Classic Aircraft profiled every issue -

Issue 14 - P-38 Lightning: Tricycle and twin – defining a classic

The X-608 and X-609 proposals were issued from the Wright Field Pursuit Projects Office of the US Army Air Corps. Wright Field was, at the time, the centre for testing aviation technology and developing new military aircraft, fulfilling a similar role to Martlesham Heath in the UK or Rechlin in Germany.

The two remarkable men responsible for the proposals were first lieutenants at the time – Gordon Phillip Saville, who would end his distinguished career as a major general, and Benjamin Scovill Kelsey, who would become a brigadier general. Both were keenly aware of the shortcomings of the aircraft in US service at the time, particularly in comparison to those entering service in Germany, Japan and the UK.

This awareness also extended to the limitations of the official Army Air Corps rules and regulations which governed the acquisition of new aircraft, rules that were the inevitable result of decades of inter-service rivalry and distrust of aviation as a viable arm of the US forces. Both men became adept at finding loopholes in those rules to draw up recommendations and proposals for developing and acquiring the kind of systems and high performance aircraft they foresaw would be essential in the near future.

They may be said to be as responsible for the P-38 as Hall Hibbard and Kelly Johnson. Saville and Kelsey often only receive a mention as the architects of the proposals, yet they influenced aviation history to an incredible degree.

Gordon Phillip Saville

Born in Macon, Georgia, on September 14, 1902, Gordon P Saville attended the Universities of Washington and California, before being commissioned as a second lieutenant of infantry in the US Army Reserve in November 1923. While on active duty, he spent time at Crissy Field in San Francisco, which began his lifelong interest in military aviation.

He joined the Army Air Service as a flying cadet in March 1926, completing his basic and advanced flying training by September the following year. He was posted to the 5th Observation Squadron at Mitchel Field, New York, and served as adjutant to the field’s commander, the future Chief of the Air Corps, Lieutenant Colonel Benjamin Foulois. Foulois was impressed with Saville, and recommended that he attend the Air Corps Tactical School at Maxwell Field in Alabama, from which Saville graduated, top of his class, in May 1934.

He spent the next three years as a fighter instructor at the school, during which time he worked with Claire Lee Chennault (later of Flying Tigers fame). He was an advocate of an air defence strategy led by a strong fighter force, as he felt fighters were undervalued in the defensive role and that too much emphasis was being placed on the bomber force. Saville also developed new control methods for fighter forces to improve co-operation between ground forces and tactical air power.

In February 1937, as the leading proponent of fighter tactics, he worked with Benjamin Kelsey on the two proposals for new fighters. In 1938, he attended the Command and General Staff School at Fort Leavenworth, Kansas, before being assigned to the plans division of the Army Air Corps in June 1939. This staff post was followed by his return to Mitchel Field as the assistant intelligence and operations officer of the Air Defense Command. Here he put his theories of radar ground control for fighters into practice, proving the validity of a co-ordinated fighter defence in exercises in August 1940. Now a major, he visited Britain and studied the RAF’s defences that had worked so well in the Battle of Britain and drafted a doctrine for air defence which, while widely admired, was not adopted. From this point on, Saville was to be one of the architects of an effective US air defence, becoming director of air defence at the US Army Air Forces Headquarters in Washington DC.

In January 1942, as a lieutenant colonel, he formed one of the first operations research groups utilising the skills of mathematicians and defence experts to solve the problems of defending the Panama Canal. In recognition of his expertise and management skills, Saville was promoted again in November 1942 to brigadier general. Once it was clear that no direct threat to the US existed, he was appointed as director of tactical development at the Army Air Forces School of Applied Tactics at Orlando, Florida, where he reorganised and co-ordinated the efforts of the research, development, testing and acquisition branches of the USAAF.

In July 1943, he became the chief of staff of the Mediterranean Air Command, then the commander of the XII Fighter Command in October 1943. In January 1944, he became the deputy commander of the XII Air Support Command and implemented close air support tactics between air and ground forces which were used in the Allied invasion of Normandy, and again in the invasion of Southern France.

Given his outstanding record, Saville was promoted to major general in June 1944 and given command of the First Tactical Air Force in January 1945. He returned to the US, becoming the commander of the III Tactical Air Command in March 1945. Several other posts followed, including two years in Brazil, before he was called back to Mitchel Air Force Base in June 1948 where he became the commanding general of the newly formed US Air Force’s Air Defence Command in November. His plans for a new radar network to cover the continental US were adopted in March 1949, alongside which Saville was responsible for such programmes as the F-86 Sabre to replace outdated Second World War fighter types.

In January 1950, he became deputy chief of staff for Development, Air Research and Development Command, and initiated a new fighter interceptor, fire control system and missile programme that resulted in the Convair F-102 Delta Dagger and the Hughes AIM-4 Falcon missile of 1956. In July 1951, after a distinguished and creative military career, Saville retired from the USAF.

His civilian career was just as impressive. He joined Ramo-Wooldridge, the company formed from the fire control system development team at Hughes, becoming director of military requirements. A merger with Thompson formed military electronics and systems giant TRW, and Saville became vice president until he retired in 1963, becoming a consultant and a cattle rancher.

Saville was to remain in these last two careers until his death on January 31, 1984. He was one of the visionary leaders who taught the value of co-ordinated defensive and tactical air forces to the world, and must be seen as one of the architects of modern air power.

Benjamin Scovill Kelsey

Benjamin Kelsey had flying in his blood. Born on March 9, 1906, in Waterbury, Connecticut, by the age of 15 he had completed a pilot’s course at the Curtiss Flying Service of Garden City, New York.

His fascination was not just with flying, but with the mechanics of aviation too. To this end, he attended the Massachusetts Institute of Technology (MIT), graduating with a bachelor of science degree in mechanical engineering in June 1928. Kelsey remained at MIT, teaching and continuing his research as part of the aeronautics department, but managed to obtain his transport pilot’s licence at the same time by flying privately and for a variety of commercial concerns.

Military developments attracted his interest, so he joined the US Army Air Corps (USAAC) as a second lieutenant in May 1929, his first posting being to Mitchel Field. Here, he worked on developing blind flying systems as part of the Guggenheim Fog Flying Laboratory, acting as the safety pilot on Jimmy Doolittle’s first full instrument flight on September 24, 1929. In 1930, Kelsey completed his USAAC primary and advanced flying courses, which must have been somewhat redundant considering his already extensive flying experience.

The following year he finished his master of science degree in aeronautical engineering at MIT, before being posted to the 20th Pursuit Group at Mather Field in California. In 1934, Kelsey was posted to Materiel Command’s engineering section at Wright Field near Dayton, Ohio, to become the fighter project officer. As the sole officer responsible for USAAC fighter development, this was a busy position, liaising between the Corps and aircraft companies, as well as acting as test pilot – a position he was exceptionally well qualified for.

Despite all of this, Kelsey also maintained his work on developing blind flying hardware and techniques, being promoted to first lieutenant on October 1, 1934. With his engineering background, Kelsey was interested in many of the innovations coming from industry, particularly in the field of aero engines as this had been a speciality of his at MIT. He took a personal interest in the development of the Allison V-1710 engine, as he recognised the potential for greater high altitude performance and speed from liquid-cooled rather than air-cooled aero engines.

In November 1936, Kelsey began an extensive flight test programme on the Consolidated XA-11A, an A-11 two seat attack aircraft converted by Bell Aircraft to take the 1000bhp Allison XV-1710-7 engine. The engine passed all of the exacting trials required by the USAAC to test reliability and performance, and Kelsey would use his knowledge of the engine’s capabilities to make it a feature of a number of future fighter specifications.

In his official capacity, Kelsey was to fly an enormous number of prototypes and trial aircraft, and this experience increased his frustration with the USAAC restrictions regarding fighter aircraft, known as pursuit aircraft at the time. When he teamed up with Gordon P Saville, the two first lieutenants wrote the X-608 and X-609 proposals for new fighters using very careful wording to get around the regulations.

The results of these proposals were so successful that Kelsey was promoted to captain in May 1939, continuing his work on fighter development. He next drew up a proposal for an aircraft to replace the Curtiss P-40, the Curtiss XP-46, placing an order for two prototypes in September 1939. This was to be an advanced fighter with a laminar flow wing, but was cancelled at a higher level when it was realised that it would disrupt mass production of the P-40.

The engine installation from the XP-46 was used in the P-40D, which was found to perform better than the XP-40 anyway. Kelsey was unwilling to let the project concepts go, and approached his boss, Colonel Oliver Echols, about offering the design to the Anglo-French Purchasing Commission, which was in the US looking for new combat aircraft. The commission was told that if it could find an aircraft manufacturer with the excess capacity to build the new aircraft, then the NACA aerodynamic data collected for the XP-46 would be made available.

This data was later sold to North American Aviation, which accepted the commission’s proposal for a new fighter. The data was used, although to what extent remains unclear, in the firm’s already well advanced NA-73 design which had certain similarities to the XP-46. The NA-73 became a fighter the British were later to name Mustang.

Between May and July 1940, Kelsey travelled widely in France and the UK as an assistant military air attaché, to assess British, French and German fighter designs. He also saw the need for a long range ferry route to bring aircraft directly to Europe, as he correctly assessed the German naval threat to transatlantic shipping convoys. He returned to Wright Field as the head of the pursuit branch of the production engineering section, and in March 1941 was promoted to major.

The next year saw Kelsey extensively test flying the Supermarine Spitfire. He found its lack of range a great drawback and asked Lockheed to design drop tanks for the P-38 to extend its range, even though this went against USAAC doctrine at the time. In January 1942, Kelsey was promoted to acting lieutenant colonel and was attached to the VIII Fighter Command at Bangor, Maine, to plan and prepare for the first transatlantic ferry flights.

He flew a P-38F of the 14th Fighter Group across to the UK in July, on the first ferry flight by fighters. He returned to his post at Wright Field in September, and became the chief of the flight research branch of the flight test division in July 1943. November saw Kelsey back in Europe, where he became the deputy chief of staff of the IX Fighter Command, before moving to the headquarters of the 8th Air Force as chief of operational engineering.

In February 1945, he returned to the materiel division at Air Corps Headquarters, then moved back to Wright Field as chief of all weather operations. After a number of staff positions, in 1948 he was promoted to colonel and attended the National War College, graduating in June and remaining on the staff as an instructor for nearly four years. He was posted to become deputy director of research and development in the Office of the Deputy Chief of Staff for Development at Air Force Headquarters in June 1952, and was promoted to brigadier general in September.

While he was in the post, Kelsey was a member of Hugh Dryden’s Research Airplane Committee and was instrumental in bringing the X-15 research aircraft programme to fruition, the contract being awarded to North American on September 30, 1955. In December that year, Kelsey retired from the US Air Force, but kept flying privately, lecturing on aeronautical subjects and producing many papers on engineering and aviation matters.

A range of awards followed, including tenure of the Lindbergh Chair at the National Air and Space Museum between 1979 and 1980. During this time he wrote the definitive work on US military aircraft development before and during the Second World War called The Dragon’s Teeth?: The Creation of United States Air Power for World War II.

After a lifetime in a huge variety of cockpits, making ground breaking decisions and driving innovation at the cutting edge of military aviation, not to mention being responsible for the success of a number of iconic and important aircraft, Ben Kelsey died of cancer on March 3, 1981.

Circumventing the rules

To return to the story of the P-38, the Circular Proposals X-608 and X-609, were issued in February 1937. Saville and Kelsey pooled their knowledge of air defence needs on the one hand, and current fighter performance and the likely capabilities of new aircraft on the other. The biggest problem they faced in ensuring the new aircraft had adequate performance was the set of rules and regulations which governed the acquisition of new aircraft.

One of the most onerous restrictions placed on new fighter aircraft by the USAAC was a limitation of 500lb (227kg) on the total weight of their armament, their guns and their ammunition. Kelsey wanted to at least double this to 1000lb (454kg) to allow future fighter pilots a chance to achieve air superiority. Saville wanted greater range and endurance to allow for longer fighter patrol and support missions.

Neither of these characteristics were allowable under the rules for fighters, but Kelsey and Saville found a way to get round them which was both simple and ingenious. They didn’t call the new aircraft specified in the proposals fighters, they called them interceptors. It was the first time that word had been used to describe a US military aircraft programme. This change in terminology also allowed them to specify that one of the aircraft was to be twin-engined – another characteristic banned by the fighter regulations.

This pedantic application of strict statutes to new aircraft may seem ludicrous to modern eyes, where multirole platforms rule the day, but it must be remembered that at the time aviation was viewed with suspicion by the armed forces as the unwanted child of technology. Nobody knew exactly what to do with it, but at the same time, no one wanted another service to have sole ownership of it or responsibility for it.

For example, the US Navy tried to clip the wings of the USAAC bomber force, as it considered the battleship fleet to be the long range defender of America. In turn, the USAAC bomber force viewed with great suspicion any attempt to develop a long range aircraft that was not a bomber, as it saw long range operations as solely within its purview. The competitiveness that resulted in these restrictions was not just inter-service, but also between elements of the same service. Although these costly and inefficient rivalries may seem a relic of their time, it is a sad truth of military life that they continue to a greater or lesser extent today.

The interceptors defined

With the interceptor designation in place, Kelsey and Saville were free to specify the aircraft they saw as vital to the future defence of the US. The X-608 proposal was for a twin engined interceptor, the X-609 for a single engined machine. Both aircraft were to use the Allison V-1710 engine, but fitted with the General Electric turbosupercharger to give them the high altitude performance both men foresaw as being vital.

Both aircraft were also to be fitted with tricycle undercarriages to simplify ground handling, another innovation not then widely in use. The performance requirements in both proposals were exacting for the day. The two aircraft had to be capable of a top speed of at least 360mph (579kph), with a preferred maximum speed of 400mph (644kph) if possible.

They also had to be able to operate at or above 20,000ft (6096m) and they had to be capable of reaching that altitude in under six minutes. As the mission stated in both proposals was the high altitude interception of hostile aircraft, cannon armament was also specified to destroy bombers. Lastly, both aircraft had to have sufficient endurance to be able to operate at full throttle for longer than an hour. These two proposals were without doubt the toughest issued by the USAAC to date, and would sorely stretch the capabilities of the aviation industry.

To put these requirements into perspective, the two fighters being evaluated for the USAAC in 1937, the Curtiss P-36 and Seversky P-35, had maximum speeds of 313mph (504kph) and 290mph (467kph) respectively. Whoever responded to the new proposals was going to have to make a quantum leap in design performance.

Three companies put forward designs for the two interceptors. The first, Vultee, forwarded a twin engined design to X-608, the XP-1015, but this was not considered sufficiently advanced and was dismissed. The single engined X-609 proposal was won by Bell with the unusual mid-engined P-39 Airacobra, which met all the requirements in its initial form and was designed around the massive 37mm Oldsmobile T9 cannon. It is worth noting that when Kelsey was posted to the UK, the P-39 development programme was handled by other members of his department. For a variety of reasons, mostly related to drag, the turbosupercharger was removed from the design, effectively destroying its usefulness as a high altitude fighter. Its performance was limited to below 12,000ft (3658m) but it did prove to be an excellent ground attack aircraft and was widely used by Allied air forces, especially that of Russia.

The Model 22 takes shape

The twin engined proposal, X-608, was answered by Lockheed with a radical approach, known internally as the Model 22. Hall Hibbard and Kelly Johnson considered a number of designs and layouts, Johnson beginning with a series of sketches of all the possible twin engined layouts in February 1937. These ideas were gradually whittled down to one which, even at this early stage, was recognisable as the aircraft we know today. This layout of twin booms and a central fuselage pod had a number of advantages regarding the requirements of X-608. It made the fitting of a tricycle undercarriage logical and enabled the turbosuperchargers to be mounted in the booms along with the main undercarriage legs and radiators. The mounting of all these components, along with the rudders and elevator directly behind the engines meant that the powerful slipstream produced by the propellers flowed over the control surfaces, improving their effectiveness.

This airflow also increased the efficiency of the radiators, oil coolers and turbosupercharger intakes. Lockheed was already gaining valuable experience with the operations and requirements of the General Electric B2 turbosuperchargers, as they were fitted to the XC-35, a one-off pressurised cabin version of the Model 10 Electra. This had been ordered by the USAAC in 1935 as a flying laboratory to prove the feasibility of pressurisation systems – research that was to be valuable to the later Boeing 307 Stratoliner and B-29 Superfortress designs, among others.

This aircraft also provided Lockheed with useful experience of high altitude operations, such as the effect of cold on hydraulic systems, which proved useful in developing the new fighter.

The Model 22 was innovative in terms of construction too, as it was the first US fighter to feature stainless steel in its structural members and flush riveted, butt-joined aluminium skin panels throughout the airframe to minimise drag. Johnson had learned a great deal about form, parasite and surface drag from the Model 12 and put this knowledge into practice in every detail of the fighter design.

Finally, the armament would be concentrated in the nose, increasing the effective range of the weapons as there would be no need to harmonise them at a pre-set range as was required with the wing mounted guns of other types. This arrangement also greatly simplified aiming for the same reason; the pilot only had to point the aircraft at the target and did not have to allow for the offset of his weapons. Lastly, the close grouping of the guns meant the firepower of the aircraft was devastating right out to their maximum range.

The initial armament was intended to be two .30in (7.62mm) Browning machine guns with 500 rounds per gun, two .50in (12.7mm) Browning M2 machine guns with 200 rounds per gun and a single .90in (23mm) Army Ordnance T1 cannon.

Kelsey was sufficiently impressed with the design to recommend that the USAAC purchase a single example for evaluation. Army Air Corps contract 9974 was awarded on June 23, 1937, for the construction of a prototype to be known as the XP-38, along with a full scale engineering mock up, a static structural load test airframe and wind tunnel models to provide aerodynamic data. The total value of the contract was $163,000, but Lockheed was to expend $761,000 in fulfilling it, such was its faith in the concept.

XP-38 Lockheed Model 022-64-01

After a year spent refining and defining the detail of the design, work commenced on the sole hand built XP-38, serial number 37-457, in July 1938. A small team of engineers put the aircraft together in a small workshop away from the main Burbank plant buildings, in order to maintain the secrecy requested by the USAAC.

During the construction, project engineer Jim Gerschler added an important refinement in ‘handing’ the engines, so they drove their propellers in opposite directions, rotating inwards at the top. This nullified the torque effects of the propellers, and gave the aircraft the excellent handling it was to become famous for. The XP-38 was powered by a pair of Allison V-1710s, a 1710-11 on the left and a -15 on the right, which drove their propellers via epicyclic gears. The only change needed to reverse the direction of a V-1710 engine was to change the firing order of the spark plugs.

The XP-38 had a number of unique features, the most obvious being the retractable entry door that controlled airflow over the oil coolers under each engine and the very small radiators with separate outlet flaps on the upper surfaces of the tail booms. The wing had a span of 52ft (15.85m) and an area of 327.5sq ft (30.4sq m). These dimensions remained constant throughout all models of the P-38, although the XP-38’s wing loading of 42.6lb per square foot (207.9 kg per sq m) naturally increased as the type grew progressively heavier throughout its development. Lastly, the modified armament of a single T1 23mm cannon and four M2 0.5in (12.7mm) machine guns was never installed.

Construction was completed by the end of the year, when delivery was made by road to March Field, California, at midnight on December 31, 1938, again to preserve secrecy. At the time, Lockheed did not have a fighter qualified test pilot, so Lieutenant Ben Kelsey, who of course was already project test pilot for the USAAC, undertook the testing of the prototype. A number of problems occurred during the ground trials, not least of which was with the brakes.

During one aborted take off trial, Kelsey actually bent one of the brake pedals trying to get the heavy fighter to stop. Fixes to this and other minor problems meant that it was not until January 27, 1939, that Kelsey took the XP-38 into the air for the first time. For 34 minutes, Kelsey fought the XP-38, as not long after take off a failure in the flap actuators left the fowler flaps fully extended. Despite the severe buffeting, he managed to land the aircraft back at March Field.

Further modifications to both the flap and brake systems followed, after which the remaining test flights were trouble free. The performance of the XP-38 exceeded the USAAC requirements comfortably. It was capable of 413mph (665kph) and had a service ceiling of 38,000ft (11,580m). Kelsey found the light XP-38 to be manoeuvrable and easy to fly, so he suggested to USAAC commander General Henry “Hap” Arnold that the prototype be used to set a new coast to coast speed record, since it was to be taken to Wright Field in Ohio for continued testing anyway.

This was approved, and if by the time Kelsey reached Wright Field all was well with the XP-38, he would be allowed to continue on to Mitchel Field on Long Island, New York. Flying via Amarillo, Texas, to Wright Field, Ohio, on February 11, Kelsey was given permission to continue, and made Mitchel Field in seven hours and two minutes flying time – a new record.

On arriving over Mitchel Field, he was cleared to land behind a trio of Consolidated PB-2A training aircraft. Making a long, flat circuit to take spacing on the trainers, Kelsey suddenly found both engines lost power, and he was forced to land on a golf course some 2000ft (610m) short of the runway. Clipping a tree on landing, the XP-38 was wrecked in the ensuing crash, but Kelsey escaped unharmed. A board of enquiry into the crash of the XP-38 determined that, due to local weather conditions, carburettor and/or fuel system icing was the likely cause of the power loss.

Such was the obvious potential of the new fighter that the loss of the prototype did not delay the programme excessively. In fact, Kelsey believed that by cutting the initial test programme short, the loss actually speeded up the development of the fighter. On April 27, 1939, the USAAC ordered 13 YP-38 service test prototypes at a cost of $134,284 each. These were the first batch of the only US fighter to remain in production throughout the Second World War.