The Hidden Accident and the Outbreak of War with the U.S. and Britain: How Did Japan Deal with the Problem?

The Shibuya archives are enormous, consisting of more than 4,000 materials on various subjects including casualties of the atomic bomb.[1] Even though we chose only the materials directly concerning the Rinkichoaccident, it is impossible to present here a full analysis of all the details gleaned from these voluminous materials. Among these, this chapter focuses on the special examination committee established in January 1938. The purpose of the committee was as follows [39]:

Problems were found with the turbines of Asashio-class destroyers…. It is necessary to work out remedial measures and study the design of the machinery involved and other related matters, so that such studies will help improvements. These research activities must be performed freely without any restrictions imposed by experience and practice in the past. The special examination committee has been established to fulfill this purpose.

Its organization was as follows [40]:

• General members who did not attend subcommittee meetings

– Chair: Isoroku Yamamoto, Vice Admiral, Administrative Vice Minister of the Navy

– Members: Rear Admiral Inoue, Director of the Bureau of Naval Affairs, the Ministry of the Navy and five other members

• First subcommittee for dealing with engine design and planning

– Members: Leader: Shipbuilding Vice Admiral Fukuma, Director of the Fifth Department (including the turbine group), the Technical Headquarters of the Navy; and nine other members

• Second subcommittee for dealing with the maximum engine power and suitable

load/volume

– Members: Leader: Rear Admiral Mikawa, Director of the Second Department, the Naval General Staff; and eleven other members

• Third subcommittee for dealing with prior studies/experiments/systems and

operations

– Members: Leader: Rear Admiral Iwamura, Director of the General Affairs Department, the Technical Headquarters of the Navy; and ten other members

Table 10.5 Members of the special examination committee by section

Ignoring duplication of members belonging to different subcommittees and arranging the net members by section, we obtain the following result (see Table 10.5).

The accident, as mentioned above, concerned the breakage of turbine blades. Tracing back the history of the development of the marine turbine in Japan since 1918 when the Navy began to adopt geared turbines, we find that various failures occurred with main turbines. When we classify these failures during the period from 1918 to October 1944 by location, failures involving turbine blades account for 60 % of the total (see Table 10.6).[2]

The Imperial Japanese Navy had thus had many problems with turbine blades for many years and accumulated experience in handling them. Accordingly, it is unsurprising that the special examination committee took the accident as merely a routine problem from the outset based upon such a long and rich experience. In fact, the special examination committee drew a conclusion made up of two points, both of which were in line with such accumulated experience. First, the accident was caused by insuffi blade strength. Second, turbine rotor vibration made the insuffi strength emerge as a problem [41]. On the basis of this conclusion, a plan was worked out to improve the design of the blades and rotors of the Kanpon type turbines for all naval vessels. It was decided to change the form of the blades so as to make their stress concentration lower to enhance their strength [42]. The improvement of 61 naval vessels' turbines was indicated as the fi step, in accordance with the voluminous previous reports of 66 committee meetings held over a period of 10 months [43].

However, the blade breakage in the accident was significantly different from that in the past. In impulse turbines, for instance, blades in most cases were broken at the base where they were fixed to the turbine rotor. In contrast, one of the salient

features of the Rinkichoaccident was that the tip of the blade was broken off. The

broken off part amounted to one third of the total length of the blade.[3] Figure 10.2 is a photograph showing the locus of the breakage.

Table 10.6 Turbine failures on naval vessels classified by location: 1918–1944

Location

Incidents

Percentage

Cumulative

Impulse blade and grommet

368

46.8

46.8

Reaction blade and binding strip

111

14.1

60.9

Reduction gear and claw coupling

80

10.2

71.1

Bearing and thrust bearing

66

8.4

79.5

Casing

46

5.9

85.7

Casing partition and nozzle

34

4.3

89.7

Blade wheel and spindle

22

2.8

92.5

Steam packing

20

2.5

95.0

Others

39

5.0

Total

786

100.0

100.0

Source Based on [25, pp. 1–2]. Reaction blade means the blade of a traditional Parsons turbine (Cf., [25, p. 4].)

These facts indicated that the accident was significantly different from any previous routine problem. Yoshio Kubota, an Engineering Captain of the Navy who happened to be transferred to the Military Affairs Bureau in November 1938 when the special examination committee reported its conclusion, eventually noticed this point. It was not really permissible for a newcomer to the Military Affairs Bureau of the Navy to utter an objection to the latest conclusion of the special committee. In addition, six months before his transfer to the bureau, the Japanese government enacted the Wartime Mobilization Law on April 1, 1938 for the purpose of “controlling and organizing human and material resources most efficiently… in case of war” (Clause 1). Naval vessels came first in the specification of the law as “resources for wholesale mobilization” (Clause 2). Against this background of wartime mobilization, a naval engine failure caused by small tip fragments of the main standard engine was a very delicate matter for anyone to raise.[4] Despite the circumstances, Kubota strongly recommended that confirmation tests should be conducted again for naval vessels of the same type. He argued that if turbine rotor vibration was the true cause, then the failure would be repeatable when the engine was run continuously at the critical speed causing rotor vibration (nearly 6/10 to 10/10 of the full speed).[5]

The Navy finally decided to initiate continuous-run tests equivalent to tenyear runs on April 1, 1939. No failure occurred. This provided the Navy with the simplest practical rationale for cancelling the overall remedial measures for all naval vessels, which were expected to require huge amounts of extra money and

Fig. 10.2 Broken part of a blade in the Rinkichoaccident (Source [42])

time.[6] An order was issued promptly to postpone the modification to the turbine blades and rotors of the Kanpon turbines for all naval vessels. At the same time, however, there was obviously an urgent need to consider the possibility of another cause and a study to identify the cause was restarted. The Maizuru Naval Dockyard conducted preliminary on-land tests and a more thorough one followed at the Hiro Naval Dockyard to confirm the conditions that would make the failure recur. However, the test was extremely difficult to carry out. There were two reasons for this. First, the complete test required the Dockyard to construct from scratch a full-scale experimental apparatus for a load test of vibration, which was only completed in December 1941, the month the war with the U.S. and Britain broke out. Second, the test turned out to be so large-scale, eventually extending to more than 35 main items, that it took far more time than expected. As a result, the schedule for identifying the cause, which was originally expected to be completed in November 1940, was extended to mid-1943.[7] Thus it is probable that all of Japan's naval vessels had turbines which were imperfect for some unknown reason when the country went to war with the U.S. and Britain in 1941. What, then, was the true cause for the accident? The true cause was binodal vibration. Previous efforts to avoid turbine vibration had been confined to onenode vibration at full speed since multiple-node vibration below full speed had been assumed to be hardly serious and unworthy of attention based on rule of thumb.[8] The final discovery of the true cause of the Rinkichoaccident drastically changed the situation. It revealed that marine turbines are susceptible to a serious vibration problem below full speed. It was in April 1943 that this true cause was eventually identified by the final report of the special examination committee—

almost one and half years after war broke out (see Fig. 10.3).[9]

Only three months before the submission of the report, a theoretical study made at the Hiro Naval Dockyard supported the conclusion that the true cause was binodal vibration.[10] The results of theoretical calculation, on-land confirmation testing, and the characteristics of the actual failure matched. The complete mechanism creating binodal vibration itself was still left for further studies. Even so,

Fig. 10.3 The front page of the final report of the special examination committee (Source [47])

every result from the special examination committee that finally concluded in 1943 pointed to the same single cause: binodal vibration [52].[11]

Strictly in terms of the technology involved in the accident without hindsight, therefore, all the evidence suggests that the Japanese government went to war in haste in 1941 notwithstanding the fact that it had unaccounted for, highly intricate, and serious problems with the main engines of all its naval vessels. And that fact was kept secret by the military sector from other sectors involved in the militaryindustrial-university complex, not to speak of the general public. The rarity of breakdowns of naval vessels due to turbine troubles during the war is a completely different matter, one of hindsight. Thus, the Rinkichoaccident strongly suggests that practical results alone (for example, rarity of breakdowns of naval vessels due to turbine troubles) during wartime, possibly in peacetime as well, do not prove the essential soundness of the development trajectory of technology, and that of the science-technology-society interface and national decision-making along the trajectory.

  • [1] When Japan was defeated in 1945, most military organizations were ordered to burn documents they had kept. Many documents of the Imperial Navy were burned before the General Headquarters of the U.S. Occupation Forces ordered the government to submit documents regarding the war. Ex-managers and ex-directors of the Imperial Japanese Navy then held meetings and decided to undertake a research project to collect, examine, and preserve technical documents to the extent possible. The Shibuya archives were the result of this project and came into the hands of Ryu¯ taroShibuya. The description of the background of the Shibuya archives is based on Shibuya Bunko Cho-sa Iinkai, Shibuya Bunko Mokuroku (Catalogue of the Shibuya Archives), March 1995, Commentary.
  • [2] This classification assumes that if a problem at one location produces another problem at another location, the latter problem is not counted separately, but is considered part of the former.
  • [3] The breakage as described in the record written at that time is as follows: “Moving blades and the rivets on the tip of the 2nd and 3rd stages of the intermediate-pressure turbines were broken…. The break in every moving blade was located at 40–70 mm from the tip” [42].
  • [4] Reference [44, p. 412]. The author was in charge of drafting the national mobilization plan at the Cabinet Planning Board (Kikaku In) in the prewar period. For the Navy, war preparation updates started from August 1940. See [45, pp. 93–94]. Sugiyama was the Chief of the General Staff at that time.
  • [5] Records of an interview with Yoshio Kubota made by the Seisan Gijutsu Kyo-kai (Association for Production Technology) on March 19, 1955 [46].
  • [6] These original remedial measures are kept in the Shibuya archives.
  • [7] The descriptions here are based on [47]. This is the final report of the special examination committee.
  • [8] In general, such was the standard of turbine design in the prewar period [48–50].
  • [9] According to this report, “Binodal vibration occurs when the product of the number of nozzles and the revolution of blades … equals the frequency of the blades at binodal vibration [47].” This means that a forced vibration caused by steam pulsation and a specific binodal frequency of blades resonate with each other, as a result of which binodal vibration occurs.
  • [10] It proved that even if uniform vertical and horizontal sections were assumed for the purpose of simplification, binodal vibration could produce the maximum stress at places less than threefifths of the distance from the tip of a blade, which matched the place of the actual breakage in the failures [51]. Dr. Yasuo Takeda discovered this document on March 3, 1997, and it was added to the Shibuya archives.
  • [11] Shigeru Mori, a contemporary Navy engineer who graduated from the Department of Physics of the Imperial University of Tokyo seems to have tried to construct a model to identify the mechanism, whose details are not available now See [53]. When we look at other circumstantial evidence such as the fact that the blade breakage was limited to a relatively small number of turbines of particular newly built destroyers, it was still plausible that the strength of particular blades had something to do with the failure. The Navy therefore revised its design directive to ensure an enormous increase (from 0.4 to 1.5 mm) in the thickness of turbine blades just after the submission of the final report of the committee in April 1943. The original design directive had been issued on May 1, 1931, the documents of which are collected in the Shibuya archives. In interpreting this circumstantial evidence, the author is indebted to Dr. Ryo-ichiroAraki for technical advice. Considering this circumstantial evidence together, there were possibly two closely associated aspects in the failure. One is a universal aspect leading to the detection of binodal vibration. The other is a more local aspect possibly due to the testing and quality control of the strength of the particular broken blades. Whatever weight may be given to each aspect in the description and analysis of the failure, however, as the date of the final report indicates, it was only after April 1943 that both aspects were finally noticed. By then, about one year and a half had already passed since the outbreak of the war with the U.S. and Britain in 1941.
 
< Prev   CONTENTS   Next >