Section XX

20-1. Leakage from air systems in submarines as a result of war damage can be a matter of great concern, particularly in the case of air leakage which occurs in main ballast tanks or at other points external to the pressure hull. The high pressure air system naturally constitutes the greatest potential air leakage hazard on submarines since that system is normally maintained at a pressure of about 3000 pounds per square inch, the piping extends throughout the entire length of submarines, and the major portion of the air reservoir is contained in air banks located in ballast tanks external to the pressure hull.

20-2. Depth charge vulnerability tests conducted with Tambor (SS198), Trout (SS202), and Gar (SS206) early in 1941 demonstrated conclusively that cone joint connections in high pressure air systems did not satisfactorily withstand attack at close range. In so far as shock is concerned, cone joints are fundamentally undesirable since such joints are relatively loose and will pound when subjected to sudden and violent movements. This will cause large stresses to be built up and, as the tightness of a cone joint is dependent upon metal to metal contact of the "belled" portion, any accompanying distortion of the sealing surfaces will result in leakage. The Bureau therefore authorized an extensive program to convert the high pressure air system joints on all fleet type submarine to silver-brazed connections and special union-type fittings having plastic annular ring seals.1 Although this program commenced in November 1941, the magnitude of the work involved required that the changes be accomplished on submarines during their regular overhaul periods, and consequently many of the submarines then in service were not modified until late in 1942 or early in 1943. In addition, the high pressure air systems on submarines in service and those in late stages of construction were in most cases only modified to cover the joints in the ballast tanks, battery compartments and easily accessible internal fittings, so that cone joints still remained in numerous locations on many ships even up to the end of the war. As a corollary alteration to the joint modifications, it was directed that all high pressure air flasks in fuel tanks be removed and reinstalled in main ballast tanks to eliminate the possibility of air leaks in fuel tanks causing oil slicks or tank rupture.

20-3. Throughout the war, there were at least twenty-five instances in which high pressure air systems on U.S. submarines developed leakage following depth charge or bomb attacks. In most of these cases, the air leakage was relatively small and had no serious consequence. The Kingfish, Tunny, Tang, and Halibut actions are among the few instances were there was important internal air leakage.2 On


Kingfish (SS234) and Tunny (SS282), the leakage occurred at cone joints of the high pressure air receiving and distributing manifolds. The resulting increase in internal air pressure on Kingfish was reported to be one of the factors which eventually forced her to surface. In the case of the air leakage which occurred on Halibut (SS232) one of the "tee" fittings at No. 1 high pressure air bank in the forward battery compartment carried away and the entire bank bled down rapidly.

20-4. Although many instances were reported where high pressure air lines within ballast tanks developed leakage, in none of these cases was a submarine forced to surface due to increase in buoyancy through rapid accumulation of air pockets in the tanks. In at least one case, Thresher's action of 9 July 1944, air bubbling to the surface from main ballast tank high pressure air leaks (vent valves were open) is believed to have enabled detection by the enemy.

20-5. Instances where high pressure air leakage developed as the result of enemy attacks continued to occur up to the end of the war. However, these casualties were limited for the most part to older submarines in which the cone joints of existing high pressure air systems had either been modified for silver-brazing or were still installed. There were remarkably few cases of high pressure air leakage on those later vessels which were built with air systems having all connections either silver-brazed or fitted with special union joints. Silver-brazed connections are believed to be completely resistant to depth charge attack when properly installed. The most important factor in proper installation is to insure that the clearance between the fitting and the piping is correct so that the brazing material will flow by capillary action. A good installation is facilitated by the use of fittings of the approved preinserted ring type, and tests have demonstrated that failure should occur in the pipe or fitting rather than the joint. It is believed that all cases where silver-brazed high pressure air system joints failed on submarines were due to one or more of the following reasons: imperfect brazing technique; excessive clearances in joint assemblies; the use of existing fittings which perhaps were excessively weakened when modified for silver-brazing, or the use of fittings which were reheated too many times either during the initial installation or through repeated use. The resistance of silver-brazed joints to explosive attack is dependent in very large measure upon the skill of the silver-brazing operator, the quality of the supervision and the methods of inspection. The special union-type joints using plastic annular ring seals have proved remarkably resistant to damage, there being no reported case in which serious leakage an be identified as having occurred at such joints through war action.

20-6. The forward and after main ballast tank vent risers, as installed on submarines prior to SS417, pass through the forward and after torpedo rooms and are subject to full sea pressure at all times. In addition, when a submarine is under attack these risers are exposed to explosive loading from close detonations. During the war a number of cases of leakage at the flange joints were experiened. In one case, the copper riser had been dented in installation to clear an interference and under depth charge attack the dent was straightened out, fouling a passing shaft. As a result of these factors there was considerable concern in the submarine service over the hazard involved in having these risers, open to the sea, passing through the torpedo rooms. There were in


service two types of risers, steel and copper. The copper risers had been installed to avoid corrosion and loss of strength in service. Both types were stronger than the pressure hull. However, the copper risers and their bronze flanges were subject to plastic deformation and were thus more vulnerable to flange leakage than the steel risers. The flanges of the steel risers involved no greater hazard, if properly installed, than any other flanged hull connections. It was usually possible to stop any leakage by tightening the flange bolts. However, in one case, Tunny's action of 1 September 1944, damage by bomb attack caused leakage at the flanges of the internal port vent riser of main ballast tank No 7 in the after torpedo room and this leakage was the determining factor in the Commanding Officer's decision to terminate the patrol (Photo 20-1.)3

20-7. Measures were instituted during 1943 in an effort to improve the resistance of internal ballast tank vent risers to flange leakage. The copper risers and bronze flanges then in service were replaced with new risers and seamless drawn or specially welded steel tubing fitted with steel flanges.4 All new construction submarines were already being provided with steel risers. Steel is a more satisfactory material for such service since its strain energy within the elastic range is much greater than that of copper. Instructions were issued to make special efforts to secure tight joints during the assembly of internal riser piping by truing the flanges to avoid bolting strain, by using pressed asbestos gasket material rather than rubber or cloth inserted rubber gaskets, and by installing lock washers under the flange nuts to prevent backing off. In addition, branch lines were eliminated in order to reduce the number of joints. It is believed that the major cause of trouble with these risers was improper installation in which the flanges were not properly regulated and in which the bolts were overstressed or thick rubber gaskets were used to correct improper flange fit. however, it was recognized that the physchological effect of such risers should be eliminated and a rearrangement of tankage in new constriction submarines was undertaken in order to relocate the torpedo room main ballast tanks so that no vent risers would be required. Commencing with the SS417 Class, the forward ballast tank was located ahead of the forward trim tank, and the after ballast tank was eliminated completely, displacing it with a variable fuel oil tank. Although the internal risers on submarines previous to SS417 will continue to be a source of concern under attack, and are therefore undesirable, it is considered that the hazard is one of flange leakage, usually in controllable amounts, rather than of sudden rupture leading to the loss of the vessel. The strength of the steel vent risers is considerably greater than that of the hull and consequently the risers are less liable to rupture than the adjacent pressure hull. War experience demonstrated that the strength of the risers was sufficient to prevent failure even in cases where adjacent pressure hull structure was considerably deformed.


20-8. The hydraulic systems of U.S. submarines extend throughout the entire length of the vessels but, unlike the high pressure air system, are wholly confined to the spaces within the pressure hull. Hydraulic power has been extensively developed by the U.S. Navy for submarine applications and on vessels now in service the following vital equipment is hydraulically operated: rudder, diving planes, periscopes, radar masts, negative and safety tank floods, main engine air induction outboard valves, ballast tank vents, engine exhaust valves, sonar hoists and torpedo tube outer doors. Hydraulic system piping must therefore be extremely resistant to depth charge attack. War experience demonstrated that such piping frequently developed minor leakage at joints and connections as a result of close attacks, but that, with few exceptions, the numerous cases where failures in hydraulic power occurred could generally be traced to causes other than loss of hydraulic oil through leakage. By far the most frequent cause of hydraulic power failure was the loss of auxiliary electrical power supplying the main hydraulic IMO pump motors and the motors driving the Waterbury hydraulic pumps for the diving planes and steering gear. These failures occurred when the submarines concerned were not rigged for silent running and in such cases hydraulic control of equipment was usually quickly regained by shifting to hand hydraulic power. Although there were numerous cases where minor leakage developed at cone joint fittings and a few instances where leakage ocurred at silver-brazed or special joints as the result of close attacks, such leakage was generally of small amount and, at cone, flanged or union joints, usually could be easily stopped.

20-9. There are several instances where hydraulic piping derangements caused loss of power to vital equipment. For example, Salmon (SS182), in her action of 30 October 1944, temporarily lost steering control when the hydraulic supply line from the Waterbury pump tore loose at the steering hydraulic manifold.5 However, steering was soon restored by shifting to hand power and later to emergency power. Thresher (SS200) was forced to terminate her tenth patrol when, as a result of depth charging on 13 November 1943, sea water entered the hydraulic system through a leak in the external replenishing tank or piping forward of the conning tower. These external hydraulic oil replenishing tanks constituted an obvious weakness and were subsequently removed on all submarines to prevent recurrence of sea water contamination, as on Thresher, and to place hydraulic systems entirely within the protection of the pressure hull. On Redfish (SS395), as a result of a depth charge attack on 19 December 1944 during her second war patrol, the entire main hydraulic system was temporarily put out of commission by loss of hydraulic oil due to leakage at piping connections at the pump-room accumulator, steering manifold aft, and sonar and bow plane tilting piping forward.6


Photo 20-1: Tunny (SS282). View showing damage which occurred to port vent riser of MBT No. 7 in after torpedo room of Tunny as a result of her action of 1 September 1944
Photo 20-1: Tunny (SS282). View showing damage which occurred to port vent riser of MBT No. 7 in after torpedo room of Tunny as a result of her action of 1 September 1944. Patrol was terminated due to excessive leakage which occurrd around bolts of the two flanges shown, rendering submergted operations dangerous. Pulling of flanges is indicated by the absence of the cork insulation in their immediate vicinity.


20-10. As a result of the Tambor, Trout, and Gar tests in 1941, which disclosed that the cone joints in hydraulic systems would start minor leaks when subjected to shock and that such leaks could be a considerable nuisance when located in relatively inaccessible areas, the Bureau directed that hydraulic piping on new construction submarines thenceforth be provided with silver-brazed sleeve type joints in lieu of cone joints, with the exception that special joints approved by the Bureau, such as union plastic seal ring joints or Van Stone joints, were to be used where piping must be dismantled to permit overhaul of machinery. The resistance of submarine hydraulic systems to shock damage was markedly increased by the introduction of silver-brazed and union plastic seal ring joints in lieu of the cone joints. However, the fact that close attacks continued to cause hydraulic system leakage up to the end of the war certainly indicates that further improvements are required before these vital systems can be considered to be as invulnerable to damage as their importance warrants. The experience of Redfish (SS395), mentioned in paragraph 20-9 above, is illustrative of this need. Van Stone type joints, of which many remain in current hydraulic installations, appear to be fundamentally less desirable for severe shock service than silver-brazed and union plastic seal ring joints. Unfortunately, those reports received from submarines noting hydraulic system leaks have rarely identified the types of joints or connections at which the leakage occurred. Leakage has occurred at silver-brazed joints, however and the remarks in paragraph 20-5 above apply with equal force to the quality of silver-brazed joints in hydraulic systems.

20-11. Impulse air systems for the forward torpedo tubes have on several occasions developed leakage at points outside the pressure hull as the result of close attacks. The air reservoir for the forward impulse air system on U.S. fleet type submarines consists of shatter-proof flasks located in the forward superstructure. Since this location is external, the air flasks and piping to the pressure hull are subject to full explosive loading, and any leakage which may develop external to the pressure hull will result in air bubbles rising to the surface. Unfortunately, sufficient space is not available in the forward torpedo room to permit interior stowage of these bulky impulse air flasks. In at least one instance, external leakage from the forward impulse system had serious consequences. On 14 January 1943, during her sixth war patrol, Pike (SS173) was heavily depth charged by several Japanese destroyers.7 Six hours after the initial attack, Pike planed up to periscope depth but was promptly bombed by an enemy plane. Periscope observations at this time disclosed that large air bubbles were rising to the surface from the impulse air system forward and this evidence of Pike's location undoubtedly enabled the enemy to make the prompt attack. The external portions of the forward impulse systems are considered to be inherently vulnerable to damage for the following reasons:


  1. the flasks are in an exposed location;

  2. the joints in the piping leading from the flasks to the torpedo room are of bolted flange type construction;

  3. the bolted through-connection for each pipe at the hull is particularly vulnerable to damage should the adjoining pressure hull sustain deformation;

  4. on war patrol, the flasks are normally maintained with air at pressures in excess of 300 pounds. It is hoped that investigations now underway on methods of ejecting torpedoes without the use of air may perhaps eventually permit the removal of torpedo impulse air systems.

20-12. Main engine air induction and ship's supply ventilation outboard piping sections as installed on U.S. submarines through 1943 were connected with bolted flanged joints. These joints were the source of leakage into the induction and ventilation systems in numerous instances where close attacks are sustained. In approximately 20 cases, the leakage was serious enough to warrant mention in war patrol reports as a major defect. Sudden flooding of induction systems in several instances caused immediate loss of depth control on the submarines concerned. In addition, when the induction or ventilation hull valves were damaged at the same time the outboard piping flooded, leakage occurred into the engine and maneuvering spaces, in some cases with serious consequences.8 As a result of numerous casualty reports, authority was granted in January 1944 to eliminate all bolted flanges joints and substitute welded joints in the main induction and ship's supply ventilation piping.9 There are only a few instances on record where submarines have sustained main induction or outboard ventilation leakage after this alteration was performed, Seahorse's (SS304) action of 18 April 1945 being one notable case. It is significant to note that although three submarines received very heavy damage on patrols subsequent to being provided with welded joints, on none of these three boats was leakage into the main induction piping reported.10

20-13. In two instances, main induction piping partially or completely collapsed under close attack: Salmon (SS182), Section VII; and Scamp (SS277), Section IX. In addition, on Scamp the main induction pipe wall was torn in four places by its welded bracket supports.11 The fact that collapse of main induction piping occurred on the above mentioned ships demonstrates that the strength under explosive loading of the induction


and ventilation outboard piping on those two Classes was less than that of the pressure hulls. There are no known cases where induction piping has collapsed on SS285 Class and subsequent submarines. It should be added, however, that these large tubes running for long distances outside the pressure hull with large exposed areas and rather light connections to the pressure hull constitute a hazard, the elimination of which would make an important improvement in the resistance of a submarine to damage. Future designs will probably incorporate an internal induction line leading directly from a snorkel intake to the engine spaces.

20-14. Close underwater detonations have frequently caused numerous valves in submarines to leak due to rotation of the disc and spindle or distortion of the valve discs or stems. For example, Cod (SS224), during her sixth war patrol, reported "During the depth charge attack on 24 April 1945 the first string of charges was of such intensity as to cause the discharge sea and stop valves in the officers' head to open violently; the sea valve operating handwheel was observed to spin rapidly and the stop valve operating lever was thrown to the open position with such force as to shear the pin joining the hand lever to the vertical operating shaft. The stop valve was closed with difficulty, using a wrench." Cobia (SS245) similarly reported that, on 14 May 1945 during her fifth patrol, the water closet valves spun to full open on each of several close detonations. War patrol reports occasionally contain mention of valves which were reported to have opened by as much as three to five full turns as the result of depth charge or bomb attack. In most instances, the opening of valves or leakage due to valve deformation has constituted a nuisance value only. The usual consequences were that full sea pressure was transmitted to internal piping systems, sanitary tanks flooded, inboard vents sprayed water or oil, etc. One of the most frequent damage control measures which submarine personnel were forced to take during attack was the simple task of closing valves which opened. The numerous war patrol reports which mention that valves backed off indicate globe valves were particularly susceptible to rotation and that generally valves throughout the length of the boats were affected rather than those confined to localized areas. The action was not limited to valves exposed to sea pressure for there were many reports where hydraulic and air system valves were affected.

20-15. In a few instanced, the chain of events caused by the opening of a valve had serious effects. The following experiences are illustrative:

(a) On 26 October 1944 during her first patrol, Icefish (SS367) lost the use of her entire hydraulic system when the 225-pound air supply valve to the hydraulic system vent and replenishing tank jarred open and caused the loss of much of the oil in the system.12 This in turn disabled the diving planes and steering until the hydraulic system could be refilled with oil. In addition, a fuel tank inboard vent valve on Icefish jarred open in the after battery compartment, causing profuse leakage of oil into that space, and the forward sanitary tank flooded from the sea when the discharge line valves opened.


(b) On 4 November 1944, during her twelfth war patrol, the No. 7 torpedo tube inboard vent valve on Sailfish (SS192) jarred open as the result of a bomb attack. The large volume of water which entered the after torpedo room through this vent caused the boat to assume a 25 degree up angle and depth control was lost.

(c) On 26 June 1945, during her sixth war patrol, No. 2 sanitary tank on Parche (SS384) flooded through the outboard vent which jarred open under attack, and a considerable amount of sea water entered the crews' mess and washroom through the inboard vent of the same tank.

20-16. A program of development and design of explosion resistant valves and locking devices to prevent rotation is currently underway at the Portsmouth Naval Shipyard. One of the features of this program is the actual explosive testing of various valve designs under simulated conditions. It is interesting to note that no large amount of valve rotation has yet been observed as a result of the tests, a 30 degree or one-twelfth full turn rotation being the maximum so far obtained. However, it has been found that deformation of valve stems and discs occurs such that a one-fourth turn of valve handwheels is sometimes necessary to secure tightness after a test. It is of course extremely desirable that valves remain tight during depth charging, but the most important requirement is that submarine valves be so designed that although leakage may occur due to deformation or rotation under attack, they will be capable of being made tight again by simply taking up hard on the handwheel.

20-17. In the case of a globe valve, the factors that keep the valve closed under normal conditions are the friction of the thread surfaces in combination with the lead angle of the screw thread, and rotational inertia, the latter being a negligible influence. Under explosive attack, the valve and is supporting structure are set in vibration and it is believed that the friction loads will from time to time be zero or appreciably reduced during the vibrating process. At such a time, it is believed that asymmetrical rotational accelerations can easily make the valve rotate. For example, one way to unscrew an obdurate bolt is to vibrate it while simultaneously applying a torque. In practice, a depth charge attack consists of a series of impulses and the vibrations in submarines are prolonged. One factor which would certainly increase the prospect of rotation of a globe type sea valve would be where the valve disc is deformed by direct pressure pulse loading so that it no longer seats, thus reducing the friction on the threads and rendering it relatively easy for later rotational forces to spin the valve. It is considered that the Portsmouth tests may have failed to produce appreciable valve rotation because the test structure used was more rigid than that of a submarine hull, the test valves may have been initially tightened to a greater degree than would normally be the case with many valves aboard a submarine in actual service, and consecutive explosions were not employed.

20-18. In an effort to reduce the possibility of war damage, various alterations were authorized in order to reduce the number of piping connections passing through the hull and to provide various piping connections subject to sea pressure with both sea and stop valves where such had not been originally installed and war experience indicated to be desirable. Among the many items eventually eliminated on submarines


in service and omitted from the designs for new construction were the following: secondary drainage lines external to the pressure hull; external piping systems penetrating more than one tank; outboard battery exhaust lines; external voice tubes; external conning tower ventilation lines; magnetic compass binnacle tubes; salvage air connections to normal and reserve fuel oil tanks; conning tower flooding line and escape gage sea connections; trim line deck hose connections; various water closet overboard connections; differential gage lines on main ballast tanks; and auxiliary engine exhaust piping. Hull stop valves were provided for the following lines where only sea valves had been previously installed: fuel ballast tank differential gage lines; fuel ballast tank filling and transfer lines; and fuel oil filling connections. The cumulative benefit of these and similar alterations, most of them minor, certainly reduced the vulnerability of our submarines to damage and decreased the number of items to which damage of "nuisance" value was commonly experienced.

20-19. In summing up the lessons learned from war experience with regard to submarine piping systems, the following general observations can be made:

  1. silver-brazed, welded or special union joints and connections should be used wherever possible on all piping where subjected to severe shock loading, both within and without the pressure hull, and even at the expense of increasing the time and cost of repairs;

  2. further development work is required to obtain flanged type joints which will be as shock resistant as silver-brazed, welded or special union joints under war service conditions;

  3. proper sectionalization of vital piping systems to isolate damage, and training of the crew in the use thereof, are of great importance;

  4. the number of hull openings should be held to an absolute minimum;

  5. explosion resistant sea valves and anti-rotational locking devices must be developed;

  6. piping should not penetrate external tank divisional bulkheads;

  7. sea chests or piping leading to sea chests should not be located in fuel tanks;

  8. brackets should not be welded directly to exposed steel piping as in the case of the induction piping on Scamp;

  9. the utmost care must be taken in the design, installation and maintenance of piping systems and detail fittings to prevent weak points;

  10. war experience has demonstrated that piping systems on submarines have been more subject to damage, although usually of a nuisance type, than the hull structure, main propulsion plant and principal auxilarly machinery.


Table of Contents
Previous Section (XIX) * Next Section (XXI)


1. Authorized by ShipAlt SS62 of November 1941. The modification of fittings in tanks external to the pressure hull was classed as a Class "A" Alteration. Modification of fittings within the pressure hull was classed as a Class "B" Alternation.

2. See Kingfish narrative, Section V; Tunny narrative, Section X; and brief of Halibut damage, 14 November 1944, Appendix I.

3. See brief of Tunny damage of 1 September 1944 in Appendix I.

4. ShipAlt SS171 of 16 March 1943.

5. See Paragraph 7-31.

6. See brief of Redfish damage in Appendix I.

7. See brief of Pike damage in Appendix I.

8. See Whale (SS239), 30 October 1942, Appendix I; Permit (SS178), 13 February 1944, Appendix I; Grenadier (SS210), Section VIII; and Kingfish (SS234), Section V.

9. BuShips conf. ltr. C-SS/S48-23(515); EN28/A2-11 of 10 January 1944.

10. Icefish (SS267), 26 October 1944; Halibut (SS232), 14 November 1944; Redfish (SS395), 19 December 1944. See Appendix I for briefs of damage.

11. See Scamp, Paragraph 9-22 and Photos 9-5 and 9-6.

12. See brief of Icefish damage in Appendix I.

Transcribed and formatted for HTML by Patrick Clancey, HyperWar Foundation