During the Second World War, the German nightfighters very successfully used upward-firing guns against British bombers. This form of armament, named Schräge Musik armament was introduced by the Germans in late 1942. It allowed German nightfighters, approaching from below and behind, to fire their shots accurately and from a direction where the bombers had weak defenses and many vulnerable spots. Typically, they aimed for the wing fuel tanks.
The RAF seems to have remained ignorant of its existence for many months. As far as countermeasures were concerned that was probably of limited importance, because RAF officers knew very well that the preferred direction of attack for a nightfighter was from below, even with conventional armament. The great difficulty was in finding an effective defense: The available ventral turret designs were too ineffective to be worth carrying. The American Sperry ball turret could have been an effective solution, but would have required major redesign of the bombers. In any case, the nightfighter was difficult to spot against the dark background of the earth, while the bomber was often clearly silhouetted against the night sky.
Still, the failure to recognize the nature of this threat has puzzled many observers, who knew that between the two World Wars, the RAF itself experimented with fixed, upward-firing guns in a number of aircraft. This disconnect in thinking probably occurred because the RAF saw the upward-firing guns in a different context: That of no-allowance shooting.
The idea behind no-allowance shooting was to balance the body lift of the projectile against the force of gravity, with the goal of giving the projectile, at least for some distance, a straight flight path instead of one that curved downwards. The diagram below makes this a little clearer.
The projectile, fired upwards from a fixed position, would leave the barrel with a muzzle velocity represented here by the red arrow: The length of this arrow corresponds to the speed of the projectile, and its orientation to the angle of firing. A projectile fired from a rifled barrel is spin-stabilized: It is a little gyroscope, that resists attempts to change its initial orientation. On firing, the axis of the projectile is parallel to the red arrow, and it will (if we ignore the finer details of projectile ballistics) remain so.
In the air, the aircraft flies at a speed corresponding to the blue arrow. This speed was, until well after WWII, always much smaller than that of the projectile; so for clarity, the length of the blue arrow has been exaggerated here. When the gun is fired from an aircraft, the projectile will move with a speed that is the combination of the speed of the aircraft (blue) and the muzzle velocity of the gun (red). The two motions have different directions, so the final path of the projectile (green) is at some intermediate angle.
The resulting direction of travel is along the green arrow. But the orientation of the projectile remains (at least initially) parallel to the orientation of the barrel, which is the direction of the red arrow. Relative to its flight path, the projectile thus has a small nose-up attitude. This gives it a certain amount of body lift. If the angle of the gun is chosen correctly for the speed of the aircraft and the muzzle velocity, this lift will compensate for gravity, and the projectile will travel in a straight line. This is the theory of no-allowance shooting. Of course drag will slow down the projectile, so the desired straight flight path can only be achieved for part of the trajectory.
This diagram has an important consequence. It is fairly obvious from it that for a faster aircraft (a longer blue arrow) the angle between the red and green arrows will become larger. But only a small angle between projectile and flight path is desirable to get the required lift, and it needs to be even smaller if the projectile flies faster. The logical way to reduce the nose-up attitude again is to make the angle in which the gun is installed in the aircraft smaller: In our diagram, tilting the red arrow towards the blue arrow. Therefore, the faster the aircraft flies, the shallower the angle at which the guns should be installed on the aircraft.
At the end of the first World War, the Coventry Ordnance Works were producing a 37-mm cannon for aircraft use, known as the 1½ pdr COW gun. This was a fairly large weapon, weighing 95 kg, and firing a 37 × 190 cartridge. It was relatively slow-firing at 100 to 120 rpm, and the muzzle velocity was 580 to 610 m/s. This was a modest performance, similar to that of the 37 mm M4 cannon which armed many P-39 Airacobra and P-63 Kingcobra fighters during WWII.
The design of the COW gun (and the later M4) was compromised by the necessity to keep both the weight and the recoil forces down, to allow installation in aircraft. The COW gun was a long-recoil design, with the recoil spring wrapped around the barrel, and the recoil blow was only about 2000 lb (910 kg). Nevertheless, the weapon was too big and heavy for the conventional fighters of the period. More so because, unlike the Vickers machine guns used as standard fighter armament at the time, the cannon could not be synchronized to fire through the propeller.
In October 1918, the COW gun was installed in the rear cockpit of an Airco D.H.4, fixed and aimed up. The angle of the installation in at least one (A2168) was about 80 degrees. The gun fired through a hole in the upper wing. The idea was to use this weapon against the large German bombers and Zeppelins, and after some reinforcement to prevent blast damage, testing was successfull. Three D.H.4s with COW guns were put into service. They carried a crew of two; the gunner in the aft cockpit not only loaded the weapon – the breech was inconveniently close to the cockpit floor – but also fired it on a signal from the pilot, who aimed with a ring-and-bead sight. Though two were sent to France, they did not see any combat before the war ended. (A2168 is recorded as flying a sortie against German Gotha bombers in August 1917, but that was presumably still without the COW gun.)
The British specification 4/24 for a Twin-Engined Home Defence Fighter (the code implies that it was the 4th specification released in 1924) called for a twin-engined, three-seat aircraft designed to destroy enemy bombers at night. On closer consideration, the RAF decided to install the COW gun on this aircraft, and amended the specification accordingly. Two designs were offered, the Bristol Bagshot monoplane and the Westland Westbury biplane. The Bagshot had serious structural problems and was very little flown. The COW gun was successfully test-fired from the nose gun position of a Westbury, where it was installed on a flexible mounting. Armament tests continued for years, and at one time the COW gun was installed fixed, firing upwards, in the middle cockpit. The large number of structural damage reports filed between 1927 and 1930 suggests that the recoil force was a bit too much for the airframen to take. Anyway, the Westbury was still slow, ponderous, and of highly dubious military usefulness; it did not enter production.
The low performance of the 4/24 aircraft condemned them to flying patrols on station, because with a top speed of only 125 mph (201 km/h) they were simply too slow to catch up with an enemy bomber otherwise. Specification 27/24 started from the healthier concept of a twin-engined, single-seat aircraft, again intended to intercept enemy bombers at night. Boulton & Paul offered the P.31 Bittern, by standards of the time a good-looking and advanced monoplane with a semi-cantilevered wing -- although struts had to be added to the second prototype to prevent the wing twisting when the ailerons were used. Two prototypes were built, one with fixed Vickers guns firing forward, the other with a Lewis gun on either side of the nose, on a mount that allowed the inclination to be changed from 0 to 45 degrees. For aiming, a ring-and-bead sight was provided on a frame in front of the cockpit, which was elevated together with the guns. This interesting concept was not properly tested, because the aircraft was written off after the undercarriage collapsed.
Meanwhile, thought had centred on a “no-allowance” installation of the COW gun, fixed, in a smaller and better performing single-seat fighter. This concept was formalized in specification F29/27. The weapon would be installed at an angle of about 45 degrees in an aircraft designed to intercept, in the shortest possible time, an enemy bomber flying at 150 mph (240 km/h) and 20,000 ft (6100 m). This incidentally, was the same target described in specification F20/27 for a conventionally (but heavily) armed fighter. The aircraft also had to be a steady gun platform. Two prototype aircraft were built to this specification, and they were usually called the “COW gun fighters” as neither was given a proper name.
Westland took advantage of the overlap between specifications F20/27 and F29/27. Its design for F20/27 was a fairly attractive low-wing, strut-braced monoplane, with fixed landing gear. Although it did not attract orders (the F20/27 competition was won by the Gloster Gauntlet), the aircraft was an advanced design for its time. The offering for F29/27 was an obvious derivative, with the COW gun installed in the right-hand side of the cockpit. Unfortunately, the Westland COW gun fighter failed to live up to its promise: The handling qualities were apalling, with a strong and at high speed uncontrollable tendency to roll, a “disquieting” vibration of the tail surfaces, and sluggish aileron response.
Vickers' submission for F20/27 was even more advanced than Westland's offering. The Vickers 151 Jockey was a low-wing monoplane with a cantilever wing, although it still had fixed landing gear. Technically, it was well ahead of both the Westland F20/27 and the contemporary Boeing P-26. Yet the Vickers 161 offered for F29/27 was not only entirely different, it was a throwback to a completely outdated concept. The Vickers submission was a heavily strut-braced, pusher biplane; as in the Vickers F.B.5 “Gun Bus” of the First World War, the engine was put behind the pilot, with a propeller caged between tail booms, to leave the front fuselage available for the weapon. There were some concessions to modernity: Light alloy was used for the construction, the numerous struts were broad-chord and nicely faired, and a curious conical fairing behind the pusher engine and propeller attempted to reduced drag.
The COW Mk.III gun was installed at an angle of 45 degrees, to the right of the pilot, whose cockpit was set off to the left to keep the nose short. Fifty rounds of ammunition would be carried (thirty on the prototype), all but six in horizontal racks at the sides of the cockpit from which they would have to be loaded manually. Concerns about the effect of that the recoil would have on the handling of the aircraft were addressed by Vickers, that showed them to be well within limits. Vibration was addressed by mounting the instrument panel on rubber sponges. The spent cases were collected in a box, for if they were dumped overboard they would hit and damage the propeller; maintenance crews disliked the lack of accessibility of the box and the gun. Barr & Stroud provided a periscopic gunsight, but test pilots were disatisfied by this because they had to peer down into the cockpit to use it. A gun trigger was provided on the control column, but it was so stiff that the force required to use it affected aiming.
The completion of the Vickers 161 was slow, leading to some irritation at the Air Ministry. After it made its first flight in December 1930 or January 1931, modifications to address structural and handling problems continued, and official trials were completed only in December 1932. They revealed the Vickers 161 to be a much more satisfactory aircraft than the Westland F29/27, but still one that suffered from a disappointing performance. As a combat type, it was a non-starter, and reportedly only 24 shells were fired by its gun. Ironically, the gun in the Westland F29/27 saw much more use, if only during ground trials.
In 1932 the Air Staff had the interesting idea to ask all people in the RAF for their ideas on the best design for a “Novel Fighter”. The notion behind this competition, for which a comittee was created to judge submissions, was that what Sir Hugh Dowding called “the united intelligence of the Service” was more likely to come up with relevant ideas than the aircraft industry. The concept the comittee selected from the submissions was a two-seat fighter with a nose turret, and one or two pusher engines. This remained a paper concept.
But the committee made an interesting comment on the designs submitted to it that featured “no-allowance” gun installations: It judged that these would be particularly useful at night. This was of course in 1933, ten years before the RAF would be surprised by the German use of upward-firing guns, but well after the evaluation of the aircraft designed to specifications 4/24, 27/24, and F29/27.
The no-allowance concept was not just limited to the big COW gun with its modest muzzle velocity, but seemed to hold promise for machine guns installations as well. In 1934, two Bristol Bulldog fighters were modified to carry guns elevated at 60 degrees; one carried Lewis guns, the other one Vickers. During tests, 90% hits were achieved on a target towed 1000 ft above the aircraft.
In 1937 Sir Hugh Dowding, Air Officer Commanding of Fighter Command, again drew attention to the possible usefulness of guns installed at the no-allowance angle. In response, the Armaments Branch reviewed the concept.
Its conclusion was that, at the increased speeds of the new monoplane fighters, the elevation of the guns required to enter the “no-allowance” condition had greatly decreased and that for all practical purposes, the method blended in conventional gunnery. The first draft of the new fighter specification F18/37 still demanded guns that could be adjusted for no-allowance shooting, “i.e. from 0° to 15° elevation”. But this provision was dropped before the specification was officially issued.
The concept still wasn't dead yet. Gloster received a specification F9/37 for a twin-engined fighter, with nose cannon installed at the no-allowance angle of 15° (another source says 12°) and a retractable dorsal turret with four .303 guns. Later, the turret was deleted from the specification and replaced by more fixed cannon: The angle of 15 degrees made it possible to install three additional cannon in the fuselage behind the pilot, firing over his head. The F9/37 was a neat aircraft, but the prototypes were powered by Bristol Taurus radials or liquid-cooled Rolls-Royce Peregrine engines, both rather unsatisfactory powerplants. When designer W.G. Carter tried to reawaken interest in the type, the Air Ministry suggested using the Merlin XX engine instead, and the F9/37 was duly reworked as the Gloster ‘Reaper’. The design was judged favourably, but the success of the Beaufighter and Mosquito removed the operational need for it.
A drawing dated September 1937 by Vickers Supermarine showed the outlines of a model 319, a twin-engined fighter with a similar outline as a Mosquito, featuring two fixed 20-mm cannon firing forward and two firing 65 degrees upwards. The second crew member would set behind the wing leading edge and the guns, perhaps with the replacement of ammunition drums and/or clearing of gun jams in mind. This does not seem to have progressed beyond a sketch.
So by the start of World War II, the concept of no-allowance sighting had been largely discredited, and it could be said that it never had been given a fair chance. The impressive series of prototypes mentioned above have this in common, that they all failed testing or were rejected for service for reasons that were essentially unconnected with their armament. Outdated airframe design, structural failures, aerodynamic problems, accidents during testing, a poor choice of engines: A sorry list.
The interest of the RAF in upward-firing guns was not limited to fixed guns: After the reasonably successful experience with the nightfighter versions of the Defiant, the staffs were repeatedly tempted to install dorsal turrets or nose turrets on nightfighters. Apart from the experimental Beaufighter Mk.V withs its dorsal turret (four .303 guns), a mock-up was made of a Mosquito with a dorsal turret, and turreted guns of calibres up to 40 mm appeared on paper. A dorsal turret for the 40 mm Vickers S cannon was test flown on a Wellington bomber. But in all these cases, the turret added substantial weight and drag, and either the firepower that could be installed in it was limited, or the size of the aircraft that had to carry it grew too much. The American P-61 Black Widow entered service with a dorsal turret with four .50 guns, but with hindsight this was a feature that ought to have been omitted from the design.
An opportunity and a threat had been overlooked, not because the RAF was unaware of the usefulness of upward-firing guns to a nightfighter, but because somehow officers never hit the right concept. They were close to it, not just one but repeatedly, for a period of ten years or more; and it still managed to escape them.