Barely a footnote in any listing of World War Two military aircraft, the Boeing (nee Stearman) XAT-15 has nevertheless been ill-served in terms of accurate reporting, as we shall see. According to Edward Phillips’ book “Stearman Aircraft: A Detailed History” (Specialty Press, 2006) it was the prototype of a twin-engined crew trainer, “
However, special schools for gunners, bombardiers, navigators, and pilots developed during the war, and the need for the XAT-15 disappeared. As a result, production was halted and materials salvaged.” (see (a) below).
Looking elsewhere, the generally reliable website “1000 Aircraft Photos” (
https://1000aircraftphotos.com/Contribu ... e/8041.htm) states for the XAT-15 that,
“Two aircraft were built…and orders for 1,045 were issued, however, these were cancelled when the USAAF decided against composite construction for advanced trainers." (see (b) below)
Finally, the XAT-15 Wikipedia pages (
https://en.wikipedia.org/wiki/Boeing_XAT-15) tells us that,
“Due to shortage of materials, the aircraft was built of welded steel tube covered with plywood, with wooden wings and tail unit. The two aircraft were delivered to the USAAC, but after the country's entry into the war a change in priorities resulted in the planned order for more than 1,000 aircraft not being placed.” (see (c) and (d) below)
To take these points in turn,
(a) – cancellation of the XAT-15 was not brought about because “…the need for [it] disappeared”.
(b) – cancellation of the XAT-15 was not because, “…the USAAF decided against composite construction for advanced trainers”.
(c) – cancellation of the XAT-15 was not due to, “…a change in priorities”.
(d) – there is no evidence that “1000 aircraft” on AT-15 contracts were planned.
To begin with, an explanation of the XAT-15’s primary manufacturer is in order. Initially designed by the Stearman Aircraft Corporation of Wichita, Kansas, since 1934 this company had been operating as a division of Boeing though still retaining its original name. This remained the status quo until September 1941 when Stearman was redesignated “Wichita Division, Boeing Aircraft Company”. Thus, though much of the initial XAT-15 design and contractual work was done under the auspices of “Stearman”, by the time that production contracts were agreed for service AT-15s, the aircraft was very much a Boeing product. For the purpose of this text, I have attempted to place “Stearman” and “Boeing” in their correct context but for clarity’s sake both terms are interchangeable.
The XAT-15 story begins in early 1941 when Air Corps chief Major General Henry H ‘Hap’ Arnold canvassed the opinions of various aircraft manufacturers regarding the manufacture of training types. On 24 February he sent a memo to Brigadier Gen George H Brett (Assistant Chief of Air Corps) as follows.
“…my attention has been brought to the fact that if the manufacture of this type [of]
airplane were limited to those types using wood and fabric, or tube and fabric construction, considerable saving of material would result of the type that is sorely needed for tactical airplane construction. The saving of material for tactical types is not so important now as it will be six months to a year hence. However, steps must be taken now looking toward the future.” “With the above situation in mind, prepare a study of the relative merits of limiting training plane construction to the use of wood and fabric or tube and fabric as against the construction of these types in metal, taking into consideration cost, man hours, availability of material, safety, durability, maintenance, etc.” Subsequently, on 7 March 1941 the Army Air Force issued instruction CTI-209 inviting the development and procurement of Bombardment Trainers (or “Bombing Crew Trainers” to use terminology also in use at this time). On 18 March a teletype (Brig Gen Oliver P Echols, Chief, Materiel Division to Assistant Chief, Materiel Division) advised that Brigadier General Carl Spaatz (Chief, Plans Division, Office Chief of the Air Corps) had placed his weight behind the push toward alternative construction methods and had recognized that the likely attrition of tactical aircraft (crews) would result in a greater need for combat crew trainers. Echols referred to the Stearman Aircraft Company, which had begun developing a suitable type, the Model X-120E7. He stressed in his message that,
“…we will probably have to place production orders for these airplanes much earlier than we expected.” Concurrently, adherence to Hap Arnold’s proposal for alternative construction methods was incorporated into the requirement for new combat crew trainer types; the request for a study into the relative merits (toward which $200,000 had been agreed in FY41 budgets) was thus absorbed into the development of service types.
But Stearman was not alone in this field, for less than a week after issuing CTI-209, the Experimental Engineering Section at Wright Field had been able to respond to Executive, Materiel Division in Washington DC that,
“Action has been initiated for completion of preliminary ground work for procurement of two XAT-13 combat crew trainers from Fairchild. These airplanes will conform to desired type of construction with minimum use of critical materials.” The competing Fairchild XAT-13 project differed from the Stearman design in being largely constructed from a moulded wood composite material, formed in a novel “Duramold” process described at the time as ‘plastic’. As such its new technology promised great savings in terms of cost, weight and availability of raw materials and in this respect the Fairchild project had something of an advantage over the more conventional Stearman design. Later in 1941 when higher-than-expected costings for the XAT-13 were being challenged, General Echols replied that,
“The Fairchild airplane will be procured. The Duramold process must be developed as it is an important development project”. The Fairchild product also benefitted in having several months’ head start over Stearman’s design, and as a result a fixed-price contract was approved on 9 April 1941 for one XAT-13 prototype with Change Order CO-1 of 29 April adding a second example, later re-designated XAT-14 as a result of a change from radial to inverted-vee engines on the latter example.
Meanwhile Contract ac-18967 was approved on 19 May 1941 for Stearman to supply two of its Model X-120E7. These would be designated XAT-15 with serial numbers 41-23162 and 41-23163 and though the XAT-15 was not a ‘plastic’ aircraft in the true sense (though it was described as having ‘composite’ construction) it did offer the benefits of using a minimum of metallic structure. It comprised a fabric-covered steel tube fuselage with plywood wings in a tail-dragger configuration with power derived from two Pratt & Whitney R-1340-AN-1 radials. The purpose of all three types (XAT-13, XAT-14 and XAT-15) was described by the Office of Chief of Air Corps as for,
“the training of a bombardment crew as an integral operating unit”. The XAT-15 was given a “Restricted” classification and assigned Air Materiel Command experimental project number MX-123. Below the two XAT-15s circa late 1942:
But as the two manufacturers geared up for production, the relative costings for the two distinctly different types of airframe construction became clearer. For while Fairchild submitted a $60,000 unit price for 75 production AT-13s, Stearman quoted just $37,000 for its AT-15. More worrying, Fairchild estimated that each AT-13 would take 22,000 man-hours to construct while Stearman quoted almost half of that figure. In August 1941, Contracts Section at Wright Field recommended that the Fairchild type should not be procured, highlighting the XAT-13’s method of construction as a risk by noting that it,
“…incorporates the so-called plastic construction throughout utilizing the Duramold process”. Going further, Contracts made the point that,
“…it appears that the Fairchild plastic construction as utilized in this airplane will always cost considerably more than the conventional type [of]
construction”. Nevertheless, as noted above, General Echols was keen to push ahead with developing the Duramold process and on 31 October 1941, fixed-price Contract ac-22236 for 75 production AT-13s plus spares for nine equivalent aircraft was approved and valued at $4,329,908.75. It must however have raised some eyebrows that the same day, Boeing (Stearman) had been awarded cost-plus fixed fee Contract ac-20858 for 75 AT-15s valued at just $3,066,838.15.
With the need to expand production beyond the limited capacity available in Wichita, Contract ac-25601 was approved on 30 March 1942 for the manufacture of 360 examples by the McDonnell Aircraft Corporation in Robertson, Missouri. The estimate for this cost-plus fixed fee contract was $14,437, 366.00 with a fixed fee of $851,241.96. On 11 April a further contract, ac-24609, was approved for 360 AT-15s to be built by the Bellanca Aircraft Corporation of New Castle in Delaware. This contract was priced identically to the McDonnell agreement and called for first delivery within six months, ramping up to a production rate of 30, eleven months from approval. The first Boeing XAT-15 prototype had in the meantime taken to the air from Wichita on 9 April 1942 and thus, by the time of its maiden flight the AT-15 had received a total of 795 orders. It is worth noting that in the same timeframe, Fairchild’s AT-13 had received just 75 production orders, and this demonstrates the differing priorities between the two types as well as indicating a sensible reluctance to commit too heavily on the unproven construction processes inherent in the AT-13 design. Nevertheless, Fairchild had received a considerable amount of funding in order to set up a devolved production line in Burlington (rather than at the company’s home field in Hagerstown) so the AAF was playing a sensible game in prioritising a ‘conventional’ type whilst at the same time providing the resources to take advantage of Fairchild’s aircraft if it showed sufficient promise. With prototype construction underway at Hagerstown (where the first few aircraft would be constructed), on 4 May 1942, the Secretary of War approved cost-plus fixed fee Contract ac-25806 for the manufacture of 400 further AT-13s, described at that time as, “Advance Trainer Type (Crew)” but still totalling just over half the amount of AT-15s on order.
And for a while it looked like the AAF had played a canny game, for despite being delayed in development compared to the XAT-13, Boeing’s ‘conventional’ XAT-15 had taken to the air three months ahead of the first Fairchild type (the #1 XAT-13 would fly for the first time on 10 July 1942 and the XAT-14 on 25 November the same year). Moreover, the XAT-13 would experience a number of incidents and accidents during its short life, the first occurring at Hagerstown on 28 July when a down-lock mechanism failed on the right main gear.
Sadly the summer of 1941 marked high tide for the AT-15 and the first hints of problems with Boeing’s design began to surface. On 1 June the company contacted Materiel Center at Wright Field to present a revised delivery schedule which amounted to a four-month delay. In particular the newly-flown #1 XAT-15 would not be available for USAAF Engineering and Flight Test Boards until 15 June and the #2 aircraft three months later. The first pair of Wichita-built production AT-15s would then follow during December before a peak of twenty aircraft per month was reached at the Boeing plant in April 1943. The full estimate was presented as follows:
First Model XAT-15 Airplane – June 15, 1942 (Delivery to Engineering & Flight Test Boards)
Second Model XAT-15 Airplane – August 1, 1942
Two Model AT-15BO Airplanes, plus static test article, during December, 1942
Five Model AT-15BO Airplanes during January, 1943
Ten Model AT-15BO Airplanes during February, 1943
Fifteen Model AT-15BO Airplanes during March, 1943
Twenty Model AT-15BO Airplanes during April, 1943 and
20 per month thereafter until contract quantities have been exhausted.
Boeing blamed these delays on the loss of key engineers to the B-29 project, exacerbated by a local lack of qualified staff in general. But it seems likely that another reason was the knowledge (not disclosed at this time) that the XAT-15 was struggling to meet its performance requirements.
In a 21 November 1942 letter to Production Division at Wright Field, Franklin O Carroll (Chief, Experimental Engineering Section) went into great detail on the various failings of the XAT-15 and the “bombing crew trainer” concept in general. Addressing the basic concept, Carroll wrote that,
“It appears that the requirements for this type of trainer were optimistically set up, as evidenced by the facts that the gross weight of the XAT-13 increased approximately 1600 pounds from an estimated weight of 10,839 pounds to an actual gross weight of 12,401 pounds, and the gross weight of the XAT-15 increased approximately 2300 pounds from an estimated weight of 12,061 pounds to an actual gross weight of 14,355 pounds.”Though neither type came in for any praise, the situation for the XAT-13 was at least more favourable, for despite its weight increase the Fairchild aircraft was still around 2,000 lb lighter than the Boeing type, using the same engines. But for the XAT-15 the position looked bleak and Brigadier General Carroll delivered the worrying news that even at a lighter weight of 13,954 lb, the type had been unable to maintain altitude on a single engine whilst being tested at Wichita. Further trials at Wright Field (now at 14,355 lb) merely confirmed this fact. Carroll did concede that some of the aircraft’s overweight was due to the installation of Government Furnished Equipment and late changes requested by the Wright Field Materiel Center, but this only amounted to 700 lb; the remaining 1600 lb of overweight was put down to,
“…a very poor weight control system coupled with a very inefficient structural design from a strength-to-weight ratio viewpoint.” However the bad news did not end there, and Brigadier General Carroll went on to list the many failings which had been noted since the XAT-15 had arrived at Wright Field,
“A structural failure of the centre section wing during the structural demonstration flights occurred at a positive load factor only 55 per cent of the ultimate design and although the exact cause for the failure is not known, it is probable that it was due to poor design and workmanship. The Contractor has made certain reinforcements to the production model design, but this can only result in a still further increase in the weight of an already underpowered airplane. The Contractor estimates the weight of the redesigned airplane at 13,500 pounds but analysed the structure for an alternate gross weight of 14,100. This estimate should be considered optimistic, and even if correct would still result in an underpowered airplane.
“The Bendix brakes were very unsatisfactory when first installed and Boeing claimed insufficient power was being furnished by the master cylinders. Various combinations of master cylinders and linings were tried over a period of more than four (4) months with no improvement in braking action. Finally, at the specific request of the Materiel Center, the brake operating linkage was changed to a more rigid type with a 4 to 1 mechanical advantage, and the braking provided thereby ample.
“During the first series of demonstration flights, the side panel on the right side of the bomb bay pulled loose from its fasteners causing considerable damage as it swept by the horizontal stabilizer.
“During the first series of demonstration flights, the outer panel wing skin cracked in several places, and it was decided that the 90-degree [alternating ply layers] should be replaced with 45-degree plywood.
“During a test flight at Wright Field, several inspection doors came off. An inspection afterwards disclosed that they were attached by means of fasteners intended for use in attaching [electrical] junction box covers.
“During another flight, the negative pressure around the landing gear well doors caused them to gap open and the doors on the left side were stripped away with parts of the nacelle. The slip-stream carried the torn-away parts against the wing and caused severe damage to both the wing and tail.”And then the bombshell; Carroll deemed that the production AT-15 would be unsuitable for its intended purpose of training bomber crews, concluding that,
“Consequently it is recommended that all contracts for AT-15 airplanes be cancelled, and that all engineering and production efforts now being expended on this project be discontinued.” On a positive note, Carroll stated that cancellation would allow Boeing to concentrate on the vital B-29 project, thereby alleviating a shortage of personnel in Wichita.
Strangely, Carroll’s 21 November letter failed to disclose one rather important point: that the #1 XAT-15 had crashed just ten days earlier. Despite being assigned to Wright Field, 41-23162 was on 11 November being flown by a Boeing crew comprising David “Deed” Levy (35) as pilot and Glen Chambers (27) co-piloting. As the aircraft approached the town of New Carlisle (just 10 miles NNW of Wright Field) it caught fire and crashed into a cornfield on the Homer Scott farm; fortunately both crew were able to bail out though Levy suffered a broken ankle on landing.
In fact, the overall bombing crew trainer programme had just received a further setback when on 19 November the Joint Aircraft Committee revised production outputs. The AT-15 line at Wichita would reduce by ten per month in order to place emphasis on the production of 275 PT-17s and would start AT-15 deliveries in February 1943 before reaching peak output of 20 per month in June. Its two subcontract lines at Bellanca and McDonnell would thereafter reach peaks of thirty AT-15s per month in October and September 1943, respectively. Fairchild was also affected by this revised planning, in this case so that priority could be given to production of PT-19 primary trainers. Below is a 25 November 1942 Joint Aircraft Committee estimate of planned production for all three AT-15 lines up to the end of 1943:
But on 28 November 1942 the axe finally fell on the Boeing’s AT-15 project. Citing the poor results of flight testing, Major General Muir S Fairchild (Director of Military Plants) made it official that no AT-15s would be contracted for, and that all work on this aircraft would cease. It took some time for contractual revisions to be finalised but on 23 December Boeing, Bellanca and McDonnell were advised that their contracts for AT-15 aircraft had been cancelled.
As a result of this cancellation, Fairchild’s AT-13 project now became the sole candidate for the bombing crew trainer requirement, and the Bellanca and McDonnell AT-15 contracts were effectively revised for similar quantities of the Fairchild type. But this did not bring an end to the troubled programme, for already in November 1942 the role requirement had changed to “Gunnery Training Airplane”, thereby marking the point at which the idea of a bombing crew trainer was finally dropped. Further, the Bellanca and McDonnell contracts were soon amended again, this time for production of AT-14s with the inverted V12 engine. By February of 1943 the radial-engined AT-13 idea had been fully dropped and a revised AT-14 (now known as the AT-21) was settled upon. Large orders were approved for this aircraft, which still employed the Duramold type of construction but sadly the AT-21 would also prove to be a failure.
One final AT-15-related action occurred in early 1943 when Bellanca and McDonnell were asked by the AT-21 Production Committee to assess if any parts from the cancelled programme could be used in the manufacture of the Fairchild product. The response is not known but it appears that minimal, if any, items would have been suitable for repurposing.
The surviving XAT-15, s/n 41-23163, appears to have been returned to Boeing at Wichita and though officially accepted by the USAAF on 10 July 1943 it was immediately placed in Class 26 (ground instructional airframe). The aircraft was declared ‘reclamation complete’ on 14 July 1945 but had likely been scrapped sometime prior to that.
For those wishing to research into the XAT-15 in more detail, there is a lot of primary-source information contained in Air Force Historical Research Agency Roll No. A2064; much of that detailed above came from this source. Other primary-source documents are available as part of the NASM Giuseppe M Bellanca Collection. Various AMC reports and correspondence are also available at The National Archives.
Contracts and dataStearman XAT-15: Contract ac-18967 approved 17 May 1941; (2 a/c); P&W R-1340-AN-1; 4-place, mid-wing, full cantilever monoplane intended primarily for training bombing crews. Has a plywood wing and welded steel fuselage. Armament consists of one .30 cal machine gun mounted in a power operated turret with 500 rounds of ammunition and ten 100-lb bombs. Radio command set SCR-283, radio compass SCR-269-A, marker beacon RC-43-A, interphone RC-36, filter equipment RC-32.
Boeing AT-15: Contract ac-20858 approved 21 October 1941; (75 a/c); P&W R-1340-AN-1; basically the same as XAT-15 with some changes in equipment. Armament consists of one .30 cal machine gun, Type M-2 with 500 rounds of ammunition mounted in fuselage turret. Ten 100-lb training bombs, Type M38A2 carried in fuselage bomb bay. Radio command set SCR-274-N; radio compass SCR-269; marker beacon RC-43-A, interphone RC-36, microphone T-30A; filter RC-32.
Boeing AT-15-BL (Bellanca): Contract ac-24609; (360 a/c) approved 11 April 1942; P&W R-1340-AN-1; same as AT-15-BO except manufactured by Bellanca
Boeing AT-15-MC (McDonnell): Contract ac-25601; (360 a/c) approved 30 March 1942; P&W R-1340-AN-1; same as AT-15-BO except manufactured by McDonnell
Login
Register
FAQ
Search
