From the time the first rude boat was fashioned from a log, this being the first great step from the raft type where buoyancy was obtained by the use of solid timbers, water-tightness has been the prime requisite of water-borne carriers. Without it buoyancy and stability become relative, decreasing as water gains entrance and vanishing at some point set by the initial design. Where leaks are well within the capacity of the ship’s pumps and drainage facilities, there is nevertheless the discomfort of a “wet ship,” spoilage of cargo or stores, contamination of fresh water or fuel, and kindred ills. The tightness of the steel hull was one factor that drove the wooden ship from ocean trade. Absolute water-tightness of the shell, of fuel and water tanks, and of weather decks, is of first importance and receives such close attention that no further mention need be made herein. If leaks develop, as they will, they are discovered immediately and if beyond the capacity of the ship’s force to repair, receive high priority during yard overhaul.
Early man progressed from his first adventures in sheltered waters or streams until he required larger and larger boats; finally his dugouts would no longer serve. Instead of a single timber, hollowed by fire or axe, a large number of pieces were used, fitted together to form the hull, and secured by lashings and trenails, finally made watertight by rude caulking or by the use of gums. This type required some means of stiffening to hold the shell to its work, and maintain the ship form against water pressure and load. In the small boat the thwart served this function, but as size increased, and the need of subdivision of the space within the shell developed, these dividing bulkheads served to hold the sides in place. At this stage we find that bulkheads, both longitudinal and transverse, had two functions: the body of the ship was subdivided to provide better use of the space, forming holds, storerooms, staterooms, and other compartments; and further, these bulkheads in conjunction with the framing supported the shell and held the ship as a whole in shape. So far as water-tightness was concerned, the first and only line of defense still remained the planking or shell.
If these bulkheads, or certain of them, are built strongly enough to withstand water pressure at-a head equal to or greater than that which would result from flooding, and further, are made water-tight, they can perform a third and very important function. In the event of damage to the shell, only the space bounded by adjacent and intact watertight bulkheads need be flooded. The bilging is restricted and, assuming a sufficient degree of subdivision in the design, the ship should be able to make port, either under her own power or, at least, at the end of a towline. Structurally, and speaking in rather broad terms, a bulkhead designed to meet water pressure under bilged conditions will have ample strength to meet the requirements of its other functions. Water-tightness in such bulkheads presents a different and much more complicated problem than in the shell, due to the necessity of running pipe lines, electrical leads, or shafts through them, and providing access with doors, manholes, or bolted plates.
Obviously, it is possible to use a great many of these water-tight bulkheads, and make the subdivision so minute as to make a ship practically unsinkable. In the commercial vessel, however, subdivision is complicated by the demands of service. In the cargo-carrying class, holds must be of relatively large size, or the loading and unloading, as well as economical stowage, of cargo becomes too difficult. In passenger-carrying ships, similar complications are encountered; comfort is highly rated, and as a result there can be few divisions in the spaces used by passengers. Nevertheless, there is a growing opinion that better subdivision to keep the damaged ship afloat is worth more than an elaborate outfit of boats, some of which, perhaps, cannot be launched when the emergency occurs, and which are unsafe at best, if weather is bad and a high sea running. In this commercial type of vessel the probable damage which necessitates some degree of water-tight division would be due to the normal hazards of the sea: grounding, collision, opening of seams due to racking in heavy storms, or failure of structure under some unusual stress; less frequently, to leaky or faulty ports and hatches, failure of piping or sea chests, or mishandling of valves.
Referring now to combatant types, the naval vessel is subject to all of the foregoing, and in addition, must be designed and kept in continual readiness to receive damage in battle. Whether this damage be due to shell fire, torpedoes, mines, or bombs, the ship must be able to sustain the maximum amount of damage, according to type, and not only remain afloat but fight, as well. Flooding must be restricted to a minimum at point of damage, for list and excessive trim must be corrected by counter flooding, in order that the ship may continue to fire. The punishment which the ship can take will depend first on the resistance to penetration of armor over vital areas, and secondly, to the extent of flooding, intentional or otherwise, which takes place in the unprotected portions. Compartments must still remain of reasonable size, consistent with the use to which they are to be put, and the total weight of hull structure cannot exceed certain limits. Subject to these limitations, naval ships of combatant type must be, and are, subdivided as minutely as possible.
Credit is generally given to very efficient water-tight subdivision, as well as protection, in enabling the German ships at Jutland to receive tremendous punishment and still remain in the fight, crippled more or less, but still in action and able to make port later. According to Captain Schelbe, ships with as high as twenty-five severe hits and extensive damage below the water line succeeded in making home ports. The Goehen, renamed the Sultan Selim, in another theater of the war, furnished a striking example. This ship suffered severely in an early action with Russia, and not long after struck a mine. In both cases she made port and could be repaired. At Imbros in 1918, she hit one or more mines, but was still able to proceed slowly until hit several times by aircraft. She was then beached, and while stranded, was subjected to several hundred attacks by aircraft. In spite of it all, the ship was refloated, made Constantinople under her own steam, and was repaired. Other instances are many where proper subdivision, in conjunction with proper upkeep, and smart handling of water-tight doors and hatches, have sustained a badly damaged ship. It is a proved fact that a large amount of comfort, ease of passage, and space is well sacrificed to obtain positive security against extensive flooding.
Water-tight subdivision can be, however, one of the greatest of snares and delusions. The notation “W. T.” on the plan, or gleaming paint and glistening bright work may be of very little help when a compartment floods, and a thirty-five-foot head of water is working on a bulkhead. A surprising quantity of water can flow through a comparatively small hole when there is a little pressure behind it. And when these small holes and leaks are legion, as they are too apt to be after the ship has seen a number of years of service, the bulkhead for all its well-kept appearance is likely to look and behave like a kitchen sieve when the emergency occurs. There are too many instances where just that has occurred, although in some the flow was sufficiently dammed to permit the pumps to make some headway, or docking facilities and help were near enough to make salvage possible. In such cases the report may read “… after engine-room slowly flooding …,” but the point is that those bulkheads were not damaged in the accident, and there should have been no flooding at all, following that immediately after the collision. In one case where magazine spaces were flooded intentionally to prevent fire spreading to powder, the water-tight integrity proved so poor that flooding had to be discontinued to avoid serious consequences.
A little calculating may serve to further emphasize the point. The quantity of water which will flow through a given orifice is expressed by:
Q = k A v = k A \/2gH
in which:
k = coefficient of discharge
A = area of hole
H = static head
g = acceleration due to gravity
Using a coefficient of .9, and a head of 9 feet, and a hole 1 square foot in area, the quantity of water is found to be 9.500 gallons per minute, or, in terms comparable with the displacement of ships, at the rate of 2,200 tons per hour. If this one-foot hole is 25 feet below the water line, the rate will be found to be 3,700 tons per hour. Calculations of this nature form the basis of argument against the installation of a main drain, intended to free the ship of water in the event a large breach is made by explosion or collision. In modern installations, the main circulating pumps are usually connected to drainage lines in the engineering spaces, so that their large capacity could be utilized in case of serious leaks in this part of the ship. The remaining pumping facilities are intended to handle very much smaller quantities, service drainage, or leakage due to strained, but not ruptured, bulkheads. For example, in one case the combined capacity of the main circulating pumps is 60,000 gallons per minute but the main drain is only nine inches in diameter, so that only a small part of this capacity is available, and that only in the boiler- and engine-rooms. The combined capacity of all other pumps which could be used for pumping other spaces on the ship is somewhat less than 2,000 gallons per minute and this total is not available at any one location.
Reliance is therefore placed on the watertight subdivision to localize the flooding, losing the buoyancy of one or more compartments, rather than on freeing the ship of water altogether. But assume that one of the bulkheads bounding this area is in the following condition: a voice-tube lead has been changed, and the original four-inch hole was not blanked. A hatch stanchion stowage was also removed, and the four bolt holes remain. Various small leaks in the structure, doors, and vents will bring the total area to about fifteen square inches. Further assuming a mean head of only ten feet, water will enter at the rate of over 1,000 gallons per minute. This is a rather bad case, but it is not impossible, nor unheard of, as it should be.
When a naval ship leaves the builder’s yard it can safely be assumed that every part of water-tight structure is, for all practical purposes, water-tight. This is assured by the rigid schedule of compartment testing and inspection. Except in water and oil tanks, and shell, bottle-tightness is not required, nor, in many places, can it be obtained. Most fittings, doors, glands, and portable plates cannot be made absolutely tight, or if they were made so for test, would develop leaks after operation. The amount of leakage has been noted in a great many experiments, and a practical limit ascertained, which forms the basis of “allowable drops” in air testing. A compartment which tests within this limit will never cause trouble in the event it is flooded. The test requirements are set forth in the “General Specifications for Building Ships,” and are amplified, with test heads and pressures given, in the “Detail Specifications” of each individual ship. In addition, each ship is furnished, when commissioned, with the results of the builder’s tests. With these, and a Bureau of Construction and Repair booklet, Air Test of Compartments, every officer, and in particular the first lieutenant and his assistants, should be familiar.
The requirements of the specifications are carried out in a private shipyard as follows. As soon as the structural work of a compartment is completed, the contractor requests preliminary inspection by the superintending constructor’s office. A special mechanic in his force goes over the tank in detail, checking against the approved plans to make certain that the work is in exact accordance. From his own experience in the trade, he is able to detect any evidence of poor workmanship, which must be corrected before he will certify the tank ready for pressure test. At this stage permanent doors, hatches, and manholes must be in place, but no piping, ventilation, or electrical leads are yet installed, and other holes, if any, are blanked. A “test sheet” prepared by the contractor from the data given in the “Detail Specifications,” and approved by the bureau, gives the test head in water, or pressure if air is used, which is then put on in the presence of the inspector, and held until he is satisfied that the work is tight and shows no undue deflection. This constitutes the “Strength and Tightness” test, which is intended to check the structural work, both in design and workmanship.
Much later in the building period, and after every fitting, pipe line, cable, label plate and permanent article is installed, the inspector is called for a check on completion. The compartment is again checked against the plans to .make sure that all work connected to, or through, the boundaries is complete, and properly done, ready for the “Completion Air” test. The pressure in this test is usually somewhat lower than for the strength test, and is held for a definite period, usually twenty minutes. During this period, the air hose is disconnected, and the drop in ounces read on a mercury gauge. The inspector then submits a report to the constructor’s office, giving the time, drop, and a list of the fittings, etc., if the drop is more than a few ounces. From this information, an “allowable drop” is calculated as described in the booklet Air Test of Compartments, and if this is greater than the actual drop, the test is passed. The compartment is then locked, and opened only with the inspector’s knowledge for completion of painting, stowage of portable articles, and cleaning. Breaking a pipe line, running a new cable, or any alteration affecting the boundaries requires a retest. Finally, just before delivery, an inspection is made to determine cleanliness, condition of paint, stowage, etc., and not until then does the government accept that compartment as satisfactory. The routine has been described briefly here to indicate the thoroughness with which the new ship is made safe and tight.
It is unfortunate that the upkeep of the ship in service is apt to run more to goodlooking paint and bright work on these bulkheads than to maintaining the original watertight integrity. In fact, the one may assist in destroying the other. A water-tight door dog is not an easy matter to fit and make tight without working unduly hard. However, it is no easy matter to polish one in place. On how many ships are these dogs frequently removed for cleaning, and how many times are they replaced with sufficient care and skill to make sure that every part is in place and reasonably tight? Are all drop bolts on hatches and water-tight covers in place? Regardless of the condition of a rubber gasket, doors will not be remotely tight unless the knife edge is both clean and true. The easiest way to clean a knife edge is to use emery cloth, the coarser the quicker, but the result is usually a wavy edge after a short time. Frequent chalk tests on doors selected at random will indicate this condition. The knife edge is rubbed with chalk, and the door closed and dogs set up by hand as it would be in an emergency or regular use. On opening, the chalk transfer on the rubber gives a rough check on the efficiency of the door. And it frequently tells an interesting story.
The regulations prohibit the use of pneumatic scaling hammers on light plating, either bulkheads or shell. Even where the plating is relatively heavy, the tightness of the riveting is not improved by the careless use of the gun. Scaling hammers should be as light as possible, and should not be held at right angles to the work. The paint can be removed by this very common method, but only at the expense of the bulkhead. Blows squarely placed on the center of a rivet point will loosen the best-driven rivet. The best rule is to scale as infrequently as possible, and by training and rigid inspection, to insure that it is done properly.
Although the foregoing faults may cause extensive leakage and in the aggregate destroy water-tightness entirely, much more serious are the openings left by the removal of pipe lines, change in the lead of cables or in the location of riveted or bolted clips, racks, and miscellaneous fittings. Frequently such changes are made by the ship’s force, leaving an opening which requires a riveted or welded patch plate. Until small welding outfits are part of our equipage, permanent repairs are beyond the capacity of the ship in many cases, but such an opening can be made practically water-tight by a well bolted patch plate, with a good canvas gasket soaked in thick red lead, and a grommet and washer under each nut. A piece of plate, bolted up without these precautions, will look just as well under a number of coats of paint, but it will not hold water. Holes left by old cable leads, or removal of bolts or rivets, can be effectually if not neatly closed by a bolt of proper size, two washers and the ever-necessary grommet, soaked in paint. With regard to electric cables, the screw holes at old clip locations must not be forgotten. The only satisfactory method is to cut the screw off flush, and peen it on both sides of the bulkhead.
The danger involved in leaving off bolted and portable plates provided for access, or removal of large or cumbersome pieces of machinery or equipment, while at sea, is so obvious as to require no elaboration. Yet it has been known to happen, leaving openings several square feet in area in bulkheads bounding large and important spaces. Not so evident is the lack of closing plates at points where light ventilation ducts pierce W. T. bulkheads. A sliding or portable section is provided in the duct in such cases, which appears to close the duct effectually. It is not designed or intended to withstand water pressure, however, and the closure plate, or other device, should be secured in place near at hand and ready for instant use. Moreover, the sliding or portable part must not be gummed with paint, but must work freely.
When alterations are made by a navy yard, permanent repairs are presumed to be made, and the compartments air tested if any work, new or old, penetrated the boundary. With one eye on the cost, and the other cocked at the completion date, this last precaution may be omitted. The ship personnel are well within their rights and are, in fact, only performing their necessary duty as an inspection force, when they insist that the job be properly completed and water-tightness made as certain as in the case of a new ship. On the other hand, in writing job orders, compartment testing instructions should form part of every job affecting water-tight integrity.
A careful surface inspection will locate the worst and most serious defects. But after all visible openings are found and corrected there are apt to be enough leaks remaining to render the bulkhead almost useless as a water-tight structure. These can be found only by filling the tank or compartment with some fluid, under sufficient pressure to force it through the leak into an adjoining compartment, thus indicating its presence and nature. The use of water is prohibited in most cases by the contents or nature of the space; furthermore it involves much time, is difficult to remove, and if the bulkheads are in poor shape, it is difficult to control until tightness is attained. For tightness testing, air is the easiest and best means to use, the leaks being located with soapy water or a candle flame, and a final check on tightness being obtained by noting the loss in pressure after the supply is cut off.
In order that the water-tight integrity of naval ships will be maintained, air pressure tests at intervals not greater than one year are now required. On battleships, cruisers, and submarines, these are to be performed by the ship’s force, the ship being supplied with compartment air-testing apparatus. To further facilitate these tests, new ships are equipped with fittings on doors, hatches, or manholes, to which the air hose and gauge may be quickly attached, eliminating the necessity of drilling and plugging test holes. Installation of these fittings is also in the modernization programs of older ships. Chapter 12, of the Bureau of Construction and Repair Manual, gives in detail the requirements of these periodical tests, the pressures to be used, and the methods of performing the tests. If completely and thoroughly carried out, they impose a large amount of work on the ship’s repair party. On a new 10,000-ton cruiser, for example, some 187 compartments are tested for tightness in the course of construction. Of this number, tanks and fuel-oil spaces in use, boiler- and engine-rooms, and certain other spaces, are excepted from test, leaving approximately one hundred spaces to be tested every twelve months. This is no small task for the party on a cruiser, and if bulkheads are permitted to get into bad shape, may well be hopeless. Due to gunfire and the normal working of the ship in service, the degree of tightness required of a new ship can scarcely be obtained, nor is it required. The fall of pressure must be reasonably close to that of the last previous satisfactory test, or, where data is missing, and in any case, not greater than 10 per cent of the initial gauge pressure in ten minutes. In the new ship building, this would be considered very high, and the requirements are much more severe in order that the ship may get a fair start in its service life. In a storeroom of average size, well below the water line, tested at a pressure of five pounds gauge, a 10 per cent drop in ten minutes represents an equivalent total aperture of but 0.04 square inches. Not many leaks can be overlooked and permit the drop to remain less than this figure. A general rule cannot be followed in all cases, of course, particularly in small compartments where a very small leak will occasion a large drop of pressure. In such cases, the allowable drop should be calculated, following the simple method demonstrated in the pamphlet on air testing referred to before. The importance of this work must be emphasized, and a definite program laid out and adhered to, for, if this is done, and close attention paid to all changes and need for repair, testing becomes a much easier, routine job.
No reference has been made to gas-tightness, not because of lack of importance, but because it is very nearly synonymous with water-tightness. Since the gas will not be under pressure ordinarily, except possibly due to wind, or in ventilation systems, tightness against gas is more easily secured. It behooves the first lieutenant, however, to look into the condition of the lighter bulkheads above the water line and the closed-in, and supposedly gas-tight, portions of the super-structure.
The official publications referred to in the course of this article contain all the information needed by the first lieutenant and his assistants. It is desired here to put down, in brief form, only certain points that seem to the writer to be of greatest importance.
(1) Close openings due to changes and alterations at the time the change is made, and see that it is properly done before painting.
(2) Inspect as carefully and frequently for holes and other visible leaks, as for dirt and poor paint.
(3) See that dogs, drop bolts, and other water-tight fittings are never removed for polishing. It can be done in place.
(4) See that rubber gaskets are not only free of paint but are live, and that knife edges are not only clean, but also true.
(5) Use scaling hammers only to remove rust, and paint that is blistered and dead, and then use with care.
(6) Lay out a program of tests and carry it out, keeping the test gang clear of other work if possible.
(7) Make sure that the proper pressure is being used. Too little will not show leaky glands and fittings, and the test will be misleading. Too much pressure may cause extensive damage, and be highly dangerous. In case of doubt, consult the bureau.
(8) Never use a spring gauge. A mercury manometer is provided with all test outfits, is sensitive, reliable, and nearly foolproof.
(9) Give the same attention to the water-tightness of bulkheads and decks that is demanded by the shell, fuel, and water tanks.
(10) Make sure that all closing devices, plates or valves on piping, air escapes, ventilation ducts and so forth, are always in place and workable.
A prominent professor of naval architecture, whose technical training is supplemented by years in a shipyard from apprentice to superintendent, once made the remark, "The more you learn about ships the less inclined you are to go to sea in one." This has some basis of truth, perhaps, but certainly the more you know about the tightness of your ship, the better you will feel when an emergency arises. The designer did his job when he laid out the subdivision and designed the structure according to the best practice; the shipbuilder did his when he followed those plans, and made the ship sound and tight; it remains the duty of the forces afloat to keep the ship in that condition.