Diving Dragonflies

The Dornier Libelle and the 1920’s studies to equip the submarines of the Italian Royal Navy with submersible aircraft. 
By Achille Vigna.
(Translated and edited by L. Pavese)

With a contract signed on October 20, 1924, the Italian Ministry of Aeronautics ordered S.A.I.C.M. of Marina di Pisa (Società Anonima Italiana Costruzioni Meccaniche) the building of two minuscule single-seat, single-engine Dornier Libelle (Dragonfly) airplanes. The first was to be delivered no later than April 1925 and the second in the following May. The design of the aircraft had been entered in an early 1924 ministerial contest for a removable wing seaplane destined to be embarked on the Barbarigo Class submarines, and it had come out the winner. S.A.I.C.M. (Italian Anonymous Mechanical Manufacturing Company) had been set up on December 17, 1921, with the purpose of building aircraft designed and certificated by the German company Dornier Mertallbauten GmbH, which could not build its aircraft in-house due to the restrictions imposed by the Treaty of Versailles.

A German-built dragonfly

            The Italian company was located at the estuary of the Arno River. It was equipped with German machinery and tools. It was staffed largely by technicians, managers and workers from Germany and it had begun building the twin-engine, all-metal flying boat Dornier “Wal,” which at the time was the most technologically advanced aircraft of that kind. The Wal was Dornier’s feather in the cap; it was continuously improved; it was produced in military and civilian versions and it enjoyed a considerable commercial success, with more than one hundred aircraft sold.
            During the first four years of activity, the head company had directed its Italian sub-contractor to build or simply assemble other Dornier designed airplanes, with the intention to widen its market. Among these there were the “Falke” fighter, the single-engine “Merkur,” the “Delphin” seaplane and the aforementioned Libelle.

A Libelle on the beach in Marina di Pisa

 The Libelle (whose complete designation was Dornier Do A Libelle II) was a small sport seaplane with rearward-folding wings, and propelled by a Le Rhone rotary engine. The presence of the Libelle in the Italian inventory was due to the fact that it answered to the above-mentioned specifications, except for the removable wing, which was a later modification.
The participation of the Libelle in the Air Ministry contest was imposed to S.A.I.C.M. by Dornier. The Germans had to overcome the reluctance of the Italian company that was already overloaded by work. Before the signing of the contract, the D.G.C.A. (the Italian government agency that decided what aircraft to procure and build) had requested a few modifications to the original design. The alterations were mandated by the size of the cylindrical hangar aboard the submersible boat, the diameter of which had been reduced from m 2.5 to m 2. The Italian company, which was saturated by the manufacturing orders of the Wal flying boat, had shown very little interest in the construction of the two Libelles, to the point that the D.G.C.A. (Direzione Superiore Genio e Costruzioni Aeronautiche) had considered consulting with other companies.
Nevertheless, according to the D.G.C.A.’s July 1924 monthly report, the agency renewed its pressure on S.A.I.C.M. to take on the work, and the contract was finally signed on October 20. The agency’s May 1925 report states that the first Libelle’s fight test was imminent, and that a static test of the second one had been requested. According to the September 1925 report, on the preceding July 23 the company had carried out a few test flights with the first airplane, and on the second aircraft there remained only to mount the engine. The company had suspended the work and was waiting for directions.
Actually, the Le Rhone engine had proved unsuitable; therefore, it was decided to re-engine both prototypes of the Libelle with a 55 hp Siemens Halske SH-5 radial engine, with an addendum to the contract issued on July 7, 1926. Meanwhile, the military registrations M.M. 56 and 57 had been assigned to the two aircraft. The decision to opt for another engine caused further delay, to the point that the second prototype was flight-tested only on August 12, 1926. The test flights were carried out by test-pilot Tullio Crosio and by Dr. Guido Guidi (manager of S.A.I.C.M.), who, out of a sense of responsibility, personally tested almost all aircraft that came out of his plant.
Once this cycle of tests ended (by now, it was the fall of 1927) the final phase of the acceptance tests began, that is, the operations of disassembly, stowing and reassembly and launching from vessel at sea, take off and retrieval. One aircraft was sent to the seaplane base of Cadimare (La Spezia), and the tests were carried out in December of 1927, aboard the submarine Provana, which had been equipped with a provisional hangar.

The Andrea Provana

The Andrea Provana was one of the four vessels of the Barbarigo class, the construction of which had begun in 1915. The boats were m 67 long, and displaced 762 metric tons on the surface and 924 tons submerged. They could sail at 16 knots on the surface and at 9 knots submerged, and they were armed with two mm 76 guns and six mm 450 torpedoes tubes.
The hauling aboard of the airplane was accomplished by means of a sort of dolly that ran on tracks, which were mounted on a long turning sled. The sled came out of the hangar and protruded on the water, reaching the bottom of the hull of the seaplane. These operations are described in the December 28, 1927 report, written by the commanding officer of the Provana Lieutenant Commander Carlo Balsamo, entitled: “Report concerning the experiments with the seaplane.” It was a classified document, illustrated by 9 photographs. A copy of the report, signed by second in command, Lieutenant Garinei was found devoid of the indication of the addressee; but in all likelihood the report was destined to the General Staff of the Italian Royal Navy. This is the text, which was taken from a typed copy of the report:

R. Submersible ship PROVANA
[…] As we were ordered by the High Ministry, a removable wing seaplane has been taken aboard and recovered in a provisional sheet-metal framed, fabric covered hangar.
The hangar, which was supposed to be a m 2 diameter, m 8 long cylinder, was actually much larger, as shown by the attached drawing. In fact, the maximum height from the deck was m 2.6.
These over-sized dimensions were due to the need of placing the guide tracks of the dolly at a certain height from the deck, in order to pass over various obstructions, such as the washboards of the hatches, the raising handle, etcetera.

The Provana's cylindrical hangar

If a special dolly were devised, and if the tracks were placed on the bottom of the cylinder, the 2 m diameter would be sufficient to accommodate the aircraft “Libellula.” (Dragonfly, in Italian).

Photograph #1 shows the aft end of the hangar, where the water-tight door should be. The nose of the airplane, pointed in the direction opposite the bow of the boat, is visible.

Photograph #2 clearly shows the tracks for the launch of the aircraft, which reach all the way to the far stern.

Photograph #3 shows the beginning of the assembly, that is, the setting up of the central section of the aircraft, with the engine and the empennage. 

After these operations, the airplane is turned athwartship (sideways) and the wings and the fins (the sponsons) are mounted at the same time. Afterwards, the aircraft is aligned with the keel, the controls are set, the engine started and the airplane is launched, as shown by photograph # 6.

To launch the aircraft, the stern of the boat is submerged flooding the stern tanks. As soon as the seaplane floats, it is able to leave the boat on its own power.
To retrieve the aircraft, two sailors wearing diving suits guide the airplane on its sled, as shown by photograph #7. 

As soon as the seaplane is on the dolly, it is hauled aboard while at the same time the stern ballast tanks are emptied, so that the aircraft is immediately out of the water (photograph #8).

Then, the aforementioned operations are carried out in the reverse order, disassembling the airplane and stowing it in its hangar.

The average operation time was about 18 minutes for the assembly and about 13 minutes for the disassembly. These times are often liable to increase considerably, due to weather conditions and the difficulties that very likely will arise during such delicate operations with material subject to wear and tear.
On December 23, in the presence of the Admiral, Commandant of the Submarine Division, the above-described operations were carried out within the confines of the dyke, and they respectively took 22 and 18 minutes. Nevertheless, it must be pointed out that the weather conditions were particularly adverse, with strong gusty wind, rough seas and low temperature which greatly impeded the work of the men.
Even from this summary description of the operation, it can be ascertained that with the “Libellula” type aircraft, the assembly, the launch, the retrieval and the stowing of the aircraft are very time consuming operations, which, in the best case scenario, with very well trained and practiced personnel, can only be accomplished in no less than about a quarter of an hour.
In an attempt to facilitate the operations, I have personally travelled to Marina di Pisa to the aircraft manufacturing plant, in order to request a few minor modifications that were immediately made and produced a gain of several minutes. But, in any case, in my opinion the time required to launch the Libellula is absolutely prohibitive. Moreover, the flight characteristics of the aircraft make it unpractical and useless. The flight endurance of the airplane (a little more than one hour) is absolutely insufficient.
At the present level of technology, I believe it would be possible to build a seaplane of adequate performance, while remaining within the limits of weight and dimensions imposed by the size of the cylinder; but it would be necessary to design a purpose-built aircraft. The airplane should have an engine that could be started from the cockpit, which now is impossible. The dolly with the rotating sled is too heavy and cumbersome. Its height was about cm 50. Therefore, it would be necessary to build a new type of dolly with externally placed wheels, so that the dolly would remain within the tracks in one piece, that is, without the above-mounted rotating sled. The dolly could be made to rotate instead on a suitably placed platform. Naturally, all the obstacles on the deck should be eliminated.
If I may, I would like to point out to Your Lordship that perhaps the operation of a seaplane from a submarine could be made much more practical if we had an aircraft that could remain submerged. In that case, it could be stowed in the vessel’s inter-space, eliminating the cylinder altogether, with great advantage and making the launch operations much simpler. I don’t think that would be technically impossible, using a completely metallic aircraft with special paint, and in view of the fact that totally watertight engines already exist.                            

Lieutenant Commander
C. Balsamo


The purpose of employing mid-size vessels like the Provana and a small airplane like the Libelle for the tests was to gather the fundamental elements on which to base the specifications for an aircraft designed to operate on the large oceanic submarine Ettore Fieramosca.
The Fieramosca had been designed by Naval Engineers’ General Curio Bernardis, and at the time it was being built in the Tosi shipyards of Taranto. It was a large m 84 vessel that displaced 1556 metric tons on the surface and 2128 tons submerged. It was intended to be a true submersible oceanic cruiser, and it was armed with a mm 203 gun. It turned out to be a poorly manueuverable ship. It served only for about ten years, and it was decommissioned in 1941, even when the war was still on-going.

   The launch of the giant Fieramosca. The cylindrical hangar can be seen abaft of the sail.

Besides the performance of the Dornier Libelle (which turned out to be inadequate for the purpose), it was necessary to evaluate the operational limits of the entire concept, especially the difficulties related to launching and retrieving the airplane in the open sea; and last but not least the difficult problem of starting the engine, caused by the space constraints and by the danger of the spinning propeller in an unstable environment such as the deck of a submarine.

Two new aircraft, the Piaggio P.8 and the Macchi M. 53, were entered in a new contest that was announced in 1927. Both were single engine, single seat airplanes powered by the 75 hp A.D.C. Cirrus II, with a removable wing, according to the specifications. As far as the Macchi is concerned, the approval was given the same year, and construction began. The first M.53 flew on October 25, 1928. The evaluation of the aircraft was still going on, when the Ettore Fieramosca was launched on April 15, 1929, equipped with a cylindrical hangar mounted abaft of the conning tower.

Macchi M.53

 The Piaggio proposal, designed by Dr. Giovanni Pegna, initially followed the same course, but the production that began in 1927, in the Finalmarina plant (Finale Ligure), suffered serious delays. The Piaggio P.8 was a parasol wing monoplane, particularly suited for the reconnaissance role. It was a modern design with an almost completely metallic airframe. For that time, it could be considered a technological marvel.

                     The P.8

The P.8 folded and stowed in its cylinder

Nevertheless, although the purchase had been finalized on February 22, 1928, the P.8 was flight tested only on November 12, 1929, when by then the entire project of the aircraft-carrying submarine had been abandoned.
 The main reason of the decision to terminate the project, though, was not due to the inadequate performance of the two prototype aircraft, or to any operational difficulties that eventually could have been overcome. Lieutenant Commander Balsamo’s idea of water-proof aircraft that could be stowed in the vessel’s inter-space could be considered odd, even with today’s technology, but the real problem at the time consisted precisely in the large water-tight cylinder mounted on the deck. The cylindrical air filled hangar required a continuous compensatory re-trimming of the ship to maintain the proper longitudinal attitude. Furthermore, the maximum depth that the boat could reach and the maneuverability of the ship were greatly reduced.
While the Ettore Fieramosca was being fitted out, the cylindrical hangar for the seaplane was off-loaded shortly before the sea trials, and the submersible ship was commissioned on April 1, 1930 in a conventional configuration, that is, without the hangar; and it remained the only ship of its class. 

        The article was originally published in June 2012, on the # 225 issue of the Italian magazine "Storia Militare," and it was relayed to me by the good folks of the Italian group Avia (whom I'd like to thank). Several countries experimented with submarine-borne aircraft. The first that comes to mind is the Focke-Achgelis Fa-330, which was a German tethered (unpowered) gyroplane that the Kriegsmarine employed, with some success, to extend the visibility of the U-boats in WWII (after experimenting with a folding-wing airplane); and I know the Japanese also built aircraft-carrying submarines, but I don't remember the details, and I'll let you do the research. In any case, I'm sure that the experiences of the Italian Royal Navy are not well known, and the original article featured the previously unpublished report, complete with pictures, of  the trials conducted aboard the Andrea Provana; therefore I thought it could be of some interest. I hope you thought so too.
Your comments will be greatly appreciated. Thank you. (And I'd like to thank Poul Webb from whose blog I took the title picture).
Leonardo Pavese     


Popular posts from this blog

Neanche in Slovacchia