the original XB-48, was a dead issue.
The Air Force accepted the first XB-48 on 26 October 1948, but only conditionally. The acceptance became final in 1950, when Martin paid the $25,000 penalty assessed by the Air Force because of the aircraft's several defects. The second XB-48, also conditionally accepted on 26 October 1948, was finally accepted on 23 February 1949, after the contractor completed various modifications.
The total cost of the XB-48 development program reached $11.5 million. Of this amount, less than $500,000 pertained to the letter contract of December 1944. The rest covered the final contract of December 1946 and represented an increase of about $100,000, justified by various changes ordered by the Air Force.
In the fall of 1949, the first XB-48 was cannibalized to provide parts for the second XB-48. The latter aircraft was scheduled for many tests, including tests on the F-1 autopilot, jet engine cooling system, and a hydraulic system for jet engines. The proposed tests, however, were canceled. The Air Force decided to use the second XB48 as a test-bed for "badweather" flight items, including a badly needed deicing system. Completion of the thermal anti-icing survey test program in mid-1951 paved the way for the second XB-48's end. In September, the aircraft was flown to Phillips Field, Aberdeen Proving Ground, Maryland, where the strength of the XB-48 structure was tested until the aircraft was totally destroyed.