1. Introduction
One of the authors, Nakahara, was involved in the development of a novel gas turbine system so-called Pescara Cycle, in the early days of his career, and spent several years of 1950s in developing a free piston gas generator. The history was traced and reported in publications⁽¹⁾.
Prior to it, there was a high pressure air compressor operating in the same principle developed by Junkers AG in Germany. As soon as it was presented in VDI Journal and exhibited in Leibzig Messe in 1935, the Imperial Japanese Navy intended to introduce it to Japan, and Kobe Steel, Ltd and Mitsui Seiki, Ltd. had been receivers of the technology.
After the War, Kobe Steel, Ltd. (later abbreviated Kobe) continued production of the same type compressors, and Nakahara also experienced jobs of their operation and maintenance on site. The history was interested by the authors and now reported. However it should be granted in advance that they were too young to attend all of actual scenes of the history, therefore, most of facts were obliged to survey in written and remained documents⁽²⁾.

2. Features of the technology
Prof. Hugo Junkers, a professor of TU Aachen and the founder of Junkers Flugzeug Werke AG and Junkers Luftverkehr AG, later consolidated to become Deutsches Lufthanza AG, invented an engine consisted of an opposed piston working in two strokes Diesel principle and obtained the patent in 1907. According to the principle, he developed a famous aeronautical Diesel engine named JUMO 205 burning heavy oil. It was installed in aeroplane type G38 and Ju52 and proved successful performance.
In 1912, Prof. Junkers reported at a conference of German Shipbuilding Association, on the principle of a free piston air compressor, and test results of an unit to investigate heat transfer in the combustion chamber of the engine. Later, the air compressor was improved and developed to meet actual operations, and became to be widely used by German Navy especially in submarines, so-called U-boats, through World War 2 performing successful results. The structure is presented in Fig. 1 as a cross-sectional drawing to show its operation. It is consisted of multi-stage compressors and an engine directly connected without any crankshaft, its bearings and connecting rod as conventional engines have. The engine is consisted of an opposed piston moving symmetrically by a set of synchronizer using rack and pinion mechanism about the fuel injection nozzle located at the center of the cylinder. On the cylinder wall there are each scavenging and exhaust ports. It operates in two strokes Diesel principle. Each piston is directly connected with reciprocating compressors. In the Figure, the left hand side is connected with the 1st and 4th stage, and the right hand is with the 2nd and 3rd.

Fig. 1 Hight pressure 4 stages air compressor.

3. Potential of Kobe Steel, Ltd. as a receiver of the technology
Kobe initiated manufacture of steel castings and forgings used for parts of ships in 1905. Since then it expanded its activity in manufacture of various heavy machineries. It has a strong connection with the Imperial Japanese Navy to supply machineries and equipments necessary for naval use. In 1914, vertical high pressure air compressors driven by steam engines to generate compressed air for torpedoes launch and propulsion, were manufactured under license agreement of Brotherhood Co. of U.K. In 1918, 2 stroke Diesel engines of 450KW to drive submarines were manufactured under license agreement of Sulzer Brothers AG in Switzerland. After these machineries, Kobe continued to supply such kinds of machineries to the Navy, accumulating skills of the technology of manufacturing. It is not only for naval service, but also for general purpose of commercial navigation service. The outbreak of the World War 1 in 1914 lasting to 1918 caused shortage of European products offering chances for infant Japanese heavy industries to grow rapidly.
Several examples of the product line were as follows. As for the main engine for a submarine in 1919, a Sulzer type 2 strokes air injection engine of 6 cylinders generating 975KW. In 1925, the same type engine of 8 cylinders generating 2550KW, and a 4 strokes air injection single acting engine of Kobe's own design. In 1927, an airless (so-called solid) injection engine first developed in Japan by Kobe. In 1932, a Sulzer type 2 strokes double acting airless injection engine of 7 cylinders generating 5700KW. In 1938, a 2 strokes airless injection double acting engine of 8 cylinders generating 8250KW of Kobe's design for the main engine of a commercial ship. It was the biggest output of that time in Japan.
For successful supply of these machineries, there was intimate human communication between Kobe and the Navy. The Navy owned a big dockyard and shipbuilding factory in Kure located at the northwestern coast of Seto Naikai, the Inland Sea, where many skilled engineers were working. In case of need, the experienced engineer was called to Kobe to instruct the way of manufacturing in shops. Sometimes it was a temporary service, the other time a permanent job having a position in factories or design rooms. In progress of time, Kobe became well qualified to manufacture newly developed machineries of complicate construction.

4. Process of transfer of the technology
VDI Zeitschrift, the Journal of the Union of German Engineers, presented a paper named "Junkers Freikolbenverdichter", Junkers free piston compressor, reported by Prof. Kurt Neumann, a professor of TU Hannover on the theoretical analysis of Junkers free piston air compressors in the February issue of 1935. A senior officer of the Navy, Captain Sakamoto, interested in it and announced in report inside of the office to attract attention of its importance. At Leibzig Messe in following March, the developed compressors were exhibited. They were 2 kinds, one was for high pressure air of 20.5 MPA, and the other was for low-pressure air of 0.7 MPa. He visited the exhibition with a liaison officer of the Navy in Germany, Commander Kimoto, to look and examine them on site.
They reported to Tokyo that the performance of the compressors was acceptable for use in the Navy. The decision was made to produce them in Japan and to purchase the license. Negotiation started. The officer of Junkers AG agreed to promote the high pressure compressors because their performance was proven to be stable during actual services in German cruisers, destroyers and submarine. However, the low pressure was still under development, and exporting was not acceptable for Junkers AG. The Japanese officers understood the situation and decided to purchase the high pressure only. The contract was made to pay DM 500,000. for licensing fee, and each 2 sets were to be imported to Mitsui Seiki, Ltd. and Kobe respectively.
In May 1936, officers of Mitsui were in Dessau to attend the witness operation, and the compressors arrived in Tokyo in following August. Kobe's officers, Ichou et al., were in June 1936 in Munich, because the compressors were constructed in a new factory. Hasegawa was working for Ichou in charge of receiving the technology. He disassembled the imported compressor and after examining and sketching all parts, reassembled it for trial operation. The operation was successful to receive it. He arranged drawings for manufacturing according to his examination and sketching, and constructed a new one. Its first operation was satisfactory, but unfortunately he could not start it again. By dismantling it to examine, he found that all parts inside were burned and stained together. It was because of selection of materials of parts especially exposed to high temperature and severe friction. There was, he felt, a deep difference between German and Japanese technology at that time.
The Imperial Japanese Navy used pure oxygen instead of air for propulsion of torpedoes. It served to extinguish bubbling traces of torpedoes, to reach long distance and to cause more intense explosion with more explosives by less space of oxygen than air, which was peculiar to the Imperial Japanese Navy. On the contrary, other navies' torpedoes were driven by air. The compressors were used to launch torpedoes and to generate high pressure air as the material of oxygen to be separated from nitrogen by deep frozen method. For the purpose, the outlet pressure should be 25 to 30 MPa. Modification was made in the specifications to adapt the condition. In production and trial operation, there caused many difficulties of engine liners, piston rods, piston rings, fuel pumps, and connectors of cooling water pumps. After removing such difficulties, the compressor was installed in a destroyer named Asashio at Sasebo Naval Dockyard in July 1939 to execute witness operation on board. In September 1939, a meeting was held in the Navy, and after confirming all of the difficulties in the early stage were solved, it was decided to adopt the compressor officially as a standard equipment of the Navy. It was named "Type Yu Air Compressor Mark 1". Mass production in factories started.
An engineer who began his career in 1939 in Kobe recalls the tremendous noise of combustion excited in the witness operation of the compressors. On the operation yard, several sets were always in witness operation. 12 hours continuous operation delivering the highest air pressure was the norm of the witness operation. However in the early days, it was difficult to keep continuous operation because of plugging of fuel nozzles, sticking of fuel pump plungers, etc. Operators must be always ready to exchange such parts. At the compressor side, cracking of valve plates especially at the high pressure stage was suffered. Curved section of valve plates, suction and delivery either, was most likely to cause cracks. Once occasionally, 36 hours continuous operation was demanded by the Naval officer attending. Just after it had finished, all of workers in the yard relieved breathing deeply and freely.
Some hundreds sets were manufactured and supplied to the Navy. The production rate was estimated 50 sets per year. They were installed in cruisers and destroyers for launching torpedoes and to generate oxygen for propulsion agent of torpedoes. On the other hand, torpedoes launched by aeroplane did not need oxygen but high pressure air of 18-30 MPa. Some hundreds sets were also manufactured for the purpose. Several sets for aerial-torpedoes were installed on automobiles to work in air bases. Fig. 2 shows the configuration.

Fig. 2 Automobile installed with high pressure air compressor.

5. Further development of the technology
As previously explained, Junkers AG had low-pressure compressors but refused to export them. There had been requirements of low-pressure air sources for various purposes. In 1944, Mark 4 of 6 MPa and Mark 5 of 1 MPa were planned, but it ceased in the early stage. Prof. Fujio Nagao of Kyoto Imperial University then was entrusted by the Navy to consult the plan in fundamental experiments. No further development was planned in Wartime.
After War, Kobe preserved the technology to manufacture the compressors, and to revive it Kobe made its market research except military use. 0.7 MPa for civil, construction and mining and 3 MPa for starters of marine Diesel engines, came in the sight. Kobe developed a series of compressors of 2 stages for 0.7 MPa, 3 stages for 3 MPa and 4 stages for 30 MPa. They were compact, light-weighted, easy to operate and especially suitable for use in remote areas.

Fig. 3 Low pressure 2 stage air compressor.

Fig. 4 Low pressure air compressor.

6. Closing
Kobe manufactured Junkers high pressure air compressors about 20 years. It proved successful performance through 1940s and 50s in naval and civil services. It served also in Kobe effectively to expand the ability of manufacturing technology concerned. Based on experiences obtained during the time, Kobe promoted to the development of a novel gas turbine system. Its history is reported in papers already published in Japanese⁽¹⁾.

References

1. Nakahara M., Free Piston Kikai-Kouzou to Rekisi, Shinko-Techno Gihou, Vol.13, No.25 & 26 (2001) (Free Piston Machineries Features and History).
2. Documents in reference are listed in Shinko-Techno Gihou, Vol. 12, No. 24 (2000) reported by Nakahara M.