US NAVY TORPEDOESby Frederick J Milford
Part Three: WW II devlopment of conventional torpedoes 1940-1946
Reproduced with permission from the January 1997 issue of THE SUBMARINE REVIEW,
a quarterly publication of the Naval Submarine League, P.O. Box 1146, Annandale, VA 2200
As we have noted earlier, the entry of the United States into WW II led to major changes in the torpedo situation. Huge quantities were required, operational experience exposed problems in service torpedoes and there were needs for new kinds of torpedoes. In this part we consider the new conventional, by which we shall mean non-homing, torpedoes that were developed as part of the WW II research and development effort. The explosive growth in the number of torpedoes under development, twenty-one distinct marks, during the four years of US involvement in WW II, was remarkable. The pace was much slower, both before and after, twenty in the entire 50 years from 1889 through 1940 and only thirteen since 1950. Another change was the involvement of the National Defense Research Committee (NDRC) in torpedo studies, which marked the beginning of the end of the Newport Torpedo Station's monopoly on torpedo research and development. Through the NDRC university and industrial laboratories became involved. These organizations greatly expanded both the industrial capabilities and the intellectual scope devoted to torpedo research and development and became the primary performers in this realm. Torpedo production was expanded by using manufacturing firms and GOCO (Government Owned Contractor Operated) plants as well as the traditional navy facilities. Of the roughly 64,0001 torpedoes produced during WW II the Naval Torpedo Stations produced about 46%, the GOCO's about 31% and the industrial firms about 23%. The Navy did not, however, dominate WW II torpedo research and development. Of the new homing torpedoes, which will be discussed in a subsequent part of this series, only one, the Mk.34, was developed entirely by a US Navy activity. Two others were developed in joint navy/contractor programs. In the realm of conventional (non-homing) electric torpedoes the navy led the NTS-Newport, GE, Exide team that developed the Mk.20 and worked with GE to develop the Mk.36. In addition to the Mk.34, the navy was solely responsible for the development of the Mk.23, single speed version of the Mk.14, and the Navol torpedoes Mk.16 and Mk.17. A Navy (NTS-Newport) monopoly of the torpedo business such as existed with steam torpedoes from 1922 to 1941 disappeared and has not been re-established in the years since WW II2. Full scale production of torpedoes at NTS Newport ended in 1946 and the Goat Island facility was totally closed by 1951. Navy torpedo research and development did continue in the Newport area at a new facility at Coddington Cove.____________________
1The number quoted in Buford Rowland and William Boyd "U.S. Navy Bureau of Ordnance in World War II" Washington: GPO, n.d., 57,653, appears to exclude homing torpedoes possibly as a security measure.
2Occasional comments imply that the navy was not entirely happy with the torpedo establishment. The use of "mine" rather than "torpedo" for the Mk.24 and several other weapons is sometimes claimed to have had a secondary objective of avoiding involvement of the torpedo establishment. BuOrd also delayed Bell Telephone Laboratory access to torpedoes and torpedo technology as it existed in late 1941, presumably in order to get a fresh perspective. M.D. Fagen, ed., "A History of Engineering and Science in the Bell System: National Service in War and Peace (1925- 1975)", Murray Hill: Bell Telephone Laboratories, 1978.
In addition to modifications of existing torpedoes, entirely new and significantly changed conventional torpedoes were developed. The two major areas where new developments were made were propulsion and warheads. The major propulsion developments were the use of Navol ( a 70% solution of hydrogen peroxide, H2O2, in water) to supply the oxygen for combustion in steam torpedoes and the development of successful electric torpedoes. The most important, but often overlooked, warhead development was the conversion from TNT to Torpex with the attendant increase in underwater damage by over 50% for a fixed weight of high explosive. Altogether nine of the eleven conventional torpedoes shown in Table 1 were under development during WW II. The other two were the last two conventional torpedoes developed by the US Navy and are included to complete the history of conventional torpedoes.
| Table 1|
NON-HOMING TORPEDOES UNDER DEVELOPEMENT DURING WW II
Service Torpedoes in bold
Design and Development
Service Dates/Total Production
NTS-Newport & NRL
Navol (high test peroxide)
NTS-Newport & NRL
Mk. 18 w/electric controls
NTS-Newport, GE &
Final version of EL. Mk.2
|Mk.23||NTS-Newport||1943-1946/9600||SS||Single speed Mk.14|
|Mk.25||Columbia U., Div. War Res||NIS/25||A/C||Improved Mk.13|
|Mk.26||Westinghouse||NIS/25||SS||Mk.18 w/electric controls|
and seawater battery
|Mk.36||TS-Newport & GE||NIS/0?||SS||All electric, seawater|
battery, pattern running
The following two torpedoes were not WW II developements, but they were the last two conventional
torpedoes considered by the US Navy. Neither was developed into a service weapon.
|Mk.40||NOTS Pasadena||NIS/0?||A/C or|
|Test vehicle, solid|
propellant driving gas
|Mk.40||NOTS Pasadena, NOL,|
NTS-Newport, Penn State
ORL & Stevens Institute of
|Turbine powered, pattern|
running, etc. The ultimate
Japanese shipping in the last years of WW II and the Mk.16, though not used in combat during WW II, was a standard submarine weapon until 1975.
Ever since Robert Whitehead invented the self propelled torpedo, a key problem has been how to carry enough energy on board to provide the desired range and speed. Burning organic fuels, hydrocarbons or alcohols, represented a huge improvement over compressed air alone, but further progress required improved oxidants. There are two obvious problems in using compressed air as the oxidant, air is only 23% oxygen and storing enough air for reasonable range and speed requires air pressures over 2500 psi and consequently a heavy, high performance air flask. Two workable solutions to the oxidant problem were found before the end of WW II, the use of pure oxygen (or a mixture of oxygen and air) and the use of a concentrated solution of hydrogen peroxide in water. Each of these has been tried with varying degrees of success by several navies and HTP (high test peroxide) torpedoes are still being produced, particularly in Sweden. The US Navy experimented with pure oxygen3, but did not go very far with it. Experiments with "chemical" propulsion, that is, propulsion using energy derived from exothermic reactions, started with internal funding in 1915 at Westinghouse Electric and Manufacturing Co. and continued there with Navy funding from about 1920 until late 1926. The Navy returned to the study of "chemical" propulsion in 1929 with a program at the Naval Research Laboratory. By 1934 Navol, a concentrated solution of hydrogen peroxide in water, and alcohol became the preferred energy source. This system produced some thermal energy from the exothermic decomposition of the hydrogen peroxide, which also yielded free oxygen. Additional energy was produced by using the oxygen to burn alcohol. The first Navol or "chemical" torpedo was a converted Mk.10 which was subjected to tank dynamometer testing and ranged at Newport. It achieved a range almost three times that of a conventional Mk.10. With this success, a Mk.14 was converted and achieved an almost four fold increase in range. These results led to plans for the production of Mk.17 torpedoes as armament for new destroyers. The program was interrupted shortly after Pearl Harbor by the need to produce standard torpedoes, especially Mk.13 and Mk.14, in an attempt to satisfy urgent fleet requirements. There was no further progress until 1943 when a re-examination of the program determined that the supply of Navol was inadequate. Plans were made for a new production plant, but it was delayed and not finally started until the fall of 1944. Also in 1943 the design of the submarine launched Mk.16 Navol torpedo, with the same envelope as the Mk.14, was begun. Solid knowledge and speculation about the very long range, high speed Japanese 24" Type 93 destroyer launched torpedo4 probably fueled the development of Navol torpedoes. Several hundred each of Mk.16 and Mk.17 torpedoes were completed before the end of WW II, but neither saw use in combat. The virtues of hydrogen peroxide are that it is a liquid, over 90% oxygen by weight as compared to air which is 23% oxygen, and has a specific volume (volume per pound) about one-fifth that of 2800 psi air. In the decomposition of the peroxide,
3The USN program apparently ran for about two years, 1929-1930, and produced a power plant that was dynamometer tested. The program was discontinued in favor of "chemical" power sources. Other navies also had short lived programs, but the Imperial Japanese Navy developed and issued for service several torpedoes that used pure oxygen as the oxidant. The best known of these was the 24" Type 93, known as the Long Lance which had a range of over 29,000 y. at speeds of 48-50 k and carried 1080 lbs. of Type 97 high explosive (roughly equivalent to TNT in performance) in its warhead.
4 BuOrd OP 1507 "Japanese Underwater Ordnance" April 1945 indicates that at the time of writing only one Type 93 had been recovered by the US Navy. The Type 93 became famous as the Long Lance a name that seems to have been coined by Samuel E. Morison.
over 48% of the oxygen becomes available. Thus about 34% of the oxygen in standard Navol (70% hydrogen peroxide dissolved in water with stabilizer added) is available for combustion. Navol will provide oxygen to burn about 50% more fuel than the same weight of air. In addition the decomposition is exothermic and the heat so produced is also useful for propulsion. The water in the Navol and that produced as a decomposition product are converted to steam reducing the amount of fresh water that must be carried. Essentially the entire weight of Navol is used for propulsion. Also, Navol is a liquid and requires only about one pound of steel tankage to store one pound, whereas 2800 psi air requires about four lbs of air flask per pound of air. When all of these factors are taken into account, Navol can, for a torpedo of fixed range/speed performance and size, dramatically reduce the weight and volume devoted to fuel and oxidant. The same amount of energy as provided by a pound of alcohol, air, water and tankage can be supplied by about a quarter of a pound of alcohol, Navol, water and tankage and the volumetric saving is even greater. The weight and volume so saved can be used to increase the range and/or provide for a much larger warhead. In addition, there is no inert nitrogen, the principal component of torpedo wakes, in the fuel or oxidant. The combustion products themselves are very soluble in water and so the torpedo is practically wakeless. Unfortunately, there is a risk of uncontrolled decomposition of Navol and the attendant explosive hazard. HMS Sidon was lost in 1955 to just such an accident. The comparison between the Mk.14 and Mk.16 is shown in Table 2.
| Table 2|
CONVENTIONAL TORPEDO PERFORMANCE
Service Torpedoes in bold
|Mk.14 Mod.3A||21"x 246"||3282 lb||660 lb TPX||Steam||4,500y @ 46.3k||9.6|
|Mk.16 Mod.1||21"x 246"||3922 lb||920 lb TPX||Navol||11,000y @ 46.2k||23.4|
|Mk.18||21"x245"||3154 lb||575 lb TPX||Electric||4,000y @ 29.0k||3.4|
|Mk.26||21"x246"||3200 lb||Approx 900 lb||Seawater Bat||6,000y @ 40k||9.6|
|Mk.36 Mod.0||21"x246"||4000 lb||800 lb HBX-1||Seawater Bat||7000y @ 47k||15.5|
* Maximum speed in knots squared times range in yards at that speed times 10-6 --
a sometimes useful figure of merit for propulsion comparison.
Both a larger warhead and greater range were provided in the Mk.16 with no sacrifice of speed. Some other components of the Mk.16 differed slightly from those of the Mk.14, in particular, the turbine axis was horizontal rather than vertical and gearing consisted entirely of spur gears. High pressure air, to pressurize expendables containers and power the controls, was provided by a five cubic foot, 2800 psi air flask, a little over two feet long. Subsequent Mods. of the Mk.16 had slightly larger warheads, substantially increased range and in some cases a pattern running capability. After WW II the Mk.16 family was extended through Mod.8 and remained in use in submarines until the mid-1970's. Its performance made it a truly formidable weapon. There were occasional problems with spontaneous decomposition of the Navol, and opinions about safety differed with some individuals feeling it was too risky for submarine service. The Mk.17 destroyer torpedo was a larger version of the Mk.16. Both of the Navol torpedoes were good weapons, but their development programs were slow and erratic. One must wonder what impact they would have had especially in view of their larger warheads if they had been available in 1943 or 1944.
Electric propulsion systems have two apparent advantages: they are wakeless so they do not provide either warning of attack or indication of the location of the attacker5 and they require both less manufacturing effort (estimated for Mk.18 at 70%
5 The US Navy Operations Research Group compared the effectiveness of Mk.18 electric and Mk.14/Mk.23 steam torpedoes. The conclusion were that for attacks at ranges under 4000 y: 1) the percentage of successful attacks against enemy ships of all types except large combatants was higher for Mk.14/Mk.23 than for Mk.18. This was attributed to better lookouts in the large combatants and consequent evasive maneuvering by the target. There was no correlation between the torpedo Mark and the occurrence of counterattacks in attacks on merchantmen. In the case of attacks on large warships there were more counterattacks when Mk.14/Mk.23 torpedoes were used. Overall, it was concluded that "...if in 1944 all U.S. submarines had carried full loads of Mk.18 torpedoes the enemy would have lost about 100 fewer merchant ships ... the exclusive use of the Mk.18 would not have prevented a single U.S. submarine casualty." These comments clearly omit consideration of both morale and manufacturing. Philip M. Morse and George E. Kimball "Methods of Operations Research" New York: Technology Press and John Wiley, 1950 (An unclassified version of Vol.2A of the NDRC Division 6 Summary Report which bears the same title).
of that required for a comparable steam torpedo, Mk.14) and a lower average manufacturing skill level. These advantages are, however, purchased at the price of significantly shorter range and lower maximum speed; Mk.18 had a range of 4000 y at 29 k.6 US Navy interest in electric torpedoes began in 1915 with a project at Sperry Gyroscope Co. Successor in-house projects, again sporadic, produced designs and development models designated "EL" and Electric Torpedo Mk.1. Interest was, however, limited by the inferior speed-range characteristics of electric torpedoes. Shortly before US entry into WW II, possibly stimulated by knowledge obtained from British sources that the German navy was using electric torpedoes, work resumed on electric torpedoes. The resulting design was first designated Electric Torpedo Mk.2 (1941) and later Mk.20 (1943). Twenty of these torpedoes were eventually produced by the General Electric Co. Slow progress on the Mk.20 led to the Mk.18 project which came to be based on the German G7e and was ready for production significantly sooner than the Mk.20.
The major problems in building electric torpedoes are storing enough energy on board to give adequate range and speed and providing, within stringent weight and space constraints, a sufficiently powerful electric motor to achieve the speed. A 21" torpedo requires about 100 hp to make 30 k and well over 300 hp to make 45 k. Thus at 30 k a 5 minute (5000 y) run requires a power plant capable of delivering about 75 kilowatts of power for five minutes -- 6250 watt-hrs. Even with the inevitable losses and taking into account the rapid discharge, batteries that could deliver the required power for four to six minutes could be designed with late 1930's technology, but their weight, about 1500 lbs. or roughly half the weight of a Mk.14 torpedo, and volume, over ten feet of a 21" torpedo envelope, were serious constraints. This constraint was not significantly lifted until the advent of seawater batteries which enabled US electric torpedo speeds and ranges to exceed 35 k. and 5000 y. Severe though the battery problem was, the motor problem was even more difficult. Conventional design of a 100 hp motor might have produced a machine that would fit into a torpedo, but it would have weighed 500 to 1000 lbs. What was required was relaxation of some of the design rules. The critical point was the recognition of the fact that the torpedo motor needed to run only five or so minutes after which it was either lost or, in exercise shots, could be refurbished. Thus severe but short term heating, e.g., 100oC in 5 minutes, and sparking commutators, among other engineering anathemas, could be accepted. With these and other concessions it became possible to build motors in the 100 hp range that weighed about 250 lbs, a weight that the 21" x 21' envelope could accommodate.
The first knowledge of German electric torpedoes came from recovered fragments of the four that sank HMS Royal Oak in September 1939. Additional information was obtained from the torpedo that struck SS Volendam. The first complete German G7e torpedoes were acquired when the German submarine U- 570 was captured by the RAF on 27 August 1941. One of these was made available to the US Navy in January 1942 and other G7e torpedoes were found, at about the same time, on the East Coast US beaches. This information stimulated US Navy interest in quickly obtaining electric torpedoes. Following a preliminary meeting on 10 March 1942, Westinghouse was placed under contract to produce an electric torpedo, which, it was quickly agreed, would be an American version of the G7e. The new torpedo was designated Mk.18. This project was of course competitive with the Mk.2/Mk.20 project and so got little help from NTS-Newport. Never-the-less, in late June of
6 Note also the propulsion figure of merit given in Table 2.
1942, just fifteen weeks after starting work on the project, the first Mk.18 was delivered to Newport for testing. The testing did not go well, Newport was unhelpful if not obstructionist and production was delayed. Again, as a result of pressure from the operating forces action came from CNO/Cominch Admiral King, who ordered an Inspector General investigation on 5 April 1943. The much quoted report of that investigation, which was issued in June 1943, says in part:
The second major development, new warheads, involved the switch from TNT to Torpex as the high explosive. Torpex is a mixture rather than a pure chemical compound as TNT is. The components are TNT 41%, RDX (Cyclonite, Hexogen) 41% and aluminum powder 18%8. Torpex is attractive because of the increased explosive energy and higher detonation velocity of RDX as compared to TNT and the prolongation of the pressure wave by the aluminum. On a weight basis, Torpex is conservatively about 50% more effective than TNT as an underwater explosive against ships. Torpex is, however, more sensitive than TNT and RDX was expensive and difficult to make safely. The process of converting to Torpex torpedo warheads (and depth charge loadings) began with an order for 20 million pounds in early 19429. The first Torpex loaded warheads10 followed late the same year. The 640 pounds of Torpex in a Mk.14 warhead was at least the equivalent of 960 pounds of TNT11 almost twice the destructive power of the original Mk.14.
7 Quoted in Theodore Roscoe "United States Submarine Operations in World War II", Annapolis: US Naval Institute Press, 1949 p.262. In addition to these problems Westinghouse seems, albeit with Navy concurrence, prematurely to have turned their attention to the all electric Mk.19 and allowed the Mk.18 to languish.
8 Torpex ranges from 45% TNT, 37% RDX, 18% Al to 41% TNT, 41% RDX, 18% Al
9 Interestingly, the US Army was willing to produce cyclonite, RDX, for the Navy's use in Torpex, but was reluctant to use it for Army munitions because of safety concerns.
10 Torpex and TNT warheads were interchangeable. If there was a substantial change in weight, some adjustment to the depth gear was required.
11 Comparisons with Japanese torpedoes often neglect the difference in high explosives. Japanese torpedoes used Type 97 high explosive, which is not significantly more powerful as an underwater explosive than TNT.
The reaction of the submariners to Torpex is apparent from an entry for 19 March 1943 in the fourth war patrol report of USS Wahoo:
Other developmentsSeveral other interesting and important developments were incorporated into WW II conventional torpedo development programs. The most prominent of these were electric controls, seawater batteries and pattern running. Electric controls were standard in homing torpedoes, but the control system dynamics are different for gyroscopic course control. The Mk.18 electric torpedo, as we have noted, used pneumatic controls for several reasons: The German G7e used pneumatic controls; the reliability of pneumatic controls was well established; and there was a risk that using an electric control system might introduce instabilities that would be time consuming to resolve. The Mk.19 torpedo was a Mk.18 with an electrical proportional servomechanism for depth control and solenoid positioned vertical (course control) rudder. The Mk.19 gave way to the Mk.26 which had similar controls and a seawater battery. About twenty-five Mk.26 torpedoes were produced but large scale production was deferred in favor of the NTS-Newport and General Electric Mk.36 which was also an all electric and seawater battery powered design that was an outgrowth of the Mk.20 program and incorporated a pattern running capability. One or two developmental models of the Mk.36 torpedo may have been built, but it too was deferred in this case in favor of the Mk.42.
The seawater battery was important in that it made possible electric torpedo performance comparable with that of the Mk.14 steam torpedo. Two developmental seawater battery powered torpedoes have been included in Table 2 for comparison purposes. The seawater battery powered Mk.26 was a little slower but longer ranged than the Mk.14 and had the same propulsion figure of merit. The projected Mk.36 represented a substantial improvement over the Mk.14 and had a figure of merit exceeded only by that of the Navol Mk.16.
The basic idea of the seawater battery is to construct a primary battery using seawater as the electrolyte. With this electrolyte a magnesium anode and a silver chloride cathode make a useful 1.55 volt cell. It required some development effort to produce a satisfactory cathode--the principal problem was the high electrical resistance of silver chloride, but these problems were solved. Bell Telephone Laboratories designed and the General Electric Company built the battery for the Mk.26 torpedo. These batteries were evacuated to keep the electrodes dry before use and to provide for rapid filling when the torpedo was launched. They delivered about three times as much energy as the lead acid batteries in the Mk.18 and weighed significantly less. With this sort of performance seawater battery powered torpedoes became very competitive and, though none of those under development during WW II became service weapons, both the Mk.44 and Mk.45 post-war service torpedoes used this propulsion scheme. The consumption of expensive silver and the attendant high cost, $6000 to $8000 per unit, was an obvious drawback.
12 RADM M..H. Rindskopf letter 3 June 1996.
For completeness, we now briefly consider pattern running. The concept is to program a torpedo to make a straight run to a target rich area, for example, the middle of a convoy, and then execute a pattern hoping to hit a target. This is obviously distinct from homing although some homing torpedoes have been programmed to run a straight course and then execute a search pattern for the purpose of acquiring a target on which to home. The pattern running concept has some instinctive appeal in that it would appear to improve the probability of hitting some target. This appeal was enough to induce the German navy to mount two programs FAT and LUT13. The US Navy included pattern running in the Mk.36 and Mk.42 development programs, but neither of these entered service. Some Mods. of the Mk.16 were equipped with pattern running controls which caused the torpedo to run in circles of 300 yard radius after a straight run of preset length. Pattern running mechanisms in the days of electromechanical, as opposed to electronic, controls involved complex arrays of cams, gears and levers that were difficult and expensive to design and build. Furthermore pattern running seems to be much less effective than instincts would predict. Roessler sums up the situation in very few words "This appears unprofitable."
The remaining new non-homing torpedoes comprise the Mk.25 which was an improved Mk.13 air launched torpedo and the clearly post-WW II Mks.40 and 42. Mk.25 was a successful design that completed development late in the war. It was not produced in quantity because of huge existing stocks of Mk.13 torpedoes. Before these stocks had been consumed the anti-surface ship mission of air launched torpedoes had disappeared. The Mk.40 propulsion system was interesting in that it used a multibase solid propellant to produce gas to drive a turbine, which, in turn, drove a pump jet propulser. Such systems became important much later when targets became fast nuclear submarines and will be discussed in more detail in a subsequent part of this series. Mark 42 was an attempt to consolidate into one torpedo all that had been learned about torpedo sub-systems. The program seems to have toppled from its own weight, five organizations had significant involvement in the program, and it was abandoned in favor of a pattern running Mod. of the Mk.16. Mark 42 was, however a significant milestone in that it was the last mark assigned to a US Navy non-homing torpedo. While it does not represent a new torpedo, the large scale research and development program aimed at understanding the dynamics of air launched conventional torpedoes and improving their performance deserves note. This program, carried out mainly at Columbia University and the California Institute of Technology, developed an understanding of the air flight of torpedoes and the problems of water entry. The most visible results were frangible wooden tail extensions and nose drag rings, which were ugly, but stabilized the air flight and reduced the water entry speed. Less visible were the structural changes in the Mk.13 torpedo that were developed to accommodate the large and complex forces associated with water entry14.
In the next part of this series we will examine the radically new development of homing torpedoes during WW II.
13 FAT and LUT are discussed in Eberhard Roessler "Die Torpedos der deutschen U-Boote" Herford: Koehler, 1984 Chapter 9, pp.114-127 (In German)
14 This work is summarized in "Torpedo Studies, Volume 21 of Summary Technical Report of Division 6, NDRC" Washington: NDRC, 1946.