Author: qaz

  • How Many Type 98 Light Tank (Ke-Ni) Were Actually Produced?

    How Many Type 98 Light Tank (Ke-Ni) Were Actually Produced?

    The Type 98 Light Tank (Ke-Ni) was developed as the successor to the Japanese Army’s primary light tank, the Type 95 Light Tank (Ha-Gō). Conceptualized in 1938 with the first prototypes appearing in late 1939, the Type 98 Light Tank was a significant improvement over its predecessor in every respect. It featured thicker armor, a lower profile, a more effective two-man turret, and a coaxial machine gun complementing its main armament. Despite these advantages, the Type 95 Light Tank would remain the only Japanese light tank deployed overseas, seeing widespread combat during the Pacific War.

    The Type 98 Light Tank only entered “mass” production in 1942 due to a somewhat protracted development process, but more-so because of a lack of strategic impetus. The existing Type 95 was producing satisfactory results in its infantry-support role in the invasion of China, and with the initiation of the Pacific War, the Army’s production priority for tanks plummeted relative to aircraft, cannons, and infantry equipment. It’s difficult to say that this was a poor decision.

    The production priority of tanks would not rise until 1945 in anticipation of the defense of the Japanese home islands, by which point Japan’s capability to produce weapons was deteriorated, and the production of new light tanks in general had ceased in favor of more powerful vehicles to counter Allied armor.

    The production run of the Type 98 Light Tank, by essentially all post-war accounts, amounted to over 100 complete vehicles. For example, in Japanese Tanks by Takeuchi Akira & Hara Tomio, it is written that “From 1942 to 1943, 100 vehicles or more were produced, and were deployed to newly established units.” Western historian Steven Zaloga wrote in Japanese Tanks 1939-1945 that 104 vehicles were completed. US technical intelligence, in postwar reporting on Japanese armor, similarly stated that “It is believed that slightly more than 100 of these tanks were completed.”

    As an indication for the “general understanding”, at the time this article was written, the English Wikipedia page on the tank cites the 104 figure, and the Japanese page states 113.

    There is nothing peculiar about these claims. This general figure of “about 100” is a consensus that originates from primary documents compiled by the US Strategic Bombing Survey’s investigation into Japanese ordnance production immediately after the war. However, closer examination of these postwar records, along with surviving wartime documents, reveals that the actual quantity of production vehicles was far lower than what is widely accepted.


    The Origin of the “100” Figure

    Following the surrender of Japan on August 15th, 1945, the US Strategic Bombing Survey launched a thorough investigation into Japanese ordnance production during the war. The USSBS ordered tabulated production results from the Army Ordnance Administrative Headquarters, which oversaw the development, production, and supply of ordnance. Below the Administrative Headquarters was the Army Arsenals, each of which oversaw the production of ordnance at their own facilities and subservient civilian factories.

    In the days immediately following the announcement of Japan’s surrender, many documents were burned on high orders before their preservation was enforced by the US. Many were also lost to the effects of strategic bombing before Japan had surrendered. Even without considering the destruction of records, it is indicated that the production recordkeeping by the Ordnance Administrative Headquarters and the arsenals subservient to it were often imprecise at any detail finer than yearly, as noted in one report:

    The Army Ordnance Bureau and the arsenals kept most of their output records for bookkeeping purposes rather than to follow production closely, and they had a tendency to enter production from the preceding months in the fiscal records of the last 2 or 3 months of the fiscal year. Then they would slacken up on their bookkeeping entries for the first 1 or 2 months of the new fiscal year. This practice tends to overemphasize the disparity between output in March, the last month in the fiscal year, and April, the first month in the fiscal year.

    This led to a situation where, rather than being able to refer to accurate records, the relevant organizations had to piece together estimates of monthly production counts after the war from sparse, imprecise data. As a result, there were not many cases where the monthly production counts were anything more than estimates or even just pure guesswork. Nonetheless, these production investigations were carried out at the level of the Army Ordnance Administrative Headquarters and at the level of each arsenal.

    The Army Ordnance Administrative Headquarters Investigation

    On November 3rd, 1945, the Army Ordnance Administrative Headquarters submitted the first monthly production count of the fiscal years 1941 to 1945, titled「昭和一九四一~一九四五年 兵器生産状況調査表」[1941~1945 Ordnance Production Status Investigation Charts]. This document was not precise, generalizing several types of vehicles into one, and did not even include the Type 98 Light Tank.

    Subsequently, a revised version of this document with much more precision was submitted on the 15th of November by the HQ, titled「昭和十六~二十年 月別兵器生産状況調査表」[1941~1945 Monthly Ordnance Production Status Investigation Charts]. The production of the Type 98 Light Tank in this document was recorded as follows:

    Type 98 Light Tank Production According to the Army Ordnance Administrative HQ
    AprMayJunJulAugSepOctNovDecJanFebMarTotal
    FY1942572332224
    FY194365656584710121589
    Total113

    This document is the origin of the figure “113“. Now let’s look at the production investigation charts submitted by the Sagami Army Arsenal itself, which directly oversaw the grand majority of tank production during the Pacific War.

    The Sagami Army Arsenal Investigation

    The Sagami Army Arsenal submitted the English document “Production Chart of Vehicles with Catapillar by Mfg Plants in 1941~1945” [sic] on November 17th, 1945.

    Type 98 Light Tank Production According to the Sagami Army Arsenal
    AprMayJunJulAugSepOctNovDecJanFebMarTotal
    FY194111
    FY194257235224
    FY1943(2)(12)(41)(7)(4)34613+(66)
    Total38 + (66) = 104

    The 38 vehicles without parenthesis were listed as “M98 Light Tank” and were produced by Mitsubishi Heavy Industries, Hino Heavy Industries, Kobe Steel Works, and the Sagami Arsenal itself. It should be noted that the additional 66 units listed with parenthesis added by myself are written in the document as “M98 Light Tank (Repair)” and were all carried out by the Sagami Arsenal.

    It seems that the origin of the figure “104” is the assumption that “M98 Light Tank (Repair)” was a typo for what was actually 66 newly produced Type 98 Light Tanks. This even appears to be a sensible conclusion on the surface, as there were not 66 Type 98 Light Tanks in existence to “repair” during that period, it would seem to fill in a strange gap in the production record, and it brings the total production count somewhat in-line with the figure reported by the Ordnance Administrative HQ (113).


    Problems with the Established Figures

    Of these two sources of production counts detailed above, the easiest to dispel is the notion that the Sagami Army Arsenal recorded the production of 104 Type 98 Light Tanks.

    While the entry “M98 Light Tank (Repair)” was indeed a typo, it was not a typo that would create a production tally of 104 for the Type 98 Light Tank. Rather, the original Japanese manuscript that was created on November 10th, 1945 lists these vehicles as 「八九式軽戦車(修理)」[Type 89 Light Tank (Repair)].

    This work carried out at the Sagami Arsenal in 1943 was not the repair, much less construction, of 66 Type 98 Light Tanks. It was the repair of 66 Type 89 Light Tanks, or what was more appropriately known as the Type 89 Medium Tank (I-Gō) at that time. (The Type 89 was once a light tank, but was redesignated as a medium tank in 1935).

    This means that what the Sagami Arsenal actually reported in its investigation was the production of only 38 Type 98 Light Tanks – 1 in FY1941, 24 in FY1942, and 13 in FY1943. Going forward, we will not be considering the single unit built in FY1941, as this was a prototype, and I do not hope to solve the question of “how many Ke-Ni prototypes were made” in this article.

    Both the Sagami Arsenal and Ord Admin HQ investigations agree that 24 of the Type 98 were produced in fiscal year 1942. We can confirm the legitimacy of this figure with a surviving document that recorded the year’s production of ordnance in fiscal year 1942, which was published on March 31st, 1943, titled「昭和十七年度主要兵器整備状況調査表」[FY1942 Major Ordnance Production Status Investigation Table]. This document was reprinted in Senshi Sōsho Volume 33.

    Thus we can be quite confident that the total production of the Type 98 in fiscal year 1942 amounted to 24 vehicles.

    The discrepancy that remains is now just fiscal year 1943, where Sagami’s investigation reported 13 tanks completed, but the Ord Admin HQ’s investigation reported a far more substantial 89. Luckily, there is another surviving document which clears the air on which of these counts is more reliable.

    A table created by the Sagami Army Arsenal on February 4th, 1944 titled「完成車両整備現況表」[Complete Vehicles Production Status Table] details the production status of the Type 98 Light Tank in FY1943 quite clearly.

    This document reveals that 15 Type 98 Light Tanks (labeled “98/Se-K (Light)”, their arsenal designation) were completed in fiscal year 1943, with no more anticipated before the year’s end. The number ordered for that year had been 80, while a total of 96 were arranged for (16 that were not completed the previous year were carried over), making the production shortage for this year 81 vehicles.

    15 vehicles produced in 1943 is very close to the arsenal’s postwar count of 13, but how did the Army Ordnance Administrative Headquarters arrive at such a grossly incorrect tally of 89? I have determined that this mistake was a simple oversight related to how the in-house production at the Sagami Army Arsenal was recorded.

    In 1942, the Sagami Arsenal reported their production of vehicles in this incidence:

    • Type 97 Medium Tank
    • Type 98 Light Tank
    • Type 98 6t Tractor
    • ……

    Subsequently, in 1943, there was no production of the Type 98 Light Tank in-house at the Sagami Arsenal itself, and I believe that when tabulating Sagami’s data, the Ord Admin HQ mistakenly counted that year’s production of the Type 98 6t Tractor within the arsenal as a “Type 98 Light Tank”.

    Sagami Arsenal reported the completion of 74 Type 98 6t Tractors in 1943. If we remove 74 vehicles from the Admin HQ’s postwar count, we get 15 Type 98 light tanks completed in 1943 – which is in perfect agreement with the wartime document created by Sagami.

    Loose Ends

    A handful of prototypes of the Ke-Ni were naturally also completed (at least 3), which I plan to investigate in a more comprehensive article on the Type 98 (if I do ever manage to write something again).

    There is evidence that the Type 98 Light Tank was planned to be produced in some quantity by the Kokura Arsenal, and the production numbers 2001~3000 were allocated, but I have not found substantial evidence that the Kokura Arsenal actually produced any vehicles in the end.


    Conclusion

    The long-accepted statement that over 100 Type 98 Light Tanks were produced is a myth caused by clerical errors in the confused aftermath of the war. Evidence definitively shows that the production run of this tank from 1942 to 1943 only amounted to a mere 39 vehicles.

    The breakdown is as follows:

    • FY1942: 40 vehicles were ordered, split between the Sagami Arsenal itself, and subservient civilian factories of Mitsubishi Heavy Industries, Hino Heavy Industries, and Kobe Steel Works.
      • 24 vehicles were actually completed.
    • FY1943: 80 vehicles were ordered, and 16 incomplete orders were carried over from the previous year, split between the Sagami Arsenal itself, and subservient civilian factories of Mitsubishi Heavy Industries and Kobe Steel Works (Hino was no longer producing this vehicle).
      • 15 vehicles were actually completed.

    After this, production shifted to the Type 2 Light Tank (Ke-To) solely under the responsibility of Kobe Steel Works, but this is another can of worms that I do not wish to open just yet.

    The Type 98 Light Tank was hardly a vehicle that saw limited mass production, but one that barely entered production at all. What was a capable design when developed was ultimately sidelined by the Japanese Army’s shifting strategic priorities, with the focus falling on the mass deployment of existing capabilities. Tanks, in general, were a lesser concern for island warfare. The Type 98 Light Tank would not see combat, and all examples were deployed within the Japanese home islands when the war ended.

    Incidentally, after the war, a total of 39 Type 98 Light Tanks appear to have been accounted for in the US Eighth Army occupation zone.


    Sources
    • 作業課長会同に関する書類綴(2分冊の2) 昭和19年2月
      • 17.造兵廠現況説明/作業課長会同提出書類 相模陸軍造兵廠 昭和19年2月6日(2)
    • 兵器固有番号に関する綴 昭和20.4~6
      • 兵器固有番号打刻に関する件通牒 等(1)
    • 昭和一九四一~一九四五年 兵器生産状況調査表
    • 昭和十六~二十年 月別兵器生産状況調査表
    • 相造 合衆国戦略爆撃調査団調査資料(予備)造兵課
    • PRODUCTION SCHEDULE (PART JAPANESE). REPORT NO. 45A(1)
    • CHARTS (SUPPLEMENTARY DATA TO QUESTIONNAIRE). REPORT NO. 45A(3)
    • PRODUCTION CHARTS. REPORT NO. 45A(4)
    • TABLE OF PRODUCTION AND MILITARY SUPPLY(REPORT BY ARMY ORDNANCE BUREAU). REPORT NO. 45V
    • 386.3: Report of Disposition of Surrendered Enemy Equipment
    • Ordnance Technical Intelligence Report No. 21
    • 高崎正男『戦史叢書第033巻 陸軍軍需動員<2>実施編』(1970)
    • 竹内昭、原乙未生『日本の戦車』出版協同社(1978)
    • Zaloga, Steven. Japanese Tanks 1939-1945. Osprey Publishing (2007)
  • The Uprating Capacity of the Ha-45 / Homare Engine

    The Uprating Capacity of the Ha-45 / Homare Engine

    It is well established that the performance of Japanese aircraft engines in WWII was limited by the suboptimal fuels available at the time, among other factors. As a result, in order to achieve boost pressures in the ballpark of high-power Allied engines, it was necessary to rely on water injection even at nominal operation.

    Nonetheless, aviation enthusiasts often speculate on the performance of Japanese aircraft if they were supplied with high-octane fuels and up-rated appropriately. This is partially due to the fact that there are widespread myths of superior performance numbers being achieved with Japanese aircraft using American high-octane fuel, when in reality, in most if not all cases these numbers are wartime calculations using inadequate data.

    Nonetheless, it is an interesting question. In a past article on the Ki-84, I wrote that “it’s unknown how much additional pressure a Japanese mass-produced Ha-45 could even handle.” Thanks to an article I recently read in the book「日本航空学術史 1910-1945」(Aeronatical Researches in Japan 1910-1945), there is a satisfactory answer to this.


    Single-Cylinder Test Units

    From April 1939 to March 1944, the Engine Department of the Navy Aviation Technical Arsenal (Kūgishō) manufactured single cylinder testers and conducted tests on the power enhancement of aircraft engines using these machines.

    In the past, these experiments relied on foreign imported single-cylinder testers manufactured by the American SPE Company (Self-Priming Pump and Engineering Company) and the German DVL (Deutsche Versuchsanstalt für Luftfahrtforschung). However, this posed a problem when experiments resulted in a broken part of the test units, as they were not domestically produced.

    To solve this problem, it was decided to manufacture a domestic single-cylinder tester that could use Japanese mass-produced engine components. The single-cylinder tester was to have easily changed experimental parameters and be able to withstand harsh operation.

    For changing the engine compression ratio, the DVL method was adopted in which the engine cylinder mounting base could be vertically adjusted with a handle. An SPE type balancing rod was used to manage the balancing of dynamic forces at high operating RPM. The valvetrain was designed so that it could be adjusted with a dolly lever so the angle of the valves and pushrods would not have to be changed according to whichever cylinder of an actual engine was installed, and the cams and dolly levers could be made to match each engine. Crankshafts were made in several different strokes with a shared diameter. The auxiliary shaft had a large amount of attachment points so that many engine accessories could be installed.

    About 20 units were manufactured in total, with examples made for the Homare, Kinsei, Amakaze, and Kamikaze engines. These single-cylinder testers were distributed to fuel depots and the Central Aviation Research Institute. Later models were being developed for water-cooled engines and prototype engines, but were not completed due to the deteriorating war situation.

    Power Enhancement Tests

    For tests on enhancing the power of engines, the Kinsei and Homare units were chiefly employed. Various factors would be adjusted individually to increase power output such as the compression ratio, RPM, boost pressure, air/fuel mixture, ignition timing, and type of fuel. The results of testing were plotted on a graph with power as the horizontal axis and cylinder temperature as the vertical axis. The objective was to increase power with as little rise in cylinder temperature as possible, so a shallow curve was desired. Generally speaking, the most promising results were achieved by increasing the RPM or boost pressure.

    With the Homare test unit, a two-stage roots supercharger with a supercharging capacity of up to +1500 mmHg boost (2.974 atm, 88.98 inHg) was attached, and tests were conducted using highly detonation-resistant fuel consisting of iso-octane mixed with 0.15% tetraethyl lead, benzol, toluol, and other aromatics, in conjunction with water-methanol injection. It was confirmed that the Homare could withstand being supercharged up to about +800 mmHg (2.053 atm, 61.42 inHg). This is the mechanical limitation of the engine.

    As designed, the nominal operating boost pressure of the Homare Model 21 / Ha-45 engine was +350 mmHg (1.461 atm, 43.70 inHg), or +500 mmHg (1.658 atm, 49.61 inHg) at takeoff/emergency operation.


    Source

    • 日本航空学術史 1910-1945. (2021). 三樹書房.
  • About the Performance of Type 4 Fighter (Ki-84) ー Pt.2

    About the Performance of Type 4 Fighter (Ki-84) ー Pt.2

    This is an update to the previous article about the performance of the Type 4 Fighter (Ki-84), addressing some errors and new information that I have found since writing it.

    It is not so much an article as a set of separate points.

    The original article is here:



    Ha-45 Special is Ha-45-11, not Ha-45-12

    The engine fitted to the first three prototypes of the Ki-84 was designated the “Ha-45 Special” by the Japanese Army. Despite the name, it is in fact an inferior engine to the “Ha-45” (Ha-45-21). The “Ha-45 Special” is a Ha-45-10 series engine.

    In the first article, I wrote that the “Ha-45 Special” was the Ha-45-12. In fact, it seems that the Ha-45 Special was actually the Ha-45-11. This is stated in the document「発動機名称一覧表」[Engine Designation List] created by the Navy Aviation Headquarters on April 11, 1945.

    This is further supported by the document “Specifications of Special Army Planes” created by the Army Aviation Examination Dept on August 1, 1943. Here, the projected performance of Ki-84 is written to be 660km/h at 5700 meters, which is the critical altitude of the Ha-45-11. This document was captured and translated by the United States, and the relevant page is pictured below.

    This is an important distinction, because the Ha-45-11 and Ha-45-12 have different altitude performance. The supercharger’s first and second speeds were stepped up 5.47⇒5.81 and 7.49⇒7.95, respectively, in the Ha-45-12. This raised the second speed critical altitude from 5700m to 6550m. Below is an excerpt from the Homare manual showing the specifications of the Model 11, 12, 21, and 22, as of December 1943.

    It appears to check out that the initial prototypes were equipped with the Ha-45-11. The official performance of the Ki-84, which originates from a test with one of these prototypes, shows that the top speed was achieved at 6550 meters. This indicates that the critical altitude of the engine was increased by perhaps 850 meters due to ram pressure at the Ki-84’s top speed. It would make less sense if this plane was using a Ha-45-12, because that would indicate that there was no increase in the critical altitude from ram pressure.

    Furthermore, we can see in the climb test obtained from the same aircraft that the engine began to lose manifold pressure just before 6000 meters when flying at low speed.


    Ki-84 Prototype vs. Service Plane

    The performance achieved by a Ki-84 prototype is considered to be the “official performance” of the Ki-84, and is written in its piloting manual. However, knowing that the Ki-84 prototype probably used a Ha-45-11 engine, we would expect the performance of a service plane to actually be somewhat different.

    Mass production examples of the Ki-84 that were put into service used a Ha-45-21 engine (which the Japanese Army simply called Ha-45, or Type 4 1850 HP Engine). The Ha-45-21 was restricted to prevent engine failures, as is well known. The RPM was limited to 2900, and the manifold pressure to +250mm, just like that of the 10 series engines. With this, the critical altitude rose to about 6500m, which is essentially the same as the Ha-45-12.

    Because the Ha-45-21 in its de-rated condition has (on paper) a critical altitude 800 meters higher than the Ha-45-11, as well as 60 more horsepower at critical altitude, we would expect a service Ki-84 to actually have a slightly higher top speed than the 624km/h of the Ki-84 prototype if it was operating in ideal condition.


    Every Ki-84 Had Exhaust Thrust

    This is not an error with the previous article but an elaboration. It is sometimes said that the early Ki-84s did not generate exhaust thrust. All units of the Ki-84 had exhaust pipes which produced thrust.

    The initial prototypes and early pre-production units had a thrust-generating collective exhaust. Later, it was changed to individual thrust-generating exhaust pipes. We can be sure that the thrust force was improved to some extent, but it will not be as marked as the change from standard exhaust pipes to thrust-generating individual exhaust pipes.

    For reference, the change from standard (non-thrust generating) exhaust to individual thrust-producing exhaust pipes netted the A6M5 about 10 kts (19km/h) of top speed, with its Sakae 21 engine that provided 980 HP at its top-speed altitude.


    The Full-Rated Performance Gap

    With a fully-rated Ha-45-21 engine, there exists a record showing a top speed of 634km/h for the Ki-84, which is referenced in the first article.

    The idea that the Ki-84 only gained about 10km/h of top speed (624km/h ⇒ 634km/h) when using a “Ha-45” (Ha-45-21) engine seems difficult to believe to some. The fact that the prototype which achieved the lower speed probably had a Ha-45-11, rather than Ha-45-12 engine, makes this performance gap even more egregious.

    It is relatively well accepted that the Ki-84’s propeller was inadequately sized for its class of engine, which greatly hindered its ability to make use of greater power. The Ha-45-21 had an increase in output of at least 120 horsepower compared to the Ha-45-12 their critical altitudes. However, the critical altitude of the 21 is about 400 meters lower than that of the 12 (ultimately, it is thought that the critical altitude of the fully-rated 21 was perhaps just 6100 m).

    On the other hand, the Ha-45-11 not only has even less power (1440 HP at 2nd speed) than the Ha-45-12, but its critical altitude is lower than even the Ha-45-21, as previously noted. This means that the Ki-84 with a fully-rated Ha-45-21 engine had not only greater power, but at a greater altitude, and still only marked 634km/h.


    Additional Evidence for Full-Rated Top Speed

    The origin and condition of the fully-rated Ki-84’s 634km/h top speed record has been questioned, and though it is said to be a translated document, there is little contextual information about it.

    However, there seems to be additional evidence supporting this fully-rated top speed. This comes from the document「陸軍航空技術沿革史」[History of Army Aviation Technical Development] which was published by the 1st Demobilization Bureau in May, 1947. Though it is post-war, this document was written by former staff of the Japanese Army. Within this document is the table「陸軍制式飛行機諸元表 (A)」[Army Service Airplanes Specification Table (A)].

    This table lists performance for both the “Type 4 Fighter (Ki-84I)” and the “Type 4 Fighter Performance Improved“. The engine column shows that the former represents the Ki-84 with a de-rated engine, while the latter represents a plane with a fully-rated engine (referred to here as Type 4 1850HP Kai). These differing names are purely informal.

    The top speed listed with the fully-rated Ki-84 is 635km/h at 6400m. The climb time is 5’50” to 5000m.

    This compares closely with the values from the other fully-rated test mentioned in the last section, which showed a top speed of 634km/h at 6650m, and a climb time of 5’37” to 5000m.

    I believe that 635km/h is a reasonable value for the “true performance” of the fully-rated Ki-84, on rated power. Use of “takeoff power” (essentially war emergency power, WEP) at altitude may have yielded a slightly higher top speed, but not markedly, as the critical altitude would be around 1000 meters lower.

    If the Ha-45 engine was able to be produced with a compression ratio of 8.0, as was originally planned, the performance would have been greater, but this was not the case.

    (The reduction of the Ha-45-21 CR from 8.0 to 7.2 is most likely the explanation for its fully-rated 2nd-speed horsepower output being reduced from the ballpark of the 1700s to the 1600s).


    Expected vs. Achieved Performance

    According to the “Famous Airplanes of the World” issue on the Type 4 Fighter, the required top speed issued for the Ki-84 was 680km/h, and the climb time was 4’30” to 5000m.

    On the other hand, the “Specifications of Special Army Planes” document shows what seems to be a calculated or required top speed performance of 660km/h, along with the same climb time of 4’30” to 5000m.

    Was the requirement reduced to 660km/h? Did Nakajima calculate 660km/h? Or was 660km/h simply the expected performance when equipped with the weaker 10-series engine? The latter would seem to be the obvious guess, but it’s not clear because the climb time remains the same.

    If we interpret the expected performance of Ki-84 to be 660km/h, the actual plane seems to have fallen short by about 25km/h. The climb time, more clearly, was more than 1 minute slower than the requirement to 5000m.


    This last section is to be taken less seriously, and is not anything more than an idea. An awfully inadequate propeller could be the sole explanation for the lack of performance in the fully-rated Ki-84.

    Anyway, the first prototype of the Ki-84 was initially completed with a smaller 19m2 main wing according to numerous secondary sources, such as the “Famous Airplanes of the World” issue. All subsequent planes had a larger 21m2 main wing, and it seems that the first unit was later modified to use the final wing.

    I have not located a source which specifies which of the initial 3 Ki-84 prototypes achieved the 624km/h at 6550m top speed. If we consider the possibility that it could have been the first prototype with its original, smaller wing, it makes sense that the performance gap between this record and the fully-rated, longer-wing Ki-84 would be more negligible in terms of top speed.

    It seems unlikely that the 19m2 wing unit recorded this test, however, because this performance was later used as the “official performance” of the Ki-84 in general.


    Sources

    • Famous Airplanes of the World No. 19: Army Type 4 Fighter ‘Hayate’. Tōkyō: Bunrindo, 2019.
    • Rep. CINCPAC-CINCPOA Translations & Interrogations No. 8. 1943.
    • Rep. 誉発動機取扱説明書 [Homare Engine Operating Manual]. 1943.
    • Rep. 「キ八十四」操縦法 [Ki-84 Piloting Manual]. 1944.
    • Rep. 発動機名称一覧表 [Engine Designation List]. 1945.
    • Rep. ADVATIS Translation 92: Specifications and Performance Data of Ki-84. 1945.
    • Rep. 陸軍航空技術沿革史 [History of Army Aviation Technical Development]. 1947.
  • What Was the A6M4?

    What Was the A6M4?

    The A6M4 is an “unknown” variant of the Zero Fighter that has been described as a variety of things over the years. The most common theory in English writing is that A6M4 was a designation for a type of Zero fitted with a turbocharger to its Sakae engine. Another common theory is that the number was skipped to avoid using the unlucky number “4”, which can be pronounced the same way as “death” (shi) in Japanese language.

    A handful of original Japanese documents exist which can be used to paint a vague outline of what the A6M4 actually was. This article serves as a summary of the various wartime references to an “A6M4”, or otherwise a “Model 40” Zero Fighter, that are known to me.


    A6M4 as a Turbocharged Zero Fighter

    In the February 1942 Arsenal Gazette of the Navy Aviation Technical Arsenal (Kūgishō), a report dated February 4th outlined a research meeting that was to take place on the 9th of the same month at 15:00. The subject of the meeting was a structural examination of a partially wooden model of a Sakae Model 10 engine fitted with a turbocharger. The purpose was for the eventual installation of a turbocharger in the Zero Fighter.

    The next reference in the Arsenal Gazette was dated February 10th, and it outlined the schedule of the first research meeting concerning the fitting of the Zero Fighter with an Ishikawajima turbocharger. This meeting was to take place on the 19th of the same month at 13:00.

    It should be noted that these brief meeting schedules did not state which model of Zero Fighter was to be equipped with the turbocharged engine, nor was a designation for this prospective variant given. However, at this time, the engine under examination was a Sakae Model 10-series, which is of the same series as the engine installed in the A6M2.

    The next reference to the turbocharged Zero Fighter is more well-known: an English translation of a Japanese document that was captured on Saipan, titled “Quarterly Report on Research Experiments,” and dated October 1, 1942. This is where the first mention of the “A6M4” as a turbocharged Zero variant is known to appear, and the relevant text is in the images below.

    Assuming that the translator did not make a typing error (which is more common in translated documents than you may hope), at this time, the “A6M4” was a designation for a development of the A6M3 with an intercooler — almost certainly indicating that it would have a turbocharged engine. The next step in development was for wind tunnel testing to be carried out, according to the record.

    We can be fairly sure that this “A6M4” was a development of the A6M3, because it is written under the general section concerning the “Type 0 Mark 2 Carrier Fighter”, which was an earlier designation of the Zerosen Model 32 (A6M3).

    On the other hand, Francillon wrote the following entry in his 1967 title “The Mitsubishi A6M3 Zero-Sen (Hamp).”

    Francillon stated that Jirō Horikoshi (the head designer of the Zero Fighter) personally informed him that the A6M4 designation referred to two A6M2s that were fitted with a turbocharger in 1943. Unsurprisingly given the situation of turbocharger implementation in Japan at that time, this testing was described as a failure.

    Whether it is the case that the turbocharged Zero Fighters were converted from the A6M2 or A6M3, there is reasonable evidence here to state that the “A6M4” designation was most likely applied to the turbocharged Zero Fighter project by late 1942. However, it seems that upon the failure and abandonment of these prototypes, the designation was later re-used for other projects.


    Turbocharged Sakae Engine

    A Sakae Model 11 engine with a wooden mockup of a turbocharger installation attached to it is pictured below. This is likely to be the same model that was examined at the Kūgishō in early 1942. It can be observed that the compact installation model did not actually include an intercooler at this time.

    The turbocharger represented by this model was the Ishikawajima Model IET4. The Model IET4 was designed to maintain the full pressurization of the Sakae engine up to an altitude of 7000 meters.


    Fake A6M4

    The image on the left below is supposed to show an A6M2 with a turbocharger. In fact, to this day it is a top image result when searching online for “A6M4” or “Zero with turbocharger”. However, it’s just a photoshop of an image of a standard A6M2, compositing a picture of the turbocharger from the Army prototype fighter Ki-87.


    A6M4 (?) as an A6M2 with Belt-Fed Guns

    Another document possibly related to the “A6M4,” titled “Matters Pending Approval Regarding Airplane Remodeling Experiments,” was created on April 28th, 1943. This document shows various tentative decisions regarding aircraft models in development, including future variants of the Zero fighter.

    The following relevant text is quoted:

    1. Type 0 Fighter
      • (a) Zerosen Model 21
        • Changed 20mm fixed machine gun ammo capacity from 100 rounds per gun to 150 rounds per gun (belt feed).
          Provisionally designated as Zerosen Model 41 and ordered to Nakajima.
      • (b) Zerosen Model 22
        • (1) Changed 20mm fixed machine gun ammo capacity from 100 rounds per gun to 150 rounds per gun (belt feed).
          (2) Abolished wingtip folding mechanism and shorten wingspan by about 1 meter.
          Provisionally designated as Zerosen Model 52 and ordered to Mitsubishi.

    Therefore, we can say that as of April 28, 1943, it had been tentatively decided to give the designation “Zerosen Model 41” to a Zerosen Model 21 (A6M2) fitted with belt-fed 20mm machine guns. The designation “Zerosen Model 52” was to be tentatively provided to a Zerosen Model 22 (long-wing A6M3) with belt-fed 20mm machine guns and shortened wings.

    In the Japanese Navy’s aircraft naming nomenclature at the time, the first numeral of the model number represented airframe modifications, while the second number represented engine modifications. According to this system, the reasoning behind these provisional designations is briefly as follows:

    • Model 21 + 1 airframe modification (belt-fed guns) ⇒ Model 41
      (3_ is skipped because it is occupied by the shortened square wing modification of Model 32).
    • Model 22 + 2 airframe modifications (belt-fed guns, shortened round wingtips) Model 52
      (3_ is skipped for the same reason as prior, because Model 32 had shortened wings with square wingtips, while Model 52 had round wingtips).

    No code names are listed in this document, and just because it was a “Model 4_” Zero Fighter does not mean that it would necessarily be designated A6M4.

    Furthermore, there is no known information which would suggest that the Model 41 was ever built. The Model 52, of course, would go on to be mass produced.

    However, when the Model 52 was actually adopted into service on August 23rd, 1943, it was recorded that its prior tentative designation had still been the “Type 0 Ship-Based Fighter Model 22 Kai”, as can be seen in the following document (which also shows the adoption of the Gekkō Model 11).

    Type 0 Ship-Based Fighter Model 22 Kai (shortened wingtips of main wing) is adopted as a weapon and designated as Type 0 Ship-Based Fighter Model 52.


    A6M4 as an Early Name for A6M5

    To quickly recap from the previous section, this is what had been tentatively decided as of April 1943:

    • Model 2X: 12 meter wingspan with folding wingtips.
    • Model 3X: 11 meter wingspan with square wingtips, no folding mechanism.
    • Model 4X: 12 meter wingspan (Model 2X airframe) with belt-fed 20mm MGs.
    • Model 5X: 11 meter wingspan with round wingtips, no folding mechanism, and belt-fed 20mm MGs.

    In reality, the belt-fed 20mm MG (that is, the Type 99 20mm Mark 2 Fixed Machine Gun Model 4) was not actually ready in time for the mass production of the Model 52.

    So while the Model 5X had been defined as having 2 modifications over the Model 2X, belt-fed 20mm guns (which is what brought it to 4X) and clipped, rounded wingtips, the actual first production Model 52 or A6M5 only had the latter modification.

    This is important to consider when looking at the next and final document regarding the A6M4, which is a translated document about aircraft armament, captured on Peleliu. The data comes from a Japanese notebook and was probably created in mid-1944. Here the “Model 42” is listed, with the code name “A6M4” specified.

    Above the Model 42 is two Model 52 (A6M5) with differing armaments. The top Model 52 would later be known as the Model 52 Otsu, or A6M5b, and the lower Model 52 is the standard initial production model without belt-fed MGs – its designation would not change.

    Relevant trivia: the designation system that introduced the ability to define minor “subvariants” of Navy aircraft with “Kō, Otsu, Hei…” was only introduced in November 1944. Before this, all Model 52 armament varieties were simply “Model 52”.

    As we can see in the document above, the A6M4 has the same armament as the initial A6M5: the Type 99 20mm Mark 2 Fixed MG Model 3, which is not belt-fed. Unfortunately this table is focused purely on armament, so there are no other details to compare.

    So, what is the A6M4?

    Considering that this Model 42 clearly does not have belt-fed MGs, we can assume that the concept of the “Model 4X” constituting “a Model 2X airframe with belt-fed MGs” had been abandoned at this time. So we may forget that the “Model 41” was ever proposed, which leaves an empty space for an airframe modification in the Zero Fighter’s designation list.

    The Model 52 was originally defined as having both belt-fed guns and the clipped, rounded wingtips, but was at first produced with only the latter modification. When changing to belt-fed guns justified an increase in airframe model number in the first place, logically, losing the belt-fed guns would seem to justify regressing the model number by one. As the Model 41 was abandoned without being constructed, there is no conflict.

    Therefore my theory is that the Model 42 (A6M4) is the initial Model 52 (A6M5). In my assumption, it was most likely a tentative designation, and was renamed as “A6M5” to avoid unnecessary confusion.


    CONCLUSION

    In consideration of all of the above, I would say that the “A6M4” is:

    A provisional designation that was never officially adopted, and that was used by at least two models of the Zero Fighter at different points in time.

    The sparse available evidence suggests that “A6M4” once referred to both the turbocharged Zero, and later the initial model of the A6M5, but neither was ever set in stone.

  • Identifying Japanese Jets Captured by the US in 1945

    Identifying Japanese Jets Captured by the US in 1945

    Following the surrender of Japan in August 1945, the American occupational authorities sought to gather whatever aircraft developments were of potential interest. Naturally, the most cutting edge planes and engines developed in Japan were high on this list. More than a few of these projects had already been destroyed by Japanese orders immediately issued to prevent that from occurring.

    None of the most powerful Japanese turbojets, Ne-130, Ne-230, or Ne-330, were left in Japan to recover. These prototype axial-flow jet engines were to be comparable in performance to the late German models, but each had been destroyed or hidden, one way or another, by September. Even the well known ‘Ne-20’, the turbojet of the Kikka, had risked destruction. The units under the Navy’s direct jurisdiction were sabotaged on such orders; luckily, a few survived at other organizations.

    A couple of photos taken on October 16, 1945, show us some of the more obscure Japanese engines that were seized by the US. These photos display a group largely consisting of prototype jet engines awaiting preparation for shipment to the US by the 7th Air Service Area Command of the USAAF. The purpose of this article is to identify each engine shown in these photos and their subsequent fate.

    Photos

    Almost every engine visible and identifiable in this roundup was designed, if not manufactured, at the Navy 1st Air Technical Arsenal (Kūgishō). The Kūgishō was a center of Japanese jet engine development until the end of the war. Here, pioneering efforts led by Tokiyasu Tanegashima from the year 1941 resulted in a variety of test engines. Only in mid-1944 did the jet engine receive appropriate attention from the upper brass of the Navy, after which increased funding and restless development by Tanegashima’s group managed to yield the successful flight of Kikka with its twin Ne-20 turbojets just before the end of the war.


    Engine Identifications – Left Side

    ‘Sakae IPR’ Blower

    First on this list is a very obscure, almost unknown jet engine project. No doubt inspired by the engine of the Caproni Campini N.1 which flew in 1940, the ‘Sakae IPR’ was a motorjet using a Sakae Mod.11 piston engine to drive a five-stage axial blower, which was followed by a burner. ‘IPR’ stood for “Internal Propeller Rocket”.

    It was one of the very first Japanese air-breathing jet engines, built during 1943. This engine produced 604 kgf of thrust, and had large dimensions of 4.17 m length by .91 m diameter. It was designed and probably manufactured at the Navy’s Kūgishō.

    Only the blower seems to have been recovered at the time of this photo. Though the Sakae IPR blower was possibly transported to the USA, it was likely of little interest, and no piece of it is known to survive today.


    Ne-12

    Next we have the Ne-12B, the last model in the original series of Japanese turbojets. This was a follow-on design to the original centrifugal ‘Ne-10 series’ turbojet. The four-stage axial compressor, clearly visible in this photo, leads into the main centrifugal compressor.

    The Ne-10 Kai was the first model to implement this axial compressor, followed by the Ne-12 which featured reinforcements to improve durability and features for fixing to an aircraft. The final Ne-12B was a model which lightened the weight of the Ne-12 by as much as 70 kg. It is most likely that the engine in this photo is in fact the Ne-12B based on the date, but the visual differences from Ne-12 are unknown. All of these engines were designed at the Kūgishō.

    The Ne-12B had a weight of 315 kg and dimensions of 1.80 m length by .86 m diameter. It ran at 15,000 rpm and aimed to produce 320 kgf of thrust. This engine only was able to demonstrate a lifespan of perhaps 1 hour by the end of the war, and had been abandoned in April 1945 after a production run of 12 units, split between the Kūgishō and the Yokosuka Arsenal.

    This is, to my knowledge, the only known photo of the Ne-12/B. The engine did not survive to this day, but the 1st stage of a Ne-12B axial compressor is displayed at the National Museum of Nature and Science in Tokyo.


    YE3B

    Behind the Ne-12 is not a jet, but the YE3B piston engine, a very unusual design. This is a 24-cylinder liquid-cooled engine of the X arrangement, which aimed to produce 2,500 horsepower. It had been abandoned by the end of the war per US intel in favor of the YE3E, a 3,200 horsepower development of the same engine.

    Based on very little documented data, the YE3B had cylinders with 145 mm bore and 160 mm stroke. The total displacement was 63.4 liters. It was designed and built by the Kūgishō. In this photo, the engine faces with the back side towards the camera, featuring the supercharger. The upper left row of cylinders and exhaust pipes are visible.

    The engine survives today in storage under the ownership of the Smithsonian. Dimensional data of 2.29 m length by 1.75 m width is provided, which contrasts with the (likely erroneous) documented US data of 1.10 m length.


    Tsu-11 / Hatsukaze Rocket

    The small piston engine here is actually part of a jet. This is the ‘Hatsukaze’ engine portion of the ‘Hatsukaze Rocket’, a Japanese informal name of the Tsu-11 motorjet. Only the Hatsukaze itself is visible in this photo, but based on later photos, it can be confirmed that this is a Tsu-11 setup.

    The Tsu-11 consisted of a Hatsukaze piston engine driving at 3,000 rpm, which was stepped up to 9,000 rpm to rotate a single-stage axial fan, followed by a burner. The overall setup weighed 200 kg and produced a very modest 220 kgf of thrust. The dimensions were 2.20 meters length by .64 meters width. Tsu-11 was designed by the Kūgishō but produced at Hitachi Aircraft.

    This engine was only ever fitted to the Ōka Model 22 piloted missile, and the ‘Ginga’ bomber as a test auxiliary power unit. It had been contemplated as a temporary engine for the Kikka (in a quad installation of two per wing) if the main turbojets were not ready in time. The performance was very poor, the engine could not be started in the air, and would spontaneously seize at altitudes higher than around 4,000 meters. It did, however, provide the Ōka with a better standoff range than the previously used rocket engines. This would have improved the survive-ability of the mother plane, but perhaps not the missile itself.

    A Tsu-11 survives today, installed in the sole Ōka Model 22 preserved at the Smithsonian National Air and Space Museum.


    Unidentified

    The last two clear objects in this photo I have not identified, although it should be possible. These are a radial engine and an unknown engine to the right of it with an exposed propeller fitting, sitting behind the IPR blower and the Ne-12. If you can help identify these, you could leave a comment on this post.

    The rest of the engines to the lower right of the whole photo are not included in this section, as they are more clearly visible in the right-side photograph.


    Engine Identifications – Right Side

    Ka-10 / Maru-Ka, & Small Model

    Here is not only the Ka-10 pulsejet (also named ‘Maru-Ka’ ㋕), but almost hidden behind it, its smaller initial prototype version. These were the only pulsejets built by Japan during World War II. They were directly based on the German As 014 pulsejet used to power the V1 flying bomb.

    The full-sized Ka-10 had dimensions of 3.70 meters length by .58 meters width, weighed 150 kilograms and was designed to produce 300 kilogram-force of thrust. The specifications of the smaller test version are unknown.

    According to Japanese records, the small test model was completed in early 1945 and tested until June; the full-sized version followed it at the end of July, and remained under testing when the war ended in August. According to a member of the Army special weapons team, five units were built. This engine was to power the ‘Baika’: an ultra low-cost, manned flying bomb designed by Kawanishi Aircraft, which was similar in concept to the Fieseler Fi 103R ‘Reichenberg’, though somewhat more sophisticated in airframe design. However, the Baika had only lapsed one month of design progress when the war ended.

    These are the only known photos of the Maru-Ka, which does not survive to this day.


    KR10

    To the right of the Maru-Ka is the rather well known KR10 liquid-rocket engine, which powered the Shūsui, Japan’s version of the Me 163B Komet rocket interceptor. Technically speaking, this could be the variant ‘KR20’, or ‘KR22’, which differed by thickening the turbopump shaft or increasing its structural support respectively. It is impossible to determine from this photo, but all versions of the engine are typically referred to as ‘KR10’ informally.

    KR10 had dimensions of 2.52 meters length by .90 meters width, and weighed 170 kg. It produced 1500 kilogram-force of thrust, identical to the initial model of the German Walther HWK 109-509. The liquid fuel used consisted of the ‘Kō’ and ‘Otsu’ liquids, analogous to the ‘T-Stoff’ and ‘C-Stoff’ used in Germany.

    The engine was designed by the Kūgishō with contribution from Mitsubishi, and was manufactured at the Kūgishō and various naval arsenals. The engine ‘KR22’ made by Hiro Naval Arsenal was the unit actually fitted to the Shūsui which flew on July 7th, 1945. This flight met with failure and death of the pilot due to the layout of the fuel system, which failed to feed with a reduced fuel load in a steep climb angle.


    Ne-20

    The Ne-20 turbojet is the most famous Japanese jet engine from World War II. It is typically referred to as “the first Japanese jet engine”; though it was not the first built by any measure, it is true in the sense that it was the first successful unit.

    This engine was designed at the Kūgishō under the leadership of Osamu Nagano. Ne-20, based on the BMW 003A format, had an incredibly rapid developmental pace – advancing from merely a concept to an initial prototype in only three months, and passing trials in another three. Due to this impressive feat, the special attack plane ‘Kikka’ was able to successfully fly on August 7th, 8 days before the end of the war.

    Ne-20 had dimensions of 2.70 meters length by .62 meters diameter. It weighed 470 kilograms, rotated at a maximum of 11,000 rpm, and produced thrust from 475 to 490 kilogram-force. The prototypes and first production engines were built at the Kūgishō, with additional production units being made at the Yokosuka Naval Arsenal. Around 20 examples are known to have been completed in total.

    The engine in this photo appears to be marked ’19’. Perhaps this was the 19th engine, one of the production examples built at the Yokosuka Naval Arsenal. Some of these engines had been rejected due to poor workmanship related to the lack of experience building jets at the naval yard.

    Three Ne-20 turbojets survive to this day: two at the Smithsonian National Air and Space Museum (one on display), and the other example at the Ishikawajima-Harima company museum.


    ‘Ne-201’ or ‘GTPR’ Turbine-Nozzle Mockup

    The Ne-201 and the GTPR are practically unknown engines, especially in English sources. These were both turboprops, designed by the Army and Navy respectively from about the same time (~1942).

    Ne-201 was designed by the Kogiken (Army Aero Tech Lab) and Kōken (Tokyo Imperial Uni Aero Dept), manufactured by Ishikawajima Shibaura Turbine. GTPR was designed at the Kūgishō, to be manufactured by Ishikawajima Shibaura Turbine as well. GTPR stood for ‘Gas Turbine Propeller Rocket’.

    Both of these engines have been listed here due to the ambiguity of their history. I can state with certainty that the Ne-201 and GTPR were, at the outset, independent projects. However, a few accounts from first-hand suggest that they were the same thing, and a US report identifies this mockup as the ‘GTPR’ component, even though it almost exactly matches a known ‘Ne-201’ design drawing.

    Currently, I’ve theorized that at the time of August 1944, when jet development between the Army and Navy was unified, the more developed turboprop project was probably taken (Army Ne-201), but placed under unified leadership. Thus, what was once just Ne-201 likely came to be known by either designation, and developed a bit further until the end of the war. This is only an assumption.

    The Ne-201 had been built in 1944, and the original GTPR was ordered but never completed. Both designs were to be converted to turbojets in 1944, as priority was placed upon that type of engine, but ultimately Ishikawajima Shibaura Turbine created the turbojet ‘Ne-130’ from scratch. In December 1944 the Ne-201/GTPR damaged itself, by April 1945 it was ready for a re-test, but due to focus on the Ne-130 it received little attention until the end of the war.

    Ne-201 had dimensions of 5.75 meters length by 1.10 meters diameter, and weighed 2,500 kilograms. It rotated at 4,200 rpm and produced 862 kilogram-force of thrust (prop 1870 shp/280 kgf + 582 kgf exhaust thrust). An iteration of the GTPR had dimensions of 5.50 meters length by .85 meters diameter and weighed 2,500 kilograms. It rotated at 5,000 rpm and aimed to produce 5,000 equivalent horsepower.

    After the war, Tanegashima prepared a report on the GTPR for the US, although the details he provides are unlike either design. It is likely that as a personal passion project, he continued to work on the GTPR design aside more pressing matters until the end of the war, incorporating data learned from BMW 003A and Ne-20.

    This turboprop mockup did not survive to this day.


    Ne-30 & Ne-30 Mockup

    The Ne-30 was one engine in the initial ‘Ne-10 series’ of Navy turbojets. It was a unique departure as an attempt to gain high thrust from the relatively low-performance engine design by upscaling it in size. Effectively, this was a larger Ne-12, with the same features. The left engine installed on a stand is the actual prototype, while the object to the right is the mockup. “Mock of TR30” is possibly written on the side.

    The Ne-30 had dimensions of 2.47 meters length by 1.03 meters width, and weighed 900 kilograms. It rotated at 15,000 rpm and aimed to produce 850 kilogram-force of thrust. It was designed by the Kūgishō and built there in November 1944. However, the Ne-30 never demonstrated its intended performance and was abandoned, as with the other Ne-10 series engines. This engine had been contemplated as the original engine of the R2Y2, Keiun Kai, and the Tenga, a jet-version of the Ginga bomber.

    Both the engine and its mockup were brought to the USA and still remain in the storage of the Smithsonian today.


    Ne-10 & Ne-10 Exhaust Nozzle

    This engine on the right seems to be the Ne-10, the first functional Japanese turbojet. This can be deduced by the apparent lack of axial compressor stages extending from the front side (which faces away from the camera). The visible side is the turbine at the rear.

    The design of the entire Ne-10 series was, put simply, a huge turbocharger converted to a turbojet by installing a folded combustion chamber. First built in mid-1943 as the ‘TR’ (Turbine Rocket), the design was renamed ‘TR10’ in 1944 and prepared for mass production to perform trial-and-error testing. By the end of 1944, it had again been renamed as the ‘Ne-10’ due to unifying development with the Army, which created shared nomenclature.

    On the left is presumably the exhaust nozzle to produce thrust from the Ne-10. You can see the attachment points both on the edges of the nozzle and the engine, circling the turbine. An early problem with the Ne-10 had been the nozzle warping into the turbine under heat due to a lack of resistant materials.

    Ne-10 had dimensions of 1.60 meters length by .85 meters diameter. It rotated at 16,000 rpm and was designed to produce 300 kilogram-force of thrust. It was designed by Tanegashima’s group at the Kūgishō and only a handful were built, perhaps less than ten. It is somewhat surprising that an original Ne-10 survived to the end of the war, as these engines had a very short lifespan, and a tendency to fail disastrously.

    No example of the Ne-10 survived to this day, nor any known components.


    YE2H

    Lastly, this engine is rather hard to spot. Behind the Ne-10 nozzle is the YE2H prototype – an 18-cylinder, liquid-cooled piston engine of the W-layout. Luckily, a view from the same side of the surviving engine provided by the Smithsonian shows identical features which can be compared to this image.

    YE2H has dimensions of 2.46 meters length by 1.12 meters width, and weighs about 1,200 kilograms. The cylinders had the same 145 mm bores and 160 mm stroke as the YE3B shown prior, with a 47.5 liter total displacement. YE2H was to produce 2,500 horsepower. It was designed and built by the Kūgishō, and undergoing a breakdown test there when the war ended.

    As mentioned, the YE2H survives today in the storage of the Smithsonian.


    Unidentified

    The only object I cannot readily identify in the right-side view is this large jet exhaust nozzle. It is clearly considerably wider than the exhaust nozzle of the Ne-20 in the foreground, and even the centrifugal Ne-10. Unlike known exhaust sections from larger Japanese jets, the exit cone protrudes far from the end of the nozzle. It does not appear to be the exhaust nozzle for the Ne-30, nor Ne-130, or Ne-330.

    I have speculated that this could be the turbine and exhaust section of the Ne-140, which was the huge turbojet converted from the GTPR turboprop design. However, it is unlikely that any part of the Ne-140 was built before development was apparently terminated in late 1944 or early 1945. The only source (of few overall) that contradicts this is Senshi Sōsho 87, which suggests that Ne-140 was tested by the end of the war, although I suspect that this is a mistake.

    Alternatively, it could be possible that this is the turbine and exhaust nozzle of the MTPR, a compound “engine-turbojet” consisting of an Atsuta (DB601) piston engine which drove a prop, and also transferred some power to the compressor of a turbojet linked to it. According to limited information, MTPR was under construction from 1943 before being canceled in mid-1944.


    *July 9th 2023: Correction on completion/test dates of Maru-Ka

  • Ōka Trainers: MXY7-K1, MXY7-K2, & Wakazakura

    Ōka Trainers: MXY7-K1, MXY7-K2, & Wakazakura

    This article is an attempt to clarify the details of the different training aircraft that were developed for the Japanese Navy’s late-war special attacker ‘Ōka’. The designations and purposes of the Ōka trainers are often confused, not only in English but even in Japanese publications. Using a few historical materials, we can correctly identify them and better understand their true details.

    Although the focus of this article is on the Ōka trainers, a brief about the Ōka Model 11, the main mass production model, follows.


    The Ōka Model 11 was a manned missile for attacking naval vessels developed by the Japanese Navy in August 1944 under the dire situation of the late war period. The project first received the secret designation ‘Maru-Dai’ (A circle or ‘maru’ around the kanji ‘dai’「大」), and although it was atypical to give Navy code-names to suicide aircraft, it was also designated ‘MXY7’ due to being created by the Navy 1st Air Technical Arsenal (Kūgishō). The Navy formal name ‘Ōka’「櫻花」(Cherry Blossom) was granted for service.

    Ōka Model 11 in front of a pile of other dilapidated Ōka, warheads, and bombs.

    This plane consisted of a tiny 6.06-meter-long by 5.12-meter-wide airframe with mid wings and a twin tail. It was constructed from duralumin, steel, and wood to conserve resources, and was designed to resist speeds up to 1000 km/h. Its sole armament was a 1.2 ton semi armour piercing warhead in the nose delivered by a ramming suicide attack, and the power plant consisted of three Type 4 Mk.1 Rocket providing 800 kg of thrust each with burn times of about 9 seconds, which were contained in the rear of the fuselage.

    Ōka Mod.11 could not take off by its own power and was brought up to the target area by a G4M2e attack plane. The pilot had only the basic instruments and controls necessary to arrive at the target, and the operational range was poor: as little as 20 km when dropped from 3500 m. Due to this latter fact, Ōka could not be deployed effectively, and the mother planes were often intercepted before reaching the drop point. Later developments were centered around extending the range of the Ōka by using alternative jet power plants, but the war ended before any could be utilized.

    The first examples of the Ōka were manufactured in September 1944, one month after the start of design. 755 examples of the Mod.11 had been constructed overall by March 1945 when mass production was terminated.


    MXY7-K1 ー Single Seat Ōka Trainer

    Naturally, the training variant of the Ōka was developed at the outset of the project, as unlike conventional aircraft, the manned missile could not take off or land in its operational configuration. This initial training model received the code name ‘MXY7-K1‘, ‘K’ being the Navy code for training aircraft, and ‘1’ denoting that it was the first of this type (a successive two-seat trainer was already planned).

    The MXY7-K1 had a few differences from the base aircraft in order to temper its flying characteristics for trainees. The wingspan was extended slightly by 12 centimeters, and the wings were equipped with flaps to decrease the landing speed. Inside the fuselage, which was extended by about 4 centimeters, the warhead & power plant areas were replaced by two water ballast tanks to maintain the proper weight and center of gravity. At the time of landing, these tanks were both dumped by the pilot to reduce weight and further decrease the landing speed. Even with the aforementioned measures, the landing speed was still a rather quick 203 km/h.

    MXY7-K1

    As for the means of landing, the MXY7-K1 was equipped with a central landing skid below the nose, similar to the method used with the Shūsui rocket fighter. The wings had a guard extending below each wingtip to stop the underside of the wing’s surface from being damaged as the plane leaned to one side and scraped against the ground after touching down.

    The first manned test flight of the Ōka was via an MXY7-K1 and took place on October 31st, 1944, with test pilot Kazutoshi Nagano in control. At the drop altitude of 3,500 meters, the G4M released the trainer. The K1 immediately fell sharply from its mother plane but began to glide as the airspeed increase generated more lift from its small wings. Nagano quickly ignited the twin wing-mounted powder rockets, but due to unequal thrust causing the plane to yaw, he released them from their mounts almost instantly. The rockets, still burning powder, flew ahead of his aircraft while spewing smoke, which attracted alarm from the observers on the ground until the plane continued to fly as normal.

    Nagano emptied the water ballasts as intended on the approach to the runway, and made a successful landing in front of the crowd of onlookers. Nagano had mainly praises for the aircraft, giving the opinions that the stability and control authority were perfect, that there were no problems with flight while emptying the ballasts, and that it could be used for training without issue. The wing-mounted rockets, however, did not function correctly due to unequal thrust, and were eventually abandoned.

    K1 trainers arrived at the 721st Naval Air Group (God Thunder Corps) at Kōnoike Air Base in November 1944, which would later become the first unit to operate the Ōka in combat. Here, the first landing-training test flight was conducted by Lieutenant Tsutomu Kariya on November 13th. The drop altitude this time was 3,000 meters, as the previous drop at 3,500 meters had initially frozen the ballast water. However, when Kariya began to dump the ballasts on his approach, the K1 immediately pitched sharply upwards, stalled, and fell from the sky. He could not recover flight, and crashed into the sand, flipping the K1 trainer end over end.

    Lieutenant Kariya was still conscious when recovered from the trainer, but he died just hours later.

    It was ascertained that Kariya’s crash was caused by pilot error: the front ballast was emptied before the rear one, the incorrect order, and so the accident occurred. But from this point onward, the water ballasts were no longer loaded during training. It was said that, in the official report, the true cause of the accident was likely the nose ballast leaking into the cockpit and blinding the pilot.

    Training with the K1 continued immediately and through to the end of the war, initially for the combat operations of the Ōka Model 11, and later for the expected deployment of the Ōka Model 22 (a development to extend the range by using a motorjet engine). By the end of the war, 86 MXY7-K1 trainers had been produced, and out of the few hundred trainees, two deaths and two injuries occurred.


    MXY7-K2 ー Two Seat Interim Trainer?

    The two-seat Ōka trainer is far less understood than its single-seat counterpart, and lots of misinformation floats around this aircraft. The names ‘K1 Kai’ and ‘Wakazakura’ are frequently used to designate this plane in English, but its actual name is ‘MXY7-K2‘. This can be verified by the original nameplate on the rear-left of the surviving example, which is under the ownership of the National Air and Space Museum.

    MXY7-K2. A two-seat Ōka trainer produced experimentally.

    MXY7-K2 is almost universally described by secondary sources (English and Japanese alike) as the trainer for the ‘Ōka Model 43 Otsu’. The Model 43 Otsu variant of the Ōka was much larger than the preceding models (8.16 meters long by 9.00 meters span), and operated completely independently by launching from land catapults. Utilizing a single Ne-20 turbojet engine for propulsion, it also had a far superior range. The Ōka Model 43 Otsu was expected to correct the problems with the previous models and become the primary special attacker for the final defense of the Japanese home islands. Coastal catapults were constructed around various expected areas of the US invasion fleet. But the war ended before a single Model 43 Otsu finished construction.

    The K2 does exhibit some specific features that would seem to imply that it was developed for this task. The wing span of K2 is ~7 meters, significantly larger than the ~5-meter wings of the Model 11 & K1. It could easily be assumed that this wingspan was chosen to emulate flight characteristics closer to the Model 43, which had a ~9-meter wingspan. Also, it is plausible to speculate that the two-seater layout was chosen for safely instructing trainees with the unfamiliar takeoff method of rocket catapulting. Lastly, it could be equipped with a single Type 4 Rocket in the rear for extending the glide range.

    Based on contemporary evidence, however, I would like to present the theory that the MXY7-K2 was only a two-seat trainer for the Ōka Model 11, and was not developed for the Model 43.


    To start off, when did the K2 originate? As the development of the Ōka Model 43 Otsu only began in March 1945, this would seem to be an easy point to immediately separate K2 from being a ‘Model 43 Trainer’. While it’s not totally clear from the materials available to me, it is certain that a two-seater Ōka was planned from essentially the very beginning of development:

    In September [1944], 9 dummy planes, 1 actual single-seater, 2 two-seaters, and 5 trainers will be manufactured.

    Results of a General Staff meeting on August 28th, 1944, quoted in ‘Senshi Sōsho 45’

    The precise date of when the K2 was actually completed is yet unknown, but based on this schedule, it appears that the construction of two double-seat Ōka was thought to be imminent just before September. It’s known that initial Ōka prototype constructions proceeded smoothly. This also coincides with the fact that only two examples of K2 are known to have been completed overall.

    Both MXY7-K2 trainers seized by US authorities.

    Before continuing with the historical analysis, there are also physical features of K2’s airframe relevant to this theory. There is clearly a loop for mounting K2 to a mother plane located between its two canopies. Such a feature would be unnecessary on a dedicated trainer for the Model 43, which only launched from catapults. Furthermore, the scale of the K2, though larger than K1, is not consistent with the giant Model 43 – at roughly 6.4 meters long by 7 meters wide, it’s almost 2 meters shorter in length and span. The fuselage in particular is clearly a direct adaption of the Model 11 design.

    The point where K2 seems to become related to the Ōka Model 43 is during June 1945. At this time, the design of Model 43 was already completed, and production plans were progressing. Starting on June 27th and lasting two days, the K2 was launched for a series of very successful flights using the rocket catapult for Mod.43 constructed on the shore of Takeyama. The pilot was Commander Hiromitsu Ito, and the observer seat was occupied by Ōka’s chief designer Tadanao Miki.

    “How about starting an aerial sight-seeing company with this plane after the war is over!”

    -Commander Ito quoted in ‘Thunder Gods’

    One fact that seems to be disregarded, however, is that in the recollections of this event, the K2 is described as “a two-seater Model 11 training plane”. This poses another question, though: If the K2 was truly built prior to the design of Model 43 as a ‘Model 11 trainer’, why would it have been designed with the capability to launch from Model 43 catapults, and utilized in these tests?

    The answer to this can be derived from the text of a slightly later document, Av HQ Aero Secret No. 5392 from July 24, 1945, which concerns the development of a two-seat trainer for the Ōka Model 43:

    In relation to Chiefs of Staff Aero Secret No. 823, conduct testing research after modifying to allow launching from the experimental rocket catapult, evaluate the practical two-seater, and obtain improvement data.

    Extract from ‘Av HQ Aero Secret No. 5392’, quoted in ‘Mysterious Ōka Model 43 Otsu Turbojet Special Attacker (First Part)’.

    The wording of this document is a bit vague, but it seems to state that the decision to develop a two-seater Mod.43 trainer only occurred in July 1945, after the catapult test of the MXY7-K2. Furthermore, this document concerns modifying a single-seat Ōka Model 43 trainer design to the newly decided twin-seat type, but also seems to state that an aircraft should be modified to allow catapult launching, to ‘evaluate the practical two-seater’. This likely just means to adjust the future Mod.43 trainer for launching as necessary — Regardless, it definitively separates the two-seater Mod.43 trainer as a later aircraft from K2.

    In summary, based on the existing evidence from the period, my theory is that ‘MXY7-K2’ was only a prototype two-seat trainer for the Ōka Model 11. Due to the rapid construction of Ōka Mod.43 rocket catapult sites in 1945 before aircraft could be completed, modifications were done to allow K2 to be catapulted from these sites for early evaluations. For this purpose it was ideal due to having two crew. As K2 was not the true Model 43 trainer, mass production did not proceed afterward. At the end of the war, the US recovered the sole two MXY7-K2 trainers at the Kūgishō, of which the most intact example was sent to the US and remains in the ownership of the Smithsonian NASM.


    Wakazakura ー Ōka Model 43 Otsu Catapult Trainer

    *Dec 13, 2023: Information on Wakazakura updated.

    Having tentatively concluded that the trainer for the Ōka Model 43 Otsu was not the MXY7-K2, let’s establish what the Mod.43 trainer actually was. In truth, there are almost no materials in my possession to define the Mod.43 trainer with, save for one primary document which coincides with the previous data, and largely is the reason I am confident in this theory. It’s the ‘Navy Prototype Planes Performance Chart’ from August 22, 1945, submitted by the Kūgishō to the US authorities following Japan’s surrender.

    At the bottom of this document, a two-seat trainer named ‘Wakazakura’「若櫻」(Young Cherry) is vaguely outlined, which seems likely to be the two-seat trainer of Ōka Model 43 Otsu.

    NameExperimental Wakazakura
    MakerKūgishō
    FormatHigh[wing] – Mono[plane]
    Crew2
    Span (m)9.000
    Length (m)9.000
    Height (m)3.200
    Empty Weight (t).600
    Gross Weight (t).750
    EnginePowder Rocket
    SummaryTraining glider for catapulting
    Progress (Schedule)Start: 07/1945 | Unit 1: 11/1945 | Finish: 03/1946
    StatusBeing designed

    The Wakazakura is otherwise briefly described by a few Japanese secondary sources. It was a modification of the Navy glider ‘Chikara’ (Power). The Chikara was designed and first built by Japan Small Airplanes in 1941 as a two-seat trainer for the experimental ‘MXY5’ transport glider developed by the Kūgishō. Two pilots are seated in tandem for the purpose of training to be towed, gliding, and landing. The Chikara was rather large for a glider, with a wingspan of 11.25 meters, and a length of 8.8 meters. The empty weight of the airframe was 326 kilograms, while loaded it weighed 516 kilograms. The structure was made of wood but designed to a high strength for unlimited aerobatic potential. The main wheels for landing were semi-recessed into the fuselage, while a skid was positioned below the nose.

    Therefore, if the specifications given in the previously mentioned Kūgishō table are correct, the design of the ‘Wakazakura’ differed in the following manner from the Chikara. The wingspan was reduced from 11.25m to 9.0m, which coincides with the wingspan of the Ōka Model 43 Otsu. The length slightly increased to 9.0 meters, probably due to the installation of rocket(s) (unclear if a single or multiple). Also due to the powder rockets, the empty weight had increased by about 274kg, and the loaded weight by about 234kg.

    It is stated in the 88th volume of Senshi Sōsho that the Wakazakura was to be used as an intermediate trainer for not only the Ōka Mod.43, but also the Kikka. This is not clarified by other sources. Regardless, the Wakazakura was also to utilize the same powder rocket catapult system as the K2 and the Ōka Mod.43. The training grounds was to be at the Mt. Hiei catapult site, where the training of the Ōka Mod.43 was being organized.

    Rather important for this theory is the design starting date listed as July 1945 in the Kūgishō table, which coincides with the documented decision to create a two-seater Ōka Model 43 trainer, as explained in the previous section. Furthermore, the initial prototype was only expected to be completed by November 1945, which is quite late. Under this situation, it seems apparent why it was necessary to modify the K2 for testing the rocket catapults beforehand.

    A cliff-side catapult for launching Ōka Model 43 Otsu was completed on Mt. Hiei near the end of the war.

    For training the pilots of the Ōka Model 43 Otsu, the 725th Naval Air Group was formed at Shiga on July 1, 1945, to operate from the Mount Hiei catapult site. Just before the end of the war, a wooden model of the Mod.43 Otsu was loaded onto this catapult and launch-tested with rocket propulsion. The necessary adjustments were made to the catapult system, the glider landing zone was constructed, and the pilots waited for the arrival of Wakazakura trainers.

    Thankfully, the war reached its conclusion before the deployment of Ōka Model 43 Otsu. Not a single prototype of the actual plane nor its Wakazakura trainer was fully completed by the end of hostilities on August 15th, leaving behind little material evidence for researchers. With such little clarifying data and prominent misinformation, it’s easy to see how the unusual ‘MXY7-K2’ and the scarcely documented ‘Wakazakura’ are typically conflated even to this day.


    Sources

    • Nomura, Minoru. Senshi Sousho 45, Imperial General Headquarters Navy Department/Combined Fleet (6), Third Stage Operations Late Period. Tokyo: Asagumo. 1971.
    • Senshi Sousho 88.
    • Naito, Hatsuho. Thunder Gods. New York: Kodansha International. 1989.
    • Kaigun Kōkū Gijutsu-shō. Tōkyō: Gakken Plus, 2008.
    • Tokko Issue 96: Kawamura, Iwao. Mysterious Ouka Model 43 Otsu Turbojet Special Attacker (First Part). 2013.
    • Tokko Issue 97: Kawamura, Iwao. Mysterious Ouka Model 43 Otsu Turbojet Special Attacker (Last Part). 2013.
    • Katō, Hiroshi. God Thunder Corps Record. Tokyo: Hobby Japan, 2021.
    • Rep. 海軍試作機性能要目一覧表 [List of Navy Prototype Planes Performance Specifications], 1945.
    • Rep. Japanese Power Plants For Jet Propulsion. 1946.
    • Rep. Desc of Experimental Aircraft and Experimental Engines Under Development by the Japanese Army and the Imperial Japanese Navy. 1946.
    • Rep. Full Picture of Navy Military and War Preparations Part 6 (War Preparations and Special Attack Preparations on Defeat). 1952.

    *December 13th 2023: Corrected information about Wakazakura