Part Four: WWII Development of Homing Torpedoes 1940-1946
Important as the WWII improvements in conventional torpedoes were, the real revolution was in the development of homing torpedoes, i.e., torpedoes which autonomously seek their targets at least during the final portions of their trajectories. The exact date when the homing concept first occurred to torpedo developers is lost, but the general idea was discussed early in the 20th century when torpedo ranges got long enough that very accurate aiming was required and relatively small angular dispersion could cause misses. Not, however, until the mid-1930s, when electronic technology provided the means for implementing the concept, was it possible to begin serious development of homing torpedoes. Programs were initiated by the German Navy in the mid-1930s and by the Royal Navy in the late 1930s. The German program suffered a hiatus from 1939 to 1942 because the expectation of a short war lowered its priority, but two torpedo types for U-boat use against surface vessels were produced during 1943. Royal Navy results, mainly dealing with acoustics, were not pursued, but were made available to the U.S. Navy. U.S. programs, as we shall relate, began in December 1941 and produced an air launched anti-submarine torpedo that entered service and sank submarines 17 months later, in May 1943. Several other important homing torpedoes were developed for the U.S. Navy before the end of the war and two of these were used against enemy targets.
Homing torpedoes are dramatically different from the gyrocontrolled, set-depth torpedoes used against surface ships in that once they acquire their target, they home on it autonomously using onboard controls. In addition to the obvious advantage of homing in the horizontal plane in attacking surface targets, homing can operate in the vertical plane thus providing an important capability against submerged submarines or shallow draft escorts. The homing concept is obviously very attractive, so attractive in fact that only one new non-homing torpedo has entered service with the U.S. Navy since 1944 and that was the wire guided Mk 45 to which special constraints applied.
A successful homing torpedo must:
- detect the target and indicate its direction relative to the torpedo axis
- process this directional information to generate orders to the vertical and horizontal rudders
- be provided with propulsion machinery and other mechanisms that do not interfere with the homing system
- be provided with adequate safety features to prevent attacking the launch platform or other friendly forces
- be sufficiently rugged to withstand launching, water entry and other challenges inherent in its use.
Rational analyses of target signatures and probes that might provide information about target location for use in homing torpedoes have been made many times. The result, even today, is invariably that the best, and possibly the only practical, possibilities are acoustic. Target detection and tracking using underwater sound had, of course, been developed during the interwar years for surface vessel anti-submarine purposes and for defensive and offensive use by submarines. These sonar systems were of two types, passive, which simply listened for noise generated by the target, and active, which detected the reflection or echo of a probing sound pulse emitted by the system. Such shipboard systems provided starting points for torpedo homing systems, but their size and weight were both much too large for torpedoes. Developing equipment that satisfied the size, weight and performance constraints associated with installation in a torpedo body was a challenging task. The first U.S. homing torpedoes used passive systems that detected ship noise, primarily cavitation noise from the screws. The directivity needed to generate homing rudder orders was provided either by mounting the hydrophones around the circumference of the torpedo and using body shadow and hydrophone directivity to provide direcivity or by mounting an array of hydrophones in the nose of the torpedo and relying primarily on hydrophone directivity. Soon after development of passive homing began, U.S. work was started on active homing based on a miniature active sonar. The problems associated with fitting an entire sonar system, using vacuum tube technology. into a torpedo body while leaving room for the propulsion system and a meaningful warhead were very severe. It was, in fact, not until early 1944 that the first active homing torpedo made a three dimensional acoustically controlled run. Ultimately, however, acoustic torpedoes incorporated passive homing for target acquisition and active homing for the attack phase.
Detecting a target and indicating its direction are not enough. This information must be converted to rudder motions that will direct the torpedo to the target. Conceptually this is relatively simple. In the case of passive homing, amplified signals from say the left and right hydrophones can be compared and the control circuits arranged to move the rudders to steer in the direction of the stronger signal. A similar, but slightly more complicated, system can be used for control in the vertical plane. This approach was used in the Mk 24 torpedo, also known as FIDO, discussed below. Simple as the process sounds, there were many problems that were important in these early days of electronics. For example balancing the left and right amplifiers was enough of a problem that the early systems used a single amplifier, which was switched back and forth between the left and right channels. Stability of the control system also required study. In 1942 these were problems at the cutting edge of engineering technology. That they were solved expeditiously in the face of similar demands for communications, radar, sonar, fire control and nuclear weapons, to mention some of the competitors for electronic development was a tremendous triumph.
An acoustic homing system can work only if the torpedo is quiet enough that its self noise does not mask the noise or echo that is the target signal. This means minimizing both the hydrodynamic noise, especially that originating in cavitation, and the propulsion machinery noise. These issues and the constraints of electrical propulsion, which was used with most WWII homing torpedoes, led to rather slow, short range torpedoes, in many cases so slow that they were effective only against submerged submarines or slow moving actively searching escorts.
As with conventional torpedoes, there were, during WWII, three launch platforms for acoustic torpedoes, aircraft, submarines and surface vessels, and two classes of targets, surface vessels and submarines. These platform-target combinations impose constraints or design requirements on homing torpedoes that are not operative, or at least much less important, in the case of conventional torpedoes. The major new safety requirement was that the torpedo should not home on the launching platfonn2 or other friendly vessel. This requirement was satisfied in a variety of ways. To protect surface vessels, ceiling switches disabled the homing system of air launched weapons when the depth was less that a preset value, say 40 feet. Floor switches similarly protected submerged submarines from their own anti-escort torpedoes. Straight enabling runs to the vicinity of the target; anti-circular run devices and other safety features were also added to some of these new torpedoes. Further, during WWD Allied aircraft did not drop homing torpedoes when operating in conjunction with surface ASW forces. Incidents did, however, occur. HMS BITER was chased by a homing torpedo giving rise to the doggerel “BITER bitten by FIDO.”
U.S. Navy Homing Torpedo Development During WWII-An Overview
The development of homing torpedoes during WWII was done almost entirely under the auspices of the Office of Scientific Research and Development (OSRD) and its subsidiary the National Defense Research Committee (NDRC). Wartime production of homing torpedoes was accomplished by standard Buord procurement contracts with industrial firms, primarily Western Electric, Westinghouse and General Electric. Major research and development contracts were issued under the authority of the Office of Emergency Management (OEM) to Harvard University, Western Electric Co. (Bell Telephone Laboratories), General Electric Co. and Westinghouse Electric Corporation with smaller contracts to other universities and commercial firms. Many subcontractors
worked for the major contractors on special aspects of torpedoes. Each of the major contractors and Brush Development Co-developed one or more homing torpedoes through the prototype stage. In several cases two contractors developed competing models designated by the same Mark, for example, the Bell Telephone Laboratories (BTL) and the Harvard Underwater Sound Laboratory (HUSL) developed competing versions of the Mk 24 and HUSL and GE developed competing versions of the Mk 32. In other cases competing torpedoes had different Marks. (The Brush Mk 30, for example, was developed, as a backup, in parallel with the Mk 24.) Thus, there was significant competition, but also a great deal of cooperation. This combination helped to produce the first operational U.S. homing torpedo in the remarkably short time of 17 months from initial concept to first combat success. One estimate suggests that the competition saved a full year in the development cycle.
Homing torpedoes developed along two lines: torpedoes based on straight runners (primarily Mk 13, Mk 18 and Mk 19) with standard 21 inches x 246 inches or 22.S inches x 161 inches envelopes and smaller torpedoes with 10 inch or 19 inch diameter envelopes seven to eight feet in length. The principal technologies that were newly incorporated to make homing torpedoes were underwater acoustics (hydrophones); hydrodynamic and mechanical quieting; electronic controls and servomechanisms. Though such items are commonplace today, in the early 1940s they were revolutionary.
The number of torpedoes under development was large as indicated by Table 1, but only three, Mk 24, Mk 27 and Mk 28, saw service during WWII. All but one, Mk 21 Mod 2 (a homing version of Mk 13), used electronic propulsion and this was the dominant mode of propulsion for new U.S. Navy homing torpedoes until high submerged speed nuclear submarines forced a return to thermal, albeit advanced thermal, propulsion in the Cold War era.
Selected U.S. Navy Homing Torpedoes-WWII Era
Among the acoustic torpedoes developed during WWII there were two that represented critical milestones. The MK 24 was the first passive homing torpedo developed for the U.S. Navy and the Mk 32 was the first active homing torpedo. The Mk 35 was the first active-passive homing torpedo and it was based on research and development started during WWII. The actual Mk 35 torpedo development program seems to have begun quite late in the war and more properly belongs to the post WWII era. We will focus here on the Mk 24 and Mk 32 torpedoes and comment briefly on some of the others the new homing torpedoes was a response to the damage being done to Allied shipping by German U-boats. From the beginning of WWII through 1941 Allied shipping losses to submarines averaged over 170,000 tons/month and aircraft were proving to be remarkably ineffective in destroying submarines.’ One consequence was that even before the U.S. entered WWII, parts of the Navy were reconsidering homing torpedoes as air launched ASW weapons. In “the fall of 1941″ (probably late November or early December), the Navy asked NDRC to consider the feasibility of a small, relatively slow-speed, acoustically controlled, air launched, anti-submarine torpedo. 6 Submarines were thus specifically added to the torpedo target list rather than being incidentally included when surfaced or at periscope depth as Engineering and Science in the Bell System: National Service in War and Peace (1925-1975)”, Murray Hill: Bell Telephone Laboratories, 1978 contains some information that i.1 not included in Gardner’s paper. These publications focus on the BTL/Wcstem Electric projects, but clearly indicate that important contributions were made by other organizations. More recent is Tom Pellet “FIDOThe First U.S. Homing Torpedo”, The Submarine Review, January 1996 and correspondence by Milford and Polmar in the April 1996 issue of The Submarine Review. Robert Gannon “Hellions of the Deep” University Parle, PA: PcM State University press, 1996 tells more of the Harvard story. The primary documentation is contained in reports submitted to NDRC by HUSL and BTL/WE.
In response to the Navy request NDRC convened a meeting at Harvard on 10 December 1941. Two weeks later at a second meeting the following requirements were outlined:
- size to fit 100 pound bomb rack, i.e., smaller than 19 inches x 90 inches
- droppable from 200 to 300 feet at about 120 knots
- electric propulsion using lead acid storage battery
- 12 knots for 5 to 15 minutes
- 100 pound high explosive charge
- acoustic homing with greatest possible range
The participants in the meeting responded as follows: General Electric agreed to design and fabricate the propulsion and steering motors. David Taylor Model Basin (DTMB) would assist in any way possible, primarily hydrodynamics and propulsion. DTMB actually supplied the propeller and shell designs and the first few actual shells used in the Mk 24 program. HUSL and BTL each undertook the independent, but cooperative and information sharing, development of experimental torpedoes with their main contributions being acoustic control systems and integration. The entire project proceeded very rapidly. Some of the key events in the development of Mine Mk 24′ (FIDO), are shown in the almost unbelievable schedule which follows.
The entire development from conception to first kill was accomplished during the general time period in which the previously described Mk 14 problems were solved. The contrast in the rate of progress on the two problems is striking. Mk 24 also established the four hydrophone acoustic sensor arrangements that were the dominant passive homing system for U.S. acoustic torpedoes in the period 1941-1950.
The Mk 24 that emerged was 84 inches long, 19 inches in diameter and had a total weight of 680 pounds. It was propelled by a General Electric five horsepower, 48 volt electric motor using an Exide lead acid storage battery for power. The warhead, containing 92 pounds of high explosive, occupied the forward 14- 1 /2 inches of the weapon. These features were substantially different from those of early torpedoes, but more significant differences were to be found in the control system.
Target detection was accomplished by four hydrophones symmetrically arranged around the circumference of the torpedo mid-section in the left, right, up and down positions. Such an array is useful for target acquisition because the four hydrophones together cover essentially all directions from the torpedo and for homing because body shadow, meaning that the hydrophone on the right side, for example, being in the acoustic shadow of the torpedo body could not hear a target on the left side, provides directionality. The basic idea is to compare the signals from the left and right hydrophones and move the rudder in such a way as to steer towards the stronger signal. In the BTL implementation of this scheme, the hydrophone signals were amplified, rectified and subtracted. The combined signal drove a DC amplifier which, in turn, controlled a differential relay that caused the rudder motor to move in the appropriate direction to reduce the input voltage (hydrophone derived voltage plus rudder potentiometer voltage) to zero. The vertical control circuit was identical except for including inputs from a hydrostat that measured depth and a pitch pendulum, which were also voltages derived from potentiometers.
These signals caused the torpedo to operate at a fixed depth until a sufficiently strong acoustic signal was received. When such a signal was detected, the hydrostat/pendulum control reestablished if the torpedo rose above a ceiling set at about 40 feet. This prevented the torpedo from attacking surface vessels including surfaced submarines. These control systems produced rudder angles that were proportional to the difference in strength between the signals from the right and left (or up and down) hydrophones Such proportional control was distinctly different from the bang· bang (rudder hard left or hard right) controls that had been uses ever since the Obry gyro was introduced, but detailed analysis and experimental work at HUSL showed that the bang-bang (no ruddes position feedback) controls would perform equally well.
The Mk 24 development program was notable not only because of the speed with which it was completed, but also because of the thorough development testing and subsequent quality control. During subsystem development there was a continuing series of tests to measure and verify essential performance characteristics. Testing included drop tests, checking fitting to aircraft and occasional drops from aircraft in addition to the usual laboratory testing of the mechanical, electrical and electronic designs. BTL alone conducted 192 in-water test runs with their experimental models between 16 April and 20 October 1942 and a comparable number of tests was conducted by BUSL. Later, BUSL conducted an extensive series of tests on Western Electric production torpedoes dropped by PBY aircraft.
Both the BUSL and the BTL programs produced successful prototypes. The BTL Mk 24 production design, which started from the BTL experimental model, used important features from the BUSL model and incorporated a number of improvements suggested by development testing. The design was frozen in October 1942. At that time Western Electric was given a sole source contract for production of the torpedoes. Subcontractors included General Electric, Electric Storage Battery Co.. and interestingly enough, a bathtub manufacturer for the shells. The first production model was delivered in March 1943 and 500 had been delivered by May 1943. The first U-boat attack using the Mk 24 was U-640 which was attacked and sunk on 14 May 1943 by a PBY from U.S. Navy VP 84.’ The Mk 24 was eventually responsible for sinking 37 enemy submarines, 10 about I5 percent of the submarines sunk by air escort or air ASW operations between May 1943 and the end of the war. This torpedo was a major success whose achievements have long gone unheralded.
Reflecting the perceived urgency of the requirement for an air dropped, homing ASW weapon, another passive homing torpedo, Mk 30, was developed by Brush Development Co. under a BuOrd contract as a backup for the Mk 24. This 10 inch diameter torpedo progressed through the successful prototype stage, but because of the success of the Mk 24 it was never put in service. It was, however, a precursor to the active homing Mk 43 Mods 1 and 3 which were in service from 1951 to 1957.
Two other passive homing torpedoes saw service in WWII. The Mk 27 torpedo was a submarine launched anti-escort weapon based on the Mk 24. The original Mk 27 Mod 0 was a minimally modified Mk 24 with wooden rails to fit 21 inch torpedo tubes, a floor switch (instead of a ceiling switch) so it would not attack the launching submarine, and various arming, warm-up and starting controls to suit a torpedo tube, swim-out launch mode. Eleven hundred Mk 27 Mod 0 torpedoes, known as CUTIE, were built by Western Electric and delivered between June 1944 and April 1945. Production on a subsequent order for 2300 torpedoes continued until the end of the war. One hundred and six were fired against enemy escorts. Thirty-three hits sank 24 ships and damaged nine others. Later versions of the Mk 27 were longer and heavier. Mod 3, which was slightly over 10 feet long and faster, had a 200 pound warhead and a gyro for straight run out before beginning to search for its quarry. Only six were completed before the project terminated at the end of the war. The post war Mk 27 Mod 4 was different from the wartime versions, especially in that it could attack submerged submarines, and is discussed in the next part of this series. The Mk 28 was a 21 inches x 246 inches, 20 knot, submarine launched anti-surface vessel torpedo with a 585 pound warhead. It was equipped with passive homing and gyroscopic control which competed for rudder control. About 1750 of these torpedoes were produced by Westinghouse and Western Electric. Only 14 were fired with four hits during WWII, but the torpedo remained in service until 1960.
The remaining passive homing torpedoes developed during WWII were generally and perhaps surprisingly successful, but were overshadowed by earlier successes or reached production readiness too late in the war to be used. Some of these programs did, however, influence post war torpedoes. The Mk 29, in particular, was the first torpedo designed to use a sea water battery for propulsion and offered other improvements that were used in later torpedoes. The Mk 33 appears to have been the first submarine launched anti-submarine torpedo developed by the U.S. Navy, but only 30 of them were built for test and evaluation.
Active Homing and the Mk 32 Torpedo. Active homing, the second milestone, is significantly more complex than passive homing and only two torpedoes of this kind, Mk 22 and Mk 32, were developed during WWII. Mk 22 began as an effort to add active homing to the Mk 14 torpedo but ended up as a standard Mk 18 electric torpedo design modified by Westinghouse and BTL to include active homing in azimuth only. The homing system transmitted a pulse of 28 KHz sound using both halves of a leftright split transducer. Echoes received by the two halves were processed separately and their relative phase was used to determine the direction of the target. From the relative phase a course correction signal was generated and this signal controlled a change in the gyro angle. The gyro maintained course control between pings of the sonar. The implementation of this scheme with minimal modification of the basic Mk 18 torpedo required a great deal of ingenuity including, in particular, a complex mechanical device called the translator which took signals from the servo amplifiers and power from the propeller shaft to drive the course input for the gyro. One of the problems that is encountered in active acoustic homing systems, but not in passive systems, is reverberation, i.e., reflections of the transmitted sound pulse from random features in the surface, body and bottom of the ocean. Reverberations are effectively false targets and without special features an active acoustic torpedo would often home on them. Fortunately, reverberations die out quickly. In the Mk 22 system, the receiver was blanked for 40 milliseconds after the transmitted pulse and the amplifier gains programmed to increase with time, (time variation of gain, TVG) in order to avoid the reverberation problem. The guidance system was successful, but by 1944 azimuth only homing, even for 21 inch torpedoes, was less attractive than the combination of vertical and horizontal homing offered by competing systems. Work on the Mk 22 was terminated before production designs were completed.
Two competing designs were developed for the other WWII active homing torpedo, Mk 32. One design was developed by BUSL and the other by General Electric both beginning in 1942. The Mk 24 body was used, in fact Mk 32 was designed as a conversion of that weapon 12 with the passive homing system replaced by a small active sonar. Size and weight constraints were severe. The total available volume was less than two cubic feet in the mid-section of the torpedo, space for the transducers in the nose and the space occupied by the Mk 24 depth control in the tail section. Weight was limited to less than 50 pounds. These space and weight constraints meant that the best options could not be used if there were a lighter or smaller option that could do the job satisfactorily. The second problem was to devise a control system that functioned on the basis of short, 30 millisecond, widely spaced, 0. 7 second separation, inputs rather than continuous inputs characteristic of passive homing systems.
The GE system that emerged used a magnetostrictive transducer, four elements wide and eight elements high, that was split into an upper half and a lower half. This configuration made it possible to use phase comparison and proportional control in the vertical plane where it was necessary to home on a submarine hull that measured around seven meters from keel to deck. In the horizontal plane, where the target was about 70 meters wide, a simpler on-off was used. In the absence of an echo the rudders were hard over to port and the torpedo circled in that direction. When an echo was received the rudder was shifted to hard starboard and remained in that position until about one second after the last echo was received. At this point the rudder was reversed and the process repeated. The torpedo thus apparently homed on either the bow or stem of the target, but the dynamics of the torpedo and the electronic time constants shifted the actual homing point toward the center of the target. The main virtue of this homing system was that it used the same amplifiers as the vertical control system without adding complex circuitry and so saved weight and space.
Homing signals in the vertical plane were derived by comparing the phase of the signals from the two halves of the transducer. The up or down signals were used to drive a pendulum frame in which the pendulum was suspended. Electrical contacts connected the horizontal (diving) rudder motor to its power source in such a way as to keep the pendulum centered in the frame. The system thus controlled the pitch angle, and consequently the rate of climb, directly. A hydrostat was installed, but it was used only to control the mode of operation, e.g., set the depth ceiling, and did not provide servo inputs that affected the horizontal rudder.
Reverberation and other false target problems were dealt with by a combination of time variation of gain and blanking. It is interesting that this system also switched between a search mode and a pursuit mode presaging the on-board logic of modem torpedoes.
An experimental Mk 32 produced by General Electric made a successful sound controlled three dimensional run in February 1944, 22 months after the concept was first presented to NDRC. Tests against target submarines began in July 1944 and were successful. Leeds Northrup was selected to produce the GE version of Mk 32 and 10 pre-production units were completed and tested before the project was canceled at the end of WWII. Later, with deliveries beginning in 1950, Philco produced a substantial number (about 3300) of the somewhat different Mk 32-2 torpedoes for fleet use by destroyer type vessels. This torpedo is discussed in a subsequent part of “U.S. Navy Torpedoes”.
The HUSL system was different. The transducer was symmetrically divided into four quadrants. The echo signals in these four quadrants were processed in an ingenious electronic system to obtain rudder orders. The system also contained a Doppler enabling system that prevented homing on reverberation and other false targets including wakes. While the HUSL system was not selected for the Mk 32 torpedo, many of its features were incorporated into the Penn State Ordnance Research Laboratory Project 4 system which was the basis for the very successful Mk 37 torpedo.
Homing torpedoes ascended to paramount importance during WWII and the principal practical techniques, active and passive acoustic homing, were well established by the end of the war. The stage for subsequent U.S. Navy torpedo development was thus, as we shall see in the next part, set during WWII.