Photograph of Mark 15 torpedo being launched from

National Archives #80-G-413484. Cropped by author.

Torpedoes are essentially self-propelled mines. The weapon of choice for submarines, they were also employed by surface ships, such as destroyers, and by aircraft. They were the only weapon that gave aircraft much chance of sinking a battleship. Because even the smallest ships could launch torpedoes capable of seriously damaging even the largest ships, torpedoes were the great equalizers of naval combat.

The first practical torpedo was the Whitehead torpedo, invented in 1866. It was capable of carrying a 118 lb (54 kg) warhead 700 yards at 7 knots. The weapon attracted enough attention to be rapidly improved, and torpedoes had transformed naval tactics by the time of the First World War. By the time of the Second World War, the Allies had torpedoes that could carry a quarter ton of explosive 5000 yards at a speed of 45 knots, while the Japanese had fielded superb torpedoes capable of carrying half a ton of explosive 28,000 yards at 49 knots.

Epstein (2014) has argued that the race to improve torpedo technology was the genesis of the military-industrial complex. No private manufacturer was eager to take the risk of developing sophisticated torpedo technology without the guarantee of a market, and this led to increasingly close collaboration between manufacturers and navies.

Torpedo Characteristics

A torpedo must have a means of propulsion yielding the maximum feasible speed and endurance, ideally without producing much of a wake; a depth keeping mechanism allowing the torpedo to strike the target at the ideal depth; a guidance mechanism to keep the torpedo on the desired course; a powerful warhead; and an exploder to detonate the warhead when the torpedo reaches its target. These mechanisms together make a torpedo a complex and expensive weapon, requiring considerable time and resources to develop and produce. Even the relatively primitive torpedoes of the First World War had over 500 parts in their guidance systems alone. Such weapons required extensive testing to achieve acceptable reliability, but the high cost of torpedoes meant that Britain, Germany, and the United States went to war with torpedoes that had not been adequately tested, and it took the United States Navy a scandalous two years after war broke out to work the bugs out of its torpedo designs. Only the Japanese went to war with reliable torpedoes.

Propulsion. Early torpedoes used a large flywheel or compressed air to provide power for propulsion. Compressed air was stored in a large bottle, called the air flask, that took up most of the space within the torpedo. This needed high tensile strength to hold air at extreme pressure and toughness to ensure that the flask would tear rather than shatter if hit during combat. Thus, the air flask was typically constructed from high-quality nickel steel. The flow of compressed air was regulated with a reducer, to ensure that the torpedo maintained a uniform speed, essential for effective fire control. Later designs greatly increased range and speed by using the compressed air to burn kerosene or alcohol. The resulting high-pressure gas was directed into a compact engine that drove the propellers. However, the heat produced by combustion was too much for the engine to endure for long, and by the time of the First World War, most torpedoes were "wet heater" designs that injected water into the combustion chamber to reduce the temperature while increasing the volume of gas. Further cooling was provided in some designs, such as the U.S. Mark 14, by making the midsection of the torpedo open to the sea. The U.S. Navy standardized early on a two-stage impulse turbine engine for all its torpedoes, while the British and Japanese preferred various configurations of piston engines. All the powers experimented with pure oxygen in place of air, but only Japan persevered to bring pure oxygen torpedo designs into operational use. The other major powers experimented with hydrogen peroxide fuel, which, while highly promising, did not result in operational designs in time for the war.

An alternative to wet heater engines was an electric motor driven by a large bank of batteries. Though electric motors could not be made as powerful as wet heater engines, and thus yielded a slower torpedo, they had the considerable advantage of producing virtually no wake. Electric torpedoes were also easier to construct. One design challenge was to ensure that the torpedo maintained a uniform speed, since the battery output was sensitive to temperature and decreased as the batteries ran down during the torpedo run. Both the Japanese and the Americans reduced the temperature variability by preheating their torpedoes before launch, and the American designers put considerable effort into developing a governor to maintain constant speed at varying battery voltages. Germany, Japan and the United States all fielded electric torpedoes for their submarine forces.

A single propeller produced torque that tended to rotate the torpedo around its long axis. Most torpedoes of the Second World War eliminated this problem by using twin counter-rotating propellers based on the Davison design.

Depth keeping was accomplished using a hydrostatic sensor stabilized by a pendulum and coupled to control vanes. This required a somewhat delicate adjustment of the vanes, and all the major powers experienced problems with depth keeping with their prewar torpedo designs. Only the Japanese recognized the problem and took corrective measures before war broke out. The Americans were very slow to realize that their torpedoes were running too deep, in part because their torpedoes were designed to run under the target and be detonated by a magnetic proximity exploder. It was not until well into the war that proper tests using target nets were conducted under Lockwood in Australia, which demonstrated that the American torpedoes ran about 11 feet (4 m) deeper than their nominal setting. Part of the difficulty was that the Americans tested their torpedoes using dummy practice heads that were significantly lighter than combat warheads, to ensure the expensive torpedoes would have enough buoyancy to be recovered after testing. This threw off the torpedo trim more than was realized. Another part of the difficulty was poor placement of the hydrostatic sensor at the rear of the torpedo body, where hydrodynamic flow reduced the apparent depth reading. This required relocation of the sensor.

Directional control. Directional control was provided by a sensitive and delicate gyro guidance system, while rotational control was provided by pendulums. This was the one part of the torpedo that proved reliable in wartime, since any design flaws in the directional guidance were immediately obvious even with limited testing. The gyroscope was usually kept spinning at a high rate using a jet of compressed air. Some torpedoes could be programmed to execute simple course changes, though this was limited to a short turn immediately after leaving the torpedo tube in most submarine torpedoes. Guided torpedoes came into use later in the war in the form of the acoustic homing torpedoes produced by the Germans and Allies. These followed a programmed course until they detected a source of propeller noise, at which point a set of acoustic sensors steered the torpedo towards the source of noise. Wire-guided torpedoes were experimented with but not operationally deployed before the end of the war.

Warheads varied greatly in explosive power, with the Japanese having the most powerful warhead designs, in part because they had standardized on 24" (61cm) torpedoes while the Allies continued to use 21" (53.3 cm) torpedoes. Japanese torpedoes used Type 97 explosive, a mixture of 60% TNT and 40% hexanitrodiphenylamine pioneered by the German navy in 1907 that was highly insensitive. The Allies switched from pure TNT to Torpex, a mixture of RDX, TNT, and powdered aluminum whose explosive characteristics were ideal for use in mines and torpedoes. 

U.S. Navy Mark 6 Mod 1 exploder
Mark 6 Mod 1 Exploder. U.S. Navy. Via Wildenberg and Polmar (2010)

Exploders were another source of difficulty. All torpedoes were equipped with inertial contact exploders that used some arrangement of a heavy weight that was slammed into a firing pin by its own inertia when the torpedo hit the target. These ought to have been simple and reliable, but both the Germans and the Americans discovered that their contact exploders were not robust enough to withstand the heavy impact and jammed rather than hitting the firing pin. The Germans quickly corrected the problem; the Americans took much longer to acknowledge that the problem existed and to produce a reliable contact exploder. Part of the difficulty was that both powers regarded the contact exploder as secondary to a magnetic proximity exploder. The Japanese made little effort to produce a magnetic exploder and put greater effort into developing a more reliable contact exploder.

The magnetic exploders developed by the Germans and Americans were designed to detect the magnetic field of a steel ship and detonate under the keel, breaking the back of the ship. Both powers found that their magnetic exploders were highly unreliable in practice, and abandoned them in favor of contact exploders — though much too reluctantly in the case of the Americans. The magnetic field of a ship proved too difficult to characterize adequately, particularly over a large range of magnetic latitude.

American Torpedoes

The U.S. Navy had set up its own Torpedo Station prior to the First World War, and it worked closely with its principal torpedo manufacturer, E.W. Bliss Company (which had licensed manufacturing rights for the original Whitehead torpedo) to improve torpedo designs. The free exchange of ideas between the Navy and Bliss eventually soured over intellectual property rights and concerns about national security, culminating in two lawsuits that the Navy successfully took all the way to the Supreme Court. Thereafter the Navy designed and manufactured its own torpedoes until the Second World War, when electric torpedo manufacture was contracted out to General Electric and Westinghouse.

The U.S. Navy had standardized its torpedo designs to use wet heater propulsion based on alcohol fuel and a two-stage turbine engine. The two stages were designed to rotate in opposite directions to prevent unbalanced torque that might rotate the torpedo. The choice of turbine engine had originally been driven by the need for high temperature tolerance in dry heater torpedoes, and was retained even after wet heaters were introduced. The Navy experimented with pure oxygen in place of air in their torpedoes, but abandoned this in favor of Navol, highly concentrated hydrogen peroxide. Navol is a liquid that decomposes into steam and oxygen in the presence of a catalyst, and a flask containing Navol in place of compressed air or even pure oxygen could produce a much greater volume of working fluid for the torpedo engine, yielding unprecedented range. However, Navol is a powerful oxidizing agent, difficult to handle safely, and the Navy was forced to drop development of the Navol torpedo in 1941 in order to ramp up conventional torpedo production. Development was resumed late in the war, but no units reached the fleet before the Japanese surrender.

The most distinctive feature of American torpedoes was the Mark 6 magnetic exploder. The U.S. Navy had officially eschewed the use of submarines against commerce and instead emphasized their use in a fleet engagement. This required a torpedo that could overcome the underwater protection of heavily armored warships. The Navy concluded that the best approach was to avoid the side protection completely and design a torpedo that would detonate  directly under the keel of a ship, where it would break the ship’s back. Proposals for electric proximity fusing, for a water kite streamed above the torpedo, and even for a torpedo that dove under the ship after it slammed into its side looked less promising than a magnetic proximity fuse similar to those developed for mines during the First World War. Hundreds of measurements of the magnetic fields of Navy warships were carried out and used to design detector coils that would fire at the proper moment. The complexity of the mechanism is illustrated by Newpower's (2006) simplified description:

The exploder mechanism received its power from the generator which itself received power from the rotation of an impeller in the base of the mechanism. This impeller turned as a result of the torpedo's motion through the water. The generator powered the electrical components inside the Mk. 6. The core rod and the pickup coil acted, collectively, as the exploder's eyes and ears, responding to changes in the direction or intensity of the earth's magnetic field. The designers tuned the pickup coil's sensitivity so that a ship's steel hull created sufficient perturbations to induce a voltage on a triode in the grid circuit. this enabled the flow of current in the grid circuit to a large thyratron tube. At a preset voltage, the tube acted as a switch, closing a circuit to a solenoid valve. The solenoid lifted a mechanical arm attached to the same impeller that powered the generator. A lever connected to the pawl then made contact with the firing ring, a circular metallic enclosure around the exploder mechanism. This is turn triggered the exploder. All of this happened within a fraction of a second. Even this simplified description speaks to the complexity of the Mk. 6's design.

The design team under Christie fired over 100 exercise shots in equatorial waters, using a photocell to verify that the exploder activated precisely under the target hull and taking some 7000 magnetic field readings. However, frugal budgets and concerns for security meant that the Mark 14 torpedo equipped with the Mark 6 exploder and an actual combat warhead was live-tested exactly twice before being issued to the fleet. One of the tests failed, which ought to have raised a red flag, but the Navy was reluctant to fund further live-firing tests.

The high cost of torpedoes played a major role in this debacle. The torpedo manufacturing process at Newport Torpedo Station more closely resembled hand crafting than modern assembly line production, and unit costs were correspondingly high. The relatively small numbers of torpedoes that could be manufactured and stocked argued for a torpedo that did as much damage as possible so that there was no need to fire large spreads.

The upshot was that the United States entered the war with highly unreliable torpedoes. It was not just a matter of the Mark 6 proving defective. Because the Navy was counting on the Mark 6 to detonate the warhead under the target keel, the Navy always set its torpedoes to run under the target, and this concealed the fact that the torpedoes were running too deep. The dummy practice warheads were kept light to ensure that the expensive torpedoes could be recovered after being test fired, and it was not realized how much the heavier combat warhead threw off the trim adjustment. In addition, the hydrostatic sensor for depth control was poorly placed, on the tail cone of the torpedo, where hydrodynamic effects at high speed settings (preferred by the combat fleet) gave a spuriously low depth reading. Finally, because of the emphasis on the Mark 6 exploder, the backup contact exploder was also inadequately tested.

Correcting the faults of American torpedoes was bound to be frustrating, since after each defect was recognized and corrected, another defect appeared. But the problem was further aggravated by institutional arrogance. The monopoly on torpedoes held by Newport, backed by the Maine congressional delegation, was one cause. In addition, both force commanders and the Torpedo Station were happy to place the blame for faulty torpedoes on ship and submarine commanders and crews. Innovations such as those of Paul Schratz of Scorpion, who claimed that tightening the packing glands on the impeller of the Mark 6 eliminated premature detonations, never made it through channels to the rest of the fleet. It took the Navy two years to officially recognize that their torpedoes ran eleven feet (over 3 meters) too deep, that the magnetic exploder was almost useless anywhere but the North Atlantic (and probably there as well), and that the contact exploder usually failed on normal impacts. In one notorious incident, the commander of submarine Tinosa crippled a large Japanese whaler, Tonan Maru #3, with two spreads of six torpedoes, then carefully squared off his boat and fired no less that nine additional torpedoes at a theoretically perfect angle of impact at the theoretically perfect range. Not one detonated. The problems with the Mark 14 were finally resolved by 1944, and U.S. submarines began to take a crippling toll of Japanese warships and merchantmen; but the Mark 14 never became a great torpedo.

One reason why the Mark 6 magnetic detonator was retained for so long was that the Mark 14 had a relatively small warhead, little over half the weight of the Long Lance warhead. This was inadequate against the armor belts of capital ships, and could not even guarantee the sinking of a large merchantman. It was felt that the torpedo had to explode under the keel of the target to ensure destruction. The firing of large spreads of torpedoes was seemingly ruled out by the high costs of torpedo production. Occasional successes with the magnetic exploder, such as those reported by "Mush" Morton on his highly successful patrol with Wahoo in January 1943, created impetus to try to improve the reliability of the exploder rather than abandon it. The opinion of Bureau of Ships that detonations under the hull were no more effective than contact detonations, except for targets with torpedo protection systems, was one of the final nails in the coffin of the Mark 6.

The American Mark 15 destroyer torpedo suffered from the same deficiencies as the Mark 14 submarine torpedo. It was launched from tube mounts of from three to five tubes, which were trained as a unit but could be individually fired. The mount crew typically consisted of a mount captain, gyro setter, and mount trainer, who could fire on local control if director control failed.

The American aircraft torpedo, the Mark 13, was much inferior to its Japanese counterpart at the start of the war, requiring a very low-speed, low-altitude drop that was nearly suicidal for the attacking aircraft. Later versions were much better, being equipped with a nose drag ring to reduce the rate of fall and a box air tail to keep the torpedo from oscillating in flight. This allowed the torpedo to be dropped from higher altitude and at a higher speed.

Chart of torpedo production

U.S. Navy

Not only did the Americans suffer from poor torpedo quality early in the war; their torpedo production facilities were inadequate to wartime demand. Some 300 torpedoes were lost when the Japanese bombed Cavite, which was a substantial fraction of the Navy's entire reserve. American submarines occasionally went to sea armed with mines instead of torpedoes, and were instructed to fire only one or two torpedoes against merchant targets. Fortunately for the Americans, production ramped up later in the war, ending the shortage. Production peaked in early 1944 at about 2000 torpedoes per month of all types. 

With the Torpedo Station fully occupied manufacturing conventional torpedoes, new torpedo designs were largely left to the National Defense Research Committee and various private contractors. The first new wartime torpedo was the Mark 18, an electric torpedo based loosely on the German G7e, examples of which were captured by the British in August 1941. This torpedo was slower than the Mark 14, but it was easier to produce and left no wake. The Mark 18 became popular with submarine commanders out of proportion to its actual capabilities and accomplishments, which included a lower hit percentage than that of the debugged Mark 14.

However, the Mark 18 became the basis for the first acoustic homing torpedoes. The Mark 24 "Fido" homing torpedo was deadly effective against German submarines, and was credited with sinking five Japanese submarines. A submarine version, the Mark 27, was used only in the Pacific. This torpedo was slow and carried a very small charge, but it was very effective in its intended role as an anti-escort torpedo. The torpedo would home in on and destroy the screws of an escort vessel, rendering it hors de combat if it was not sunk outright.

Japanese Torpedoes

Japanese torpedoes were superb. The Japanese were employing 24” (610 mm) torpedoes on their surface ships at the start of the war, versus 21” (530 mm) torpedoes for the Western powers. This reflected the Japanese determination to make their warships, whose numbers were limited by the naval disarmament treaties and Japan's own limited production capacity, superior on a unit-for-unit basis to Western warships. The Japanese also persevered in the development of pure oxygen propulsion after the Western powers had abandoned the idea. Ironically, this was driven at least in part by mistaken intelligence that the British had successfully fielded pure oxygen torpedoes. The Japanese were also more willing to conduct dangerous experiments and accept the resulting casualties, an attitude that was also reflected in their highly realistic fleet exercises. 

The Japanese found that the source of accidental explosions in their pure oxygen torpedoes was compression heating of the oxygen in sharp bends in tubes contaminated with traces of machine oil. The heated oxygen reacted explosively with the oil traces. The solution was to eliminate all sharp bends in the oxygen tubing and to thoroughly clean the interiors of the tubes with a strong alkaline solution. The Japanes also carried out the extensive live-fire testing denied to Western navies on tight budgets, and by the time war broke out in the Pacific, the Japanese oxygen-powered torpedoes were highly reliable and reasonable safe to handle. All used kerosene as their fuel and were powered by Whitehead two-cylinder piston engines.

The best Japanese 24” pure oxygen torpedo, the Long Lance, had a half-ton warhead and had the incredible range of 30 miles (50 km) at 36 knots. The Japanese 21" submarine torpedo, the Type 95, which was essentially a smaller version of the Long Lance, was also superb. The Japanese aerial torpedo, the Type 91, was much more rugged than its American counterpart, giving the Japanese a huge advantage in torpedo bomber effectiveness early in the war.

The Japanese also produced an electric torpedo for their submarine forces, the Type 92, but this offered no particular advantage over the oxygen-powered Type 91 except ease of manufacture.

Torpedo tactics played an important role in Japanese Decisive Battle Doctrine. In the event of war with the United States, the Japanese anticipated a 3:2 advantage in battleships in favor of the Americans, as established by the naval disarmament treaties. Under the square law of combat effectiveness, this all but guaranteed an American victory in the final decisive battle. The Japanese therefore put heavy emphasis on attrition tactics (zengen sakusen) to weaken the American battle line during its long voyage across the Pacific. These were based on night torpedo attack by destroyers and cruisers, and, as they became more capable, aircraft. This doctrine emphasized development of torpedoes that outranged their enemy counterparts.

The Great Decisive Battle never took place (although Midway came close, ironically ending in Japanese defeat.) The decisive campaign of the war was instead fought in the South Pacific, mostly in the Solomon Islands, and included numerous surface night engagements. The Japanese had a decided advantage early in the campaign, using their Long Lance torpedoes to deadly effect. Even when ambushed, the Japanese often prevailed, since their almost instinctive response was to launch shoals of torpedoes in the general direction of the American force. However, the Americans gradually gained control of the air, and learned from hard experience how to use their radar and improved tactics to neutralize the Japanese torpedo advantage.

Photograph of last Mark 14 torpedo produced during the war

Naval History and Heritage Command #NH 94117

British Torpedoes

British torpedoes were intermediate in quality between the Americans and the Japanese. British torpedoes were wet heater designs using kerosene fuel and a compact four-cylinder radial engine. The Mark VIII torpedo had been in service since September 1932 and was thoroughly debugged and highly reliable by the time war broke out in the Pacific. Initial depth keeping problems had long been ironed out and early experiments with magnetic fusing had been abandoned in favor of a completely reliable contact detonator. However, British torpedoes had nothing like the range of the Long Lance.

Torpedo Damage

Morison quotes the engineering officer of Hornet on the effects of a torpedo hit:

A sickly green flash momentarily lighted the scullery compartment and seemed to run both forward towards Repair Station 5 and aft into the scullery compartment for a distance of about 50 feet. This was preceded by a thud so deceptive as to almost make one believe that the torpedo had struck the port side. Immediately following the flash a hissing sound as of escaping air was heard followed by a dull rumbling noise. The deck on the port side seemed to crack open and a geyser of fuel oil which quickly reached a depth of two feet swept all personnel at Repair 5 off their feet and flung them headlong down the sloping decks of the compartment to the starboard side. Floundering around in the fuel oil, all somehow regained their feet and a hand chain was formed to the two-way ladder and escape scuttle leading from the third deck to the second deck....

The damage from a torpedo was not limited to the hole blasted in the side of the ship and the resulting flooding. Shock from the explosion could damage equipment located far from the point of impact. A 1943 British study examined nineteen cases of torpedo damage to cruisers and found that the average repair time was 9.5 months.

Japanese prewar estimates were that four or five torpedo hits would sink a battleship and three or four hits would sink a carrier. Actual war experience showed a considerably wider variation in outcomes, with Taiho succumbing to a single torpedo due to inept damage control, while Hornet remained afloat after an incredible ten torpedo hits from Japanese torpedo bombers and the carrier's own escorts (which were attempting to scuttle her.)

Japanese torpedoes

Long Lance
Type 89
Type 91
Type 92

Type 93 "Long Lance"
Type 95
Type 97

Type 98

U.S. torpedoes

Mark 8
Mark 10
Mark 13
Mark 14

Mark 15
Mark 18

Mark 24

Mark 27

British torpedoes

Mark IX
Mark XV


Blair (1975)

Branfill-Cook (2014)

Campbell (1985)

Epstein (2014)

Evans and Peattie (1997)

Friedman (2013)

Morison (1949, 1951)

Newpower (2006)

Ordnance Pamphlet #946 (1943-4; accessed 2015-1-17)

Roscoe (1953)

Wildenberg and Polmar (2010)

Zimm (2011)

Valid HTML 4.01 Transitional