One Hundred Years of U.S. Navy Air Power

by Smith, Douglas V.

  5.The 1920 program as drafted in September 1918 envisaged carriers (NARA RG 38: GB [General Board] 420-2, 10 September 1918); in its discussion of auxiliaries the General Board wrote that “the need for airplane carriers of high speed to accompany the fleet for the purpose of extending its scouting area has arisen from the experiences of the present war. The General Board recommends that the design and construction of such vessels be not delayed.” A list of ships to be completed by 30 June 1925 (i.e., by the end of FY25) included six carriers. This paper included a large aircraft program, but it was limited to large seaplanes, dirigibles, and kite-balloons, plus land-based aircraft then being operated in England, Ireland, and France. No carriers were included in the FY20 program. A 12 October 1919 General Board (GB 420-2) summary of future U.S. naval policy called for one carrier per squadron of capital ships, for a total of seven to work with twelve first-line battleships, and sixteen battle cruisers (seven squadrons, four ships each). That is, at this point carriers were envisaged as supporting arms for capital ships. They were still described as auxiliaries, on a par with destroyer and submarine tenders. The October summary ended with a proposed FY21 building program including two carriers (as well as two battleships, a battle cruiser, ten more scout cruisers, and lesser units). The carriers were expected to be expensive: hull and machinery would cost $20 million, compared to $21 million for a battleship or $23 million for a battle cruiser (but the armor and armament of the gun ships would add considerably; total cost for a carrier was $23 million, compared to $39.5 million for a battleship). Nothing was bought. In November 1920, looking toward the FY22 program, the General Board also vigorously pressed for carriers, arguing that “it is now perfectly evident that the Navy skilled in the use of airplanes and well provided with the most modern types will have a great advantage in war over a Navy not trained in their use or not well supplied [with them]. It is not too much to say that the influence of airplanes upon scouting and information gathering duties may revolutionize former naval practice.” On this basis the General Board asked for two to be laid down in FY22 and then one in each of FY23 and FY24 (its recommendations were in the context of a proposed three-year program). In July 1921 the board associated the figure of two with the need for at least one carrier with each fleet, Atlantic and Pacific, urgently recommending that two carriers by authorized in the FY23 program.

  6.GB 420-7 File 1916-24, dated 6 May 1920, supporting plans to build fast U.S. carriers. According to the memo, “recent reports” show that the Royal Navy is developing carrier tactics in the Mediterranean with three ships (including the seaplane carrier Pegasus), the carrier Argus is assigned to the Atlantic Fleet, and a carrier (the seaplane carrier Ark Royal) is assigned to the China Fleet; Furious is inactive at Rosyth, and Hermes is being completed. Until well into the 1930s the Royal Navy operated the most numerous carrier fleet in the world—but by far not the largest sea-based air arm.

  7.The conversion was proposed in May 1920; a converted scout would support a 500 × 60-foot flight deck, and her hangar could accommodate twelve fighters and six torpedo bombers. Increased tonnage would reduce speed from thirty-five to thirty-one knots and the battery would be cut to four 6-inch guns. Conversion plans were submitted on 12 November 1920. The General Board rejected the idea in December 1920.

  8.GB (General Board) 420 of 3 February 1921 (in NARA RG 80), calling for a hull 812 × 80 × 30 feet, supporting an 800 × 106-foot flight deck, 46 feet above water. Armament (16 × 6-inch in twin mounts and two triple torpedo tubes) and cruiser-level protection were specified, but not the number of aircraft. However, a 20 November 1920 paper in the same file estimates aircraft capacity for either monoplanes or biplanes, which might be fully or partly assembled (in the hangar). The hangar was to be 290 × 80 feet, i.e., a cavity within the hull rather than a long open space, as in later U.S. carriers, with 45 × 50-foot hatches or elevators at either end. Given assumed airplane dimensions, capacity was estimated. The stowage (hangar) deck could accommodate thirty-five fighters and twenty-three torpedo bombers; half as many again would be carried disassembled in the hold, as replacements for crashed airplanes, for a total of fifty-two fighters and thirty-six torpedo planes. Nothing in these papers suggests any attempt to work out the tactics and hence the ideal numbers of these aircraft. Prior to writing the characteristics, the board issued a memo on carriers, noting that they were a new kind of ship and summarizing British experience. Both the flush-deck Argus and the island-equipped HMS Eagle had shown excellent results. Their flight decks were 540 × 68 feet and 660 × 100 feet, respectively, and their respective speeds twenty and twenty-four knots. The board noted that HMS Hermes had been designed with an island in 1917 before any experience had been gained (it presumably did not know that Argus had been tested with a dummy island). In the board’s view, none of the British carriers was entirely satisfactory; “it is desired to give the Naval Air Service the highest type of carrier possible with which to experiment in the construction and operation of heavier than air aircraft.” General Board files also include a 20,000-tonner sketched in November 1920 as a basis for discussion: 660 × 69 × 23 feet (20,000 tons) with a flight deck 650 × 86 feet (104 feet over sponsons) and two stowage spaces 125 × 64 feet forward and 300 × 55 feet aft, both with 20-foot clear height, each with one elevator, making thirty knots using cruiser machinery (90,000 SHP). She would have a 5-inch belt, considerable by light cruiser standards, and a 2.5-inch protective deck. Stowage on the handling (hangar) deck was given as forty-eight fighters (sixteen ready for flight) and twenty-four torpedo bombers (eight ready for flight), with 40 percent more aircraft stowed in the hold disassembled (figures were developed by filling the spaces involved with airplane silhouettes). Speed was thirty-five knots.

  9.GB 420-7 dated 27 June 1921. C&R had submitted alternative sketches with a flush deck (A) and with an island (B) on 12 May. In support of its island design, the bureau commented that so long a ship might not need a flush deck. It now seemed that a carrier should be able to fly aircraft on and off simultaneously, and the deck was so long that she could do so while leaving a neutral area abeam the island. The bureau also pointed out that a flush-deck carrier presented real design problems, not only of disposing of smoke but also of leading air into the boilers from a position forward of the uptakes, which required ducting that would interfere with the forward elevator and also reduce crew spaces. The need to provide telescoping masts, housing radio masts, search lights, and the like (including a pilot house) might well demand unattractive compromises. The same issues arose when the U.S. Navy tried to build huge flush-decked carriers after World War II. The bureau pointed out that permanent masts would give longer radio range, an important consideration. These designs had clipper bows carried up to the flight deck, to cut through the waves (i.e., keep the flight deck dry). The forward end of the flight deck was squared off so that both of an airplane’s wheels would leave the deck at the same time. The open stern was offered to provide cranes to handle seaplanes (a solution the Royal Navy had already adopted in Eagle and Hermes); it would be closed by a roller curtain 20 × 30 feet. The bureau doubted that a lee could be created aft for seaplane handling, and it pointed out that the open stern would lead directly into the large unobstructed hangar; in a following sea water might easily pour in. The curtain would provide a degree of safety, but the ship would have to be maneuvered carefully if it was open. An island design could provide a big crane abaft the island. A flush-decker could also have cranes, but they would be more difficult to place. Based on hangar area, the B design would accommodate thirty-eight fighters and nineteen torpedo bombers, the former fully assembled and the latter with wings folded (Type A had slightly less hangar area, due to the ducting for uptakes and downtakes).

  10.In July 1921 Senator William Borah of Idaho introduced a bill to stop construction of six battleships and three battle cruisers and to convert two of the battle cruisers already under construction to carriers (and also to buy four fleet submarines already authorized).
This was advertised as disarmament by example, the emphasis being on the termination of contracts rather than on the new carriers and submarines, and Secretary of the Navy Theodore Roosevelt (a cousin of the late president) urged that the bill be abandoned for fear that it would affect the disarmament conference the United States had just called for the following November. It is significant here because it shows that the idea of converting two battle cruisers was current well before it was inserted into the Washington Treaty.

  11.Tonnage was a tricky yardstick, because different navies demanded different amounts of fuel to meet their strategic needs. The solution adopted at Washington was a new standard displacement, defined as the ship’s displacement ready for battle less the two main consumables of fuel and reserve feed water. It was difficult to predict standard displacement because ships were designed to operate at a normal displacement including most of their fuel and reserve feed water; one consequence was that many early Treaty designs came out lighter than expected (and many later ones came out heavy, as designers sought to get closer to the limit). The phrase “ready for battle” also caused problems, as navies sought to shave nominal displacement by measures such as including only part of a ship’s ammunition or excluding peacetime equipment such as ships’ boats. Thus Taylor’s 36,000 tons was an estimate based on a design that displaced over 40,000 tons in normal condition.

  12.The initial planned aircraft complement was two fighter, two torpedo bomber, and half an observation squadron (the other half of which would be on board Scouting Force battleships). Each fighter squadron consisted of twenty-seven aircraft (eighteen operating, nine reserve), each torpedo squadron of twenty-four aircraft (sixteen operating, eight reserve), and the observation squadron of eighteen aircraft (twelve operating, six reserve). On this basis the ship would operate thirty-six fighters, thirty-two torpedo bombers, and six observation planes, a total of seventy-four aircraft. Undated memo in GB 420-7 files for 1925–1931 from RADM W. A. Moffett, BuAer Chief. This was soon greatly exceeded.

  13.Thomas Wildenberg, All The Factors of Victory: Admiral Joseph Mason Reeves and the Origins of Carrier Airpower (Washington, DC: Potomac Books, 2003), is Reeves’ only biography. It provided the dates used here.

  14.This innovation seems to have been made almost immediately, but Wildenberg does not mention it. It may have been associated with the creation of the Landing Signal Officer (LSO) position, as short landings required special assistance. The Royal Navy did not use LSOs. The guess as to the date comes from the way in which carrier design was described in the March 1926 C&R memo. As a motivation for increasing carrier capacity, Wildenberg emphasizes the idea that only a large fighter force could fend off attacks.

  15.Memo by BuAer Carrier Section, 1940, in GB 420-7. The equation of high performance and landing deck length helps explain why short Royal Navy carriers using high-performance fighters such as the Corsair and Seafire suffered so many barrier crashes compared to longer U.S. carriers.

  16.There was also considerable interest in rigid airships as fleet scouts, in some cases carrying fighters for self-protection. Although these aircraft were retired in the mid-1930s after some spectacular accidents, the idea survived, and the 1940 “Estimate of the Situation” looking toward the FY42 program included a big rigid airship (ZR, 3-million-cubic-feet capacity) in addition to the blimps used for ASW during World War II. It was never built.

  17.General Board files include a pair of design studies for maximum-size (27,000-ton) carriers dated 24 May 1924 showing that demanding high speed cost aircraft capacity (seventy-two aircraft in a 27.5-knot carrier, sixty in a 32.5-knot ship). At the other end of the scale, the Washington Treaty did not limit carriers displacing less than 10,000 tons. On 23 March 1925 the Bureau of Aeronautics suggested that such unlimited carriers should be the next considered by the U.S. Navy. It guessed that such a ship could carry sixty-four fighters or twenty-four torpedo bombers. On 31 March 1925 the Secretary of the Navy asked the General Board to consider a ship combining the attributes of a scout cruiser and a light cruiser, a new type of ship. To the General Board, this was much the same as the Omaha question raised five years earlier, since Congress would pay either for cruisers or for cruiser-size carriers, but not for both. The BuAer letter actually raised the wider question of carrier size versus aircraft capacity.

  18.Contemporary discussions refer to a non-watertight flight deck, well adapted to fittings such as arrester gear. Apparently this was not the superstructure flight deck adopted in the Ranger design; it was assumed that a watertight flat would be built a foot or two below the flight deck proper. Ranger originally had an open hangar deck mainly to accommodate the planned pair of athwartships catapults, which did not materialize.

  19.C&R planned a total of sixteen studies, varying different factors to show their influence on the design, as it had successfully done in the run-up to the first U.S. heavy cruiser design (Pensacola class). This systematic approach proved impossible partly because aircraft capability depended more on space and dimensions than on weight, and partly because there was no consensus as to which key factors were involved. The bureau therefore preferred not to submit the studies to the General Board, as they could not be used to draw conclusions accurately enough; however, they did indicate some key limiting conditions. Given the two 33,000-ton ex–battle cruisers, the 135,000-ton treaty total for carriers left 69,000 tons (if the experimental Langley were discarded, as the treaty permitted) for two 27,000-tonners, or three 23,000-tonners, or five 13,800-tonners, or six 10,000-tonners. The three tonnages investigated offered, respectively, totals of 192, 220, and 144 aircraft—showing that below a certain point, the ship was too small. C&R also took into account the possibility that the two big carriers would be modified so that their standard displacement could be given as 27,000 tons, leaving 81,000 tons for new carriers (this did not happen), or that the 36,000-ton displacement would be accepted, leaving 62,000 tons. This tonnage could be distributed among up to six carriers displacing between 10,800 tons (for the 62,000-ton total) and 27,000 tons (two ships in each case). No carrier with a mixed air group could be built on about 10,000 tons. Smaller carriers had about 15 percent more flight deck per airplane than larger ones. Although a larger number of smaller carriers would accommodate more aircraft, a smaller number of larger ones would cost much less per airplane; the first cost of five 13,800-tonners would be about 25 percent more than that of three 23,000-tonners, and the cost per airplane would be about 20 percent more. The smaller the carrier, the worse her protection. Although that might not count for above-water weapons that might cause a massive explosion, it certainly did count for torpedo attack.

  20.Characteristics were dated 1 November 1927. There had recently been a proposal to convert Langley’s sister collier, Neptune, into another second-line training carrier, the resulting discussion probably helping to prompt the decision to build a first-line ship instead.

  21.Plans called for one to be an athwartships catapult on the hangar deck; cross-wind launches had been tested at the Naval Aircraft Factory (Philadelphia Navy Yard). The hangar deck location was attractive partly because space under the flight deck near the bow (needed for the machinery of a bow catapult) was so limited by the bridge, by the forward anti-aircraft guns, and by hangar equipment. Moreover, a flight-deck catapult would interfere with flight-deck operations. BuAer therefore proposed two double-ended battleship-type powder catapult on the hangar deck, one just abaft the after elevator and one abaft the uptakes, which themselves were abaft the two elevators, BuAer memo dated 10 March 1928 in GB 420-7. The hangar deck catapults were sometimes described as equivalent to the British and Japanese practice of providing a secondary short takeoff deck on the hangar deck level; catapults did not sacrifice as much hangar space. As detail design proceeded, the two elevators were brought together until they were about forty feet apart (General Board 420-7, 16 March 1928). By March 1930 a third elevator had been added, right aft, to make aircraft handling more flexible. This addition was sugg
ested by aviators aboard Lexington and Saratoga. After aircraft landed, they were immediately re-spotted to the holding area amidships and prepared for another mission; sometimes that required flight-deck crews to work far into the night. The aft elevator could bring some of those aircraft down to the hangar deck, where they could also be prepared, and then back up into the spotting area amidships. Commander Aircraft Squadrons, Admiral Reeves, strongly recommended the third elevator. It became much less useful when operating practice changed to spot aircraft forward rather than amidships.