Armour of the Air

First Concepts

From the first days of military aviation, and even before military aviation became a reality, people understood that the protection of combat aircraft against enemy fire would be both desirable and very difficult. Armour plate is heavy, and its presence ate deeply into the very limited useful load and fuel capacity of the first aircraft. Moreover, the additional weight could make the aircraft unsafe, and present greater danger to the pilot than enemy fire would. And if the aircraft became too sluggish, the probability of it being hit by enemy fire might even increase more than could be offset by the additional protection. Therefore, even when combat aircraft became bigger and more powerful, armour was usually designed only to protect the crew and the vital systems. The goal could not be to prevent damage to the aircraft; only to limit the effects, and give the crew a better chance of surviving and returning to base. 

More armour could be desirable, however, for close support aircraft. By definition, these operated in the close proximity of ground troops, and after their initial amazement and the occasional scare, soldiers began to use their weapons to fire at overflying aircraft. The effective range of an infantry rifle was sufficient to make this a serious threat to low-flying aircraft. Even the air above the own lines was far from safe: pilots soon discovered that ground troops — and ships — had a strong tendency to fire indiscriminately at any aircraft. Infantry caught in open terrain was highly vulnerable to air attack, but on the static frontline of the Western Front during WWI the soldiers were dug in and well prepared. For any attacker this meant that there was a very high probability of being hit, which justified a high level of protection.

Well before WWI broke out, the larger countries created air services, initially quite small but soon expanding rapidly, and these devoted much thought and experiment to the development of combat aircraft armed with machine guns, bombs, and even torpedoes. These concepts were largely untried and the available aircraft could carry only very light loads. Nevertheless, requirements for armoured aircraft preceded the war. The French general Bernard formulated one in 1913, and armoured aircraft of six different types, with 3 mm thick plates, were presented on the airfield of Villacoublay in June 1913. But reportedly four of the six were considered so overweight and even dangerous, that their pilots refused to fly them. Perhaps a typical effort was represented by an armoured version of the Nieuport IV monoplane, that had its front fuselage wrapped in a steel cylinder and the front of the rotary engine protected by armoured “persian blinds”. The result was an affront to aerodynamics and must have created serious cooling problems.

Armour plate was not only considered for ground attack aircraft. Another Nieuport effort of 1913, the so-called «Destroyer», was reportedly designed as an armoured anti-airship aircraft. For this role the two-seat aircraft featured not only armour plate (2.5 mm to 3 mm thick) but also a machine gun sliding on a rail around the rear cockpit, one of the first such installations. Despite a 160 hp engine, the ability of this aircraft to climb to the altitudes reached by airships seems extremely dubious, and when the type was used in combat in 1914, the armament was reduced to a carbine. Among the more useful armoured aircraft available in 1914 were three modified Voisin 1913 pusher biplanes, for short-range reconaissance.

One of the first moderately successful attempts to produce an armoured combat aircraft entered service in the summer of 1916. This was a variation on the Caudron G.4, a small biplane powered by two 80 hp Le Rhone engines, and therefore able to carry some additional weight. Nevertheless the performance of the G.4 IB suffered seriously, making it vulnerable to enemy aircraft. It was suggested that the aircraft was best operated only under heavy cloud cover. Because production was limited by a shortage of armour plate, the armoured aircraft were reserved for the best units, but this did not imply that they were grouped in specialised units. Instead the G.4 IBs were distributed among the standard G.4s. This type of aircraft soon became obsolete, whether armoured or not.

By 1917, ground attack missions had developed into a large component of the bloody offensives on the Western front, such as the battles of Cambrai and Passchendaele. The RAF commander in France, Sir John Salmond, attached high importance to such missions, confident that they would play a large role in stopping the expected German 1918 spring offensive. But any such achievements came at horrible cost; at Cambrai the average loss ratio for units engaged in strafing missions was 30 % per day of operations. In a series of post-war lectures, Wing Commander Slessor would use such statistics to defend the new RAF view, that aircraft should not be used as battlefield weapons at all — that strategic bombing was to be the RAF’s core mission.

The J Class

The German air forces first committed themselves to ground-attack missions on a large scale during the Battle of Verdun, in 1916, and late in that year they initiated the development of specialized ground attack aircraft. Some C-class general-purpose aircraft, mainly built by L.V.G. and D.F.W., were adapted by adding armour around crew positions and fuel tanks. From the autumn of 1917 onward the CL class of multi-role two-seat biplanes became important. The CL aircraft were designed to be lighter than the C-class aircraft, and better able to defend themselves against fighters; among their intended roles was also that of escort fighter. They were effective ground-attack aircraft, but not heavily armoured.

In contrast the J class were armoured, dedicated two-seat attack aircraft; the first ones were nevertheless derived from existing multi-role types. The A.E.G. J.I was essentially a C.IV with an armoured fuselage and a 200 hp Benz Bz IV engine instead of the 160 hp Mercedes D III. Two machine guns were fixed to fire downwards, and the observer had a single gun to defend the rear. Experimentally, some were flown as single-seaters with six fixed guns. The J.I was followed by the improved J.II. In total, A.E.G. delivered 609 of these attack aircraft. The Albatros J.I and J.II were similar in concept: An armoured fuselage mated to the wings of the Albatros C.XII. The engine was again the Benz Bz IV, a liquid-cooled powerplant. The Albatros J.II had additional protection for the engine, but few were completed.

Junkers J.I (Ray Wagner Image Collection, via WikiMedia)

In contrast, the Junkers J.I was an all-new design, and a quite revolutionary one at that. It was built of steel tube with a skinning of corrugated Duraluminium alloy instead of fabric. The biplane wings had inner but no outer struts; the engine was again a Bz IV. The armoured steel nose section was 5 mm thick and weighed 470 kg. The J.I was an extremely rugged aircraft and popular with its crews. The armament conformed to the standard of two Spandau guns fixed to fire forward and a flexible Parabellum. Despite the success of the aircraft, only 207 were completed. The innovative construction of the Junkers designs had great advantages, but at the time it made them difficult to produce in large numbers. The Idflieg tried to get around the problem by asking Junkers to associate with Fokker, and in October 1917 the Junkers-Fokker Werke A.G. was created, but it was a reluctant co-operation at best. 

The only engines of sufficient power available to the Germans were liquid-cooled, but this must have caused some major headaches for the designers, because the highly vulnerable cooling systems required extensive protection, while they still had to be exposed to an airflow. In 1918 a specification was issued for an armoured single-seat aircraft, intended to both fly ground attack missions and provide fighter protection to the heavy attack aircraft. Fokker’s chief designer, Reinhold Platz, devised an unique solution to the problem. The radiator of the 195 hp Bz IIIb engine was buried behind a large propeller spinner, that incorporated a 2.5 mm metal plate. Six curved fan blades where mounted behind the spinner, attached to the propeller; they drew in air from the sides and directed it into the radiator. The system was said to work well, but no orders for the Fokker V.38 parasol monoplane came forth. In comparison with the Sopwith Salamander, discussed below, the pilot would have enjoyed less protection, for the armoured box of the front fuselage was constructed out of plate only 2.5 mm thick. On the other hand the parasol configuration provided an excellent downward and forward view.

The winner in the Panzer-Einsitzer category was delivered by A.E.G. This firm initially flew a triplane in March 1918, but the Idflieg rejected this aircraft because of insufficient performance and poor handling characteristics. Work was already ongoing on a streamlined biplane, which featured wings with I-struts, a metal-skinned fuselage, and modest armour. Prototypes of the DJ.I were built with the 195 hp Benz Bz IIIb and the 240 hp Maybach Mb IVa. The armament consisted of the usual pair of LMG 08/15 Spandau guns. Flight testing had not yet been completed at the end of the war.

No specialized armoured aircraft were produced for the Austro-Hungarian air forces, and any experiments in that direction were frustrated by a shortage of suitable armour plate. But experiments were conducted with downward-firing gun batteries on the Brandenburg C.I. Two aircraft equipped with six Schwarloze guns in the rear cockpit, aimed 45 degrees down, were used in combat on the Italian front from August 1917 onwards. The initial success induced not only the Austrians, but also the Germans to investigate this form of armament more closely. However, after further frontline trials it was judged that the lack of a rear gunner made such aircraft too vulnerable. It was thought a better solution to carry a rear gunner, and provide him with an additional ventral gun that could be used in strafing attacks.

Trench Fighters

In 1918 Sopwith converted one example of the famous F.1 Camel fighter into the T.F.1 (for “Trench Fighter 1”) armoured ground attack aircraft. The armour protection was limited to a sheet of steel below the fuselage and the wing roots. Armament included two Lewis guns, angled 45 degrees down, between the landing gear legs, and one more on top of the upper wing. The problem of aiming the downward-firing guns was obvious, so one Camel was experimentally fitted with a periscopic sight. The downward-firing armament did not find favour with operational pilots, and the next development had conventional forward-firing armament. The Camel had unforgiving flight characteristics and seems a poor candidate for ground attack missions, but its high manoeuverability was considered an asset for operations close to the ground, and its air-cooled rotary engine resisted damage better than liquid-cooled engines. The T.F.1 was to remain an experiment. 

The Sopwith T.F.2 Salamander owed much to the Snipe fighter, but its entire forward fuselage was new. It was an armoured box with an 11 mm bottom plate and 6 mm side plates. The rear end had 10 mm and 6 mm plates spaced 95 mm apart, and an armoured headrest was fitted. The front plate was only 8 mm thick, because the mass of the engine was reckoned to provide additional protection for the pilot. The Salamander was an early example of the use of “structural armour”, that is, instead of being bolted on, the armour was designed as a part of the aircraft’s main structure, carrying loads as well as protecting the occupant. For decades such structures remained quite difficult to manufacture with sufficient precision. The implication for the Salamander was that the attachment points for the lower wing spars were often not correctly aligned, because the hardening process distorted the elements of the armoured box to which they were directly attached. The upper wing, which was carried on struts, could be correctly rigged.

Prototype of the Sopwith Salamander (WikiMedia)

An air-cooled engine was preferred to a liquid-cooled type. The 230 hp Bentley B.R.2 had reached the limit of what could be achieved with rotary engines, but the air-cooled engine was less vulnerable to damage, and its greater frontal area offered more protection to the pilot. The fuel system was redesigned to include an emergency hand-pump, installed on the inside of the armoured box. The designers also sought to make the rest of the structure more damage resistant. The tailplane had a metal instead of a wooden main spar, and later the bracing wires for the tailplane were replaced by steel tubes.

Initially two downward-firing Lewis guns (at an angle adjustable between 35 and 55 degrees) and a single gun firing forward were planned, but as mentioned above, the pilots preferred two forward-firing Vickers guns, and this armament was installed. On production aircraft the guns were staggered, so that ammunition boxes for 1000 rpg could be accomodated. Empty cases were dumped overboard. The aircraft could also carry four 20-lb bombs, but such bombs were a standard fitting for RAF fighters for a long period. The Salamander was not well received by pilots, indeed it was described by one officer as “clumsy” and “more lethal to its pilot than to the enemy.” Inevitably the T.F.2 was much heavier than a Snipe, with an empty weight of 835 kg against 592 kg, but it did not have more engine power. The rate of climb at 3050 m was only half that of a Snipe. The problems in rigging the lower wing must have contributed to the unsatisfactory handling characteristics. The Salamander picked up speed quickly in a dive, and landed at a fairly high speed for the period. Over 400 were built, most of them after WWI, but they were never assigned to an operational unit. 

The concept of a structural armour box was also employed in the Sopwith Buffalo two-seat fighter, a design that was never adopted for production. The addition of a second crew member required an extension of the armour, and the engine was still the B.R.2, too low-powered to make the aircraft really competitive. Although the aircraft was in fact very compact and only a little heavier than a Bristol F.2 Fighter, performance was mediocre. The pilot had a single synchronised Vickers and the observer a Lewis. The official mission was “contact patrol,” a low altitude reconnaissance over the battlefield. 

Also in 1918 the Royal Aircraft Factory produced the A.E.3 Ram. The designation stood for “Armoured Experimental”, but it was less a purpose-built design than an attempt to find a suitable role for an aircraft with an obsolete configuration. The A.E.3 was a derivative of the N.E.1 which itself was based on the F.E.9; they were the last in a line of pusher biplanes. The nacelle for the crew of two was designed as a steel armoured box, the front and floor having double walls. The two Lewis guns in front had a restricted field of fire, but could be aimed vertically downwards. A third, defensive, gun was installed on a tall mount that allowed the gunner to fire rearwards, over the upper wing and pusher propeller. The ammunition supply consisted of 32 pans of 97 rounds. As engine the 200 hp liquid-cooled Sunbeam Arab or the 230 hp Bentley B.R. air-cooled rotary were considered, prototypes with these engines being known as the Ram I and Ram II, respectively. Service trials in France failed to evoke any enthusiasm, and the design was abandoned.

The most outrageous venture among the armoured British aircraft was inspired by the Admiralty. The Wight-built AD Type 1 (AD for Air Department), also known as the Admiralty Type 1000, was designed to attack enemy ships with, depending on the version, a 12-pdr Davis recoilless gun, twenty-two 100-lb bombs, or a single 18-inch Whitehead torpedo. (Because the 12-pdr Davis was not yet available, consideration was given to installation the conventional 12-pdr Quick Firing Gun, also known as the Naval Landing Gun because it was originally carried by cruisers as a field gun for landing parties.) These large biplanes were powered by two tractor engines and one pusher, and protection against rifle-calibre fire was to be provided for the crew, the engines, the radiators, and the floats. For the radiators, which were on top of the engines in front of the two tail booms and at the end of the central nacelle, this meant that they were shielded by triangular metal plates to each side, contributing much to the curious look of the AD Type 1. For the crew the design provided a large, unaerodynamic, box-like “greenhouse” with numerous bullet-proof glass panels. The floats had steel planing bottoms. It was inherent in the design that the aircraft would barely be able to take-off with a full load, gradually climbing higher as fuel was burnt. During the few tests that were made, it inevitably proved seriously overweight, and probably it would have been unable to take off with a war load. The second prototype was completed without the armour plating for the radiators. The ponderous aircraft was developed from late 1914 to March 1917, when the Air Department came to long-overdue conclusion that it was useless.


In the final year of the war the French military defined several categories of armoured ground support aircraft. The most conventional of these was the Ab2 category, a two-seat tactical reconnaissance aircraft with light armour, designed to protect it from small calibre fire at distances over 300 m. Compared with the standard A2 two-seat observation aircraft, it was accepted that the useful load would be reduced from 450 kg to 350 kg, and operational ceiling would be reduced from 7000 m to 4500 m. 

The S2 was to be an armoured two-seat aircraft for low-altitude attack missions, able to carry 150 kg of bombs or a machine gun. The S3 requirement was the most unconventional, because it specified a three-seat aircraft able to carry a 75 mm cannon with 50 rounds. If possible, this was to be installed in a turret with a field of fire 60 degrees wide. And as if this was not yet hard enough to achieve, the S3 was also to carry fore and aft turrets each capable of mounting a 37 mm cannon or a pair of machine guns!

For the Ab2 specification, Salmson offered the Salmson 4, a fairly straightforward development of the successful Salmson 2 A2 observation aircraft. The additional weight of the armour was compensated by giving the Salmson 4 a three-bay biplane wing of larger span. The engine, however, remained the 260 hp Salmson 9Z, one of the rare kind of liquid-cooled radial engines. Predictably, maximum speed at sea level dropped from 188 to 168 km/h and climb time to 4000 m almost tripled, from 7 minutes 13 seconds to 20 minutes 14 seconds. A dozen were in service when the war ended; no more were built. A competing Ab2 design by Vendôme remained a prototype.

Competition for the S2 requirement was much wider. Lioré et Olivier offered the model 5, a biplane powered by two 175 hp Gnome-Rhône rotary engines. These engines, chosen because they were considered the least vulnerable to enemy fire, where given large spinners, perhaps as additional protection. The LeO 5 was a fairly well streamlined and neat design, with 300 kg of armour. The short nose did not extend much beyond the propellers, and contained a fixed machine gun; the rear gunner had a pair of guns. When flown in 1919, the LeO 5 was thought to fit better in the Ab2 category. It did not enter production, but was developed into the LeO 7, a multi-role aircraft which featured a longer nose for an additional gunner. Following the rejection of the LeO 5, Lioré et Olivier prepared another S2 design, but it was never built.

Canton offered an S2 biplane powered by two 230 hp Canton-Unné engines. The main armament was to be four Lewis machine guns, firing forwards and downwards; a fifth gun was provided to defend the rear. This aircraft was found totally unacceptable. The armour plate was relatively thin at 4 mm, but nevertheless the Canton aircraft struggled to get off the ground when fitted with it. According to test reports, it could not climb higher than about 30 m, probably barely out of ground effect. The Hochart S2 did not even get that far. This aircraft also featured four machine guns firing downwards and forward, plus two machine guns for the defense; the engines were to be two 200 hp Clerget 11E rotaries. The military lost interest before the prototype was completed and it was never flown. 

Schneider Henri-Paul (WikiMedia)

Considering the dismal results of the S2 competition, there was every reason to fear for the S3. The tentative supplier of this aircraft was the Schneider heavy industry group, a large concern founded by Eugène and Adolphe Schneider. It attempted to move into aircraft manufacture with the Schneider “Henri-Paul”, named after a pilot son of Eugène who had been killed in combat. The “Henri-Paul” was a biplane with a 30 m wing span, powered by four 370 hp Lorraine 12DA engines, mounted in tandem in nacelles between the wings. The slab-sided fuselage had rings for the gunners with their 37 mm cannon in the nose and aft of the wing, and the 75 mm cannon was intended to be placed in a balcony under the fuselage. This armament was never installed. Construction was slow, perhaps because of its innovative all-metal structure (much of it covered with fabric), and when the aircraft was finally completed in 1922, the notion of using it as an attack aircraft had long been abandoned; it was then only considered as a BN4 four-seat night bomber. But the huge aircraft was already obsolete. The only other attempt to meet the S3 specification was made by Voisin, who fitted a cannon to the Voisin 12 unarmed night bomber in an attempt to meet the S2 or S3 specificiations, but without success.

Lessons of the Great War

Thus at the end of the First World War, a number of armoured ground attack had been designed, flown, and in some cases used in combat. Only a few, such as the Junkers J.I, were entirely practical and effective. Some others designs were, with hindsight, obvious dead ends, but in 1918 this was not clear to those involved. The obvious flaw of many of the latter designs was that excessive armour and firepower had been piled on, resulting in aircraft that were either unairworthy or just too big and complex. Finding the right balance between protection, armament, and the other desirable qualities of an aircraft was not easy. 

A major stumbling block was the lack of rugged and sufficiently powerful engines. The Germans had to use liquid-cooled powerplants that were vulnerable and not quite powerful enough. The Allied resorted to using rotary engines, which were much less vulnerable, but had no development potential left to them. After the war, the air-cooled radial engine was perfected, and in many ways this was an ideal powerplant for an attack aircraft. 

However, after WWI most powers gave relatively little attention to armoured attack aircraft. Because of budgettary constraints, air forces spent their money on less specialized types of aircraft. Perhaps more decisive, the war had also seen the first strategic bombing offensives, even if they were primitive and had very limited results. Strategic bombardment promised to be a military revolution that could break the bloody stalemate of trench warfare; in comparison close air support appeared to hold only very limited promise.

Lumbering Monsters

In 1921 the US Army Engineering Divison demonstrated that the concept of an armoured combat aircraft could be extended way too far. Design work on the GAX, for Ground Attack Experimental, had been begun at the very end of the war. It corresponded to the highly theoretical concept of an heavily armoured, heavily armed “aerial battleship.” The result was far too big and too heavy for its engines. The GAX emerged as a twin-engined triplane with a weight of 3.4 ton, of which over a ton was armour plate. The aircraft was to be flown by a pilot and two gunners, and carried a 37 mm Baldwin cannon and eight .30 inch machine guns. The Baldwin cannon was an unimpressive weapon, and the performance of the grossly underpowered aircraft was unacceptable. Nevertheless Boeing completed 10 production aircraft, known as GA-1s. 

Boeing GA-1 (WikiMedia)

Boeing tried to improve on the concept with the GA-2, a considerably smaller biplane. The heavily armoured front fuselage culminated in a towering structure to protect a gunner, seated aft of the pilot. The armament again included the 37 mm Baldwin cannon and two .50 inch Browning guns, in a fairing between the landing gear legs, angled down. There were also two .30 guns. The GA-2 was not as underpowered as the GA-1, but nevertheless it was a failure, and only two were built. 

In 1922 another attack aircraft was flown, a single-suit “Pursuit and Ground Attack” type, the Aeromarine PG-1. The PG-1 too, was designed by the Engineering Division. It was a fairly compact sesquiplane, destined to be powered by a Wright K-2 engine that would have a 37-mm Baldwin cannon installed between the cylinder banks. The cockpit was protected with ¼ inch thick armour. However, the development of the K-2 engine was never completed, and the PG-1 was initially delivered with a mock-up engine installed. Flight tests were made with Wright H or Packard engines, but the PG-1 was a poor performer and suffered from excessive levels of vibration. To be fair, this was a very difficult period in American fighter development, because there was very little previous experience in this field in the USA, and such disappointments were the rule rather than the exception. Only two of the three PG-1s built were flown. 

The apparent result of these experiments was to discourage further development of armoured attack aircraft. Instead, the attack units were equipped with the Curtiss A-3, a straightforward development of the Curtiss O-1 observation biplane. Modifications were limited to installing an additional pair of .30 inch machine guns and bomb racks. In the 1930s the US Army produced a new specification for an attack aircraft, calling for a modern all-metal monoplane. General Aircraft offered its XA-7 and Curtiss the A-8, named Shrike by the company (but not by the Air Corps). These rather hideous aircraft, monoplanes with fixed landing gear, were modern designs for their time. The A-8 was declared the winner and Curtiss produced 46 of the improved A-12 version, but their service life was fairly short. These were light attack bombers, not armoured ground support aircraft. They were armed with four fixed .30 inch guns and a single flexible gun, and could carry a load of small bombs. The Northrop A-17 of 1935 represented the fast progress in aerodynamic design that characterized the period, but had a similar role and armament.

As WWII approached, the Air Corps abandoned the concept of the single-engined attack aircraft altogether. Instead it adopted the fast twin-engined light bomber, exemplified by the Douglas A-20. This was an excellent aircraft, but hardly suitable for close air support; it was more useful for anti-shipping operations and interdiction missions behind the enemy lines.

Soviet Armour

In the Soviet union the development of specialized attack aircraft was initiated in the late 1920s. From the start, it followed a number of different tracks, because the authorities defined no less than three different categories: LSh (“Legkyi Shturmovik”, light attack aircraft), TSh (“Tiazely Shturmovik”, heavy attack aircraft) and TShB (“Tiazely Shturmovik Bronirovanny”, heavy armoured attack aircraft). In theory the light attack aircraft were intended to harass enemy forces with surprise attacks, the heavy attack aircraft would cooperate with the own forces, and the armoured attack aircraft would be used to attack enemy forces that were prepared to defend themselves. It simplified things a bit when the LSh category was dropped in 1930. 

In 1929 the Polikarpov R-5, a single-engined two-seat biplane, entered service in the USSR. It belonged to the then very common category of multi-role two-seat types, and accordingly performed many duties, including those of light bomber and attack aircraft. For the latter role, Polikarpov developed a version, known as the R-5LSh, with additional armour, four additional 7.62 mm PV-1 guns fixed in packs on the upper wing, and bomb carriers for increased loads. This flew in 1931, after the air force abandoned the LSh concept; but nevertheless a limited number were built for the VVS, with four ShKAS machine guns on the lower wings. Replacing the PV-1 with the ShKAS of similar calibre more than doubled firepower, because of the 1800 rpm rate of fire of the newer weapon. 

Another improved version of the R-5 was known as the R-5SSS. Aerodynamically refined, this normally had two fixed and one flexible ShKAS, and could be fitted with four to eight additional guns in containers attached to the lower wing. Further development produced the R-Z. These aircraft served as light bombers and attack aircraft, but they were not armoured. 

Grigorovich had attempted to develop an aircraft in the LSh category employing as much elements from the ubiquitous R-5 as possible; when the demand for this fell away he converted his design to meet the TSh requirements. The three prototypes of the TSh-1 differed in the attachment of the armour plate around the engine, coolant radiator, and crew. Like others before him, he experienced great difficulties in accurately manufacturing a construction out of armour plate. Attempts to weld the plate, instead of rivet or bolt it into place, provoked cracks. The TSh-1 was also underpowered. Similar problems beset the slightly modified TSh-2, of which about ten were completed.

The TSh-3, designed by Kotcherigin, was a considerably more ambitious project. The configuration was that of a strut-braced low-wing monoplane with fixed landing gear, with eight fixed ShKAS machine guns installed in the quite thick wing, which also had internal bomb bays for six small anti-personnel bombs. (On some plans two more ShKAS are indicated in the landing gear leg covers.) The front fuselage was constructed as a steel tube structure to which the armour plates, 5 to 8 mm thick, were attached. Armour plating with a weight of 576 kg protected the entire front fuselage except the upper cowling of the liquid-cooled AM-34F engine. To reduce the vulnerability of the large coolant radiatior, located between the landing gear legs and surrounded by 6 mm plate, it was made retractable; a narrow channel under the engine provided some cooling during the attack phase. Flight tests began in 1934, but the TSh-3 was overweight and underpowered, with a disappointing performance, and was not put in production.


So far the experience was discouraging. But Soviet aircraft and aircrew participated in the Spanish Civil War, and in March 1937, during the Battle of Guadalajara, air attack played an important role in the defeat of the Italians. The success of this operation, flown mostly by the R-5, R-Z, I-15 and I-16, convinced Stalin of the need to develop a dedicated ground support aircraft. The VVS duly prepared new specifications, this time for an armoured ground attack aircraft, a Bronirovannyi Shturmovik

For Stalin this was a change of direction, for before the experience in Spain the Kremlin had seen more advantages in a light, fast multi-role aircraft, the so-called Ivanov programme. Some of the aircraft developed for this specification were now offered in ground attack versions with additional armour. Sukhoi’s successful BB-1 (also known as the Su-2) was the basis for the BB-2, or ShB (attack bomber). Similarly, Kotcherigin developed his R-9 into the LBSh (Legky Bronirovannyi Shturmovik or “lightly armoured attack aircraft”). The American Vultee V-11G light bomber, for which a license was acquired in 1936, was the basis for the BSh-1. Such aircraft had increased armour, armament optimised for ground attack missions, and more powerful engines to compensate for the increased weight. Kotcherigin also replaced the retractable landing gear of the R-9 by a supposedly more rugged fixed landing gear. 

Designs like this did not satisfy Sergei Ilyhusin. In 1938 he proposed a different design philosophy in a letter to Stalin. This included the use of extensive structural armour, in contrast to the prevalent practice of local installation of armour plate. All vital parts of the aircraft were to be located inside a armoured shell that also carried structural loads. He envisaged that the aircraft would operate below an altitude of 500 m, most of the time within range of small arms fire, and therefore needed to be armoured against hits from these weapons. Evidently, protecting the aircraft against direct hits from heavier anti-air weapons would deteriorate its flying qualities too much, so a balance needed to be found. (Later careful calculations were made, involving the likely angles of impact and relative velocities of projectile and aircraft, to determine what thickness of armour would be sufficient.) The Kremlin approved of Ilyushin’s ideas, and in May 1938 he was ordered to devote all his efforts to the development of his Shturmovik. A mock-up was ready in January 1939, and after its inspection and approval by the VVS the construction of two prototypes, designated TsKB-55 or BSh-2, began.

On first sight structural armour is a very efficient concept, but a serious problem with its application is that an aircraft’s needs for strength and protection do not always coincide. One location were they do is the engine, which needs to be well protected but also needs strong support structures; the most powerful aircraft engines of this period weighed between 600 and 800 kg. For lack of an alternative, Ilyushin had to use the Mikulin AM-35, originally developed for high altitude bombers, and liquid-cooled. The engine was installed in a bath of nickel-molybdenum steel, 4 to 6 mm thick, and covered on top with easily removable decking of 5 mm thick duraluminium. The protection of the vital cooling systems was a serious hurdle to overcome. At first the designers considered to make the oil and coolant radiators retractable, so that they could be pulled into the armoured fuselage for the duration of the attack. This would also reduce drag, but the engine could not be expected to function in this way for longer than six to eight minutes. It was decided to bury the radiators entirely whitin the fuselage instead, behind the engine, in a duct that had its inlet above the engine and an exhaust below the cockpit. A price was paid in cooling efficiency, but the new design was simpler and better protected.

In the TsKB-55 the pilot and rear gunner sat back to back in a cramped, but well protected cockpit. The armour thickness was 4 to 8 mm on the sides and 13 mm for the rear plate. The armoured windscreen was 65 mm thick. Against the opposition of Ilyushin, the VVS decided to eliminate the rear gunner, in the belief that the speed and armour of the Il-2, and a fighter escort, would be sufficient. The weight saving allowed the TsKB-57 prototype to carry more fuel and heavier armour for the pilot. To improve the pilot’s view, his seat was raised by 5 cm and the engine mounted 17.5 cm lower. Production saw further increases in protection, the cockpit windows becoming 55 mm armourglass in front and 49 mm to the sides, with a 65 mm window at the rear. Combat experience later lead to the installation of more steel armour plate instead of glass panels, reducing the field of view of the pilot.

The TsKB-57 was powered by the AM-38, a development of the AM-35 optimised for low altitudes, that was to power all production Il-2s. In the summer of 1941 shortages of this engine prompted the design of the Il-2-M-82, also (confusingly) known as the Il-4, powered by an air-cooled radial, the Shvetsov M-82. This was an excellent engine, but because of its larger diameter and a smaller weight it had to be installed ahead of the original engine position, and unprotected. The Il-2-M-82 had a redesigned, roomier two-seat cockpit, but nevertheless there was little room for the gunner to manipulate his UB machine gun, and the field of fire remained limited. The aircraft was usefully lighter than the Il-2 with AM-38 engine, but when the tests were finally completed enough AM-38 engines were available and converting factories to build the redesigned aircraft would have involved considerable loss in production.

The fixed armament of the prototypes was limited to a pair of ShKAS guns in each wing. The second TsKB-55 prototype was brought to TsKB-57 standard and equipped with the armament desired by the VVS, a Taubin PTB-23 cannon and a ShKAS in each wing. The Taubin cannon revealed itself unreliable and hampered by a very strong recoil blow, so it soon gave way to a 20 mm ShVAK with 210 rpg. This was accepted for production (with 250 rpg), but alongside a much more powerful weapon, the 23 mm VYa. The VYa fired a projectile that was more than twice the weight at a slightly higher muzzle velocity; armour penetration was much superior. A disadvantage was that the ammunition capacity for the VYa was limited to 150 rounds. The production rate of the VYa remained low until 1942, so the first batches of Il-2s had ShVAK cannon. 

The Il-2 could also carry small bombs in bomb bays in the wing roots, and rockets under the outer wing panels. The latter could be the 82 mm RS-82 or, from early 1942 onwards, the heavier RS-132. This rocket was powerful enough to defeat the armour of a medium tank, and later the Soviets produced improved versions of these rockets with shaped-charge warheads, the RBS-82 and RBS-132. But like the rockets used on the Western front, these were insufficiently accurate for use against point targets, although they could be fired at armour concentrations. More effective were small shaped-charge bombs, the PTAB series, which could penetrate armour up to 70 mm thick, and be dropped in large numbers. An Il-2M could carry 192 in its internal bombs bays, enough to saturate an area 15 m wide and 70 m long.

Ilyushin Il-2 (WikiMedia)

Flown in combat under conditions of enemy air superiority in 1941, the Il-2 suffered high losses. The installation of wooden guns, or later of real but fixed “scare” guns for rear defence, did not bring much improvement. In several regiments, a position for a rear gunner was improvised, wielding a single ShKAS on a mount taken from R-5 biplanes. A demonstration of a modified Il-2 to VVS commanding officers convinced them that the idea was sound, and a large-scale modification programme was ordered. In February 1942 a conference was held in the Kremlin to discuss the issue, and ordered the development of a two-seat Il-2, on the strict condition that production would not suffer. This prevented major modifications, and effectively excluded Ilyushin’s attempt to provide the rear gunner with a protection of 6 mm armour plate and an MV-3 turret with a 12.7 mm UB machine gun.

So inevitably, the rear gunner found himself outside the armoured shell, although protected by a 13 mm rear plate, and wielding an UBT on a quite simple mount. His ability to warn the pilot of the presence and actions of enemy fighters was as valuable as his return fire. The canopy over the rear position was often disposed of to improve the field of view and fire; later the introduction of the VUB-3 mount improved the field of fire. The gunner could not fire against enemy fighters attacking from below and behind; and such attacks often killed the rear gunners, whose life expectancy in combat was extremely short: A 7 to 1 loss rate for gunners vs. pilots has been reported. On the Il-2 Type 3, a DAG-10 grenade launcher was installed in the tail. This released small parachute-retarted bomblets to dissuade pursuing fighters. 

The two-seat Il-2M entered combat in October 1942. The Il-2 Type 3 (or Il-2M3) of early 1943 had a more powerful AM-38F engine and wings that were swept back slightly, moving the center of lift aft relative to the center of mass, to improve the stability of the aircraft. 

The February 1942 conference mentioned above also came to the conclusion that the ShVAK was too weak, and higher priority was given to equipping aircraft with the VYa, while awaiting the arrival of 37 mm cannon. At first the Shpitalnyi Sh-37 cannon was installed in a substantial fairing under the wings, with 40 rounds per gun. Only two ShKAS guns were retained in the wings. High recoil forces and unreliability made the Sh-37 a generally unsuccessful weapon, so it was supplanted by the Nudelman-Suranov NS-37 after brief service. The NS-37 was adapted to fire the same ammunition, but it was more reliable and more compact; 50 rounds were carried in the wing, while the cannon was a smaller fairing. Because of the weight of this installation, 237 kg per gun including ammunition, this so-called Il-2 Type 3M did not carry rockets and its bomb load was reduced. Production of this version remained limited. The cannon were effective against the top armour of even the heaviest tanks, but the high recoil forces and the deteriorated flying characteristics of this version made accurate aiming difficult.

Huge numbers of Il-2s were produced; in 1943 the Il-2 accounted for one third of the total aircraft production in the USSR. Its combat record reveals the dangers that await low-flying attack aircraft, even if they are well protected. The average loss rate during 1943 was one in 26 sorties, about half to enemy aircraft and half to fire from the ground. Weak points were the aft fuselage and the wooden outer wing panels; reinforcement of the aft fuselage and the standardization of all-metal wings helped to reduce losses. So did better tactics and training, and the wearing down of the Luftwaffe. In the second half of 1944 the loss rate dropped to 2.8%.


Ilyushin did not have a monopoly on the design of Shturmovik aircraft. In 1941 Sukhoi contributed the Su-6(A), a small armoured aircraft powered by the new M-71 radial. Structural armour was employed for the cockpit, but it was not as extensive as on the Il-2. Single-seat from the start, smaller and lighter, with less armament, this was more akin to an armoured fighter. The Su-6(A) carried only four ShKAS guns in the wings, the second prototype Su-6(SA) could carry two VYa cannon and four ShKAS guns. The type also had a small internal bomb bay for 120 kg. The Su-6(SA) found approval with the VVS; it outperformed the Il-2 by a considerable margin. But mass production of the latter was already underway, and late 1942 was obviously the wrong time to switch factories. Besides, the M-71 engine was never perfected. In 1942 the Su-6(SA) was modified to have a ShKAS and a 37 mm 11-P-37 cannon in each wing.

Sukhoi too was instructed to convert his design into a two-seater. As in the Il-2, the rear gunner wielded an UBT (with 196 rounds), but his armour protection was better. The gunner enjoyed bottom and side armour as well as a rear plate, while the gun mount took the form of a rotating, armoured disc. This version was known as the Su-6(S2A) and was considered superior to the Il-2 Type 3, but still could not displace it from the production lines. Struggling to convert praise into production, Sukhoi then studied the M-82 radial before deciding to adopt the liquid-cooled AM-42 engine, to create a competitor for the Il-10. This engine was protected by plate 2 to 4 mm thick and the radiator was in a bath directly below it. During testing the prototype reverted to VYa cannon in the wings, in addition to two ShKAS guns. It was all to no avail. 

Also in 1942 designer Dmitri Thomashevitsh, at the time semi-imprisoned in a factory in Siberia, was ordered to design a very light and cheap attack aircraft, which could replace the Po-2 on night harassment operations. The Pegas flew before the end of the year. It was a very angular monoplane constructed out of wood, powered by two small M-11F radials. To protect the pilot, armour plate was crudely bolted on the exterior. Four prototypes were built; the small and underpowered aircraft was intended to carry two 23 mm VYa or a single 37 mm cannon. But the war situation improved sufficiently to remove the demand for this aircraft.


In Germany a requirement for a specialized, armoured ground attack aircraft was issued in May 1937. Henschel designed the twin-engined Hs 129, powered by Argus As 410A engines. A neat, compact aircraft, the Hs 129 was characterized by an abbreviated nose and a small cockpit covered with armoured steel and glass panels. The armour plate was 6 to 12 mm thick, with a 75 mm armoured glass windscreen; the armour protection weighed over a ton. The As 410 was an inverted, air-cooled V-12; it was less vulnerable than liquid-cooled engines and not much in demand for other combat aircraft. It was at the time a modern engine, but also a relatively small one, with a swept volume of 12 liters and delivering 465 hp at take-off. It turned out to be far too weak for the Hs 129, and the small batch of Hs 129A-0 pre-production aircraft were useless. Pilots also complained about a cramped cockpit with a poor view.

Focke-Wulf competed with attack versions of the Fw 189, a tactical reconnaissance aircraft that was to replace Henschel’s Hs 126. The basic Fw 189 was a twin-boom design, powered by two As 410 engines, and with an extensively glazed central nacelle, offering good all-round view. The V-1 prototype was modified into an attack version with a heavily armoured and very small central nacelle, just large enough for the pilot and gunner, seated back to back. Again, complaints about the field of view from the armoured cockput were strong; the prototype was also heavy and handled sluggishly. Focke-Wulf completed the sixth prototype as an improved attack version, with a larger nacelle and a crew of three. Two MG-FF cannon and four MG 17 machine guns were installed in the wing roots. The Luftwaffe still preferred the smaller and cheaper Hs 129.

But the Hs 129 needed a more powerful engine to become a viable combat aircraft. The problem was solved after the fall of France, when substantial stocks of the Gnome-Rhône 14M radial became available. This 14-cylinder French radial had a swept volume of 18.9 liters and an output of 700 hp. Unfortunately, the 14M was not a very reliable engine, nor did it deliver enough power to allow the Hs 129 to fly well on one engine. French Ratier propellers were used on 14M4 and 14M5 engines rotating in opposite directions, to eliminate torque and slipstream effects. The underside of the engines was protected by some 5 mm plate. Though still slow, the Hs 129B was manoeverable and rugged. The cockpit was improved, but remained very cramped; the pilot’s seat was just aft of the thick windscreen, leaving no room inside the cockpit for the Revi C12 gunsight, so it was installed externally, on the nose. Some engine instruments were installed on the nacelles. Of course this reliance on French engines was a stop-gap measure, indicative of the rather low priority given to the type; no new engines were likely to be forthcoming. (In July 1941 an agreement was signed with the Vichy government, which allowed the French aviation industry to resume production in return for very large deliveries of equipment, but such aircraft and engines as were delivered, after much delay, were of highly suspect quality.) Only 867 of the Hs 129B aircraft were delivered before the production was terminated in September 1944. A plan to install Italian Isotta-Fraschini Delta IV engines had to be abandoned. 

Captured Hs 129B being evaluated in the USA. (USAAF via WikiMedia)

The Hs 129B-1 baseline model was armed with two MG 151/20 cannon and two MG 17 machine guns, installed in the fuselage sides. Four more machine guns could be carried in a belly gun pack. The capability of the type was transformed when the high-velocity MK 101 30-mm cannon (a weapon that had been languishing unused in the Luftwaffe’s inventory) was installed in a ventral pack on the B-1/R2 model. Although limited to only 30 rounds, this turned out to be a very effective anti-tank weapon, especially when firing Hartkernmunition with a tungsten penetrating core. This projectile penetrated up to 42–52 mm of armour at a distance of 300 m and a striking angle of 60 degrees. In tests, the combination of the Hs 129B and MK 101 was very accurate, some of this being attributed to additional side area of the ventral fairing acting as a keel.

The MK 101 was no longer in production. But a belt-fed cannon had been ordered to replace it, and the B-2/R2 model carried this MK 103 instead, with a belt of 100 rounds. The 30 mm projectile was effective only against the rear of the Soviet tanks, so other models of the Hs 129B were equipped with the 37 mm BK 3,7. About 25 were delivered of the B-3 model, armed with the 75 mm BK 7,5, a weapon that weighed about 1000 kg, not including ammunition and installation! Although the MG 151/20 cannon were deleted to save weight, the effect on handling qualities could not fail to be severe. In an emergency the heavy gun pod could be dropped.

The Hs 129 is often mentioned as the German equivalent of the Il-2, but this is only valid up to a point. The Soviet production of the Il-2 was about 40 times larger than the German production of the Hs 129. To produce enough engines for the Il-2s the USSR sacrificed the production of the MiG-3, one of its best fighters; the Germans powered their attack aircraft with a captured French engine that had a rather poor reputation. And this despite an urgent need for an aircraft to destroy some of the very large number of tanks in the Red Army. The Luftwaffe relied more on the old Ju 87 dive-bomber, which found a new career as low-level attack aircraft, and the new Fw 190 fighter-bomber. The Fw 190 was of course unable to carry the large cannon of a dedicated tank-buster, or the armour of a dedicated ground support aircraft; but its high performance made it much less vulnerable to attacks by enemy aircraft than the Ju 87.

The fighter-bomber models of the Fw 190 were fairly highly specialized. The first attack missions were flown by a small number of Fw 190A-4/U3 fighter-bombers in September 1942, and the equipment standard for a dedicated fighter-bomber series was agreed in November. The decision was taken to develop the Fw 190F and G series as dedicated fighter-bombers, and the Fw 190A-4/U3 was subsequently renamed the Fw 190F-1, while the A-4/U8 became the G-1. The F was primarily a Schlachtflugzeug for close-air support, while the G was a fighter-bomber with a longer range, a so-called Jabo-Rei, with less armour and more fuel. Significant numbers of these aircraft became available in the summer of 1943. To save weight, the F models did not carry guns in the outer wing panels; the G models dispensed with the fuselage machine guns as well, and only the MG 151/20 cannon in the wing roots remained.

Despite these weight-saving measures, the attack versions had inferior flying characteristics compared to the much lighter fighter models. In November 1942 Galland had called for a reduction in the armour to less than the 480 kg then installed, and at the time it was considered to remove the side armour and retain only the protection of the engine and the bottom of the fuselage. Standard protection of the F series included 5 mm plate to the sides of the cockpit, a similar plate on the bottom of the fuselage, and additional 6 mm plate to protect the engine. Against small arms fire this was effective, but the aircraft were still vulnerable to anti-aircraft cannon.

When enemy armour formations were found, the attack was concentrated on their soft-skinned supply vehicles, in the knowledge that the tanks would not get far without fuel. Against those small bombs and fire from 13 mm and 20 mm weapons were very effective. But if the enemy tanks were engaged in combat with German ground forces, their destruction did become an immediate goal. Without a suitable cannon, the fighter bombers resorted to “skip-bombing”, dropping their 250 kg bombs at high speed and an altitude of only 10 m or less, so that they skipped off the ground into their targets. Against tanks in open country, this was very effective. The MK 103 cannon was too heavy and had a too high recoil to be carried on the wings of the radial-engined Fw 190 fighter-bombers. Although a handful of aircraft were so equipped, they were not successful. At the end of the war the efforts were concentrated on launchers for anti-tank rockets.

Before the war, the Fw 189 had a competitor for the role of tactical reconnaissance aircraft: The Blohm und Voss Bv 141. The characteristically unusual design of Dr. Vogt had a slender fuselage on the left containing the engine and carrying the tail surfaces, while the crew was accommodated in abbreviated nacelle on the right. A lot of development work separated the Ha 141V1 prototype from a small number of Bv 141B pre-series aircraft, but Luftwaffe saw no advantage over the Fw 189, and declined the type. In 1943, however, Dr. Vogt was permitted to develop a dive-bomber and attack aircraft around the same concept. The Bv 237 design featured an armoured fuselage nacelle, and would have been flown by the pilot alone in the dive-bomber configuration and by a pilot and a tail gunner in the attack configuration. In the latter version the armament would have included three MK 103 cannon, besides two fixed MG 131 machine guns and a pair of similar weapons for the tail gunner. The B-1 series was projected with an auxiliary turbojet engine under the fuselage. Although design work reached an advanced stage, the RLM decided against construction of an Bv 237 prototype; instead production efforts were concentrated on a smaller number of less specialized aircraft. 

The final German effort towards a ground-support aircraft was the little Henschel Hs 132, a jet aircraft of configuration similar to the He 162, with the engine carried on top of a small and streamlined fuselage. Designed to operate as dive bomber or low-level attack aircraft, the small Hs 132 was intended to be flown from a prone position, to reduce drag and to help the pilot to resist the expected accelerations, up to 12 g in the pull-out from the dive. The cockpit was to be protected by steel plate and a windscreen 75 mm thick under a transparant streamlining cap. The Hs 132A dive bomber would carry no guns, but Hs 132B attack version would be armed with four MG 151/20 cannon, and the Hs 132C was even projected to carry two MG 151/20 and two MK 103 cannon, a heavy load for such a small aircraft. Three prototypes were under construction when Soviet forces captured the factory.

Desultory Attempts

During most of the war, the USAAF and RAF relied on fighter-bombers for close air support. But in November 1942 the USAAF ordered a specialized dive bomber, the Vultee XA-41. This was a single-seat aircraft powered by the new four-row R-4360 radial. At the time of the mockup inspection in the spring of 1943, the air force had decided against dive bombers, and ordered the type to be reworked as an attack aircraft. The XA-41 had aerodynamic cleanliness as well as impressive bulk, but when it made its first flight on 11 February 1944, it was already too late to have any hope of it reaching the front before the end of the war. Its wings had bays for four .50 inch machine guns and two 37 mm cannon. It is reported that it was planned to install the M9 cannon, a much more powerful, but also much heavier, cannon than the M4 and M10 carried by the P-39 and P-63. Pictures indicate that fairings for four cannon barrels were present at one time, but four M9 cannon seems an impossible load for this aircraft. 

Vultee XA-41 prototype

Another attack aircraft built for the air force was the Brewster XA-32, also a single-seat type, with an internal bomb bay and four 20-mm cannon in the outer wing panels. The XA-32 had compact, if rather brutish looks, and was hopelessly underpowered by its R-2800 engine; it would have needed the R-4360. Everybody seemed to prefer to forget about it, and Brewster was soon out of business.


In 1942 Ilyushin’s team began to study advanced Shturmovik designs, to be put in production when the war situation allowed it. The Il-8 was a heavy armoured attack aircraft, with an AM-42 engine, an increased bomb load, and a redesigned cockpit that included both pilot and gunner within the armoured nacelle. The engine installation and the armament remained similar to that of the Il-2. The second prototype was aerodynamically cleaned up and of all-metal construction. The wing guns were replaced by 23 mm NS-23 cannon and the rear gun by a 20 mm UB-20. 

The Il-8 did not enter production; the lighter, faster Il-10 was preferred. This design returned to the original Il-2 concept by seating the pilot and gunner back to back in an armoured shell. Again, the need to keep this small resulted in a very unconfortable cockpit. Armour protection for the crew was improved by using spaced sheets, each 8 mm thick. On the other hand armour was removed or reduced where operational experience indicated that it wasn’t necessary. A more efficient engine installation was designed for the AM-42: The oil and coolant radiators were installed inside the lower fuselage, and the coolant air inlets were in the wing roots. Fixed armament remained at two VYa cannon and two ShKAS machine guns, the rear gun was a 20 mm BT-20, and the DAG-10 grenade launcher was still present. The internal bomb capacity was enlarged. The Il-10 was considerably faster and handled much better than the Il-2. It was put in production in 1945.

After the end of the war, most Il-10s were re-armed with four 23 mm NS-23 cannon with 150 rpg. The NS-23 lacked the hitting power of the high-velocity VYa, but the higher rate of fire of the new installation was more effective against soft-skinned targets. Its reduced recoil affected the stability of the aircraft less.

The type was further improved after the end of the war. The Il-10M had a completely new wing, an updated structure, and many detail improvements. The armament was standardised on four Nudelman-Rikhter NR-23 cannon, with 150 rounds each. The NR-23 fired the same ammunition as the NS-23, but rate of fire was raised from 550 to 850 rpm. The rear gun was installed in an electrically powered turret. Production ran until 1954. The ASCC gave it the codename “Beast.” Combat experience in Korea was not very favourable, because the Il-10M was vulnerable to modern fighters. By 1956 they had been retired from the VVS. 

Ilyushin Il-10 at the Monino air museum. (WikiMedia)

An radically different configuration had been tested as well. During WWII the Soviet Union received a large number of Bell P-39 Airacobras and P-63 Kingcobras, fighter aircraft that had the engine amidships, behind the pilot, driving the propeller with a long extension shaft that ran under the pilot’s seat. A 37-mm M4 or M10 cannon was installed in the nose, firing through the propeller hub. Contrary to what has often been reported in the Western media, the P-39 and P-63 were employed as fighter aircraft, not in the attack role; the M4 and M10 were low-velocity weapons and not very suitable for anti-tank missions. But in 1942 Tupolev made drawings for an attack aircraft with a similar layout, the MSh. Never built, this design emphasized the view from the cockpit; the cockpit was therefore above the gun bay in the nose.

The Il-20 prototype of 1948 again repositioned the pilot as far forward as possible, but this time the engine was left in the conventional position, so that the pilot sat directly above the AM-42 engine. The result was of course a very blunt, high nose. A second crew member, separated from the pilot by a large fuel tank, controlled a dorsal gun turret with two NS-23 cannon. Four more of these weapons were in the wings, and four more in the fuselage, aimed 30 degrees down. The wing guns could be depressed 22 degrees on their mounts, allowing the Il-20 to strafe targets in level flight instead of shallow dives. The VVS chose the more conventional Il-10, which had better performance overall.

Heavy Twins

Most attack aircraft of the Second World War were relatively small and single-engined; the Hs 129 is only an apparent exception, being a small aircraft with two relatively light engines. Many were tempted by the concept of a larger aircraft, with two powerful engines, which would be able to carry more weapons, more armour and more fuel. An especially attractive feature of the large attack aircraft was its potential to carry high-velocity anti-tank cannon of large calibres, 50 mm or even 75 mm.

Already in the early 1930s Tupolev prepared the ANT-17 or TCh-B design in response to a VVS requirement. This aircraft was never built and its design features remain mysterious, but it was intended to be armed with a 75 mm DRP recoilless cannon, or alternatively with a pair of conventional 37 or 45 mm cannon. 

The VIT series developed by Polikarpov attracted much more interest. Polikarpov had created the basic design as a twin-engined multi-role fighter, a concept that enjoyed great popularity during the late 1930s, but the VIT-1 emerged in 1937 as an anti-tank aircraft. Built mostly of wood and highly streamlined, the aircraft seems to have incorporated little armour for the crew or the M-103 engines, but was armed with long-barreled ShK-37 cannon in the wing roots and a flexible 20 mm ShVAK in the nose. The VIT-2 featured increased armament: Two additional ShVAK would be fixed in the wing, and upper and lower defensive stations were introduced, firing towards the rear with a ShVAK and a pair of ShKAS, respectively. After long trials, the type was not adopted.

The first experiences in combat with the Il-2 revealed that the short range of the type prevented it from operating behind the German lines. And in 1944-45, when the Germans were on the retreat, the speed of the Soviet advance moved them beyond the range of Il-2s operating from bases in the rear. A requirement for a Shturmovik with a longer range was passed to Sukhoi in 1941. The result of his first study was the ODBSh, a twin-engined aircraft offered in single-seat and two-seat versions. Two 37 mm and two 12.7 mm cannon would be installed in the lower fuselage. It was not built, but in 1942 he got a second opportunity with the DDBSh, also known as the Su-8. Like the Il-2, this design had a forward fuselage constructed in structural armour; the rear gunner, who operated a turret with a 12.7 mm UBT, and a ventral ShKAS, was protected by armour plate. The aircraft was compact and neat, powered by M-71F engines rated at 2200 hp at take-off. The lower fuselage housed four 45 mm or 37 mm cannon; eight ShKAS guns were installed in the wings. The Su-8 was successfully tested in 1944, but as the end of the war was obviously near, there was no longer an operational need for the type.

For the most part, twin-engined attack aircraft tended to be modified bombers or fighters, with often very powerful armament but less armour protection than dedicated attack types. None of them was much of a success. The Germans developed the Ju 88P series, armed with, dependent on the model, one 75 mm PaK 40, one 50 mm KwK 39, or two 37 mm BK 3,7 cannon; but the type was too unwieldy and its armour protection, although better than that of the bomber, was insufficient to survive combat at low altitude. The British developed the Mosquito Mk.XVIII, nicknamed “Tse-Tse”, armed with the powerful 57 mm Molins 6pdr, but decided not to employ it in the anti-tank role; the handful of aircraft built were given to Coastal Command. In 1946 Soviets installed a 57 mm RShR cannon in the lower fuselage of a Tu-2 bomber, but despite successful tests it was not adopted for service.

The Japanese did not develop a dedicated single-engined ground support aircraft; the Japanese army relied on light bombers, such as the Ki-30 (‘Ann’), Ki-32 (‘Mary’), Ki-36 (‘Ida’) and Ki-51 (‘Sonia’). These were all obsolescent. However, the Kawasaki Ki-45-KAI Toryu (‘Nick’), although primarily designed as a twin-engined long-range fighter, turned out to be a quite useful attack aircraft. The Ki-45-KAIb version was armed with a 37 mm Type 98 tank gun, which fired the same ammunition as the Type 94 anti-tank gun (not to be confused with the less powerful Type 94 tank gun). The Type 98 was manually loaded. The Ki-45-KAIc instead carried a 37 mm Ho-203, less powerful than the Type 98 but equipped with a 15-round belt feed. The Ho-203 was later scaled up to the Ho-401 57 mm cannon, and this weapon (with 17 rounds) was installed in the attack version of the Ki-102 (‘Randy’) fighter, the successor of the Ki-45. Of this Ki-102b (also known as the Army Type 4 Assault Aircraft) about 200 seem to have been completed. The Ho-401 with its 520 m/s muzzle velocity was a suitable weapon for use against soft targets, but not much use against armour. Rikugun, the army aeronautical research institute, designed the Ki-93 with the Ho-402 in a belly fairing; this was also a 57 mm weapon but much larger and more powerful, firing its projectiles at 700 m/s. However, only one Ki-93 was ever flown. These Japanese aircraft were no longer as unprotected as most Japanese combat aircrafthad been at the start of the conflict, but they were not heavily armoured either, the designers’ priorities being performance and handling. 

The American industry had a little more success with dedicated heavy attack aircraft. Some experience had been gained with the installation of 75 mm cannon in B-25 bombers, although this was a relatively low-powered, manually-loaded gun, and the aircraft clearly a medium bomber rather than a dedicated attack aircraft. The Douglas A-26 Invader, smaller and faster, was tested with various cannon but entered service only with 12.7 mm machine guns. But at the end of 1942 the air force ordered two prototypes of the Beech A-38 Grizzly. The aircraft carried a 75 mm M10 (experimental designation T15E1) cannon in the nose, fully automatic and provided with 20 rounds. The pilot also controlled two fixed 12.7 mm Browning machine guns, and remote-controlled ventral and dorsal turrets contained two each of these weapons. The structure was conventional, with internal armour plates to protect pilot and gunner. With two R-3350 radials, the XA-38 prototypes demonstrated excellent performance; but the fact that these engines were urgently needed for the B-29 programme caused the cancellation of the A-38. Anyway, in service these early R-3350s caused a lot of trouble.

Beechcraft XA-38 Grizzly (WikiMedia)

Lessons of the Second World War

When the war ended, the importance of tactical air support for army operation was no longer in doubt. However, doctrinal differences remained, and have persisted to this day, as to its best use. Broadly speaking, armies tend to stress close air support, attacking the enemy while he is also engaged by ground troops. Air forces instead prefer interdiction, attacking the enemy and his supplies while they are moving to the front line. The former is a more direct way of giving support to the soldiers on the ground, and presumably better for their morale; of course it is also much more dangerous. The latter approach hopes to catch the enemy force at its most vulnerable and before it can do damage. 

For interdiction fighter-bombers, light bombers and medium bombers were likely to be most suitable. For close air support, the armoured attack aircraft was in competition with the fighter-bomber. What an aircraft carried in armour, it could not carry in bombs or fuel; the dedicated attack aircraft also needed a measure of air superiority to protect it against enemy fighters. The main choices were to use fairly standard fighter-bombers, such as the British and American air forces did; to use fighter-bombers adapted for their task with extra armour, which became the German preference; or to develop a dedicated armoured attack design, which was the Soviet choice. 

So what approach delivered the best results? Even with hindsight, it is very hard to say. After Normandy, the most effective air support appears to have been delivered on the Western front. However, this may very well have been a consequence of number of aircraft relative to front length. The quality and speed of communications between aircraft and ground units also made a huge difference. British and American aircraft had much better radios than contemporary Soviet designs.

There is, however, one role in which a specialized attack aircraft certainly performed best, and that was the destruction of tanks. As a rule, fighter-bombers did not carry the powerful cannon that were the only weapons that were both accurate and powerful enough to destroy tanks. The only fighter-bombers dedicated to a tank-hunting rule with cannon were the versions of the Hurricane armed with the 40 mm Vickers S. These were very successful, until the armour of German tanks became too thick for these weapons, but even in that case a dedicated attack design would probably have been superior.


Piston-engined attack aircraft such as the Il-10M flew missions in the Korean War, but in absence of air superiority they were too vulnerable. The Ilyushin design bureau made an attempt to reconcile the traditional features of the Shturmovik with jet power and high speed-aerodynamics. The result was the Il-40, codename ‘Brawny’ for NATO, flown in 1953.

The Il-40 posessed an armoured cockpit for a crew of two, four forward-firing NR-23 cannon on a mount that could be depressed for strafing runs, and two remote-controlled tail cannon of the same type. It also had swept wings and two AM-5F jet engines, installed in the wing roots. After the first trails, the inlets of these engines were extended all the way up to the nose of the aircraft, to avoid gun gas ingestion problems. Up to 1500 kg of bombs could be carried in internal bays and under the wings. Although construction of a pre-series batch was started, only the single prototype was flown; the VVS cancelled the Il-40 in 1956 because it was considered obsolete. 

A decade later, interest in a dedicated attack aircraft was renewed, and Ilyushin used the same general configuration for the Il-102. This time the aircraft had a GSh-23 tail cannon (600 rounds) and a GSh-30/II cannon (500 rounds) under the fuselage centreline; there was no need to extend the jet inlets forward. Although it was not larger, the aircraft was considerably heavier than the Il-40, and the weapons load rose to 7250 kg. The Il-102 made its first flight in 1978 and was praised in test reports, despite its rather unfashionable configuration and ugly looks. But the VVS chose the competing Su-25.

Ilyushin Il-102 (Piotr Butowski via WikiMedia)

In 1968 Sukhoi had started work on studies of a Shturmovik for the modern age. This was somewhat ahead of the LSSh requirement of the armed forces, issued in March 1969, and a formal order did not arrive until 1973. The T-8 prototypes flew in 1975.

From the start, the programme aimed at a lighter, smaller and less specialised aircraft than the American AX concept, which would result in the A-10. The warload was smaller, the range requirements resulted a much smaller internal fuel capacity, and the gun was the twin-barrel GSh-30. Though an effective weapon, this does not have the anti-armour performance of the American GAU-8/A, and only 250 rounds are carried. (It has been reported that at one time the intent existed to arm the Su-25 with the TKB-700, a 45 mm smoothbore gun, probably intended to fire anti-tank ammunition.)

The general layout of the Su-25 is conventional, with a straight wing, two R-95Sh engines under the wing roots, and a single tail. The R-95Sh is a non-afterburning turbojet, and on first sight a less than ideal choice for an attack aircraft, because turbojet engines perform best at high altitude. On the other hand turbojet engines are simpler than turbofans, and the development of the R-95Sh from the R-13-300 turbojet involved making the engine resistant to small arms fire. The R-95Sh is also able to run on a variety of low-grade fuels. Later developments had R-195Sh engines.

The pilot of the Su-25 sits in the titanium ‘bath’ armour typical of modern attack aircraft, in this case 24 mm thick. Combat experience in Afghanistan showed that the aircraft was quite rugged, but an engine fire could spread from one engine to the other, as they were located fairly closely together in the fuselage. An additional internal shield, 5 mm thick, alleviated the problem. In general the Su-25 was highly valued, its performance and weapons load were well matched to the needs of irregular warfare in mountainenous terrain. 

Combat experience also underlined the need for a better avionics fit. The concept of a CAS aircraft as it was understood in the 1960 and 1970s involved making it as simple as possible, but this often restricted operations: Enemy forces often respond to the danger of air attack by moving around at night or in bad weather. The original Su-25 did not even have a head-up display (HUD), only a gyroscopic gunsight was provided. During the 1980s Sukhoi worked on a major upgrade of the Su-25, resulting in the Su-25T. This uses the airframe of the Su-25UT trainer, but with the space of the rear cockpit allocated to fuel and avionics. This includes an inertial navigation system, an electro-optical system with a laser target designator, a low-light level camera (LLTV), a forward-looking infra-red (FLIR) scanner, and a HUD. The gun was moved from the lower nose to a position externally on the fuselage, with 200 rounds. Emphasis in the Su-25T was on anti-tank missions, using missiles such as the AT-12 Vikhr. An upgraded version, the Su-25TM, carried a radar set in a pod under the wing. This entered small-scale production, and it and its derivatives are also referred to as the Su-39.

With uncertain funding for the Russian armed forces, development plans for the Su-25 were ambitious, but actual progress extremely slow. As an alternative to new and (too) expensive Su-25TM, it was proposed to upgrade existing Su-25s to Su-25SM configuration. The upgrade gives the aircraft some of the capability of the Su-25TM and allows it to carry and deliver modern precision weapons. Plans to install a multimode radar in the nose are apparently not being proceeded with. This would have given the modified aircraft the ability to carry a much wider range of weapons, including medium-range AAMs, and fly a wider range of missions. 

The future of the Su-25 remains uncertain. The developed versions are capable, but highly specialized combat aircraft, while air forces increasingly show an interest in multi-role types. Like other Soviet designs of the 1960s and 1970s, the Su-25 needs much maintenance work and replacement parts to remain operational, and in the post-Soviet era is an expensive aircraft to operate, at least in terms of man-hours.


In the late 1960s, the USAF mulled over its experience in Vietnam. An evaluation at Eglin AFB in 1962 had determined that the best close air support then available was the Douglas A-1 Skyraider, and subsequently the aircraft had proven its qualities in combat. But the A-1 was a World War II design that had made its first flight in 1946; every Skyraider delivered to the Air Force was a refurbished surplus aircraft from the Navy! Production of the Skyraider had ended in 1957. The need for a new aircraft was obvious, and officers formulated the requirement AX for a close air support aircraft. A first RFP was released in March 1967.

The AX was a to be a rugged, simple, cheap aircraft, able to take off from a short airfield, fly to a combat zone up to 250 miles (400 km) away, and loiter there up to two hours. To achieve survivability and carry a combat load of up to 16,000 lb (7,250 kg), two engines were specified. Because the AX would operate at low altitude, turboprop engines were initially considered the best choice, but the development of efficient high-bypass turbofan engines resulted in a preference for these, in the form of the General Electric TF34 or Lycoming F102. Such engines move a relatively high fraction of ‘cold’ air through the front fan blades, in comparison to the ‘hot’ flow through the engine core and combustion chamber, making them efficient at low altitude. It also reduces the temperature of the exhaust, making the aircraft less visible to an IR missile seeker. However, later the slow throttle response of the TF34 proved a handicap in combat, and although it is bigger and heavier than the Su-25, the A-10A does not have more power. Other engines have been proposed but the USAF is unwilling to invest much in re-engining the A-10, although the possibility is kept open.

Armour protection and redundancy of the vital systems of the aircraft were required to make it survivable. Of course Vietnam had demonstrated that attack helicopters could give very efficient air support, but they were relatively vulnerable and limited in the amount of weaponry and fuel they could carry. The emphasis on survivability, payload and loiter time over the target put the AX into a different niche than helicopters or fast jets.

But the specification soon shifted away from close air support in jungle warfare. The war planners in the Pentagon were looking for a weapon to combat the numerical superiority in tanks that the Soviet Union would have in a conventional war in Europe. To meet this need the AX was optimized as an anti-tank aircraft, designed around a powerful 30 mm cannon. The final Request For Proposals was issued in May 1970. A RFP for the development of the cannon followed in November of that year. In June 1971 contracts were to General Electric and Philco Ford to develop the rotary GAU-8, while Hughes was given a backup contract to study the GAU-9, an adaptation of the Oerlikon 304RF revolver cannon. The winner was the seven-barrel General Electric GAU-8; as finally installed in the A-10, this weighs 1828 kg with a full ammunition drum.

Northrop designed its A-9 with much attention to cost reduction, as the Secretary of the Air Force had announced that he would not buy the aircraft if it cost more than US$ 1.4 million apiece. The A-9 had a straight wing, set high on the fuselage. The F102-LD-100 engines were tucked below the wing roots, where they were easily accessible to maintance crews, if perhaps in some risk of foreign object ingestion. The cockpit was to be seated into a titanium armoured “bathtub” (on the prototypes this was made from aluminium, so save money). The structure was designed to have enough redundancy in it, so that even the loss of a major structural member would be survivable. The control surfaces were moved by two independent hydraulic circuits, with a manual backup system. The fuel tanks were divided into three tanks in each wing (five in the prototypes) and filled with foam to prevent fire, although not all were self-sealing. (Fuel loss was accepted, provided that there remained enough to return to base.) The design was much influenced by the need to install the big GAU-8/A cannon on the centreline. (The gun was not tested in the air until February 1974, and on the prototypes the M61 was installed for firing trials.) To make aiming corrections easier, Northrop developed the Side Force Control system, which used the rudder and split-aileron wingtip speed brakes to change the direction of flight without banking the aircraft, which would upset the aim. To save cost, many parts were “off the shelf” items. The first YA-9 prototype flew in March 1972.

Northrop YA-9 (USAF)

Fairchild appears to have been less economic in its design of the A-10, and its development contract was substantially more expensive. The fuselage is similar in shape, with the cannon, or more precisely the cannon barrel that fires, on the centreline. The ammunition drum for up to 1350 rounds is behind the armoured cockpit (up to 38 mm of titanium), and aft of that most of the fuselage is filled by fuel tanks; the A-10 does also have wing fuel tanks, but these are unprotected integral tanks, and expected to be emptied before the aircraft reaches a combat zone. The wing is straight, and set low on the fuselage. The A-9 had a high wing but stood low on the ground, low enough to ensure that all maintenance could done without ladders or platforms; the landing gear legs of the A-10 are sufficiently high to allow a not too tall person to walk under the wing. The TF34 engines, more expensive than the F102 but also more powerful and already proven in service, were put on pylons on the aft fuselage. This unusual location ensured that the wings, the tailplane, and the twin tail fins shielded the IR emissions of the engines from a missile seeker on the ground. The high location also reduces the risk of foreign object ingestion, and because the engines are widely separated a hit in one engine is relatively unlikely to result (directly or indirectly) in damage to the other one. 

Combat experience with the A-10 so far has involved a relatively small number of close air support missions. Due to the speed of ground operations, and perhaps also because of the doctrine of the air force, interdiction missions behind the enemy lines have been the majority. For this the A-10 is well-suited, despite being rather slow, because its endurance and large weapons load allow it to pursue several different targets in one mission. With precision-guided weapons targets can be hit from altitude, instead of from the treetop heights envisaged by the designers, and aircraft have mostly stayed out of reach of small weapons and missiles. The gun has nevertheless been much used, mainly against small targets; as the GAU-8/A cannon has a range of up to 3 km the pilot does not necessarily need to get within the range of enemy return fire. (Outranging the powerful ZSU-23-4 anti-aircraft vehicle was an important tactical consideration in its originally intended role.) Nevertheless several pilots had reason to be very grateful for the ruggedness of the A-10 and its ability to survive combat damage. 

As in the case of the Su-25, experience with the A-10 indicated a need for better avionics. One was converted to A-10B configuration, with a crew of two, and equipment for night/adverse weather operations. This was rejected in favour of installing podded systems, such as the Litening ER, on single-seat A-10s. In February 2004 a contract was awarded to install the new “Sniper XR” targeting pod, with IR and optical sighting systems and a target designation system. The avionics of the A-10s were also upgraded, as it was (despite frequent reports and rumours of impending retirement) planned to retain the aircraft in service until 2028.

The Future?

Is there a future for the armoured attack aircraft, and if so, what is it? The concept of structural armour has already been abandoned in favour of a more limited protection of vital systems and a high level of built-in redundancy. The nature of the threat has changed from small arms fire to small missiles, which commonly have warheads that rely on blast and fragmentation rather than armour penetration. In recent conflicts attack aircraft have also found it safer to operate from altitudes that are beyond the reach of man-portable air defence missiles. Doing this, however, depends on the disabling of the longer-ranged air defense systems first, primarily by taking out their radars. 

Modern precision weapons have an uncertain influence. Attack aircraft are being equipped with increasingly sophisticated equipment, to enable them to use these weapons effectively. But on the other hand such weapons could be delivered with sufficient precision on targets close to the own forces also by strike aircraft, by unmanned combat vehicles (UCAVs), or even by large bombers. This removes something of the need for a dedicated attack type.

However, all the effective use of most of this sophisticated technology depends entirely on the accurate gathering, interpreting and redistribution of intelligence. There are limits to the remote planning and executing of operations, both technical and human, and errors are inevitably made. After the NATO intervention in Kosovo, the Serbian commanders claimed that most of the “tanks” destroyed by air attack had been dummies. More recently in Iraq, air attacks on apparently poorly identified targets have created great controversy. Thus the armoured, manned attack aircraft may very well survive. Perhaps not primarily as a weapons carrier, but as a data collection platform with a secondary attack capability. The USAF has experimented with a two-seat A-10B well equipped for this purpose, but in the end decided against it. The idea may well be revived in the future.


  • Ashley, Mark, Russian Upgrade Wars, in AIR International, April and May 2003.
  • Avram, Valeriu, translated by Zaganesco, Georges, 15 missions quotidiennes par pilote, in Le Fana de l’Aviation No.325 of December 1996.
  • Baugher, Joe, American Military Aircraft
  • Boniface, Patrick, Beech’s Destroyer, in Aeroplane Monthly Vol.25 No.4 of April 1997
  • Braybrook, Roy, Rapidly going nowhere? in AIR International, August 1992.
  • Brown, Eric, Testing for Combat, Airlife, 1994.
  • Butowski, Piotr, New spirit in old flesh, in AIR International, March 2000.
  • Collier, Basil, Japanese Aircraft of World War II, Mayflower Books, 1979.
  • Danel, Raymond, and Cuny, Jean, L’aviation Française de bombardement et de renseignement (1918/1940), Editions Larivière, 1978.
  • Davilla, Dr. James J, and Soltan, Arthur M., French Aircraft of the First World War, Flying Machines Press, 1997.
  • Duffy, Paul, Ilyyshin’s Shturmovik, in AIR International, May 1996.
  • Filley, Brian, Junkers Ju 88 in action Part 2, Squadron/Signal, 1991.
  • Francillon, Rene J., Rooks & Hogs, in AIR International, January 1999.
  • Goodall, Michael H., The Wight Aircraft, Gentry Books Ltd., 1973.
  • Green, William, and Swanborough, Gordon, The Complete Book of Fighters, Salamander Books, 1994.
  • Griehl, Manfred, Junkers Ju 88 — Star of the Luftwaffe, Arms and Armour, 1990.
  • Griehl, Mangred, and Dressel, Joachim, Focke-Wulf Fw 190/Ta 152, Motorbuch Verlag, 1997.
  • Herwig, Dieter, and Rode, Heinz, Lufwaffe Secret Projects — Ground Attack & Special Purpose Aircraft, Midland, 2002
  • Jesse, William, and Engbrecht, Bradley, Not Quite Ten — Northrop’s A-9, A-X runner-up in Air Enthusiast, July/August 1996.
  • King, H.F., Armament of British Aircraft 1090–1939, Military Book Society, 1971.
  • King, H.F., Sopwith Aircraft 1912–1920, Putnam, 1980.
  • Leaman, Paul Fokker Aircraft of World War One, Crowood, 2001.
  • Léonard, Herbert Les Chtourmovik, Larivière, 1999
  • Léonard, Herbert Stukas!, Heimdal, 1997
  • Liron, Jean, Borget, Michel, and Bénichou, Michel, Quand le Groupe Schneider faisait de l’aviation de guerre, in Le Fana de l’Aviation, Février 2000. 
  • Mermet, Jean-Claude Le Blohm und Voss 141 — Un mouton à cinq pattes, in Aéro Journal No.25, Juin–Juillet 2002
  • Mikesh, Robert C., Japanese Aircraft — Code Names & Designations, Schiffer, 1993
  • Mikesh, Robert C., and Abe, Shorzoe, Japanese Aircraft 1910-1941, Putnam, 1990
  • Paris, Michael, Winged Warfare — The literature and theory of aerial warfare in Britain 1859–1917, Manchester University Press, 1992.
  • Pelletier, Alain, Beech XA-38 «Grizzly», un ours bien léché, in Le Fana de l’Aviation No.330 of May 1997.
  • Pelletier, Alain, L’Air Force dont l’Amerique ne voulait pas, Le Fana de l’Aviation Hors Série No.27, Decembre 2004.
  • Philpott, Bryan, The Encyclopedia of German Military Aircraft, Arms and Armour Press, 1980.
  • Price, Alfred, Focke-Wulf Fw 190 in Combat, Sutton Publishing, 1998.
  • Price, Alfred, Focke-Wulf’s Dienstpferd, in AIR International, May and June 1992.
  • Price, Alfred, translated by Bénichou, Michel, Des Sangliers dans le Désert, in Le Fana de l’Aviation, October 1998.
  • Rocher, Alexis, Northrop YA-9, rapace resté poussin, in Le Fana de l’Aviation, Novembre 2001.
  • Rosenthal, Léonard; Marchand, Alain; Borget, Michel; and Bénichou, Michel, Nieuport 1909 – 1950, Editions Larivière, 1997.
  • Smith, Peter C., Dive Bomber! An Illustrated History, Naval Institute Press (USA) and Moorland Publishing Co (UK), 1982
  • Thompson, Jonathan, Vultee Aircraft 1932-1947, Narkiewicz/Thompson, 1992.
  • Wallace, James, Frogfeet and Scorpions, in AIR International, September 2001.
  • Williams, Anthony G., and Gustin, Emmanuel, Flying Guns — World War I, Airlife 2004.
  • Williams, Anthony G., and Gustin, Emmanuel, Flying Guns — World War II, Airlife 2003.
  • Williams, Anthony G., and Gustin, Emmanuel, Flying Guns — The Modern Era, Airlife 2004.
  • Yenne, Bill, The World’s Worst Aircraft, Bison, 1990.
  • anonymous, AirData File — Northrop A-9A, in Air Enthusiast, September 1972.