The Military Aviation History channel on YouTube now has this informative discussion on Japanese bombers and policy: https://youtu.be/nstaj1mHFHU
Japanese Competitors
As mentioned before, inter-service rivalry between the Imperial Japanese Army (IJA) and Navy (IJN) took extreme forms in absence of effective government oversight. What both had in common was the need to operate over vast areas: China, the Pacific, and potentially Siberia. The Navy needed to invest in carrier aircraft, which at the time were nearly exclusively single-engined, but also operated land-based bombers, such as the G3M, discussed earlier. The Army fielded a wider array of multi-engined bombers. Navy bombers tended to have a longer range than Army bombers, to cover the vast distances in the Pacific, and a smaller offensive load, because their primary targets were ships, against which torpedoes or armour-piercing bombs might be used. Whether that difference in priorities justified the duplication of designs is something else.
European ideas about strategic bomber offensives appear to have been less influential in Japan, perhaps because its strategists were confronted with longer distances and a paucity of worthwhile industrial targets within reach. During the war in China, the IJA and IJN both attacked Chinese cities, in particular China’s wartime capital Chonqin, but without achieving much strategic effect. In the case of war with the US, a sustained attack on US industry was unthinkable. The enemy’s fleet, shipping lines, and forward bases were far more realistic targets.
The IJA’s standard bomber at the outbreak of the war was the Mitsubishi Ki-21-II, which was later given the allied recognition name “Sally”. The official designation was Type 97 Heavy Bomber, which reflects that the type, first flown in December 1936, had been ordered for service in 1937. That original version, known as the Ki-21-Ia and powered by 850 hp Nakajima Ha-5 radials, had been a very modern bomber in its day. Despite the “heavy bomber” designation, and the goal of replacing the Ki-1 and Ki-2, the specification had demanded a relatively modest bomb load, 1000 kg over short ranges and 750 kg with full fuel. That fuel capacity had been 1,840 liter in the prototypes but was increased to 2,635 liter in the production aircraft. Combat experience in China resulted in the production of the Ki-21-Ib and and Ki-21-Ic with upgrades in crew protection, fuel tank protection, and defensive armament, though the latter remained relatively modest.
The IJA did see the need for a substantially improved aircraft, and the Ki-21-II entered service in December 1940 with 1450 hp Mitsubishi Ha-101 radials. This was essentially an Army version of the 14-cylinder Kasei engine that powered the Navy’s G4M, in redesigned nacelles. Defensive armament remained weak with 7.7 mm machine guns in nose, dorsal, ventral, beam, and tail positions. This quickly supplanted the Ki-21-I in frontline service and equipped most of the Army’s heavy bomber force at the start of the Pacific War. The final Ki-21-IIb of 1942 replaced the dorsal gunnery positions with a turret for a 12.7-mm gun. The Ki-21-II was in the same size class as the IJN’s G4M, with an empty weight of 6070 kg, and a normal bomb load of 750 kg, although it could lift 1000 kg. The Ki-21 was reliable and had good handling characteristics, but by 1943 it was obsolete, and very vulnerable to allied fighters. Production still continued well into 1944, completing 2064 plus some 500 transport versions, known as the MC-21.
While Mitsubishi built the Ki-21-II for the IJA, it was building a different aircraft for the IJN. As soon as the G3M entered service, the Navy had sought a replacement for it with better performance and greater operational effectiveness: A roomier fuselage would allow for an internal bomb-bay and improved defensive armament. A contract for the development of the 12-Shi land-based attack aircraft was awarded to Mitsubishi, which produced a prototype of the new G4M in September 1939. However, not until the spring of 1941 were the first G4M1 production aircraft delivered to operational units, as the Type 1 land-based attack plane model 11. It was a fairly large aircraft, with a wing span of 25 m, an empty weight of 6800 kg, and two 1530 hp Kasei 11 radial engines. The bomb bay was roomy enough for a 800 kg torpedo or the equivalent weight in bombs, but in the early version it did not have opening doors, and a covering panel was only installed if no ordnance was carried. In striving for an impressive maximum range of 4815 km, the designers omitted armour and self-sealing fuel tanks. The wing fuel tanks were integral, i.e. the skin of the wings between the two spars formed the outer walls of the the wing tanks. In combat the bomber would earn unflattering nicknames because of its propensity to catch fire, and its fire-extinguishing equipment was inadequate. This risk was understood during the design work, but accepted. On the positive side, the G4M1 did have a sting in the tail in the form of a 20-mm Type 99 model 1 cannon, while 7.7 mm Type 92 machine guns were installed in the nose, a dorsal position, and lateral blisters.
More than most foreign air forces at this time, the IJN was aware of the need to provide bombers with a fighter escort, thanks to combat experience acquired over China. It would soon acquire an effective long-range fighter in the A6M “Zero”, which would put it in the forefront of tactical thinking about bomber operations. But with much less wisdom or success, it also made an attempt to derive an effective bomber escort from the G4M itself. The G6M1, or “Type 1 wing-tip convoy fighter”, had its armament increased to four 20-mm cannon plus one machine gun, and had some protection for the wing fuel tanks added. Although 30 were built, their inferior performance and handling made them undesired in that role, and they were converted to trainers and transports. Unfortunately for the IJN, the development of the G6M1 had been allowed to delay service entry of the G4M by a year.
As it was able to fly nearly 4300 km with a war load and was well matched in performance to the escorting A6Ms, the G4M1 was effective over China. At the outbreak of the Pacific war 120 were in service, and the type’s role in the destruction of HMS Prince of Wales and HMS Repulse in December 1941 appeared to confirm the usefulness of land-based bombers against fleet units. But on that occasion the enemy ships had no air cover and their anti-aircraft firepower was much less formidable than that fitted to capital ships later in the war. The key weakness of the G4M1 was that it was too easily brought down by a few hits in the wing fuel tanks. Possibly, if projectiles it the integral fuel tanks, the hydraulic shock caused important structural damage even before fire was ignited. From the 663rd aircraft a rubber sheet was applied to the lower surface of the wings, a measure that seems far more effective against leaks than against fire. The G4M1 remained in production until January 1944.
The G4M2 series of 1943 had redesigned wing that allowed the aircraft to carry even more fuel, boosting maximum range to 5615 km. It had more powerful 1850 hp Kasei 21 engines, a dorsal turret with a 20-mm Type 99-1 cannon, and additional of 7.7 mm machine guns in the beam wings and the sides of the nose. Later G4M2 variants had further increased armament, with 20 mm cannon installed in the beam positions and a 13.2 mm machine gun in the nose. And from late in 1943, the G4M2a-Ko was equipped with search radar to find naval targets more easily. But the changes were insufficient to address the type’s high loss rate. In 1944 additional work to improve the survivability of the G4M produced the G4M3. The wing was redesigned again, so that the fuel tanks were no longer integral, and protected fuel tanks were introduced at the expense of range. The tail gun position was redesigned as well. But only about 90 production G4M3s were completed, against 1170 G4M1s and 1158 G4M2s.
A notorious use of the G4M2, specifically the model 24 Tei, was as carrier for the MXY-7 Ohka suicide attack plane, a small aircraft with a 1200 kg warhead. This was ahead of a future in which naval bombers would carry guided cruise missiles as stand-off weapons to engage fleets from a safe distance. But the MXY-7 was essentially a glide bomb that used rocket power only for its final dive, and it still required the carrier aircraft to approach the enemy fleet too closely. And as it weighed 2140 kg, the performance and handling of the carrier aircraft were badly affected and they became very vulnerable to interception. Losses were very heavy, and hits were scored only against ships that formed the defensive perimeter screen of the American fleet. Only ten such missions were flown.
On the whole, although they were built and used in substantial numbers for the duration of the Pacific War, the Ki-21-II and G4M cannot be considered really successful. They fared well in the first phase of the conflict, when the Japanese held the advantage and Allied defences were weak. But in the second half of 1942, as the G4M units were drawn into the protracted Guadalcanal campaign, they were forced to fly repeated long missions to attack defended targets. It could avoid interception by some allied fighters, notably the P-39, by flying at high altitude, but is still suffered steady attrition. When attacks in daylight against American fleet units became suicidal because of their growing fighter and anti-aircraft strength, the remaining expert crews demonstrated a proficiency in night torpedo attacks that came as a nasty surprise to the US Navy, and it was in this role that the G4M would fight until the end.
Already in 1938 the IJA had attempted to put a better defended bomber in service, and it began to be delivered in September 1941. Compared to the older Ki-21, the new bomber was required to be faster, with a top speed of 500 km/h, and more survivable. The Nakajima Ki-49 Donryu featured a manned tail turret, instead of the remote-controlled machine gun that was retrofitted to the Ki-21. The Ki-49 also had a better designed dorsal gun position with a 20-mm Ho-1 cannon, was better armoured, and had basic protected fuel tanks. Allied investigators found that the aluminium fuel tanks were covered in three different layers of rubber (of which the middle one would dissolve and expand in contact with fuel) and then wrapped in felt and covered with silk. This design was not as effective as allied self-sealing tanks, but it provided some protection. The crew also enjoyed the protection of a number of armoured plates 12.5 or 16.5 mm thick, and a bullet proof windscreen. The problem was that all this added weight, and with its 1250 hp Ha-41 engines the Ki-49-I was underpowered. Though the Ki-49-I was nominally designed for the same 1000 kg bomb load as the Ki-21, on operations it could carry less than the older bomber. The army accepted only 128. The Allied forces gave it the reporting name “Helen”.
From September 1942 the Ki-49-II was built with the 1520 hp Ha-109. This was to be the major production model, with 366 completed of the Ki-49-IIa and 301 of the Ki-49-IIb. The latter had improved defensive armament, with the 7.7 mm Type 89 machine guns in the nose, ventral and tail positions being replaced by the substantially harder hitting 12.7 mm Ho-103. The Ki-49 reportedly had good flying characteristics, though not as good as the Ki-21. But the air force was still disappointed by its performance and never considered it a satisfactory replacement for the Ki-21. Nakajima started development of a version with 2420 hp Ha-117 engines, designated Ki-49-III, but only six of these would be built.
Incidentally, in a not atypical example of army-navy parallelism, the army also tried to develop its own “wingtip convoy fighter” from the Ki-49, which resulted in the Ki-58. This was recognised as a bad idea and abandoned after the completion of three prototypes. Nakajima also tried to design an improved Ki-49 as the Ki-82, but this project came to nothing.
The best of the IJA’s medium bombers was the Mitsubishi Ki-67. Mitsubishi had been given a contract for a next-generation bomber on 17 February 1941. The engine it chose for its new design was the 1900 hp Ha-104, an 18-cylinder development of the Kasei. This turned out to be a powerful and reliable engine, and it was neatly installed in streamlined cowlings with a cooling fan in front. Attention to aerodynamics also ensured that the relatively powerful defensive armament was faired in as much as possible. The Ki-67 quickly showed its potential, with good performance and excellent agility for an aircraft this size. Development was slow, however. The first prototype flew on 17 December 1942, the second in February, and the 16 pre-series aircraft were completed between May and March 1944. In April 1944 the first production aircraft were delivered, and it reached combat units in the summer. The type had been carefully designed to be easy to build in quantity, but production was affected by the worsening of the war economy, increasing US bombing, and an earthquake that disrupted production of the Ha-104 engine. By the end of the war, only 688 were delivered.
The Ki-67 Hiryu was a relatively small bomber, with a wing span of 22.50 m and an empty weight of 8650 kg. A top speed of 537 km/h at 6000 m and a cruising speed of 400 km/h at 8000 m were not good enough to avoid interception. But it was not an easy prey, and was praised for its good handling and fighter-like manoeuvrability, being able to fly aerobatic manoeuvres with ease. (The IJA unsuccessfully tried to develop the Ki-109 bomber interceptor from it.) It had self-sealing fuel tanks, armour plate for the crew, and decent defensive armament, with a 20-mm Ho-5 cannon in a dorsal turret, and 12.7-mm Ho-103 machine guns in the nose, the tail, and lateral blisters. (Later models had twin tail guns.) At low level it was quite fast, though the performance of the Ha-104 decreased at altitude. It carried a relatively small bomb load of up to 800 kg but was soon modified to carry a 1070 kg torpedo. Because, in a surprising development, Army Ki-67 units were selected to fly attacks on US fleet in conjunction with Navy bombers, and even more strangely, the Navy received a number of Ki-67 bombers as well. Some of the best crews of the Army and Navy manned the respective “Taifun Attack Force” and “Yasukuni Naval Air Group”, but their celebrated claims of sinking multiple major allied warships turned out to be false. To avoid the numerous fighters that established overwhelming allied air superiority in daylight, units trained to attack at night or in bad weather, and some Ki-67s were fitted with radar. The Ki-67 was also one of the aircraft converted to suicide attack use. More conventionally, the “Peggy” (as Allied intelligence officers dubbed it) was used to strike back against the B-29 bomber bases in the Marianas.
The Ki-67 was too little, too late, but it has been considered the best Japanese bomber of the Pacific war. It was clearly superior to the Ki-21, Ki-49, and G4M. Meanwhile, the development of new bombers for the IJN ran on two seemingly contradictory tracks.
One of those was an investment in four-engined long-range bombers to replace the G4M in a conventional bombing role. The Nakajima G5N Shinzan of 1941 was not a success, as it was overweight and underpowered, and the Navy cancelled its production order. As a replacement Nakajima developed the G8N Renzan, but when this made its first flight in October 1944, Japan’s situation had already deteriorated so much that it had to be cancelled. Much like Italy, Japan was not in a technological and industrial position to successfully build a meaningful fleet of four-engined bombers. (It did have an excellent four-engined long-range reconnaissance type in the Kawanishi H8K flying boat, but it completed only 167 of those.)
The other track was the development of a small, high-performance type for tactical missions. Late in 1944, the P1Y Ginga made its combat debut. Designed at the naval arsenal in Yokosuka, the P1Y had been designed as a fast, long-range multi-role aircraft optimised for low-level anti-shipping operations, capable of level bombing, dive-bombing and torpedo-bombing. With an emphasis on speed over defensive armament, the P1Y only had a gun in the nose and one in a dorsal position (depending on the model either 20-mm Type 99 cannon or 13.2 mm Type 2 machine guns) and a crew of three. The initial engine chosen was the 1800 hp Nakajima NK9B Homare 11, which was compact and powerful but unfortunately unreliable. The P1Y1 made its first flight in the middle of 1943 and entered production soon after, but actual service entry had to be delayed until the engine problems were solved and other reliability and maintenance issues addressed. About a year passed until the 1825 hp NK9B Homare 12, with improved cooling, was available. The bomber had a maximum range of 5370 km, a normal operational range of 1920 km, and a top speed of 547 km/h. That was good enough that the P1Y was also produced in a night-fighter version, the P1Y-1S. As alternative to the Homare engine, 1850 hp Mitsubishi Kasei 25a engines powered the P1Y2-S model which entered service in the middle of 1944, but most of these were reportedly converted to P1Y2 bombers because of poor performance at altitude. The IJN finally received about a thousand P1Ys. Fast and agile at low level, the P1Y was potentially a deadly anti-ship weapon, but because of the long development delays it entered service when the operational capabilities of the IJN were in sharp decline, and US carrier task forces were now too large and too well defended.
Arguably, at the outset of the war, Japan had a need for a good multi-role twin-engined bomber, able to fly anti-shipping missions but also suitable to attack Allied bases and deliver support to its ground units. As long as its armed forces were on the offensive, the benefits of having a compact and easy to operate multi-role bomber were great, because the size of the force that could be sustained on remote bases in the Pacific was limited. Also, before December 1941 the Japanese were primarily engaged in a land war in China and undetermined whether to start another land war with the USSR or a naval war with Britain and the USA. That the IJA and IJN developed two parallel bombing forces was inherently wasteful, but arguably for 1942 their best bomber was the G4M, despite the type’s vulnerability. As the tide shifted against Japan and it was forced in a defensive position on the home islands, its primary need was for a more specialised naval bomber that could operate from home bases against enemy fleet units. Both the P1Y and Ki-67 met the need but probably the Ki-67 was the better choice. Either aircraft could have been in service significantly earlier if their development had been less prolonged. The G4M and Ki-21 both remained in service too long because their successors were delayed.
Despite the enormous difference in size of the Pacific Ocean and the Mediterranean Sea, the Japanese and Italian experience had parallels. Both powers had to focus on twin-engined medium bombers because they lacked the industrial capacity to build four-engined heavy bombers in meaningful numbers. And for both, controlling the sea lanes was more critically important and more achievable than hitting out at enemy industry or populations.
Flying Guns 